Skip to main content

Table 3 Results of species distribution model (SDM) performance for Bioclim, generalized linear model (GLM), MaxEnt, boosted regression tree (BRT), and random forest (RF) algorithms in predicting occurrence of Northern Bobwhite (NOBO), Eastern Meadowlark (EAME), and Cassin’s Sparrow (CASP) on East Foundation’s Coloraditas Grazing Research and Demonstration Area (CGRDA) in the summer of 2016. Model performance metrics (area under curve [AUC] and true sensitivity statistic [TSS]) are compared for SDMs using environmental predictors only, and environmental predictors stacked with a raster representing localized grazing pressure (denoted by “+”).

From: Modeling the influence of livestock grazing pressure on grassland bird distributions

  BioClim GLM MaxEnt BRT RF
AUC TSS AUC TSS AUC TSS AUC TSS AUC TSS
NOBO   0.54 0.032 0.64 0.034 0.61 0.056 0.59 0.049 0.81 0.42
+ 0.58 0.96 0.67 0.24 0.67 0.18 0.60 0.57 0.84 0.48
EAME   0.81 0.41 0.78 0.62 0.79 0.61 0.84 0.46 0.91 0.62
+ 0.81 0.49 0.78 0.80 0.78 0.75 0.89 0.80 0.95 0.67
CASP   0.58 0.13 0.44 0.18 0.69 0.24 0.62 0.21 0.78 0.22
+ 0.62 0.15 0.44 0.18 0.71 0.29 0.64 0.67 0.81 0.23