Evaluation of Different Strategies to Incorporate Absolute Abundance into Habitat Selection Modeling for the...

Developing habitat selection models is crucial for predicting the potential distribution of rare and endangered insects; however, current modeling approaches largely ignore absolute abundance data from sampling records. Integrating absolute abundance as sample weights involves two contrasting ecological hypotheses: one assumes high abundance reflects superior habitat quality (direct weighting), while the other attributes high abundance to spatially biased sampling effort (inverse weighting). Both hypotheses are plausible for Cheirotonus jansoni (Jordan, 1898), a highly phototactic endangered species, complicating the selection of an appropriate weighting strategy. This study evaluated four weighting strategies (unweighted, linear, inverse, and inverse square) derived from these hypotheses. Results indicated that the sampling effort correction strategy (inverse weighting) outperformed habitat-quality-driven strategies (linear weighting) in habitat suitability mapping, achieving a higher AUC (93.41) compared to the linear weighting (91.32). Notably, the linear weighting strategy performed worse than the unweighted baseline, yielding a lower AUC (91.32) than the baseline's 93.11. These findings provide methodological guidance for modeling the distribution of phototactic insects and offer scientific support for the in situ conservation, ex situ site selection, and supplementary field surveys of C. jansoni.