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

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 with four modeling methods run with multiple scenarios of (1) sources of occurrences and geographically isolated background ranges for absences, (2) approaches to drawing background (absence) points, and (3) 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 by 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 Partenium 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 hysterophorus L. (Asteraceae; parthenium). However, discrepancies between model predictions and parthenium invasion in Australia indicate successful management for this globally significant weed. This article is protected by copyright. All rights reserved.

Bactrocera frauenfeldi (Diptera: Tephritidae), an invasive fruit fly in Australia that may have reached the extent of its spread due to environmental variables

Bactrocera frauenfeldi (Schiner), the ‘mango fruit fly’, is a horticultural pest originating from the Papua New Guinea region. It was first detected in Australia on Cape York Peninsula in north Queensland in 1974 and had spread to Cairns by 1994 and Townsville by 1997. Bactrocera frauenfeldi has not been recorded further south since then despite its invasive potential, an absence of any controls and an abundance of hosts in southern areas. Analysis of cue-lure trapping data from 1997 to 2012 in relation to environmental variables shows that the distribution of B. frauenfeldi in Queensland correlates to locations with a minimum temperature for the coldest month >13.2°C, annual temperature range <19.3°C, mean temperature of the driest quarter >20.2°C, precipitation of the wettest month >268 mm, precipitation of the wettest quarter >697 mm, temperature seasonality <30.9°C (i.e. lower temperature variability) and areas with higher human population per square kilometre. Annual temperature range was the most important variable in predicting this species' distribution. Predictive distribution maps based on an uncorrelated subset of these variables reasonably reflected the current distribution of this species in northern Australia and predicted other areas in the world potentially at risk from invasion by this species. This analysis shows that the distribution of B. frauenfeldi in Australia is correlated to certain environmental variables that have most likely limited this species' spread southward in Queensland. This is of importance to Australian horticulture in demonstrating that B. frauenfeldi is unlikely to establish in horticultural production areas further south than Townsville.

Halosulfuron-methyl: A selective herbicide option for the control of the invasive 'Cyperus aromaticus' (Ridley) Mattf. and Kukenth (Navua sedge)

There are currently limited options for the control of the invasive tropical perennial sedge 'Cyperus aromaticus' (Ridley) Mattf. and Kukenth (Navua sedge). The potential for halosulfuron-methyl as a selective herbicide for Navua sedge control in tropical pastures was investigated by undertaking successive field and shade house experiments in North Queensland, Australia. Halosulfuron-methyl and adjuvant rates, and combinations with other herbicides, were examined to identify a herbicide regime that most effectively reduced Navua sedge. Our research indicated that combining halosulfuron- methyl with other herbicides did not improve efficacy for Navua sedge control. We also identified that low rates of halosulfuron-methyl (25 g ha-1 a.i.) were just as effective as higher rates (73 g ha-1 a.i.) at controlling the sedge, and that this control relied on the addition of the adjuvant Bonza at the recommended concentration (1% of the spray volume). Pot trials in the controlled environment of the shade house achieved total mortality under these regimes. Field trials demonstrated more variable results with reductions in Navua sedge ranging between 40-95% at 8-10 weeks after treatment. After this period (16-24 weeks after treatment), regrowth of sedge, either from newly germinated seed, or of small plants protected from initial treatment, indicated sedge populations can rapidly increase to levels similar to pre-application, depending on the location and climatic conditions. Such variable results highlight the need for concerted monitoring of pastures to identify optimal treatment times. Ideally, initial treatment should be done when the sedge is healthy and actively growing, with follow up-treatments applied when new seed heads are produced from regrowth.

Impact of the invasive rust Puccinia psidii (myrtle rust) on native Myrtaceae in natural ecosystems in Australia

The invasive rust Puccinia psidii (myrtle rust) was detected in Australia in 2010 and is now established along the east coast from southern New South Wales to far north Queensland. Prior to reaching Australia, severe damage from P. psidii was mainly restricted to exotic eucalypt plantations in South America, guava plantations in Brazil, allspice plantations in Jamaica, and exotic Myrtaceous tree species in the USA; the only previous record of widespread damage in native environments is of endangered Eugenia koolauensis in Hawai’i. Using two rainforest tree species as indicators of the impact of P. psidii, we report for the first time severe damage to endemic Myrtaceae in native forests in Australia, after only 4 years’ exposure to P. psidii. A 3-year disease exclusion trial in a natural stand of Rhodamnia rubescens unequivocally showed that repeated, severe infection leads to gradual crown loss and ultimately tree mortality; trees were killed in less than 4 years. Significant (p < 0.001) correlations were found between both incidence (r = 0.36) and severity (r = 0.38) of P. psidii and subsequent crown loss (crown transparency). This provided supporting evidence to conclude a causal association between P. psidii and crown loss and tree mortality in our field assessments of R. rubescens and Rhodomyrtus psidioides across their native range. Assessments revealed high levels of damage by P. psidii to immature leaves, shoots and tree crowns—averaging 76 % (R. rubescens) and 95 % (R. psidioides) crown transparency—as well as tree mortality. For R. psidioides, we saw exceptionally high levels of tree mortality, with over half the trees surveyed dead and 40 % of stands with greater than 50 % tree mortality, including two stands where all trees were dead. Tree mortality was less prevalent for R. rubescens, with only 12 % of trees surveyed dead and two sites with greater than 50 % mortality. Any alternative causal agents for this tree mortality have been discounted. The ecological implications of this are unclear, but our work clearly illustrates the potential for P. psidii to negatively affect Australia’s biodiversity.