To elaborate our model we removed uncertain localities such as 'Bogota skins' and all records with coordinates laying down in sites with estimated elevations above 1,144 m. Also four records in BioMap, two from Cundinamarca, one from Santander and one from Quindio, and a further record in DatAves from Putumayo but with coordinates misplaced in Choco, which represent corrupted data were removed previously to modelling.
The habitat suitability model generated in Maxent showed areas suitable in climatic terms for this species west of the Andes. These areas are not known to be occupied by the species and were excluded from the potential distribution map. Otherwise, prediction of areas east of the Andes was relatively poor above the EETOD threshold, below it but above the 5 percentile areas predicted in both Orinoco and Amazon regions increased, although there were extensive areas not predicted by the model.
Regularized training gain is 1.362, training AUC is 0.902, unregularized training gain is 1.861.
Algorithm converged after 780 iterations (21 seconds).
The follow settings were used during the run:
33 presence records used for training.
10033 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: 0.241, categorical: 0.250, threshold: 1.670, hinge: 0.500
Feature types used: hinge linear quadratic
'Fixed Cumulative Value 5%', 'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.553-0.185-0.255-Fractional predicted area
0.030-0.182-0.182 Training omission rate