4.4 Model performance and verification
Habitat suitability modeling techniques have inherent limitations, but we have taken several measures to ensure the accuracy of our research findings. We employed ensemble modeling to reduce uncertainties associated with SDMs. However, it is important to acknowledge that SDM modeling algorithms still have their own limitations. Previous studies have emphasized the presence of uncertainties in SDMs, which can be minimized through a combination modeling approach (Guisan et al., 2017; Thuiller et al., 2019; Zhang et al., 2019). By integrating multiple models, we can address variable importance and capture habitat changes resulting from different factors, thereby advancing the study of species’ spatial distribution more efficiently and accurately (Fig. 3-4). Furthermore, during model validation, techniques for validating presence/absence models are well-developed and offer higher interpretability of prediction results (Palialexis et al., 2011). Therefore, this study utilized presence/absence data for habitat analysis. Moreover, existing evidence suggests that T. maximaexhibits two evolutionary branches within our study area’s spatial scale (Nuryanto & Kochzius 2009; Hui et al. 2016). However, a more detailed classification would likely lead to more accurate predictions. Future research can explore this hypothesis by conducting further genetic studies on T. maxima .
Conclusion
Post-2020, biodiversity conservation has become a shared concern among all parties to the Convention on Biological Diversity. Giant clams hold significant economic and ecological importance as coral reef species. In the present study, we employed SDMs with a dataset comprising 213 occurrence records and nine environmental variables to assess the potential distribution of T. maxima , a giant clam species, in the Indo-Pacific core region. We identified that land distance and light intensity are the dominate factors influencing the distribution ofT. maxima . Our analysis encompassed both current and future climate scenarios. Our study revealed that the potential distribution area of T. maxima is 1,519,764.73 km2, constituting 16.10% of the total protected areas. Additionally, through an overlay analysis, we evaluated the alignment between potential suitable areas and existing protected areas, enabling us to identify gaps in conservation efforts. Our findings provide insight into the spatial distribution patterns ofT. maxima , offering scientific guidance for effective conservation management and recommendations for the establishment of future protected areas.
Data availability statement
The raw data used in this study are available through the following links: the Global Biodiversity Information Facility (GBIF, https://www.gbif.org/), iNaturalist (https://www.inaturalist.org/), and the Ocean Biogeographic Information System (OBIS, https://obis.org/). Bio-ORACLE (https://bio-oracle.org/downloads-to-email.php) and Global Marine Environment Datasets (GMED, https://gmed.auckland.ac.nz/download.html). If the manuscript is accepted, then the code for the analysis will be available at datadryad.org.
Author contributions
Shenghao Liu: Conceptualization; Formal analysis; Funding acquisition; Writing – original draft. Tingting Li: Data curation; Formal analysis; Methodology; Visualization. Bailin Cong: Data curation; Formal analysis; Methodology. Leyu Yang: Data curation; Formal analysis. Zhaohui Zhang: Conceptualization; Supervision; Project administration. Linlin Zhao: Conceptualization; Funding acquisition; Writing – review & editing.