Abstract:
Over the past few decades, species distribution modelling has been increasingly used to
monitor invasive species. Studies herein propose to use Cellular Automata (CA), not only to model
the distribution of a potentially invasive species but also to infer the potential of the method in risk
prediction of Reticulitermes grassei infestation. The test area was mainland Portugal, for which an
available presence-only dataset was used. This is a typical dataset type, resulting from either
distribution studies or infestation reports. Subterranean termite urban distributions in Portugal
from 1970 to 2001 were simulated, and the results were compared with known records from both
2001 (the publication date of the distribution models for R. grassei in Portugal) and 2020. The
reported model was able to predict the widespread presence of R. grassei, showing its potential as a
viable prediction tool for R. grassei infestation risk in wooden structures, providing the collection of
appropriate variables. Such a robust simulation tool can prove to be highly valuable in the decisionmaking
process concerning pest management.
