The importance of interspecific competition in the actual and future distributions of plant species assessed by a 2-D grid agent modelling

- We report a model to link plant distribution, climate and biotic interactions.
- It is used to simulate the future plant distribution under two climatic scenarios.
- Methodology: integration between Cellular Automata and statistical tools.
- Comparison with other studies that considered single species only.
- Mutualistic relationships among plant species impede abrupt climate change-induced movements to more suitable sites.

Currently, potential distribution of plant species is represented by different uses of presence/absence indicators or by density-dominance-based ones such as the Importance Value (I.V.), and their geographical representation is based on statistical models (Random forest model, General Regression Models etc.) relating these indicators with climate and physical features of a given territory. Here, we have built an integrated model which is able to link climate, physical features of territory and inter-specific competition at aiming to simulate potential distribution of eight plant species (Quercus cerris, Quercus pubescens, Fraxinus ornus, Fraxinus excelsior, Acer campestre, Acer obtusatum, Ostrya carpinifolia and Carpinus betulus) either at actual climate conditions or future ones (B1 and A2 climatic scenarios; IPCC, 2013). The integrated model is based on innovative methodological approach, which combined statistical tools (Principal Component Analyses and Discretization) and two-dimensional grid based model to consider the interactive effects of climate and inter-specific competition on plant species distribution. The study case was given by geographical spatialisation of the I.V. for each considered plant species, in the overall Italian territory. The simulations had always occurred by taking into consideration both expected climate change and inter-specific competition under highly complexity terrains. The results substantially differed from other studies that used only environmental predictors for estimating potential plant distributions. Variations in the I.V. distribution seemed to be due to variations of relative competitive abilities of plants, reducing both local extinctions of some species in Southern and Central Italy and migration toward North. Furthermore, our simulations suggested that under pressure of climate change the competitive ability of plant species will be likely affected by reducing the effects of competitive asymmetry.