To elaborate our model we removed uncertain localities such as 'Bogota skins', those localities with no coordinates and all records with coordinates laying down in sites with estimated elevations above 2,200 m.
The habitat suitability model generated in Maxent showed areas suitable in climatic terms for this species in most of the eastern slope of the Andes. Exception the northern Eastern Andes, these areas are not known to be occupied by the species and were excluded from the potential distribution map of this hummingbird. Some areas in the Pacific lowlands (likely suitable) were unpredicted and were noted as such in our maps. Likewise, lowland areas north of the Western and Central Andes between Uraba and serrania de San Lucas possibly are in this same category.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 133,845 km2, which corresponds to a loss of 58 % of its potential original distribution due to deforestation. Nonetheless, this species favours edges, secondary vegetation and gardens and therefore possibly deforestation has not negatively affected greatly its populations.
Geographical distribution limits of subspecies are relatively unclear, with wide areas where appears there is intergradation or hybridisation between the several subspecies present in the country. Distribution of specimens according to BioMap suggest a possible area (46,887 km2) of intergradation between subspecies colombica, subtropicalis and fannyae in most of the Cauca Valley. Also a possible area (19,700 km2) of intergradation between subspecies fannyae and verticeps in west Cauca and central Nariño.
Regularized training gain is 1.344, training AUC is 0.923, unregularized training gain is 1.549.
Algorithm terminated after 2000 iterations (73 seconds).
The follow settings were used during the run:
551 presence records used for training.
10546 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:
0.157-0.262-Fractional predicted area
0.158-0.045-Training omission rate