Aims Studies integrating phylogenetic history and large-scale community assembly are few, and many questions remain unanswered. Here, we use a global coastal dune plant data set to uncover the important factors in community assembly across scales from the local filtering processes to the global long-term diversification and dispersal dynamics. Coastal dune plant communities occur worldwide under a wide range of climatic and geologic conditions as well as in all biogeographic regions. However, global patterns in the phylogenetic composition of coastal dune plant communities have not previously been studied.
Methods The data set comprised vegetation data from 18463 plots in New Zealand, South Africa, South America, North America and Europe. The phylogenetic tree comprised 2241 plant species from 149 families. We calculated phylogenetic clustering (Net Relatedness Index, NRI, and Nearest Taxon Index, NTI) of regional dune floras to estimate the amount of in situ diversification relative to the global dune species pool and evaluated the relative importance of land and climate barriers for these diversification patterns by geographic analyses of phylogenetic similarity. We then tested whether dune plant communities exhibit similar patterns of phylogenetic structure within regions. Finally, we calculated NRI for local communities relative to the regional species pool and tested for an association with functional traits (plant height and seed mass) thought to vary along sea–inland gradients.
Important findings Regional species pools were phylogenetically clustered relative to the global pool, indicating regional diversification. NTI showed stronger clustering than NRI pointing to the importance of especially recent diversifications within regions. The species pools grouped phylogenetically into two clusters on either side of the tropics suggesting greater dispersal rates within hemispheres than between hemispheres. Local NRI plot values confirmed that most communities were also phylogenetically clustered within regions. NRI values decreased with increasing plant height and seed mass, indicating greater phylogenetic clustering in communities with short maximum height and good dispersers prone to wind and tidal disturbance as well as salt spray, consistent with environmental filtering along sea–inland gradients. Height and seed mass both showed significant phylogenetic signal, and NRI tended to correlate negatively with both at the plot level. Low NRI plots tended to represent coastal scrub and forest, whereas high NRI plots tended to represent herb-dominated vegetation. We conclude that regional diversification processes play a role in dune plant community assembly, with convergence in local phylogenetic community structure and local variation in community structure probably reflecting consistent coastal-inland gradients. Our study contributes to a better understanding of the globally distributed dynamic coastal ecosystems and the structuring factors working on dune plant communities across spatial scales and regions.

Aims While using phylogenetic and functional approaches to test the mechanisms of community assembly, functional traits often act as the proxy of niches. However, there is little detailed knowledge regarding the correlation between functional traits of tree species and their niches in local communities. We suggest that the co-varying correlation between functional traits and niches should be the premise for using phylogenetic and functional approaches to test mechanisms of community assembly. Using functional traits, phylogenetic and environmental data, this study aims to answer the questions: (i) within local communities, do functional traits of co-occurring species co-vary with their environmental niches at the species level? and (ii) what is the key ecological process underlying community assembly in Xishuangbanna and Ailaoshan forest dynamic plots (FDPs)?
Methods We measured seven functional traits of 229 and 36 common species in Xishuangbanna and Ailaoshan FDPs in tropical and subtropical China, respectively. We also quantified the environmental niches for these species based on conditional probability. We then analyzed the correlations between functional traits and environmental niches using phylogenetic independent contrasts. After examining phylogenetic signals of functional traits using Pagel's λ, we quantified the phylogenetic and functional dispersion along environmental gradients within local tree communities.
Important findings For target species, functional traits do co-vary with environmental niches at the species level in both of the FDPs, supporting that functional traits can be used as a proxy for local-scale environmental niches. Functional traits show significant phylogenetic signals in both of the FDPs. We found that the phylogenetic and functional dispersion were significantly clustered along topographical gradients in the Ailaoshan FDP but overdispersion in the Xishuangbanna FDP. These patterns of phylogenetic and functional dispersion suggest that environmental filtering plays a key role in structuring local tree assemblages in Ailaoshan FDP, while competition exclusion plays a key role in Xishuangbanna FDP.

Aims Recent mechanistic explanations for community assembly focus on the debates surrounding niche-based deterministic and dispersal-based stochastic models. This body of work has emphasized the importance of both habitat filtering and dispersal limitation, and many of these works have utilized the assumption of species spatial independence to simplify the complexity of the spatial modeling in natural communities when given dispersal limitation and/or habitat filtering. One potential drawback of this simplification is that it does not consider species interactions and how they may influence the spatial distribution of species, phylogenetic and functional diversity. Here, we assess the validity of the assumption of species spatial independence using data from a subtropical forest plot in southeastern China.
Methods We use the four most commonly employed spatial statistical models—the homogeneous Poisson process representing pure random effect, the heterogeneous Poisson process for the effect of habitat heterogeneity, the homogenous Thomas process for sole dispersal limitation and the heterogeneous Thomas process for joint effect of habitat heterogeneity and dispersal limitation—to investigate the contribution of different mechanisms in shaping the species, phylogenetic and functional structures of communities.
Important findings Our evidence from species, phylogenetic and functional diversity demonstrates that the habitat filtering and/or dispersal-based models perform well and the assumption of species spatial independence is relatively valid at larger scales (50×50 m). Conversely, at local scales (10×10 and 20×20 m), the models often fail to predict the species, phylogenetic and functional diversity, suggesting that the assumption of species spatial independence is invalid and that biotic interactions are increasingly important at these spatial scales.

Aims The coastal Brazilian rainforest on white-sand (restinga) ranks among the most fragmented forest types in the tropics, owing to both the patchy distribution of sandy soils and widespread coastal development activities. Here we study the environmental and evolutionary determinants of a forest tree assemblage at a single restinga forest in Southeastern Brazil. We also explore the ability of competing hypotheses to explain the maintenance of species diversity in this forest type, which includes contrasting extremes of edaphic conditions associated with flooding stress.
Methods The study was conducted in a white-sand forest permanent plot of 10.24 ha on the coastal plain of Southeastern Brazil. This plot was divided into 256 quadrats of 20×20 m, which were classified into two main edaphic habitats (flooded and drained). Trees with a diameter ≥1cm at breast height were identified. We assembled DNA sequence data for each of the 116 morphospecies recognized using two chloroplast markers (rbcL and matK). A phylogenetic tree was obtained using the maximum likelihood method, and a phylogenetic distance matrix was produced from an ultrametric tree. We analyzed similarity in floristic composition and structure between habitats and related them to cross-plot distances using permutation procedures. Null model torus shift simulations were performed to obtain a statistical significance level for habitat association for each species. The phylogenetic structure for the two habitats and for each 20×20 m quadrat was calculated using the mean phylogenetic distance weighted by species abundance and checked for significance using the standardized effect size generated by 5000 randomizations of phylogenetic tip labels.
Important findings Our results indicate that partitioning among edaphic habitats is important for explaining species distributions and coexistence in restinga forests. Species distributions within the plot were found to be non-random: there was greater floristic similarity within than between habitats, and>40% of the more abundant species were positively or negatively associated with at least one habitat. Patterns of habitat association were not independent of phylogenetic relatedness: the community was overdispersed with respect to space and habitat type. Closely related species tended to occur in different habitats, while neighboring trees tended to belong to more distantly related species. We conclude that habitat specialization is important for the coexistence of species in restinga forests and that habitat heterogeneity is therefore an essential factor in explaining the maintenance of diversity of this unique but fragile and threatened type of forest.

Aims Environmental gradients are drivers of species diversity; however, we know relatively little about the evolutionary processes underlying these relationships. A potentially powerful approach to studying diversity gradients is to quantify the phylogenetic structure within and between assemblages arrayed along broad spatial and environmental gradients. Here, we evaluate the phylogenetic structure of plant assemblages along an environmental gradient with the expectation that the habitat specialization of entire lineages is an important evolutionary pattern influencing the structure of tree communities along environmental gradients.
Methods We evaluated the effect of several environmental variables on the phylogenetic structure of plant assemblages in 145 plots distributed in northwestern South America that cover a broad environmental gradient. The phylogenetic alpha diversity was quantified for each plot and the phylogenetic beta diversity between each pair of plots was also quantified. Both the alpha and beta diversity measures were then related to spatial and environmental gradients in the study system.
Important findings We found that gradients in temperature and potential evapotranspiration have a strong relationship with the phylogenetic alpha diversity in our study system, with phylogenetic overdispersion in low temperatures and phylogenetic clustering at higher temperatures. Further, the phylogenetic beta diversity between two plots increases with an increasing difference in temperature, whereas annual precipitation was not a significant predictor of community phylogenetic turnover. We also found that the phylogenetic structure of the plots in our study system was related to the degree of seasonal flooding and seasonality in precipitation. In particular, more stressful environments such as dry forests and flooded forests showed phylogenetic clustering. Finally, in contrast with previous studies, we find that phylogenetic beta diversity was not strongly related to the spatial distance separating two forest plots, which may be the result of the importance of the three independent mountain ranges in our study system, which generate a high degree of environmental variation over very short distances. In conclusion, we found that environmental gradients are important drivers of both phylogenetic alpha and phylogenetic beta diversities in these forests over spatial distance.

Aims Understanding what drives the variation in species composition and diversity among local communities can provide insights into the mechanisms of community assembly. Because ecological traits are often thought to be phylogenetically conserved, there should be patterns in phylogenetic structure and phylogenetic diversity in local communities along ecological gradients. We investigate potential patterns in angiosperm assemblages along an elevational gradient with a steep ecological gradient in Changbaishan, China.
Methods We used 13 angiosperm assemblages in forest plots (32×32 m) distributed along an elevational gradient from 720 to 1900 m above sea level. We used Faith's phylogenetic diversity metric to quantify the phylogenetic alpha diversity of each forest plot, used the net relatedness index to quantify the degree of phylogenetic relatedness among angiosperm species within each forest plot and used a phylogenetic dissimilarity index to quantify phylogenetic beta diversity among forest plots. We related the measures of phylogenetic structure and phylogenetic diversity to environmental (climatic and edaphic) factors.
Important findings Our study showed that angiosperm assemblages tended to be more phylogenetically clustered at higher elevations in Changbaishan. This finding is consistent with the prediction of the phylogenetic niche conservatism hypothesis, which highlights the role of niche constraints in governing the phylogenetic structure of assemblages. Our study also showed that woody assemblages differ from herbaceous assemblages in several major aspects. First, phylogenetic clustering dominated in woody assemblages, whereas phylogenetic overdispersion dominated in herbaceous assemblages; second, patterns in phylogenetic relatedness along the elevational and temperature gradients of Changbaishan were stronger for woody assemblages than for herbaceous assemblages; third, environmental variables explained much more variations in phylogenetic relatedness, phylogenetic alpha diversity and phylogenetic beta diversity for woody assemblages than for herbaceous assemblages.

Aims The last decade has seen many plant ecologists integrating phylogenetic analysis into ecology to explain patterns of species co-occurrence and compositional similarity across assemblages. Despite the advances in this area, there are still some challenges that need to be addressed. One challenge is that most of the phylogenetic studies of plant assemblages have focused only on a small proportion of all of the vascular plants that co-occur (e.g. woody plants), while much of the remaining co-occurring flora has been ignored.
Methods Here we introduce an analytical approach that we term phylofloristics that analyzes the compositional similarity of floras in relation to spatial and environmental gradients to understand their assembly. As an illustration, we assembled a large phylogenetic tree for the flora of the Lesser Antilles and evaluated the patterns of floristic and phylofloristic similarity among the island-specific floras. We analyzed the relationship of these similarities with spatial and environmental distance and compared the results for non-endemic and endemic lineages.
Important findings The results show a major influence of environmental heterogeneity on the assembly of island floras and far less evidence for the importance of dispersal limitation of lineages and species. This study highlights the importance of incorporating broader taxonomic sampling to improve our understanding of assembly processes in ecology. We expect future phylofloristic studies will improve the approach we have taken by generating more refined phylogenetic trees and by incorporating phylogeographic information.

Aims The relationship between biodiversity and ecological stability is a long-standing issue in ecology. Current diversity–stability studies, which have largely focused on species diversity, often report an increase in the stability of aggregate community properties with increasing species diversity. Few studies have examined the linkage between phylogenetic diversity, another important dimension of biodiversity, and stability. By taking species evolutionary history into account, phylogenetic diversity may better capture the diversity of traits and niches of species in a community than species diversity and better relate to temporal stability. In this study, we investigated whether phylogenetic diversity could affect temporal stability of community biomass independent of species diversity.
Methods We performed an experiment in laboratory microcosms with a pool of 12 bacterivorous ciliated protist species. To eliminate the possibility of species diversity effects confounding with phylogenetic diversity effects, we assembled communities that had the same number of species but varied in the level of phylogenetic diversity. Weekly disturbance, in the form of short-term temperature shock, was imposed on each microcosm and species abundances were monitored over time. We examined the relationship between temporal stability of community biomass and phylogenetic diversity and evaluated the role of several stabilizing mechanisms for explaining the influence of phylogenetic diversity on temporal stability.
Important findings Our results showed that increasing phylogenetic diversity promoted temporal stability of community biomass. Both total community biomass and summed variances showed a U-shaped relationship with phylogenetic diversity, driven by the presence of large, competitively superior species that attained large biomass and high temporal variation in their biomass in both low and high phylogenetic diversity communities. Communities without these species showed patterns consistent with the reduced strength of competition and increasingly asynchronous species responses to environmental changes under higher phylogenetic diversity, two mechanisms that can drive positive diversity–stability relationships. These results support the utility of species phylogenetic knowledge for predicting ecosystem functions and their stability.

Aims To examine if and how species and phylogenetic diversity change in relation to disturbance, we conducted a review of ecological literature by testing the consistency of the relationship between phylogenetic diversity and disturbance and compared taxonomic groups, type of disturbance and ecosystem/habitat context. We provide a case study of the phylogenetic diversity–disturbance relationship in angiosperm plant communities of a boreal forest region, compared with types of natural and anthropogenic disturbances and plant growth forms.
Methods Using a large-scale sampling plot network along a complete (0–100%) anthropogenic disturbance gradient in the boreal biome, we compared the changes of angiosperm plant community structure and composition across plots. We estimated natural disturbance with historical records of major fires. We then calculated phylogenetic diversity indexes and determined species richness in order to compare linear and polynomial trends along disturbance gradients. We also compared the changes of community structure for different types of anthropogenic disturbances and examined how the relationships between species and phylogenetic diversity and disturbance regimes vary among three different life forms (i.e. forbs, graminoids and woody plants).
Important findings Phylogenetic diversity was inconsistently related to disturbance in previous studies, regardless of taxon, disturbance type or ecosystem context. In the understudied boreal ecosystem, angiosperm plant communities varied greatly in species richness and phylogenetic diversity along anthropogenic disturbance gradients and among different disturbance types. In general, a quadratic curve described the relationship between species richness and anthropogenic disturbance, with the highest richness at intermediate anthropogenic disturbance levels. However, phylogenetic diversity was not related to disturbance in any consistent manner and species richness was not correlated with phylogenetic diversity. Phylogenetic relatedness was also inconsistent across plant growth forms and different anthropogenic disturbance types. Unlike the inconsistent patterns observed for anthropogenic disturbance, community assembly among localities varying in time since natural disturbance exhibited a distinct signature of phylogenetic relatedness, although those trends varied among plant growth forms.

Aims Theory predicts that the success of introduced species is related to the diversity of native species through trait-based processes. Abiotic site characteristics may also affect a site's susceptibility to invasion. We quantified resident plant species richness, phylogenetic diversity and several abiotic site characteristics for 24 oak forests in Minnesota, USA, to assess their impact on the abundance of a widespread, introduced terrestrial plant species, common buckthorn (Rhamnus cathartica L.). Specifically, we asked (1) whether resident species richness and phylogenetic diversity affected the abundance of R. cathartica and (2) what site characteristics explained the overall abundance of R. cathartica .
Methods Our survey included 24 oak-dominated stands in Minnesota's deciduous forests. In each stand, we identified all species in 16 plots. We also measured a series of environmental site characteristics, including canopy openness (a proxy for light availability), percent bare soil, soil pH, percent sand, an index of propagule availability, duff layer thickness (a proxy for earthworm activity), an index of insolation and slope. For all species present in at least one site, we estimated a community phylogeny. We combined all site-level characteristics, including phylogenetic diversity of the resident plant species, in a multiple regression model to examine site level drivers of community invasibility.
Important findings Results indicate that sites with higher overall plant phylogenetic diversity harbor less R. cathartica, even though native species richness was not significantly related to R. cathartica abundance. Regression analyses indicated that, in addition to resident species phylogenetic diversity, the most important predictors of R. cathartica abundance were canopy openness and the amount of bare soil, both positively related to the abundance of the invader. By combining the effects of abiotic site characteristics and resident species phylogenetic diversity in a model that predicted the abundance of R. cathartica, we were able to simultaneously account for a wide range of factors that might influence invasibility. Overall, our results suggest that management strategies aimed at reducing disturbances that lead to increased bare soil and light levels may be more successful if they also maximize phylogenetic diversity of the resident plant community.