To elaborate the model of this species we removed five records in BioMap from Cundinamarca; three of them Bogota skins and two that represent corrupted data. Additionally, we did not use records for which their locality coordinates lay down at elevations out of the range of altitudes recorded in the specimens' labels (i.e. 800-3,100 m).
The habitat suitability model generated in Maxent predicted extensive areas in the three Andean cordilleras; except the northern end of the Eastern Andes and the eastern slope of the same cordillera in Boyaca. Most areas predicted in the Eastern Andes are not known to be occupied by this wood-quail and were trimmed off from the final version of our potential distribution map. Additionally, we trimmed areas below 800 m and above 3,100 m of elevation.
Assuming that the distribution of the species may have filled the areas predicted as suitable (i.e. marginally suitable, suitable and highly suitable), its potential distribution today in remnants of forest is about 43,205 km2, which corresponds to a loss of ≈ 61 % of its potential original distribution due to deforestation. However, if we consider a more restricted distribution to this species, excluding potential areas north of Andalucia (Huila) from were there are no known records for this species, its potential distribution today in remnants of forest is about 35,257 km2, which corresponds to a loss of ≈ 64 % of its potential original distribution due to deforestation.
Regularized training gain is 2.108, training AUC is 0.970, unregularized training gain is 2.464.
Algorithm converged after 800 iterations (22 seconds).
The follow settings were used during the run:
68 presence records used for training.
10067 points used to determine the Maxent distribution (background points and presence points).
Environmental layers used (all continuous): bio10co bio11co bio12co bio13co bio14co bio15co bio16co bio17co bio18co bio19co bio1co bio2co bio3co bio4co bio5co bio6co bio7co bio8co bio9co
Regularization values: linear/quadratic/product: 0.141, categorical: 0.250, threshold: 1.320, hinge: 0.500
Feature types used: linear quadratic hinge
'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.08-0.122-Fractional predicted area
0.074-0.015-Training omission rate