Mapping impacts of alien species on biodiversity in the face of climate change

doi:10.21203/rs.3.rs-4437291/v1

Download PDF

Research Article

Mapping impacts of alien species on biodiversity in the face of climate change

https://doi.org/10.21203/rs.3.rs-4437291/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Risk maps are a useful tool to prioritise sites for management and allocate resources where they are most needed as they can show us where impacts of biological invasions are most likely to happen or expected to be largest. Given the pace of global changes, we need to understand not only the risks under current conditions, but future risks taking into account aspects such as climate change. In this study, we use Australian acacias alien to South Africa as a case study to model their potential distribution under climate change scenarios, and map their potential impacts and the uncertainty related to variability across socio-economic pathways and climatic models, to help us set better priorities for the future. We found that although climatic suitability and therefore the potential high risk area is predicted to decrease under climate change, the highly vulnerable fynbos biome remains an area with high projected impacts. We found that potential impacts would decrease under climate change within protected areas of the Western Cape. However, in this province, national parks and other areas with high levels of protection would experience greater impacts than private protected areas of lower protection status. This can help us prioritise management actions and aid the development of suitable plans to protect biodiversity not only under current, but also future climate conditions.

EICAT

biological invasions

Australian Acacia

SDM

protected area

risk map

Biological invasions and climate change are two of the main drivers of global biodiversity loss (Master and Norgrove 2010; Johnson et al. 2017; IPBES 2023). Impacts of biological invasions are manifold, including changes to ecosystem functions such as fire regimes and hydrology and changes to biodiversity including extirpations and extinctions of native species (Vilà et al. 2011; Smith 2020; Vilà et al. 2024; IPBES 2023). Similarly, climate change can have severe impacts on native species, for example increasing their risk of extinctions (e.g., Thomas et al. 2004; Manes et al. 2021) and causing range changes (e.g., Lenoir et al. 2008; Doak et al. 2010). While these drivers can act in isolation, climate change can also affect the invasion of alien species as climatic conditions play an important role in the growth, survival and distribution of species (Walther et al. 2009; Master and Norgrove 2010). Climate change can not only lead to changes in native species distributions, but also those of alien species (Bellard et al. 2013).

Member states of the Convention on Biological Diversity (CBD) are mandated to improve the protection of natural resources under the Kunming-Montreal Global Biodiversity Framework (GBF). Target 1 asks to minimize the loss of areas with high biodiversity. To achieve this, we need to understand the pressures on these areas from drivers such as climate change and biological invasions, and map them to prioritise control. Risk maps have been used for various purposes in invasion science. For example, they can be useful to improve detection of new incursions or individuals of species with limited populations (Kaplan et al. 2014). They have also been used to map the potential impacts to assess sites experiencing the highest current or potential impact, translating to sites at risk of invasion impacts (Nentwig et al. 2010). However, as species’ distributions are expected to change with climate change, considering only current climatic conditions in such risk maps does not prepare us well for future risk scenarios. Therefore, to improve our ability to manage biological invasions, as stipulated under Target 6 of the GBF, we need to understand how biological invasions and their impacts could change under climate change. Simultaneously, this addresses Target 8, which aims to minimize the effect of climate change on biodiversity. Still, climate change is not routinely included in risk assessments for alien taxa (but see Marchioro and Krechemer 2021 for some insect pests).

Correlative species distribution models (hereafter SDM) are popular in various fields of ecology and are increasingly used in invasion science (Guisan et al. 2014). They are statistical methods which correlate known occurrences of species with environmental variables and predict a species' potential distribution in other regions over space and time (Guisan and Zimmermann 2000). The advantages of SDMs are manifold as they are easy to implement and can cover large geographic areas. Therefore, they are widely used to produce maps of potential invasion by delineating areas based on climatic suitability for a species (e.g. Bradley et al. 2010; Jiménez-Valverde et al. 2011; Tingley et al. 2017). Despite known limitations (see for example Hui 2023), they have shown potential for accurate prediction of alien species spread (Barbet-Massin et al. 2018) and response of biological invasions to climate change (Sheppard et al. 2014) if properly calibrated.

In this study, we use Australian Acacia species alien to South Africa as a case study to assess i) the current distribution of the species and their potential impacts, ii) the potential distribution under current climate, and iii) the projected future distribution under climate change scenarios. By overlaying the outputs of the models with the species potential impacts, we produce risk maps for current and projected future climatic conditions and assess potential impacts for protected areas of the Western Cape Province in South Africa.

We detail below how we selected Australian Acacia species and fitted SDM models per species that were then used to project the habitat suitability across South Africa per time horizon, accounting for various climatic scenarios. Then, we explain how we aggregated an impact score from the species-wise SDM projections and the documented impacts per species.

Species selection

We selected, as a case study, Australian Acacia species (Acacia Mill. sensu stricto – synonym: Acacia subg. Phyllodineae (DC.) Seringe) which are alien in South Africa. This includes 33 species (Magona et al. 2018; Jansen and Kumschick 2022; Kumschick and Jansen 2023; Supplementary Material Appendix S4.1), of which three species (Acacia crassiuscula, A. acuminata and A. koa) had to be excluded from all analysis because they had less than 10 presence records after the filtering steps detailed below. Acacias are among the most highly impacting alien taxa in South Africa with a diverse range of impacts on native biodiversity (le Maître et al. 2011; Kumschick & Jansen 2023).

Occurrence data

We gathered distribution data from the native and introduced ranges of the 30 Acacia species, and used it along with bioclimatic variables to build individual ensemble models which were then projected based on current and future climate in South Africa. For each species, worldwide occurrence data were obtained from the Global Biodiversity Information Facility (GBIF, http://www.gbif.org/) using the rgbif package (Chamberlain et al. 2022). For this extraction, we kept all records which had complete coordinates, which were tagged as “presence”, and which were either human or machine observations or coming from the literature. No time limit was specified. Then we filtered occurrences to reduce the uncertainty arising from questionable records and only kept records which came from a list of 65 trusted GBIF datasets. This list was previously developed for the global distribution of Acacia species (Botella et al. 2023) and is composed of datasets which include at least one documented step of taxonomic verification in the recording of the occurrences. The documented native and introduced countries of each species were also obtained from that study. Besides, we extracted the global land cover raster (approx. 300 m resolution) from the European Space Agency GlobCover product. We excluded records located outside the native and introduced countries of the species (obtained from Botella et al. 2023) or whose 4 km x 4 km cell was located in urban areas based on the GlobCover 2009 global land cover raster (ESA and UCLouvain GlobCover 2009 Project, last accessed 29/01/2024). The latter step was implemented to exclude planted trees. More precisely, we excluded cells classified as “Artificial surfaces and associated areas (Urban areas > 50%)” by GlobCover. We then resampled our remaining records to keep only up to one record per 4 km × 4 km cell for each species to reduce the oversampling of some areas (Boria et al. 2014). We obtained a total of 37,940 filtered records (Figure S4.1 & S4.2). The number of filtered records were imbalanced across the 30 considered species, with 27 for the least sampled (A. adunca) and 3,738 for the most sampled (A. dealbata).

Climate variables

To estimate the climatic suitability for each Acacia species from their occurrence data, we extracted four bioclimatic variables from the Climatologies at High Resolution for Earth’s Land Surface Areas database (CHELSA, Karger et al. 2017). We chose the mean annual air temperature (bio1), mean diurnal air temperature range (bio2), annual precipitation (bio12) and precipitation seasonality (bio15) as environmental variables as they are likely to affect plant productivity and survival across the landscape (Mod et al., 2016), and the correlation between each of these variables was moderate (Pearson’s |r| < 0.55). These data are available globally at a 4 km × 4 km resolution for the period 1981–2010.

Then, we predicted the suitable climatic range of each species per time period in South Africa. The “current” South African climate was derived from the 1981–2010 period and the “future” climate was decomposed in two horizons: mid-century (2041–2070) and end of the century (2071–2100). For both time horizons, we extracted the four climatic variables based five Global Climate Models from CHELSA (“gfdl-esm4”, “ipsl-cm6a”, “mpi-esm1”, “mri-esm2”, “ukesm1”) for three shared socio-economic pathway scenarios (“ssp1.26”, “ssp3.70” and “ssp5.85”) for the South African region: ssp1.26: scenario with low greenhouse gas (GHG) emission; ssp3.70: scenario with high GHG emission; ssp5.85: scenario with very high GHG emission (Fig. 1 and Supplementary Material, Appendix S1).

Species distribution models

Pseudo-absences selection

For each species we drew ten datasets of pseudo-absences composed of the same number of pseudo-absences as presences. For species with less than 100 presences, we increased the number of pseudo-absences to 100 in each dataset, to reduce the variance (over-fitting) of each model. These pseudo-absences were drawn in the countries being part of the native and introduced ranges of the species assuming that the species likely had an opportunity to establish at pseudo-absence locations. Following recommendations of Barbet-Massin et al. (2012), pseudo-absences were selected based on the surface range envelope method (Busby 1991), by randomly selecting pseudo-absences outside the usual climatic range of the species occurrences (Thuiller et al. 2009). Given that pseudo-absences were drawn in the native and introduced ranges but outside of the climatic envelope where the species occurrences were recorded (PA.sre.quant = 0), we can assume false absences to be rare. Therefore, the final model predictions will approximate the probability of presence conditional to the species presence in the area, i.e. an establishment probability.

Model implementation

For each species we built a species distribution model using functions and workflow of the biomod2 package (Thuiller et al. 2009). We applied a random forest classifier (Liaw and Wiener 2002) on the presence/pseudo-absence datasets and the four climatic variables to decipher which climatic conditions are suitable for each species. Each forest was composed of 500 trees, where each branch tests two randomly picked variables (mtry = 2), and we impose a minimum of five sampled presence/pseudo-absences per terminal node. We implemented a step of cross-validation for which the presence/pseudo-absence dataset was divided into four blocks along the longitude axis with presences equally balanced between each block (Wenger and Olden 2012). Then, three blocks were used for model training and one for testing. This approach allows for spatial and environmental independence between training and testing sets. Kappa and TSS metrics were computed each time on the testing dataset. As we had ten pseudo-absence datasets and four blocks, we computed 40 random forest models for each species.

For each species we kept only the random forest models with a Kappa and a TSS over 0.7, indicating a robust predictive capacity, and which we refer to as sub-models. We then derived for each species an ensemble of its sub-models that was a “committee averaging” of the binarized responses given by the sub-models (Thuiller et al. 2009). We used a probability threshold of 0.5 to binarize each submodel response, corresponding to a majority vote among trees of the corresponding random forest. Committee averaged values are the proportion of sub-models predicting the species as present. Here, we considered that committee averaged values can be regarded as a proxy for the probability of a species to establish. Thus, we obtained one projection of establishment probability for South Africa under the current climate and 15 scenarios (five global climate models x three socio-economic pathways) for each time horizon (2041–2070 and 2071–2100) as represented in Fig. 1. So, for each time horizon and each spatial cell, we averaged the 15 establishment probabilities of the 15 scenarios to obtain the final establishment probability of the species in that cell. This final establishment probability gives an equal weight to the 15 plausible scenarios. Finally, for each time horizon, we produced a potential Acacia species richness map summing the 30 species establishment probability maps, and an associated uncertainty map. The uncertainty map shows standard deviation of the potential richness estimate based on the variability standard deviation of the predicted establishment per species and its variability across the 15 scenarios.

Risk maps

As we aimed to map the accumulated risk of Acacia species potentially causing impacts in different regions in South Africa, we firstly derived a potential impact score per species based on its documented impacts. Data on the impacts of the Acacia species were taken from Jansen and Kumschick (2022). They assigned categories of Minimal Concern to Massive to each species, based on the impacts reported in the literature and following the International Union for the Conservation of Nature (IUCN) standard for impact classification, the Environmental Impact Classification for Alien Taxa, EICAT (Blackburn et al. 2014; Hawkins et al. 2015; IUCN 2020a, b; Volery et al. 2020). We used the maximum impact category per species as a precautionary measure, which is the recommended approach used by the IUCN EICAT Standard (IUCN 2020a; Kumschick et al. 2024), noting that other calculation methods are possible depending on the aim of the study (see also Kumschick et al. 2024, Boulesnane-Genguant et al. in prep). We transformed the impact categories into numerical values as follows: Minimal Concern (MC) = 0, Minor (MN) = 1, Moderate (MO) = 2, Major (MR) = 3, Massive (MV) = 4. 19 species classified as Data Deficient (DD), i.e. no data on impact was found in a comprehensive literature search, and they were assigned a score of 0 (Supplementary Material Appendix S4.1).

As a prediction of local impact risk per species, we multiplied the predicted establishment probability derived from the SDMs in each grid cell by the species impact score. Finally, we summed up the local impact risks across species in each grid cell to obtain the impact risk map (see also Nentwig et al. 2010). Hence, each species occurring adds to the risk of the potential impacts actually manifesting at that site, again using a precautionary approach. Our maps therefore do not represent sites where impacts are definitely going to occur, as the potential establishment of each species alone does not necessarily happen and, even if it does, it does not necessarily result in the worst impact documented for that species. Other factors such as the recipient community and the abundance of the alien species play important roles in the manifestation of impacts. Note also that even if data deficient species are present, they do not affect the impact risk value, as for species whose maximum impact is MC. This might underestimate the impact of some species which actually cause higher impacts but which have not been studied.

Changes in impact risk in protected areas

To assess the specific change in risk across time on protected areas, we assessed the potential changes of impact risk for protected areas in the Western Cape Province of South Africa. For each protected area, we calculated the mean impact risk score and compared it under current and future (2071–2100) projected climate. To see whether differences could be found based on the protected area status, we also calculated the mean impact risk score per protected area type. In the Western Cape, protected areas are classified as follows (with increasing level of protection): private nature reserves, provincial nature reserves and national parks. We obtained protected areas data from the South African National Biodiversity Institute (www.bgis.sanbi.org).

Firstly, we present a method to map the potential cumulative impacts of alien species under climate change projections (Fig. 1, Supplementary Material, Appendix S1). It combines the recorded impacts alien species have had with the projected species richness. Applying this method to Acacia species in South Africa, we find that across all species currently present in the country, the suitable area declines over time and only the south coast in the Eastern Cape province and parts of the Western Cape province still remain suitable for a considerable number of species (Fig. 2). Large parts of the north eastern (NE) part of the country are predicted to become unsuitable for most wattles in future. Indeed, while the current projected impact score lies between 15 and 25 in a relatively large part of the NE area, it is inferior to 5 in nearly all of the area for the end of the century (Fig. 2). Yet, this decrease should be interpreted with caution, as the standard deviation of the end-of-century impact score, due to the variability across socio-economic pathways and climate models, is around 2 in the NE area that currently shows a high risk (Figure S3). Comparing the current impact risk map to a risk map based on actual records for Acacia species (Fig. 3), the general pattern looks similar to potential impacts under current climate, with a less continuous distribution along the east coast, and a marked region of high risk in the Gauteng province. The latter is probably due to the fact that in the map showing impacts based on recorded occurrences, urban areas were not excluded, and much of the Gauteng province would fall under this category (Fig. 3). Furthermore, a positive sampling bias in this highly populated area might be responsible for higher detection rates of the species present and explain the relatively high impact risk when based on actual occurrences.

[insert Fig. 2 here]

[insert Fig. 3 here]

Although species richness of Acacias is not projected to be as high in the southwestern part of the country as along the south coast, the projected risks are similar in both areas (Fig. 2). That shows that the areas around Cape Town and Hermanus, although not suitable for as many species, might remain suitable for the high impacting species. The impact score in this area remains most likely high compared to other areas even when considering its standard deviation (Figure S3).

We found that potential impacts would decrease under climate change within protected areas of the Western Cape. Our results showed a lower impact risk with climate change for protected areas along the coast line, almost no change for northern protected areas and a higher impact risk in few isolated protected areas (Fig. 4). However, the impact risk differed according to the protection status. The impact risk remains high in areas of high protection status (such as national parks) and low for private protected areas (with lower protection status) (Fig. 5).

This study presents, to the best of our knowledge, a first attempt at combining impacts of alien taxa with their projected distribution under climate change to produce risk maps for future climatic scenarios. Our results indicate that the projected suitable areas for alien Acacia species in South Africa are reduced under climate change (Fig. 2). This is similar to a study looking at A. mangium and A. auriculiformis in Brazil, where they found a shift in potential distribution and a reduction in suitable area (Heringer et al. 2019). More generally, this is in line with Bellard et al. (2018) who showed in a review of 71 papers covering 423 alien species that climate change is more frequently projected to contribute to a decrease in range size than an increase. However, to assume that the problem of Acacia invasions is going to sort itself out over time might be a bit too optimistic. For once, the sites at highest risk are also the regions where the highly vulnerable fynbos biome is located. This biome is already under pressure from the effects of climate change, including increased temperature and drought, and therefore more vulnerable to other pressures (Slingsby et al. 2017; Skowno et al. 2021).

Furthermore, Acacia species which are not currently alien in the country (not yet introduced) were not modelled in this study. That means that other species not in our dataset might behave differently from what we present here. Therefore, this does not exclude the possibility of other Acacia species, and other alien species in general, becoming more problematic in future (see also Sheppard et al. 2016). Furthermore, Acacias and trees in general are long lived species which can result in a long lag between reduced climate suitability and decline in populations or observed reduction in impacts (Kowarik 1995; Robeck et al. 2024). Besides, perennial and predominantly allogame plants like Acacia species often exhibit lag phases of several decades before the beginning of an invasion (Robeck et al. 2024). Hence, several Acacia species could still be in a lag phase preceding an invasion in South Africa, and this invasion debt was not accounted for in the trends presented here (e.g., Rouget et al. 2016).

As with any model, there are uncertainties linked to the projections. SDMs are based on several strong assumptions (Guisan et al. 2017; Hui 2023). First, occurrence records should reflect the true performance of the species. However, sampling biases are present in most if not all record databases (Beck et al. 2014). Moreover, low sample size can highly impact the performance of SDMs and data available may not be sufficient to fully inform the models (Wisz et al. 2008; Stockwell et al. 2002). To address these shortcomings, we designed a conservative methodology to keep only the most trustworthy records and limit oversampling in some areas. We also kept records from both the native and the alien range (Broennimann and Guisan 2008).

Second, the species’ performance should respond directly to the variation of the selected predictors. In this study we chose four climate variables for their known link with plant species survival and development (Mod et al. 2016) and their availability in both fine spatial and temporal scales. Moreover, Sheppard et al. (2014), showed that predictions of similar SDMs of the response of three invasive plant species to climate change was highly correlated with field experiments. However, SDMs are correlative and may yield incorrect estimates of habitat suitability if climatic variables are correlated to other unknown variables in the training area (Guisan et al. 2017; Jarnevich et al. 2015). Future projections from our fitted models of climatic suitability could be biased by potential confounding factors of climate, as we do not take into account other factors which determine the success of alien species in a new region, such as dispersal capabilities and biotic interactions. Our model fitting procedure implicitly assumed that a species had the opportunity to colonise a large part of any country where it was recorded. Regarding interactions, Australian Acacia species often have competitive advantages over native plant species and tend to become dominant among plant communities, especially after disturbance (Morris et al. 2011), suggesting that the spatial extent of the realised niche would not be strongly restricted by competitive interactions compared to the potential niche. Yet, other interactions, such as mutualistic and trophic interactions with soil fungi (Birnbaum et al. 2018) may contribute to constrain the actual range of Acacias and act as confounding factors of climate in fitted SDMs, inducing bias in future projections. Furthermore, factors such as topography could be taken into account to improve the models (Bradley and Mustard 2006). Thus, our model predictions could be improved by considering non-climatic drivers (such as soil composition) if these variables were available at a fine resolution.

Third, the species’ distribution, represented by recorded occurrences, should be stable and fill any available niche in the study environment. Several studies have shown violations of the niche conservatism hypothesis during invasion with niche shifts between the native niche and the introduced niche (e.g. Parravicini et al. 2015; Broennimann et al. 2007; Guisan et al. 2014). Moreover, predictions based on extrapolations on a new territory and with future climatic conditions may not be robust because the data used for model parameterization cannot represent all conditions in the extrapolated region (Elith and Leathwick 2009; Barbet-Massin et al. 2010; Sinclair et al. 2010). Thus, it must be kept in mind that our suitability and risk maps may be underestimated and should not be taken as a prediction of true future species richness and impacts.

Fourth, some future climatic conditions may not have an analog amongst the historical climates of the study area (Williams & Jackson 2007). Yet, SDMs cannot predict exactly how species will respond to conditions that were not used for calibration (Pouteau et al. 2021). This could partly explain why most alien species (Bellard et al. 2018) including Acacias alien to South Africa, are projected to experience a decrease in the size of their potential range according to our current knowledge. Future work should consider the identification of novel climates so as to avoid putting too much confidence in climates with no current analogues.

Caution is also advised when interpreting the risk maps including the sum of potential impacts of Acacias. Firstly, there are many ways to aggregate impacts, both, within species (calculating one impact value taking into account all impact records for the species) and across species (calculating an impact score for a site where several alien species are present) (Boulesnane-Genguant et al. in prep). Some of the most prominent methods to get one impact value per species have been to sum scores (e.g., Nentwig et al. 2016), to calculate a mean value (e.g., Rumlerova et al. 2016), and to take a maximum value (e.g., Blackburn et al. 2014) (see also Kumschick et al. 2024). They each come with underlying assumptions which can affect the results. The maximum score per species, as used here, was chosen for range shifting species in the US to anticipate if any high impacting invaders are likely to arrive under climate change (Rockwell-Postel et al. 2020). However, the applications of scores aggregated across species are scarce (for an example, see Nentwig et al. 2010).

Furthermore, climate change can not only affect the potential distribution of species, but also modify their invasion behaviour and impacts (e.g., Le Maitre et al. 2020). Changes to fire regimes could have large impacts on alien species and native ecosystems, especially in the fynbos biome of South Africa where fires have been increasing in frequency and intensity due to alien invasions and climate change (Le Maitre et al. 2020; Slingsby et al. 2017). Furthermore, CO₂ concentration could lead to woody plant densification, which is already shown for native woody plants in some southern African habitats (Skowno et al. 2017). In the fynbos biome, increased CO₂ could also favour the alien wattles as they are nitrogen fixers as opposed to the native flora which are adapted to low nitrogen conditions (Richardson et al. 2014). In other examples, the synchronisation of the flowering period of native and alien plants may favour the latter, through increased interactions with pollinators to the detriment of the pollination of native species. This is the case in New Zealand with Calluna vulgaris whose greater phenological plasticity compared to the native species Dracophyllum subulatum means that it can reproduce more easily in areas with a high floral density (Giejsztowt et al. 2020). Morphological responses to climate change can also increase the competitiveness of alien species. For example, milder winter temperatures in China facilitate the survival of the water hyacinth Pontederia crassipes, and also allow it to develop a greater biomass, forming denser foliage that excludes submerged native plants the following season (You et al. 2013). In terrestrial environments, rising temperatures can also encourage the development of alien plants, reducing the availability of water for native plants, which are then at a disadvantage when it comes to coping with dry spells. This is the case of Tamarix spp. introduced in the United States, which develops greater capacity to capture and use water resources than native riparian species under the effect of drought (Hellmann et al. 2008).

Given all these potential interactions between climate change and impacts of alien species, what we are showing here is not an accurate representation of sites with future impacts of alien Acacia species in South Africa. However, risk maps like these of sites where impacting species could occur under climate change projections can be valuable in helping us prioritise sites for future protection from invasion impacts. Despite the potential shortcomings of the models presented here, our study contributes to our understanding of the impacts of climate change on the risk of alien Acacia invasions, including their impacts. Furthermore, it can aid the prioritisation of clearing actions for alien Acacia species in South Africa and feed into strategies for protected area management and expansion (e.g., Department of Environmental Affairs 2016). Protected areas are set up to safeguard biodiversity and ecosystem services for the future. We show that for the Western Cape province, most protected areas should experience less impact risk in future. However, the areas with increased risk are of lowest protection, which makes them more vulnerable to negative changes and biodiversity loss. Elevating the protection of these sites and controlling harmful Acacias can help us reduce the loss of high biodiversity areas, as stipulated in the GBF Target 1. Understanding where some of the most damaging invaders might occur under climate change is important so our management is tackling not only current, but also potential future problems.

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Funding

SK acknowledges the support of the Centre for Invasion Biology (CIB) at Stellenbosch University and the South African Department of Forestry, Fisheries and the Environment (DFFE). SK also acknowledges funding from the B3 project. B3 (Biodiversity Building Blocks for Policy) receives funding from the European Union’s Horizon Europe Research and Innovation Programme (through grant no .584 101059592). LJ was supported by the OUSNEXT project funded by the Structure fédérative de recherche BioST of the Université de La Réunion (PI: RP). OBG was supported by the Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD), Fonds Vert and by the Université de La Réunion. This work was co-funded by the European Union: Agricultural Fund for Rural Development (EAFRD), by the Conseil Départemental de La Réunion, and by the Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD).

Author contributions

SK and MR conceptualised and designed the study. CB provided distribution data. CB, LJ, OBJ and MR performed the data analyses. The first draft of the manuscript was written by SK and all authors contributed to the writing of the submitted draft and read and approved the final manuscript.

Acknowledgements

We thank Cally Jansen for data collection on impacts of alien Acacia species, and the participants of the workshop on “Managing invasions in protected areas” held in Stellenbosch from 3–5 April 2024 for valuable comments received on an earlier draft of the manuscript. SK acknowledges the support of the Centre for Invasion Biology (CIB) at Stellenbosch University and the South African Department of Forestry, Fisheries and the Environment (DFFE), noting that this manuscript does not necessarily reflect the opinions of the DFFE or its employees. SK also acknowledges funding from the B3 project. B3 (Biodiversity Building Blocks for Policy) receives funding from the European Union’s Horizon Europe Research and Innovation Programme (through grant no .584 101059592). The views and opinions expressed in the present article are, however, those of the authors only and do not necessarily reflect those of the European Union or the European Commission. Neither the EU nor the EC can be held responsible for them. LJ was supported by the OUSNEXT project funded by the Structure fédérative de recherche BioST of the Université de La Réunion (PI: RP). OBG was supported by the Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD), Fonds Vert and by the Université de La Réunion. This work was co-funded by the European Union: Agricultural Fund for Rural Development (EAFRD), by the Conseil Départemental de La Réunion, and by the Centre de Coopération Internationale en Recherche Agronomique pour le Développement (CIRAD).

Barbet-Massin M, Rome Q, Villemant C, Courchamp F (2018) Can species distribution models really predict the expansion of invasive species? PLoS ONE 13(3):e0193085. https://doi.org/10.1371/journal.pone.0193085
Barbet-Massin M, Thuiller W, Jiguet F (2010) How much do we overestimate future local extinction rates when restricting the range of occurrence data in climate suitability models? Ecography 33:878–886
Barbet‐Massin M, Jiguet F, Albert CH, Thuiller W (2012) Selecting pseudo‐absences for species distribution models: How, where and how many?. Methods in Ecology and Evolution 3(2):327-338
Beck J, Böller M, Erhardt A, Schwanghart W (2014) Spatial bias in the GBIF database and its effect on modeling species’ geographic distributions. Ecological Informatics 19:10–15. https://doi.org/10.1016/j.ecoinf.2013.11.002
Bellard C, Thuiller W, Leroy B, Genovesi P, Bakkenes M, Courchamp F (2013) Will climate change promote future invasions?. Global Change Biology 19(12):3740-3748
Bellard C, Jeschke JM, Leroy B, Mace GM (2018) Insights from modeling studies on how climate change affects invasive alien species geography. Ecology and Evolution 8:5688–5700. https://doi.org/10.1002/ece3.4098
Birnbaum C, Morald TK, Tibbett M, Bennett RG, Standish RJ (2018) Effect of plant root symbionts on performance of native woody species in competition with an invasive grass in multispecies microcosms. Ecology and Evolution 8(17):8652-8664.
Blackburn TM, Essl F, Evans T, Hulme PE, Jeschke JM, Kühn I, Kumschick S, Mrugała A, Marková Z, Nentwig W, Pergl J, Pyšek P, Rabitsch W, Ricciardi A, Richardson DM, Sendek A, Vilà M, Wilson JRU, Winter M, Genovesi P, Bacher S (2014) A unified classification of alien species based on the magnitude of their environmental impacts. PLoS Biology 12(5): e1001850. DOI: 10.1371/journal.pbio.1001850
Boria RA, Olson LE, Goodman SM, Anderson RP (2014) Spatial filtering to reduce sampling bias can improve the performance of ecological niche models. Ecological Modelling 275:73-77
Botella C, Marchante H, Celesti-Grapow L, Brundu G, Geerts S, Ramirez-Albores JE, González-Moreno P, Ritter M, Richardson DM (2023) The global distribution of Acacia. Wattles: Australian Acacia Species Around the World, 131. https://doi.org/10.1079/9781800622197.0009
Bradley BA, Wilcove DS, Oppenheimer M (2010) Climate change increases risk of plant invasion in the Eastern United States. Biological Invasions 12:1855–1872
Bradley BA, Mustard JF (2006) Characterizing the landscape dynamics of an invasive plant and risk of invasion using remote sensing. Ecological Applications 16(3):1132-1147
Broennimann O, Guisan A (2008) Predicting current and future biological invasions: both native and invaded ranges matter. Biol Lett 4:585–589
Broennimann O, Treier UA, Müller-Schärer H, Thuiller W, Peterson AT, Guisan A (2007) Evidence of climatic niche shift during biological invasion. Ecol Lett 10:701–709
Busby JR (1991) BIOCLIM-a bioclimate analysis and prediction system. Plant Protection Quarterly 6(1):8–9
Chamberlain S, Barve V, Mcglinn D, Oldoni D, Desmet P, Geffert L, Ram K (2022) rgbif: Interface to the Global Biodiversity Information Facility API. R package version 3.7.2. https://CRAN.R-project.org/package=rgbif
Department of Environmental Affairs (2016) National protected areas expansion strategy for South Africa 2016. Department of Environmental Affairs, Pretoria, South Africa.
Doak D, Morris W (2010) Demographic compensation and tipping points in climate-induced range shifts. Nature 467:959–962. doi:10.1038/nature09439
Elith J, Leathwick J (2009) Species distribution models: Ecological explanation and prediction across space and time. Evolution and Systematics 40:677–697
Giejsztowt J, Classen AT, Deslippe JR (2020) Climate change and invasion may synergistically affect native plant reproduction. Ecology 101(1):e02913
Guisan A, Petitpierre B, Broennimann, et al (2014) Unifying niche shift studies: insights from biological invasions. Trends in Ecology and Evolution 29:260–269. https://doi.org/10.1016/j.tree.2014.02.009
Guisan A, Thuiller W, Zimmermann NE (2017) Habitat suitability and distribution models: with applications in R. Cambridge University Press, Cambridge
Guisan A, Zimmermann NE (2000) Predictive habitat distribution models in ecology. Ecological Modelling 135(2):147–186. https://doi.org/10.1016/S0304-3800(00)00354-9
Hawkins CL, Bacher S, Essl F, Hulme PE, Jeschke JM, Kühn I, Kumschick S, Nentwig W, Pergl J, Pyšek P, Rabitsch W, Richardson DM, Vilà M, Wilson JRU, Genovesi P, Blackburn TM (2015) Framework and guidelines for implementing the proposed IUCN Environmental Impact Classification for Alien Taxa (EICAT). Diversity and Distributions 21(11):1360-1363
Hellmann JJ, Byers JE, Bierwagen BG, Dukes JS (2008) Five potential consequences of climate change for invasive species. Conservation Biology 22(3):534-543
Heringer G, Bueno ML, Meira-Neto JA, Matos FA, Neri AV (2019) Can Acacia mangium and Acacia auriculiformis hinder restoration efforts in the Brazilian Atlantic Forest under current and future climate conditions?. Biological Invasions 21(9):2949-2962
Hui C (2023) The dos and don’ts for predicting invasion dynamics with species distribution models. Biological Invasions 25:947–953. https://doi.org/10.1007/s10530-022-02976-3
IPBES (2023) Summary for policymakers of the thematic assessment report on invasive alien species and their control of the Intergovernmental science-policy Platform on Biodiversity and Ecosystem Services. Roy HE, Pauchard A, Stoett P, Renard Truong T, Bacher S, Galil BS, Hulme PE, Ikeda T, Sankaran KV, McGeoch MA, Meyerson LA, Nuñez MA, Ordonez A, Rahlao SJ, Schwindt E, Seebens H, Sheppard AW, Vandvik V (eds.). IPBES secretariat, Bonn, Germany. https://doi.org/10.5281/zenodo.7430692
IUCN (2020a) IUCN EICAT Categories and Criteria. The Environmental Impact Classification for Alien Taxa, 1st edn. IUCN, Gland, Switzerland and Cambridge, UK. https://doi.org/10.2305/IUCN.CH.2020.05.en
IUCN (2020b). Guidelines for using the IUCN Environmental Impact Classification for Alien Taxa (EICAT) Categories and Criteria: First edition. Version 1.1. Gland, Switzerland and Cambridge, UK. IUCN.
Jansen C, Kumschick S (2022) A global impact assessment of Acacia species introduced to South Africa. Biological Invasions 24(1):175-187. https://doi.org/10.1007/s10530-021-02642-0
Jarnevich CS, Stohlgren TJ, Kumar S, Morisette JT, Holcombe TR (2015) Caveats for correlative species distribution modeling. Ecol Inform 29:6–15
Jiménez-Valverde A, Peterson AT, Soberón J, Overton JM, Aragón P, Lobo JM (2011) Use of niche models in invasive species risk assessments. Biol Invasions 13:2785–2797
Kaplan H, van Niekerk A, Le Roux JJ, Richardson DM, Wilson JR (2014) Incorporating risk mapping at multiple spatial scales into eradication management plans. Biological Invasions 16:691-703
Karger DN, Conrad O, Böhner J, Kawohl T, Kreft H, Soria-Auza RW, Zimmermann NE, Linder HP, Kessler M (2017) Climatologies at high resolution for the earth’s land surface areas. Scientific Data 4(1):1-20
Kowarik I (1995) Time lags in biological invasions with regard to the success and failure of alien species. Plant invasions: general aspects and special problems, 15-38.
Kumschick S, Jansen C (2023) Evidence-based impact assessment for naturalized and invasive Australian Acacia species. Wattles: Australian Acacia species around the world (eds. D.M. Richardson, J.J. Le Roux & E.M. Marchante). CABI, Wallingford, pp. 359-381.
Kumschick S, Bertolino S, Blackburn TM, Brundu G, Costello KE, de Groot M, Evans T, Gallardo B, Genovesi P, Govender T, Jeschke JM, Lapin K, Measey GJ, Novoa A, Nunse AL, Probert AF, Pyšek P, Preda C, Rabitsch W, Roy HE, Smith KG, Tricarico E, Vilà M, Vimercati G, Bacher S (2024) Using the IUCN Environmental Impact Classification for Alien Taxa (EICAT) to inform decision-making. Conservation Biology 38. DOI:10.1111/cobi.14214
Le Maitre DC, Blignaut JN, Clulow A, Dzikiti S, Everson CS, Görgens AH, Gush MB (2020) Impacts of plant invasions on terrestrial water flows in South Africa. In Biological Invasions in South Africa (pp. 431-457). Cham: Springer International Publishing.
Le Maitre DC, Gaertner M, Marchante E et al. (2011) Impacts of invasive Australian acacias: implications for management and restoration. Diversity and Distributions 17:1015–1029
Lenoir J et al. (2008) A significant upward shift in plant species optimum elevation during the 20th century. Science 320:1768–1771
Liaw A, Wiener M (2002) Classification and regression by randomForest. R news 2(3):18-22
Magona N, Richardson DM, le Roux JJ, Kritzinger-Klopper S, Wilson JRU (2018) Even well-studied groups of alien species might be poorly inventoried: Australian Acacia species in South Africa as a case study. NeoBiota 39:1-29. https://doi.org/10.3897/neobiota.39.23135
Manes S et al. (2021) Endemism increases species’ climate change risk in areas of global biodiversity importance. Biological Conservation 257. https://doi.org/10.1016/j.biocon.2021.109070.
Marchioro CA, Krechemer FS (2021) Prevention is better than cure: Integrating habitat suitability and invasion threat to assess global biological invasion risk by insect pests under climate change. Pest Management Science 77(10):4510-4520
Mod HK, Scherrer D, Luoto M, Guisan A (2016) What we use is not what we know: environmental predictors in plant distribution models. Journal of Vegetation Science 27(6):1308-1322
Morris TL, Esler KJ, Barger NN, Jacobs SM, Cramer MD (2011) Ecophysiological traits associated with the competitive ability of invasive Australian acacias. Diversity and Distributions 17(5):898-910
Nentwig W, Bacher S, Pyšek P, Vilà M, Kumschick S (2016) The Generic Impact Scoring System (GISS): a standardized tool to quantify the impacts of alien species. Environmental Monitoring and Assessment 188:315. doi:10.1371/journal.pbio.1001850.
Nentwig W, Kühnel E, Bacher S (2010) A generic impact‐scoring system applied to alien mammals in Europe. Conservation Biology 24(1):302-311
Parravicini V, Azzurro E, Kulbicki M, Belmaker J (2015) Niche shift can impair the ability to predict invasion risk in the marine realm: an illustration using Mediterranean fish invaders. Ecol Lett 18:246–253
Pouteau R, Biurrun I, Brunel C, Chytrý M, Dawson W, Essl F, Fristoe T, Haveman R, Hobohm C, Jansen F, Kreft H, Lenoir J, Lenzner B, Meyer C, Moeslund JE, Pergl J, Pyšek P, Svenning J-C, Thuiller W, Weigelt P, Wohlgemuth T, Yang Q, van Kleunen M (2021) Potential alien ranges of European plants will shrink in the future, but less so for already naturalized than for not yet naturalized species. Diversity and Distributions 27:2063–2076. https://doi.org/10.1111/ddi.13378
Richardson DM, Allsopp N, D’Antonio CM et al. (1994) Plant invasions – the role of mutualisms. Biol Rev 75:65–93. https://doi.org/10.1111/j.1469-185X.1999.tb00041.x
Robeck P, Essl F, van Kleunen M, Pyšek P, Pergl J, Weigelt P, Mesgaran MB (2024) Invading plants remain undetected in a lag phase while they explore suitable climates. Nature Ecology & Evolution, 1-12.
Rockwell-Postel M, Laginhas BB, Bradley BA (2020) Supporting proactive management in the context of climate change: prioritizing range-shifting invasive plants based on impact. Biological Invasions 22:2371-2383
Rouget M, Robertson MP, Wilson JRU, Hui C, Essl F, Renteria JL, Richardson DM (2016) Invasion debt - quantifying future biological invasions. Diversity and Distributions 22(4):445-456
Rumlerová Z, Vilà M, Pergl J, Nentwig W, Pyšek P (2016) Scoring environmental and socioeconomic impacts of alien plants invasive in Europe. Biological Invasions 18:3697-3711
Sheppard CS, Burns BR, Stanley MC (2014) Predicting plant invasions under climate change: are species distribution models validated by field trials? Glob Change Biol 20:2800–2814
Sheppard CS, Burns BR, Stanley MC (2016) Future-proofing weed management for the effects of climate change: is New Zealand underestimating the risk of increased plant invasions?. New Zealand Journal of Ecology 40(3):398-405
Sinclair SJ, White MD, Newell GR (2010) How useful are species distribution models for managing biodiversity under future climates? Ecology and Society 15(1):8.
Skowno AL, Thompson MW, Hiestermann J et al (2017) Woodland expansion in South African grassy biomes based on satellite observations (1990-2013): general patterns and potential drivers. Glob Chang Biol 23:2358–2369. https://doi.org/10.1111/gcb.13529
Skowno AL, Jewitt D, Slingsby JA (2021) Rates and patterns of habitat loss across South Africa's vegetation biomes. South African Journal of Science 117(1-2):1-5
Slingsby JA, Merow C, Aiello-Lammens M, Allsopp N, Hall S, Kilroy Mollmann H, Turner R, Wilson AM, Silander Jr JA (2017) Intensifying postfire weather and biological invasion drive species loss in a Mediterranean-type biodiversity hotspot. Proceedings of the National Academy of Sciences 114(18):4697-4702
Smith K (2020) The IUCN Red List and invasive alien species: An analysis of impacts on threatened species and extinctions. International Union for the Conservation of Nature.
Stockwell DRB, Peterson AT (2002) Effects of sample size on accuracy of species distribution models. Ecol Modell 148:1–13.
Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BF, De Siqueira MF, Grainger A, Hannah L, Hughes L, et al. (2004) Extinction risk from climate change. Nature 427(6970):145-148
Thuiller W, Lafourcade B, Engler R, Araújo MB (2009) BIOMOD–a platform for ensemble forecasting of species distributions. Ecography 32(3):369-373
Tingley R, García-Díaz P, Arantes CRR, Cassey P (2017) Integrating transport pressure data and species distribution models to estimate invasion risk for alien stowaways. Ecography
Vilà M, Espinar JL, Hejda M, Hulme PE, Jarošík V, Maron JL, Pergl J, Schaffner U, Sun Y, Pyšek P (2011) Ecological impacts of invasive alien plants: a meta‐analysis of their effects on species, communities and ecosystems. Ecology Letters 14(7):702-708
Vilà M, Trillo A, Castro-Díez P, Gallardo B, Bacher S (2024) Field studies of the ecological impacts of invasive plants in Europe. NeoBiota 90:139-159
Volery L, Blackburn TM, Bertolino S, Evans T, Genovesi P, Kumschick S, Roy HE, Smith KG, Bacher S (2020) Improving the Environmental Impact Classification for Alien Taxa (EICAT): a summary of revisions to the framework and guidelines. NeoBiota 62:547–567
Wenger SJ, Olden JD (2012) Assessing transferability of ecological models: an underappreciated aspect of statistical validation. Methods in Ecology and Evolution 3(2):260-267
Williams JW, Jackson ST (2007) Novel climates, no-analog communities, and ecological surprises. Frontiers in Ecology and the Environment 5(9):475–482. https://doi.org/10.1890/070037
Wisz MS, Hijmans RJ, Li J, Peterson AT, Graham CH, Guisan A (2008) Predicting Species Distributions Working Group. Effects of sample size on the performance of species distribution models. Diversity and Distributions 14:763–773
You W, Yu D, Xie D, Yu L (2013) Overwintering survival and regrowth of the invasive plant Eichhornia crassipes are enhanced by experimental warming in winter. Aquatic Biology 19:45-53

Download PDF

Reviewers agreed at journal
13 Jun, 2024
Reviewers invited by journal
26 May, 2024
Editor assigned by journal
21 May, 2024
First submitted to journal
17 May, 2024

You are reading this latest preprint version

Mapping impacts of alien species on biodiversity in the face of climate change

Status:

Version 1

Abstract

Figures

Introduction

Methods

Species selection

Occurrence data

Climate variables

Species distribution models

Risk maps

Changes in impact risk in protected areas

Results

Discussion

Declarations

Competing Interests

Funding

Author contributions

Acknowledgements

References

Supplementary Files

Status:

Version 1