To elaborate our model we removed uncertain localities such as those with no coordinates and all records with coordinates laying down in sites with estimated elevations below 913 m and above 3,262 m.
The habitat suitability model generated in Maxent showed areas suitable in climatic terms for this species in serrania de San Lucas and 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 hummingbird.
Distribution of specimens, according to BioMap, suggests at least two possible areas of intergradation. One possible area (≈ 5,534 km2) of intergradation between subspecies connectens and cervina, located between Cauca and Huila and another possible area (≈ 2,763 km2) of intergradation between subspecies maculata and cervina, located in central southern Cauca.
Two specimens in BioMap from serrania de la Macarena identified as subspecies maculata suggest great variability within subspecies or possibly a different race not described in that area, which requires further research. Another specimen in BioMap from La Paila (Valle del Cauca) suggest great variability within subspecies, although this needs likewise further research.
Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 75,424 km2, which corresponds to a loss of 63 % of its potential original distribution due to deforestation.
Regularized training gain is 1.752, training AUC is 0.949, unregularized training gain is 1.967.
Algorithm converged after 1580 iterations (55 seconds).
The follow settings were used during the run:
175 presence records used for training.
10175 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:
0.114-0.173-Fractional predicted area
0.114-0.029-Training omission rate