Spatial modelling of the habitat suitability for the colour-polymorphic millipede Bicoxidens flavicollis in Zimbabwe

Résumé fourni par la source

Millipedes form a vital component of terrestrial ecosystems. Though the spatial ecology, habitat suitability, and the environmental factors governing their distribution remain poorly understood, despite the availability of occurrence records from field surveys, museum collections, and biodiversity databases in Southern African countries including Zimbabwe. Colour-polymorphic millipede ( Bicoxidens flavicollis ) presents a unique model for examining environmental heterogeneity influence on species distribution and phenotypic variation. Spatial modelling techniques incorporating nineteen (19) WorldClim bioclimatic variables, together with additional environmental predictors including vegetation cover, soil pH, soil texture and moisture, land use / land cover, and elevation, were used to predict the current and potential spatial distribution of B. flavicollis . MaxEnt model’s permutation importance analysis for B. flavicollis revealed that vegetation cover emerged as the most important variable when used in isolation whilst Bio 13 (Precipitation of Wettest Month), Bio 3 (Isothermality), Bio 16 (Precipitation of Wettest Quarter), Bio 6 (Mean Temperature of Coldest Month), Bio 12 (Annual Precipitation), Soil Moisture, Bio 10 (Mean Temperature of Warmest Quarter), Soil pH, Bio 18 (Precipitation of Warmest Quarter) exhibited higher model contributions. Variables with high percent contribution, such as Bio5 (Maximum Temperature of Wettest Month) (56.9%) and Bio18 (Precipitation of Warmest Quarter) (26.8%), showed zero permutation importance, indicating potential multicollinearity. Other variables, including elevation, soil texture made no significant contribution. There were no significant differences in the distribution of the species relative to previous studies. Integrating multidisciplinary techniques encompassing molecular, artificial intelligence and traditional ecological (millipede) knowledge and niche modelling methods is crucial for understanding millipede distribution. These interdisciplinary techniques must interfuse contribution and permutation importance metrics to disentangle valid spatial distribution patterns and location specific evolutionary adaptive traits of the B. flavicollis .