Discussion
Existing literature reports both, positive as well as negative effects of non-native plant species on resident biodiversity. However, negative effects have been noted more frequently especially for insect diversity (see Wohlgemuth et al. 2022 for a recent meta-analysis on the effects). Contradicting findings always raise the question about global drivers of the observable effects, that might help to reconcile such contrasting findings existent in the published literature. Geographic, climatic and phylogenetic distance between the non-native plant species and its native congeners or area of introduction have been named as three dominant drivers of the strength and direction of the effects of non-native plant species on biodiversity. In our study, we identified phylogenetic distance between non-native and native tree species to be the main diver of resident insect abundance and diversity with lower insect abundance and diversity being recorded on more phylogenetically distant, non-native species. This predominance of the phylogenetic effect in contrast to geographic or climatic factors might relate to the dominant role of evolutionary forces shaping plant-insect interactions such as the process of coevolution.
There is increasing consensus, that diversity of phytophagous arthropods resulted from balanced rates of successful colonization of the new host, speciation on the respective host and local extinction (Farrell et al. 1992). The main mechanism driving the positive relationship between phylogenetic relatedness among host plants and their herbivore colonization is proposed to be a high similarity of functional traits involved in host plant selection or host resistance (Pearse et al. 2013; Ricciardi & Ward 2006). Whether a shared phylogenetic history between plant species can explain similarities in inhabited insect communities thereby strongly depends on the degree of conservatism of such herbivore related traits (Fine et al. 2006). Insect-plant associations are reported to be generally very conservative with respect to the evolution of new host affiliations and, thus, may persist over extensive periods of the phylogenetic history (Farrellet al. 1992; Farrell & Mitter 1994). Closely related host plant species are therefore expected to be more likely to be colonized by the same, specialized insect species due to a high degree of phylogenetic conservatism being reported for morphological and chemical plant traits that regulate interactions with herbivores (Brändle et al. 2008; Gilbert & Webb 2007; Whitfeld et al. 2012). Proportion of specialized, herbivorous insects are generally reported to be high (Ali & Agrawal 2012; Price et al. 2011). However, many plant functional traits affecting plant-insect interactions are strongly shaped by the abiotic environment (e.g. leaf thickness and hairs, thickness of the cuticula as a response to low water availability and/or high radiation etc.) – a fact that might challenge plant-insect trait matching dynamics acting over evolutionary time scales in variable environments.
Already Darwin proposed in his naturalization hypothesis that non-native plants are more likely to invade if they lack close relatives in their new range (Darwin 1859). Several studies highlight a positive relationship between the abundance and phylogenetic relatedness of non-native plants with their native congeners and the likelihood that the non-native species get colonized by native herbivores (Goßneret al. 2009; Kirichenko & Kenis 2016; Neuvonen & Niemelä 1981; Roques et al. 2006). This phylogenetic effect on plant-herbivore interactions has been shown to be especially strong for tropical plant-insect herbivore interactions, where 25% of the variance in herbivore community similarity between tropical tree species was explained by the phylogenetic similarity of the host tree species (Weiblen et al. 2006). Furthermore, host switches of phytophagous insects have been reported for the temperate region to be more probable to happen between related host tree species than between more unrelated host species which co-occurred in the same habitat (Brändle & Brandl 2006). Another impressive though anecdotal example of a high degree of phylogenetic conservatism in host-plant insect interactions is reported for Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco), a non-native, economically important tree species with no close relatives in Europe. In the 130 years since this species has been introduced to Europe, it has recruited only 87 arthropod species, despite being the most planted, non-native tree species in European forest systems (Roqueset al. 2006). These are only 33.9% of the arthropod species richness found on Douglas-fir in its native range in North America. Our results show that such a strong phylogenetic effect on plant-insect interactions of non-native trees is not a peculiarity of single tree species but a general phenomenon with extensive geographic and taxonomic coverage. In our study these effects have been detected for non-native species of native tree genera, showing that this effect even holds true on lower taxonomic levels.
Our results furthermore show that the effect of phylogenetic distance on resident insect abundance, species richness and diversity is visibly stronger for phytophagous insect species in comparison to non-phytophagous species. This reflects theoretical assumptions proposed for the enemy release hypothesis (ERH), one of the most prominent hypotheses in plant-related invasion biology. It states that host-specific herbivores are left behind when a plant species is transferred to a new area that haven’t been occupied before (Keane & Crawley 2002). This lack of native antagonists in a plant host’s introduced range can favour its vitality and spread (Adams et al.2009; Cincotta et al. 2009). Lower trait similarity between phylogenetically distant, non-native plants and their native congeners in the area of introduction has been proposed as an underlying, prominent mechanism explaining the establishment success of such non-natives (Goßner et al. 2009). In this framework, our results would basically mean, that such release effects might be stronger, the more phylogenetically distant non-native plant species will be to their native counterparts on place (for empirical evidence see e.g. Nesset al. 2011).
Although phylogenetic distance turned out to be the predominant, global driver in our study, other context-specific drivers of the effects of non-native plants on resident biodiversity have to be expected to be important on local scale (Wohlgemuth et al. 2022). Local drivers such as density effects (i.e. trees planted in mono- vs. mixed stands; Adams et al. 2009; Yang et al. 2019) or differences in species pools of arthropods (Miles et al. 2019 in the context of urbanisation) are named to be important. However, these effects might not play out in a common garden setting such as ours, where all tree species are experiencing a very similar abiotic and biotic environment and, thus, genetic differences might predominate the responses (Allenet al. 2017). Local differences in environmental conditions might overwrite the effects of phylogenetic differences between non-native species and their native congeners when comparing different, local study sites. Other exogenic factors such as time since introduction in a new biogeographic realm (i.e. residence time) can play an additional, crucial role for native insect - non-native host plant interactions (Brändle et al. 2008; Sheppard & Brendel 2021). Further investigations on other tree genera and species, as well as in other geographical areas will be needed to test the relative importance of local vs. global drivers of non-native tree species on resident insect communities. Coordinated, spatially distributed, common garden experiments would help to advance in this direction. Tree collections in botanical gardens can provide a fruitful basis for such kind of investigations.