An impact assessment of avrdc tomato grafting in vietnam by Bharath Kroshnan

Research

Action

An impact assessment of AVRDCâ&#x20AC;&#x2122;s tomato grafting in Vietnam

Christian Genova Pepijn Schreinemachers Victor Afari-Sefa

atinka Weinberger Christian A. Genova II Antonio L. Acedo Jr.

An impact assessment of AVRDC’s tomato grafting in Vietnam

Christian Genova, Pepijn Schreinemachers and Victor Afari-Sefa

AVRDC – The World Vegetable Center Shanhua, Taiwan

AVRDC – The World Vegetable Center is an international nonprofit research institute committed to alleviating poverty and malnutrition in the developing world through the increased production and consumption of nutritious, health-promoting vegetables.

About Research in Action The Research in Action series disseminates the practical applications of the Center’s work in vegetable breeding, production, marketing, and nutrition. The series aims to encourage vegetable-based enterprise through the extension of information, ideas, technologies, and skills.

AVRDC – The World Vegetable Center P.O. Box 42 Shanhua, Tainan 74199 TAIWAN Tel: +886 6 583 7801 Fax: +886 6 583 0009 Email: info@worldveg.org Web: www.avrdc.org AVRDC Publication: 13-773 ISBN 92-9058-203-0

Editor: Maureen Mecozzi Publishing Coordinator: Kathy Chen

Suggested citation Genova C, Schreinemachers P, Afari-Sefa V. 2013. An impact assessment of AVRDC’s tomato grafting in Vietnam. AVRDC – The World Vegetable Center, Shanhua, Taiwan. AVRDC Publication No. 13-773. 52 p. (Research in Action; no. 8).

Table of Contents List of Tables ............................................................................................................................ iv List of Figures ............................................................................................................................ v Acknowledgements ................................................................................................................... vi Executive Summary .................................................................................................................vii Acronyms .................................................................................................................................. ix

INTRODUCTION ............................................................................................................. 1 1.1.

Bacterial wilt and tomato grafting in Vietnam ............................................................ 2

1.2.

Objectives of the study ................................................................................................ 3

THE ROLE OF AVRDC – THE WORLD VEGETABLE CENTER IN PROMOTING

TOMATO GRAFTING ............................................................................................................. 4

2.1.

AVRDC’s effort to address bacterial wilt ................................................................... 5

2.2.

AVRDC's tomato grafting technology dissemination in Vietnam .............................. 8

STUDY METHODS AND DATA COLLECTION ........................................................ 11 3.1.

Selection of location and data sources ...................................................................... 11

3.2.

Data collection........................................................................................................... 11

RESULTS AND DISCUSSION ...................................................................................... 15 4.1.

Nurseries producing grafted seedlings in Lam Dong province ................................. 15

4.2.

Farm households ....................................................................................................... 18

4.2.1.

Socio-demographic characteristics .................................................................... 18

4.2.2.

Adoption of tomato grafting .............................................................................. 20

4.2.3.

Yield and yield stability of grafted tomato ........................................................ 28

4.2.4.

Production function analysis .............................................................................. 32

4.2.5.

Effect of tomato grafting on profit and household income ................................ 37

CONCLUSIONS AND RECOMMENDATIONS .......................................................... 38

REFERENCES ................................................................................................................ 39

List of Tables Table 1. Framework for integrated crop management (ICM) practices for quality tomato .................................... 4 Table 2. Survey sites and number of respondents................................................................................................. 11 Table 3. Proportionate stratified sampling by location in Vietnam ...................................................................... 13 Table 4. Average socioeconomic characteristics of nursery operators in three study sites .................................. 15 Table 5. Nursery operatorsâ&#x20AC;&#x2122; perception about why farmers buy grafted tomato seedlings................................... 16 Table 6. Marketing information for grafted tomato seedlings .............................................................................. 17 Table 7. Profitability of grafted tomato at nursery (VND per seedling) ............................................................... 17 Table 8. Socioeconomic profile of sampled respondents ..................................................................................... 19 Table 9. Average characteristics of the most recently harvested tomato field ...................................................... 20 Table 10. Ranking of the three most important reasons for planting grafted tomato............................................ 22 Table 11. List of rootstock and scion varieties released in Lam Dong province from 2004-2011 ....................... 24 Table 12. List of rootstock and scion varieties released in the Red River Delta from 2004-2011 ........................ 25 Table 13. Main rootstock and scion varieties used by farmers, and reasons for selection.................................... 26 Table 14. Seed source of main rootstock and scion varieties ............................................................................... 26 Table 15. Yield of grafted versus non-grafted tomatoes by month, 2010-2011 (t/ha) .......................................... 29 Table 16. Coefficient of variation of grafted vs non-grafted tomatoes by month, 2010-2012.............................. 29 Table 17. Farmerâ&#x20AC;&#x2122;s perceptions of changes in selected production practices with the adoption of tomato grafting .............................................................................................................................................................................. 31 Table 18. Variables used in the ordinary least squares (OLS) regression............................................................. 33 Table 19. Econometric estimation results of production inputs on yield .............................................................. 36 Table 20. Sample means of inputs and labor between grafted and non-grafted tomato production by location, 2011/2012 (in million VND/ha) ........................................................................................................................... 36 Table 21. Profitability of grafted versus non-grafted tomato production in the Red River Delta, 2011-2012 (in million VND/ha) ................................................................................................................................................... 37 Table 22. Estimated effect of tomato grafting on net profits in Lam Dong province ........................................... 37

List of Figures Figure 1. Timeline of AVRDC's research on bacterial wilt and tomato grafting, 1970-2012 ................................ 6 Figure 2. Location of Red River Delta and Lam Dong province in Vietnam ..................................................... 14 Figure 3. Adoption of tomato grafting in Lam Dong province and the Red River Delta ..................................... 21 Figure 4. Tomato production in Lam Dong province, Vietnam ........................................................................... 27 Figure 5. Seasonal calendar of production activities by location.......................................................................... 28

Acknowledgements Special appreciation goes to Ngo Quang Vinh, Deputy Director of the Institute for Agricultural Science in Southern Vietnam (IAS), and To Thi Thu Ha, Head of the Department of Vegetables and Spicy Crops of Fruit and Vegetable Research Institute (FAVRI), for their excellent logistical support during the survey implementation. We are also grateful to Trinh Khac Quang, Director General, and Bui Quang Dang, Head of the Department of Science and International Cooperation of FAVRI for their interest and hospitality during the courtesy visit. Our sincerest gratitude also goes to the survey enumerators and data encoders for their patience and diligence in collecting information from farmers and nursery operators in the field: Le Thi Ha, Nguyen Xuan Diep, Le Thi Thuy, Nguyen Tuan Dung, Duong Kim Thoa and Nguyen Trung Dung from FAVRI; Nguyen The Nhuan, Ngo Minh Dung, Dao Trong Duc, Dinh Thi Hong, Vo Thi Ngoc, Pham Thi Luyen, Ton Chat Sang, Tran Anh Thong, and Tuong Thi Ly from the Potato, Vegetable and Flower Research Center (PVFRC). We also wish to acknowledge the support of the commune leaders and district agricultural extension officers who assisted us in compiling the list of tomato farmers in our target locations, and to thank the farmers and nursery operators for their warm reception and willingness to share their time and knowledge in this research study. Finally, we are grateful to our colleagues at AVRDC â&#x20AC;&#x201C; The World Vegetable Center: Kartini Luther, Sheila de Lima and Olivia Liang for their administrative assistance; Jin-tien Hu for providing journal articles and other publications; Chih-hung Lin for answering queries related to AVRDCâ&#x20AC;&#x2122;s research on bacterial wilt; and the Global Technology Dissemination team (Greg Luther, Lydia Wu, Mandy Lin and Willie Chen) for providing information on the off-season vegetable production training workshop and answering questions about grafting.

Executive Summary AVRDC â&#x20AC;&#x201C; The World Vegetable Center started working on tomato grafting in 1992 and introduced the technique to Vietnamese scientists in September 1998, during a one-month training course at AVRDC headquarters in Taiwan. From 2002-2006, the technique was introduced to Lam Dong province in southern Vietnam in collaboration with the Potato, Vegetable and Flower Research Center, and to the Red River Delta in northern Vietnam in collaboration with the Fruit and Vegetable Research Institute in Hanoi. This evaluation study assessed the impact of the tomato grafting technique ten years after its introduction in Vietnam. It measures change in four impact indicators: the adoption of tomato grafting, yield and yield stability of grafted tomatoes, input factors affecting yield, and profit from tomato grafting. Data were collected in August 2012 from 300 tomato farmers using a semi-structured questionnaire survey. The results show 100% (n=225) adoption in Lam Dong province and 48% (n=36) adoption in the Red River Delta. For rootstock, tomato variety Vimina (HW7996) was selected in Lam Dong province as it is resistant to bacterial wilt, and eggplant variety EG203 was used in the Red River Delta as it is resistant to bacterial wilt and tolerates waterlogging, which are major problems affecting farmers in the Delta. Results show yield of grafted tomato (73.3 t/ha in Lam Dong Province and 81.4 t/ha in the intensive production area in the Red River Delta) was significantly greater than non-grafted tomato (56.5 t/ha in the Red River Delta), which confirms the findings of an earlier survey. The coefficient of variation (CV) suggests that yield of grafted tomatoes was more stable in Lam Dong province (0.20-0.30) than in the Red River Delta (0.33-0.80), but for the Red River Delta there was no clear difference in yield stability between grafted and non-grafted tomato. Further studies are required to confirm this observation, given the small sample size (n=25) for nongrafted tomatoes in the Delta. A comparison of profits between grafted and non-grafted tomato was made only for the Red River Delta because all tomato farmers in Lam Dong province had adopted the technology at the time of the survey. The average yield (81.4 t/ha) and farm gate price (8,447.8 VND/kg) of grafted tomato were significantly greater by 31% and 39%, respectively, compared with nongrafted tomato. As a result, there was a significant (p<0.100) difference in average revenues of

vii

409 million VND/ha. Although the total costs, particularly labor costs, were significantly greater for grafted tomato (189.6 million VND/ha) than for non-grafted tomato (106.6 million VND/ha), the difference in revenue was large enough to make grafted tomato significantly more profitable. The benefit-cost ratio for grafted tomato was 4.6 in comparison to 3.5 for nongrafted tomato. Nonetheless, further studies would be useful considering the relatively small sample size in the Red River Delta and the high variation observed for some of the variables. Based on the average difference in profits between grafted and non-grafted tomato, the 100% adoption rate and the total area under tomato in Lam Dong province, the estimated total profit for tomato farmers was US$ 41.7 million higher than if the same area had been planted with non-grafted tomato. The study clearly shows that in places where bacterial wilt and other soilborne diseases affecting tomato are a problem, tomato grafting offers very significant monetary benefits to farmers.

viii

Acronyms ACIAR

Australian Centre for International Agricultural Research

ADB

Asian Development Bank

AVNET

Southeast Asian Vegetable Research Network

AVRDC

AVRDC â&#x20AC;&#x201C; The World Vegetable Center

AVRDC-ARC HRDP

AVRDC-Asian Regional Center Swiss Agency for Development and Cooperation-funded Human Resource Development Project for the Mekong Region Project

AVRDC-ARC HRDP-IV

AVRDC-ARC HRDP project Phase IV

CIP

International Potato Center

CLVNET I and II

Cambodia, Lao PDR, and Vietnam Network I and II

coefficient of variation

DFID

Department for International Development (formerly the Overseas Development Administration or ODA)

EG195,203,210

AVRDC eggplant varieties EG 195, EG 203 and EG 210

FAVRI

Fruit and Vegetable Research Institute (formerly the Research Institute for Fruits and Vegetables or RIFAV)

FGD

focus group discussion

HSC

Hanoi Seed Company

HW 7996,7997,7998

AVRDC tomato varieties Hawaii 7996, Hawaii 7997 and Hawaii 7998

IAS

Institute of Agricultural Science For Southern Vietnam

ICRISAT

International Crops Research Institute for the Semi-Arid Tropics

IDM

integrated disease management

IPM

integrated pest management

PVFRC

Potato, Vegetable and Flower Research Center

SAVERNET I and II

South Asia Vegetable Research Network I and II

SUSPER

Sustainable Development of Peri-urban Agriculture

VND

Vietnamese Dong

INTRODUCTION

Grafting is the union of two plant parts, a rootstock and a scion.1 The rootstock is the base portion of the union that provides the root system while the scion is the upper portion that carries the harvestable yield. Vegetable grafting is an ancient practice with the earliest evidence dating back as early as 1800 BC (Mudge et al., 2009). However, wide adoption of the practice began only in the 20th century. It began in Japan and Korea in the 1920s using resistant rootstocks to control soil-borne diseases caused by Fusarium oxysporum (Davis et al., 2008; Mudge et al., 2009) and has since expanded to include vegetables such as eggplant (Solanum melongena L.), cucumber (Cucumis sativus L.), tomato (Solanum lycopersicum), pepper (Capsicum annum L.) and several other solanaceous crops. Vegetable grafting can provide a high level of tolerance to soil-borne diseases, such as those caused by Ralstonia solanacearum, Fusarium, Verticillium, Phytophthora, Pseudomonas, Didymella bryoniae, Monosporascus cannonballus and nematodes, but the degree of tolerance varies considerably with the type of rootstock used (Lee, 1994; Venema et al., 2008; Louws et al., 2010; Mohamed et al., 2012). Grafting can also improve quality attributes of scionproduced fruits due to the increase in fruit index (Davis et al., 2008), number of fruits/truss and fruit weight (Turhan et al., 2011), fruit yield (Burleigh et al., 2005; Khah et al., 2006; Palada and Ali, 2006; Qaryouti et al., 2007; Davis et al., 2008), and the longer harvest duration (Lee, 1994). Several studies have reported that vegetable grafting can increase resistance to abiotic stresses such as drought, salinity, heat and low soil temperatures, and improve water use efficiency (Lee, 1994; Martorana et al., 2006; Venema et al., 2008; Schwarz et al., 2010; Mohamed et al., 2012). The purpose of grafting in tomato production is to combine a flood- and bacterial wilt-resistant rootstock with a high-yielding tomato scion (Aganon et al., 2002). It is an alternative crop management strategy to control bacterial wilt when high-yielding resistant tomato varieties are unavailable (Wang and Lin, 2005). The use of grafting is widespread across Asia, parts of Europe and the Middle East (Mohamed et al., 2012). Grafted seedlings are commonly used in

This is to differentiate from what is termed as â&#x20AC;&#x2DC;double workingâ&#x20AC;&#x2122; in some plants, which consists of a three-part

grafted plant, i.e., rootstock, interstock, and scion.

An impact assessment of AVRDC's tomato grafting in Vietnam

hydroponic farming, as diseases infecting the plants through their roots can spread quickly in such systems (Lee et al., 2010).

1.1. Bacterial wilt and tomato grafting in Vietnam Tomato is a major crop in Vietnam, but during the hot-wet season yields are low due to poor fruit setting caused by high temperatures and high incidence and severity of disease problems, particularly bacterial wilt caused by R. solanacearum (Doan and Nguyen, 2005; Nguyen and Ranamukhaarachchi, 2010). Bacterial wilt has been reported for all eight administrative regions of Vietnam in varying degrees of severity. It is usually more severe during the wet season (April-October) than during the drier months (November-March) (Tung, 1985; Vinh and Ngo, 2006). Prior to the 1990s, the Red River and Mekong River deltas did not have problems with bacterial wilt in crop production during the drier months of the year (Tung, 1985). Yet, in the mid-1990s, Dung (1997) found that bacterial wilt had become prevalent all year-round in Hanoi and adjacent areas, reaching peak levels from March to June, and from September to November. Bacterial wilt affects tomato plants severely and can lead to 100% yield loss (AfariSefa, 2012). In Ho Chi Minh City, bacterial wilt has become prevalent with the rapid expansion of vegetable cultivation including tomato, eggplant and pepper, which are all highly susceptible to the disease. Farmers have few options for managing bacterial wilt once the soil is infested with the bacterium (Wang and Lin, 2005). Disease-resistant varieties can be overwhelmed by the pathogen due to its genetic diversity and complex genotype-environment interactions. The usefulness of crop rotation is limited due to the pathogen’s wide range of host plants (Nguyen and Ranamukhaarachchi, 2010). Chemical control of soil-borne diseases is costly and usually unsuccessful (Lin et al., 2008). AVRDC – The World Vegetable Center (AVRDC) introduced tomato grafting to Vietnam through a one-month training course for Vietnamese scientists at AVRDC Headquarters in Taiwan in 1998. Dr. Ngo Quang Vinh, one of the trainees and current Deputy Director of the Institute of Agricultural Science of Southern Vietnam (IAS), introduced the technique to vegetable farmers in southern Vietnam in 2002-2003. The technique was introduced to the Red River Delta in 2002-2006 through the AVRDC project, “Sustainable Development of Periurban Agriculture (SUSPER)” (Palada and Wu, 2005). An initial rapid assessment conducted recently suggested that the tomato grafting approach has been widely adopted across Vietnam

AVRDC - The World Vegetable Center

and was very profitable (Afari-Sefa, 2012). In Lam Dong province, it has been successful for managing bacterial wilt of tomato year-round, even in severely infected soils resulting from the very humid conditions associated with the agroclimatic zones of the Central Highlands. Based on interviews with nursery operators, Afari-Sefa (2012) found that grafted seedlings had a survival rate of 95% while non-grafted seedlings had a survival rate of only 20%, and that yield of grafted tomato plants was double. In the Red River Delta, soil-borne diseases are presently a problem only during the wet season from May to October. In the major dry season (November-April), farmers typically plant nongrafted tomato seedlings, as these are much cheaper. Grafting also offers benefits beyond disease control by helping growers manage abiotic stresses such as excess moisture and soil temperature extremes, and allowing for the extension of the growing season. Field trials by scientists at the Fruits and Vegetable Research Institute (FAVRI) showed that grafting can increase stress tolerance and productivity while maintaining high fruit quality. In addition, grafted plants have higher yields and show increased water and nutrient uptake (Gomi and Masuda, 1981; Heo, 1991; Jang et al., 1992 as cited by Lee, 1994).

1.2. Objectives of the study The aim of this study was to assess the impact of AVRDCâ&#x20AC;&#x2122;s tomato grafting on productivity and profitability of farmers and nursery operators 10 years after the technique was introduced in Vietnam (2002 - 2012). To the best of our knowledge, there are no previous studies on the adoption of tomato grafting in Vietnam, and there are few data available on the impact of grafting on the productivity and well-being of farmers and other market actors in Vietnam or elsewhere. This report provides a detailed characterization and descriptive analysis of tomato farmers and nursery operators with emphasis on the use of grafting as an alternative production method, and examines constraints to the adoption of this method. The following impact indicators are employed: the diffusion of tomato grafting technology, yield and yield stability of grafted tomatoes, input factors affecting yield, and the contribution of grafting to farm profitability.

An impact assessment of AVRDC's tomato grafting in Vietnam

2. THE ROLE OF AVRDC – THE WORLD VEGETABLE CENTER IN PROMOTING TOMATO GRAFTING AVRDC – The World Vegetable Center is an international nonprofit research and development institution that promotes environmentally sound, sustainable, and profitable vegetable production and marketing systems to increase the supply of nutritious vegetables to consumers. AVRDC’s integrated research disciplines develop vegetable lines and technologies that increase production and encourage consumption, thereby increasing income opportunities and healthier diets for the poor in developing economies. These are offered as international public goods to partners, such as national agricultural research systems, non-governmental organizations, universities, and private seed companies, and through training programs at AVRDC headquarters in Taiwan, its regional centers in Thailand, India, Tanzania and Mali, and offices in Cameroon, Uzbekistan, Bangladesh, Indonesia and Fiji. AVRDC is renowned for breeding heat-tolerant tropical tomato lines, and has developed and adapted integrated crop management systems for safer vegetable production (Table 1) as a model project that includes grafting, among others (Chadha, 2010). Table 1. Framework for integrated crop management (ICM) practices for quality tomato AVRDC-developed/ adapted ICM component healthy seedlings

fertilizer management water management integrated pest/disease management

crop management

Subcomponent Media Watering avoid shading IPM and IDM* starter solution balanced fertilizer use drip irrigation furrow irrigation resistant varieties grafting onto bacterial wilt or Fusarium wilt-resistant rootstocks nethouses or net tunnels sex pheromones biofumigation for bacterial wilt colored sticky traps biopesticides (neem, Bt) minimum and efficient use of pesticides Staking Mulching raised beds pruning and binding branches

* IPM – integrated pest management, IDM – integrated disease management Source: Adapted from Chadha (2010)

AVRDC - The World Vegetable Center

2.1. AVRDC’s effort to address bacterial wilt The problem of bacterial wilt has been at the forefront of AVRDC’s research since its establishment in 1973 (Opeña and Tschanz, 1987), and finding solutions has been an objective in nearly all tomato breeding programs in Southeast Asia (Hanson et al., 1996). AVRDC led several research networks starting with the Collaborative Vegetable Research Program in Southeast Asia (also known as Southeast Asian Vegetable Research Network [AVNET]) in 1989, which created a platform for bacterial wilt research and facilitated the exchange of research results and germplasm. In addition, AVRDC collaborated with the International Potato Center (CIP), and the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) to determine options for controlling the disease in other crops affected by bacterial wilt such as potatoes, peanuts, eggplants and peppers. The partnership started in 1976 with the “Planning Conference and Workshop on the Ecology and Control of Bacterial Wilt caused by Pseudomonas solanacearum” supported by grants from the US Agency for International Development (USAID), the University of California/AID Pest Management and Related Environmental Project, and CIP (Hayward et al., 2005) (Fig. 1). In 1992, AVRDC together with the Australian Centre for International Agricultural Research (ACIAR), ICRISAT, CIP and the UK Department for International Development (DFID) initiated the International Bacterial Wilt Symposia in Kaohsiung, Taiwan, which became a regular event every four years (Hayward et al., 2005). The fifth and most recent symposium was held at the Oil Crops Research Institute of the Chinese Academy of Agricultural Sciences in China in 2011. There was also the South Asia Vegetable Research Network (SAVERNET I and II) and the Cambodia, Lao PDR and Vietnam Network (CLVNET I and II), which promoted grafting as one of its component technologies in these regions (AVRDC, 2000b; ADB, 2005). To date, AVRDC has worked on host plant resistance to manage the disease, but the usefulness of resistant varieties is limited in geographical scope due to large variations in virulence of bacterial wilt and the location-specific nature of resistance related to soil physical and chemical properties, cropping systems, host genotypes, local weeds and microbes (Lin et al., 2008).

An impact assessment of AVRDC's tomato grafting in Vietnam

Figure 1. Timeline of AVRDC's research on bacterial wilt and tomato grafting, 1970-2012 In 1992, AVRDC began conducting simple experiments in Taiwan to develop quick and inexpensive grafting procedures to ascertain graft compatibility between tomato, pepper and eggplant scions onto eggplant rootstocks (AVRDC, 1994). The following year, grafting F1 fresh market tomato onto tomato or eggplant rootstocks tolerant to waterlogging was shown to extend options for growing tomato under tropical hot-wet summer conditions in Taiwan (AVRDC, 1995; Midmore et al., 1997). A follow-up study on the effects of different grafting techniques (use of tube and sticker) on the success of tomato/eggplant rootstock combinations identified tube grafting as the most successful method for grafting tomato onto eggplant (Attanayaka et al., 1997), which became part of AVRDCâ&#x20AC;&#x2122;s grafting procedures from 1998 onwards (Ya-Juia et al., 2001). AVRDC recommended tomato line Hawaii 7996 to be used as rootstock because it has a high level of resistance to bacterial and Fusarium wilt (Wang et al., 1998). In 1998, AVRDC further experimented on different eggplant rootstock varieties for grafted tomato by evaluating yield, growth and fruit quality parameters. Eggplant varieties

AVRDC - The World Vegetable Center

EG190, EG203 and EG219 were found to be resistant to bacterial wilt, Fusarium wilt and root knot nematode (AVRDC, 2000a). It was found that cherry tomato (ASVEG#6) grafted onto these rootstocks produced similar or better quality fruits compared with non-grafted plants used as control. In 1999, AVRDC studied the effect of rain shelters and grafting on tomato yields, but results were not significant due to low rainfall during the experiment (AVRDC, 2000a). Several sweet pepper and chili pepper varieties also were identified as resistant to bacterial wilt (AVRDC, 2002). A farmer field day and a four-day training course on grafting techniques for summer tomato production were held in Taipao City, Chiayi County, Taiwan in August and September 1999 to transfer the grafting technique to the national agricultural research system and local farmers (AVRDC, 2000a). Follow-up observations indicated rapid adoption of the technology as local nursery operators started using grafting the following year with three nurseries in Chiayi County alone producing about 330,000 grafted tomato seedlings (AVRDC, 2001). AVRDC also began promoting the grafting of tomato onto eggplant rootstock and transplanting seedlings onto raised beds protected by rain shelters to enable production during the hot-wet months when flooding and bacterial wilt are major production constraints. Training materials on grafting tomatoes for summer production in the hot, wet season were published in the same year (AVRDC, 2001). A 2001 study on planting media found that coconut coir gave the highest percentage of seedlings with a uniform stem diameter (AVRDC, 2002). Field experiments conducted over several years (1998-2007) at AVRDCâ&#x20AC;&#x2122;s headquarters in Taiwan and other locations in Asia (Philippines, Cambodia, Bangladesh and Lao PDR) demonstrated the benefits of grafting tomato onto eggplant or tomato rootstocks during unfavorable environmental conditions in the hot-wet season. Tomato grafting was shown to improve resistance against flooding and bacterial wilt, gave a high plant survival rate and increased the marketable fruit yield (Midmore et al., 1997; Rashid et al., 1999; 2000; Black et al., 2002; AVRDC, 2004; De la PeĂąa and Hughes, 2007). Results of the field trials led to the recommendation of tomato rootstock varieties with good bacterial wilt resistance but otherwise poor horticultural traits such as Hawaii 7996, Hawaii 7997 and Hawaii 7998, while eggplant varieties EG195, EG203 and EG210 were recommended for areas affected by flooding or excess soil moisture (Wang et al., 1998; Black, et al., 2003).

An impact assessment of AVRDC's tomato grafting in Vietnam

2.2. AVRDC's tomato grafting technology dissemination in Vietnam In Vietnam, grafting previously had been used for fruit tree propagation, especially for orange and longan (IFPRI, 2002). In 1998, Vietnamese scientists attended a one-month training on off-season vegetable production organized by AVRDC in Taiwan. The training was attended by Dr. Ngo Quang Vinh, the former Regional Coordinator of the AVRDC-Asian Regional Center Swiss Agency for Development and Cooperation-funded project, “Human Resource Development Project for the Mekong Region” (AVRDC-ARC HRDP) (Ooi, Canillas and Yambao, 2007), and Pham My Linh, a researcher from FAVRI. Training topics included "Grafting and its benefits for off-season production" and "Studies on new grafting as a method to overcome flooding damage to tomatoes" which were taught by AVRDC staff including Dr. Toshio Hanada, Deng-lin Wu, Roan Yu-chi and Dr. David Midmore (Vinh and Ngo, 2006). In 2001, Vu Thanh Hai, a faculty member of Hanoi Agricultural University, received training on vegetable production technologies for the hot-wet season including tomato and pepper grafting (AVRDC, 2002). His research has contributed to the development of new training materials and the AVRDC International Cooperator’s Guide on grafting. In 2002, another FAVRI staff member, Le Thi Thuy, attended a training course in off-season production of grafted tomato at AVRDC headquarters as part of the SUSPER Project. From the late 1980s until 2007, AVRDC implemented the AVRDC-ARC HRDP (AVRDC, 2008) in Lao PDR, Cambodia, Vietnam, Myanmar and China. In Vietnam, AVRDC worked with the Institute of Agricultural Science for Southern Vietnam (IAS), Research Institute of Fruits and Vegetables (RIFAV) in Hanoi, Hanoi Agricultural University, Can Tho University, and the Faculty of Agronomy of Hue University of Agriculture and Forestry. During the implementation of this project, IAS through Dr. Vinh conducted three experiments in Hoc Mon District, Ho Chi Minh City and Duc Trong District, Lam Dong province from 2002-2003 to identify tomato rootstock varieties resistant to bacterial wilt when grafted onto normal susceptible tomato varieties. The experiment found HW96 or Hawaii7996 (Vimina) resistant to bacterial wilt (Vinh and Ngo, 2006). Lam Dong province, the largest producer of tomatoes in Vietnam, was the first province to apply tomato grafting. Dissemination of the technology in the province began in 2003 when IAS trained 56 lead farmers and 59 extension staff from provincial and district extension offices and relevant agricultural institutions under the AVRDC-ARC HRDP-IV (AVRDC, 2008). The

AVRDC - The World Vegetable Center

method to produce grafted seedlings and procedures for managing a crop of grafted tomato were developed and published during this period (Black et al., 2003). Two TV broadcasts were produced highlighting tomato grafting for off-season planting and home vegetable gardens. IAS also produced and distributed one-page handouts and leaflets in Vietnamese (AVRDC, 2008) on the following topics: 

Technical process for grafting tomatoes, 200 copies



Technical process for growing grafted tomatoes, 200 copies



Guide for establishing a mini-base (tent) for home-sized tomato grafting (4,000-5,000 seedlings/grafting time), 50 copies



Establishing a model of a large base (house) for farm-sized grafting and post-grafting tomatoes (700,000-800,000 seedlings per month), 10 copies



Guide for compost-making, 50 copies



Guide for establishing home-sized (tent) watermelon grafting (20,000 seedlings/grafting time), 50 copies



Guide for growing grafted watermelon, 100 copies



Design of nethouses (500 m2 and 1000 m2), 50 copies

As noted by Baumuller (2012), farmers will easily adopt a technology if they have: (a) knowledge of the existence of the technology, the ability to assess its suitability for the farming system as well as potential risk, and the ability to obtain and finance the technologies; (b) requisite knowledge on how to use them, the ability to use them (e.g. sufficient labor or other resources), and the ability to manage any associated risk; and (c) ability to accrue reasonable profits from sales and to save and reinvest from the resulting returns. Indeed, the various training, extension and information campaigns spearheaded by Dr. Vinh as the Regional Coordinator of the HRDP project led to the widespread adoption of tomato transplants grafted onto resistant rootstocks for the control of bacterial wilt in Vietnam (Vinh and Ngo, 2006). The use of diverse media to deliver information and messages fostered better understanding among local plant protection extension staff and farmers about the role of resistant rootstocks in the control of diseases caused by soil-borne pathogens (Dau et al., 2009). An online newspaper, Nong Nghiep Vietnam, and the online journal New Agriculturist credited the role of AVRDC and that of local scientific research institutes in the research, testing and dissemination of the

An impact assessment of AVRDC's tomato grafting in Vietnam

method to tomato farmers in Vietnam.2 By 2007, grafted tomatoes were planted on 4,000 ha of arable land in Lam Dong province and Vinh and Ngo (2006) estimated that this gave farmers an additional profit of US$ 6 million each year. In north Vietnam around Hanoi, the Hanoi-CLV Peri-urban agriculture project conducted field experiments by setting up rain shelters, simple grafting chambers and grafting nurseries in 2002 (AVRDC, 2003). The project tested three scions (local cultivars VL2000 and HS902, and AVRDC line CHT501) grafted onto two AVRDC recommended rootstocks (EG203 and Hawaii 7996), and found no significant advantage of grafting due to the lack of environmental stresses during the experiment. However, grafting became popular among farmers and attracted the attention of the Hanoi Seed Company (HSC), which requested technical training from FAVRI for rapid and mass production of grafted seedlings (Palada and Wu, 2010). In 2008 -2009, the Ministry of Agriculture and Rural Development (MARD) commissioned a project to improve the tomato grafting protocol to develop off-season tomato in the Red River Delta and help farmers successfully produce grafted seedlings by themselves (Thuy, 2010). Since 2010, FAVRI has set up three seedling stations that supply an aggregated volume of two million grafted tomato seedlings annually; the stations are located at FAVRI, Hanoi; Moc Chau district, Son La province; and Tam Dao district, Vinh Phuc province. Currently, the area devoted to grafted tomatoes is around 50-60 hectares (ha), up from 1-2 ha in 2007.

http://nongnghiep.vn/nongnghiepvn/vi-vn/72/45/67/18428/Ky-thuat-ghep-ca-chua-.aspx (accessed December

17, 2012); and http://www.new-ag.info/en/focus/focusItem.php?a=38 (accessed August 19, 2013).

AVRDC - The World Vegetable Center

3. STUDY METHODS AND DATA COLLECTION 3.1. Selection of location and data sources This study was conducted in six provinces comprising eight districts and 20 communes in the Red River Delta and Lam Dong province that are considered to be the main tomato-producing areas in Vietnam (Table 2; Fig. 2). The national collaborators from FAVRI and the Potato, Vegetable and Flower Research Center (PVFRC) selected the study areas where tomato is a major crop. The study used both primary and secondary data. A structured questionnaire was pre-tested in both locations and translated into Vietnamese prior to data collection in August 2012. Secondary data came from official statistics and documents provided by FAVRI and PVFRC on tomato production. Table 2. Survey sites and number of respondents Item Provinces/districts

Farmers No. of respondents No. of grafting adopters No. of focus group discussion sessions conducted (and participants) Nursery operators No. of respondents

Lam Dong province Doc Truong District Don Duong District

Red River Delta Bac Ninh Province: Tu Son District Ha Noi Province: Hoai Duc District Hai Duong Province: Hai Duong City and Nam Sach District Nam Dinh Province: Nghia Hung Vinh Phuc Province: Vinh Tuong District

Total 6 provinces and 8 districts

225 225 0

75 36 1 (5 males, all adopters)

300 261 1

3.2. Data collection Twenty nursery operators and 300 tomato growers were interviewed. Seventy-five respondents (representing 25% of the total sample) were selected from the Red River Delta, while 225 respondents (representing 75%) were selected from Lam Dong province. All nursery operators

An impact assessment of AVRDC's tomato grafting in Vietnam

were from Lam Dong province. The sample size and regional distribution were pre-determined based on the estimated population of tomato farmers and the relative importance of grafted tomato in each region, as well as the time and resources available. After identifying the main tomato-producing provinces, districts and communes, a list of tomato farmers in each selected production area was compiled by agricultural extension officers in Lam Dong province and FAVRI staff, assisted by the commune leaders. Compiling the list of tomato farmers in both locations proved a daunting task given the time constraints and the challenges faced in identifying tomato farmers; it is therefore likely that the total number of tomato farmers in both regions (1,440) is an underrepresentation. A two-stage stratified random sampling was used to identify the sample units, i.e. farm households. Areas were stratified by district and by commune in Lam Dong province; and by province, district and commune in the Red River Delta. Tomato growers were allocated across the provinces/districts/communes so that the proportion of farmers sampled for each district/commune was identical to the proportion of farmers in each district/commune in the total population (Table 3). Two semi-structured questionnaires were developed and administered by trained enumerators. Both the farmer and nursery operator questionnaires were pre-tested in each location (Lam Dong province and Red River Delta) and adapted for local conditions. For the farmer questionnaire, household heads were surveyed to elicit both quantitative and qualitative data. Data included socio-demographic information, farm assets, farmerâ&#x20AC;&#x2122;s perceptions on the use of grafting in growing tomato, rootstock and scion varieties used, cost and returns of non-grafted and grafted tomato cultivation, changes in crop management practices following the adoption of grafting, pest and disease management, marketing information, training and extension needs, and household income and welfare indicators. A focus group discussion was also conducted with selected farmers in one district in the Red River Delta to generate supplementary qualitative data for triangulation purposes, given the lack of appropriate baseline data. The nursery operator questionnaire included socio-demographic information, agricultural land endowment, nature of grafting nursery operations, most commonly used rootstock/scion combination, cost and returns of non-grafted and grafted tomato seedlings, changes in nursery operations following adoption of grafting technique, pest and disease incidence, market information, training and extension needs, enterprise and personal assets, and questions on the most important challenges being faced by and reforms most needed in the tomato grafting industry.

AVRDC - The World Vegetable Center

Table 3. Proportionate stratified sampling by location in Vietnam Province/District/Commune Number of tomato growers

Population Frequency 1,440

% -

Proportionate stratified sample Frequency % 300 -

Lam Dong province a. Communes in Don Duong district Da Ron D'ran Ka Do P'ro Tu tra Lac XuĂ˘n Sub-total b. Communes in Duc Trong district Lien Nghia Hiep TháşĄnh Gia Chanh Phu Hoi Tan Hoi Tan Thanh Sub-total

169 96 131 101 63 212 772

21.9 12.4 17.0 13.1 8.2 27.5 100.0

39 22 30 23 15 49 178

21.9 12.4 17.0 13.1 8.2 27.5 100.0

36 14 10 69 46 27 202

17.8 6.9 4.9 34.2 22.8 13.4 100.0

8 3 3 16 11 6 47

17.8 6.9 4.9 34.2 22.8 13.4 100.0

Red River Delta a. Bac Ninh province b. Hai Duong province c. Nam Dinh province d. Vinh Phuc province e. Ha Noi province Sub-total

18 97 241 100 10 466

3.9 20.8 51.7 21.5 2.1 100.0

3 16 38 16 2 75

3.9 20.8 51.7 21.5 2.1 100.0

An impact assessment of AVRDC's tomato grafting in Vietnam

Figure 2. Location of Red River Delta and Lam Dong province in Vietnam

AVRDC - The World Vegetable Center

4. RESULTS AND DISCUSSION 4.1. Nurseries producing grafted seedlings in Lam Dong province Operating the nursery was the main occupation for all respondents, yet 35% produced vegetables as well. About 85% operated a single nursery with the mean land area being nearly half a hectare. The average nursery operator in our sample was in his mid-40s with 12 years of education (Table 4). Table 4. Average socioeconomic characteristics of nursery operators in three study sites Variable Number of respondents General characteristics Age (years) Sex (% male) Respondent is owner of nursery (%) Education (years) Nursery operator as main occupation (%) Secondary occupation (%) a. none b. vegetable producer Household size (person) Land area Operate a single nursery (%) Total area of operation (‘000 m2) Nursery area (‘000 m2) Vegetable cultivated area (‘000 m2) Nursery experience Member of nursery operator organization (%) Experienced in non-grafting tomato seedlings at nursery (% yes) Experienced in commercial operation of non-grafted nursery before (% yes) Training Attended vegetable grafting training (%) Avg. number of tomato grafting trainings attended Received certificates or letters of recognition after attendance to this tomato grafting training (%)

Lam Dong Province 20

Don Duong District 15

Duc Trong District 5

43.8 (6.2) 80.0 85.0 12.0 100.0

42.4 (6.3) 73.3 80.0 12.0 100.0

48.2 (3.9) 100.0 100.0 10.0 100.0

65.0 35.0 4.3 (0.8)

60.0 40.0 4.3 (0.9)

80.0 20.0 4.4 (0.5)

85.0 5.6 (5.7) 5.0 (4.1) 1.2 (2.6)

80.0 5.6 (6.4) 4.9 (4.5) 1.6 (2.9)

100.0 5.6 (2.7) 5.2 (2.8) 0.2 (0.4)

47.4 30.0

57.1 20.0

20.0 60.0

30.0

20.0

60.0

85.0 1.3 (0.7) 17.6

93.3 1.2 (0.7) 21.4

60.0 1.7 (0.6) 0.0

Note: Values in parenthesis are standard deviations (SD).

Many operators (85%) had already attended vegetable grafting training, most of them on tomato (n=15/17, 88.2%), from 2007 to 2010, with three nursery operators (17.6%) having received a certificate or letter of recognition for attending such training. Before engaging in the production of grafted tomato seedlings, 70% had no prior experience with tomato seedlings.

An impact assessment of AVRDC's tomato grafting in Vietnam

The use of grafting and the type of planting material used from an operatorâ&#x20AC;&#x2122;s perspective are mainly to protect crop from soil-borne and other diseases, and to extend the growing season throughout the year (Table 5). More than half of the operators thought that grafted tomato was more suitable to meet market demand and gave a higher output (35%) and a higher income and profit (40%). All operators use Vimina (Hawaii 7996) as rootstock and Anna as scion, which were priced around VND 3.4-3.5 million and VND 38.9 million per kilogram of seed, respectively. Table 5. Nursery operatorsâ&#x20AC;&#x2122; perception about why farmers buy grafted tomato seedlings Reasons of farmers for growing grafted tomato Protects from soil-borne diseases (caused by fungi, bacteria, or nematodes) and other diseases Helps control other abiotic stressors, excess moisture (waterlogging) and possibly soil temperature extremes as well as allowing the extension of the growing season Increases tomato yield and not necessarily for disease or abiotic stress control

Lam Dong province Rank % 1 100.0

Don Duong District Rank % 1 100.0

Duc Trong District Rank % 1 100.0

53.8

50.0

100.0

60.0

75.0

66.7

Note: Percent of cases.

All nursery operators conferred and discussed with each other at least twice a year to set the price of seedlings for farmers (Table 6). Grafted seedlings were sold at an average price of VND 612.0 per seedling, giving operators a mean profit of VND 126.0 per seedling (Table 7), which was nearly twice the profit from non-grafted tomato seedling of VND 73.3 per seedling; however, the profit margin was smaller. This requires further verification given the small sample size of nursery operator respondents in this study. The additional income from grafted tomato helped a few operators (n=3) to invest in other business ventures, such as opening a shop for selling agricultural inputs.

AVRDC - The World Vegetable Center

Table 6. Marketing information for grafted tomato seedlings Variable N Base seedling price on the price charged by other nurseries Discuss price information with other nursery operators (%) When do you seek price information prior to selling? a. 2x a year b. 2 wks before c. one month before d. 1 wk before Satisfied with your current seedling market information (% yes) Profits from nursery invested in other business ventures (%)

Lam Dong Province 20 100.0

Don Duong District 15 100.0

Duc Trong District

100.0

45.0 25.0 20.0 10.0 80.0

53.3 20.0 13.3 13.3 80.0

20.0 40.0 0.0 40.0 80.0

15.0

20.0

0.0

5 100.0

Note: Percent of cases. â&#x20AC;&#x2DC;n/aâ&#x20AC;&#x2122; not applicable.

Table 7. Profitability of grafted tomato at nursery (VND per seedling) Variable Non-grafted N Estimated cost Selling price Profit Profit margin (%)

Lam Dong Province

Don Duong District

Duc Trong District

6 151.7 (60.5) 225.0 (98.7) 73.3 (43.7) 32.6

3 140.0 (52.0) 216.7 (115.5) 76.7 (63.5) 35.4

3 163.3 (77.7) 233.3 (104.1) 70 (26.4) 30.0

Grafted N Estimated cost Selling price Profit Profit margin (%)

20 486.0 (85.5) 612.0 (106.8) 126.0 (45.1) 20.6

15 486.0 (84.1) 610.0 (108.9) 124.0 (35.6) 20.3

5 486.0 (99.9) 618.0 (112.3) 132.0 (71.9) 21.3

Note: Values in parenthesis are standard deviations (SD). Profit margin = (selling price - estimated cost) / (selling price * 100).

Nursery operators perceived that the major problems of farmers in terms of grafted tomato production and marketing were: (1) heavy rainfall leading to high incidence of pests and diseases; (2) seasonal fluctuations in market prices; (3) a lack of traders operating in the area leading to low market demand and producer prices; (4) challenges to meet consumer preference for quality produce; and (5) a lack of technical know-how in the production of grafted tomato. Challenges to nursery operators themselves mostly relate to input supplies: high and/or unstable prices of inputs and seeds (90%), poor quality of seeds and agricultural inputs (40%), and the lack of seed sources (35%). Collecting payments from farmers was also a problem for

An impact assessment of AVRDC's tomato grafting in Vietnam

45% of the operators. Improvements to the tomato grafting sector, as suggested by nursery operators, could include controlling the quality of inputs (seeds, fertilizers, pesticides, etc.), providing credit to and finding new markets for farmers. A few suggested â&#x20AC;&#x153;building a wholesale marketâ&#x20AC;? and processing facilities for tomato.

4.2. Farm households 4.2.1. Socioâ&#x20AC;?demographic characteristics In the Red River Delta, tomato production takes place from September to February, while in Lam Dong province tomato is cultivated all year round. Most tomato farms are small, with an average farm size of 3,188 m2 (Table 8). Men, in their mid-40s with 7 years of education, typically head tomato-producing households (80%). Most cultivate tomato as their sole crop and main source of livelihood. The average farmer has nearly 20 years of experience as an independent farmer and 11 years of experience growing tomato. Most farmers are members of a farmer-based organization or association, but only half of the respondents had attended an average of two vegetable grafting training programs that included tomato grafting methods. The average tomato growing farm household has 4-5 family members, which is slightly greater than the average household size in the Red River Delta (3.5) and in Lam Dong province (3.9) as reported in the 2009 Vietnam Population and Housing Census (Central Population and Housing Census Steering Committee, 2010). The average farm size in Lam Dong province is roughly twice that in the Red River Delta. In the Red River Delta, non-adopters of tomato grafting have a larger area under tomato than adopters. About 75% of the adopters in Red River Delta attended vegetable grafting training, while 49% of the non-adopters did. In comparison, only 52% of the adopters in Lam Dong province had received training in grafting techniques. This is not surprising given the presence of well-developed private commercial tomato grafting nurseries supplying grafted seedlings to farmers in Lam Dong province compared to the Red River Delta, where seedlings are provided to farmers by FAVRI and there is minimal involvement from the private sector.

AVRDC - The World Vegetable Center

Table 8. Socioeconomic profile of sampled respondents Item

All

Number of farmers (n) General profile Age (year) Gender (% male) Respondent is HH head (%) Education level (mode) Main occupation (%) Farming Other Secondary occupation (%) None Non-agricultural labor Trading Household size (persons) Farm experience Member of a farmer organization (%) Farm decision-making experience (years) Experience in tomato cultivation as HH head (years) Land area (â&#x20AC;&#x2DC;000 m2) Total farmland Land under tomato Tomato area under screen house production Land under other vegetables Training Attended vegetable grafting training (%) Number of tomato grafting trainings attended

Red River Delta

300

Lam Dong province Adopters 225

Adopters 36

Non-adopters 39

44.2 (9.3) 80.0 87.1 7

42.6 (8.6) 81.8 89.3 9-10

45.9 (9.1) 83.3 88.0 7

51.7 (9.6) 66.7 70.0 7

98.7 1.3

99.1 0.9

94.4 5.6

100.0 0.0

79.8 9.0 3.0 4.4 (1.4)

87.6 7.1 0.9 4.3 (1.2)

46.1 11.5 15.4 4.6 (1.2)

25.0 31.2 12.5 4.7 (2.1)

81.7

80.9

88.9

79.5

19.7 (10.6)

17.2 (8.9)

22.1 (9.0)

32.1 (11.9)

11.4 (6.8)

12.5 (7.7)

10.1 (5.9)

6.9 (6.3) 3.2 (2.8) 0.1 (1.2)

8.0 (6.8) 3.8 (2.9) 0.1 (1.4)

3.2 (2.2) 0.9 (0.7) 0.0

4.2 (3.9) 1.6 (0.8) 0.0

1.9 (2.9)

2.4 (3.2)

0.6 (0.8)

0.2 (0.4)

52.3

49.3

75.0

48.7

1.9 (1.9)

1.7 (1.0)

3.0 (3.9)

1.8 (1.1)

Note: Standard deviations in parentheses.

Most farmers in both locations grow tomatoes on single plots of land. Fifteen per cent of farmers in the Red River Delta owned two tomato plots compared with 20% in Lam Dong province. Most farmers grow tomatoes in a monocropping system but a few adopters in the Red River Delta intercrop tomatoes with melons, pumpkins, onions, garlic, lemongrass or beans. The open field system is the main tomato production system in both locations. All farmers in Lam Dong province use grafted seedlings while only a third in Red River Delta did at the time of the survey. These observed patterns of adoption of tomato grafting form the basis

An impact assessment of AVRDC's tomato grafting in Vietnam

for characterization of input and output data in the following sections of this report. The majority of farmers in the Red River Delta obtain seeds for non-grafted tomato production from private dealers/stockists or farmer groups (Table 9). Table 9. Average characteristics of the most recently harvested tomato field Characteristic Number of farmers (n) Average plot size (â&#x20AC;&#x2DC;000 m2)

Lam Dong province 224 3.3

Red River Delta 75 1.5

Production technology Sole stand (%) Open field system (%) Using plastic sheets/shades (%)

100.0 96.0 2.6

98. 7 100.0

Planting material Use grafted seedling (%) Use seeds (%) Use non-grafted seedling (%)

100.0 0.0 0.0

32.9 43.0 24.1

Source of planting material Private nursery operator (%) Own nursery (%) Private dealer/stockist (%) FAVRI (%) Farmer group (%) Other source (%)

83.0 15.6 1.3 0.0 0.0 0.0

7.5 13.8 31.3 25.0 17.5 5.0

Source: Survey conducted by AVRDC in collaboration with FAVRI and PVFRC (2012), n=299.

4.2.2. Adoption of tomato grafting The survey results show that adoption of grafted tomato commenced in 2002 in Lam Dong province and in 2007 in the Red River Delta. By 2012, the adoption rate was 100% in Lam Dong province and 45% in the Red River Delta (Fig. 3). A significant increase in the number of grafted tomato adopters was observed from 2002-2005 in Lam Dong province. In the delta, farmers only started using the grafting technique from 2009 onwards. Seven farmers continued to use non-grafted tomato until 2010/2011, even though they had started grafting one year or two years earlier (2009/2010).

AVRDC - The World Vegetable Center

100

% Adoption

Lam Dong province 60

20 Red River Delta 0 2000

2002

2004

2006

2008

2010

2012

Figure 3. Adoption of tomato grafting in Lam Dong province and the Red River Delta

The high adoption rate in Lam Dong province could be due to the influence of Dr. Vinhâ&#x20AC;&#x2122;s position as Regional Coordinator of the HRDP-IV project, which allowed him to prioritize tomato grafting in the research activities, and design training and information dissemination strategies for Lam Dong province. About 2,000 farmers from 20 provinces were trained in seedling production, in addition to lead farmers and staff from provincial and district extension centers and relevant agricultural institutions (AVRDC, 2008). Training courses included safe vegetable production, off-season vegetable production, vegetable greenhouse production, vegetable home gardens, organic fertilizer processing, grafted tomato production to control bacterial wilt, proper irrigation techniques, vegetable breeding, and setting up vegetable seed production enterprises. About 50 households in Tra Vinh province also received training and technical support for setting up greenhouse vegetable production while more than 80 vegetable nurseries were established in Lam Dong, Tra Vinh and Vinh Long provinces. According to Afari-Sefa (2012), farmers in the Red River Delta only use grafted seedlings during the off-season (May-October) when the incidence of soil-borne diseases is high. Most farmers typically plant non-grafted tomato seedlings in the dry season (November-April). Consequently, grafting in the Red River Delta is mostly beneficial in the off-season and offer benefits beyond disease control by increasing yield and helping to control other abiotic stresses,

An impact assessment of AVRDC's tomato grafting in Vietnam

excess moisture and soil temperature extremes as well as allowing the extension of the growing season. In Lam Dong province, the problem of bacterial wilt is experienced all year round due to high humidity; this explains why 89% of the farmers used grafting to address soil-borne diseases (Table 10). It may also be one reason why adoption in the south (Lam Dong province) started earlier than in the north (Red River Delta). Moreover, it was only in 2004-2005 that the grafting technology was transferred to 16 farmers in Dong Anh district near Hanoi (AVRDC, 2008) after field trials conducted at FAVRI under the SUSPER project implemented from 2002-2004 (Palada and Wu, 2005) showed very promising results. Table 10. Ranking of the three most important reasons for planting grafted tomato All Rank 1

% 88.6

Lam Dong province Rank % 1 92.4

52.9

57.2

58.3

60.1

65.8

39.3

Reason for adopting grafted tomato To protect from soil-borne diseases (caused by fungi, bacteria, or nematodes) and other diseasesa To increase tomato yield and not necessarily for disease or abiotic stress control Helping to control other abiotic stressors, excess moisture (waterlogging) and possibly soil temperature extremes as well as allowing the extension of the growing season.

Red River Delta Rank % 1 54.8

Notes: n=261. a Throughout the year for Lam Dong province and during the off-season period for the Red River Delta.

The majority of respondents in the Red River Delta sourced their grafted seedlings from FAVRI, which has an experimental station and production hub for seedling production in Moc Chau province in the highlands, about 130 km northwest of Hanoi. For farmers this can be a source of inconvenience compared with buying from a private nursery located within the Hanoi area, such as the Hanoi Seed Company. Several options are available to farmers in the Red River Delta to access and purchase grafted seedlings from FAVRI: (a) the farmer can agree to a seedling supply contract at the Hanoi office, (b) the farmer can place a purchase order with FAVRI via a local agricultural extension office or broker; or (c) the farmer can buy the seedlings from the FAVRI nursery in Moc Chau province. To better understand how adoption took place in the Red River Delta, five farmers were invited to participate in a focus group discussion in Tho Tang commune, Vinh Tuong district, Vinh Phuc province. One participant adopted grafted tomato in 2011, the rest in 2012. Several farmers benefited from training programs organized by the Ministry of Agriculture and Rural Development (MARD) on safe vegetable production, but none of the programs were

AVRDC - The World Vegetable Center

specifically about tomato grafting. Their knowledge of grafting came from a television broadcast on a channel that aired agricultural programs. It was only in 2010 when the agricultural provincial authority established a 722-m2 demonstration plot in their commune to highlight the benefits of grafting that these farmers learned about it at close proximity. Although initial results were unsuccessful due to the farmers' lack of experience in grafting and unfavorable hot weather conditions, Mr. Hoa, a lead farmer, adopted the technology in 2011. He had received grafted seedlings free of charge from Mr. Do Tat Chanh, head of the Agricultural Division of Vinh Tuong district. In 2012, all five farmers received 800 grafted seedlings and 80-90 non-grafted seedlings for a 360-m2 parcel of land, with sufficient supply for additional requests from the commune farmers. Mr. Hoa found that grafted tomatoes allowed tomato fruits to ripen on the plant, used less pesticides than non-grafted tomato, extended the harvesting period, and produced better fruit color and fruit appearance. Other farmers, basing their opinions and perceptions on the field demonstrations in 2010, also confirmed resistance of grafted tomato to bacterial wilt, late blight, and other pests and diseases in the off-season, which allowed them to plant early in the season, harvest for a longer period, obtain high yields and get a better market price. Additional benefits of planting grafted tomato included savings from applying less pesticide, and reducing the number of laborers needed for field operations. However, grafted tomato required more fertilizer and labor for staking and harvesting. Further research is needed to validate this perception. Adoption proved beneficial to Mr. Hoa; his income increased 5-7 times per unit area with yield reaching 138.9 t/ha. Factors that impeded the widespread adoption of grafting, according to him, were the limited supply of grafted seedlings and the higher investment cost affecting poorer farmers in Vinh Phuc province. In Lam Dong province, farmers only use one rootstock tomato variety, Vimina (or HW7996), which IAS tested for bacterial wilt resistance in 2002/2003 and released after evaluation from 2004-2011 (Ngo Quang Vinh, personal communication, November 7, 2012) (Table 11). Eggplant EG203, the main rootstock variety used in the Red River Delta from 2007-2011, is also resistant to bacterial wilt and other soil-borne diseases (Table 12). Both are AVRDCdeveloped varieties released by local partners in Vietnam.

An impact assessment of AVRDC's tomato grafting in Vietnam

Popular scion varieties are Anna (in Lam Dong province) and Savior (in the Red River Delta) due to their high yield performance, good appearance, popularity among consumers, and higher number of fruits harvested (Table 13). Table 11. List of rootstock and scion varieties released in Lam Dong province from 20042011 Year

Grafted tomato scions

Grafted tomato rootstocks

Districts

Area (ha)*

2004 2005 2006 2007 2008 2009 2010 2011

TG105 (386) TG105 (386) TG105 (386) Anna Anna Anna Anna Anna

Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996) Vimina (Hawaii 7996)

Duc Trong, Don Duong Duc Trong, Don Duong Don Duong, Duc Trong Don Duong, Duc Trong Don Duong, Duc Trong Don Duong, Duc Trong Don Duong, Duc Trong Don Duong, Duc Trong

100 500 2000 4000 5000 6500-7000 6500-7000 6500-7000

Source: Personal communication with Dr. Vinh (December 11, 2012). Note: Anna is a hybrid variety of Monsanto. (*) Estimated based on volume of seeds and rubber tubes sold.

AVRDC - The World Vegetable Center

Table 12. List of rootstock and scion varieties released in the Red River Delta from 20042011 Years 2004 2005 2006 2007

2008

2009

2010

2011

Grafted tomato scions HS902, P375 TN005, MonganT11 HS902, P375 TN005, VL2910, HS902, P375 TN005, VL2910, Savior VL2910, FM29 P375 P375, VL642, VL3500 Savior Savior PP 89 Emural, DV2962 Savior Savior Savior, P375, VL3500 Savior, GS901 Emural PP 89, Magic Savior, DV2962 VL642 GS901 DV2962, Gadeeva Savior, Mongan T11 VL2200 Savior Savior, VL3500 Savior Savior Savior, Mongan T11 Savior VL3500 Savior Savior Savior, Anna Savior

Grafted tomato rootstocks EG203, EG219, HW7996 EG203, EG219, HW7996 EG203, EG219, HW7996 EG203, EG219, HW7996 EG203, HW7996 EG203, HW7996 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203 EG203, Hawaii EG203

Provinces Ha Noi Hai Phong Ha Noi Hai Phong Ha Noi Hai Phong Bac Ninh Hai Phong Ha Noi Ha Noi Bac Ninh Vnh Phuc Nam Dinh Thai Binh Hoa Binh Hai Phong Ha Noi Bac Ninh Vinh Phuc Nam Dinh Thai Binh Ninh Binh Phu Tho Hung Yen Hai Duong Hoa Binh Hai Phong Ha Noi Bac Ninh Vinh Phuc Hai Duong Other provinces Ha Noi Bac Ninh Vinh Phuc Hai Duong Other provinces

Areas (ha) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 4.0 2.0 0.5 2.0 1.0 1.0 2.0 6.0 3.0 0.5 1.0 1.0 1.0 2.0 2.0 1.0 3.0 1.0 6.0 3.0 8.0 3.0 1.0 8.0 3.0 8.0 3.0

Source: Personal communication with Dr. To Thi Ha from FAVRI (October 10, 2012). “-” no information.

An impact assessment of AVRDC's tomato grafting in Vietnam

Table 13. Main rootstock and scion varieties used by farmers, and reasons for selection. Reason for adoption Rootstock Resistant to bacterial wilt Resistant to other soil-borne diseases Other (high yield, high producer price, good appearance, longer shelf-life, and others) . Scion High yield Market preference Resistant to bacterial wilt Higher number of fruits High producer price Longer harvesting period Good appearance Preferred color Thick flesh Good taste Long shelf-life (freshness) Other

Lam Dong province N % Vimina (n=147) 104 70.7 39 26.5 4 2.8

Red River Delta* N % EG203 (n=32) 10 31.3 17 53.1 22 68.7

Anna (n=219) 107 48.9 81 37.0 15 6.8 5 2.3 4 1.8 3 1.4 1 0.5 1 0.5 1 0.5

Savior (n=26) 18 69.2 8 30.8 3 11.5 8 30.8 7 26.9 3 11.5 19 73.1 9 34.6 2 7.7 1 3.8 5 19.2 3 11.5

0.5

Source: Survey conducted by AVRDC in collaboration with FAVRI and PVFRC (2012), n=261. Note: (*) For Red River Delta, multiple-response question.

In Lam Dong province, farmers purchased almost all seedlings for rootstocks and scions from specialized nursery operators, and only about 14-16% from farmer-managed nurseries (Table 14). In the Red River Delta, about 61-67% of EG203 and Savior came from FAVRI and 1519% from farmer nurseries, with the rest sourced from farmer groups and specialized nurseries. Specialized nurseries are still not as common in the Red River Delta as in Lam Dong province. Table 14. Seed source of main rootstock and scion varieties Source

Lam Dong province

Red River Delta*

Rootstock FAVRI Specialized nurseries Farmer nurseries Farmer group

Vimina (n=161) 86.3 13.7 -

EG203 (n=33) 60.6 12.1 15.2 12.1

Scion FAVRI Specialized nurseries Farmer nurseries Farmer group nursery

Anna (n=216) 83.8 16.2 -

Savior (n=27) 66.7 18.5 14.8

Source: Survey conducted by AVRDC in collaboration with FAVRI and PVFRC (2012), n=261.

AVRDC - The World Vegetable Center

(A) Mr. Linh, a nursery operator in Don Duong district, Lam Dong province using Hawaii 02 (or HW7996 from AVRDC)

(B) A farmer showing her small nursery of nongrafted tomato in Nghia Hung district, Nam Dinh province, Red River Delta

(D) Dr. Ngo Quang Vinh looking at a field of grafted tomatoes in Dâ&#x20AC;&#x2122;Ran commune, Don Duong district, Lam Dong province

Figure 4. Tomato production in Lam Dong province, Vietnam

In Lam Dong province, the main planting months are from January to April but some farmers also plant tomatoes in May or June and in August or September (Figure 5). The harvest starts in March and peaks in May. In the Red River Delta, farmers using grafted tomato seedlings are able to plant tomatoes as early as June-July and harvest as early as September.

An impact assessment of AVRDC's tomato grafting in Vietnam

Lam Dong province, adopter

100

Planting Harvesting

Dec

Nov

Oct

Sep

Aug

Jul

Jun

May

Apr

Jan

Mar

Feb

No. of farmers

150

Management (bet. planting and harvesting)

Source: Survey conducted by AVRDC in collaboration with PVFRC (2012), n=225

Red River Delta, adopter

Red River Delta, non-adopter

No. of farmers

25 20 15 10

Planting Harvesting

Nov Dec

Sep Oct

Jul Aug

May Jun

Mar Apr

Jan Feb

Dec

Oct Nov

Aug Sep

Jun Jul

Apr May

Jan

Feb Mar

Management (bet. planting and harvesting)

Source: Survey conducted by AVRDC in collaboration with FAVRI (2012), n=75

Figure 5. Seasonal calendar of production activities by location

4.2.3. Yield and yield stability of grafted tomato During the main harvesting months, yield of grafted tomato ranged between 70.5-73.8 t/ha in Lam Dong province and 62.9-87.3 t/ha in the Red River Delta (Table 15). Yield for grafted tomato was at its lowest in October at 62.9 t/ha in the Red River Delta. Non-grafted tomato yield was highest in April at 62.3 t/ha. The coefficient of variation (CV) showed a more stable

AVRDC - The World Vegetable Center

yield from April to August for grafted tomato compared to non-grafted tomato, with CV ranging from 0.30-0.62 in the Red River Delta (Table 16). Table 15. Yield of grafted versus non-grafted tomatoes by month, 2010-2011 (t/ha) Month harvested

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Yield (production) Lam Dong Red River Delta province Grafted Grafted Non-grafted (n=65) (n=223) (n=25) 46.8 (32.4)

62.3 (19.1)

72.6 (14.4) 73.8 (14.7) 72.9 (19.8) 72.3 (20.3) 70.5 (21.4) 78.8 (39.5) 62.9 (49.5) 87.3 (28.6)

Yield (quantity sold) Lam Dong Red River Delta province Grafted Grafted Non-grafted (n=223) (n=25) (n=65) 35.5 (22.9)

70.4 (14.9) 71.7 (14.9) 70.6 (19.1) 70.2 (20.0) 68.6 (20.7)

50.7 (17.5)

71.4 (38.5) 56.8 (46.7) 83.2 (28.4)

44.2 (27.3) 50.5 (24.3)

36.4 (26.6) 44.8 (24.2)

Note: n=300. Standard deviations in parentheses.

Table 16. Coefficient of variation of grafted vs non-grafted tomatoes by month, 2010-2012 Month harvested

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Yield (production) Lam Dong Red River Delta province Grafted Grafted Non-grafted (n=65) (n=223) (n=25) 0.69

0.30

0.20 0.20 0.27 0.28 0.30 0.50 0.79 0.33

Yield (quantity sold) Lam Dong Red River Delta province Grafted Grafted Non-grafted (n=223) (n=25) (n=65) 0.64

0.21 0.21 0.27 0.28 0.30

0.62 0.48

0.34

0.54 0.82 0.34

0.73 0.54

Source: Field survey conducted by AVRDC in collaboration with FAVRI and PVFRC (2012), n=300.

Farmers were asked about the changes in their cultural practices following the adoption of grafting for the same piece of land (per 1,000 m2). The purpose was to assess if farmers perceived a significant change in their production cycle by way of input costs, productivity in terms of yield per hectare, demand and sales volume, and contribution of grafted tomato

An impact assessment of AVRDC's tomato grafting in Vietnam

production and sales to total household income. While some farmers provided estimates of changes in their production practices involving grafting and non-grafting techniques for the selected variables, others had difficulty recalling their experiences encountered as far back as up to 10 years ago. We therefore divided the responses into four categories: higher, lower, no change, or cannot tell for missing cases (Table 17).

In Lam Dong province, most farmers (>90%) perceived that grafting increased their yieldâ&#x20AC;&#x201D; possibly due to the cropâ&#x20AC;&#x2122;s tolerance to flooding, which also lessened their dependence on herbicides for weed control (Table 20). Nearly 80% thought that the number of tomato production cycles per year remained unaffected by the adoption of grafting, while over 90% thought that grafting increased the total cost of production per unit area. While the majority of farmers from this cohort reported an increase in both the quantity and frequency of pesticide application due to the adoption of grafting, about 30% of them experienced a decrease in the time spent on spraying pesticides.

AVRDC - The World Vegetable Center

Table 17. Farmerâ&#x20AC;&#x2122;s perceptions of changes in selected production practices with the adoption of tomato grafting Lam Dong province N Decreased No change % % 117 35.9 6.8

Increased

Red River Delta N Decreased

% 57.3

Number of production cycles per year

130

76.9

23.1

Cost per seed/seedling produced/bought

131

0.8

Quantity of pesticide applied

128

17.2

Frequency of pesticide application (spraying)

129

Amount spent for inorganic fertilizers

% 86.7

No change % -

% 13.3

18.7

81.3

99.2

100.0

4.7

78.1

42.3

19.2

38.5

33.3

5.4

61.2

37.5

18.7

43.8

131

4.6

1.5

93.9

11.5

15.4

73.1

Amount spent for organic manure

118

17.8

16.9

65.2

7.7

30.8

61.5

Amount spent on mulching

2.7

4.1

93.1

52.2

47.8

Amount spent on weed control

124

94.3

0.8

4.8

4.5

36.4

59.1

Amount spent on irrigation/watering

119

39.5

8.4

52.1

41.7

58.3

Tolerance to flooding

129

0.8

3.9

95.3

3.8

11.5

84.6

Miscellaneous costs (grafting clips, rootstock, seed, etc.)

12.1

15.5

72.4

9.5

33.3

57.1

Total labor cost for other activities

13.5

44.2

42.3

8.3

12.5

79.2

Total cost of production

120

1.7

0.8

97.5

8.3

91.7

Yield per hectare (estimate)

127

1.6

98.4

8.0

4.0

88.0

Demand and sales volume Contribution to total household income

125

100.0

20.0

80.0

122

4.1

1.6

94.3

6.9

93.1

Variable Land area

Increased

An impact assessment of AVRDC's tomato grafting in Vietnam

A slightly different picture emerged for the Red River Delta. While the majority (>90%) reported an increase in their costs for seedlings and total production costs, about 30-50% of the farmers experienced no significant change in the costs for mulching, irrigation, herbicides, other miscellaneous items, and organic manure per unit area. More farmers reported that they decreased their pesticide use after adopting grafting in the Red River Delta than in Lam Dong province. The majority of respondents also reported higher yields, and higher demand and sales for their tomato crop, which contributed to an increase in the household income. 4.2.4. Production function analysis We used a Cobb-Douglas production function to assess the effect of grafting on tomato yield. Ideally, a profit function should be used to estimate the effect of grafting not just on crop yield but also on net profit. However, a profit function requires detailed data on input and output prices and because our study area was relatively small and relied on a single cross-sectional data set, there was not enough inter-household variation in input prices to estimate such a function. The general expression of the Cobb-Douglass functional form is: (1) where Y is total production (endogenous or dependent variable); L is labor input; K is capital input (L and K are the exogenous or independent variables); A is the total factor productivity; and α and β are the output elasticities of capital and labor, respectively. The model can be linearized as:

(2)

where Yi denotes the yield of the ith farmer, Xij the vector of jth input used in the ith farm in the production process, αi represents coefficients of inputs which are estimated from the model (α is a constant term), and ei is the error term of the ith farm. The αj is the set of parameters to be estimated that reflect the impact of change on yield given a change in the levels of each input, ceteris paribus. This implies an ideal division of yield due to each factor input of production (e.g. seed, fertilizer, labor).

AVRDC - The World Vegetable Center

Estimating the Cobb-Douglas function also allows testing for endogeneity between fungicide use and the use of grafting (Mutuc, Rejesus and Yorobe, 2011). Because fungicide typically is applied in response to fungus attacks such as Fusarium wilt (i.e. disease severity that is unobserved by the analyst), it is possible that the residuals of the production function are correlated with fungicide cost. Hence, the fungicide cost variable can be endogenous in this case and this would cause inconsistent parameter estimates in the production function. One approach in dealing with this is simply to use a disease severity variable that would eliminate the unobservability of disease/fungi shocks that can potentially cause bias. To check for endogeneity, exogenous variables that are correlated (e.g., gender and membership to farm organization) with fungicide cost but not with disease severity were determined. A reduced form model in which all exogenous variables are included was then estimated. The residual from the ordinary least squares (OLS) estimation was then added to the production function in Equation 2 as an additional regressor (Table 18). If the residual is statistically insignificant, then endogeneity is not severe and the actual fungicide cost variable could still be used in an OLS estimation of Equation 2. Table 18. Variables used in the ordinary least squares (OLS) regression Variables LNYIELD LNSEED LNMA_IFE LNFUNG LNINSE LNMULC LNIRRI LNSTAK LNAG_OTH LNTOMDHA GT_NGT LOC_N MODE S1_A4_SEX S1_A6_ORG

Description Yield level of the ith farmer Natural log of seed/seedling expenditures Natural log of manure and inorganic fertilizer expenditures Natural log of fungicide expenditure Natural log of insecticide expenditure Natural log of mulching expenditure Natural log of irrigation expenditure Natural log of staking expenditure Natural log of other expenditures Natural log of labor Dummy variable: =1 if grafted, =0 non-grafted Dummy variable: =1 if Lam Dong province (north), =0 Red River Delta (north) Binary pest severity variable: =1 if less severe, =0 otherwise Dummy variable: =1 if male, =0 female Dummy variable: =1 if yes, =0 no

Tomato yield (dependent variable) was assumed to be a function of seed, manure and inorganic fertilizer, fungicide, insecticide, mulching, irrigation, staking, other input costs, labor and three dummy variables (the use of grafting, regional difference, and pest/disease severity) as independent variables. A dummy variable for the use of grafting is included to evaluate the impact of tomato grafting. A location dummy variable was furthermore included to capture the

An impact assessment of AVRDC's tomato grafting in Vietnam

yield variation between the agro-climatic conditions in the two different locations. A pest/disease severity dummy variable was included because previous studies have shown that insecticide/fungicide productivity is underestimated if pest/disease severity is not specified in the production function (Norwood and Marra, 2003). The binary pest/disease severity variable was measured in terms of farmersâ&#x20AC;&#x2122; estimation of the severity of most critical pests/diseases for the most recently harvested produce in percentage terms (i.e., 1= <25%, 2=25-50%, 3=50-75%, 4=76-90% and 5=>90%). As this is a multiple-response (multiple entries of pest/disease identified) variable, the binary pest/disease severity was created by getting the mode per farmer per non-grafted and grafted tomato, wherein it is equal to one when the farmerâ&#x20AC;&#x2122;s response is less severe (<25%) and zero otherwise. In cases of multiple or zero modes, we used the maximum level of pest/disease severity. We used various regression diagnostics to verify that the data did not violate any of the assumptions underlying the OLS regression. Outliers and extreme values were excluded in the model using Cookâ&#x20AC;&#x2122;s D conventional cutoff point of 4/306 (4/n). The model passed tests for linearity, normality of residuals (Shapiro-Wilk W test), homogeneity of variance of the residual (Breusch-Pagan / Cook-Weisberg test for heteroskedasticity, and using robust standard error), multicollinearity (variance inflation factor), and model specification (Ramsey RESET test for omitted variables). The Hausman test suggested that fungicide endogeneity might not be severe enough in the context of our data set to cause significant bias in the results [F(1, 292) = 0.38; Prob > F =0.5378]. We therefore continued using the actual fungicide cost rather than its instrumented values. Three alternative models were estimated for both locations combined (Model 1), and separately for Lam Dong province (Model 2) and the Red River Delta (Model 3). Table 19 shows the results. The coefficients for seedlings, fungicide, insecticide, mulching, labor (person day per hectare [MD/ha]) and grafting are highly significant (p<0.01) for Model 1. A 100% increase in input use would increase yield by 4% for fungicides, 2% for insecticides, 1% for mulching, 9% for labor and 30% for the use of grafting. The substantial effect of grafting is also confirmed by a simple comparison of the sample means of grafted (73.3 t/ha) and non-grafted (56.5 t/ha) tomato in Table 20. Increasing seedling expenditures by 100% would result to a yield reduction of 1%. The model shows a decreasing returns to scale of 0.403 (p<0.01) as suggested by the sum of regression coefficients.

AVRDC - The World Vegetable Center

Although location was not significant in Model 1, differences between the two locations were observed by comparing Models 2 and 3. In Lam Dong province (Model 2), all observations involved grafted tomato and hence the grafting dummy could not be included. Seed and insecticide costs were found to be highly significant (p<1%). A 100% increase in seed expenditures corresponded to a 28% increase in yield of grafted tomato. The effect of insecticide use on yield was small at only 1.5% change for a 100% change in insecticide use. In the Red River Delta (Model 3), the impact of grafting on yield was highly significant, as was the expenditure on seedlings, fungicides, insecticide, mulching, and the use of labor. Controlling for all other factors, the use of grafted seedlings led to a 31% increase in yield (p<0.01). However, since grafting was relatively new in Red River Delta, a 100% change in seedling expenditures would lead to a 1.3% reduction in yield (p<0.01). This could be due to a number of reasons, including low grafted seedling survival rate due to laborersâ&#x20AC;&#x2122; lack of grafting experience; poor handling of grafted seedlings after purchase and prior to transplanting; and improper field management practices such as graft joints planted below the soil or farmersâ&#x20AC;&#x2122; failure to remove suckers, which defeat the purpose of grafting. A 100% increase in the use of other inputs in the specified model resulted in a significant (p<0.05) increase in yield of 4% for fungicides, 2% for insecticides, 1% for mulching and 6% for labor. The results from Models 1 and 3 suggest that the use of grafting provides statistically significant yield improvement over the use of non-grafted seedlings. The estimates also suggest that the marginal effect of fungicides, insecticides, and mulching on tomato yields is lower for grafted than for non-grafted tomato. This is consistent with the idea that applying fungicide, insecticide and mulching is redundant when a technology that inherently controls for bacterial wilt, nematodes and other soil-borne diseases is already being used; grafting therefore substitutes for fungicide use as it controls Fusarium wilt. Table 20 shows the sample means of all exogenous variables used in the estimations. One of the assumptions prior to model specification was that the sample means for fungicide and insecticide would be lower for grafting compared with non-grafting. However, this was not the case. The large standard deviations of each variable for both grafted and non-grafted seedlings, especially for Red River Delta, may explain these unexpected results compounded by an imbalanced and relatively smaller sample size. Seedling and mulching costs were higher for grafted tomato as expected. Seed costs amounted to 38.6 million VND/ha with grafted tomato

An impact assessment of AVRDC's tomato grafting in Vietnam

accounting for 20% of the total cost of production compared to only 4.5 million VND/ha for non-grafted. The share is lower at 15% when we aggregate all adopters as a whole. Table 19. Econometric estimation results of production inputs on yield Dependent variable: LNYIELD Variable

Pooled (Model 1)

LNSEED LNMA_IFE LNFUNG LNINSE LNMULC LNIRRI LNSTAK LNAG_OTH LNTOMDHA GT_NGT LOC_N MODE _CONS

Coef. -0.011 0.004 0.036 0.016 0.009 0.002 0.000 0.004 0.086 0.301 -0.114 0.070 2.612

Number of cases F-ratio Prob > F R-squared

279 23.09 0.000 0.433

SE 0.004 0.007 0.009 0.005 0.003 0.002 0.003 0.002 0.014 0.077 0.074 0.040 0.218

Sig. ** ns *** ** ** ns ns ns *** *** ns ns ***

Lam Dong province (Model 2) Coef. SE Sig. 0.277 0.051 *** 0.066 0.037 ns 0.013 0.011 ns 0.015 0.005 ** 0.002 0.004 ns -4.360 0.002 ns 0.001 0.002 ns 0.004 0.003 ns 0.048 0.042 ns

Red River Delta (Model 3) Coef. SE Sig. -0.013 0.005 ** -0.005 0.011 ns 0.037 0.014 * 0.018 0.008 * 0.012 0.005 * -0.009 0.009 ns -0.001 0.011 ns 0.001 0.006 ns 0.060 0.025 * 0.307 0.084 **

0.104 -2.414

0.025 2.972

0.053 0.880

ns **

215 12.14 0.000 0.270

0.070 0.318

ns ***

64 10.61 0.000 0.692

Note: ***,**,* means significant at p<0.000, p<0.01, and p<0.05, respectively.

Table 20. Sample means of inputs and labor between grafted and non-grafted tomato production by location, 2011/2012 (in million VND/ha) Variable

Yield (t/ha) Seed/seedling cost Manure and inorganic fertilizer Fungicide Insecticide Mulching Irrigation Staking Other inputs Labor (person-days/ha)

Lam Dong province (n=215 obs) Mean SD % to Total 73.3 17.1 19.8 4.1 14.4 32.8 15.9 23.9

Red River Delta (grafted) (n= 16 obs) Mean SD % to Total 81.4 28.4 38.6 24.5 20.4 42.7 49.3 22.5

Red River Delta (nongrafted) (n=48 obs) Mean SD % to Total 56.5 18.2 4.5 5.3 4.2 28.5 70.7 26.8

14.7 5.1 7.8 3.3 18.9 19.3 331.1

8.1 8.7 5.4 10.6 11.4 220.9 138.4

10.7 3.7 5.7 2.4 13.8 14.1 -

Labor costs Total input costs

15.3 137.0

7.6 222.8

11.2 100.0

15.5 7.4 8.1 5.2 48.2 10.4 1,314. 2 13.6 189.6

Note: n=279. “-” not applicable.

AVRDC - The World Vegetable Center

19.2 12.8 7.7 20.8 20.4 36.2 633.1

8.2 3.9 4.3 2.7 25.4 5.5 -

10.4 2.2 2.5 1.1 42.7 1.2 896.0

10.8 2.6 4.5 5.4 113.3 2.0 519.6

9.7 2.1 2.3 1.0 40.1 1.2 -

7.7 92.1

7.2 100.0

13.4 106.6

10.6 139.7

12.6 100.0

4.2.5. Effect of tomato grafting on profit and household income Profitability analysis was done for both grafted and non-grafted data in the Red River Delta only (Table 21 & 22), given the 100% adoption rate in Lam Dong province. Total sales amounted to 648 million VND/ha for farmers using grafted tomato but only 238.8 million VND/ha for farmers who used non-grafted tomato (p<0.01). Yet the total cost, including labor cost, was significantly higher for grafted tomato than for non-grafted tomato. Nevertheless, because of the significantly higher revenues, farmers using grafted tomato obtained profits that were 326.1 million VND/ha higher. Based on these results, the benefit-cost ratio was 4.6 for grafted tomato compared with 3.5 for non-grafted tomato. Further research would be useful to generalize these results for the whole of the Red River Delta, given the small sample size for the non-adopter group. Table 21. Profitability of grafted versus non-grafted tomato production in the Red River Delta, 2011-2012 (in million VND/ha) Variable Yield (t/ha) Tomato price (VND kg) Sales Input cost Labor cost Total cost Profit Benefit-cost ratio

Grafted (n=16) Mean SD 81.4 28.4 8,447.8 2,392.7 648.0 324.7 176.0 89.3 13.6 7.7 189.6 92.1 458.4 284.5 4.6 4.8

Non-grafted (n=48) Mean SD 56.5 18.2 5,162.8 1,960.7 238.8 118.1 93.1 138.3 13.4 10.6 106.6 139.7 132.3 188.4 3.5 2.5

Sig.

p-value

** *** *** ** ns ** *** ns

0.0039 0.0001 0.0001 0.0086 0.9257 0.0098 0.0004 0.3967

Notes: n=64. T-test with unequal variance used.

Table 22. Estimated effect of tomato grafting on net profits in Lam Dong province Item

Lam Dong province

Adoption rate Yield (t/ha) Tomato price (VND/kg) Sales (in million VND/ha) Total cost (in million VND/ha) Total profit (in million VND/ha) Total area under tomato production (ha, 2011)a Total profit from grafting (million US$)b

100% 73.3 3,867.0 272.0 137.0 135.0 6,388.0 41.7

Notes: n=225 a Agroâ&#x20AC;&#x201C;Forestry Department, General Statistics Office, Vietnam. b Foreign exchange rate in August 2012 average (VND/US$) = 20,703.4

Based on average revenues of 272 million VND/ha, average production costs of 137 million VND/ha, and an estimated tomato area of 6,388 ha, total profit was higher by US$ 41.7 million

An impact assessment of AVRDC's tomato grafting in Vietnam

compared to the situation if the same land area had been cultivated with non-grafted tomatoes. However, this is not the net economic effect of grafting, because it is likely that the large increase in tomato output would affect the average tomato price (which reduces the economic impact) while the profitability of tomato production would have led to an expansion of the tomato area (which would have increased the economic impact). Moreover, it excludes the employment effect of tomato grafting on nursery operations, which might be considerable as seedling production is labor-intensive.

5. CONCLUSIONS AND RECOMMENDATIONS AVRDC introduced tomato grafting to Vietnam in 2002 through the training of Vietnamese scientists who then took a prominent role in disseminating the technology locally. In Lam Dong province in southern Vietnam, Dr. Ngo Quang Vinh, who was Regional Director of the AVRDC-ARC HRDP project, organized numerous training and extension activities that facilitated the adoption process. Dr. Vinh trained about 2,000 staff and extension workers and farmers in Lam Dong province and contributed to the construction and development of more than 80 vegetable nursery farms in Lam Dong, Tra Vinh and Vinh Long. Because of these efforts as well as the profitability of the technique to farmers, the adoption of tomato grafting by farmers in Lam Dong province rapidly increased since its introduction in 2002 to 100% in 2012. It is likely that the grafting technique led to a substantial expansion of the tomato area in the province. The case was different in the Red River Delta, where bacterial wilt affects tomato production only in the wet season and flooding is a major production constraint. Nurseries in Lam Dong province used tomato rootstock variety Vimina, while in Red River Delta eggplant variety EG203 was used as a rootstock due to its tolerance to flooding. Dissemination efforts in the Red River Delta were led by FAVRI; being primarily a research organization, its dissemination efforts were not as extensive as those in Lam Dong province. Farmer participation in field trials in the Red River Delta, for example, only started in 2007. Specialized private nurseries and individual farmer-managed nurseries have emerged in Lam Dong province, generating local employment opportunities, but have not emerged to a similar extent in the Red River Delta. In both locations, the average yield from grafted tomato plants was substantially higher than that of non-grafted plants. Results of the production function analyses suggest that while

AVRDC - The World Vegetable Center

controlling for all other production factors, the use of grafted seedlings increases the average tomato yield by about 30%. This result appeared when pooling the data for the two locations (while adding a locational dummy, but also when estimating the effect separately for the Red River Delta. Comparing the coefficient of variation, it appeared that the yield of grafted tomato was more stable in Lam Dong province than in the Red River Delta. As a result of the significantly higher average yields, the average revenues were significantly higher for farmers using grafted tomato seedlings. The increase in revenues was large enough to offset the significantly higher production costs. Grafted tomatoes had a benefit-cost ratio of 4.6, which compares to a ratio of 3.5 for non-grafted tomato. The farm benefits of tomato grafting can be achieved only if soil-borne disease or flooding constrains tomato production, as grafted seedlings are more costly and have no other yield benefit over non-grafted seedlings, as shown by several studies conducted by AVRDC in Taiwan and Vietnam. Tomato grafting is therefore not suitable for all farmers. The extent of soil-borne disease and abiotic stresses should be investigated first before promoting tomato grafting in a particular location. We also note that the disease resistance of current rootstock varieties may break down, and that alternative rootstocks will need to be selected. This study showed the critical role played by certain change agents, such as extension services and entrepreneurial private investors, in the adoption and spread of tomato grafting, and also highlighted the importance of private nurseries that developed in response to new market opportunities. In the case of Lam Dong province, the dissemination of tomato grafting can largely be credited to a single person. Both examples provide important lessons for the future dissemination of AVRDC technologies.

6. REFERENCES ADB. (2005). ADB RETA No.6011. Collaborative Vegetable Research Network for Cambodia, Lao PDR, and Vietnam (CLVNET) Phase II April 2002â&#x20AC;&#x201C;October 2005. Progress Completion Report. Afari-Sefa, V. (2012). Vietnam Grafting Success Story: Unpublished Report on Preliminary Field Visit to Vietnam to Develop Protocols for Ex-post Impact Assessment Study of AVRDCâ&#x20AC;&#x2122;s Tomato Grafting Technology Reporting Period: [April 25, 2012] to [May 5, 2012]. Arusha, Tanzania: AVRDC - The World Vegetable Center: Regional Center for Africa. Aganon, C. P., Mateo, L. G., Cacho, D., Bala, A., Jr., and Aganon, T. M. (2002). Enhancing off-season production through grafted tomato technology. Philippine Journal of Crop Science, 27(2), 3-8.

An impact assessment of AVRDC's tomato grafting in Vietnam

Attanayaka, A., Perera, M., Chen, J., and Hanson, P. (1997). Effects of different grafting techniques on the success of tomato/eggplant rootstocks combination. Shanhua, Taiwan: Asian Vegetable Research and Development Center. AVRDC. (1994). AVRDC 1993 Progress Report. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, pp.298-301, 318-327. AVRDC. (1995). AVRDC 1994 Progress Report. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, pp.257-265. AVRDC. (2000a). AVRDC Report 1999. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, pp.44-46, 61-63, 109-110. AVRDC. (2000b). Vegetable Research in South Asia: Proceedings of the South Asia Vegetable Research Network (SAVERNET-II) Mid-term Review Meeting, 5-9 February 1999, AVRDC, Taiwan. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, p.52. AVRDC. (2001). AVRDC Report 2000. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, pp.98, 106. AVRDC. (2002). AVRDC Report 2001. Shanhua, Tainan, Taiwan: Asian Vegetable Research and Development Center, pp.59-62, 99, 138. AVRDC. (2003). AVRDC Report 2002. Shanhua, Taiwan: AVRDC - The World Vegetable Center, pp.142-143. AVRDC. (2004). AVRDC Report 2003. Shanhua, Taiwan: AVRDC - The World Vegetable Center, p.109. AVRDC. (2008). HRDP IV: Final Report of the SDC-funded Human Resource Development Project for the Mekong Region Phase IV (AVRDC Publication No 08-706 ed.). Bangkok: AVRDC-The World Vegetable Center Asian Regional Center, pp. 33-34, 41-42, 55, 64. Baumuller, H. (2012). Facilitating agricultural technology adoption among the poor: The role of service delivery through mobile phones. Bonn: Center for Development Research, University of Bonn. Black, L., Palada, M., & Roan, Y. (2002). Yield of grafted and non-grafted tomato under plastic rain shelters during the hot-wet season in Taiwan. Paper presented at the 30th National Agricultural Plastics Congress, San Diego, California, USA. Black, L., Wu, D., Wang, J., Kalb, T., Abbass, D., and Chen, J. (2003). International Cooperatorsâ&#x20AC;&#x2122; Guide: Grafting tomatoes for production in the hot-wet season. Asian Vegetable Research and Development Center. AVRDC Publication No. 03-551. p. 6.Burleigh, J. R., Black, L. L., Mateo, L. G., Cacho, D., Aganon, C. P., Boncato, T., Arida, I.A., Ulrichs, C., and Ledesma, D. R. (2005). Performance of grafted tomato in Central Luzon, Philippines: A case study on the introduction of a new technology among resourcelimited farmers. Online. Crop Management. DOI:10.1094/CM-2005-0701-01-MG. Central Population and Housing Census Steering Committee. (2010). The 2009 Vietnam population and housing census: Completed results. Hanoi: Department of Population and Labor Statistics, the General Statistics Office. Chadha, M. L. (2010). Vegetable research and development programs to alleviate poverty and malnutrition in South Asia. AVRDC - The World Vegetable Center. pp. 21. Dau, V. T., Dang, N. V., Nguyen, D. H., Pham, L. T., Le, T. T. M., Phan, H. T., and Burgess, L. W. (2009). A simplified technique for grafting watermelon onto resistant cucurbit rootstocks for control of Fusarium wilt of watermelon in Nghe An Province, Vietnam. Australasian Plant Disease Notes, 4(1), 114-116. Davis, A. R., Perkins-Veazie, P., Hassell, R., Levi, A., King, S. R., and Zhang, X. (2008). Grafting Effects on Vegetable Quality. HortScience, 43(6), 1670-1672. De la PeĂąa, R., & Hughes, J. (2007). Improving vegetable productivity in a variable and changing climate. Journal of SAT Agricultural Research, 4(1), 1-22. Doan, T., and Nguyen, T. (2005). Status of research on biological control of tomato and groundnut bacterial wilt in Vietnam. Paper presented at the Mitteilungen Aus Der Biologischen Bundesanstalt Fur LandUnd Forstwirtschaft Berlin-Dahlem, 23-26 October 2005, Seeheim/Darmstandt, Germany. Dung, D. T. (1997). Bacterial wilt of some upland crops in Hanoi and its surrounding areas, 2nd International Bacterial Wilt Symposium. Guadeloupe, French West Indies.

AVRDC - The World Vegetable Center

Hanson, P. M., Wang, J.-F., Licardo, O., Hanudin, Mah, S. Y., Hartman, G. L., Lin, Y.-C., and Chen, J.-t. (1996). Variable Reaction of Tomato Lines to Bacterial Wilt Evaluated at Several Locations in Southeast Asia. HortScience, 31(1), 143-146. Hayward, A., Allen, C., and Prior, P. (2005). Research on bacterial wilt: a perspective on international linkages and access to the literature. Bacterial wilt disease and the Ralstonia solanacearum species complex, pp. 1-8. IFPRI. (2002). Fruits and vegetables in Vietnam: Adding value from farmer to consumer. Washington, DC: IFPRI. Khah, E., Kakava, E., Mavromatis, A., Chachalis, D., and Goulas, C. (2006). Effect of grafting on growth and yield of tomato (Lycopersicon esculentum Mill.) in greenhouse and open-field. J. Appl. Hortic, 8(1), 3-7. Lee, J.-M. (1994). Cultivation of Grafted Vegetables I. Current Status, Grafting Methods, and Benefits. HortScience, 29(4), 235-239. Lee, J.-M., Kubota, C., Tsao, S. J., Bie, Z., Echevarria, P. H., Morra, L., and Oda, M. (2010). Current status of vegetable grafting: Diffusion, grafting techniques, automation. Scientia Horticulturae, 127(2), 93-105. Lin, C. H., Hsu, S. T., Tzeng, K. C., and Wang, J. F. (2008). Application of a preliminary screen to select locally adapted resistant rootstock and soil amendment for integrated management of tomato bacterial wilt in Taiwan. Plant Disease, 92(6), 909-916. Louws, F. J., Rivard, C. L., & Kubota, C. (2010). Grafting fruiting vegetables to manage soil-borne pathogens, foliar pathogens, arthropods and weeds. Scientia Horticulturae, 127(2), 127-146. Martorana, M., Giuffrida, F., Leonardi, C., & Kaya, S. (2006). Influence of rootstock on tomato response to salinity. Paper presented at the VIII International Symposium on Protected Cultivation in Mild Winter Climates: Advances in Soil and Soilless Cultivation. No. 747. Midmore, D., Roan, Y., and Wu, M. (1997). Management practices to improve lowland subtropical summer tomato production: yields, economic returns and risk. Experimental Agriculture, 33(02), 125-137. Mohamed, F., El-Hamed, K., Elwan, M., and Hussien, M.-A. (2012). Impact of Grafting on Watermelon Growth, Fruit Yield and Quality, Volume 76, p. 99. Mudge, K., Janick, J., Scofield, S., and Goldschmidt, E. E. (2009). A history of grafting. Horticultural Reviews, Volume 35, pp. 437-493. Mutuc, M.E., Rejesus, R.M., & Yorobe, Jr., J.M. (2011). Yields, insecticide productivity, and Bt corn: Evidence from damage abatement models in the Philippines. AgBioForum, 14(2), 35-46. Available on http://www.agbioforum.org/v14n2/v14n2a01-mutuc.pdf. Nguyen, M., and Ranamukhaarachchi, S. (2010). Soil-borne antagonists for biological control of bacterial wilt disease caused by Ralstonia solanacearum in tomato and pepper. Journal of Plant Pathology, 92(2), 395-406.Norwood, F., & Marra, M. (2003). Pesticide productivity: Of bugs and biases. Journal of Agricultural and Resource Economics, 28(3), 596-610. Ooi, P.A.C, Canillas, M.D.B., and Yambao, R.T. 2007. SDC-AVRDC Partnership – Promoting Sustainability: Beyond Capacity Building. AVRDC-ARC, Bangkok, Thailand 71 pp. Opeña, R., and Tschanz, A. (1987). Bacterial wilt resistance program on tomato at AVRDC. ACIAR Bacterial Wilt Nwsl (2), 1-2. Palada, M. C., and Ali, M. (2006). Evaluation of Technologies for Improving Year-Round Production of Safe Vegetables in Peri-Urban Agriculture of Southeast Asia. Paper presented at the XXVII International Horticultural Congress-IHC2006: International Symposium on Horticultural Plants in Urban and PeriUrban. No. 762. Palada, M. C., and Wu, D. L. (2005). Increasing off-season tomato production using grafting technology for peri-urban agriculture in Southeast Asia. Paper presented at the International Conference and Exhibition on Soilless Culture: ICESC 2005. No. 742. Palada, M. C., and Wu, D. L. (2010). Grafting techniques for tomato and pepper under rice-based cropping system. In IPM in vegetables: enhancing its implementation in rice-based cropping systems (pp. 91123). Metro Manila: Department of Agriculture - Bureau of Agricultural Research. PVFRC. "Project implementation of PVFRC in recent years under management organization and implementation of the applicant to bidding package (translated from CÁC ĐỀ TÀI DỰ ÁN CỦA VIỆN

An impact assessment of AVRDC's tomato grafting in Vietnam

KHKTNNMN ĐÃ THỰC HIỆN TRONG NHỮNG NĂM GẦN ĐÂY DƯỚI SỰ TỔ CHỨC QUẢN LÝ VÀ THỰC HIỆN CỦA CÁC ỨNG VIÊN THAM GIA ĐẤU THẦU GÓI THẦU). pvfcdalat.org. Potato, Vegetable and Flower Research Center, n.d., Web, 28 Dec. 2012. <http://pvfcdalat.org/t.aspx?t=5>. Qaryouti, M., Qawasmi, W., Hamdan, H., and Edwan, M. (2007). Tomato fruit yield and quality as affected by grafting and growing system. Acta Horticulturae, 741, p. 199. Rashid, M. A., Hossain, M. M., Rahman, A., Alam, S., & Luther, G. (1999/2000). Evaluation of grafting compatibility of cultivated eggplant/tomato varieties on different Solanum rootstocks, IPM CRSP Annual Report. pp. 374-375. Schwarz, D., Rouphael, Y., Colla, G., & Venema, J. H. (2010). Grafting as a tool to improve tolerance of vegetables to abiotic stresses: Thermal stress, water stress and organic pollutants. Scientia Horticulturae, 127(2), 162-171. Thuy, L.T. 2010. Results of investigation and transferring technique of grafted tomato seedling on eggplant rootstock in the north of Vietnam. Agriculture and Rural Development Magazine. Special issue of 20th anniversary of FAVRI, 3/2010. pp 80-87. Tung, P. X. (1985). Bacterial wilt in Vietnam. Paper presented at the Workshop on Bacterial Wilt Disease in Asia and the South Pacific, Los Banos, Laguna (Philippines), 8-10 Oct 1985. Turhan, A., Ozmen, N., Serbeci, M., and Seniz, V. (2011). Effects of grafting on different rootstocks on tomato fruit yield and quality. Hortic. Sci, 38, 142-149. Venema, J. H., Dijk, B. E., Bax, J. M., van Hasselt, P. R., & Elzenga, J. T. M. (2008). Grafting tomato (Solanum lycopersicum) onto the rootstock of a high-altitude accession of Solanum habrochaites improves suboptimal-temperature tolerance. Environmental and Experimental Botany, 63(1–3), 359-367. Vinh, N. Q., and Ngo, X. C. (2006). Grafting and growing tomatoes by grafted seedlings to protect from bacterial wilt (Ralstonia solanacearum) on 3500-4,000 ha each year in Lam Dong Province, Viet Nam. Viet Nam: Institute of Agricultural Sciences for Southern Viet Nam. Wang, J. F., Hanson, P., and Barnes, J. A. (1998). Worldwide evaluation of an international set of resistance sources to bacterial wilt in tomato. In P. Prior, C. Allen and J. Elphinstone (eds.), Bacterial Wilt Disease: Molecular and Ecological Aspects. (pp. 269-275). Berlin, Germany: Springer Verlag. Wang, J. F., and Lin, C. H. (2005). Integrated management of tomato bacterial wilt. Shanhua, Taiwan: AVRDC The World Vegetable Center. Publication No. 05-615. Ya-Juia, S., Deng-Lin, W., and Black, L. (2001). Controlled environment studies to improve survival and quality of grafted tomato, sweet pepper, and watermelon transplants - Summer 2001. Shanhua, Taiwan: Asian Vegetable Research and Development Center.

AVRDC - The World Vegetable Center

www.avrdc.org AVRDC - The World Vegetable Center Headquarters PO Box 42 Shanhua, Tainan 74199 Taiwan

T +886 (0) 6 583-7801 F +886 (0) 6 583-0009 E info@worldveg.org