Aim

To determine how reptile populations respond to anthropogenic habitat modification and determine if species traits and environmental factors influence such responses.

Location

Global.

Time period

1981–2018.

Major taxa studied

Squamata.

Methods

We compiled a database of 56 studies reporting how habitat modification affects reptile abundance, and calculated standardised mean differences in abundance (Hedges’ g). We used Bayesian meta-analytical models to test whether responses to habitat modification depended on body size, clutch size, reproductive mode, habitat specialisation, range size, disturbance type, vegetation type, temperature and precipitation.

Results

Based on 815 effect sizes from 376 species, we found an overall negative effect of habitat modification on reptile abundance (mean Hedges’ g = -0.43, 95% credible intervals = -0.61 to -0.26). Reptile abundance was, on average, one-third lower in modified compared to unmodified habitats. Small range sizes and small clutch sizes were associated with more negative responses to habitat modification, although the responses were weak and the credible intervals overlapped zero. We detected no effects of body size, habitat specialisation, reproductive mode (egg-laying or live-bearing), temperature, or precipitation. Some families exhibited more negative responses than others, although overall there was no phylogenetic signal in the data. Mining had the most negative impacts on reptile abundance, followed by agriculture, grazing, plantations and patch size reduction, whereas the mean effect of logging was neutral.

Main conclusions

Habitat modification is a key cause of reptile population declines, although there is variability in responses both within and between species, families, and vegetation types. The effect of disturbance type appeared to be related to intensity of habitat modification. Ongoing development of environmentally sustainable practices that ameliorate anthropogenic impacts is urgently needed to prevent reptile population declines.

Based on published literature, we compiled a database of 56 studies reporting how habitat modification affects reptile abundance. We extracted data from the text, tables, figures and appendices of papers. We used the means, standard deviations and sample sizes to calculate standardised mean differences (Hedges’ g and log response ratio) and sampling variances. For each data point, we recorded a number of ecological and environmental traits predicted to be important determinants of population sensitivity to habitat disturbance. The ecological traits were body mass, clutch size, reproductive mode, habitat specialisation and range size. We calculated an index of habitat specialisation by counting the number of major habitat types (e.g. forest, savanna, wetlands, rocky areas) listed in each species’ IUCN Red List profile. We derived range size from species distribution maps. A full set of trait data was not available for all study species. We recorded vegetation types as either forest, woodland, shrubland or grassland. We calculated the mean temperature of the warmest quarter of the year and mean annual precipitation within a 5-km radius around each study location.