This species was recently reported in Colombia by Olaciregui & Guzman (2011, Cons. Col. 15) in the valley of the Chingual river in the border between Colombia and Ecuador, between 1,300-1,700 m. Original coordinates reported by the authors were moved less than 3/1,000 of a degree to the west to agree with the elevation given in Cons. Col. Later to elaborate the model, these coordinates were moved about 14 km southeast to fit the locality within the mask of Colombia used in our analyses. This was necessary given a small inherent geographical uncertainty in the representation of the border between Colombia and Ecuador in our maps. Thus, we moved coordinates to the nearest site where elevation was similar to that given by Olaciregui & Guzman (2015). Additionally, we derived three more pseudo-localities placing them at approximately 1/100 of a degree to the southwest, south and southeast; all within the range of elevations specified by the authors.
The habitat suitability model generated in Maxent predicted areas in the southeastern portion of the Eastern Andes and south. Nevertheless, areas north of extreme northeastern Putumayo are unlikely occupied by this species and were trimmed off from our final potential distribution map. Similarly, we trimmed areas below 1,300 m and above 1,700 m of elevation. Otherwise, we highlighted in our Google map the immediate area of the Chingual river where the species was reported and towards the contiguous valley to the east, which is excluded from our mask for the reasons explained, but that very likely have populations of this species. Most of the area predicted as suitable for the species in Colombia is covered by unbroken forest.
Regularized training gain is 4.997, training AUC is 1.000, unregularized training gain is 6.371.
Algorithm converged after 180 iterations (1 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:
0.001-0.007-Fractional predicted area
0-0-Training omission rate