4. Discussion
Our study showed how non-invaded bird communities retained higher taxonomical and functional diversity (TD and FD respectively) thaninvaded communities in a highly human-modified landscape, supporting the idea that higher species diversity shapes community resistance to invasions (Gerhold et al., 2011; Lososová et al., 2015). Moreover, we found that it is more likely that alien birds occupy novel parts of the functional space in bird communities characterised by low TD and FD (invaded vs invaded no alien ), but that they do not fully compensate for the taxonomic and functional biodiversity loss caused by the absence of native species in invaded sites. Our findings also highlighted how these trends were stronger in less human-modified and more heterogeneous areas, whereas the diversity metrics were negatively affected in more human-modified areas. Finally, these results were similar (but stronger) when we considered those communities where the Feral Pigeon was present as invaded (i.e. it was treated as an alien species), suggesting that this ubiquitous bird has more characteristics of an alien than native species.
4.1. Diversity metrics and community types
In our study, the higher TD and FD found in non-invaded thaninvaded bird communities (Test 1) across land-uses likely suggests a general negative impact of alien species on native bird communities inhabiting the area. TD and FD were also higher innon-invaded than invaded no alien communities (i.e.,invaded communities considering only native species; Test 2). This impact could reflect in a loss of ecosystem functioning in the invaded bird communities (e.g. through loss of functional groups; Flynn et al., 2009; Cadotte et al. 2011). For example, high functional divergence (FDiv) infers a high niche differentiation in the community (Cosset & Edwards, 2017), suggesting that the non-invadedcommunities provide a more diverse pool of functional processes thaninvaded communities (Edwards et al., 2013). Similar reasoning could be applied when accounting for functional evenness (FEve), since high values of FEve infer efficient use of resources by species in the community (Lee & Martin, 2017). Therefore, non-invaded showed higher resilience (i.e. FEve) than invaded communities (Test 1), but lower susceptibility to disturbance than invaded no aliencommunities (Test 2). This is consistent with the biotic resistance hypothesis stating that higher diversity should promote resistance to invasion (Ordonez, 2010).
Following Test 3 (invaded vs invaded no alien ), our results showed how alien species were functionally different from native species of the resident invaded community assemblages. In fact, FDiv was higher in invaded than invaded no alien communities. This suggests that alien species provide a set of functional traits that are different from those of the native species. For example, in Portugal, the common waxbill (Estrilda astrid ) has successfully established in unsaturated communities, occupying a marginal niche (Batalha et al., 2013). Therefore, most alien species may show different adaptations to native species, leading to higher niche differentiation of theinvaded communities, suggesting that alien species usually occupy a different ecological niche in those communities (Hejda & de Bello 2013). Nonetheless, the FEve showed no significant difference betweeninvaded vs invaded no alien communities. Thus, despite that alien species have a higher niche differentiation in invadedcommunities, they do not provide a more efficient use of resources than the native species in the community.
From a PD point of view, we found that invaded communities had similar Mean Pairwise Distance (MPD) to both non-invaded andinvaded no alien communities (Test 1 and 3), while species ininvaded no alien communities were more clustered (lower MPD) than species in non-invaded communities (Test 2). The latter, along with the TD and FD results, supports the resistance hypothesis. In fact, higher PD could infer a community characterized by several adaptations, leading to increased competition and hampering alien species spread (Ketola et al., 2017). Finally, since the addition of alien species closes this phylogenetic gap with non-invaded communities (Test 1), it is more likely that alien species are favoured in more clustered communities (Lososová et al., 2015), occupying novel phylogenetic space in the invaded community (e.g. Psittacula krameri ).
These patterns were also partially true when we compared the diversity metrics within land-uses to exclude the effect of environmental filtering on bird community assemblage (Loiola et al., 2018). Nonetheless, MPD was higher for invaded than non-invadedbird communities of urban areas, suggesting a higher vulnerability to alien species invasion (i.e. a phylogenetic gap). Moreover, no differences were found between community types within forest, either due to the low proportion of invaded communities, or because alien bird species seem to be weakly influenced by environmental filtering compared to its effect on the structure of native bird communities (Lazarina et al., 2022).