To elaborate the model of this species we removed seven Bogota skins' records in BioMap and one more record in this same dataset from San Juan de Rio Seco (Cundinamarca) west of Bogota, allegedly collected by brother Niceforo in 1923. Apparently this last record is correct, suggesting it well might be the species once inhabited subtropical forests west of Bogota, where therefore most of Bogota skins may have been collected. BirdLife International (2015) points out that the species has not been recorded in Cundinamarca since 1954. Some authors believe the elevational range of this species is between 1,500 and 2,500 or 3,000 m (BirdLife International, 2015; Mcmullan & Donegan, 2014), which is most likely incorrect. The records we have suggests this wood-quail is mostly subtropical and must have been restricted to the belt of forest between approximately 1,300 m to 2,300 m.
The habitat suitability model generated in Maxent predicted extensive areas in the three Andean cordilleras and in some lowlands in the Caribbean, Amazon and Pacific regions, which constitute extremely odd predictions. Thus, our final potential distribution map was limited to areas between 1,300-2,300 m in the western slope of the Eastern Andes in Santander, Boyaca and Cundinamarca. Furthermore, we restricted the distribution to the most likely areas in Santander and Boyaca, pointing out to other climatically suitable areas that possibly were occupied in the past.
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 3,376 km2, which corresponds to a loss of ≈ 57 % of its potential original distribution due to deforestation; this figure increases to 62 % if we include those areas climatically suitable that possibly were occupied in Cundinamarca and Santander in the past.
Regularized training gain is 1.258, training AUC is 0.943, unregularized training gain is 1.590.
Algorithm converged after 80 iterations (0 seconds).
The follow settings were used during the run:
4 presence records used for training.
10004 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: 1.000, categorical: 0.590, threshold: 1.960, hinge: 0.500
Feature types used: linear
'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.087-0.284-Fractional predicted area
0-0-Training omission rate