Assessment of habitat factors and development of a species distribution model for the long nosed potoroo (Potorous tridactylus tridactylus) in SEQ

Trent, Stephen Wade (2015) Assessment of habitat factors and development of a species distribution model for the long nosed potoroo (Potorous tridactylus tridactylus) in SEQ. [USQ Project]

[img]
Preview
Text
StephenTRENT-2015.pdf

Download (5MB) | Preview

Abstract

A species distribution model for the long-nosed potoroo (Potourous tridactylus tridactylus) was developed for South East Queensland based upon known occurrence locations using Maxent software (3.3.3k). Nine environmental predictor datasets reflective of bioclimatic, biophysical and anthropogenic elements were initially compiled and developed for the purpose of comparison against known occurrences of the species. To minimise issues associated with high localised survey bias and spatial autocorrelation resulting in discrete clusters of record locations, occurrence records were initially spatially rarefied. Residual broad geographic survey bias was then addressed via development of a bias grid, based upon 1,106 surveyed sites from a target group species background sample consisting of 26 small native mammal species.

Model performance was based upon the threshold independent measure, the area under the curve (AUC) value, developed from the receiver-operator characteristic curves (ROC). Of the initial nine predictors, a subset of four which excluded strongly correlated variables was found to produce the highest level of discrimination between observed occurrence locations and random background locations. The four variables which were retained were, the 'mean annual temperature', 'low undergrowth vegetation cover', 'potential vegetated habitat extent within 1km of each cell location' and 'mean annual precipitation'. Whilst this combination of predictor variables was found to be highly significant when assessed against 1,000 bias corrected null models, a number of competing models were developed which also exhibited high levels of performance. Further work is required to validate the final suite of variables used to model the focal species distribution.


Statistics for USQ ePrint 37623
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Additional Information: Masters of Spatial Science (Geographic Information Systems) project.
Faculty/School / Institute/Centre: Current - Faculty of Health, Engineering and Sciences - School of Civil Engineering and Surveying (1 July 2013 -)
Faculty/School / Institute/Centre: Historic - International Centre for Applied Climate Sciences (ICACS) (1 Jan 2014 - 31 July 2018)
Supervisors: Apan, Armando
Date Deposited: 06 Jan 2020 00:11
Last Modified: 06 Jan 2020 00:11
Uncontrolled Keywords: species distribution model; long-nosed potoroo
Fields of Research : 09 Engineering > 0909 Geomatic Engineering > 090903 Geospatial Information Systems
06 Biological Sciences > 0602 Ecology > 060207 Population Ecology
URI: http://eprints.usq.edu.au/id/eprint/37623

Actions (login required)

View Item Archive Repository Staff Only