EFFECT OF SELECTED INSECTICIDE ON WHITEFLY (Bemisia tabaci) INFESTING BRINJAL PLANTS © 2016 Asian Economic and Social Society. All rights reserved ISSN (P): 2304-1455/ISSN (E):2224-4433 Volume 6(8), 152-162 152 THE STATUS OF THE ADOPTION OF MANGROVE SWAMP RICE VARIETIES IN SIERRA LEONE Moses Gborie, Nazir A. Mahmood, Adam S. Kanu, Finda B. Jenkins, Henry M.S. Kargbo Sierra Leone Agricultural Research Institute (SLARI), Rokupr Agricultural Research Center (RARC), Freetown, Sierra Leone Article History: Received: 15-Oct-2016 Revised received: 10- Nov-2016 Accepted: 15-Nov-2016 Online available: 30- Nov-2016 Keywords: Mangrove, rice, adoption, adopters, non-adopters, probit model, Sierra Leone Abstract Mangrove swamp rice varieties developed at the Rokupr Agricultural Research Centre (RARC), Sierra Leone were evaluated in 2014 to determine their adoption status in Kambia, Port Loko and Moyamba districts of Sierra Leone. The study evaluated farmers’ knowledge about the varieties, the varieties grown by farmers at least once, adopters, non-adopters, loss of varieties, abandonment of varieties, and the determinants of the adoption of mangrove rice varieties. Data were obtained from a sample survey of 600 mangrove rice farmers using the multistage stratified random sampling technique. Farmers in the Kambia district were more knowledgeable about the mangrove varieties: ROK 5 and ROK 10 are the most adopted varieties while ROK 5, ROK 10 and CP 4 are the most cultivated and have been grown at least once by some farmers in all the districts – ROK 5 (380 farmers); ROK 10 (315 farmers); and CP 4 (99 farmers). Probit regression estimation shows that farmers’ adoption of the mangrove rice varieties was affected positively and significantly by farmers’ level of education, training, access to credit, farm size, contact with extension, and farming experience. 1. INTRODUCTION The mangrove ecology has significant potentials for rice production in West Africa. Mangrove swamp rice farming is among the oldest rice farming systems in West Africa and accounts for 10% of the total regional rice production. The average mangrove rice farmers’ annual yield is 2 t/ha of paddy compared to 1 t/ha for most other rice ecologies in the region (Agyen-Sampong, 1984). This suggests that improving rice production in the mangrove ecology can enhance the production of enough rice to meet the increasing demand in West Africa in general and Sierra Leone in particular. Corresponding author’s Name: Moses Gborie Email address: mosesggborie@yahoo.co.uk Asian Journal of Agriculture and Rural Development http://aessweb.com/journal-detail.php?id=5005 DOI: 10.18488/journal.1005/2016.6.8/1005.8.152.162 Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 153 Such improvement can be facilitated through technologies generated at research centers, disseminated and validated at the farm level and communicated to influence adoption by farmers. Adoption is most often a self-decision process based on the perceived knowledge and appreciation of the said technology in terms of its specific physical characteristics and intrinsic qualities. In addition to these factors, the related cost of the technology, its simplicity, adaptability to the farming circumstances, and its comparative advantage over existing local technologies constitute a range of potential factors that can affect adoption decisions made by farmers. This study was undertaken to obtain an inventory of the disseminated and adopted mangrove swamp rice varieties in Sierra Leone. 1.1. Objectives of the study The key objective was to evaluate the disseminated mangrove swamp rice technologies (mostly rice varieties) in the various rice growing ecologies and assess their adoption status in Sierra Leone. The specific objectives were to:  List the mangrove swamp rice technologies and their present adoption status.  Identify the key determinants of adoption of the improved mangrove swamp rice technologies.  Identify key policy issues that will guide the strategy for out-scaling the improved mangrove swamp rice technologies to ensure high rates of their adoption. 2. METHODOLOGY 2.1. Sampling design A multi-stage stratified random sampling procedure was adopted for the study. First, Kambia, Port Loko and Moyamba districts were purposively selected based on the ecology to be surveyed. The mangrove ecology is not spread across the country but is predominant in the selected districts. Then, two chiefdoms 1 were purposively selected from each district, based on the spread of the mangrove ecology therein. Finally, 100 farmers were randomly selected and interviewed per chiefdom. Thus 600 respondents were interviewed for the entire survey. 2.2. Data collection and analysis Primary data were collected from selected samples of rice farmers within the various chiefdoms. For that purpose, a data collection tool was developed, field tested, and administered to farmers. Qualitative and quantitative methods were used to analyze the data with Microsoft Excel 10 and STATA 13, a robust analytical tool for socio-economic studies. 2.3. Probit regression model Probit regression, also called probit model, is used to model dichotomous or binary outcome variables. Data gathered in the 2014 cropping season on the cultivation of mangrove swamp rice varieties developed at the Rokupr Agricultural Research Centre (RARC) were used to define two categories of rice farmers:  Adopter - a farmer who has planted a RARC variety at least once and continues to use such a variety in his/her field in the 2014 cropping season (the period when the survey took place), and  Non-adopter - a farmer who previously planted a RARC variety but did not continue growing such a variety during the 2014 season. 1 The third geographical division of Sierra Leone Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 154 Probit model is one of the most popular statistical methods developed for analyzing dichotomous response-dependent variables (Donkoh & Awuni, 2011). Most of the adoption literature estimate a probit model to ascertain the factors that affect the adoption of a new technology or innovation (Feder et al., 1985). For instance, Foltz (2003) provided a theoretical explanation of the probit model and also summarized the factors that explain the rate of adoption/diffusion of a soil and water conservation technology under four hypotheses, viz resource scarcity, capital constraint, learning costs and risk aversion. According to Foltz (2003), increasing scarcity of a natural- resource endowment (like fertile lowland) leads to higher shadow prices for the resource, causing farmers to switch to a resource-conserving technology (like soil and water conservation technology). The capital constraint hypothesis implies that new technologies would spread faster among farmers with better access to capital to pay for the new technology than farmers with little or no access. Moreover, the learning-cost-hypothesis suggests that technologies will spread fastest in areas where information about the innovation is readily available and most easily evaluated by potential adopters. This means that farmers who have access to extension services, better education (able to read and understand information about the technology), the opportunity to attend useful workshops and participate in on-farm experimentation stand a better chance of adopting the technology than the less privileged ones. Lastly, risk aversion implies that farmers would not like to invest in unknown technologies because of uncertainties with regards yield and income. Similarly, an expensive technology is relatively more risky because farmers are not sure if they would be able to recoup the money invested into the technology. In this case, the chances of adoption/diffusion would be slim. But, if a new technology is risk-reducing in the sense that farmers are familiar with it or it is relatively cheap, then farmers would readily adopt it, other things being equal. The probit model was used in this study to empirically quantify the relative influence of various factors in the decision of the respondents to adopt the mangrove rice varieties developed by RARC. From previous adoption studies, many factors influence farmers’ adoption decision. For instance, Tiamiyu et al. (2009) found that age, education, land tenure status, farming experience, farm income, farm family size, family labour, contact with extension agents, membership of an association, credit use, and extent of commercialization influence farmers’ adoption decisions. In India, Sita and Ponnarasi (2009) analysed variables like age, literacy level, farm size, income of households, number of earners in the family, and number of contacts with extension agencies as determinants of adoption. In the current study, it is postulated that the probability of a farmer adopting the RARC mangrove varieties (Pi) depends on farm and farmers’ specific attributes like age, education, farmer’s income, training, access to credit, membership of an association, farm size, sex, extension contact and farming experience. The index variable (Pi), which indicates whether a farmer adopts the RARC mangrove varieties, is a function of the independent variables (Xi). In this study, age was measured in years, a dummy variable was used for education - a literate farmer was assigned a score of 1 and an illiterate farmer a score of 0. Farmers’ income was determined based on farm income in the previous year. Training was determined based on rice farming training acquired and a dummy variable was used - farmers who acquired training were assigned a score of 1 and a score of 0 otherwise. Access to credit was measured as a dummy - farmers with access to credit were assigned a score of 1 and a score of 0 otherwise. A dummy variable was used for membership of an association - farmers who are members of an association were assigned a score of 1 and a score of 0 otherwise. Farm size (hectare) was defined as the size of the farm allocated to rice cultivation during the growing season under survey. A dummy variable was used to specify the sex of the respondent - a male respondent was assigned a score of Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 155 1 and a female respondent a score of 0. Extension services to the farmers were defined as a dummy variable in the model - 1 for farmers who received at least one contact with the extension service and 0 otherwise. Farming experience was incorporated into the model as the number of years for which the farmer has been a rice farmer. Assuming that a farmer is an adopter of the RARC varieties, the probit regression model has been specified as follows: Pi = f(X₁, X₂, X₃, X₄, X₅, X₆, X₇, X₈, X₉, X₁₀) Pi= β₀ + β₁X₁ + β₂X₂ + β₃X₃ + β₄X₄ + β₅X₅ + β₆X₆ + β₇X₇ + β₈X₈ + β₉X₉ + β₁₀ X₁₀ + Ԑi Where: X₁ = age of respondent X₂ = education X₃ = farmers income X₄ = training X₅ = access to credit X₆ = membership of an association X₇ = farm size in hectares X₈ = sex X₉ = extension contact X₁₀ = farming experience β0 = constant βis = parameters to be estimated, and Ԑi = error-term 3. RESULTS Nine RARC mangrove rice varieties (CP 4, ROK 4, ROK 5, ROK 8, ROK 9, ROK 10, ROK 21, ROK 22 and ROK 23) were identified and the 600 farmers were interviewed about their knowledge and use of these varieties. Data and information were gathered using the data collection tool developed for this purpose. Both qualitative and quantitative data were collected and analysed. The list of mangrove rice varieties was provided by the breeding unit at RARC, so that all mangrove varieties developed at RARC by the breeding unit were captured in the research. 3.1. Knowledge of the improved mangrove rice varieties  CP 4 A large proportion of the farmers interviewed know about the RARC-improved rice varieties. Most of the farmers in Kambia district know about all the RARC mangrove varieties; this is because the research centre (RARC) is located in this district and this enables them to easily know about and have access to these improved rice varieties. About 81% of farmers in Mambolo Chiefdom and 61% of those in Samu Chiefdom know about CP 4.  ROK varieties It is estimated that 66% of farmers in Mambolo and Samu Chiefdoms know about ROK 4. Also, ROK 5 is known to 97% of farmers in Mambolo Chiefdom and 98% of farmers in Samu Chiefdom. ROK 8 is known by 36% of farmers in Mambolo Chiefdom and 34 % of farmers in Samu Chiefdom. ROK 9 is known by a relatively smaller proportion of farmers - 34% in Mambolo and Samu Chiefdoms. About 98 % of farmers in Mambolo Chiefdom and 88% of those in Samu Chiefdom are familiar with ROK 10. These compare with 51% of farmers in Mambolo Chiefdom and 26% in Samu Chiefdom who know about ROK 21. ROK 22 is known by 60% of Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 156 farmers in Mambolo Chiefdom and 22 % of those in Samu Chiefdom. ROK 23 is known by 48% of farmers in Mambolo Chiefdom and 24 % of farmers in Samu Chiefdom. Farmers in Port Loko and Moyamba districts are more knowledgeable about ROK 4, ROK 5 and ROK 10 because these varieties are more popular in mangrove rice farming communities than the other mangrove rice varieties. Farmers in Kaffu Bullom Chiefdom know about ROK 4 (61%), ROK 5 (92 %), and ROK 10 (34%). 3.2. Mangrove rice varieties grown at least once by farmers Table 2 shows that ROK 5, ROK 10 and CP 4 are the most cultivated of the mangrove rice varieties and have been grown at least once by some farmers in all the districts – ROK 5 (380 farmers); ROK 10 (315 farmers); and CP 4 (99 farmers) in 2014. In Kambia district, CP 4 was grown at least once by 44% farmers in Mambolo Chiefdom and 33% in Samu Chiefdom. However, ROK 8, ROK 9, ROK 21, ROK 22 and ROK 23 were not widely grown by farmers because they were not aware of their existence. Also, ROK 9, ROK 21 and ROK 22 have never been grown by farmers in Moyamba district because these varieties are not accessible. Finally, ROK 21 and ROK 23 have not been grown by farmers in Port Loko district due to issues relating to awareness, exposure and accessibility. 3.3. Adopters of RARC mangrove rice varieties The number of farmers who cultivated the varieties at least once and who have continued to cultivate them up to 2014 is provided in Table 3. ROK 5 and ROK 10 are the most cultivated rice varieties (163 farmers still cultivate ROK 5 and 86 farmers still cultivate ROK 10). ROK 4 and CP 4 were also grown by quite a reasonable number of farmers in all the districts surveyed (30 farmers for ROK 4 and 23 farmers for CP 4). In 2014, ROK 8 was not cultivated in Samu, Lokomasama and Bumpeh Chiefdoms; ROK 21 and ROK 22 were not cultivated in Port Loko and Moyamba districts; ROK 23 was cultivated by only one farmer in Samu Chiefdom in Port Loko district and one farmer in Ribbi Chiefdom in Moyamba district. The reasons for varieties not being cultivated are a lack of awareness, availability and accessibility. In this regard, ROK 5 and ROK 10 are the varieties that are mostly readily available to farmers and are, therefore, the most adopted ones. 3.4. Non-adopters of RARC mangrove varieties Table 4 shows the number of farmers who cultivated the varieties at least once but discontinued their cultivation - ROK 5 (217 farmers); ROK 10 (127 farmers); CP4 (76 farmers); and ROK 4 (68 farmers). The reasons given for non-adoption were the lack of skilled labour for the various farm operations and the lack of training and extension service. The low levels of adoption of the remaining mangrove rice varieties were due to the lack of awareness of the existence of the varieties. These results reveal that a large number of mangrove farmers did not adopt RARC mangrove rice varieties because of the lack of skilled labour, training, extension visits and awareness of the varieties. 3.5. RARC mangrove rice varieties lost by farmers Table 5 provides information on farmers who cultivated the varieties at least once but stopped their cultivation because they lost the varieties. This is a serious problem because it is difficult for farmers to regain lost varieties. Farmers reported that varieties were lost due to admixture and the civil war. ROK 5 and ROK 10 were the most affected - 172 farmers lost ROK 5 while 112 lost ROK 10 – but other varieties were also lost (Table 5). Farmers indicated their interest in cultivating the lost varieties should they become available again. 3.6. RARC mangrove rice varieties abandoned by farmers Table 6 shows the number of farmers that cultivated the RARC varieties at least once but abandoned them because of high bird damage and the lack of access to seed. The abandonment Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 157 rate was high for ROK 5 (43 farmers), CP 4 (22 farmers), ROK 4 (19 farmers) and ROK 10 (16 farmers). 3.7. Probit model results The usage of the rice varieties in the 2014 cropping season was regressed against farmers’ socio- economic variables, which include farmers’ age , education, income, training, access to credit, membership of an association, farm size, sex, extension contact and farming experience. These variables were chosen based on the findings of previous studies on technology adoption in developing countries (see Tiamiyu et al., 2009; Sita & Ponnarasi, 2009). The influence of age on adoption decision depend on experience, wealth, authority, risk taking and consumers/workers ratio in the household. Older farmers may have more experience, authority and wealth, while younger farmers might be less risk-averse, may have greater access to information and have higher consumer/workers ratio but lack experience and resources. Thus, older farmers are expected to be positively related to adoption. Age can either be positively or negatively correlated with adoption decision. The coefficient of age carries a negative sign and is significant in adoption decision. This indicates that younger farmers are more likely to be interested in adopting new technologies if they are not constrained by cash resources, while older farmers are less likely to be able to use new technologies if they require extra physical labour and/or older farmers may be less interested because they are more risk-averse. Education is expected to affect the level of technology adoption positively through effective skill acquisition in choosing better inputs. In this study, the coefficient of education carries the expected sign and is significant in influencing the adoption of the varieties. Therefore, farmers that are educated (literate) are more likely to adopt new technologies than those that are not. Income derived from farming activities indicates the level of profit of the farmers. The expectation is that farmers will be able to plough back profit from their previous farming into future production process in order to increase profit. Farmers with more income are likely to be able to afford and apply expensive inputs aimed at increasing productivity; hence income is expected to influence technology adoption positively. The result from the Probit does not carry the expected sign and is not significant. Training acquired for rice farming is expected to affect adoption of technology positively, since farmers acquired better skills to carry out their farming activities. The result carries the expected sign and is significant. This proves that farmers who benefitted from training are more likely to adopt technologies than those who did not benefit from any training. Access to credit is expected to assist farmers to purchase necessary inputs for rice production and pay for hired labour. Many sources of credit give the farmer more chances of securing improved inputs. They also provide farmers with an additional source of investment in new ideas and therefore is expected to be positively related to technology adoption. In this study, access to credit carries the expected sign and is significant. This indicates that the more farmers have access to credit facilities the better their chances of adoption of improved technologies. Membership of an association is expected to assist farmers to have easy access to credit and other production inputs. It can also enhance access to technological information. The sign of the parameter of this variable is expected to be positive. The result carries the expected sign but is not significant. This indicates that gaining membership of an association is not enough reason for farmers to adopt a technology. Information on the influence of farm size on adoption decision is important in order to identify the appropriateness of a technology for a range of farm sizes. The influence of farm size on the Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 158 decision to adopt depends on the technology (Jha et al., 1991). Farm size carries a positive coefficient and is significant, suggesting that farm size has a significant positive influence on adoption decision. In the surveyed districts, the improved varieties are cultivated for both subsistence and commercial purposes, thus making adoption of the varieties more attractive to farmers. The role of gender in rice farming is important. A dummy variable was used to specify the sex of the respondent, 1 for a male respondent and 0 for a female respondent. Male farmers have better access to the resources needed to use improved technologies. They are, therefore, expected to be positively related with the adoption of improved varieties. Women contribute about 70% of the labour used for agricultural production in Sierra Leone but farming resources are mostly controlled by men (Lakoh, 1996). The coefficient of the result is negative and significant in influencing adoption decision. This means that gender significantly influences adoption decision negatively. The result obtained indicates that female farmers were therefore more likely to adopt improved rice varieties, probably because NGOs favour women in the distribution of seed of improved rice varieties. Extension contact is a very important determinant of technology adoption because newly developed technology is introduced to farmers through the activities of extension agents. A farmer whose contact with extension agents is very high is expected to be more familiar with and more knowledgeable about the use of improved agricultural innovations. This variable is expected to be positively related to technology adoption. The result has the expected sign and is significant in determining the adoption of the varieties. Farming experience could take on either a negative or positive sign depending on the length of the period involved. It is expected to demonstrate increasing returns up to a stage and later diminish as more elderly farmers have been reported to be more risk averse, hence less likely to experiment with new technologies. In this study, farming experience carries the expected sign and also has a significant influence on adoption. 4. CONCLUSIONS Nine RARC mangrove rice varieties (CP 4, ROK 4, ROK 5, ROK 8, ROK 9, ROK 10, ROK 21, ROK 22 and ROK 23) were identified. Most farmers in Kambia district are knowledgeable about all the RARC varieties but farmers in Port Loko and Moyamba districts are more knowledgeable about ROK 4, ROK 5 and ROK 10. In all the districts surveyed, ROK 5 and ROK 10 are the most known varieties. A significant proportion of farmers (528) know about ROK 5 and 321 farmers know about ROK 10. The percentages on the knowledge and use of the varieties indicate that some varieties are specific to some areas while others are grown in several localities. A constraint was, however, the fact that improved rice varieties are given local names linked to the source of access, the performance of the variety or physical traits. In the 2014 cropping season, for farmers who are still cultivating the varieties (i.e. the adopters), ROK 5 and ROK 10 are the most adopted RARC varieties. ROK 4 and CP4 were adopted by quite a reasonable number of farmers in all the districts surveyed - ROK 4 by 30 farmers and CP 4 by 23 farmers. The number of non-adopters is high for ROK 5 and ROK 10. The reasons given were the lack of skilled labour for various farm operations, lack of training and extension service. Loss of variety was recorded for all the varieties but ROK 5 and ROK 10 recorded the highest loss due to admixture and the civil war. Farmers indicated their interest in cultivating the lost varieties if they have access to them. Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 159 The cultivation of RARC varieties at least once is high particularly for ROK 5, ROK10 and ROK4. In 2014, however, the cultivation rate dropped, indicating the abandonment and loss of varieties. Reasons given were the lack of access to seed of RARC varieties, admixture and the high rate of bird damage to the varieties. Seed of RARC varieties with awns should be made available and accessible to mangrove rice farmers. The availability of and easy access to improved varieties with awns can increase production and help reduce bird damage. All the variables in the probit model, except farm income and sex, have the expected signs and eight are significant in explaining the use of the RARC improved rice varieties. The positively related and significant variables are farmers’ level of education, training, access to credit, farm size, extension contact and farming experience, as expected. This means that improvements in these major factors would lead to higher levels of technology adoption. 4.1. Recommendations To improve the adoption of improved rice varieties, the following recommendations can be useful. 1. Seed of RARC mangrove varieties should be made available and accessible to mangrove rice farmers 2. Farmers should be provided with credit 3. Extension services should not be biased against women 4. Adult literacy facilities should be establish in farming communities Funding: The study was funded by West Africa Agricultural Productivity Programme (WAAPP) – Sierra Leone Competing Interests: The authors declare that they have no conflict of interests. Contributors/Acknowledgements The authors acknowledge the West Africa Agricultural Productivity Programme (WAPP)-Sierra Leone for financial support and the Rokupr Agricultural Research Center (RARC) of the Sierra Leone Agricultural Research Institute for providing the enabling environment for the study. We are grateful to Dr. C. S. Kamara (Director RARC) for his moral support during the study. We are also grateful to Dr. Ali A. Toure (Agricultural Economist) and Dr. Olupomi Ajayi (Rice Research Coordinator), both out-posted by the Africa Rice Centre to RARC, for backstopping the study and for editing the original manuscript. Views and opinions expressed in this study are the views and opinions of the authors, Asian Journal of Agriculture and Rural Development shall not be responsible or answerable for any loss, damage or liability etc. caused in relation to/arising out of the use of the content. References Agyen-Sampong, M. (1984). Mangrove swamp rice production in Africa. Pp 185-188. In: Cornier-Salem, M-C (Ed.). Dynamics of mangrove use in southern rivers (From Senegal to Sierra Leone). Proceedings of the workshop of Dakar from 8 to 15 May 1984. Orstom Edition Paris 1994. Donkoh, S. A., & Awuni, J. A. (2011). Adoption of farm management practices in lowland rice production in Northern Ghana. Journal of Agricultural and Biological Science, 2(4), 084- 093. Feder, G. R., Just, R. E., & Zilberman, D. (1985). Adoption of agricultural innovations in developing countries: A survey. Economic Development and Cultural Change, 33(2), 255- 298. Foltz, J. D. (2003). The economics of water-conserving technology adoption in Tunisia: An empirical estimation of farmer technology choice. Economic Development and Cultural Change, 51(2), 359-373. Jha, D., Hojjati, B., & Vosti, S. (1991). The use of improved agricultural technology in Eastern Province. In Adopting Improved Farm Technology: A Study of Smallholder Farmers in Eastern Province, Zambia, ed. R. Celis, J.T Milimo, and S. Wanmali, 173-201”. Washington, D.C: International Food Policy Research Institute. Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 160 Lakoh, A. K. (1996). The effectiveness of agricultural research in Sierra Leone. ATPS working paper No.6: African Technology Policy Studies, Nairobi. Sita, D. K., & Ponnarasi, T. (2009). An economic analysis of modern rice production technology and its adoption behaviour in Tamil Nadu. Agricultural Economics Research Review, 22, 341-347. Tiamiyu, S. A., Akintola, J. O., & Rahji, M. A. Y. (2009). Technology adoption and productivity difference among growers of new rice for Africa in Savanna zone of Nigeria. Tropicultura, 27(4), 193-197. Appendix Table 1: Knowledge of RARC mangrove rice varieties District Chiefdom Knowledge of Variety CP4 ROK 4 ROK 5 ROK 8 ROK 9 Count % Count % Count % Count % Count % Kambia Mambolo 81 81.0 66 66 97 97 36 36 34 34 Samu 61 61.0 66 66 98 98 34 34 34 34 Port Loko Kaffu Bullom 26 26.0 61 61 92 92 15 15 14 14 Lokomasama 32 32.0 46 46 90 90 17 17 17 17 Moyamba Bumpeh 22 22.0 21 21 89 89 6 6 9 9 Ribbi 20 20.0 31 31 62 62 8 8 7 7 Total 242 291 528 116 115 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 98 98 51 51 60 60 48 48 Samu 88 88 26 26 22 22 24 24 Port Loko Kaffu Bullom 34 34 1 1 5 5 4 4 Lokomasama 61 61 15 15 11 11 11 11 Moyamba Bumpeh 24 24 5 5 4 4 4 4 Ribbi 16 16 6 6 4 4 5 5 Total 321 104 106 96 Table 2: Mangrove rice varieties grown at least once by farmers District Chiefdom Grown Variety at least once CP 4 ROK 4 ROK 5 ROK 8 ROK 9 Count % Count % Count % Count % Count % Kambia Mambolo 44 44 22 22 88 88 6 6 5 5 Samu 33 33 20 20 84 84 8 8 8 8 Port Loko Kaffu Bullom 6 6 28 28 57 57 4 4 2 2 Lokomasama 7 7 12 12 57 57 1 1 1 1 Moyamba Bumpeh 7 7 4 4 61 61 0 0 0 0 Ribbi 2 2 12 12 33 33 2 2 0 0 Total 99 98 380 21 16 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 94 94 11 11 29 29 12 12 Samu 68 68 7 7 5 5. 6 6 Port Loko Kaffu Bullom 10 10 1 1 1 1 0 0 Lokomasama 33 33 2 2 0 0 1 1 Moyamba Bumpeh 4 4 0 0 1 1 0 0 Ribbi 6 4 1 1 0 0 1 1 Total 215 22 36 17 Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 161 Table 3: Adopters of RARC mangrove rice varieties Table 4: Non-adopters of RARC mangrove varieties District Chiefdom Non-adopters CP4 ROK 4 ROK 5 ROK 8 ROK 9 Count % Count % Count % Count % Count % Kambia Mambolo 33 75 17 77.3 48 54.6 5 83.3 5 100 Samu 29 87.9 18 90 73 86.9 8 100 8 100 Port Loko Kaffu Bullom 5 83.3 14 50 34 59.6 3 75 2 100 Lokomasama 7 100 11 91.7 38 66.7 1 100 1 100 Moyamba Bumpeh 2 28.6 2 50 16 26.2 0 0 0 0 Ribbi 0 0 6 50 8 24.2 0 0 0 0 Total 76 68 217 17 16 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 41 43.6 10 90.9 25 86.2 10 100 Samu 59 86.8 7 100 5 100 5 83.3 Port Loko Kaffu Bullom 7 70 0 0 1 100 0 0 Lokomasama 18 54.5 0 0 0 0 0 0 Moyamba Bumpeh 2 50 0 0 0 0 0 0 Ribbi 0 0 0 0 0 0 0 0 Total 127 17 31 15 Table 5: RARC mangrove rice varieties lost by farmers District Chiefdom Varieties Lost CP4 ROK 4 ROK 5 ROK 8 ROK 9 Count % Count % Count % Count % Count % Kambia Mambolo 24 54.5 12 54.5 37 42 5 83.3 3 60 Samu 26 78.8 12 60 69 82.1 6 75 8 100 Port Loko Kaffu Bullom 1 16.7 13 46.4 27 47.4 3 75 1 50 Lokomasama 3 42.9 5 41.7 22 38.6 0 0 0 0 Moyamba Bumpeh 1 14.3 1 25 10 16.4 0 0 0 0 Ribbi 0 0 6 50 7 21.2 0 0 0 0 District Chiefdom Adopters - Still Grown the Variety CP 4 ROK 4 ROK 5 ROK 8 Count % Count % Count % Count % Kambia Mambolo 11 17.5 5 7.9 40 63.5 1 1.6 Samu 4 21.1 2 10.5 11 57.9 0 0 Port Loko Kaffu Bullom 1 3.3 14 46.7 23 76.7 1 3.3 Lokomasama 0 0 1 4 19 76 0 0 Moyamba Bumpeh 5 10.6 2 4.3 45 95.7 0 0 Ribbi 2 7.4 6 22.2 25 92.6 2 7.4 Total 23 30 163 4 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 53 84.1 1 1.6 4 6.3 0 0 Samu 9 47.4 0 0 0 0 1 5.3 Port Loko Kaffu Bullom 3 10 0 0 0 0 0 0 Lokomasama 15 60 0 0 0 0 0 0 Moyamba Bumpeh 2 4.3 0 0 0 0 0 0 Ribbi 4 14.8 0 0 0 0 1 3.7 Total 86 1 4 2 Asian Journal of Agriculture and Rural Development, 6(8)2016: 152-162 162 Total 55 49 172 14 12 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 35 37.2 6 54.5 22 75.9 11 91.7 Samu 59 86.8 7 100 5 100 5 83.3 Port Loko Kaffu Bullom 2 20 1 100 1 100 0 0 Lokomasama 13 39.4 2 100 0 0 1 100 Moyamba Bumpeh 1 25 0 0 1 100 0 0 Ribbi 2 33.3 1 100 0 0 0 0 Total 112 17 29 17 Table 6: RARC mangrove rice varieties abandoned by farmers District Chiefdom Abandonment Rate CP4 ROK 4 ROK 5 ROK 8 ROK 9 Count % Count % Count % Count % Count % Kambia Mambolo 9 20.5 5 22.7 9 10.22 0 0 2 400 Samu 3 9.1 6 30 4 4.8 2 25 0 0 Port Loko Kaffu Bullom 4 66.7 1 3.6 7 12.3 0 0 1 50 Lokomasama 4 57.1 6 50 16 28.1 1 100 1 100 Moyamba Bumpeh 2 28.6 1 25 6 9.8 0 0 0 0 Ribbi 0 0 0 0 1 3 0 0 0 0 Total 22 19 43 3 4 District Chiefdom ROK 10 ROK 21 ROK 22 ROK 23 Count % Count % Count % Count % Kambia Mambolo 6 6.4 4 36.4 3 10.3 1 8.3 Samu 0 0 0 0 0 0 1 16.7 Port Loko Kaffu Bullom 4 4 0 0 0 0 0 0 Lokomasama 5 15.1% 0 0 0 0 0 0 Moyamba Bumpeh 1 25.0 0 0 0 0 0 0 Ribbi 0 0 0 0 0 0 0 0 Total 16 4 3 2 Table 7: Parameter estimates for probit model Parameter Estimates Parameter Estimate Std. Error Z Sig. PROBIT a Age -0.126*** 0.187 -0.673 0.005 Educational level 0.014*** 0.047 0.291 0.010 Farmer's income -0.166 0.056 -2.964 0.255 Training 0.055* 0.023 2.391 0.065 Access to credit 0.202** 0.192 1.052 0.038 membership of association 0.010 0.159 0.060 0.352 Farm size 0.044*** 0.022 2.000 0.008 Sex -0.167*** 0.183 -0.912 0.009 Extension contact 0.289*** 0.325 0.889 0.007 Farming experience 0.072** 0.096 0.750 0.041 Intercept 0.369 1.080 0.341 0.733 a. PROBIT model: PROBIT(p) = Intercept + BX Note: ***, ** and * indicate significance at 1 percent, 5 percent and 10 percent probability levels, respectively