To elaborate the habitat suitability model we removed a corrupted record from Munchique in BioMap. Additionally, we added during the modelling exercise a record of a specimen collected in 'Estacion Biologica El Guafal' (Arauca/Casanare) and deposited in the ornithology collection of the IAvH.
The habitat suitability model generated in Maxent showed extensively areas that are suitable in climatic terms for this species in most of tropical central, north, east and northern southeast Colombia. Most of these areas are not known to be occupied by this curassow and were excluded from its potential distribution map. Using the known localities we included in the final potential distribution map the areas predicted as marginally suitable, suitable and highly suitable in the serrania de Perija, the eastern slope of Sierra Nevada de Santa Marta and the Catatumbo. Equally, we included areas predicted as suitable and highly suitable in the northern portion of the Orinoco region.
Assuming that the distribution of the species may have filled the areas predicted as suitable (i.e. marginally suitable, suitable and highly suitable) in the zone between serrania de Perija-Sierra Nevada de Santa Marta-Catatumbo, its potential distribution today in remnants of forest is about 9,086 km2, which corresponds to a loss of 66 % of its potential original distribution due to deforestation. It is interesting to note that most of this deforestation has occurred from where it came 75 % of the records of this species in the country (Perija and Sierra Nevada de Santa Marta). This curassow is mostly distributed in Venezuela, north of the Orinoco (HBW Alive, 2015) and the areas where it has been registered or possibly occurs in Colombia represents the outermost portion of its distribution to the west, making it apparently a very rare species in Colombian territory.
Regularized training gain is 0.436, training AUC is 0.890, unregularized training gain is 0.901.
Algorithm converged after 80 iterations (1 seconds).
The follow settings were used during the run:
6 presence records used for training.
10006 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: 0.560, threshold: 1.940, hinge: 0.500
Feature types used: linear
'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:
0.186-0.647-Fractional predicted area
0.167-0-Training omission rate