Introduction

Sthenoteuthis oualaniensis, a warm-water oceanic species, is widely distributed in the equatorial and subtropical waters of the Indian and Pacific Oceans1. This species exhibits distinct habitat preferences, typically thriving in warm surface waters (27–29 °C) with moderate chlorophyll-a concentrations (0.5–1.1 mg/m3)2. Additionally, Sthenoteuthis oualaniensis demonstrates pronounced diel vertical migration behavior, rising to surface layers during dusk for feeding. Consequently, its distribution is closely associated with the vertical thermal structure of upper water layers3.

The Indian Ocean, separated by continents, experiences significant thermal differences between land and sea, rendering it a globally typical monsoon region4,5. The most typical interannual modes in the Indian Ocean, such as the basin-wide sea surface temperature (SST) mode and the Indian Ocean Dipole, occur during the monsoon transition periods, leading to distinct seasonal variations in the oceanic environment. Specifically, the northwestern Indian Ocean is dominated by seasonal reversal of monsoonal winds: the northeast monsoon prevails in winter (December–February), driving westward surface currents, while the southwest monsoon dominates in summer (June–August), reversing currents eastward4. This seasonal circulation reversal directly affects water mass distribution, nutrient transport, and thermal structure, thereby shaping the spatiotemporal distribution of Sthenoteuthis oualaniensis and further influencing the habitat selection of Sthenoteuthis oualaniensis. Sthenoteuthis oualaniensis is highly sensitive to its oceanic habitat, and its population distribution is greatly influenced and driven by environmental factors6,7. Therefore, examining the habitat model of Sthenoteuthis oualaniensis in the northwestern Indian Ocean based on environmental factors can offer a scientific foundation for fishery forecasts and management of this species in the region.

The habitat suitability index (HSI) model, initially proposed by the US Fish and Wildlife Service’s National Wetlands Research Center in the 1980s8, simulates organisms’ responses to their surrounding habitat factors. This model has been extensively utilized in species management and fish distribution9,10. For example, Liu et al.11 employed the HSI model with production and environmental data of Cololabis saira from the high seas of the northwest Pacific Ocean between 2013 and 2016 to explore fishery forecasts for this species. Wu et al.12 analyzed the spatiotemporal changes of Scomber japonicus habitats in the East China Sea under three future climate scenarios using the HSI model, based on environmental data from 2015 to 2100 sourced from the World Climate Research Program. Zou et al.13 studied the spatial distribution characteristics of Pseudosciaena polyactis in Haizhou Bay by applying the HSI model, combining fisheries data from 2011, 2013, and 2014 with corresponding environmental data. Despite the HSI model’s widespread application in fisheries, there have been no reported studies, domestically or internationally, on the relationship between the seasonal distribution of Sthenoteuthis oualaniensis habitats and environmental factors in the Northwestern Indian Ocean. Based on the latest production statistics and environmental data of Sthenoteuthis oualaniensis collected by Chinese squid-jigging vessels in the Northwestern Indian Ocean from 2021 to 2024, this paper utilizes the generalized additive model (GAM) to identify the top three marine environmental factors that significantly impact the habitat suitability of Sthenoteuthis oualaniensis in this region. By assigning different weights to these three environmental factors, HSI models with varying seasonal weights are established. The optimal model is then selected to predict the habitats of Sthenoteuthis oualaniensis, providing a theoretical basis for the rational utilization and scientific management of this species’ resources.

Materials and methods

Data sources

The fisheries data (Private Dataset) for this study originated from the 3,326 light purse seine fishing logbooks of Sthenoteuthis oualaniensis, captured by Fujian Zhengguan Fisheries Development Co., Ltd. from September of each year spanning from 2021 to 2024, up until the following May. These logbooks encompassed detailed information regarding fishing duration, operating longitude and latitude, as well as the daily catch volume measured in tons. The spatial scope of the study extended across latitudes 12°N–22°N and longitudes 60°E–69°E.

The oceanic environmental data were exclusively sourced from the Copernicus Marine Environment Monitoring Service (CMEMS), specifically the Global Ocean Physics Reanalysis (GLO-PHY) and Biogeochemistry Reanalysis (GLO-BGC) datasets, accessible at https://resources.marine.copernicus.eu. This comprehensive dataset included Chlorophyll-a Concentration (Chl-a), Sea Surface Height (SSH), Sea Surface Salinity, vertical temperature profiles within the 0–600 m water layer, as well as wind speed (W) and current speed (U) measured in both eastward and northward directions. The data featured a monthly temporal resolution and a spatial resolution of 0.25° × 0.25°.

Methods

Calculation of CPUE and fishing ground centroid

The nominal Catch Per Unit Effort (CPUE) and the fishing ground centroid were calculated for the period spanning from each September to following May from 2021 to 2024. The calculation formulas utilized for this purpose are referenced in14,15,16.

$$CPUE = \frac{C}{E}$$
(1)
$$X_{i} = \sum\limits_{i = 1}^{k} {(D_{ij} \times } X_{ij} )/\sum\limits_{i = 1}^{k} {D_{ij} } ,$$
(2)
$$Y_{i} = \sum\limits_{i = 1}^{k} {(D_{ij} \times Y_{ij} )/\sum\limits_{i = 1}^{k} {D_{ij} } } .$$
(3)

In the formula, CPUE is measured in tons per trip. Here, C and E respectively represent the monthly catch and the corresponding number of fishing trips for a single fishing vessel. Xi and Yi, on the other hand, denote the longitude and latitude of the centroid of the fish aggregation, respectively. Additionally, Dij stands for the catch in the jth month of the ith fishing area, while Xij and Yij represent the longitude and latitude of the centroid of the ith fishing area during the jth month, respectively. Finally, k signifies the total number of fishing areas considered.

Selection of characteristic variables

Due to the varying magnitudes among diverse datasets, errors may occur during modeling and analysis. To guarantee the precision of the model, it is imperative to normalize these disparate datasets. In this paper, we utilize the Min–Max normalization method to scale the CPUE and environmental factor data of Sthenoteuthis oualaniensis in the northwestern Indian Ocean to the range of [0,1], as referenced in15. The specific calculation formula for this normalization technique is detailed in15.

$$P = \frac{{F - F_{{{\text{Min}}}} }}{{F_{{{\text{Max}}}} - F_{{{\text{Min}}}} }}$$
(4)

In the formula, P denotes the normalized value, F signifies the actual value, while FMax and FMin represent the maximum and minimum values of the actual value, respectively.

To prevent potential high correlation among oceanic environmental factors and spatio-temporal variables, a multicollinearity diagnosis is performed on both sets of factors. This diagnosis relies on the Variance Inflation Factor (VIF) as its metric15. A VIF value of 10 or above indicates the presence of a collinearity issue between two variables15; conversely, a VIF below 10 suggests the absence of such an issue. During the multicollinearity diagnosis process, environmental factors exhibiting a VIF greater than 10 are progressively removed until all remaining factors have a VIF of less than 10. The specific calculation formula for the Variance Inflation Factor is referenced in15 and detailed below:

$$VIF = \frac{1}{{1 - W^{2} }}$$
(5)

In the formula, Wi represents the variance proportion of the i-th variable, while VIF stands for the Variance Inflation Factor.

Generalized additive model (GAM)

The GAM is employed to analyze the relationship between the CPUE of Sthenoteuthis oualaniensis in the northwestern Indian Ocean and its influencing factors. Taking the CPUE of Sthenoteuthis oualaniensis as the response variable and various environmental and spatial factors as explanatory variables, a GAM model is constructed to explore the impacts of these factors on the spatio-temporal distribution of the CPUE. The construction of this GAM model is facilitated through the use of the mgcv package in R language (Version 4.0.3), where a Gaussian distribution is assumed for the model’s errors and the natural logarithm is chosen as the link function17. To assess the degree of influence of the explanatory variables on the CPUE, the F-test is utilized, following the methodology outlined by Xu et al.18. The formulation of the GAM model, as referenced in17, is presented below:

$$Z = s(Chla) + s(U) + s(W) + s(SSH) + S(SSS) + s(SST) + \beta$$
(6)

where Z represents the logarithmic value of the CPUE; s denotes the smoothing function using a natural cubic spline; and β stands for the residual terms.

Habitat model construction

Fishing effort is commonly employed as an indicator of the presence or utilization of fish species19,20. In this study, we integrated environmental and fisheries data spanning from September 2021 to May 2023 to compute the Comprehensive Habitat Suitability Index (HSI) values for each quarter under various weighted models. Furthermore, we predicted the suitable habitat areas for Sthenoteuthis oualaniensis in the northwestern Indian Ocean for each quarter from September 2023 to May 2024. Suitability indices (SIs) were derived using catch data from September 2021 to May 2023, along with the top three environmental factors based on their contribution rates. The study assumes that the highest fishing effort observed in the corresponding quarter at each latitude and longitude signifies the area with the highest distribution of Sthenoteuthis oualaniensis resources, assigned an SI value of 1; conversely, an SI value of 0 is assigned to areas where fishing effort is 0, indicating the lowest distribution of Sthenoteuthis oualaniensis resources20. The formula for calculating SI21,22 is provided below:

$$SI_{i} = \frac{{CPUE_{i} }}{{CPUE_{i,MAX} }}$$
(7)

where, SIi stands as the suitability index for the ith quarter, while CPUEi,MAX indicates the peak CPUE value documented during the ith quarter, and CPUEi signifies the average or typical CPUE for that same quarter.

Using the established SI model for a single environmental factor as a foundation, various weights are assigned to the environmental factors according to the weight setting scheme outlined in Table 120. Based on these weights, HSI models for habitat suitability are developed and compared under different weighting scenarios. The formula for calculating the HSI model is provided as follows20:

$$HSI = W_{1} \times SI_{1} + W_{2} \times SI_{2} + W_{3} \times SI_{3}$$
(8)

where W1, W2, and W3 represent the respective weight values assigned to the SI models for the three most significant environmental factors. Similarly, SI1, SI2, and SI3 denote the corresponding SI values for these top three environmental factors.

Table 1 Diffferent scenarios for the various factors.
Full size table

Model validation and selection

Utilizing the HSI models developed under the aforementioned diverse weighting schemes, HSI values for each quarter spanning from September 2021 to May 2023 were computed, ranging from 0 to 1. These HSI values were categorized into ten grades: (0, 0.1], (0.1, 0.2], …, (0.9, 1]. Sea areas with HSI values of 0.6 or above were deemed suitable habitats for Sthenoteuthis oualaniensis22. To ascertain the optimal HSI model for each quarter, the proportions of fishing effort allocated within different HSI intervals were calculated independently for each weighting model. ArcGIS software was employed to map the spatial distribution of HSI, based on the optimal model selected, from September 2023 to May of the subsequent year. This map was then overlaid with actual CPUE data to verify the efficacy of the optimal model in predicting quarterly fishing ground distributions.

Spatio-temporal variations of habitats under abnormal climatic events

Using the definitions of El Niño and La Niña events provided by NOAA [An El Niño (or La Niña) event is deemed to occur when the Oceanic Niño Index (ONI) remains above 0.5 °C (or below − 0.5 °C) for a sustained period of 5 months.]23,24 (http://ggweather.com/enso/oni.htm), comparative analyses were conducted for La Niña years (spanning from December of the preceding year to May), El Niño years (from December to May of the subsequent year), and normal years (also from December to May of the subsequent year). Notably, from June to August, influenced by the summer southwest monsoon, most fishing vessels are non-operational, resulting in a lack of fishing data for Sthenoteuthis oualaniensis25. In September, the southwest monsoon has just ended; the catch from September to November is relatively low, and the ONI indices during September to November indicate no occurrence of climatic events—this deviates from the climatic events represented by the period from December to the following May. Therefore, the climatic events represented by December to the following May were selected for analysis. Specifically, the periods analyzed were December 2021 to May 2022 for La Niña, December 2022 to May 2023 for El Niño, and December 2023 to May 2024 for normal years. The optimal model for each quarter was employed to calculate the average monthly HSI values for the corresponding months during El Niño and La Niña years (Table 2). Spatial distribution maps of HSI were then generated to investigate the spatio-temporal variations in suitable habitats for Sthenoteuthis oualaniensis in the northwestern Indian Ocean during these abnormal climatic events.

Table 2 Oceanic Niño Index data spanning 2021 to 2024.
Full size table

Results

Impact of marine environmental factors on CPUE

Based on the multicollinearity diagnosis results outlined in Table 3, the VIF values for all eleven variables—chlorophyll-a concentration (Chl-a), eastward surface current velocity (EU) and northward surface current velocity (NU), northward wind speed (NW), sea surface height (SSH), sea surface salinity (S0), sea surface temperature (T0), temperature at 50 m depth (T50), temperature at 100 m depth (T100), temperature at 300 m depth (T300), and temperature at 600 m depth (T600)—are beneath the threshold of 10. Consequently, these eleven variables have been chosen as the explanatory variables for the GAM.

Table 3 Multicollinearity analysis among variables.
Full size table

The results of the GAM for Sthenoteuthis oualaniensis in the northwestern Indian Ocean are presented in Table 4. Based on the statistical significance of the factors (P < 0.05), the F-test reveals that T0 (surface temperature), chlorophyll-a concentration (Chl-a), and eastward current velocity (EU) are the three most influential environmental factors affecting the CPUE of Sthenoteuthis oualaniensis in this region. EU transports Sthenoteuthis oualaniensis' prey (phytoplankton, zooplankton) to aggregation zones to aid foraging, influences paralarvae dispersal to guide them to suitable habitats with optimal temperature and nutrients, and maintains coastal-offshore habitat connectivity for access to spawning/feeding grounds. These directly affect the species’ distribution and abundance, making EU a key driver of CPUE variations. These are followed by temperature at 50 m depth (T50), temperature at 100 m depth (T100), sea surface salinity (S0), northward current velocity (NU), northward wind speed (NW), temperature at 300 m depth (T300), sea surface height (SSH), and temperature at 600 m depth (T600). According to the findings of the GAM analysis, T0, Chl-a, and EU have been selected to establish the HSI model for Sthenoteuthis oualaniensis in the northwestern Indian Ocean.

Table 4 Statistics of each factor of the GAM model.
Full size table

Seasonal variations in CPUE and the fishing ground centroid in latitude and longitude

As depicted in Fig. 1, regarding latitude, the CPUE of Sthenoteuthis oualaniensis in the northwestern Indian Ocean demonstrates a positive correlation with latitude during autumn (September–November) and winter (December–February), and a negative correlation during spring (March–May). The CPUE reaches its peak in autumn, with the fishing ground centroid positioned in the northernmost waters. Conversely, in spring, the CPUE tends to increase compared to winter, as the fishing ground centroid migrates southwards. In terms of longitude, the CPUE of Sthenoteuthis oualaniensis exhibits a positive correlation. Specifically, the fishing ground centroid shifts westward from autumn to winter, accompanied by a decline in CPUE. In spring, the fishing ground shifts eastward, leading to an increase in CPUE.

Fig. 1
figure 1

Seasonal changes in CPUE and latitude and longitude.

Full size image

Comparison of HSI models

By constructing 30 models (Tables 5, 6, 7) for Sthenoteuthis oualaniensis in the northwestern Indian Ocean across three seasons—autumn, winter, and spring—we have calculated the proportions of yield and fishing effort within various HSI ranges. The optimal models vary across seasons, reflecting differing influence weights of the environmental factors T0, Chl-a, and EU in different quarters. In autumn, Model 4 is optimal, with weight proportions of 0.1, 0.1, and 0.8 for T0, Chl-a, and EU, respectively, highlighting EU as the factor with the highest weight. In winter, Model 9 is optimal, with weight proportions of 0.8, 0.1, and 0.1 for T0, Chl-a, and EU, respectively, emphasizing T0 as the most influential factor. In spring, Model 5 is optimal, with weight proportions of 0.25, 0.5, and 0.25 for T0, Chl-a, and EU, respectively, indicating Chl-a as the dominant factor. When comparing the optimal models across the three seasons, it is observed that under these models, the proportions of catch and fishing effort peak within the HSI range of 0.6–1.0. However, due to seasonal variations in the primary environmental factor influencing the optimal model, the spatial distribution of habitat suitability areas differs across seasons, leading to variations in the size of these areas.

Table 5 Percentage of catch and effort at different HSI class intervals sourced from different weighting models in the winter from 2021 to 2023.
Full size table
Table 6 Percentage of catch and effort at different HSI class intervals sourced from different weighting models in the spring from 2021 to 2023.
Full size table
Table 7 Percentage of catch and effort at different HSI class intervals sourced from different weighting models in the summer from 2021 to 2023.
Full size table

Model validation

The optimal HSI models, developed using quarterly data spanning from September 2021 to May 2023, were employed to forecast the suitable habitats for Sthenoteuthis oualaniensis in the northwestern Indian Ocean for each quarter from September 2023 through May of the subsequent year. These forecasts were then validated by comparing the fisheries data from the same period (September 2023 to May of the next year) with the predicted HSI values. As depicted in Fig. 2, the catch data for Sthenoteuthis oualaniensis during the winter season encompasses a broader geographic range compared to spring and autumn, while the suitable habitat area in spring is notably larger than that in autumn. The majority of the higher fishing effort is concentrated within the range of suitable habitats. Figure 3 reveals that the proportions of both catch and fishing effort increase in tandem with the HSI value. Specifically, in autumn, the proportion of the best habitat (HSI > 0.8) exhibits a declining trend when compared to the more suitable habitat (0.6 < HSI < 0.8). Across all three seasons, the proportions of fishing effort and catch with HSI values exceeding 0.6 surpass 70%, and in spring, they exceed 90%. These findings indicate that the seasonal optimal models can effectively predict the habitat suitability of Sthenoteuthis oualaniensis in the northwestern Indian Ocean, with reliable prediction results.

Fig. 2
figure 2

Spatial distribution of CPUE overlapped with the HSI values in each season in September 2023 to May 2024.

Full size image
Fig. 3
figure 3

Percentage of catch and fishing effort within different class intervals of HSI in September 2023 to May 2024 based on the best HSI model.

Full size image

Response of habitat changes to abnormal climatic events

As illustrated in Fig. 4, the habitat suitability distribution of Sthenoteuthis oualaniensis in the northwestern Indian Ocean demonstrates notable variations across different climatic years. In comparison to El Niño events, the spatial extent of suitable habitats during La Niña events in winter and spring occupies the largest proportion, whereas during El Niño events, it is the smallest. Specifically, in winter, the range of suitable habitats during La Niña events expands eastward, southward, and westward in contrast to El Niño events. Conversely, in spring, while the range of suitable habitats during La Niña events also extends eastward, it contracts in the southwest compared to El Niño events. Under normal climatic conditions, the characteristics of suitable habitats vary across seasons. In winter, the area of suitable habitats is smaller compared to La Niña events and extends southeastward relative to El Niño events. In spring, the area of suitable habitats expands westward and contracts eastward when compared to La Niña events, while it extends southwestward and contracts northeastward in relation to El Niño events. This phenomenon is particularly evident for habitats with an HSI value greater than 0.8.

Fig. 4
figure 4

Spatial distribution maps of habitats in winter and spring during El Niño years, La Niña years, and normal years.

Full size image

Discussion

Selection of marine environmental factors

The activities, distribution, reproduction, and growth of marine organisms are inextricably tied to the marine environment26. This environment serves as the cornerstone for marine life’s survival and significantly impacts the interplay between fish and their habitats23. When constructing the HSI model, the judicious selection of impact factors and the determination of their respective weights are paramount. Typically, environmental factors are chosen based on existing literature, prior expertise, and the availability of data. In this study, considering the profound influence on the activities of Sthenoteuthis oualaniensis in the northwestern Indian Ocean, along with the squid’s inherent life habits, eleven marine environmental factors, such as chlorophyll-a concentration and sea surface temperature, were initially chosen for model building27,28. However, acknowledging that the impact of the same environmental factor on the distribution of a species can vary across regions, seasons, and growth stages29,30, it is imperative to screen the initially selected factors to ensure an optimal combination and enhance the model’s efficacy31. This study utilized the GAM model to screen these factors based on their contribution to the CPUE of Sthenoteuthis oualaniensis in the northwestern Indian Ocean, ultimately identifying sea surface temperature, chlorophyll-a concentration, and eastward current velocity as the marine environmental factors for HSI model construction.

The analysis indicates that these three factors contribute most significantly to Sthenoteuthis oualaniensis in this region, which is bordered by Africa, Asia, and the Arabian Peninsula—geographic features that amplify monsoonal effects, making it a quintessential monsoon sea area. Driven by seasonal monsoons (northeast monsoon in winter and southwest monsoon in summer), the region experiences strong monsoonal currents and coastal upwelling: winter northeasterlies push surface waters westward, inducing upwelling along the coasts that draws nutrient-rich deep waters to the surface; summer southwesterlies reverse this circulation, enhancing vertical mixing and further ventilating deep water into the euphotic zone4,32. This dynamic water exchange elevates surface chlorophyll-a concentrations by stimulating phytoplankton blooms, as nutrients fuel primary productivity33. For Sthenoteuthis oualaniensis, this nutrient cycling and phytoplankton abundance are tightly linked to its life cycle: phytoplankton supports a robust food web, providing prey for juvenile squids, while adults rely on these productive zones for foraging to accumulate energy for spawning34. Additionally, monsoon-driven temperature fluctuations (modulated by currents) influence egg development and larval survival, as the species exhibits thermal preferences for early life stages1. Collectively, these monsoon-mediated processes—currents, upwelling, nutrient supply, and food web dynamics—profoundly shape the spatial distribution and resource abundance of Sthenoteuthis oualaniensis in the region.

Notably, this spatiotemporal pattern of CPUE occurs against a backdrop of limited seasonal variability in key oceanographic factors, consistent with general regional oceanographic patterns, where SST varies within a relatively narrow range across seasons([e.g., 26.6–28.5 °C in autumn, 26.1–28.4 °C in winter, 27–28.7 °C in spring]), chl-a concentrations show similarly constrained fluctuations([e.g., − 0.19 to 1.34 mg/m3in autumn, − 0.16 to 0.29 mg/m3in winter, − 0.18 to 1.39 mg/m3in spring]), and eastward currents([e.g., − 0.13 to 0.34 m/s in autumn, − 0.25 to 0.35 m/s in winter, − 0.08 to 0.33 m/s in spring]), which may influence larval dispersal and prey distribution, exhibit moderate seasonal differences in intensity; this limited seasonal variability in SST and chl-a levels emphasizes that the observed shifts in Sthenoteuthis oualaniensis distribution and CPUE are likely driven by fine-scale habitat selection within these constrained ranges, the absence of large seasonal fluctuations in these factors also highlights the importance of other variables in shaping habitat suitability, and our study clarifies how the species responds to subtle environmental cues, which would be less distinguishable in regions with more dramatic seasonal changes.

Model construction

The HSI model is grounded on the SI models formulated for individual environmental factors, representing the cumulative impact of these factors on the habitat. Since each environmental factor exerts varying degrees of influence on the habitat, different weights are assigned to each when constructing the HSI model to analyze the seasonal impact of diverse environmental factors on the habitat. In cephalopod habitat modeling studies, multiple methodologies exist for computing the HSI by amalgamating all single-factor SI models. In this study, HSI models tailored to each season were developed based on differing weight proportions. The optimal model for each season was chosen through comparative analysis and validated using data spanning from September 2023 to May of the subsequent year. The predictive results revealed that prediction accuracy exceeded 70% in autumn and winter, and surpassed 95% in spring. The study underscored that incorporating the weights of various environmental factors in the HSI model enhanced prediction accuracy across all seasons. Comparable findings were also observed in studies by Feng et al.20 on the habitat of Dosidicus gigas in Peru’s offshore waters and by Zhou et al.35 on the habitat of Todarodes pacificus in the northwestern Pacific Ocean. Consequently, the influence of environmental factors on the habitat distribution of Sthenoteuthis oualaniensis in the northwestern Indian Ocean fluctuates across different seasons.

The spatial distribution and proportion of HSI calculated under the optimal models for each season are depicted in Figs. 5 and 6. In the spring’s optimal model, Chl-a exhibited the highest weight contribution rate, potentially linked to nutrient reproduction during this season. The increased variability in nutrient acquisition by Sthenoteuthis oualaniensis in spring amplified the impact of Chl-a on its habitat. In the autumn’s optimal model, EU had the highest contribution rate, analyzed to be associated with the seasonal shift in monsoon wind direction in the Indian Ocean. Specifically, eastward surface currents in autumn shape Sthenoteuthis oualaniensis distribution and abundance by transporting prey to aggregation zones, dispersing paralarvae to suitable nurseries, and connecting coastal-offshore habitats, making EU the dominant driver and directly impacting CPUE by concentrating squid in resource-rich areas. In the winter’s optimal model, T0 demonstrated the highest contribution rate, indicating that the temperature decrease in the northwestern Indian Ocean during winter has a notable correlation with changes in the habitat of Sthenoteuthis oualaniensis. This could be attributed to the species’ preference for specific thermal conditions (27–29 °C), as cooler waters during winter may influence their vertical or horizontal distribution, potentially driving them to areas within this favorable ranges2,36. Additionally, the seasonal shift in oceanographic features, such as thermocline depth and primary productivity, may further interact with temperature changes to impact their habitat preferences37. Using data from September 2021 to May of the following year and validating under the optimal model for each season, the proportion of HSI values exceeding 0.6 was above 75% across all seasons. Therefore, it is viable to investigate the relationship between the habitat distribution of Sthenoteuthis oualaniensis and environmental factors in the northwestern Indian Ocean by selecting the three environmental factors: T0, Chl-a, and EU.

Fig. 5
figure 5

Spatial distribution of HSI values of Sthenoteuthis oualaniensis from September 2021 to May 2024 calculated by the optimal model of different quarters.

Full size image
Fig. 6
figure 6

The percentages of catch and fishing effort at different HSI class intervals in each season in September 2021 to May 2024.

Full size image

Response of Sthenoteuthis oualaniensis to climate change

The spatiotemporal distribution of Sthenoteuthis oualaniensis is significantly shaped by climate change. Existing research27 has demonstrated that ENSO events exert a pivotal influence on the dynamics of its habitat. Compared to El Niño conditions, Sthenoteuthis oualaniensis exhibits greater adaptability to the climatic regimes of La Niña events, which typically result in a more extensive range of suitable habitats, whereas El Niño events contract its viable environment. This suggests that La Niña occurrences, whether in winter or spring, facilitate the survival of Sthenoteuthis oualaniensis in the northwestern Indian Ocean, driving the expansion of its suitable habitat range. Such conclusions align with those of38, and the present study further confirms that the area of suitable habitat reaches its maximum during La Niña events.

The disparity in habitat range arises from the synergistic interplay of multiple environmental factors. During La Niña events, a combination of relatively cool sea surface temperatures, elevated chlorophyll-a concentrations, and robust upwelling collectively promotes the expansion of suitable habitat for Sthenoteuthis oualaniensis39, creating optimal living conditions. In contrast, El Niño events disrupt this balance: the marine environment deviates from the species’ adaptive range, with weakened upwelling leading to reduced chlorophyll-a concentrations and diminished nutrient availability. This aligns with Yu et al.’s findings on Dosidicus gigas in the southeastern Pacific, where El Niño was shown to reduce suitable habitat area and lower yields. Additionally, Waluda et al.40 noted that intense La Niña years, characterized by strong upwelling, correlate with higher resource abundance and catch rates of Dosidicus gigas—further validating the critical role of upwelling and related factors in shaping habitat extent.

During El Niño events, the suitable habitat of Sthenoteuthis oualaniensis in the northwestern Indian Ocean exhibits distinct spatial shifts: northward in latitude, westward in winter (longitude), and eastward in spring (longitude). This pattern is hypothesized to be linked to the region’s monsoonal dynamics—specifically, the winter northeast monsoon and spring southwest monsoon25. Against the backdrop of global warming, Sthenoteuthis oualaniensis generally exhibits a trend toward higher latitudes25,41 though La Niña events may temporarily reverse this, driving shifts toward lower latitudes. This pattern is supported by findings from other studies25,42: for instance, one year after the 2016–2017 La Niña event, the species shifted southward in latitude during 2017–2018; conversely, the 2015–2016 super El Niño event was followed by an unusually large northward shift in 2016–2017 compared to other years25.

These observations collectively reveal a clear response mechanism of Sthenoteuthis oualaniensis habitat in the northwestern Indian Ocean to ENSO: by modulating key factors such as water temperature, upwelling, and monsoon-driven currents, ENSO induces predictable changes in both the size and spatial distribution of the species’ habitat.

Conclusion

This study analyzed the latest fishing data for Sthenoteuthis oualaniensis collected from September 2021 to May of the subsequent year in the northwestern Indian Ocean. The seasonal variations in CPUE and the latitudinal and longitudinal centroids of fishing grounds centroid for Sthenoteuthis oualaniensis in this region were examined. The findings revealed a positive correlation between CPUE and latitude during autumn and winter, whereas a negative correlation was observed in spring. In terms of longitude, a positive correlation between CPUE and longitude was noted. Through multicollinearity analysis and the application of the GAM model, three marine environmental factors with the greatest influence on the abundance of Sthenoteuthis oualaniensis in the northwestern Indian Ocean were identified.

For each quarter, ten HSI models with varying weights were proposed. The optimal model was selected based on the principle of achieving the highest proportion of HSI values greater than 0.6. In the optimal autumn model, weights of 0.1, 0.1, and 0.8 were assigned to T0, Chl-a, and EU, respectively, with EU having the highest weight. During winter, the optimal model assigned weights of 0.8, 0.1, and 0.1 to T0, Chl-a, and EU, respectively, with T0 carrying the greatest weight. For spring, the optimal model assigned weights of 0.25, 0.5, and 0.25 to T0, Chl-a, and EU, respectively, with Chl-a having the highest weight. The models were validated using data from September 2023 to May of the subsequent year for each quarter, achieving prediction accuracies exceeding 70% and reaching over 95% in spring.

By comparing the habitat trends of Sthenoteuthis oualaniensis in the northwestern Indian Ocean under different climate events, it was observed that the habitat area was largest during La Niña events and smallest during El Niño events, with a noticeable northward shift occurring during El Niño events. Additionally, the habitat of Sthenoteuthis oualaniensis in the northwestern Indian Ocean shifted in the east–west direction under the influence of different monsoon winds, moving westward during the northeast monsoon in winter and eastward during the southwest monsoon in spring.

This study primarily focused on developing optimal monthly models for the habitat of Sthenoteuthis oualaniensis in the northwestern Indian Ocean, aiming to further enhance prediction accuracy based on previous research. The validation results demonstrated that the models’ prediction performance was relatively reliable, albeit with certain limitations. The environmental factors affecting the habitat of Sthenoteuthis oualaniensis are complex and multifaceted, with intricate interactions. This study considered a limited number of environmental factors. For instance, sea surface height has also been shown to impact Sthenoteuthis oualaniensis43. Moreover, cephalopods, as short-lived species, exhibit significant variations on meso- and small scales. Therefore, short-term and highly variable marine dynamic environmental factors, such as mesoscale eddies, should be incorporated and quantified as indicators for constructing habitat models in future research endeavors.