The habitat suitability model generated in Maxent showed areas that are suitable in climatic terms for this species in Sierra Nevada de Santa Marta. These areas are not known to be occupied by the species and were excluded from the potential distribution map of this guan.
To elaborate our model, we removed all records with coordinates laying down in sites with estimated elevations out of the range within 1,700-3,700 m.
There are three records from Subia (Cundinamarca) in BioMap that have been identified erroneously as the subspecies brooki and must be the nominate subspecies; this needs revision. Collected in 1913, they have been certainly reidentified erroneously to brooki because this last subspecies was described in 1917. It is interesting to note that one of these specimens have a note that reads 'Only genus...', and therefore it is not clear if this identification was done in the collection or if in some way data were corrupted in the database while compiling the datasets.
Distribution of specimens, according to BioMap, suggests a possible narrow area (≈ 4,544 km2) of intergradation between subspecies atrogularis and montagnii in southwestern Huila to central southeastern Nariño. Most specimens in this area have been collected after the description of both subspecies involved and we believe it is unlikely they are erroneously identified. Additionally, there are a few notes in BioMap that seem to indicate several birds identified as the nominate in that area have some features of atrogularis. On the other hand, its is not clear how far north goes the distribution of the subspecies brooki in Colombia. Its limit must be between the furthest record to the north (Santiago, Putumayo) and the limit assigned in its distribution map in this study.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 42,570 km2, which corresponds to a loss of 67 % of its potential original distribution due to deforestation.
Regularized training gain is 2.255, training AUC is 0.971, unregularized training gain is 2.503.
Algorithm converged after 660 iterations (18 seconds).
The follow settings were used during the run:
55 presence records used for training.
10055 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: 0.179, categorical: 0.250, threshold: 1.450, 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:
0.091-0.105-Fractional predicted area
0.091-0.055-Training omission rate