Trait-based ecological indices for evaluating mammalian species responses to habitat fragmentation

• We assessed responses of 10 mammalian species to habitat fragmentation. • Trait-based indices (SDI, IDI, FSI) provide framework to assess fragmentation risk. • Patch size increased species occurrence, while other effects varied among species. • Species with low SDI or high FSI favored well-connected and simple patches. • FSI best predicted patch connectivity dependence, supporting trait-based planning. Species-specific traits influence the response of mammals to habitat fragmentation, yet trait-based approaches remain rare in landscape-scale assessments. This study assesses this gap by providing a quantitative framework to identify species vulnerable to fragmentation and guide conservation planning. We aimed to quantify species-level structural dependence and evaluated whether it could be explained by ecological trait indices, supporting evidence-based conservation prioritization. We compiled nationwide occurrence records for ten mammalian species in the Republic of Korea, selecting species with sufficient records. Logistic regression models quantified species-specific responses to three patch attributes: patch size, connectivity, and boundary complexity. To assess whether the quantified responses could be explained by ecological traits, we examined whether the variations were predicted by three ecological trait indices: the Shannon diversity index (SDI) representing habitat specialization, the interior dependency index (IDI) reflecting road proximity, and the fragmentation sensitivity index (FSI) integrating rarity, home range size and road avoidance. Our analysis revealed that patch size consistently promoted species occurrence, whereas the effects of connectivity and boundary complexity varied among the species. For instance, Martes flavigula responded most strongly to all three patch attributes and had the lowest SDI, moderate IDI and highest FSI, indicating high structural dependence. Contrastingly, Hydropotes inermis showed minimal sensitivity to fragmentation and exhibited the highest SDI, lowest IDI and FSI. Among the three indices, FSI showed the strongest correlations with species-level structural dependence (ρ = 0.823, p = 0.003), suggesting it may serve as a useful tool for identifying vulnerable species and informing conservation planning.