Crypturellus soui

NOTES

We elaborated a first habitat suitability model in Maxent for the whole species, the predictions of this model performed fairly well in the Caribbean and Andean regions, whilst missed areas in the Pacific and predicted very poorly potential areas east of the Andes. We obtained better results producing a collage of that first model joined together with the results from individual models generated for each of the five subspecies confirmed in the country.

There are a few records from the area northwest of Sierra Nevada de Santa Marta in BioMap that have been identified erroneously as the nominate subspecies and must be mustellinus; those records were eliminated for the modelling exercises since they represented anyway replicated localities.

Distribution of specimens, according to BioMap, suggests a possible area (≈ 17,593 km²) of intergradation between subspecies harterti and caucae in northwest Colombia. Most specimens in this area have been collected well after the description of both subspecies involved and we believe it is unlikely they are erroneously identified. Similarly, there are three small possible areas (≈ 5,218 km²) of intergradation between subspecies caucae and mustellinus in northwestern half of the Eastern Andes which need further study.

This species is common in secondary forest and riparian forests, and therefore we did not overlapped its potential distribution with forested areas since it might appear almost anywhere in the country where there are forest remnants in different stages of succession.

Some authors have reported the subspecies panamensis as present in extreme northwest Colombia in the border with Panama (HBW Alive). However, there are no specimens or confirmed records pointing to this.

MODEL METADATA

Crypturellus soui

Regularized training gain is 0.757, training AUC is 0.869, unregularized training gain is 1.026.

Algorithm converged after 1520 iterations (52 seconds).

The follow settings were used during the run:

225 presence records used for training.

10224 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: 0.050, categorical: 0.250, threshold: 1.000, hinge: 0.500

Feature types used: product linear quadratic hinge threshold

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

31.453-7.302-Cumulative threshold

C. soui soui

Regularized training gain is 0.027, training AUC is 0.727, unregularized training gain is 0.151.

Algorithm converged after 80 iterations (0 seconds).

The follow settings were used during the run:

4 presence records used for training.

10004 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: 0.590, threshold: 1.960, hinge: 0.500

Feature types used: linear

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

C. soui mustellinus

Regularized training gain is 1.927, training AUC is 0.971, unregularized training gain is 2.429.

Algorithm converged after 660 iterations (19 seconds).

The follow settings were used during the run:

32 presence records used for training.

10032 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: 0.244, categorical: 0.250, threshold: 1.680, 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

C. soui caucae

Regularized training gain is 1.244, training AUC is 0.916, unregularized training gain is 1.462.

Algorithm converged after 560 iterations (16 seconds).

The follow settings were used during the run:

56 presence records used for training.

10055 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: 0.176, categorical: 0.250, threshold: 1.440, 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

C. soui harterti

Regularized training gain is 1.189, training AUC is 0.921, unregularized training gain is 1.478.

Algorithm converged after 180 iterations (1 seconds).

The follow settings were used during the run:

12 presence records used for training.

10012 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: 0.714, categorical: 0.429, threshold: 1.880, hinge: 0.500

Feature types used: 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

C. soui caquetae

Regularized training gain is 0.505, training AUC is 0.908, unregularized training gain is 0.923.

Algorithm converged after 200 iterations (1 seconds).

The follow settings were used during the run:

14 presence records used for training.

10014 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: 0.629, categorical: 0.357, threshold: 1.860, hinge: 0.500

Feature types used: 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