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 2,426 m and above 3,871 m.
The habitat suitability model generated in Maxent showed very good fit not showing suitable areas out of the main Andean Ranges.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 11,365 km2, which corresponds to a loss of 77 % of its potential original distribution due to deforestation.
This species has a large extent of occurrence, estimated in 233,000 km2 and therefore it is considered of Low Concern given it does not approach to the thresholds for Vulnerable under the range size, population trend or population size criterion (BirdLife International 2016). Nevertheless, interestingly, considering the major proportion of its habitat that has been degraded and looking at it at subspecies level we can see that in Colombia it may emerge a different picture. The nominate subspecies has a potential extent of occurrence of 25,669 km2 and of this it may occur in some 4,404 km2 remnants of forest, which corresponds to a loss of 83 % of its potential original distribution; slightly overestimated given areas of paramo above 3,500 m? Subspecies caeruleus has a potential extent of occurrence of 23,519 km2 and of this it may occur in some 6,961 km2 remnants of forest, which corresponds to a loss of 70 % of its potential original distribution. Giving the fact that the extent of forested areas are well below the threshold of Vulnerable (VU) that is 20,000 km2 and that the habitat of this species still continues to be threatened by deforestation and fragmentation in Colombia, possibly independently as subspecies both the nominate and caeruleus may be considered of certain importance for conservation in Colombia. Particularly, the nominate subspecies, which apparently may have been extirpated from a larger area.
Regularized training gain is 2.992, training AUC is 0.986, unregularized training gain is 3.245.
Algorithm converged after 1780 iterations (54 seconds).
The follow settings were used during the run:
116 presence records used for training.
10116 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:
0.041-0.050-Fractional predicted area
0.043-0.017-Training omission rate