Projecting future expansion of invasive species: comparing and improving methodologies for species distribution modeling

Abstract

Modeling the distributions of species, especially of invasive species in non-native ranges, involves multiple challenges. Here, we developed some novel approaches to species distribution modeling aimed at reducing the influences of such challenges and improving the realism of projections. We estimated species-environment relationships for Parthenium hysterophorus L. (Asteraceae) with four modeling methods run with multiple scenarios of (i) sources of occurrences and geographically isolated background ranges for absences, (ii) approaches to drawing background (absence) points, and (iii) alternate sets of predictor variables. We further tested various quantitative metrics of model evaluation against biological insight. Model projections were very sensitive to the choice of training dataset. Model accuracy was much improved using a global dataset for model training, rather than restricting data input to the species' native range. AUC score was a poor metric for model evaluation and, if used alone, was not a useful criterion for assessing model performance. Projections away from the sampled space (i.e., into areas of potential future invasion) were very different depending on the modeling methods used, raising questions about the reliability of ensemble projections. Generalized linear models gave very unrealistic projections far away from the training region. Models that efficiently fit the dominant pattern, but exclude highly local patterns in the dataset and capture interactions as they appear in data (e.g., boosted regression trees), improved generalization of the models. Biological knowledge of the species and its distribution was important in refining choices about the best set of projections. A post hoc test conducted on a new Parthenium dataset from Nepal validated excellent predictive performance of our 'best' model. We showed that vast stretches of currently uninvaded geographic areas on multiple continents harbor highly suitable habitats for parthenium. However, discrepancies between model predictions and parthenium invasion in Australia indicate successful management for this globally significant weed.

Climatic niche shift of aquatic plant invaders between native and invasive ranges: a test using 10 species across different biomes on a global scale

“…1) and considering previous studies, we suggest the use of occurrence locality records of both native and invasive ranges in ENMs to increase the accuracy of projections of climatic niche distributions of APIs on a global scale (Broennimann and Guisan, 2008;Kolanowska, 2013;Kelly et al, 2014;Fernández and Hamilton, 2015;Mainali et al, 2015). For example, Mainali et al (2015) reported improved prediction performance of ENMs of Parthenium hysterophorus L. (Asteraceae) with the distribution occurrences from both native and invasive ranges. Therefore, it is important to obtain a large volume of occurrences to generate accurate predictions with ENMs (Donaldson et al, 2014;Mainali et al, 2015).…”

“…Hence, the transferability of ENMs supports the necessity for an accurate determination of climatic niche distributions of APIs in invasive ranges based on observed habitat suitability in both native and invasive ranges (Mainali et al, 2015;Wan et al, 2017).…”

“…For example, Mainali et al (2015) reported improved prediction performance of ENMs of Parthenium hysterophorus L. (Asteraceae) with the distribution occurrences from both native and invasive ranges. Therefore, it is important to obtain a large volume of occurrences to generate accurate predictions with ENMs (Donaldson et al, 2014;Mainali et al, 2015). Additionally, our study may provide an important theoretical basis for the future prediction of APIs in invasive ranges based on ENMs and aid further studies that study invasion mechanism(s) of APIs on a global scale.…”

Tracking Invasive Alien Species (TrIAS): Building a data-driven framework to inform policy

“…Assigning alien status to a species is not always straightforward and may be subjected to different types of errors such as misidentification and taxonomic uncertainty (McGeoch et al 2012). The TrIAS partnership brings together the owners of various databases considered to be authoritative sources in Belgium, such as the Manual of Alien Plants of Belgium (Verloove 2016) and the VLIZ Alien Species List for the Belgian part of the North Sea and the Scheldt Estuary (Vandepitte et al 2012).…”

“…Information derived from these data should ideally feed into the risk evaluation process for AS and therefore form a sound scientific basis to guide decision making. Transparent use of available data in risk assessment and decision making is crucial to guarantee reliability, credibility and endorsement of the outcomes by stakeholders and the public (Hattingh 2011, McGeoch et al 2012 and to ensure efficient allocation of available biodiversity conservation budget. In recent years, Belgium has developed valuable decision support tools to inform IAS policy, including information systems, early warning initiatives (waarnemingen.be, observations.be) and risk assessment protocols .…”