Fishery livelihoods and an adaptation to climate change-induced threats at the Bhomra Wetland: a case study through a stakeholder-driven approach

Floodplain wetlands are one of the most productive but vulnerable ecosystems that offer various commodities and services (Winter 2000; Sarkar et al. 2021a) to their local stakeholders. India is bestowed with 0.5 million ha of floodplain wetlands, distributed especially in the River Ganga and Brahmaputra Basins. These wetlands cater to fish production besides providing other ecological goods and services that benefit millions by providing food, nutritional and livelihood security. However, these resources are threatened by anthropogenic factors and climate change, resulting in degraded ecosystem health, loss of biodiversity and decreased fish production. The wetlands of West Bengal are typical in their characteristics, as they are part of India's largest floodplain wetland system. The state has a sizable amount of wetland, i.e., 504,949.06 ha, with 5.86% of its physical area scattered across the Districts of Cooch Behar, Hooghly, Malda, Midnapore, Murshidabad, Nadia and North 24-Parganas (Vinci et al. 2000). In broad ecological aspects, the wetlands of this region can be categorized into two main types: freshwater and brackish-water. Within West Bengal, Nadia District has the largest area in terms of the coverage of the wetlands (3.13% of the total district area). Culture-based fisheries (CBF) management through a cooperative society is widely practiced in the wetlands of West Bengal. CBF is effective for fishery enhancement and conservation.

The effects of extreme climatic events like unpredictable shifts in rainfall patterns, recurrent floods, droughts and cyclones are the major setbacks for ecology, fisheries and thus the livelihoods of associated people (Sarkar & Borah 2018). The extreme climatic events render these ecosystems more vulnerable (Dhanya & Ramachandran 2016; Uddin et al. 2019; Tasnim et al. 2023). It is anticipated that the fluctuations in the global mean surface temperature will have an impact on tropical developing countries like India (Rathore et al. 2013; IPCC 2021). Just within one decade (years 2011–2022), the current global surface temperature has risen to +1.09 °C (IPCC 2021). The impacts are evident in India, as manifested by an increase in heavy rainfall events and a decrease in low and medium rainfall events (Goswami et al. 2006), along with altered geographic distribution, loss of species, change in breeding period, life cycle and physiological behaviour of inland fishes (Sharma et al. 2015; Sarkar & Das 2021). According to Naskar et al. (2022) and Das et al. (2019), the status of wetland fisheries in the concerned area was severely vulnerable. Although Sarkar et al. (2016, 2018) have discussed some scientifically established climate-smart adaptive strategies, the scientific reports on assessing the overall health and importance of tropical floodplain wetlands in India are scanty and poorly understood (Pratyashi & Ranjan 2014). Also, there is a dearth of comprehensive studies and literature on the spatiotemporal change analysis of floodplain wetland resources, although ecological degradation, encroachment and shrinkage are concerning in these sensitive ecosystems. In order to mitigate the climate change-driven scenario in wetland ecosystems, there is an urgent need to produce region- or location-specific studies to develop advanced generation adaptation strategies.

To date, a number of studies have been carried out in the Bhomra Wetland focussing on its ecology (Das 1998; Sugunan et al. 2000; Pramanik & Nandi 2004; Biswas et al. 2005; Roy et al. 2016a, 2021), fishery practices and fish diversity (Samanta et al. 2005; Nandi & Pramanik 2013; Samadder et al. 2016; Ghosh et al. 2018; Sandhya et al. 2019) and socio-economics of the wetland users (Biswas et al. 2005; Biswas et al. 2010; Roy et al. 2012, 2016b). Some studies were also made to identify the vulnerability of the fisheries associated with the wetland (Naskar et al. 2018; Sarkar et al. 2020, 2021). Recently, the prokaryotic community structure of the concerned wetland was studied by Kumari et al. (2021). However, a consolidated study comprising the climate, ecology and fisheries of the selected wetland is lacking to date. The present study is the solution to this research gap, combining stakeholder perception analysis with an assessment of the current ecological and fisheries status. The incorporation of stakeholders’ perceptions is very important or sometimes the only way to assess critical links between long-term climate change and ecosystems where systematic historic data is not available (Naskar et al. 2018; Schulz et al. 2019; Benansio et al. 2022).

The aim of the study was to systematically assess the effect of a climatic anomaly on the wetland, the current status of the ecology, including fish species abundance, diversity, richness and evenness pattern of the wetland. The study also attempted to utilize stakeholders’ perceptions on climate change, ecosystem health and services to assess the impact of changing climates on wetland fisheries. The paper also recommends some of the climate-resilient adaptive strategies for enhancing the adaptive capacity of fishers.

Primary data were collected from March 2020 to February 2021. Fish and water samples were collected in triplicate from each site on a monthly basis (Sarkar et al. 2020). Water temperature (WT), pH, salinity, total dissolved solid (TDS) and conductivity were determined with the multi-parameter Testr™ 35 series (OAKTON). The depth of the water column was estimated through the Hondex™ digital depth sounder. Transparency was measured using a 20 cm diameter standard black-and-white Secchi disk. Dissolved oxygen (DO) was measured by a digital DO meter (Ultron DO5510). Total alkalinity (TA), hardness, available phosphate (AP), nitrate (N), free carbon dioxide (FCO₂), chlorophyll-a (Chl-a) and primary productivity (gross primary productivity: GPP and net primary productivity: NPP) were measured using standard methods (APHA 2012). A questionnaire (Supplementary material, Appendix 1) comprising 18 different items (broadly classified into independent and dependent items) in relation to the impact of climate change and anthropogenic stress upon ecosystem services, ecosystem health and locally pursued adaptive techniques in wetland fisheries was designed for the collection of stakeholders’ perceptions (Naskar et al. 2022; Sarkar et al. 2022). Data were collected through surveys and face-to-face interviews with the stakeholders. A total of 60 members who were directly engaged with the fisheries' activities for more than 20 years (age > 40 years) were considered the population of the study during the years 2020–2021. It was found that the persons in the selected age group knew better about wetland fisheries owing to large experience (>20 years). The responses were recorded on a five-point Likert scale (1 = strongly agree, 2 = agree, 3 = undecided, 4 = disagree and 5 = strongly disagree). Only complete surveys were included in the analysis.

Time series grid data (annual) of the selected wetland in the form of temperature (1° x 1°) and rainfall (0.25° x 0.25°) for a period of 20 years (2001–2020) were collected from the Indian Meteorological Department (IMD), Alipore, West Bengal, India (Srivastava et al. 2009; Pai et al. 2014). Finally, the annual fish production data were collected from the records of the wetland's fishermen cooperative society.

The Mann–Kendall test is a non-parametric test recommended by the World Meteorological Organization (2018) to determine a trend in climate (rainfall and temperature) data within a specific time range. Sen's slope is part of the trend detection process with the Mann–Kendall test. It has an advantage over linear regression as it is not affected by the number of outliers present within the data (Aditya et al. 2021). With the Sen's slope, a 3-year moving average was shown for better visualization of the climate data.

The Shapiro–wilk test was chosen for the normality test when the numbers of the sample were <50. Pearson correlation was used to determine the direction and magnitude of the relationship between ecological parameters, which followed a normal distribution. A two-way analysis of variance (ANOVA) followed by Tukey's post hoc test was carried out to estimate the effect of both the season and sampling site on the collected data. Eta squared (η²) represents the amount of variance associated with each main effect and interaction effect in an ANOVA model.

Fish diversity data were first recorded and then arranged according to the taxonomic orders. These data were presented as per the abundance of the species at different sites. Site-specific fish species diversity and evenness were also calculated following standard calculations (Table 1). Canonical Correspondence Analysis (CCA) at a significance level of p < 0.05 using PAST (version 3.26) was performed to explore the linkage between environmental factors and fish community structure (Toham & Teugels 1999). To assess the relative importance of each site, we used the CCA on each site, linking fish abundance and environmental parameters. The CCA was applied to the overall fish data matrix and environmental data matrix, obtaining a direct environmental interpretation of the extracted ordination axes. The annual fish production data were also presented as a time series with a linear trend line and a 3-year moving average.

Table 1

Different species diversity and evenness indices used in the study

Index .	Equation .	Reference .
Shannon–Weiner diversity index (HI)		Weaver & Shannon (1949)
Simpson diversity index (DSim)		Whittaker (1965)
McIntosh diversity index (DMC)		McIntosh (1967)
Menhinick diversity index (DMn)		Whittaker (1977)
Margalef diversity index (DMg)		Margalef (1958)
Pielou evenness index (JI)		Pielou (1975)
McIntosh evenness index (EMC)		McIntosh (1967)

Index .	Equation .	Reference .
Shannon–Weiner diversity index (HI)		Weaver & Shannon (1949)
Simpson diversity index (DSim)		Whittaker (1965)
McIntosh diversity index (DMC)		McIntosh (1967)
Menhinick diversity index (DMn)		Whittaker (1977)
Margalef diversity index (DMg)		Margalef (1958)
Pielou evenness index (JI)		Pielou (1975)
McIntosh evenness index (EMC)		McIntosh (1967)

View Large

The acquired data from the stakeholders’ perceptions were collectively subjected to a reliability test (determination of Cronbach's alpha >0.70). All reliable data were then tested for normality (Kolmogorov–Smirnov test) and analysed accordingly. The ordinal logistic regression-proportional odds model (cumulative logit model) was used to find any relation between the dependent and independent variables. Unlike the linear regression model, this method has the benefit of treating the outcome variable as an ordinal and not a continuous variable (Lombardo et al. 2018). Ordinal regression analysis was performed for the analysis of the questionnaire. All the statistical analysis, unless stated specifically, was done through SPSS 26.00.

The water depth varied (1.79–2.78 m) seasonally (p = 0.00, F = 98.766, η² = 0.804). Depth can significantly affect effective water volume. Most wetlands in West Bengal are vulnerable to water stress and the relatively low rainfall is the prime cause for water balance-related problems in the closed wetlands (Mukhopadhyay 1997). The DO varied (7.20–9.7 mg/L) spatially (p = 0.00, F = 475.78, η² = 0.906) (Figure 3). It was less in the shallow sampling site. DO positively correlated (p < 0.05) with nitrate and hardness (Figure 4). It is considered a prime factor for regulating the metabolic processes of aquatic communities as well as indicating water quality. Minimum DO was recorded in pre-monsoon probably due to higher temperatures (Bhat & Pandit 2014; Bouslah et al. 2017). The TA of water was primarily influenced by different seasons (p = 0.00, F = 257.42, η² = 0.839). It varied from 51.00 to 152.20 mg/L. The maximum values were recorded during the post-monsoon. The high alkalinity value has been recorded in wetlands infested with macrophyte-associated fauna and benthic biomass (Sugunan et al. 2000). According to Adebisi (1981), the amount of TA is inversely correlated with the water level. The results obtained in the present study were parallel with the findings of Mishra et al. (2014) and Arya et al. (2012).

The combined effect (p = 0.00, F = 119.93, η² = 0.829) of different sites and seasons was slightly higher than their individual effect on the AP in water and varied between the sites as well as the seasons. The seasonal variations of phosphate values were observed at their maximum during May (monsoon) and at their minimum during January (pre-monsoon). Phosphorus is the most critical nutrient affecting aquatic productivity (Ziauddin et al. 2013). The major input of phosphorus into the water-body comes from the leaching of soils from the catchment area by rain (Heron 1961). In the present study, the value of phosphate was relatively higher than in other studies (Vass & Zutshi 1979; Biswas et al. 2011; Ziauddin et al. 2013) and the concentration of this nutrient was the highest in the monsoon compared to other seasons. This can be due to the cause of extensive jute ratting at the studied wetland during this season (Das et al. 2009, 2011; Ghosh & Biswas 2018). On the basis of the phosphate concentration the wetland is oligo-mesotropic (Goldman & Horne 1983). The nitrate content (N) in water significantly varied spatially and seasonally (p = 0.00, F = 701.50, η² = 0.934) (Figure 3). The shallow site (site 3) has maximum nitrate content in the water during monsoon seasons. The nitrate values varied from 0.00 to 0.45 mg/L. The maximum nitrate values were recorded during September (monsoon) and the minimum during March (pre-monsoon). It was positively correlated (p < 0.05) with the productivity parameters (Chl-a, NPP and GPP) (Figure 4). Goldman & Horne (1983) identified this particular type of fluctuation trend in nitrate level as an indicator of mesotrophic to eutrophic wetland conditions. The nitrate contents in the studied wetlands were within the productive range (Banerjea 1967) and most favourable for the growth of plankton, which is in agreement with the study. In the present investigation, the values of available nitrate varied from system to system and with seasonal changes. According to Goldman & Horne (1983), a nitrate level of up to 1 mg/L is not harmful for aquatic fauna and flora. However, the present study revealed higher nitrate concentrations within the water due to a higher rate of nitrification compared to other studies (Joo & Francko 1995; Feresin et al. 2010; Ziauddin et al. 2013). The combined effect (p = 0.00, F = 117.82, η² = 0.841) of different sites and seasons was slightly higher than their individual effect on the hardness of the water. The value of hardness varied from 38.12 to 112.36 mg/L. According to Jhingran (1991), total hardness >60 mg/L is desirable for fisheries. The maximum values of total hardness were recorded during the pre-monsoon months, with a decreasing trend in the monsoon months. The total hardness value reached its lowest value in the post-monsoon months. Previous studies showed that the crash in hardness starts at the beginning of the monsoon season (Chakrabarty et al. 1959; Goldman & Wetzel 1963) and is negatively related to the water table (Rao & Govind 1964).

The influence of both different sites and seasons on the amount of FCO₂ varies widely depending on the quantity and nature of biological activity in the water. The values varied from 3.00–9.46 mg/L. The highest value is recorded during the pre-monsoon followed by the monsoon and the post-monsoon. The higher levels of FCO₂ observed during the pre-monsoon might be due to the decomposition of organic matter after eutrophication (Chakrabarty et al. 1959). The individual influences of different sites and seasons on GPP and NPP were significant (p < 0.05). The range of estimated GPP and NPP was 90.94–201.24 and 179.78–376.18 mgC/m³/h, respectively. The highest values were recorded in the pre-monsoon season and the lowest in the post-monsoon season. Both parameters were positively correlated (p < 0.05) with Chl-a (Figure 4). Seasonal and site-specific fluctuations in the rates of GPP and NPP were recorded, which indicates that the production rate did not remain the same throughout the course of the study.

The fish species diversity and evenness indices for all three different sites are shown in Table 2. The Shannon–Weiner diversity index (H^I) and Simpson diversity index (D_Sim) show the highest fish diversity at site-1, followed by sites-3 and 2. However, both the McIntosh diversity index (D_MC) and the Menhinick diversity index (D_Mn) depict the order of highest to lowest diversity as site-3 > site-1 > site-2. This is probably because of the microenvironment of site-3, which was a macrophyte-dominated refuge site for the fish species. On the other hand, the Margalef diversity index (D_Mg) is the same at all three sites. A high value of these indices characterises a diverse community, while a lower value characterises a less diverse community. In the present result, the recorded species diversity was low, as it was well known that H^I ≥ 3.5 depicts a healthy and rich diversity (Clarke & Warwick 2001). D_Mg was also low in the present study, representing lower species richness (Hanif et al. 2015; Khanom et al. 2016; Rahman et al. 2016). Evenness is the degree of relative density and the high value of evenness tells us that the total studied area holds similar densities of species, i.e., all species are distributed identically in a population (Aziz et al. 2021). Both the Pielou evenness index (J^I) and the McIntosh evenness index (E_MC) show the highest value at site-1, followed by site-3 and 2. However, the Simpson evenness index (E_Sim) is equal at sites-1, 3, and site-2 has the lowest value. It may be the cause of frequent fishing activity by applying a cast-net. This kind of fishing process may alter the natural habitat of the aquatic fauna and promote the uniform distribution of organisms. The present findings are similar to Sarkar et al. (2020).

For the estimation of the effect of climate change and anthropogenic stress on ecological services, the goodness of fit was determined by the results of the Deviance chi-square test [χ²(47) = 1,784.10, p > 0.05] and the Pearson chi-square test [χ²(47) = 42.64, p > 0.05], which indicated that the model fits the data well (Petrucci 2009; Field 2018). The result of the Omnibus test indicates a significant improvement in the new model, which contains the full set of explanatory variables, over the baseline model [χ²(1) = 6.657, p < 0.05]. This test uses a chi-square test to see if there is a significant difference between the –2Log-likelihoods of the two models. To test whether there was a significant improvement in the fit of the final model compared to the intercept only model, a likelihood chi-square test was performed, and it was found to be significant. The odds (ordinals) ratio of independent variables indicates that the odds of being at a higher level of stress in ecosystem health increase by a factor of 1.528 and 0.363, respectively, for every one-unit increase in the climatic stress due to climate change and anthropogenic stress (Table 3).

For the estimation of the effect of climate change and anthropogenic stress on ecosystem services, the goodness of fit was determined by the results of the Deviance chi-square test [χ² (11) = 10.66, p > 0.05] and the Pearson chi-square test [χ² (11) = 9.05, p > 0.05], which indicated that the model fits the data well (Petrucci 2009; Field 2018). The result of the Omnibus test indicates a significant improvement in the new model, which contains the full set of explanatory variables, over the baseline model [χ² (1) = 4.23, p < 0.05]. This test uses the chi-square test to see if there is a significant difference between the –2Log-likelihoods of the two models. A likelihood chi-square test was performed to see if there was a significant improvement in the fit of the final model over the intercept only model, and it was found to be significant (p < 0.05). The odds (ordinals) ratio of independent variables indicates that the odds of being at a higher level of stress in ecosystem services increase by a factor of 1.40 and 0.30, respectively, for every one-unit increase in climatic stress due to climate change and anthropogenic stress (Table 3).

It has been found that the majority of the stakeholders, who are active fishermen, experienced parallel effects of climate change on the ecosystem health and fisheries (Naskar et al. 2018, 2022; Sarkar et al. 2022). Considering these congruencies, it can be stated that climate change is likely to increase the vulnerability of wetland fisheries and productivity in this region. When the overall perceptions were analysed through ordinal regression, it was found that both the ecosystem health and services were mostly affected by climate change. However, anthropogenic stressors also play a significant role in the degradation of the wetland ecosystem.

According to the stakeholders, both climate change and anthropogenic stressors are responsible for food insecurity and reduced household income, as these deleterious situations are directly linked with the productivity of the wetland. Household income of the fishers is mainly affected due to stunted self-recruitment of naturally occurring fishes, reduced diversity, premature death of fishes because of jute retting and illegal fishing activities, which altogether affect the total fish production adversely. On the other hand, the adaptive capacity of the whole system in this context is still in its naïve experimental phase, as the maximum number of stakeholders are unaware of definite climate-smart strategies for wetland fisheries. The scientific community, however, has been able to discuss and formulate adaptation and mitigation strategies in order to combat climate change (Sarkar et al. 2016, 2018; Sarkar & Borah 2018; Sarkar & Das 2021). One must also note that here we measure the perceptions of climate change by the stakeholders of the wetland and its impacts on ecosystem health and services; it is important for future studies to be able to directly associate climatic events with these outcomes and behaviours in more detail. Semi-intensive carp culture within the pens was previously advised in the Bhomra Wetland along with conventional CBF post-monsoon. The carp species are stocked into the pens and subsequently to open waters in the advanced fingerling stage. Other fishes, such as catfish and snakeheads, are not intentionally stocked because they are naturally sustained from autochthonous wetlands stocks and are an important component of capture fisheries. In the context of continued water stress, the fishermen have applied some adaptive approaches. Before summer, as the water level begins to fall, a net-mediated enclosure (mesh size 10 mm) is constructed around the deepest part of the wetland. The main focus of enclosure culture in a wetland was to protect the stocked carp from illegal fishing in the shrinking wetland, permit more time for the carps to mature, and allow proper harvesting at an appropriate time based on size for maximum revenue. Therefore, this enclosure is kept under daily supervision by fishermen, which includes observation of fish health, observation of water quality through a necked eye, recruitment of a night guard, etc. The commercially important fish are incorporated from time to time within the enclosure and subsequently from the adjacent waters with the help of the cast and drag nets. However, the threat to the carp within the enclosure through natural predation remains a researchable issue. This altered MSMH (multiple stocking multiple harvesting) procedures for high-value fishes into the confined zone within the wetland and subsequent collection of sizable individuals is constant from January to March, after which fishing is terminated and the residual stocks are left as they are. Before the onset of monsoon months from April to May, the temporary pre-monsoon enclosure is removed in a strategic manner that allows gradual but restricted movement of fish between the fringe area, i.e., the arena, for feeding and breeding. The fishing or no-fishing decisions are determined and imposed by the management body of the fishery's cooperative society. In line with our study, Naskar et al. (2018) identified some of the most important climate change-induced pressures on wetland fisheries, which were water stress, sedimentation, aquatic weed proliferation and loss of wetland connectivity. The present study, based on the stakeholders’ perceptions, also pointed out the adverse climatic effect on the wetland ecosystem health and subsequently on the ecosystem services, including fish diversity and production. Sarkar et al. (2018) identified six climate-smart fisheries strategies, among which temporary pre-monsoon enclosure was one of the most effective techniques and is also applied at the Bhomra Wetland. Other recommended mitigation and adaptation strategies for coping with the effects of climate change are listed in Table 4.

Presently, research is generally concentrated on understanding the effects of a single stressor on an ecosystem or fish. However, how diverse and multiple stressors interact in an ecosystem needs to be studied to understand the actual impact on fish. Research needs to be done to determine critical alterations such as a shift to eutrophic conditions, dominance by cyanobacteria or other harmful floating plants, or the increase of predator fish species in the context of climate change. In addition to conventional observation of species abundances and abiotic ailments, researchers should look for indicators of ecological resilience. Recent studies have looked at how the distance to a threshold is caused by generic modifications of spatial as well as temporal patterns, reflecting the loss of resilience (Pace et al. 2015; Scheffer et al. 2015). These methods make use of precise time series and spatial information that is now easy to get, using automatic sensors and satellite-based data. Wetlands are complex ecological systems that need effective management through adaptable governance structures and organisations with the goal of adapting to change (Palomo et al. 2014). This type of governing system should have collaborations among scientists, socio-economic stakeholders, local communities and government organizations.

The present study clearly showed the relationship of the regulators of wetland fisheries to changing climate. The climatic data show increased temperature and decreased rainfall in the studied region over the last two decades. After studying the ecology of the wetland, it can be concluded that the wetland is suitable for carp culture, but the ecosystem's health has degraded over time. The wetland is dominated by Cypriniformes species, followed by Perciformes species. However, recent data of the year 2020 shows that the overall production of fish from the wetland is below the average production of the last 10 years. CCA revealed a strong relationship between the fish assemblage and ecological parameters (mainly TA and available nitrate) of the wetland. Though the species diversity and richness were both moderate, the evenness was high throughout the wetland. Though the ecological parameters of the wetland varied spatially and temporally, their relationship with the fish assemblage is not constant. The stakeholders of the wetland were aware of the climatic threats. However, local stressors had an additive effect in lowering the ecosystem's resilience. Their knowledge about climate-smart fisheries should be enhanced through frequent training programmes. It has been found that the reduction of depth at various sites of the water-body and the frequent infestation of macrophytes are predominant. With the proper stocking of fish seed and management, the fish yield can be increased many-fold. The macrophyte management needs to be done specially for the common water hyacinth, which can be utilized for making compost. The wetland is now receiving greater attention for rejuvenation, which is possible through the implementation of scientific management practices through stakeholders’ participation.

S.K. did sampling, investigated the study and performed manuscript revision. S.D. did sampling, investigated the study, performed data generation and analysis and wrote the original draft. U.K.S. conceptualized the study, did project administration, acquired funds, developed the methodology, investigated and supervised the study, validated the data, prepared and revised the manuscript. L.L., M.P., and G.K. did sampling and data generation and performed the manuscript revision. B.K.D. did project administration, acquired funds, developed the methodology, supervised the study and validated the data. B.D.G. and A.D. did sampling and performed data generation.

The study was funded by the NICRA, Indian Council of Agricultural Research.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.