Metallura williami


To elaborate our model we removed uncertain localities such as those localities with no coordinates and all records with coordinates laying down in sites with estimated elevations below 2,100 m and above 4,100 m.

The habitat suitability model generated in Maxent showed some areas suitable in climatic terms for this species in the southern half of the Eastern Andes. Those areas are not known to be occupied by this species and were deleted from our final potential distribution map.

We have portrayed in our maps the distribution of the nominate subspecies as continuous since all Central Andes has Suitable areas climatically for the species. Nevertheless, there is a belt of areas in the central south portion of the cordillera from where there are no known records of the species.

Collections suggest the presence of subspecies primolina in south Nariño, we extended the distribution further north. However, there are no specimens in collection confirming that sightings from those areas are in fact subspecies primolina.

Sensu stricto, subspecies recisa is known from Frontino, there are no more specimens in collection confirming if it is further south. However, we are assuming sightings south to Tatama are this subspecies.

Assuming that the distribution of the species may have filled the complete climatic model generated, its distribution today in remnants of forest is about 10,008 km2, which corresponds to a loss of 67 % of its potential original distribution due to deforestation. This proportion may be slightly overestimated since in some areas of its distribution this species includes Paramo.


Regularized training gain is 3.329, training AUC is 0.990, unregularized training gain is 3.608.

Algorithm converged after 1400 iterations (45 seconds).

The follow settings were used during the run:

100 presence records used for training.

10099 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: hinge product linear threshold quadratic

responsecurves: true

jackknife: true

maximumiterations: 2000

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

10.491-7.203-Cumulative threshold

0.201-0.137-Logistic threshold

0.030-0.036-Fractional predicted area

0.030-0.030-Training omission rate