Ortalis ruficauda

NOTES

We elaborated a first habitat suitability model in Maxent for this species using all data available. The predictions of this model showed suitable areas in climatic terms for this species in a few pockets in west Colombia, the high Magdalena valley and the extreme northeastern Orinoco region. Also, good part of the central and western Caribbean region were predicted as suitable. These areas are certainly not occupied by this chachalaca and were trimmed off from the final potential distribution map of the species. Nevertheless, we kept areas in the central and western Caribbean region in the model when detailing the marginally suitable, suitable and highly suitable areas for this species. These are of particular interest since that range is occupied by O. garrula, which indicates other factors different to climate have been the underlying cause of their different distributions.

EBIRD distributional data for this species shows a few records in the mid-high Magdalena valley and several in the Orinoco region in Casanare. To improve our model predictions in those areas, we selected five localities that we thought represent the most different or distant sites in comparison to those we already had and repeated the modelling exercise. Those new localities from EBIRD are placed in Puerto Boyaca (Boyaca), El Paujil Natural Reserve (Boyaca/Santander), serrania de Las Quinchas (Santander), Ecolodge Juan Solito-La Aurora and Guaracura (Casanare). New results expanded slightly the distribution of this chachalaca in the mid-high Magdalena valley, east Norte de Santander and from Arauca south into Casanare.

It is interesting to note that in DatAves there are a couple of records from Pivijay (Magdalena) and Mompos (Bolivar) that expand slightly the known distribution of this species to the west. When we elaborated the final distribution maps for this chachalaca we used those to set its limit to the west. Though there is no apparent reason to think these records are erroneous, if they are not correct the distribution of this species would be even more restricted to the east possibly not surpassing the meridian 74° W.

Some authors do not recognise the subspecies lamprophonia from serrania de Macuira (Guajira) as a valid taxon, merging it into ruficrissa (HBW Alive, 2015). However, they consider this still needs further research and for this reason we decided to include it here as a separate subspecies.

According to Hilty & Brown (1986) there are intermediate forms between ruficauda and ruficrissa north of Cucuta. However, distribution of specimens according of BioMap does not give any hint on this; it needs further research.

This species occupies deciduous forest, arid scrub, riparian forest, borders and second growth (Hilty & Brown 1986) and may appear almost anywhere in its delimited potential area of distribution. Therefore, we did not elaborate a map overlapping its distribution with forested areas.

MODEL METADATA

First Model

Regularized training gain is 2.290, training AUC is 0.980, unregularized training gain is 3.019.

Algorithm converged after 780 iterations (21 seconds).

The follow settings were used during the run:

25 presence records used for training.

10025 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: 0.346, categorical: 0.250, threshold: 1.750, 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:

ETSS-EETOD-Description

Second model

Regularized training gain is 1.835, training AUC is 0.965, unregularized training gain is 2.492.

Algorithm converged after 860 iterations (24 seconds).

The follow settings were used during the run:

30 presence records used for training.

10030 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: 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:

ETSS-EETOD-Description