This article was downloaded by: [Canadian Research Knowledge Network] On: 26 June 2009 Access details: Access Details: [subscription number 783016891] Publisher Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Sustainable Agriculture Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t792306915

Factors Influencing the Adoption of Pasture Crop Rotation in the Semiarid Area of China's Loess Plateau Xuefeng He ab; Huhua Cao c; Fengmin Li a a MOE Key Laboratory of Arid and Grassland Ecology, School of Life Science, Lanzhou University, Lanzhou, China b R&D Center for Finance of Chongqing, Chongqing Institute of Technology, Chongqing, China c Department of Geography, University of Ottawa, Ottawa, Ontario, Canada Online Publication Date: 11 June 2008

To cite this Article He, Xuefeng, Cao, Huhua and Li, Fengmin(2008)'Factors Influencing the Adoption of Pasture Crop Rotation in the

Semiarid Area of China's Loess Plateau',Journal of Sustainable Agriculture,32:1,161 — 180 To link to this Article: DOI: 10.1080/10440040802121551 URL: http://dx.doi.org/10.1080/10440040802121551

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf This article may be used for research, teaching and private study purposes. Any substantial or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Factors Influencing the Adoption of Pasture Crop Rotation in the Semiarid Area of China’s Loess Plateau

1540-7578 1044-0046 WJSA Journal of Sustainable Agriculture Agriculture, Vol. 32, No. 1, May 2008: pp. 1–31

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao,ofand Journal Sustainable Li Agriculture

Xuefeng He Huhua Cao Fengmin Li

ABSTRACT. Alfalfa (Medicago Sativa L.) pasture crop rotations (PCR) have been promoted in the semiarid region of northern China for many years, with seemingly limited success. Using a logistic regression model, this paper evaluated the determinants of adoption of PCR among smallholder farmers in rain-fed farming systems, based on a survey of 238 farmers in the hilly area of the Loess Plateau in 2004 and 2005. Results indicated that farmers’ socio-economic background, farm-related conditions, and institutional factors significantly affect their decision of whether or not to adopt PCR. First, the age of farmers, the size of their family, the number of ruminants owned, and farmers’ attitude toward PCR have significant Xue-Feng He is affiliated with the MOE Key Laboratory of Arid and Grassland Ecology, School of Life Science, Lanzhou University, Lanzhou, 730000 China and the R&D Center for Finance of Chongqing, Chongqing Institute of Technology, Chongqing, 400050, China (E-mail: [email protected]). Huhua Cao is affiliated with the Department of Geography, University of Ottawa, 60 University St. Ottawa, Ontario, Canada K1N 6N5. Feng-Min Li is affiliated with the MOE Key Laboratory of Arid and Grassland Ecology, School of Life Science, Lanzhou University, Lanzhou, 730000 China (E-mail: [email protected]). The authors thank the reviewers for their valuable comments on the manuscript. A “973” project (2007CB106804) Ministry of Education (MOE) Innovation Team Program and PhD candidate project from MOE (20060730027) support this research. Address correspondence to: Fengmin Li (E-mail: [email protected]). Journal of Sustainable Agriculture, Vol. 32(1) 2008 Available online at http://jsa.haworthpress.com © 2008 by The Haworth Press. All rights reserved. doi:10.1080/10440040802121551

161

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

162

JOURNAL OF SUSTAINABLE AGRICULTURE

positive effects on the adoption of PCR, where the level of the family income is low. Second, the slope of the field and fallow practice variables also play a critical role in the adoption of PCR, suggesting the importance of targeting the promotion of PCR among farmers in areas affected by soil erosion. The analysis also showed that farmers’ perception of the problem of soil erosion is positively correlated with adoption of PCR, having the most significant impact with the highest estimated elasticity. The likelihood a farmer who perceived soil erosion as a problem would adopt PCR was 225 greater than that of a farmer who did not have such perceptions. Third, the institutional variables, which are represented by contacts with extension workshops, a participatory approach, the credit and training facilities obtained, and participation in the Grain-for-Green project (PGFG), all had significant positive effects on the adoption of PCR. On the other hand, participation in the rainwater-harvesting project and in the supplementary irrigation project did not show any correlation with the adoption of PCR by farmers. Variables such as education level, the size of farms, farm tenure, off-farm employment, and distance between the farm and the farmers’ residence did not significantly influence the adoption of PCR. This information will help prioritize factors that affect decision-making in the adoption of PCR and will provide insights into ways of increasing the adoption of PCR by farmers.

KEYWORDS. Adoption, Loess Plateau, logistic regression, pasturecrop rotation

INTRODUCTION Many parts of northern China are known for their substantial soil erosion. This is particularly the case of the semi-arid Loess Plateau. Soil erosion is mainly due to an increase in human and livestock population pressures, resulting in inappropriate soil and crop management practices, including continuous cropping, the increased occurrence of reduced fallow periods, and overgrazing for subsistence requirements. Jiang (1997) reported that the surface soil loss in most hilly areas of the Loess Plateau reaches approximately 5000 to 10,000 Mg km−2 y−1. Unfortunately, the eco-environmental situation in this area continues to worsen, especially due to accelerated rates of soil erosion (Long et al., 2005), which is not only a major obstacle to productive agriculture, but also an important factor in environmental deterioration. In this area, farmers rely heavily on livestock to reduce the use of manpower, both as a source of cash income and to meet their requirements for

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

163

nutritious food and compost. There is a scarcity of feed, given the limited arable land available for subsistence requirements, as upland agriculture is rendered unsustainable by soil erosion. This, in turn, leads to overgrazing. In response, research institutions developed agro-techniques for soil conservation. One such technique is pasture crop rotation (PCR) farming (Wang et al., 2006). PCR is promoted as a sustainable agricultural technology that can increase, sustain, and improve the physical properties of the soil, as well as prevent soil erosion (Lu, et al., 1999; Lazányi, 2000; Yang et al., 2001; Jeffrey et al., 2002; Tang & Ranst, 2005; Benediktas and Genovaite, 2002; Li et al., 2002). In the late 1990s, sustainable alfalfa (Medicago Sativa L.) crop rotation (3-year crop/3-year alfalfa) was introduced in the Loess Plateau. The crop rotation system consisted of 3 years of continuous crops such as spring wheat (Triticum aestivum L.), potato (Solanum tuberosum L.), and maize (Zea mays L.) and 3 years of continuous alfalfa (Medicago saliva L.). In the rain-fed agricultural production zone of the Loess Plateau, the introduction of Medicago Sativa L. into this system was the optimal solution to improve the quantity and depth distribution of soil organic matter due to the high biomass production of roots. According to Stinner and colleagues (1992), the use of forage legumes such as Medicago Sativa L. in a rotation sequence is a cornerstone of successful sustainable agriculture, and is practiced in many parts of the world. During the 20th century, farmers around the world replaced forage legume rotations with synthetic nitrogen (N) fertilizers. Some researchers, such as Crews and Peoples (2004) have concluded that obtaining N from legumes, rather than from industrial sources, is potentially more sustainable, offering a more environmentally sound and sustainable source of N to crop systems. This is because it can support biological nitrogen N2 fixation. In the Loess hilly area, in most cases, farmers are encouraged to plant alfalfa (Medicago Sativa L.) for the purpose of nitrogen N2 fixation. It plays a role in reducing soil erosion and soil amelioration, as it has deep roots for nutrient capture and recycling. In addition, the hedgerow species of alfalfa was chosen for rotation because it is useful as fodder for livestock and produces substantial amounts of biomass. A more recent study by Adesina and coworkers (2000) showed the role of alfalfa not only as livestock feed but also in soil conservation. In the semiarid area of China’s Loess Plateau, there is ample evidence to show that PCR has the potential to increase substantially the amount and quality of forage supply and to promote sustainability in different agro-ecosystems (Wang et al., 2005). Lu and colleagues

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

164

JOURNAL OF SUSTAINABLE AGRICULTURE

(2004) reported that crop rotation with alfalfa may be the best option for soil conservation. Hence, alfalfa crop rotation could have a role in meeting farmers’ objectives of doubling crop production and improving livestock-production. The development of a 3-year corn/3-year alfalfa rotation compatible with individual farming systems could be more profitable and need fewer fertilizer inputs. It could contribute to the sustainable production of crops and livestock in the Loess hilly area. According to Wang and coworkers (2006), PCR is regarded as a key planting system, and is especially important for food security in the semiarid area. Despite the clear benefits in the context of soil quality and agricultural production, the extent of PCR adoption is still poor in these ecological and socio-economic settings. A large number of studies have been examined crop rotation, focusing more on the purely agronomic aspects. Unfortunately, very little attention has been given to studying factors influencing the adoption of PCR, particularly in the Loess Plateau. The reasons for the poor response to the adoption of PCR are unclear, and without a better understanding, attempting any further introduction of PCR would be difficult. Therefore, the aim of this study is to examine the determinants of farmers’ decisions regarding PCR.

MATERIALS AND METHODS Research Site The study area is located in the northern mountainous region of the province of Gansu in the western part of the semiarid Loess Plateau, which belongs to the rain-fed agricultural production zone, where the ecosystem is very fragile and subject to soil erosion. The study area is a part of the Loess hilly area characterized by a medium, temperate, semiarid continental climate, with an annual mean air temperature of 6.5°C (Jia et al., 2005). The mean annual precipitation in the area is approximately 350 mm, and the ratio of rainfall to free evaporation is 3.36 (Jia et al., 2005). Hence, the scarcity of rainfall and drought represent the main constraints to production. Major crops in the area are spring wheat (Triticum aestivum L.), beans (Phaseolus vulgaris), potato (Solanum tuberosum L.), millet (Panicum miliaceum), and increasingly alfalfa (Medicago Sativa L.). Agricultural activities that require low-input crop management are carried out mostly in upland fields. The rapid growth in population and the pressure on land use have led to a reduction of fallow duration below the

He, Cao, and Li

165

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

minimum threshold required for the systems’ sustainability. Research and development agencies have implemented several agro-techniques for soil conservation, including cropping in rotation with alfalfa, improved crop variety, water catchments, supplementary irrigation projects, and the ‘Grain-for-Green’ project among farmers. Despite research interest for more than 20 years, the adoption of PCR has been lower than expected. Hence, understanding the factors affecting farmers’ adoption of PCR is critical to success in the implementation of PCR.

Sample Selection and Survey For the purpose of this study, the heads of the selected households were interviewed in 2004 and 2005, using a structured and semi-structured questionnaire. To guide the development of the formal questionnaire, an informal survey was conducted using interviews with key informants including 10 farmers, government officials, nongovernment organizations, and other research agencies. A visual assessment was used to gather site information. A two-stage sampling technique was used to select the sample of farmers. In the first stage, five villages were purposefully selected out of the 12 found in the mountainous northern region, on the basis of the severity of soil erosion. A complete list of 12 villages where PCR had been previously introduced was available. The five villages selected were Yejiuchuan, Zhonglianchuan, Dadayao, Salagou, and Gaoyaogou. These villages are representative of rain-fed farming systems in the semiarid Loess Plateau and of the different agro-ecological and socio-economic environments under which farmers operate. In the second stage, farmers in the five sampled villages were listed. The survey covered 238 farmers randomly selected in proportion to the number of farmers (1:3) in each village. Two farmers were eliminated from the analysis because of missing values in their data. The type of data collected concerned variables of the crop rotation farming system; farmer and village factors such as age, education, income, farmers’ perception on soil erosion, contact with research or extension agencies, and farmers’ attitude toward PCR; farm characteristics; external and institutional factors, and farmers’ preferences with regard to PCR. The results from this survey indicate a low adoption rate among the farmers who use PCR. About 34% of surveyed farmers have adopted PCR. The rate of adoption varied from village to village and ranged from 23% (Dadayao) to 48% (Yejiuchuan). Table 1 shows the characteristics’ profile of enumerated farmers.

166

JOURNAL OF SUSTAINABLE AGRICULTURE

TABLE 1. Descriptive statistics of variables used in the empirical econometric model

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

Variables

Minimum

Adoption 0 Farmers socio-economic variables AGE 25 EDUCATION 0 FAMILYSIZE 1 INCOME 1 PARTICIPATORY 0 ATTITUDE 0 Farm structural variables FARMSIZE 0.40 OFF-FARM 0 RUMINANT 1 DISTANCE 0 FALLOW 0 Institutional policy variables CREDIT 0 EXTENSION 0 TRAINING 0 GFG PROJECT 0 RHSIT PROJECT 0 TENURE 0 Ecological variables SLOPE 0 PERCEPTION 0

Maximum

Mean

Standard Deviation

1

0.34

0.47

76 12 9 5 1 1

42.74 6.18 4.39 3.00 0.68 0.38

8.67 2.56 1.22 1.06 0.47 0.49

2.24 0.57 4.70 7.63 0.08

0.77 0.50 3.18 5.46 0.27

1 1 1 1 1 1

0.56 0.29 0.25 0.66 0.39 0.82

0.50 0.46 0.43 0.48 0.49 0.39

1 1

0.58 0.79

0.50 0.41

4.67 18 30 1 1

Modeling PCR Adoption The farmers’ decision to adopt or reject PCR at any time is influenced by a complex set of socio-economic, farm-related, institutional, and biophysical factors. In this study, a dichotomous dependent variable (ADOP) was identified. The binary variable has a value of 1 for farmers who adopt PCR, and 0 for those who do not. Estimation of this type of relationship requires the use of qualitative response models. The technology-adoption decision of a farmer is assumed to be motivated by utility maximization (Kebede et al., 1990; Adesina et al., 2000). The assumption is that the desire to maximize the expected utility influences whether the farmer adopts PCR. If the utility perception of PCR is greater than that of continuous cropping, then the adoption of PCR is expected.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

167

As a popular statistical technique in which the probability of a dichotomous outcome (such as adoption or nonadoption of PCR) is related to a set of explanatory variables, the binary logistic regression has been widely applied in adoption studies because it represents a close approximation to the cumulative normal distribution. It is relatively simple from a mathematical viewpoint and lends itself to a meaningful interpretation (see Adesina et al., 2002; Somda et al., 2002; Asfawa et al., 2004; and Chianu et al., 2004). The Hosmer and Lemeshow goodness-of-fit test is one of the most reliable tests of model fit for binary regression, estimated by the maximum likelihood estimates of the coefficients. The probability of farmers’ decision to adopt PCR, Pi is a cumulative density function F evaluated at XiB, where Pi the probability that an event occurs (i.e., the farmer adopts PCR) for an observed set of variables Xi. Thus, the probability of adoption can be written as (He & Li, 2006). (1)

Pi = e zi /1 + e zi

Zi, the utility of PCR, is a linear function of explanatory variables (X), and is expressed as n

Z i = b 0 + ∑ b i X ki

(2)

i =1

Where β0 is the intercept term, and β1, β2, β3, … βi are the coefficients associated with each explanatory variable X1, X2, X3, … Xki. The estimation form of logistic transformation of the probability of farmers’ decision to adopt PCR can be represented as n

ln( Pi / 1 − Pi ) = ln(odds) = b 0 + ∑ b i X ki = Z i

(3)

i =1

These estimates show the relationship between the independent variables and the dependent variable, where the latter is on the logit scale. Pi denotes the probability of adoption by the ith farmer, and (1 – Pi) is the probability of nonadoption. The odds to be used can be defined as the ratio of the probability that a farmer (Pi) to the probability that not (1 – Pi), namely odds = (Pi/1 – Pi). The odds ratio is the exponentiation of the B coefficient, where Zi is also referred to as the log of the odds ratio in favor of adoption.

168

JOURNAL OF SUSTAINABLE AGRICULTURE

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

EXPLANATORY VARIABLE SELECTION AND HYPOTHESES Decision-making in PCR adoption is largely the result of the behavior of individual farmers, which is influenced by internal and external factors. Similar to any other new agricultural technologies or soil conservation adoption research (Gillespie et al., 1996; Adesina et al., 2000; Sheikh et al., 2003; Herath et al., 2003), the adoption of PCR is a complicated process that may be influenced by a set of explanatory variables such as socio-economic, farm structural, institutional, and ecological variables, which are hypothesized to determine farmers’ decisions.

Socio-Economic Variables and Hypotheses The AGE variable measures age of the household head. The EDUCATION variable measures the level of education of the household head. According to the theory of human capital, young heads of households have a greater chance of being updated on innovations in terms of agricultural methods; thus, new technologies are more likely to be adopted by young farmers than by older ones, who have a tendency to stick to traditional methods (Adesina, et al., 2000; Comer et al., 1999; Sidibé, 2005). Hence, it is hypothesized that AGE will have a negative influence on the decision to adopt PCR, as older farmers are less likely to accept it. In most adoption studies, it has been shown that agricultural technological practices are positively related to education. Hence, it is expected that better educated farmers are more likely to adopt PCR than less educated ones. FAMILYSIZE measures the size of the household. In our area of interest, this might be a proxy for consumption pressure. A large household with a greater number of dependants aged less than 10 years or above 65 years, who are unproductive and yet consume a significant proportion of the household income, is likely to be poorer. According to Enyong and colleagues (1999), a major concern of the farmers is that of ensuring a minimum level of food production to meet the family’s needs, thus causing them to adopt only such technologies that they consider as having minimum risk and highest payoff. It is hypothesized that it will have a negative influence. INCOME measures the level of the family income (rankings of total income: (US $) y−1: 1, <500; 2, 500 ≤625; 3, 625 ≤1000; 4, 1000 ≤ 1500; 5, ≥1500). A higher family income implies the ability to invest in PCR and to bear the risk associated with its adoption. A positive relationship should be expected.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

169

PARTICIPATORY is a dummy variable that assume, the value of 1 if the farmer is a member of a participatory society or participatory evaluation of PCR and 0 if otherwise. In Southeast Asia, Cramb (2005) reported that participatory evaluation can increase the likelihood of improved management practices. It is hypothesized to influence the decision on the adoption of PCR positively. ATTITUDE is a dummy variable that represents the farmers’ subjective assessment of yield performance using PCR instead of continuous cropping (1 if positive assessment, 0 if otherwise). In this study, attitude toward PCR was evaluated by accessing the farmers’ opinion directly through interviews. It is assumed that attitudes largely depend on actual profitability of PCR as perceived by farmers. In this model, farmers with positive feelings toward PCR were hypothesized to have a positive impact on the outcome.

Farm Structural Variables and Hypotheses FARMSIZE variable is the total size of cultivated land in hectares. Chianu & Tsujii (2004) noted that the effect of farm size on farmers’ likelihood of adopting soil conservation has been mixed. According to Sidibé and colleagues (2005), farm size has different effects on the probability of adoption, depending on technological characteristics and the institutional setting. Therefore, this variable is difficult to predict. OFF-FARM is a dummy variable that measures the likelihood of farmers’ off-farm activities. It will reduce the resources available to the farmer for the adoption of PCR, but may have positive effects due to the access the farmer has to outside information. According to Dimara & Skurass (1998), an increase in the off-farm annual work units decreases the probability of adopting flue-cured tobacco varieties in Greece, but this relationship is not statistically significant. Herath (2003) also noted that the role of off-farm activities on the decision to adopt is not very clear. Hence, it is difficult to predict the sign of this independent variable in the model. RUMINANT measures the number of ruminant livestock owned by the farmer. About 90% of households in the study area are involved in livestock keeping. The most important types of livestock are little-fat-tail sheep, donkeys, or mules. The number of ruminant animals being raised by a farmer influences his demand for forage for animal feed, such as alfalfa, and also increases the availability of manure in the rotation system. Shortage of fodder for ruminant animals is likely to increase the probability of adoption of the planted alfalfa. PCR can significantly

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

170

JOURNAL OF SUSTAINABLE AGRICULTURE

increase the quantity of forage yield of alfalfa (Jia et al., 2005), and farmers may have greater incentive to adopt PCR. It is hypothesized that increased numbers of ruminant animals will affect the adoption of PCR positively. DISTANCE is the distance of a farmer’s field from the dwelling and is measured by the total amount of minutes of walking. The closer a farmer’s fields are from his dwelling, the greater the possibilities of supervision and attention from the farmer’s families, and the lower the costs for transportation of plant material, crop residues, and compost (Nkamleu & Adesina, 2000). In this case, farmers are more likely to adopt PCR. It is hypothesized to influence decision-making negatively. FALLOW is a dummy variable, which takes a value 1 if a farmer has adopted fallow systems in the past, and 0 in the opposite case. In many parts of the Loess Plateau, farmers periodically fallow their land, allowing land to lie idle for one or more seasons primarily to improve soil fertility. As the population increases, however, cropping periods may increase and fallow periods decline. While continuous cropping can significantly decrease soil fertility, the trend can be reversed by crop rotation. Hence, fallow systems adopted in the past, which constitute a proxy variable for sustainable agricultural technology awareness, are hypothesized to have a positive effect on the adoption of PCR.

Institutional Policy Variables and Hypotheses CREDIT measures the accessibility of the households to fertilizers, seeds, and credit. This variable takes the value of 1 if the farmer obtained credit between 2004 and 2005, and 0 if otherwise. In these poor areas, most farmers cannot afford recommended inputs without credit. It may affect rotation adoption positively. EXTENSION is a dummy variable that measures the contact of farmers with research and development agencies or extension agencies. It takes a value of 1 if the farmer had contact with such agencies, and 0 if otherwise. Farmers who have frequent contact with extension workshops and experts have better access to information about problems, potential training services, and performance of PCR, to upgrade their knowledge on technology. Therefore, this variable is expected to be positive in this model. TRAINING is a dummy variable that represents training services obtained and takes the value of 1 if yes, and 0 if no. We would also expect this variable to exert a positive influence on the adoption of PCR in this case.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

171

GFG PROJECT is a dummy variable that indexes whether or not the farmer has participated in the Grain-for-Green project and takes the value of 1 if yes, and 0 if no. China’s Grain-for-Green (GFG) policy of converting steep cultivated land to grassland is one of the most important initiatives to develop western regions. Due to limited empirical information, the relative impact of this variable is difficult to assess. RHSIT PROJECT is a dummy variable that represents participation in the project of rainwater harvesting and supplementary irrigation technology (RHSIT), and takes the value of 1 if the farmer has participated in the project, and 0 if he has not. Rainwater harvesting uses an integrated technological system to supply supplemental irrigation to crops, and generates a significant increase in crop productivity (Li et al., 2004). Rainwater harvesting has been promoted by research and development agencies in the study area, since 1995, to tackle water shortages and rainfall fluctuations and to increase water availability to crops. Supplementary irrigation provided at critical stages of pasture or crop growth is an ideal choice, which can generate a significant increase in pasture or crop productivity (Li et al., 1999). This is also expected to exert a positive influence on the adoption of PCR. TENURE is a dummy variable that indexes land tenure rights. It takes the value of 1 if the farmer has permanent tenure rights and 0 if otherwise (e.g., if the crop field is a rented one). There is mixed evidence about the impact of land ownership on the adoption of agricultural technology. Accordingly, the impact of this variable is difficult to predict.

Ecological Variables and Hypotheses SLOPE is defined as slope of the field, which if flat takes a value of 1, and if very steep takes a value of 0. The slope of a field is an important indicator of erosion potential (Lapar et al., 1999), particularly in the Loess hilly area. PCR is proven to reduce run-offs and should be adopted on steeper plots, where erosion is seen to be more serious than on flat plots. PERCEPTION measures farmers’ perception of the soil erosion problem, and it takes the value of 1 if erosion is perceived, and 0 if otherwise. Perception of the soil erosion problem is considered vital in influencing conservation decisions. Previous research found a positive role of this variable on conservation decisions (e.g., Shiferaw and Holden, 1998; Sidibé, 2005). Ervin & Ervin (1982) and Norris & Batie (1987) also showed that farmers’ awareness of soil erosion problem is the first step in the process of adopting PCR. Thus, farmers who perceived the problem

172

JOURNAL OF SUSTAINABLE AGRICULTURE

more clearly are expected to invest more in conservation, suggesting a positive hypothesized effect of perception on farmers’ PCR decisions.

RESULTS AND DISCUSSION

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

Validity of the Model The results of the binary logistic regression model estimated by the maximum likelihood estimates of the coefficients using SPSS 13.0 software are given in Table 2. Table 2 shows the signs, magnitudes, and TABLE 2. Parameter estimates of the binary logistic regression model for factors influencing the adoption of PCR in the semiarid Loess Plateau of China Variable

B

Farmers socio-economic variables AGE*** 0.103 EDUCATION 0.058 FAMILYSIZE** 0.576 INCOME* −0.715 PARTICIPATORY** 2.107 ATTITUDE** 1.543 Farm structural variables FARMSIZE −0.190 OFFFARM 0.434 RUMINANT* 0.211 DISTANCE −0.023 FALLOW** 2.407 Institutional policy variables CREDIT* 1.421 EXTENSION*** 2.591 TRAINING*** 3.463 GFG PROJECT** 01.527 RHSIT PROJECT −0.866 TENURE 01.140 Ecological variables SLOPE* −1.161 PERCEPTION*** 5.418 Constant*** −17.683

S.E.

WALD

Sig.

Exp(B)

0.037 0.126 0.290 0.371 0.920 0.759

7.820 0.212 3.950 3.718 5.243 4.135

0.005 0.645 0.047 0.054 0.022 0.042

1.109 1.060 1.778 0.489 8.226 4.678

0.488 0.686 0.110 0.068 1.107

0.152 0.399 3.666 0.110 4.728

0.697 0.527 0.056 0.740 0.030

0.827 1.543 1.235 0.978 11.106

0.730 0.710 0.895 0.749 0.811 1.018

3.795 13.334 14.984 4.159 1.140 1.254

0.051 0.000 0.000 0.041 0.286 0.263

4.142 13.345 31.927 4.604 0.421 3.128

0.657 1.990 3.945

3.125 7.413 20.090

0.077 0.006 0.000

0.313 225.339 0.000

***Significant at 1%; **Significant at 5%; *Significant at 10%. Chi-square using the Hosmer and Lemeshow test, 5.162; df = 8; Significance, 0.740; −2Log likelihood, 91.33; Cox & Snell R2, 0.591; Nagelkerke R2, 0.819; Overall Percentage of right predictions, 92.4%.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

173

statistical significance of the estimated parameters. The model is highly persuasive, with an overall predictive accuracy of 92.4%. The p value of 0.74 uses Hosmer and Lemeshow goodness-of-fit test, which is computed from the chi-square distribution with 8 df. As planned, we fail to reject the null hypothesis that there is no difference, implying that the model’s estimates very well fit the data at an acceptable level. The column, exp (B), in Table 2, gives the exponential of expected value of B raised to the value of the logistic regression coefficient. These values represent the odds of an event occurring when the ith independent variable increases by one unit (Sheikh et al., 2003). Of the 19 variables hypothesized that explain farmers’ decision to adopt PCR, 4 were found to be significant at p < 0.01. These are the following variables: AGE, EXTENSION, TRAINING, and PERCEPTION. Table 2 also shows that 5 variables were significant at p < 0.05. These variables include FAMILYSIZE, GFG PROJECT (participation in Grainfor-Green project), ATTITUDE, FALLOW, and PARTICIPATORY. Variables such as SLOPE, CREDIT, INCOME, and RUMINANT were found to be significant at p < 0.1.

Farmers Socio-Economic Variables The analyses show that at 1% level, farmers’ age (AGE) has a positive impact on the adoption of PCR, suggesting that the probability adopting the technology is higher among older farmers than younger ones. Likewise, education level of farmers (EDUCATION) in reality does not have the expected positive impact on the adoption of the technology. It is insignificant. FAMILYSIZE is significant, probably due to the fact that larger family size would include more children, who are unproductive members and PCR is essentially a laborsaving technology. This result corroborates an earlier finding by Enyong and coworkers (1999) that farmers’ rationale behind their decisions to adopt technologies for the enhancement of soil fertility is influenced by food security concerns. Higher consumption pressure faced by their family (especially if large) may influence farmers’ decisions to adopt PCR. Larger family size is generally associated with a greater availability of labor. Larger human capital may positively influence the decision towards adopting PCR, encouraging participation in rotation works. INCOME exhibits a negative impact on the outcome. The result indicates that high-income farmers whose primary source of income is not

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

174

JOURNAL OF SUSTAINABLE AGRICULTURE

agriculture are less concerned about land conservation compared with low-income farmers, whose livelihood is derived mainly from agriculture. PARTICIPATORY has a very significant correlation with adoption. The results imply that “participatory approaches to PCR technology development” have a substantial influence on decisions to adopt the technology. Participatory methods hold the greatest potential for integrating farmers into the process of designing PCR systems. Recently, participatory training and development is seen as a new concept, with farmers’ own assessments through participatory evaluation exercises. Farmers are now seen as partners in research activities, further encouraging wider adoption PCR. In fact, Whyte (1981) reported that the top-down approach is faulty in its perception of farmers as passive recipients of technology. Participatory research methods have been advocated as a means of improving relevance and adoption of new technologies (Kuntashula & Mafongoya, 2005). In Southeast Asia, Cramb (2005) has also shown that a participatory approach can increase the likelihood of improved management practices for acid upland soils. In the central Sal forests in Bangladesh, results were observed by Salam and colleagues (2005), indicating that participation is positively and significantly correlated with: (i) participants’ satisfaction; (ii) participants’ confidence; (iii) the provision of training; and (iv) the financial contribution of participants. Participatory assessment of technologies prior to on-farm experimentation is now widely recognized as an important factor in determining whether farmers adopt technologies (Dorward et al., 2003). Farmers’ attitude toward PCR (ATTITUDE) had a significant impact on the probability of their adopting the technology (p < 0.01). Exp(B) shows that the likelihood that a farmer who has a positive impression adopts RHSIT is 4.678 times that of a farmer who has a negative impression. The results imply that the responsiveness to PCR depends heavily on the strength of individuals’ attitude toward and impressions of, the technology. In Gansu Province, PCR has been proposed to farmers for many years as an inexpensive way to improve soil fertility and production. Thus, in this model, a positive subjective assessment of PCR has a positive effect on its adoption. Similar findings in some empirical studies show that farmers with a generally positive attitude are keen on implementing agricultural technologies (e.g., Shiferaw and Holden, 1998; Herath and Takeya, 2003; Somda et al., 2002). The implication is that the benefits of PCR over continuous cropping must be clearly perceived by farmers in the context of their own socio-economic conditions, which in turn provides them an incentive to adopt PCR.

He, Cao, and Li

175

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

Farm Structural Variables RUMINANT is consistent with our hypothesis. In the Loess hilly area, where grazing and pasturelands are slowly disappearing because of soil degradation, on-farm production of improved alfalfa is considered a lowcost route to improved animal performance. The more ruminant animals the farmer has, the more likely he or she is to adopt such dual-purpose forages that could be used as livestock feed in addition to being effective barriers to soil erosion. Shortage of fodder for livestock is likely to increase the probability of adopting planted forages. The planting of forage species on contour bunds thus responds to both demands for fodder and soil conservation functions in farming (Lapar and Ehui, 2004). Simultaneously, farmers in the study area have been trained in applying alternative fertilizer options, such as compost and green manure. Ruminant animals provide manure for crop rotation. Hence, the more ruminant animals the farmer has, the more likely he or she is to adopt PCR. As explained by Burgers and coworkers (2005), in Southeast Asia, integrating livestock with crop-fallow rotation incorporates both the soil fertility and the economic return aspects. As expected, FALLOW variable is positively and significantly related to the dependent variable. This result implies that those farmers who know the available options for tackling soil erosion will be more likely to adopt PCR. The results of FARMSIZE, OFF-FARM, and DISTANCE do not influence decisions significantly in the study area at 10% probability level.

Institutional Policy Variables As hypothesized, the impact of institutional factors such as CREDIT, EXTENSION, and TRAINING have significant effects. This finding is consistent with our expectations. Similar results were observed by Ervin and Ervin (1982), Norris and Batie (1987), Adesina and associates (2000), as well as Tenge and colleagues (2004). EXTENSION and TRAINING are both positively and significantly correlated (p < 0.01). Exp(B) shows that holding all other explanatory variables constant, for every one-unit increase in EXTENSION, TRAINING score, we expect 13.345 and 31.927 fold increases in the log-odds of adoption (the probability of adoption), respectively. The same positive correlation is found with regard to CREDIT. It is significant at p < 0.05 level and positively correlated to adoption, suggesting that the higher the availability of credit services, the greater the likelihood of adopting PCR.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

176

JOURNAL OF SUSTAINABLE AGRICULTURE

The new findings of this study is that the odds of a farm household involved in the Grain-for-Green (GFG) program adopts RHSIT is 4.604 times the odds of a farm household which was not involved in the program. It is possible that participation in the GFG project triggers the adoption of PCR, such as being innovative or aware of soil quality issues. During the interviews, some government officials and farmers mentioned that the GFG not only improved livelihoods (farmers involved in the program receive cash and food compensation, which generally exceed net production), but also improved the agro-ecological status and ensured sustainable development. PCR not only aimed at minimizing soil erosion, but also at achieving other household objectives, such as the improvement of soil fertility, increase in yield, and increase in fodder for animals. This is consistent with the goal of China’s GFG policy as well. As Gong and colleagues (2005) reported, “as for the ‘GFG policy’ in the Loess hilly area, grass planting is an optimal choice for improving soil conditions.” Findings of such research will be helpful for devising appropriate GFG policies to promote PCR. Moreover, RHSIT PROJECT was not statistically significant, despite showing a negative impact on the outcome. The insignificance may be due to the fact that the source of water for irrigation in the area is very important. Although RHSIT, as an innovation in agriculture, has been considered both economically and environmentally beneficial (e.g., increasing agricultural productivity, improving soil fertility, and reducing soil erosion) source of water for irrigation is not used in the production of pasture. Deng et al. (2004) indicated that if water is used in the production of high-value cash crops like watermelon, greenhouse vegetables, and fruit trees, it will provide an even greater economic benefit to the farmer. Leib and coworkers (2002) also indicated that producers were willing to pay for irrigation to ensure good quality of high-value cash crops. The results of TENURE imply that land security had a positive but nonsignificant correlation with the adoption of PCR. Asfawa and coworkers (2004) also indicated similar findings on chemical fertilizer adoption in Ethiopia.

Ecological Variables As hypothesized, SLOPE is correlated to the dependent variable. This may be explained by the positive relationship between the slope and severity of soil erosion. The higher the slope, the greater the severity of soil erosion. The impact of soil erosion would be more visible to the farmers and would force them to take remedial actions such as crop rotation

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

177

system. PCR improves soil conditions by including deep-rooted plants and plants with a fibrous root system to improve the stability of soil aggregates (Lazányi, 2000). Ervin and Ervin (1982), Norris and Battie (1987), Pender and Kerr (1998), Lapar and colleagues (1999), and Asrat and coworkers (2004) also reported similar results. As expected, farmers’ perception (PERCEPTION) is positively and significantly correlated (p < 0.01). Exp(B) is shown to have the maximum impact with the highest estimated level of elasticity, indicating that the likelihood that a farmer who perceived a soil erosion problem would adopt PCR is 225 times greater than that for a farmer who is unaware of soil erosion. Accelerated soil erosion is one of the major constraints to agricultural production in the area of study. Farmers who have already perceived problems of soil erosion are more likely to adopt PCR. Results were supported by findings in studies on the adoption of soil conservation measures (see, for example, Shiferaw and Holden, 1998; Tenge et al., 2004; Asrat et al., 2004; and Sidibé, 2005).

CONCLUSIONS AND IMPLICATIONS PCR is considered a cornerstone of integrated farming, particularly in view of its ability to maintain soil fertility and to increase productivity (Roth et al., 2002; Jagtap & Adeleye, 1999). PCR, based on the inclusion of polyannual legumes such as alfalfa, is to be considered among the most powerful management practices that are also economically and environmentally beneficial in the semiarid area. To plan effective interventions to facilitate the adoption of PCR, it is important to understand the factors that influence the decision of small-hold farmers. The analyses in this study arrive at several conclusions that provide an insight into ways of increasing the adoption of PCR. First, variables representing farmers’ age, family size, number of ruminants owned, perception of erosion problems, and attitudes toward PCR all have significant positive effects on their choice to adopt PCR. However, the level of family income exhibits a negative effect on the outcome. In fact, many of the earlier efforts to target crop rotation are based on purely agronomic characteristics. The conclusion is that the absence of consideration of socio-economic, farm structural, institutional, and ecological factors can lead to inappropriate targeting of farmers. The benefits of PCR must be clearly perceived by farmers to help them develop a positive assessment. Second, the results of the study also reveal that adoption was higher for farmers having contacts with extension agencies, access to

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

178

JOURNAL OF SUSTAINABLE AGRICULTURE

credit and training, participatory evaluation exercises, and participation in the GFG project. Therefore, there is a need to design and develop alternative policy instruments, institutions, and projects for extension, assistance, training, credit services, and GFG and RHSIT projects that facilitate the adoption of farmer-participatory practices to better fit the needs of farmers. Third, the slope of the field and the fallow practice variables also play a critical role in adoption, suggesting the importance of targeting the promotion of PCR among farmers in areas with problems of soil erosion. Information obtained from this research can be used by policy-makers, researchers, and extension staff to enhance the adoption of crop rotation in the semiarid Loess Plateau as well as other areas with similar conditions.

REFERENCES Adesina, A.A., D. Mbila, G.B. Nkamleu, and D. Endamana. 2000. Econometric analysis of the determinants of adoption of alley farming by farmers in the forest zone of southwest Cameroon. Agriculture, Ecosystem and Environment 80:255–265. Asrat, P., K. Belay, and D. Hamito. 2004. Determinants of farmers’ willingness to pay for soil conservation practices in the southeastern highlands of Ethiopia. Land degradation and Development 15:423–438. Burgers, P., Q.M. Ketterings, and D.P. Garrity. 2005. Fallow management strategies and issues in Southeast Asia. Agriculture, Ecosystem and Environment 110:1–13. Caporali, F., and A. Onnis. 1992. Validity of rotation as an effective agroecological principle for a sustainable agriculture. Agriculture, Ecosystem and Environment 41:101–113. Chianu, J.N., and H. Tsujii. 2004. Determinants of farmers’ decision to adopt or not adopt inorganic fertilizer in the savannas of northern Nigeria. Nutrition Cycle in Agroecosystems 70: 293–301. Comer, S., E. Ekanem, S. Muhammad, S. P. Singh, and F. Tegegne. 1999. Sustainable and Conventional Farmers: A Comparison of Socio-Economic Characteristics, Attitude, and Beliefs. Journal of Sustainable Agriculture 15 (1): 29–45. Cramb, R.A. 2005. Farmers’ strategies for managing acid upland soils in Southeast Asia: an evolutionary perspective Agriculture. Agriculture, Ecosystem and Environment 106:69–87. Crews, T.E., and M.B. Peoples. 2004. Legume versus fertilizer sources of nitrogen: ecological tradeoffs and human needs Peoples. Agriculture, Ecosystem and Environment 102:279–297. Dimara, E., and D. Skurass. 1998. Adoption of new tobacco varieties in Greece: impact of empirical findings on policy design. Agricultural Economics 19:297–307. Dorward, P., M. Galpina., D. Shepherd. 2003. Participatory farm management methods for assessing the suitability of potential innovations. A case study on green manuring options for tomato producers in Ghana. Agricultural Systems 75:97–117.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

He, Cao, and Li

179

Enyong, L.A., S.K. Debrah, and A. Bationo. 1999. Farmers’ perceptions and attitudes towards introduced soil-fertility enhancing technologies in western Africa. Nutrition Cycle in Agroecosystems 53: 177–187. Ervin, C.A., and E.D. Ervin. 1982. Factors affecting the use of soil conservation practices: hypothesis, evidence, and policy implications. Land Economics 58 (3) :277–292. Feder, G., R.E. Just, and D. Zilberman. 1985. Adoption of agricultural innovations in developing countries: A survey. Economic Development and Cultural Change 33: 255–297. Gillespie, G.W.J., T.A. Lyson, and A. Power. 1996. Crop rotation patterns among New York potato growers: insights from conventional and sustainable agricultural theory. Journal of Sustainable Agriculture 7 (1): 5–18. Gong, J., L. Chen, B. Fu, Y. Huang, Z. Huang, and H. Peng. 2005. Effect of land use on soil nutrients in the loess hilly area of the Loess Plateau, China. Land Degradation and Development 17 (5): 453–465. He, X.F. and F.M. Li. 2006. Econometric analysis of the determinants of adoption of rainwater harvesting and supplementary irrigation technology (RHSIT) in the semiarid Loess Plateau of China. Agricultural Water Management 89 (3): 243–250. Jagtap, S.S., and O. Adeleye. 1999. Nutrient recovery and productivity of a soybeanmaize rotation in thederived savanna ecology of West Africa. Journal of Sustainable Agriculture 15 (1): 75–85. Jeffrey, S., J.D.L. Young, and W.F. Schillinger. 2002. Economics of conservation tillage in a wheat–fallow rotation. Agronomy Journal 94:527–531. Jia, Y., F.M. Li, and X.L. Wang. 2005. Soil quality responses to alfalfa watered with a field micro-catchment technique in the Loess Plateau of China. Field Crops Research 95(1) : 64–74. Jiang, D.S. 1997. Soil erosion and control models in the Loess Plateau. China Hydroelectricity Press: Beijing (in Chinese). Kebede, Y., K. Gunjal, and G. Coffin. 1990. Adoption of new technologies in Ethiopian agriculture: The case of Tegulet-Bulga District, Shoa Province. Agricultural Economics 4:27–43. Kuntashula, E., and P.L. Mafongoya. 2005. Farmer participatory evaluation of agroforestry trees in eastern Zambia. Agricultural Systems 84:39–53. Lapar, M.L.A., and S.K. Ehui. 2004. Factors affecting adoption of dual-purpose forages in the Philippine uplands. Agricultural Systems 81: 95–114. Lapar, M.L.A., and S. Pandey. 1999. Adoption of soil conservation: the case of Phillippine uplands. Agricultural Economics 21:241–256. Lazányi, J. 2000. Sustainable Rye production in the Westsik crop rotation experiment. Acta Agronomica Hungarica. 48 (3): 271–277. Li, F.M., J. Wang, and S.L. Zhao. 1999. The rainwater harvesting technology approach for dryland agriculture in semi-arid Loess plateau of China. Acta Ecologia Sinica 19 (2): 259–264. Li, F.M., P. Wang, J. Wang, and J. Zh., Xu. 2004. Effects of irrigation before sowing and plastic film mulching on yield and water uptake of spring wheat in semiarid Loess Plateau of China. Agricultural Water Management 67:77–88. Li, F.R., C.Y. Gao, H.L. Zhao, and X.Y. Li. 2002. Soil conservation effectiveness and energy efficiency of alternative rotations and continuous wheat cropping in the Loess Plateau of northwest China. Agriculture, Ecosystem and Environment 91:101–111.

Downloaded By: [Canadian Research Knowledge Network] At: 14:46 26 June 2009

180

JOURNAL OF SUSTAINABLE AGRICULTURE

Lockie, S., A. Mead, F. Vaneley, and B. Butler. 1995. Factors encouraging the adoption of more sustainable crop rotations in South-East Australia: profit, sustainability, risk and stability. Journal of Sustainable Agriculture 6 (1): 61–79. Long, H.L., G.K. Heilig, J. Wang, X.B. Li, M. Luo, X.Q. Wu, and M., Zhang. 2006. Land use and soil erosion in the upper reaches of the yangtze river: some socio-economic considerations on China’s Grain-For-Green Programme. Land Degradation and Development 17 (6): 589–603. Lu, C.H., M.K. van Ittersum, and R. Rabbinge. 2004. A scenario exploration of strategic land use options for the Loess Plateau in northern China. Agricultural Systems 79:145–170. Lu, Y.C., and B. Watkins, Teasdale, J. 1999. Economic analysis of sustainable agriculture cropping systems for Mid-Atlantic States. Journal of Sustainable Agriculture 15 (2–3):77–93. Nkamleu, G.B., and A.A. Adesina. 2000. Determinants of chemical input use inperiurban lowland systems: bivariate probit analysis in Cameroon. Agricultural Systems 63: 111–121. Norris, E.P., and S.S. Batie. 1987.Virginia farmers’ soil conservation decisions: an application of Tobit analysis. Southern Journal of Agricultural Economics 19 (1):89–97. Pender, J.P., and J.M. Kerr. 1998. Determinants of farmers’ indigenous soil and water conservation investments in semi-arid India. Agricultural Economics 19:113–125. Place, F., and P. Dewees. 1999. Policies and incentives for the adoption of improved fallows. Agroforestry Systems 47: 323–343. Roth, G.W., J.K. Harper, A.A. Hower, and R.A. Kyper. 2002. Potential of shorter corn/ alfalfa rotations for dairy farms. Journal of Sustainable Agriculture 20 (2): 41–52. Sheikh, A.D., T. Rehman, and C.M. Yates. 2003. Logit models for identifying the factors that influence the uptake of new ‘no-tillage’ technologies by farmers in the rice—wheat and the cotton—wheat farming systems of Pakistan’s Punjab. Agricultural Systems 75:79–95. Shiferaw, B., and S.T. Holden. 1998. Resource degradation and adoption of land conservation technologies in the Ethiopian Highlands: a case study in Andit Tid, North Shewa. Agricultural Economics 18: 233–247. Skaggs, R.K. 2001. Predicting drip irrigation use and adoption in a desert region. Agricultural Water Management 51: 125–142. Stinner, D.H., I. Glick, and B.R. Stinner. 1992. Forage legumes and cultural sustainability: lessons from history. Agriculture, Ecosystem and Environment 40:233–248. Tang, H.J., and E.V. Ranst. 2005. Is highly intensive agriculture environmentally sustainable. A case study from Fugou County, China. Journal of Sustainable Agriculture 3: 97–102. Tenge, A.J., J. De. Graaff, J.P. Hella. 2004. Social and economic factors affecting the adoption of soil and water conservation in West Usambara Highlands, Tanzania. Land Degradation and Development 15: 99–114. Wang, J., F.M. Li, and Y. Jia. 2006. Responses of soil water, nitrogen and organic matter to the alfalfa crop rotation in semiarid Loess Area of China. Journal of Sustainable Agriculture 28 (1): 117–130. Whyte, W.F. 1981. Participatory approaches to agricultural research and development: a state-of-the-art paper. Cornell University, Ithaca, New York. Yang, X.M., and B.D. Kay. 2001. Rotation and tillage effects on soil organic carbon sequestration in a typic Hapludalf in Southern Ontario. Soil and Tillage Research 59:107–114.