There are two sightings in DatAves that allegedly were collected in the Pacific lowlands of Antioquia and Valle del Cauca, which clearly are erroneous records in the database. These records were excluded when modelling the distribution of this species. In spite of that, the habitat suitability model generated in Maxent showed a few zones in the northern portion of the Pacific and in the Caribbean where exist suitable areas in climatic terms for the species. Since it is well established this zones are not occupied by the species, those areas were removed from the potential distribution map. On the other hand, the model failed to predict areas south of Vaupes, possibly due to the lack of records coming from that particular area in Colombia. Nevertheless, it is suspected the species definitely can be there since there are abundant sightings in Ecuador near the border (EBIRD), and therefore we decided to include that area in the potential distribution for the species in the country.
Assuming that the distribution of the species may have filled the complete climatic model generated in the Amazon region plus the area south of Vaupes that was not predicted by the model, the distribution today in remnants of forest is about 476,805 km2, which corresponds to a loss of ≈ 10 % of its potential original distribution due to deforestation.
Regularized training gain is 0.878, training AUC is 0.912, unregularized training gain is 1.469.
Algorithm converged after 540 iterations (15 seconds).
The follow settings were used during the run:
20 presence records used for training.
10020 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: 0.442, categorical: 0.250, threshold: 1.800, 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:
0.2-0.416-Fractional predicted area
0.2-0-Training omission rate