To elaborate our model we removed uncertain localities such as those localities with no coordinates. Otherwise, we tried to locate the headwaters of Rio Salaqui based on a paper of Haffer (1970) in Caldasia vol. X - No. 50 at 1,000 m, although assigned locality is relatively uncertain. From there we produced four pseudolocalities around the original locality at approximate 2 km north, east, south and west.
The habitat suitability model generated in Maxent showed areas suitable in climatic terms for this species in the northwestern slope of the Sierra Nevada de Santa Marta and in the northern section of the eastern slope of the Eastern Andes, also in the zone of cerro Tacarcuna in the border with Panama. These areas are not known to be occupied by the species and were excluded from the potential distribution map of this hummingbird.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 299 km2, which corresponds to a loss of 11 % of its potential original distribution due to deforestation.
BirdLife International has categorised this species as Near Threatened (NT) because it has a very restricted range, where the habitat is still relatively pristine. Its range has been estimated in 5,100 km2 (BirdLife International 2017). Our maps suggests an extent of occurrence of 337 km2 in Colombia. It is considered uncommon and very local along its distribution (McMullan & Donegan, 2014; BirdLife International, 2017). This species is in great need of further research of its ecology to define better if there are any threats to its populations. There is not suggestion that its habitat is threatened at present time. Nonetheless, given the highly restricted range in Colombia and how little is known of its ecology, we consider it must be uplisted as Vulnerable (VU), until it is clarified the status of its populations.
Regularized training gain is 3.697, training AUC is 0.998, unregularized training gain is 4.155.
Algorithm converged after 300 iterations (5 seconds).
The follow settings were used during the run:
5 presence records used for training.
10005 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: 1.000, categorical: 0.575, threshold: 1.950, hinge: 0.500
Feature types used: linear
'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.003-0.025-Fractional predicted area
0.000-0.000-Training omission rate