To elaborate our model we removed uncertain localities such as those localities with no coordinates and all records with coordinates laying down in sites with estimated elevations below 850 m and above 3,100 m.
The habitat suitability model generated in Maxent showed some areas suitable in climatic terms for this species in the serranias del Baudo, San Lucas and La Macarena. Those areas are not known to be occupied by this species and were deleted from our final potential distribution map.
Distribution of specimens according to BioMap suggest a possible zone of intergradation (15,068 km2) between subspecies caucensis and the nominate subspecies in several portions along the Central Andes, particularly in the central north and southern sections of this cordillera.
Our model suggest there are extensively areas suitable in climatic terms for this species in the southwestern Andes of Colombia in Cauca and Nariño. Nevertheless, there are no known records in collection that can confirm, if present, which subspecies might occur.
There are four records in BioMap from cerro Tacarcuna identified as subspecies caucensis, which most likely are misidentified an must be referable to subspecies floccus. Additionally, our model predicted a few suitable areas in the mountainous section of the border with Panama in the area of cerro Pirre, where it may occur subspecies galindoi. Both subspecies are not yet recorded in our checklist and must be added (ABC 2016).
Our checklist (ABC 2016) suggests the unconfirmed possibility that subspecies russata is present in extreme south Colombia in the border with Ecuador. Differently, McMullan & Donegan (2014) suggest the presence of russata in their distribution maps as certain. Nevertheless, distribution of specimens determined as the nominate subspecies according to BioMap suggest it is unlikely that russata exists in Colombia.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 61,110 km2, which corresponds to a loss of 59 % of its potential original distribution due to deforestation.
Regularized training gain is 1.959, training AUC is 0.958, unregularized training gain is 2.148.
Algorithm converged after 1140 iterations (44 seconds).
The follow settings were used during the run:
382 presence records used for training.
10380 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: 0.050, categorical: 0.250, threshold: 1.000, hinge: 0.500
Feature types used: hinge product linear threshold quadratic
'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.097-0.141-Fractional predicted area
0.097-0.024-Training omission rate