Alien plant invasion hotspots and invasion debt in European woodlands
Prepared by Viktoria Wagner, Martin Večeřa, Borja Jiménez-Alfaro, Jan Pergl, Jonathan Lenoir, Jens-Christian Svenning, Petr Pyšek, Emiliano Agrillo, Idoia Biurrun, Juan Antonio Campos, Jörg Ewald, Federico Fernández-González, Ute Jandt, Valerijus Rašomavičius, Urban Šilc, Željko Škvorc, Kiril Vassilev, Thomas Wohlgemuth & Milan Chytrý
Woodlands cover a third of the terrestrial European territory and provide important ecosystem services, including carbon sequestration, timber production, and biodiversity preservation. Over the last five centuries, European woodlands have also been colonized by at least 386 alien plant species, of which some have become invasive and transformative, such as black locust (Robinia pseudoacacia) and black cherry (Prunus serotina).
The factors that facilitate local alien species richness and abundance across European woodlands are not well understood. To gain a better insight, we analysed a stratified selection of vegetation-plot data from the European Vegetation Archive (final dataset size: 16,211 plots). We calculated relative alien species richness and sum of alien species covers for each plot and extracted 75 environmental variables that characterize regional, landscape and local attributes, such as climate and soil conditions, elevation, human pressure, habitat type and region. We asked how levels of invasion were connected to these predictors. Two data properties made our statistical analysis tricky: Our response variables had an excess number of zeroes, i.e. many plots without any alien plant. In addition, a visual inspection of relationships between response and predictor variables indicated that trends could be caused by plots close to maximum values. Hence, we tackled our data analysis with both general additive models (GAMs) and quantile generalized additive models (qGAMs) – with the latter being a flexible modelling technique that can fit functions across different quantiles.
Our results showed that relative alien species richness and the sum of alien species covers varied across habitat types, with highest levels of invasion found in Temperate and boreal mountain Betula and Populus tremula woodlands, Acidophilous Quercus woodlands, Fagus woodlands on acid soils and Temperate and boreal softwood riparian woodlands. This result was not surprising as many other studies found differences in the levels of invasion across habitats. A more unexpected result was that the differences in levels of invasion varied across regional grid cells, with highest values observed in north-west Europe and in parts of Central Europe and lowest in the Mediterranean region. Our results support a previous study by Ronk et al. (2017) which found similar regional patterns in Europe’s alien species pool. However, their study was conducted at a much coarser scale, whereas our study demonstrated regional patterns in local levels of invasion. Furthermore, the levels of invasion declined with increasing elevation and distance to the nearest road or railroad and increased with the amount of sealed soil, which confirmed our initial hypothesis that human pressure is driving invasions. Trends between response and predictor variables were always more pronounced at high quantiles, rather than at average quantiles, which suggests dispersal limitations among alien plants.
Our results imply an invasion lag in European woodlands, i.e., a time lag between the introduction and spread of alien species, regional invasion hotspots and invasion debt. Based on the outcome of this study, we predict that alien species will continue to spread further through European woodlands, facilitated by ecological differences associated with habitat types, transportation corridors and urban areas.
References
- Ronk, A., Szava‐Kovats, R., Zobel, M. & Pärtel, M. (2017) Observed and dark diversity of alien plant species in Europe: estimating future invasion risk. Biodiversity and Conservation, 26, 899–916. https://doi.org/10.1007/s10531‐016‐1278‐4
This is a plain language summary for the paper of Wagner et al. published in the Journal of Vegetation Science (https://doi.org/10.1111/jvs.13014).
