An analysis of the joint adoption of water conservation and soil conservation in Central Chile

Land Use Policy 32 (2013) 292–301

Contents lists available at SciVerse ScienceDirect

Land Use Policy journal homepage: www.elsevier.com/locate/landusepol

An analysis of the joint adoption of water conservation and soil conservation in Central Chile Roberto Jara-Rojas a,∗ , Boris E. Bravo-Ureta a,b , Alejandra Engler a , José Díaz a a b

Department of Agricultural Economics, Universidad de Talca, Chile Department of Agricultural and Resource Economics, University of Connecticut, Storrs, 06269-1182 CT, USA

a r t i c l e

i n f o

Article history: Received 29 August 2011 Received in revised form 29 October 2012 Accepted 4 November 2012 Keywords: Water conservation Soil conservation Bivariate Probit model Technology adoption Smallholder agriculture Chile

a b s t r a c t Studies reveal that 80% of the world’s agricultural land is showing signs of moderate levels of soil erosion. On the other hand, it is a fact that water is becoming a more scarce resource jeopardizing food security. Thus, conserving both water and soil are two of the most pressing issues in international agriculture and food production. This article examines the impact of natural, social, human, and financial capital variables on the adoption of water conservation and soil conservation (WC&SC) as a joint decision, using a bivariate model. Socioeconomic and production information was collected by surveying a random sample of 319 small-scale irrigated farms in central Chile in 2005. The results suggest that the adoption of WC&SC is a joint and complementary decision. The results also indicate that farm size, production system, access to credit, and government incentives are important variables associated with the adoption of conservation measures. From a policy stand point, the institutions in charge of providing incentives and administering instruments intended to promote conservation should take into account the complementarity of the adoption decisions. Program designs should incorporate incentives that jointly promote the adoption of WC&SC in order to enhance effectiveness. © 2012 Elsevier Ltd. All rights reserved.

Introduction Land degradation consists of the deterioration of soil quality and thus of its productive capability. Erosion is a key culprit in land degradation and can play a major role, since it is often of significant magnitude, irreversible and, in extreme cases, creates the total loss of soil (Hugo, 2008). Rapid erosion in Chile is associated primarily with farming practices that degrade the soil such as compaction, and the loss of organic matter and soil structure. Erosion is a concern given the hilly landscape of the country particularly in areas with high annual precipitation where much of the rain falls in short periods of time (Ellies, 2000). According to the international literature, 70–80% of the land dedicated to agricultural activities worldwide exhibits moderate to severe erosion (Blaike and Brookfield, 1987; Pimentel et al., 1995). Soil erosion is a challenging issue not only because it leads to productivity losses, but also because it is strongly linked to desertification and rural poverty (Barbier and Bishop, 1995; Ruben et al., 2004). The causes of land degradation are varied and complex and can be grouped in three major categories: (1) climatic (e.g., rain fall, drought); (2) bio-geophysical (e.g., slope, soil type); and (3)

∗ Corresponding author. Tel.: +56 71 200222; fax: +56 71 200212. E-mail addresses: [email protected] (R. Jara-Rojas), [email protected] (B.E. Bravo-Ureta), [email protected] (A. Engler), [email protected] (J. Díaz). 0264-8377/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.landusepol.2012.11.001

managerial (e.g., farmer education, experience, access to extension services). These three groups of variables are critical in determining the likelihood and speed of soil erosion (Muchena et al., 2005). Many traditional agricultural practices contribute to soil degradation (Solís et al., 2009) while technologies designed to improve or conserve soil are not always adopted, even when their usefulness has been demonstrated (Amsalu and de Graaff, 2007). Furthermore, reduced productivity is usually mitigated in the short run with intensive use of inputs, which leads to additional degradation problems in the longer run (Solís and Bravo-Ureta, 2005). The severity of current degradation has inspired significant efforts to develop and promote the adoption of conservation strategies. However, the results are not always positive, and soil degradation continues to be a major problem worldwide (Pimentel et al., 1995). In addition, irrigation water makes an important contribution to agricultural productivity and food security, but it is becoming an increasingly scarce resource (FAO, 1992; Bruinsma, 2009). This growing scarcity becomes even more significant in a context of climate change where global precipitation patterns are already altered (Reilly et al., 2007). This issue is critical considering that the productivity of 82% of the world’s arable land depends on precipitation (Schultz et al., 2005). In areas where precipitation might decline, agriculture would face growing competition for water from higher-valued uses such as domestic, industrial and hydropower—all of which are rising. Thus, the agricultural sector will need to produce more food with less water.

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

Dealing with expected water shortages represents an important technological and policy challenge. The careful management and increased efficiency of irrigation water is a relevant component of any plan of action that helps to understand the benefits of irrigation and conservation. Such benefits include higher land productivity and increased yields, a lower risk of crop failure, and increased yearround farm and nonfarm employment (Hussain and Hanjra, 2004). Despite the potential complementarity of water conservation and soil conservation (WC&SC), very few studies have modeled the determinants of farm-level decisions to conserve water and soil simultaneously, and no such study has been carried out in Chile. Most of the literature treats the adoption WC&SC as separate decisions (Staal et al., 2002; Kim et al., 2005; Anley et al., 2007; Calatrava-Leyva et al., 2007; Kabubo-Mariara, 2007). Another approach, which depicts these decisions as interdependent, uses multivariate Probit regression models to estimate the adoption of different conservation strategies (Marenya and Barrett, 2007). The U.S. Environmental Protection and Agency (1997) defines WC&SC as managerial, vegetative and structural practices to reduce the loss of soil and water, respectively. Several WC&SC strategies are described by Lee (2005) and analyzed by Knowler and Bradshaw (2007). Natural resource conservation has been found to have a positive impact on the productivity of annual crops (Gupta and Seth, 2007) and even to increase farm income (Bravo-Ureta et al., 2006). Limited evidence from Chile indicates that the implementation of crop rotations, minimum tillage and the use of legumes have improved soil quality and increased crop yields over time (Vidal et al., 2002). Considering that conservation has beneficial effects in the long run while possibly increasing profitability in the short run, gives rise to the following questions: what are the factors promoting or limiting the adoption of such strategies; and is the WC&SC a joint or an independent decision? In economic terms, farmers adopt technologies and conservation strategies that they perceive to be profitable (Ellis, 1993; de Graaff et al., 2008); however, socioeconomic, cultural, and natural resource factors affect the rate at which farmers adopt conservation strategies (Lapar and Pandey, 1999; Soule et al., 2000). Moreover, the process of adopting interrelated conservation strategies is more complex than the decision to adopt a single technology, such as use of fertilizers or an improved seed variety. The single decision is usually based on short-term profitability considerations, while interrelated adoption implies a more substantial and longer-lasting change in farming conservation (Caswell et al., 2001; Boyd et al., 2000). The objective of this paper is to contribute to the literature on resource conservation by analysing the factors that influence the simultaneous adoption of WC&SC using a bivariate modeling approach. The results could help to improve the understanding of farmers’ behavior regarding conservation, and therefore help the development of incentives and/or instruments focusing on soil and water programs. More generally, the analysis will help to fill the gap that exits in the literature concerning the drivers of adoption of conservation in the country. Despite the prominence of agricultural production in the country’s economy, Wandel and Smithers (2000) point out that the relevant factors associated with adoption have a high degree of locational and technological specificity. Therefore, for Chilean policy makers and research institutions it is important to have a localized understanding of the factors that influence the adoption of WC&SC. The rest of the paper is organized as follows: “Methodological framework” section describes the methodological framework; “Materials and methods” section discusses the study area, the data and the empirical model. “Results and discussion” section includes the results. “Policy implications” section contains policy implications and the final section is dedicated to concluding remarks.

293

Methodological framework Utility maximization theory has often motivated the methodological framework used to study technology adoption (e.g., Rahm and Huffmann, 1984; Adesina and Zinnah, 1993). According to this theory, a new technology will be adopted if the expected utility (U) derived from its use is higher than that of the current technology. Since U is not directly observable, the model can be expressed as a binary choice in terms of observable adoption or non-adoption, which implies the use of a Probit or Logit model (e.g., Lapar and Pandey, 1999; Feder et al., 1985; Foltz, 2003). The underlying U function of the ith farmer, expressed in terms of farm and farmer-specific attributes, X, and a disturbance term having a mean of 0, can be written as: Uij = ˇi (Xi ) + εij

j = 1, 0 i = 1, 2, . . . , n

(1)

where Xi is a vector of explanatory variables and ˇi is a vector of parameters. Adoption of the new technology (j = 1) by the ith farmer occurs when Ui1 > Ui0 . According to Greene (2008), the bivariate Probit model, which considers two dichotomous decisions simultaneously, is a natural extension of the individual Probit model that features only one decision and a single equation. In our case, the decisions are the adoption/non-adoption of soil conservation, denoted by s, and of water conservation, denoted by w. In the model specification s, w = 1 for adoption and s, w = 0 for non-adoption, and the disturbance terms of the two equations can be correlated. The specification of the bivariate Probit model, following Greene (2008), is given by: s = ˇs x + εs w = ˇw x + εw

(2)

E[εs , εw ]∼BVN[0, 0, 1, 1, ] where E(εs ) = E(εw ) = 0, Var(εs ) = V (εw ) = 1, Cov(εs , εw ) =  and the distribution is bivariate normal. Thus, the bivariate Probit model is a Seemingly Unrelated Regression (SUR) specification, where it is assumed that all regressors are exogenous, and estimation is done using maximum likelihood (Greene, 2007). This approach has been used in explaining chemical input use, such as the joint adoption of fertilizer and pesticides (Nkamleu and Adesina, 2000), the adoption and dis-adoption of technologies as a sequential and conditional process (Amsalu and de Graaff, 2007; Neill and Lee, 2001), and the adoption of improved groundnut varieties along with chemical fertilizer (Thuo et al., 2013). In the first two studies,  was found to be significant, which indicates that the error terms across equations are correlated. In the present study, we argue that the adoption of soil conservation is likely to condition positively the decision to adopt water conservation, and vice versa; thus, treating these decisions separately would generate biased parameter estimates and thus we hypothesize that  is significant and positive. Once the bivariate model is estimated, the next step is to consider the Marginal Effects (MEs) of the covariates. According to Greene (1996), the marginal means reported for the model assume that soil and water adoption are equal to 1, as follows: E[s|w = 1, xs , xw ] =

Prob[s = 1|w = 1, xs , xw , ] Prob[w = 1|xs ]

(3)

Then, the conditional mean for these two models would be identical: E[s |w = 1] =

˚w [˛s , ˛w , ] ˚(˛w )

(4)

294

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

where Фw is the bivariate normal cumulative distribution function (CDF), Ф is the univariate normal CDF, and ˛s and ˛w are the estimates of the MEs. This model allows multiplicative heteroscedasticity in either or both equations (Greene, 2007); therefore, ˛s is calculated (similarly for ˛w ) as: ˛s =

ˇs xs , exp(s hs )

(5)

where exp(s hs ) is the variance-covariance matrix. In the homoscedastic model,  s and/or  w is a zero vector.

Materials and methods Sampling and data description The data used in this study were generated from a farm-level survey of small-scale farmers. The sample was drawn by first randomly selecting 32 of the 169 existing water communities of the project area, and then randomly selecting 9–11 farmers per water community. The study area was subdivided into 32 groups, defined by the boundaries of the water communities registered in the Chilean General Water Administration Office. A total of 319 farmers were interviewed, and this sample size is statistically representative of the population under study at the 5% level of confidence with a 5.5% sampling error.1 A few surveys with missing or incomplete data were excluded from the analysis, yielding a final data set of 307 observations. All surveys were geo-referenced using GPS instruments (Staal et al., 2002). Soil information, such as soil depth, slope and pH (see Table 1), for the farms in the sample was obtained from integrated Geographic Information Systems (GIS) analysis. On average, 81.5% of the farms have a soil with moderate to light depth and 15.4% of them have a poor quality soil in terms of depth. Almost 100% of the farms have flat land and very few stones. According to CIRENCORFO (1997), 69% of the farms in the sample have soils with moderate restrictions for crops (Class III) while 26% of the farms have more severe restrictions regarding potential soil use (Class IV, VII, VIII, and/or IX). The overall agro-climatic condition in the area is Template Mediterranean with an average annual precipitation varying between 600 and 750 mm. The water deficit between December and March, which corresponds to the summer season, reaches ˜ 440 mm and the average annual temperature is 13.8 ◦ C (Santibánez, 2012). According to Pérez and González (2001), soil and water erosion, and insufficient amounts of land to allow farmers to pursue recommended follow and crop rotations are key environmental challenges in the area under study. Although soil losses are common throughout Chile (Francke, 1997; Bonilla et al., 2010), a high percentage of smallholders in the study area are considered to be particularly vulnerable to increased degradation given that most of these farmers live in less-favored areas (Ruben et al., 2004; Kessler, 2007). Thus, the environment facing these farmers regarding their soil and water resources provides an interesting context to study the adoption of conservation strategies.

1 The sample size is calculated using the following equation (Nyariki, 2009): n = 2 (p(1 − p)/d2 ), where Z is the statistical certainty usually set at the 95% confidence Z˛/2

level (Z = 1.96), p is the estimated level/coverage to be investigated, usually a p = 0.5 is chosen, and d is the precision, maximum permissible error in terms of proportion (0.055). These parameters yielded a sample size of 320. The result was corrected using the equation n = n /(1 + n /U) where U is the finite Universe of the study area (8772 small-farms). Thus, the final sample size is 309, 319 farmers were interviewed and the sample used includes 307.

Table 1 Agroecological conditions of the farms in the study area. Category

Description

Frequency

Percentage

Soil depth Depth Moderate depth Lightly depth Thin soil Extremely thin

>100 cm 75.1–100 cm 50.1–75 cm 25.1–50 cm ≤25 cm

10 124 136 27 22

3.1 38.9 42.6 8.5 6.9

≤1% 1.1–3% 3.1–5% >5%

224 85 2 8

70.2 26.6 0.7 2.5

≤5% 5.1–15% 15.1–35% >35%

262 40 12 5

82.1 12.5 3.8 1.6

pH Moderate acidity Light acidity Neutral Lightly alkaline

5.6–6.0 6.1–6.5 6.6–7.3 7.4–7.8

160 46 97 16

50.2 14.4 30.4 5.0

Land use Class I Class II Class III Class IV Class VII Class VIII Class IX

Without restrictions Low restrictions Moderate restrictions Severe restrictions Not suitable for crops Apt for forestry Wild life

1 15 220 52 17 1 13

0.3 4.7 69.0 16.3 5.3 0.3 4.1

Soil slope Flat Hardly flat Lightly flat Hill soils % stone in soil No stones Light stones Moderate stones Many stone

Study area The study site is in an irrigated valley fed by five rivers—the Melado, Putagán, Ancoa, Achibueno, and the Longaví—in the SouthMaule watershed of the Linares Province in Central Chile (see Fig. 1). The water in each case is managed by a board of directors, and each river supplies several water communities.2 The area encompasses a total of 9590 farms, which as indicated above have mostly flat land, while some have moderate slopes (CIREN-CORFO, 1997). Farms in Chile fall into three main categories: small-scale farms (further subdivided into subsistence farms and small farm enterprises); medium-sized farms; and large farms. This classification takes into account agroecological conditions and various other factors such as access to capital and technology, market orientation, cultivated land area, and agricultural potential. Some of the characteristics of small-scale farming systems in Chile are summarized by the OECD (2008). In the study area, average farm size is 42 ha but 91% of the farms (8772 farms) have less than 50 physical hectares; thus, the majority are smallholdings (INE, 2007). These 8772 farms constitute the total population for the present study. Since the area under study is irrigated, farmers can pursue a variety of farming systems. However, small-holders in the area exhibit three primary systems: (a) a mono culture based on raspberry production or a single annual crop; (b) a diversified farming system based on annual crops where the most common are wheat, beans,

2 The Water User Associations (WUAs) in Chile are charged with the distribution of water, the enforcement of its appropriate use, fee collection, and the administration and maintenance of the irrigation infrastructure. There are three different types of associations considered in the Chilean Water Code: (a) juntas de vigilancia, which are supervisory committees in charge of monitoring the use of natural sources of water or rivers; (b) asociaciones de canalistas, which are associations of channel water users in charge of administering primary infrastructure such as main artificial irrigation channels and dams; and (c) comunidades de agua, which are water communities responsible for secondary infrastructure or distribution channels (Ríos Brehm and Quiroz, 1995).

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

295

Fig. 1. Map of the area under study.

maize and vegetables; or (c) raspberries along with annual crops. Table 2 shows a complete description of the farming systems found in the sample. A typical small-scale farmer in the sample cultivates 0.5–1.0 ha of raspberries, which provides most of the household’s cash income, and 2.0–4.5 ha of various crops. In some cases, all or part of the farm is dedicated to pastures for cattle. The most frequent irrigation systems are furrow and flood, and only a few farms have drip irrigation (0.7% of the sample). Management practices include conventional tillage and two or three fertilizer applications per season. Crop stubble is usually burned, causing a reduction in soil quality (McCool et al., 2008), which in turn forces an increase in inputs and leads to soil degradation. Less commonly, the stubble is incorporated into the soil or used for compost. In Chile, soil conservation policies are the responsibility of the Vice-Ministry of Agriculture (Subsecretaría de Agricultura) and of the Agricultural and Livestock Service (ServicioAgrícola y Ganadero or SAG), while irrigation policies fall under the National Irrigation Commission (Comisión Nacional de Riego or CNR). These institutions not only generate policy strategies but also provide incentives to promote desirable soil and water management practices. These incentives are allocated to producers throughout the country according to the requirements of each municipality and region. In the case of small-scale farmers, these subsidies are administered by the National Institute for Agricultural Development (INDAP), which is the agency responsible for that segment of the farm population.

Empirical model A Seemingly Unrelated Bivariate Probit Regression (SUBPR) model is used to estimate the joint adoption of WC&SC.3 The decision to adopt is conditioned by a set of variables related to natural, social, human, physical, and financial capital associated with the farm and the farmer (Boyd et al., 2000). The dependent variable was constructed based on a set of conservation strategies used by farmers, as outlined in Table 3. The most adopted soil conservation is “crop rotation” (62.8%) while “removal of weeds in channels” (87.1%) is the most prevalent water conservation. The SUBPR model requires two dichotomous dependent variables; thus, a panel of Chilean experts in soil and water conservation was set up to rank the strategies according to their importance vis a vis conservation. This ranking was then used to define the adoption

3 The model was estimated under two additional specifications of the dependent variables: (1) Two Seemingly Unrelated Poisson equations where the dependent variables are the number of soil and water conservation adopted, respectively; and (2) Two Seemingly Unrelated Tobit models, where a weighted dependent variable was used according to the adoption level and opinions from a panel of experts. The econometric results of all three models are consistent and lead to similar implications. Thus, the analysis is based on the Seemingly Unrelated Bivariate Probit Regression (SUBPR) because the interpretation of the results is more straight forward compared to the other two alternatives. The econometric results for the two alternative models are available from the authors upon request.

296

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

Table 2 Alternative farming systems in the sample. Farming system

Crop(s)

One crop

Raspberry Corn Rice Pasture Wheat Beans Chacraa Tomato Sweet Corn Beet Othersb

Sub-total one crop Two crops

Corn–beans Raspberry–otherc Raspberry–corn Wheat–beans wheat–corn Raspberry–pasture Raspberry–wheat Wheat–rice Wheat–otherd Corn–otherd Wheat–pasture Raspberry–beans Otherd

Sub-total two crops Three crops

Raspberry–CORN/wheat–otherd Wheat–corn–otherd Rice–corn/wheat–otherd Beans–corn/wheat–otherd Wheat–corn–beans Wheat–corn–raspberry Raspberry–blackberry–otherd Othere

Sub-total three crops Four crops

Beans–corn–raspberry–otherd 4 annual crops or vegetables Corn–wheat–raspberry–otherd 3 annual crops and pasture 3 annual crops and raspberry

Sub-total four crops Total a b c d e

Frequency

Percentage (%)

23 19 17 13 11 8 8 2 2 2 3

7.5 6.2 5.5 4.2 3.6 2.6 2.6 0.7 0.7 0.7 1.0

108

35.2

13 9 8 8 8 7 6 6 5 5 4 4 17

4.2 2.9 2.6 2.6 2.6 2.3 2.0 2.0 1.6 1.6 1.3 1.3 5.5

100

32.6

14 14 10 10 7 6 4 13

4.6 4.6 3.3 3.3 2.3 2.0 1.3 4.2

78

25.4

6 6 3 3 3

2.0 2.0 1.0 1.0 1.0

21 307

6.8 100

Onion, lettuce, oats, kiwi, sunflower. Intercropping: sweet corn–beans–potatoes. Chacra, sweet corn, lettuce, blackberry. Annual crop, pasture or vegetables. Include pasture, raspberry, annual crops and vegetables.

variable for each model. Table 3 shows the mean score given by the experts for each conservation. In the case of water, there are three conservation with low impact, according to the scores given by the experts (“removal of weeds in channels”, “cleaning sediments in channels” and “stone bunds”) and three conservations with high impact (“work to improve water conduction”, “new channels infarm” and “pressurized irrigation system”). Thus, the variable water conservation is represented by the binary variable WATER, which is equal to 1 if a farmer adopts at least two conservation strategies with a low score or at least one conservation with a high score, or 0 otherwise. Table 3 also shows the mean score of the soil conservation. “Fallow pasture” has a low score, while “pasture for cutting” “manure” and “improved pasture” have middle scores, meaning that they

have a medium impact on conservation. The rest of the soil conservation strategies have a high impact. Thus, adoption of soil conservation is represented by the variable SOIL, which takes the value of 1 if the farmer adopts at least two conservations in an annual cropping system and at least one practice in a fruit system or 0 otherwise. This definition takes into account the fact that crop rotation, one of the most commonly used soil conservation for annual crops, is not appropriate for perennial crops, such as raspberries, that is present as a monoculture crop in 14% of the sample (includes farmers that only have raspberries as a cash crop). Although the definition of the dependent variables SOIL and WATER is somewhat arbitrary, the aim is to classify a farmer as an adopter when adoption can be expected to have a significant impact on conservation.

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301 Table 3 WC&SC and their ranking based on expert opinion. Soil conservation

Fallow pasture Crop rotation Cover crop Stubble incorporation Pasture for cutting Stubble cover Compost Mulch Manure Ridge crops Improved pasture

% Adoption

The empirical models can be written as: Score of impact on soil conservation (scale 1–7)a

14.7 62.8 7.4 57.1 4.5 5.1 0.6 0.3 3.8 0.3 13.8

SOILi = ˛si + ˇs1 AGEi + ˇs2 EDUCi + ˇs3 FAMLABi + ˇs4 LANDi +ˇs5 QSOILi + ˇs6 SELFCONi + ˇs7 RASPBi + ˇs8 LGRASSi +ˇs9 OFFINCi + ˇs10 CREDi + ˇs11 PARTi + ˇs12 EXTENi + εsi

2.7 6.3 6.0 5.7 5.3 5.7 6.3 6.0 5.0 6.0 5.5

WATERi = ˛wi + ˇw1 AGEi + ˇw2 EDUCi + ˇw3 FAMLABi +ˇw4 LANDi + ˇw5 QSOILi + ˇw6 SELFCONi + ˇw7 OFFINCi

(6)

+ˇw8 CREDi + ˇw9 INCENTi + ˇw12 WCOMi + εwi

Water conservation

% Adoption

Score of impact on water conservation (scale 1–7)

Removal of weeds in channels Cleaning of sediments in channels Work to improve water conduction New channels in-farm Stone bunds Pressurized irrigation

87.1 26.3 2.6 20.9 1.0 0.6

4.7 4.7 6.7 6.0 4.0 6.7

% Non-adopters of soil conservation % Non-adopters of water conservation % Non-adopters soil and water conservation

18.3 9.9 5.5

a

297

Score values given by a panel of experts ranging from 1 (low) to 7 (high).

The explanatory variables were selected based on data availability and on the relevant literature. A recent comprehensive metaanalysis of Conservation Agriculture in farming by Knowler and Bradshaw (2007) provides a useful reference for model formulation. Table 4 shows the description and definition of all dependent and independent variables. Results and discussion Table 4 shows the descriptive statistics for all variables included in the models. According to the definition of adoption used in this study, 62% of the farmers adopt soil conservation, 51% adopt water conservation and 37.5% adopt both. Hence, the distribution of adopters and non-adopters in the sample is well balanced, although, and as mentioned in the previous section, “crop rotation” and “removal weeds in channels” are the most recurrent WC&SC observed, respectively. Average farm size is 11.7 ha, and average age of the head of household is 57, which is consistent with the widely held view that young people tend to migrate from rural to

Table 4 Description and definition of the variables used in the econometric model on WC&SC. Variables

Definition

Continuous variables

Mean SOIL

WATER AGE EDUC FAMLAB LAND QSOIL HOMECON RASPB

1 if the farmers adopt at least one conservation in fruit crops systems and at least two conservations in annual crops system, 0 otherwise 1 if the farmers adopt at least two conservations with low score or at least one conservation with high score, 0 otherwise Age of head of household in years Educational level of head of household in years Number of family members working on-farm Farm size in hectares 1 for farms with low fertility soil, 0 otherwise 1 if the farmer does not sell products, 0 otherwise 1 if the farm has only raspberries as a cash crop, 0 otherwise 1 if the farm has only pasture, 0 otherwise 1 if the farm has mixed crops, 0 otherwise (omitted category) 1 if the farmer received off-farm income, 0 otherwise

Dummy variables (% = 1) S.D. 62

51 57.1 6.3 2.1 11.7

14.1 3.7 0.9 12.3 30 12 14

4

PASTURE 1 if the farmer used credit, 0 otherwise

72

MIXCROP

OFFINC CRED INCENT WCOMa PARTb EXTEN a

1 if the farmer has received subsidies that improve the irrigation infrastructure, 0 otherwise 1 if the farmer participates in water community meetings, 0 otherwise. Index of social activities (varies between 0 and 1). 1 if the farmer reported having had the visit or assistance from an extensionist, 0 otherwise

41 48 6 66 0.4

0.24 33

Participation in a water community implies only being registered in that community. It does not imply participating in meetings or being part of the decision making. b The Index is estimated as follows: number of social activities that each farmers participate/total number of activities (membership in agricultural cooperatives, agricultural associations, neighborhood boards, church participation, and sport clubs, among others).

298

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

Table 5 Seemingly unrelated bivariate Probit results of WC&SC (Robust standard errors in italics). Variables Constant AGE EDUC FAMLAB LAND QSOIL HOMECON RASPB PASTURE OFFINC CRED INCENT WCOM PART EXTEN Log-likelihood Rho () N

SOIL −1.1408* 0.0100 0.0286 0.1659* 0.0156* −0.1485 −0.5529** −0.6509*** −1.2884** 0.2768 0.5154*** – – 0.0479 0.3033*

WATER 0.6395 0.0074 0.0259 0.0933 0.0090 0.1870 0.2808 0.2572 0.6156 0.2277 0.1929 – – 0.3668 0.1819

−0.2305 −0.0033 0.0055 0.0775 0.0297*** −0.5447*** −0.4793* – – −0.0165 −0.2533 0.7906** 0.3276** – –

0.5958 0.0074 0.0251 0.0843 0.0100 0.1767 0.2973 – – 0.2174 0.1791 0.4111 0.1759 – –

−344.04 0.262*** 307

Robust standard errors used to correct for heteroscedasticity (Greene, 2007). * Significant at 10%. ** Significant at 5%. *** Significant at 1%. –: variables not included. Table 6 Percent of correct predictions of WC&SC. Dependent variable

Percent of correct predictions (%)

Soil conservation Water conservation Joint soil and water adoption

74.6 65.5 51.1

urban areas in search of better opportunities. The educational level of the farmers in the sample is rather low; the average farmer has 6.3 years of schooling and only 5.3% have a college degree or higher. In terms of land use, 14% of the farmers grow only raspberries as a cash crop, 4% have only pastures4 and 72% have diversified cropping patterns. It is relevant to mention that 12% of the farmers grow crops only for home consumption, and receive no income from their farming activities. Average farm size is only 1.4 ha for these farmers, they grow mostly vegetables and 70% have off-farm income, whereas the latter figure is 41% for the sample as a whole. As shown in Table 5, the econometric results reveal that, overall, the parameter estimates are highly significant, with a Log-likelihood value of −344. Moreover, the observed value of the dependent variable is predicted correctly 74.6% and 65.5% of the times in the SOIL and WATER equations, respectively. For joint adoption, the percentage of correct predictions is 51.1% (see Table 6). In addition, the rho () coefficient is positive (0.262) and statistically significant at the 1% level, indicating that the disturbance terms of the two equations are correlated, which lends support to the use of the SUBPR model. Moreover, the positive sign for  suggests that WC&SC are complementary (Rahim et al., 2005). In the SOIL equation, seven coefficients (out of 13) are significant at the 10% or better. FAMLAB, LAND, CRED and EXTEN which measure Family labor, farm size, access to credit and support from extension, respectively, all have a positive and significant effect on the likelihood of adopting conservation. In contrast, RASPB and PASTURE both binary variables for farm orientation (raspberries

4 Pasture for cattle are included in the sample because they are irrigated by flooding with fertilization every year and occasionally farmers improve pasture by sowing new seeds.

Table 7 Marginal Effects (MEs) from the SOIL and WATER adoption equations (Standards errors in italicsa ). Variables

Marginal Effects

AGE EDUC FAMLAB LAND QSOIL HOMECON RASPB PASTURE OFFINC CRED INCENT WCOM PART EXTEN

0.0036 0.0096 0.0532* 0.0036 −0.0197 −0.1645* −0.2269** −0.4491** 0.0974 0.1946 0.0465 0.0193 0.0167 0.1057*

0.0024 0.0089 0.0326 0.0032 0.0661 0.0995 0.0905 0.2192 0.0769 0.0662 0.0302 0.0124 0.1278 0.0608

Note: The Marginal Effects for dummy variables are computed as: E[y1|y2 = 1,d = 1] − E[y1|y2 = 1,d = 0] where d is the dummy variable. a The standard errors are computed using the delta method (Greene, 1996). * Significant at 10%. ** Significant at 5%.

and pastures), and HOMECON (production exclusively for home consumption), have a negative and significant influence on adoption. In the WATER equation, five coefficients (out of 11) are significant at least at the 10% level. LAND, INCENT (incentives for irrigation investments) and WCOM (a binary variable that captures an active participation in water communities) have a positive and significant effect on adoption. In contrast, QSOIL (a binary variable capturing low soil fertility) and HOMECON have a negative and significant influence. The variables LAND and QSOIL represent natural capital at the farm level. As expected, the parameters for LAND are positive and significant in both the SOIL and WATER equations, suggesting that families with larger farms are more likely to adopt conservation. A graphical analysis confirms a linear relationship between farm size and the probability of adoption (SOIL = 1, WATER = 1). This finding is consistent with results reported by Asafu-Adjaye (2008), Marenya and Barrett (2007), Cramb et al. (2007) and Westra and Olson (1997). Table 7 shows Marginal Effects (MEs), calculated using Eq. (5), for the case of joint adoption (i.e., when SOIL and WATER are both equal to 1). Following Greene (1996), the values reported are percent changes. The ME for LAND is low (0.003) suggesting that a 10% increase in the amount of land cultivated increases the probability of adopting WC&SC by 0.3%. The parameter for QSOIL is negative in both equations but significant only in the WATER model. This result suggests that farmers with poor soil have a lower probability of adopting conservation. The ME for this variable is 0.019, implying that farmers with low quality soil are 1.9% less likely to adopt WC&SC than farmers with better soils. The parameters for EDUC are not significant although this variable is expected to have a positive effect on adoption while the same lack of significance is found for AGE. These findings are consistent with mixed interpretations reported in the literature. For example, Amsalu and de Graaff (2007) argue that age is an indicator of experience and thus is expected to have a positive association with adoption. On the other hand, Norris and Batie (1987), Lapar and Pandey (1999), Lichtenberg (2001) and Cramb (2005) show that young farmers have a higher probability of adopting innovative technologies. The parameter for FAMLAB is positive and significant at the 10% level in the SOIL equation and positive but not significant in the WATER equation. The ME indicates that a 10% increase in family labor leads to a 0.5% rise in the probability of a farmer adopting

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

WC&SC. The parameter for HOMECON is negative and significant in both equations. The ME for HOMECON reveals that the probability of adoption is 16% lower for farmers who produce only for home consumption than for those who produce and sell for the market and thus generate cash income. As indicated earlier, the HOMECON farmers represent about 12% of the sample and most have off-farm income. We also included a binary variable to capture whether or not the farmer receives off-farm income (OFFINC), and the results indicate that the associated parameters are not significant in both equations. By contrast, Mbaga-Semgalawe and Folmer (2000), and Bravo-Ureta et al. (2006) have reported a negative association between off-farm work and technology adoption, while Tenge et al. (2004) also found that off-farm work has a negative effect on WC&SC. Off-farm income reduces the available time for agricultural work, and farmers may be less concerned about improving natural resource quality due to their orientation toward off-farm employment (Bravo-Ureta et al., 2006). An attempt is also made to capture the influence of monoculture farming on adoption by including the variables RASPB and PASTURE in the SOIL equation. The parameters of both variables are negative and highly significant, indicating that farmers that only produce raspberries or pastures adopt fewer WC&SC than farmers with diversified cropping patterns. The MEs for both variables are relatively high. In the case of pastures, the probability of adoption drops by 45% and in the case of raspberries by 23%. The financial capital variable CRED is highly significant in the SOIL equation and this is in line with the findings of Caviglia-Harris (2003, 2004) who states that credit gives an incentive for farmers to adopt sustainable activities. However, in the WATER equation, the parameter of CRED shows inconclusive results. Two additional variables are included in the WATER equation: INCENT and WCOM.5 As expected, the parameters for both variables are positive and significant. About 7% of the farmers in the sample received subsidies for the improvement of irrigation canals and infrastructure to avoid water losses. The estimated MEs reveal that these subsidies increase the chances that a farmer will adopt conservation by 4.7%, while participation in a water community does so by 1.7%. Chile has two major laws related to government investment in irrigation infrastructure that have bearing on water investment and conservation (Ríos Brehm and Quiroz, 1995). One of these laws provides the framework for funding major projects including the construction of dams and inter-basin channels, and the improvement of existing irrigation systems. The second law establishes subsidies that encourage smaller private investments for the construction and improvement of irrigation infrastructure. It subsidizes up to 75% of the costs related to the design and installation of minor works. Thus, our results are consistent with a positive impact of the incentives that promote conservation. However, the availability of credit is found to be relevant only for soil conservation. These findings could be interpreted in two ways: first, since subsidies for soil conservation are more limited, credit becomes more relevant; and second, the most adopted water conservation, “removal of weeds in channels”, does not require capital, therefore CRED is less important. The variable CRED includes access to credit from INDAP, but also from other institutions such as banks, credit agencies, and input suppliers. To account for social capital, the SOIL equation includes an index that measures participation in social activities, PART, and EXTEN. The parameter for PART is positive but non-significant, while the one for EXTEN is positive and significant, which is in line with Shultz

5 As we state in “Materials and methods” section, the sample was drawn from water communities. However, farmer participation in water community meetings is not a compulsory activity.

299

et al. (1997). The estimated MEs show that extension increases the likelihood of adoption by 10.5%. Although INDAP is in charge of extension and credit, these programs are managed independently of each other.6 Thus, it is possible to find farmers who are involved in both, one or neither of the two programs. Policy implications The results of our analysis are highly consistent with the literature in terms of the influence of several of the variables considered on the adoption of conservation. As stated in the introduction, the adoption of a technology is not only related to its expected profitability, but also to socioeconomic, cultural, and natural resource factors (Lapar and Pandey, 1999; Soule et al., 2000). Regarding the latter, the analysis showed that farmers with poor quality soils exhibit low adoption rates suggesting that they expect limited productivity gains from conservation and therefore conclude that adoption is not profitable. In other words, poor soil quality becomes a disincentive for conservation, exacerbating the problem of future sustainability of the agricultural system. The results shown have several implications for agricultural and conservation policy. First, since extension proved to have a high influence on adoption, one way to increase the probability of adoption is to strengthen extension education programs focusing on WC&SC through water communities, which are established organizations with presence in all irrigated areas in Chile. Also, promoting the engagement of farmers in such organizations should be the focus of extension programs. Second, targeting economic incentives in coordination with water communities appears to be a desirable policy to increase the efficiency of such funding. Third, the promotion of conservation among monocultural farmers, represented here by raspberries and pastures, requires special efforts given the low levels of adoption observed among these producers. These findings are consistent with the fact that the most widely adopted soil conservation is crop rotation, which is not applicable to raspberry or pasture systems. However, strategies that are appropriate for raspberries, such as the use of cover crops, stubble cover, compost, and the use of manure, still exhibit a low likelihood of adoption. The same is true for the use of manure on pastures. Therefore, the incentives to invest in conservation should be coupled with on-going training programs designed to educate producers, particularly in the more complex conservation technologies, and INDAP has a crucial role to play in this matter. Fourth, special efforts are needed to focus on the poorest/subsistence farmers which are unlikely to adopt WC&SC without special incentives. Therefore, creative programs, specifically targeted to this group of producers, combining social and economic incentives to encourage conservation, deserve serious consideration. Generally, the most vulnerable farmers who cannot support their households with the income they generate in their farms see younger family members leaving the rural areas in favor of urban alternatives seeking better opportunities. This phenomenon, which is more evident among producers with lower human, social and natural capital, has led to urban unemployment, social tension and environmental problems in many cities around the developing world (Bilsborrow, 1992). These challenges could be mitigated with appropriate incentives to promote adequate returns to poor farm families and thus make farming an attractive way of life for the present and future generations. Fifth, given that the adoption of WC&SC are complementary, agricultural extension and economic incentives to improve conservation of both resources should be considered as a package, which requires the articulation of the work

6 It is worth noting that the correlation between CRED and EXTEN is positive but not significant (0.277).

300

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301

done by several different units within and outside the Ministry of Agriculture. Again, INDAP has a leading role given that it finances technical assistance and management plans for smallholders and implements general assistance programs for subsistence farmers. Moreover, this Institute has the capacity and the means to develop and better target incentives to increase access to credit that could improve the adoption of conservation techniques. It is also in a position to strengthen the function of social capital through water communities, which can be vital in facilitating technology adoption. Hence, INDAP should target extension efforts on water community groups rather than on individual producers in order to increase the participation of poor farmers in innovative activities that would conserve resources while increasing welfare. Conclusions This study examines the joint decision of adopting WC&SC by small-scale farmers in the Province of Linares, Chile. The analysis is based on a Seemingly Unrelated Bivariate Probit (SUBPR) model, given that the disturbance terms in the two adoption equations (SOIL and WATER) are positively and significantly correlated. The results reveal that the overall statistical significance and predictive ability of the bivariate Probit model is relatively high, implying that this is a suitable approach for the case at hand, and that WC&SC are complementary. It is important to underscore that even though the sample includes only smallholders, farm size is a significant variable in increasing the probability of adopting conservation. Moreover, farms with cash crops and diversified farming systems are more likely to adopt WC&SC. At present, it is the poorest farmers—those who cultivate the smallest plots and produce primarily for home consumption—who are the least likely to adopt WC&SC. Access to credit and extension services play an important role, while incentives to improve the irrigation infrastructure and participation in water communities are also crucial in the adoption of water conservation. References Adesina, A.A., Zinnah, M.M., 1993. Technology characteristics, farmers’ perceptions and adoption decisions: a Tobit model application in Sierra Leone. Agricultural Economics 9, 297–311. Amsalu, A., de Graaff, J., 2007. Determinants of adoption and continued use of stone terraces for soil and water conservation in an Ethiopian highland watershed. Ecological Economics 61, 294–302. Anley, Y., Bogale, A., Haile-Gabriel, A., 2007. Adoption decision and use intensity of soil and water conservation measures by smallholder subsistence farmers in Dedo District, Western Ethiopia. Land Degradation & Development 18, 289–302. Asafu-Adjaye, J., 2008. Factors affecting the adoption of soil conservation measures: a case study of Fijian cane farmers. Journal of Agricultural and Resource Economics 33 (1), 99–117. Barbier, E.D., Bishop, J.T., 1995. Economic values and incentives affecting soil and water conservation in developing countries. Journal of Soil and Water Conservation 50 (2), 133–137. Bilsborrow, R., 1992. Rural Poverty, Migration, and the Environment in Developing Countries. Three Case Studies. Working paper WPS 1017. World Development Report. Blaike, P., Brookfield, H., 1987. Land Degradation and Society. Methuen & Co. Ltd., London. Bonilla, C.A., Reyes, J.L., Magri, A., 2010. Water erosion prediction using the Revised Universal Soil Loss Equation (RUSLE) in a GIS Framework, Central Chile. Chilean Journal of Agricultural Research 70 (1), 159–169. Boyd, C., Turton, C., Hatibu, N., Mahoo, H.F., Lazaro, E., Rwehumbiza, F.B., Okubal, A.P., Makumbib, M., 2000. The contribution of soil and water conservation to sustainable livelihoods in semi-arid areas of sub-Saharan Africa. Agricultural Research and Extension Network. Network Paper N◦ 102. Bravo-Ureta, B., Solís, D., Cocchi, H., Quiroga, R., 2006. The impact of soil and output diversification on farm income in Central American hillside farming. Agricultural Economics 35, 267–276. Bruinsma, J., 2009. The Resource Outlook to 2050: By How Much Do Land, Water and Crop Yields Need to Increase by 2050? Food and Agriculture Organization of the United Nations Economic and Social Development Department.

Caviglia-Harris, J., 2003. Sustainable agricultural practices in Rondônia. Brazil: do local farmer organizations affect adoption rates? Economic Development and Cultural Change 52 (1), 23–49. Caviglia-Harris, J., 2004. Household production and forest clearing: the role of farming in the development of the Amazon. Environment and Development Economics 9, 181–202. Calatrava-Leyva, J., Franco Martínez, J.A., González Roa, M.C., 2007. Analysis of the adoption of soil conservation practices in olive groves: the case of mountainous areas in southern Spain. The Spanish Journal of Agricultural Research 5 (3), 249–258. Caswell, M., Fuglie, K., Ingram, C., Jans, S., Kascak, C., 2001. Adoption of Agricultural Production Practices. Lessons Learned from the U.S. Economic Research Service/USDA Chapter 1: Introduction/AER-792 1. Centro de Información de Recursos Naturales (CIREN)-CORFO, 1997. Estudio Agroecológico VII Región, Materiales y símbolos, Chile, pp. 635–659. Cramb, R., 2005. Social capital and soil conservation: evidence from the Philippines. Australian Journal of Agricultural and Resource Economics 49, 211–226. Cramb, R.A., Catacutan, D., Culasero-Arellano, Z., Mariano, K., 2007. The “landcare” approach to soil conservation in the Philippines: an assessment of farm-level impacts. Australian Journal of Experimental Agriculture 47, 721–726. de Graaff, J., Amsalu, A., Bodnár, F., Kessler, A., Posthumus, H., Tenge, A., 2008. Factors influencing adoption and continued use of long-term soil and water conservation measures in five developing countries. Applied Geography 28, 271–280. Ellies, A., 2000. Soil erosion and its control in Chile—an overview. Acta Geologica Hispanica 35, 279–284. Ellis, F., 1993. Peasant Economics. Cambridge University Press, Cambridge. FAO, 1992. The Use of Saline Waters for Crop Production. FAO Irrigation and Drainage. Paper 48. Rome, Italy. Feder, G., Just, R.E., Zilberman, D., 1985. Adoption of agricultural innovations in developing countries: a survey. Economic Development and Cultural Change 33 (2), 225–298. Foltz, J., 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. Francke, S., 1997. Lineamientos de políticas para la ordenación de cuencas hidrográficas y conservación de suelos. Chile Forestal. DocumentoTécnico N◦ 127.7 p. Greene, W.H., 1996. Marginal effects in the bivariate Probit model. Working Paper EC-96-11. Greene, W.H., 2008. Econometric Analysis. Prentice Hall, Upper Saddle River, NJ. Greene, W.H., 2007. NLOGIT Version 4.0. Reference Guide. Econometric Software, Inc. Gupta, R., Seth, A., 2007. A review of resource conserving technologies for sustainable management of the rice–wheat cropping systems of the Indo-Gangetic plains (IGP). Crop Protection 26, 436–447. Hugo, G., 2008. Trends in land degradation in South America. Management of natural and environmental resources for sustainable agriculture. In: Stefanski, R., Pasteris, P. (Eds.), Proceedings of the USDA/WMO Workshop on Management of Natural and Environmental Resources for Sustainable Agricultural Development, United States Department of Agriculture/World Meteorological Organization, 13–16 February 2006, Portland, OR, USA/Washington, DC, USA/Geneva, Switzerland; Technical Bulletin WAOB-2008, NRCS-2008, and AGM-10, WMO/TD No.1428, 211 pp. Hussain, I., Hanjra, M.A., 2004. Irrigation and poverty alleviation: review of the empirical evidence. Irrigation and Drainage 53, 1–15. INE, 2007. VII Censo Agrícola. Número y superficie de las explotaciones agropecuar˜ según Región, Provincia y Comuna. ias con tierra, según tamano, Kabubo-Mariara, J., 2007. Land conservation and tenure security in Kenya: Boserup’s hypothesis revisited. Ecological Economics 64, 25–35. Kessler, C.A., 2007. Motivating farmers for soil and water conservation: a promising strategy from the Bolivian mountain valleys. Land Use Policy 24, 118–128. Kim, S., Gillespie, J.M., Paudel, K.P., 2005. The effect of socioeconomic factors on the adoption of best management practices in beef cattle production. Journal of Soil and Water Conservation 60 (3), 111–120. Knowler, D., Bradshaw, B., 2007. Farmers’ adoption of conservation agriculture: a review and synthesis of recent research. Food Policy 32, 25–48. Lapar, M.L.A., Pandey, S., 1999. Adoption of soil conservation: the case of the Philippine uplands. Agricultural Economics 21, 241–256. Lee, D., 2005. Agricultural sustainability and technology adoption: issues and policies for development countries. American Journal of Agricultural Economics 87 (5), 1325–1334. Lichtenberg, E., 2001. Adoption of Soil Conservation Practices: A Revealed Preference Approach. Department of Agricultural and Resource Economics. The University of Maryland, College Park. Marenya, P.P., Barrett, B.B., 2007. Household-level determinants of adoption of improved natural resources management practices among smallholder farmers in western Kenya. Food Policy 32, 515–536. Mbaga-Semgalawe, Z., Folmer, H., 2000. Household adoption behavior of improved soil conservation: the case of the North Pare and West Usambara Mountains of Tanzania. Land Use Policy 17, 321–336. McCool, D.K., Pannkuk, C.D., Kennedy, A.C., Fletcher, P.S., 2008. Effects of burn/lowtill on erosion and soil quality. Soil & Tillage Research 101, 2–9. Muchena, F.N., Onduru, D.D., Gachini, G.N., de Jager, A., 2005. Turning the tides of soil degradation in Africa: capturing the reality and exploring opportunities. Land Use Policy 22, 23–31.

R. Jara-Rojas et al. / Land Use Policy 32 (2013) 292–301 Neill, S.P., Lee, D.R., 2001. Explaining the adoption and disadoption of sustainable agriculture: the case of cover crops in Northern Honduras. Economic Development and Cultural Change 49 (4), 793–820. Nkamleu, G.B., Adesina, A.A., 2000. Determinants of chemical input use in peri-urban lowland systems: bivariate probit analysis in Cameroon. Agricultural Systems 63, 111–121. Norris, E.P., Batie, S.S., 1987. Virginia farmers’ soil conservation decisions: an application of Tobit analysis. Southern Journal of Agricultural Economics 19 (1), 89–97. Nyariki, D., 2009. Household data collection for socio-economic research in agriculture: approaches and challenges in developing countries. Journal of Social Science 19 (2), 91–99. OECD, 2008. OECD Review of Agricultural Policies, Chile. ISBN 978-92-64-04223-0. Pérez, C., González, J., 2001. Diagnóstico sobre el estado de degradación del recurso suelo en el país. Instituto de Investigaciones Agropecuarias (INIA), Chillán, Chile, 196 pp. Pimentel, D., Harvey, C., Resosudarmo, P., Sinclair, K., Kurz, D., McNair, M., Crist, S., Shpritz, L., Fitton, L., Saffouri, R., Blair, R., 1995. Environmental and economics costs of soil erosion and conservation benefits. Science 267, 1117–1123. Rahim, A.H., Ruben, R., van Ierland, E.C., 2005. Adoption and abandonment of gum Arabic agroforestry in Sudan. In: AFTA 2005 Conference Proceedings. Rahm, M., Huffmann, W., 1984. The adoption of reduced tillage: the role of human capital and other variables. American Journal of Agricultural Economics 66, 405–413. Reilly, J., Paltsev, S., Felzer, B., Wang, X., Kicklighter, D., Melillo, J., Prinn, R., Sarofim, M., Sokolov, A., Wang, C., 2007. Global economic effects of changes in crops, pasture, and forests due to changing climate, carbon dioxide, and ozone. Energy Policy 35, 5370–5383. Ríos Brehm, M., Quiroz, J., 1995. The market of water rights in Chile: major issues. Cuadernos de Economía 32, 317–345. Ruben, R., Pender, J., Kuyvenhoven, A., 2004. Editorial: Development strategies for less-favoured areas. Food Policy 29, 295–302. Shultz, S., Faustino, J., Melgar, D., 1997. Agroforestry and soil conservation: adoption and profitability in El Salvador. Agroforestry Today 9, 16–17.

301

Schultz, B., Thatte, C.D., Labhsetwar, V.K., 2005. Irrigation and drainage. Main contributors to global food production. Irrigation and Drainage 54, 263–278. ˜ P., 2012. Informe Agrometeorológico para el Ministerio de Agricultura, Santibánez, Región del Maule, 14 pp. Solís, D., Bravo-Ureta, B., 2005. Economic and financial sustainability of private agricultural extension in El Salvador. Journal of Sustainable Agriculture 26, 81–102. Solís, D., Bravo-Ureta, B., Quiroga, R., 2009. Determinants of household efficiency among small-scale hillside farmers in El Salvador and Honduras. Journal of Agricultural Economics 60 (1), 202–219. Soule, M., Tegene, A., Wiebe, K., 2000. Land tenure and the adoption of conservation practices. American Journal of Agricultural Economics 4, 993–1005. Staal, S.J., Baltenweck, I., Waithaka, M.M., De Wolff, T., Njoroge, L., 2002. Location and uptake: integrated household and GIS analysis of technology adoption and land use, with application to smallholder dairy farms in Kenya. Agricultural Economics 27, 295–315. Tenge, A.J., de Graaff, J., Hella, J.P., 2004. Social and economic factors affecting the adoption of soil and water conservation in West Usambara highlands, Tanzania. Land Degradation & Development 15, 99–114. Thuo, M., Bravo-Ureta, B.E., Obeng-Asiedu P., Hathie, I., 2013. Adoption of agricultural input packages by smallholder farmers: the case of groundnut hybrid seed and chemical fertilizer in the Senegalese groundnut basin. Journal of Developing Areas, in press. U.S. Environmental Protection Agency. 1997. Terms of Environment: Glossary, Abbreviations and Acronyms [Online], Washington, DC. Available from: http://www.epa.gov/OCEPAterms/ Vidal, I., Etchevers, J., Fischer, A., 2002. Dinámica del nitrógeno bajo diferentes rotaciones, sistemas de labranza y manejo de residuos en el cultivo de trigo. Agricultura Técnica 62 (1), 121–132. Wandel, J., Smithers, J., 2000. Factors affecting the adoption of conservation tillage on clay soils in Southwestern Ontario. American Journal of Alternative Agriculture 15 (4), 181–188. Westra, J., Olson, K., 1997. Farmers’ decision processes and adoption of conservation tillage. Staff Paper P97-9, University of Minnesota.