Potential global distribution of Aleurocanthus woglumi considering climate change and irrigation

Historical climate data The meteorological files for projecting the potential distribution of A. woglumi contain site cli- mate data for daily minimum temperature (Tmin), daily maximum temperature (Tmax), monthly precipitation (Rainfall), relative humidity at 9 am (RH 0900) and relative humidity at 3 pm (RH 1500). Historical meteorological data (period 1981–2010) were obtained from Cli- Mond [50]. The CliMond climate dataset consists of gridded historical climate data and some future climate scenario data at 10’ or 30’ spatial resolution. In this study, 30’ climate data were used. Climate change scenario There are substantial, irreducible uncertainties in estimating future climates. The most sensi- tive uncertainty appears to be the rate of emission of greenhouse gases through time [51]. The realised emission path will depend on factors such as the pattern of economic activity, geopo- litical factors such as the willingness to decarbonise economies and technological and eco- nomic developments affecting the cost of renewable energy. History has underlined how these factors cannot be reliably predicted; hence, a future climate scenario that is internally consis- tent but relies on these uncertain factors cannot, in turn, be considered a reliable prediction of future climatic conditions. Put simply, we cannot reliably pick which emission scenario is the “correct” one, so we should reframe the problem from trying to predict the future conditions to trying to identify potential emerging problems by stress-testing the biophysical model with

a climate model based on an extreme scenario. In this study, in addition to modelling the potential distribution of A. woglumi under his- toric climate conditions and irrigation, we choose to use irrigation incorporated in a near- term (averaging period 2040–2059) business as usual emission scenario, Representative Con- centration Pathway 8.5 (RCP8.5), realised with a popular global climate model (ACCESS1-0). The RCP8.5 pathway arises from little effort to reduce emissions and represents a failure to curb warming by 2100. It is characterized by increasing greenhouse gas emissions over time and represents scenarios in the literature leading to high greenhouse gas concentration levels [52]. As suggested above, we do not use the model to try to predict the future potential distri- bution of A. woglumi, but rather to understand the direction of range changes and the relative sensitivity of the potential distribution to expected future climatic changes. In essence, we are undertaking a sensitivity analysis of our CLIMEX model results to expected climate changes. This allows us to comment on what invasion threats are likely to emerge or wane in the near to medium term future. Hence, the model results should not be considered a prediction of what will come to pass. Rather, they are intended to be used as a basis for stress-testing biosecurity policies. Location records The historical distribution of A. woglumi used to build the CLIMEX model was assembled mostly from published studies and databases such as EPPO-Q-bank, and the American

Museum of Natural History (S1 Table). Province/region-level occurrence records of A. woglumi were retrieved also from the literature and represent administrative regions where the pest has been reported, limiting their value for precise model calibration. Fuzzy logic is required to compare these records to the model results. The provincial/regional records are in S1B Table, while the point locations (cities, town, municipalities) are presented in S1A Table. The coordinates listed in S1B Table represent the centroid of each polygon. The native range data was mostly used for model-fitting, with data elsewhere used for model validation. In total, the dataset consisted of 179 locations. PLOS ONE Potential global distribution of Aleurocanthus woglumi PLOS ONE | https://doi.org/10.1371/journal.pone.0261626 December 20, 2021 4 / 18