Functional similarity and dissimilarity facilitate alien plant invasiveness along biotic and abiotic gradients in an arid protected area

Abstract

Two contradictory hypotheses have been put forth to forecast alien invasiveness: being either functionally similar, or dissimilar, to resident natives along environmental gradients. The ‘try-harder’ hypothesis predicts that alien plants will be functionally dissimilar to natives and should thus exhibit exaggerated trait values when compared to natives in respect to resource extraction or stress tolerance. In contrast, the ‘join-the-locals’ hypothesis, which is compatible with “environmental filtering”, predicts functional similarities among alien and native species in richer, but not in resource-limited environments. Here, we propose a framework that links the successful alien plant strategy, i.e. being functionally similar or dissimilar to natives, to the harshness of the environment and the availability of resources. We tested these two hypotheses using a trait-environment dataset of 33 alien and 130 native plants in 96 sites covering a gradient of soil resources (organic matter, nitrogen, soil moisture) in Saint-Katherine, an arid protected area in Egypt. We estimated 18 interaction coefficients between three candidate traits (specific leaf area, above-ground biomass, height) of alien and native plants as well as soil resources using linear mixed-effects models. Additionally, we calculated the mean and the hierarchical functional distance among aliens and natives along soil gradients. Our results revealed that in extreme resource-limited environments, aliens and natives were functionally similar and had relatively equal trait values consistent with environmental filtering, thus supporting the ‘join-the-locals’ hypothesis. However, in environments richer in resources, aliens and natives were functionally dissimilar with aliens exhibiting more exploitative trait values (taller, higher SLA and biomass) than natives, providing support for the ‘try-harder’ hypothesis. While demonstrated in only one arid system, results suggest that linking soil resource availability with functional divergence and convergence among native and alien plants could be used as an informative strategy to predict alien invasiveness. Hence, future studies should investigate the functional response of alien and native plants in different environments against different resource gradients to test for the generality of the patterns we found.

Appendix: Estimation of native SLA and biomass data using allometric equations

Because there was no allowance of destructively measuring native biomass and SLA in SKP, we estimated the above-ground biomass and SLA of native plants using allometric equations (Basuki et al. 2009). First, we used four non-destructive traits directly measured in the field per native plant individual. These traits were plant height (cm), leaf area (cm²) (drawn in the field then scanned using IMAGEJ), number of leaves and number of flowers and fruits. Second, we constructed series of multivariate regression models to calibrate a predictive model of aliens above-ground biomass and SLA as a function of their height (cm), number of leaves, flowers, fruits and leaf area (cm²). Then, we validated the predictive power of this model through model selection criteria using Akaike Information Criterion (AIC) and Akaike Weights (AW) in addition to diagnostic regression plots (e.g. QQ plots) to infer the best regression models. Additionally, we found high correlation between the actual observed values of biomass and SLA and the predicted values from the best regression models. Based on this validation, the best fitted model was used to predict natives’ biomass and SLA in each stand based on their empirical, field-measured trait values (see Table 3).

Initially, we considered using biomass and SLA data from global databases such as TRY or LEDA, because trait values were found for only 24 native species (out of 133). Likewise, database trait values of these native species that could also be measured non-destructively were much larger compared to the ones that were measured in the field. This difference might indicate the effects of the hyper arid conditions prevailing in Saint-Katherine. We also considered imputing missing values from less than 20% of the native species as extremely unreliable. Thus, we refrained from using trait values derived from databases and believe that the field - measured trait values (and their allometric derivatives) are more representative and relevant for this study.

Table 3 List of the 130 native and 33 alien plants with mean trait values and SD per species that were recorded in 96 stands

Fig. 5

PCA showing the contribution of the soil resources and functional traits of the two axes of the PCA in explaining the co-occurrence of aliens and natives that resulted in observed overlapped niche (similarity) and differential niche (dissimilarity)

Fig. 6

Observed trait differences (Above-ground biomass-specific leaf area (SLA)—height) among alien (red dots) and native (blue dots) plant species in response to soil resources (nitrogen, organic matter, soil moisture) in SKP