Adoption Determinants of Modern Rice Cultivars among Smallholders of Northern Iran

: Several agronomic innovations and precision technologies have been developed and disseminated in rice cultivation, but adoption is often lagged, particularly in developing countries. The objective of this work was to collect information on the use of improved rice cultivars and factors a ﬀ ecting adoption through a farmers’ survey in Guilan Province, northern Iran. About 4 out of 10 farmers (42.8%) used improved rice cultivars. Above half of the farmers (55.2%) recognized the high proﬁtability of this technology and half (49.8%) realized that modern rice cultivars are of great importance. However, there were some farmers who perceived no proﬁtability (14.8%) or no importance (12.5%) regarding this technology. Additionally, 9.8% were not aware of the proﬁtability, and 16.2% were not aware of the importance of the technology. The majority of the farmers (60.5%) collaborated with fellow farmers about common production problems. Nevertheless, low cooperation with extension agents was noted for 58.5% of the farmers and low attendance of local agricultural o ﬃ ces activities was noted for 59.8% of the farmers. A logistic regression analysis showed that perceived proﬁtability and perceived importance of modern varieties, background in rice farming, and size of livestock holdings were positively associated with the adoption of improved cultivars. It can be concluded that the low acceptance of modern rice cultivars in the study area is related to current farming socio-economic conditions. Dissemination of information on improved cultivars from experienced farmers to other farmers should be considered to promote adoption, along with ﬁnancial incentives to low-income farmers.


Introduction
Rice (Oryza sativa L.) is a major crop in Iran along with wheat and barley. Based on statistics of the Ministry of Jihad-e-Agriculture of Iran, the country produced about 2.7 million metric tonnes of rice in 2011 [1]. Rice is mainly produced in the northern part of Iran near the Caspian Sea. More than 80% of the rice belt is spread in the provinces of Mazandaran and Guilan. Guilan Province is the most important province in this regard, followed by Mazandaran Province and much smaller areas in the rest of the country [2]. Despite the low productivity of local varieties (ranging between 2.5 to 3.5 t/ha), more than 80% of the total rice area in Iran is cultivated with these varieties owing to their high-quality traits (i.e., long slender grain, head rice recovery of 60% to 63%, intermediate amylose calculations [26]. The farm structure model emphasizes profitability as well as economic incentives for innovation adoption, while farmers' ability to bear the risk involved in trying an innovation is also an important determinant of adoption. According to the rationale of this model, the existence of these incentives increases the profitability and consequently the acceptance of an innovation. An assumption of the model is that acceptance behavior is initially subordinate to the ability to act on innovation, and that the acceptance of innovation is limited by the lack of economic resources. Another assumption of the model is that individuals are unable to adopt technologies because they do not have the financial resources to operate them. The farm structure model states that the most important limiting factors for the adoption of innovations are access to capital and land [27], while the role of information and attitude factors has received less attention [28].
As already explained above, understanding the factors affecting the adoption of modern cultivars is essential for shaping appropriate policy measures to improve the livelihoods of smallholders following the assumptions of the farm structure model. By identifying what factors affect adoption, it is possible to identify how to induce adoption among non-adopters. Thus, the objective of this work was to highlight the drivers of the use of modern varieties among rice growers in Guilan Province, Iran.

Study Area and Sample Selection
Farmers of Rudsar County of northern Iran (Guilan Province) were selected for the study. All rice farmers living in the county were considered, irrespective of the adoption of improved varieties. The smallest sample size was determined to be 370 people according to Bartlett et al. [29], but an extra 10% was added to avoid non-response errors. The survey was conducted in 2015.

Data Collection
Data were collected on the basis of a questionnaire that was developed after a literature review [15,18,19]. The questionnaire assessed the adoption of improved varieties (non-adoption = 0, adoption = 1) as the dependent variable and various attributes as the independent variables. Among the independent variables, profitability referred to the capability of generating profits from the use of modern cultivars, while importance referred to adaptability, stability, and productivity of these cultivars under diverse growth conditions, reflecting farmers' level of necessity of this material. Independent variables were assessed with a 5-point scale: very low (= 1), low (= 2), moderate (= 3), high (= 4), and very high (= 5). The content validity of the research tool was tested by experts in agricultural extension and agronomy from the Jihad-e-Agriculture Organization. Before the study, a pilot study was conducted with 30 participants outside the study sample to assess the reliability of the questionnaire. Cronbach's alpha was 0.91, which indicated good internal consistency of the questionnaire. Face-to-face interviews were conducted for data collection, with participants being recruited using simple random sampling based on farmers' lists from the local authorities.

Data Analysis
Descriptive statistics were calculated for the independent and dependent variables. The independent t-test (for continuous variables) and Mann-Whitney test (for ordinal variables) were used to compare characteristics of adopters and non-adopters. In addition, logistic regression was implemented to determine the association of the independent variables with the adoption of the modern varieties. Logistic regression was used because of the dichotomous nature (no, yes) of the dependent variable. The independent variables in the model included perceived profitability and perceived importance (i.e., necessity) of the modern varieties, social participation of growers, demographic data, economic data, and technical data (related to farm structure). In statistical model building, it is common to minimize variables until the most parsimonious model fits the data, which also results in numerical stability and generalizability of the results; therefore, the backward stepwise selection approach was used. Data analysis was performed using Statistical Package for the Social Sciences (SPSS Inc, Chicago, USA).

Farmer Profile
Participant ages ranged from 25 to 80 years, with a mean age of 51.26 (Table 1). Most farmers were in the age group of 40-50 years. Almost one-third of the farmers (31.5%) had a university degree and 26.2% had graduated high school. Nevertheless, the greatest part (40.0%) had primary or medium education, and a small fraction (2.3%) were illiterate. Farmers had variable experience in rice cultivation from 5 to 65 years with a mean of 29.73 years. Most farmers (34.2%) were in the category of 20-30 years of experience. Most farmers did not have livestock, but the average livestock of the farmers in the sample was 4.23 heads (Table 2). Moreover, most farmers were in the agricultural income group of <50 million Iranian rials (IRR) and non-agricultural income group of <50 million IRR. Rice production averaged 2153.25 kg/year. Farmers in the sample had an average of 48,632,500 IRR agricultural expenses and most were in the group of >30 million IRR. The average farm area was calculated to 1.60 ha, but most farmers (67.2%) had farm area <1 ha ( Table 3). The highest proportion (53.0%) was associated to >3 field plots. The average distance of the household from farm was 2.10 km, but most households (52.5%) were within a distance of <1 km and few (17.4%) were within a distance of >3 km. The average distance of the household from the road was 0.92 km. For most farmers (56.2%) the distance of the household from the road ranged between 0.1 and 1 km.  High levels of cooperation with colleague farmers, with Dehyari (local municipality), and with rural entities and cooperatives were noted among participants (Table 4). Nevertheless, there was low participation in Jihad-e-Agriculture programs and poor collaboration with extension services. It is noted that 59.8% of the farmers had very low or low participation in Jihad-e-Agriculture programs and a similar proportion had very low or low co-operation with extension agents.

Utilization of Improved Cultivars
Most farmers (57.2%) did not adopt the technology of improved cultivars-this technology was adopted by 42.8% of the respondents. Nevertheless, about half of the farmers (55.2%) perceived the improved cultivars as highly profitable and half of the farmers (49.8%) perceived them as highly important ( Table 5). The most convenient information source among the rice farmers surveyed was information from other farmers (mean = 2.95), followed by TV and radio (mean = 2.51), while extension agents, education movies, and research centers showed low mean values (1.60, 1.59, and 1.28, respectively) ( Table 6).

Differences between Adopters and Non-Adopters of Improved Cultivars
Adopters of improved cultivars perceived higher importance and profitability of improved cultivars and had higher social participation than non-adopters (Table 7). Moreover, adopters were slightly older and had lower education than non-adopters. Adopters were more experienced farmers, had greater number of livestock and poultries, higher agricultural income, and greater number of land parcels than non-adopters (Table 7).

Logistic Regression Analysis
The logistic regression model showed high fit to the data with correct prediction 81.0% (Table 8), indicating the capability of the model to predict a high level of the amount of variance of the dependent variable. Perceived profitability and perceived importance of the improved varieties, background in rice cultivation, and size of livestock were positively associated with the adoption. Experience in rice cultivation, perceived importance, and perceived profitability of the improved varieties shared the maximum contribution in the prediction of adoption according to Wald values (Table 8). Perceived profitability showed the greatest contribution in forecasting adoption according to Exp (B) (i.e., the exponentiation of the B coefficient) odds ratio values.

Discussion
This study determined drivers of the adoption of modern rice cultivars among smallholders of northern Iran, for which no relevant data exist in the literature. Significant relationships between the use of modern rice cultivars and several independent variables were identified with logistic regression analysis. Perceived profitability and perceived importance of improved seeds, background in rice farming, and livestock holdings were positively associated with the adoption of improved cultivars. Today, farmers, advisors, and policy makers face a wide range of technologies with uncertainties in the agri-food chain which they must deal with. Therefore, our findings assist in better understanding the adoption of this technology and thus may be important for seed delivery systems, pointing to policies for promoting acceptance of modern cultivars among non-adopters. The novelty of this research stems from the paucity of data in the literature on modern cultivar adoption concerning rice, as very few studies have assessed the adoption of modern rice cultivars, while no study has identified factors affecting the adoption of modern rice cultivars in Guilan Province, a major rice producing-province of Iran. Moreover this study provides a detailed regional representation of farmers included in the project, based on well-established data sources. Therefore, it adds useful information on the adoption of modern rice cultivars in the study area, and by identifying what factors affect adoption, one can determine how to induce the adoption of modern cultivars among non-adopters.
The perceived profitability of the technology was positively associated with the acceptance of modern rice cultivars (p < 0.01), in agreement with previous literature [30,31], which showed that stating the advantages and benefits of modern technologies to rice growers can encourage their use. A technology is useful to the extent that it improves production and profit, and it is applied to the extent that it generates and satisfies users' needs. Economic profitability is a crucial factor before farmers make adoption decisions [32]. Therefore, a lag in the spread of improved seeds use could be due to the perception of low profitability among farmers [33,34]. Other studies have shown that the adoption of different technologies was favored by positive perceptions of profitability of each technology among farmers [35,36]. Additionally, doubts about the benefits of variable fertilization rate among grain growers of Western Australia hindered adoption of the technology [37]. In our study, the perceived importance of improved seeds favored the adoption of modern cultivars (p < 0.01), showing that technologies that are tailored to farmers' needs are more likely to be accepted. The importance and usefulness of the technology have been reported to be significant determinants of the adoption of a technology in previous studies [38,39]. From this point of view, research centers should determine farmers' needs and provide technologies that are more interesting to them.
The adoption of improved varieties was positively favored by rice cultivation experience (p < 0.01), as also reported elsewhere [15,40,41]. Similarly, the adoption of current technology in rice management in India was constrained by lack of experience [42]. Farming experience reflects knowledge that farmers have gained, and can be used to tackle farm production problems [33]. Farming experience can play an important role early in the adoption because the advantages of the technology are still being assessed. From this point of view, to promote adoption there is a need for progressive growth of the technology and constant farmer training [43]. This finding is rational, as experienced farmers are generally more competent with better access to necessary information about modern technologies, implying that accumulated knowledge may help in better evaluating information and thus impacting adoption decision [40]. It is noteworthy that apart from solely being adopters, farmers are involved in the generation of innovations [44].
The adoption of modern cultivars was positively associated with number of livestock (p < 0.05); that is, farmers with large number of livestock were more willing to adopt the improved varieties.
In agreement with the current finding, a previous study reported that owning oxen favored the probability of adopting modern varieties [45]. It seems that animal power is a significant factor in small cropped areas because it is available for plowing where tractor mechanization does not exist or perhaps is unprofitable. Moreover, livestock can be a source of off-farm income that can be exploited for supplying farming inputs. Livestock shows opportunity of working capital for farming tasks [46]. From this point of view, livestock can contribute to the diversification of income sources, and can thus allow the reallocation of other resources for improving farm productivity [47]. However, it is also possible that farmers with income from livestock may be reluctant to accept a new technology unless non-farm alternatives are more rewarding to them [35,48]. Additionally, the number of livestock held likely affects access to animal manure and hence soil fertility. From this point of view, the number of livestock may also be a proxy for wealth, and hence can increase farmers' capacity to bear the risk of trying a new cultivar. However, smaller and poorer farmers operating on less-favorable plots may correctly perceive that the new cultivars are not more profitable for them and thus that the new technology is "not important" for addressing their most pressing constraints, which could vary from labor availability at peak periods of the production cycle for the new varieties to access to finance for inputs.
The current study also showed that adopters had significantly higher agricultural income and greater number of parcels than non-adopters. Farm income promotes technology adoption by overcoming the credit constraints of rural households in developing countries [49]. Farm income offers liquid capital to farmers, thus enhancing capacity for purchase of inputs such as improved seed and fertilizers. It is likely that a great number of plots may allow for some kind of experimentation with the new technology by spreading out risk. In this context, the new technology, as developed by the researchers in the study area, may be suitable for only a subset of the farmers in the area (e.g., those who are wealthier and on more favorable plots). This problem seems to lie in the technology, not necessarily in those who do not adopt, as farmers who do not adopt may be acting very rationally given their circumstances. This point was made in much of the farming systems research of the 1970s and 1980s, when researchers at CIMMYT (International Maize and Wheat Improvement Center) developed the notion of "recommendation domains" for different types of technology (i.e., tailoring different products for different groups of farmers, even farmers living in the same area) [50].

Conclusions
This work highlighted drivers of the adoption of improved rice varieties among farmers of Guilan Province. Farmers in the study area are expected to use improved rice varieties in their fields if they perceive that the investment is important and will be profitable. On the other hand, low adoption seems likely to be connected with the inconsistency of the modern varieties with current farming socio-economic conditions. Where benefits from the adoption of sustainable technologies are expected to accrue only to people outside agriculture and where there are no markets for the benefits of those technologies, the levels of adoption could be sub-optimal. Based on the findings of this study, the most likely way to convince farmers to accept an innovation is the provision of basic information about the advantages of the technology through extension services, demonstration projects, participatory research, and close cooperation with other farmers. Using experienced farmers for information transfer to other farmers should be considered to promote adoption, along with financial incentives to low-income farmers. It is possible that tailoring different modern varieties to different groups of farmers would be required, as there may be a range of different recommendation domains. Appropriate policies facilitating the adoption of improved varieties, such as incentives, subsidies, or price supports as well as growth of the number of people involved in extension and their quality, could be put forth according to findings. Nevertheless, it should be kept in mind that while such incentives may be appropriate in the short term to mitigate the risk associated with adoption and to help farmers learn more about the technologies by trying them out, they may not be a financially sustainable strategy in the long run, as the budgetary costs of such measures can be high.
Author Contributions: D.A. conducted the survey, gathered data, and analyzed data; M.S.A. conceived the idea, planned the research, analyzed data, interpreted data, and wrote the article; A.B. contributed to data interpretation and edited the article; C.A.D. contributed to data interpretation and edited the article. All authors approved the final version of this article.