Submitted as: methods for assessment of models 31 Aug 2021

Submitted as: methods for assessment of models | 31 Aug 2021

Review status: this preprint is currently under review for the journal GMD.

A validation standard for Area of Habitat maps for terrestrial birds and mammals

Prabhat Raj Dahal1,2, Maria Lumbierres1,2, Stuart H. M. Butchart2,3, Paul F. Donald2,3, and Carlo Rondinini1 Prabhat Raj Dahal et al.
  • 1Global Mammal Assessment Program, Department of Biology and Biotechnologies, Sapienza University of Rome, Viale dell’Università 32, 00185 Rome, Italy
  • 2BirdLife International, David Attenborough Building, Pembroke Street, Cambridge CB2 3QZ, UK
  • 3Department of Zoology, University of Cambridge, Downing Street, Cambridge CB2 3EJ, UK

Abstract. Area of Habitat (AOH) is a deductive model which maps the distribution of suitable habitat at suitable altitudes for a species inside its broad geographical range. AOH maps have been validated using presence-only data for small subsets of species for different taxonomic groups, but no standard validation method exists when absence data are not available. We develop a novel two-step validation protocol for AOH which includes first a model-based evaluation of model prevalence (i.e, the proportion of suitable habitat within a species’ range), and second a validation using species point localities (presence-only) data. We applied the protocol to AOH maps of terrestrial birds and mammals. In the first step we built logistic regression models to predict expected model prevalence (the proportion of the range retained as AOH) as a function of each species’ elevation range, mid-point of elevation range, number of habitats, realm and, for birds, seasonality. AOH maps with large difference between observed and predicted model prevalence were identified as outliers and used to identify a number of sources of systematic error which were then corrected when possible. For the corrected AOH, only 1.7 % of AOH maps for birds and 2.3 % of AOH maps for mammals were flagged as outliers in terms of the difference between their observed and predicted model prevalence. In the second step we calculated point prevalence, the proportion of point localities of a species falling in pixels coded as suitable in the AOH map. We used 48,336,141 point localities for 4889 bird species and 107,061 point localities for 420 mammals. Where point prevalence exceeded model prevalence, the AOH was a better reflection of species’ distribution than random. We also found that 4689 out of 4889 (95.9 %) AOH maps for birds, and 399 out of 420 (95.0 %) AOH maps for mammals were better than random. Possible reasons for the poor performance of a small proportion of AOH maps are discussed.

Prabhat Raj Dahal et al.

Status: open (until 19 Feb 2022)

Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor | : Report abuse
  • RC1: 'Comment on gmd-2021-245', Anonymous Referee #1, 15 Oct 2021 reply
  • CC1: 'Comment on gmd-2021-245', Matthew Rubino, 28 Oct 2021 reply

Prabhat Raj Dahal et al.

Prabhat Raj Dahal et al.


Total article views: 618 (including HTML, PDF, and XML)
HTML PDF XML Total BibTeX EndNote
532 70 16 618 2 3
  • HTML: 532
  • PDF: 70
  • XML: 16
  • Total: 618
  • BibTeX: 2
  • EndNote: 3
Views and downloads (calculated since 31 Aug 2021)
Cumulative views and downloads (calculated since 31 Aug 2021)

Viewed (geographical distribution)

Total article views: 593 (including HTML, PDF, and XML) Thereof 593 with geography defined and 0 with unknown origin.
Country # Views %
  • 1
Latest update: 17 Jan 2022
Short summary
This paper describes the validation of Area of Habitat (AOH) maps produced for terrestrial birds and mammals. The main objective of the validation was to asses the accuracy of these maps based on independent data. We used the open access data from repositories like ebird and gbif to check if our maps were a better reflection of species' distribution than random. When points were not available we used logistic models to validate the AOH maps. We found majority of AOH maps have high accuracy.