PS 65-60
Potential global spread of a large predator, the Nile monitor lizard (Varanus niloticus)

Thursday, August 13, 2015
Exhibit Hall, Baltimore Convention Center
Hannah R. Perkins, Biology, University of Central Florida, Orlando, FL
David G. Jenkins, Biology, University of Central Florida, Orlando, FL

The pet trade is an infamous source of invasive species; the Nile monitor lizard (Varanus niloticus) is no exception. This large, generalist carnivore is native to sub-Saharan Africa and the Nile River but is frequently exported and bred across the globe. Nile monitors burrow, swim, climb trees and consume many different invertebrate and vertebrate species. The predator established in Florida in 1990 and may establish in other regions given release from captivity (as occurred in Florida) and its wide African range (30.9°N to -33.9°S). Here we analyze the potential current and future (2050) spread of the Nile monitor on a global scale using ensemble modeling (Biomod2 in R), based on 507 geo-referenced African observations. Projections represent alternative hypotheses based on climatic (temperature and precipitation, from BIOCLIM), vegetation (from MODIS) and/or elevation (from SRTM) variables. Alternative hypotheses were evaluated using TSS, sensitivity, specificity and ROC scores at a quality threshold of 0.7. Background information was incorporated through the appropriate selection of pseudo absences based on the models used.


ROC and TSS scores for current and future models indicate successful results (each >0.9) that effectively accounted for omission (sensitivity) and commission (specificity) errors (each >90%). Consensus range projections to date indicate the Nile monitor will spread into many subtropical and tropical regions of North America, South America, Australia and Asia, consistent with its African range. Models for 2050 show similar spread with limited additional expansion. Projected ranges overlap greatly with multiple biodiversity hotspots where the Nile monitor could threaten multiple vulnerable and endemic species. Based on these results, we recommend subtropical and tropical countries ban the import and internal sales of this large predator before introductions and spread occur. Further model refinements may modify range projections and will identify factors that best predict the future range of this large predator if it continues to be spread via the pet trade.