Pipile cumanensis


There are six 'Bogota' skins in BioMap probably collected near Villavicencio and one sighting in Antioquia in ProAves Database, which is clearly erroneous. These accessions were removed when modelling the distribution of this species. On the other hand, to produce the model we added one more record from San Teodoro (Vichada) found in EBIRD.

The habitat suitability model generated in Maxent showed a few areas that are suitable in climatic terms for this species west of the Andes in Sierra Nevada de Santa Marta, the northern end of the three cordilleras, the high Patia valey and some areas in the mid-high Magdalena valley. These areas are known not to be occupied by the species and were excluded from the potential distribution map of this guan.

Assuming that the distribution of the species may have filled the complete climatic model generated in the Amazon region, its distribution today in remnants of forest is about 475,080 km2, which corresponds to a loss of 9 % of its potential original distribution due to deforestation in that region of Colombia.


Regularized training gain is 1.170, training AUC is 0.911, unregularized training gain is 2.279.

Algorithm converged after 980 iterations (26 seconds).

The follow settings were used during the run:

16 presence records used for training.

10015 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: 0.543, categorical: 0.286, threshold: 1.840, hinge: 0.500

Feature types used: linear quadratic hinge

responsecurves: true

jackknife: true

maximumiterations: 2000

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


33.258-17.564-Cumulative threshold

0.304-0.198-Logistic threshold

0.164-0.31-Fractional predicted area

0.188-0.125-Training omission rate