Table 1 Potential predictors considering environmental conditions and resources within climate, habitat, and soil variables.

Climate

Based on averaged monthly air temperature in C°: annual mean (mean), annual maximum (max), annual standard deviation (sd)

Landsat-8 scene, 30 m spatial resolution

After ref.⁸⁰

Humidity in %: mean annual

Landsat-8 scene, 30 m spatial resolution

After ref.⁸¹

Habitat

Normalized difference vegetation index (NDVI)

Landsat-8 scene, 30 m spatial resolution

(near infrared − red)/(near infrared + red)

Image textural metric: NDVI correlation

Landsat-8 scene, 30 m spatial resolution

‘glcm’ package in R; using all directions^82,83

Red-Blue normalized difference vegetation index (RBVI)

Landsat-8 scene, 30 m spatial resolution

(red − blue)/(red + blue)

Forest cover in %

Landsat-8 scene, 30 m spatial resolution

Spectral unmixing in IDRISI Andes after⁹²

Topographical position index (TPI)

Sigtierras image, 6 m spatial resolution

After ref.⁹³ with a surrounding of 17 pixels

Leaf area index (LAI)

Sentinel-2 scene, 10 m spatial resolution

S2 SNAP (Sentinel Application Platform) Toolbox Biophysical Processor⁹⁴

Soil

pH in Bv-Horizon

30 m spatial resolution

Model prediction; Extended Data Tables E1, E2

Total phosphorus soil in kg/ha in Ah-Horizon (phosphorus)

30 m spatial resolution

Model prediction; Extended Data Tables E1, E2

Organic layer depth in cm

30 m spatial resolution

Model prediction; Extended Data Tables E1, E2