Introduction

Climate change’s effects on agriculture exacerbate intricate economic, social, and political challenges globally. The interconnection between climate change, agriculture, and food security is becoming increasingly complex, leading to significant repercussions for developing nations1,2,3. As global climate change progresses, coastal systems, along with low-lying areas, will face escalating risks throughout the twenty-first century and beyond4. Tropical cyclones, the most deadly coastal disasters globally, cause significant fatalities and property destruction5,6. India has been identified as one of the most susceptible developing nations to climate change hazards7,8,9. The Indian subcontinent, with its extensive coastline of 7516 kms (including 5400 kms of mainland), has 8% of its geographical region exposed to cyclone-led climate disasters10. This heightens the threatened condition of the agrarian scenario in coastal India. Climate change is an ongoing phenomenon; without mitigating measures, its severity and impact will likely escalate in the future. In this context, adaptation has been identified by several studies as the essential policy strategy for addressing the inevitable effects of climate change11,12,13.

Despite accounting for only 7% of global cyclone activity, the impact of cyclones in India is often severe, especially when they make landfall along the northern coastline of the Bay of Bengal14,15. The Sundarban region in coastal West Bengal, India, has been designated as the “Cyclone Capital of India” by the India Meteorological Department16 because it frequently experiences intense cyclones and has a high score on the Normalised Cyclone Vulnerability Index17. Coastal West Bengal is experiencing significant shoreline erosion, with 60.50% of its coastline under threat18. The region’s sea level is rising at a rate of 4.04 ± 0.44 mm per year19. The impact of this sea level rise on the Sundarbans is exacerbated by the simultaneous land subsidence occurring at a rate of 2.9 mm per year20. Besides, it also covers around 38% of Indian coastal saline soils and ranks second after Gujarat21.

The smallholder farming community in Coastal West Bengal is predominantly dominated, and livestock plays a significant role in their livelihood. It plays a crucial role in ensuring food security, providing organic matter to the fields, and generating additional income. Indigenous cattle, which account for nearly 13% of West Bengal’s total livestock, are the most common farm animal in the region22. The consequences of cyclone-induced climate disasters and persistent climate change on livestock are intricate. The direct consequences of the cyclone disaster include significant livestock mortality due to the absence of climate-resistant shelters23, the inability to relocate animals to safer areas, scarcity of fodder24, consumption of contaminated water25, and an increased prevalence of diseases24. The primary fodder source for livestock includes crop residues, particularly those from paddy and grazing areas. However, after cyclones, salinity intrusion can severely reduce crop productivity25 and damage grazing land, thereby limiting fodder availability and extending the indirect effects over time. Although coastal salinity typically diminishes with the monsoon rains25, declining annual precipitation (Fig. 1a) limits the natural leaching of salinity from the soil. Additionally, rising average temperature and increased humidity have led to a higher temperature-humidity index (Fig. 1b), causing heat stress that negatively affects dairy productivity and reproductive performance26. Reference27 has projected a loss of 15 million tonnes of milk production by 2050 due to the consequences of climate change. It was also observed that a unit increase in THI results in a reduction of 0.42–0.67% in milk productivity of livestock on a fortnightly basis28. And it is more prevalent in crossbreds (0.61%), followed by impacts on buffaloes (0.5%) and indigenous cattle (0.4%), projected to exceed THI above 80 during April to September by 2030, increasing the vulnerability of the livestock sector in coastal regions29. Some of the pictorial depictions of the impact of climate change on livestock are shown in supplementary material (Fig. 2).

Fig. 1
figure 1

Trend analysis of (a) Annual precipitation values (b) Temperature.

Full size image
Fig. 2
figure 2

(Source: Author’s own compilations).

(a) Dilapidated housing conditions of livestock due to cyclones and floods; (b) Lack of maintenance of river dykes; (c) Increased dependency of livestock on crop residue as feed; (d) Poor animal health due to lack of fodder and grazing owing to increased salinity

Full size image

According to the report of the Intergovernmental Panel on Climate Change (IPCC), “adaptation” involves adjusting to actual or expected climate effects to reduce harm or exploit opportunities30. It was observed in the study that adoption of proper adaptation measures can minimise two-thirds of the loss in agricultural profits, potentially resulting from the ill impacts of climate change12. The adaptive capacity of farmers could be enhanced by aligning policies to provide credit, timely inputs and information, proper extension and advisory services, and other necessary support. Climate change adaptations exhibit regional variations, which are influenced by socio-economic, demographic and factors related to the farm. For instance, determinants such as age, gender, education, non-farm income, household size, location, farming experience, land size, access to credit and extension services influence adaptation strategies globally31,32,33. Reference34 has also emphasised the importance of local and indigenous knowledge of farmers in varying adaptation contexts, and thereby has highlighted its inclusion in policy discussions.

Consequently, it is essential to comprehend how local farmers interpret and perceive the effects of climate change, climate variability, and the following elements that influence their adaptation decisions and behaviours. This would enhance our understanding of how farmers adapt to the adverse effects of climate change. Despite the growing body of literature on climate change adaptation in the coastal Indian agricultural context, most studies have focused primarily on climate adaptation strategies in the crop sector, such as altering sowing dates35, crop diversification36,37, crop rotation38,39,40 or adoption of stress-tolerant varieties41,42. Despite the livestock sector being equally or more vulnerable than crops, getting directly and disproportionately exposed to cyclone-led climate disasters, there remains a distinct paucity of research on livestock-specific climate adaptation strategies, particularly in the coastal Indian context.

Existing studies that have touched upon the livestock sector have often been descriptive, cataloguing strategies without systematically analysing the determinants influencing farmers’ adoption behaviour43,44. Furthermore, adoption behaviour in the existing studies has generally been treated as a binary choice, i.e., adopt or not adopt, and thereby, questions related to the intensity of adoption were often not addressed45,46,47. In reality, farmers adopt strategies at varied levels of intensity based on their socio-economic and institutional conditions, ranging from partial to full adoption. This heterogeneity has significant implications, as partial versus full adoption can lead to substantial differences in the degree of resilience achieved. Hence, it becomes crucial to measure adoption behaviour across both dimensions, i.e., initial decision of whether to adopt (adoption decision) and subsequent decision of how much to adopt (intensity of adoption)48,49,50. Therefore, a modelling approach needs to be considered that can address the heterogeneity in farmers’ adoption, where adaptation strategies might be implemented partially or fully. In this backdrop of these thematic and methodological gaps, there remains a pressing need for empirical analysis that would (1) focus explicitly on livestock adaptation strategies in cyclone prone coastal regions, (2) simultaneously apply rigorous econometric models which is capable of disentangling the determinants influencing the dual behavioural dimensions of adoption i.e., decision of ‘whether to adopt’ and the subsequent decision of ‘how much to adopt’.

Integrating both qualitative and quantitative approaches, the study’s significance lies in two key aspects. Firstly, from a scientific perspective, it attempted to contribute to the scientific literature by identifying a crucial research gap and providing systematic evidence on livestock-specific adaptation strategies in cyclone-prone coastal regions of West Bengal. Using a qualitative approach, the study explored the importance and relevance of livestock-specific adaptation strategies to cope with cyclone-led climate disasters in coastal West Bengal. The study also methodologically advanced adaptation research by applying a double hurdle econometric approach, and attempted to disentangle the dual behavioural aspects of adoption, capturing determinants influencing this farmer’s heterogeneity more accurately than conventional models. Secondly, the study also holds firm policy and developmental implications. The identification of key determinants affecting the adoption behaviour of smallholder livestock rearers provides actionable insights for strengthening institutional services, promoting mechanisms for risk transfer such as insurance, and advocating for investments in cyclone-resilient livestock infrastructure. By combining these efforts, the results of this study will be crucial for formulating informed, evidence-based policies that bolster adaptation efforts, boost livestock resilience, and assure the overall viability of the livestock sector in the context of climate change.

Conceptual framework of the study

Adaptation decision of farmers depends upon underlying dimensions, viz., purpose (whether to prevent harm or have benefits), timing (whether it is proactive or reactive), time scale (i.e., short term or long term), as well as who acts (i.e., individual alone or with others), which determines its heterogeneity among the farmers51. A farmer’s adoption decision is a personal phenomenon, but several other factors also influence it. Empirical research suggests that the adoption and intensity of climate change adaptation strategies are affected by household socio-demographic variables and other institutional aspects31,48,49,52. Farmers’ decisions regarding technology adoption are driven by the maximisation of expected utility, which is influenced by the perceived costs and benefits associated with the technology. A multitude of variables influence producers’ preferences. The concept of climate change adaptation strategies for livestock is conceptualised in this study using the theory of utility maximisation. The potential for sustained productivity is a desirable outcome of selecting a specific strategy, as it minimises the adverse effects of climate change. Risk-averse farmers optimise their utility by selecting strategies in which the anticipated benefits of adaptation exceed the costs of adoption compared to the benefits that would be achieved without intervention. The farmer employs a strategy if the utility obtained from it surpasses that of not adopting it48. Figure 3 shows the conceptual diagram of the study.

Fig. 3
figure 3

Schematic conceptual frameworks of farmer’s adaptation behaviour.

Full size image

Methods

Study area

The study was conducted in coastal West Bengal, comprising three districts: Purba Medinipur, South 24 Parganas, and North 24 Parganas (Fig. 4). It spanned from 87° 26′ to 89° 00′ East and from 21° 35′ to 22° 36′ North. Coastal West Bengal is located between the cuspate-formed Subarnarekha delta in the west and the Hariabhanga River in the East53. South and North 24 Parganas form the Indian part of the Sundarbans, recognised as the world’s largest contiguous mangrove forest and deltaic ecosystem. Geographically, this region is bounded by the Bay of Bengal and intersected by major rivers, including the Hooghly, Rupnarayan, Haldi, and Matla. Agriculture is dominated by paddy, pulses, oilseeds, fruits, and vegetables54,55, while the livestock sector, dominated by indigenous cattle, plays a crucial role in maintaining food and income security of the region56. Persistently high temperatures and relative humidity characterise climatic conditions throughout the year, with heavy monsoonal rainfall and intermittent dry spells extending from November to April57.

Fig. 4
figure 4

Locale of study.

Full size image

Data collection

Both qualitative and quantitative methods were employed to collect data. The study is based on primary cross-sectional data, collected through a pre-tested and structured questionnaire in coastal West Bengal. Data were collected covering all three cyclone-prone districts in the region: Purba Medinipur, South 24 Parganas, and North 24 Parganas. The respondents of the study were smallholder livestock rearers, with livestock holdings of 1–5 milch animals as part of a sustenance farming system58. Primary data was collected through a multistage sampling technique. In the first stage, all three districts of coastal West Bengal were purposively selected for the study. There are six coastal blocks in Purba Medinipur district, thirteen coastal blocks in South 24 Parganas district and six coastal blocks in North 24 Parganas district59. Proportionate random sampling was employed, and one-third of the coastal blocks were selected from each district, ensuring spatial representativeness to maintain proportionality and comparability, and to reflect the distribution of coastal exposure. In the third stage, three villages were randomly selected from each block using a lottery method without replacement. It ensured the risk of selection bias and ensured diversity in socio-economic and agro-ecological contexts. Finally, from each village, 15 smallholder livestock rearers who met the study’s criteria were selected as respondents, resulting in a total sample size of 360 smallholder livestock rearers.

The required sample size was estimated using Cochran’s formula (Eq. 1), which is used when the population is large or infinite60,61.

$$n = \frac{{Z^{2} *p*\left( {1 - p} \right)}}{{e^{2} }}$$
(1)

where Z = confidence level, e = margin of error, and p = assumed population proportion.

The data were collected from respondents using a structured questionnaire during face-to-face interviews. Additionally, focus group discussions were conducted to gather information on farmers’ knowledge of climate change, its related hazards, adaptation strategies they follow, and the role of local institutions. Climate data was obtained from gridded data of the India Meteorological Department (https://cdsp.imdpune.gov.in) with the help of a Python package ‘IMDLIB’62, and NASA Power (https://power.larc.nasa.gov/).

Adoption and intensity of adoption

The adoption decision was measured as a binary outcome, with a value of 1 indicating adoption and 0 indicating non-adoption. Specifically, if a farmer chose to implement any identified adaptation strategy, they were categorised as having adopted the strategy (scored as 1); otherwise, they were categorised as not having adopted (scored as 0).

Additionally, the study aimed to assess the intensity of adoption among individuals who had adopted at least one strategy. This was operationalised by distinguishing between partial and complete adoption of adaptation strategies. The importance of adoption was measured using scores, where partial adoption was assigned a score of 1 and full adoption a score of 2. For instance, if a farmer partially adopted strategies A, B, C, and D, and fully adopted strategies E, F, and G, their total adoption score would be calculated as 10 (with 1 point for each partially adopted strategy and 2 points for each fully adopted strategy). The dependent variable for the analysis of the determinants that influence the intensity of adoption among farmers was this scoring system, which ranges from a minimum of 1 to a maximum of 14. The conceptual distinction between partial and full adoption in this study is grounded in Diffusion of Innovation Theory63. It described adoption as a sequential process extending over a period of time, from awareness of a technology to its final adoption. A partial adopter was thereby operationalised as households that adopted a strategy only to a limited extent. In contrast, a full adopter was operationalised as households that comprehensively adopted a strategy for their entire livestock unit. This operational conceptualisation was consistent to Ref.64.

The double hurdle model was used in the study as it better reflected the two distinct stages of adoption behaviour of farmers48,49. Alternative models, such as the Poisson or the ordered probit model, would not have separated these two stages and thereby might overlook the practical realities of farmers’ adoption behaviour.

Model specification

An analysis of the adoption behaviour of climate adaptation strategies for livestock involves a scenario in which an event may or may not occur at each observation. A continuous, non-negative variable is associated with the occurrence of an event (adoption), whereas a variable with a value of zero is the result of non-occurrence (non-adoption)65. A continuous distribution over positive values is feasible when the event transpires. However, an accumulation at zero exists (as a result of non-occurrence), which serves as a corner solution to the adoption problem66. Recent studies48,49,67 had suggested the use of Cragg’s double hurdle model65 in such scenarios, where separate vectors of independent variables influence the farmers decision of adoption. Following the assumption of the double hurdle model, farmers’ adoption was analysed in a two-step approach. First step postulating the adoption decision of the farmers, i.e., ‘whether to adopt’, and second step measuring the intensity of adoption, i.e., ‘how much to adopt’. The model is a generalisation of Tobit, in which two distinct stochastic processes determine quantitative decisions and participation. Compared to the Tobit model, the double hurdle model, as proposed by Ref.65 provides a more adaptable framework. This method enables the independent estimation of the adoption decision and intensity of adoption68. The model entails the sequential or simultaneous implementation of probit and truncated regression models49. The probit model serves as the theoretical foundation of the double hurdle estimation framework. The initial obstacle is represented by the probit model, which calculates the likelihood of a farmer implementing any climate adaptation strategies. The second obstacle, the intensity of adoption, is a non-negative decision that can only be quantified for non-zero adoption levels. As a result, it is estimated using a truncated regression model.

The double hurdle model is expressed as;

(a) First hurdle: decision to adopt:

$$\begin{aligned} & d_{i}^{*} = {\varvec{z}}_{{\varvec{i}}}^{\prime} {\varvec{\alpha}} + {\varvec{\varepsilon}}_{{\varvec{i}}} ; \\ & {\varvec{d}}_{{\varvec{i}}} = \{ {\mathbf{1}}\;\;{\varvec{if}}\;\;{\varvec{d}}_{{\varvec{i}}}^{{*}} > {\mathbf{0}};\;{\mathbf{0}}\;{\varvec{otherwise}} \\ \end{aligned}$$

(b) Second hurdle: intensity of adoption:

$${\varvec{Y}}_{{\varvec{i}}}^{*} = {\varvec{x}}_{{\varvec{i}}}^{\prime} {\varvec{\beta}} + {\varvec{\mu}}_{{\varvec{i}}} \mathop \int \limits_{{{\varvec{y}}_{{\varvec{i}}} = 0,\;{\varvec{Otherwise}}}}^{{{\varvec{y}}_{{\varvec{i}}} = {\varvec{y}}_{{\varvec{i}}}^{{*}} \;{\mathbf{if}}\;{\varvec{y}}_{{\varvec{i}}}^{{*}} > {\mathbf{0}}\;{\mathbf{and}}\;{\varvec{d}}_{{\varvec{i}}}^{{*}} > {\mathbf{0}}}} {\mathbf{0}}$$

where \(d_{i}^{*}\): a latent variable that defines the decision to adopt, \(d_{i}\): observed household decision to adopt and takes a value of 1 if the farmers adopt and zero otherwise, \(y_{i}^{*}\) = Latent variable describing the intensity of adoption, \(y_{i}\): observed response on intensity of adoption, z and x are vectors of variables explaining the decision to adopt and intensity of adoption respectively (Table S1), α and β are vectors of the parameters to be estimated, εi are μi are the respective error terms assumed to be independent and normally distributes as εi ~ N (0,1) and μi ~ N (0, σ2).

All methods used in the study were carried out in accordance with relevant guidelines and regulations. The research procedure and survey schedule were approved at both the departmental and institutional levels by the Division of Dairy Extension, ICAR-National Dairy Research Institute, Karnal, India. Informed verbal consent was obtained from all respondents before data collection. Respondents were also informed about their voluntary participation and the confidentiality of their information regarding the study. No minors were included as respondents in the survey.

Results

Descriptive statistics of smallholder farmers in the study area

The demographic and socioeconomic characteristics of the surveyed households in the coastal villages of West Bengal are described in this section. The descriptive statistics of the households are shown in Table 1.

Table 1 Descriptive statistics of socio-economic characteristics of surveyed households in coastal West Bengal.
Full size table

Regarding the sample data, it portrayed a mid-aged population who were engaged in agricultural activities in the region. The mean age of the respondents was found to be almost 53 years, indicating an ageing population involved in farming practices in the study locale. The eldest farmers among the respondents were 82 years old, while the youngest was 24. The family education status for most (47.50%) of the respondents’ households was in the group of eight to ten years of formal schooling. About 39.72% of the respondents had operational land holding between 0.2 and 0.5 ha, followed by 32.77% of the respondents having their operational land holding between 0.5 and 1 ha. The average landholding size was found to be 0.622 ha among the respondents. The study found that all surveyed households owned livestock; however, the number of livestock holdings varied. Coastal West Bengal is majorly dominated by indigenous cattle in the livestock sector. The average holding of indigenous cattle among the respondents was 2.156. Compared to indigenous cattle, the holding of crossbred cattle was found to be much lower. Mean holding of crossbred cattle was only 0.317. Mean holding of small ruminants (goats and sheep) among the surveyed households was 1.861. Combining all types of livestock, the mean Standard Animal Unit was 1.930.

Adaptation strategies adopted by the farming community

The study identified seven prominent strategies adopted by the livestock rearers of coastal West Bengal against climate extremes, such as cyclones and climate variabilities, such as erratic precipitation and an increase in temperature. They included improved feeding practices, shifting from large ruminants to small ruminants, availing of insurance, climate-proof shelter, moving animals to a safe place during a disaster and preservation of fodder (Fig. 5). On average, respondents implemented 3.04 adaptation strategies, with a standard deviation of 2.05.

  1. (a)

    Improved feeding practices Under intense unavailability of proper fodder and water after cyclone incidents, livestock come under high risk25. The risks multiply with an increase in the frequency of heat waves, leading to heat stress. These increases in climate extremity and uncertainty led to metabolic and behavioural changes in the livestock of coastal West Bengal. This included negative energy balance due to loss of sodium and potassium with excessive sweating, reduced feed intake, an increase in disease proneness, and thereby limiting milk production44. Conjointly, an increase in energy deficits further led to a decline in cattle fertility, fitness and longevity. Approximately 49.17% of the respondents had adopted improved feeding practices. It included the provision of extra concentrates to the animals, as well as the addition of mineral supplements such as common salt, mineral mixture, oilcake, and boiled crop residues to the diet (Figs. S1 and S2). Inclusion of mineral mixture in the diet helps in maintaining the mineral profile of animals and helps in increasing the utilisation of nutrients, mineral bioavailability and milk production69. Providing extra concentrates to the animals helped maintain milk productivity and alleviate heat stress.

  2. (b)

    Shifting from large ruminants to small ruminants Majority of the respondents adopted shifting from large ruminants to small ruminants as an essential climate adaptation strategy in coastal West Bengal (Fig. S3). A similar trend was also observed by several other studies44,59. The increase in climate vulnerability has also resulted in increased economic losses due to the death of a large number of cattle during disasters, an increase in disease incidences, a feed and fodder crisis, and an unorganised market. This had gradually initiated a shift from large ruminants to small ruminants, specifically goats (mainly the Black Bengal breed) and sheep (primarily the Garole breed). The breeding tracts of both breeds lie in the coastal Sundarban region, which has the capacity to adapt and survive in difficult environments70. The shifting facilitated livestock rearers in diversifying their risks, reducing huge monetary losses, and enabling them to quickly convert their assets into cash, as the animals contributed to the working capital of the farmers43.

  3. (c)

    Availing of livestock insurance The study found 43.89% of the respondents to avail livestock insurance as a climate adaptation strategy. Livestock insurance is acclaimed as a promising mechanism for enhancing resilience of livestock rearers to climate change71 and decreases their vulnerability towards climate change72. Insurance of livestock helped the rearers minimise their losses, particularly due to the death of animals after the disaster (Fig. S4). Insurance is an external policy intervention being introduced into a diversity of rural community adaptation strategies that are crucial components of resilient livelihood development73. However, it was also observed that awareness and knowledge among farmers regarding livestock insurance were low, resulting in a relatively lower adoption rate. Therefore, it is recommended that the government and other stakeholders advocate policies and programs that will increase awareness and evaluate the actual demand for the attributes that farmers prefer in livestock insurance74.

  4. (d)

    Well-ventilated housing Farmers perceived the need for well-ventilated housing to remove moisture and odours from the shed and replace them with fresh air. Coastal West Bengal also faces severe heatwaves during the summer, which often result in animals experiencing heat stress. Proper ventilation of the animal shed would help in creating a comfortable microclimate inside44,75. The study found 49.17% of the respondents had adopted this climate adaptation strategy. Farmers resorted to the use of electrical fans to facilitate cross-ventilation and maintain an optimum Temperature Humidity Index during the summer. Cross ventilation of air positively impacts maintaining hygiene in the animal shed by helping to remove microorganisms, dust, and harmful gases, thereby reducing the occurrence of animal diseases59. Figures S5, S6 and S7 show examples of improved housing adopted by smallholder livestock rearers in coastal West Bengal.

  5. (e)

    Moving of animals to a safe place during disaster Livestock rearers used to shift their animals to safer places during disaster incidents such as cyclones and floods76. Smallholder livestock rearers, who lacked proper climate-proof shelters to keep their animals safe during disasters, had to relocate their animals to safer areas, such as cyclone shelters or elevated regions, during floods. This was done to prevent animal deaths and reduce the substantial monetary losses associated with them. This strategy was reported to be adopted by 46.39% of the respondents.

  6. (f)

    Preservation of fodder As mentioned earlier, crop residues of paddy and grasses maintained the fodder security of the region. However, during the disaster, both elements were seriously affected, resulting in a severe fodder crisis43. Fodder scarcity was also witnessed during drier months. To prevent this, farmers used to store fodder to meet unforeseen circumstances. Paddy straw was stored by farmers for a continuous supply of dry matter, even during crises (Fig. S8). However, the adoption rate of this climate adaptation strategy was found to be low (25.56% of the respondents). Farmers reported a decline in paddy production in the region, and a gradual shift towards a vegetable and fisheries-based cropping system had also reduced the availability of paddy residues. Lack of infrastructure for storage and preservation of fodder was also one of the main reasons77. Instead of fodder preservation, farmers are aligning with the “buying of fodder from market” when it is needed or “shifting from large ruminants to small ruminants” to reduce maintenance costs.

  7. (g)

    Providing more healthcare practices to the livestock Para-veterinarian workers had the leading role to play in the provision of healthcare practices to the livestock in coastal West Bengal. With an increase in the intensity of cyclones and climate change in recent years, the biotic stress of animals has been reported to be very high43. It impacted their livestock productivity and sometimes even led to the death of animals. Thus, the provision of health care practices should be considered as an essential climate adaptation strategy44,78. However, the lack of sufficient veterinary institutions, coupled with inadequate road connectivity (particularly in the islands), was a reason for the lesser adoption (39.17% of the respondents) of this adaptation strategy in coastal West Bengal.

Fig. 5
figure 5

Adoption status of different climate adaptation strategies for livestock in coastal West Bengal.

Full size image

Determinants of smallholder livestock rearers’ decision on adoption of climate adaptation strategies

The result presented in Table 2 shows the first hurdle (probit model) explaining the determinants influencing the livestock rearers’ adoption decision, i.e., ‘whether to adopt’ any adaptation strategies. A farmer adopting any adoption strategy was scored as ‘1’, and if not, it was scored as ‘0’ to measure the adoption decision. The model was found to be statistically significant at 1% level of significance, indicating good fit. Multicollinearity was measured using the variance inflation factor (VIF), and the calculated VIF values for all the variables were below the threshold of 10. The mean VIF was found to be 2.64 (Table S2).

Table 2 Determinants of decision on adoption of climate change adaptation strategies: probit model.
Full size table

Table 2 revealed that the availability of climatic information influenced the respondents’ adoption decision (i.e., the first hurdle) at p < 0.01, indicating the significant importance of this service in farmers’ adoption behaviour. Herd size was found to have a positive significant influence on the adoption decision at p < 0.1, indicating that households with larger livestock holdings were more likely to adopt at least one strategy. Community participation showed positive significant influence on adoption decision of the respondents at p < 0.05. This might be attributed to the diffusion effect of information and strategies among the community members. Significant negative influences on the adoption decision were observed in the availability of non-institutional credit (p < 0.05), infrastructural availability (p < 0.01), and income from the livestock sector (p < 0.05).

Determinants of smallholder livestock rearers’ intensity of adoption of climate adaptation strategies

The results presented in Table 3 show that the second hurdle (truncated regression model) explains the determinants that impact the intensity of adoption of adaptation strategies by livestock rearers, specifically partial adoption or full adoption. Intensity of adoption was measured using a continuum, where for every adaptation strategy, a continuum, viz., fully adopted (2), partially adopted (1), was used. This was followed by the summation of scores for all identified strategies for each respondent. The summation score of the intensity of adoption of all seven adaptation strategies was considered the dependent variable, with a score ranging between a minimum of 1 and a maximum of 14. The model was found to be significant at p < 0.01.

Table 3 Determinants of intensity of adoption of climate change adaptation strategies: truncated regression model.
Full size table

Results revealed that herd size, community participation, and access to livestock extension services influenced the intensity of adoption among smallholder livestock rearers at p < 0.01. Availing institutional credit influenced the intensity of adoption at p < 0.05, and training received by the rearers influenced the intensity at p < 0.1. The result indicated the impact of strengthening institutional services and financial inclusivity in increasing the level of adoption of climate adaptation strategies by smallholder livestock rearers in coastal West Bengal.

Discussions

Determinants influencing the adoption decision

The herd size of the rearers, which was expressed in terms of the standard animal unit, was found to have a statistically significant positive relationship with the adoption of climate adaptation strategies at p < 0.10. This could be attributed to the fact that farmers with larger herd sizes tended to invest in adaptation strategies. With an increase in herd size, income opportunities also increased. Additionally, it increased the risk of being affected by both cyclones and ongoing climate change, resulting in a decline in income from the livestock sector and an increase in costs and financial losses. Thus, to minimise the risk posed, rearers tended to adopt climate adaptation strategies for livestock. This finding was consistent with Ref.47, who reported significant positive relationship of herd size with adaptation strategies viz. ‘keeping local breeds’, ‘change in micro-climate in the shed’ and ‘shifting from large ruminants to small ruminants’. The findings also agreed with Ref.79, who reported a significant and positive impact of cattle herd size on the adoption of ‘concentrated livestock feed’, ‘forage cropping’, ‘herd destocking’ and ‘transhumance’. Livestock ownership positively influenced the adoption of climate adaptation strategies for crops, such as ‘changing planting date’, ‘tree planting’ and ‘integrating crop with livestock’. The result was also consistent with80.

The availability of infrastructure was found to have a significantly negative relationship with the adoption decision of livestock rearers at p < 0.01. Although this result contradicts the general expectation that better infrastructure facilitates adoption, it may be explained by a substitution effect. Households located in areas with relatively better physical infrastructure, such as roads, dykes, market facilities, or veterinary infrastructure, would have a lower adoption of strategies like shifting from large to small ruminants, preserving fodder, and relocating animals to safe places during disasters. Their risk management and livelihood support may depend on external facilities available due to better infrastructure. Conversely, the absence of infrastructure created conditions where farmers were compelled to adopt these adaptation strategies to buffer against risks. Occurrences of recurrent cyclones caused huge infrastructural losses in coastal West Bengal, heavily impacting the socio-economic aspects of the smallholder farming community of the region. Loss of infrastructure included loss of road connectivity, discontinued power supply, damage to houses, storage structures, and input availability. Damage to dykes and embankments catalysed the exacerbation of the infrastructural loss in the region. It directly disrupted the feed and water supply of the animals, destroying their shelters, disrupting livestock healthcare services and leading to increased disease incidences and even deaths of the animals. To avert the risks and losses caused by infrastructural damage due to climatic disasters, farmers in poorly serviced or disaster-affected areas have decided to adopt climate adaptation strategies. A similar result, although in a different context, was observed for Ref.82, which reported a negative influence of infrastructure on the adoption decision related to e-commerce for agricultural products.

Provision of adequate information on seasonal and long-term climatic changes, as well as agro-met advisory services, decreases the downside risks of failure to adopt new technologies and adaptation measures81. The availability of climatic information services was found to have a significant and positive relationship with the adoption decision of livestock rearers at p < 0.01. Climate information services helped farmers adopt immediate measures, such as shifting animals to a safe place or preserving fodder, to protect them from the impacts of climate change or disasters. This allowed them to engage in contingency planning to enhance buffer capacity during disaster occurrence and build resilience. The findings were consistent with Ref.82,83,84,85. Reference86 found climate services for Murrah buffalo farmers to be a potential adaptation tool to enhance the resilience capacity of vulnerable dairy farmers to climate change. Climatic information services and agro-met services were provided to livestock rearers from various sources, including line departments, local government bodies, research institutes, and cooperatives. The respondents also received information from multiple mass media and social media. Information was disseminated among farmers through their social networks, which consisted of farmers’ organisations and fellow farmers. The social network had a role in enhancing the likelihood of the adoption decision of farmers87,88,89. Following similar results, the study found that community participation among farmers has a positive and significant relationship with the adoption decision of livestock rearers at p < 0.05.

The study found that income from the livestock sector had a negative and significant impact on the adoption decisions of the rearers. This may be attributed to the fact that the majority of the smallholder farmers practised an integrated farming system, with livestock as an important component. However, farmers reported that the income from the sector did not contribute a major portion of their income due to low livestock productivity, the unavailability of an organised market, and a lack of infrastructural availability, such as storage, preservation, and dairy cooperatives. And with an increase in climate sensitivity over time, income from the sector had deteriorated consecutively. In contrast, livestock rearers with comparatively high income from the sector exhibited greater risk-bearing ability and more financial resilience, which reduced their perceived vulnerability and, consequently, their likelihood of adopting new adaptation measures. This finding suggests that better-off households may rely on existing income as a buffer rather than adopting adaptation strategies. In contrast, families with lower livestock income are more vulnerable to climatic risks, and the adoption of adaptation strategies becomes a necessity to secure their livelihoods. The findings agree with Ref.46,72,90. In the same context, the study also found that availing non-institutional credit has a negative and significant impact on the adoption decision of farmers at p < 0.05. This may be because smallholder farmers availed non-institutional credit during the time of emergency, when there was a void for institutional credit91. However, due to the nature of informal credit systems, characterised by high interest rates of non-institutional credit and its unregulated sector, farmers often find themselves in a weaker position due to shorter repayment periods. This decline reduced their risk-bearing abilities, and they were unable to adopt any adaptation strategies92. Farmers who depend on moneylenders or other informal sources typically use the borrowed money for immediate consumption needs or debt repayment, rather than investing in long-term, climate-resilient practices. Although this study did not find any significant influence of availing institutional credit on the adoption decision, Ref.49, found an essential influence of credit on the adoption decision. Availing of non-institutional credit also occurs when institutional credit is unavailable.

Determinants influencing the intensity of adoption

Along with influencing adoption decisions, herd size, expressed as standard animal units, also impacted the intensity of adoption of adaptation strategies and had a significant positive relationship at p < 0.01. Livestock rearers with larger herd sizes faced greater climate risks and vulnerability from all aspects, ranging from the unavailability of fodder and shelter to declining animal health and, in some cases, even the death of animals. This puts a huge monetary deficit on the rearers. To minimise risks and avoid disastrous consequences, rearers may consider adopting more adaptation strategies.

Resource scarcity and a lack of capital among the smallholder farming community in the coastal region have long been a major constraint to the adoption of new technologies. The presence of institutional credit in this context would facilitate the adoption of climate adaptation strategies for the livestock sector. The findings showed that availing of institutional credit had a positive and significant impact on the intensity of adoption of adaptation strategies at p < 0.05. The findings agree with Ref.48,93,94,95,96,97. Utilising institutional credit helps increase the financial stability and risk-bearing capacity of farmers, and enhances the adoption of climate adaptation strategies98,99. Adoption of adaptation strategies could be capital-intensive, with some strategies requiring investment, such as building a well-ventilated house and providing improved feed. Therefore, it may be challenging for them to adopt adaptation strategies, even with access to all the necessary information and technical expertise, due to a lack of capital.

The intensity of adoption of climate adaptation strategies by livestock rearers was found to be positively and significantly influenced by the training they received, at p < 0.10. The findings were consistent with100,101,102,103. Training helped in improving the knowledge, attitude and skills of the farmers related to the use and implementation of any adaptation strategies104,105,106. It played a role in instilling confidence among livestock rearers and changing their behaviour towards the adoption of adaptation strategies. Trained farmers were also mentioned as an excellent medium for the dissemination and adoption of climate-smart practices through their social networks107. As mentioned in the earlier section, diffusion of information through social networks influenced the adoption decision of the farmers. Similarly, it also played a role in influencing the intensity of adoption of adaptation strategies by farmers. The study found community participation had a positive and significant relationship with intensity of adoption at p < 0.01. The findings agree with Ref.108, which reported that farmers who participated in collective action groups had a higher level of climate adaptation.

Similar to the training received, access to extension services plays a pivotal role in disseminating technologies to farmers and increases the likelihood of the intensity of adoption of climate adaptation strategies109,110. The study found a positive and significant relationship between access to extension contacts and the intensity of adoption at p < 0.01. The findings were consistent with Ref.46,102,103,111,112,113,114,115. Besides the dissemination of technologies, extension services included the provision of agro-met services, climate forecasting, and conducting training programmes and workshops to enhance knowledge, skills, and attitudes. Extension services were provided to livestock rearers by various sources, including line departments, research institutes, and KVKs. Additionally, agricultural input dealers, private input companies, and NGOs provide extension advisory services to farmers related to the use of adaptation strategies and improved livestock management. However, farmers had a significant dependency on local agricultural input dealers as their most connected source of extension advisory services. Researchers also observed that the provision of extension advisory services exclusively for the livestock sector was comparatively less than extension services for the crop sector in coastal West Bengal. Ref.111 also reported that farmer-to-farmer extension is an essential determinant of adaptation strategies.

Conclusion and recommendations

The study was undertaken with the dual objective of (1) identifying and exploring the climate adaptation strategies adopted by the smallholder livestock rearers in climatically vulnerable and cyclone-prone region of coastal West Bengal, (2) analysing the determinants influencing the adoption behaviour of the smallholder livestock rearers including both the dimension of adoption decision (whether to adopt?) and intensity of adoption (how much to adopt?). A cross-sectional survey was conducted, and primary data were collected from all three districts of coastal West Bengal using a multi-stage sampling approach. Data were analysed using a double hurdle econometric approach, which helped separate the estimation of the adoption decision from the intensity of adoption.

Smallholder farmers in coastal West Bengal adopted several climate adaptation strategies for livestock, including improved feeding practices, shifting from large to small ruminants, livestock insurance, ventilated housing, moving to a safe place during disasters, fodder preservation and better healthcare facilities for livestock. Among these, shifting from large to small ruminants emerged as the most widely adopted strategy, while fodder preservation was the least adopted. The likelihood of adoption was positively associated with herd size, access to climate information, and community participation, while it was negatively influenced by reliance on non-institutional credit, availability of infrastructure, and income from livestock. In contrast, the intensity of adoption was significantly enhanced by access to institutional credit, training, extension services, and community participation, in addition to herd size.

The results of this study have important policy relevance for enhancing climate change adaptation among smallholder livestock rearers and strengthening climate-resilient livestock management in cyclone-prone coastal West Bengal.

  • First, the strong influence of climate information underscored the immense need for timely, reliable, and locally contextualised advisory services. Climate services that integrate short-term forecasting (e.g., cyclone warnings) with long-term advisories (e.g., feeding practices under heat stress) should be institutionalised and disseminated through ICT platforms, extension networks, and community-based systems.

  • Second, access to institutional credit was a key enabler of both the adoption decision and the intensity of adoption. Enhancing institutional credit coverage and reducing dependency on high-interest loans can strengthen farmers’ financial resilience and investment capacity. This requires targeted financial inclusion policies and credit-linked adaptation packages.

  • Third, the study also revealed the limited reach of livestock extension services, which remain biased toward crop agriculture. Expanding livestock extension services, along with training para-veterinarians and input dealers, could cater to the needs and interests of the farmers. Adopting a participatory approach, from planning to knowledge co-production and an iterative feedback mechanism, can facilitate the adoption by farmers.

  • Fourth, infrastructural availability was found to influence adaptation decisions negatively. This indicated that farmers in remote and infrastructure-constrained areas had to rely heavily on self-driven adaptation strategies. Thus, climate policies should consider investment in climate-resilient and inclusive infrastructure. It should focus on elevated livestock shelters, community-based fodder banks, mobile veterinary units, and localised markets for livestock products.

  • Finally, scaling up livestock insurance would provide a crucial safety net against cyclone-led climate disasters and increase the risk-bearing ability of livestock rearers in coastal West Bengal.