Native forest and proximity to humans are stronger drivers of Brazilian cottontail habitat use than invasive European hare
Data files
Feb 06, 2024 version files 35.03 KB
Abstract
Human activities and biological invasions have caused unprecedented biodiversity loss over the past 500 years. Proximity to humans drives the spatial distribution of species toward less disturbed habitats. Invasive species can competitively exclude native species, but species may coexist due to different habitat preferences. Here, we investigated how proximity to farms and the presence of the non-native European hare (Lepus europaeus) influence the habitat use by the Brazilian cottontail (Sylvilagus minensis) in southeastern Brazil. We found that the probability of cottontail site use increased with native forest cover and decreased with farmhouse proximity, ranging from 0.05 (SE = 0.02) at sites close to farmhouses (≅ 900 m) with no native forest to 0.70 (SE = 0.15) at sites far from farmhouses (≅ 2500 m) dominated by native forest. Higher risk of harassment and predation by free-roaming dogs and cats may explain the negative effect of farmhouse proximity on cottontail habitat use. We found little evidence for competitive exclusion by the European hare. Instead, our results suggest that the two species spatially segregate due to different habitat preferences. While the European hare more likely uses farmland in its native and non-native range, our results suggest that the Brazilian cottontail is a forest dweller. Although we found only weak evidence of competitive exclusion, we advise caution because invasive species may delay the onset of detrimental effects due to initial low population densities in newly invaded areas as is the case of the European hare in southeastern Brazil.
README: Native forest and proximity to humans are stronger drivers of Brazilian cottontail habitat use than invasive European hare
https://doi.org/10.5061/dryad.3r2280gpc
Brazilian cottontail (Sylvilagus minensis) detection obtained in 205 sites located in southeastern Brazil and covariates used to model species habitat use and detection probabilities. This dataset was obtained after sampling each site with a camera trap and a 200-m transect.
Description of the data and file structure
The dataset (.inp file) has 209 lines. The heading (non-readable information placed within /* */) is stored in the first four lines. The remaining 205 lines represent each sampled site. The line of each site starts with site name (e.g., / * JES - B11 */), followed by species detection (0) or non-detection (1) in each of the eight sampling occasions, a dummy covariate describing sampling months (from April to September), and all covariates used to model the Brazilian cottontail habitat use and detection probabilities. Each sampling month was defined as a group in Mark.
Covariates presented in the .inp file have the following names and measurement units: Sug = Sugarcane (%); ManF = Managed Forest (%); Past = Pastures (%); NatF = Native Forest (%); Sav = Savanna (%); StrD = Stream Density (m/ha); FED = native Forest Edge Density (m/ha); FDist = Farmhouse Distance (m); H = European hare conditional occupancy probability (n/a); Rainf = Rainfall (mm); MinTemp = Daily Minimum Temperature (°C); Y = year (n/a); CT_pos = Camera-trap Position (n/a); CT_eff = Camera-trap Effort (camera days). The 'n/a' means not applied, used for dimensionless covariates.
Code/Software
Occupancy analyses were performed in program Mark 9.0 (White & Burnham, 1999; https://doi.org/10.1080/00063659909477239)