Coordinates from eight accessions in BioMap from the Munchique area and Frontino (located in places with elevations above 2,100 m) were moved slightly (centecimals of a degree) to make localities lay down in sites fitting better the elevation specified in each record or, in cases were the observed elevation was missing, to keep the specific locality within the upper elevational limit of the species (≈ 1,600 m). It is of special interest the locality 'La Costa' from where exist many specimens of K. von Sneidern. This locality it is known to be in the Pacific slope of cerro Munchique (Paynter, 1997), with specimens being collected at 1,000-1,800 m (BioMap, 2015), although it is not clear where exactly it is. It was previously approximated to the geographical centre of cerro Muchique in BioMap, but this is very high point (≈ 2,500 m). Now it has been corrected and approximated to the Pacific slope of the cerro to an elevation of 1,200 m near the road from 'El Tambo' to the Pacific where most likely von Sniedern was collecting.
The habitat suitability model generated in Maxent predicted extensive areas in Sierra Nevada de Santa Marta, serrania del Perija, the slopes above the Cauca valley, the slopes above the mid and high Magdalena valley and the slopes of the Eastern Andes. These areas are not known to be occupied by this wood-quail and were trimmed off from the final version of our potential distribution map.
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 89,705 km2, which corresponds to a loss of ≈ 50 % of its potential original distribution due to deforestation.
Regularized training gain is 1.667, training AUC is 0.949, unregularized training gain is 2.067.
Algorithm converged after 440 iterations (12 seconds).
The follow settings were used during the run:
25 presence records used for training.
10024 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: 0.346, categorical: 0.250, threshold: 1.750, 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:
0.146-0.188-Fractional predicted area
0.16-0.04-Training omission rate