Crax alector

NOTES

To elaborate the habitat suitability model we removed four corrupted records placed west of the Andes and three 'Bogota' specimens in BioMap. Also, we removed an accession in BioMap of a specimen allegedly collected in Leticia by Hno. Niceforo Maria in 1957, which doubtfully is correct; Nonetheless, this needs confirmation.

The predictions from the habitat suitability model generated in Maxent seemed relatively restricted to the areas near serrania de La Macarena, the Andean slopes north of it, and a few areas east and in the extreme northeastern Orinoco region. Most of the records for this species come from La Macarena, and apparently the model overfitted predictions to the environmental space within the surrounding areas of this serrania and extended it to the Andean slopes. Our model also predicted as suitable a few pockets west of the Andes. These areas an areas in the northern portion of the eastern slope of the Eastern Andes are not known to be occupied by this curassow and were excluded from its potential distribution map. It is interesting to note that when applied a much lower threshold (FCV1) than the usual we have used (EETOD), the marginally suitable areas increased improving predictions east of the Andes, although incurring in a higher commission error west of the Andes. Nevertheless, these unlikely areas were trimmed off from the final potential distribution map. Additionally, we highlighted not predicted areas lying north of the rio Caqueta that possibly can be occupied by this curassow (sensu Hilty & Brown, 1986).

Assuming that the distribution of the species may have filled the areas predicted as suitable (i.e. marginally suitable, suitable and highly suitable) and those not predicted in the Amazon region of Colombia, its potential distribution today in remnants of forest is about 355,791 km2, which corresponds to a loss of 9 % of its potential original distribution due to deforestation in that region. It is interesting to note that as a whole this figure must be higher, since we do not account for areas around Macarena and in the rio Guayabero where there is a very strong colonisation front that has reduced forested areas severely in the southwestern Orinoco region.

MODEL METADATA

Regularized training gain is 2.591, training AUC is 0.991, unregularized training gain is 3.882.

Algorithm converged after 820 iterations (24 seconds).

The follow settings were used during the run:

17 presence records used for training.

10017 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: 0.500, categorical: 0.250, threshold: 1.830, hinge: 0.500

Feature types used: linear quadratic hinge

responsecurves: true

jackknife: true

maximumiterations: 2000

'Fixed Cumulative Value 1', 'Equal Training Sensitivity and Specificity' and 'Equate Entropy of Thresholded and Original Distributions' thresholds and omission rates:

FCV1-ETSS-EETOD-Description

1-23.583-19.978-Cumulative threshold

0.008-0.157-0.131-Logistic threshold

0.364-0.059-0.075-Fractional predicted area

0-0.059-0-Training omission rate