Habitat loss is one of the greatest threats to species persistence. Gauging the scale of this problem requires quantitative methods that can predict the number of extinctions resulting from habitat loss. For the past three decades, the species–area relationship, an empirical relationship between the number of species present in an area and the size of that area, has been this tool. However, it fails to incorporate threats to species aside from habitat loss and the heterogeneous distribution of these threats across habitats. Recent studies have improved species–area predictions by incorporating not only direct effects of area on richness, but also indirect effects of area (through area-mediated predator abundance), on prey species richness. We extend this work to test the hypotheses that the indirect effects of the multiple threats of grazing and trampling in addition to fragmentation will amplify the effect of area on species richness and that this effect will be greatest in zones closest to the fragment edge. Further, we test for species and population level effects of fragmentation and grazing, including the non-random pattern of species loss and the decline in population sizes. We test our hypotheses with a field study of land snail richness in fragments with and without the additional threats of grazing and trampling. Our study supports the hypotheses that fragments with multiple threats in addition to habitat loss harbour fewer species than fragments without these threats, and that this effect is non-uniform across fragments, populations and species.