Understanding farmers’ adaptation to climate change: A protection motivation theory application

Abstract This study examines the psychological determinants influencing farmers’ intentions to adapt to climate change through the lens of Protection Motivation Theory. This study uses cross-sectional data comprising 302 farmers in East Java, Indonesia, analyzed using a structural equation (SEM) model. The findings indicate that perceived risk and adaptation assessment positively impact perceived climate variability but negatively impact maladaptation. Meanwhile, perceived risk and adaptation assessment directly correlate with higher intention for adaptation. Intriguingly, adaptation assessment and perceived risk indirectly impact adaptation intention via perceived climate variability, although this influence is negative. Likewise, maladaptation emerged as a mediator variable, exerting a counteractive indirect impact on adaptation assessment. These findings suggest that farmers with a robust understanding of climate change are more inclined to adapt. Conversely, greater maladaptation diminishes the adaptation intention. This study recommends continuous implementation of adaptation strategies, coupled with more accessibility to climate information and extensions facilitated by governmental bodies. This multifaceted approach will raise awareness regarding climate change impacts, hence improving their adaptive practices.


Introduction
Agriculture is one of the most vulnerable sectors to climate change.One of the effects of climate change is the long rainfall season, which stimulates the emergence of diseases and pest attacks on horticultural commodities (Kogo et al., 2021).These changes have been shown to significantly impact productivity (Asante et al., 2021;Supriyanto et al., 2020;van Oort & Zwart, 2018;Williams et al., 2019;Rahman et al., 2023).The impact of this situation makes farmers increasingly vulnerable, particularly those in developing countries such as Indonesia that heavily rely on the agricultural sector (Altieri et al., 2017;Comoé & Siegrist, 2015;Esham & Garforth, 2013).Furthermore, Supari et al. (2017) stated that several regions in Indonesia have seen significant changes in rainfall.The average annual rainfall in Indonesia reached 2300 mm during the period 1991-2005 (Aldrian & Djamil, 2008).A significant change in average rainfall was observed in several provinces in Indonesia, particularly on Java Island, during the period 1981-2016 (Rahman et al., 2021).In another case in Southeast Asia, Indonesia, the super El Niño phenomenon occurred in 1997/1998 and 2015, in which significantly lower-than-normal rainfall was observed (Kirono et al., 1999) A study in Yogyakarta, Indonesia, has shown that climate variability, including floods and droughts, could cause damage to agricultural lands and increase the risk of pest attacks on crops (Saptutyningsih et al., 2019).Meanwhile, exteme weather events such as continuous heavy rain have been known to devastate hectares of potato crops, resulting in billions of losses for farmers in the Sumber sub-district, Probolinggo Regency (Supriyanto et al., 2020).According to data documentation from Statistics Indonesia (BPS 2020), there has been a decline in potato productivity in Probolinggo and Malang Regency due to extensive rain and high precipitation in 2016.
Adaptation strategies play an important role in reducing the negative effects on crop productivity, so farmers need to adopt them to mitigate the impacts of climate change (Di Falco et al., 2011;Rahman et al., 2022Rahman et al., , 2023)).Without the implementation of adaptation strategies, the impact of climate change will continue to drive the decline in agricultural yields (van Oort & Zwart, 2018).A study conducted by Khanal et al. (2018) summarizes that the climate change adaptation strategies employed by farmers have enabled them to improve their yields.Using propensity score matching analysis to estimate the role of adaptation strategy on farms' yields and crop income, Ahmad and Afzal (2020) found that farmers who adapted produced higher yields and earned crop income than those who did not.A similar finding was also presented by Mottaleb et al. (2017), who indicated a positive and significant effect of adaptation strategies on farmers' productivity.The study revealed that the implementation of adaptation strategies can lead to an improvement in wheat yields by approximately 42-65 kg per hectare.
In the agricultural sector, adaptation to climate change can be implemented on two levels, namely, community and individual.In this study, adaptation refers specifically to individual-level adaptation, which is initiated and implemented by individuals and households based on their rational self-interest (IPCC, 2018).This sector plays a significant role in the design, finance, and implementation of adaptation strategies (Centre for Climate Change and Multilateral Policy, 2020).This adaptation can be influenced by various socio-economic factors, which have been described as key determinants of farmers' decision-making.These factors include farming experience and education level (Fadina & Barjolle, 2018), wealth, government farm support, access to fertile land, and credit (Bryan et al., 2009).Other determinants include socio-demographic characteristics and institutional accessibility (Alemayehu & Bewket, 2017;Arunrat et al., 2017).However, climate change adaptation is also influenced by the strength of belief in the reality of climate change and adaptive capacity.These factors have been found to influence the adaptation of Swedish forest owners (van Gameren & Zaccai, 2015).Belief in the reality of climate change is crucial for initiating any actions (Fielding & Hornsey, 2016).Empirical evidence has shown that the perceived risks significantly influence motivation to adapt to climate change (Osberghaus et al., 2010).Psychological factors have also been observed to influence farmers' adaptive capacity and decision-making (Grothmann & Patt, 2005).According to Dang et al. (2014), the concept of a psychological factor was built upon the protection motivation theory (PMT), which considers four elements in the cognitive mediating processes of PMT: threat appraisal, coping appraisal, maladaptive coping (the individual who denials of climate change adaptation), and protection motivation (Milne et al., 2000).The results indicate that farmers show higher adaptation intentions when they perceive higher risks of climate change and believe in the effectiveness of adaptive measures.Conversely, they are less likely to adapt when they deny the risks of climate change or have a fatalistic attitude (Dang et al., 2014).
Although studies on the role of perception in adaptation to climate change have been conducted extensively, little attention is given to psychological factors (López-Marrero & Yarnal, 2010).Most studies focus on agriculture in general (Chipfupa et al., 2021;Hitayezu et al., 2017), the mental health of farmers (Berry et al., 2011), and specific crops such as rice and maize (Albert et al., 2021(Albert et al., , 2022;;Dang et al., 2014;Truelove et al., 2015;Waldman et al., 2019;Zhang et al., 2020).No studies specifically examine the adaptation strategies of horticulture farmers, especially the potato commodity in Indonesia.Likewise, previous research on climate change adaptation has utilized the PMT (Dang et al., 2014) to understand individuals' responses to environmental challenges.However, farmers' climate perceptions have not been much considered in the studies.Recognizing that individuals' perception of climate change greatly influences their adaptive behaviors, this study aims to address this gap by introducing climate perceptions as an additional variable within the PMT framework.This research examines how farmers' understanding and interpretation of climate changes interact with motivational factors to shape their adaptation strategies.As such, this research seeks to provide a more comprehensive and nuanced understanding of the dynamics underlying climate change adaptation.Through this extension of the theoretical model, valuable insights into the interplay between cognitive factors and motivation can be gained.As such, interventions aimed at fostering sustainable agricultural practices will be more targeted and effective.
The present study makes a significant contribution to the understanding of climate change adaptation strategies within the agricultural sector, with a particular focus on horticultural farmers and the potato commodity.Agriculture, especially in developing countries like Indonesia, remains highly vulnerable to the adverse impacts of climate change, including increased disease outbreaks and pest attacks due to changing rainfall and temperature patterns (Rahman et al., 2022(Rahman et al., , 2022)).Despite the detrimental effects of these changes on productivity and livelihoods, there is a lack of comprehensive research that specifically examines the adaptation strategies employed by horticultural farmers to mitigate these challenges.This study addresses this critical gap by extending the application of the PMT to incorporate farmers' climate perceptions as an additional variable (Dang et al., 2014).Previous research has utilized PMT to understand individuals' responses to environmental challenges.This study specifically considers farmers' cognitive understanding and interpretation of climate change in the PMT framework to offer a more holistic view of the dynamics influencing farmers' adaptive behaviors.This can extend the understanding of how cognitive factors interact with motivational elements to influence adaptation strategies to deal with climate change.Another contribution of this study is the examination of adaptation strategies at a more granular level, focusing on the private adaptation initiatives by individual farmers and households.Furthermore, the study acknowledges the role of psychological factors in shaping adaptation decisions, i.e., the risk perception and belief in the efficacy of adaptive measures.Ultimately, this research extends the existing literature, which has primarily focused on general agriculture or specific crops (Bagagnan et al., 2019;Villamor et al., 2023), by specifically targeting the horticultural sector and the potato commodity.As such, the study provides valuable insights for policymakers, researchers, and practitioners seeking to enhance climate resilience within this critical agricultural subsector.Exploring the interplay between climate perception, cognitive factors, and motivation, this study can inform the development of tailored and effective interventions aimed at promoting sustainable agricultural practices in dealing with climate change.
Based on the background above, this research hypothesizes that factors in the PMT model, including Adaptation Assessment and Perceived Risk, positively influence Adaptation Intention.Conversely, Maladaptation and Perceived Climate Variability are expected to negatively impact Adaptation Intention.Additionally, the study posits that Maladaptation and Perceived Climate Variability significantly mediate the relationship between PMT factors and Adaptation Intention.
This research was conducted in Malang and Probolinggo Regency (Figure 1), which are major agricultural areas and have the highest horticultural productivity in East Java, Indonesia.These regencies are part of the Bromo Tengger Semeru National Park (BTS) and have been identified as areas highly vulnerable to climate change.
A multistage random sampling was applied in this study.In the first stage, the location, East Java Province, was determined purposively.Subsequently, two regencies were chosen, i.e., Malang and Probolinggo.In the second stage, the selection of sub-districts was based on the information obtained from the Department of Agriculture and the local agriculture agencies.Areas with major horticultural production, potatoes in particular, were selected.As a result, Poncokusumo was selected for Malang Regency, and Sumber was selected for Probolinggo Regency.In the third stage, two villages were randomly selected from each regency (Table 1).Finally, horticultural farm households were selected using simple random sampling from each village.From the compiled list, a total of around 150 horticulture farmers were randomly chosen from each sub-district.Therefore, we obtained a total of 302 horticulture farmers from the two sub-districts.
The survey data were collected from August to September 2021 through face-to-face interviews with respondents, utilizing a questionnaire as the data collection instrument.As part of the collaboration with Brawijaya University, Faculty of Agriculture, specifically in the field of socioeconomics, the survey was conducted by a team of ten trained enumerators, with five assigned to each regency.The respondents were heads of households who were responsible for the family's economy and engaged in potato cultivation.To qualify as respondents, farmers needed to have previous experience in growing potatoes and have some level of awareness regarding climate change in the respective locations.The enumerators provided an explanation of the definition of climate change and its interpretation to the respondents before proceeding with the structural questions.This study uses the definition of climate change from the Intergovernmental Panel on Climate Change (IPCC, 2007), delivered to the farmers using simplified language.Prior to the actual interview process, each enumerator conducted a pilot test to ensure clarity, relevance, and ease of conducting the interviews.
The survey questionnaire encompasses various aspects, including household characteristics, psychological factors (such as climate change perceptions, risk experience, climate change risk perception, climate change adaptation assessment, adaptation intention, maladaptation (the individual who denials climate change adaptation), and adaptation strategies.In addition to the survey data, supplementary secondary data were collected from various sources, including journal articles, books, annual reports, Intergovernmental Panel on Climate Change (IPCC) reports, and other relevant resources.

Variable measurement and estimation procedures
This study utilizes the PMT approach to understand the impact of psychological factors on farmers' adaptation to climate change.PMT has been widely applied in the agricultural sector to investigate farmers' pro-environmental behavior in the contexts of drought (Keshavarz & Karami, 2016), organic food choice (Scarpa & Thiene, 2011), and agricultural conservation (Tama et al., 2021).PMT has also been employed in the agricultural sector to investigate climate change (Feng et al., 2017;Nabara et al., 2021;Poudyal et al., 2021).Hence, the PMT is an effective framework for understanding how horticulture farmers cope with the challenges posed by climate change.
Based on previous studies (Feng et al., 2017;Nabara et al., 2021;Poudyal et al., 2021), the PMT consists of four key elements, namely threat appraisal, coping appraisal, maladaptation, and protection motivation.In this study, we aim to apply these four elements to address climate change issues, utilizing the following variables: "Perceived risk of climate change" for threat appraisal, "Adaptation assessment" for coping appraisal, "Maladaptation" for maladaptive coping, and "Adaptation intention" for protection motivation.The entire model comprises five constructs, namely risk perception of climate change, adaptation assessment, maladaptation, climate variability, and adaptation intention.Each of these constructs is measured using distinct observed indicators, as detailed in Table 2.
The first variable is the perceived risk of climate change.It examines the perceived risk across five aspects of farmers' lives, namely physical health, finance, production, social relationships, and psychology.This variable addresses the threat of climate change to individuals using two terms.Firstly, perceived vulnerability captures individuals' expectations of their susceptibility to the risks associated with climate change.Secondly, perceived severity reflects individuals' perception of the magnitude or severity of the dangers posed by climate change (Oakley et al., 2020).Therefore, respondents in this study were asked questions regarding their perceptions of vulnerability and severity in relation to climate change's impact on each aspect of their lives (physical health, finance, production, social relationships, and adaptation intention).These perceptions were assessed using indicators with a seven-point scale.Next, adaptation assessment is related to individuals' evaluation of their relative capacity to implement adaptation interventions.This assessment consists of three parts: perceived selfefficacy, which refers to individuals' belief in their ability to carry out adaptive actions; perceived adaptation efficacy, which pertains to individuals' conviction that adaptive measures are effective; and perceived adaptation cost, which refers to the anticipated costs associated with implementing adaptive strategies, such as money, time, and effort-related expenses.
Furthermore, individuals may exhibit a preference for maladaptation, characterized by fatalism, ignorance, and wishful thinking, or they may demonstrate behavioral intention to implement adaptation measures after assessing the risks associated with climate change.Fatalistic attitudes, wishful thinking, and denial of climate change risks can hinder farmers' efforts to adapt.People's perceptions play a crucial role in their adaptation to climate change (Paudel et al., 2019).These include how individuals perceive changes in climate variability, including shifts in rain intensity, temperature, and the occurrence of drought.
To estimate the impact of psychological factors on farmers' intention to adopt adaptation strategies, structural equation modeling using AMOS 22 was employed.The objective of this estimation is to identify which constructs significantly influence farmers' adaptation intention in response to climate change.The model used in this study is presented in Figure 2.This model has been widely implemented to describe relationships between several variables in climate change studies, such as (Luu et al., 2019;Nguyen et al., 2021;Tikir & Lehmann, 2011).The indicators of each latent variable in the study model are shown in Table 2, with the measurement using a Likert scale from 1 to 7.

Structural equation model
This research employs the PMT framework to explore how psychological factors influence farmers' adaptation to climate change.PMT comprises four key components: threat appraisal, coping appraisal, maladaptive responses, and protection motivation.Additionally, this investigation seeks to adapt these four components to the context of climate change, measuring the protection motivation using the "perceived risk of climate change" variable to measure threat assessment, "assessment of adaptation" to evaluate coping strategies, "maladaptation" to understand maladaptive coping, and "adaptation intention" to measure protection motivation.
The perceived climate change risk deals with the threat that climate change poses to farmers, encompassing two key aspects.The first is the element of perceived vulnerability, which denotes how individuals perceive their own risk and anticipate their vulnerability to the hazards of climate change.The second is perceived severity, where the individual may consider the extent to which the potential threat is harmful (Oakley et al., 2020).Adaptation assessment is related to how people assess their relative capacity to implement adaptation interventions, which consists of three aspects.First, perceived self-efficacy refers to how individuals perceive their ability to carry out adaptive steps.Second, the perceived adaptation efficacy refers to people's belief that an adaptation measure will be successful.Third, the perceived adaptation cost refers to the costs of implementing adaptive strategies, including money, time, and effort.Furthermore, people's inclination toward maladaptation (i.e., fatalism, ignorance, and wishful thinking) or adaptation intentions may change after assessing the danger and dealing with climate change.Fatalism, wishful thinking, and denials of climate change danger can drive farmers' efforts to adapt.Perceptions also have an essential role in making people adapt (Paudel et al., 2019), such as perceived climate variability, rain intensity change, temperature change, and drought.
The estimation of the effect of psychological factors on farmers' intention to adapt to climate change was done by structural equation modeling using Smart PLS software.PLS-SEM has advantages for dealing with a complex model, lower sample sizes, non-normal data distribution, and predictive or exploratory research (Hair et al., 2017).The software has been widely implemented to describe relationships between several variables in climate change studies (Sohail, 2023;Wang et al., 2022;Zhang et al., 2020).Figure 2 and Table 2 provide an illustration of the model applied in this study, along with descriptions of the indicators for each latent variable, which are assessed using a Likert scale ranging from 1 to 7.
Figure 2 shows the direct and indirect effects between variables (independent, mediator, and dependent variable).Perceived risk and adaptation variables have a direct effect on adaptation intention and an indirect effect on adaptation intention through maladaptation and perceived climate variability as mediator variables.Furthermore, maladaptation and perceived climate variability have two functions, as a mediator variable and a dependent variable.As a mediator variable, it modulates the indirect effect of perceived risk and adaptation to adaptation intention.In addition, it functions as a dependent variable to modulate the effect of perceived risk and adaptation assessment on maladaptation and perceived climate variability.
In addition, the structural equation modeling analysis uses a reflective measure type (Figure 2), meaning that changes in the underlying latent construct are reflected by changes in indicators.Each construct has a different indicator that is mutually correlated.If high correlations between the indicators exist, an indicator can be dropped without changing the essential nature of the underlying construct (Jarvis et al., 2003).

Composite validity test
Figure 3 presents the factor loading value.The results show the reliability and validity of all the constructs used in the theoretical model.According to Wong et al. (2013), the indicators of a latent variable are acceptable if the factor loading > 0.5.Our results indicated that the model constructs factor loadings are above 0.5.However, we decided to remove the adaptation response cost indicator of the adaptation assessment construct and the social relationship indicator of the perceived risk construct due to a low factor loading of those indicators.It can be interpreted that the low factor loading indicates that the perceived adaptation cost is not strongly/significantly related to the adaptation assessment construct.Furthermore, it is difficult to interpret the effectiveness and the cost in one element.Also, we removed several indicators of the adaptation intention with factor loading < 0.5.

Discriminant validity and convergent validity
As presented in Table 3, Cronbach's Alpha values range from 0.611-0.838,with most values higher than 0.7 except for perceived climate variability and perceived risk.While the value of composite reliability (CR) values for all constructs are greater than 0.7.The value of Cronbach's Alpha and Composite reliability at 0.7 or greater indicates good reliability (Henseler et al., 2014), as we know that CR is one indicator of convergent validity.However, Cronbach's Alpha below 0.7 is deemed an acceptable reliability (Triwidyati et al., 2020).Ab Hamid et al. (2017) and Hair et al. (2011) recommended that an Average Variance Extracted (AVE) value is greater than the minimum score of 0.5.In this study, the AVE values are greater than 0.5, so they are declared valid and the constructs have an acceptable convergent validity (Fornell & Larcker, 1981).Therefore, the square root of each construct's AVE should have a greater value than the correlations with other latent constructs, which verifies that one construct differs from the other (Table 4) (Hair et al., 2017).
Furthermore, the variance inflation factor (VIF) indicates multicollinearity and should be smaller than 5 (Hair et al., 2011).In this study, the VIF value for all of the indicators in each construct (adaptation intention, maladaptation, perceived climate variability, perceived risk, adaptation assessment) is smaller than 0.5.Based on those findings, all of the construct measurements are reliable and valid, so the structural model can be assessed for the hypothesized relationships between constructs.

The Goodness of Fit (GOF)
The goodness of fit estimation is presented in Table 5.Firstly, the Saturated Model, representing a perfect fit, is compared to the Estimated Model.The Standardized Root Mean Square Residual (SRMR) is 0.997, which exceeds the recommended threshold of 0.100, indicating a fit.The Chi-Square statistic is 1127.797,suggesting a higher value, which is generally better for model fit.
However, it should be interpreted cautiously as it is sensitive to sample size.Next, several indices are considered.The Normed Fit Index (NFI) is 0.566, signifying a moderate fit when closer to 1.00 (Putritamara et al., 2023).The Goodness of Fit Index (GFI) and Comparative Fit Index (CFI) both exceed the desired threshold of 0.90, indicating an acceptable fit (Farani et al., 2021).

Q-square (Q 2 )
Q-square (Q 2 ) predictive relevance is determined by blindfolding procedure analysis.Q 2 values greater than zero suggested that the exogenous constructs (adaptation assessment, perceived risk, maladaptation, and perceived climate variability) have predictive relevance for the endogenous construct under consideration (Hair et al., 2021).More specifically, this study has resulted in a Q 2 of 0.284, indicating that this model has a medium predictive relevance.

Hypothesis test
The determinants of farmers' adaptation intention using protection motivation theory are presented in Table 6.First, the path from adaptation assessments to perceived climate change shows a positive coefficient with the value of t-statistics 7.41, statistically significant at the 1% level, suggesting that farmers with better adaptation assessment are more likely to perceived a change in the climate indicators, such as the change of rain intensity and temperature.Yet, this finding also confirms that the farmers' awareness of adaptation assessment is associated with farmers' perception of climate change.Second, the path from adaptation assessments to maladaptation has a negative coefficient with a statistical value of about 5.58, significant at the 1% level.This result implies that maladaptation is negatively influenced by farmers' adaptation assessments.Third, the path from assessment to adaptation intention shows a positive coefficient, with a value of t-statistics of about 9.03, significant at the 1% level.This result implies that the adaptation assessment has a positive and significant direct effect on farmers' adaptation intention.
The finding suggested that farmers who believe in adaptation strategies to mitigate the impact of climate change tend to have higher adaptation intentions.The finding is in line with a previous study by Nguyen et al. (2021), which also points out the positive association between adaptation assessment and adaptation intention.On the other hand, adaptation assessment influences adaptation intention indirectly, but significantly, through perceived climate variability.This finding confirms that farmers' intentions to implement adaptation in their agricultural systems are influenced by their perceptions of climate indicators, such as rain intensity and temperature change.According to Singh (2020), farmers protect their farming systems and livelihoods by applying their expertise, guided by their perceptions of changing climate (e.g., changes in rainfall and temperature).This finding is in line with a previous study by Abid et al. (2019), revealing that farmers' perception of climate change has a significant contribution to climate change adaptation improvement.However, adaptation intention is negatively and significantly influenced by adaptation assessment indirectly through maladaptation.This finding reveals that farmers with a superior climate change adaptation assessment are less likely to engage in wishful thinking, fatalism, or denials of climate change risks.Those with less maladaptation experience are more inclined to adapt.The next path of this study shows the influence of perceived risk on climate variability, which indicates a positive coefficient with the value of t-statistics is 5.33, statistically significant at the 1% level.The finding reveals that farmers with a high perceived risk of climate change are more likely to sense changes in climate indicators, such as rainfall and temperature.This finding is in line with the previous study, suggesting a significant correlation between climate risk and temperature change (Lu et al., 2020) and rainfall (Simelton et al., 2013).The influence of perceived risk on maladaptation shows a negative coefficient, with the value of the t-statistic being 3.78, significant at the 1% level, suggesting that the perceived risk of climate change significantly and negatively influenced farmers to deny climate change.This finding aligns with a study by Dang et al. (2018), which discovered that farmers who perceive higher climate change risks are less likely to engage in wishful thinking, fatalism, or denials of risk.However, a study by Grothmann and Patt (2005) indicates otherwise.The different results may be explained by the different locations, with cultural and religious backgrounds, that shape the beliefs about climate change.Our study is in line with Nor Diana et al. ( 2022), which may be because the research location is the same southeast Asia region.
The direct effect of perceived risk on adaptation intention shows a positive coefficient with a statistical significance at the 1% level.The finding implies that farmers with a higher perception of climate change risk tend to have more intention to apply climate change adaptation.A tentative explanation is, when farmers perceive increased hazards of climate change as a threat to their physical health, financial security, production, social ties, and psychology, they are more likely to reduce climate change impacts and, hence adapt.
On the other hand, perceived risk also has an indirect effect on adaptation intention through perceived climate variability.The finding shows that perceived risk has a positive and significant effect mediated by perceived climate variability.The result suggests that farmers with a higher perception of climate change risk and perceive the climate indicators, such as rainfall and temperature, are more likely to adapt.Therefore, this study claims that farmers' perception of climate change, i.e., when they believe that they are threatened by climate change, can improve farmers' adaptation intention.This result is in line with the existing literature that highlight the important role of farmers' perception in climate change, for instance a study by Asrat and Simane (2018) in West Ethiopia, Kemausuor et al. (2011) in Ghana, and Zhang et al. (2020) in China, and Fachrista (2019) for Indonesia.
The indirect effect of perceived risk on adaptation intention through perceived climate maladaptation shows a negative coefficient and is statistically significant at a 1% level.As we have mentioned earlier, the risk perception directly and significantly improves farmers' adaptation intention.More specifically, when farmers experience a higher maladaptation, they are less likely to adapt the climate change.This finding reveals that farmers' probability of having a high adaptation intention is lower when they are in denial of the climate change phenomena.According to Dang et al. (2014), farmers' maladaptation is the effect of incorrect information that may reduce farmers' awareness of climate change phenomena.Therefore, it will reduce farmers' adaptation intention.Dang et al. (2014) also indicated a similar result, stating that maladaptation significantly reduces farmers' adaptation intention.

Conclusion and policy implication
This study investigated the influence of psychological factors on farmers' intention toward adaptation strategies by protection motivation theory.This study employs cross-sectional data from 302 farmers in East Java, Indonesia, which was analyzed by a structural equation model.The empirical results indicate that perceived risk and adaptation assessment tend to have significantly positive effects on perceived climate variability and significantly negative effects on maladaptation.Further, perceived risk and adaptation assessment has a positive and statistically significant direct effect on farmers' adaptation intention.Moreover, adaptation assessment and perceived risk influences adaptation intention indirectly but significantly through perceived climate variability.Adaptation intention is also negatively and significantly influenced by adaptation assessment indirectly through maladaptation as a mediator variable.This finding implies that farmers with a better perception of climate change are more likely to apply climate change adaptation strategies.However, higher maladaptation tends to lower the intention to adapt.
This study provides important implications for agriculture development.First, we suggested that farmers should employ adaptation strategies continuously to reduce climate change's impact on their agricultural activities.To implement this, the government can provide climate information access and climate-related extensions.Farmers' awareness of the climate change phenomena and its negative impact on climate change should be increased.As such, their intention to adapt will be higher.

Figure 1 .
Figure 1.Map of study location.

Figure
Figure 2. Conceptual model of the protection motivation theory.