Biological Invasions

, 13:2409

Predicting the potential invasive range of light brown apple moth (Epiphyas postvittana) using biologically informed and correlative species distribution models

Original Paper

DOI: 10.1007/s10530-011-0052-5

Cite this article as:
Lozier, J.D. & Mills, N.J. Biol Invasions (2011) 13: 2409. doi:10.1007/s10530-011-0052-5


The light brown apple moth (Epiphyas postvittana) is a highly polyphagous species that has invaded several geographic regions across the globe and has stimulated substantial concern over possible impacts for agriculture in the US. We aimed to predict the potential geographic range of E. postvittana to better understand the threat of this species in the US and globally. We used the mechanistic simulation modelling method CLIMEX and the correlative niche modelling method Maxent to predict the geographic distribution of E. postvittana in its native range and globally and tested model projections using known invasion data. Different predictor variable data sets and threshold dependent and independent measures of environmental suitability were considered in model evaluation. Models accurately predicted known invasive localities of E. postvittana across the globe. Overall predictions of environmental suitability were largely congruent across models, although there were some notable differences. Ephiphyas postvittana clearly has the potential to establish in many regions of the globe, although some previous analyses of the potential distribution of this species appear overly pessimistic. Additional studies of the biology of this species in invaded areas, including interactions with natural enemies and the capacity to adapt to novel climatic conditions, are ultimately needed to more fully understand its potential economic and environmental impacts.


Biological invasion CLIMEX Epiphyas postvittana LBAM Maxent Species distribution modelling 

Supplementary material

10530_2011_52_MOESM1_ESM.tif (583 kb)
Fig. S1Binary presence (black)-absence (gray) maps for Maxent-1 (a) and Maxent-2 (b) models after application of lowest presence thresholds showing the overall global projections and enlarged maps of regions with known invasive populations. Native localities used for model training are shown in pink and known invasive localities in red (TIFF 583 kb)
10530_2011_52_MOESM2_ESM.tif (350 kb)
Fig. S2Maxent models trained using 6,618 Epiphyas postvittana trap records (291 used for training, following exclusion of multiple points in the same pixel) from California (Fig. 3) with the Maxent-1 and Maxent-2 environmental variable sets. Background ‘pseudo-absences’ were sampled from a geographic extent including counties where we had information about trapping effort (left-most map in each panel), then projected to all of California or to the native range (upper and lower maps in the right of each panel). Methods otherwise follow those in the main text (TIFF 349 kb)

Copyright information

© Springer Science+Business Media B.V. 2011

Authors and Affiliations

  1. 1.Department of EntomologyUniversity of IllinoisUrbanaUSA
  2. 2.Department of Biological SciencesUniversity of AlabamaTuscaloosaUSA
  3. 3.Department of Environmental Science, Policy and Management, College of Natural ResourcesUniversity of CaliforniaBerkeleyUSA

Personalised recommendations