Animal Biodiversity and Conservation issue 24.1 (2001) by Museu Ciències Naturals de Barcelona MCNB

Formerly Miscel·lània Zoològica

2001

and

Animal Biodiversity Conservation 24.1

"La tortue greque" Oeuvres du Comte de Lacépède comprenant L'Histoire Naturelle des Quadrupèdes Ovipares, des Serpents, des Poissons et des Cétacés; Nouvelle édition avec planches coloriées dirigée par M. A. G. Desmarest; Bruxelles: Th. Lejeuné, Éditeur des oeuvres de Buffon, 1836. Pl. 7

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Consell Editor / Consejo editor / Editorial Board José A. Barrientos Univ. Autònoma de Barcelona, Bellaterra, Spain Jean C. Beaucournu Univ. de Rennes, Rennes, France David M. Bird McGill Univ., Québec, Canada Mats Björklund Uppsala Univ., Uppsala, Sweden Jean Bouillon Univ. Libre de Bruxelles, Brussels, Belgium Miguel Delibes Estación Biológica de Doñana CSIC, Sevilla, Spain Dario J. Díaz Cosín Univ. Complutense de Madrid, Madrid, Spain Alain Dubois Museum national d’Histoire naturelle CNRS, Paris, France John Fa Durrell Wildlife Conservation Trust, Trinity, United Kingdom Marco Festa–Bianchet Univ. de Sherbrooke, Québec, Canada Rosa Flos Univ. Politècnica de Catalunya, Barcelona, Spain Josep Mª Gili Inst. de Ciències del Mar CMIMA–CSIC, Barcelona, Spain Edmund Gittenberger Rijksmuseum van Natuurlijke Historie, Leiden, The Netherlands Fernando Hiraldo Estación Biológica de Doñana CSIC, Sevilla, Spain Patrick Lavelle Inst. Français de recherche scient. pour le develop. en cooperation, Bondy, France Santiago Mas–Coma Univ. de Valencia, Valencia, Spain Joaquín Mateu Estación Experimental de Zonas Áridas CSIC, Almería, Spain Neil Metcalfe Univ. of Glasgow, Glasgow, United Kingdom Jacint Nadal Univ. de Barcelona, Barcelona, Spain Stewart B. Peck Carleton Univ., Ottawa, Canada Eduard Petitpierre Univ. de les Illes Balears, Palma de Mallorca, Spain Taylor H. Ricketts Stanford Univ., Stanford, USA Joandomènec Ros Univ. de Barcelona, Barcelona, Spain Valentín Sans–Coma Univ. de Málaga, Málaga, Spain Tore Slagsvold Univ. of Oslo, Oslo, Norway

Animal Biodiversity and Conservation 24.1, 2001 © 2001 Museu de Zoologia, Institut de Cultura, Ajuntament de Barcelona Autoedició: Montserrat Ferrer Fotomecànica i impressió: Sociedad Cooperativa Librería General ISSN: 1578–665X Dipòsit legal: B–16.278–58

Animal Biodiversity and Conservation 24.1 (2001)

Tasmanitachoides Erwin glabellus n. sp. from North Queensland, Australia, with a note on Tasmanitachoides lutus (Darlington) (Insecta, Coleoptera, Carabidae, Bembidiinae) M. Baehr

Baehr, M., 2001. Tasmanitachoides Erwin glabellus n. sp. from North Queensland, Australia, with a note on Tasmanitachoides lutus (Darlington) (Insecta, Coleoptera, Carabidae, Bembidiinae). Animal Biodiversity and Conservation, 24.1: 1–7. Abstract Tasmanitachoides Erwin glabellus n. sp. from North Queensland, Australia, with a note on Tasmanitachoides lutus (Darlington) (Insecta, Coleoptera, Carabidae, Bembidiinae).— A new species of the genus Tasmanitachoides Erwin from North Queensland is described: T. glabellus n. sp. The species belongs to the T. murrumbidgensis– group of species that is characterized by its distinctly impressed clypeus, but it is distinguished from all related species by its glabrous body surface. It is the first Tasmanitachoides from northern Australia to be found in rainforest on high mountains and has thus probably preserved the original habits of the genus that are still characteristic for those species living in southern temperate regions of Australia. Tasmanitachoides lutus (Darlington) so far known from the type locality in southern New South Wales and from the holotype only, is now recorded from eastern Victoria. Key words: Tasmanitachoides, T. glabellus n. sp., Australia. Resumen Tasmanitachoides Erwin glabellus sp. n. del norte de Queensland, Australia, con una mención sobre Tasmanitachoides lutus (Darlington) (Insecta, Coleoptera, Carabidae, Bembidiinae).— Se describe una nueva especie del género Tasmanitachoides Erwin del norte de Queensland: T. glabellus sp. n. Esta especie pertenece al grupo de especies T. murrumbidgensis que se caracteriza por la impresión diferenciada del clipeo, pero que se distingue de todas las especies relacionadas por presentar una superficie corporal glabra. Es el primer Tasmanitachoides del norte de Australia encontrado en la selva pluvial de alta montaña y probablemente mantiene los hábitos originales de su género, que siguen siendo característicos de las especies que viven en las zonas templadas del sur de Australia. Tasmanitachoides lutus (Darlington) conocido hasta el momento a partir de la localidad tipo, en el sur de Nueva Gales del Sur, y únicamente por el holotipo, ha sido registrado ahora en el este del estado de Victoria. Palabras clave: Tasmanitachoides, T. glabellus sp. n., Australia. (Received: 18 IX 01; Final acceptance: 4 X 01) Martin Baehr, Zoologische Staatssammlung, Münchhausenstr. 21, D–81247 München, Germany. e-mail: Martin.Baehr@zsm.mwn.de

ISSN: 1578–665X

Introduction While examining the immense bulk of rainforest carabid beetles collected during the last decades by staff at Queensland Museum, Brisbane, the author recently detected two specimens of the genus Tasmanitachoides that he was unable to identify at once. The specimens were quite unusual, because —according to the labels— they were collected near a small creek at the highest top of a rainforest–coated mountain in far northern Queensland, presumably even at or near the source of this creek. Careful examination and comparison with all related species revealed that the specimens belong to an undescribed species that is of special interest due to its habits.

Methods Description and measurements follow the style used in the author’s revision of the genus Tasmanitachoides (BAEHR, 1990). The types are shared with Queensland Museum, Brisbane (QM) and the author’s working collection in Zoologische Staatssammlung, Munich (CBM).

Studied material Genus Tasmanitachoides Erwin Erwin, 1972: 2 (ERWIN, 1972); Moore et al., 1987: 144; (MOORE et al., 1987); Baehr, 1990: 868 (BAEHR, 1990)

Type species Bembidion hobarti Blackburn, 1901; by subsequent designation. This genus of small, elongate, Perileptus–like, sand– or gravel–inhabiting ground beetles was founded by ERWIN (1972) who included those species that were combined by DARLINGTON (1962) to the “hobarti–group” within the genus Tachys s. l. BAEHR (1990) later included additional species not mentioned by Darlington or Erwin, and described further species. At present, this genus includes 16 species which are distributed through the east (including Tasmania) and tropical north of Australia including the Kimberleys in northwestern Australia. A single species (T. arnhemensis Erwin), however, apparently ranges far inland into the west of Western Australia and also into central Australia (see BAEHR, 1990: fig. 45). The genus combines some archaic bembidiine character states as enumerated in ERWIN (1972) with characters comparable with similar states in the trechine complex. Erwin regarded these similarities as remnants of an archaic prebembidiine stock, but analyses using molecular techniques seem to indicate that Tasmanitachoides indeed belongs rather to the trechine than to the bembidiine stock (Maddison, pers. comm.).

Baehr

Within the genus, according to BAEHR (1990), the dark coloured species of the T. hobarti– subgroup in its restricted sense (T. hobarti, T. leai, T. wattsense) that occur in southeastern Australia and Tasmania are most basic phylogenetically, whereas the light–coloured, more delicate species of the T. fitzroyi–group are most advanced. If this is true, then the genus originated somewhere in temperate (montane) southeastern Australia and derivative stocks later spread to open, sometimes even rather dry lowlands of the north, west and centre.

Tasmanitachoides glabellus sp. n. (figs. 1, 2) Types Holotype: }, head of Francis Ck 12km WSW Mossman, NQ 30 Dec 1989, 1200m ANZSES Expedition (QMT, 93349). Paratype: 1 }, same data (CBM). Diagnosis Distinguished by almost glabrous surface of elytra from all other species of the T. murrumbidgensis– group that is characterized by anteriorly impressed clypeus. Further distinguished from most similar T. murrumbidgensis (Sloane) of southern New South Wales by larger size; from T. fitzroyi (Darlington) of tropical Australia by dark colour of surface and dark 2 nd –4 th antennomeres; and from T. maior Baehr of southeastern Victoria by smaller size and slightly more divergent frontal furrows. Description Measurements. Length: 2.45–2.50 mm; width: 0.95 mm; ratio width/length of pronotum: 1.32–1.33. Colour. Dark piceous, anteriorly almost black, only disk of elytra with faint brownish lustre. Antenna and palpi piceous, only 1st antennomere reddish. Legs piceous, tibiae in middle slightly lighter. Head. Slightly narrower than pronotum. Surface nitid, with scattered fine punctures and highly superficial isodiametric microreticulation. Labrum anteriorly deeply impressed. Frontal furrows deep, slightly divergent and posteriorly curved outwards. Eyes large, protruding, orbits short. Mandibles short. Antenna medium–sized, median antennomeres slightly longer than wide. Pronotum. Wide, though considerably narrower than elytra. Heart–shaped, fairly convex, distinctly narrowed to base. Widest at anterior third, sides evenly convex, shortly sinuate in front of the rectangular basal angles. Base in middle produced, anterior angles slightly projecting. Median line inconspicuous, no lateral channel developed. Transverse basal sulcus deeply impressed, laterally coarsely punctate, in middle with a longitudinal furrow. Disk nitid, sparsely punctate, with highly superficial, isodiametric microreticulation. Elytra. Rather elongate, widest at about middle,

Animal Biodiversity and Conservation 24.1 (2001)

0.1 mm

Fig. 1. Tasmanitachoides glabellus n. sp.: A. Habitus, length 2.5 mm; B. } left stylomeres 1 & 2, ventral view. Fig. 1. Tasmanitachoides glabellus sp. n.: Habitus, longitud 2,5 mm; B. } estilómeros 1 y 2 izquierdos, visión ventral.

surface depressed. Inner five striae at least in basal half deeply impressed, 5th stria near base even sulcate. Sixth and 7th striae barely impressed, becoming very weak towards apex. Third stria at position of anterior discal pore characteristically outturned and shortly interrupted to meet 4th stria. Discal pores almost foveiform. Recurrent stria short. Striae anteriorly rather coarsely punctate. Inner five intervals very gently convex, sparsely and very faintly punctate, with superficial traces of microreticulation only, surface remarkably nitid. Marginal pores reduced to four behind humerus, two in middle, and two near apex within the deeply impressed submarginal sulcus that forms the apical part of 7th stria. Legs. Anterior tibia barely excised at outer edge. Aedeagus. Unknown. Female stylomeres (fig. 1). Both stylomeres very slender and elongate. Stylomere 1 dorsoventrally curved, without any setae at apical margin. Stylomere 2 almost straight, with a very elongate and a shorter nematiform seta right on apex, and one, respectively one or two shorter nematiform setae at internal and external margins close to apex. Variation. Little variation noted.

Distribution Far northeastern Queensland. Known only from type locality. Biology Very little known. According to the label, both specimens were collected at the top of a mountain at the height of 1,200 m, most probably in montane rainforest, at the edge of a creek and perhaps even at or very close to the source of this creek. Probably the habits of this species are similar to those living in montane regions of temperate southern Australia (southern New South Wales, eastern Victoria, Tasmania) where species of Tasmanitachoides likewise occur in sand or gravel of small banks at mountain creeks and small rivers. Etymology The name refers to the rather glabrous elytral surface as compared with that of similar species. Recognition The determination key in the author’s revision of Tasmanitachoides (BAEHR, 1990, p. 869–870) has been fully revised.

Baehr

Revised key to the genus Tasmanitachoides. Clave revisada del género Tasmanitachoides

Elytral striae, except sutural, reduced. Southern New South Wales, eastern Victoria Elytra with at least 5th stria marked, others sometimes superficial Clypeus distinctly impressed at middle (doubtful species under both couplets) Clypeus not impressed at middle Larger and wider species, c. 2.5 mm long or more (doubtful species under both couplets) Smaller and narrower species, < 2.3 mm long Either reasonably smaller species (< 2.5 mm long) with elytra almost lacking microreticulation; or anterior body rufous–testaceous, elytra at apex testaceous; antennae and palpi yellow; anterior angles of pronotum produced. Northern Queensland, northwestern Australia Larger species, 2.9 mm long; colour uniformly piceous; antennae and palpi dark; anterior angles of pronotum not produced. Eastern Victoria Dark species with dark antennae and palpi; elytra almost lacking microreticulation; frontal furrows little divergent. Northeastern Queensland, near rain forest bordered creek on high mountain Light species with rufous–testaceous fore body, elytra at apex testaceous, antennae and palpi yellow; frontal furrows distinctly divergent. Northern Queensland, northwestern Australia, near rivers and creeks in open to sparsely forested lowland Fore body reddish to reddish–testaceous, elytra testaceous. Northern Territory and northwestern Australia Either completely piceous or dark reddish, or fore body dark piceous and elytra dark reddish with piceous borders, suture, base, and apex. Eastern Australia from north Queensland to southern New South Wales Larger and wider species, 1.9–2.15 mm long; frontal furrows posteriorly slightly divergent; border of pronotum convex throughout to the small, projecting basal angle; 2nd–4th elytral striae less impressed. Central and far Northern Territory, northwestern Australia Smaller and narrower species, 1.65–1.95 mm long; frontal furrows parallel; border of pronotum distinctly sinuate to the right, but non–projecting basal angle; 2nd–4th elytral striae more impressed. Northwestern Australia north of Great Sandy Desert Clypeus anteriorly deeply impressed; only 1st antennomere reddish, others piceous; colour dark piceous with elytra at most feebly lighter on disk Clypeus anteriorly lightly impressed, impression sometimes difficult to see; 1st–4th antennomeres reddish, others piceous; colour either completely dark reddish, or fore body dark piceous with contrastingly lighter elytra

T. lutus (Darlington) 2 3 11 4 6

T. maior Baehr

T. glabellus n. sp.

T. fitzroyi (Darlington) 7

T. arnhemensis Erwin

T. minor Baehr

Animal Biodiversity and Conservation 24.1 (2001)

Smaller species, < 2.2 mm long; elytra markedly microreticulate. Southern New South Wales Larger species, > 2.4 mm long; elytra barely microreticulate. Far northeastern Queensland Smaller species, 1.75–2.1 mm long; eyes large, protruding, orbits almost wanting; fore body piceous, elytra lighter on disk, colour of body and antennae rather contrasting; pronotum less narrowed to base, dorsally more convex. Northeastern Queensland Larger species, 2.0–2.3 mm long; eyes smaller, less protruding, orbits perceptible, oblique; completely dark reddish to light piceous, elytra, at most, slightly lighter, colour of body and antennae not much contrasting; pronotum rather narrowed to base, dorsally more depressed. Eastern New South Wales, . Australian Capital Territory Elytra parallel, depressed; eyes small, depressed, with well developed orbits; posterior supraorbital seta situated far behind eye; mandibles very elongate, decussate; pronotum trapezoid, widest slightly behind anterior angles; colour testaceous. Northeastern New South Wales, eastern Queensland Elytra less parallel and depressed; eyes larger, more protruding, orbits small; posterior supraorbital seta situated immediately at posterior border of eye; mandibles shorter, not decussate; pronotum laterally more convex, widest far behind anterior angles; colour reddish–testaceous to black Only 1st and 5th elytral striae well impressed, others barely recognizable; elytra nitid All elytral striae present, though 2nd–4th sometimes superficial; elytra distinctly microreticulate Short, convex species; elytra considerably less than 1.5x as long as wide; frontal furrows short; pronotum wide, base (at basal angles) as wide as apex, basal angles over 90°, not projecting. Southern Queensland More elongate, less convex species; elytra more than 1.5x as long as wide; frontal furrows elongate, conspicuous; pronotum narrower, base (at basal angles) considerably narrower that apex, basal angles acute, laterally projecting Smaller species, less than 1.7 mm long; eyes less protruding, orbits perceptible; basal angles of pronotum c. 90°, less acute and projecting. Tasmania Larger species, 1.8–2.0 mm long; eyes more protruding, orbits almost reduced, basal angles of pronotum acute, less than 90°, laterally distinctly projecting. Eastern Victoria and New South Wales Elongate, depressed, very small species, 1.5–1.7 mm long; colour testaceous to light reddish. Far Northern Territory, northwestern Australia, northeastern Queensland, northeastern New South Wales More convex, larger species, 1.7–2.6 mm long; colour dark reddish to black

T. murrumbidgensis (Sloane) T. glabellus sp. n.

T. bicolor Baehr

T. rufescens Baehr

T. obliquiceps (Sloane)

12 13 15

T. wilsoni (Sloane)

T. kingi (Darlington)

T. angulicollis Baehr

T. katherinei Erwin 16

Baehr

16 On the average smaller species, 1.7–2.3 mm long, rather depressed; clypeus faintly impressed; colour either rather uniformly dark reddish, or piceous with disk of each elytron contrastingly lighter On the average larger species, 2.1–2.6 mm long, more convex; clypeus not at all impressed; colour uniformly dark piceous to black, or piceous with elytra slightly (not contrastingly) lighter 17 Smaller species, 1.75–2.1 mm long; eyes large, protruding, orbits almost wanting; fore body piceous; elytra lighter on disk, colour of body and antennae rather contrasting; pronotum less narrowed to base, dorsally more convex. Northeastern Queensland Larger species, 2.0–2.3 mm long; eyes smaller, less protruding, orbits perceptible, oblique; completely dark reddish to light piceous, elytra, at most, slightly lighter, colour of body and antennae not much contrasting; pronotum rather narrowed to base, dorsally more depressed. Eastern New South Wales, Australian Capital Territory 18 Elytral striae, including 5th, strongly impressed Elytral striae, especially 5th, rather superficial. Eastern Victoria, southern New South Wales 19 Colour uniformly dark piceous to almost black; antennae completely dark. Tasmania Colour piceous, disk of elytra slightly lighter; basal antennomeres reddish. Northeastern New South Wales

Remarks With respect to the distinctly impressed clypeus, T. glabellus clearly belongs to the T. murrumbidgensis– group within the genus Tasmanitachoides. The combination of its dark colour and almost glabrous surface, however, at once distinguishes this species from all known species. Moreover, the dark colour is also unique within all Tasmanitachoides known so far to occur in northern tropical Australia. All those species are either completely reddish or testaceous (arnhemensis, fitzroyi, katherinei, minor, obliquiceps), or are at least bicolourous with dark fore–body though lighter elytra (bicolor). In contrast, almost all of the southern species are completely dark. According to ERWIN’s (1972), DARLINGTON’s (1962: “gravel by brooks”), and the authors observations, in temperate Australia Tasmanitachoides are commonly found at small rivers and mountain brooks, even in shaded places, and commonly also at high altitudes. The species living in tropical Australia, however, are generally found in gravels and sands of lakes, rivers and creeks of open lowlands, commonly even in comparatively arid

T. bicolor Baehr

T. rufescens Baehr 19 T. wattsense (Blackburn) T. hobarti (Blackburn)

T. leai (Sloane)

regions. Here, while exposed to the bright sun, their reddish or testaceous colour corresponds well with the colour of the substratum they live on and in —namely light coloured, at most light reddish— gravels and sands. It has been postulated that the habits near streams in temperate montane regions is regarded the original mode of life for the genus Tasmanitachoides, whereas their occurrence in the tropical regions of northern Australia is secondary (BAEHR, 1990). If this assumption is true, then the occurrence of a dark coloured species living at shaded rainforest creeks in montane northern Queensland would be quite surprising, because this would mean a relict occurrence with an ancient mode of life far north of the roots of this ancient genus that most probably originated somewhere in temperate southeastern Australia. This assumption seems rather unlikely at first glance, though within recent years a number of examples of definitely southern groups were detected that have members far north in the tropics and subtropics well outside of their recognized range. Carabid examples for this

Animal Biodiversity and Conservation 24.1 (2001)

distribution pattern are two merizodine species of the genus Sloaneana Csiki which occur on Lamington Plateau of south–eastern Queensland (BAEHR, in press), or the occurrence and remakable taxonomic radiation of the psydrine genera Raphetis Moore, Sitaphe Moore, and of amblyteline Psydrinae in the wet tropics of North Queensland (unpublished records), or even the discovery of a peculiar (yet undescribed) new genus of the definitely “antarctic” subfamily Migadopinae, likewise in tropical North Queensland. It follows from these examples, which could be complemented by certain non–carabid examples, that remnants of the southern temperate “Antarctic” faunal element of Australia are still present even in tropical northern Queensland, and furthermore that this distribution pattern is probably more common than was believed to date. If related to the geographic history of Australia, these examples demonstrate that various elements of the southern fauna were somehow trapped on mountains and tablelands of eastern and north– eastern Queensland during Australia’s drift to the north during the Tertiary period. As a result, they can now be found high up in environments which —although allowing them to survive there— prevent their contact with their southern counterparts and also prevent any further spreading. When seen in the light of the biogeographical history of north–eastern Australia, the unexpected discovery of the new Tasmanitachoides adds valuable information towards understanding the complexity of the montane fauna of the wet tropics of northern Queensland.

Tasmanitachoides lutus (Darlington) Tachys lutus Darlington, 1962: 120 (DARLINGTON, 1962) Tasmanitachoides lutus, Erwin 1972: 5 (ERWIN, 1972); Moore et al. 1987: 145 (MOORE et al., 1987); Baehr 1990: 877 (BAEHR, 1990)

This remarkably and easily recognized species that lacks all but the sutural, elytral striae was

only known to date from the holotype collected at Termeil, near the coast of southeastern New South Wales. During ecological studies on riparian gravel bank arthropods carried out by V. Framenau on rivers in eastern Victoria (FRAMENAU, et al., in press) this species has been now recorded from Cann River and Castleburn Creek, both in southeastern Victoria. At both localities, a single specimen each was found on gravel banks within closed forest. New records VIC: Cann River at Chandlers Ck Bridge, 37.20 S, 149.12 E, 8 XII 1998; Castleburn Ck, Junction with Mitchell River, 37.31 S, 147.12 E, 26 XI 1998.

References B AEHR, M., 1990. Revision of the Australian ground–beetle Genus Tasmanitachoides Erwin (Insecta: Coleoptera: Carabidae: Bembidiinae), with special regard to the tropical species. Invertebr. Taxon., 4: 867–894. – (in press). Two new species of Sloaneana Csiki from southern Queensland (Coleoptera, Carabidae, Merizodinae). Mem. Queensland Mus. DARLINGTON, P. J. JR, 1962, Australian Carabid beetles XI. Some Tachys. Psyche, Cambridge 69: 117–128. ERWIN, T. L., 1972. Two new genera of Bembidiine Carabid beetles from Australia and South America with notes on their phylogenetic and zoogeographical significance (Coleoptera). Breviora, 383: 1–19. FRAMENAU, V., MANDEBACH, R., & BAEHR, M. (in press). Riparian gravel banks of upland and lowland rivers in Victoria (South East Australia): Arthropod community structure and life history patterns in a longitudinal gradient. Aust. J. Zoology. MOORE, B. P., WEIR, T. A. & PYKE, J. E., 1987. Rhysodidae and Carabidae. In: Zoological Catalogue of Australia, 4: 17–320. Australian Government Publishing Service, Canberra.

Animal Biodiversity and Conservation 24.1 (2001)

Description of Sphaericus selvagensis n. sp. from the Salvage Islands, and new data on Sphaericus bicolor Bellés (Coleoptera, Ptinidae) X. Bellés

Bellés, X., 2001. Description of Sphaericus selvagensis n. sp. from the Salvage Islands, and new data on Sphaericus bicolor Bellés (Coleoptera, Ptinidae). Animal Biodiversity and Conservation, 24.1: 9–13. Abstract Description of Sphaericus selvagensis n. sp. from the Salvage Islands, and new data on Sphaericus bicolor Bellés (Coleoptera, Ptinidae).— Sphaericus (Sphaericus) selvagensis n. sp. is described from the Salvage islands. With Sphaericus (Sphaericus) bicolor Bellés, this new species is only the second ptinid beetle reported from these islands. S. selvagensis belongs to the Sphaericus pilula group, which also includes S. bicolor. However, the transverse shape of the pronotum (with its maximal breadth near the base) and the peculiar morphology of the aedeagus, distinguish S. selvagensis from all other members of the S. pilula group. S. selvagensis lives in all the major islands of the Selvagens archipelago: Selvagem Grande, Selvagem Pequena and Ilhéu de Fora. Key words: Coleoptera, Ptinidae, Sphaericus, Salvage Islands. Resumen Descripción de Sphaericus selvagensis sp. n. del archipiélago de las Salvajes, y nuevos datos sobre Sphaericus bicolor Bellés (Coleoptera, Ptinidae).— Se describe Sphaericus (Sphaericus) selvagensis sp. n. del archipiélago de las Salvajes. Junto a Sphaericus (Sphaericus) bicolor Bellés, esta nueva especie es el segundo coleóptero ptínido registrado en esas islas. S. selvagensis pertenece al grupo de Sphaericus pilula, que también incluye S. bicolor, aunque la forma transversa del pronoto (con anchura máxima cerca de la base) y la peculiar morfología del edeago distinguen a S. selvagensis de los restantes miembros de grupo de S. pilula. S. selvagensis vive en todas las islas principales del archipiélago de las Salvajes: Salvaje Grande, Salvaje Pequeña (o Pitón Grande) y La Salvajita (Ilhéu de Fora). Palabras clave: Coleoptera, Ptinidae, Sphaericus, Islas Salvajes. (Received: 1 X 01; Conditional acceptance: 10 X 01; Final acceptance: 20 X 01) Xavier Bellés, Dept. of Physiology and Molecular Biodiversity, Inst. de Biologia Molecular de Barcelona (CID, CSIC), c/ Jordi Girona 18, 08034 Barcelona, Espanya (Spain). e-mail: xbragr@cid.csic.es

ISSN: 1578–665X

Bellés

Introduction

Description

The Salvage Islands lie in the Atlantic Ocean between the well–known archipelagos of Madeira and Canaries (BRAVO & COELLO, 1978). Up to now, the only ptinid beetle reported from the Selvagens is Sphaericus (Sphaericus) bicolor Bellés, described from Selvagem Pequena (= Pitão Island) (BELLÉS, 1982) and later recorded by E RBER & W HEATER (1987) from Selvagem Grande and Ilhéu de Fora. However, the study of the ptinid beetles collected during a campaign carried out in the Salvages in May 1999, in the context of the Project “Macaronesia 2000” of the Museo de Ciencias Naturales de Tenerife, has lead to the discovery of a new species of Sphaericus, which is described in the present paper. The data on the arthropods collected during this expedition of 1999 have been reported by ARECHAVALETA et al. (2001). The genus Sphaericus was proposed by Wollaston as early as 1854, but has been the subject of a relatively recent synopsis by BELLÉS (1994), who divided it into three subgenera: Sphaericus , the members of which are characterized by having 11–segmented antennae, 5–segmented male metatarsi, the base of the pronotum simple and the parameres of the aedeagus slender and pubescent only at the apex; Nitpus Jacquelin du Val, whose two species have 9–segmented antennae and 4–segmented male metatarsi; and Doramasus Bellés, described in the same synopsis (B ELLÉS, 1994) as similar to Sphaericus s. str. but showing the base of the pronotum protuberant and the parameres of the aedeagus robust and evenly pubescent. With the exception of Sphaericus (Sphaericus) gibboides (Boieldieu), which is anthropophilous and nearly cosmopolitan (HINTON, 1941), and Sphaericus (Sphaericus) niveus (Boieldieu), Sphaericus (Sphaericus) exiguus (Boieldieu) and Sphaericus (Nitpus ) ptinoides (Boieldieu), which are known from sparse localities in the Mediterranean area (BOIELDIEU, 1856; P IC, 1912; BELLÉS, 1994), all the other species of these three subgenera are endemic to islands of Atlantic archipelagos. The island groups include the Canaries (10 species), Madeira (nine species), Cape Verde (two species), Salvages (two species, including that described herein), and Açores (one species) (BELLÉS, 1994). More recently, the new subgenus Leasphaericus Bellés (1998) (BELLÉS, 1998) has been proposed for two Australian species. These taxa, in contrast with the Palaearctic Sphaericus, have a triangular scutellum easily visible from above. Due to the morphology of the aedeagus and the pronotum, the number of the segments in the antennae and tarsi, and the hidden scutellum, the new species described below falls into the subgenus Sphaericus Wollaston.

Sphaericus (Sphaericus) selvagensis n. sp. Types Holotype: 1{ labelled “Islas Salvajes, Selvagem Grande, 21/26–V–1999, M. Arechavaleta leg.” (Museo de Ciencias Naturales, Santa Cruz de Tenerife). Paratypes: 84 specimens of both sexes with the same label as the holotype; 6 specimens of both sexes with the label “Islas Salvajes, Selvagem Pequena, 25–V–1999, M. Arechavaleta leg.”; 1} labelled “Selvagem Pequena, Pico Veado, 21–8–70, Maul leg.”; 18 specimens of both sexes with the label “I. Selvagens, Pitão, 5–VI–1970, Maul leg.”; 1} labelled “Islas Salvajes, Ilhéu de Fora, 25–V–1999, M. Arechavaleta leg.” (Museo de Ciencias Naturales, Santa Cruz de Tenerife; Departamento de Biología Animal, Universidad de La Laguna; Museo Nacional de Ciencias Naturales, Madrid; Museu de Zoología, Barcelona; colls. Oromí, Bellés, Arechavaleta and García Becerra). Description of the male (fig. 1) Length: 1.2–1.8 mm (n = 12) Broadly oval; pronotum black, elytra dark brownish–red, appendages and sternal part of body testaceous. Head clothed with short, recumbent, golden hairs; eyes moderately convex, round, about two–thirds as broad as first antennal segment; space between antennal fossae narrow and acute; antennae 11–segmented, short (about half as long as body) and robust, segments 3 to 10 oval, slightly longer than width. Pronotum evenly convex, transverse, sides feebly arcuate, maximal breadth near base; surface with very dense to contiguous granules each about twice as large as eye facets; covered with flat, obovate, recumbent yellowish scales (easily abraded), each about same size as granules, also sparse, short and suberect dark rufo–piceous hairs; legs relatively short, robust; tarsi 5–segmented. Elytra broadly oval, sides regularly rounded, maximal breadth at middle; surface shiny, with moderately dense, irregularly distributed punctures slightly narrower than pronotal granules; covered with scales like those of pronotum (also easily abraded), but ellipsoidal in shape, also with sparse, short and recumbent yellowish irregularly distributed hairs. Scutellum much reduced, hidden from above. Aedeagus (figs. 2–3) symmetrical, with the median lobe slender, in dorsal view slightly shorter than parameres; parameres elongate, slightly broader than median lobe, with sparse, short and erect setae at apex. Description of the female The female is externally similar to the male. Distribution and habitat S. selvagensis is known from the three major islands of the Salvage Islands: Selvagem Grande,

Animal Biodiversity and Conservation 24.1 (2001)

Figs. 1–3. Sphaericus (Sphaericus) selvagensis n. sp., a typical specimen from Selvagem Grande (21/26 V 1999, M. Arechavaleta leg.): 1. Habitus; 2. Aedeaguss, dorsal view; 3. Aedeagus, lateral view. Figs. 1–3. Sphaericus (Sphaericus) selvagensis sp. n., un ejemplar típico de isla Salvaje Grande (21/26 V 1999, M. Arechavaleta leg.): 1. Habitus; 2. Edeago, vista dorsal; 3. Edeago, vista lateral.

Selvagem Pequena (= Pitão Island ) and Ilhéu de Fora. Specimens from the campaign of May 1999 were collected with pitfall traps and sifting leafmould from different plant species. Moreover, the label of the female collected by Maul at Pico Veado, in Selvagem Pequena, in August 1970, indicates that it was collected “from sifted dry foliage and leafmould under Bassia tomentosa”. In the Salvages, B. tomentosa (Lowe) (Chenopodiaceae) is a relatively rare species observed in the Selvagem Pequena, and localized only in two spots, one on the West slope of the Pico Veado and the other one in the Eastern part of the island (PÉREZ DE PAZ & ACEBES, 1978). Comparative notes Within the subgenus Sphaericus, the general shape and the typical scaliform pubescence of S. selvagensis reminds one of the species belonging to the Sphaericus gibboides and Sphaericus dawsoni groups (sensu BELLÉS, 1994). However, the narrow interantennal space distinguishes S. selvagensis from the species of these groups. Moreover, the pronotum evenly convex and the morphology of the aedeagus, especially that of the basis of the median lobe,

easily separates S. selvagensis from the species of the Sphaericus gibbicollis group (sensu BELLÉS, 1994). The new species appears to belong to the Sphaericus pilula group (sensu BELLÉS, 1994), which includes S. bicolor, previously known from the Salvage Islands. The members of this group have the interantennal space narrow, the pronotum evenly convex and the elytra irregularly punctated. In the case of S. selvagensis, the transverse shape of the pronotum (with its maximal breadth near the base) and the peculiar morphology of the aedeagus, distinguish it from all other members of the S. pilula group. The differences are well apparent between S. selvagensis and S. bicolor, as shown by the key shown below.

New data on Sphaericus bicolor Bellés, 1982 Up to now, S. bicolor was the only known ptinid species from the Salvage Islands. It was described from Selvagem Pequena (= Pitão Island), on the basis of abundant material collected in February 1976 by P. Oromí (BELLÉS, 1982). The specimens were found in leaf mould under Suaeda vera Gmelin (Chenopodiaceae) (OROMÍ et al., 1978),

Bellés

Key to the Sphaericus of the Salvage Islands. Clave para los Sphaericus de las Islas Salvajes.

Antennae long and slender, clearly longer than half the body, with the segments 2–10 subcylindrical, nearly longer than width. Pronotum longer than width, with the maximal breadth near the middle. Legs long and slender. Elytra ellipsoidal (fig. 1 from BELLÉS, 1982). Aedeagus in dorsal view with the median lobe much shorter than the parameres; parameres clearly broader than the median lobe (figs. 3–4 from BELLÉS, 1982) Antennae short and robust, about half as long as the body, with the segments 2–10 oval, slightly longer than width. Pronotum transverse, with the maximal breadth near the base. Legs short and robust. Elytra broadly oval (fig. 1, present paper). Aedeagus in dorsal view with the median lobe almost as long as the parameres; parameres slightly broader than the median lobe (figs. 2–3, present paper)

which is one of the most abundant and typical plants of the Salvages, either in the Selvagem Grande or in the Selvagem Pequena (PÉREZ DE PAZ & ACEBES, 1978). Interestingly, no specimens of S. selvagensis were collected during this 1976 campaign. Almost simultaneously, SERRANO (1983) recorded an undetermined species of Sphaericus from the Selvagem Grande (1 specimen) and Selvagem Pequena (549 specimens). Seventy–two specimens were examined by the author from this large series and all were S. bicolor. The specimen from Selvagem Grande was collected on S. vera and those from Selvagem Pequena on Elytrigia junceiforme A. et D. Löve (Poaceae) (SERRANO, 1983). A. junceiforme is relatively rare in Selvagem Pequena, being found in a single locality on the Eastern part of the island. More recently, ERBER & WHEATER (1987) have reported the identification of 89 specimens of S. bicolor from Selvagem Pequena, 4 from Ilhéu de Fora and 1 from Selvagem Grande, which had been collected by Backhuys in 1968 and deposited in the Museum of Funchal. Materials from the expedition in 1999 studied in the present work included specimens of S. bicolor mixed with the new S. selvagensis, and was collected using pitfall traps and sifting leaf mould from different plants. The number of specimens of both species present in these and in other samples studied by the author is indicated in table 1. These data suggest that any of the two species may be very abundant

S. bicolor Bellés, 1982

S. selvagensis n. sp.

Table 1. Number of specimens of Sphaericus bicolor and Sphaericus selvagensis collected in the Salvage Islands and studied by the author: M. Maul, 5 VI 1970; O. Oromí, 26/29 II 1976; S. Serrano, 20 IV–15 V 1980; A. Arechavaleta 21/26 V 1999; SP. Selvagem Pequena; SG. Selvagem Grande. (* From a total sample of 549 specimens identified by SERRANO, 1983 as Sphaericus sp., 72 were studied by the author and identified as a S. bicolor.) Tabla 1. Número de ejemplares de Sphaericus bicolor y Sphaericus selvagensis recogidos en las Islas Salvajes, estudiados por el autor: M. Maul, 5 VI 1970; O. Oromí, 26/29 II 1976; S. Serrano, 20 IV–15 V 1980; A. Arechavaleta, 21/26 V 1999; SP. Salvaje Pequeña; SG. Salvaje Grande. (* De un total de 549 ejemplares identificados por S ERRANO , 1983 como Sphaericus sp., 72 fueron estudiados por el autor e identificados como S. bicolor.)

SP M

S. selvagensis 18

S. bicolor

SG S

72(549)* 6

Animal Biodiversity and Conservation 24.1 (2001)

depending on the time and eventually on the precise site of collection. All data (BELLÉS, 1982; ERBER & WHEATER, 1987; present results) indicate that both S. bicolor and S. selvagensis are widespread in the three main islands of the archipelago: Selvagem Grande, Selvagem Pequena (= Pitão Island) and Ilhéu de Fora.

Acknowledgements Thanks are due to Pedro Oromí for critical reading of the manuscript and for sending abundant material of Sphaericus from the Salvages, especially those collected by M. Arechavaleta during the expedition of May 1999, in the context of the Project “Macaronesia 2000” of the Museo de Ciencias Naturales de Tenerife. Keith Philips also reviewed the manuscript. Artur R. M. Serrano sent a large sample of S. bicolor from Selvagem Pequena collected during the Expedição Zoológica aos Arquipélagos da Madeira e das Selvagens (30 de Abril–15 de Maio, 1980).

References ARECHAVALETA, M., ZURITA, N. & OROMÍ, P., 2001. Nuevos datos sobre la fauna de artrópodos de las Islas Salvajes. Rev. Acad. Canar. Cienc., 12(3– 4): 83–99 (2000). BELLÉS, X., 1982. El primer representante de la familia Ptinidae (Col.) de las Islas Salvajes: Sphaericus bicolor n. sp. Vieraea, 11: 103–108. – 1994. El género Sphaericus Wollaston, 1854 (Coleoptera: Ptinidae). Boln. Asoc. esp. Ent., 18: 61–79.

– 1998. A new subgenus and two new species of Sphaericus Wollaston (Coloptera, Ptinidae) from western Australia. Eur. J. Entomol., 95: 263–268. BOIELDIEU, A., 1856. Monographie des Ptiniores. Annls. Soc. ent. Fr., (3)4: 285–315, 487–504, 629–686. B RAVO , T. & C OELLO , J., 1978. Descripción geográfica del Archipiélago de las Salvajes. In: Contribución al estudio de la historia natural de las Islas Salvajes: 9–14. Aula de Cultura de Tenerife, Santa Cruz de Tenerife. ERBER, D. & WHEATER, C. F., 1987. The Coleoptera of the Selvagem Islands, including a catalogue of the pecimens in the Museu Municipal do Funchal. Bol. Mus. Mun. Funchal, 39(193): 156–187. HINTON, H. E., 1941. The Ptinidae of economic importance. Bull. ent. Res., 31: 331–381. O ROMÍ , P., B AEZ , M. & M ACHADO , A., 1978. Contribución al estudio de los artrópodos de las Islas Salvajes. In: Contribución al estudio de la historia natural de las Islas Salvajes: 178–194. Aula de Cultura de Tenerife, Santa Cruz de Tenerife. PÉREZ DE PAZ, P. L. & ACEBES, J. R., 1978. Las Islas Salvajes: Contribución al conocimiento de su flora y vegetación. In: Contribución al estudio de la historia natural de las Islas Salvajes: 79– 104. Aula de Cultura de Tenerife, Santa Cruz de Tenerife. P IC , M., 1912. Ptinidae. In: Coleopterorum Catalogus, 41: 1–46 (W. Junk & S. Schenkling, Eds.). W. Junk, Berlin. S ERRANO , A. R. M., 1983. Os coleopteros do Arquipélago das Selvagens. In: Act. I Congr. Ibérico Ent., 2: 759–776. Servicio de Publicaciones de la Universidad de León, León.

Animal Biodiversity and Conservation 24.1 (2001)

Survival of a small translocated Procolobus kirkii population on Pemba Island A. Camperio Ciani, L. Palentini & E. Finotto

Camperio Ciani, A., Palentini, L. & Finotto, E., 2001. Survival of a small translocated Procolobus kirkii population on Pemba Island. Animal Biodiversity and Conservation, 24.1: 15–18 . Abstract Survival of a small translocated Procolobus kirkii population on Pemba Island.— A survey to evaluate the distribution of Procolobus kirkii on Pemba island (Tanzania) was conducted, 20 years after they had been translocated from Zanzibar in the Ngezi forest park. A team of both expert and trained observers, guided by the authors, censused 68.3 linear km of forest, corresponding to an estimated area of 3.5 km2 (63.6%) of the protected Ngezi forested area of 5.5 km2. Nineteen groups of Cercopithecus aethiops were observed, with a total of 166 animals and an estimated density of 47.43 individuals per km2, and only one troop of Procolobus kirkii. Supplemented by interviewing the local people we obtained an estimate of 15–30 P. kirkii, including a small troop outside the protected area. This small population survived but did not increase, possibly due to adverse relations with humans. Key word: Procolobus kirkii, Translocated population, Density, Conservation, Pemba Island. Resumen Supervivencia de una pequeña población trasladada de Procolobus kirkii en la isla de Pemba.— Se realizó un estudio para evaluar la distribución de Procolobus kirkii en la isla de Pemba (Tanzania), veinte años después de que fuera trasladada desde Zanzíbar al Parque Ngezi. Un equipo de observadores expertos y entrenados, guiados por los autores, efectuó un censo a lo largo de 68,3 km lineales de bosque, correspondiente a un área estimada de 3,5 km2 (63,6%) del área protegida del bosque de Ngezi de 5,5 km2. Se observaron 19 grupos de Cercopithecus aethiops, con un total de 166 animales y una densidad estimada de 47,43 individuos/km2, y sólo un grupo de Procolobus kirkii. Complementando los datos con entrevistas a la población local se obtuvo una estimación de 15–30 ejemplares de P. kirkii, incluyendo un pequeño grupo localizado fuera del área protegida. Este pequeño grupo sobrevivía pero no se incrementaba en número, posiblemente debido a las relaciones adversas con los humanos. Palabras clave: Procolobus kirkii, Población trasladada, Densidad, Conservación, Isla de Pemba. (Received: 23 VII 01; Final acceptance: 2 X 01) Andrea Camperio Ciani(1), Loris Palentini & Enrica Finotto, Dip. di Psicologia Generale, Universita’degli Studi di Padova, 8 via Venezia, 35139 Padova, Italy. (1)

e-mail: andrea.camperio@unipd.it

ISSN: 1578–665X

Camperio Ciani et al.

Introduction

Procolobus kirkii, member of the Colobinae family, represents one of Africa’s most endangered primate species. It is mainly an arboreal and folivorous species, sympatric but not in competition with Cercopitecus aethiops which is mainly frugivorous (S IEX & STRUHSAKER, 1999a). It has been reported that to contrast the toxins contained in certain fruit P. kirkii eats a small quantity of charcoal which allow a slower, but otherwise impossible, digestion (STRUHSAKER et al., 1997). Its ideal habitats in Zanzibar are areas with ground water, swamp forest, scrub forest or mangrove swamp. Troops are numerous and can include more than 80 individuals. They have a multi–male structure which is unusual for the Colobinae family, with a 1:2 sex ratio with adult females. Fecundity is about 1.5 new–born every 2 years and infant care is intense and shared by several related females. Infanticide is common as in most Colobinae when a new male joins the group. P. kirkii is endemic and confined to the island of Zanzibar. It is present in 3 different forests with a total population of about 1,500 individuals (Zanzibar Unpublished Government Census, 1981). Two decades ago specimens were moved to new areas, mostly small islands, in order to try to inhibit their decline leading to a rapid extinction. These animals are threatened by massive deforestation and furthermore are hunted for their meat and for pet markets (STRUHSAKER & SIEX, 1996). An assessment of the present survival rate and the diffusion of the small Procolobus population (14 individuals) translocated from Jozani Park and introduced in the region of Ngezi Forest, in 1974 (STRUHSAKER & SIEX, 1998) in the north of Pemba Island is reported in this study.

Methods Data were collected from 15th–20th October 2000. To census the region as thoroughly as possible the forest was divided into 14 transects (fig. 1) varying in length from about 2 to 8 km (totally 68.3 km). Each transect segment was identified by a 1:50,000 topographical map, and located in the field with a GPS and compass. Teams included volunteers who underwent prior training in the Jozani Forest of Zanzibar to identify the different species of monkeys until consensus with the trainers reached complete agreement. Transects were walked with a fixed departure, arrival and direction. Each transect was walked by a rotating team of 3 to 4 people, scaled in experience in the field, and randomly changed each day in order to avoid individual bias in data collection (CAMPERIO CIANI et al., 2001).

Forest quality was classified into five main habitat types: gallery forest, mangrove, savannah, swamp and cultivations. To estimate the density of monkeys in each different habitat we calculated the width of our transects in each habitat. To assess the width, as for the case of transects of indefinite width (CAUGHLEY, 1977), we used the average distance at first sighting of the Cercopithecus aethiops in that habitat. Field survey was supplemented with interviews among the local people living in villages around and within the Ngezi Forest in search of witnesses and information about the presence of P. kirkii.

Results and discussion A distance of approximately 68.3 km was walked in the five various habitats inside the park. Considering the length and the relative width of our transects, during the study about 3.5 km2, 63.6% of the total forested area of the park was monitored (5.5 km2) (table 1). Sightings almost exclusively regard C. aethiops. A total of 19 troops were located from among all habitats except swamps. A total of 166 animals were observed inside the forested region (table 1), mainly sighted in the gallery forest. The estimated total density of C. aethiops in the park area is 47.43 individuals per km2. Only an elusive sighting of P. kirkii was noted, this occurring in the gallery forest in the south

Table 1. Distribution and habitat preference of C. aethiops: Tl. Transect lenght (in km); V. Visibility (in m); Nt. Number of troops; Ni. Number of individuals; D. Estimated density; * Distance not calculated because it was a sighting from the boat. Distribución y preferencia de hábitat de C. aethiops: Tl. longitud del transecto (en km); V. Visibilidad (en m); Nt. Número de grupos; Ni. Número de individuos; D. Densidad estimada; * Distancia no calculada por tratarse de una observación realizada desde el barco.

41.5

95 57.23

Mangrove

5.1

–

Savannah

8.7

25 33.42

Swamp

3.3

–

Cultivation

9.7

19 22.78

68.3

–

166 47.43

Gallery forest

Total

D – –

Animal Biodiversity and Conservation 24.1 (2001)

39o42'20''E

Transec Sighting

04o66'90''S

Research area

Fig. 1. Pemba island with insert indicating the study area and the 14 transects walked. Circles show the location of recent Procolobus kirkii sightings. Fig. 1. Isla de Pemba con el área de estudio indicada y las 14 transecciones realizadas. Los círculos indican la localización de avistamientos recientes de Procolobus kirkii.

of the Ngezi Park near Bandarikuu village with a count of three individuals. Our field observation, however, was supplemented by frequent interviews with local people regarding recent sightings of the red monkeys (as the Procolobus monkeys are known). These interviews confirmed the presence of a small troop of 5 to 7 individuals in the Bandarikuu area corresponding to our sighting. Furthermore, most people interviewed reported recent sightings in two other locations in the park: the first in the Makangale school area, in a mosaic habitat of forest and rubber plantation, with counts of 5 to 8 individuals; and a second sighting, confirmed by most interviews, indicated an area near the east section of the Wumawimbi beach in a mosaic of mangrove and gallery forest, with counts of 5 to 7 individuals. Finally, various people interviewed reported the presence of another small troop of red monkeys, 4 to 6 individuals, 6 km south of the Ngezi Forest park, in a region with abandoned clove plantation, between the town of Conde and city of Wete. The home range of Procolobus is particularly small and all these sightings are too far from each other to be the same troop shifting around (fig. 1). A small population of P. kirkii can thus be

confirmed that still survives in the Ngezi Forest of Pemba, and some individuals have even moved outside the park area. However, the estimated abundance of the whole population in the Ngezi Forest region does not exceed 15–30 units (less than 6 individuals/km2), confirming difficulties in the diffusion of these translocated Procolobus kirkii populations (STRUHSAKER & SIEX, 1998). A sympatric cohabitation with a relatively high density of C. aethiops should not be a major problem for P. kirkii which has very different dietary preferences, and favors mangrove and swamp areas little used by C. aethiops (STRUHSAKER et al., 1997). Most problems and risks for their survival and growth in number were suggested that comes from the local people, as in the case of the Jozani park population in Zanzibar (SIEX & STRUHSAKER, 1999b). In the interviews with locals, it emerged that as the result of local superstition, farmers in Pemba fear and occasionally harass this species of monkey as they are considered to bring bad luck. To promote the conservation of this beautiful, unique and elusive Colobinae population, we suggest the interest to develop awareness amongst the local people that these animals are

not only harmless but that their protection and an increase in numbers will eventually be beneficial in attracting tourists to the Ngezi Park, as occurred in Zanzibar.

Acknowledgements We wish to thank K. Siex for suggesting this study and introducing us to the Jozani Park. We thank the direction of the Ngezi Forest Park for their enthusiastic collaboration in the field. Special thanks to all the members of the GEA Pemba expedition who funded and volunteered in this project.

References CAMPERIO CIANI, A., MARTINOLI, L., CAPILUPPI, C., ARAHOU, M. & MOUNA, M., 2001. Effect of Water Availability and Habitat Quality on BarkStripping in Barbary Macaques. Conservation Biology, 15(1): 259–265. C AUGLEY , G., 1977. Analysis of vertebrate populations. John Wiley and Sons Ltd., NY.

Camperio Ciani et al.

COONEY, D. O. & STRUHSAKER, T. T., 1997. Adsorptive capacity of charcoal eaten by Zanzibar red colobus monkeys: implications for reducing dietary toxins. International Journal of Prymatology, 18(2): 235–246. SIEX, K. S. & STRUHSAKER, T. T., 1999a. Colobus monkeys and coconuts: A study of perceived human–wildlife conflicts. Journal of Applied Ecology, 36(6): 1,009–1,020. – 1999b. Ecology of the Zanzibar red colobus monkey: demographic variability and habitat stability. International Journal of Prymatology, 20(2): 163–192. STRUHSAKER, T. T., COONEY, D. O. & SIEX, K. S., 1997. Charcoal consumption by Zanzibar red colobus monkey: its function and its ecological and demographic consequences. International Journal of Prymatology, 18(1): 61–72. STRUHSAKER, T. T. & SIEX, K. S., 1996. The Zanzibar red colobus monkey Procolobus kirkii : conservation status of an endangered island endemic. African Primates, 2(2): 54–61. – 1998. Translocation and introduction of the Zanzibar red colobus monkey: Success and failure with an endangered island endemic. Oryx, 32(4): 277–284.

Animal Biodiversity and Conservation 24.1 (2001)

New molecular challenges in animal conservation X. Domingo–Roura, J. Marmi, J. F. López–Giráldez & E. Garcia–Franquesa

Domingo–Roura, X., Marmi, J., López–Giráldez, J. F. & Garcia–Franquesa, E., 2001. New molecular challenges in animal conservation. Animal Biodiversity and Conservation, 24.1: 19–29. Abstract New molecular challenges in animal conservation.— The contribution of genetics to wildlife conservation has been stressed often forgetting the existing theoretical and empirical limitations in the use of genetic information to solve ecological and demographic problems. The possibilities of molecular analyses are extensive and the automation of procedures is increasing the efficiency and reducing the cost of molecular technology. With large amounts of molecular data already available, the interest is switching towards the analysis of these data and the interpretation of genetic variability within and across species from a functional perspective. The understanding of the link between genetic variation and fitness or survival is essential in conservation biology and this understanding needs the combination of molecular data with non–molecular (e.g. physiological, behavioural and ecological) data. Progress in this promising field will depend on the trust and collaboration between molecular and field biologists. Key words: Review, Molecular techniques, Animal conservation, Fitness, Genetic variation. Resumen Nuevos retos moleculares en la conservación animal.— La contribución de la genética a la conservación de la vida salvaje ha sido enfatizada, olvidándose a menudo que existen limitaciones teóricas y empíricas sobre el uso de la información genética para solucionar problemas ecológicos y demográficos. Los análisis moleculares ofrecen numerosas posibilidades y la automatización de los procesos está incrementando la eficiencia y reduciendo los costes de la tecnología molecular. Con grandes cantidades de datos moleculares ya disponibles, el interés se está desplazando hacia el análisis de dichos datos y la interpretación de la variabilidad genética intraespecífica e interespecífica desde una perspectiva funcional. La comprensión del vínculo entre variabilidad genética y eficacia biológica o supervivencia es esencial en la biología de la conservación, requiriendo esta comprensión la combinación de datos moleculares con datos no moleculares (por ejemplo fisiológicos, de comportamiento y ecológicos). El progreso en este campo tan prometedor debe basarse en la confianza y la colaboración entre biólogos moleculares y de campo. Palabras clave: Revisión, Técnicas moleculares, Conservación animal, Eficacia biológica, Variación genética. (Received: 17 IX 01; Final acceptance: 10 X 01) Xavier Domingo–Roura(1), J. Marmi & J. F. López–Giráldez, Unitat de Biologia Evolutiva, Dept. de Ciències Experimentals i de la Salut, Univ. Pompeu Fabra, Dr. Aiguader 80, 08003 Barcelona, Espanya (Spain).– Xavier Domingo–Roura, Wildlife Conservation Research Unit, Dept. of Zoology, Univ. of Oxford, South Parks Road, Oxford OX1 3PS, UK.– Eulàlia Garcia–Franquesa, Museu de Zoologia de Barcelona, Passeig Picasso s/n, 08003 Barcelona, Espanya (Spain). (1)

e–mail: xavier.domingo@cexs.upf.es

ISSN: 1578–665X

Rationalising the use of molecular biology The current diversity of molecular techniques offers a wide range of possibilities to support decision makers, and genetic studies are becoming a primary argument in wildlife conservation. The importance of genetic variation in biodiversity evaluation has been recognised (EHRLICH & WILSON, 1991). Molecular biology tools have already been used to guide expensive conservation programs, including risky reintroduction projects (e.g. brown bear Ursus arctos [TABERLET & BOUVET, 1994]; bearded vulture Gypaetus barbatus [NEGRO & TORRES, 1999]). The protection of genetic diversity has been incorporated into national and international legislation. To optimise the use of molecular biology in conservation, a wise rationalisation of the techniques and a realistic interpretation of the data produced are needed. Technological seduction and the availability of numerous informative techniques should not interfere with the recognition of the actual limitations of these techniques, both in the theoretical ground and in supporting the real problems that nature is facing (HEDRICK, 1996). For instance, it is important to recognise that molecular information might not be as critical for the immediate survival of a species as improving its habitat (CAUGHLEY, 1994) and reducing the exploitation of natural resources in this habitat (BEGON et al ., 1999). Current limitations are also evident from the recognition, for instance, that no agreement has yet been reached on how to incorporate genetic diversity into land–use planning (MORITZ & FAITH, 1998). It is also important to note that special care needs to be taken before reaching management conclusions in endangered species, where in spite of the urgency implied, erroneous recommendations could be detrimental to a species and ecosystem. Recommending the separate management of already–reduced populations could promote inbreeding. Proposing population intermixing could promote the hybridisation of specific adaptations to a particular environment (WAYNE et al., 1994). In this work, the wide variety of molecular techniques available to support wildlife management are reviewed and relevant examples are provided in order to better understand when these techniques are used (table 1). The gap that exists between technological possibilities and their use can thus be recognized to interpret the complexity of life is noted. Finally, molecular and non–molecular biologists are appealed to collaborate in tracing the link between genes and adaptation so as to progress in many fields of life sciences including conservation biology.

Information contained in the DNA Variation at a given DNA region is a consequence of evolutionary forces such as mutation, selection,

Domingo–Roura et al.

genetic drift or recombination that have acted over the DNA and the species (GRAUR & LI, 2000; BERTRANPETIT, 2000). Within and across populations and species the coalescense of genomic regions can be traced back and the time when genes or genome separated can be infered. Similarity relationships between DNA segments can also be evaluated to infer relationships between genes, individuals and groups of individuals. If we compare derivative characters with their geographic distribution, we can infer gene flow and colonisation events. In addition, the distribution of alleles and the structure of the genetic variation might be used to infer demographic parameters such as population size and subdivisions (LUIKART & ENGLAND, 1999). A wide variety of polymorphic DNA regions with different mutation patterns and rates have been recognised. The choice of one or another region will depend on the objectives of our research. Most nuclear genome regions are diploid and inherited in an autosomal and codominant fashion affected by recombination. They can code for RNA or be non–coding regions. In wildlife studies, microsatellites or STRs have been widely used (QUELLER et al., 1993; LUIKART & ENGLAND, 1999). They consist of a short string of one to ten base pairs repeated in tandem and are dispersed throughout the genome. They are highly polymorphic due to the variation in the number of repeat units and most behave as neutral markers. Minisatellites are also tandemly repeated strings of longer repeat units (JEFFREYS et al., 1985). The number of repeats is inherited and variable among individuals. This variability can be detected with a probe that will attach to a single or several complementary DNA fragments among all DNA fragments distributed through an electrophoresis gel, providing a pattern of bands for comparison. Some microsatellites and minisatellites are associated with mobile genetic elements, another DNA class that is currently gaining support for phylogenetic inference (BUCHANAN et al., 1999). These mobile or interspersed elements of different families and subfamilies occur throughout the genome. Short Interspersed Elements (SINEs) are excellent markers for molecular phylogeny since their integration at a particular position in the genome can be considered an unambiguous derived homologous character (TAKAHASHI et al., 1998). Mitochondrial DNA (mtDNA) sequences include the other major group of markers widely used in wildlife analyses (AVISE , 1994). Mitochondrial DNA is haploid, recombination free and maternally inherited. It has a low frequency of insertion, deletion and duplication events and an evolutionary rate 5–10 times higher than single copy nuclear genes (BROWN et al., 1979). Conclusions in animal conservation should be supported by the analyses of several independent data sets (WAYNE et al., 1994). If we use different

Animal Biodiversity and Conservation 24.1 (2001)

Table 1. References with examples on the applications of molecular biology technologies to wildlife management and conservation. Tabla 1. Referencias con ejemplos de las aplicaciones de tecnologías de biología molecular a la gestión y conservación de la vida salvaje.

Technique Allozymes Reference: MERENLENDER et al. (1989) Purpose: quantification of genetic variation and differentiation in African rhinoceroses (Ceratotherium simum and Diceros bicornis) and Asian rhinoceroses (Rhinoceros unicornis) Results: low levels of intraspecific variation found below the levels expected in comparisons among subspecies RFLPs Reference: WATKINS et al. (1988) Purpose: quantification of Major Histocompatibility Complex (MHC) polymorphism in cotton–top tamarin (Saguinus oedipus ) Results: very low levels of polymorphism found in its MHC class I DNA Fingerprinting Reference: PACKER et al. (1991) Purpose: study of the kinship structure in lion (Panthera leo) social groups Results: female within the same group are closely related, whereas males can be either related or unrelated. Reproductively active males are usually unrelated to group females. Males only act as non–reproductive helpers in coalitions composed of close relatives Sequencing Reference: BAKER et al. (2000) Purpose: determine the origins of whale products purchased from markets in Japan and the Republic of South Korea Results: some protected species, such as baleen whales and sperm whales, were identified among the commercial products analysed SSCP & Sequencing Reference: SHAFFER et al. (2000) Purpose: screening population structure and identification of management units in Yosemite toad (Bufo canorus) Results: different genetic substructure and no shared haplotypes among animals from Yosemite and Kings Canyon National Parks. Animals from the two parks should be managed as different units RAPDs, DGGE & Sequencing Reference: NORMAN et al. (1994) Purpose: analysis of population structure and identification of management units in green turtles (Chelonia mydas) Results: Indo–Pacific rookeries include a number of genetically differentiated populations, with minimal female–mediated gene flow among them RAPDs Reference: NEVEU et al. (1998) Purpose: comparison of the genetic diversity of wild and captive populations of mouse lemur (Microcebus murinus) Results: captive groups have lost genetic variation in comparison with wild groups

Domingo–Roura et al.

Table 1. (Cont.)

Technique AFLPs Reference: GIANNASI et al. (2001) Purpose: exploring the possibilities of AFLPs for phylogenetic reconstruction in the snake Trimeresurus albolabris Results: T. albolabris is not monophyletic Microsatellite analysis Reference: CIOFI & BRUFORD (1999) Purpose: assess the level of genetic variability and gene flow among populations of Komodo dragon (Varanus komodoensis) Results: high levels of genetic diversity and gene flow between Rinca and Flores Islands, highest levels of genetic divergence in Komodo Island and low levels of genetic variability and gene flow in Gili Motang Island Microarrays Reference: TROESCH et al. (1999) Purpose: genotyping and identification of Mycobacterium species Results: the array can identify species within the genus Mycobacterium and detect drug– resistance Minisequencing Reference: MORLEY et al. (1999) Purpose: assay the effectivity of fluorescent minisequencing of mtDNA for forensic use in animal, bacterial and fungal species extracts Results: the technique is reliable, reproducible and suitable for forensic uses in a wide range of organisms Quantitative PCR Reference: FELDMAN et al. (1995) Purpose: detection of malaria infection in Hawaiian birds Results: avian malaria was more widespread in Hawaii than previously thought

types of molecular data with different mutation rates we might be able to separate ancient from recent events. Another alternative is the comparison of male–inherited DNA regions (i.e. non– recombining regions of the Y–chromosome) versus female–inherited DNA regions (such as mitochondrial DNA) to understand the contribution of each sex in determining genetic diversity (MELNICK & HOELZER, 1992; PÉREZ–LEZAUN et al., 1999). This analysis can contribute to understanding how a balance is achieved between the proportion of individuals leaving the natal area and the proportion remaining philopatric to minimise inbreeding and resource competition (GOMPER et al., 1998). To identify individuals, populations or species it is often recomended to work with genetic markers that are neutral and therefore good indicators of ancestry or relationship (HEDRICK, 1996).

However, there is some concern regarding how neutral characters obtained from non–coding regions reflect the diversity of functional attributes (WILLIAMS et al., 1994; LYNCH, 1996).

Technology available The main goal of molecular techniques is to detect the variation in DNA sequences, directly through sequencing or indirectly through other methods sensitive to sequence variations. This variation can be detected using a wide range of techniques. A first group of techniques including isozymes and restriction fragment length polymorphisms (RFLP) is based on the differential mobility of proteins and DNA fragments respectively (due to their different charge or size)

Animal Biodiversity and Conservation 24.1 (2001)

in an electrophoretic field (MÜLLER–STARK, 1998; BRETTSCHNEIDER, 1998). Hybridisation between a labeled DNA fragment or probe and a target DNA is the principle involved in many other techniques (SAMBROOK et al., 1989). With the discovery of the polymerase chain reaction (PCR) (SAIKI et al., 1988), a new wave of molecular techniques appeared. One important advantage of the PCR is that a given DNA fragment can be isolated and copied millions of times reliably and quickly using temperature cycles and a thermally stable polymerase. This allows the use of minute amounts of DNA in molecular studies, such as those obtained from biological remnants obtained non-invasively (WOODRUFF, 1993). Sequencing The complete sequencing of the whole genome is the most detailed method to detect genetic variability. However, sequencing complete genomes is tedious and expensive and most studies rely on the sequencing of a minute portion of the genome and the assumption that variation within the fragment sequenced represents the variation along the whole genome. Sequencing of PCR products of up to several hundred base pairs is a widely used methodology in life sciences. During the sequencing reaction of a PCR product, a large number of fragments differing by a nucleotide in length and with the last base labelled with a specific fluorochrome depending on its identity are obtained (WEAVER & HEDRICK, 1992). When these sequencing products of different length are electrophoresed in a DNA sequencer, the ladder of fluorochrome signals obtained will indicate the nucleotide sequence of the PCR product under analysis. It is common practice to deposit the sequences obtained in public databases, facilitating both the comparison and complementation of one’s own data with the data from the same or other species obtained by other researchers. Sequencing can be combined with other methods to reduce its cost. A first group of PCR– based methods (Heteroduplex analysis, Single Strand Conformation Polymorphisms, Denaturing Gradient Gel Electrophoresis and Temperature Gradient Gel Electrophoresis) consists of screening techniques for detecting sequence variation in PCR products of identical sizes, without the need to go through sequencing. These protocols are based on the physical behaviour of DNA during electro– phoresis in acrylamide gels. The use of these methods is adequate when dealing with a large number of samples and when alleles are shared by many individuals (LESSA & APPLEBAUM, 1993). Heteroduplex analysis Heteroduplex analysis starts with the denaturing of the PCR product at 95ºC and its subsequent renaturation before electrophoresis (L ESSA &

A PPLEBAUM, 1993). Using this technique it is possible to distinguish between homozygous and heterozygous DNA fragments. If a sample contains two different alleles, heteroduplex molecules (hybrids of the two strands belonging to different alleles) are obtained. Since these heteroduplexes have one or more mismatches in their double strands, they migrate onto the gel more slowly than the homoduplex molecules obtained from the hybridization of strands containing the same allele. Single–Strand Conformation Polymorphism (SSCP) SSCP is a simple and fast method for screening DNA fragments for nucleotide sequence polymorphisms. PCR products that have been denatured by temperature and/or chemicals are loaded and run onto a non–denaturing polyacrylamide gel. The electrophoretic mobility of each single–stranded DNA fragment depends on its secondary structure, which in turn depends on its nucleotide sequence (JORDAN et al., 1998). SSCP can distinguish DNA fragments that differ only by one base-pair substitution in a fragment of up to several hundred nucleotides (ORITA et al., 1989). Denaturing Gradient Gel Electrophoresis (DGGE) and Temperature Gradient Gel Electrophoresis (TGGE) DGGE and TGGE work over double stranded DNA. In these methods, PCR products are loaded onto a polyacrylamide gel and run in a linear gradient of concentration of denaturing solvents (urea, formamide) or temperature respectively (LESSA & APPLEBAUM, 1993). The point along the gradient where the DNA fragment is partially denatured is called the melting point. This point depends on the overall base composition and the interactions across the molecule and can be modified by point mutations that will be reflected in the gel. Randomly Amplified Polymorphic DNAs (RAPDs) and Amplified Fragment Length Polymorphisms (AFLPs) The principle of the RAPD technique is the simultaneous amplification of DNA regions by using a single randomly chosen primer which acts as both forward and reverse (GROSBERG et. al., 1996). This primer is able to hybridise with many sites of target DNA, but amplification only occurs when the primer anneals at two sites on opposite strands separated by a reasonable distance for the PCR to work (20 to 2000 bp). These fragments are then separated in an electrophoresis gel and stained with chemicals such as ethidium bromide or silver nitrate. The gels can be scored as the presence or absence of a band of a specific molecular weight. Bands of different sizes usually represent independent loci. RAPDs are treated as neutral and anonymous markers, can be generated quickly and a large number of individuals can be processed in a

Domingo–Roura et al.

short time. However, results are difficult to repeat, a band can contain more than one amplification product that can not be distinguished and it is difficult to estimate allelic frequencies because homozygotes can not be distinguished from heterozygotes. In addition, it is sometimes difficult to know whether the variation is neutral or whether it follows Mendelian inheritance. In AFLPs, genomic DNA is digested with restriction enzymes and the goal is to reduce the complexity of the initial mixture of fragments. To achieve this reduction a subset of fragments is biotinylated and selected by union to streptavidin-coated paramagnetic beads (since biotine binds covalently to streptavidine) (MATTHES et al., 1998). The unbound fragments are washed and discarded. A subset of the biotinylated fragments is then amplified by PCR to further reduce complexity. Finaly, PCR products are analysed by denaturing polyacrilamide gel electrophoresis and revealed by autoradiography. AFLPs are more informative and easier to reproduce than RAPDs.

Automation required Automation is a key issue in molecular biology and the machinery used in the automated analyses of humans and model animals is later adapted to wildlife research. Automated procedures are currently used for standard procedures such as DNA isolation or library construction and spotting but also for the fast scoring of genetic variability among individuals with technologies such as microarrays or quantitative PCR. Microsatellite multiplexing Several microsatellite loci can be amplified in a single PCR reaction containing different primers (GILL et al., 1995). The primers are labeled with different fluorochromes and amplify fragments of different lengths. When the multiplex PCR reaction is run in an automated sequencer it is possible to sequentially detect the length of the different PCR products corresponding to the alleles of the different microsatellite loci. DNA array technology A DNA array consists of up to thousands of DNA strings attached in order over a solid support (SOUTHERN et al., 1999). An unknown sample is passed over the array and it will hybridize upon the immobilised probes when finding a complementary sequence. The reverse is also possible when a known probe hybridises upon unknown immobilised fragments. The full microarray equipment consists of a machine to produce the array and a machine with a fluorescence laser scanner to read the signal and translate this signal to a computer. The great

advantage of microarray technology is that it allows the fast detection of sequence information from a large number of loci or individuals at the same time. Paradoxically, one of the main problems encountered with microarray technology is that it generates such a large amount of information that results are often difficult to interpret. Microarrays are used, for instance, to monitor RNA expression and gene function (DE SAIZIEU et al., 1997; WODICKÁ et al., 1997; CHO et al. , 1998) or to detect single nucleotide polymorphisms (SNPs) (CHAKRAVARTI, 1999). All studies published used model species and, as far as we know, no study using microarray technology has yet been performed in any species with a conservation perspective. Minisequencing The technique consists of a PCR–based minisequencing reaction where the polymerase adds a single nucleotide. Primers finalise just before the polymorphic position that needs to be interrogated. The polymerase extends the first base position after the primer with labelled new nucleotides and the identity of the incorporated nucleotide can be determined with an automated sequencer. Several reactions can be performed simultaneously with primers of different sizes. It is also possible to conduct a minisequencing reaction in a DNA array (HACIA, 1999; RAITIO et al., 2001). Quantitative PCR Quantitative PCR consists of a reaction that detects and quantifies nucleic acid sequences either as a final product or while the reaction is being produced. The protocol is based on the detection of fluorescence emitted by the degradation of an internal labelled oligo complementary to our sample when the PCR proceeding is being produced. The outcome is the quantification of a PCR product that can be used in gene expression studies (DE KOK et al., 2000), to evaluate viral load (LIMAYE et al., 2001), and to detect transgenes (FAIRMAN et al., 1999), duplications and deletions (AARSKOG & VEDELER, 2000; WILKE et al., 2000) and SNPs (BREEN et al., 2000). Quantitative PCR and minisequencing can be cheaper alternatives to microarrays for the study of SNPs if a moderate number of SNPs and individuals are to be analysed.

Looking for the link between molecular data and conservation Technological resources are available, but the connection between molecular variability and the needs of endangered species is not straightforward. Gene dynamics is complex, most phenotypic characters are multigenic, and the

Animal Biodiversity and Conservation 24.1 (2001)

genetic machinery is loaded with complicated gene interactions and epistases (HEDRICK, 1996). More than one protein can be translated from a single gene due to alternative splicing (GRAUR & LI, 2000). Genetic linkage can also mask the role of important genes. In addition, the relationship between gene and environment is often difficult to discern (FALCONER, 1989). All levels of life expression and population processes are complex and manifold and quick fixes to animal management questions based on simple molecular biology analyses should be avoided. Genetic diversity has been linked to species richness and to better chances to cope efficiently with enviromental change (HEDRICK & MILLER, 1992; O’BRIEN et al., 1985). Consanguineous matings promote the existence of deleterious genes in homozygosis, which can be detrimental for survival and reproduction. In theory, fitness in small populations will decline due to the accumulation of detrimental mutations (LYNCH et al., 1995a, 1995b). However, the importance of genetic variability for species survival is not clearly defined. In practice, at least some populations can survive in spite of having low genetic variability (e.g cheetahs [ Acinonyx jubatus ] [O´B RIEN et al. , 1987], mole–rats [Heterocephalus glaber] [FAULKES et al., 1990; REEVE et al., 1990] and Eurasian badgers [Meles meles] [DOMINGO–ROURA, 2000]). The empirical relationship between genetic distance and fitness is likely to be species–specific and is unlikely to be linear (LYNCH, 1991). In the past, considerable effort has been devoted to describe key demographic numbers required to maintain the necessary genetic variability needed for species survival (SOULÉ, 1987). However, key numbers are unlikely to be applicable across populations or habitats. In the last decade, polemics concerning the existence of key numbers for survival have often given way to other discussions, not often based on molecular information. Within species, conservation strategies have been proposed on the basis of the existence of Evolutionary Significant Units (ESUs) which have been defined as population units that merit separate management and have high priority for conservation (RYDER, 1986). The use of ESUs in conservation has signified an upgrade from previous strategies that only gave importance to individual numbers without considering dif– ferences among individuals of the same species. However, a compromise has not yet been reached regarding the relative importance of ecological adaptation and genetic variability to determine these units (MACE et al., 1996). Furthermore, the inadequacies of the dichotomy implied in the ESU concept in a world ruled by a continuum of population differentiation have been noted (CRANDALL et al., 2000). Across species, molecular techniques are also at the base of new strategies to support an integrated approach to conservation, focusing on

the preservation of evolutionary diversity instead of focusing on species number (MAY, 1990; MACE et al., 1996) or single–species management. In this case, molecular data should play a predominant role in the selection of areas that contain evolutionarily distant lineages and areas of potential evolutionary novelty, such as multispecies contact zones (MORITZ & FAITH, 1998). The protection of these areas is likely to preserve large amounts of evolutionary heritage and will maximise the evolutionary–response potential to perturbations (PETIT et al., 1998). In fact, discussing key numbers and even single–species conservation strategies might be naif in the face of the immense complexity of nature.

Looking for the link between gene and function The rapid development of molecular genetics for biomedical and industrial purposes facilitates the access to molecular technology. Resolution is also increased with new techniques and a higher number of markers. The increase in the number of markers known in any species means greater probabilities to detect major loci that influence quantitative traits. As we learn more about DNA, molecular information will be better understood when used in combination with physiological, demographic, ecological and behavioural data collected in the field (HAIG, 1998). Data can originate from any parameter that can group individuals in relation to their evolutionary origin and/or ecological needs. In animal conservation it is not enough to understand and describe molecular variation or even ecological and demographic characteristics using molecular tools. We need to find loci that have variants that are responsible for low fitness and survival. Ecologically relevant heritable traits might need to be emphasised (CRANDALL et al., 2000). Nevertheless, fitness measurement might be difficult in endangered species. Since a selective difference smaller than the reciprocal of twice the effective population size (1/2Ne) is effectively neutral (KIMURA, 1979), small selective differences are unlikely to be of adaptive significance in most endangered species. A further complication arises from the possible differences between former and current selection and adaptation processes. The habitat currently used by a rare species can be marginal and might no longer reflect the environmental condition in which the traits evolved (JOHNSON et al., 2000). This is especially true for carnivores since human expansion has considerably altered their distribution and ecology (GRIFFITHS & THOMAS, 1993). To unravel the link between gene and function or adaptation is not a goal exclusive to conservation genetics. For instance, to clarify the function of genes that are likely to be responsible for diseases is a major enterprise in

Domingo–Roura et al.

current biomedicine. Since the link between genotype and phenotype is still widely unknown, the potential of molecular biology in wildlife management and conservation is still at a very early stage. At this point, still far from applying the functional interpretation of genetic variation to wild species, to advance the understanding of enhanced fitness and the evolutionary paths of physiological systems, several approaches can be considered. The genetic structure of a population can be examined to identify physiological phenotypes with highest fitness. Interindividual variation can be used to identify physiological, biochemical and molecular characters that correlate with fitness and survival. Comparative studies to trace the evolution of characters with particular phylogenies can also be useful to understand the role of these characters in radiation and extinction. Other important approaches are the experimental manipulation of genes through genetic engineering, and the experimental manipulation of the environment through controlled laboratory conditions and imposed selection pressures. Even if many wildlife biologists dislike strategies such as genetic engineering, manipulative experiments and keeping animals in captivity, scientific progress will certainly be slower and may be incomplete without using these more aggressive approaches. Unfortunately, economic progress and habitat deterioration is unlikely to be slow or incomplete.

More data, new trends The current trend towards automation and robotisation can create important shifts in the focus of wildlife research. Highly automated laboratories are expensive but open the possibility of subcontracting services to specialised companies which can offer the same protocols than a university researcher could conduct in his or her laboratory in a cheaper, faster and often more reliable way. Some biotechnology companies are even taking a further step and sequencing interesting regions, such as regions responsible for main human diseases, in a large number of individuals, with no previous order, and selling the use of the sequences as a product. Not only this, but since many journals and common sense require sequence data be deposited in public databases, the amount of sequence data is increasing steadily. Laboratory technicians can be accurately trained to develop protocols complementary to the services offered by specialised companies. The trend is switching from young researchers who can run molecular protocols towards young researchers who can analyse molecular data generated by others. Even if this high throughput trend makes better sense when considering human molecular biology, the amount of DNA sequences from wild animals that can be found in public databases is already amazing. In addition, the dog has been suggested to be a

good model for identifying the genetic control of morphologic characteristics in mammals (WAYNE & O STRANDER , 1999). The sequencing of whole genomes for conservation purposes has not yet begun. However the proposal to start sequencing the genome of chimpanzees or macaques to understand genetic and functional differences between humans and other primates (MCCONKEY & VARKI, 2000) is likely to see the light soon. When we leave molecular dynamics and start dealing with gene–environment interactions and adaptive characters, knowledge in other biological sciences such as ecology, zoology and behaviour becomes essential. Molecular differences have to be contrasted against non–molecular data in, for instance, geography, behaviour, morphology, and function. Applications in animal conservation only make sense when compared to field data, even if initially these data are just the species name (not always easy to determine) and the geographic origin of a sample. Accurate field data can also considerably improve the resolution of experiments. As noted by MACE et al. (1996), many studies that attempt to reconstruct familial relationships from molecular data are unable to resolve the relationship fully, even if this might have been feasible had observations been made on the breeding population to reduce the set of uncertainties for analysis. It would be great for the progress of biological sciences and, thus, for animal conservation if samples and field data could be as easily accessible as DNA sequences to the general public. This lack of availability of samples and data creates drawbacks such as the need to spend long periods of time searching for the material required, for instance, to review the phylogeny of a doubtful taxon. This time could be devoted to more fruitful tasks if the material and accompanying data were readily available from museums and other, often public, specialized institutions. In this regard, the link between non–molecular and molecular databases is becoming an urgent need. Successful experiments should be based in the future on a justified trust and collaboration between field and laboratory biologists. The molecular trend of research during recent years and the ease and speed with which molecular data can sometimes be published might have worked against the funding of field projects and of projects in many other areas of biology crucial for conservation biology. This trend will certainly need to be reviewed in the future when we try to translate molecular data back to nature.

Acknowledgements We thank Francesc Calafell and Luis Pérez–Jurado for their helpful comments to improve the manuscript. J. Marmi and J. F. López–Giráldez are supported by scholarships from the

Animal Biodiversity and Conservation 24.1 (2001)

Departament d’Universitats, Recerca i Societat de la Informació, Generalitat de Catalunya (Refs. 2000FI–00698 and 2001FI–00625 respectively).

References AARSKOG, N. K. & VEDELER, C. A., 2000. Real–time quantitative polymerase chain reaction. A new method that detects both the peripheral myelin protein 22 duplication in Charcot– Marie–Tooth type 1A disease and the peripheral myelin protein 22 deletion in hereditary neuropathy with liability to pressure palsies. Hum. Genet., 107(5): 494–498. AVISE, J. C., 1994. Molecular markers, natural history and evolution. Ed. Chapman and Hall, London. BAKER, C. S., LENTO, G. M., CIPRIANO, F. & PALUMBI, S. R., 2000. Predicted decline of protected whales based on molecular genetic monitoring of Japanese and Korean markets. Proc. R. Soc. Lond. B. Biol. Sci., 267: 1,191–1,199. BEGON, M., HARPER, J. L. & TOWNSEND, C. R. (Eds.), 1999. Ecología. Individuos, poblaciones y comunidades. Ed. Omega, Barcelona. BERTRANPETIT, J., 2000. Genome, diversity, and origins: the Y chromosome as a storyteller. Proc. Natl. Acad. Sci. USA, 97: 6,927–6,929. BREEN, G., HAROLD, D., RALSTON, S., SHAW, D. & ST C LAIR , D., 2000. Determining SNP allele frequencies in DNA pools. Biotechniques, 28: 464–466, 468, 470. B RETTSCHNEIDER , R., 1998. RFLP analysis. In: Molecular tools for screening biodiversity: 83– 85 (A. Karp, P. G. Isaac & D. S. Ingram, Eds.). Ed. Chapman & Hall, Cambridge. BROWN, W. M., GEORGE, M. JR. & WILSON, A. C., 1979. Rapid evolution of animal mitochondrial DNA. Proc. Nat. Acad. Sci. USA, 76: 1,967–1,971. BUCHANAN, F., CRAWFORD, A., STROBECK, C., PALSBOLL, P. & PLANTE, Y., 1999. Evolutionary applications of MIRs and SINEs. Anim. Genet., 30: 47–50. CAUGHLEY, G., 1994. Directions in conservation biology. J. Anim. Ecol., 63: 215–244. C HAKRAVARTI , A., 1999. Population geneticsmaking sense out of sequences. Nat. Genet. Suppl., 21: 56–60. CHO, R. J., FROMONT –RACINE, M., WODICKÁ, L., FEIERBACH, B., STEARN, T., LEGRAIN, P., LOCKHART, D. J. & DAVIS, R. W., 1998. Parallel analysis of genetic selections using whole genome oligonucleotide arrays. Proc. Natl. Acad. Sci. USA, 95: 3,752–3,757. CIOFI, C. & BRUFORD, M. W., 1999. Genetic structure and gene flow among Komodo dragon populations inferred by microsatellite loci analysis. Mol. Ecol., 8: S17–S30. CRANDALL, K. A., BININDA–EMONDS, O. R. P., MACE, G. M. & W AYNE, R. K., 2000. Considering evolutionary processes in conservation biology. Trends Ecol. Evol., 15: 290–295. DE KOK, J. B., RUERS, T. J., VAN MUIJEN, G. N., VAN BOKHOVEN, A., WILLEMS, H. L. & SWINKELS, D. W.,

2000. Real–time quantification of human telomerase reverse transcriptase mRNA in tumors and healthy tissues. Clin. Chem., 46: 313–318. DE SAIZIEU, A., VANKAN, P., VOCKLER, C. & VAN LOON, A. P., 1997. The trp RNA–binding attenuation protein (TRAP) regulates the steady–state levels of transcripts of the Bacillus subtilis folate operon. Microbiology, 143: 979–989. D OMINGO –R OURA , X., 2000. Biochemical and molecular ecology of Eurasian badgers, Meles meles, and Japanese macaques, Macaca fuscata: Nutritional and reproductive requirements. Ph. D. Thesis, Univ. of Oxford. EHRLICH, P. R. & WILSON, E. O., 1991. Biodiversity studies: science and policy. Science, 253: 758–761. FAIRMAN, J., ROCHE, L., PIESLAK, I., LAY, M., CORSON, S., FOX, E., LUONG, C., KOE, G., LEMOS, B., GROVE, R., F RADKIN , L. & V ERNACHIO , J., 1999. Quantitative RT–PCR to evaluate in vivo expression of multiple transgenes using a common intron. Biotechniques, 27: 566–570, 572–574. FALCONER, D. S., 1989. Introduction to quantitative genetics. 3rd Edition. Ed. Longman Scientific and Technical, Hong Kong. FAULKES, C. G., ABBOTT, D. H. & MELLOR, A. L., 1990. Investigation of genetic diversity in wild colonies of naked mole–rats (Heterocephalus glaber) by DNA fingerprinting. J. Zool., Lond., 221: 87–97. FELDMAN, R. A., FREED, L. A. & CANN, R. L., 1995. A PCR test for avian malaria in Hawaiian birds. Mol. Ecol., 4: 663–673. GIANNASI, N., THORPE, R. & MALHOTRA, A., 2001. The use of amplified fragment length polymorphism in determinng species trees at fine taxonomic levels: analysis of a medically important snake, Trimeresurus albolabris. Mol. Ecol., 10: 419–426. GILL, P., KIMPTON, C. P., URQUHART, A., OLDROYD, N., MILLICAN, E. S., WATSON, S. K. & DOWNES, T. J., 1995. Automated short tandem repeat (STR) analysis in forensic casework – a strategy for the future. Electrophoresis, 16: 1,543–1,552. GOMPER, M. E., GITTLEMAN, J. L. & WAYNE, R. K., 1998. Dispersal, philopatry, and genetic relatedness in a social carnivore: comparing males and females. Mol. Ecol., 7: 157–163. GRAUR, D. & LI, W. H., 2000. Fundamentals of molecular evolution . 2 nd ed. Ed. Sinauer, Sunderland, Massachusetts. GRIFFITHS, H. I. & THOMAS, D. H., 1993. The status of the badger Meles meles L., 1758 (Carnivora, Mustelidae) in Europe. Mammal Rev., 23: 17–58. GROSBERG, R., LEVITAN, D. & CAMERON, B., 1996. Characterization of genetic structure and genealogies using RAPD–PCR markers: a random primer for the novice and nervous. In: Molecular zoology: advances, strategies and protocols: 67–100 (J. D. Ferraris & S. R. Palumbi Eds.). Ed. Wiley–Liss, New York. HACIA, J. G., 1999. Resequencing and mutational

analysis using oligonucleotide microarrays. Nat. Genet. Suppl., 21: 42–47. HAIG, S. M., 1998. Molecular contributions to conservation. Ecology, 79: 413–425. HEDRICK, P. W., 1996. Conservation genetics and molecular techniques: a perspective. In: Molecular genetics approaches in conservation: 459–477 (T. B. Smith & R. K. Wayne, Eds.). Oxford University Press, Oxford. HEDRICK, P. W. & MILLER, P. S., 1992. Conservation genetics: techniques and fundamentals. Ecol. Appl., 2: 30–46. JEFFREYS, A. J., WILSON, V. & THEIN, S. L., 1985. Hypervariable "minisatellite" regions in human DNA. Nature, 314: 67–73. JOHNSON, D. D. P., MACDONALD, D. W. & DICKMAN, A. J., 2000. An analysis and review of models of the sociobiology of the Mustelidae. Mammal Rev., 30: 171–196. JORDAN , W., F OLEY , K. & B RUFORD, M., 1998. Single–strand conformation polymorphism (SSCP) analysis. In: Molecular tools for screening biodiversity: 152–156 (A. Karp, P. G. Isaac & D. S. Ingram, Eds.). Ed. Chapman & Hall, Cambridge. K IMURA , M., 1979. The neutral theory of molecular evolution. Sci. Am. , 241: 94–104. LESSA, E. P. & APPLEBAUM, G., 1993. Screening techniques for detecting allelic variation in DNA sequences. Mol. Ecol., 2: 119–129. LIMAYE, A. P., JEROME, K. R., KUHR, C. S., FERRENBERG, J., HUANG, M. L., DAVIS, C. L., COREY, L. & MARSH, C. L., 2001. Quantitation of BK virus load in serum for the diagnosis of BK virus–associated nephropathy in renal transplant recipients. J. Infect. Dis., 183: 1,669–1,672. LUIKART, G. & ENGLAND, P. R., 1999. Statistical analysis of microsatellite DNA data. Trends Ecol. Evol., 14: 253–256. LYNCH, M., 1991. The genetic interpretation of inbreeding and outbreeding depression. Evolution, 45: 622–629. – 1996. A quantitative genetic perspective on conservation issues. In: Conservation genetics: case histories from nature: 471–501 (J. C. Avise & J. L. Hamrick, Eds.). Ed. Chapman and Hall, New York. L YNCH , M., C ONERY , J. & B URGER , R., 1995a. Mutational meltdowns in sexual populations. Evolution, 49: 1,067–1,080. – 1995b. Mutational accumulation and the extinction of small populations. Am. Nat., 146: 489–518. MACE, G. M., SMITH, T. B., BRUFORD, M. W. & WAYNE, R. K., 1996. An overview of the issues. In: Molecular genetic approaches in conservation: 3–21 (T. B. Smith & R. K. Wayne, Eds.). Oxford University Press, New York. MATTHES, M. C., DALY, A. & EDWARDS, K., 1998. Amplified fragment lenght polymorphism (AFLP). In: Molecular tools for screening biodiversity: 183–190 (A. Karp, P. G. Isaac & D. S. Ingram, Eds.). Ed. Chapman & Hall, Cambridge.

Domingo–Roura et al.

MAY, R., 1990. Taxonomy as destiny. Nature, 347: 129–130. MCCONKEY, E. H. & VARKI, A., 2000. A primate genome project deserves high priority. Science, 289: 1,295–1,296. MELNICK, D. J. & HOELZER, G. A., 1992. Differences in male and female macaque dispersal leads to contrasting distributions of nuclear and mitochondrial DNA variation. Int. J. Primatol., 13: 379–393. MERENLENDER, A. M., WOODRUFF, D. S., RYDER, O. A., KOCK, R. & VAHALA, J., 1989. Allozyme variation and differentiation in African and Indian rhinoceroses. J. Hered., 80: 377–382. MORITZ, C. & FAITH, D. P., 1998. Comparative phylogeography and the identification of genetically divergent areas for conservation. Mol. Ecol., 7: 419–429. MORLEY, J. M., BARK, J. E., EVANS, C. E., PERRY, J. G., HEWITT, C. A. & TULLY, G., 1999. Validation of mitochondrial DNA minisequencing for forensic casework. Int. J. Legal Med., 112: 241–248. MÜLLER–STARCK, G., 1998. Isozymes. In: Molecular tools for screening biodiversity: 73–82 (A. Karp, P. G. Isaac & D. S. Ingram, Eds.). Ed. Chapman & Hall, Cambridge. N EGRO , J. J. & T ORRES, M. J., 1999. Genetic variability and differentiation of two bearded vulture Gypaetus barbatus and implications for reintroduction projects. Biol. Conserv., 87: 249–254. NEVEU, H., HAFEN, T., ZIMMERMANN, E. & RUMPLER, Y., 1998. Comparison of the genetic diversity of wild and captive groups of Microcebus murinus using the random amplified polymorphic DNA method. Folia Primatol., 69: 127–135. NORMAN, J. A., MORITZ, C. & LIMPUS, C. J., 1994. Mitochondrial DNA control region polymorphisms: genetic markers for ecological studies of marine turtles. Mol. Ecol., 4: 363–373. O´BRIEN, S. J., ROELKE, M. E., MARKER, L., NEWMAN, A., W INKLER, C. A., MELTZER, D., C OLLY, L., EVERMANN, J. F., BUSH, M. & WILDT, D. E., 1985. Genetic basis for species vulnerability in the cheetah. Science, 227: 1,428–1,434. O´BRIEN, S. J., WILDT, D. E., BUSH, M., CARO, T. M., FITZGIBBON, C., AGGUNDEY, I. & LEAKEY, R. E., 1987. East African cheetahs: evidence for two population bottlenecks? Proc. Natl. Acad. Sci. USA, 84: 508–511. ORITA, M., IWAHANA, H., KANAZAWA, H., HAYASHI, K. & SEKIYA, T., 1989. Detection of polymorphisms of human DNA by gel electrophoresis and single–strand conformation polymorphisms. Proc. Natl. Acad. Sci. USA, 86: 2,766–2,770. PACKER, C., GILBERT, D. A., PUSEY, A. E. & O´BRIEN, S. J., 1991. A molecular genetic analysis of kinship and cooperation in African lions. Nature, 351: 562–565. PÉREZ–LEZAUN, A., CALAFELL, F., COMAS, D., MATEU, E., BOSCH, E., MARTÍNEZ–ARIAS, R., CLARIMÓN, J., FIORI, G., LUISELLI, D., FACCHINI, F., PETTENER, D. & BERTRANPETIT, J., 1999. Sex–specific migration

Animal Biodiversity and Conservation 24.1 (2001)

patterns in Central Asian populations revealed by the analysis of Y–chromosome STRs and mtDNA. Am. J. Hum. Genet., 65: 208–219. PETIT, R. J., EL MOUSADIC, A. & PONS, O., 1998. Identifying populations for conservation on the basis of genetic markers. Conserv. Biol., 12: 844–855. QUELLER, D. C., STRASSMANN, J. E. & HUGHES, C. R., 1993. Microsatellites and kinship. Trends Ecol. Evol., 8: 285–288. RAITIO, M., LINDROOS, K., LAUKKANEN, M., PASTINEN, T., SISTONEN, P., SAJANTILA, A. & SYVANEN, A. C., 2001. Y–chromosomal SNPs in finno–ugric–speaking populations analyzed by minisequencing on microarrays. Genome Res., 11: 471–482. REEVE, H. K., WESTNEAT, D. F., NOON, W. A., SHERMAN, P. W. & AQUADRO, C. F., 1990. DNA "finger– printing" reveals high levels of inbreeding in colonies of the eusocial nacked mole–rat. Proc. Natl. Acad. Sci. USA, 87: 2,496–2,500. RYDER, O. A., 1986. Species conservation and systematics: the dilemma of subspecies. Trends Ecol. Evol., 1: 9–10. SAIKI, R. K., GELFAND, D. H., STOFFEL, S., SCHARF, S. J., HIGUCHI, R., HORN, G. T., MULLIS, K. B. & ERLICH, H. A., 1988. Primer–directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science, 239: 487–491. SAMBROOK, J., FRITSCH, E. F. & MANIATIS, T., 1989. Molecular cloning. A laboratory manual, 2nd edition. Ed. Cold Spring Harbor Laboratory Press, New York. SHAFFER, H. B., FELLERS, G. M., MAGEE, A. & VOSS, S. R., 2000. The genetics of amphibian declines: population substructure and molecular differentiation in the yosemite toad, Bufo canorus (Anura, Bufonidae) based on single– strand conformation polymorphism analysis (SSCP) and mitochondrial DNA sequence data. Mol. Ecol., 3: 245–257. SOULÉ, M. E. (Ed.), 1987. Viable populations for conservation. Ed. Cambridge University Press, Cambridge. SOUTHERN, E., MIR, K. & SHCHEPINOV, M., 1999. Molecular interactions on microarrays. Nat. Genet., 21: 5–9. TABERLET , P. & BOUVET, J., 1994. Mitochondrial

DNA polymorphism, phylogeography, and conservation genetics of the brown bear Ursus arctos in Europe. Proc. R. Soc. Lond ., 255: 195–200. TAKAHASHI, K., TERAI, Y., NISHIDA, M. & OKADA, N., 1998. A novel family of short interspersed repetitive elements (SINEs) from cichlids: the patterns of insertion of SINEs at orthologous loci support the proposed monophyly of four major groups of cichlid fishes in Lake Tanganika. Mol. Biol. Evol., 15: 391–407. TROESCH, A., NGUYEN, H., MIYADA, C. G, DESVARENNE, S., GINGERAS, T. R, KAPLAN, P. M., CROS, P. & M ABILAT , C., 1999. Mycobacterium species identification and rifampin resistance testing with high–density DNA probe arrays. J. Clin. Microbiol. 37: 49–55. WATKINS, D. I., HODI, F. S. & LETVIN, N. L., 1988. A primate species with limited major histocompatibility complex class I polymorphism. Proc. Natl. Acad. Sci. USA, 85: 7,714–7,718. WAYNE, R. K., BRUFORD, M. W., GIRMAN, D., REBHOLZ, W. E. E., SUNNUCKS, P. & TAYLOR, A. C., 1994. Molecular genetics of endangered species. In: Creative conservation. Interactive management of wild and captive animals: 92–117 (P. J. S. Olney, G. M. Mace & A. T. C. Feistner, Eds.). Ed. Chapman & Hall, London,. WAYNE, R. K. & OSTRANDER, E. A., 1999. Origin, genetic diversity, and genome structure of the domestic dog. BioEssays, 21: 247–257. WEAVER, R. F. & HEDRICK, P. W., 1992. Genetics. 2nd ed. Wm. C. Brown, Dubuque, Iowa. WILKE, K., DUMAN, B. & HORST, J., 2000. Diagnosis of haploidy and triploidy based on measurement of gene copy number by real–time PCR. Hum. Mutat., 16: 431–436. WILLIAMS, P. H., GASTON, K. J. & HUMPHRIES, C. J., 1994. Do conservationists and molecular biologists value differences between organisms in the same way. Biodiversity Letters, 2: 67–78. WOODRUFF, D. S., 1993. Non–invasive genotyping of primates. Primates, 34: 333–346. WODICKÁ, L., DONG, H., MITTMANN, M., HO, M. H. & LOCKHART, D. J., 1997. Genome–wide expression monitoring in Saccharomyces cerevisiae. Nat. Biotechnol., 15: 1,359–1,367.

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Editors / Editores / Editors Antonio Barbadilla Univ. Autònoma de Barcelona, Bellaterra, Spain Xavier Bellés Centre d' Investigació i Desenvolupament CSIC, Barcelona, Spain Juan Carranza Univ. de Extremadura, Cáceres, Spain Luís Mª Carrascal Museo Nacional de Ciencias Naturales CSIC, Madrid, Spain Adolfo Cordero Univ. de Vigo, Vigo, Spain Mario Díaz Univ. de Castilla–La Mancha, Toledo, Spain Xavier Domingo Univ. Pompeu Fabra, Barcelona, Spain Francisco Palomares Estación Biológica de Doñana, Sevilla, Spain Francesc Piferrer Inst. de Ciències del Mar CSIC, Barcelona, Spain Ignacio Ribera The Natural History Museum, London, United Kingdom Alfredo Salvador Museo Nacional de Ciencias Naturales, Madrid, Spain José Luís Tellería Univ. Complutense de Madrid, Madrid, Spain Francesc Uribe Museu de Zoologia de Barcelona, Barcelona, Spain Consell Editor / Consejo editor / Editorial Board José A. Barrientos Univ. Autònoma de Barcelona, Bellaterra, Spain Jean C. Beaucournu Univ. de Rennes, Rennes, France David M. Bird McGill Univ., Québec, Canada Mats Björklund Uppsala Univ., Uppsala, Sweden Jean Bouillon Univ. Libre de Bruxelles, Brussels, Belgium Miguel Delibes Estación Biológica de Doñana CSIC, Sevilla, Spain Dario J. Díaz Cosín Univ. Complutense de Madrid, Madrid, Spain Alain Dubois Museum national d’Histoire naturelle CNRS, Paris, France John Fa Durrell Wildlife Conservation Trust, Trinity, United Kingdom Marco Festa–Bianchet Univ. de Sherbrooke, Québec, Canada Rosa Flos Univ. Politècnica de Catalunya, Barcelona, Spain Josep Mª Gili Inst. de Ciències del Mar CMIMA–CSIC, Barcelona, Spain Edmund Gittenberger Rijksmuseum van Natuurlijke Historie, Leiden, The Netherlands Fernando Hiraldo Estación Biológica de Doñana CSIC, Sevilla, Spain Patrick Lavelle Inst. Français de recherche scient. pour le develop. en cooperation, Bondy, France Santiago Mas–Coma Univ. de Valencia, Valencia, Spain Joaquín Mateu Estación Experimental de Zonas Áridas CSIC, Almería, Spain Neil Metcalfe Univ. of Glasgow, Glasgow, United Kingdom Jacint Nadal Univ. de Barcelona, Barcelona, Spain Stewart B. Peck Carleton Univ., Ottawa, Canada Eduard Petitpierre Univ. de les Illes Balears, Palma de Mallorca, Spain Taylor H. Ricketts Stanford Univ., Stanford, USA Joandomènec Ros Univ. de Barcelona, Barcelona, Spain Valentín Sans–Coma Univ. de Málaga, Málaga, Spain Tore Slagsvold Univ. of Oslo, Oslo, Norway

Animal Biodiversity and Conservation 24.1 (2001)

Commercial bushmeat hunting in the Monte Mitra forests, Equatorial Guinea: extent and impact J. E. Fa & J. E. García Yuste

Fa, J. E. & Gracía Yuste, J. E. Commercial bushmeat hunting in the Monte Mitra forests, Equatorial Guinea: extent and impact. Animal Biodiversity and Conservation, 24.1: 31–52. Abstract Commercial bushmeat hunting in the Monte Mitra forests, Equatorial Guinea: extent and impact.— Understanding the exploitation of bushmeat by commercial hunters is fundamental to resolving hunting sustainability issues in African rainforests. The objective of this study was to examine the impact of hunters operating from the village of Sendje in the Monte Mitra region, Republic of Equatorial Guinea. Offtake patterns of 42 hunters were studied over a period of 16 months. A total of 3,053 animals of 58 species were hunted during 1,914 hunting days. This represented around 11,376 kg of bushmeat or 2,219 animals extracted per annum. Most captures were mammals (43 species, 79%), constituting 90% of the biomass hunted, of these 30% were ungulates and 27% were rodents. Hunters used 17 hunt camps within the 1,010 km2 total study area. Hunting activity fell from the start to the end of the study, with fewer hunting days, biomass and captures being recorded per month. Captures fell from 700 animals in the first month to less than 100 during the last month. Per hunter, returns diminished from 21 in the first month to around 13 animals from the third month. Average body mass of prey also declined throughout the study period. The principal hunting method was cable snaring —over 100 million snare nights were estimated. An average hunter extracted around 50 animals or 271 kg of bushmeat per annum. Hunter and camp differences were significant. Most carcasses were sold for the city market or to villagers, and the proportion of carcasses sold to market was positively correlated with the species body mass. Capture rates and vulnerability were dependent on prey size since medium–sized animals were more vulnerable to be caught than small or large– bodied animals. Harvest sustainability was calculated for 14 mammals and it was seen that the situation was unsustainably for 5 species due to the extent and impact of hunting. The bay duiker (Cephalophus dorsalis) was by far the most heavily exploited species. Conservation of the Monte Mitra region is impossible unless the hunting for profit issue is resolved in Sendje and adjoining villages. Key words:: Bushmeat hunting, Hunters, Sustainability, Monte Mitra, Equatorial Guinea. Resumen Caza comercial en los bosques de Monte Mitra, Guinea Ecuatorial: alcance e impacto.— Entender la explotación de la carne de selva por parte de cazadores comerciales es fundamental para resolver las cuestiones de sostenibilidad referentes a la caza en los bosques húmedos de África. El objetivo de este estudio fue examinar el impacto de la actividad de los cazadores de la aldea de Sendje, en la región del Monte Mitra, República de Guinea Ecuatorial. Se estudiaron los patrones de caza de 42 cazadores durante un periodo de 16 meses. Se cazaron un total de 3.053 animales de 58 especies en 1.914 jornadas de caza, lo que representa aproximadamente 11.376 kg de carne de selva o 2.219 animales extraídos por año. La mayoría de capturas fueron mamíferos (43 especies, 79%), que constituyeron el 90% de la biomasa cazada, y entre ellos un 30% de ungulados y un 27% de roedores. Los cazadores utilizaron 17 campos de caza dentro de un área de estudio con una extensión total de 1.010 m2. La actividad de caza fue disminuyendo desde el inicio del estudio hasta al final del mismo, con menos días de caza, biomasa y capturas registradas por mes. Las capturas disminuyeron desde 700 animales durante el primer mes a menos de 100 en el último. Por cazador, el rendimiento diminuyó de 21 animales en el primer mes a 13 en el tercero. La media de masa corporal de las presas también disminuyó a lo largo del periodo de estudio. El método de caza más utilizado ISSN: 1578–665X

Animal Biodiversity and Conservation 24.1 (2001)

fue el cepo (se estimó la existencia de alrededor de 100 millones de cepos noche). En promedio, cada cazador extrajo alrededor de 50 animales o 271 kg de carne de selva por año. Las diferencias entre campos de caza y cazadores fueron significativas. La mayoría de piezas fueron vendidas al mercado de la ciudad o a los aldeanos, y la proporción de piezas vendidas al mercado estuvo correlacionada positivamente con la masa corporal de las mismas. Los índices de captura y vulnerabilidad dependieron del tamaño de las presas ya que los animales de tamaño medio resultaron más vulnerables que los pequeños o grandes. Se calculó la sostenibilidad de la caza para 14 mamíferos en los bosques de Monte Mitra, Guinea Ecuatorial, resultando insostenible para cinco especies por su extensión e impacto. Cephalophus dorsalis fue la especie explotada con mayor intensidad. La conservación de la región del monte Mitra es imposible a no ser que el problema de la caza de carne de selva para su comercialización se resuelva en Sendje y pueblos vecinos. Palabras clave: Caza de carne de selva, Cazadores, Monte Mitra, Guinea Ecuatorial. (Received: 18 IX 01; Final acceptance: 16 X 01) John E. Fa(1), Durrell Wildlife Conservation Trust, Les Augrès Manor, Trinity, Jersey JE3 5BP, Channel Islands, UK.– Juan E. Gracía Yuste(2), Agencia Española de Cooperación Internacional (AECI), Proyecto Araucaria Amazonas Nauta, Loreto 442, Iquitos, Peru. (1) (2)

e–mail: jfa@durrell.org e–mail: jgarciay@nexo.es

Animal Biodiversity and Conservation 24.1 (2001)

Introduction In tropical areas world–wide the meat of wild animals has long been part of the staple diet of forest–dwelling peoples. However, in recent years, there has been an important transition from subsistence to commercial hunting and trading of wildlife because of accelerating population growth, modernisation of hunting techniques, and greater accessibility to remote forest areas (APE ALLIANCE, 1998; WILKIE & CARPENTER, 1999). In Africa, bushmeat is sold for public consumption either fresh or smoked. Bushmeat remains the primary source of animal protein for the majority of forest families, and can also constitute a significant source of revenue (JUSTE et al. 1995). The high demand for bushmeat and the lucrative trade associated with it is the main reason for the high extraction rates estimated for many West and Central African countries (FA & PERES, 2001). Although changes from subsistence to commercial hunting have been occurring for some time (see HART, 2000), many more hunters currently supplement their incomes with the sale of meat. Such commerce increases the amount of hunting and reduces the sustainability of numerous wildlife species largely because it enlarges the effective human population density of consumers eating meat from an area of forest (BENNETT & ROBINSON, 2000). Commercial hunters and traders supply urban markets for profit to meet the increasing demand for animal protein in urban centres. Markets in towns and cities are the main sales–point for species extracted from natural areas (FA, 2000; FA et al., 1995). Interest in markets, for estimating game extraction rates from the surrounding areas is growing (FA et al., 2000). Investigations at the supplier end are also necessary to understand the extent and limits of the commercial hunting (R OBINSON et al., 1999). From this informed perspective, it may be possible to propose sound management policies. However, despite the importance of commercial hunting in African moist forests, few studies have documented temporal and spatial activities of multiple hunters operating in a known area. Here, the extent and impact of commercial hunters in the Monte Mitra forests is examined, Rio Muni, Republic of Equatorial Guinea. Wildlife harvests were documented for a total of 42 hunters over a period of 16 months. Destination of the bushmeat, whether consumed locally or sold is also assessed. By estimating hunt catchments for a selection of hunt camps we then describe the overall impact of such hunting pressure on densities and biomass of selected mammal species in the area.

Study Area Río Muni (26,000 km2), located just north of the equator, is part of the Republic of Equatorial Guinea

(fig. 1). The city of Bata (population 55,000) is the major urban centre in the region. From the coast, elevation rises to 1,200 m at the highest peak Monte (Mount) Mitra. The Monte Mitra region (44 km2), is now part of the Monte Alén National Park (FA, 1991; GARCÍA YUSTE & ENEME, 2000), within the Niefang mountains (fig. 1). Elevations of just over 1,000 m are typical. Relief is abrupt with some flat areas along river valleys. Climate is typically hot humid equatorial (average temperature 25°C but 2– 5°C lower in the highlands), with 3,000–3,500 mm annual rainfall. Most precipitation occurs from September to December and from March to May; less rain falls from June through August. The region’s vegetation forms part of the Guineo–Congolian forest (SAYER et al., 1992). The Monte Mitra forests are dominated by Xylopia, Anthocleista, Barteria, Morinda, and Uapaca (BEUDELS, 1998). In flooded areas along river valleys, Mitragina ciliata, Anthostema aubreyanum and Raphia spp. are typical, with oil palm (Elaeis guineensis) being commonest. Secondary formations of Aframomum spp., bushes of the Rubiaceae family, and some lianas especially Tetracera and Cissus, predominate in the more disturbed areas. Intact dense tropical rainforest covers most of the study area. This forest has a closed upper canopy of Gilbertiodendron dewevrei, Brachystegia, Piptadeniastrum, Pterocarpus, Coula edulis, Santiria, Staudtia, Strephonema pseudocola, Berlinia, Dialium , and Desbordesia . Around 400 m, Olacaceae, Irvingiaceae, Myristicaceae and Euphorbiaceae are common plant families. There are some small seasonal swamps and lakes dominated by Nitragina ciliata, Pandanus candelabrum and Anthocleista. Between 400 and 700 m, the vegetation changes and Lovoa trichilioides, Guarea cedrata, members of the Meliaceae, as well as some Cesalpinaceae (Tetraberlinia bifoliata, Anthonotha cladantha and Anthonotha ferruginea) are common. Begonias, diverse species of Canthium, Acanthonema and Trachystigma are likewise characteristic. Above 700 m, one of the most abundant species is Tetraberlinia bifoliata in association with Irvingia rubur, Garcinia couriana, Staudtia sp, Pentadesma butyracea. The human population, around 1,500 inhabitants, is concentrated along the Senye–Cogo road in the villages of Sendje, Binguru, Miton and Emangos to the north and Ncoho, Basilé and Mitong in the south (fig. 1). In the past, there were human settlements within the forest interior, but these are nowadays abandoned although some are used as hunt camps, e.g. Bisun. Until recently most of the population was employed in cocoa and coffee plantations, as well as in the logging businesses in the zone. With the abandonment of plantations and cessation of logging operations due to political instability and economic decline, the population has had to turn to hunting and subsistence agriculture. Villages in the south also fish along the headwaters of the Muni River Estuary.

Fa & García Yuste

Gulf of Guinea

Cameroon

Equatorial Guinea Sendje

Gabon

Mobun-nwuom

Binguru Bisun Anvira

Mitomo

Enuc Tom–asi Avis–ncha

Ongam–nsok Aben–nam

rivers roads camps

Mitra 1200 m 0

10 km

Fig. 1. Geographical location of Rio Muni region, Equatorial Guinea, showing the position of the Monte Mitra study area, and hunt catchments for eight camps. Fig. 1. Localización geográfica de la región de Río Muni, Guinea Ecuatorial, mostrando la situación del área de estudio de Monte Mitra y las zonas de captura de ocho campos de caza.

Methods Over 16 months (1 January 1998–26 April 1999) we collected data on cable snaring and shooting activities of 42 hunters from Sendje. Hunters operated in an area approx. 1,010 km2 within 17 different hunt camps. Areas furthest away from the village were unhunted before our study. Harvested bushmeat was taken from camps to the village to be transported by intermediaries to the main city market in Bata. An assistant, a local villager, recorded all carcasses arriving in Sendje and interviewed hunters on duration of hunting trips (defined as a hunting excursion undertaken by a hunter), hunting days (days spent by the hunter in the forest), number of snares set, and hunting camp operated from.

All hunted animals were identified to species, but no attempt was made to weigh or measure animals. It was possible to determine for age class (juvenile or adult) and sex for 99% and 97% of carcasses respectively. Whether the animal had been shot, snared or caught by other means (by hand, machete or dogs) was also documented. Capture rate per species (N OSS , 1998) was calculated as the number of snare nights required to capture one animal of a particular species. Animals that were scavenged or decomposed were not recovered by hunters and were recorded as wastage. Information on whether the carcass was consumed in camp, consumed by the hunters’family, sold in the village or destined for the Bata market was also recorded. Hunt camps were geo–referenced with the aid of a GPS and altimeter. Camps were subsequently

Animal Biodiversity and Conservation 24.1 (2001)

Log. numbers/biomass (kg)

6 5

Numbers Biomas Biomasss

y = 1.1659x + 0.9195 R 2 = 0.6014

4 3 2 1

-1

0 0

y = 1.1659x + 0.9195 R 2 = 0.0296 1

-1 Log body mass (kg) Fig. 2. Relationship between the body mass (kg) of the hunted bushmeat species, the number of animals captured and the total biomass (kg) of each species extracted. Fig. 2. Relación entre la masa corporal (kg) de las especies de carne de selva cazadas, el número de animales capturados y la biomasa total (kg) de cada especie extraída.

mapped onto a 1/100,000 land use map from the CUREF Project ("Conservación y Utilización Racional de los Ecosistemas Forestales de Guinea Ecuatorial", http://www.internetafrica.com/curef/) based in Bata. CUREF maps are based on radar, Spot XS and Landsat TM (1988–1995) images. Hunt catchments, defined as the area (in km2) operated by hunters during the study, were estimated for only eight camps (Aben–nam, Anvira, Avis–ncha, Bisun, Mobun–nwuom, Enuc, Ongam–nsok, Tom–asi). This was undertaken by accompanying hunters for periods of 2 to 20 hunting days and geo-referencing the limits of their hunting territories (fig. 1). Species names follow KINGDON (1997). Biomass extracted per species was calculated by multiplying the recorded number of carcasses of a species by the mass of an "average" individual. Body masses were taken from FA & PURVIS (1997) for adults, and halved for juveniles. By using productivity and population density data in FA et al. (1995) for the same region, it was possible to evaluate sustainability of hunting for 14 mammal species (2 rodents, 6 ungulates, 5 primates, 1 pangolin) for the estimated hunt catchments. Harvest rates were calculated by FA et al. (1995) using the method of ROBINSON & REDFORD (1991). Statistical analyses were carried out using S– plus (VENABLES & RIPLEY, 1999). All means are reported with one standard deviation (±1 SD).

Results Prey Species During the study period, hunters caught 3,053 individuals of 58 species (43 mammals, 8 birds, 6 reptiles, 1 snail) or 15,169.1kg. Mammals accounted for 79% of total captures, reptiles 16%, birds 5% and snails 0.03%. By weight, 90.0% of the hunted biomass consisted of mammals, 9.2% reptiles, 0.86% birds and 0.03% of snails. Over 30% of captures were made up of ungulates (884 carcasses, 12 species), followed by rodents (27%, 826 carcasses, 7 species), reptiles (16%, 490 carcasses, 6 species) and primates (11%, 329 carcasses, 11 species). Pangolins (2 species) were represented by 224 carcasses (7%), birds by 142 (5%, 8 species), carnivores by 112 (10 species, 4%), and Tubulidentates by one animal of a single species. Nine species (2 species each of rodents, ungulates, and primates, and one species of reptile, bird and pangolin) were represented by >100 captures, but 33 species (56.89%) had less than 10 carcasses each. The most–captured species was the blue duiker, Cephalophus monticola , which represented 21.6% (658 carcasses) of all captures and 15.3% by weight. The brush–tailed porcupine, Atherurus africanus

Fa & García Yuste

Total number of snares

300

250

200

150

100

Total number of camps

350

F Mr Ap My Jn

Jl Ag S

D Ja

Mr Ap

7,000

6,000

7 5,000 6 5

4,000

3,000

2,000

Camps Snare Snaress

1 0

Total number of snares

Snares/day Hunting days

Total number of hunting days

400

1,000

F Mr Ap My Jn Jl Ag S Month

D Ja

Mr Ap

Fig. 3. Monthly changes (I 98–IV 99) in the numbers of hunting days recorded and number of camps used by hunters in Monte Mitra, Equatorial Guinea: Ja. January; F. February; Mr. March; Ap. April; My. May; Jn. June; Jl. July; Ag. August; S. September; O. October; N. November; D. December. Fig. 3. Cambios mensuales (I 98–IV 99) en el número de días de caza registrados y el número de campos usados por los cazadores en Monte Mitra, Guinea Ecuatorial. (For abbreviations see above.)

appeared in almost the same proportion (20.3%, 619 carcasses), but represented only 8.3% by weight. The bay duiker, Cephalophus dorsalis, contributed 12.4% of the total hunted biomass although it comprised only 4.09% (128 carcasses) of total captures. Larger–bodied species contributed most to hunted biomass but there was no correlation between body mass and number of animals hunted (fig. 2).

A monthly average of 25.90±44.42 hunters (range 19–34) were active in the entire study area, an average of 31.50±13.90 hunting days month-1. A total of 1,914 hunting days were recorded, but number of hunting days month-1 dropped significantly from 364 in the first month to around 100 after the eighth month (R2 = 0.50; d.f. = 14; P = 0.000) —a minimum of 26 hunting days was recorded in December 1998 (fig. 3A).

Animal Biodiversity and Conservation 24.1 (2001)

250

Number of animals hunted

200

Birds

Carnivores

Pangolins

Rodents

Ungulates

Primates

Reptiles 150

100

Mr Ap My Jn

Jl Ag S O Month

D Ja

Mr Ap

Fig. 4. Monthly changes (I 98–IV 99) in number of hunted animals within the main taxonomic groups in Monte Mitra, Equatorial Guinea. (For abbreviations see fig. 3.) Fig. 4. Cambios mensuales (I 98–IV 99) en número de animales cazados de los principales grupos taxonómicos en Monte Mitra, Guinea Ecuatorial. (Para las abreviaturas ver fig. 3.)

In contrast, the average number of snares set per month increased significantly from the start to the end of the study (R2 = 0.61; d.f. = 14; P = 0.0003). Hunters used a total of 17 camps during the study, an average of 5.25±1.48 camps per month-1. The number of camps used ranged from 3 in August 1998 to 8 in January 1998, and the total number of snares set per month correlated with the number of camps used (fig. 3B). There was a significant positive correlation between the number of operational camps and the total number of hunting days per month-1 (R2 = 0.36; d.f. = 14; P = 0.000).

observed in all main taxonomic groups (fig. 4). Number of captures per hunter also declined from 20.73±12.53 animals hunter-1 in the first month to around 10.52±4.93 animals hunter-1 by the third month (fig. 5A). Mean numbers fluctuated between 4 to 13 animals after the third month. Corresponding with captures, biomass dropped steeply from 126.82±117.44 kg hunter-1 in the first month to 20.65±20.18 kg hunter-1 in the third month (fig. 5B). Average body mass of hunted animals also declined throughout the study period (fig. 5C); larger– bodied animals were more prevalent during the earlier months of the study.

Temporal changes in bushmeat numbers and biomass

Hunter differences

Captures fell from around 700 in the first month (January 1998) to less than 100 during the last month (April 1999). This amounted to 2,663.2 kg extracted in January 1998 and 321.5 kg in April 1999. The drop was significant in the number of animals snared (R2 = 0.25; d.f. = 173; P = 1.163e–012), numbers shot (R2 = 0.06; d.f. = 83; P = 0.02) and animals killed by other means ( R2 = 0.158; d.f. = 26; P = 0.04). A fall in animals hunted between the first and the third month was

All hunters used firearms and cable snares, but snare hunting was the principal method used. The main type of snare is a noose made out of wire cable that is set along an animal trail. When the animal steps on a pressure pad, it releases a bent–over pole, which springs up to tighten the noose around the animal’s leg. During the study period, hunters deployed a total of 56,398 snares. This amounted to 107,945,772 snare–nights (the number of snares

Fa & García Yuste

35 30 25 20 15 10 5 0 Ja

F Mr Ap My Jn Jl Ag Month

O N

F Mr Ap

Ja F

Mr Ap

Ja F

Mr Ap

300 250 200 150 100 50 0 Ja F Mr Ap My Jn Jl Ag S Month

300 250 200 150 100 50 0

Mr Ap My Jn

Jl Ag S Month

Fig. 5. Mean (± SD) monthly changes (I 98–IV 99) in: A. Number of animals; B. Total animal biomass (kg) extracted per hunter per hunting day; C. Body mass (kg) of hunted animals. (For abbreviations see fig. 3.) Fig. 5. Cambios medios mensuales (± SD) (I 98–IV 99) en: A. Número de animales; B. Biomasa animal total (kg) extraída por cada cazador por día de caza; C. Masa corporal (kg) de los animales cazados. (Para abreviaturas ver fig. 3.)

Animal Biodiversity and Conservation 24.1 (2001)

times the number of nights set), an average of 1,927,603 snare–nights per hunter. An average of 112.06±57.34 snares (136,358.31±118,441.48 snare nights ranging from 28,303 snare nights in June 1998 to 408,192 snare nights in July 1998) were operational every month throughout the study area, with no significant monthly variation being detected. However, number of snares set in each camp differed significantly, from 50.0±70.7 snares hunting trip-1 in Sendje to 222.4±106.9 snares hunting trip-1 in Aben–nam (Goodness of fit test, R2 = 286.35; d.f. = 15; P = 1.7e–51). Per camp, the number of snares set was not correlated with the number of hunters operating in the area (R2 = 0.002; d.f. = 40; NS) or with the size of the hunt catchment area (R2 = 0.0002; d.f. = 6; NS). A total of 563 hunting trips was recorded for the 42 hunters in the area. Each hunter undertook 13.73±13.96 (range 1–52) hunting trips during the study period, and spent an average of 46.27±54.49 (range 1–233 days) total hunting days (table 1). Hunting trips lasted 3.81±3.21 days per hunter (range 1.00±0.00–9.43±4.34 days) during which 1,484.16±1723.92 snares were operated per hunter (range 80–5,459 snares). The hunting trip duration did not differ significantly among hunters (R2 = 12.11; d.f. = 40; P = 0.74). However, the number of snares operated per hunter varied significantly (R2 = 80025.98; d.f. = 40; P = 0.000). Number of animals hunted and biomass extracted per hunter were positively correlated with total number of hunting days (fig. 6). Biomass extracted and captures per hunter were also positively correlated with number of snares set (R2 = 0.48; d.f. = 40; P = 0.000). However, the total number of camps used was not correlated with number of hunting trips completed by each hunter. Hunters extracted 66.29±66.51 animals or 270.87±219.35 kg of animal biomass hunter-1 during the study period or a mean of 50 animals or 203.18 kg hunter -1 annum -1. Number of animals hunted (R2= 2736.22; d.f. = 40; P = 0.000) and biomass extracted (R2= 7282.53; d.f. = 40; P = 0.000) varied significantly among hunters. The most productive hunter was Hunter 8, who captured a total of 276 animals on 52 hunting trips whilst Hunter 42, the least prolific, caught a single animal on any one hunting trip. Most animals were caught by snares (60.83±67.23 animals hunter-1), and significantly fewer animals were shot (5.45±10.27 animals hunter-1). Number of animals extracted per hunting day by each hunter averaged 0.83±1.72 for the study period. The lowest monthly extraction figure was for April 1998 (0.53±0.41 animals hunter-1 hunting day-1) whilst the highest (1.78±1.82 animals hunter-1 hunting day-1) was in January 1999. There was no significant inter–monthly difference in number of animals extracted by hunters (R2= 3.74; d.f. = 14; P = 0.999). Hunters used from one to six camps, 2.29±1.50

camps per hunter (median 2 camps). Most (n = 19 hunters; 45.24% of all hunters) used only one camp, eight (19.05%) used two camps, 14 (33.33%) from three to five camps, but a single hunter (2.38%) operated in six different camps. Only one hunter used Esua–asas and Eto–mbeng, but a maximum of 20 hunters entered Bisun (table 2). Number of snares set per trip in each camp varied from 50.0±57.74 in Sendje to 222.35±106.85 in Aben–nam. An average of 112.96±54.25 snares per hunt catchment was set during each hunting trip. Camp differences For the camps surveyed, regular hunt catchments were an average of 28.3± 8.9 km from the village, ranging from 11.7 km (Bisun) to 41.7 km (Ongam– nsok) (table 2). Hunt catchments varied significantly in size from 6.2 km2 in Ongam–nsok to 314.2 km2 in Bisun, with the larger areas being found closer to the village (R2 = 0.63; d.f. = 6; P = 0.019). The size of the hunted areas was correlated with the number of hunters operating within them (R2 = 0.59; d.f. = 6; P = 0.020). Per camp, annual harvests varied from 5 animals (4.77 kg) in Esua–asas to 764 animals (4,413.53 kg) in Bisun. Biomass extracted per hunter differed significantly between camps (R2 = 957.5; d.f. = 15; P = 1.4e–193). Biomass extracted per camp per hunting day also varied significantly (R2 = 29.7; d.f. = 15; P = 0.013). For those camps for which hunt catchment area was measured, number of animals and biomass extracted per km2 differed significantly between camps (table 3). Ongam–nsok was by far the most productive with 224.47 kg of bushmeat km-2, whereas Bisun, which was also the most hunted camp, produced 14.05 kg bushmeat km -2 (R2 = 667.5; d.f. = 15; P = 1.6e132). Proportion of species sold and consumed Although some meat is for home consumption (22.87%), the largest proportion of animals hunted (67.77%) was either sold in Sendje (34.05%) or in the Bata market (33.72%). Bushmeat consumed by the hunters’families was 16.26%, and hunters themselves would consume 6.61% in forest. Only 9.35% of the total number of recorded animals was unsuitable for consumption. The number of animals sold (R2 = 0.02; d.f. = 56; NS) or consumed (R2 = 0.02; d.f. = 56; NS) was not correlated with the number of carcasses per species hunted. Similarly, biomass was not correlated with proportion sold (R2 = 0.02; d.f. = 56; NS) or consumed (R2 = 0.03; d.f. = 56; NS). Per hunter, an average of 39.02±26.98% of the animals hunted were sold to the Bata market. For the subsample of species with >100 carcasses (7 mammals, 2 reptiles, 1 bird), the proportion of animals sold to the Bata market

Fa & García Yuste

Table 1. Recorded activity and offtake of hunters in the Monte Mitra area (I 98–IV 99), Equatorial Guinea. The number of carcasses recorded with hunter information was lower than the total number (3,053 carcasses) noted during the study: H. Hunter; Ht. Duration of hunting trips (in days). Thd. Total hunting days; Tht. Total hunting trips; C. Number of camps; Sno. Number of snares operated; Sph. Number of species hunted; Ahsh. Number of animnals hunted with shotgun; Ahsn. Number of animals hunted with snares; Tah. Total animals hunted; Be. Biomass extracted (in kg). Tabla 1. Actividad registrada y productividad de los cazadores en el área de Monte Mitra (I 98–IV 99), Guinea Ecuatorial. El número real de piezas registrado mediante información de los cazadores fue inferior que el número total (3.053 piezas) anotado durante el estudio: H. cazador; Ht. Duración de las salidas de caza (en días). Thd. Total de jornadas de caza; Tht. Total de salidas de caza; C. Número de campos; Sno. Número de cepos utilizados; Sph. Número de especies cazadas; Ahsh. Número de animales cazados con armas de fuego; Ahsn. Número de animales cazados con cepo; Tah. Total de animales cazados; Be. Biomasa extraída (en kg).

Ht H

Mean

Thd

Tht

Sno

Sph

Ahsh

Ahsn

4.18

2.53

1,478

Tah 89

424.23

9.43

4.39

1,645

532.67

3.17

2.71

264

346.75

6.13

3.28

233

4,192

238

243

712.48

4.11

1.62

1,898

469.40

6.63

3.01

179

2,555

189

191

828.71

5.78

2.05

104

1,998

175

599.53

2.80

2.81

137

5,459

246

252

668.91

6.14

1.35

712

362.50

4.25

1.44

2,631

101

103

329.11

2.36

2.09

2,469

108

398.30

4.40

1.82

437

116.16

2.00

–

197

23.89

3.35

1.23

104

5,955

166

498.46

5.00

3.10

512

127.06

3.10

1.86

130

3,615

159

165

455.33

2.80

2.33

4,812

107

127

505.56

6.75

1.89

612

311.37

4.67

1.53

1,012

240.89

3.75

2.25

747

191.61

5.00

–

85.97

2.63

1.19

353

103.56

5.11

6.85

2,829

146.32

3.00

1.00

356

128.31

4.33

2.52

472

48.79

4.00

1.41

440

42.84

3.39

5.09

129

5,216

133

486.34

4.00

–

68.91

1.30

0.48

–

221.56

4.00

–

120

37.54

Animal Biodiversity and Conservation 24.1 (2001)

Table 1. (Cont.) Ht Mean

Tht

Sno

Sph

Ahsh

Ahsn

Tah

1.86

1.07

124

531.23

4.33

2.31

600

22.51

1.50

0.71

140

153.83

2.20

1.10

270.90

1.35

1.14

1,772

169.57

2.20

1.64

182.90

3.00

–

130

11.90

1.00

–

97.01

1.00

0.00

276

22.15

3.75

1.24

–

332.56

1.25

0.50

–

34.90

3.00

–

33.97

Totals

3.80

3.21

1,916

563

56,398

229

Log. number of animals/biomass (kg) extracted

Thd

2,555 2,784 11,376.49

3.5 y = 0.6419x + 1.3858 R 2 = 0.6465 3.0

2.5

2.0

1.5

1.0

Biomass Numbers

0.5

0.0 0.0

y = 0.8446x + 0.4799 R 2 = 0.8363 0.5

1.0 1.5 2.0 Log. hunting days per hunter

2.5

Fig. 6. Relationship between number of hunting days per hunter and number of animals captured, and total animal biomass (kg) extracted per hunter. Fig. 6. Relación entre el número de días de caza por cada cazador, el número de animales capturados y la biomasa animal total (kg) obtenida por cada cazador.

Fa & García Yuste

Table 2. Details of hunting intensity within camps in the Monte Mitra area, Equatorial Guinea: Hd. Hunting days hunter–1; Sht. Snares hunting trip–1; Dv. Distance from village (in km); Hc. Hunt catchment (in km 2); N. Species recorded; A. Average body mass of recorded prey (in kg); H. Hunters. Tabla 2. Detalles de la intensidad de caza en los campos del área de Monte Mitra, Guinea Ecuatorial: Hd. Días de caza para cada cazador; Sht. Salidas para colocación de cepos; Dv. Distancia hasta el pueblo (en km); Hc. Área de caza (en km2); N. Especies registradas; A. Masa corporal media de las presas registradas (en kg); H. Cazadores.

Hd Camp

Mean

9.08

48.8

3.89

1.23

222.35

106.85

20.83

38.48

28.7

4.48

2.94

86.16

41.85

Avis–ncha

27.50

75.43

27.1

5.05

1.82

105.71

47.34

Bisun

11.67

314.16

26.6

Ebang

–

Echun–ndje

–

Aben–nam

Anvira

Enuc

Mean

Sht

3.18

3.91

125.88

47.93

1.20

0.45

80.60

13.99

25.8

3.60

0.70

175.10

75.61

2.26

75.41

48.72

28.33

12.57

24.8

3.28

Esua–asas

–

23.4

Eto–mbeng

–

23.3

Evuadulu

–

23.1

Kong

–

Mandjana

–

Mitong–evina

–

50.27

41.67 – 32.50

Mobun–nwuom 33.33 Ongam–nsok Sendje Tom–asi Grand total

–

3.25

3.17

5.67

2.31

21.5

4.00

–

21.1

6.18

5.40

140.91

15.14

20.6

6.93

1.46

100.64

32.65

6.16

19.0

6.98

2.62

108.26

–

10.3

7.07

9.7

4.17

1.99

109.13

39.30

–

23.9

4.20

3.38

112.96

54.25

was significantly positively correlated with body mass of the species. The relationship was polynomial (y = 0.0209x3 - 0.9194x2 + 11.769x + 4.3471; R2 = 0.80; d.f. = 8; P = 0.000). The total percentage of animals sold per camp averaged 71.67±14.15%. The proportion sold varied from 33.3% in Mandjana to 88.2% in Ebang. The proportion of game sold or consumed was not related to the distance of the camp to the village (Sold R2 = 0.28; d.f. = 6; P = 0.1818; Consumed R2 = 0.28; d.f. = 6; P = 0.1818). There was no correlation between the number of animals hunted and percentage sold. The number of animals consumed in forest was correlated with wastage (R2 = 0.74; d.f. = 15; P = 9.288e–006). The number of animals sold in the village was also correlated with the number sold in Bata market (R2 = 0.78; d.f. = 15; P = 2.661e–006).

53.35

34.45

94.67

4.16

120

–

41.74 57.74

Capture rates and vulnerability Most animals (n = 2,636; 86.3%) were caught by snares, 7.9% (n = 241) were killed with shotgun, and 5.8% (n = 176) were taken by other methods. Per hunting day, 15.5±16.17 animals were snared, but significantly fewer were shot (4.2±3.5) or taken by other methods (2.9±2.2). The proportion of animals shot was significantly lower during all months of the study (fig. 7). Over one–half of all species (32 species) encountered was caught only in snares (table 4). Of 42 species (73.68%), over 50% of individuals caught were snared. Ungulates, rodents and carnivores were relatively more vulnerable to snares than to firearms; 10 of the 12 ungulates, 6 of the 7 rodents, and 7 of the 9 carnivores were caught exclusively with snares. The species most

Animal Biodiversity and Conservation 24.1 (2001)

Table 3. Captures, wastage and hunting method for bushmeat species in Monte Mitra, Equatorial Guinea: Hm. Hunting method (%); C. Captures; Sn. Snares; Sh. Shotgun. Tabla 3. Capturas, piezas desaprovechadas y métodos de caza para carne de selva en Monte Mitra, Guinea Ecuatorial: Hm. Método de caza (%); C. Capturas; Sn. Cepos; Sh. Armas de fuego.

Groups Species

Capture rate Mean

Wastage

Hm (%)

Other

–

100

66.7

Snails

Achatina spp.

–

Reptiles

Bitis gabonica

2,991.25

2,292.82

Chamaleo cristatus

4,333

2,907.62

Kynixis erosa

33.3

211.92

219.63

337

60.2

39.8

Osteolaemus tetraspis

1,095.93

1,367.91

6.7

Python sebae

7,272.83

3,845.01

100

Varanus niloticus

1,028.78

994.66

3.3

95.7

4.3

Group total

2,822.29

2,650.63

487

1.2

66.6

4.6 28.8

Ceratogymna atrata

2,209.67

3,619.46

33.3

Francolinus lathanmi

3,562.75

2,026.69

22.2

Birds

Gypohierax angolensis Haliaetus vocifer Numida meleagris Obom (unidentified bird) Psittacus erithacus

–

908.47 29,975 551.93

66.7

100

118

14.4

97.5

–

714.55

–

143

13.3

1,117.57

2.5

100

Stephanoaetus coronatus

29,975

Group total

11,197.14

14,583.91

Bdeogale nigripes

3,375.67

2,738.74

100

Civicttis civetta

3,667.58

6,331.87

15.4

100

Crossarchus obscurus

5,452.19

6,419.79

100

Felis aurata

4,744.46

6,009.13

100

Genetta tigrina/servalina

1,822.29

790.6

7.7

100

Herpestes sanguinea

7,394.83

6,644.5

14.3

100

Lutra maculicolis

5,837.50

100

Panthera pardus

9,091

8,338.91

66.7

33.3

Poiana richardsoni

2,555.46

2,438.79

26.1

90.9

9.1

Group total

4,882.33

2,340.02

112

14.3

96.9

3.1

Dendrohyrax dorsalis

4,150.44

570.72

11.1

100

Group total

4,150.44

570.72

11.1

100

276.13

302.86

222

14.4

100

94.4

4.9

0.7

Carnivores

779.94

9.1

Hyrax

Pangolins

Phataginus tricuspis Smutsia gigantea

–

Group total

–

224

–

99.1

0.9

Fa & García Yuste

Table 3. (Cont.) Groups Species

Capture rate Mean

Wastage

Hm (%)

Sh Other

Primates

Cercocebus torquatus

Cercopithecus cephus

2,192.93

2,221.31

17.9

82.1

Cercopithecus nictitans

1,452.37

1,537.8

2.3

30.2

69.8

Cercopithecus pogonias

6,118.79 10,597.63

36.4

63.6

22.5

Colobus satanas Galago alleni Gorilla gorilla Mandrillus sphinx Miopithecus onguensis Pan troglodytes

–

434.49

–

111

77.5

–

100

–

100

601.13

479.94

103

3.9

47.6

52.4

3,194.5

3,003.59 17,630.5

521.4

100

1,964.32

46.2

53.8

17,457.76

Perodicticus potto

4,907.67

1,339.97

99.5

0.5

Group total

4,392.89

5,307.74

329

1.8

33.6

66.4

Rodents

Atherurus africanus Cricetomys emini Funisciurus lemniscatus

90.78

43.2

619

12.1

100

437.61

446.17

177

11.9

100

3,460.15

100

17,975.36

100

517.6

100

–

100

4,496.55

Heliosciurus rufobrachium 17,264.5 Myosciurus pumilio

4,920

Protoxerus stangeri

–

Thryonomys swinderianus

1,317.42

1,125.89

Group total

4,754.48

6,462.1

826

11.6

99.5

0.5

Tubulidentate

Orycteropus afer

6,389

–

35.6

2.2 62.2

Group total

6,390

–

35.6

2.2 62.2

Ungulates

Cephalophus callipygus Cephalophus dorsalis Cephalophus montícola

1,707.34

1,678.91

17.9

100

665.34

1,041.98

128

11.7

82.19

39.46

658

11.9

0.9

100

33.3

100

98.4 99.1

1.6

Cephalophus nigrifrons

1,138.5

Cephalophus ogilbyi

3,374.67

Cephalophus sylvicultor

3,425.5

913.27

13.3

100

Hyemoschus aquaticus

2,143.94

1,064.01

100

Neotragus batesi

5,471.5

1,297.54

100

Potamochoerus porcus

2,814.98

2,358.77

4.5

100

Syncerus caffer

8,843.75

8,688.57

Tragelaphus scriptus

6,472.46 13,146.1

Tragelaphus spekei

6,941.38

Group total

3,590.13

All groups

19.16

– 2,701.5

100

781.18

28.6

100

2,761.09

919

109

11.9

99.1

0.9

11.18 3,050

285

9.3

86.3

7.9

5.8

Animal Biodiversity and Conservation 24.1 (2001)

800

Number of hunted animals

700 600

Other Shotgun Snares

500 400 300 200 100 0

F Mr Ap My Jn Jl

Ag S O Month

D Ja

F Mr Ap

Fig. 7. Monthly changes (I 98–IV 99) in number of animals hunted by cable snaring and shotgun in the Monte Mitra, Equatorial Guinea. Fig. 7. Cambios mensuales (I 98–IV 99) en número de animales cazados mediante cepos y armas de fuego en el Monte Mitra, Guinea Ecuatorial.

5.0

Log capture rate

4.5 4.0 3.5 3.0 2.5 2.0

-1.0

-0.5

1.5 0.0

y = 0.2956x2 – 0.3963x + 3.3505 R 2 = 0.1266 0.5 1.0 1.5 Log body mass (kg)

2.0

2.5

3.0

Fig. 8. Relationship between body mass (kg) and estimated mean snare capture rate for bushmeat species hunted in Monte Mitra, Equatorial Guinea. Fig. 8. Relación entre masa corporal (kg) e índice estimado de capturas medias con cepo para especies cazadas y comercializadas como carne de selva en Monte Mitra, Guinea Ecuatorial.

Fa & García Yuste

Table 4. Details of hunting output within hunt camps in the Monte Mitra area, Equatorial Guinea: Taa. Total annual of animals; Tab. Total annual biomass (in kg); Bh. Biomass hunter–1; Bhd. Biomass hunting day–1; B. Biomass km–2; N. Number of animals km–2. Tabla 4. Detalles de producción cinegética en el área de Monte Mitra, Guinea Ecuatorial: Taa. Total anual de animales; Tab. Biomasa total anual (en kg); Bh. Biomasa por cazador; Bhd. Biomasa por día de caza; B. Biomasa por km-2; N. Número de animales por km-2.

Taa

Tab

Bhd

420.49

140.16

6.01

10.44

46.21

Anvira

273

1,348.92

84.31

5.31

7.09

35.04

Avis–ncha

396

1,868.85

133.49

6.87

5.25

24.79

Bisun

764

4,413.53

220.68

7.30

2.43

14.05

Ebang

37.92

12.64

6.32

–

Echun–ndje

128.02

64.01

3.56

–

131

716.79

79.64

5.19

4.77

–

71.75

–

Aben–nam

Enuc Esua–asas Eto–mbeng

10.4

56.89

Evuadulu

167.17

33.43

4.78

–

Kong

149.37

74.69

8.79

–

4.75

1.19

–

144.31

2.12

–

Mandjana Mitong–evina Mobun–nwuom

63.52

3.27

1.71

6.31

233

1,391.73

278.35

5.01

37.58

224.47

Sendje

64.06

32.03

16.02

–

Tom–asi

31.70

4.37

5.21

17.86

–

270.88

5.94

–

Ongam–nsok

Grand total

2,219

317.6

11,376.8

vulnerable to snaring was the blue duiker (table 4) and the least the crowned eagle (Stephanaeotus coronatus). The proportion of animals snared or shot was not correlated with body mass of the species. However, mean capture rates were correlated with body mass; smaller and larger– bodied species were significantly less vulnerable than medium–sized animals (fig. 8). All taxonomic groups, except reptiles and primates, were caught mainly with snares. A significant proportion (28%) of reptiles was caught by other means (gathered by hand). In the case of primates, most individuals were shot (66.4%), but number of animals snared varied among species; nocturnal primates (Allen’s squirrel galago Galago alleni and potto Perodictus potto) being vulnerable only to snares. Because snares are non–selective, captures should reflect sex and age ratios in the population (ALVARD, 1994). For the most hunted species (>100 captures)–two duikers ( C. montícola, bay duiker Cephalophus dorsalis), two rodents (A. africanus, giant pouched rat

Cricetomys emini), two primates (black colobus Colobus satanas, mandrill Mandrillus sphinx), a reptile (hinged tortoise Kynixis erosa), pangolin (tree pangolin Phataginus tricuspis) and a bird (guinea fowl Numida meleagris) —sex ratios were not significantly different to 1:1 in any species (table 5). In contrast, the age class ratios were significantly biased towards juveniles in all species (table 5). Sustainability of harvests For the 14 mammal species for which harvest information could be calculated, mean total extraction rate was 41.75±45.56 animals km-2 ranging from 10.27 animals km-2 in Bisun, to 148.2 animals km-2 in Ongam–nsok. In all camps, A. africanus and C. dorsalis were extracted at significantly higher rates than other species (table 6). The average extraction rates for these species were 12.22±14.50 for A. africanus and 10.44±11.05 for C. dorsalis. In four camps, A.

Animal Biodiversity and Conservation 24.1 (2001)

Table 5. List of bushmeat species recorded in the Monte Mitra area, Equatorial guinea, indicating average individual body mass (Abm, in kg) and numbers of animals of each age: J. Juveniles; A. Adults; T. Total; R. Ratio; Gt. Grand total. Tabla 5. Relación de las especies de carne de selva registradas del área de Monte Mitra, Guinea Ecuatorial, con indicación de la masa corporal individual media (ABM, en kg) y número de animales de cada edad: J. Jóvenes; A. Adultos; T. total; R. Ratio; Gt. Gran total.

Groups Species

Age class Abm

–

Sex

}

{

–

Snails Achatina spp. Reptiles

Bitis gabonica Chamaleo cristatus

0.7

0.5

Kynixis erosa

3.5

259

314

4.7

140

193

333

0.7

339

Osteolaemus tetrapis

31.8

3.1

4.1

Python sebae

35.2

3 1

Varanus niloticus

5.5

Total

–

3.9

376

462

4.4

225

253

478

0.9

4 92 490

Birds

Ceratogymna atrata

1.2

–

Francolinus lathanmi

0.3

3.5

Gypohierax angolensis

1.5

–

Haliaetus vocifer

2.8

–

Numida meleagris

1.4

104

114

Psittacus erithacus

0.4

Stephanoaetus coronatus

0.8

Unidentified spp.

0.5

–

10.4

114

–

120

136

12.4

2.3

1.3

Total

– 7.5

– 1.6

–

1 118

135

1.5

142

0.5

0.3

Carnivores

Bdeogale nigripes Civicttis civetta Crossarchus obscurus Felis aurata

1.8

0.8

Genetta servalina

2.2

9.5

0.9

Genetta tigrina

2.5

–

1.5

Herpestes sanguinea

0.6

–

0.8

Lutra maculicolis

5.3

–

Panthera pardus

47.5

–

3.0

Poiana richardsoni

0.6

Total

112

179

220

4.9

0.6

112

0.8

112

1.3

120

101

221

1.2

222

Hyrax

Dendrohyrax dorsalis

Pangolins

Phataginus tricuspis Smutsia gigantea Total

1.5 32.5

181

222

4.4 – 4.4

121

102

223

1 1.2

2 224

Fa & García Yuste

Table 5. (Cont.) Age class

Group Species

Sex

Abm

}

{

Cercocebus torquatus

7.8

–

Cercopithecus cephus

3.5

1.1

Cercopithecus nictitans

6.6

0.8

Cercopithecus pogonias

3.8

4.5

1.2

12.5

110

3.2

110

1.1

111

0.3

Primates

Colobus satanas Galago alleni Gorilla gorilla Mandrillus sphinx Miopithecus onguensis Pan troglodytes Perodicticus potto

133

–

101

0.3

101

0.9

103

1.3

1.6

0.9

1 0.5

1.2

Total

–

17.4

203

118

321

1.7

159

167

326

329

326

285

611

1.1

619

Rodents

Atherurus africanus

2.8

501

100

601

Cricetomys emini

1.1

129

174

2.9

173

1.2

177

Funisciurus isabella

0.2

–

1.5

Heliosciurus rufobrachium

0.3

–

Myosciurus pumilio

0.2

–

Protoxerus stangeri

0.8

Thryonomys swinderianus

5.1

–

657

148

805

440

374

814

61.0

–

Cephalophus callipygus

20.1

Cephalophus dorsalis

20.4

100

120

4.9

558

643

6.6

Cephalophus nigrifrons

13.9

Cephalophus ogilbyi

19.5

Cephalophus sylvicultor

52.5

6.5

Hyemoschus aquaticus

2.7

–

67.5

1.8

Total

4.4

1.2

826

Tubulidentate

Orycteropus afer

–

Ungulates

Cephalophus montícola

Neotragus batesi Potamochoerus porcus Syncerus caffer

2.5

128

1.1

128

337

320

657

1.1

658

0.5

1.5

2.3

1.4

–

2 22

285

0.5

Tragelaphus scriptus

4.6

1.4

Tragelaphus spekei

100

1.3

0.4

757

138

895

5.5

478

440

918

1.1

919

4.2 1,562

1,455

Total All groups Grand total

2,395

567 2,962

3.17

1.1 3,052

Animal Biodiversity and Conservation 24.1 (2001)

Table 6. Estimated productivity of mammal species in hunt camps in the Monte Mitra area, Equatorial Guinea. Comparisons of the extraction rates for all camps and production figures estimated in FA et al. (1995) are also given: *Species considered to be hunted unsustainably; Abn. Aben–nam; An. Anvira; Avn. Avis–ncha; Bi. Bisun; En. Enuc; Mon. Mobun–nwoum; Onn. Ongam–nsok; Toa. Tom–asi; Ahr. Annual harvest rate (Nos km2 yr–1); Cps. All camp ps; P. Production (Nos km2 yr–1) Tabla 6. Productividad estimada de especies de mamíferos en campos de caza de Monte Mitra, Guinea Ecuatorial. También se incluyen valores estimativos comparativos: *Especies cuya caza se considera insostenible; Abn. Aben–nam; An. Anvira; Avn. Avis–ncha; Bi. Bisun; En. Enuc; Mon. Mobun–nwoum; Onn. Ongam–nsok; Toa. Tom–asi; AHR. Índice de captura anual (número km2 yr–1); Cps. Todos los campos ps; P. Producción (número km2 yr–1).

Ahr Species

Abn

0.88

0.62

–

0.1

Anv

Cps

Mon

Onn

Toa

Mean

Ungulates–Red duikers

Cephalophus callipygus Cephalophus ogilbyi

Cephalophus sylvicultor 0.44

0.21 0.06 –

0.96

–

– 0.14

*3.25

0.57

0.82

1.12 1.31

–

0.65

–

*0.09

0.24 2.02

–

*0.65

–

*0.18

0.25 0.29

6.23

6.97

8.47 8.57

0.21

4.26

3.87 1.27

6.37 0.48 25.35

5.52 2.39

10.83 2.07 *43.55

Ungulates–Blue duiker

Cephalophus monticola 7.93 Rodents

Atherurus africanus Cricetomys emini

18.51 10.4 2.64

0.94

1.38 1.26

–

0.16

9.1

4.53 12.22 14.5 27.12 1.7

2.15

3.16 81.49

Monkeys

Cercopithecus cephus

0.44

0.42

0.05 0.22

0.64 0.24

0.65

–

0.33

0.22 1.18

Cercopithecus nictitans

0.44

0.62

0.27 0.09

*2.23 0.08

1.3

–

0.63

0.79 1.55

Colobus satanas

1.32

2.18

1.43 1.36

*2.87 0.72 *11.7

3.96 *3.19

3.86 2.72

Mandrillus sphinx

0.44

1.35 *1.33 0.57

*2.23 0.8

0.57 *1.97

2.84 0.79

Phataginus tricuspis

1.76

1.66

0.8

Potamochoerus porcus

0.88

0.1

0.05 0.1

–

*8.45

Pangolins 0.2

4.14 0.32

5.85

1.7

2.05

2.12 6.63

0.64

–

3.9

–

0.71

1.4

1.89

–

0.001

–

0.03

Pigs Apes

Gorilla gorilla All species

–

0.01

49.34 30.14 20.06 10.27 45.56 7.26 148.2

africanus was the most harvested species, but rates differed significantly from 5.52 animals km-2 in Avis-ncha to 43.55 animals km-2 in Ongamnsok. In two camps, Bisun and Enuc, C. dorsalis was the most heavily extracted species followed by A. africanus. Comparison between extraction rates and estimated production (table 6) showed that C. dorsalis was hunted unsustainably in all camps, the mandrill Mandrillus sphinx in four camps, black colobus Colobus satanas in two camps, and three other species (Peter’s duiker

23.22 41.751 50.02 136.09

Cephalophus callipygus, yellow–backed duiker Cephalophus sylvicultor, and spot–nosed guenon Cercopithecus nictitans) in one camp.

Discussion The aim of this study was to document the process of faunal extraction in a representative area of African moist forest. Through relatively unobtrusive and cost–effective means we were

able to gather data for an unprecedented number of hunters in an equally unprecedented number of hunting areas. This study also examines the impact of hunting on forest vertebrate communities over a reasonably long period and offers a new insight into ways of collecting valuable data for assessing sustainability. Our results point to trends in hunting performance and outcomes which have been observed elsewhere. For example, the number of captures and biomass extracted per hunter were correlated with the amount of time dedicated to hunting by each hunter and to size of areas operated by them. Equally, extraction rate was positively correlated with distance from the village since interference levels and hunting pressure decreases as distance from human habitation increases (INFIELD, 1988; LAHM, 1993; MUCHAAL & NGANDJUI, 1999). Animals killed per hunter per hunting day did not vary significantly throughout the study but total numbers killed declined during the same period. This effect can be explained by the amount of time hunters spent in the forest. During the early part of the study, number of days dedicated to hunting was high but this declined later on. Biomass extracted per month was observed to drop dramatically from the start to the end of the study. Whether there is overt feedback between returns during one month and the number of days spent hunting the following month is difficult to know. However, it is likely that this is happening, given that hunters knew each other and would discuss the state of the game in the forest. Perhaps an indication that previous knowledge of the possible condition of prey populations was present is the fact that number of snares set per day increased dramatically as number of days spent in the forest declined. Hunters would be attempting to maximise or keep constant their daily hunting returns by intensifying snaring activities. One of the most pervasive conclusions of our study is the importance of cable snares in supporting commercial hunting activities in the Monte Mitra region. Cable snares are probably the most widely used hunting method in African forests today (NOSS, 1998, 2000) because the method is affordable, easy to implement and very effective. Hunter return rates are high as a result, but not without severe consequences. Cable snares are indiscriminate and wasteful. Prey–selectivity exercised by other methods of hunting, especially more traditional techniques, is severely reduced. Species of any age or sex, exhibiting any terrestrial activity, of any speed and of mid–range body size are vulnerable to capture by cable snare. Only very small species, with insufficient body mass to trip the cable wire, and very large species, likely to overpower the mechanism, are left non–targeted. Large animals may be injured by the snare, which may in turn have implications for their survival and reproduction. Elephants, for example, may trigger

Fa & García Yuste

the cable snare with their trunks. Estimated wastage in our study was 9.7%, substantially lower than the 26.7% reported by NOSS (1998) for Bayanga hunters in the CAR. In the 1,010 km2 of the Monte Mitra study area, we estimated bushmeat offtake of over 2,000 total captures, around 10,000 kg of animal biomass annum-1. This amounts to 56 captures or 10 kg of bushmeat km-2 annum-1. This is a substantially larger extraction rate than elsewhere in Central African forests. For example, for a similar–sized hunting range, NOSS (1998) estimated only 9 captures km-2. The explanation for this, may be found in the density of cable snares used, since this was significantly higher in the entire Monte Mitra area (56 snares km-2) than in Bayanga (4.2 snares km-2). However, considerable between–hunt catchment differences existed in biomass and number of animals extracted. What determines variation in game productivity, within what is apparently the same forest, is not known and requires further investigation. Between–site disparity in hunter–kill profiles may be influenced by both the effort of hunters and the "catchability" of their prey (FA et al., submitted). Investigating human hunting behaviour may shed light on patterns of prey selectivity and how variations in habitat, prey availability and hunting methods influence the impact of hunting on prey populations. Studying the social organisation and behaviour of prey may enable predictions to be made concerning the response of species to different levels of harvesting (FITZGIBBON, 1998). Most studies report that hunters prefer large or medium–sized prey (FITZGIBBON, 1998; FA & PERES, 2001). Hunting in this study occurred throughout the year and no clear seasonal patterns in harvest rates were detected (although more long–term data are required). Furthermore, number of animals killed and biomass extracted declined dramatically in the first three months and then gradually until the end of the study. There was a clear decline in average body mass of prey since the start of the study. Even though overall hunter effort dropped during the study, biomass and number of animals per hunter also declined. This is indicative of depletion of the sites since extraction rates per hunter would have increased with a decline in hunter pressure. Hunters would select large animals in order to maximize the quantity of meat extracted from an area, per unit of hunter effort, in accordance with models of optimal foraging. The pattern emerging from this study indicates that larger prey is indeed taken first, but this is not hunter–led since most animals are caught by snares (although there is a body mass effect on vulnerability to snaring). Large prey are generally more profitable to hunters, as long as handling costs do not increase in proportion to body mass. With increasing hunting pressure, more of the smaller sized species are depleted (NEWING, 2001). The loss of these species, important in seed dispersal, will have serious long–term consequences

Animal Biodiversity and Conservation 24.1 (2001)

on the forest ecosystem (WRIGHT et al., 2000; MOORE, 2001). Inter–hunter variation in number of animals hunted was considerable in the Monte Mitra area. Essentially, extraction of game was directly proportional to the amount of time dedicated to hunting. All hunters in the study hunted game for profit and were dedicated full–time to this activity. An average of around 70% of all game hunted was sold by the hunter. Because of the detrimental effects of cable snares on wildlife, most Central African nations have banned this method. In the case of Equatorial Guinea this is not the case, but if cable snares had not been used in Monte Mitra during this study, only less than 9% of the documented prey would have been taken. The importance of snare hunting in increasing profitability for the hunter is then clear. Wildlife populations in Monte Mitra declined under the heavy hunting pressure during the study period. If they stabilised at new and lower levels, then current hunting may be sustainable, although this is unlikely given the emphasis on selling the meat to the Bata market. Alternatively, if hunting pressure of a site is not too intense, adjacent large tracts of undisturbed forest can buffer and replenish hunted areas, restocking game populations and therefore contributing to the sustainability of hunting in an area (FA & PERES, 2001). However, heavy hunting pressure, deforestation and habitat fragmentation of many areas disrupt the source-sink dynamics (NOVARO et al., 2000), leading to potential over–exploitation of populations. Our estimates of sustainability of a number of game species indicate that currently most species are overharvested. Conservation of the Monte Mitra region is impossible unless the hunting for profit issue in Sendje and adjoining villages is resolved. Conservationists will need to work with local residents, who have few alternative methods for finding food and earning an income, to find a solution to game exploitation. Bans on cable snares may be totally unenforceable by the reduced number of park guards operating in the Monte Alén national park, and equally such measures will generate considerable antagonism. Firearms may be permissible but are not a good alternative because of the costs involved and because of the much lower returns. The challenge is to reduce current levels of hunting and integrate human needs and expectations within conservation objectives for the region (NOSS, 1997; EVES & RUGGIERO, 2001).

Acknowledgements The fieldwork reported in this study was undertaken as part of the Proyecto CUREF in Rio Muni. We are most grateful to staff at the Monte Alén National Park, especially wardens José Ndong and Julián Nsihi, for field assistance. We

would also like to acknowledge the help and support of Adolfo Ncogo in mapping hunt catchments, and in providing us with invaluable logistic support during our visits. Manuel collected offtake data during the entire study period. We would also like to thank the hunters who contributed information to the study. We thank Guy Cowlishaw, John Oates and Helen Newing for comments on the manuscript.

References A LVARD , M. S., 1994. Conservation by native peoples: prey choice in a depleted habitat. Human Nature, 5 : 127–154. A PE A LLIANCE, 1998. Bushmeat: A recipe for extinction . Fauna and Flora International, Cambridge. BENNETT, E. L. & ROBINSON, J. G., 2000. Hunting for sustainability: The start of a synthesis. In: Hunting for sustainability in tropical forests: 499–519 (J. G. Robinson & E. L. Bennett, Eds.). Columbia University Press, New York. BEUDELS, R., 1998. Appui au développement d’une typologie de habitats et pour le rassemblement des donées faunistiques. Rapport de Synthèse, Informe Técnico CUREF–SUC.26, Bata, Equatorial Guinea. EVES, H. E. & RUGGIERO, R. G., 2000. Socioeconomics and the sustainability of hunting in the forests of northern Congo (Brazzaville). In: Hunting for sustainability in tropical forests: 427–454. (J. G. Robinson & E. L. Bennett, Eds.) Columbia University Press, Columbia. FA, J. E., 1991. Conservación de los ecosistemas forestales de Guinea Ecuatorial . IUCN, Cambridge and Gland. – 2000. Hunted animals in Bioko Island, West Africa: sustainability and future. In: Hunting for sustainability in tropical forests: 165–195. (J. G. Robinson & E. L. Bennett, Eds.) Columbia University Press, Columbia. FA, J. E., GARCÍA YUSTE, J. E. & CASTELO, R., 2000. Bushmeat markets in Bioko Island as a measure of hunting pressure. Conserv. Biol., 14: 1,602–1,613. FA, J. E., JUSTE, J., PEREZ DEL VAL, J. & CASTROVIEJO, J., 1995. Impact of market hunting on mammal species in Equatorial Guinea. Conserv. Biol., 9: 1,107–1,115. FA, J. E. & PERES, C. A., 2001. Game vertebrate extraction in African and Neotropical forests: an intercontinental comparison. In: Conservation of exploited species: 203–241 (J. D. Reynolds, G. M. Mace, J. G. Robinson, & K. H. Redford, Eds.). Cambridge University Press, Cambridge. FA, J. E. & PURVIS, A., 1997. Body size, diet and population density in Afrotropical forest mammals: a comparison with Neotropical species. J. Anim. Ecol., 66: 98–112. FA, J. E., RYAN, S. & BELL, D. J. (submitted). General patterns and inter–sote variation in bushmeat hunting in Afrotropical Forests. Conserv. Biol.

FITZGIBBON, C., 1998. The management of subsistence harvesting: behavioural ecology of hunters and their mammalian prey. In: Behavioral ecology and conservation biology: 449–473 (T. Caro, Ed.). Oxford University Press, Oxford. GARCIA YUSTE, J. E. & ENEME, F., 1997. Diagnóstico de las áreas críticas para la conservación . CUREF–SUC.2, Bata: Equatorial Guinea. H ART , J., 2000. Impact of sustainability of indigenous hunting in the Ituri forest, CongoZaire: A comparison of unhunted and hunted duiker populations In: Hunting for sustainability in tropical forests: 106–153 (J. G. Robinson & E. L. Bennett, Eds.). Columbia University Press, Columbia. INFIELD, M., 1988. Hunting, trapping and fishing in villages within and on the periphery of the Korup National Park. WWF Publication 3206/ A9.6, Goldaming, UK. JUSTE, J., FA, J. E., PEREZ DEL VAL, J., & CASTROVIEJO, J., 1995. Market dynamics of bushmeat species in Equatorial Guinea. J. Appl. Ecol., 32: 454–467. KINGDON, J., 1997. The Kingdon field guide of African mammals. Academic Press, London. LAHM, S. A., 1993. Ecology and economics of human/wildlife interaction in northeastern Gabon. Ph. D. Thesis, Univ. of New York. MOORE, P. D., 2001. The rising cost of bushmeat. Nature, 409: 775–777. MUCHAAL, P. K. & NGANDJUI, G., 1999. Impact of village hunting on wildlife populations in the western Dja Reserve, Cameroon. Conserv. Biol., 13: 385–396. N EWING , H., 2001. Bushmeat hunting and management: implications of duiker ecology and interspecific competition. Biodivers. and Conserv., 10: 99–118.

Fa & García Yuste

NOSS, A. J., 1997. Challenges to nature conservation with community development in central African forests. Oryx, 31: 180–188. – 1998. The impacts of cable snare hunting on wildlife populations in the forests of the Central African Republic. Conserv. Biol., 12: 390–398. – 2000. Cable snares and nets in the Central African Republic. In: Hunting for sustainability in tropical forests: 106–153 (J. G. Robinson & E. L. Bennett, Eds.). Columbia University Press, Columbia. NOVARO, A. J., REDFORD, K. H. & BODMER, R. E., 2000. Effect of hunting in source–sink systems in the Neotropics. Conserv. Biol., 14: 713–721. ROBINSON, J. G. & REDFORD, K. H., 1991. Sustainable harvest of Neotropical forest animals. In: Neotropical wildlife use and conservation: 415–429 (J. G. Robinson & K. H. Redford, Eds.). Chicago University Press, Chicago. ROBINSON, J. G., REDFORD, K. H. & BENNETT, E. L., 1999. Wildlife harvest in logged tropical forests. Science, 284: 595–596. SAYER, J. A., HARCOURT, C. S. & COLINS, N. M., 1992. The conservation atlas of tropical forests: Africa. Macmillan Publishers, Basingstoke. VENABLES, W. N. & RIPLEY, B. D., 1999. Modern applied statistics with S–Plus. Springer–Verlag, Berlin. WILKIE, D. S. & CARPENTER, J. F., 1999. Bushmeat hunting in the Congo Basin: an assessment of impacts and options for mitigation. Biodiv. and Conserv., 8: 927–955. WRIGHT, S. J., ZEBALLOS, H., DOMÍNGUEZ, I., GALLARDO, M. M., MORENO, M. C. & IBAÑEZ, R., 2000. Poachers alter mammal abundance, seed dispersal, and seed predation in a Neotropical forest. Conserv. Biol., 14: 227–239.

Animal Biodiversity and Conservation 24.1 (2001)

Composición y calidad de la dieta del ciervo (Cervus elaphus L.) en el norte de la península ibérica I. Garin, A. Aldezabal, R. García–González & J. R. Aihartza

Garin, I, Aldezabal, A., García–González, R. & Aihartza, J. R., 2001. Composición y calidad de la dieta del ciervo (Cervus elaphus L.) en el norte de la península ibérica. Animal Biodiversity and Conservation, 24.1: 53–63. Abstract Plant composition and quality of the red deer (Cervus elaphus L.) diet in the northern Iberian peninsula.—The feeding pattern of red deer varies greatly among the different European populations. The aim of our study was to elucidate the plant composition and the quality of the red deer diet in the Pyrenees (Northern Iberian peninsula). Over a one–year period, the red deer fed mainly on browse, pines being the main food. However, unlike other populations on the Iberian peninsula, consumption of herbaceous plants was higher than browse in the spring–summer period. Nevertheless, the diet of Pyrenean red deer shared some features with the Mediterranean populations such as browsing on woody legumes. Fecal nitrogen content, as an index of diet quality, showed low annual values with a marked decrease in winter. The overall feeding pattern was similar to that of other Central European populations. The large size of the surveyed population probably affected its high level of browse consumption and poor quality diet. Key words: Diet, Red deer, Plant composition, Quality. Resumen Composición vegetal y calidad de la dieta del ciervo (Cervus elaphus L.) en el norte de la península Ibérica.— El patrón de alimentación del ciervo es muy variable entre las diferentes poblaciones europeas. El objetivo de nuestro estudio fue determinar la composición vegetal y la calidad de la dieta del ciervo en los Pirineos (norte de la península Ibérica). Durante un periodo de un año, el ciervo se alimentó principalmente de plantas leñosas, siendo los pinos su principal alimento. Sin embargo a diferencia de otras poblaciones de la península Ibérica, el consumo de plantas herbáceas fue superior al de leñosas en primavera y verano. No obstante, la dieta del ciervo del Pirineo comparte algunas características con la de las poblaciones mediterráneas, como el consumo de plantas leguminosas leñosas. El contenido en nitrógeno fecal, como índice de calidad de la dieta, presenta valores anuales bajos con una marcada disminución en invierno. El patrón de alimentación global fue similar al de otras poblaciones centroeuropeas. El gran tamaño de la población estudiada influye probablemente en el alto nivel de consumo de leñosas y la baja calidad de su dieta. Palabras clave: Dieta, Ciervo, Composición vegetal, Calidad. (Received: 31 I 01; Conditional acceptance: 13 VI 01; Final acceptance: 3 VII 01) I. Garin & J. R. Aihartza, Zoologia eta Animali Zelulen Dinamika Saila, UPV/EHU, 644 PK, 48080 Bilbo, Basque Country, España (Spain).– A. Aldezabal, Landare–Biologia eta Ekologia Saila, UPV/EHU, 644 PK, 48080 Bilbo, Basque Country, España (Spain).– R. García–González, Instituto Pirenaico de Ecología (CSIC), Apdo. 64, 22700 Jaca, España (Spain).

ISSN: 1578–665X

Introducción La valoración del impacto del ciervo (Cervus elaphus) sobre los ecosistemas ha dado lugar a numerosos trabajos sobre sus hábitos alimentarios. Las primeras revisiones constataron la enorme variabilidad en la preferencia por las distintas especies vegetales, a partir de las cuales se intentó generalizar una clasificación de especies preferidas por el ciervo (KAY & STAINES, 1981). Así, algunos autores han constatado que el alimento más importante en la dieta del ciervo a lo largo del año son las especies leñosas de hoja caduca (DZIECIOLOWSKI, 1969); por otra parte, otros autores han enumerado una serie de especies clave para la dieta del ciervo (GOFFIN & DE CROMBRUGGHE, 1976). Probablemente, la plasticidad en el comportamiento alimentario del ciervo, definido como pastador– ramoneador (HOFMANN, 1989), le permita adecuar fácilmente el consumo de las diferentes categorías vegetales a los cambios tanto temporales como espaciales o geográficos de la disponibilidad de los recursos alimentarios. Las herbáceas son generalmente más abundantes en la dieta primaveral, mientras que la importancia de las leñosas aumenta a medida que nos adentramos en el otoño y el invierno (MITCHELL et al., 1977). Los diferentes regímenes de precipitación y temperatura en Europa dan lugar a variados patrones fenológicos de los grupos de plantas consumidas: por ejemplo, el comienzo tardío del periodo vegetativo de las latitudes más septentrionales condiciona que las herbáceas alcancen su máximo en la composición de la dieta durante el verano (MITCHELL et al., 1977); mientras que en regiones mediterráneas el máximo puede ser a finales de invierno o comienzos de primavera (RODRÍGUEZ– BERROCAL, 1978). La nieve llega a hacer anecdótica la presencia de herbáceas en la dieta invernal (DZIECIOLOWSKI, 1969), mientras que en zonas de influencia oceánica o mediterránea, de inviernos más templados, la proporción de herbáceas durante la época llamada desfavorable supera el 20% y puede alcanzar el 50% de la dieta (JENSEN, 1968; RODRÍGUEZ–BERROCAL, 1978; VENERO, 1984; GROOT–BRUINDERINK & HAZEBROEK, 1995). En las zonas deforestadas de Escocia su dieta es mayoritariamente herbácea (MITCHELL et al., 1977) y en las poblaciones polacas más forestales, la hierba es cuantitativamente poco importante en la dieta media anual (DZIECIOLOWSKI, 1969). La mencionada plasticidad permite a la especie mantenerse en un mismo lugar a pesar de que la disponibilidad de alimento varíe notablemente de una época a otra del año y habitar ambientes vegetales radicalmente opuestos. La variación estacional en la alimentación está acompañada además de cambios en la función ruminal, que ayudan a reducir el impacto de la variación de la calidad de la dieta en la efectividad de la digestión (JIANG & HUDSON, 1996; LENTLE et al., 1996).

Garin et al.

En los ungulados domésticos el descenso de la calidad de la ingesta coincide con la reducción de las ganancias corporales de los individuos (LENG, 1990; VOGEL et al., 1993). Si la calidad de la dieta del ciervo desciende por debajo de un umbral, el individuo no satisface sus requerimientos, comenzando a movilizar sus reservas y a perder peso (GATES & HUDSON, 1981). Por ello, la monitorización de la calidad de la dieta puede ofrecer un diagnóstico poblacional. En condiciones naturales y con animales salvajes, los índices de calidad nutricional utilizados de forma habitual se estiman a partir de las características químicas de las heces (PUTMAN, 1984). Los altos requerimientos de proteína (Nx6,25) en los animales en relación a su disponibilidad en las plantas, convierte al nitrógeno ingerido en un indicador importante de la calidad de la dieta de los herbívoros (MATTSON, 1980). La posibilidad de que el contenido de nitrógeno en las heces pueda estar a su vez relacionado con el contenido en nitrógeno de la dieta y la facilidad en la obtención de muestras fecales, han convertido el nitrógeno fecal en un índice muy utilizado (GOGAN & BARRET, 1994; MASSEI et al., 1994; MERRILL et al., 1995). El patrón de variación temporal del nitrógeno fecal está determinado principalmente por el desarrollo fenológico de las plantas consumidas, ya que el contenido de nitrógeno en las plantas depende de dicho factor. Así, cuando el material vegetal se encuentra en las primeras etapas del desarrollo, contiene el pico máximo de nitrógeno, y decae a medida que nos acercamos al invierno (MATTSON, 1980). La misma tendencia temporal ha sido observada en el nitrógeno fecal (VAN SOEST, 1994). Al igual que otras muchas especies de cérvidos, el ciervo ha sido objeto durante los últimos siglos de numerosas reintroducciones y traslados a lo largo del mundo, y hoy en día podemos encontrarlo fuera de su área natural de distribución, como por ejemplo en Argentina, Nueva Zelanda o Australia (WHITEHEAD, 1972; PUTMAN, 1988). Como en otros macizos montañosos de Europa, el ciervo también ha sido reintroducido en ambas vertientes de los Pirineos (TEILLAUD et al., 1991). Las poblaciones de Euskal Herria, Aragón y Cataluña proceden de animales provenientes del sur de España y reintroducidos en los años 50 y 60. Aunque la alimentación del ciervo en las poblaciones de la región mediterránea ha sido ampliamente estudiada (MARTÍNEZ 1996; ÁLVAREZ & RAMOS, 1991; SORIGUER et al., 1994; GARCÍA– GONZÁLEZ & CUARTAS, 1992), la dieta en el norte de la península ibérica es prácticamente desconocida. El objetivo del presente estudio ha sido cubrir ese vacío en el conocimiento del ciervo, investigando su dieta y calidad en una población pirenaica. Por otro lado, se pretende determinar el modelo trófico (centroeuropeo o Mediterráneo) al que pertenece el ciervo en el, Pirineo dado

Animal Biodiversity and Conservation 24.1 (2001)

que las características geográficas de ambas regiones confluyen en este área. Asímismo, se ha valorado y discutido sus hábitos alimentarios estacionales en relación a las características nutritivas de la vegetación. Área de estudio El presente trabajo se ha realizado en los valles de La Garcipollera y Cenarbe (norte de Huesca), que conforman la Reserva de Caza de La Garcipollera de 55,8 km2 de extensión. La precipitación anual es 1.051 mm y la temperatura media anual 9,7°C. Los meses de verano son los más secos y las temperaturas medias de los meses invernales superiores a los 0°C. La zona se encuentra entre los 850 m y los 2.200 m de altitud y está prácticamente deshabitada en la actualidad. La vegetación está constituida principalmente por repoblaciones de negral de Austria (Pinus nigra subsp. nigra) y de pino albar (P. sylvestris), abetares (Abies alba) de reducida extensión, quejigares (Quercus humilis), algunos prados y los pastos supraforestales. La población de ciervos procede de reintroducciones realizadas en los años 60 con 37 individuos procedentes de los Montes de Toledo. A mediados de los años 90 se estimó que la densidad era de 35 ind./km2, la sex– ratio de tres hembras por macho y la fertilidad del 51% (MARCO et al., 1996; GARIN, 1997). No se han observado corzos (Capreolus capreolus) en la reserva y los jabalíes (Sus scrofa) son probablemente abundantes debido a la ausencia de caza (GARIN, 1997).

Material y métodos Dieta La composición de la dieta del ciervo se determinó mediante el análisis microhistológico de las heces (SPARKS & MALECHEK, 1968; HOLECHEK et al., 1982; CUARTAS & GARCÍA–GONZÁLEZ, 1996). Este método permite estudiar la dieta de una especie animal sin causar ninguna molestia a los individuos y es apto para medios con escasa visibilidad, donde existen dificultades para observar directamente los hábitos alimentarios de la especie. La aparición de fragmentos vegetales en las heces depende de su resistencia a la degradabilidad durante los procesos digestivos del herbívoro. Por ello, las epidermis de las especies menos resistentes se ven subestimadas respecto a otras con mayor grado de resistencia. Aquellas partes de la planta que por carecer de características específicas son difícilmente identificables en las heces (p.e. las partes leñosas), se subestiman igualmente (H OLECHEK & VALDEZ, 1985). En consecuencia, este método no refleja de forma precisa la cantidad relativa de cada especie o grupo de especies consumida por el animal (P UTMAN, 1984). La técnica micro-

histológica permite, sin embargo, clasificar las plantas consumidas por el herbívoro según su orden de importancia en la dieta y seguir su variación temporal y su diversidad (CUARTAS & GARCÍA–GONZÁLEZ, 1996). Desde mayo de 1993 hasta abril de 1994 se recogieron mensualmente en 4 estaciones de muestreo un mínimo de 9 deposiciones fecales frescas por estación, correspondiente a individuos adultos (según ÁLVAREZ, 1994). En una de las estaciones no pudieron recogerse excrementos en mayo y tampoco pudieron recolectarse en ninguna de ellas durante junio. La localización de las estaciones de recogida reflejó la diversidad de ambientes altitudinales y vegetales del área de estudio. Los excrementos se guardaron en bolsas de plástico, que se congelaron a –20°C hasta su procesamiento. Tras la homogeneización de las deposiciones, se separaron 5cc de cada deposición que se mezclaron de acuerdo a su estación y mes de muestreo. La contribución de material fecal de cada deposición a la mezcla fue similar. Las mezclas se prepararon siguiendo el protocolo de C UARTAS (1992) y se guardaron en acetoformol antes de la identificación de las epidermis vegetales. Para el conteo de cutículas se prepararon cinco portaobjetos por mezcla y mes, y en la medida que fue posible, se identificaron hasta el nivel de especie todas las epidermis interceptadas a lo largo de tres transectos dispuestos regularmente en el porta (SEBER & PEMBERTON, 1979). Las epidermis se clasificaron en tres grupos: 1. Graminoides, en las que se incluyen las gramíneas, las ciperáceas y las juncáceas; 2. Dicotiledóneas, en las que se incluyen todas las plantas herbáceas correspondientes a este grupo más las monocotiledóneas no gramínoides (es decir, liliaceas, iridáceas, orquídeas,…); 3. Leñosas, tanto de porte arbustivo como arbóreo. Debido a que la determinación de forma unificada de la disponibilidad del estrato herbáceo y del estrato leñoso no estuvo al alcance del equipo de trabajo, no se ha podido realizar un análisis cuantitativo de la selección de la dieta. Nitrógeno fecal Hemos valorado la calidad de la dieta del ciervo a través del nitrógeno de las heces asumiendo que el nitrógeno fecal y el nitrógeno de la dieta están relacionados (L ESLIE & S TARKEY , 1985; PUTMAN & HEMMINGS, 1986; IRWIN et al., 1993). El nitrógeno fecal tendría una respuesta sigmoidal a las variaciones del nitrógeno de la dieta: cuando el contenido del nitrógeno de la dieta es alto el nitrógeno puede ser tóxico y por ello se excretaría relativamente más; por el contrario, cuando el nitrógeno de la dieta es muy bajo se incrementa la capacidad de retención. En tramos cortos de variación del nitrógeno en la dieta la relación entre el nitrógeno de la dieta

Garin et al.

Graminoides

Dicotiledóneas

Leñosas

100

11 Mes

Fig. 1. Variación mensual del porcentaje de los grupos de plantas en los excrementos de ciervo de La Garcipollera. Fig. 1. Monthly variation of the percentage of plant groups in the fecal pellets of red deer from La Garcipollera.

y nitrógeno fecal sería lineal (PUTMAN , 1984). Esta relación lineal depende, sin embargo, de varios factores, como el contenido en taninos, la digestibilidad del alimento o el estado fisiológico del animal (H OBBS, 1987; ROBBINS et al., 1987; HOWERY & PFISTER , 1990). Sin embargo, su efecto sobre la relación entre el nitrógeno fecal y el nitrógeno de la dieta en ambiente natural puede no ser tan importante (M OULD & ROBBINS, 1981; H ANLEY et al., 1992; CAUGHLEY & SINCLAIR , 1994) lo que permite el uso del nitrógeno fecal como valor indicativo de la calidad de la dieta. Debido a que las inclemencias meteorológicas y el lavado del nitrógeno de la muestra fecal previo a su recolección afecta el valor del nitrógeno fecal, la recolección de heces debe realizarse antes de 24 días a partir de su deposición (JENKS et al., 1990). En nuestro caso, la apariencia de las muestras (color, brillo y mucosidad) aseguró su recolección en menos de una semana desde su deposición. Parte de las deposiciones se secaron en estufa a 90°C, se molieron hasta obtener partículas menores a 1 mm, y se guardaron en bolsa de plástico previo a su análisis, según el procedimiento de Kjeldahl.

Resultados Composición florística de la dieta En total se recogieron 791 deposiciones y se identificaron 18.093 fragmentos vegetales, es decir, una media de 421 por mes y estación de muestreo. En el total anual, el 51% de los fragmentos vegetales corresponde a la fracción leñosa (Intérvalo de Confianza, IC = 57%÷45%), siendo el grupo de plantas más consumido, s e g u i d o d e l a s g r a m i n o i d e s (x = 36%, IC = 40%÷32%) y por último las dicotiledóneas fueron las menos representadas en la dieta (x = 10%, IC = 13%÷7%) La variación mensual de los tres tipos de plantas consumidos se ha representado en la figura 1. El alimento leñoso tiene un peso importante a lo largo de todo el año, con máximo invernal —durante enero y febrero las plantas de origen leñoso sobrepasaron las tres cuartas partes de las epidermis determinada— y mínimo entre abril y septiembre —únicamente una cuarta parte en julio— que corresponde a la época en la que las herbáceas cobran mayor importancia en la dieta del ciervo de La

Animal Biodiversity and Conservation 24.1 (2001)

0 4

9 10 11 12 1 Mes

Brachypodium spp.

2 3

Carex spp.

10 11 12 1 Mes

E. horridum y G. scorpius Enebro

Festuca rubra

Pinos Zarza

Fig. 2. Variación mensual del porcentaje de las herbáceas (A) y las leñosas (B) más importantes en la dieta del ciervo de la Garcipollera. Fig. 2. Monthly variation of the percentage of most important grass (A) and browse (B) species in diet of red deer from La Garcipollera.

Garcipollera (> 50%). Las dicotiledóneas fueron importantes durante la primavera y el verano, y a partir de entonces disminuyeron en otoño hasta casi desaparecer en invierno. En La Garcipollera el máximo anual de las herbáceas dicotiledóneas fue en julio, pasando a ser el grupo más representado. Además en ese mes las graminoides mostraron una pequeña disminución. El nivel de degradación general de las epidermis impidió que numerosos fragmentos (42,6%) fueran identificados a nivel de especie, aunque si lo fueron a nivel de categoria vegetal. Entre las especies o taxones identificables se han distinguido 15 graminoides, 20 dicotiledóneas y 14 leñosas. Entre ellas sólo 7 exhibieron una frecuencia de aparición anual mayor que 1%: las graminoides Carex spp. (x = 6,3%; IC al 95% = 8,2÷4,7%), Festuca rubra (x = 6,1%, IC = 9,2÷3,5%) y Brachypodium spp. (2,7%, IC = 3,6÷1,9%), y entre las leñosas, los pinos (x = 14,2%, IC = 22,8÷7,4%), el enebro (x = 4,5%, IC = 7,0÷2,6%), Echinospartum horridum / Genista scorpius (x = 4,8%, IC = 8,2÷2,0%) y Rubus sp. (x = 1,0%, IC = 1,7÷0,5%). La tendencia mensual de esas plantas mostró el aumento de las leñosas (excepto la zarza) y la disminución de las herbáceas durante el invierno (fig. 2). Además, las leguminosas leñosas también incrementaron puntualmente su presencia en la dieta de julio y el consumo de Festuca rubra aumentó mucho entre marzo y mayo.

Nitrógeno fecal Los valores obtenidos del nitrógeno fecal dependieron significativamente del mes de recogida (ANOVA, F = 214,1, p < 0,001, g.l. = 9). La variación de la media mensual del nitrógeno fecal a lo largo del año se muestran en la figura 3. Las diferencias encontradas entre los meses (test a posteriori PLSD al 1% del nivel de significación) muestran un patrón cíclico del nitrógeno fecal a lo largo del año.

Discusión Por lo general, las especies y grupos de plantas más importantes encontrados en las heces coinciden con las plantas más abundantes en el área de estudio (GARIN, 1997). El pino (Pinus nigra, P. sylvestris y P. uncinata) ha sido la planta más abundante encontrada en la dieta del ciervo de La Garcipollera. Las variaciones en la frecuencia mensual de las leñosas en la dieta están asociadas en gran medida a los cambios en el consumo de los pinos. En general, cuando están disponibles, las coníferas pueden alcanzar un alto nivel de consumo (JODRA, 1986; GROOT–BRUINDERINK & HAZEBROEK, 1995), aunque sobre todo constituyen un alimento invernal (JENSEN, 1968; GEBZCYNSKA, 1980; HOMOLKA, 1993, 1995). La relevancia del enebro para la dieta del ciervo está sujeta a cierta controversia por que la

Garin et al.

%NF 2,75 a

2,50 2,25 c

2,00 d

d 1,75

e f

f 1,50 1,25 1,00

11 Mes

Fig. 3. Variación mensual del % del nitrógeno fecal (%NF) de ciervo en La Garcipollera. Las barras muestran el intervalo de confianza al 95% y las letras representan las diferencias entre los meses (p < 0,01). Fig. 3. Monthly variation of the % of red deer fecal nitrogen (%NF) in La Garcipollera. Rods show 95% confidence intervals and letters show differences between months (p < 0.01).

preferencia por él varía con la población estudiada (DZIECIOLOWSKI, 1969; GOFFIN & DE CROMBRUGGHE, 1976). El enebro es en La Garcipollera un elemento significativo en la alimentación del ciervo. Al igual que sucede con los pinos, su mayor consumo durante el otoño/invierno no coincide con el óptimo anual en su valor nutritivo (GARIN et al., 1996). La abundancia de pinos y enebro en la dieta parece determinado por la limitación en la disponibilidad de alimentos más nutritivos en el área de estudio. Los brotes de los arbustos leguminosos, Echinospartum horridum y Genista scorpius, se consumieron con mayor intensidad durante la primavera y a comienzos del verano. No obstante el consumo de estas especies, sobre todo de su parte leñosa, se produjo durante todo el año y de forma importante durante enero y febrero, lo que pudo ser debido a la reducción drástica de alimento por las persistentes nevadas en aquellos meses. La calidad de ambos arbustos es baja durante el invierno a causa de la elevada lignificación de las ramillas y la ausencia de hojas. El consumo de leguminosas de porte arbustivo parece característico del área mediterránea, sobre todo durante la primavera (ÁLVAREZ & RAMOS, 1991). La zarza (Rubus spp.) es uno de los alimentos más comunes en la dieta del ciervo en Europa, presente desde el sur de España (VENERO, 1984) hasta Polonia oriental (GEBCZYNSKA, 1980).

Su incidencia en la dieta es desigual entre las poblaciones, aunque parece complementarse en la dieta con las especies arbóreas de hoja caduca, como el roble, allí donde ambas están presentes. Y al igual que éstas, por regla general, el consumo de la zarza disminuye durante el invierno. Tanto la zarza como en general el resto de las rosáceas arbustivas contienen poca cantidad de lignina y fibra neutrodetergente durante el periodo vegetativo, y por lo tanto pueden constituir un aporte de alimento potencialmente muy digerible, a pesar de su elevada cantidad de taninos (GARIN et al., 1996). Cuando son abundantes, los ciervos prefieren las gramíneas perennes de hoja estrecha a las de hoja más ancha (CLUTTON–BROCK et al., 1982), más bastas y menos digeribles por lo general (KAY & STAINES, 1981). Así, en la zona europea de mayor influencia oceánica, Deschampsia flexuosa , Festuca sp. o Agrostis sp. son la gramíneas más consumidas (JENSEN, 1968; SHERLOCK & FAIRLEY, 1993; GROOT–BRUINDERINK & HAZEBROEK, 1995). En La Garcipollera Festuca rubra es la gramínea de hoja estrecha más abundante (GARIN, 1997) y una de las más nutritivas (KAY & STAINES, 1981; GarcíaGONZÁLEZ & ALVERA, 1986). En el área de distribución centroeuropea del ciervo la gramínea forestal más abundante parece ser Calamagrostis arundinacea (DZIECIOLOWSKI, 1969; BOROWSKI & KOSSAK, 1975; HOMOLKA, 1995), que es poco apete-

Animal Biodiversity and Conservation 24.1 (2001)

Primavera

Verano

Otoño

Invierno Paises Bajos G ROOT–BRUINDERINK & H AZEBROEK ((1995) 1995)

Escocia M ITCHELL et al. (1977)

Escocia Colquhoun, 1971 en M ITCHELL et al. (1977)

Polonia 1969) D ZIECIOLOWSKI ((1969)

A ndalucía V ENERO (1 984) (1984)

Castilla–La Mancha 1991) Á LVAREZ & RAMOS ((1991)

La Garcipollera

Graminoides

Dicotiledóneas

Leñosas

Bellotas/hayucos

Fig. 4. Proporción de los tipos de alimento más importantes en la dieta estacional del ciervo en Europa. Fig. 4. Proportion of most important food types in the seasonal diet of European red deer.

cible (HEROLDOVÁ, 1993). Esta planta es consumida principalmente en primavera (DZIECIOLOWSKI, 1969; HOMOLKA, 1995), al igual que otras gramíneas de conocida poca calidad en otras poblaciones, como Molinia coerulea (KAY & STAINES, 1981), o incluso, en menor medida, Brachypodium spp. (CABALLERO, 1985; ASCASO, 1990). Las ciperáceas (Carex sp. mayoritariamente) son un alimento con una fre-

cuencia muy baja en la dieta (JENSEN, 1968; D ZIECIOLOWSKI , 1969; PICARD & GEGOUT , 1992; SHERLOCK & FAIRLEY, 1993), aunque aumenta ligeramente su importancia en el invierno, al igual que en nuestro estudio. Su inferior calidad respecto a gramíneas más apetecidas (DZIECIOLOWSKI, 1969; KAY & STAINES, 1981) condiciona su relevancia en la dieta, sobre todo, cuando este tipo de gramíneas

abundan (p. e. D. flexuosa o F. rubra). La persistencia vegetativa de los cárices durante el invierno favorece probablemente su consumo, sobre todo cuando otras graminoides están secas. Por otro lado, la aparición de la nieve en La Garcipollera a finales de diciembre de 1993 y su persistencia durante los meses de enero y febrero de 1994 estuvo ligado a un descenso brusco en la frecuencia de los alimentos herbáceos. En las zonas más continentales, cuando la nieve cuaja y cubre el suelo las especies herbáceas disminuyen en la dieta (DZIECIOLOWSKI, 1969; JODRA, 1986; PICARD & GEGOUT, 1992). La aparición de las quercíneas en las heces ha sido reducida. A pesar de que la extensión de los bosques formados por el quejigo no es pequeña (399 Ha), el ramoneo en los años previos a este estudio puede ser la causa de su baja disponibilidad actual, tal y como se ha sugerido para las especies arbóreas más preferidas en otras poblaciones de ciervos (DZIECIOLOWSKI, 1969; SORIGUER et al., 1994). En la región mediterránea, cuando los ciervos habitan bosques naturales maduros, el alimento leñoso más consumido proviene de las quercíneas (ÁLVAREZ & RAMOS, 1991; CUARTAS, 1992; MARTÍNEZ, 1996). En la Europa continental también el roble pedunculado (Quercus robur) es consumido todo el año cuando está presente (BOBEK et al., 1972), incluso durante el invierno (GEBCZYNSKA, 1980; MÁTRAI & KABAI, 1989; PICARD & GEGOUT, 1992). Además de los tipos de alimento utilizados en este estudio, algunas poblaciones aprovechan también los frutos (de fagáceas) sobre todo en otoño e invierno (VENERO, 1984; PALACIOS et al., 1989; PICARD & GEGOUT, 1992; GROOT–BRUINDERINK & HAZEBROEK, 1995). La diferencias en la disponibilidad de ese alimento pueden explicar la desigual incidencia de la montanera en la dieta del ciervo en Europa (fig. 4), aunque la coexistencia del ciervo con alguna(s) de las muchas especies que también utilizan los frutos forestales puede ser en parte responsable de esas diferencias. La ausencia de bellotas en la dieta del ciervo en La Garcipollera puede deberse a que su producción falló en el otoño de 1993 (obs. pers.). Aunque, por otro lado, hay que considerar que la técnica de determinación microhistológica de la dieta puede no detectar los frutos de las fagáceas (ÁLVAREZ & RAMOS, 1991). El consumo otoñal de los frutos de quercíneas, que es un alimento de alta digestibilidad y contenido energético (ROBBINS, 1993; FOCARDI et al., 1995), puede ayudar a mejorar la condición corporal de los individuos al comienzo del invierno (JACKSON, 1974). Es más que probable que la población de La Garcipollera recurra a las bellotas como alimento otoñal en los años de buena fructificación. El valor anual del nitrógeno fecal observado en La Garcipollera es bajo en relación a otras poblaciones de ciervos (GOGAN & BARRET, 1994; LESLIE & STARKEY, 1985). La reducida calidad de las leñosas consumidas por el ciervo en La

Garin et al.

Garcipollera (GARIN et al., 1996) mantendría los niveles anuales de nitrógeno ingerido relativamente bajos. La relación entre las leñosas y un nitrógeno fecal reducido ya ha sido sugerida por otros autores (HODGMAN & BOWYER, 1986). En general, la dieta del ciervo en La Garcipollera se caracteriza por un elevado consumo de leñosas en el otoño/invierno y de las herbáceas en primavera/verano. El tipo de alimento varía entre poblaciones de ciervos, lo cual es debido a variaciones en la disponibilidad de los recursos (MARTÍNEZ, 1996). Asímismo, el patrón estacional de la dieta varía enormemente entre las distintas poblaciones de ciervos (fig. 4). Es posible distinguir tres patrones en la alimentación estacional del ciervo en Europa. El oceánico, situado en las Islas Británicas y en toda la franja costera de la Europa Occidental, con preponderancia de las herbáceas, y los brezos (familia Ericaceae) como representantes principales de las leñosas (JENSEN, 1968; MITCHEL et al., 1977; G ROOT –B RUINDERINK & H AZEBROEK , 1995). El centroeuropeo, que se extiende por el interior del continente y que en las poblaciones forestales alterna la preponderancia de herbáceas y leñosas (Sablina, 1955 en BOROWSKI & KOSSAK, 1975; DZIECIOLOWSKI, 1969). Por último, el tipo mediterráneo, distribuido en el área del mismo nombre, con preponderancia de especies leñosas todo el año (VENERO, 1984; ÁLVAREZ & RAMOS, 1991; GARCÍA–GONZÁLEZ & CUARTAS, 1992). El patrón alimentario general observado en el presente estudio sugiere que la dieta del ciervo en La Garcipollera está encuadrada en el modelo centroeuropeo (fig. 4). Además, la inversión en la relación entre dicotiledóneas y graminoides observada en la dieta de La Garcipollera también se repite en otras poblaciones forestales centroeuropeas (DZIECIOLOWSKI, 1969; HOMOLKA, 1995). La mayor calidad de las dicotiledóneas durante el verano, coetánea a la maduración y reducción en la calidad de las gramíneas (HANLEY & MCKENDRICK, 1983; FILLAT et al., 1989), puede explicar su alternancia en la dieta durante primavera y verano. Por otro lado, el patrón mensual del nitrógeno fecal observado en La Garcipollera es similar al encontrado en los grandes ungulados de regiones templadas (GATES & HUDSON, 1981; LESLIE & STARKEY, 1985): disminución en el otoño/invierno y incremento durante la primavera/verano. Por el contrario, en las regiones mediterráneas, las concentraciones mínimas se adelantan al verano/otoño y las máximas al invierno/primavera (GOGAN & BARRET, 1994; MASSEI et al., 1994). Sin entrar en consideraciones sobre la incidencia que la disponibilidad de alimento puede ejercer sobre la calidad de la dieta en las poblaciones de ciervos, parece evidente una tendencia general: el descenso del componente herbáceo en la dieta durante el o los periodos desfavorables. En ese sentido, el periodo desfavorable en la alimentación del ciervo en La Garcipollera se

Animal Biodiversity and Conservation 24.1 (2001)

produce durante el invierno, aunque la ingestión ininterrumpida de alimentos de poca calidad (pinos y enebro principalmente) a lo largo del año indica una situación no óptima, motivada probablemente por la elevada densidad.

Agradecimientos El Gobierno Vasco subvencionó este trabajo mediante una beca predoctoral al primer autor y el Instituto Pirenaico de Ecología (CSIC) aportó su infraestructura y personal. El Dr. J. Carranza aportó interesantes comentarios al manuscrito original.

Referencias ÁLVAREZ, G., 1994. Morphological variability and identification of deer pellets in central Spain. Folia Zoologica, 43: 25–37. Á LVAREZ , G. & R AMOS , J., 1991. Estrategias alimentarias del ciervo (Cervus elaphus L.) en Montes de Toledo. Doñana, Acta Vertebrata, 18: 63–99. ASCASO, J., 1990. Estudio fitocenológico y valoración de los recursos pastorales de las zonas forestales y arbustivas del Prepirineo aragonés. Institución Fernando el Católico, Zaragoza. BOBEK, B., WEINER, J. & ZIELINSKI, J., 1972. Food supply and its consumption by deer in a decidous forest of southern Poland. Acta Theriologica, 17: 187–202. BOROWSKI, S. & KOSSAK, S., 1975. The food habits of deer in the Bialowieza Primeval Forest. Acta Theriologica, 20: 463–506. CABALLERO, R., 1985. Hábitat y alimentación del ciervo en ambiente mediterráneo. ICONA, Madrid. CAUGHLEY, G. & SINCLAIR, A. R. E., 1994. Wildlife ecology and management. Blackwell Science, London (UK). CLUTTON–BROCK, T. H., GUINNESS, F. E. & ALBON, S. D., 1982. Red deer: behaviour and ecology of the two sexes. University of Chicago Press, Chicago. CUARTAS , P., 1992. Herbivorismo de grandes mamíferos en un ecosistema de montaña mediterránea. Tesis doctoral, Universidad de Oviedo. CUARTAS, P. & GARCÍA–GONZÁLEZ, R., 1996. Review of available techniques for determining the diet of large herbivores from their faeces. Oecologia Montana, 5: 47–50. DZIECIOLOWSKI, R., 1969. The quantity, quality and seasonal variation of food resources available to red deer in various environmental conditions of forest management. Forest Reseach Institute, Warsaw. FILLAT, F., VILLAR, L., CHOCARRO, C., FERNÁNDEZ DE PINEDO, M. C., GONZÁLEZ, E. & IRIGOYEN, M. T., 1989. Estudio sobre el estado y evolución de

cinco pastizales en la Zona Media de Navarra. Gobierno de Navarra, Iruña. FOCARDI, S., POLI, B. M. & TINELLI, A., 1995. The nutritional carrying capacity of four mediterranean habitats for fallow deer (Dama dama). Revue d’Ecologie. La Terre et la Vie, 50: 97–107. GARCÍA–GONZÁLEZ, R. & ALVERA, B., 1986. Relaciones entre la composición mineral de plantas abundantes en pastos supraforestales pirenaicos y su utlización por los rumiantes. In: XXVI Reunión Científica de la SEEP : 249–265. Consejería de Agricultura y Pesca, Principado de Asturias, Oviedo. GARCÍA–GONZÁLEZ, R. & CUARTAS, P., 1992. Food habits of Capra pyrenaica, Cervus elaphus and Dama dama in the Cazorla Sierra (Spain). Mammalia, 56: 195–202. GARIN, I., 1997. Ecología del ciervo (Cervus elaphus) en la Reserva de Caza de la Garcipollera (Huesca). Tesis doctoral, UPV/EHU, Leioa. GARIN, I., AZORÍN, J., ALDEZABAL, A. & GARCÍA– GONZÁLEZ, R., 1996. Implicaciones nutritivas del contenido en taninos de varias especies leñosas. In: Actas XXXVI Reunión Científica de la SEEP: 292–297. Consejería de Agricultura, Ganadería y Desarrollo Rural, Gobierno de La Rioja, Logroño. G ATES , C. C. & H UDSON , R. J., 1981. Weight dynamics of wapiti in boreal forest. Acta Theriologica, 26: 407–418. GEBCZYNSKA, Z., 1980. Food of the roe deer and red deer in the Bialowieza Primeval Forest. Acta Theriologica, 25: 487–500. GOFFIN, R. A. & DE CROMBRUGGHE, S. A., 1976. Régime alimentaire du cerf (Cervus elaphus) et du chevreuil (Capreolus capreolus) et critères de capacité stationnelle de leurs habitats. Mammalia, 40: 355–376. GOGAN, P. J. P. & BARRETT, R. H., 1994. Roosevelt elk dietary quality in northern coastal California. California Fish and Game, 80: 80–83. GROOT–BRUINDERINK, G. W. T. A. & HAZEBROEK, E., 1995. Ingestion and diet composition of red deer (Cervus elaphus L.) in the Netherlands from 1954 till 1992. Mammalia, 59: 187–195. HANLEY, T. A. & MCKENDRICK, D. J., 1983. Seasonal changes in chemical composition and nutritive value of native forages in a Spruce–Hemlock Forest, Southeastern Alaska. United States Department of Agriculture, Anchorage (USA). HANLEY, T. A., ROBBINS, C. T., HAGERMAN, A. E. & M C A RTHUR , C., 1992. Predicting digestible protein and digestible dry matter in tannin– containing forrages consumed by ruminants. Ecology, 73: 537–541. HEROLDOVÁ, M., 1993. The food of red deer (Cervus elaphus ) in a part of the Krusné Hory Mountains affected by emission. Folia Zoologica, 42: 381–382. HOBBS, N. T., 1987. Fecal indices to dietary quality: a critique. Journal of Wildlife Management, 51: 317–320. HODGMAN, T. P. & BOWYER, R. T., 1986. Fecal crude

protein relative to browsing intensity by whitetailed deer on wintering areas in Maine. Acta Theriologica, 31: 347–353. HOFMANN, R. R., 1989. Evolutionary steps of ecophysiological adaptation and diversification of ruminants: comparative view of their digestive system. Oecologia, 78: 443–457. HOLECHEK, J. L. & VALDEZ, R., 1985. Magnification and shrub stemmy material influences on fecal analysis accuracy. Journal of Range Management, 38: 350–352. HOLECHEK, J. L., VAVRA, M. & PIEPER, R. D., 1982. Botanical composition determination of range herbivore diet: a review. Journal of Range Management, 35: 309–315. H OMOLKA , M., 1993. The Food Niches of 3 Ungulate Species in a Woodland Complex. Folia Zoologica, 42: 193–203. – 1995. The diet of Cervus elaphus and Capreolus capreolus in deforested areas of the Moravskolezske Beskydy Mountains. Folia Zoologica, 44: 227–236. HOWERY, L. D. & PFISTER, J. A., 1990. Dietary and fecal concentrations of nitrogen and phosphorous in penned white–tailed deer does. Journal of Wildlife Management, 54: 383–389. IRWIN, L. L., COOK, J. G., MCWHIRTER, D. E., SMITH, S. C. & ARNETT, E. B., 1993. Assessing winter dietary quality in Bighorn sheep via fecal nitrogen. Journal of Wildlife Management, 57: 413–421. JACKSON, J. E., 1974. Feeding habits of deer. Mammalian Review, 4: 93–101. JENKS, J. A., SOPER, R. B., LOCHMILLER, R. L. & LESLIE, D. M., 1990. Effect of exposure on nitrogen and fiber characteristics of white–tailed deer feces. Journal of Wildlife Management, 54: 389–391. JENSEN, P. V., 1968. Food selection of danish red deer (Cervus elaphus L.) as determined by examination of the rumen content. Danish Review of Game Biology, 5: 1–44. JIANG, Z. & HUDSON, R. J., 1996. Digestive responses of wapiti Cervus elaphus canadensis to seasonal forages. Acta Theriologica, 41: 415–423. JODRA, P. J., 1986. Censo de la población de ciervo (Cervus elaphus L.) y estudio de su alimentación en los terrenos de la Reserva Nacional de Caza de los Montes Universales (Teruel). Diputación General de Aragón, Zaragoza. KAY, R. N. B. & STAINES, B. W., 1981. The nutrition of the red deer (Cervus elaphus). Nutrition Abstracts and Reviews., 51: 601–622. LENG, R. A., 1990. Factors affecting the utilization of “poor–quality” forages by ruminants particularly under tropical conditions. Nutrition Research Reviews, 3: 277–303. LENTLE, R. G., HENDERSON, I. M. & STAFFORD, K. J., 1996. A multivariate analysis of rumen papillary size in red deer (Cervus elaphus). Canadian Journal of Zoology, 74: 2.089–2.094. LESLIE, D. M. & STARKEY, E. E., 1985. Fecal indices to dietary quality of cervids in old–growth forests. Journal of Wildlife Management, 49: 142–146.

Garin et al.

MARCO, J., GORTÁZAR, C., FERRERES, J. & MARTÍN, M., 1996. Parámetros poblacionales y distribución de la población de ciervo (Cervus elaphus) en la Reserva de Caza de La Garcipollera. Diputación General de Aragón, Zaragoza. MARTÍNEZ, T., 1996. Estrategia alimentaria del ciervo (Cervus elaphus) en la Sierra de Cazorla. In: Actas XXXVI Reunión Científica de la SEEP: 319–322. Consejería de Agricultura, Ganadería y Desarrollo Rural, Gobierno de La Rioja, Logroño. MASSEI, G., RANDI, E. & GENOV, P., 1994. The dynamics of the horn growth in Bulgarian chamois Rupicapra rupicapra balcanica. Acta Theriologica, 39: 195–199. MÁTRAI, K. & KABAI, P., 1989. Winter plant selection by red and roe deer in a forest habitat in Hungary. Acta Theriologica, 34: 227–234. MATTSON, W. J. J., 1980. Herbivory in relation to plant nitrogen content. Ann. Rev. Ecol. Syst., 11: 119–161. MERRILL, E. H., CALLAHAN–OLSON, A., RAEDEKE, K. J., TABER, R. D. & ANDERSON, R. J., 1995. Elk (Cervus elaphus roosevelti) dietary composition and quality in the Mount St. Helene blast zone. Northwest Science, 69: 9–18. MITCHELL, B., STAINES, B. W. & WELCH, D., 1977. Ecology of red deer: a research review relevant to their management in Scotland. Institute of Terrestrial Ecology, NERC, Banchory (Scotland). MOULD, E. D. & ROBBINS, C. T., 1981. Nitrogen metabolism in elk. Journal of Wildlife Management, 45: 323–334. PALACIOS, F., MARTÍNEZ, T. & GARZÓN–HEYDT, P., 1989. Data on the autumn diet of the red deer (Cervus elaphus L. 1758) in the Montes de Toledo (Central Spain). Doñana, Acta Vertebrata, 16: 157–163. PICARD, J. F. & GEGOUT, J. C., 1992. L’alimentation du cerf (Cervus elaphus) en hiver dans une forêt feuillue de plaine. Gibier Faune Sauvage, 9: 127–136. PUTMAN, R. J., 1984. Facts from faeces. Mammal Review, 14: 79–97. – 1988. The natural history of deer. Christopher Helm, London (UK). PUTMAN, R. J. & HEMMINGS, G. J., 1986. Can dietary quality of free–ranging ungulates be simply determined from faecal chemistry. Acta Theriologica, 31: 257–270. ROBBINS, C. T., 1993. Wildlife feeding and nutrition. Academic Press, New York (USA). ROBBINS, C. T., HANLEY, T. A., HAGERMAN, A. E., HJELJORD, O., BAKER, D. L., SCHWARTZ, C. C. & M AUTZ , W. W., 1987. Role of tannins in defending plants against ruminants: reduction in protein availability. Ecology, 68: 98–107. RODRÍGUEZ–BERROCAL, J., 1978. Introducción al estudio y valoración de recursos forestales y arbustivos para el ciervo en el área ecológica de Sierra Morena: I. Estudio de la dieta del ciervo. Archivos de Zootecnia, 27: 73–82. SEBER, G. A. F. & PEMBERTON, J. R., 1979. The line intercept method for studying plant cuticles

Animal Biodiversity and Conservation 24.1 (2001)

from rumen and fecal samples. J. Wildlife Management, 43: 916–925. SHERLOCK, M. G. & FAIRLEY, J. S., 1993. Seasonal changes in the diet of red deer Cervus elaphus in the Connemara National Park. In: Biology and Environment: 85–90. Irish Royal Academy, Galway (Ireland). SORIGUER, R. C., FANDÓS, P., BERNÁLDEZ, E. & DELIBES, J. R., 1994. El ciervo en Andalucía. Junta de Andalucía, Sevilla. SPARKS, D. R. & MALECHEK, J. C., 1968. Estimating percentage dry weight in diets using a microscopic technique. J. Range Management, 21: 264–265. TEILLAUD, P., BON, R., GONZÁLEZ, G., SCHAAL, A.,

BALLON, P. & CAMPAN, R., 1991. The cerf. Revue d’Écologie, Supplement 6: 185–217. VAN SOEST, P. J., 1994. Nutritional ecology of the ruminant. Cornell Univerity Press, New York. VENERO, J. L., 1984. Dieta de los grandes fitófagos silvestres del Parque Nacional de Doñana– España. Doñana, Acta Vertebrata, 11: 93–112. V OGEL, K. P., G ABRIELSEN, B. C., W ARD , J. K., ANDERSON, B. E., MAYLAND, H. F. & MASTERS, R. A., 1993. Forage quality, mineral constituents, and performance of beef yearlings grazing two crested wheatgrasses. Agronomy Journal, 85: 584–590. W HITEHEAD, G. K., 1972. Deer of the world. Constable & Co, London (UK).

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Animal Biodiversity and Conservation 24.1 (2001)

Calling songs of certain orthopteran species (Insecta, Orthoptera) in southern Portugal P. A. P. Oliveira, P. C. Simões & J. A. Quartau

Oliveira, P. A. P., Simões, P. C. & Quartau, J. A., 2001. Calling songs of certain orthopteran species (Insecta, Orthoptera) in southern Portugal. Animal Biodiversity and Conservation, 24.1: 65–79. Abstract Calling songs of certain orthopteran species (Insecta, Orthoptera) in southern Portugal.— The calling songs produced by males of the Orthoptera occurring at the field station “Herdade da Ribeira Abaixo” (Centre for Environmental Biology), in Grândola (southern Portugal), are described. The songs were recorded in the field with a portable professional DAT recorder and were analysed in the form of oscillograms and sonagrams. Except for the interesting Gryllotalpa vineae Bennet–Clark, these are the 12 first descriptions of the acoustic parameters and behaviour of the Portuguese populations of the 13 species occurring at the field station and which belong to the following genera: Conocephalus Thunberg, Tettigonia Linnaeus, Platycleis Fieber, Thyreonotus Serville and Uromenus Bolívar (Tettigoniidae), Gryllus Linnaeus, Nemobius Serville and Oecanthus Serville (Gryllidae), Gryllotalpa Latreille (Gryllotalpidae), and Omocestus Bolívar and Euchorthippus Tarbinskii (Acrididae). All species, including pairs and closely related groups, can be readily separated by temporal and frequency parameters of the calling songs through oscillogram and sonagram analyses. Platycleis sabulosa Azam is a new record for Portugal. Key words:: Orthoptera, Calling songs, Oscillograms, Sonagrams, New record, Portugal. Resumen Cantos de llamada en algunas especies de ortópteros (Insecta, Orthoptera) del sur de Portugal.— Se describen los cantos de llamada producidos por machos de ortópteros en el centro de observación “Herdade da Ribeira Abaixo” (Centro de Biología Ambiental), de Grândola (sur de Portugal). Los cantos fueron registrados mediante una grabadora portátil profesional DAT, analizándose en forma de oscilogramas y sonogramas. A excepción del interesante Gryllotalpa vineae Bennet–Clark, se dan las 12 primeras descripciones de los parámetros acústicos y de comportamiento de las poblaciones portuguesas de las 13 especies presentes en el centro de observación y que pertenecen a los siguientes géneros: Conocephalus Thunberg, Tettigonia Linnaeus, Platycleis Fieber, Thyreonotus Serville y Uromenus Bolívar (Tettigoniidae), Gryllus Linnaeus, Nemobius Serville y Oecanthus Serville (Gryllidae), Gryllotalpa Latreille (Gryllotalpidae), Omocestus Bolívar y Euchorthippus Tarbinskii (Acrididae). Todas las especies, incluidas parejas y grupos de especies muy próximas, pueden ser fácilmente identificadas a través del análisis de sus oscilogramas y sonogramas. El registro de Platycleis sabulosa es nuevo en Portugal. Palabras clave: Orthoptera, Cantos de llamada, Oscilogramas, Sonogramas, Nuevo registro, Portugal. (Received: 12 VI 01; Final acceptance: 17 IX 01) P. A. P. Oliveira, C. Simões & J. A. Quartau(1), Centro de Biologia Ambiental, Depto. de Zoologia e Antropologia, Fac. de Ciências, C2, Campo Grande, 1700 Lisboa, Portugal. (1)

e–mail: jquartau@fc.ul.pt

ISSN: 1578–665X

Oliveira et al.

Introduction Airborne vibrations are used for communication purposes by many insects, such as the majority of Orthoptera which use these signals as the most widespread method of intraspecific communication (ALEXANDER, 1968; OTTE, 1977; ELSNER, 1983; EWING, 1984; RAGGE & REYNOLDS, 1998). Acoustic signals in Orthoptera are produced through stridulation, a process whereby vibration or sound results from the friction of one body structure against another. This may be achieved by two main methods: a tegminal mechanism, in which the sound is mainly produced during the closing stroke of the tegmina (suborder Ensifera such as Tettigoniidae, Gryllidae and Gryllotalpidae); or a femoro–tegminal mechanism, where sounds may be produced on both the up and down strokes of both hind femora against the tegmina (suborder Caelifera such as Acrididae) (e. g., RAGGE & REYNOLDS, 1998; CHAPMAN, 2000). In general, each species can produce different song patterns depending upon behavioural context (DUMORTIER, 1963; HASKELL, 1974). In the present study only the most commonly produced sound, the “calling song”, is described. Typically this acoustic signal is produced by males in order to attract conspecific females. In the Gryllidae and most Tettigoniidae the females are silent and if receptive they are induced to perform phonotaxis towards the singing conspecific male, so that individuals are brought together for mating. In other cases, both sexes may walk towards each other until visual contact is established (RAGGE & REYNOLDS, 1998). Thus, the calling song functions as a premating isolation mechanism and its structure is an important component of the specific mate recognition system (PERDECK, 1957; PATERSON, 1985). Therefore, the analysis of the calling songs may provide important taxonomic information at a specific level, namely for deciding on the status of allopatric populations showing small morphological differences or in cases of sibling species, i.e. species which have diverged without showing clear–cut external morphological differences (CLARIDGE, 1985; RAGGE & REYNOLDS, 1998). Calling songs are species specific and since they provide the basis for a mate recognition system they are a particularly reliable indication of the species limits. RAGGE & REYNOLDS´s (1998) comprehensive account of the songs of European orthopterans, for example, puts emphasis on the taxonomy indicated by the song, for which the authors proposed the term phonotaxonomy. Despite some previous knowledge on the acoustic behaviour of some Portuguese insects such as leafhoppers and cicadas (e.g., QUARTAU et al., 1992; QUARTAU & REBELO, 1994; QUARTAU, 1995; QUARTAU & BOULARD, 1995; QUARTAU et al., 1999c), the Orthoptera have been a neglected group in this respect. The present study gives the first descriptions of the calling songs of Portuguese

populations of certain species of Orthoptera occurring in southern Portugal, Platycleis sabulosa Azam being a new record for Portugal. It represents part of a larger project covering the faunistics of the Homoptera and Orthoptera (Auchenorrhyncha) occurring at the field station of the Centre for Environmental Biology, in the area of Grândola (Alentejo) (QUARTAU et al., 1999a, 1999b).

Material and methods Sound recordings were made at the field station of the Centre for Environmental Biology “Herdade da Ribeira Abaixo” near Grândola, in Alentejo (southern Portugal) from June to October of 1997 and during March of 1998. The acoustic recordings were made in the field using digital techniques in the sonic range between 50 Hz and 18 kHz with a SONY DAT recorder TCD-D10 Pro II (tape speed: 0.85 cm/s; unidirectional microphone SONY C76). Some of the recordings were also made through a Report Monitor UHER 4200 (tape speed: mostly 19 cm/s, but also 9.5 cm/s; dynamic microphone AKG D202). The ambient air temperature at the time of the recording was always taken and is referred to in the song descriptions. Recordings and collected specimens are kept in the Department of Zoology and Anthropology with one of the authors (J. A. Quartau). Sound recordings were analysed at the Department of Zoology and Anthropology, in Lisbon, using the PC software Cool EditTM V96 and Avisoft–SAS Lab Light 97. The songs were visualized as oscillograms with 1 min., 10 s and 1 s and, where necessary, other time expansions. Moreover, for a more thorough description of the calling songs, sonagrams were also produced. The terminology as well as its interpretation in connection with the leg– or wing–movements of the singer (i.e., the functioning of the stridulatory apparatus) follow RAGGE & REYNOLDS (1998): (i) calling song, the song produced by an isolated male; (ii) syllable, the sound produced by the opening stroke followed by the closing stroke of the tegmina (Ensifera) or the upstroke followed by the downstroke of the hind femur (Gomphocerinae grasshoppers); (iii) some Tettigoniidae produce two contrasting kinds of syllable, the longer ones are termed macrosyllables and the shorter ones, which usually last less than 10 ms, are the microsyllables; (iv) in Gomphocerinae grasshoppers there are momentary breaks in the sound during the louder part of the syllable, of at least 1.25 ms duration, which are called gaps; (v) diplosyllable, a syllable in which sound is generated by both directional movements of the stridulatory apparatus; (vi) hemisyllable, the sound produced by one unidirectional movement of the elytra or hind femora; (vii) echeme, a first–order assemblage of syllables; (viii) echeme–sequence, a first–order

Animal Biodiversity and Conservation 24.1 (2001)

assemblage of echemes; and (ix) carrier wave, the fundamental wave of a resonant song, i.e., a song with an almost pure dominant frequency. Whenever available, the duration of echemes, syllables and gaps were based on at least ten measurements involving one or two males.

Results Fieldwork provided a total of 21 acoustic recordings of the male calling song from 13 distinct species: six bush–crickets, four crickets, one mole–cricket and two grasshoppers. Specimens were collected on low grass, on top of shrubs and on trees, and on the riparian vegetation of the field station.

Song descriptions

Conocephalus discolor Thunberg, 1815 (Tettigoniidae, Conocephalinae) This species was always found near the stream of the field station, on Scirpus holoshoenus L., and called during the day–time and at sunset on the warm days in August. The song sounded like a quiet or faint sizzling, audible at a distance of 4 to 5 m. The calling song, recorded at 28ºC, consisted of trisyllabic echemes (figs. 1A–1F). The syllables consisted of short opening hemisyllables, followed by longer closing hemisyllables (fig. 1F); the first diplosyllable lasted about 16 ms, was followed by a gap of about 2 ms, and then the second one lasted about 15 ms. The third diplosyllable lasted about 25 ms and followed the

A 10 s B

1 s

0.1 s

0.01 s

ohm G

chm

ohm

chm

ohm

chm

0.01 s

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

Fig. 1. Oscillograms and sonagram of the calling song of a male of Conocephalus discolor at 28ºC: A. 1 min.; B. 10 s; C. 1 s; D. 0.5 s; E. 170 ms; F. One echeme; s1. First syllable; s2. Second syllable; s3. Third syllable; ohm. Opening hemisyllable; chm. Closing hemisyllable; G. Sonagram, showing the audible frequencies of the calling song ranging from about 8 to 19 kHz. Fig. 1. Oscilogramas y sonograma del canto de llamada del macho de Conocephalus discolor a 28ºC: A. 1 min.; B. 10 s; C. 1 s; D. 0,5 s; E. 170 ms; F. Un "echeme"; s1. Primera sílaba; s2. Segunda sílaba; s3. Tercera sílaba; ohm. Hemisílaba inicial; chm. Hemisílaba final; G. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 8 y 19 kHz.

Oliveira et al.

10 s B

1 s

0.1 s D

0.1 s

0.01 s F

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

Fig. 2. Oscillograms and sonagram of the calling song of a male of Tettigonia viridissima at 20ºC: A. 1 min.; B. 10 s; C. 1 s; D. 0.5 s; E. 121 ms; F. Sonagram, showing the audible frequencies of the calling song ranging from about 6 to 13 kHz. Fig. 2. Oscilogramas y sonograma del canto de llamada del macho de Tettigonia viridissima a 20ºC: A. 1 min.; B. 10 s; C. 1 s; D. 0,5 s; E. 121 ms; F. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 6 y 13 kHz.

second after a gap of about 6 ms. The first opening hemisyllable was very prominent, contrasting with the third one which can be quite faint (figs. 1A–1F). The audible frequencies of the calling song ranged from about 8 to 19 kHz (fig. 1G).

Tettigonia viridissima (Linnaeus, 1758) (Tettigoniidae, Tettigoniinae) This loud calling song was heard mainly in the end of June and the beginning of July. It was produced by males in the late afternoon and at night on top of shrubs and trees. The calling song, recorded at 20ºC, consisted of echemesequences interrupted at irregular intervals by pauses shorter than one second (figs. 2A–2B). The echemes were disyllabic (with two closing hemisyllables), lasting 80–90 ms, with gaps of 20–40 ms between them and were produced at the repetition rate of about 8/s (figs. 2C–2D). The first closing hemisyllable, with a duration of

about 30 ms, was slightly shorter than the second, which lasted about 40 ms (fig. 2E).The dominant audible frequencies of the calling song ranged from about 6 to 13 kHz (fig. 2F).

Platycleis sabulosa Azam, 1901 (Tettigoniidae, Decticinae) This species was recorded in the afternoon and after dark, during August on low grasses and small shrubs. The calling song, recording at 25ºC, consisted of long sequences of echemes repeated regularly at the rate of about two per second (figs. 3A–3B). Each echeme lasted 246– 318 ms and consisted of five to six syllables repeated at the rate of about 18/s (figs. 3C–3D). The syllables present on the oscillogram are closing hemisyllables, lasting 40–50 ms each, with the opening hemisyllables absent (RAGGE & REYNOLDS, 1998).. The closing hemisyllables usually increased slightly in amplitude throughout the

Animal Biodiversity and Conservation 24.1 (2001)

10 s

1 s

0.5 s

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 3. Oscillograms and sonagram of the calling song of a male of Platycleis sabulosa at 25ºC: A. 1 min.; B. 10 s; C. 2.3 s; D. 1 s; E. Sonagram, showing the audible frequencies of the calling song ranging from about 6 to 19 kHz. Fig. 3. Oscilogramas y sonograma del canto de llamada del macho de Platycleis sabulosa a 25ºC: A. 1 min.; B. 10 s; C. 2,3 s; D. 1 s; E. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 6 y 19 kHz.

echeme (figs. 3A–3D), appearing slightly longer than those given by RAGGE & REYNOLDS (1998). The audible frequencies of the calling song ranged from about 6 to 19 kHz (fig. 3E).This is a new record for Portugal.

Platycleis affinis Fieber, 1853 (Tettigoniidae, Decticinae) This species was found on low grass, in the evening and early at night in August. The calling song, recorded at 28ºC, consisted of a typical mixture of short echemes (figs. 4C–4F), lasting less than one second and usually composed of three to four macrosyllables, and longer echemes (figs. 4C–4D) lasting from four to eight seconds and composed of 33–53 macrosyllables; these latter, therefore, slightly longer than those given by RAGGE & REYNOLDS (1998). Each echeme usually ends with a series of 7–10 microsyllables. The echemes are often grouped into two to five short ones followed by a long one (fig. 4A–4B). Macrosyllables are repeated at the rate of

about eight per second and microsyllables at about 21/s. Similarly to P. sabulosa, the audible frequencies of the calling song ranged from about 7 to 19 kHz (fig. 4G).

Platycleis intermedia (Serville, 1838) (Tettigoniidae, Decticinae) This bush–cricket called on low grasses and on small shrubs mostly in the late evening and at night in August and September. The calling song, recorded at 28ºC, consisted of long sequences of disyllabic echemes, consisting of two closing hemisyllables, and repeated at the rate of about three per second (figs. 5A–5B); the duration of each echeme varied between 150 and 160 ms and the gap between consecutive echemes varied between 160 and 180 ms. The second closing hemisyllable of each pair was slightly longer and louder than the first (figs. 5C–5D). As with the two previous species, the audible frequencies of the calling song ranged from about 6 to 19 kHz (fig. 5E).

Oliveira et al.

10 s

1 s

0.1 s F 0.1 s G

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2 .

1.4

1.6

Fig. 4. Oscillograms and sonagram of the calling song of a male of Platycleis affinis at 28ºC: A. 1 min.; B. 30 s; C. 10 s; D. 3.9 s; E. 1 s; F. 0.7 s; G. Sonagram, showing the audible frequencies of the calling song ranging from about 7 to 19 kHz. Fig. 4. Oscilogramas y sonograma del canto de llamada del macho de Platycleis affinis a 28ºC: A. 1 min.; B. 30 s; C. 10 s; D. 3,9 s; E. 1 s; F. 0,7 s; G. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 7 y 19 kHz.

Thyreonotus bidens Bolívar, 1887 (Tettigoniidae, Decticinae) This species is camouflaged by a colouration closely matching the trunk and branches of the Quercus spp. trees where it was found. The song was produced late afternoon and at night in September and October. The calling song, recorded at 28ºC, consisted of 1–28 single syllables repeated at the rate of about two per second (figs. 6A–6F). Each syllable varied in duration between 195 ms and 320 ms and consisted of a very small opening hemisyllable and a much longer closing hemisyllable, most of the sound being produced in the closing hemisyllable

(figs. 6B–6E). The gaps between syllables ranged from 278 ms to 822 ms. The audible frequencies of the calling song ranged from about 7 to 19 kHz (fig. 6F).

Gryllus campestris Linnaeus, 1758 (Gryllidae, Gryllinae) This cricket can be heard at the entrance to a burrow at any time of the day or night mostly from May to July. The calling song, recorded at 20ºC, consisted of long and loud echeme– sequences, produced at the rate of about three per second (figs. 7B–7C). Each echeme varied in duration between 104 ms and 113 ms and was

Animal Biodiversity and Conservation 24.1 (2001)

10 s

1 s

0.5 s

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 5. Oscillograms and sonagram of the calling song of a male of Platycleis intermedia at 28ºC: A. 1 min.; B. 10 s; C. 2.3 s; D. 1 s; E. Sonagram, showing the audible frequencies of the calling song ranging from about 6 to 19 kHz. Fig. 5. Oscilogramas y sonograma del canto de llamada del macho de Platycleis intermedia a 28ºC: A. 1 min.; B. 10 s; C. 2,3 s; D. 1 s; E. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 6 y 19 kHz.

composed of four closing hemisyllables at the repetition rate, within an echeme, of about 24/s. The gaps between echemes ranged from 176 ms to 745 ms. The first hemisyllable was in general quieter than the remaining three and in general a crescendo continued throughout the echeme (figs. 7C–7D) as referred to by RAGGE & REYNOLDS (1998). The frequency of the carrier wave was about 4 kHz (fig. 7E).

Gryllus bimaculatus De Geer, 1773 (Gryllidae, Gryllinae) Like the previous species, this cricket called from the entrance to a burrow during the afternoon and mainly during the night from August until October. The calling song, recorded at 27ºC, was very similar to that of G. campestris (fig. 8): it consisted of long sequences of tetrasyllabic echemes of closing hemisyllables, repeated within each echeme at the rate of

about 24/s. The echeme repetition rate was similar to that in the previous cricket with about three per second but, in contrast with G. campestris, all the hemisyllables were similar in duration and amplitude (figs. 8C–8D). Each echeme lasted from 113 ms to 132 ms, appearing therefore slightly longer than in G. campestris. The gaps between echemes varied from 176 ms to 235 ms. The frequency of the carrier wave was about 4.5 kHz, therefore slightly higher than in G. campestris (fig. 8E).

Nemobius sylvestris (Bosc, 1792) (Gryllidae, Nemobiinae) This small cricket called any time of the day or night during September and October, and was found in moist places in the ground under leaves, e.g. the bed and shores of dried temporary brooks of the field station. The presumable calling song, recorded at 29ºC, consisted of sequences of long

Oliveira et al.

10 s

1 s

0.5 s

0.1 s

E 0.1 s F

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 6. Oscillograms and sonagram of the calling song of a male of Thyreonotus bidens at 28ºC: A. 1 min.; B. 10 s; C. 2.3 s; D. 1 s; E. 0.5 s; F. Sonagram, showing the audible frequencies of the calling song ranging from about 7 to 19 kHz. Fig. 6. Oscilogramas y sonograma del canto de llamada del macho de Thyreonotus bidens a 28ºC: A. 1 min.; B. 10 s; C. 2,3 s; D. 1 s; E. 0,5 s; F. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 7 y 19 kHz.

echemes repeated at the rate of about 0.5/s (figs. 9A–9B). The duration of each echeme depended on the number of syllables, which were between 12 and 62 (fig. 9C), therefore considerably longer than those described by RAGGE & REYNOLDS (1998). It is possible that this atypical sound might represent instead a courtship sound. The syllables were repeated at the rate of about 32/s, lasting about 9 ms each and with a gap of 15 to 20 ms between them. The first 8–12 syllables showed an increase in amplitude (figs. 9C–9D). The frequency of the carrier wave was about 3.8 kHz (fig. 9E).

Oecanthus pellucens (Scopoli, 1763) (Gryllidae, Oecanthinae) This slender and yellowish cricket called on top of low grass or on small bushes in July and August.

The calling song, recorded at 22ºC, consisted of long echemesequences (figs. 10A–10B). The duration of each echeme varied between 800 and 1300 ms, at a repetition rate of about 0.6/s, and the gap between consecutive echemes varied from 400 to 500 ms. Each echeme was composed of 18– 29 syllables at the repetition rate of about 22/s (figs. 10C–10D). Each syllable lasted about 32 ms and the gap between successive syllables was about 10 ms. The frequency of the carrier wave was about 2.3 kHz (fig. 10E).

Gryllotalpa vineae Bennet–Clark, 1970 (Gryllotalpidae, Gryllotalpinae) Like most mole–crickets, this species produces its song from a specially made singing burrow, with a pair of short horn–shaped passages, leading to two entrance holes. The song is very

Animal Biodiversity and Conservation 24.1 (2001)

10 s

1 s

C 0.5 s D

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 7. Oscillograms and sonagram of the calling song of a male of Gryllus campestris at 20ºC: A. 1 min.; B. 10 s; C. 1.6 s; D. 1 s; E. Sonagram, showing the frequency of the carrier wave at about 4 kHz. Fig. 7. Oscilogramas y sonograma del canto de llamada del macho de Gryllus campestris a 20ºC: A. 1 min.; B. 10 s; C. 1,6 s; D. 1 s; E. Sonograma que muestra la frecuencia de la onda portadora a 4 kHz.

loud and shrill, audible from a distance of even 600 m. The calling song, recorded at 17ºC, was composed of syllables lasting 10–20 ms each and repeated at the rate of about 54/s (fig. 11). The gap between the syllables was 7–9 ms. The dominant frequency was about 3 kHz (fig. 11E). This mole–cricket is very close to Gryllotalpa gryllotalpa (Linnaeus), from which it is easily separated by the frequency of the carrier wave, which is below 2 kHz in G. gryllotalpa.

Omocestus raymondi (Yersin, 1863) (Acrididae, Gomphocerinae) This grasshopper was recorded in September, in the low grasses and called during daylight of warm days. The calling song (fig. 12), recorded at 30º C, consisted of isolated echemes, repeated at irregular intervals, normally from 8 to 10 s, and lasting 1.7–1.8 s each. Each echeme was composed of about 24 syllables (downstroke hemisyllables) repeated at the rate of about 15/ s. Each echeme begins quietly, rapidly increasing in amplitude (fig. 12C). Each downstroke

hemisyllable, lasting 60–70 ms, had a characteristic pattern of 3–4 gaps, which became obscured towards the end of the echeme (figs. 12D–12E). The range of audible frequencies of the calling song varied at the start of the echeme from about 6-10 kHz to 4–19kHz at a later phase (fig. 12F).

Euchorthippus pulvinatus gallicus Maran, 1957 (Acrididae, Gomphocerinae) The calling song of this grasshopper was recorded on low grass and during the daytime in June. It was recorded at 27ºC and consisted of sequences of echemes regularly repeated at the rate of about two per second and lasting 106–130 ms each (figs. 13A–13C). Each echeme was composed of about seven syllables, which slowly increased in amplitude towards the end; the syllables were repeated at the rate of about 45/s, the last two or three syllables having two gaps each (figs. 13D–13F). The audible frequencies of the calling song ranged from about 4 to 13 kHz (fig. 13G).

Oliveira et al.

10 s

1 s

C 0.5 s

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 8. Oscillograms and sonagram of the calling song of a male of Gryllus bimaculatus at 27ยบC: A. 1 min; B. 10 s; C. 1.6 s; D. 1 s; E. Sonagram, showing the frequency of the carrier wave at about 4.5 kHz. Fig. 8. Oscilogramas y sonograma del canto de llamada del macho de Gryllus bimaculatus a 27ยบC: A. 1 min; B. 10 s; C. 1,6 s; D. 1 s; E. Sonograma que muestra la frecuencia de la onda portadora a 4,5 kHz.

A 10 s B 1 s C 0.5 s

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 9. Oscillograms and sonagram of the calling song of a male of Nemobius sylvestris at 29ยบC: A. 1 min.; B. 10 s; C. 2.4 s; D. 1 s; E. Sonagram, showing the frequency of the carrier wave at about 3.8 kHz. Fig. 9. Oscilogramas y sonograma del canto de llamada del macho de Nemobius sylvestris a 29ยบC: A. 1 min.; B. 10 s; C. 2,4 s; D. 1 s; E. Sonograma que muestra la frecuencia de la onda portadora a 3,8 kHz.

Animal Biodiversity and Conservation 24.1 (2001)

A B

10 s

1 s 0.1 s

0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

Fig. 10. Oscillograms and sonagram of the calling song of a male of Oecanthus pellucens at 22ยบC: A. 1 min.; B. 10 s; C. 1.3 s; D. 1 s; E. Sonagram, showing the frequency of the carrier wave at about 2.3 kHz. Fig. 10. Oscilogramas y sonograma del canto de llamada del macho de Oecanthus pellucens a 22ยบC: A. 1 min.; B. 10 s; C. 1.3 s; D. 1 s; E. Sonograma que muestra la frecuencia de la onda portadora a 2,3 kHz.

10 s

1 s C 0.1 s D 0.1 s E

kHz 20 15 10 5 0.2

0.4

0.6

0.5

1.2

1.4

1.6

1.8

Fig. 11. Oscillograms and sonagram of the calling song of a male of Gryllotalpa vineae at 17ยบC: A. 1 min.; B. 10 s; C. 1 s; D. 0.4 s; E. Sonagram, showing the frequency of the carrier wave at about 3 kHz. Fig. 11. Oscilogramas y sonograma del canto de llamada del macho de Gryllotalpa vineae a 17ยบC: A. 1 min.; B. 10 s; C. 1 s; D. 0,4 s; E. Sonograma que muestra la frecuencia de la onda portadora a3 kHz.

Oliveira et al.

A 10 s

1 s

0.5 s

uhm F

0.1 s

dhm

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

Fig. 12. Oscillograms and sonagram of the calling song of a male of Omocestus raymondi at 30ºC: A. 1 min.; B. 10 s; C. 3.5 s; D. 1 s; E. 0.4 s; s6. Sixth syllable; uhm. Upstroke hemisyllable; dhm. Downstroke hemisyllable; F. Sonagram, showing the audible frequencies of the calling song varying from about 6–10 kHz at the beginning of the echeme to 4–19 kHz at a later phase. Fig. 12. Oscilogramas y sonograma del canto de llamada del macho de Omocestus raymondi a 30ºC: A. 1 min.; B. 10 s; C. 3,5 s; D. 1 s; E. 0,4 s; s6. Sexta sílaba; uhm. Hemisílaba ascendente; dhm. Hemisílaba descendente; F. Sonograma que muestra las frecuencias audibles del canto de llamada que varían entre 6 y 10 kHz al principio del "echeme" y entre 4 y 19 kHz en la fase final.

Discussion There are a few previous descriptions of the songs of Portuguese Orthoptera in Portuguese males e.g., in RAGGE & REYNOLDS (1998). However, the songs here presented are the first descriptions of the acoustic parameters and behaviour of the Portuguese populations of 12 species of the 13 found at the field station “Herdade da Ribeira Abaixo”, nearby Grândola (southern Portugal). Thesse descriptions are of special value at the species level, since calling songs are species specific and are important components of the specific mate recognition system in orthopterans (PERDECK, 1957; PATERSON, 1985; REYNOLDS, 1988; RAGGE & REYNOLDS, 1998). Therefore, these songs can be a quick, easy and very practical way of species identification in the field, without the need for collecting, killing and mounting

specimens. Calling songs are also of particular value for determining the geographic boundaries of the species ranges (GREEN, 1995). Three closely related species of the genus Platycleis were found at the field station: P. sabulosa, P. affinis and P. intermedia, the first being a new record for Portugal. The temporal patterning of their calling songs differs greatly among species, offering good characters for their taxonomic separation, as RAGGE (1990) has also emphasized: for instance, the echemes vary from disyllabic (P. intermedia), penta– or hexasyllabic (P. sabulosa) to comprising several scores of syllables (P. affinis). Differences in the temporal patterning also occur in the remaining Tettigoniidae and in the Acrididae here studied. Oscillograms are thus very convenient for portraying the calling songs and for discriminating most of these species.

Animal Biodiversity and Conservation 24.1 (2001)

10 s

1 s

0.5 s

D E

0.1 s

F 0.1 s G

kHz 20 15 10 5 0.2

0.4

0.6

0.8

1.2

1.4

1.6

Fig. 13. Oscillograms and sonagram of the calling song of a male of Euchorthippus pulvinatus gallicus at 27ºC: A. 1 min.; B. 10 s; C. 3.5 s; D. 1 s; E. 0.2 s, an echeme with 6 gaps; F. 0.2 s, an echeme with 4 gaps; g. Sonagram, showing the audible frequencies of the calling song ranging from about 4 to 13 kHz; ¤Gap. Fig. 13. Oscilogramas y sonograma del canto de llamada del macho de Euchorthippus pulvinatus gallicus a 27ºC: A. 1 min.; B. 10 s; C. 3,5 s; D. 1 s; E. 0,2 s, un "echeme" con seis intervalos; F. 0,2 s, un "echeme" con 4 intervalos; g. Sonograma que muestra las frecuencias audibles del canto de llamada que oscila entre 4 y 13 kHz; ¤Intervalo.

In the genus Gryllotalpa, the situation is, however, different. Two species occur in western Europe: G. gryllotalpa (Linnaeus) and the species here found G. vineae Bennet–Clark. These two mole-crickets are, in fact, close morphologically and it is probable that the previous citations of G. gryllotalpa to Portugal (e.g., AIRES & MENANO, 1916; SEABRA, 1939, 1942) should refer instead to G. vineae. Temporal parameters of the calling song are rather variable in both species, being temperature dependent. However, the dominant frequency carrier wave appears to be much more constant within each species, being below 2 kHz in G. gryllotalpa and at least 3 kHz in G. vineae (RAGGE & REYNOLDS, 1998). As such, this is a nice example of frequency seemingly being of great taxonomic importance for discrimination of sibling species. Sonagrams are thus quite useful here for specific identification and delimitation.

In the pair of closely related species of the genus Gryllus the temporal patterning of the calling songs is very similar. However, in G. campestris the four hemisyllables tended to increase in a crescendo throughout the echeme, in contrast with G. bimaculatus, where all four were of the same amplitude. G. campestris also showed a dominant frequency slightly lower than in G. bimaculatus. In the two remaining gryllids, Nemobius sylvestris and Oecanthus pellucens, which are quite distinct morphologically, there was also some similarity in the temporal patterning of the songs. They are, however, readily separated by temporal parameters such as the structure and duration of the echemes. Moreover, the carrier frequency is also discriminatory, since it is about 3.8 kHz in N. sylvestris and 2.3 kHz in O. pellucens In short, this study suggests that oscillograms were in general very useful for portraying the

calling songs of the species here studied. On the other hand, except for Gryllidae and Gryllotalpidae, sonagrams from audio recordings may give a misleading impression of the frequency spectrum of the song, which in Acrididae and Tettigoniidae extends well into the ultrasonic range, and in some Tettigonniidae is mainly ultrasonic. For Gryllidae and Gryllotalpidae, the carrier frequency can be measured more accurately from fast oscillograms than from sonagrams (D. R. Ragge, pers. comm.). Nevertheless, sonagrams are a useful addition to song analysis, especially if they include the full frequency spectrum of the song. Therefore, the simultaneous use of both oscillograms and sonagrams is encouraged. Finally, as previously mentioned, one species found at the field station —Platycleis sabulosa Azam—, is here recorded for the first time from Portugal. Moreover, considering the small geographic area investigated many more new records are yet to be found in this country and hence further studies dealing with these insects should be encouraged.

Acknowledgements For their help in the field work, as well as for other support the authors would like to express their sincere thanks to Dr. Arabolaza (Escola Superior Agrária, Bragança), Dr. Bívar de Sousa (Sociedade Portuguesa de Entomologia, Lisboa), and Mr. Genage André (Departamento de Zoologia e Antropologia, Faculdade de Ciências, Universidade de Lisboa). Finally, the authors greatly appreciate the comments of Dr. D. R. Ragge, who kindly accepted to critically reading the manuscript.

References AIRES, B. & MENANO, H. P., 1916. Catálogo sinóptico dos ortópteros de Portugal. Univ. de Coimbra, Coimbra. ALEXANDER, R. D., 1968. Arthropods. In: Animal Communication: Techniques of Study and Results of Research: 167–216 (T. A. Sebeok, Ed.). Indiana University Press, Bloomington. CHAPMAN, R. F., 2000. The Insects. Structure and Function, 4th edition. Cambridge University Press, Cambridge. CLARIDGE, M. F., 1985. Acoustic signals in the Homoptera: behaviour, taxonomy and evolution. Annual Review of Entomology, 30: 297–317. DUMORTIER, B., 1963. Ethological and physiological study of sound emission in Arthropods. In: Acoustic Behaviour of Animals: 583–684 (R. G. Busnel, Ed.). Elsevier Pub. Co., Amsterdam. E LSNER , N., 1983. Insect stridulation and its neurophysiological basis. In: Bioacoustics, a comparative approach: 69–92 (B. Lewis, Ed.). Academic Press, London.

Oliveira et al.

EWING, A. W., 1984. Acoustic signals in insect sexual behaviour. In: Insect communication: 223–240 (T. Lewis, Ed.). Academic Press, London. GREEN, S. V., 1995. Song characteristics of certain Namibian grasshoppers (Orthoptera: Acrididae: Gomphocerinae). African Entomology, 3(1): 1–6. HASKELL, P. T., 1974. Sound Production. In: The physiology of Insecta., vol. 2: 354–410 (M. Rockstein, Ed.). Academic Press, New York and London. OTTE, D., 1977. Communication in Orthoptera. In: How Animals Communicate, Part. I: 334– 361 (T. A. Sebeok, Ed.). Indiana University Press, Bloomington and London. PATERSON, H. E. H., 1985. The recognition concepts of species. In: Species and Speciation: 21–29 (E. S. Vrba, Ed.). Transval Museum, Pretoria. PERDECK, A. C., 1957. The isolating value of specific songs patterns in two sibling species of grasshoppers (Chorthippus brunneus Thunb. and C. biguttulus L.). Behaviour, 12: 1–75. QUARTAU, J. A., 1995. Cigarras esses insectos quase desconhecidos. Correio da Natureza, 19: 33–38. QUARTAU, J. A. & BOULARD, M., 1995. Tettigetta mariae n. sp., nouvelle Cigale lusitanienne (Homoptera, Cicadoidea, Tibicinidae). EPHE, Biologie et Évolution des Insectes, 7/8: 105–110. QUARTAU, J. A, CLARIDGE, M., MORGAN, J., & REBELO, T., 1992. Os sinais acústicos de Jacobiasca lybica (Homoptera; Cicadellidae). Boletim da Sociedade Portuguesa de Entomologia, supl. no. 3, 1: 247–252. QUARTAU, J. A., PICCIOCHI DE OLIVEIRA, P. A., REBELO, M. T. & S IMÕES , P. C., 1999a. Ortópteros (Insectos). In: Caracterização da Flora e da Fauna do Montado da Herdade da Ribeira Abaixo (Grândola–Baixo Alentejo): 61–68 (M. Santos–Reis & A. I. Correia, Eds.). Centro de Biologia Ambiental, Lisboa. QUARTAU, J. A. & REBELO, M. T., 1994. Sinais acústicos em Cicadidae e Cicadellidae (Homoptera, Auchenorrhyncha) que ocorrem em Portugal. In: Actas do I Congresso Nacional de Etologia: 137–142 (V. Almada & R. Oliveira, Eds.). Instituto Superior de Psicologia Aplicada, Lisboa. QUARTAU, J. A., REBELO, M. T. & SIMÕES, P., 1999b. Cicadídeos (Insectos, Homópteros). In: Caracterização da Flora e da Fauna do Montado da Herdade da Ribeira Abaixo (Grândola–Baixo Alentejo): 69–74 (M. Santos– Reis & A. I. Correia, Eds.). Centro de Biologia Ambiental, Lisboa. QUARTAU, J. A., REBELO, M. T., SIMÕES, P. C., FERNANDES, T. M., CLARIDGE, M. F., DROSOPOULOS, S. & MORGAN, J. C., 1999c. Acoustic signals of populations of Cicada orni L. in Portugal and Greece (Hemiptera: Auchenorrhyncha: Cicadomorpha: Cicadidae). Reichenbachia, Staatliches Museum fuer Tierkunde Dresden, 33(9): 71–80. RAGGE, D. R., 1990. The songs of the western European bush–crickets of the genus Platycleis in relation to their taxonomy (Orthoptera:

Animal Biodiversity and Conservation 24.1 (2001)

Tettigoniidae). Bulletin of the British Museum of Natural History (Entomology), 59(1): 1–35. RAGGE, D. R. & REYNOLDS, W. J., 1998. The songs of the grasshoppers and crickets of Western Europe. Harley Books, London. REYNOLDS, W. J., 1988. The use of insect sounds in taxonomy. British Journal of Entomology and Natural History, 1: 147–152. SEABRA, A. F., 1939. Contribuição para a história

da Entomologia em Portugal. Catálogo das colecções entomológicas do Laboratório de Biologia Florestal em 1937. Direção Geral dos Serviços Florestais e Aquícolas, 6(2): 155–304. SEABRA, A. F., 1942. Contribuições para o inventário da fauna lusitânica: Insecta, Orthoptera (Saltatoria, Phasmida, Dermaptera, Blattaria e Mantodea). Memórias e Estudos do Museu de Zoologia da Universidade de Coimbra, 127: 1–13.

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Animal Biodiversity and Conservation 24.1 (2001)

Corncrake Crex crex census estimates: a conservation application of vocal individuality T. M. Peake & P. K. McGregor

Peake, T. M. & McGregor, P. K., 2001. Corncrake Crex crex census estimates: a conservation application of vocal individuality. Animal Biodiversity and Conservation, 24.1: 81–90. Abstract Corncrake Crex crex census estimates: a conservation application of vocal individuality.—Vocal individuality could be used to estimate numbers of individuals in species otherwise difficult to monitor. However, the usefulness of this technique in providing conservation information is little studied. The vocalisations of the Corncrake show a high level of individual distinctiveness. This fact was used to examine current counting methods and estimate movement patterns within one breeding season. Information on individual identity gained from vocalisations increased census estimates by 20–30% and showed that male Corncrakes called less frequently than previous studies had suggested. Males moved greater distances in areas with lower availability of suitable habitat. The conservation implications of these results are discussed. Key words: Vocal individuality, Census accuracy, Corncrake. Resumen Estimación del censo de guiones de codornices Crex crex: una aplicación de la individualidad vocal a la conservación.— La individualidad vocal puede utilizarse para estimar el número de individuos en especies que de otra manera sería difícil controlar. Sin embargo, la utilidad de esta técnica para obtener información acerca de la conservación ha sido poco estudiada. La vocalización del guión de codornices muestra un alto grado de diferenciación individual, lo que se utilizó para examinar métodos de recuento y estimar patrones de movimiento durante la época de reproducción. La información basada en la identificación individual a partir de las vocalizaciones incrementó la estimación del censo en un 20–30% y mostró que el macho del guión de codornices efectuaba cantos de llamada con menor frecuencia de lo que otros estudios previos sugerían. Los machos recorren grandes distancias en áreas que presentan una escasa disponibilidad de hábitats adecuados. Se discuten las implicaciones de estos resultados en cuanto a la conservación. Palabras clave: Individualidad vocal, Exactitud en los censos, Guión de codornices. (Received: 16 V 01; Final acceptance: 22 V 01) T. M. Peake & P. K. McGregor, Behaviour and Ecology Research Group, Dept. of Life Science, Univ. of Nottingham, NG7 2RD, UK.– Current address, Dept. of Animal Behaviour, Univ. of Copenhagen, Tagensvej 16, DK 2200 Copenhagen N, Denmark.

ISSN: 1578–665X

Introduction A wide variety of techniques is used to count and monitor the fates of animal populations (SOUTHWOOD, 1978; KREBS, 1989; BIBBY et al., 1992; POLLARD & YATES, 1993; SUTHERLAND, 1996), most of which do not require the identification of individuals. Techniques that involve the ability to identify individual animals can provide ecological information that alternative techniques cannot. Such information generally falls into three categories (MCGREGOR & PEAKE, 1998): a. The assessment of census error; b. The estimation of population parameters including age structure, survival and migration rates; c. The detection of individual behavioural differences. Census errors are due either to random sampling error, which leads to an imprecise estimate, or to systematic bias, which leads to an inaccurate estimate (BIBBY et al., 1992). Census precision can be increased by taking more samples. Census accuracy is more difficult to determine as the extent of bias is frequently unknown even if the sources and directions of bias are understood. Often it may be reasonable to assume that census estimates from different areas or times are subject to the same sources of bias and are therefore comparable. However, the extent of bias in census estimates for many endangered species is particularly likely to differ between years and areas due to the effects of habitat change, large fluctuations in density due to stochastic variation in small populations or human exploitation. Differences in bias between areas may be of particular importance if habitat management decisions are made based upon census estimates. For example, GIBBS & WENNY (1993) found that unpaired males of two bird species were three to five times more likely to be detected than paired males; thus, one area could appear to contain many fewer individuals than another while actually providing a better breeding habitat. Hypothetically, management decisions based upon measures of habitat in an apparently high–density site could result in a reduction in the amount of available breeding habitat. Measurement of bias can only be achieved if the actual number of animals within an area is known; this requires intensive study and frequently involves the use of individually identifiable animals (BIBBY et al., 1992). The majority of identification techniques involve capture and the addition of external marks (STONEHOUSE, 1978; MCGREGOR & PEAKE, 1998); both capture and marking are capable of producing biased data. Appreciation of the potential biases and welfare implications of capture has led to an increasing interest in the use of naturally occurring marks to identify animals. For example, photo–identification techniques are routinely used in cetacean censuses (DUFAULT & WHITEHEAD, 1995).

Peake & McGregor

While the majority of techniques based upon natural variation utilise variation in visual features, many studies have found a degree of individuality in bird vocalisations, to the extent that the potential for use as a monitoring tool is established (e.g. SAUNDERS & WOOLLER, 1988; DAHLQUIST et al., 1990; GILBERT et al., 1994). Some studies have utilised this level of individuality; e.g. GALEOTTI (1994) used individual characteristics of Tawny Owl (Strix aluco) hoots (GALEOTTI & P AVAN , 1991) to determine owl territories. However, few studies show that such techniques can provide information useful to conservation. A recent survey of bird survey techniques lists only one census that routinely uses this technique, that of the European Bittern (Botaurus stellaris) in the UK by the Royal Society for the Protection of Birds (RSPB; GILBERT et al, 1998). The Corncrake (Crex crex) is an endangered land–rail (COLLAR & ANDREW, 1988) that presents monitoring problems due to its tendency to occupy areas of tall, dense vegetation. The census technique currently employed for Corncrakes in Britain and Ireland is based upon findings from studies of radio–tagged individuals (STOWE & HUDSON, 1988, 1991). Results from radio–tracking suggest that, although they forage over a larger area, males rarely move more than 250 m between calling sites. STOWE & HUDSON (1988) visited males carrying radio–tags on a number of nights and found that males called on 75–80% of nights. Based on these findings, the current census technique involves mapping calling males within an area on two nights, separated by one to two weeks, between 20th May and 10th July (HUDSON et al., 1990; GREEN, 1995). These maps are then combined on a single map. A problem arises when a male calls from one site (A) on the first visit, and on the second visit a male calls from a site nearby (B) with no calling heard from site A; is this one male or two? Currently this problem is solved using the rule: if site A is < 250 m from B then count one male, otherwise count two males (henceforth "the 250 m rule"). Violations of the 250 m rule provide an obvious source of bias in census estimates as males moving more than 250 m are consistently over–counted if encountered on both census nights. Another potential source of bias is between–male variation in the incidence of calling, males calling on fewer than 75% of nights being consistently undercounted owing to a lower likelihood of encounter. Recently, a high level of individuality in the "crake" vocalisation of male Corncrakes has been demonstrated (MAY, 1994; PEAKE et al., 1998). The aim of this study is to examine how this level of individuality might provide useful conservation information only otherwise available through extensive and/or expensive ringing and radio–tracking studies. Here two types of information are considered: 1. Estimation of the accuracy of the current counting technique;

Animal Biodiversity and Conservation 24.1 (2001)

2. Assessment of the ability of vocal individuality to provide information on within–season movements in a small study area and how these movements relate to vegetation characteristics.

Methods Fieldwork took place at Balranald RSPB reserve, North Uist, Scotland, an area of approximately 400 hectares. The entire study area was visited on 80 nights from 5 V 95 to 27 VII 95 between 23.00 h and 03.00 h. Positions of calling males were mapped after close approach (< 20 m) on each visit and recordings made whenever weather conditions allowed. Recordings were collected using Sennheiser MKH–816T microphones, AKB– 11 pre–amplifiers and Uher 4000 Report Monitors. In comparing census methods, only nights between 20 V 95 and 10 VII 95 were included in accordance with the standard census method outlined previously. Between these dates the area was visited on 48 nights; recording was possible on 33 of 48 nights. In order to determine individual identity, five calls with a high signal–to–noise ratio and no obvious calling from neighbouring males were chosen from each recording on each night. On nights on which recording was not possible, it was assumed that calling sites were occupied by the same individual as had been recorded there on the most recent night possible. Following the methods of PEAKE et al. (1998), measurements of call structure were taken from waveform representations of recorded calls. For each pair of recordings, Pearson correlation coefficients (r) were then calculated as a measure of similarity. PEAKE et al. (1998) found that calls recorded from the same individual had r values > 0.7, while over 80% of calls from different individuals had r values < 0.7. On occasions where two sets of calls had r > 0.7 it was assumed that the two sets were recorded from the same individual; in this way, over–estimation of the number of males present was avoided. Three counting methods were examined, in each case comparing results derived from mapping data alone with those derived from mapping data and individual identification based on recordings. The current (two–night) census technique was assessed by randomly selecting pairs of nights that were separated by a minimum seven and a maximum 14 days; each night only appeared in one pair. For each pair of nights, calling sites were plotted on a single map and the 250 m rule used together with information on males calling simultaneously to decide how many males were present. This process was then repeated adding information on identity obtained from recordings. The value of information gained by using three nights of data rather than two was then assessed. From 48 nights, 16 sets of three nights

were randomly chosen such that no two nights were separated by fewer than seven or more than 28 nights. As with the two night census assessment, data from each triplet was plotted on the same map and the total number of Corncrakes calculated: a. Using the 250 m rule, b. Adding information on individual identity. Finally, the information gained by using four nights of data was assessed. Twelve sets of four nights were chosen randomly such that no two nights were within seven days of one another; the maximum separation between nights was 39 days. Again, results obtained with and without knowledge of individual identity were compared. The total number of males present on the study site over the whole season was estimated from the combined map of calling sites over 80 nights. On each night several calling sites were used simultaneously and hence those sites represented different birds. As the season progressed some sites were no longer used and new calling sites were used. Based on the timings of site abandonment and new site occupation and the distances between sites, the pattern of movements of individuals and hence the total number of males present was estimated. This exercise was then repeated with the addition of information on individual identity. Here, distance between calling sites was ignored when estimating movement patterns. Again the threshold criteria for acceptance of a correct match between recordings was r > 0.7, which could conceivably lead to the spurious matching of different individuals. However, the chances of this occurring were small due to the large number of recordings collected from individuals on different nights, the number of possibly incorrect matches that could be ruled out due to simultaneous calling and thus the small number of comparisons that needed to be made relative to the number that could possibly be made. Vegetation characteristics were measured for the entire study area on three occasions during the season (15 May, 15 June and 15 July). In each case vegetation was categorised according to GREEN & STOWE (1993) and the extent of each category plotted on a large scale map (1 cm = 20 m). Based on vegetation suitability indices calculated by GREEN & STOWE (1993), vegetation types were split into two simple categories: those having a positive effect on Corncrake presence (usable habitat: Iris pseudacorus patches, areas of the grasses Phragmites australis and/or Phalaris arundinacea and hay meadows) and those having a negative effect on the presence of Corncrakes (unusable habitat: short dry pasture, wet pasture and unsuitable areas such as roads, open water and buildings). Areas of nettles (Urtica dioica) and umbellifers were also considered "usable habitat" as these are heavily used by Corncrakes (TYLER, 1996). For each calling site, habitat suitability was estimated as the percentage of usable habitat

Peake & McGregor

Results Over the season, calling male Corncrakes were mapped at 40 sites within the study area. Sites were occupied by a calling male from 1 to 28 nights (mean = 6.7, s.e. = 1.06). The number of calling males present on a single night varied from 0 to 14 (mean = 6.1, s.e. = 0.55). From a total of 48 nights, 24 pairs of nights were randomly generated. The mean (±s.e.) census figure derived using mapping data alone was 11.2±0.56 calling males. When information on individual identity was included, this figure increased by 28.6% to 14.4±0.44. During fieldwork it was noted that considerably fewer Corncrakes were heard on nights when weather conditions were poor (usually when the wind was particularly strong). As an arbitrary measure of adverse weather conditions, the season was separated into nights on which recording was possible and nights on which weather conditions made recording impossible (strong winds, heavy rain). Unpredictability of weather conditions at the study site meant that observer effort was approximately equal on all nights. In order to minimise the chance of failing to detect males due to poor weather, all sites where males had previously called were visited to within 20 m at least once on each night. Significantly more Corncrakes called on nights when weather conditions allowed recording (mean = 8.1, s.e. = 0.45) than on nights when weather conditions were poor (mean = 2.1, s.e. = 0.48, H = 27.7, d.f. = 1, p < 0.001). Thus, the two–night census analysis was repeated, excluding nights on which recordings could not be made. The mean (±s.e.) census figure derived from mapping data alone was 12.8±0.61; the addition of information on individual identity increased this figure by 28.1% to 16.4±0.82. Estimates derived using three nights of data found an average 12.9±0.67 (s.e.) males without the use of individual identification and 16.9±0.84

40 m 40 m

surrounding the site. Percentage habitat was estimated by overlaying the vegetation map with a grid of 21 squares each 40 m by 40 m arranged in the form of a five by five grid with the corner squares removed (fig. 1). Giving a total area of 33,600 m2 this grid approximates a circle of radius 100 m. Although it is likely that males travelled outside this area in order to forage (STOWE & HUDSON, 1988), TYLER & GREEN (1996) found that the average distance between a male’s calling site and subsequent nesting attempts was 101m, thus the area chosen should be of approximately the size relevant to both males and females when choosing sites in which to make a breeding attempt. The grid was placed such that the centre of the grid corresponded with the male’s calling site and aligned so that vertical grid lines ran North/South.

Fig. 1. Diagram showing the grid used to measure habitat characteristics of Corncrake calling sites. The central dot indicates the calling site of a male Corncrake. Fig. 1. Diagrama que muestra la rejilla utilizada para medir las características del hábitat de los lugares de llamada del guión de codornices. El punto central indica el lugar de llamada del macho del guión.

males when individual identification was included, an increase of 31.0%. Excluding poor weather nights resulted in figures of 14.6±0.80 and 19.9±0.45 respectively, an increase of 36.3%. With four nights of data, estimates obtained using only mapping data averaged (±s.e.) 14.2±0.76 males, those including individual information 18.2±0.84 males, an increase of 28.2%. Excluding poor weather nights gave figures of 15.7±0.55 and 21.6 ± 0.91 respectively, an increase of 37.6%. Thus, increasing the number of census nights and including information from recordings both increase census estimates (fig. 2). Based on mapping data and using the 250 m rule, the total number of male Corncrakes in the study site was estimated as 24. When information on individual identity was included, this figure increased by 20.8% to 29. This number is unlikely to represent the number of males present on the study site at any one time as birds likely moved to and from the study site throughout the season. It is conservative to assume that all males arrive from migration and begin calling before 31 May. Nine (32%) males recorded on the study site were not recorded before this date, suggesting that they moved onto the study site during the season from surrounding areas. It is also conservative to assume that no males leave on migration before 1 July. Seven (24%) males were not recorded on the study site after this date, suggesting that they

Animal Biodiversity and Conservation 24.1 (2001)

Discussion The current census method for Corncrakes appears to underestimate the true number of birds present, at least within the confines of this study. The addition of information on the identity of individuals increased census estimates by nearly 30% in all three situations tested. However, even estimates derived using vocal individuality fell short of the mean number of males present during the study period. The main source of bias resulting in undercounting would appear to be a lower than expected incidence of calling. STOWE & HUDSON (1988) found that radio–tagged males called on 75-80% of visits, a result backed up by later radio–tracking studies carried out on Coll, Inner Hebrides (TYLER & GREEN, 1996). Thus, on a two– night census, the probability of encountering any given male is between 0.94 and 0.96. In this study, males called on an average of 41.5% of nights, giving a probability of encounter of 0.66. With, on average, 20.7 males present and a probability of encounter of 0.66 over two nights, the expected census estimate would be 13.6 males, close to the observed two night census figure using mapping data alone of 11.8 males. This would suggest that

24 20 Census estimate

had either moved from the study area to another site or died. Assuming that males remained in the study area on the nights between that on which they were first recorded and that on which they were last recorded, it is possible to calculate the number of males present on each night; the mean number of males present on each night was 20.7±0.5 (s.e.). Figure 3 shows the patterns of movements between the 40 calling sites derived from: mapping data alone and mapping data in conjunction with information on individual identity. There are a number of differences between the two maps with a number of birds moving more than 250 m between sites. Distances moved by Corncrakes between calling sites are greater in some areas of the study site than in others. This seems to coincide with the amount of suitable habitat shown in figure 4. There was a significant negative correlation between the distance moved by birds following site abandonment and the percentage of usable habitat in the area surrounding the abandoned site (rs = – 0.58, n = 15, p < 0.05). Using information on individual identity, we then investigated the radio-tracking result that males call on 75–80% of nights (STOWE & HUDSON, 1988). As we could only be certain that a male was present if he called, we assumed that if a male had previously called, became silent and subsequently re–appeared (either at the same site or at a new site) he had been present in the study area for the silent period; on average males called on 41.5% of nights (fig. 5).

16 12 8 4 0 1

2 4 3 Number of census nights

Fig. 2. Results of various methods used to count male Corncrakes. As the number of census nights increases the census estimate derived increases. Census estimates which included information on individual identity (closed circles) are greater than those which use only mapping data (open circles). Fig. 2. Resultados de distintos métodos utilizados para contar machos de guión de codornices. Si el censo nocturno se incrementa, también se incrementa la estima del censo derivado. Los censos estimados que incluyen información de la identidad individual (círculos negros) son mayores que los que sólo utilizan datos de mapeo (círculos blancos).

variation in the proportion of males in an area that call on a given night represents a significant source of bias in the census method. The second radio–tracking result contributing to the current census method (STOWE & HUDSON, 1988) was that males rarely move more than 250 m between calling sites. Within our study area, four of 29 (13.7%) males moved distances greater than this (4–500 m). In addition, nine (32%) arrived at the study site late in the season, the assumption being that they must have arrived from migration elsewhere and moved onto the study site later. While birds that moved to and from the site may introduce bias to a census of the North Uist population, they do not represent a source of bias in the estimates made within the study site as they will not have been recorded at two sites. It is unlikely therefore that movements greater than 250 m are a major source of bias in the census estimates derived in this study.

Peake & McGregor

A 200m

B 200m

Fig. 3. The pattern of movements of Corncrake surmised from: (A) mapping data alone and (B) mapping data in conjunction with knowledge of individual identity obtained from recordings of calls. Mapping alone arrives at a total census figure of 24 whilst the inclusion of information on the identity of individuals increases this estimate to 29. One site (indicated by an asterisk on the lower map) was used at different times by two males. Fig. 3. Patrón de movimientos del guión de codornices a partir de: (A) datos de mapeo únicamente y (B) datos de mapeo conjuntamente con el conocimiento de la identidad individual obtenida a partir de grabaciones de llamadas. Mediante el mapeo sólo se obtiene un censo total de 24, mientras que incluyendo información de la identidad de los individuos se incrementa a 29. Una localización (indicada con un asterisco en el mapa inferior) fue usada en diferentes ocasiones por dos machos.

Animal Biodiversity and Conservation 24.1 (2001)

< 5% 5–14% 15–24% >25%

Fig. 4. Map showing the percentage cover of habitat suitable for Corncrake in 100x100 m2 squares throughout the study area. Closed circles represent calling sites of male corncrake; lines connecting sites indicate movements between sites. Fig. 4. Mapa que muestra el porcentaje de cobertura de hábitat disponible para el guión de codornices en cuadrículas de 100x100 m2 del área de estudio. Los círculos negros representan los sitios de llamada de machos de guión; las líneas que los conectan indican movimientos entre estos sitios.

However, the effects of the movement of individuals on census estimates may be greater in some areas than others. It appears from this study that where suitable habitat is scarce or patchily distributed, male Corncrakes move greater distances throughout the season than in areas where habitat is more homogenous. Where areas differ greatly in habitat availability or continuity, census figures may be difficult to compare directly. Perhaps more importantly, where habitat within a given site changes over time, whether for better (positive habitat management) or for worse (habitat destruction), movement rates, and hence census inaccuracy, are likely to change. This could result in trends become obscured or accentuated depending on the direction of habitat change. Having an idea of the levels of inaccuracy within and between given areas and/or years would enable this problem to be counteracted to some extent. There are several reasons why the results of this study might differ from those achieved by radio-tracking. Radio–tracking studies carried out by STOWE & HUDSON (1988, 1991) took place in two areas, three km and 50 km south of our study site. It is possible that geographical variation in habitat between these areas and our study area may have resulted in behavioural differences between

the males in each study. This is perhaps less likely than the fact that these previous studies were carried out between eight and ten years before this study; during that time, changes in agricultural practices may have had a greater effect on habitat than subtle geographic differences. It is also possible that radio–tracking itself affected the behaviour of male Corncrakes studied previously. There are two main sources of potential bias involved in radio–tracking studies. Firstly, the method of capturing male Corncrakes may not capture a random sample of the population. Males are usually captured using playback of calls to stimulate approach. It is possible that this procedure biases capture rates towards a certain behavioural subset of male Corncrakes, e.g. those that are more strongly territorial and thus more vocally active. Second, radio-tags themselves may have caused changes in the behaviour of males. Male Corncrakes appear only to call nocturnally while attempting to attract a mate, becoming silent at night once successful (TYLER & GREEN, 1996). Radio–tags may reduce mating opportunities, resulting in an increase in the amount of time spent calling. However, TYLER (1996) found no difference in the attraction of radio–tagged females to tagged and untagged males.

Peake & McGregor

Number of {

6 5 4 3 2 1 0

20 30 40 50 60 70 80 90 100 Nights calling as % of nights present

Number of periods without calling

90 80 70 60 50 40 30 20 10 0

2 3 4 5 6 7 8 9 10 >10 Number of consecutive nights without calling

Fig. 5. Likelihood of calling by male Corncrakes: A. Number of nights on which males were heard to call as a percentage of nights on which they were assumed present; B. Frequency distribution of the number of consecutive nights on which males were not heard to call yet were assumed present at the study site between 5 V 95 and 27 VII 95. Fig. 5. Probabilidad de llamada del macho de guión de codornices: A. Número de noches en que se escucharon los cantos de llamada de los machos con relación al número de noches en que se supuso que estaban presentes; B. Distribución de frecuencias del número de noches consecutivas en las que no se escucharon los cantos de llamada de los machos aunque se supuso que estaban presentes entre el 5 V 95 y 27 VII 95.

Another explanation for the difference may be that some birds left the study site for short periods and later returned. Inspection of figure 5B shows that some males do indeed stop calling for periods of a week or more during the study period. While a number of these undoubtedly represent breeding attempts, some may represent temporary movements to other sites. However, in the majority of instances males become silent for only one or two nights at a time, increasing the likelihood that they remained on the study site over these periods. If the low incidence of calling found in this study is due to males leaving the

site for short periods, then movement rates of males between sites must be considerable. Although census workers counted three males calling within 1 km of the study site, the nearest area with large numbers of calling males was approximately 3 km away (R. E. Green, pers. comm.). The level of increased census accuracy achieved by recording individual birds may not be enough to warrant issuing recording equipment to census workers. Indeed, the predictions of movement patterns based on mapping data alone often corresponded with those ascertained by the analysis of recordings (fig. 4). However, our results

Animal Biodiversity and Conservation 24.1 (2001)

suggest that the sources and extents of bias in census estimates may differ between areas, potentially making comparison difficult. The relative accuracy of a census in a given area at a given time can be assessed relatively quickly and easily using the vocal individuality technique without the need to capture or otherwise disturb individual males. Individually distinct vocalisations can have a very direct application in the detection of breeding attempts. Results of recent studies of radio-tagged individuals suggest that males cease nocturnal calling when accompanied by a female (TYLER & GREEN, 1996) and thus that cessation of nocturnal calling gives an indication of attempted breeding. However, this method relies on the assumption that subsequent calling at a given site is produced by the previously resident male and cannot detect males that move large distances following breeding attempts. It is only possible to confirm the identity of a resident male if that male is either ringed (requiring capture on at least two occasions), radio–tagged (expensive and requires capture) or through analysis of recorded calls (requires sophisticated equipment but not capture). Despite a growing number of studies that examine the potential of vocal individuality to provide ecological information, there is little published evidence that such techniques are used or indeed are useful in practice. This study has demonstrated that individual distinctiveness present in the calls of the Corncrake (MAY, 1994; PEAKE et al., 1998) can provide information that is of comparable accuracy to information collected by other techniques. More importantly, individually distinct calls provide an opportunity to assess relative accuracy.

Acknowledgements TMP was supported by a CASE studentship between the BBSRC and the RSPB. Analysis equipment was funded by NERC and the Royal Society. We are grateful to Rhys Green, Ken Smith, Ken Otter and Andrew Terry for comments that improved the manuscript considerably. We are also grateful to Phil Benstead, Cath Jeffs, Alex Turner and Dave Hodson for help and hospitality during fieldwork in North Uist. Fieldwork was carried out under licence SCB:02:95 from Scottish Natural Heritage.

References BIBBY, C. J., BURGESS, N. D. & HILL, D. A., 1992. Bird Census Techniques. Academic Press, London. COLLAR, N. J. & A NDREW, P., 1988. Birds to watch: The ICBP World checklist of threatened birds. ICBP, London. DAHLQUIST, F. C., SCHEMNITZ, S. D. & FLACHS, B. K.,

1990. Distinguishing individual male wild turkeys by analyzing vocalisations using a personal computer. Bioacoustics, 2: 303–316. DUFAULT, S. & WHITEHEAD, H., 1995. An assessment of changes with time in the marking patterns used for photo–identification of individual sperm whales, Physeter macrocephalus. Marine Mammal Science, 11: 335–243. GALEOTTI, P., 1994. Patterns of territory size and defence level in rural and urban tawny owl (Strix aluco) populations. J. Zool. (Lond.), 234: 641–658 G ALEOTTI , P. & P AVAN , G., 1991. Individual recognition of male tawny owls (Strix aluco) using spectrograms of their territorial calls. Ethol. Ecol. Evol., 3: 113–126. GIBBS, J. P. & WENNY, D.G., 1993. Song output as a population estimator: effect of male pairing status. J. Field. Ornith., 64: 316–322. GILBERT, G., GIBBONS, D. W. & EVANS, J., 1998. Bird Monitoring Manual: a compendium of monitoring and survey techniques for birds of conservation concern and others in the UK. RSPB/BTO/WWT/JNCC/ITE/Seabird Group, Sandy, UK. GILBERT, G., MCGREGOR, P. K. & TYLER, G., 1994. Vocal individuality as a census tool, practical considerations illustrated by a study of two rare species. J. Field Ornith., 65: 335–348. GREEN, R. E., 1995. The decline of the Corncrake Crex crex in Britain continues. Bird Study, 42: 66–75. GREEN, R. E. & STOWE, T. J., 1993. The decline of the corncrake Crex crex in Britain and Ireland in relation to habitat change. J. Appl. Ecol., 30: 689–695. HUDSON, A.V., STOWE, T. J. & ASPINALL, S. J., 1990. Status and distribution of corncrakes in Britain in 1988. British Birds, 83: 173–187. KREBS, C. J., 1989. Ecological Methodology. Harper and Row, New York. MAY, L., 1994. Individually distinctive Corncrake Crex crex calls: A pilot study. Bioacoustics, 6: 25–32. MC GREGOR, P. K. & PEAKE , T. M., 1998. The role of individual identification in conservation biology. In: Behavioural Ecology and Conservation Biology: 31–55 (T. M. Caro, Ed.). Oxford University Press, Oxford. PEAKE, T. M., MCGREGOR, P. K., SMITH, K. W., TYLER, G., GILBERT, G. & GREEN, R. E., 1998. Individuality in Corncrake Crex crex vocalisations. Ibis, 140: 120–127. P OLLARD , E. & YATES, T. J., 1993. Monitoring butterflies for ecology and conservation . Chapman Hall, London. SAUNDERS, P. A. & WOOLLER, R. D., 1988. Consistent individuality of voice in birds as a management tool. Emu, 88: 25–32. SOUTHWOOD, T. R. E., 1978. Ecological Methods. Chapman Hall, London. STONEHOUSE, B., 1978. Animal marking of animals in research. Macmillan, London.

STOWE, T. J. & HUDSON, A. V., 1988. Corncrake studies in the western isles. RSPB Cons. Rev., 2: 38–42. – 1991. Radio–telemetry studies of corncrakes in Great Britain. Vogelwelt, 112: 10–16. S UTHERLAND , W. J., 1996. Ecological census techniques . Cambridge University Press, Cambridge.

Peake & McGregor

TYLER, G., 1996. The ecology of the corncrake with special reference to the effect of mowing on breeding production. Ph. D. Thesis, University College Cork. T YLER, G. & G REEN, R. E., 1996. The incidence of nocturnal song by male Corncrakes Crex crex is reduced during pairing. Bird Study, 43: 214–219.

Animal Biodiversity and Conservation 24.1 (2001)

Aligning conservation goals: are patterns of species richness and endemism concordant at regional scales? T. H. Ricketts

Ricketts, T. H., 2001. Aligning conservation goals: are patterns of species richness and endemism concordant at regional scales? Animal Biodiversity and Conservation, 24.1: 91–99. Abstract Aligning conservation goals: are patterns of species richness and endemism concordant at regional scales?— Biodiversity conservation strategies commonly target areas of high species richness and/or high endemism. However, the correlation between richness and endemism at scales relevant to conservation is unclear; these two common goals of conservation plans may therefore be in conflict. Here the spatial concordance between richness and endemism is tested using five taxa in North America: butterflies, birds, mammals, amphibians, and reptiles. This concordance is also tested using overall indices of richness and endemism (incorporating all five taxa). For all taxa except birds, richness and endemism were significantly correlated, with amphibians, reptiles, and the overall indices showing the highest correlations (rs = 0.527–0.676). However, “priority sets” of ecoregions (i.e., the top 10% of ecoregions) based on richness generally overlapped poorly with those based on endemism (< 50% overlap for all but reptiles). These results offer only limited support for the idea that richness and endemism are correlated at broad scales and indicate that land managers will need to balance these dual, and often conflicting, goals of biodiversity conservation. Key words: Conservation goals, Species richness, Endemism, Regional scales. Resumen Ajuste de los objetivos de conservación ¿Son concordantes a escala regional los patrones de riqueza de especies y de endemismos?— Las estrategias de conservación de la biodiversidad se centran habitualmente en áreas con una gran riqueza de especies y/o un alto nivel de endemicidad. Sin embargo, la correlación entre riqueza y endemismo a escala relevante para la conservación es poco clara; por consiguiente, estos dos objetivos comunes de los planes de conservación pueden entrar en conflicto. En este trabajo se estudia la concordancia espacial entre riqueza y endemismo en Norteamérica utilizando cinco taxones: mariposas, aves, mamíferos, anfibios y reptiles. Esta concordancia se estudia también empleando índices globales de riqueza y endemismo (incorporando los cinco taxones). Para todos los taxones, excepto para las aves, riqueza y endemismo aparecen correlacionados significativamente, mostrando para los anfibios y reptiles una alta correlación de todos los índices (rs = 0.527–0.676). Sin embargo, las “actuaciones prioritarias” de las ecoregiones (por ejemplo, el 10% de ecoregiones de vanguardia) basadas en la riqueza de especies normalmente se solapan poco con las basadas en endemismos (< 50% de solapamiento para todos los taxones excepto para los reptiles). Estos resultados apoyan limitadamente la idea de que riqueza y endemismo están correlacionados a gran escala e indica que los gestores del territorio deberán tener en cuenta estos objetivos duales, y a menudo en conflicto entre sí, de conservación de la biodiversidad. Palabras clave: Objetivos de conservación, Riqueza de especies, Endemismo, Escala regional. (Recieved: 4 X 01; Final acceptance: 10 X 01) Taylor H. Ricketts, Center for Conservation Biology, Dept. of Biological Sciences, Gilbert Hall / 371 Serra Mall, Stanford Univ., Stanford, CA 94305–5020 (USA).

ISSN: 1578–665X

Ricketts

Introduction It is well recognized by conservation biologists that there are limited resources available to address intensifying anthropogenic threats to biodiversity (EHRLICH & WILSON, 1991; MYERS et al., 2000). Geographic priorities must therefore be established, so that these resources and effort can be allocated to areas with high biodiversity value, such as high species richness and/or endemism (C EBALLOS et al., 1998; O LSON & DINERSTEIN, 1998). While in theory this is a sound strategy, its implementation has encountered two major difficulties. First, a lack of high–quality species distribution data, especially at broad scales, has made it difficult to identify priority areas with confidence (WILLIAMS & GASTON, 1994). Second, there is frequently a difference of opinion among conservationists over which aspects of biodiversity are most important in setting priorities. Some authors have emphasized species richness, while others argue that areas of high endemism should be targeted most (PRENDERGAST et al., 1993; KERR, 1997; CEBALLOS et al., 1998). A popular response to this first problem has been to propose indicator taxa: well–studied groups of organisms whose richness patterns can be used as surrogates for other taxa or for overall species richness. Many recent studies have either proposed indicator taxa (e.g., PEARSON & CASSOLA, 1992), assumed them to indicate overall richness and based conservation plans on them (e.g., SCOTT et al., 1993), or tested their utility directly (e.g., DAILY & EHRLICH, 1996; CARROLL & PEARSON, 1998; RICKETTS et al., 1999a; RICKETTS et al., in press). To date, tests of indicator taxa for species richness have produced mixed results, suggesting the utility of this conservation tool depends on context, taxon, and scale (WEAVER, 1995). Even if suitable indicator taxa can be found, however, the second problem remains. Priorities set on the basis of species richness may not successfully conserve areas of high endemism, which are clearly important to biodiversity conservation at any scale. Data on endemism are typically less available than on species richness, and patterns of endemism are thus less well understood (BIBBY, 1992; KERR, 1997). Therefore, biologists still have a relatively poor understanding of whether patterns of species richness and endemism are concordant, and thus whether

these two common goals of conservation plans are in conflict or alignment. This second problem is addressed here, using a large North American dataset to examine the concordance of richness and endemism patterns in five animal taxa (butterflies, birds, mammals, reptiles, and amphibians). Two specific questions are asked. First, are levels of species richness and endemism correlated across the United States and Canada? this correlation is tested for each taxon individually as well as for indices of overall richness and endemism that incorporate all five taxa. Second, to what extent do areas selected for conservation priority on the basis of richness overlap with areas selected on the basis of endemism? Answers to these questions will help determine whether the two primary goals of biodiversity conservation plans will tend to reinforce or compete with each other for limited resources.

Methods Species data The species distribution data are based on the 110 ecoregions of the continental United States and Canada (fig. 1). These ecoregions were first developed by Ricketts et al (RICKETTS et al., 1999b), and are based largely on three established ecoregion mapping projects (ESWG 1995; GALLANT et al., 1995; O MERNIK , 1995). Ecoregions are relatively coarse biogeographic divisions of a landscape that delineate areas with broadly similar environmental conditions and natural communities. They are nested within eight major biomes in North America (fig. 1). Because of the complexity with which environmental and ecological factors vary across a landscape, ecoregion boundaries are necessarily approximate and represent areas of transition rather than sharp divisions. RICKETTS et al. (1999b) compiled presence/ absence data for butterflies, birds, mammals, reptiles, and amphibians on these ecoregions. The same dataset was used after performing further checks for quality and accuracy. From presence/absence data, the number of species (hereafter “richness”) and the number of endemic species (hereafter “endemism”) were calculated of each taxon in every ecoregion. Following

Fig. 1. A. Map of the 110 terrestrial ecoregions of the United States and Canada; B. Map showing the eight biomes represented by these ecoregions. Fig. 1. A. Mapa de las 110 ecorregiones terrestres de Estados Unidos y Canadá; B. Mapa que muestra ocho biomas representados por estas ecorregiones.

Animal Biodiversity and Conservation 24.1 (2001)

Temperate broadleaf and mixed forests

Mediterranean scrub and savanna

Temperate coniferous forests

Xeric shrublands/deserts

Temperate grasslands/savannas

Boreal forest/taiga

Flooded grasslands

Tundra

Ricketts

Table 1. Spearman rank correlations between richness and endemism for the five animal taxa considered. Results given for all 110 ecoregions, and for each biome separately: * Significance level at p < 0.05 (missing entries indicate that in the corresponding taxon, no endemic species are found in any ecoregion of the corresponding biome); All. Includes six ecoregions from minor biomes that are not included in any of the biome analyses. Tabla 1. Correlaciones del rango de Spearman entre riqueza y endemismo para los cinco taxones considerados. Los resultados se indican para cada una de las 110 ecorregiones y para cada bioma por separado: * Nivel de significancia para p < 0,05 (los datos que faltan indica que en el correspondiente taxón no se han encontrado especies endémicas en ninguna ecorregión del bioma que le corresponde); All. Incluye seis ecorregiones de biomas pequeños que no están incluidas en ninguno de los biomas analizados.

Biome

Butterflies

Birds

0.304*

0.011

All Temperate broadleaf

–0.153

Temperate coniferous

0.298*

Overall indices

0.527*

0.676*

0.588*

110

0.153

–0.112

0.795*

0.430

0.635*

0.471*

–

0.186

0.632*

0.526*

0.479*

–

0.407

–0.098

0.704*

0.796*

0.504*

0.327

0.412

0.218

–0.127

0.788*

0.753*

Temperate grasslands Xeric shrublands

Mammals Amphibians Reptiles

Boreal forest / Taiga

–

0.584*

0.333

–

0.442

Tundra

–

0.398

–0.212

–

0.047

RICKETTS et al. (1999b), a species to be endemic in an ecoregion was counted if it either (i) was found in no other ecoregion, including Mexico and other continents or (ii) occupied a range totaling less than 50,000 km2 (BIBBY, 1992). Thus species with exceptionally small ranges that crossed an ecoregion boundary were considered endemics in both ecoregions. To examine more general patterns of biodiversity, overall indices of richness and endemism that incorporate information from all five taxa were also calculated. The richness index was defined as 5

1/5

✟ i

Ri / Ti

where Ri is the richness of taxon i in the ecoregion, and Ti is total number of species of taxon in the database (SISK et al., 1994; RICKETTS et al., 1999a). This index normalizes the richness of each taxon by the number of North American species in that taxon and then averages those fractions across all five taxa. It therefore weights taxa evenly, preventing speciose groups from dominating measures of overall richness. The endemism index is defined as 5

1/5

✟E / R i

where Ei is the number of endemic species of taxon i in the ecoregion, and Ri is as above. This index computes, for each taxon, the fraction of

species in an ecoregion that is endemic there, and then averages these fractions across all five taxa. Again, the index thus normalizes counts of endemics by the taxon’s richness in each ecoregion. Analyses Correlation between richness and endemism measures were tested using Spearman rank correlations, because data were seldom normally distributed (ZAR, 1999). Since ecoregions vary widely in area (fig. 1, RICKETTS et al., 1999b) and both richness and endemism are typically expected to increase with area (ROSENZWEIG, 1995), any correlations found may be driven by these area effects. To examine this possibility, the Spearman rank correlation between the richness and endemism measures and ecoregion area was computed. Finally, to examine whether the degree of concordance between richness and endemism differs among biomes, the richness/ endemism correlations were tested for each biome independently (fig. 1, table 1). To determine the overlap between richness– based and endemism–based priority sets of ecoregions, the ecoregions in the top decile were identified (i.e., 90th percentile and above) for each measure. The percent overlap of these sets for each taxon, and for the overall indices were then calculated (PRENDERGAST et al., 1993). The top decile of 110 ecoregions typically contains

Animal Biodiversity and Conservation 24.1 (2001)

B 3

8 7 Endemism

6 2

5 4 3

2 1 0

0 0

50 100 150 200 250 300 D

Endemism

50 100 150 200 250 300

10 20 30

25 20

5 15

4 3

2 5

1 0

0 0

100 120

40 50 60 70

F 18 0.1

16 Endemism

0.08

12 0.06

10 8

0.04

6 4

0.02

2 0

0 0

40 60 Richness

80 100 120

0.1

0.2 0.3 Richness

0.4

Fig. 2. Relationship between richness and endemism across all 110 ecoregions. Each circle represents an ecoregion: A. Butterflies; B. Birds; C. Mammals; D. Amphibians; E. Reptiles; F. Overall richness and endemism indices. Dashed lines in panel f delineate the top decile on each axis; note only two ecoregions lying above both lines (i.e., in the upper right quadrant). These are the only two ecoregions that are members of both richness-based and endemism–based priority sets, and they are coded in green in figure 3F. Fig. 2. Relación entre riqueza y endemicidad en las 110 ecorregiones. Cada círculo representa una ecorregión: A. Mariposas; B. Aves; C. Mamíferos; D. Anfibios; E. Reptiles; F. Índices globales de riqueza y endemicidad. Las líneas discontinuas en la figura F delimitan el decilo superior de cada eje; obsérvese que únicamente dos ecorregiones se encuentran por encima de ambas líneas (por ejemplo en el cuadrante superior derecho). Éstas son las dos únicas ecorregiones que optan a la vez por las actuaciones prioritarias basadas en la riqueza y en el endemismo, están indicadas en negro en la figura 3F.

Ricketts

11 ecoregions. In some cases, however, ties between the 11th–ranked ecoregion and those ranked below it forced inclusion of more than 11 in the priority set. Overlap between richness and endemism in these cases was calculated by dividing the number of shared ecoregions by the number of ecoregions in the smaller of the two priority sets (PRENDERGAST et al., 1993).

Results Across all North American ecoregions, species richness and endemism were in general positively correlated (table 1, top row). For all taxa except birds, richness and endemism were significantly correlated, with amphibians, reptiles, and the overall indices showing the highest correlations. There is a large amount of scatter in bivariate plots for all taxa, however (fig. 2), indicating a low degree of predictive power in these relationships. Richness and endemism for most taxa were not significantly correlated with ecoregion area (table 2). The only three significant relationships found (i.e., involving endemism in butterflies,

Table 2. Spearman rank correlations between ecoregion area and measures of richness and endemism for the five taxa and for the overall indices: * Significance level p < 0.05 (n = 110). Tabla 2. Correlaciones de rango de Spearman entre área de ecorregión y medidas de la riqueza y endemicidad para los cinco taxones y para la totalidad de índices: * Nivel de significación p < 0,05 (n = 110).

Taxon

Richness

Endemism

Butterflies

0.09

-0.21*

Birds

0.07

0.19*

Mammals

0.15

-0.26*

Amphibians Reptiles Overall indices

0.05

-0.08

-0.02

0.00

0.09

-0.16

Table 3. Percent overlap between priority sets of ecoregions based on richness and endemism. Tabla 3. Porcentaje de solapamiento entre prioridades de ecorregiones basado en riqueza y endemicidad.

Amphibians

Reptiles

Overall indices

3 (27%)

5 (45%)

7 (64%)

2 (18%)

Butterflies

Birds

Richness set

Endemism set

18 4 (36%)

Overlap

Mammals

Fig. 3. Maps showing the distribution of, and overlap between, richness and endemism priority sets. Light gray ecoregions are in the top decile for richness, medium gray ecoregions are in the top decile for endemism, and black ecoregions are in the top decile for both: A. Butterflies; B. Birds; C. Mammals; D. Amphibians; E. Reptiles; F. Overall richness and endemism indices. Fig. 3. Mapas que muestran la distribución y la coincidencia de las acciones prioritarias en riqueza y endemicidad: Gris claro, ecorregiones situadas en el decilo superior en cuanto a riqueza; Gris medio, ecorregiones situadas en el decilo superior en cuanto a endemicidad; Negro, ecorregiones situadas en el decilo superior para ambas prioridades: A. Mariposas; B. Aves; C. Mamíferos; D. Anfibios; E. Reptiles; F. Índices globales de riqueza y endemicidad.

Animal Biodiversity and Conservation 24.1 (2001)

birds, and mammals) were weak and inconsistent in their sign (table 2). Therefore, the correlation results in table 1 are unlikely to be caused by the commonly–expected effects of area on richness and endemism. When correlations were tested within each biome independently, the results generally reflected those found using all ecoregions (table 1). Amphibians, reptiles, and the overall indices again tended to show strong correlations in all but the tundra and taiga biomes. Correlations for butterflies, birds and mammals, which showed weak or no correlation using all ecoregions, remained generally non–significant in the biome–by–biome analyses. Overlap between richness-based and endemism– based priority sets were generally low, varying between 27% (birds and mammals) to 64% (reptiles) (table 3). In addition, priority ecoregions for richness and endemism were often found on opposite sides of the continent and often in different biomes (fig. 3).

Discussion These results offer mixed support for the idea that richness and endemism patterns are correlated at broad scales. On one hand, two taxa and the overall indices showed quite strong and consistent correlations across the 110 ecoregions and within each major temperate biome (table 1). On the other hand, three of the five taxa showed much weaker or no correlations, and the scatter in all of these relationships (and thus their unpredictability) was high for all taxa. On the more practical level of choosing areas for conservation investment, the results are even less encouraging. Because of the scatter mentioned above, the statistical correlations found, even when strongly significant, do not translate into high overlap between priority sets based on richness and endemism (table 3, fig. 2). A good example is the relationship between the overall richness and endemism indices (fig. 2F); the statistical correlation between them is quite high (table 1), but their priority sets overlap in only 2 out of a possible 11 ecoregions (table 3). This contradiction is best understood by examining figure 2F; although the two variables are correlated overall, only two ecoregions fall in the top decile for both richness and endemism. Indeed, for all taxa except reptiles this overlap is less than 50% (table 3). Basing conservation strategies on richness, therefore, will seldom effectively conserve areas of high endemism. Previous studies on this topic also show a mixture of results. In North America, KERR (1997) found relatively strong correlations between richness and endemism in four taxa: mammals, a bee genus, a moth subfamily, and a butterfly family. PRENDERGAST (1993), however, reported little concordance between species–rich hotspots and

Ricketts

rare species in Great Britain, using birds, butterflies, dragonflies, liverworts, and aquatic angiosperms. Similarly, CEBALLOS et al. (1998) found “very low correspondence” among areas of high mammalian richness and endemism in Mexico. What accounts for the differences in results among these studies? Among other factors, results may be influenced by the taxa and region considered, the scale of observation (both extent and resolution, LEVIN, 1992; PRENDERGAST et al., 1993), the definition of endemism used, and the choice of geographic units. For example, PRENDERGAST (1993) based their analyses on 10 km grid squares in Great Britain, while KERR (1997) used much larger (2.5º of latitude and longitude) grids over a much larger extent in North America (in addition to testing different taxa). Clearly the four studies (i.e., the three mentioned above and mine) differ among themselves in several of these factors, making it difficult to glean general lessons from the collective results. Perhaps of most interest are the contrasting findings between my study and that of KERR (1997). These two studies were performed in the same region at similar scales, with one taxon in common (mammals). Nevertheless, KERR (1997) found high correlation in mammals (r = 0.807, p < 0.001), while the results presented here show quite a weak relationship (table 1). This difference may be due to differences in the definition of endemism. KERR (1997) calculates the endemism value of a given square by summing, over all species present in the square, the inverses of the number of squares occupied by each species (e.g., 1/24+1/137+1/3…). This measure, however, is not independent of richness; the more species present, the more inverses are added to the sum. In contrast, counting the simple number of true endemics in an area (i.e., species found nowhere else) is not statistically related to richness measures, and thus may better reveal the actual relationship between these two measures of conservation priority. One caveat deserves mention here. Since a typical species range overlaps with several ecoregions (and thus ecoregions do not accrue their richnesses independently), these richness data probably contain a certain degree of spatial autocorrelation (JONGMAN et al., 1995). This problem tends to inflate the degrees of freedom used in significance testing, and therefore the probabilities reported here should be interpreted with caution. However, these results remain useful for comparing strengths of relationships among taxa, because the correlation coefficients themselves are unaffected (only the significance tests). In addition, endemism, by definition, does not suffer this same problem. In conclusion, the results presented here and in other studies (PRENDERGAST et al., 1993; KERR, 1997; C EBALLOS et al., 1998) indicate that conservation biologists may not have the luxury of assuming that management plans based on

Animal Biodiversity and Conservation 24.1 (2001)

“hotspots” of species richness will also capture important centers of endemism. Additional studies undertaken at different scales and with different taxa may yield a better understanding of the factors that determine the degree of concordance between richness and endemism patterns. Until then, however, conservation biologists and land managers will need to continue to balance these dual, and often conflicting, goals of biodiversity conservation.

Acknowledgements I thank the Conservation Science Program at World Wildlife Fund–U.S. for their collaboration in originally compiling these data. K. Bowen, J. Fay, M. Mayfield, and J. Schwan helped to improve, error check, and manage the databases. C. Boggs, K. Carney, J. Hellmann and J. Hughes provided valuable discussions and comments on the manuscript. Finally, the support of U.S. NASA and The Summit Foundation are gratefully acknowledged.

References BIBBY, C. J., 1992. Putting biodiversity on the map: priority areas for global conservation. ICBP, Washington, DC. CARROLL, S. S. & PEARSON, D. L., 1998. Spatial modeling of butterfly species richness using tiger beetles (Cicindelidae) as a bioindicator taxon. Ecological Applications, 8: 531–543. CEBALLOS, G., RODRIGUEZ, P. & MEDELLIN, R. A., 1998. Assessing conservation priorities in megadiverse Mexico: Mammalian diversity, endemicity, and endangerment. Ecological Applications, 8: 8–17. DAILY, G. C. & EHRLICH, P. R., 1996. Nocturnality and species survival. Proceedings of the National Academy of Sciences of the United States of America, 93: 11,709–11,712. EHRLICH, P. R. & WILSON, E. O., 1991. Biodiversity studies: science and policy. Science, 253: 750–752. ESWG, 1995. A national ecological framework for Canada. Agriculture and Agri–food Canada, Research Branch, Centre for Land and Biological Resources Research; and Environment Canada, State of the Environment Directorate, Ecozone Analysis Branch, Ottawa/Hull. GALLANT, A. L., BINNIAN, E. F., OMERNIK, J. M. & SHASBY, M. B., 1995. Ecoregions of Alaska. U.S. Geological Survey, Washington, DC. JONGMAN, R. H. G., TER BRAAK, C. J. F. & VAN TONGEREN, O. F. R. (Eds.), 1995. Data analysis in community and landscape ecology. Cambridge University Press, New York. KERR, J. T., 1997. Species richness, endemism, and the choice of areas for conservation.

Conservation Biology, 11: 1,094–1,100. LEVIN, S. A., 1992. The problem of pattern and scale in ecology. Ecology, 73: 1,943–1,967. MYERS, N., MITTERMEIER, R. A., MITTERMEIER, C. G., FONSECA, G. A. B. DA & KENT, J., 2000. Biodiversity hotspots for conservation priorities. Nature, 403: 853–858. OLSON, D. M. & DINERSTEIN, E., 1998. The Global 200: A representation approach to conserving the Earth’s most biologically valuable ecoregions. Conservation Biology, 12: 502–515. OMERNIK, J. M., 1995. Level III ecoregions of the continental US. US Environmental Protection Agency, Washington, DC. PEARSON, D. L. & CASSOLA, F., 1992. World–wide species richness patterns of tiger beetles (Coleoptera: Cicindelidae): Indicator taxon for biodiversity and conservation studies. Conservation Biology, 6: 376–391. PRENDERGAST, J. R., QUINN, R. M., LAWTON, J. H., EVERSHAM, B. C. & GIBBONS, D. W., 1993. Rare species, the coincidence of diversity hotspots and conservation strategies. Nature, 365: 335–337. RICKETTS, T. H., DAILY, G. C. & EHRLICH, P. R. (in press). Does butterfly diversity predict moth diversity? Testing a popular indicator taxon at local scales. Biological Conservation. RICKETTS, T. H., DINERSTEIN, E., OLSON, D. M. & L OUCKS, C., 1999a. Who’s where in North America: patterns of species richness and the utility of indicator taxa for conservation. Bioscience, 49: 369–381. RICKETTS, T. H., DINERSTEIN, E., OLSON, D. M., LOUCKS, C., EICHBAUM, W., KAVANAGH, K., HEDAO, P., HURLEY, P. , CARNEY, K. M., ABEL, R. & WALTERS, S., 1999b. Terrestrial ecoregions of North America: A conservation assessment. Island Press, Washington, DC. ROSENZWEIG, M. L., 1995. Species diversity in space and time . Cambridge University Press, Cambridge. SCOTT, M. J., DAVIS, F., CSUTI, B., NOSS, R., BUTTERFIELD, B., GROVES, C., ANDERSON, H., CAICCO, S., D’ERCHIA, F., EDWARDS, T. C. J., ULLIMAN, J. & WRIGHT, R. G., 1993. Gap analysis: protecting biodiversity using geographic information systems. Wildlife Monographs, 123: 1–41. SISK, T. D., LAUNER, A. E., SWITKY, K. R. & EHRLICH, P. R., 1994. Identifying extinction threats: Global analyses of the distribution of biodiversity and the expansion of the human enterprise. BioScience, 44: 592–604. WEAVER, J. C., 1995. Indicator species and scale of observation. Conservation Biology, 9: 939–942. WILLIAMS, P. H. & GASTON, K. J., 1994. Measuring more of biodiversity: Can higher–taxon richness predict wholesale species richness? Biological Conservation, 67: 211–217. ZAR, J. H., 1999. Biostatistical analysis. 4th Edition. Prentice Hall, Upper Saddle River, NJ.

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Animal Biodiversity and Conservation 24.1 (2001)

101

A survey of a pampas deer, Ozotoceros bezoarticus leucogaster (Arctiodactyla, Cervidae), population in the Pantanal wetland, Brazil, using the distance sampling technique W. M. Tomás, W. McShea, G. H. B. de Miranda, J. R. Moreira, G. Mourão & P. A. Lima Borges

Tomás, W. M., McShea, W., Miranda, G. H. B. de, Moreira, J. R., Mourão, G. & Lima Borges, P. A., 2001. A survey of a pampas deer, Ozotoceros bezoarticus leucogaster (Arctiodactyla, Cervidae), population in the Pantanal wetland, Brazil, using the distance sampling technique. Animal Biodiversity and Conservation, 24.1: 101–106. Abstract A survey of a pampas deer, Ozotoceros bezoarticus leucogaster (Arctiodactyla, Cervidae), population in the Pantanal wetland, Brazil, using the distance sampling technique.— The pampas deer is an endangered South American species which occurs in open grasslands and savannas. This aim of this survey was to evaluate the use of the distance sampling technique to estimate densities of the species in the Pantanal wetland, as well as to analyze the applicability of the method for a monitoring program. The surveys were conducted on roads from vehicles and also on foot along 26 parallel transects in November 1999 and 2000 at Campo Dora ranch, south-central Pantanal, Brazil. Deer densities were estimated using the program DISTANCE, and the program MONITOR was used to run a power analysis to estimate the probability of detection of a decline in the population. The deer density estimated from vehicles, with data from both years, was 9.81±3.8 individual/km2, and 5.53±0.68 individuals/km2 from transects sampled on foot. The power analysis of these data revealed a monitoring program would require at least two surveys per year over seven years to obtain a 90% chance of detecting a 5% decline in the population. Our results also indicate surveys from roads are not recommended for pampas deer counts as the animals appear to keep a relatively safe distance from cars. Key words: Pampas deer, Ozotoceros, Distance sampling technique, Pantanal wetland, Population survey. Resumen Estudio de una población de venados de la Pampa Ozotoceros bezoartcus leuogaster (Artiodactyla, Cervidae) en el Pantanal, Brasil, mediante la técnica del muestreo a distancia.— El venado de la Pampa es una especie sudamericana en peligro de extinción que se encuentra en praderas abiertas y sabanas. El objetivo de este trabajo es evaluar el uso de la técnica de muestreo a distancia para estimar densidades de esta especie en el Pantanal, así como analizar la aplicabilidad de este método a un programa de monitoreo. Los estudios se realizaron desde caminos, con vehículos y a pie, a través de 26 transectos paralelos en noviembre de 1999 y 2000, en la hacienda Campo Dora, Pantanal, Brasil. Las densidades de venados se estimaron con el programa DISTANCE, empleándose el programa MONITOR para efectuar un análisis de poder estimativo para la detección de un descenso en la población de venados. La densidad de venados estimada desde los vehículos fue de 9.81±3.8 individuos/km2, mientras la obtenida desde transectos realizados a pie fue de 5.53±0.68 individuos/km2. Ambas densidades incluyen datos de los dos años de estudio. El análisis potencial de estos datos señala que un programa de monitoreo precisaría como mínimo de dos muestreos por año, durante siete años, para obtener una probabilidad del 90% de detectar un descenso del 5% en la población. Los resultados de este estudio indican asimismo que las observaciones efectuadas desde caminos no son recomendables para el recuento de venados de la Pampa, ya que se observó que éstos tienden a mantener una distancia de seguridad respecto a los coches. Palabras clave: Venados de la Pampa, Ozotoceros, Técnica de muestreo a distancia, Pantanal, Estudio de población. (Rebut: 20 VII 01; Final acceptance: 1 X 01) W. M. Tomás, EMBRAPA Recursos Genéticos e Biotecnologia, PqEB, Final W5 Norte, 70770–900 Brasília DF, Brasil. e–mail: tomasw@cenargen.embrapa.br ISSN: 1578–665X

102

Introduction The pampas deer (Ozotoceros bezoarticus L., 1758) is a species characteristic of open habitats in South America, with historic distribution ranging from central Argentina to mid–western and northeastern Brazil, eastern Bolivia, Paraguay and Uruguay (CABRERA, 1943; CARVALHO, 1973; JUNGIUS, 1976; JACKSON & GIULIETTI, 1988; JACKSON & LANGGUTH, 1987; TOMÁS, 1995). The species is included in the International Union for Conservation of Nature (IUCN) Red Data Book as a lower risk, near–threatened species (WEMMER, 1998); it is also considered endangered by the United States Department of Interior–USDI, is in the Appendix I of Convention International Trade Endangered Species–CITES (CITES, 1995), and is listed as endangered in Brazil (FONSECA et al., 1994). Population declines in this species have been attributed to habitat destruction related to agricultural expansion, poaching, and diseases transmitted by cattle (M ERINO et al., 1997). Although surveys and monitoring programs have been recommended in conservation action plans for the species (CBSG, 1993; WEMMER, 1998), little has been published on population size estimates for this species in Brazil (e.g., LEEUWENBERG & LARA RESENDE, 1994; RODRIGUES, 1996; MOURÃO et al., 2000). The largest population is known to occur in the Pantanal wetland, and is estimated at 60,000 individuals (MOURÃO et al., 2000). MOURÃO et al. (2000) called for long–term monitoring of pampas deer populations in the Pantanal by means of ground surveys, but we know of no concerted effort to evaluate the appropriate techniques to accomplish this goal. Distance sampling techniques offer potential for a monitoring program because the assumptions are relatively robust and the protocols can be quickly taught to survey staff (ANDERSON et al., 2001). This survey aims to evaluate the use of the distance sampling technique (BURNHAM et al., 1980) to estimate densities of pampas deer through transects sampled on foot and/or from a vehicle, as well as to analyze the applicability of the method and sampling protocol for a monitoring program of this species.

Material and methods The survey was conducted in an area of 8,400 ha of the Campo Dora ranch (40,000 ha) located 90 km from Corumbá, Mato Grosso do Sul State, Brazil, in the south–central Pantanal wetland. The average annual rainfall is 1,182 mm and the average temperature varies from 31.6°C to 20.2°C (SORIANO, 1999). The Pantanal vegetation consists of a mosaic of several forested and open habitats that vary in topography and flooding regime (PRANCE & SCHALLER, 1982). The open habitat is flooded from January to June, with a draining period from July to August.

Tomás et al.

Lower areas may retain water until October, and some permanent ponds are scattered throughout the study area. During the flooding period the grassland is substituted by a massive formation of aquatic macrophytes, which is gradually replaced by grasses as the water recedes. The principal economic activity in the study area is cattle ranching. Pampas deer were simultaneously counted from vehicles in three different, non–intercepting transects (roads) on November, 1999, and in the same month of 2000, between 7.30 and 11.00 a.m., at a speed of 20 km/h. In each car, one observer standing in the back of the vehicle recorded the presence of deer clusters on both sides of the road. For each sighting, the vehicle stopped and the perpendicular distance to the road was measured by counting steps, which were then converted into meters. The conversion factor had been previously established for each observer. The number of individuals was recorded in each cluster as observed from the vehicle without optical instruments, as well as the actual number of deer per cluster, which included any additional individual observed afterwards during the perpendicular distance estimation and/or with binoculars. Deer were also counted on foot from 12 parallel east–west oriented transects in November, 1999, and 14 transects in 2000, between 7.30 and 11.00 a.m. In 1999, fifteen observers, divided into six groups of two or three people, surveyed the transects starting from a road (with approximately north–south orientation) that traversed the Campo Dora ranch. In the 2000 survey, seven groups of at least 2 observers sampled the south–north and east–west oriented transects The transects were separated by 2 km, with lengths varying from 3 to 11 km. Because pampas deer do not use forested habitats (MERINO et al., 1997), we excluded the interception with forest patches from the total length of each transect. Deer clusters were recorded using the same protocol defined in the survey from vehicles. Deer cluster densities were estimated using the program DISTANCE (L AAKE et al., 1993; BUCKLAND et al., 1993) by selecting the model that best fit the data (BURNHAM et al., 1980). The data were analyzed separately for each year. The histograms of observation distributions were examined visually and truncated as necessary. To determine average cluster sizes and calculate densities, truncation was based on the definition of the effectively sampled area given by the program DISTANCE to avoid any bias of cluster size being related to the sighting distance. DISTANCE produces a variance estimate that has 3 components: the first is the proportion due to the observer’s ability to detect animals along the transect; the second due to the variability between transect lines; and the third due to variance in group size observed. The program MONITOR (GIBBS, 1995) was used

103

Animal Biodiversity and Conservation 24.1 (2001)

to run a power analysis with the data obtained from the transects surveyed on foot in 1999. To perform this analysis we need to know the number of observations expected along each transect and their variance. The number of deer/km of transect for each of the 12 transects surveyed were used in order to estimate number of observations and calculate variance. Simulations were made with a one–tailed test, and the amount of effort needed to establish a 90% probability of detecting a population decline was estimated as to avoid error type II. In these simulations we varied: 1. The number of transects per year; 2. The number of times the sampling would be repeated per year; 3. The number of years of monitoring necessary to detect decline.

Results A total of 58.7 km of transects were surveyed by vehicle in 1999, and 29.5 in 2000. Twenty–seven deer clusters were detected, with a total of 58 individuals in the first year, and 31 clusters (79 individuals) in 2000. The pooled data obtained by vehicle displayed in figure 1B demonstrate that few deer were observed close to the road, contrasting with the data obtained from the surveys made on foot (fig. 1A). One critical assumption with the distance techniques is that the further the distance from the survey line, the lower the count (BURNHAM et al., 1980; BUCKLAND et al., 1993; LAAKE et al., 1993). In order to meet this assumption we had to truncate observations up to 100 m from the road, thus reducing the analysis to 58 clusters. The model which best fitted our data was a half normal adjustment. The density estimate was 3.63±1.31 clusters/km2 with an average cluster size of 2.38±0.28 deer/ cluster. The deer density was estimated to be 9.81±3.8 deer/km2. For the vehicular survey, the probability of detection accounted for 43.9% of the variance, the encounter rate 38.7%, and the cluster size 17.4%. The population size was estimated as 824±318.68 pampas deer. A total of 77.6 km of transects was surveyed on foot in 1999. Seventy–eight deer clusters were recorded in 1999. Unlike the vehicle survey, examination of the data indicated no truncation along the survey line was necessary. A half normal model was found to best fit our data, and effective sampled width was 163.65±14.75 m. Cluster density was estimated as 3.07±0.59 clusters/km2 and the average cluster size as 2.23±0.18 individuals. The deer density for our study area was estimated to be 6.85±1.43 individuals/km2 and the population size was estimated as 575±120.16 pampas deer for the Campo Dora ranch. The encounter rate (differences between transect lines) accounted for 67.3% of observed variance, leaving 18.6% for detection along the transect line and 14.1% for cluster size. A total of 106.51 km of transects was surveyed

on foot in 2000. Ninety–eight deer clusters were recorded in 2000. A half normal model was found to best fit our data, and effective sampled width was 175.96±14.02 m. Cluster density was estimated as 2.61±0.32 clusters/km2 and the average cluster size as 1.91±0.13 individuals. The deer density for our study area was estimated to be 4.99±0.70 individuals/km2 and the population size was estimated as 419±59.84 pampas deer for the Campo Dora ranch. The encounter rate accounted for 45.8% of the variance, with 32.2% of the variance due to detection probability and 22.0% to cluster size. A total of 186 deer clusters was recorded during the two sampling periods. The sighting of clusters was rare 500 m beyond the transects (fig. 1A), with a positive correlation between the log of cluster size and perpendicular distance from the transect (r = 0.037, t = 2.65, Df = 184, P = 0.009). The data at this distance was therefore truncated. Analysis of the pooled data from 1999 and 2000 indicated that the best model fit was a half normal key (fig. 2), and the effective sampled width was 181.12±10.76 m. The estimated cluster density was 2.68±0.30 clusters/km 2, and the average cluster size was 2.06±0.10 individuals. The deer density for our study area was estimated to be 5.53±0.68 individuals/km2 and the population size was estimated as 465±57.11 pampas deer. The encounter rate accounted for 59.8% of the variance, leaving 23.4% for detection probability and 16.8% for cluster size. Our power analysis of the 1999 data revealed that to obtain a 90% chance of detecting a 5% annual decline in the studied population, at least two surveys per year for 7 years would be necessary. On the other hand, it would take at least 10 years with one survey per year to obtain a 90% chance of detecting the same annual decline. In a shorter time period, three surveys per year would be necessary for 5 years to detect a 7% decline (table 1).

Discussion Reviewed survey information revealed few studies of pampas deer whose survey protocols offered viable data for comparison. RODRIGUES (1996) found 1.97±1.38 deer/group and 0.1 deer/km2 for Emas National Park, applying the distance sampling technique to analyze counts obtained from a vehicle using roads as transects. LEEUWENBERG & LARA RESENDE (1994) found 1.26 (SD = 0.65) deer/km2 in night counts using strip transects 100 m wide, in the environmental protection area of Gama– Cabeça de Veado, near Brasilia. In northern Argentina, MERINO & BECCACECI (1999) counted pampas deer from an airplane defining a strip of 300 m in each side, and found an average group size of 1.75±0.78 deer/group, and a density of 0.39±0.35 deer/km2. The authors also surveyed deer from roads using a strip of 300 m on each

Tomás et al.

104

60 50 40 30 20 10 0

E F G H I Distance classes

Fig. 1. Distribution of observed clusters of pampas deer (Ozotoceros bezoarticus) at different distances from the transect lines surveyed on foot (A) and from a vehicle (B), at Campo Dora ranch, Pantanal, Brazil. Distance classes: A. 0–49; B. 50–99; C. 100–149; D. 150–199; E. 200–249; F. 250–299; G. 250–299; H. 300–349; I. 350–399; J. 400–449; K. 450–499; L. 500–549; M. 550–599. Fig. 1. Distribucion de los grupos observados del venado de la Pampa (Ozotoceros bezoarticus) en diferentes distancias de la transección realizada a pie (A) y en vehiculo (B), en la Hacienda Campo Dora, Pantanal, Brasil. (Para las clases de distancias, ver arriba.)

side of the vehicle at 40 km/h, but the estimates were not reported. For the Pantanal, MOURÃO et al. (2000) found an overall density of 0.25 groups/ km2 for the entire floodplain and an average group size of 1.67±0.85 deer, using aerial survey techniques. In areas of slightly higher elevation in the Central Pantanal, MOURÃO et al. (2000) found a density of 0.57 groups/km2. The survey results presented in this study produced the highest population density reported

to date for this species, with 2.68±0.30 clusters/km2, and an average cluster size of 2.06±0.10 individuals. This result is due in part to our survey of only grasslands and not the intervening forest, which is included in any aerial survey. Campa Dora is also high quality pampas deer habitat and probably represents one of the highest density limits for pampas deer within the Pantanal (pers. obs.). MOURÃO et al. (2000) indicate that the relatively small deer is difficult to monitor from aerial surveys.

105

Animal Biodiversity and Conservation 24.1 (2001)

P 1.10 0.99 0.88 0.77 0.66 0.55 0.44 0.33 0.22 0.11 0 0

Table 1. Probabilities of detecting declines in the pampas deer (Ozotoceros bezoarticus) population from Campo Dora ranch, Pantanal, Mato Grosso do Sul, Brazil, using distance sampling technique in transects conducted on foot: Pd. Percent decline; N. Number of surveys per year.

100 150 200 250 300 350 400 Perpendicular distance (m)

Fig. 2. Distribution of observed pampas deer (Ozotoceros bezoarticus) clusters at different distances from the transect line surveyed on foot, pooled from 1999 and 2000, at Campo Dora ranch, Pantanal, Brazil; and the fitted curve of detection probabilities (P). Fig. 2. Distribución de grupos de venados de la Pampa (Ozotoceros bezoarticus) a diferentes distancias de la transección realizada a pie, datos de 1999 y 2000 agrupados, en la Hacienda Campo Dora, Pantanal, Brasil; y la curva ajustada de las probabilidades de detección (P).

Tabla 1. Probabilidad de detección del descenso en la población de venados de la Pampa (Ozotoceros bezoarticus) en la Hacienda Campo Dora, Pantanal, Mato Grosso do Sul, Brasil, usando la técnica del muestreo a distancia en transectos a pie: Pd. Descenso del porcentaje; N. Número de observaciones por año.

N Period 5 years

7 years

Ground surveys are more labor intensive, but may supplement a more broad–scale aerial survey. The present study is the first to make direct comparisons between vehicle and foot surveys for this species and indicate that surveys from roads should be avoided. This recommendation makes no distinction between the survey being made from a vehicle or walking, because the large variance about the estimate along roads produced no viable monitoring schedule in a power analysis. Roads in the Pantanal tend to be constructed in higher areas, avoiding obstacles, channels and marshy areas. This may influence the location of the deer clusters in relation to the roads in such a way that no representative sampling of the population would be obtained. Additionally, it is possible that pampas deer tend to keep a relatively safe distance from roads, as a means of avoiding the movements of cars, even if this movement is not intense in the Pantanal. The results of the power analysis indicate that an adequate monitoring program, using the distance sampling technique, to detect population declines is feasible. As with previous surveys using distance techniques (ANDERSON et al., 2001), teams of students and volunteers were utilized to complete the survey. The comparable results between the 2 survey years, despite using different teams of students, indicate that the protocols can

10 years

0.15

0.18

0.21

0.3

0.4

0.52

0.6

0.49

0.68

0.78

0.86

0.64

0.86

0.95

0.97

0.75

0.93

0.98

0.23

0.25

0.31

0.37

0.48

0.69

0.8

0.87

0.74

0.92

0.97

0.91

0.99

0.97

0.99

0.3

0.44

0.5

0.6

0.78

0.95

0.99

0.97

0.99

be sufficiently basic for use by non–professionals or people with litle experience. For large areas, such as the Pantanal, we suggest several areas such as Campo Dora, should be established and distributed throughout the region, covering a gradient of habitat types used by pampas deer. Each of these sampling areas could be monitored after a power analysis to establish a suitable local survey program. As recommended by MOURÃO et al. (2000), ground surveys may be a necessity to accurately monitor trends in pampas deer abundance in the Pantanal. By utilizing teams of students and volunteers within select ranches the present study indicates it is feasible to monitor population trends using standard distance sampling techniques.

106

Acknowledgements We thank the Smithsonian Institution/National Zoological Park/Conservation and Research Center, the IUCN/Deer Specialist Group, the EMRAPA/ Recursos Genéticos e Biotecnologia, the EMBRAPA/ Pantanal, and the Conservation International for financial and logistic support given to the Training Course on Population Survey Techniques, during which these data were collected. We thank also the Campo Dora ranch owners for permission to conduct the surveys on their property. Finally, we are grateful to the students who helped us in data collecting: Angela Begrow, Elisabeth Burkhardt, Humberto Gómez, Angelika Jüncke, Stewart Klorfine, Bernardo Lartigau, Angela Nuñez, Boris Ríos, Cézar Scheide, Diego Varela, Gustavo Porini, Tarcisio S. Santos Jr, Marcelo D. Beccaceci, Alexander V. Christianini, Milene M. Martins, Luis F. Pacheco, Lila A. Sainz Bacherer, Glaucia H. F. Seixas, Ana C. R. Lacerda, and Hector A. Regidor.

References ANDERSON, D. R., BURNHAM, K. P., LUBOW, B. C., THOMAS, L., CORN, P. S., MEDICA, P. A. & MARLOW, R. W., 2001. Field trials of line transect methods applied to estimation of desert tortoise abundance. Journal of Wildlife Management, 65: 583–597. BUCKLAND, S. T., ANDERSON, D. R., BURNHAM, K. P. & LAAKE, J. L., 1993. Distance Sampling. Estimating Abundance of Biological Populations. Chapman & Hall, London. BURNHAM, K. P., ANDERSON, D. R. & LAAKE, J. L., 1980. Estimation of Density from Line Transect Sampling of Biological Populations. Wildlife Monographs, 72: 1–202. CABRERA, A., 1943. Sobre la sistemática del venado y su variación individual y geográfica. Rev. Mus. La Plata, 3: 5–41. CARVALHO, C. T., 1973. O veado mateiro, situação e distribuição. Bol. Tec. do I. F. São Paulo, 7: 9–22. CBSG, 1993. Population and Habitat Viability Assessment for the Pampas Deer Ozotoceros bezoarticus . IUCN/SSC Captive Breeding Specialist Group, Gland. CITES, 1995. Appendix I. Convention International of Trade Endangered Species, Gland, Switzerland. FONSECA, G. A. B., RYLANDS, A. B., COSTA, C. M. R., MACHADO, R. B. & LEITE, Y. L. R., 1994. Livro Vermelho dos Mamíferos Brasileiros Ameaçados de Extinção . Fundação Biodiversitas, Belo Horizonte. GIBBS, J. P., 1995. Monitor Users Manual. Yale University, New Haven, Connecticut. JACKSON, J. E. & GIULIETTI, J. D., 1988. The food of pampas deer Ozotoceros bezoarticus celer in relation to its conservation in a relict natural grassland in Argentina. Biological Conservation,

Tomás et al.

45: 1–10. JACKSON, J. E. & LANGGUTH, A., 1987. Ecology and status of the pampas deer in the Argentinian Pampas and Uruguay. In: Biology and Management of Cervidae: 402–409 (C. M. Wemmer, Ed.). Smithsonian Institution Press, Washington, DC. J UNGIUS, H., 1976. Status and distribution of threatened deer species in South America. In: World wildlife Yearbook 1975–76: 203–217. World Wildlife Fund, Morges, Switzerland. LAAKE, J. L., BUCKLAND, S. T., ANDERSON, D. R. & BURNHAM, K. P., 1993. Distance User’s Guide V 2.0. Colorado Cooperative Fish & Wildlife Research Unit., Colorado State University, Fort Collins, CO. LEEUWENBERG, F. & LARA RESENDE, S., 1994. Ecologia de cervídeos na Reserva Ecológica do IBGE– DF: manejo e densidade de populações. Cad. Geoc., 11: 89–95. MERINO, M. L. & BECCACECI, M. D., 1999. Ozotoceros bezoarticus (Artiodactyla, Cervidae) en Corrientes, Argentina: distribución, población y conservación. Iheringia, Ser. Zool., 87: 87–92. MERINO, M. L., GONZALES, S., LEEWENBERG, F., RODRIGUES, F. H. G., PINDER, L. & TOMÁS, W. M., 1997. Veado– campeiro (Ozotoceros bezoarticus). In: Biologia e Conservação de Cervídeos Sulamericanos: Blastocerus, Ozotoceros e Mazama: 42–58 (J. M. Barbanti Duarte, Ed.). FAPESP/FUNEP/UNESP, Jaboticabal, SP. MOURÃO, G. M., COUTINHO, M., MAURO, R., CAMPOS, Z., TOMÁS, W. M. & MAGNUSSON, W., 2000. Aerial Surveys of Caiman, Marsh Deer and Pampas Deer in the Pantanal Wetland of Brazil. Biological Conservation, 92: 175–183. PRANCE, G. T. & SCHALLER, G. B., 1982. Preliminary Study of Some Vegetation Types of the Pantanal, Mato Grosso, Brazil. Brittonia, 34: 228–251. RATTER, J. A., POTT, A., CUNHA, C. N. DA & HARIDASAN, M., 1988. Observations on Wood Vegetation Types in the Pantanal and at Corumbá, Brazil. Notes RGB Edinb., 45: 503–525. RODRIGUES, F. H. G., 1996. História Natural e Biologia Comportamental do Veado–campeiro (Ozotoceros bezoarticus) em cerrado do Brasil Central. Ph. D. Thesis, Univ. Estadual de Campinas. SORIANO, B. M. A., 1999. Caracterização climática da sub–região da Nhecolândia, Pantanal, MS. Anais do II Simpósio sobre Recursos Naturais e Socio–econômicos do Pantanal, EMBRAPA– Pantanal, Corumbá, MS. TOMÁS, W. M., 1995. Seasonality of the antler cycle of Pampas deer (Ozotoceros bezoarticus leucogaster) from Pantanal wetland, Brazil. Studies on Neotropical Fauna and Environment, 30: 221–227. WEMMER, C. (Ed.), 1998. Deer. Status Survey and Conservation Action Plan . IUCN/SSC Deer Specialist Group. IUCN, Gland, Switzerland and Cambridge, UK.

107

Animal Biodiversity and Conservation 24.1 (2001)

Does foraging habitat quality affect reproductive performance in the Little Egret, Egretta garzetta? C. Tourenq, C. Barbraud, N. Sadoul, A. Sandoz, K. Lombardini, Y. Kayser & J.–L. Martin

Tourenq, C., Barbraud, C., Sadoul, N., Sandoz, A., Lombardini, K., Kayser, Y. & Martin, J.–L., 2001. Does foraging habitat quality affect reproductive performance in the Little Egret, Egretta garzetta? Animal Biodiversity and Conservation, 24.1: 107–116. Abstract Does foraging habitat quality affect reproductive performance in the Little Egret, Egretta garzetta?— In order to understand the role of foraging habitat quality on fecundity parameters we measured habitat use, breeding parameters, and body condition of chicks in six colonies of Little Egrets in southern France. The foraging habitat available differed between colonies; it was mainly natural marshes around the Carrelet colony, agricultural lands (rice fields and dry crops) around the Agon colony, a mix of agricultural and natural lands around the Redon and Fiélouse colonies, a mix of natural and urbanised/industrial lands around the Palissade colony, and mainly cultivated and urbanised lands around the Chaumont colony. The habitat attractiveness to adult Little Egret breeding was higher for natural marshes than for other habitat types. Agricultural marshes (rice fields) came next. Other human–made habitats came last. Clutch size and body condition index of chicks did not differ between colonies. Brood size was influenced by both the association of the proportion of natural marshes in the foraging area and clutch size, and the association of clutch size and the total number of heron pairs in the colony. The effect of the proportion of natural marshes could not be distinguished from the effects of the colony size. The potential influence of other parameters not taken into account in this study is discussed. Key words: Egretta garzeta, Foraging habitat, Reproductive parameters, Body condition, Natural marshes. Resumen ¿Afecta la calidad del hábitat alimentario a la capacidad reproductiva de la garceta común, Egretta garzetta?— Con la finalidad de conocer el papel que ejerce la calidad del hábitat alimentario sobre los parámetros de fecundidad, se evaluaron el uso del hábitat, los parámetros reproductivos y las condiciones físicas de los polluelos de seis colonias de garceta común en el sur de Francia. El hábitat alimentario disponible variaba de unas colonias a otras, siendo principalmente marismas naturales en el entorno de la colonia de Carrelet, terrenos agrícolas (campos de arroz y cultivos de secano) alrededor de la colonia de Agon, una combinación de terrenos agrícolas y naturales alrededor de las colonias de Redon y Fiélouse, una combinación de terrenos naturales y urbanizados/industriales alrededor de la colonia de Palissade, y principalmente terrenos cultivados y urbanizados alrededor de la colonia de Chaumont. En la época de reproducción, los adultos de garceta común se sienten atraídos principalmente por las marismas naturales, en detrimento de otros tipos de hábitat. Las tierras agrícolas anegadas (campos de arroz) siguen en orden de preferencia, mientras los hábitats construidos por el hombre ocupan el último lugar. El tamaño de la puesta y el índice de estado físico de los polluelos no mostraron diferencias entre las colonias. El tamaño de la nidada estuvo influenciado tanto por la asociación de la proporción de marismas naturales en el hábitat alimentario y el tamaño de la puesta, como por la asociación del tamaño de la puesta y el número total de parejas de garzas de la colonia. El efecto de la proporción de marismas naturales no se puede diferenciar del ejercido por el tamaño de la colonia. Se discute también la influencia potencial de otros parámetros que no se han tenido en cuenta en este estudio. Palabras clave: Egretta garzeta, Hábitat alimentario, Parámetros reproductores, Condiciones físicas, Marismas naturales. (Received: 29 VI 01; Final acceptance: 30 IX 01) Christophe Tourenq(1), Christophe Barbraud, Nicolas Sadoul, Alain Sandoz, Katia Lombardini & Yves Kayser, Station Biologique de la Tour du Valat, Le Sambuc, 13200 Arles, France.– Christophe Tourenq & Jean–Louis Martin, CEFE/CNRS, 1919 route de Mende, F–34293 Montpellier Cedex 5, France. (1)

e–mail: tourenq@tour-du-valat.com

ISSN: 1578–665X

Tourenq et al.

108

Introduction The distribution of food and the ability of individuals to exploit it have major implications on animal population dynamics (STEPHEN & KREBS, 1986). The exploitation of favourable patches with good foraging efficiency is linked to the structure of the habitat (CARTER & ABRAHAMS, 1997) and, on a broader scale, to the landscape (DOOLEY & BOWERS, 1998). Human activities, such as agricultural practices (soil management or pesticides and fertiliser use), affect the abundance, quality and availability of food resources for birds (e.g. AEBISCHER, 1990; TUCKER, 1992; W ILSON et al., 1996; P ETERSEN , 1998; CHAMBERLAIN et al., 1999). Although farmland has often been viewed in a rather dichotomic way as the juxtaposition of more or less isolated patches of suitable (non cultivated) areas within a matrix of non–suitable (cultivated) habitats (FARINA, 1997; PETERSEN, 1998), farmland can be considered, from an animal species’perspective, as a whole made up of a mosaic of habitat patches providing resources of varying abundance and quality. This raises the question of the function and value of the farmed component of the landscape in the functioning of these animal populations. In birds, some fitness components related to reproductive parameters (e.g. clutch size, brood size, breeding success) or chicks’condition and growth, may be directly connected to the availability and quality of foraging habitats in the vicinity of nesting sites (e.g. CLAMENS & ISENMANN, 1989; TIAINEN et al., 1989; BURGER & GOCHFELD, 1991; HAFNER et al., 1993) and can, therefore be used as estimators of habitat quality. One of the main crops on a global scale is rice (Oryza spp.), which covers over 11% of the farmed lands (FASOLA & RUIZ, 1997). Rice fields often replaced natural wetlands and numerous wetland bird species use the agricultural wetlands provided by rice cultivation (FASOLA & RUIZ, 1997). In some regions (e.g. the Ebro Delta in Spain or the Po Valley in Italy), rice fields have actually become the only significant wetland available for waterbirds (F ASOLA & R UIZ , 1997). This prompted several studies to understand the consequences of rice field use on the ecology and population dynamics of such species (see FASOLA & RUIZ, 1997; TOURENQ et al., 2001). As one of the main wetland complexes of the Western Palearctic, the Camargue, southern France, is a major area of rice production in Europe (FASOLA & RUIZ, 1997). It also contains one of the largest industrial salt pans in the Mediterranean and is bordered by a large industrial complex (BATTY et al., 1996; BERNY et al., in press). This mosaic consists of natural, agricultural and industrial wetlands and offers a rather unique opportunity to study the respective value of man–made and more natural wetlands on the health of water bird populations.

Among the species which extensively use rice fields, the Little Egret ( Egretta garzetta ), a common heron in the Camargue, provides a good study model. The Little Egret is a colonial species that uses a wide range of habitats for foraging, including all types of wetlands (TOURENQ et al., 2000). In this context, artificial wetlands such as rice fields may consequently provide food resources, especially during the breeding period in the Camargue (HAFNER et al., 1986; HAFNER & FASOLA, 1992). TOURENQ et al. (2000) have shown that egret numbers have increased over the past decades together with an increase in the area cultivated in rice. At first glance, rice cultivation seems therefore to have been beneficial to egrets. However, BENNETTS et al. (2000) and LOMBARDINI et al. (2001) showed that this correlation may be misleading. Respectively, these authors found that reproductive parameters have decreased during the past decades and that this species preferentially used natural marshes rather than anthropized habitats for foraging. To investigate this point further: 1. The foraging habitat use of adult Little Egrets around breeding colonies was investigated in order to identify the habitat selected in relation to its availability; and 2. Clutch size, brood size and the condition of Little Egret chicks were hypothesized that were influenced by the proportion of this habitat.

Material and methods Study area The Camargue deltaic complex, southern France (43°40’–43°30’ N, 4°10’–4°30’ E; ca. 1,450 km²), is renowned as one of the most important wintering and breeding grounds in Europe for water birds (HEATH & EVANS 2000). Natural habitats cover some 60,000 ha (±41% of total surface) and salt pans some 21,000 ha (±15% of total surface) in the southern region. Some 24,000 ha (±16% of total surface) are devoted to rice farming, whereas dry crops cover 26,000 ha (±18% of total surface). Located in the south– eastern area, the industrial complex of Fos–sur– Mer (metal transformation and refineries) covers about 9% of the total surface of the delta (CHAUVELON, 1996). This study was carried out in 1998 and 1999 in six colonies of tree–nesting herons located within, or adjacent to, the Camargue: Agon, Fiélouse and Chaumont in 1998, and Carrelet, Redon, Palissade in 1999 (fig. 1). Palissade is situated in the south–eastern part of the delta, between the industrial complex of Fos–sur–Mer and the industrial salt pans of Salin de Giraud. Colonies of Agon, Fiélouse, Redon and Carrelet are located in the semi–natural central area of the delta. The Chaumont colony is situated outside the delta within a vineyard cultivation area near a coastal tourist resort.

109

Animal Biodiversity and Conservation 24.1 (2001)

N St. Gilles 0

Aigues Mortes

e ôn h R

Rh ôn e

A C

France

Arles

10 km

Vaccares Lagoon F R

Stes Maries Salin de Giraud Port St. Louis

Mediterranean Sea

Fig. 1. Study area with the location of Little Egret colonies sampled in 1998 and 1999: C. Carrelet, A. Agon; F. Fiélouse; R. Redon; P. Palissade; ch. Chaumont. (Inset shows the location of the study area in France.) Fig. 1. Área de estudio con la localización de las colonias de garceta común muestreadas en 1998 y 1999: C. Carrelet; A. Agon; F. Fiélouse; R. Redon; P. Palissade; ch. Chaumont. (El mapa del recuadro muestra la localización del área de estudio en Francia.)

Foraging habitat composition and habitat use by egrets Digitised aerial photographs of the study area (scale 1:20,000) were ortho–rectified, georeferenced and gathered (software Geoimage, on a UNIX workstation; SANDOZ & CHAUVELON, 1998). A total of 18,500 agricultural plots and natural areas and marshes were digitised with GIS MapInfo. Land–use was updated each year during the study through integration of classified satellite images. MapInfo was used to calculate the area of habitats available within the foraging range of colonies (radius = 10 km; HAFNER & FASOLA, 1992). For each colony, the proportions of each habitat type were estimated: rice fields (RICE), dry cultivation lands such as vineyards, corn, pastures,… (DRY), urbanised and industrial areas (URB), flooded surfaces of natural marshes (NM) and other natural lands such as salt flats called “sansouïre” (SAN). The proportion of

woodlands, sea, as well as central parts of lagoons and Rhône river arms were not included in the analysis, as they are not used by Little Egrets when foraging. The one meter shore zone of lagoons and river arms which is used by foraging Little Egrets were only considered (C. Toureng, pers. obs.). For the five colonies within the delta (Agon, Carrelet, Fiélouse, Redon, Palissade), foraging individuals were counted by aerial surveys (at 400 ft above ground) of the foraging range of each colony in the morning. Each flock or individual counted on one of the habitat types was recorded and plotted on a map. In order to take into account the variation in foraging habitat use that may occur during the reproductive season (LOMBARDINI et al., 2001), aerial surveys were conduced at the egg–laying stage, the brooding stage and at the fledging stage. In 1998, nine aerial surveys (three per reproductive stage) were made for Agon and Fiélouse but in 1999, due to

Tourenq et al.

110

meteorological and technical hazards and airforce exercises over the study area, were only able to perform eight surveys for Redon, Carrelet and Palissade colonies. Due to flight authorizations (proximity of the Montpellier international civil and Nîmes military airports), we could not carry out aerial survey over the Chaumont colony foraging range. Breeding parameters and chicks measurements For the six colonies, we measured clutch size, brood size and nest success. Little Egrets nests were individually tagged at the beginning of the season (early May) and monitored weekly. The clutch size was defined as the number of eggs per nest after laying was completed. The brood size was defined as the number of chicks remaining alive at 20–25 days of age for successful nests. After 20–25 days, chicks are capable of escaping by walking and may not be present in the nest (BENNETTS et al., 2000). Nest success was estimated as the proportion of active nests (i.e. nests with at least one chick) that were successful (i.e. nests that fledged at least one chick). Measurements included tarsus length (mm), taken from the middle of the mid–tarsal joint to the distal end of the tarso–metatarsus, and body mass (g). Chicks were aged according to tarsus length (T HOMAS et al., 1999). All birds were released unharmed at the site of capture. Statistical analysis Habitat selection by Little Egrets was determined using a chi–square goodness–of–fit test to compare the observed distribution of foraging adults with that expected, relative to the proportion of each suitable habitat available within the foraging range of colonies. The body condition index (BCI) was calculated as the residuals from the model II (reduced major axis) regression of body mass (W) on the tarsus length (T) (G REEN, 2000). The use of model II regressions is likely more appropriate to study body condition since the use of residuals from ordinary least square linear regression of body mass against a linear measure of size is subject to errors due to measurements and violation of assumptions (SOKAL & ROHLF, 1997; GREEN, 2000). Clutch size and brood size were compared between colonies using one–way analyses of variance (ANOVA) and Tukey–Kramer (HSD) post-hoc tests (SOKAL & ROHLF, 1997). The effects of habitat types selected by foraging adult Little Egrets and their interactions on the clutch size and brood size of egrets for each colony were assessed using a generalised linear model (GLM) approach. Because the brood size is at least partially limited by clutch size (H AFNER et al., 2001), the clutch size (CS) was considered as an explanatory variable of brood size. Little

Egrets nest in mixed colonies with Cattle Egret (Bubulcus ibis), Squacco Heron (Ardeola ralloides), Black–crowned Night Heron (Nycticorax nycticorax) and Grey heron (Ardea cinerea) in the Camargue (TOURENQ et al., 2000). Because of possible density– dependent effects on brood size (BENNETTS et al., 2000; HAFNER et al., 2001), the total number of breeding pairs of herons (PAIR) was also considered as an explanatory variable of brood size. Since 1967, all heron colonies have been censused each year in the Camargue. The census is based on direct nest counts and counting error increases with colony size, with up to 10% under–estimation of larger colonies, while over–estimation is unlikely (TOURENQ et al., 2000). Using a generalized linear model procedure with a identity link function, we explored the effects of clutch size, total number of breeding pairs and habitat types selected by foraging breeding adults on the breeding parameters. Models with non–identifiable or non– estimable effects were ignored. Model selection was based on Akaike’s Information Criteria (AIC AKAIKE, 1973; SHIBATA, 1989) and multi–model inference (MMI; ANDERSON et al., 2000). AIC is defined as: –2ln(L) + 2np where –2ln(L) represents the deviance and np is the number of parameters estimated in the model. Models with AIC scores differing by < 2 were not considered statistically different (SAKAMOTO et al., 1986). Multi–model inference is based on the entire set of models, using AIC differences (≅i ) between the best model, i.e. with the minimum AIC, and each model and using Akaike weights ( ∆ i ; A NDERSON et al., 2000). Akaike weights are calculated as: ∆i = exp(–0.5≅i) / ΣRr=1 exp(–0.5≅i) where exp (–0.5≅ι) is the likelihood of a model i given the data for i = 1, 2,... R models.

Results Foraging habitat and habitat use The foraging habitat composition differed significantly among the five colonies aerialsurveyed (table 1). The foraging range of the Agon colony mainly consisted of pastures, dry cultivation lands (sunflowers) and rice fields. The Fiélouse colony was mainly surrounded by salt flats ("sansouïres") and pastures. The Palissade colony was located between the industrial salt pans of Salin de Giraud and the industrial zone of Fos–sur–Mer, but near the large natural marshes of they de Roustan and Palissade. The foraging range of the Redon colony mainly consisted of "sansouïres" from the Vaccarès lagoon system, rice fields, salt pans and some natural marshes. More than 50% of the Carrelet foraging habitats were natural marshes, the rest being pastures, rice fields and "sansouïres"(table 1).

111

Animal Biodiversity and Conservation 24.1 (2001)

Table 1. Composition of the foraging habitat and selection of habitats within the foraging range by Little Egrets during the reproductive seasons 1998 and 1999. Number of individuals observed (N) and number of individuals expected relative to the proportion of habitat available (Exp) are given. * Indicate the highest selected habitats in a greater proportion than their availability for a given colony. Indices indicate the preference rank of habitats by increasing order: RICE. Rice fields; DRY. Dry cultivated lands; URB. Urbanised lands; NM. Natural marshes; SAN. "Sansouïres"; § To avoid problems due to empty cells and oversmoothing the data, the constant 10-8 was added to all the cells for ✪2 computations (AGRESTI, 1990). Table 1. Composición del hábitat alimentario y selección de hábitats dentro de la gama alimentaria de la garceta común durante las estaciones reproductoras de 1998 y 1999. Se presentan el número de individuos observados (N) y el número de individuos esperados (Exp) según la proporción de hábitat disponible. * Indican los hábitats seleccionados en mayor proporción que su disponibilidad para una colonia indicada. Los índices indican en rango de preferencia de hábitats y en orden creciente: RICE. Campos de arroz; DRY. Tierras cultivadas de secano; URB. Tierras urbanizadas; NM. Marismas; SAN. "Sansouïres"; § A fin de evitar problemas debidos a celdas vacías y uniformizar los datos, se añadió la constante 10-8 a todas las celdas para los cálculos de las ✪2 (AGRESTI, 1990).

Habitat Colony Agon Palissade Fiélouse Carrelet Redon

DRY

URB

ha (%)

5,177 (21.49)

RICE

10,169 (42.22)

16 (0.07)

3,958 (16.43)

4,765 (19.78)

SAN

N (Exp)

177 (38.03)2

1 (0.44)3

0 (0)4§

1396 (229.36)1*

0 (0)5§

ha (%)

345 (2.13)

1,113 (6.87)

7,403 (45.72)

2,209 (13.64)

5,122 (31.63)

N (Exp)

20 (0.42)2

16 (1.10)3

24 (10.97)4

635 (86.61)1*

9 (2.84)5

ha (%)

3,209 (13.04)

6,739 (27.37)

16 (0.07)

2,090 (8.49)

12,564 (51.04)

N (Exp)

140 (18.25)2

124 (33.93)3

0 (0)5

ha (%)

3,489 (14.88)

4,159 (17.74)

16 (0.07) §

(124.54)1*

181 (92,.38)4

13,446 (57.34)

2,339 (9.97)

1467

N (Exp)

47 (7)3

37 (6.56)4

0 (0)5

ha (%)

3,877 (14.37)

3,617 (13.41)

3,256 (12.07)

10,121 (37.50)

9,378 (37.74)

N (Exp)

49 (7.03.)2

18 (2.41)3

0 (0)5§

1293 (484.87)1*

61 (21.97)4

The foraging habitat surrounding the Chaumont colony consisted of a few natural marshes (2,060 ha; 7.03%) dispersed in an intensive dry cultivation area (vineyards, asparagus; 12,000 ha; 41%), near the industrial salt pans of Aigues–Mortes and a highly developed tourist coastal zone (6,357 ha; 21.7%). The rest of the habitat consisted of salt flats and pine woods (8,755 ha; 29.87%). Overall, adult Egrets from aerial–surveyed colonies selected natural marshes first and agricultural marshes (rice fields) second. Foraging adult egrets were found in higher numbers than expected in natural marshes in all the colonies considered (table 1). Rice fields were used more often in relation to "sansouïres" and dry cultivated lands (mostly pastures) in the Agon, Carrelet and Palissade colonies whereas "sansouïres" were used more often than rice fields and dry cultivated lands by egrets in the Redon and Fiélouse colonies.

2619 (1501.73)1

38 (3.78)2

Breeding parameters Clutch size, brood size and nest success were checked for 123 nests in the five colonies within the Rhône delta and the Chaumont colony. The mean clutch size per colony ranged from 3.74 to 4.22 eggs per nest (Agon and Redon colonies, respectively). There was no significant difference of mean clutch size between colonies (ANOVA, F[5,122] = 1.45, P = 0.21; table 2). The mean brood size ranged from 1.12 to 2.63 chicks per nest (Carrelet and Agon colonies, respectively). A significant difference of brood size was observed between colonies (ANOVA, F [5,122] = 4.20, P = 0.001; table 2). Post–hoc tests (Tukey–Kramer HSD) indicated that Carrelet nests had a lower brood size than Agon, Redon and Fiélouse nests, but this was not significantly different from Palissade and Chaumont (table 2).

Tourenq et al.

112

Table 2. Number of nests sampled in the studied colonies and corresponding estimates of total number of pairs (Np), mean clutch size (CS ± SE), mean brood size (BS ± SE) and proportion of nests successful (NS). Same letters indicates that brood size values are not different (Tukey– Kramer HSD test, P<0.05). Tabla 2. Número de nidos muestreados en las colonias estudiadas y correspondientes estimaciones del número total de parejas (Np), tamaño medio de la nidada (CS ± desviación estándar), tamaño de las crías (BS ± desviación estándar) y porporción de nidos llenos (NS). Las mismas letras indican que los valores del tamaño de las crías no son diferentes (test de Tukey–Kramer HSD, P < 0,05).

Colony

NS a,

0.92

Agon

1,237

3.74 (±0.13)

2.63 (±0.21)

Redon

1,108

4.22 (±0.16)

2.55 (±0.26)a

0.88

3.76 (±0.16)

0.88

1.60 (±0.26)

a,b

0.65

a,b

0.90

Fiélouse Palissade

17 20

1,452 118

4.15 (±0.15)

2.41 (±0.27)

Chaumont

548

3.93 (±0.12)

2.10 (±0.30)

Carrelet

470

3.87 (±0.17)

1.12 (±0.28)b

The highest value of nest success was obtained for the Agon colony (0.92) and the lowest for the Carrelet colony (0.63; table 2). The total number of pairs of herons also varied between colonies (table 2). The Fiélouse colony (1,452 pairs) was the largest while the smallest was the Palissade colony (118 pairs). Chick body condition For both years (1998 and 1999), a total of 172 chicks aged between 6 and 21 days (mean = 12.97±0.24 SE) were measured in the five colonies within the Rhône delta (Agon, Carrelet, Fiélouse, Redon, Palissade) and the Chaumont colony. Body mass (W) was significantly correlated with tarsus length (r2 = 0.89, N = 167, P < 0.001). The body condition index (BCI) was obtained from the residuals of the model II regression: W = 7.49T–153.26 The corrected body condition index was not significantly different between colonies (ANOVA, F[5,161] = 1.25, P = 0.28). Subsequently, preliminary univariate tests showed that the mean age of chicks varied between colonies (ANOVA, F[5,166] = 8.61, P < 0.001). Post–hoc test (Tukey–Kramer HSD) indicated that Palissade chicks were older than chicks from other colonies (table 3). Relation between environmental and colonial parameters As the body condition and the clutch size (CS) were not significantly different between colonies, we concentrated our analysis on the brood size (BS) as the variable to investigate. The habitat

0.63

selected by adult breeding Little Egrets being natural marshes (NM), we considered the effect of this habitat on the brood size. Using a generalised linear model procedure, we obtained the best (i.e. lowest) scores of AIC for the model PAIR+CS+NM (AIC = 366.62; table 4). Second and third models PAIR+CS and NM+CS showed identical AIC (376.62). The examination of Akaike weights suggests that the model PAIR+CS+NM is the best. However ∆i was increased by 10 units by the subtraction of one effect (PAIR or NM). Finally, the model PAIR+CS+NM was rejected on the basis of the large number of parameters not estimated. The fact that the model PAIR+NM had a higher AIC score (382.26) and that PAIR+CS and NM+CS models had identical scores suggests that the proportion of natural mashes (NM) was not estimable in presence of the size of colony (PAIR). This is linked with the fact that these effects are confounded; i.e. each level of NM is included exclusively in a level of PAIR.

Discussion In regions where natural marshes are scarce, the proportion of rice fields available in the foraging range affects the distribution and size of colonies. In this context, rice fields provide suitable foraging habitats for tree–nesting herons (HAFNER & FASOLA, 1992; FASOLA & RUIZ, 1997). HAFNER et al. (1986) showed that in the Camargue region, rice fields were intensively used by the Little Egret during the reproductive season, especially when adults were feeding chicks. However, natural habitats (i.e. marshes) in this area still cover a great proportion of the landscape mosaic.

113

Animal Biodiversity and Conservation 24.1 (2001)

More recent studies revealed that cultivated habitats such as rice fields were avoided (in relative terms) throughout the year, whereas natural marshes were the habitat preferred by foraging Litte Egrets (LOMBARDINI et al., 2001). It was thus suggested that rice fields might be of lower value than natural marshes. In this study, aerial surveys revealed that for all monitored colonies, natural marshes were the habitat preferred (selected more than expected regard to its proportion in the landscape) by adult Little Egrets during the breeding season, even in an agricultural or industrial environment. Most natural marshes (about 40% of the Camargue surface area) are presently situated in protected areas but also in private properties devoted to waterfowl hunting (TOURENQ et al., 2000). Thus, valuable foraging habitats for the Little Egret depend on the maintenance of these natural marshes and wildfowling in the Camargue. Nevertheless, rice fields were the anthropised habitats most preferred by egrets. The importance of ricefields for waterbirds may be most pronounced during extremely dry years when natural marshes are dry. Our study included a relatively dry year (1998, with a total annual rainfall = 471.10 mm) and a relatively wet year (1999, with a total

Table 3. Number of individuals (N), mean age ± SE of Little Egret chicks from the colonies studied in 1998–1999 in the Camargue, France. Same letters indicates that ages are not different (Tukey–Kramer HSD test, P < 0.05). Tabla 3. Número de individuos (N), edad media ± desviación estándar de los pollos de las colonias de garceta común estudidas en 1998–1999 en la Camarga francesa. Las mismas letras indican que las edades no son diferentes (test de Tukey–Kramer HSD, P < 0,05).

Colony

Age (days)

Palissade

16.33 (0.66)a

Chaumont

13.10 (0.52)b

Fiélouse

13.08 (0.53)b

Redon

11.21 (0.61)b

Carrelet

12.43 (0.44)b

Agon

12.48 (0.44)b

Table 4. Modelling the influence of total number of pair of herons (PAIR), clutch size (CS), natural marshes (NM) on the brood size of Little Egrets in the Camargue: Dev. Deviance of the model; Np. Number of parameters; AIC. Akaike Information Criteria; ≅i AIC – minAIC; ∆i Akaike weight; * The model had higher number of estimates non–estimated; (1) Model including two– way interactions; (2) Model including three–way interactions. Tabla 4. Modelos de influecia del número total de parejas de garzas (PAIR), tamaño de la nidada (CS), marismas (NM) en el tamaño de las crías de garceta común en la Camarga: Dev. Desviación del modelo; Np. Número de parámetros; AIC. Criterio de información de Akaike; ≅i AIC – minAIC; ∆i Peso de Akaike; * Modelo con gran número de valores no estimados; (1) Modelo con dos tipos de interacciones; (2) Modelo con tres tipos de interacciones.

Dev

AIC

≅i

∆i

PAIR+CS+MN*

358,62

366,62

0.98

CS+MN

358,62

376,62

0.006

PAIR+CS

358,62

376,62

0.006

370,26

382,26

15,64

0.0001

Model

PAIR

370,26

382,26

15,64

0.0001

PAIR+MN

370,26

382,26

15,64

0.0001

382,52

390,52

23,9

< 0.0001

PAIR/CS/MN∆2*

(1)

352,96

392,96

26,34

< 0.0001

PAIR/CS/MN∆3*

(2)

352,96

392,96

26,34

< 0.0001

Tourenq et al.

114

annual rainfall = 719.10 mm) compared to the mean annual rainfall of 625.40 mm for the 1963– 99 period (Chauvelon, unpub. data). However, colonies surveyed in 1998 (Agon, Fiélouse and Chaumont) presented among the highest reproductive performances (see table 2). Moreover, extremely dry years have been recorded in the Camargue (e.g., 252.00 mm and 325.60 mm for 1989 and 1992, respectively; Chauvelon unpub. data). Our results are therefore applicable to a relatively dry year, but do not allow us to extend reliable inference to extreme conditions which may periodically occur. T IAINEN et al. (1989) suggested that the intensification of farming induced a decrease in the Finish Starling (Sturnus vulgaris) population as the result of a reduction in nestling fitness (i.e. growth and survival). In Little Egrets, chick body condition and reproductive success are said to be a function of food quality and abundance in the foraging range and of food quantity collected by adults (HAFNER et al., 1993). Despite the conversion of many natural marshes into rice fields and other artificial habitats in the past 30 years, a general increase in population numbers of Little Egrets has been observed in the Camargue (TOURENQ et al., 2000). The analysis of a series of reproductive parameters in this species revealed no clear stresses in habitats of little appeal to egrets. The mean clutch size and chick body condition did not differ between colonies. In birds, the number of eggs laid by the female is partly related to the body condition of the female before ovulation (e.g. DRENT & DAAN, 1980; MONAGHAN et al., 1989; CHASTEL et al., 1995). However, Little Egrets may be “income breeders” (sensu MEIJER & DRENT, 1999) like purple herons, (Ardea purpurea ; MOSER, 1986): eggs are relatively small in relation to female body weight (5%, Hafner et al., unpubl. data), thus requiring small reserves before their production (MOSER, 1986), and both sexes participate in incubation during laying (HAFNER et al., 1993). Moreover, the Little Egret is a partial migrant (TOURENQ et al., 2000). Thus, females may feed on breeding areas just before and/or during egg production. Further studies are needed to confirm this hypothesis. Brood size among birds is supposed to reflect local conditions (DRENT & DAAN, 1980) and our results suggest that brood size may vary with the proportion of natural marshes around colonies. However, we could not separate the effect of the proportion of natural marshes from the effect of the colony size. Whereas the Carrelet colony was surrounded mainly by natural marshes, it was also the colony with one of the lowest number of breeding pairs, the lowest mean brood size and the lowest nest success. The brood size of Carrelet was similar to the mean brood size of colonies with a significant amount of anthropised habitat in their foraging range (Chaumont, Palissade). One possible stressor not taken into account

in this study could be the ingestion of pesticides through the consumption of food collected in rice fields. Little Egrets are mainly insectivorous– piscivorous (T OURENQ et al., 2000) and non– negligible concentrations of organochlorine pesticides typically used in rice farming were detected in tissues of fish from the Camargue (ROCHE et al., 2000). Organochlorines were also found in the eggs of Little Egrets in colonies with a «rice environment» (BERNY et al., in press). Through the quality of the food ingested (GRASMAN et al., 1998), contaminants are known to influence reproductive parameters (NEWTON, 1986; BURGER & GOCHFELD, 1991; BERNY et al., 2001). This might also be the case for the Carrelet colony where brood size and nest success were low. The presence of industrial areas set upstream and near the mouth of the Rhône river (Fos sur Mer complex), generates exposure to heavy metal or polychlorinated byphenyls (PCBs) and may account for the presence of contaminants found in eggs (BATTY et al., 1996; BERNY et al., in press.). This is especially valid for the Palissade colony where brood size and nest success were low. Studies are in progress to confirm the impact of contaminants on egrets reproduction in the Camargue.

Acknowledgements We thank the Station Biologique de la Tour du Valat, especially H. Hafner and F. Mesléard for their support in this study and S. Befeld, A. Berceaux, G. Bertault, R. Cambag, C. Caritey, L. Dami, L. Dietrich, M. Gonzalez, V. Lemoine, A. Mora, and C. Pin for their help in collecting and compiling the data. We are indebted to B. Blohorn, Mas d’Agon, and J. C. Briffaud, Conservatoire du Littoral–La Palissade, for allowing us to collect data in their property. We thank A. Besnard, CEFE–CNRS, M. Fasola, University of Pavia (Italy), A. Green, Estación Biológica de Doñana (Spain), and P. Heurteaux for their constructive comments. This study was funded by the Station Biologique de la Tour du Valat, the Sansouïre Foundation, the MAVA Foundation, and the Centre Français du Riz. The dexterity of the pilots V. Heurteaux and J. Toutain, Aeroclub de Montpellier, was of great value.

References AEBISCHER, N. J., 1990. Assessing pesticide effects on non–target invertebrates using long–term monitoring and time–series modelling. Functionary Ecology, 4: 369–373. AGRESTI, A., 1990. Categorical data analysis. Whiley & Sons, New York. A KAÏKE , H., 1973. Information theory as an extension of the maximum likelihood principle. In: Second International Symposium on

Animal Biodiversity and Conservation 24.1 (2001)

Information Theory: 267–281 (B. N. Petrov & F. Csaki, Eds.). Akademiai Kaido, Budapest, Hungary. ANDERSON, D. R., BURNHAM, K. P. & THOMPSON, W. L., 2000. Null hypothesis testing: problems, prevalence and an alternative. Journal of Wildlife Management, 64: 912–923. BATTY, J., PAIN, D. & CAURANT, F., 1996. Metal concentrations in eels, Anguilla anguilla, from the Camargue Region of France. Biological Conservation, 76: 17–23. BENNETTS, R. E., FASOLA, M., HAFNER, H. & KAYSER, Y., 2000. Influence of environmental and density–dependent factors on reproduction of Little Egrets. Auk, 117: 637–639. BERNY, P., SADOUL, N., DOL, S., VIDEMAN, B., KAYSER, Y. & HAFNER, H. (in press). Impact of local agricultural and industrial practices on egg contamination of the Little Egret in the Rhône delta, southern France. Environmental Toxicology Chemistry. BURGER, J. & GOCHFELD, M., 1991. The Common Tern, its breeding biology and social behavior. Columbia University Press, New York. CARTER, R. V. & ABRAHAMS, M. V., 1997. Predicting the distribution of organisms among a few patches: problems with detecting departures from the ideal free distribution. Oikos, 78: 388–393. CHAMBERLAIN, D. E., WILSON, A. M., BROWNE, S. J. & VICKERY, J. A., 1999. Effects of habitat type and management of skylarks in the breeding season. Journal of Applied Ecology, 36: 856–870. CHASTEL, O., WEIMERSKIRCH, H. & JOUVENTIN, P., 1995. Boy condition and seabird reproductive performance: a study of the petrel species. Ecology, 76: 2240–2246. CHAUVELON, P., 1996. Hydologie quantitative d’une zone humide méditerranéenne aménagée: le bassin du Fumemorte en Grande Camargue, delta du Rhône. Ph. D. Thesis, Université de Montpellier II. CLAMENS , A. & I SENMANN , P., 1989. Effect of supplemental food on the breeding of Blue and Great Tits in Mediterranean habitats. Ornis Scandinavica, 20: 36–42. DOOLEY, J. L. & BOWERS, M. A., 1998. Demographic responses to habitat fragmentation: ex– perimental tests at the landscape and patch scale. Ecology, 79: 969–980. DRENT, R. H. & DAAN, S., 1980. The prudent parent: energetic adjustments in avian breeding. Ardea, 68: 225–252. FARINA, A., 1997. Landscape structure and breeding bird distribution in a sub–Mediterranean agro– ecosystem. Landscape Ecology, 12: 365–378. FASOLA, M. & RUIZ, X., 1997. Rice farming and waterbirds: integrated management in an artificial landscape. In: Farming and Birds in Europe: 210–235 (D. Pain & M. W. Pieckowski, Eds.). Academic Press, London. GRASMAN, K. A., SCANLON, P. F. & FOX, G. A., 1998. Reproductive and physiological effects of

115

environmental contaminants in fish–eating birds of the Great Lakes: a review of historical trends. Environmental Monitoring Assessment, 53: 117–145. GREEN, A. J., 2000. Mass/length residuals: measures of body condition or generators of spurious results? Ecology, 82: 1,473–1,483. HAFNER, H., BENNETTS, R. E. & KAYSER, Y., 2001, Changes in clutch size, brood size and numbers of nesting Squacco Herons (Ardeola ralloides) over a 32–year period in the Camargue, South France. Ibis, 143: 11–16. HAFNER, H., DUGAN, P. J. & BOY, V., 1986. Use of artificial and natural wetlands as feeding sites by Little Egrets (Egretta garzetta L.) in the Camargue, southern France. Colonial Water– birds, 9: 149–154. HAFNER, H., DUGAN, P. J., KERSTEN, M. & WALLACE, J. P., 1993. Flock feeding and food intake in Little Egrets (Egretta garzetta L.): the effects of food provisioning on reproductive success. Ibis, 135: 25–32. HAFNER, H. & FASOLA, M., 1992. The relationship between feeding habitat and colonial nesting Ardeidae. In: Managing Mediterranean wetlands and their birds: 194–201 (C. M. Finlayson, G. E. Hollis & T. J. Davis, Eds.). Proceedings of the Symposium of Grado, Italy, IWRB Special Publication 20, Slimbridge, UK. HEATH, M. F . & EVANS, M. I. (Eds.), 2000. Important Bird Areas In Europe: Priority Sites for Conservation. Vol. 2: Southern Europe. BirdLife International Conservation Series N°8, Cambridge, UK. LOMBARDINI, K., BENNETTS, R. E. & TOURENQ, C., 2001. Foraging success and foraging habitat use by Cattle egrets and Little Egrets in the Camargue. Condor, 103: 38–44. MEIJER, T. & DRENT, R., 1999. Re–examination of the capital and income dichotomy in breeding birds. Ibis, 141: 399–414. MONAGHAN, P., UTTLEY, J. D., BRUNS, M. D., THAINE, C. & BLACKWOOD, J., 1989. The relationship between food supply, reproductive effort and breeding success in Arctic Terns, Sterna paradisea. Journal of Animal Ecology, 58: 261–274. MOSER, M. E., 1986. Breeding strategies of purple herons in the Camargue, France. Ardea, 74: 91–100. NEWTON, I., 1986. The Sparrowhawk. T. & A. D. Poyser, Calton. PETERSEN, B. S., 1998. The distribution of Danish farmland birds in relations to habitats characteristics. Ornis Fennica, 75: 105–118. ROCHE, H., BUET, A., JONOT, O. & RAMADE, F., 2000. Organochlorine residues in european eel (Anguilla anguilla), crucian carp (Carassius carassius) and catfish (Ictalurus nebulosus) from Vaccares lagoon (French National Nature Reserve of Camargue): effects on some physiological parameters. Aquatic Toxicology, 48: 443–459. SAKAMOTO, Y., ISHIGURO, M. & KITAWAGA, G., 1986.

116

Akaike information criterion statistics. KTK Scientific Publishers, Tokyo, Japan. SANDOZ, A. & CHAUVELON, P., 1998. La géomatique comme outil de suivi des conditions hydrologiques de la Camargue, zone deltaïque. Méditerranée, 4: 73. SHIBATA, R., 1989. Statistical aspects of model selection. In: From data to model: 215–240 (J. C. Willems, Ed.). Springer Verlag, London, UK. SOKAL, R. R. & ROHLF, F. J., 1997. Biometry. Third Edition. Freeman and Co., New York. STEPHEN, D. W. & KREBS, J. R., 1986. Foraging theory. Princeton University Press, UK. THOMAS, F., KAYSER, Y. & HAFNER, H., 1999. Nestling size rank in the Little Egret (Egretta garzetta) influences subsequent breeding success of offspring. Behavioural Ecology and Sociobiology, 45: 466–470. TIAINEN, J. T., HANSKI, I. K., PKKALA, T., PIIROINEN, J. & YRJÖLÄ, R., 1989. Clutch size, nestling growth and nestling mortality of the Staling Sturnus

Tourenq et al.

vulgaris in south Finnish agroenvironments. Ornis Fennica, 66: 41–48. TOURENQ, C., BENNETTS, R. E., KOWALSKI, H., VIALET, E., LUCCHESI, J.–L., KAYSER, Y. & ISENMANN, P. , 2001. Are ricefields a good alternative to natural marshes for waterbird communities of the Camargue, southern France? Biological Conservation, 100: 335–343. TOURENQ, C., BENNETTS, R. E., SADOUL, N., MESLEARD, F., KAYSER, Y. & HAFNER, H., 2000. Long-term Population and Colony Patterns of Four Species of Tree–nesting Herons in the Camargue, South France. Waterbirds, 23: 236–245. T UCKER , G. M., 1992. Effects of agricultural practices on field use by invertebrate–feeding birds in winter. Journal of Applied Ecology, 29: 779–790. WILSON, J. D., TAYLOR, R. & MUIRHEAD, L. B., 1996. Field use by farmland birds in winter: an analysis of field type preferences using resampling methods. Bird Study, 43: 320–332.

117

Animal Biodiversity and Conservation 24.1 (2001)

Spatial relations of American bison (Bison bison) and domestic cattle in a montane environment D. H. Van Vuren

Van Vuren, D. H., 2001. Spatial relations of American bison Bison bison and domestic cattle in a montane environment. Animal Biodiversity and Conservation, 24.1: 117–124. Abstract Spatial relations of American bison Bison bison and domestic cattle in a montane environment.— Restoration of American bison (Bison bison) to montane environments where they once occurred requires information on ecological relationships with domestic cattle (Bos taurus) that now live there. Comparisons of the foraging distributions of sympatric bison and cattle in a 375–ha basin revealed that cattle were constrained by slope and distance from water, especially vertical distance, whereas bison responded mostly to forage availability. Cattle appeared to be central place foragers oriented around water and followed a strategy of meeting their energetic needs with the least cost. Bison, in contrast, appeared to be energy maximizers that moved often in response to forage availability. The result was relatively little overlap (29%) in spatial distributions. If bison replace cattle in montane environments, managers can expect a more even distribution of grazing pressure. Bison and cattle might be managed sympatrically; their spatial distributions may be sufficiently different to minimize competition for food, and the risk of interspecific disease transmission as well. Key words: Bison, Cattle, Foraging Ecology, Montane environments, Spatial relations. Resumen Relaciones espaciales entre el bisonte americano Bison bison y el ganado vacuno en un medio de montaña.— La reintroducción del bisonte americano (Bison bison) en un ambiente de montaña donde ya había vivido antes requiere información acerca de las relaciones ecológicas con el ganado vacuno (Bos taurus) que ahora habita en ese lugar. La comparación de las distribuciones de forrajeo del bisonte con las de la vaca en una cuenca de 375 ha demostraron que la vaca estaba limitada por la inclinación del terreno y la distancia al agua, especialmente la distancia vertical, mientras que el bisonte lo estaba principalmente por la disponibilidad de pasto. La vaca mostró clara orientación a pacer principalmente alrededor del agua y siguió una estrategia de obtención de sus necesidades energéticas con el mínimo coste. En contraste, el bisonte se mostró maximizador de energía, efectuando frecuentes desplazamientos en función de la disponibilidad de pastos. El resultado dio una coincidencia relativamente reducida (29%) en las distribuciones espaciales. Si el bisonte sustituye al ganado en medios de montaña, puede esperarse una mejor distribución de la presión sobre los pastos. Los bisontes y el ganado vacuno pueden convivir en la misma área geográfica, puesto que sus distribuciones espaciales son suficientemente diferentes para minimizar la competencia por el alimento, así como el riesgo de transmisión interespecífica de enfermedades. Palabras clave: Bisonte, Ganado, Ecología de forrajeo, Medios de montaña, Relaciones espaciales. (Received: 29 VI 01; Final acceptance: 18 VII 01) D. H. Van Vuren, Dept. of Wildlife, Fish, and Conservation Biology, Univ. of California, Davis, CA 95616, USA.

ISSN: 1578–665X

Van Vuren

118

Introduction Most grasslands around the world once supported vast herds of migratory ungulates (FRYXELL et al., 1988; FRANK, et al., 1998). Over the past 150 years, however, these herds have been drastically reduced by overhunting, and much of their grassland habitat has been cultivated for crops or converted to pasture for domestic livestock (FRYXELL & SINCLAIR, 1988; FRANK et al., 1998). Conservation of remaining herds is threatened by continued encroachment by an increasing human population and conflicts with domestic livestock (REYNOLDS & HAWLEY, 1987; NORTON– GRIFFITHS, 1995; FRANK et al., 1998). The American bison, a large migratory ungulate that was once distributed throughout much of North America, is no exception. Densities were greatest in the grasslands of the Great Plains, which supported numbers in the tens of millions (ROE, 1970), but bison also lived at lower densities in meadows and shrub–steppe communities to the west, in the Rocky Mountains and in the mountains and valleys beyond (ROE, 1970; VAN V UREN , 1987; M EANEY & V AN V UREN , 1993). Indiscriminate slaughter during the 1800s led to near–extinction; by 1900, only a few hundred bison remained, almost all of them in captivity. Intensive conservation efforts narrowly averted extinction, and numbers have recovered to about 200,000. Most bison today, however, are intensively managed on private lands for commercial purposes or are confined by fences on wildlife refuges (DARY, 1989; MANNING, 1996). The extensive range formerly inhabited by bison has been plowed and converted to crops or is grazed by domestic cattle. Consequently, bison have recovered from near– extinction, but the ecological role they once played has not been restored. Recent research has demonstrated the ecological importance of bison in a variety of the biotic communities that they once inhabited (KRUEGER, 1986; F RANK & MC NAUGHTON, 1993; CAMPBELL et al., 1994; KNAPP et al., 1999; STEINAUER & COLLINS, 2001). Consequently, there has been increasing interest in restoring bison to a functional role in natural areas, both in the grasslands of the Great Plains and in montane environments to the west (PLUMB & DODD, 1993; WUERTHNER, 1993; CALLENBACH, 1996; HAMILTON, 1996; STEPHENSON & FLEENER, 1998; KNAPP et al., 1999). However, domestic cattle, often considered the ecological equivalent of bison (N OSS & C OOPERRIDER , 1994; H A RTNETT et al., 1997; WUERTHNER, 1998), now occupy many of these areas, raising concerns about the consequences of restoring bison. If bison and cattle have similar niches and occur in the same area, then competition for food may result in management conflicts (WAGNER, 1978). Further, the possibility of disease transmission between sympatric bison and cattle has caused controversy at several localities (VAN VUREN & SCOTT, 1995).

Comparisons of bison and cattle in the Great Plains have revealed interspecific differences in foraging ecology (P EDEN et al., 1974; P LUMB & D ODD , 1993; H ARTNETT et al., 1997), but corresponding studies have not been done in montane environments, where abiotic factors may be particularly important in influencing foraging distribution. Cattle distribution is strongly constrained by slope and distance from water (M UEGGLER , 1955; R OATH & K RUEGER, 1982; G ILLEN et al., 1984; G ANSKOPP & V AVRA , 1987; P INCHAK et al., 1991; T ELFER , 1994), and elevation may have an effect as well (S ENFT et al., 1983). In contrast, several authors have claimed that bison are relatively unaffected by these factors (W ARREN , 1927; F RYXELL, 1928; C AHALANE , 1947; C ALLENBACH , 1996), suggesting t h e p o t e n t i a l f o r n i c h e d i ff e r e n t i a t i o n between bison and cattle. These claims, however, remain unsubstantiated. The Henry Mountains, a semi–arid range west of the Rocky Mountains, support free–ranging bison and cattle that co–exist in an area characterized by rugged topography and limited water. Although located about 100 km outside of the former range of bison (Van Vuren, unpublished data), the Henry Mountains are typical of montane environments where bison are known to have occurred (MEANEY & VAN VUREN, 1993). Foraging distributions of bison and cattle were studied in relation to slope, distance from water, and elevation. A preliminary analysis suggested that bison and cattle responded differently to these factors (VAN VUREN, 1982). In the present report those preliminary findings are confirmed and extended by showing that differential responses to environmental factors, perhaps stemming from differing evolutionary histories, result in spatial segregation of bison and cattle, with implications for conservation.

Study area L ocated in southeastern Utah, USA, the Henry Mountains (38o 5’ N, 110o 50’ W) rise abruptly above the Colorado Plateau (ca. 1,500 m elevation) to 3,540 m elevation at the summit of Mount Ellen, the highest peak. Precipitation increases with elevation, ranging from 15 cm in the surrounding deserts of the plateau to > 50 cm on the higher slopes. The lower slopes of the range (ca. 1,800–2,400 m elevation) support extensive pinyon pine (Pinus edulis) and juniper (Juniperus spp.) woodlands. Slopes above 2,400 m are an interspersion of Douglas fir (Pseudotsuga menziesii), spruce (Picea engelmannnii), and fir (Abies spp.) forests, groves of quaking aspen ( Populus tremuloides ), and shrub–steppe openings dominated by shrubs (Artemisia spp., Symphoricarpos alba), forbs (Penstemon spp., Oxytropis spp., Astragalus spp.) and perennial grasses (Poa spp., Festuca spp., Nassella spp.).

119

Animal Biodiversity and Conservation 24.1 (2001)

The highest elevations support subalpine grasslands dominated by Festuca thurberi and Carex spp. Bison were introduced to the area in 1941 as part of early efforts to restore bison and numbered about 200 at the time of this study (VAN VUREN & BRAY, 1986). The bison were free-ranging and migrated seasonally, spending summers on the upper slopes of the mountains and moving to lower elevations, mostly to the west and southwest, during winter. Domestic cattle have grazed the Henry Mountains since the late 1800s; the two species were sympatric during summer but allopatric during winter, when cattle were herded to ranges apart from bison. The study area comprised the 375–ha basin that formed the headwaters of South Creek, which drained the west side of Mount Ellen. Elevation ranged from 2,800 m on the west side of the basin to 3,400 m at the head. Slopes were gentle in much of the central and western portion of the basin but increased steadily toward the east, reaching 40o at the head of the basin. Vegetation in the lower portion of the basin consisted of conifer or aspen groves interspersed with shrub-steppe openings, the middle portion was mostly shrub–steppe, and the upper portion supported subalpine grassland. Both bison and cattle frequently grazed the basin during the study; about 30 cattle lived there all summer, and groups of up to 60 bison were often present. Drinking water was available all summer at seven springs and catchment basins. There were no fences, and cattle were not herded after being moved into the basin during late spring; thus, the entire basin was physically accessible to both species.

Methods The study was conducted July and August 1977 and June through September 1978. Spatial distributions were determined visually; bison and cattle were observed and classified as foraging or not, and their locations were plotted at 30–minute intervals on a 1:24,000 topographic map of the basin. The map was overlaid with a grid scaled at 100–m intervals (thus, 1 ha per grid cell), and each observation was assigned to the grid cell that it fell within. Bison and cattle were observed from several vantage points, none of which allowed a view of the entire basin, so vantage points were rotated among to ensure that all parts of the basin were surveyed. Observations were distributed throughout daylight hours. The numbers of observations of foraging bison and cattle were summed for each grid cell. Environmental factors were measured at the center of each grid cell. A clinometer was used to measure slope and a topographic map was used to determine elevation and distance from the nearest source of drinking water; horizontal

and vertical distance from water were recorded as separate factors. Frequency distributions were used to compare the spatial distributions of bison and cattle in relation to each environmental factor. The range of each factor was divided into intervals and all observations of bison or cattle that fell within each interval for slope (4o intervals), horizontal (100–m intervals) and vertical (30–m intervals) distance from water, and elevation (50–m intervals) were totaled. Frequency distributions of bison and cattle were compared using a G–test of independence. Observations were not independent, but this is a concern only if analyses yield marginally significant results, which did not happen. To compare the spatial distributions of bison and cattle graphically a three–dimensional mesh plot was used, in which x– and y– coordinates corresponded to the axes of the grid that overlay the basin, and the z–coordinate represented the frequency of use of each grid cell. To compare spatial distributions numerically, Kulczynski’s similarity index was calculated (OOSTING, 1956), which compares frequencies in each grid square, then sums for all grid squares. The index ranges from 0 (completely different use of space) to 1 (identical use of space). The foraging distribution of cattle often declines exponentially with increasing distance from water (MUEGGLER, 1955; ROATH & KRUEGER, 1982; NASH et al., 1999), creating a “piosphere”, a zone of attenuating impact away from each watering point (A NDREW , 1988). A similar exponential decline may exist with increasing slope (GANSKOPP & VAVRA, 1987). This relationship was evaluated for bison and for cattle by regressing the square root of proportion of observations on each of three variables, horizontal distance from water, vertical distance from water, and slope. Because only a negative relationship was expected, one–tailed tests were used.

Results Among > 22,000 observations of bison and >3,000 of cattle recorded in the study area, 9745 were of foraging bison and 1,196 were of foraging cattle. The species differed markedly in the slopes they grazed (G = 1,264.0, P < 0.001); cattle occurrence declined rapidly as slope increased beyond 4o, whereas bison exhibited a bimodal pattern with the highest peak in occurrence at 28–32o (fig. 1). Cattle observations fit a negative exponential relationship with slope (r2 = 0.86, P < 0.001), but bison observations did not (r2 = 0.03, P > 0.50). The species were less differentiated according to horizontal distance from water, yet differences were evident (G = 492.0, P < 0.001). Cattle occurrence declined with increasing distance, while bison occurrence was unrelated to distance until beyond 700 m. Cattle observations fit a

120

negative exponential relationship with horizontal distance from water (r2 = 0.88, P < 0.001); so did bison observations, but the fit was poor (r2 = 0.33, P = 0.04) and the slope was less negative than that of cattle (t = 2.53, P < 0.05). The species differed strikingly according to vertical distance from water (G = 1,260.7, P < 0.001). Cattle occurrence declined steeply with increasing distance, whereas bison occurrence decreased only slightly. Cattle observations fit a negative exponential relationship with vertical distance from water (r2 = 0.78, P < 0.001); the decline was so steep, however, that examination of the residuals indicated a fourth–root transformation was a better representation of the relationship (r 2 = 0.83, P < 0.001). Bison observations also fit a negative exponential relationship (r2 = 0.72, P < 0.001), but the slope was less negative than that of cattle (t = 2.75, P < 0.02). Bison and cattle differed in the elevations they grazed (G = 1,354.0, P < 0.001), with bison grazing more often at higher elevations. The spatial distributions of bison and cattle in the basin were not uniform, in part because both species grazed almost entirely in shrubsteppe or subalpine grassland communities and used conifer or aspen groves mostly for resting. However, within shrub–steppe and grassland communities, differential response of bison and cattle to slope, distance from water, and elevation translated into differential use of space. Bison and cattle were recorded in similar numbers of grid squares (163 and 16, respectively), but only 66 grid squares were grazed by both species. Intensity of use differed as well. Two pronounced peaks in cattle distribution, in the north and northwest portions of the basin, occurred at large “flats” with level or gentle slopes and drinking water nearby (fig. 2). Most of the lesser peaks in cattle occurrence were located at smaller flats with adjacent water. Some bison also grazed these areas, but most bison observations were distributed in an arc that extended across the highest (and steepest) slopes at the head of the basin and along the north slope of the ridge that formed the southern boundary of the basin (fig. 2). Graphical differences in distribution were supported numerically; the similarity index was only 0.286, indicating that spatial distributions of foraging bison and cattle were largely dissimilar.

Discussion Distribution of cattle was strongly constrained by slope, although not as severely as in other studies, which reported that cattle seldom used slopes greater than 11o (GILLEN et al., 1984; GANSKOPP & VAVRA, 1987; PINCHAK et al., 1991; TELFER, 1994). Bison, in contrast, frequented much steeper slopes than did cattle, both in the Henry

Van Vuren

Mountains and elsewhere, exhibiting a peak in distribution at about 30o and confirming earlier assertions (FRYXELL, 1928). Horizontal distance from water seemed less important to cattle distribution than did slope, probably because all parts of the basin were relatively close to water (< 1000 m) and because response of cattle to proximity of water is influenced by other factors, especially topography (HOLECHEK et al., 1989). Nonetheless, bison were relatively unaffected by availability of water compared with cattle and in particular were less likely to graze close to water, again confirming previous claims (CAHALANE, 1947; CALLENBACH, 1996). In contrast to horizontal distance from water, vertical distance from water sharply constrained cattle distribution. These findings parallel those of ROATH & KRUEGER (1982), who found that cattle rarely grazed sites more than 80 m above water. Bison were much less affected than cattle by vertical distance from water. Bison grazed at higher elevations than cattle; however, they probably were responding to forage availability rather than elevation. Precipitation in the Henry Mountains increases linearly with elevation (VAN VUREN & BRAY, 1986), and so does forage availability; production of graminoids, the primary food of both bison and cattle (VAN VUREN, 1984), ranged 31–179 kg/ha (dry basis) in the study area and increased with elevation (r = 0.85, P < 0.05; Van Vuren, unpublished data). This gradient may have been exacerbated by depletion of forage at lower elevations by cattle concentrated on gentle slopes near water (ANDREW, 1988; HOLECHEK et al., 1989). These findings agree with those from the Great Plains, where availability of graminoids was a more important factor for bison than cattle in determining foraging distribution (PLUMB & DODD, 1993). Thus, these results suggest that slope and distance from water, especially vertical distance, are most important in influencing cattle distribution, whereas availability of forage is more important for bison. The result was little spatial overlap between the species. An alternative explanation, that spatial segregation resulted from behavioral avoidance, is unlikely; the species sometimes grazed close to each other, and neither species altered its behavior in response to the other until about 4 m apart, whereupon cattle always avoided bison (VAN VUREN, 1980). Bison and cattle are closely related and are generally similar in size, appearance, and food preference (WUERTHNER, 1998), consequently their differing spatial distributions are somewhat surprising. Perhaps the explanation lies in their differing evolutionary histories. Cattle originated in mesic environments of Eurasia whereas bison evolved in the semi–arid Great Plains, consequently cattle may have a greater requirement for water (NOSS & COOPERRIDER, 1994; WUERTHNER,

121

Animal Biodiversity and Conservation 24.1 (2001)

Cattle Bison A

0.4

0.3

0.3 0.2 0.2 0.1 0.1

0.0

6 10

14 18 22 26 30

34 38

0.0

50 150 250 350 450 550 650 750 850 950

D 0.3

0.5 0.4

0.2

0.3 0.2

0.1

0.1 0.0

30 60

90 120 150 180 210 240 270

0.0

2,800

2,900

3,000

3,100

3,200

Fig. 1. Proportion of observations of foraging bison and cattle in relation to: A. Slope, in degrees; B. Horizontal distance from water, in m; C. Vertical distance from water, in m; D. Elevation, in m, at the headwaters of South Creek, Henry Mountains, Utah. Fig. 1. Proporción de observaciones de forrajeo del bisonte y del ganado vacuno en relación con: A. Inclinación del terreno, en grados; B. Distancia horizontal al agua, en m; C. Distancia vertical al agua, en m; D. Altura, en m, de la cabecera del río South Creek, Henry Mountains, Utah.

1998). Moreover, cattle have undergone intense artificial selection for traits that maximize productivity. Maximizing fat storage for overwinter survival probably is less important because cattle are herded to pastures known to provide sufficient forage, or they are supplementally provisioned. Bison, in contrast, face the risk of starvation during harsh winters (MEAGHER, 1986), thus they exhibit adaptations for overwinter survival superior to those of cattle (HAWLEY, 1987; PLUMB & DODD, 1993). Cattle distribution was most constrained by slope and vertical distance from water. Both involve movement in a vertical plane, which is roughly 10 times more expensive energetically than horizontal movement (CLAPPERTON, 1964;

BROCKWAY & GESSAMAN, 1977; PARKER et al., 1984), suggesting that cattle were minimizing their foraging costs. Increased travel costs result in reduced productivity (HOLECHEK et al., 1989). Consequently, cattle appeared to be following a strategy of meeting their energetic needs with least overall cost, a strategy reported for kudu (Tragelaphus strepsiceros; OWEN–SMITH, 1994). The result was that cattle behaved as central place foragers, with grazing activity centered on a water source or perhaps on thermal cover, but not on the feeding site (ROATH & KRUEGER, 1982; STUTH, 1991; GUTHERY, 1996). Forage is depleted on gentle slopes near water, but cattle foraging there evidently can meet their energetic needs with minimal energy expenditure. Bison, in

Van Vuren

122

Cattle

Bison

W es

h ut o S

Fig. 2. Three–dimensional mesh plots of the foraging distributions of bison and cattle at the headwaters of South Creek, Henry Mountains, Utah. The z–coordinate indicates frequency of occurrence in each grid square. Slope and elevation increase from west to east, and water sources are denoted by black circles. Fig. 2. Gráfico en malla tridimensional de la distribución de pasto del bisonte y del ganado vacuno en las cabeceras del río South Creek, Henry Mountains, Utah. La coordenada z indica la frecuencia de forrajeo en cada cuadrícula. La inclinación y la altura aumentan de oeste a este, las fuentes de agua se indican con círculos negros.

contrast, behaved as energy maximizers, willing to expend energy to obtain richer rewards, with grazing activity oriented on the feeding site rather than on a water source. Bison typically rested within or adjacent to the feeding site, traveled once per day to water and drank briefly

(x = 21 minutes), then returned to the feeding site or moved to a new one (VAN VUREN, 1980). Bison moved much more than cattle; most cattle remained within the 375–ha basin throughout the summer, whereas bison roamed about home ranges that averaged 5,220 ha (VAN VUREN, 1983),

Animal Biodiversity and Conservation 24.1 (2001)

rarely remaining in the same area longer than 3 days (VAN VUREN, 1980). Spatial segregation of sympatric bison and cattle on montane ranges during summer, resulting from differential response to environmental factors, has important implications for conservation. If bison replace cattle, managers can expect a more even distribution of grazing pressure in response to forage availability, with little evidence of a piosphere, instead of a clumped distribution in response to slope and distance from water. Bison and cattle might even be managed sympatrically; although their diets are generally similar, their spatial distributions may be sufficiently different not only to minimize competition for food, but also to reduce the risk of transmission of diseases requiring close spatial proximity.

Acknowledgments I thank the U. S. Bureau of Land Management, Utah Division of Wildlife Resources, Mzuri Safari Foundation, and Sigma Xi for financial and logistical support; Marty Bray, Bruce Coblentz, and Golden and Keith Durfey for assistance throughout the study; and Vickie Bakker, Chris Floyd and Rob Klinger for helpful comments on the manuscript.

References ANDREW, M. H., 1988. Grazing impact in relation to livestock watering points. Trends Ecol. & Evol., 3: 336–339. BROCKWAY, J. M. & GESSAMAN, J. A., 1977. The energy cost of locomotion on the level and on gradients for the red deer (Cervus elaphus). Quart. J. Exp. Physiol., 62: 333–339. C AHALANE , V. H., 1947. Mammals of North America. MacMillan Company, New York. CALLENBACH, E., 1996. Bring back the buffalo! Island Press, Washington, DC. CAMPBELL, C., BLYTH, C. B. & MCANDREWS, J. H., 1994. Bison extirpation may have caused aspen expansion in western Canada. Ecography, 17: 360–362. CLAPPERTON, J. L., 1964. The energy metabolism of sheep walking on the level and on gradients. Brit. J. Nutr., 18: 47–54. DARY, D. A., 1989. The buffalo book. Swallow Press/Ohio University Press, Athens, Ohio. FRANK, D. A. & MCNAUGHTON, S. J., 1993. Evidence for the promotion of aboveground grassland production by native large herbivores in Yellowstone National Park. Oecologia, 96: 157–161. FRANK, D. A., MCNAUGHTON, S. J. & TRACY, B. F., 1998. The ecology of the Earth’s grazing ecosystems. Bioscience, 48: 513–521. FRYXELL, F. M., 1928. The former range of the

123

bison in the Rocky Mountains. J. Mammal., 9: 129–139. FRYXELL, J. M., GREEVER, J. & SINCLAIR, A. R. E., 1988. Why are migratory ungulates so abundant? Am. Nat., 131: 781–798. FRYXELL, J. M. & SINCLAIR, A. R. E., 1988. Causes and consequences of migration by large herbivores. Trends Ecol. & Evol., 3: 237–241. GANSKOPP, D. & VAVRA, M., 1987. Slope use by cattle, feral horses, deer, and bighorn sheep. Northwest Sci., 61: 74–81. GILLEN, R. L., KRUEGER, W. C. & MILLER, R. F., 1984. Cattle distribution on mountain rangeland in northwestern Oregon. J. Range Manage., 37: 549–553. G UTHERY , F. S., 1996. Upland gamebirds. In: Rangeland wildlife: 59–69 (P. R. Krausman, Ed.). Society for Range Management, Denver. HAMILTON, R. G., 1996. Using fire and bison to restore a functional tallgrass prairie landscape. Trans. N. Am. Wildl. & Nat. Resour. Conf., 61: 208–214. HARTNETT, D. C., STEUTER, A. A. & HICKMAN, K. R., 1997. Comparative ecology of native and introduced ungulates. In: Ecology and conservation of Great Plains vertebrates: 72– 101 (F. L. Knopf & F. B. Samson, Eds.). Springer– Verlag, New York. HAWLEY, A. W. L., 1987. Bison and cattle use of forages. In: Bison ecology in relation to agricultural development in the Slave River lowlands, NWT. Occasional Paper 63,: 49–52 (H. W. Reynolds & A. W. L. Hawley, Eds.). Canadian Wildlife Service, Edmonton. HOLECHEK, J. L., PIEPER, R. D. & HERBEL, C. H., 1989. Range management . Prentice Hall, Englewood Cliffs. K NAPP, A. K., B LAIR, J. M., BRIGGS, J. M., COLLINS, S. L., H ARTNETT , D. C., JOHNSON, L. C. & TOWNE, E. G., 1999. The keystone role of bison in North American tallgrass prairie. Bioscience, 49: 39–50. KRUEGER, K., 1986. Feeding relationships among bison, pronghorn, and prairie dogs: an experimental analysis. Ecology, 67: 760–770. MANNING, R., 1996. The buffalo is coming back. Defenders, 71(1): 6–15. MEAGHER, M. M., 1986. Bison bison. Mammal. Spec., 266: 1–8. MEANEY, C. A. & VAN VUREN, D., 1993. Recent distribution of bison in Colorado west of the Great Plains. Proc. Denver Mus. Nat. Hist., 3(4): 1–10. MUEGGLER, W. F., 1955. Cattle distribution on steep slopes. J. Range Manage., 18: 255–257. NASH, M. S, WHITFORD, W. G., DE SOYZA, A. G., VAN ZEE, J. W. & HAVSTAD, K. M., 1999. Livestock activity and Chihuahuan Desert annual–plant communities: boundary analysis of disturbance gradients. Ecol. Applic., 9: 814–823. NOSS, R. F. & COOPERRIDER, A. Y., 1994. Saving nature’s legacy. Island Press, Washington, DC. NORTON–GRIFFITHS, M., 1995. Economic incentives to develop the rangelands of the Serengeti:

124

implications for wildlife conservation. In: Serengeti II: dynamics, management and conservation of an ecosystem: 588–604 (A. R. E. Sinclair & P. Arcese, Eds.). University of Chicago Press, Chicago. OOSTING, H. J., 1956. The study of plant communities, 2nd ed. W. H. Freeman, San Francisco. OWEN–SMITH, O., 1994. Foraging responses of kudus to seasonal changes in food resources: elasticity in constraints. Ecology, 75: 1,050–1,062. PARKER, K. L., ROBBINS, C. T. & HANLEY, T. A., 1984. Energy expenditures for locomotion by mule deer and elk. J. Wildl. Manage., 48: 474–488. PEDEN, D. G., VAN DYNE, G. M., RICE, R. W. & HANSEN, R. M., 1974. The trophic ecology of Bison bison L. on shortgrass plains. J. Appl. Ecol., 11: 489–497. PINCHAK, W. E., SMITH, M. A., HART , R. H. & W AGGONER , J. W. J R ., 1991. Beef cattle distribution patterns on foothill range. J. Range Manage., 44: 267–275. PLUMB, G. E. & DODD, J. L., 1993. Foraging ecology of bison and cattle on a mixed prairie: implications for natural area management. Ecol. Applic., 3: 631–643. REYNOLDS , H. W. & H AWLEY, A. W. L., 1987. Introduction. In: Bison ecology in relation to agricultural development in the Slave River lowlands, NWT, Occasional Paper 63: 10–12 (H. W. Reynolds & A. W. L. Hawley, Eds.). Canadian Wildlife Service, Edmonton. ROATH, L. R. & KRUEGER, W. C., 1982. Cattle grazing and behavior on a forested range. J. Range Manage., 35: 332–338. ROE, F. G., 1970. The North American buffalo, 2nd ed. University of Toronto Press, Toronto. SENFT, R. L., RITTENHOUSE, L. R. & WOODMANSEE, R. G., 1983. The use of regression models to predict spatial patterns of cattle behavior. J. Range Manage., 36: 553–557. STEINAUER, E. M. & COLLINS, S. L., 2001. Feedback loops in ecological hierarchies following urine deposition in tallgrass prairie. Ecology, 82: 1,319–1,329. STEPHENSON, R. O. & FLEENER, C. L., 1998. Biopolitical perspectives on a proposed wood bison reintroduction to Alaska. In: International

Van Vuren

symposium on bison ecology and management in North America: 251–257 (L. Irby & J. Knight, Eds.). Montana State University, Bozeman. STUTH, J. W., 1991. Foraging behavior. In: Grazing management: 65–83 (R. K. Heitschmidt & J. W. Stuth, Eds.). Timber Press, Portland. TELFER, E. S., 1994. Cattle and cervid interactions on a foothills watershed in southwestern Alberta. Can. Field–Nat., 108: 186–194. VAN VUREN, D., 1980. Ecology and behavior of bison in the Henry Mountains, Utah. M. S. Thesis, Oregon State University. – 1982. Comparative ecology of bison and cattle in the Henry Mountains, Utah. In: Proceedings of the wildlife–livestock relationships symposium: 449–457 (J. M. Peek & P. D. Dalke, Eds.). University of Idaho Forestry, Wildlife & Range Experiment Station, Moscow, Idaho. – 1983. Group dynamics and summer home range of bison in southern Utah. J. Mammal., 64: 329–332. – 1984. Summer diets of bison and cattle in southern Utah. J. Range Manage., 37: 260–261. – 1987. Bison west of the Rocky Mountains: an alternative explanation. Northwest Sci., 61: 65–69. VAN VUREN, D. & BRAY, M. P., 1986. Population dynamics of bison in the Henry Mountains, Utah. J. Mammal., 67: 503–511. VAN V UREN, D. & SCOTT , C. A., 1995. Internal parasites of sympatric bison, Bison bison, and cattle, Bos taurus. Can. Field–Nat., 109: 467–469. WAGNER, F. H., 1978. Livestock grazing and the livestock industry. In: Wildlife and America: 121–145 (H. P. Brokaw, Ed.). Council on Environmental Quality, Washington, DC. WARREN, E. R., 1927. Altitude limit of bison. J. Mammal., 8: 60–61. W UERTHNER , G., 1993. Envisioning wildland restoration. Wild Earth, 3(3): 70–74. – 1998. Are cows just domestic bison? Behavioral and habitat use differences between cattle and bison. In: International symposium on bison ecology and management in North America: 374–383 (L. Irby & J. Knight, Eds.). Montana State University, Bozeman.

Animal Biodiversity and Conservation 24.1 (2001)

Animal Biodiversity and Conservation Animal Biodiversity and Conservation (abans Miscel·lània Zoològica) és una revista interdisciplinària publicada, des de 1958, pel Museu de Zoologia de Barcelona. Inclou articles d'investigació empírica i teòrica en totes les àrees de la zoologia (sistemàtica, taxonomia, morfologia, biogeografia, ecologia, etologia, fisiologia i genètica) procedents de totes les regions del món, amb especial énfasis als estudis que d'una manera o altre tinguin relevància en la biología de la conservació. La revista no publica catàlegs, llistes d'espècies o cites puntuals. Els estudis realitzats amb espècies rares o protegides poden no ser acceptats tret que els autors disposin dels permisos corresponents. Cada volum anual consta de dos fascicles. Animal Biodiversity and Conservation es troba registrada en la majoria de les bases de dades més importants, de manera que permet una difusió mundial dels seus articles. Tots els manuscrits són revisats per l'editor executiu, un editor i dos revisors independents, triats d'una llista internacional, a fi de garantir-ne la qualitat. El procés de revisió és ràpid i constructiu. La publicació dels treballs acceptats es fa normalment dintre dels 12 mesos posteriors a la recepció. Una vegada hagin estat acceptats passaran a ser propietat de la revista. Aquesta es reserva els drets d’autor, i cap part dels treballs no podrà ser reproduïda sense citar–ne la procedència.

Normes de publicació Els treballs s'enviaran preferentment de forma electrònica (mzbpubli@intercom.es). El format preferit és un document Rich Text Format (RTF) o DOC que inclogui les figures (TIF). Si s'opta per la versió impresa, s'han d'enviar quatre còpies del treball juntament amb una còpia en disquet a la Secretaria de Redacció. Cal incloure, juntament amb l'article, una carta on es faci constar que el treball està basat en investigacions originals no publicades anteriorment i que està sotmès a Animal Biodiversity and Conservation en exclusiva. A la carta també ha de constar, per a aquells treballs en que calgui manipular animals, que els autors disposen dels permisos necessaris i que compleixen la normativa de protecció animal vigent. També es poden suggerir possibles assessors. Quan l'article sigui acceptat, els autors hauran d'enviar a la Redacció una còpia impresa de la versió final acompanyada d'un disquet indicant el programa utilitzat (preferiblement en Word). Les proves d'impremta enviades a l'autor per a la correcció, seran retornades al Consell Editor en el termini de 10 dies. Aniran a càrrec dels autors les despeses degudes a modificacions substancials introduïdes per ells en el text original acceptat. ISSN: 1578–665X

El primer autor rebrà 50 separates del treball sense càrrec a més d'una separata electrònica en format PDF. Manuscrits Els treballs seran presentats en format DIN A–4 (30 línies de 70 espais cada una) a doble espai i amb totes les pàgines numerades. Els manuscrits han de ser complets, amb taules i figures. No s'han d'enviar les figures originals fins que l'article no hagi estat acceptat. El text es podrà redactar en anglès, castellà o català. Se suggereix als autors que enviïn els seus treballs en anglès. La revista els ofereix, sense cap càrrec, un servei de correcció per part d'una persona especialitzada en revistes científiques. En tots els casos, els textos hauran de ser redactats correctament i amb un llenguatge clar i concís. La redacció del text serà impersonal, i s'evitarà sempre la primera persona. Els caràcters cursius s’empraran per als noms científics de gèneres i d’espècies i per als neologismes intraduïbles; les cites textuals, independentment de la llengua, seran consignades en lletra rodona i entre cometes i els noms d’autor que segueixin un tàxon aniran en rodona. Quan se citi una espècie per primera vegada en el text, es ressenyarà, sempre que sigui possible, el seu nom comú. Els topònims s’escriuran o bé en la forma original o bé en la llengua en què estigui escrit el treball, seguint sempre el mateix criteri. Els nombres de l’u al nou, sempre que estiguin en el text, s’escriuran amb lletres, excepte quan precedeixin una unitat de mesura. Els nombres més grans s'escriuran amb xifres excepte quan comencin una frase. Les dates s’indicaran de la forma següent: 28 VI 99; 28, 30 VI 99 (dies 28 i 30); 28-30 VI 99 (dies 28 a 30). S’evitaran sempre les notes a peu de pàgina. Format dels articles Títol. El títol serà concís, però suficientment indicador del contingut. Els títols amb designacions de sèries numèriques (I, II, III,...) seran acceptats previ acord amb l'editor. Nom de l’autor o els autors. Abstract en anglès que no ultrapassi les 12 línies mecanografiades (860 espais) i que mostri l’essència del manuscrit (introducció, material, mètodes, resultats i discussió). S'evitaran les especulacions i les cites bibliogràfiques. Estarà encapçalat pel títol del treball en cursiva. Key words en anglès (sis com a màxim), que orientin sobre el contingut del treball en ordre d’importància. Resumen en castellà, traducció de l'Abstract. De la traducció se'n farà càrrec la revista per a aquells autors que no siguin castellanoparlants. Palabras clave en castellà. © 2001 Museu de Zoologia

Adreça postal de l’autor o autors. (Títol, Nom, Abstract, Key words, Resumen, Palabras clave i Adreça postal, conformaran la primera pàgina.) Introducción. S'hi donarà una idea dels antecedents del tema tractat, així com dels objectius del treball. Material y métodos. Inclourà la informació pertinent de les espècies estudiades, aparells emprats, mètodes d’estudi i d’anàlisi de les dades i zona d’estudi. Resultados. En aquesta secció es presentaran únicament les dades obtingudes que no hagin estat publicades prèviament. Discusión. Es discutiran els resultats i es compararan amb treballs relacionats. Els suggeriments de recerques futures es podran incloure al final d’aquest apartat. Agra decimientos (optatiu). Agradecimientos Refer enc Referenc enciia s. Cada treball haurà d’anar acompanyat de les referències bibliogràfiques citades en el text. Les referències han de presentar–se segons els models següents (mètode Harvard): * Articles de revista: CONROY, M. J. & NOON, B. R., 1996. Mapping of species richness for conservation of biological diversity: conceptual and methodological issues. Ecological Applications, 6:: 763–773. * Llibres o altres publicacions no periòdiques: SEBER, G. A. F., 1982. The estimation of animal abundance. C. Griffin & Company, London. * Treballs de contribució en llibres: MACDONALD, D. W. & JOHNSON, D. P., 2001. Dispersal in theory and practice: consequences for conservation biology. In: Dispersal: 358–372 (T. J. Clober, E. Danchin, A. A. Dhondt & J. D. Nichols, Eds.). Oxford University Press, Oxford. * Tesis doctorals: MERILÄ, J., 1996. Genetic and quantitative trait variation in natural bird populations. Tesis doctoral, Uppsala University. * Els treballs en premsa només han d’ésser citats si han estat acceptats per a la publicació: RIPOLL, M. (in press). The relevance of population studies to conservation biology: a review. Anim. Biodivers. Conserv.

La relació de referències bibliogràfiques d’un treball serà establerta i s’ordenarà alfabèticament per autors i cronològicament per a un mateix autor, afegint les lletres a, b, c..., als treballs del mateix any. En el text, s’indicaran en la forma usual: “...segons WEMMER (1998) ... ”, “...ha estat definit per ROBINSON & REDFORD (1991)...”, “...les prospeccions realitzades (BEGON et al., 1999)...” Quan en el text s’anomeni un autor de qui no es dóna referència bibliogràfica el nom anirà en rodona: “...un altre autor és Caughley...” Taules. Les taules es numeraran 1, 2, 3, etc. i han de ser sempre ressenyades en el text. Les taules grans seran més estretes i llargues que amples i curtes ja que s'han d'encaixar en l'amplada de la caixa de la revista. Figures. Tota classe d’il·lustracions (gràfics, figures o fotografies) entraran amb el nom de figura i es numeraran 1, 2, 3,... i han de ser sempre ressenyades en el text. Es podran incloure fotografies si són imprescindibles. La mida màxima de les figures és de 15,5 cm d'amplada per 24 cm d'alçada. S'evitaran les figures tridimensionals. Tant els mapes com els dibuixos han d'incloure l'escala. Els ombreigs preferibles són blanc, negre o trama. S'evitaran els punteigs ja que no es reprodueixen bé. Peus de figura i capçaleres de taula. Els peus de figura i les capçaleres de taula seran clars, concisos i bilingües en la llengua de l’article i en anglès. Els títols dels apartats generals de l’article (Introducción, Material y métodos, Resultados, Discusión, Conclusiones, Agradecimientos y Referencias) no aniran numerats. No es poden utilitzar més de tres nivells de títols. Els autors procuraran que els seus treballs originals no passin de 20 pàgines (incloent–hi figures i taules). Si a l'article es descriuen nous tàxons, caldrà que els tipus estiguin dipositats en una institució pública. Es recomana als autors la consulta de fascicles recents de la revista per tenir en compte les seves normes.

III

Animal Biodiversity and Conservation 24.1 (2001)

Animal Biodiversity and Conservation Animal Biodiversity and Conservation (antes Miscel·lània Zoològica ) es una revista interdisciplinar, publicada desde 1958 por el Museo de Zoología de Barcelona. Incluye artículos de investigación empírica y teórica en todas las áreas de la zoología (sistemática, taxonomía, morfología, biogeografía, ecología, etología, fisiología y genética) procedentes de todas las regiones del mundo, con especial énfasis en los estudios que de una manera u otra tengan relevancia en la biología de la conservación. La revista no publica catálogos, listas de especies sin más o citas puntuales. Los estudios realizados con especies raras o protegidas pueden no ser aceptados a no ser que los autores dispongan de los permisos correspondientes. Cada volumen anual consta de dos fascículos. Animal Biodiversity and Conservation está registrada en todas las bases de datos importantes, lo que permite una difusión mundial de sus artículos. Todos los manuscritos son revisados por el editor ejecutivo, un editor y dos revisores independientes, elegidos de una lista internacional, a fin de garantizar su calidad. El proceso de revisión es rápido y constructivo, y se realiza vía correo electrónico siempre que es posible. La publicación de los trabajos aceptados se realiza con la mayor rapidez posible, normalmente dentro de los 12 meses siguientes a la recepción del trabajo. Una vez aceptado, el trabajo pasará a ser propiedad de la revista. Ésta se reserva los derechos de autor, y ninguna parte del trabajo podrá ser reproducida sin citar su procedencia.

Normas de publicación Los trabajos se enviarán preferentemente de forma electrónica (mzbpubli@intercom.es). El formato preferido es un documento Rich Text Format (RTF) o DOC, que incluya las figuras (TIF). Si se opta por la versión impresa, deberán remitirse cuatro copias juntamente con una copia en disquete a la Secretaría de Redacción. Debe incluirse, con el artículo, una carta donde conste que el trabajo versa sobre investigaciones originales no publicadas anteriormente y que se somete en exclusiva a Animal Biodiversity and Conservation. En dicha carta también debe constar, para trabajos donde sea necesaria la manipulación de animales, que los autores disponen de los permisos necesarios y que han cumplido la normativa de protección animal vigente. Los autores pueden enviar también sugerencias para asesores. Cuando el trabajo sea aceptado los autores deberán enviar a la Redacción una copia impresa de la versión final junto con un disquete del manuscrito preparado con un procesador de textos e indicando el programa utilizado (preferiISSN: 1578–665X

blemente Word). Las pruebas de imprenta enviadas a los autores deberán remitirse corregidas al Consejo Editor en el plazo máximo de 10 días. Los gastos debidos a modificaciones sustanciales en las pruebas de imprenta, introducidas por los autores, irán a cargo de los mismos. El primer autor recibirá 50 separatas del trabajo sin cargo alguno y una copia electrónica en formato PDF. Manuscritos Los trabajos se presentarán en formato DIN A–4 (30 líneas de 70 espacios cada una) a doble espacio y con las páginas numeradas. Los manuscritos deben estar completos, con tablas y figuras. No enviar las figuras originales hasta que el artículo haya sido aceptado. El texto podrá redactarse en inglés, castellano o catalán. Se sugiere a los autores que envíen sus trabajos en inglés. La revista ofrece, sin cargo ninguno, un servicio de corrección por parte de una persona especializada en revistas científicas. En cualquier caso debe presentarse siempre de forma correcta y con un lenguaje claro y conciso. La redacción del texto deberá ser impersonal, evitándose siempre la primera persona. Los caracteres en cursiva se utilizarán para los nombres científicos de géneros y especies y para los neologismos que no tengan traducción; las citas textuales, independientemente de la lengua en que estén, irán en letra redonda y entre comillas; el nombre del autor que sigue a un taxón se escribirá también en redonda. Al citar por primera vez una especie en el trabajo, deberá especificarse siempre que sea posible su nombre común. Los topónimos se escribirán bien en su forma original o bien en la lengua en que esté redactado el trabajo, siguiendo el mismo criterio a lo largo de todo el artículo. Los números del uno al nueve se escribirán con letras, a excepción de cuando precedan una unidad de medida. Los números mayores de nueve se escribirán con cifras excepto al empezar una frase. Las fechas se indicarán de la siguiente forma: 28 VI 99; 28, 30 VI 99 (días 28 y 30); 28–30 VI 99 (días 28 al 30). Se evitarán siempre las notas a pie de página. Formato de los artículos Título. El título será conciso pero suficientemente explicativo del contenido del trabajo. Los títulos con designaciones de series numéricas (I, II, III, etc.) serán aceptados excepcionalmente previo consentimiento del editor. Nombre del autor o autores. Abstract en inglés de 12 líneas mecanografiadas (860 espacios como máximo) y que exprese la esencia del manuscrito (introducción, material, métodos, resultados y discusión). Se evitarán las © 2001 Museu de Zoologia

especulaciones y las citas bibliográficas. Irá encabeza do por el título del trabajo en cursiva. Key words en inglés (un máximo de seis) que especifiquen el contenido del trabajo por orden de importancia. Resumen en castellano, traducción del abstract. Su traducción puede ser solicitada a la revista en el caso de autores que no sean castellano hablantes. Palabras clave en castellano. Dirección postal del autor o autores. (Título, Nombre, Abstract, Key words, Resumen, Palabras clave y Dirección postal conformarán la primera página.) Introducción. En ella se dará una idea de los antecedentes del tema tratado, así como de los objetivos del trabajo. Material y métodos. Incluirá la información referente a las especies estudiadas, aparatos utilizados, metodología de estudio y análisis de los datos y zona de estudio. Resultados. En esta sección se presentarán únicamente los datos obtenidos que no hayan sido publicados previamente. Discusión. Se discutirán los resultados y se compararán con otros trabajos relacionados. Las sugerencias sobre investigaciones futuras se podrán incluir al final de este apartado. Agradecimientos (optativo). Referencias. Cada trabajo irá acompañado de una bibliografía que incluirá únicamente las publicaciones citadas en el texto. Las referencias deben presentarse según los modelos siguientes (método Harvard): * Artículos de revista: CONROY, M. J. & NOON, B. R., 1996. Mapping of species richness for conservation of biological diversity: conceptual and methodological issues. Ecological Applications, 6:: 763–773 * Libros y otras publicaciones no periódicas: SEBER, G. A. F., 1982. The estimation of animal abundance. C. Griffin & Company, London. * Trabajos de contribución en libros: MACDONALD, D. W. & JOHNSON, D. P., 2001. Dispersal in theory and practice: consequences for conservation biology. In: Dispersal: 358–372 (T. J. Clober, E. Danchin, A. A. Dhondt & J. D. Nichols, Eds.). Oxford University Press, Oxford. * Tesis doctorales: MERILÄ, J., 1996. Genetic and quantitative trait variation in natural bird populations. Tesis doctoral, Uppsala University.

* Los trabajos en prensa sólo se citarán si han sido aceptados para su publicación: RIPOLL, M. (in press). The relevance of population studies to conservation biology: a review. Anim. Biodivers. Conserv. Las referencias se ordenarán alfabéticamente por autores, cronológicamente para un mismo autor y con las letras a, b, c,... para los trabajos de un mismo autor y año. En el texto las referencias bibliográficas se indicarán en la forma usual: "...según WEMMER (1998)...", "...ha sido definido por ROBINSON & REDFORD (1991)...", "...las prospecciones realizadas (BEGON et al., 1999)..." Cuando en el texto se mencione un autor no incluido en la bibliografía el nombre irá en redonda: "...otro autor es Caughley..." Tablas. Las tablas se numerarán 1, 2, 3, etc., y se reseñarán todas en el texto. Las tablas grandes deben ser más estrechas y largas que anchas y cortas ya que deben encajarse en la caja de la revista. Figuras. Toda clase de ilustraciones (gráficas, figuras o fotografías) se considerarán figuras, se numerarán 1, 2, 3, etc., y se citarán todas en el texto. Pueden incluirse fotografías si son imprescindibles. El tamaño máximo de las figuras es de 15,5 cm de ancho y 24 cm de alto. Deben evitarse las figuras tridimensionales. Tanto los mapas como los dibujos deben incluir la escala. Los sombreados preferibles son blanco, negro o trama. Deben evitarse los punteados ya que no se reproducen bien. Pies de figura y cabeceras de tabla. Los pies de figura y cabeceras de tabla serán claros, concisos y bilingües en castellano e inglés. Los títulos de los apartados generales del artículo (Introducción, Material y métodos, Resultados, Discusión, Agradecimientos y Referencias) no se numerarán. No utilizar más de tres niveles de títulos. Los autores procurarán que sus trabajos originales no excedan las 20 páginas incluidas figuras y tablas. Si en el artículo se describen nuevos taxones, es imprescindible que los tipos estén depositados en alguna institución pública. Se recomienda a los autores la consulta de fascículos recientes de la revista para seguir sus directrices.

Animal Biodiversity and Conservation 24.1 (2001)

Animal Biodiversity and Conservation Animal Biodiversity and Conservation (formerly Miscel·lània Zoològica) is an interdisciplinary journal which has been published by the Zoological Museum of Barcelona since 1958. It includes empirical and theoretical research in all aspects of Zoology (Systematics, Taxonomy, Morphology, Biogeography, Ecology, Ethology, Physiology and Genetics) from all over the world with special emphasis on studies that stress the relevance of the study of Conservation Biology. The journal does not publish catalogues, lists of species (with no other relevance) or punctual records. Studies about rare or protected species will not be accepted unless the authors have been granted all the relevant permits. Each annual volume consists of two issues. Animal Biodiversity and Conservation is registered in all principal data bases, thus assuring world–wide access to articles published therein. All manuscripts are screened by the Executive Editor, an Editor and two independent reviewers in order to guarantee the quality of the papers. The process of review is rapid and constructive. Once accepted, papers are published as soon as practicable, usually within 12 months of initial submission. Upon acceptance, manuscripts become the property of the journal, which reserves copyright, and no published material may be reproduced without quoting its origin.

Information for authors Electronic submission of papers is encouraged (mzbpubli@intercom.es). The preferred format is a document Rich Text Format (RTF) or DOC, including figures (TIF). In the case of sending a printed version, four copies should be sent together with a copy in a computer disc to the Editorial Office. A cover letter stating that the article reports on original research not published elsewhere and that it has been submitted exclusively for consideration in A n i m a l Biodiversity and Conservation is also necessary. When animal manipulation has been necessary, the cover letter should also make explicit that the authors follow current norms on the protection of animal species and that they have obtained all rellevant permissions. Authors may suggest referees for their papers. Once an article has been accepted, authors should send a printed copy of the final version together with a disc of the manuscript prepared on a word processor. Please identify software (preferably Word). Proofs sent to the authors for correction should be returned to the Editorial Board within 10 days. Expenses due to any substantial alterations of the proofs will be charged to the authors. ISSN: 1578–665X

The first author will receive 50 reprints free of charge and an electronic version of the article in PDF format. Manuscripts Manuscripts must be presented on A–4 format page (30 lines of 70 spaces each) with double spacing. Number all pages. Manuscripts should be complete with figures and tables. Do not send original figures until the paper has been accepted. The text may be written in English, Spanish or Catalan. Authors are encouraged to send their contributions in English. The journal provides a FREE service of correction by a professional translator specialized in scientific publications. Care should be taken in using correct wording and the text should be written concisely and clearly. Wording should be impersonal, avoiding the use of the first person. Italics must be used for scientific names of genera and species as well as untranslatable neologisms. Quotations in whatever language used must be typed in ordinary print between quotation marks. The name of the author following a taxon should also be written in small print. The common name of the species should be written in capital letters. When referring to a species for the first time in the text, both common and scientific names must be given when possible. Place names may appear either in their original form or in the language of the manuscript, but care should be taken to use the same criteria throughout the text. Numbers one to nine should be written in full in the text except when preceding a measure. Higher numbers should be written in numerals except at the beginning of a sentence. Dates must appear as follows: 28 VI 99, 28,30 VI 99 (days 28th and 30th), 28–30 VI 99 (days 28th to 30th). Footnotes should not be used. Formatting of articles T itle. The title must be concise but as informative as possible. Part numbers (I, II, III,...) should be avoided and will be subject to the Editor’s consent. Name of author or authors. Abstract in English, no longer than 12 typewritten lines (840 spaces), covering the contents of the article (introduction, material, methods, results and discussion). Speculation and literature citation must be avoided. Abstract should begin with the title in italics. Key words in English (no more than six) should express the precise contents of the manuscript in order of importance. Resumen in Spanish, translation of the Abstract. Summaries of articles by non–Spanish speaking authors will be translated by the journal on request. Palabras clave in Spanish. © 2001 Museu de Zoologia

Address of the author or authors. (Title, Name, Abstract, Key words, Resumen, Palabras clave and Address should constitute the first page.) Introduction. The introduction should include the historical background of the subject as well as the aims of the paper. Material and methods. This section should provide relevant information on the species studied, materials, methods for collecting and analysing data and the study area. Results. Report only previously unpublished results from the present study. Discussion. The results and their comparison with related studies should be discussed. Suggestions for future research may be given at the end of this section. Acknowledgements (optional). References. All manuscripts must include a bibliography of the publications cited in the text. References should be presented as in the following examples (Harvard method): * Journal articles: CONROY, M. J. & NOON, B. R., 1996. Mapping of species richness for conservation of biological diversity: conceptual and methodological issues. Ecological Applications, 6:: 763–773. * Books or other non-periodical publications: SEBER, G. A. F., 1982. The estimation of animal abundance. C. Griffin & Company, London. * Contributions or chapters of books: MACDONALD, D. W. & JOHNSON, D. P., 2001. Dispersal in theory and practice: consequences for conservation biology. In: Dispersal: 358–372 (T. J. Clober, E. Danchin, A. A. Dhondt & J. D. Nichols, Eds.). Oxford University Press, Oxford. * Ph. D. Thesis: MERILÄ, J., 1996. Genetic and quantitative trait variation in natural bird populations. Ph. D. Thesis, Uppsala University. * Works in press should only be cited if they have been accepted for publication: RIPOLL, M. (in press). The relevance of population studies to conservation biology: a review. Anim. Biodivers. Conserv.

References must be set out in alphabetical and chronological order for each author, adding the letters a, b, c,... to papers of the same year. Bibliographic citations in the text must appear in the usual way: "...according to W EMMER (1998)...", "...has been defined by ROBINSON & REDFORD (1991)...", "...the prospections that have been carried out (BEGON et al., 1999)..." When an author is mentioned in the text but no bibliographical reference is given, the name must appear in ordinary print: "...another of these authors is Caughley..." Tables. Tables must be numbered in Arabic numerals with reference in the text. Large tables should be narrow (across the page) and long (down the page) rather than wide and short, so that they can be fitted into the column width of the journal. Figures. All illustrations (graphs, drawings or photographs) must be termed as figures, numbered consecutively in Arabic numerals and with reference in the text. Glossy print photographs, if essential, may be included. Maximum size of figures is 15.5 cm width and 24 cm height. Figures will not be tridimensional. Both maps and drawings must include scale. The preferred shadings are white, black and bold hatching. Avoid stippling, which does not reproduce well. Legends of tables and figures. Legends of tables and figures must be clear, concise, and written both in English and Spanish. Main headings (Introduction, Material and methods, Results, Discussion, Acknowledgements and References) should not be numbered. Do not use more than three levels of headings. Manuscripts should not exceed 20 pages including figures and tables. If the article describes new taxa, type material must be deposited in a public institution. Authors are advised to consult recent issues of the journal and follow its conventions.

Animal Biodiversity and Conservation 24.1 (2001)

VII

Animal Biodiversity and Conservation Subscription Form G Please enter my subscription to Animal Biodiversity and Conservation G 11,000 Ptas. (66.11 € ) Spain G 11,400 Ptas. (68.52 € ) Europe G 11,500 Ptas. (69.12 € ) rest of world G Personal subscription: G 3,500 Ptas. (21.04 € ) Spain G 3,900 Ptas. (23.4 4 € ) Europe G 4,000 Ptas. (24.04 € ) rest of world G Please despatch my issues by air mail (supplement of 1,000 Ptas. (6.01 € ) for outside Europe)

G Please send me the Instructions to authors

Name Institution

Address

Payment method G Cheque payment made payable to Associació d'Amics del Museu de Zoologia de Barcelona and drawn against a Spanish bank G Visa credit card Card number Valid until G Please send a proforma invoice and wait the payment

Date

Send this order form to: Associació d'Amics del Museu de Zoologia de Barcelona Museu de Zoologia Psg. Picasso s/n 08003 Barcelona, Spain Fax: +34-93-3104999 E-mail m zbpubli@intercom.es mzbpubli@intercom.es

Signature

Editor executiu / Editor ejecutivo / Executive Editor Joan Carles Senar

Secretaria de Redacció / Secretaría de Redacción / Editorial Office

Secretària de Redacció / Secretaria de Redacción / Managing Editor Montserrat Ferrer

Museu de Zoologia Passeig Picasso s/n 08003 Barcelona, Spain Tel. +34–93–3196912 Fax +34–93–3104999 E–mail mzbpubli@intercom.es

Consell Assessor / Consejo asesor / Advisory Board Oleguer Escolà Eulàlia Garcia Anna Omedes Josep Piqué Francesc Uribe

Les cites o els abstracts dels treballs d’aquesta publicació es resenyen a / Las citas o los abstracts de los trabajos de esta publicación se mencionan en / This publication is cited or abstracted in: Abstracts of Entomology, Agrindex, Animal Behaviour Abstracts, Anthropos, Aquatic Sciences and Fisheries Abstracts, Behavioural Biology Abstracts, Biological Abstracts, Biological and Agricultural Abstracts, Current Primate References, Ecological Abstracts, Ecology Abstracts, Entomology Abstracts, Environmental Abstracts, Environmental Periodical Bibliography, Genetic Abstracts, Geographical Abstracts, Índice Español de Ciencia y Tecnología, International Abstracts of Biological Sciences, International Bibliography of Periodical Literature, International Developmental Abstracts, Marine Sciences Contents Tables, Oceanic Abstracts, Recent Ornithological Literature, Referatirnyi Zhurnal, Science Abstracts, Serials Directory, Ulrich’s International Periodical Directory, Zoological Records.

Animal Biodiversity and Conservation 24.1 (2001)

ISSN 1578–665X

Índex / Índice / Contents 1–7 Baehr, M. Tasmanitachoides Erwin glabellus n. sp. from North Queensland, Australia, with a note on Tasmanitachoides lutus (Darlington) (Insecta, Coleoptera, Carabidae, Bembidiinae) 9–13 Bellés, X. Description of Sphaericus selvagensis n. sp. from the Selvage Islands, and new data on Sphaericus bicolor Bellés (Coleoptera, Ptinidae) 15–18 Camperio Ciani, A., Palentini, L. & Finotto, E. Survival of a small translocated Procolobus kirkii population on Pemba Island 19–29 Domingo–Roura, X., Marmi, J., López–Giráldez, J. F. & Garcia–Franquesa, E. New molecular challenges in animal conservation 31–52 Fa, J. E. & García Yuste, J. E. Commercial bushmeat hunting in the Monte Mitra Forests, Equatorial Guinea: extent and impact 53–63 Garin, I., Aldezabal, A., García–González, R. & Aihartza, J. R. Composición y calidad de la dieta del ciervo (Cervus elaphus L.) en el norte de la península ibérica

65–79 Oliveira, P. A. P., Simões, P. C. & Quartau, J. A. Calling songs of certain orthopteran species (Insecta, Orthoptera) in southern Portugal 81–90 Peake, T. M. & McGregor, P. K. Corncrake Crex crex census estimates: a conservation application of vocal individuality 91–99 Ricketts, T. H. Aligning conservation goals: are patterns of species richness and endemism concordant at regional scales? 101–106 Tomás, W. M., McShea, W., Miranda, G. H. B. de, Moreira, J. R., Mourão, G. & Lima Borges, P. A. A survey of a pampas deer, Ozotoceros bezoarticus leucogaster (Arctiodactyla, Cervidae), population in the Pantanal wetland, Brazil, using the distance sampling technique 107–116 Tourenq, C., Barbraud, C., Sadoul, N., Sandoz, A., Lombardini, K., Kayser, Y. & Martin, J.–L. Does foraging habitat quality affect reproductive performance in the Little Egret, Egretta garzetta? 117–124 Van Vuren, D. H. Spatial relations of American bison (Bison bison) and domestic cattle in a montane environment