Overlap between species and administrative boundaries (%)
Analyses revealed pervasive biases throughout ERMs. For all groups high proportions of species ranges overlapped with administrative borders (Table 1; Figure S1). Non-coastal borders show that administrative boundaries disproportionately impact range limit delineation, averaging 20-30% of non-coastal species range boundaries coinciding national administrative boundaries alone. Provincial boundaries are also used, increasing the average overlap by upto 10%.
To assess how species turnover relates to administrative areas, we calculated the standard deviations of richness maps (Figure 1), then identified what percentages of national administrative layers were included in the higher richness classes, to see what proportion of administrative boundaries fall within these hotspots. For reptiles, the upper four classes were retained, covering 40% of national boundaries (4% low turnover, 13% medium turnover 24% high turnover). This means that the hotspots for reptiles are disproportionately delimited by global administrative boundaries. In total, the overlap between administrative boundaries and these upper classes equates to 37% of all richness classes, but equals only 10% of the lowest quartile, 33% of medium turnover and 82% of the highest turnover. However, as low-diversity areas will necessarily have low turnover, this was then extracted for areas with >3 reptiles species, and for these regions 68% of national boundaries were on upper richness classes (43% highest turnover).
For amphibians, 34% of national boundaries overlapped with richness classes (two top richness classes, 27% high turnover), though equating to only 10% of high-turnover areas (16% of the highest levels). For birds, 52% of national boundaries coincide with turnover (3 upper richness classes, 21% at highest level), this equates to 29% of the highest bird turnover. For Mammals, 60% of national boundaries coincided with high turnover areas (3 top richness classes, 29% at high turnover), representing 26% of high turnover areas, and 35% of highest levels. For Odonata, 40% of national boundaries were covered (three top richness classes, with 25% at the highest turnover levels), representing 9% of high turnover areas but 16% of the highest.
In terms of individual countries, countries with longer coastlines obviously exhibited higher diversity changes, as coastlines typically mark absolute distributional boundaries. Yet, despite this, some countries with little coastline, and even landlocked countries, showed comparable levels of turnover at political boundaries. For reptiles, Nepal had levels of turnover at political boundaries roughly comparable to that of coastlines, and China also shows exceptional turnover, especially on its southwestern border (Figure 2); Bolivia showed some similar patterns.
Without coastal boundaries included Northern Southeast Asian boundaries are visible across taxa (excepting birds-Figure S1), with the Chinese side of borders show much lower diversity than neighboring Thailand, Vietnam and Myanmar. Iran also shows high turnover along administrative boundaries, considerably below the areas it borders. What is striking is for some countries, especially Southeast Asian countries, 100% of non-coastal boundaries show overlap with peak areas for species turnover, highlighting the disproportionately large role these borders play in mapping species distribution (Figure S1).