Heliangelus clarisse

NOTES

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 2,200 m and above 3,244 m. Also a record in BioMap from Bahia de Malaga (at low elevations) was deleted previous to modelling. It is important to highlight that three further records from Munchique (Cauca) in BioMap, which represent corrupted information were left by mistake during the modelling process. In spite of this, Maxent predictions were very robust and these area was excluded from presence predictions.

The habitat suitability model generated in Maxent showed a few areas suitable in climatic terms for this species in Sierra Nevada de Santa Marta, the northern Central Andes and south of the Colombian Masiff. 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 7,109 km2, which corresponds to a loss of 77 % of its potential original distribution due to deforestation. Although this species uses edges and secondary vegetation it needs forest to certain extent and therefore possibly deforestation has affected its populations. However, there is no clarity on the status of its populations. BirdLife International (2016) recognises the species has a restricted range (Colombia and Venezuela) but does not believe it approaches to the thresholds to consider it a threatened species. In the specific case of Colombia it is possible that this species approaches the thresholds to be considered Near Threatened given the small range and fragmented habitat, and possibly its status needs re-evaluation.

The Perija subspecies 'violiceps' is a highly threatened form since its range is extremely small and fragmented in Colombia.

MODEL METADATA

Regularized training gain is 3.191, training AUC is 0.990, unregularized training gain is 3.538.

Algorithm converged after 660 iterations (24 seconds).

The follow settings were used during the run:

30 presence records used for training.

10029 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: 0.250, categorical: 0.250, threshold: 1.700, hinge: 0.500

Feature types used: hinge linear quadratic

responsecurves: true

jackknife: true

maximumiterations: 2000

'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:

ETSS-EETOD-Description

10.112-8.367-Cumulative threshold

0.178-0.15-Logistic threshold

0.037-0.041-Fractional predicted area

0.033-0.033-Training omission rate