The diversity of gendered adaptation strategies to climate change of Indian farmers: A feminist intersectional approach

International initiatives, particularly the United Nations Framework Convention on Climate Change (UNFCCC) and the Sustainable Development Goals, recognize that understanding the vulnerability and adaptation responses of local communities to climate change is critical to promote mechanisms for climate change-related planning. In the context of climate change, adaptation refers to adjustments in natural or human systems in response to actual or expected climatic stimuli or their effects, which moderates harm or exploits beneficial opportunities (IPCC 2014). It involves adjustments in lifestyle, behaviours and socio-economic structures (i.e. livelihood security-based responses) as well as in land use and management of biodiversity and ecological processes (i.e. ecosystem-based responses) (Smit et al. 2000; Ojea 2015; Vignola et al. 2015). The social and institutional dimensions of vulnerability and adaptation to climate change have been increasingly addressed by scholars (Smit and Wandel 2006; McLaughlin and Dietz 2008), specifically opening questions related to understanding matters of social differentiation, equity and power (Füssel and Klein 2006). However, the multiple determinants that shape differentiated context-specific vulnerabilities and, particularly, adaptive capacity responses are still largely unexplored (Terry 2009; Carr and Thompson 2014; Kaijser and Kronsell 2014). Specifically, scholars recognize that such determinants are gendered and mediated by social, cultural, institutional and economic structures and processes (Morton 2007; Kaijser and Kronsell 2014; Sugden et al. 2014). In this sense, a significant body of literature on gender and climate change shows that women and men perceive and experience climate change differently, and usually women are more vulnerable due to their dependence on natural resources and structural inequity in their access and control of such resources (Dankelman 2010; FAO 2011; Nellemann et al. 2011; WEDO-IUCN 2014). However, recent critiques from feminist political ecology perspectives question the validity of the conventional binary male–female view of gender in climate change studies as they can lack sufficient consideration of power relations determined by the social context (Demetriades and Esplen 2008; Rodenberg 2009; Arora-Jonsson 2011; Resurreccion 2011; Carr and Thompson 2014). Firstly, presenting women as a passive victim of change may misinterpret the causes of vulnerability and obscure the role of women as proactive agents of adaptation (Mitchell 2007; Dankelman 2010). Secondly, climate change debates related to gender are shifting from a simple view of women as a homogenous group toward a more complex view of identities within gender categories (Tschakert 2012). Indeed, multiple social, economic, and cultural characteristics interact with gender in influencing power inequities and explaining how and why people face and manage climate change in different ways (Carr 2008). Thirdly, several authors recognize that climate change occurs together with other concomitant direct and indirect drivers, which together affect local livelihoods, gender relations and management decisions (O’Brien et al. 2004; Nielsen and Reenberg 2010; Sugden et al. 2014).

In this context, it is useful to introduce intersectionality—a feminist sociological concept first introduced by Crenshaw (1991)—which leads us to critically understand the nature of differentiated vulnerability and adaptation due to the interactions among multiple dimensions of social and power relations (McCall 2005; Davis 2008). Recently, a growing body of research from critical feminist political ecology has raised awareness about the importance of intersectionality on adaptation to climate change in agrarian settings (Carr and Thompson 2014; Kaijser and Kronsell 2014; Sultana 2014; Ogra and Badola 2015). In the Asian context, for instance, several empirical studies clearly confirmed the interplay of multiple identities, such as caste, economic class and gender, shaping differentiated vulnerability to risks and disasters (Ahmed and Fajber 2009; Ray-Bennett 2009; Onta and Resurreccion 2011). Other studies found that social class, household head gender, age and stage of life may determine women’s ability to respond to water scarcity (Huynh and Resurrección 2014). Similar interactions have been identified in Africa, where the joint effects of gender, access to education, land and credit are analysed as local determinants of the capacity to adapt to decreasing precipitation (Below et al. 2012; Fosu-Mensah et al. 2012). The intersectionality framework, thus, has potential application as an analytical tool for understanding how interactions among multiple social dimensions of power can determine the development of adaptation strategies to climate and other concomitant drivers of change (Adger and Kelly 1999; Nightingale 2011; Marino and Ribot 2012). However, methodological discussions on how to implement such a framework are open and empirical applications remain to be undertaken (Davis 2008; Lykke 2010; Kaijser and Kronsell 2014).

By adopting an intersectional analysis approach, this paper aims to investigate the role of gender and its interaction with cultural, social and economic factors in determining the adaptive responses to climate and other multiple drivers of change in two empirical cases in India, located in two regions extremely vulnerable to climate change. We specifically aim to answer three questions: (i) How do the responses to climate change and other concomitant agrarian pressures manifest themselves in different biophysical and socio-economic regions? (ii) Are local vulnerability and adaptation strategies gender-differentiated in their perception and adoption? (iii) Are there other cultural, social and economic factors interacting within the different contexts with gender, contributing to shape adaptation responses?

The first case study was located in the Kumaon region of the Himalayan State of Uttarakhand, representing a mid-high hills agro-ecological zone (ca. 1200–1600 m above sea level). The second case study was located in an agro-ecological zone characterized by plains in the State of Bihar, in the mid Indian Gangetic region (ca. 30–50 m a.s.l) (Fig. 1).

At a first tier of analysis, our findings highlight how bundles of strategies are prioritized and adopted depending on the socio-ecological context, but mediated by a gendered nature of perceptions and decision-making within the household. In the more remote areas of Uttarakhand, a sustainable management of ecological functions and processes and agro-biodiversity are prioritized as a question of livelihood security. These findings are similar to those found in other mountain contexts in the world, for example in the Himalayan mountains (Regmi et al. 2009), in the Pamir mountains (Partoev 2012; Zimmerer 2014) and in the Andean mountains (Narloch et al. 2012). As noted elsewhere, such strategies are risk-smoothing in the context of climatic variability (Kotschi 2007; Lipper and Cooper 2009; Pascual et al. 2015), especially in low-intensity agro-ecosystems (Pascual et al. 2011; Hellin et al. 2014). Women are especially concerned with complementing agro-biodiversity management with social ties and knowledge system management mainly through exchanges of knowledge and planting materials (crops and varieties in different fields). This is reflected by a major adoption of such strategies when women are making decisions within the family. However, the data show several shifts in the strategies related to agro-biodiversity and ecological functions management. These include changes towards permanent crops and the introduction of high-yield improved seeds. Such changes are more evident in the mid-plains of Bihar, where there is a long history of external interventions and technological strategies which complement the range of strategies adopted to ensure livelihood security. This result suggests that ecosystem-based strategies are not always preferred among small-holders farmers, possibly due to the trade-offs between the benefits of adaptation in the long term and the possible transition costs in the short term, such as those related to labour intensity (as suggested by Bhattarai et al. 2015). Additionally, our findings suggest that there are differences in the awareness of individual family members to change and thus different responses are preferred and adopted to cope with shock events. Hence, a priori assumptions on the basis of a male/female dichotomy are unable to lead to a comprehensive understanding of farmers’ choices, vulnerability, adaptation processes and barriers to adoption.

Therefore, at a second tier of analysis we attempt to disentangle how multiple and fragmented dimensions of identity (including gender) intersect, revealing systems of power and cross-cutting institutions, to shape situation-specific interactions between farmers and ecosystems in the context of climate change. By helping represent complex realities, such feminist intersectional approach can potentially improve our understanding on how gender relationships are crosscut by ethnicity, caste, age, wealth class and capabilities, but also shaped by factors such as knowledge, access to communication networks, risk perception and awareness and social mobilization, ultimately influencing the ability to undertake adaptive measures. Additionally, the approach demonstrates its usefulness to visualize how categories are changing and renegotiated under new drivers. Specific examples emerge from our study. In Uttarakhand, as a result of the phenomenon of economically driven male out-migration, middle-aged women with good access to land and diversified incomes (which together may be used as a proxy for measuring wealth) are not only more involved in agriculture and marketing, but are also playing a greater role in decision-making in their domestic spheres (Fig. 4, axis 2). These women have demonstrated common adaptive responses to the main drivers of change, mostly through traditional management of agro-biodiversity, networks and knowledge. Our results thus firstly confirm that there are women who especially use, conserve and manage agro-biodiversity in mountain contexts (as per Padmanabhan 2011). Secondly, we recognize that such adaptation strategies promote a female collective agency, i.e. a capacity of groups of women to make choices in order to meet their cultural and biological needs under change. These findings are consistent with other findings elsewhere (Ray-Bennett 2009; Gabrielsson and Ramasar 2013). Indeed, previous explorations in the same area (unpublished data) noted that where women have control over decision-making and agricultural tasks, they tend to cultivate more diversified fields and home gardens, capitalize on the pooling of natural and human resources and join forces in groups of collective action. This is realized including by sharing tasks in agriculture and promoting extensive networks of seed exchange that guarantee the flow of genetic material from trusted sources (as per Bellon et al. 2011 in Mexico), as well as through experimentation, conservation of genetic material and the maintenance of “safety nets”. Mostly, local varieties are conserved and exchanged, which are adapted to the local agro-ecological conditions, and have coevolved through a long process of human selection based on farmers’ preferences and traditional knowledge. Such increased ability to benefit from the unique characteristics of a range of landraces reinforces the individual and collective adaptive capacity. Nevertheless, despite these general findings on the role of women in agro-biodiversity conservation for adaptation, our quantitative analysis also indicates that gender intersects with wealth indicators in jointly shaping the decisions on how to deal with climate change (Badola et al. 2014). Therefore, as suggested by Bhattarai et al. (2015), it is not always the case that significant ecological resilience achieved through the management of ecological functions and agro-biodiversity management may be associated with gender equity. In fact, we observe that the traditional socio-economic unequal access to land and assets is still a barrier for other groups of women in the adoption of such a range of proactive strategies, increasing the vulnerability of the majority of households who face perturbations and crisis. Finally, we also highlight that age plays a role in the capacity to adopt such complementary strategies, with old men and women mainly adopting a limited number of reactive strategies related to livelihood security-based adaptation. However, similar weaknesses in our study are also evident. Further research should focus on additional factors associated with the hidden dynamics of power within groups, such as power dynamics between tenants and the landless, those related to the gendered nature of age classes or differences between migrant and non-migrant households within wealth classes. Additionally, the results leave unclear how preferences for coping mechanisms or adaptation strategies would be negotiated by younger generations in terms of their degree of participation in decision-making within the household, especially for those families where men are more likely to have migrated and roles and activities are redefined (Lambrou and Nelson 2010). Our data also suggested that gender and wealth may also depend on the context-specific biophysical and socio-economic conditions that may catalyse or constrain decisions on adaptation pathways (Fig. 4, axis 1). For instance, in the specific socio-ecological context of Bageshwar (site 2), much more connected to markets and in general to external interventions, women with more access to land and irrigation and a greater involvement in agricultural tasks favour the adoption of technological adaptation strategies. In contrast, under more marginal biophysical and socio-economic conditions, such as in Nainital (site 1), participation in training and informal networks, access to animals and diversified income, as well as awareness of multiple drivers of change, but not gender, mainly influence the priorities on management of ecological functions and proactive livelihood strategies.

As suggested by Chant (2008), stereotypes around the relationships between gender, as a homogeneous group, and the environment mask inequalities and obscure complex environmental and gender challenges. Some other interesting findings from our empirical work reinforce such an argument and evidence the dynamics of constant re-negotiation of women’s roles and responsibilities. For instance, in Bihar, our results show that a small group of women who belong to higher castes and have higher levels of education and wealth (i.e. good access to access, information, credit and animals and multiple sources of on-farm and off-farm income) are more able to re-negotiate their roles in decision-making and develop a range of proactive ecosystem-based management strategies, decreasing their vulnerability to crisis. Further, the results suggest that the participation in training and extension programmes promoted by NGOs and intergovernmental organizations and programmes has a positive influence on such adaptation process (Fig. 5, axis 1). However, this result appears to be location sensitive. For instance, higher caste women from the other site (site 2, Muzzafarpur) are neither involved in such training nor in decision-making, probably as a result of more rigid system values leading to the greater seclusion of women from public activity (including outdoor work) (Fig. 5, axis 2). By contrast, lower caste, less-educated women are less subject to such seclusion and are therefore more likely to work as agricultural labourers. These women consequently adopt mainly reactive livelihood security strategies, but, to some extent, also technological adaptation strategies, especially those related to the introduction of hybrid seeds (Fig. 5, axis 1). Our findings show lower adoption rates of ecosystem-based strategies by such group of poor women, who are net labour sellers in the area, because they are more likely than rich women, who are net labour buyers, and men to face time constraints, as a response to the increased burden of day-to-day activities as workers. This may affect their ability to participate in accessing information and community-based climate adaptation initiatives (Behrman et al. 2014) such as the ‘climate smart agriculture’ promoted by CCAFS which requires high labour inputs in the first years. As a consequence, they show a higher exposure to perturbations and crisis. In the other site (site 1, Vaishali), the access to schooling, the participation in training and the awareness of climate change are driven by external interventions and influence mainly the adoption of agro-biodiversity management and social ties and knowledge management as relevant strategies (Fig. 5, axis 2). As suggested by Mehar et al. (2016) in a recent study in the same area, our findings confirm that the awareness on a wide range of ecosystem-based strategies through extension programmes may influence their adoption. Therefore, in line with the findings of Agarwal (2000, 2009), we argue that is important to ensure the institutional integration of gender and power analysis in planning ecosystem based adaptation processes. Such integration may avoid recommendations that lead to adaptation pathways for some groups which carry potential maladaptation for others (Barnett and O’Neill 2010), thereby burdening the most vulnerable groups.

In summary, the study suggests a fragmented and complex interplay of existing and unequal gendered context-specific multiple identities and socio-economic structural differences that might shape adaptation strategies (cp. Ray-Bennett 2009; Onta and Resurreccion 2011). Access by both women and men to training and official programmes may promote divergent but complementary bundles of opportunities that households are using for adaptation. Whether this ultimately translates into the most vulnerable women contributing more to decision-making will, however, depend on how local, national and international initiatives facilitate empowerment and policy enhancement (Djoudi and Brockhaus 2011).

This paper contributes to facilitating the recognition of diverse and multiple adaptation responses and the link with gender in the context of climate change. Firstly, it demonstrates that, despite the evidence of the growing impact of climate and other socio-economic drivers, there is little recognition of geographically determined and gender-sensitive preferences and adoption of options related to ecosystem-based management, livelihoods security and technology development. Secondly, the paper offers interesting insights into the complex intersection of multiple factors which differently influence farmers’ choices on the range of adaptation options and clarifies how roles, responsibilities and power dynamics are renegotiated within the household and the community, (un)empowering women. Indeed, the empirical lessons from the more remote mountain region show the emergence of a collective agency of women to decrease common vulnerabilities to climatic variability and shocks through ecosystem-based and especially agro-biodiversity-based strategies associated with knowledge and ties management. However, the household’s capabilities (i.e. access to land, assets and diversified income), but also the age of farmers are the main barriers to the adoption of such a diversity of proactive strategies. In contrast, in the middle plains of the Indian Gangetic region a complex interaction of gender and ability to accede to training and education, depending on the specific context, the caste and the rigidity of gender norms of seclusion, mainly influences the major adoption of a mixed range of agro-biodiversity-based and technological strategies. Our results thus make clear that adaptation is not a homogenous process agreed upon by all parties and there are risks of reproducing gender bias and inequalities in development policies and interventions if it is not carefully addressed. Acknowledging the possibility of maladaptation or divergent adaptation must be a first step for researchers, policy makers and civil society to facilitate, support and fund mechanisms for adaptation at local level.

Overall, this paper also contributes to recent methodological discussions about how to conceive and implement research and development projects related to climate change adaptation under a gender transformative focus. Our work suggests how to implement a feminist intersectional approach through a two-tier interdisciplinary research approach which integrates qualitative and quantitative methods and tools. Such approach and methods should be further explored, particularly as we consider that future research could benefit from more emphasis on a nuanced analysis of the intra-gender differences that shape adaptive capacity to climate change.