Table 1.
Results of multiple regression on distance matrices and additional parameters derived from commonality analysis
| ka | r | β | U | C | |
|---|---|---|---|---|---|
| FCA | |||||
| Cattle density [C]b | 1000 | 0.1733 | 0.1551 | 0.0221 | 0.0079 |
| Grassland [R]c | 1000 | 0.1080 | 0.0637 | 0.0037 | 0.0079 |
| aR | |||||
| Elevation [R]c | 100 | 0.0254 | − 0.9945 | 0.00044 | 0.0002 |
| Elevation [R]c | 1000 | 0.0291 | 0.9860 | 0.00048 | 0.0004 |
| Grassland [R]c | 100 | 0.0448 | 0.2279 | 0.00005 | 0.0020 |
| Grassland [R]c | 1000 | 0.0539 | − 0.1483 | 0.00002 | 0.0029 |
| Urban areas [R]c | 100 | 0.0725 | − 0.1279 | 0.00014 | 0.0051 |
| Urban areas [R]c | 1000 | 0.0879 | 0.2156 | 0.00049 | 0.0072 |
Pearson’s correlation coefficient (r), significant regression coefficient (β), as well as unique (U) and common (C) contributions of environmental distances to the variance in the dependent variable are shown; FCA factorial correspondence analysis, aR Rousset’s distance
aRe-scaling parameter used to transform the initial raster file
bThe considered environmental raster was treated as a conductance [C] factor
cThe considered environmental raster was treated as a resistance [R] factor