ECONOMIC VALUATION OF AGROBIODIVERSITY: FARMERS'CONTRIBUTION FOR ON FARM CONSERVATION OF RICE LANDRACES IN NEPAL

DRAFT ECONOMIC VALUATION OF AGROBIODIVERSITY: FARMERS' CONTRIBUTION FOR ON FARM CONSERVATION OF RICE LANDRACES IN NEPAL Diwakar Poudel1† and Fred Johnsen2 1Local Initiatives for Biodiversity, Research and Development (LIBIRD), Post Box 324, Mahendrapool, Pokhara, Kaski, Nepal email: dpoudel@libird.org 2Norwegian University of Life Sciences (UMB), Department of International Environment and Development studies (Noragric), Post Box 5003, N-1432 Aas, Norway email: fred.johnsen@umb.no ABSTRACT This paper documents total economic value (TEV) of crop genetic resources (CGRs) in monetary terms by eliciting farmers' willingness to conserve the genetic resources on farm. The study employs contingent valuation method (CVM) to assess the farmers’ willingness to pay (WTP) in the form of foregone benefit. We did questionnaire survey in 107 households in Kaski, Nepal. It was found that 93 percent of the surveyed farmers willing to contribute for on-farm conservation of agrobiodiversity. The mean WTP was $4.18 per annum per household meaning that farmers were ready to forgo $4.18 benefit by growing landraces. Several socio-economic factors such as landholding size, household size, education level, resource endowment status, sex of respondent, number of crop landraces grown, and knowledge on diversity influenced farmers' WTP. The study indicates that farmers value crop genetic resources (such as rice landraces) for use, non-use, and existence options and there is considerable support for such genetic resources conservation. Such economic valuation of agrobiodiversity can draw attention of the policy makers to formulate appropriate policy mechanism, raise public and political awareness of the importance of agrobiodiversity, and also support to set conservation priorities. Keywords: agrobiodiversity, conservation, landraces, valuation, willingness to pay ___________________________ † Corresponding address: Local Initiatives for Biodiversity, Research and Development (LIBIRD), Post Box: 324, Mahendrapool, Pokhara, Kaski, Nepal Tel: +977 61 5 32912/ +977 61 5 26834 Direct email: dpoudel@libird.org ECONOMIC VALUATION OF AGROBIODIVERSITY: FARMERS' CONTRIBUTION FOR ON FARM CONSERVATION OF RICE LANDRACES IN NEPAL INTRODUCTION The crop genetic resources (CGRs) valuation is to establish what the weight should be if the interest of the whole of the society is taken into account for its conservation (UNEP, 1995). A value measured in monetary terms will show the importance of a project in a given society and the society’s ability to pay for it. It can be seen as the willingness to commit resources to biodiversity conservation (Bräuer, 2003). Economic valuation helps authorities to make informed decisions about the diversity conservation (UNEP, 1995). Apart from acting as a decision aid, the estimation of benefits of conservation programs helps to promote acceptance for these programs. The economic evaluation is necessary in order to calculate the incremental costs for biodiversity conservation enshrined in the Convention of Biological Diversity (Bräuer, 2003). The monitory value of the biodiversity acts as an indicator for the transformation of environmental problems into policy actions. Various studies have shown that the policy decisions supported without having economic valuation are contradicted with economic rationality (Bräuer, 2003). There is also need for economic valuation of landrace diversity to generate information and knowledge for resource allocation in appropriate decision making, derive information for benefit sharing out of use of genetic resource, and identify least cost strategy to conserve landraces diversity (Gauchan, 1999). Cardoso de Mendonca et al., (2003) explicitly mentioned that if biodiversity can not be measured then there is no way to make rational decisions at what needs to be preserved or conserved. Many studies have been done to value the environmental services and biodiversity (see Nunes and van den Bergh, 2001). but the economic valuation of the agricultural genetic resources is almost a lack. The Convention on Biological Diversity (CBD) gave the legitimacy for specifying the value of such crop genetic resources because of two reasons: the conservationist need measures of value to justify budget and farmers’ right activists want to obtain compensation for farmers or for less developed countries (Brush and Meng, 1998). Although, there is a range of valuation techniques available to estimate non-marketed goods (Bann, 2003), the valuation of the genetic resources in the farmers’ field either suffers from inadequate data or appropriate methodology (Brush and Meng, 1998), thus lacking monitory valuation of CGRs. Valuation of landraces in Nepal was done using cost benefit analysis of improved variety and local landrace and cost foregone was compared (Gauchan, 1999). Recently Gauchan and Smale (2003) have conducted economic analysis of agrobiodiversity, particularly use values of rice landraces using marginal analysis based on partial budget and they have also recommended tools for valuation of non-use values of such agrobiodiversity. Assigning monitory values for landraces diversity is complicated in traditional subsistence farming systems like in Nepal (Gauchan, 1999) and therefore, is a challenging area of study. This study is aimed at assessing the total economic value (TEV) The total economic value of crop genetic resources is the direct use value, ecological function value, option value, existence value and bequest value or TEV= ƒ (UV+EFV=OV=EV=BV) (Edward and Abivardi 1998; Young 1992). of crop genetic resources in monetary terms by eliciting farmers' willingness to conserve the genetic resources on farm Conservation of the plant genetic resources in the place where they originated, variation created and developed, is the on farm or in situ conservation (Brush, 1995). using contingent valuation method (CVM). MATERIALS AND METHODS The Contingent Valuation Method (CVM) The contingent valuation is a method of valuing environmental goods for which no surrogate market exists. This method involves setting up a hypothetical market, obtaining bids, estimating mean WTP, constructing bid curves, aggregating data and evaluating the CVM exercise (Hanley and Spash (1993). The CVM has been extensively used to value environmental and natural resources in developing countries (Mbelwa, 2002; Navrud, and Mungatana, 1994). Therefore, I employed CVM to value CGRs in this study. While using the CVM, the recommendations by the National Oceanic and Atmospheric Administration (NOAA) panel (Arrow et al., 1993) were followed wherever possible in designing the survey and implementation. In setting the hypothetical market for CVM, the evaluated program was clear in its aims and consequences and the valued goods (CGRs) was well known and of interest to the people. Although the current trend in contingent valuation (CV) survey is towards the referendum formats using dichotomous choice (Arrow et al., 1993; Whittington, 2002), bidding game approach is also used for bid collection (Dong at al., 2003; Mbelwa, 2002; Mitchell and Carson, 1989). This study used open ended bidding game approach to elicit willingness for contribution. Two main reasons of using the open ended question are, first the dichotomous choice question tends to give higher estimates compared with the open ended (Bateman et al., 1995; White and Lovett, 1999) especially for studies conducted in low income countries. Second, the sample required for the dichotomous choice method for the statistical reliability is higher (Mitchell and Carson, 1989; OECD, 2002; White and Lovett, 1999) which was restricted in the present study due to time and resources constraints. The form and frequency of payment was clearly explained to the respondents. The payment vehicle in the study was the annual contribution for conservation program by forgone opportunity costs. As the CVM is always accompanied with doubts about its viability because of various potential biases such as hypothetical bias, information bias, strategic bias, design bias (bid vehicle, starting point) as well as embedding bias (Hanley and Spash, 1993; Hausman, 1993; Mbelwa, 2002; Mitchell and Carson, 1989), this study was designed to reduce such potential biases as far as possible. Efforts were made to deliver clear and equal information to all the respondents, payment vehicle was clear, pre-tested questionnaire used, and four different local enumerators were employed to reduce the strategic bias. To test the response validity, the questionnaire was designed to generate information on demographic and social background and embedded biodiversity knowledge of the respondents. The study area The study area is one of the in situ conservation project site implemented by Nepal Agriculture Research Council (NARC), Local Initiatives for Biodiversity, Research and Development (LIBIRD) and International Plant Genetic resource Institute (IPGRI), Kaski district in mid hill region (800 m to 1500 masl) of Nepal. The topography of the region is ancient lake and river terraces in moderate to steep slopes. The area experiences high rainfall (>3900 mm per annum) with warm temperate to subtropical climate. Mean daily minimum temperature of the coldest month is 7 and the mean daily maximum of the hottest month is 30.5 with monthly mean of 20.9C (Rana et al., 2000). The area is reported to be the hot spot in terms of crop diversity (Rijal et al., 1998). The major crop is rice, for which the farmers are maintaining 63 local landraces (Rana et al., 1999; Rijal et al., 1998). Therefore, the study used rice landraces as an example of crop genetic resources. Population, Sample and Sampling The study population was selected from 12 hamlets Hamlets are smaller village or are the parts of the bigger village, also called Tol. of the project area. Proportionate random sampling was done to draw the samples by stratifying the hamlets and the socio-economic status of the people (Nichols, 2000). The total samples were 20 percent constituting 107 households drawn proportionally from 3 socio-economic strata The socio-economic strata (rich, medium and poor) were identified using farmers’ own criteria. The main parameters for the categorisation are food self sufficiency of households, landholdings, government services, business, house and homestead properties, possessing house and/ or land in major cities (Rana et al., 1999; 2000). People having more than 0.75 ha of land, almost food self sufficient, have permanent monthly income source or remittances, have land and house in cities, are socio-economically ‘rich’. Similarly, people having 0.25 to 0.75 ha of land, more than 6 months food self sufficiency, have off farm income sources are ‘medium category’ and people with less than 0.25 ha of land, less than 6 months food self sufficiency, and labourers are categorised as ‘poor’. . As each household represented a sampling unit, one of the household members taking the decisions in agriculture was selected for survey. The survey and the questionnaire The instrument used for the study was a survey questionnaire. Local experienced enumerators were trained and employed to conduct the household survey. The open-ended structured questionnaire was pre-tested with 8 respondents out of the sample frame and revised before the administration. The questionnaire consisted of three sections. The first section was eliciting the household and the respondents’ demographic and socio-economic information. The second section elicited the information regarding the resources (land, genetic diversity), and the perception towards general knowledge on biodiversity and particularly on agrobiodiversity and rice landraces. The third section consisted of the explanation of the study hypothesis and eliciting the willingness to pay (WTP) for landraces conservation. The explanation of the subject matter was done before asking WTP questions. The respondents were briefed on the need of conservation using hypothetical market value of the landraces and asked whether they are willing to contribute for landraces conservation on farm. The question was: Are you willing to grow local landraces for the sake of conservation instead of growing your preferred variety? If yes, state the area (ropani A local unit for area measurement (20 ropani = 1 hectare).) and expected yield and the price of those conserved landraces (n=6) in your own land. Please provide the productivity and market price of your preferred landrace or variety. The data were entered in the Excel worksheet and individual’s WTP bids were then calculated from the willingness for contribution of area to grow landraces and expected market price and production, using the following equation. n WTP =Σ(AiYiPi) - Σ (aiyipi)………………………………………….…(1) i =1 Where, WTP = Total willingness to pay for in situ conservation of landraces, Ai = Total area for conservation = a1+a2…an, Yi = yield of preferred variety, Pi= price of the preferred variety, ai = Area for conservation of landraces i (i= 1to n where, n =6), yi= Expected yield of landrace i, and pi= Price for the landrace i. Data analysis and model specification The WTP bids were transferred to Minitab statistical software for analysis (MINITAB, 2000). During data analysis, 14 protest bids were omitted from the data set. These bids contained negative value due to higher expected market price of the said landraces. Mean willingness to pay (WTP), standard deviation, and the relationship between WTP and categorical variables were analysed using descriptive statistics, 2-sample t-test and ANOVA. The WTP bids were also regressed with various explanatory variables. The bid function was arrived at using general regression analysis, starting from the all-potential explanatory variables, removing the non-significant, re-estimating the model and so on until all remaining variables were significant at 95% level (for example see Horton et al., 2003). The valuation function is WTP = ß0 +ß1 X1 + ß2 X2 + ß3 X3+ …+ ßn Xn +i ………………………………..(2) Where, WTP =Farmers willingness to pay for landraces conservation, ß0 = Constant, ß1-ßn= Coefficients, X1 - Xn = Variables contributing in WTP, and εi = Random error ~N (0, 1) FINDINGS AND DISCUSSIONS Sample characteristics The demographic data obtained from the 107 households were proportionately stratified within socio-economic status and randomised for sampling. The mean age of the respondents was 46 years. The respondents were nearly equally represented males and females. Fifteen percent of the respondents were illiterate and less than 5 percent attended the university level. The average household size was 6.5 which is higher than national average 5.44 (CBS.2003) but exactly similar reported by Rana et al., (2000) in the same study area. The mean household income was USD1153 in the study area, which is slightly lower than the national level (CBS, 2003). The total land capital of the respondents was 0.8926 ha which is almost equal to the national average 0.96 ha (p>0.05) (Rana et al., 2000; CBS, 2003). The average number of land parcels was 7.75 in the study area, which is very high compared to the national average 4.0 (p<0.01) (Rana et al., 2000; CBS, 2003). The average food self-sufficiency was 9.95 months, which shows majority of the people are relatively better off in terms of food production. It was found that 25% of the households are producing food for less than 6 months whereas 25% producing food which is sufficient for more than 12 months. The average number of landraces (excluding MVs) grown was 3.2 per household (Table 1). (Table 1 here) The sampling data are almost representative of the study area with the gender, household size, the educational status of people, the annual income and landholdings of the people, although the average diversity of landraces is higher than in Terai and high hills of the country (Rana et al. 1999; Paudel et al., 1999). Therefore, the study is representative of the similar agro-ecological, and socio-economic conditions of country. Farmers’ knowledge on biodiversity Since the study was conducted in the In situ conservation project area, the farmers of the study area have been found very much informed by the project on the need of conservation. They were familiar with the term ‘agrobiodiversity’ and different methods of biodiversity conservation. Majority (67 %) of the farmers were familiar with the concept of agrobiodiversity. Most of the farmers explained the use value of the rice landraces relating to conservation. Although few farmers (11%) perceived that landraces must be conserved for the future generations need (intrinsic value), most of them acknowledged the need of the conservation for direct use value, ecological function value and option use value. Farmers have been very enthusiastic to explain the need for landraces conservation and all the respondents participated to explain why the local rice landraces should be conserved. The farmers’ perceived knowledge on the ‘need for rice landraces conservation’ is presented in Table 2. (Table 2 here) Willingness to pay (WTP) for on farm conservation of rice landraces Mean WTP per annum Ninety three percent of the respondents show willingness to pay for the landraces conservation on-farm. Rest of the respondents (farmers) were either landless or not willing to pay for conservation. The contribution of mean area to be grown for such landraces for the purpose of the landraces conservation was 0.03 ha per farmer. The mean WTP for conservation of landraces in the form of opportunity costs was equivalent to USD 4.18 per farmer per annum (calculated including zero bids). This contribution is the farmers’ willingness to forgo benefits from the actual expected farm income by growing the landraces. The aggregate contribution per annum by the community of the study area will be USD 2153, assuming that all the farmers will participate in the conservation activities. The mean willingness to contribute for in situ conservation of landraces in the form of opportunity cost foregone was almost similar to estimated cost to conserve one local landrace Nakhisaro in Bara Nepal. Gauchan (1999) estimated that the opportunity cost of conservation of Nakhisaro in 0.028 hectares is USD 3.136 (i.e. USD 112 per hectare). The present value is USD 3.802 The present value was calculated using consumer price index from the base year 1999. , which is non-significantly different (p>0.05) with the value in this study. This study indicates that CVM could be employed to assess the farmers' willingness to pay for landraces conservation and this is relatively easy and efficient if carried out carefully. Factors affecting WTP The variables hypothesised to affect the farmers’ willingness to contribute for conservation of landraces were regressed with WTP bids using multiple linear regression. Most of the tested variables were non-significant with the willingness to pay (Table 3). The WTP bids were significantly associated with the education, food self sufficiency and number of landraces grown. WTP bids increased with the increased education level and increased food sufficiency whereas, the farmers already growing a large number of landraces in their farm were less willing to contribute for conservation more landraces. (Table 3 here) Educated people might have more knowledge on biodiversity conservation, thus they could have higher WTP for conservation. The higher food sufficiency level means more cultivated land (which is different from the total landholding that might include the forest or fallow). People having a high food sufficiency level can grow landraces in more areas even when the production of the said landraces is low, thereby contributing in conservation. Similarly, the farmers already growing a large number of landraces are not interested to grow even more for conservation or the contribution for the conservation is less for those households, which is theoretically acceptable. These diversity managing households are custodians for the on farm conservation of crop genetic resources (Subedi et al., 2003a). Younger people with bigger household sizes and higher landholding size were also found more willing to contribute for conservation although there was no significant difference (p>0.05). Household income was not significantly associated with WTP bids which is inconsistent with the economic theory. However, the empirical result by the World Bank (1993) on WTP for quality water supply also does not depend on income. Similarly, it has been found that the low-income population disproportionately has higher environmental risk than the wealthy or high-income population (Aldy et al., 1999). The low income population having marginal growing area requires landraces for their subsistence because modern varieties do not perform well in such conditions. The low income population, therefore, might have expressed higher willingness to conserve the CGRs. Apart from these explanatory variables, other categorical variables caste, gender, socio-economic status and the knowledge on the diversity were tested for the effect on the willingness to contribute for conservation using ANOVA. Sex of the respondents, socio-economic status and the knowledge on agrobiodiversity were significantly influencing the willingness to contribute for the conservation. The mean willingness for contribution by male (USD 5.71) was significantly more than the females’ contribution of USD 2.61 (p < 0.05). The educated people and usually male decision makers are found willing to contribute more because of the individual income effect which is obvious in case of developing countries (Horton et al., 2003). They are also the decision makers of the households and can decide quickly as much as they want to pay but the female are usually not the decision maker and they have low WTP. Similarly, richer people were willing to pay more than the poor people (p<0.01) and people having knowledge on agrobiodiversity were willing to contribute more (p<0.05). The farmers who maintain on farm agrobiodiversity are socially resource rich, having more cultivated lands (Rana et al., 1999; Subedi et al., 2003a; 2003b) and educated and male-headed households (Gauchan et al, 2001). Similarly, the people having knowledge on agricultural diversity are interested to contribute for the conservation program because they generally maintain a large number of landraces in their farm for their diverse land parcels (Subedi et al., 2003a; 2003b) ) and they also consider the intergenerational equity. The low R2 value (0.28) indicates that there are factors external to the regression models, which influence the willingness to pay (WTP) of the farmers. However, Mitchell and Carson (1989) mentioned that the CV study showing R2 more than 0.15 using few key variables is acceptable. Therefore, this study explains that the CVM can reliably be used in the valuation of the crop genetic resources. CONCLUSION When questions are carefully constructed and respondents are well informed on valuing goods, willingness to pay values can be useful means of quantifying preferences. The study generated some empirical evidence that the farmers are willing to contribute for conservation of agrobiodiversity on farm and identifies the factors that affect farmers’ willingness to pay. The values obtained in this study are quantified indications of the value placed by the society on the crop genetic resources particularly rice landraces measured in terms of their willingness to pay for conservation. Contingent Valuation (CV) generated monetary value gives an empirical idea about the benefits of the conservation for society when they state their use and non-use values. The study indicates that there is considerable support for the conservation which is relevant for the research and development agencies in planning for a conservation project because the marginal conservation cost could be borne by the people, thus reducing the total cost. Although there are some criticism for the valuation and value generated by CVM, it can be argued that ‘it is better to be roughly right than to be perfectly wrong’ without value of something to make it rational. Even if stated willingness to pay is inaccurate, it should be used because no alternative estimate exists for public policy purposes (Diamond and Hausman, 1994) realising that ‘some number is better than no number’. Implicitly, the economic valuation of agrobiodiversity can draw attention of the policy makers to formulate appropriate policy mechanisms; it will raise public and political awareness on the importance of agrobiodiversity, and also helps to set conservation priorities. ACKNOWLEDGEMENT This paper is output of Masters Thesis research. This study was conducted under the research grant provided by Norwegian Agency for International Development and Cooperation (NORAD), Norway. Therefore, the authors would like to express their sincere gratitude to the organization. Authors are also grateful to the workshop organising committee for accepting this paper. REFERENCES Aldy, J.E., Kramer, R.A. and Holmes, T.P. 1999. Environmental equity and conservation of unique ecosystems: An analysis of the distribution of the benefits for protecting Southern Applachian spruce-fir forests. Society and Natural Resources. 12, 93-106. Arrow, K.J., Solow, R., Leamer, E., Portney, P., Radner, R. and Schuman, H., 1993. Report of the NOAA Panel on Contingent Valuation. 58 Fed. Reg. 4601. Washington, DC, Government Printing Office. Bann, C., 2003. An overview of valuation techniques: advantages and limitations. Asean biodiversity, 2 (2), 8-15. Bateman, I.J., Langford, I.H., Turner, R.K., Willis, K.G. and Garrod, G.D., 1995. Elicitation and truncation effects in contingent valuation studies. Ecological Economics, 12, 161-179. Bräuer, I.,2003. Money as an indicator: to make use of economic evaluation for biodiversity conservation. Agriculture, Ecosystems and Environment, 98, 483-491. Brush, S. B. 1995. In situ conservation of landraces in centres of crop diversity, crop science 5. 418-135. Brush, S.B. and Meng, E., 1998. The Value of Wheat Genetic Resources to Farmers in Turkey. In R.E. Evenson, D. Gollin and V. Santaniello (Eds.), Agricultural Values of Plant Genetic Resources: CAB International. Cardoso de Mendonca, M.J., Sachsida, A. and Loureiro, P.R.A., 2003. A study on the valuing of biodiversity: the case of three endangered species in Brazil. Ecological Economics, 46(1), 9-18. CBS, 2003. Statistical Year Book of Nepal. Central Bureau of Statistics: Kathmandu, Nepal Diamond, A.P. and Hausman, J.A., 1994. Contingent Valuation: Is Some Number is Better than No Number? Journal of Economic Perspectives, 8(4), 45-64. Dong, H., Kouyate, B., Carins, J. and Sauerborn, R., 2003. A comparison of the reliability of the take-it-or-leave-it and the bidding game approaches to estimating willingness-to-pay in a rural population in West Afroca. Social Science and Medicine, 56, 2181-2189. Edwards, P.J. and Abivardi, C., 1998. The value of Biodiversity: where ecology and economy Blend. Biological Conservation., 83(3), 239- 246. Gauchan, D. and Smale, M., 2003. Choosing the 'right' tools to assess the economic costs and benefits of growing landraces: an example from Bara District, Central terai, Nepal. Plant Genetic Resource Newsletter, 134, 18-25. Gauchan, D., 1999. Economic valuation of rice landraces diversity: a case study of Bara ecosite, Terai, Nepal. In B.Sthapit,M. Upadhyay and A. Subedi (Eds.), A scientific basis of in situ conservation of agrobiodiversity on farm: Nepal's contribution to global project.: NP Working paper No. 1/99. NARC/LIBIRD/IPGRI, Rome Italy. Gauchan, D., Chaudhary, P., Smale, M., Sthapit, B.R. and Upadhyay, M.P., 2001. An analysis of market incentives for on-farm conservation of rice landraces in central Terai Nepal., Poster presented in International Symposium on 'Managing biodiversity in agriculture ecosystems' organised by United National University, IPGRI, and the Secretariat for convention on biological diversity,. Monterial, Cancada, Quebec, Canada. Hanley, N. and Spash, C.L., 1993. Cost benefit analysis and the environment. United Kingdom: Edward Elgar Publishing Limited. Hausman, J., 1993. Contingent Valuation-A Critical Assessment: Amsterdam. North-Holland Press. Horton, B., Colarullo, G., Bateman, I.J. and Peres, C.A., 2003. Evaluating non-user willingness to pay for a large-scale conservation programme in Amazonia: a UK/Italian contingent valuation study. Environmental Conservation, 30(2), 139-146. Mbelwa, J.R., 2002. Improving Beach Management: An analysis of the role of the government and local community in the management of the beach areas in Dar es Salaam., Noragric: MSc Thesis. Agricultural University of Norway, Ås. MINITAB, 2000. MINITAB for Windows: MINITAB statistical software, MINTIAB Inc. Mitchell, R.C. and Carson, R.T., 1989. Using Surveys to Value Public Goods: The Contingent Valuation Method. Washington D.C., Resource for the future. Navrud, S. and Mungatana, E.D., 1994. Environmental valuation in developing countries: The recreational value of wildlife viewing. Ecological Economics, 11, 135-151. Nichols, P., 2000. Social Survey Methods: A Field guide for Development Workers: Oxfam, Oxford, UK, Oxfam publishing. Nunes, P. and van den Bergh, J.C.J.M., 2001. Economic valuation of biodiversity: sense or nonsense? Ecological Economics, 39(2), 203-222. OECD, 2002. Handbook of Biodiversity Valuation: a guide for policy makers: Paris, France. Paudel, C.L., Chaudhary, P., Rijal, D.K., Vaidya, S.N., Tiwari, P.R., Rana, R.B., Gauchan, D., Khatiwada, S.P., Gupta, S.R. and Sthapit, B.R., 1999. Agro ecosystem factors for in situ conservation of agrobiodiversity in different eco-zones of Nepal. In B. Sthapit, M. Upadhyay and A. Subedi (Eds.), A scientific basis of in situ conservation of agrobiodiversity on farm: Nepal's contribution to global project.: NP Working paper No. 1/99. NARC/LIBIRD, Nepal /IPGRI, Rome Italy. Rana, R.B., Gauchan, D., Rijal, D.K., Khatiwada, S.P., Paudel, C.L., Chaudhary, P. and Tiwari, P.R., 1999. Cultural and socio-economic factors influencing farmer management of local crop diversity: Experience from Nepal. In B. Sthapit,M. Upadhyay and A. Subedi (Eds.), A scientific basis of in situ conservation of agrobiodiversity on farm: Nepal's contribution to global project.: NP Working paper No. 1/99. NARC/LIBIRD/IPGRI, Rome Italy. Rana, R.B., Rijal, D.K., Gauchan, D., Sthapit, B., Subedi, A., Upadhyay, M.P., Pandey, Y.R. and Jarvis, D.I., 2000. In situ Crop Conservation: Findings of agro-ecological, crop diversity and socio-economic baseline survey of Begnas ecosite Kaski, Nepal.: NP Working paper No. 2/2000. NARC/LIBIRD, Nepal /IPGRI, Rome Italy. Rijal, D.K., Rana, R.B., Sherchand, K.K., Sthapit, B.R., Pandey, Y.R., Adhkari, N., Kadayat, K.B., Gauchan, Y.P., Chaudhary, P., Poudel, C.L., Gupta, S.R. and Tiwari, P.R., 1998. Strengthening the Scientific basis of in situ conservation of agrobiodiversity: Finding of site selection Kaski, Nepal.: NP Working paper No. 2/1998. NARC/LIBIRD, Nepal /IPGRI, Rome Italy. Subedi, A., Chaudhary, P., Baniya, B.K., Rana, R.B., Tiwari, R.K., Rijal, D.K. and Sthapit, B.R., 2003b. Who maintains crop genetic diversity and how: Implication for on farm conservation and utilization? Culture and Agriculture 25 (2), 14-50. Subedi, A., Chaudhary, P., Baniya, B.K., Rana, R.B., Tiwari, R.K., Rijal, D.K., Jarvis, D. and Sthapit, B.R., 2003a. Who maintains genetic diversity and how? Policy implications for agrobiodiversity management. In D. Gauchan, B. Sthapit, and D. Jarvis (Eds.), Agrobiodiversity conservation on farm: Nepal's contribution to scientific basis for national policy recommendations. Kathmandu: IPGRI. Subedi, A., Chaudhary, P., Baniya, B.K., Rana, R.B., Tiwari, R.K., Rijal, D.K. and Sthapit, B.R., 2003b. Who maintains crop genetic diversity and how: Implication for on farm conservation and utilization? Culture and Agriculture 25 (2), 14-50. UNEP, 1995. Global Biodiversity Assessment.: Cambridge. Cambridge University Press. White, P.C.L. and Lovett, J.C., 1999. Public preferences and willingness-to-pay for nature conservation in the North York Moors National Park, UK. Journal of Environmental Management, 55, 1-13. Whittington, D., 2002. Improving the performance of contingent valuation studies in developing countries. Environmental & Resource Economics, 22(1-2), 323-367. World Bank, 1993. World development report. Oxford University Press, New York. Young, M.D., 1992. Sustainable investment and resource use: equity, environmental integrity and economic efficiency. Carnforth and Unesco, Paris, Parthenon Press. Table 1: Characteristics of sample (n=107) Age (years; mean, SD) 45 (11.78) Gender (n,%) Female Male 53 (49.5) 54 (50.5) Education of the respondents (n, %) Illiterates Literates Primary Secondary Higher secondary University education 16 (15) 33 (31) 16 (15) 28 (26) 10 (9) 4 (4) Socio-economic status (n,%) Rich Medium Poor 41 (38.3) 40 (37.4) 26 (24.3) Household size (mean, SD) 6.5(2.9) Household income USD (mean, SD) 1153 (1647) Landholdings (ha) (mean, SD) 0.90 (0.66) Parcel of Land (mean, SD) 7.75 (5.1) Food self sufficiency months (mean, SD) 9.95 (4.46) Number of varieties grown (mean, SD) 3.2 (1.9) Knowledge on agrobiodiversity (n, %) Yes No 72 (67.3) 35 (32.7) n= Number, SD= Standard Deviation Table 2: Farmers’ reason for conservation of rice landraces in Kaski districts, Nepal Need for landraces conservation Responses (n = 107) n % The local landraces have diverse adaptability 64 59.8 Local landraces have diverse taste, use, straw need and importance 57 53.3 Possibility of extinction from the habitat 33 30.8 Medicinal values 27 25.2 Possesses cultural value 23 21.5 These have yield stability over years 13 12.1 Security for future 12 11.2 Improved variety have no guarantee of yield in marginal growing conditions 10 9.3 Local landraces are suitable in adverse climatic conditions such as epidemic etc 8 7.5 Need for choice in seed in future 7 6.5 For improvement of new varieties 6 5.6 The basis for subsistence are landraces 2 1.8 These should be conserved because of international or global importance 1 0. 9 These are being grown from ancient time 1 0. 9  Table 3: Estimates of the multiple regression model: factors influencing the respondents responses (n=93) Explanatory variables Regression model: with all variables Best fitting regression model Coefficient T values Coefficient T values Constant 0.321 0.10 -0.63 -0.39 Household size 0.403 1.63 - - Age of respondent -0.079 -1.24 - - Education level 0.867* 1.70 1.01** 2.11 Household income ($) -0.00012 -0.28 - - Food sufficiency 0.620** 2.79 0.63*** 3.79 Landholdings 0.569 0.37 - - Land parcel -0.046 -0.28 - - Number of landraces grown -1.04** -2.54 -1.03*** -2.86 R2 0.28 0.25 * p < 0.10 , ** p <0.05 and *** p <0.01 13