Environmental Management DOI 10.1007/s00267-016-0759-2

Farmers’ Options to Address Water Scarcity in a Changing Climate: Case Studies from two Basins in Mediterranean

1 2 3 4 Lisandro Roco ● David Poblete ● Francisco Meza ● George Kerrigan

Received: 17 December 2015 / Accepted: 10 August 2016 © Springer Science+Business Media New York 2016

Abstract Irrigated agriculture in Mediterranean areas faces groups: investments for water accumulation, modernization tremendous challenges because of its exposure to hydro- of irrigation systems, rationalization of water use, and climatic variability, increasing competition for water from partnership activities. Using a multinomial logit model these different sectors, and the possibility of a climatic change. In strategies were related to socioeconomic and productive this context, efficient management of water resources characteristics. Results show that gender and farm size are emerges as a critical issue. This requires the adoption of relevant for investments, implementation and improvement technological innovations, investment in infrastructure, of irrigation systems. For all the strategies described, access adequate institutional arrangements, and informed decision to weather information was a relevant element. The study makers. To understand farmers’ perceptions and their provides empirical evidence of a recent increase in the implementation of climate change adaptation strategies with importance assigned to climate factors by producers and regards to water management, primary information was adaptation options that can be supported by agricultural captured in the Limarí and basins in Chile. policy. Farmers identified stressors for agriculture; climate change, droughts, and lack of water appeared as the most relevant Keywords Agricultural production ● Chile ● Drought ● stressors compared to others productive, economic, and Multinomial logit ● Stressors ● Water scarcity institutional factors; revealing a rising relevance of climate related factors. While most producers perceived climate changes in recent years (92.9 %), a significant proportion (61.1 %) claim to have experienced drought, whereas only a fraction (31.9 %) have implemented a strategy to deal with Introduction this situation. Identified actions were classified in four Water availability plays a key role in agricultural production systems. Agriculture uses—consumed and non-consumed * Lisandro Roco fractions—more than two-thirds of global fresh water and [email protected] 90 % of this share is used by developing countries. Fresh

1 water use is expected to rise another 25 % by 2030 due to an Department of Forestry, Faculty of Agriculture and Forestry, increase in population from 6.6 billion to about 8 billion by Universidad Católica del Maule, PO Box 617, Av. San Miguel # 3605, , Chile 2030 (Misra 2013). Additionally, climate change is emer- ging as one of the main threats for food security in devel- 2 School of Civil Engineering, Universidad de Valparaíso, Valparaíso, Chile oping countries (Comoé and Siegrist 2015). The 3 Intergovernmental Panel on Climate Change defines this Centro Interdisciplinario de Cambio Global, Departamento de “ Ecosistemas y Medioambiente, Facultad de Agronomía e phenomenon as a change in the state of the climate that can Ingeniería Forestal, Pontificia Universidad Católica de Chile, be identified by changes in the mean and/or the variability , Chile of its properties, and that persists for an extended period, 4 Independent Consultant on Agricultural Economics, Talca, Chile typically decades or longer. It refers to any change in Environmental Management climate over time, whether due to natural variability or as a argue that droughts will be longer and more severe in the result of human activity” (IPCC 2007; 2014). Unevenness future; therefore, water governance in water-scarce areas of in water availability and quality will affect society in many Chile will be the most important issue to address in future ways; changes in climatic variables, such as temperature water management plans in order to appropriately adapt to and precipitation, have significant impacts on water potential impacts of climate change in Chile. Water man- resources and hence on societies and ecosystems (Mata and agement decisions must be based on a sound knowledge of Budhooram 2007). Wang et al. (2014) argue that water the hydrology of the system that will be affected. A lack of demand in agriculture will be affected more heavily than understanding of hydrological connections may produce other sectors, given that irrigation includes the major por- unforeseen effects that could affect third parties such as tion of global consumptive water use, the increase of water small farmers (Arumí et al. 2013) and worst of all, could led demand in irrigation may cause severe stress on water to maladaptation processes. In Chile, The Water Resources resources. Directorate (Direccioń General de Aguas, DGA), is the These conditions are especially relevant for Mediterra- government organization in charge of the water resources of nean ecosystems. While all Mediterranean-type climates the country, the DGA and other relevant public and private present bi-seasonality in temperature and precipitation, institutions presents competition across different scales and there are important differences between northern and levels of interaction associated with water resources gov- southern hemispheres, it is expectable that climatically ernance (Valdes-Pinedá et al. 2014). induced land degradation would be less rapid in Chile, The recent study of Vicuna et al. (2014) demonstrates South Africa, and Australia than in California and the that in recent years, an unusually long drought has changed Mediterranean basin (Davis et al. 1996). Nevertheless, the expectations with regard to climate variability in the region. Mediterranean condition is a result of the global climatic This situation has led to dramatically lower reservoir levels forcing on the climate system of the Central Chile region, and continuous reductions in water supply. In addition, due the corresponding interannual variability is quite important to the already high levels of irrigation efficiency and large for water resources management, especially for agriculture amount of acreage devoted to crops with high water in river basins with high prevalence of smallholder farming requirements, the effectiveness of the portfolio options systems. In this regard, García de Jalón et al. (2014) argue available to farmers to adapt to these stressed conditions is that agricultural water management needs to evolve given limited. the possible increase of water scarcity in Mediterranean On the other hand, adaptive capacity is defined as the basins. capacity of actors to respond to variability and create In Chile, the agricultural sector represents 73 % of con- changes in the state of the system (Nelson et al. 2007; sumptive water use, allowing the irrigation of 1.1 million Adger et al. 2005). The presence of adaptive capacity leads hectares and generating exports for 22 % of the national to a better ability to cope with climate risk (Clarvis and total (2011 data) and employing about 9 % of the workforce Allan 2014). As is stated by Zilberman et al. (2012), (Chilean Ministry of Public Works 2012). The agricultural adaptation can be defined as changes in the private and sector accounted for 3 % of national GDP in 2011. A major public decision-making process with respect to resource challenge in this sector is the increase of water productivity, allocation; therefore, adaptation strategies will include both which represents an improvement on irrigation technology public and private actions. In this sense, Kiparsky et al. and the development of infrastructure for conduction and (2012) argue that water decision makers at all levels are storage of water, considering that improvements in irriga- already familiar with managing a source strongly affected tion technology can reduce the use of water by 50 %. Fol- by weather and climate. Additionally, it is expectable that lowing Pereira et al. (2012 and 2009), it is necessary to less demanding options, in terms of time and investment, increase the beneficial water use and to reduce the non- can be implemented in the first stages. Berger and Troost beneficial water use in agriculture. A combination of tech- (2014) differentiate short-term and long-term adjustments; nical, managerial, legal, and investment options are needed the former are typically extensions of existing schemes to help farmers produce more with less water (FAO 2013). dealing with climate variability and do not carry mayor Additionally, Chile is highly impacted by climate costs. Conversely, the latter imply large changes in pro- variability associated to El Niño Southern Oscillation duction systems and land use; they often require new (ENSO) and the Pacific Decadal Oscillation (Ancapichun institutional arrangements or investments in infrastructure and Garces-Vargas 2015); these climatic phenomena pro- and may thereby lead to structural breaks. duce multi-year droughts that affect Chilean agriculture. According to the analysis of Brechét et al. (2013) some Beyond these somewhat predictable oscillations, climate adaptation measures are undertaken by the government to change is introducing a high degree of hydrological contribute to the public good (the macro level), while others uncertainty (Arumí et al. 2013). Valdes-Pinedá et al. (2014) are motivated by the self-interest of individual economic Environmental Management agents (the micro level). In the first group we can mention assistance, may both exacerbate this tendency and increase public investment in irrigation infrastructure and the the scale of future disasters. Additionally, it is recognized breeding of drought-tolerant, heat-tolerant, and early that smallholder farming is among the most vulnerable maturing plant varieties (Deressa et al. 2009). Such actions sectors due to its great social and economic sensitivity; are long-term adaptations. The second group includes the current climate variability is already an issue of concern that adoption by individual farmers of practices such as justifies adaptation efforts (Lindoso et al. 2014). increased irrigation, crop diversification, tree planting, and Under the hypothesis of a rising importance of climate soil and water conservation measures, among others related stressors, given the prevalence of droughts in the (Tambo and Abdoulaye 2012; Manandhar et al. 2011;Di study area, the objective of this study is twofold: first, to Falco et al. 2011; Sofoluwe et al. 2011; Gbetibouo 2009; analyze the position of several sources of stress for agri- Hageback et al. 2005). In this context, water management cultural production; and second, to correlate the options to practices appear as relevant for climate change adaptation; address lack of water implemented by farmers with demo- there is no single approach to drought adaptation, nor does graphic and productive factors affecting their adoption. In one solution fit all regions or countries (Alam 2015). this paper, the aim is to capture the factors influencing the This study focused on farm-level practices, using tech- implementation of differentiated strategies by producers. It nology adoption modeling to describe water scarcity adap- is important to note that this is one of the few studies from tation at the micro level, based on the assumption that a Latin American Region identifying farmers’ responses to farmer will adopt a new technology only if the expected droughts and its results may serve as useful information for utility of that technology is superior to the current utility the design of adaptation policies in developing countries. (for details on the assumption of utility maximization, see Rahm and Huffmann 1984), this rationale provides room for the use of binary dependant variable models in technology Study Area and Methods adoption studies. Most of the scientific information related to water scar- The Study Area and Sample city and climate change and their effect on agriculture comes from developed countries. In general in developing The study area is comprised of two river basins in Chile: countries, there are high levels of uncertainty and vulner- Limarí and Maule, located in the administrative Regions of ability to climate change; and a lack of connection between Coquimbo and Maule, respectively. In Fig. 1, it is possible adaptation to drought and agricultural policy. According to to appreciate the location of river basins, as well the dis- Alam (2015) studies specifically focusing on adaptation to tribution of land uses. These Mediterranean regions present drought and lack of water are scarce. an important dependency on agricultural production and In recent decades Chilean agriculture has experienced consequently on water management. important changes: there is an advantage taken of the A random sample covering 12 municipalities in both Mediterranean climate of the area, the demand in the river basins (Table 1) was obtained considering the differ- Northern Hemisphere for fresh agricultural products in ences in farm size; this sample comprises 665 field surveys winter, and the growing markets for Chilean wines. As a applied in 2013. The survey sections were: identification, consequence, there is increasing production of vegetables, agricultural production, production purposes, use of inputs fruits, and wine in the Central Valley, crops that are and expenditures, livestock production, use of labor, water heavily dependent on water resources (Arumí et al. 2013). rights, investment and water storage, water management, Conversely, Chilean agricultural policy includes a rural risk sources and information access. In the section of program but lacks of an integrated territorial approach for investment and water storage, farmers were asked specifi- rural areas (OECD 2014). In this regard, the researches of cally if they have experienced droughts, and what actions Torres et al. (2015) and Maraseni et al. (2012) support the they implemented to face this situation; then answers were integrated approach to avoid possible conflicts in the design classified into categories. of climate change policies related with water use in The Limarí river basin is in a semi-arid climate and agriculture. comprises more than 80,000 irrigated hectares (Regional With regard to climate risk policy, Vicuna et al. (2014) Government of Coquimbo 2013) distributed into 7,400 argue that the management of expectations about future farms devoted to fruits and vineyard production (DGA droughts or flood assistance is particularly important for 2004a). The Limarí basin drains the western side of the governments. There is always a tendency for companies, Cordillera along a stretch between 30°S and 31.5°S households, and farmers to ignore flood and drought risk within the so-called Norte-Chico, bounded by the extremely and to imprudently invest in new homes and cropland. dry Atacama Desert to the north and Mediterranean Central Climate change, as well as the likelihood of government Chile farther south. In this region, the Andes rise to more Environmental Management

Fig. 1 The study area and distribution of land uses in both river basins

Table 1 Distribution of farmers surveyed in both river basins than 5,000 m within 200 km of the coastline. Consistently, River basin Municipality n % mean annual precipitation exhibits strong spatial variations, ranging from about 100 mm on the coast to about 300 mm Limarí (29.7 %) Monte Patria Monte Patria 123 18.5 at the top of the Andes and a marked decline from south to Ovalle 129 19.4 north (Vicuña et al. 2011). This hydrographic system poses Punitaqui 12 1.8 a big agricultural importance due to its availability of water Total 264 39.7 and soils that allows growing a great diversity of crops; Maule (60.3 %) Linares San Javier 38 5.7 there are land destined to irrigated grasslands, horticulture, Villa Alegre 58 8.7 and vines (DGA 2004a). 53 8.0 The Maule river basin has a temperate climate and Colbún 78 11.7 comprises a surface of more of 460,000 irrigated hectares Total 227 34.1 (INE 2010) divided into 40,000 farms with annual crops, Talca San Clemente 85 12.8 grasslands and fruit production (DGA 2004b). Unlike Talca 35 5.3 Limarí, in Maule the water demand is connected not just to San Rafael 17 2.6 irrigation but also to hydropower generation. The climate in 11 1.6 this area is of the Mediterranean type, considerable wetter Maule 26 3.9 than the Limarí basin; annual precipitation varies between approximately 830 mm in the coastal area at the Nirivilo Total 174 26.2 station (35°S–72°W) and approximately 2,300 mm at Total 665 100.0 the Armerillo station (35.5°S–71°W) in the Andes Environmental Management

(Urrutia et al. 2011). The main economic activities in the where Yi represents the option or strategy to deal with water river basin are agriculture and forestry; agriculture is the scarcity implemented (j) by producer i, w represents a typical of a temperate climate with warm summers and dry, vector of specific characteristics of producer i, in our case cold, and wet winters; in Maule river exists important corresponding to demographic characteristics of the energy generation (DGA 2004b). respondents (age, gender , education, location, and other) In both basins, rain is concentrated in the austral fall and and characteristics of the farm (farm size, land tenure, crops, winter (April–August), when over 75 % of the annual pre- location, and others). Meanwhile, α represents the coeffi- cipitation occurs (Miller 1976; Vicuña et al. 2011). Climate cient matrix associated with each variable (w) and option variability in central Chile is driven by both low-altitude (j). and high-latitude climatic forcings. Extreme dry and wet The implemented strategy j (dependent variable) takes years are related to the El Niño–Southern Oscillation, where the following values: warm ENSO events are characterized by abundant rainfall ● 0, if the producer did not implement any strategy (base in winter, while cold ENSO events are related to dry con- outcome). ditions, low rainfall, and snow accumulation (Aceituno ● 1, if the producer used a strategy based on infrastructure 1988). investment. This category includes shaft sinking, building of ponds, and water purchases. The Farmers’ Decision Model ● 2, if the producer implemented a modernization of their irrigation systems. This category includes improvements Given that agricultural policy is similar for all regions in the in irrigation and pumping systems. country, we used one model for both river basins, controlled ● 3, if the producer used a strategy based on the by binary localization variables. The use of binary location rationalization of water use. This category includes variables in the specification allows the detection of dif- irrigation reduction, reduction of cropped land, and use ferences among geographical areas, but aggregates other of irrigation schedules. drivers that can be acting in a specific territory. ● 4, if the producer used a strategy based on partnership to The modeling approach is based on the estimation of an face droughts. This category includes organization of econometric specification that allows relating factors to a water committees and associations, and communication choice between several responses (Multinomial Logit with peers. Model, MNL). The regression is a classification method that generalizes logistic regression to multiclass problems, It is expected that producers react differently to adverse which is used to predict the probabilities of the different scenarios, in our case drought and water scarcity, and possible outcomes of a categorically distributed dependent implement various strategies for the use of water resources. variable, given a set of independent variables (Greene Table 2 shows the descriptive statistics of the independent 2008). This regression allows explaining the relation variables used in the model. between one dependent nominal variable and one or more The estimated parameters in the MNL indicate the continuous-level independent variables. The main advan- direction of effect of each independent variable on the tage of this model compared to a binary choice model is the response variable, but do not represent the magnitude of possibility to differentiate the performance of factors among change or probability. Therefore it is necessary to calculate the dependent variables, enriching the analyses. This model the marginal effects of the variables, which represent the has been used to understand the strategies of adaptation to expected change in the probability of an option with respect climate change in several African countries: South Africa, to one unit of change of an independent variable, usually Ethiopia, Nigeria and Kenya (Gbetibouo 2009; Deressa et calculated at the average values of the variables (Maddala al. 2009; Gebrehiwot and van der Veen 2013; Sofoluwe et 1987; Greene 2008). al. 2011; Silvestri et al. 2012) in recent years. However, studies using this formulation to model responses regarding the use and planning of water resources in agriculture are Results scarce. The MNL model was implemented for the portion of the Characteristics of Farms and Farmers sample that reported having faced drought situations. The model formulation, according to Greene (2008), for this A description of a typical farmer from the survey corre- study would be: sponds to a producer of 56.5 years of age, with a mean ′α education of 7.6 years. The use of credit by farmers in the exp wi j PYðÞ¼i ¼ jjwi P ; j ¼ 0; 1; ¼ ; 4 ð1Þ municipalities under study ranges from 15.8 % in San Javier 4 ′α j¼0exp wi j to 42.4 % in San Clemente. 70.8 % of the farmers in the Environmental Management

Table 2 General characteristics of the sample and independent variables of the MNL Variable Mean Standard Description Typea Deviation

Demographic characteristics Gender 0.25 0.43 Takes value 1 if the head of the farm is female and 0 otherwise D Age 56.46 14.60 Age of the farmer in years C Education 7.60 4.03 Formal education of the farmer in years C Participation in organizations 0.46 0.49 Takes value 1 if the farmer has participated in organizations in D last years and 0 otherwise Use of credit 0.29 0.46 Takes value 1 if the farmer is using credit and 0 otherwise D Farm characteristics Farm size 16.35 57.7 Farm size in hectares C Fruit trees 0.12 0.32 Takes value 1 if there are fruit trees in the farm and 0 otherwise D Berry bushes 0.10 0.31 Takes value 1 if there are berry bushes in the farm and 0 D otherwise Vineyards 0.20 0.39 Takes value 1 if there are vineyards in the farm and 0 otherwise D Livestock 0.18 0.38 Takes value 1 if there are livestock (cattle, sheep or goats) in D the farm and 0 otherwise Temporary workers 0.39 0.48 Takes value 1 if there are temporary workers in the farm and 0 D otherwise Weather information sources TV 0.69 0.46 Takes value 1 if the main source of weather information for the D farmer is TV and 0 otherwise Radio 0.14 0.34 Takes value 1 if the main source of weather information for the D farmer is radio and 0 otherwise Internet 0.08 0.28 Takes value 1 if the main source of weather information for the D farmer is Internet and 0 otherwise Peers 0.02 0.14 Takes value 1 if the main source of weather information for the D farmer are other farmers and 0 otherwise Location Linares 0.34 0.47 Takes value 1 if the farm is located in and 0 D otherwise Talca 0.26 0.44 Takes value 1 if the farm is located in and 0 D otherwise Limarí 0.40 0.48 Takes value 1 if the farm is located in Limarí Province and 0 D otherwise (omitted) a D is a binary variable and C is a continuous variable

sample are landowners. These characteristics are homo- Monte Patria, San Javier and Ovalle. Internet has a greater geneous for the municipalities in the study area. With regard presence in Villa Alegre and Colbún. to participation in organizations (neighborhood committees, Most producers (92.9 %) claimed that perceived changes water communities, agriculture associations, etc.), 45 % in climate in recent years, a situation that is homogeneous reported having been involved in organizations over the among the municipalities. A high portion expects climate past 5 years, but only 15 % reported to be actively involved changes in the future (93.5 %). 61.1 % of farmers inter- at present. General characteristics of farmers and farming viewed report having experienced at least one drought systems are described in Table 2. (Table 3), although only 31.9 % of the sample reported Regarding the main sources of weather information to having implemented a strategy to deal with this situation. farmers in the study area, television has the highest pro- The identified strategies were classified in four groups: portion (69.0 %), followed by radio (13.7 %), Internet investments related to water accumulation (9.8 %), moder- (8.3 %) and direct communication with other farmers (2.1 nization of irrigation systems (5.6 %), rationalization of %). Television is massively used in all municipalities, while water use at farm level (14.0 %), and partnership strategies radio is mostly used as a source of weather information in (3.0 %). Environmental Management

Table 3 Farmers who declare Municipality Farmers who Strategies employed (%) having experienced drought and declare faced strategies employed drought (%) Investment Irrigation Rationalization Partnership None technology of water use

Monte Patria 96.7 13.8 8.1 16.2 5.7 54.5 Ovalle 88.4 22.5 8.5 20.2 1.6 36.5 Punitaqui 91.7 58.3 8.3 0.0 0.0 25.0 San Javier 52.6 5.3 5.3 5.3 10.6 26.3 Villa Alegre 27.6 1.7 3.4 8.6 5.2 8.6 Yerbas Buenas 37.7 0.0 3.8 9.4 0.0 24.5 Colbún 35.9 1.3 5.1 11.5 1.3 16.7 San Clemente 48.2 5.9 1.2 17.6 0.0 23.5 Talca 40.0 2.9 5.7 8.6 0.0 22.9 San Rafael 70.6 11.8 5.9 23.5 5.9 23.5 Pelarco 54.5 0.0 9.1 9.1 18.2 18.0 Maule 19.2 0.0 0.0 11.5 0.0 7.7 Total 61.1 9.8 5.6 14.0 3.0 29.2

Stressors of Agricultural Production Additionally, we found three stressors related to the quality and access to natural resources (lack of land, In our study a stressor is defined as an element that can excessive use of chemicals and pesticides and pollution of represent a source of hazard for agricultural production. irrigation water). Access to natural resources, in terms of Results evidence the high importance assigned to climate- quantity and quality, is limiting the competitiveness of related stressors over others related to production, or eco- modern agriculture, required for the development of eco- nomic and institutional factors. Consistency in discourse for nomically and environmentally sustainable farming. each type of stressor was assessed with Cronbach’s α Exchange rate (CLP/dollar) and scarcity of labor are coefficient (α) to obtain acceptable values for the three types located as a less important category. Results support the of stressors. Table 4 contains the description of the notion that family farms would not be affected by these two assessment of stressors for agricultural production by variables. In the first case, the exchange rate directly affects producers. the commercial production for exportation and import The results show that the main stressors of farmers are substitution. In the second case the possibility of incor- related to the availability and management of water porating family labor to the production process is always a resources. This is consistent with studies indicating that more affordable option for family farmers. small-scale farmers have increasing restrictions on access to Higher values were observed for the climate related natural resources, including water (FAO 2012). Comoé stressors, while non-significant differences among farm et al. (2014) argue that climate variations (scarcity of rain, sizes were found for climate and productive factors. The strong winds, and high temperatures) are one of the main exception corresponds to the stressor of lack of labor, threats for farming activities. representing a problem with greater importance as farm size Other stressor identified as important is the uncertainty increases. caused by the high volatility of prices of agricultural pro- As indicated above, a stressor perceived differentially ducts, a fundamental aspect in the determination of net among farm sizes is the one related to access to effective income of the producer. In a context of lack of mechanisms institutions, which is a more important factor for small to assure prices for futures markets, strategies such as price producers than for large farms. This response is consistent insurance and associative marketing are difficult to apply with the notion of an increased complexity of agricultural for this segment of producers. services, as well with differences in access mechanisms of Another stressor highlighted by the producers corre- public and private institutions. Finally, although not statis- sponds to the absence of support services to meet the new tically significant, the price volatility factor is perceived as a demands and requirements of family farming, particularly major stressor for all farm sizes, but increasing with farm regarding to the insertion in value chains of family farmers size. These results are consistent with the high rate of par- and requirements that are derived from this scenario. ticipation of Chilean agricultural products in exportations. Environmental Management

Factors Affecting the Selection of a Strategy

The model presents a value of maximum likelihood func- 5.46 5.34 5.12 Lack of good institutions tion (MLF) of −473.9, a suitable value for this type of analyzes. The pseudo-R2 value, although low, is adequate (13 %) for socioeconomic analyses. The estimation results

a b b of the MNL model are detailed in Table 5.

0.64 With regard to demographic variables, respondents of Exchange rate

= female gender are more likely to modernize irrigation and α s

’ infrastructure, as well to take actions of partnership. The largest producers are less likely to implement associative actions; this can be explained due that the first Price volatility action for large farms is related to the implementation of investments. Additionally, the hiring of temporary work- ers is a variable that could indicate more intensive and market oriented agriculture, which impacts positively on irrigation technology and water rationalization actions at farm level.

and chemical Our study shows that the use of credit by producers influences positively the implementation of irrigation tech- nology. On the other hand, farmers involved in vineyard production make actions of partnerships; it is expected that such producers are associated locally and this has an effect opinion and farm size ’ over information diffusion processes. A similar situation Lack of land Use of pesticides occurs for livestock producers (cattle, goats, and sheep), since maintain livestock has a positive effect on investment in infrastructure and rationalization of water use, as well as

0.75the Cronbach actions of associativity. 4.945.15 5.075.46 5.15 4.93 5.01 5.31 4.71 5.39 5.86 4.17 5.87 5.39 5.19 Water pollution =

α Regarding access to weather information on television, it s ’ has a positive association with the implementation of infrastructure and the actions of association. Internet also a a,b c influences positively both outcomes, as well actions of Lack of labor water rationing. In relation to the location variables, it can be noted that farmers in the Province of Limarí are more likely to implement infrastructure investments than those in Linares scarcity or Talca (in Maule). The results of the estimates of the marginal effects are contained in Table 6, these coefficients represent expected changes in probabilities. Analyzing the marginal effects of 0.71 Cronbach Drought Water = demographic variables, it may be noted that a female pro- α s

’ ducer has an 11.6 % lower probability of not having a strategy in contrast to a male producer. For each year of age, it is expected that the probability of using associative stra- 5.94 5.88 5.90 5.13 Climate relatedCronbach Productive factors related Institutional and economics Climate change tegies decreases by 0.1 %. Analyzing the results for production variables, it is possible to note that for each additional hectare of farm size, the likelihood of implementing infrastructure actions increases by 0.2 %, while for the actions of irrigation modernization this value increases by 0.1 %. Production of Average importance of agricultural production stressors (scale from 1 to 7) according to farmers vineyards on the farm increases by 3 % the probability of indicate different groups at 5 % implementing actions of association for managing water Medium (between 12 ha and less than 30 ha) Large (30 ha or more)Total 6.10 6.04 6.23 6.07 6.12 5,42 6.05 4.68 5.03 5.07 5.04 5.54 4.51 5.40 Farm size Stressor type Table 4 a, b, c Small (less than 12 ha) 6.33 6.21 6.08resources. 4.44 Livestock ownership increases the likelihood of Environmental Management

Table 5 Estimation of the MNL Variables Strategies Investment (1) Irrigation technology (2) Rationalization of Partnership (4) water use (3) Coefficient P Coefficient P Coefficient P Coefficient P

Demographics Gender 0.591* 0.083 0.774* 0.080 0.311 0.317 1.016* 0.055 Age −0.014 0.226 0.004 0.809 −0.006 0.614 −0.051** 0.045 Education 0.192 0.676 0.082 0.122 0.055 0.170 −0.086 0.266 Participation in organizations 0.404 0.211 0.304 0.469 0.121 0.655 −0.960 0.115 Use of credit 0.249 0.504 0.769* 0.075 0.033 0.915 0.519 0.354 Farm system Farm size 0.020*** 0.006 0.021*** 0.002 0.006 0.434 −0.024 0.264 Fruit trees 0.316 0.455 −0.252 0.687 0.359 0.389 0.712 0.438 Berry bushes 0.700 0.461 0.057 0.947 −0.953 0.190 −0.967 0.441 Vineyards 0.335 0.378 0.467 0.331 0.094 0.790 1.981*** 0.004 Livestock 0.988** 0.029 0.617 0.364 1.060*** 0.006 1.791** 0.029 Temporary workers −0.141 0.685 0.921** 0.036 0.517* 0.071 0.421 0.440 Source of weather information TV 1.840* 0.066 1.210 0.263 0.114 0.850 14.403*** 0.000 Radio 1.525 0.156 0.524 0.669 0.541 0.402 12.855*** 0.000 Internet 2.821** 0.026 2.322* 0.062 1.398* 0.095 14.824*** 0.000 Peers −12.910*** 0.000 −13.118*** 0.000 −0.816 0.501 0.462 0.629 Location Linares −2.634*** 0.001 −0.663 0.244 0.085 0.932 1.181 0.130 Talca −1.307* 0.010 −1.076 0.127 0.356 0.338 0.187 0.827 Constant −2.588* 0.070 −4.911*** 0.004 −1.825* 0.085 −14.520*** 0.000 Base outcome None (0) Pseudo R2 0.13 N 406 MLF −473.982 Estimations using STATA SE 10.1, command mlogit *P < 0.1; **P < 0.05; ***P < 0.01 taking actions to rationalize water use by 16.5 %, while this Discussion condition reduces the probability of not having a strategy in 24.6 %. On the other hand, the hiring of temporary workers Availability and management of water resources appear as a increases the likelihood of implementing rationalization main stressor for farmers in the study area. At the national strategies by 5 %, and reduces the probability of not having level, water shortages are being addressed by the CNR a strategy by 10.4 %. (Comisioń Nacional de Riego or National Commission for Additionally, the use of meteorological information Irrigation) and INDAP (Instituto de Desarrollo Agrope- reduces the probability of not having a strategy by 35.4 % cuario or National Agency for Agricultural Development). for television, 59.6 % for radio, and 59.2 % for Internet; it The CNR allocates annually around US$125 million for also increases the likelihood of implementing actions of investments in irrigation infrastructure at the national level, association. Conversely, the probability of not having a principally targeted for larger commercial farms. INDAP is strategy is increased by 28.2 % when the information comes implementing an irrigation program for peasants with a from other farmers. These results reveal the importance of budget of approximately US$22 million per year. The information quality as an important input of decision- objective of those programs is to promote the productivity making when farmers face water shortages. and the competitiveness of farmers through the Environmental Management

Table 6 Marginal effects of the MNL estimated Variables Strategies Investment (1) Irrigation technology Rationalization of Partnership (4) None (0) (2) water use (3) Coefficient P Coefficient P Coefficient P Coefficient P Coefficient P

Demographics Gender 0.043 0.288 0.040 1.193 0.024 0.670 0.009 0.203 −0.116* 0.062 Age −0.001 0.257 0.001 0.625 −0.001 0.741 −0.001* 0.060 0.002 0.446 Education −0.001 0.969 0.004 0.181 0.009 0.225 −0.001 0.167 −0.012 0.146 Participation in 0.034 0.237 0.138 0.550 0.009 0.849 −0.009** 0.047 −0.047 0.390 organizations Use of credit 0.017 0.637 0.049 0.113 −0.017 0.761 0.004 0.472 −0.053 0.383 Farm system Farm size 0.002*** 0.006 0.001*** 0.001 0.001 0.794 −0.001 0.173 −0.003 0.109 Fruit trees 0.021 0.619 −0.021 0.430 0.064 0.439 0.006 0.585 −0.070 0.405 Berry bushes 0.117 0.434 0.008 0.878 −0.159** 0.033 −0.006 0.337 0.039 0.788 Vineyards 0.023 0.538 0.023 0.465 −0.008 0.896 0.030* 0.068 −0.068 0.321 Livestock 0.057 0.267 0.006 0.885 0.165** 0.043 0.018 0.312 −0.246*** 0.001 Temporary workers −0.034 0.268 0.049* 0.064 0.087 0.104 0.002 0.678 −0.104* 0.072 Source of weather information TV 0.066 0.213 0.017 0.648 −0.136 0.188 0.407*** 0.000 −0.354*** 0.005 Radio −0.087*** 0.000 −0.061*** 0.001 −0.251*** 0.000 0.995*** 0.000 −0.596*** 0.000 Internet −0.095*** 0.000 −0.057*** 0.000 −0.252*** 0.000 0.996*** 0.000 −0.592*** 0.000 Peers −0.137*** 0.000 −0.081*** 0.000 −0.079 0.643 0.012 0.502 0.285* 0.100 Location Linares −0.156*** 0.000 −0.026 0.277 0.068 0.381 0.019 0.232 0.095 0.247 Talca −0.096*** 0.000 −0.048** 0.026 0.117 0.128 0.002 0.796 0.025 0.758 *P < 0.1; **P < 0.05; ***P < 0.01 incorporation of improved irrigation technologies that bring probably be more costly and less relevant to the needs (FAO new areas under irrigation and enhance the management of 2012). Existing support services to producers are char- water resources (Roco et al. 2015). acterized by their high specialization, lack of moderniza- Additionally, an important point regarding stressors for tion, and low adaptation to current requirements of family agricultural production is the fact that our results differ from farming systems, with a clear focus, in some cases, on the analysis of the research by Toledo et al. (2011), who commercial agriculture. detected that in the year 2006, the most relevant risk factors The results of the stressor related to lack of labor are for Chilean agricultural producers were price and cost consistent with the idea of a low demand of employment volatility; climate change appeared as one of the lowest and a dominance of self-employment in family farming source of risk. Thus it is possible to argue that the factors systems, and with higher levels of employment required by associated with climate have become important for produ- commercial farms. Furthermore, with regard to economic cers, given recent experiences of drought and lack of water. and institutional stressors, it can be noted that the exchange The value assigned to the stressor related to the lack of rate is especially relevant for medium and large farms. For institutions reflects the need of provide support services in a this reason, it is important to establish mechanisms to comprehensive manner for producers and in accordance ensure against significant declines of this monetary variable. with their particular insertion in value chains allowing the Small-scale farmers experience difficulties with global access and reducing the transaction costs of small-scale markets (Torres et al. 2015) and public institutions plays a farming. For example, if a producer must apply for different key role in improving access to existing innovation. The services and public institutions to access credit, productive low level of association or existing cooperation in agri- technical assistance, plant and animal health programs, culture could aggravate this situation, so this condition market information and marketing support, the result will should be changed to address this stressor and, in Environmental Management consequence allowing family farmers to have a more of investments or irrigation improvements having peers as complete perception of productive problems in order to main source of weather information; this situation can be promote adaptation to climate change (Roco et al. 2015). explained due to this being an option for small-scale farmers In relation to the analyses of results derived to the model, who cannot access more efficient information media. gender appears as an important element for drought adap- Farmers in Limarí river basin are more likely to imple- tation. Basu et al. (2015) claimed that women are likely to ment investment for water management, which could be implement several strategies in the face of increasing explained by the more restrictive conditions regarding water uncertainties in order to meet water demand. In this sense, resources of this area, as is documented by Vicuna et al. Bhattarai et al. (2015) argue that gender is a key variable in (2014). Additionally, it is necessary to mention that farmers agrobiodiversity adaptation processes, especially for agrar- in Talca are more likely to implement infrastructure actions ian contexts where climate change is exacerbated. than famers form Linares, this fact can be explained given Our results show that farm size is an important element that Talca corresponds to the Province of the Regional for the implementation of investments. Farm size reflects capital city, and it is expectable that this fact can affect the the financial capital of the producers; for this reason, it is technology diffusion and the implementation of actions. expected that market-oriented farms are able to afford Farmers located near to markets tend to have more access to higher payments demanded by infrastructure improvements. information and extension services from public and private These outlays are related to investment in infrastructure and sources. In the same way, Alam (2015) argue that high modernization of irrigation systems. In this sense, Mussetta institutional access is an important element for crop diver- and Barrientos (2015) argue that the diversity of strategies sification among farmers in Bangladesh. to cope with lack of water is correlated with the financial In this sense, the role of information and training is capacity of producers. In this sense, Gebrehiwot and van der highlighted by Gebrehiwot and van der Veen (2015) who Veen (2015) found that perceived costs of drought risk- argue that public drought risk management strategies aim- reduction measures are a key element for their imple- ing at stimulating farmers’ adaptive capacity should raise mentation at farm-level. awareness on the consequences of drought risk, as well as to Credit access appears as a determinant for irrigation provide information and training addressing the effective- improvement. This result is consistent with the findings of ness of available farm-level adaptation measures. Farmers Zorom et al. (2013), who found that credit access is an utilize their own experience, resources and ascribed infor- important element to implement diversification and adap- mation from their own level and also from the organiza- tation strategies to climate variability by farmers. Similarly, tional level to cope with drought (Habiba et al. 2012). Reyes and Lensink (2011) evidence credit constraints among market-oriented farmers in Chile and the importance of improving access to financial capital. Conclusions According to our results, farmers involved in the produc- tion of some products (vineyards and livestock), generally are From the results, the growing importance of climatic factors involved in agricultural associations and are more likely to for Chilean agricultural production in recent years was evi- implement rationalization actions. Similarly, Giuliani (2013) dent. Climate change, drought and water shortage become shows that one of the elements of success for regional wine relevant in the opinion of the producers of both basins. The production clusters is that they facilitate the formation of local relation of dependence on weather, and consequently of inter-organizational networks, which act as conduits of water resources in agriculture is verified, drought events are knowledge and innovation. Teshome et al. (2015); Eriksen increasing uncertainty in a frame of climate variability. et al. (2005), and Finan and Nelson (2001) highlight the Producers that adopted a strategy for the management of importance of social networking to implement conservation of their water resources given the occurrence of droughts are soil and water and increase adaptive capacity to face droughts. female; young producers, in the case of partnership actions; Producers who present access to weather information are and producers who use credit in the case of improvement of more likely to implement investment actions. This result is irrigation systems. Additionally, farms with greater exten- consistent with the research of Roco et al. (2014) who sion are more likely to invest and modernize their irrigation claimed that access to weather information is a key element systems. This point is necessary to take into account that the for climate change adaptation among farmers in Chile. In farm size is closer to the level of wealth of the producer and the same way, Bojovic et al. (2015) argue that online par- both actions, infrastructure and modernization, are plaintiffs ticipation of farmers can improve the decision-making in investment. For farms with livestock, the implementation processes for climate change adaptation, based on results of actions on infrastructure, rationalization, and associativ- for Italian agricultural producers. It is important to note the ity appear as necessary; livestock means a greater need for existence of a negative relation between the implementation adaptive management at farm level. Environmental Management

Weather information seems relevant for all actions of Adger WN, Arnett N, Tompkins E (2005) Successful adaptation, especially in the case of associativity. This is adaptation to climate change across scales. Glob Environ Chang 15:77–86 especially relevant for Internet users when they implement Alam K (2015) Farmers’ adaptation to water scarcity in drought-prone strategies for rationalization of the use of water and environment: a case study of Rajshahi Disctrict, Bangladesh. Agr improvement of irrigation. The model also shows that Water Manag 148:196–206 producers located in the Limarí province are much more Ancapichun S, Garces-Vargas J (2015) Variability of the southeast pacific subtropical anticyclone and its impact on sea surface likely to carry out actions of infrastructure given the more temperature off north-central Chile. Cienc Mar 41(1):1–20 restrictive conditions in that area regarding the availability Arumí JL, Rivera D, Holzapfel E, Muñoz E (2013) Effect of drought and use of water resources. on groundwater in a Chilean irrigated valley. Water Manag – Kiparsky et al. (2012) argue that adaptation to climate in 166:231 241 Basu M, Hoshino S, Hashimoto S (2015) Many issues, limited the water sector is driven not only by knowledge and responses: coping with water insecurity in rural India. Water Res information on projected changes but also by mediating and Rural Dev 5:47–63 factors, such as politics, culture, and institutions. Previous Berger T, Troost C (2014) Agent-based modelling of climate adapta- literature points out the importance of networks and social tion and mitigation options in agriculture. J Agric Econ 65(2):323–348 capital that family farmers have to innovate in strategies for Bhattarai B, Beilin R, Ford R (2015) Gender, agrobiodiversity, and natural resources management (Jara-Rojas et al. 2012, climate change: a study of adaptation practices in the Nepal Knowler and Bradshaw 2007, Sidibé 2005); it is interesting Himalayas. World Dev 70:122–132 to understand how producers are included in these networks Bojovic D, Bonzaningo L, Giupponi C, Maziotis A (2015) Online participation in climate change adaptation: a case study of agri- in relation to demographic and productive characteristics. cultural adaptation measures in Northern Italy. J Environ Manag Finally, this study contributes to the identification of 157:8–19 elements for public policies to promote the development of Brechét T, Hritonenko N, Yatsenko Y (2013) Adaptation and miti- a climate-smart agriculture. Actions that contribute to face gation in long-term climate policy. Environ Resour Econ 55: 217–243 the constraints arising from water scarcity were described. Chilean Ministry of Public Works (2012) Estrategia Nacional de Stressors for agricultural production were identified in order Recursos Hídricos 2012–2025. http://www.mop.cl/documents/ to understand the capacity of action and governance of the ENRh_2013_OK.pdf. Accessed Dec 2014 different groups of agricultural producers to improve their Clarvis MH, Allan A (2014) Adaptive capacity in a Chilean context: a questionable model for Latin America. Environ Sci Policy capacity of adaptation. 43:78–90 Additional research is needed, as is recognized by Comoé H, Finger R, Barjolle D (2014) Farm management decision and Pahl-Wostl (2009): governance failures are at the origin of response to climate variability and change in Côte d’Ivore. Mitig – many resource management problems, climate change and Adapt Strategies Glob Chang 19:123 142 Comoé H, Siegrist M (2015) Relevant drivers of farmers’ decisions the concomitant increase of extreme weather events has behavior regarding their adaptation to climate change: a case exposed the inability of current governance regimes to deal study of two regions in Côte d’Ivore. Mitig Adapt Strategies Glob with present and future challenges. Chang 20:179–199 Davis GW, Richadson DM, Keeley JE and Hobbs RJ (1996) Chapter 7. Mediterranean-type ecosystems: the influence of biodiversity Acknowledgments This work was developed under the project on their functioning. In: Mooney HA, Cushman JH, Medina E, “Support system for decision making to reduce vulnerability in irri- Sala OE and Schulze ED (eds) 1996. Functional roles of biodi- gated agriculture facing climate change and variability”, code N° versity: a global perspective, John Wiley & Sons Ltd, New York, – D10i1051, founded by FONDEF Chile and implemented by Centro USA, pp 151 183 fi Interdisciplinario de Cambio Global UC. The authors thank the DGA (2004a) Diagnóstico y clasi cación de los cursos y cuerpos de farmers who answered the survey. The authors also thank the feedback agua según objetivos de calidad, Cuenca del Río Limarí. Chilean obtained from Mrs. Melanie Oertel, The LACEEP 10+ Workshop Ministry of Public Works. p 130 fi (México, July 2015) and The Workshop on Water Scarcity of Exceed- DGA (2004b) Diagnóstico y clasi cación de los cursos y cuerpos de Swindon Program (Argentine, September 2015). agua según objetivos de calidad, Cuenca del Río Maule. Chilean Ministry of Public Works. p 145 Deressa TT, Hassan RM, Ringler C, Alemu T, Yesuf M (2009) Determinants of farmers’ choice of adaptation methods to climate Compliance with Ethical Standards change in the Nile Basin of Ethiopia. Glob Environ Chang 19:248–255 Conflict of interest The authors declare that they have no com- Di Falco S, Veronesi M, Yesuf M (2011) Does adaptation to climate peting interests. change provide food security? A micro-perspective from Ethio- pia. Am J Agric Econ 93(3):829–846 Eriksen SH, Brown K, Kelly PM (2005) The dynamics of vulner- References ability: locating coping strategies in Kenya and Tanzania. Geogr J 171:287–305 Aceituno P (1988) On the functioning of the Southern Oscillation in FAO (2012) Marco estratégico de mediano plazo de cooperación de la the South American sector. Part I. Surface climate. Mon Weather FAO en agricultura familiar en América Latina y el Caribe 2012- Rev 116:505–524 2015. Rome. p 45 Environmental Management

FAO (2013) Climate smart agriculture sourcebook. Rome. p 557 Miller A (1976) The climate of Chile. In: Schwerdtfeger W (ed) World Finan TJ, Nelson DR (2001) Making rain, making roads, making do: survey of climatology: Climates of central and South America, public and private adaptations to drought in Ceará, northeast Elsevier, Amsterdam. Brazil. Clim Res 19:97–108 Misra AK (2013) Climate change impacts, mitigation and adaptation García de Jalón S, Iglesias A, Cunningham R, Perez Díaz JI (2014) strategies for agricultural and water resources, in Ganga Plain Building resilience to water scarcity in southern Spain: a case (India). Mitig Adapt Strategies Glob Chang 18:673–689 study of farming in Doñana protected wetlands. Reg Environ Mussetta P, Barrientos MJ (2015) Vulnerabilidad de productores rur- Change 14:1229–1242 ales de Mendoza ante el cambio ambiental global: clima, agua, Gbetibouo GA. 2009. Understanding farmers’ perceptions and economía y sociedad. Revista de la Facultad de Ciencias Agrarias adaptations to climate change and variability: the case of the —Universidad Nacional del Cuyo 47(2):145–170 Limpopo Basin, South Africa. IFPRI discussion paper No. 849. Nelson DR, Adger WN, Brown K (2007) Adaptation to environmental pp 36 change: contributions of a resilience framework. Annu Rev Gebrehiwot T, van der Veen A (2013) Farm level adaptation to climate Environ Resou 32:395–419 change: the case of farmer’s in the Ethiopian Highlands. Environ OECD (2014) OECD Rural Policy Reviews: Chile 2014. OECD Manag 52:29–44 Publishing, Paris, France, pp 210 Gebrehiwot T, van der Veen A (2015) Farmers prone to drought risk: Pahl-Wostl C (2009) A conceptual framework for analyzing adaptive why some farmers undertake farm-level risk-reduction measures capacity and multi-level learning processes in resource govern- while others not?. Environ Manag 55:588–602 ance regimes. Glob Environ Chang 19:354–365 Giuliani E (2013) Network dynamics in regional clusters: evidence Pereira LS, Cordery I, Iacovides I (2012) Imporved indicators of water from Chile. Res Policy 42(8):1406–1419 use performance and productivity for sustainable water con- Greene WH (2008) Econometric analysis, 7th Edn. Prentice Hall, servation and saving. Agr Water Manage 108:39–51 Upper Saddle River Pereira LS, Paredes P, Cholpankulov ED, Inchenkova OP, Teodoro Habiba U, Shaw R, Takeuchi Y (2012) Farmers perception PR, Horst MG (2009) Irrigation scheduling strategies for cotton and adaptation practices to cope with drought: perspectives to cope with water scarcity in the Fergana Valley, Central Asia. from Northwestern Bangladesh. Int J Disaster Risk Red 1: Agr Water Manage 96:723–735 72–84 Rahm M, Huffmann W (1984) The adoption of reduced tillage: the role Hageback J, Sundberg J, Ostwald M, Chen D, Yun X, Knutsson P of human capital and other variables. Am J Agric Econ 66:405–413 (2005) Climate variability and land-use change in Danangou Regional Government of Coquimbo (2013) Diagnostico del Plan Watershed, China: examples of small-scale farmers’ adaptation. Maestro para la Gestión de Recursos Hídricos, Región de Clim Change 72:189–212 Coquimbo. Executive Summary IPCC (2007) Climate change 2007: impacts, adaptation and vulner- Reyes A, Lensink R (2011) The credit constraints of market-oriented ability. Intergovernmental Panel on Climate Change Fourth farmers in Chile. J Dev Stud 47(12):1851–1868 Assessment Report. Geneve, Switzerland Roco L, Engler A, Bravo-Ureta B, Jara-Rojas R (2014) Farm level IPCC (2014) Climate change 2014: impacts, adaptation and vulner- adaptation decisions to face climatic change and variability: ability. Intergovernmental Panel on Climate Change Fifth evidence from central Chile. Environ Sci Policy 44:86–96 Assessment Report. Geneve, Switzerland Roco L, Engler A, Bravo-Ureta B, Jara-Rojas R (2015) Farmers’ INE (2010) Cifras Censo 2007. http://icet.odepa.cl/exp/ficha.php. perception of climate change in Mediterranean Chile. Reg Accessed Dec 2014. Environ Change 15(5):867–879 Lindoso DP, Rocha JD, Debortoli N, Parente I, Eiro F, Bursztyn M, Sidibé A (2005) Farm-level adoption of soil and water conservation Rodrigues-Filho S (2014) Integrated assessment of smallholder techniques in northern Burkina Faso. Agric Water Manag farming’s vulnerability to drought in the Brazilian Semi-arid: a 71:211–224 case study in Ceara.́ Clim Chang 127:93–105 Silvestri S, Bryan E, Ringler C, Herrero M, Okoba B (2012) Climate Jara-Rojas R, Bravo-Ureta BE, Díaz J (2012) Adoption of water change perception and adaptation of agro-pastoral communities in conservation practices: a socioeconomic analysis of small-scale Kenya. Reg Environ Change 12:791–802 farmers in Central Chile. Agr Syst 110:54–62 Sofoluwe N, Tijane A, Baruwa O (2011) Farmers’ perception and Kiparsky M, Milman A, Vicuña S (2012) Climate and water: knowl- adaptation to climate change in Osun State, Nigeria. Afr J Agric edge of impacts to action on adaptation. Annu Rev Environ Resour Econ 6(20):4789–4794 Resour 37:163–194 Tambo JA, Abdoulaye T (2012) Climate change and agricultural Knowler D, Bradshaw B (2007) Farmers’ adoption of conservation technology adoption: the case of drought tolerant in rural agriculture: a review and synthesis of recent research. Food Nigeria. Mitig Adapt Strategies Glob Chang 17:277–292 Policy 32:25–48 Teshome A, de Graaff J and Kassie M. (2015) Household-level Maddala GS (ed.) (1987) Limited-dependent and qualitative variables determinants of soil and water conservation adoption phases: in econometrics. In: Econometric society monographs. Cam- evidence from North-Western ethiopian highlands. Environ bridge University Press, Cambridge Manage. doi: 10.1007/s00267-015-0635-5 Manandhar S, Vogt DS, Perret SR, Kazama F (2011) Adapting Toledo R, Engler A, Ahumada V (2011) Evaluation of risk factors in cropping systems to climate change in Nepal: a cross-regional agriculture: an application of the analytical hierarchical process study of farmers’ perception and practices. Reg Environ Change (AHP) methodology. Chil J Agr Res 71(1):114–121 11(2):335–348 Torres R, Azócar G, Rojas J, Montecinos A, Paredes P (2015) Vul- Maraseni TN, Mushtaq S, Reardon-Smith K (2012) Climate change, nerability and resistance to neoliberal environmental changes: an water security and the need for integrated policy development: assessment of agriculture and forestry in the Biobio region of the case of on-farm infrastructure investment in the Australian Chile (1974–2014). Geoforum 60:107–122 irrigation sector. Environ Res Lett 7(3):034006 Urrutia RB, Lara A, Villalba R, Christie DA, Le Quesne C, Cuq A Mata L, Budhooram J (2007) Complementarity between mitigation (2011) Multicentury tree ring reconstruction of annual streamflow and adaptation: the water sector. Mitig Adapt Strategies Glob for the Maule river watershed in south central Chile. Water Chang 12:799–807 Resour Res 47(6):W06527 Environmental Management

Valdés-Pineda R, Pizarro R, García-Chevesich P et al. (2014) Water Wang X, Zhang J, Shahid S, Guang E, Wu Y, Gao J and He R. (2014) governance in Chile: availability, management and climate Adaptation to climate change impacts on water demand. Mitig change. J Hydrol 519:2538–2567 Adapt Strategies Glob Chang. doi: 10.1007/s11027-014-9571-6 Vicuña S, Garreaud RD, McPhee J (2011) Climate change impacts on Zilberman D, Zhao J, Heiman A (2012) Adoption versus adaptation, the hydrology of a snowmelt driven basin in semiarid Chile. Clim with emphasis on climate change. Annu Rev Resour Econ 4:27–53 Chang 105(3-4):469–488 Zorom M, Barbier B, Mertz O, Servat E (2013) Diversification and Vicuna S, Alvarez P, Melo O, Dale L, Meza F (2014) Irrigation adaptation strategies to climate variability: a farm typology for infrastructure development in the Limarí basin in Central Chile: the Sahel. Agr Syst 116:7–15 implications for adaptation to climate variability and climate change. Water Int 39(5):620–634