Top six logistic regression models for each season used to evaluate microhabitat use for Texas Alligator Lizards (Gerrhonotus infernalis) at Bamberger Ranch Preserve, Blanco County, Texas.

Contexts in source publication

Context 1

... model selection, bare ground and leaf litter were removed during variable reduction for all seasons as they did not contribute significantly to differences between used and random reference locations. Woody plant, woody debris, and rocky refuge (hereafter described as the primary covariates) were included in the top-ranked models, and model averaging indicated they were consistently important predictors of lizard microhabitat use across seasons (Tables 3 and 4). ...

Context 2

... were found at locations with greater canopy cover, slope, woody plant, woody debris, and rocky refuge compared to the random reference locations during the summer season (Table 2). Model averaging indicated the primary covariates and canopy cover were important predictors of site use (Tables 3 and 4). During the fall breeding season, lizards began spreading out away from the more commonly used OHRSF resulting in a greater concentration of locations with grass/forbs during this period. ...

Context 3

... the winter season, we saw a shift to locations with higher percentages of slope and rocky refuge (Table 2). Model averaging indicated the primary covariates were important predictors of winter microhabitat use along with slope to a lesser extent (Tables 3 and 4). For the spring season, we saw similar results to summer with greater canopy cover, slope, woody plant, woody debris, and rocky refuge percentages in used locations. ...