Table A6.
List of the remaining environmental variables after the execution of the reduceVar function (Figure A2) and relative percent contribution rounded to the first decimal place
| Variable | Percent contribution | |
|---|---|---|
| optimizeModel | gridSearch | |
| bush_fr | – | 0.9 |
| compact_limestone_fr | 7.7 | 6.8 |
| c_way_d_av | 1.7 | 1.9 |
| eastness_av | 1.4 | 1.4 |
| forest_fr | 1.3 | 1.4 |
| gneiss_fr | 1.5 | 1.7 |
| granite_fr | 1.5 | 1.3 |
| ibex_summer_density_av | 19.0 | 18.5 |
| ibex_winter_density_av | 6.8 | 7.0 |
| northness_av | 6.8 | 6.8 |
| other_rocks_fr | 2.2 | 3.2 |
| other_soi_fr | – | 0.8 |
| permanent_fr | 0.7 | 0.7 |
| prec_122_av | 24.6 | 24.2 |
| prec_57_av | 2.5 | 2.4 |
| roughness | 1.5 | 1.0 |
| rupicapra_hs_av | 3.5 | 3.5 |
| scree_fr | 9.0 | 9.9 |
| ski_d_av | 1.3 | 1.3 |
| solar_rad57_av | 5.2 | 3.4 |
| wind_100_av | 1.7 | 1.9 |
The model parsimony optimization was performed based on the output of the optimizeModel and gridSearch functions respectively, executed to tune 1200 possible combinations of hyperparameters.