Farmers’ adaptive measures to climate change induced natural shocks through past climate experiences in the Mekong River Delta, Vietnam

This study examined farmers’ adaptive measures to climate change induced natural shocks through past climate experiences in the Mekong River Delta (Vietnam) from a data set of 330 farmers. Seemingly unrelated regression model was used to identify the determinants of farmers’ adaptive measures. Results showed that male household head, education of the household head, marital status of the household head, production assets, firm size, availability of credit, access to market, temperature and rainfall had significant impacts on choices of adaptation. Results also indicated that past climate experiences was the most important determinant of adaptive measures. Policy messages enhanced access to credit, to markets, and created awareness on climate change. Other policy options could also be suggested, including: strengthening education level of farmers, facilitating cheap technologies, spurring irrigation investment through public - private partner, and supporting the land reform such as farmers’ cooperation in large-scale production.


INTRODUCTION
The Mekong River Delta, the major agricultural region of Vietnam, is identified as significantly vulnerable to climate change (Yusuf and Francisco, 2010;Asian Disaster Preparedness Center, 2003). Agricultural production remains the main source of livelihoods for most farmers in this area (Nguyen and Le, 2012;Le Anh et al., 2014;Asian Disaster Preparedness Center, 2003). Climate change has greater negative impacts on farm households as they have the lowest capacity to adapt to changes in climatic conditions (Yu et al., 2010;Asian Disaster Preparedness Center, 2003).
Adaptation measures are therefore important to help farmers to better face extreme weather conditions and associated climatic variations (Valipour et al., 2015;Adger et al., 2003;Kandlinkar and Risbey, 2000).
A better understanding of current adaptation measures and their determinants will be important to inform policy for future successful adaptation. Some related studies  937105567.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License conducted in the last few years focused on farmers' past climate experiences. Niles et al. (2015a), using farmer survey data from New Zealand, showed that limiting factors mediated the effect of past climate experiences on the adoption of adaptation strategies differently in two regions with water acting as a limiting factor in Hawke's Bay and water and temperature as a limiting factor in Marlborough. Le Dang et al. (2014) addressed the limited understanding of how rice farmers appraise their private adaptive measures and influential factors in the Mekong River Delta of Vietnam. Authors found that belief in climate change, information and objective resources were found to influence farmers' adaptation assessments. Geoff (2014) also stressed that farmers' climate change beliefs affect adaptation to climate change. Nicholas and Gina (2012) explored commercial farmers' perceptions of and responses to shifting climates in the Little Brak River area along South Africa's south coast and found that farmers' experience with shifting climates has played a large part in driving their adaptive decision-making. This study adds to these analyses by evaluating farmers' past climate experiences and distinguishing determinants of adaptation to climate change induced natural shocks among farmers in the Mekong River Delta. Evaluating past climate experiences of and response to climate change includes exploring what these perceptions are, how they are formed and how perception affects response (Bryant et al., 2000;Vedwan and Rhoades, 2001).

Research site and data
Long An, Ben Tre, Can Tho, Soc Trang, Kien Giang, and Ca Mau are the six provinces randomly selected from 13 provinces in the Mekong Delta which are defined at different agro-ecological systems which enabled representation of the Mekong Delta region. One district from each province and two communes from each district were randomly chosen. In total, there are 12 communes and commune centres in the survey. From the official household lists of the twelve communes, farm households were selected by simple random sampling. The face-to-face structured interviews were conducted in July, 2014. Four teams of 10 interviewers each had been involved in two intensive training sections, one before and one after the pre-test. The interviewers visited 335, but interviewed 330 farm households, 50 in each commune. Each interview was around two hours in duration. In this study, the farm household was the unit of analysis and the household heads or their spouses were the interviewees.
The structured questionnaire mainly covers perception of past climate change, climate change adaptation assessment, and a number of influential factors. The questionnaire was refined and finalized based on the information from three focus group discussions in provinces such as Long An, Ben Tre, Can Tho and of the questions were also tested through the pre-tests with 30 randomly chosen farm households in Ben Tre Province. The data used in this paper are specified from questions about climate change, adaptation assessment, farm characteristics, income, assets, infrastructure and institutional factors. The principal systems in the survey areas are mono-rice, shrimpmixed rice, fish, shrimp, cereal-root crop mixed, and fruit. Out of 330 farm households, 32% of them cultivate rice, 1% of them shrimp-mixed rice, 2.5% of them fish, 11.3% of them shrimp, 5.5% of them annual crops, and 10.5% of them fruit. Farmers have average land size of about 1.1 ha. About 70% of farming housholds have less than 1.2 ha. Average size of rice land, land for annual crops and land for fruit are 2.5, 2.4, and 1.9 ha, respectively. Average shrimp-mixed rice land size is 4.3 ha. Paddy yield in the study area is around 6.09 tones/ha. Average income of farmers in the study area is about 76.348 thousand Vietnam Dong (about 3.455 USD). Several Mekong River Delta related studies and reports by the (United Nations Development Programme (UNDP), 2008;EU/MWH, 2006;ADPC/GTZ, 2003) clarify the trends of climate change in terms of higher temperature, salt water intrusion, eroded shorelines, exacerbated coastal flooding, rainfall increasingly concentrated over fewer months in the rainy season, while the dry season will be more prolonged. This will lead to more frequent and intense floods and droughts simultaneously. In addition, tropical cyclone and typhoon occurrence are expected to alter and become more intense under a warmer climate as a result of higher sea-surface temperatures.

Model specification
Common approach uses a univariate technique such as probit/logit analysis for discrete choice dependent variables to model each of the adaptation measures individually as functions of the common set of explanatory variables. The shortfall of this approach is that it is prone to biases caused by ignoring common factors that might be unobserved and unmeasured and affect the different adaptation measures. In addition, independent estimation of individual discrete choice models fails to take into account the relationships between adoptions of different adaptation measures. Farmers might consider some combinations of adaptation measures as complementary and others as competing. By neglecting these common factors the univariate technique ignores potential correlations among the unobserved disturbances in adaptation measures, and this may lead to statistical bias and inefficiency in the estimates (Lin et al., 2005;Belderbos et al., 2004;Golob and Regan, 2002).
A multinomial (MNL) discrete choice model is another alternative to the multivariate model with more than two endogenous discrete choice variables. In the multinomial discrete choice model the choice set is made up of all combinations of adaptation measures. The shortfall of this technique is that interpretation of the influence of the explanatory variables on choices of each of the original separate adaptation measures is very difficult. The shortfall of this technique is that all multinomial replications of a multivariate choice system have problems in interpreting the influence of explanatory variables on the original separate adaptation measures (Golob and Regan, 2002).
This study follows Zellner's Iterative Seemingly Unrelated Regressions (ISUR) to overcome the shortfalls of using the univariate and multinomial discrete choice techniques. The ISUR technique provides parameter estimates that converge to unique maximum likelihood parameter estimates. The resulting model has stimulated countless theoretical and empirical results in econometrics and other areas (Zellner, 1962;Srivastava and Giles, 1987). The benefit of this model is that the ISUR estimators utilize the information present in the cross regression (or equations) error correlation and hence it is more efficient than other estimation methods such as the univariate and multinomial discrete choice techniques.
The methodology used in this paper for misspecification testing follows Godfrey (1988) and Shukur (2002). White (1980) test was used (including cross products of the explanatory variables) to test for heteroscedasticity and Ramsey's (1969) RESET test to test for functional misspecification (Ramsey, 1969)   Change irrigation schedule 0.07 13 Change crop rotation 0.07

E No adaptation (n) 43
Source: Climate change survey in the Mekong River Delta (2014).
following formula: Where w is the wealth index, b is the weights from PCA, a is the asset value, x is the mean asset value, and s is the standard deviation of the assets.

Dependent variables
Intergovernmental Panel on Climate Change (IPCC's) report, which is based on the social sciences, states that "adaptation refers both to the process of adapting and to the condition of being adapted" (Smit and Pilifosova, 2001). Specifically, adaptation is the process of improving society's ability to cope with changes in climatic conditions across time scales, from short term (for example, seasonal to annual) to the long term (for example, decades to centuries). TheIPCC (2007) defines adaptive capacity as the ability of a system to adjust to climate change (including climate variability and extremes), to moderate potential damages, to take advantage of opportunities, or to cope with the consequences. The goal of an adaptation measure should be to increase the capacity of a system to survive external shocks or change. Important adaptation options in the agricultural sector include: crop diversification, mixed crop-livestock farming systems, using different crop varieties, changing planting and harvesting dates, mixing less productive, drought-resistant varieties and high-yield water sensitive crops (Adger et al., 2003;Bradshaw et al., 2004). Agricultural adaptation involves two types of modifications in production systems. The first is increased diversification that involves engaging in production activities that are drought tolerant and or resistant to temperature stresses as well as activities that make efficient use and take full advantage of the prevailing water and temperature conditions, among other factors. Crop diversification can serve as insurance against rainfall variability as different crops are affected differently by climate events (Orindi and Eriksen, 2005;Adger et al., 2003). The second strategy focuses on crop management practices geared towards ensuring that critical crop growth stages do not coincide with very harsh climatic conditions such as mid-season droughts. Crop management practices that can be used include modifying the length of the growing period and changing planting and harvesting dates (Orindi and Eriksen, 2005). Use of irrigation has the potential to improve agricultural productivity through supplementing rainwater during dry spells and lengthening the growing season (Baethgen et al., 2003;Orindi and Eriksen, 2005). It is important to note that irrigation water is also subject to impacts from climate change. Use of irrigation technologies need to be accompanied by other crop management practices such as use of crops that can use water more efficiently. Important management practices that can be used include: efficient management of irrigation systems, growing crops that require less water, and optimizing of irrigation scheduling and other management techniques that help reduce wastage (Loё et al., 2001).
The adaptation measures in this study are based on asking farmers about their perceptions of past climate change induced natural shocks and the actions they take to counteract the negative impacts of these shocks. We follow Maddison (2006) and Hassan and Nhemachena (2008) by assuming that farmers' adaptation actions are driven by climatic factors.
A list of private adaptive measures to climate change was initially developed from the literature (Bradshaw et al., 2004;Bryan et al., 2009;Deressa et al., 2009;Hassan and Nhemachena, 2008;Thomas et al., 2007). To ensure the appropriateness, these measures were raised for discussion in focused group discussions. Typical farmers, participants of the focus grouped discussions, were asked to choose the measures that have been used or available in their areas. The same request was given to agricultural provincelevel officers interviewed. The adaptive measures had finally been refined by the pre-tests before they were actually included in the questionnaire. Those 14 adaptive measures in table were considered as farmers' adaptive responses to climate change induced natural shocks. These adaptive measures were grouped into five groups: (A) Water use management, (B) Adjustments of crops and varieties, (C) Adjustments of planting techniques, (D) Adjustments of planting calendar, and (E) No adaptation. Most of previous studies considered "no adaptation" as lacking of information on climate change and dropped from analysis. In this paper, we treated "no adaptation" as one of choices by farmers since climate change can also bring positive effects on agriculture production.
In general, measures such as "Build/repair cistern" (5%), "Build/ repair well" (8%), and "Water saving technology" (3%) in water use management were not very commonly used. The limited use of these adaptations could be attributed to need for more capital.
While a high proportion of farmers used measure of "Change varieties" (39%) as an adjustment of crops and varieties, a lower proportion of farmers use "Change crops/livestocks" (8%), "Change crop structure" (7%) in response to climate change. Local farmers may be lacking skills, motivation and opportunities for other crops and/or livestocks. While a high proportion of farmers used measure of "Change pesticide/herbicides" (30%) as an adjustment of planting techniques, a lower proportion of farmers use "Change crop cultivation" (14%), "Change fertilizer input/stimulus" (16%), "Change crop quantity" (13%), and "Change farmyard manure" in response to climate change. These adaptations could be associated with the less expense and ease of access by farmers than that of adjustments of crops and varieties.
Less proportion of farmers use "Change irrigation schedule" (7%), and "Change crop rotation" (7%) in response to climate change. This is probably because farmers access to climate change information is rather limit. No farmer bought insurance as an adaptive measure to climate change and only 1 percent of farmers combined agriculture and forestry. Those numbers implied a lacking of suitable mechanism that can secure farmers from extreme climate events associated with climate change in the Mekong River Delta. Moreover, about 13% of the surveyed farmers reported not to have taken any adaptation method above.

Explanatory variables
With respect to household characteristics such as household head's gender, on the one hand, various studies have shown that gender is an important variable affecting adoption decisions at the farm level. According to Bayard et al. (2007) female farmers are more likely to adopt natural resource management and conservation practices. It was also emphasized by Burton et al. (1999) that female farmers are indeed important in the choice of agricultural practices to adopt, particularly in regard to conservation or sustainable technology. According to Nhemachena and Hassan (2007) the possible reason for female to adapt is that in most rural smallholder farming communities, men are more often based in towns, and much of the agricultural work is done by women. Therefore, women have more farming experience and information on various management practices and how to change them, based on available information (Anim, 1999). On the other hand, according to Asfaw and Admassie (2004), male-headed households are more likely to get information about new technologies and Ngo et al. 1365 undertake risky businesses than female-headed households. Moreover, Tenge and Hella (2004) argue that having a female head of household may have negative effects on the adoption of soil and water conservation measures, because women may have limited access to information, land, and other resources due to traditional social barriers. There is a line of argument in between by Bekele and Drake (2003) who believe that gender has no significant factor in influencing farmers' decision to adopt climate change adaptation measures. They stressed that there is a significant difference in farmer's ability to adapt to climate change due to major differences between them in terms of access to assets, education, credit, technology and input supply. Many research works have shown that education increases one's ability to receive, decode, and understand information relevant to making innovative decisions (Maddison, 2006;Lin, 1991;Igoden et al., 1990). On the contrary, Clay et al. (1998) found that education was an insignificant determinant of adoption decisions. The influence of age on adaptation has been mixed in the literature. For example, Bekele and Drake (2003) in their study of Eastern highlands of Ethiopia found that age had no influence on farmer's decision to participate in soil and water conservation activities. Others such as Nyangena (2008) and Dolisca et al. (2006), however, found that age is significantly and negatively related to farmers' decisions to adopt in soil and water conservation and forestry management programs, respectively. Studies by Maddison (2006) and Nhemachena and Hassan (2007) indicate that experience in farming increases the probability of uptake of adaptation measures to climate change.
Regarding farm characteristics, according Nhemachena and Hassan (2007) the empirical literature shows that household size has mixed impacts on farmers' adoption of agricultural technologies. Larger family size is expected to enable farmers to take up labour intensive adaptation measures. Alternatively, a large family might be forced to divert part of its labour force into non-farm activities to generate more income and reduce consumption demands.
Farm and nonfarm income and assets represent wealth. It is regularly hypothesized that the adoption of agricultural technologies requires sufficient financial wellbeing (Knowler and Bradshaw, 2007). Other studies that investigate the impact of income on adoption found a positive correlation (Franzel, 1999). The occupation of the farmer is an indication of the total amount of time available for farming activities. Off-farm employment may present a constraint to adoption of technology because it competes for on-farm managerial time (McNamara et al., 1991).
Among institutional factors, access to credit is an important factor affecting adoption of measures. Access to affordable credit increases financial resources of farmers and their ability to meet transaction costs associated with various adaptation options they might want to take (Nhemachena and Hassan, 2007). Similarly, land tenure can contribute to adaptation, because landowners tend to adopt new technologies more frequently than tenants, an argument that has justified numerous efforts to reduce tenure insecurity (Lutz et al., 1994;Shultz et al., 1997). Land ownership is widely believed to encourage the adoption of technologies linked to land such as irrigation equipment or drainage structures. Land ownership is likely to influence adoption if the innovation requires investments tied to land.
Among infrastructure factors, it is hypothesized that as distance to output and input markets increases, adaptation to climate change decreases. Proximity to market is an important determinant of adaptation, presumably because the market serves as a means of exchanging information with other farmers (Maddison, 2006). Perceived change in climate variables and access to climatic change information are also important pre-conditions to take up adaptation measures (Niles et al., 2015a;Geoff, 2014;Nicholas and Gina, 2012;Nhemachena and Hassan, 2007;Maddison, 2006). Accordingly, it is hypothesized that farmers that perceive change in climatic conditions and farmers who have access to climate change information have higher chances of taking adaptive measures in response to observable changes (Nhemachena and Hassan, 2007). The explanatory variables in this study include household characteristics such as education, gender, age of the household head, household size. Farm characteristics include farm size, farm and nonfarm income; institutional factor such as access to credit; infrastructure includes distance to input and output markets, climate conditions, and past climate experiences (Le Dang et al., 2014;Nicholas and Gina, 2012;Patrick and Richard, 2012;Nhemachena and Hassan, 2007;Maddison, 2006). In the empirical model, each explanatory variable is included in all four equations to help test if the impacts of variables differ from one adaptation option to another. Table 3 gives the descriptive statistics of variables in the models.

Farmers' past experiences of climate change induced natural shocks
Farmers' past climate experiences and variability is a prerequisite for devising subsequent adaptation strate-gies. Therefore, it is important to understand how farmers perceive climate change and variability in the context of agriculture production in the Mekong River Delta. Studies in Mekong River Delta have found that farmers' perception of climate change corresponds with local climate data (Le Dang et al., 2014;Nguyen and Le, 2012;Nguyen, 2007;UNDP, 2008, EU/MWH, 2006ADPC/GTZ, 2003). In this study, farmers' perceptions of the changes were considered in terms of nine past climate experiences -wind storm, drought, flood, higher temperature, untimely rains, salt water intrusion, eroded shorelines, pestilent insect, and water shortages. The key informants were asked to assess the frequencies of the main climate changes they have seen/observed over the last 5 years. A list of options was provided with 5 levels of occurrences. From Figure 1, recurrent droughts, flood, changes in temperature, and pestilent insect have been confirmed by their often frequencies annually in the Mekong River Delta. Higher temperature and pestilent insect are the two most often past climate change phenomena.

Past climate change induced natural shocks' impacts
The respondents were asked to rank from "very severe"  (5) to "not severe at all" (1) from options provided by the impacts they have noticed on different agricultural activities. From Figure 2, typhoon, drought, flood, higher temperature, untimely rain, eroded shoreline, pestilent insect, and water shortages are the most severe climate change phenomena in cultivation (average points of above 3.0). Pestilent insect and water shortages are the most severe climate change phenomena in husbandry (average points of above 3.0). Drought, flood, higher temperature, untimely rain, pestilent insect, and water shortages are the most severe climate change phenomena in fishery (average points of above 3.0).

Regression results
Despite the fact that the majority of the farmers interviewed claimed that they have perceived at least one change in climatic attributes, some of the farmers who perceived climate change did not respond by taking adaptation measures. Here it is argued that farmers who perceive and responded (or did not respond) share some common characteristics, which assist in better understanding the reasons underlying their response Afr. J. Agric. Res. in Table 4 (Note that one choice of adaptation -Water use management -was not significant and dropped from the results). Four models with alternative choices are in columes, namely: (1) Adjustments of crops and varieties, (2) Adjustment of planning techniques, (3) Adjustments of planting calendar, and (4) No adaptation. The R 2 for all models indicated that the statistically significant explanatory variables can explain around 14 to 20% of the variation of farmers' adaptation assessments. Breusch-Pagan test for independent equations were highly significant with values less than 0.001, implying that equations are correlated. Goodness-of-fit test indicates that four models fit the data well. No multi-collinearity problems were detected as the variance inflation factor (VIF) for all explanatory variables were less than 1.4. Tests for model specification (Ramsey test) pass for all four models. The regression coefficients in the four ISUR regression models are presented in Table 5. Bootstrap estimates were conducted. We used bias corrected bootstrapped (n =1000) results because they have been shown to perform the best with regards to power and Type I error results (Briggs, 2006), particularly for smaller sample sizes (Preacher and Hayes, 2008).

DISCUSSION
With respect to household characteristics, the estimated coefficients for the household head's gender, years of education by household head, and marital status of household head were statistically significant. Male headed households have more probability of specifically adapting to climate change which is revealed by the fact that a unit change from being headed by a female household to male increases the probability of adapting adjustments of planting techniques to climate change by 23.0%, ceteris paribus. This result is in line with the argument that male-headed households are often considered to be more likely to get information about new technologies and take risky businesses than female headed households (Asfaw and Admassie, 2004). However, male headed households have less probability of choosing no adaptation to climate change by 22.3%, Ceteris paribus.
A unit increase in the education of the head of the household will have the impact of raising the probability of making adjustments of planting techniques to climate change by 2.2%, ceteris paribus. This is in line with studies of Maddison (2006), Lin (1991) and Igoden et al. (1990). Although a series of adaptive measures has been used by many households, the above findings may imply causes of inefficient adaptation in local areas. Poor education can be one possible cause. Our estimation also showed that education level of household head has less probability of choosing no adaptation to climate change by 1.7%, Ceteris paribus.
With respect to farming characteristics, farm size was statistically significant and positively related to adaptation. Farm households who owned more land conducted adjustments of crops and varieties more by 5.8%, Ceteris paribus. The probable reason for the positive relationship between adaptation and farm size could be due to the fact that adaptation is subject to economies of scale. The result is in accordance with study by Nhemachena and Hassan (2007). If large-scale cultivation is an advantaged in conducting adaptation measures, in general, government should support some implementations of the land reform such as farmers' cooperation in cultivating in large-scale fields. In the Mekong River Delta in recent years, this trend has turned to be common. It was also found that more land households have less probability of choosing no adaptation to climate change by 3.9%, Ceteris paribus.
Households with higher production assets (but may not necessarily have most land) increases the probability of adjustments of planting calendar to climate change by 4.3%, Ceteris paribus. This is in line with Nhemachena and Hassan (2007), Knowler and Bradshaw (2007) and Table 5. Results of ISUR probit analysis of determinants of adaptation measures.

Variable
Adjustments  Franzel (1999). As mentioned in Nhemachena and Hassan (2007), with access to technology farmers are able to vary their planting dates, switch to new crops, diversify their crop options and use more irrigation, apply water conservation techniques. However, production assets requires large capital stock that can be a constraint and thus ensuring availability of cheap technologies for farmers, especially smallholders, can significantly increase their use of other adaptation options. Our estimation finds that households with higher production assets decrease the probability of choosing no adaptation to climate change by 1.6%, Ceteris paribus.
Our results indicate that farm system types alone may not determine climate change's responses; these systems are also imbedded with institutional factors, infrastructure, climate conditions and varying climate experiences as well. With respect to institutional factors, farmers with access to credit have higher chances of adapting to changing climatic conditions as found in Nicholas and Gina (2012) and Nhemachena and Hassan (2007). Household with access to credit will have the impact of raising the probability of making adjustments of planting calendar to climate change by 10.9%, Ceteris paribus.
According to Nhemachena and Hassan (2007), access to affordable credit increases financial resources of farmers and their ability to meet transaction costs associated with the various adaptation options they might want to take. With more financial and other resources at their disposal farmers are able to change their management practices in response to changing climatic and other factors. Farmers with limiting access to market have higher probability of adjustments of crops and varieties to changing climatic conditions by 1.4%, Ceteris paribus. With access to markets farmers are able to buy new crop varieties, new irrigation technologies, and other important inputs they may need to change their practices to suit the forecasted and prevailing climatic conditions (Nhemachena and Hassan, 2007). However, household with plots in longer distance from house will have the less probability of making adjustments of crops and varieties to climate change by 3.3%, Ceteris paribus. In addition, household with plots in longer distance from house will have the more probability of choosing no adaptation to climate change by 3.6%, Ceteris paribus. Overall, the improvement of both the accessibility and usefulness of local services is deemed a necessity for adaptation strategies.
With respect to climate conditions, increasing temperature increases the probability of adapting adjustments of planting techniques to climate change by 0.1%, Ceteris paribus. The fact that adaptation to climate change increases with increasing temperature is in line with the expectation that increasing temperature is damaging to agriculture and farmers respond to this through the adoption of different adaptation methods (Nguyen and Le, 2012;Nhemachena and Hassan, 2007). It was also found that increasing temperature also decreases the probability of choosing no adaptation to climate change by 0.1%, Ceteris paribus.
Annual average precipitation is negatively related to adaptation. Increasing rainfall decreases the probability of adapting adjustments of planting techniques and adjustments of planting calendar to climate change by 0.01%, Ceteris paribus. The probable reason for the negative relationship between average annual precipitation and adaptation could be due to the fact that agriculture in the Mekong River Delta is water scarce and faces high temperature and increasing precipitation will not constrain agricultural production or does not promote the need to adapt (at least using the main adaptation options considered in this study).
ISUR estimates show that past climate experiences increases the probability of uptake of adaptation measures (Niles et al., 2015b;Nicholas and Gina, 2012;Maddison, 2006). Farmers who are aware of changes in climatic conditions have higher chances of taking adaptive measures in response to observed changes. Specifically, increasing drought increases the probability of farmers to respond to changes in terms of adjustments of planting calendar (Niles et al., 2015b andPatrick andRichard, 2012) by 17.7%, Ceteris paribus.
Increasing untimely rain increases the probability of farmers changing their management practices, in particular, adjustments of crops and varieties (including changes in varieties, crops/livestocks, and crop structure) by 26.6%, Ceteris paribus, and adjustments of planting techniques (including changes in crop cultivation, fertilizer/stimulus, pesticides/herbicides, crop quantity, and farmyard manure) by 11.2%, Ceteris paribus. Resulting water shortages leads to adjustments of planting calendar, including changes in irrigation schedule, and crop rotation (Niles et al., 2015b) by 21.8%, Ceteris paribus. However, increasing untimely rain decreases the probability of farmers making no adaptation by 8.2%, Ceteris paribus. This can be explained that farmers' experiences of climate change induced natural shocks lead their choice of no adaptation since this kind of shock benefits their agriculture production (Geoff, 2014). Generally, if perception of climate change induced natural shocks are the most salient for farmers, it likely has significant implications for assessing how short-term responses can influence longterm adaptations and the subsequent policies that may be needed to accompany such actions (Carlo et al., 2015;Le Dang et al., 2014;Park et al., 2012). In addition, because climate variability in higher temperature and accompanied by drought and water shortages, irrigation investment needs from the viewpoint of Public -Private Partner (PPP) should be reconsidered to allow farmers increased water control to counteract adverse impacts from climate variability and change.

CONCLUSION AND IMPLICATIONS
A better understanding of current adaptation measures and their determinants will be important to inform policy for future successful adaptation. This study was based on farm-level analysis of farmers' past climate experiences and distinguished determinants of adaptation to climate change induced natural shocks in the Mekong River Delta of Vietnam. This research has shown that the majority of farmers used adaptive measures that mostly related to their farming practices such as adjustments of crops and varieties (including changes in varieties, crops/livestocks, and crop structure), adjusting planting techniques (such as changes in crop cultivation, fertilizer/stimulus, pesticides/herbicides, crop quantity, and farmyard manure) and adjusting planting calendar (such as changes in irrigation schedule, and crop rotation). On top of that, no adaptation is also considered as a choice of adaptation. The adaptive measures farmers followed were those that they perceived climate change induced natural shocks such as wind storm (typhoon), drought, flood, higher temperature, untimely rain, salt water intrusion, eroded shorelines, pestilent insect, and water shortages.