Figure 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
An example of four pairs of hypothetical communities and types of phylogenetic beta diversity. The species in a single community have the same color boxes. Species that are in neither community are left blank. All branch lengths are set to one and all species are scored as present or absent in this simplified example. It is important to note that in each of the four scenarios there is a complete turnover of species between the two communities, but the degree of phylogenetic beta diversity varies. Scenario A indicates species in the blue community are closely related to one another, but distantly related to the species in the orange community. This is an example of ‘basal’ phylogenetic turnover. Scenario B also indicates species in the blue community are closely related to one another, but distantly related to the species in the orange community. The main difference in that Scenario B has a much lower level of ‘basal’ phylogenetic beta diversity than that in Scenario A. Scenario C indicates locally phylogenetically overdispersed communities that have little phylogenetic beta diversity. Scenario D also indicates local phylogenetic overdispersion and low phylogenetic beta diversity. In both scenarios phylogenetic beta diversity measured using a nearest neighbor metric will be lower than when measured using a pairwise metric that considers the basal portion of the phylogeny and this effect will be maximized in Scenario C. doi:10.1371/journal.pone.0021264.g001
Source publication
The beta diversity of communities along gradients has fascinated ecologists for decades. Traditionally such studies have focused on the species composition of communities, but researchers are becoming increasingly interested in analyzing the phylogenetic composition in the hope of achieving mechanistic insights into community structure. To date man...
Contexts in source publication
Context 1
... tree communities?; ( ii ) how much phylogenetic signal in trait data is needed for phylogenetic beta diversity metrics to reflect the functional beta diversity and how does this vary from metric to metric?; and ( iii ) are any of the eight phylogenetic beta diversity metrics used in this study redundant and which provide novel insights? The second and third questions are largely of a methodological nature, but answering these questions is critical for one to appropriately address the first question posed. That is, without addressing the statistical underpinnings and relationships of the large number of phylogenetic beta diversity metrics that are accumulating it is difficult, if not irresponsible, to address the biological questions of interest with these metrics. Thus, the work will primarily focus on the key methodological questions while trying to provide some biological insights along the way. Several metrics of phylogenetic beta diversity have been produced in recent years. In Figure 1 I present a simplified picture of different types of phylogenetic beta diversity or turnover where phylogenetic beta diversity is relatively ‘basal’ or ‘terminal’. In this hypothetical set of scenarios the species turnover between the communities being compared is complete or in other words species beta diversity is the maximum possible. In contrast the phylogenetic beta diversity is more variable. The present work seeks to analyze eight of the most commonly implemented metrics. There are undoubtedly alternative metrics that have been developed or that will be developed, but for the time being the manuscript will be constrained to the follow set of eight. The first metric I used is phylogenetic analog of Sorensen’s Index termed PhyloSor [15]: where BL k1k2 is the total length of the branches shared between community k 1 and k 2 , BL k1 and BL k2 are the total branch lengths found in communities k 1 and k 2 respectively. This metric may be considered a ‘basal’ metric upon initial inspection, but in reality most of the variability in values necessarily comes from the terminal aspects of the phylogeny unless communities turnover over almost entirely between very basal clades, but this is likely never occurring. The second metric used is a presence-absence weighted dissimilarity metric representing the unique fraction ( UniFrac ) of the phylogeny represented between two communities [12]: where n is the number of branches in the phylogeny, BL i is the length of branch l , k 1l and k 2l are the numbers of species descendent from branch l in communities k and k . Lastly k and k 2T are the total numbers of species in communities k 1 and k 2 respectively. Similar, to the PhyloSor metric this metric primarily will detect ‘terminal’ phylogenetic beta diversity. The third metric used is presence-absence weighted and calculates the mean nearest phylogenetic neighbor between two communities [21]: where min d ik 2 is the nearest phylogenetic neighbor to species i in community k 1 in community k 2 and min d ik 1 is the nearest phylogenetic neighbor to species j in community k 2 in community k 1 . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fourth metric is similar to the above nearest neighbor metric except that it is abundance weighted [21,21]: where f i and f j are the relative abundance of species i and species j . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fifth metric is a presence-absence weighted pairwise phylogenetic dissimilarity metric [21]: where d ik 2 is the mean pairwise phylogenetic distance between species i in community k 1 to all species in community k 2 and d ik 1 is the mean pairwise phylogenetic distance between species j in community k 2 to all species in community k 1 . This metric unlike those above is a ‘basal’ metric of phylogenetic beta diversity. The sixth metric is an abundance weighted version of the above pairwise phylogenetic dissimilarity [21,22]: where f i and f j are the relative abundance of species i and species j . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. The seventh metric is derived from Rao’s quadratic entropy [13,23]: where the variables are the same as those used the above nearest neighbor and pairwise metrics. This metric can be considered a ‘basal metric of phylogenetic beta diversity. The final metric standardizes Rao’s D based upon differences in alpha diversity between the two communities: where d k 1 is the mean pairwise phylogenetic distance between species in community k 1 and d jk 1 is the mean pairwise phylogenetic distance between species in community k 2 . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. In the above I describe the eight metrics as relatively ‘terminal’ or ‘basal’ metrics. To demonstrate this property I have calculated each of the presence-absence weighted metrics using the four simplified scenarios showed in Figure 1. I performed the calculations on the original tree in Figure 1 and on four transformed versions of that tree using a lambda transformation [24]. The last of these transformations generated a star phylogeny, which allowed for the comparison of the metrics when all species are equally related. The results in Table 1 provide initial insights into the similarity of some of the metrics and their ability to detect terminal versus basal phylogenetic turnover. In general the nearest neighbor metrics of Dnn, PhyloSor and UniFrac were able to detect terminal turnover in Scenarios C and D and contrast them with the basal turnover in Scenarios A and B. The pairwise metrics of Dpw, Rao’s D and Rao’s H were able to do the same except the magnitude of the beta diversity measured was the inverse of that for the terminal metrics. This suggests that these two classes of metrics are complementary, rather than redundant, and may be utilized to differentiate patterns such as Scenario C versus D. The results in Table 1 also show the behavior of the metrics when the phylogeny becomes more ‘star-like’. In particular, each metric converged on a single value across all four scenarios when a star phylogeny was utilized. In other words the phylogenetic metric could not tell the scenarios apart because all species are equally related and every scenario demonstrates maximum phylogenetic turnover. This is intuitive as phylogenetic relatedness is equal between all species and no additional information regarding similarity can be gleaned from this phylogeny. This elucidates the fact that the phylogenetic beta diversity metrics when utilized on a star phylogeny are essentially the same as most species beta diversity metrics. For example, a presence-absence metric like PhyloSor will converge on a traditional Sorensen’s Index and an abundance-weighted metric like Dpw will converge on a traditional Bray-Curtis Distance when a star phylogeny is used. Thus nearly all information is lost when a star phylogeny is utilized and this is particularly so for metrics scales between zero and one such as UniFrac and PhyloSor. Metrics that are not scaled between zero and one do provide additional information above and beyond what can be gleaned from a traditional species beta diversity metric in that they still relay branch length information in the form of the distances from the root to the tips of the tree. Whether this information is actually useful for inferences regarding community structure and assembly is another question. Lastly, it should be noted that when the phylogeny was very ‘tippy’ (I.E. lambda = 0.25) signifying an early burst of speciation followed by stasis, the metrics were still able to differentiate between the scenarios. Thus in the unlikely scenario of a star phylogeny the present metrics of phylogenetic beta diversity may convey little additional useful information to the ecologist, but even in scenarios where there was a rapid radiation followed by little net diversification the metrics still can differentiate between the patterns of importance to the ecologist. The first goal of this study was to quantify the relationship between the phylogenetic beta diversity of tropical tree communities and their spatial distance or climatic difference. The results of the Mantel tests show that species and phylogenetic beta diversity was generally more correlated with differences in annual precipitation rather than changes in altitude or spatial distance (Table 2). When comparing the phylogenetic metrics, pairwise metrics Dpw , Dpw ’, Rao’s D, and Rao’s H generally had weaker correlations with annual precipitation differences than did PhyloSor , UniFrac , Dnn , and Dnn ’ (Table 2). These results were consistent whether randomly resolved or the original less well resolved phylogeny was utilized (Table 2). The second goal of this study was to examine the relationship between different patterns of trait evolution and the ability to predict functional beta diversity from patterns of phylogenetic beta diversity. The prediction was that a high degree of phylogenetic signal in trait data should strengthen the correlation between phylogenetic and functional beta diversity values. This prediction was supported when using the PhyloSor , UniFrac and nearest neighbor metrics where the stronger the phylogenetic signal in trait data (i.e. a higher K value) the stronger the correlation between the phylogenetic and functional beta diversity patterns (Figure 2). Conversely the pairwise and Rao metrics were less likely to accurately predict the pattern of functional beta diversity even when there was moderate to high phylogenetic signal in the trait data. A final goal of the present study was to examine the statistical relationships between the eight phylogenetic beta diversity metrics. This was done by calculating Pearson’s correlations between the outputs from all metrics and using a principal components analysis. The correlation analyses show ...
Context 2
... species may be unfeasible [2,5]. These studies have focused on quantifying whether the phylogenetic diversity in an assemblage is higher or lower than that expected given the observed species richness and a species pool. Using the assumption that phylogenetic relatedness is positively correlated with ecological or functional similarity, such studies have made inferences regarding the role of abiotic and biotic interactions in structuring communities. In recent years, community ecologists have expanded the above phylogenetic approach to include analyses of phylogenetic beta diversity or turnover between communities [11–15]. This approach is potentially powerful in that it can detect phylogenetically basal or terminal turnover between communities that traditional species-based metrics do not. For example, the turnover of con-geners along an environmental gradient would be considered low phylogenetic turnover, but high species turnover. These opposing patterns may have substantial consequences for how we understand the structure of communities [14]. Despite the power of this phylogenetic beta diversity approach, few empirical examples exist and to my knowledge there are no existing examples from diverse tropical ecosystems. In particular, we do not know whether spatial distance or environmental distance is more correlated with the phylogenetic turnover between communities in diverse systems like tropical tree communities. For example, ancient divergences in habitat preferences between clades and little divergence within clades should generate high levels of phylogenetic beta diversity along environmental gradients whereas recent large habitat shifts should provide the opposite pattern. Thus instead of simply knowing that species composition turns over along environmental gradients, phylogenetic metrics can begin to provide insights into how the evolution of habitat preferences or species function has influenced the observed distributional patterns. Although analyses of phylogenetic diversity within and between communities are potentially very powerful particularly in diverse ecosystems, they both critically rely on the assumption that phylogenetic relatedness is a sound proxy for functional or ecological similarity. This assumption is routinely questioned and examples where the assumption is violated are not difficult to find particularly when examining patterns of trait evolution [16–18]. Thus phylogenetic community ecologists are tasked with quantifying the phylogenetic signal in trait data rather than assume that it is there in order to make robust inferences [3,19]. In particular, if there is phylogenetic signal in trait data, then the patterns of phylogenetic diversity in a community or between communities should mirror the functional diversity [3]. Simulation-based studies that have examined alpha diversity have generally supported this expectation [20], but similar studies of phylogenetic beta diversity have not been conducted. In particular, it is not clear whether the functional beta diversity of communities can be predicted from the phylogenetic beta diversity when there is, or is not, phylogenetic signal in functional trait data. For example, how much phylogenetic signal is needed for a phylogenetic beta diversity metric to accurately recover the functional beta diversity of two communities? Another challenge for phylogenetic analyses of beta diversity is the rapid accumulation of metrics that may or may not be redundant. Thus it will become increasingly difficult to compare and contrast the results across different studies and to determine which metrics provide novel information over others. The present study utilizes a large tree inventory plot dataset from India to address the above outstanding challenges for investigations into the phylogenetic dissimilarity of communities. Specifically, here I ask: ( i ) is spatial or environmental distance more correlated with the phylogenetic beta diversity of tropical tree communities?; ( ii ) how much phylogenetic signal in trait data is needed for phylogenetic beta diversity metrics to reflect the functional beta diversity and how does this vary from metric to metric?; and ( iii ) are any of the eight phylogenetic beta diversity metrics used in this study redundant and which provide novel insights? The second and third questions are largely of a methodological nature, but answering these questions is critical for one to appropriately address the first question posed. That is, without addressing the statistical underpinnings and relationships of the large number of phylogenetic beta diversity metrics that are accumulating it is difficult, if not irresponsible, to address the biological questions of interest with these metrics. Thus, the work will primarily focus on the key methodological questions while trying to provide some biological insights along the way. Several metrics of phylogenetic beta diversity have been produced in recent years. In Figure 1 I present a simplified picture of different types of phylogenetic beta diversity or turnover where phylogenetic beta diversity is relatively ‘basal’ or ‘terminal’. In this hypothetical set of scenarios the species turnover between the communities being compared is complete or in other words species beta diversity is the maximum possible. In contrast the phylogenetic beta diversity is more variable. The present work seeks to analyze eight of the most commonly implemented metrics. There are undoubtedly alternative metrics that have been developed or that will be developed, but for the time being the manuscript will be constrained to the follow set of eight. The first metric I used is phylogenetic analog of Sorensen’s Index termed PhyloSor [15]: where BL k1k2 is the total length of the branches shared between community k 1 and k 2 , BL k1 and BL k2 are the total branch lengths found in communities k 1 and k 2 respectively. This metric may be considered a ‘basal’ metric upon initial inspection, but in reality most of the variability in values necessarily comes from the terminal aspects of the phylogeny unless communities turnover over almost entirely between very basal clades, but this is likely never occurring. The second metric used is a presence-absence weighted dissimilarity metric representing the unique fraction ( UniFrac ) of the phylogeny represented between two communities [12]: where n is the number of branches in the phylogeny, BL i is the length of branch l , k 1l and k 2l are the numbers of species descendent from branch l in communities k and k . Lastly k and k 2T are the total numbers of species in communities k 1 and k 2 respectively. Similar, to the PhyloSor metric this metric primarily will detect ‘terminal’ phylogenetic beta diversity. The third metric used is presence-absence weighted and calculates the mean nearest phylogenetic neighbor between two communities [21]: where min d ik 2 is the nearest phylogenetic neighbor to species i in community k 1 in community k 2 and min d ik 1 is the nearest phylogenetic neighbor to species j in community k 2 in community k 1 . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fourth metric is similar to the above nearest neighbor metric except that it is abundance weighted [21,21]: where f i and f j are the relative abundance of species i and species j . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fifth metric is a presence-absence weighted pairwise phylogenetic dissimilarity metric [21]: where d ik 2 is the mean pairwise phylogenetic distance between species i in community k 1 to all species in community k 2 and d ik 1 is the mean pairwise phylogenetic distance between species j in community k 2 to all species in community k 1 . This metric unlike those above is a ‘basal’ metric of phylogenetic beta diversity. The sixth metric is an abundance weighted version of the above pairwise phylogenetic dissimilarity [21,22]: where f i and f j are the relative abundance of species i and species j . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. The seventh metric is derived from Rao’s quadratic entropy [13,23]: where the variables are the same as those used the above nearest neighbor and pairwise metrics. This metric can be considered a ‘basal metric of phylogenetic beta diversity. The final metric standardizes Rao’s D based upon differences in alpha diversity between the two communities: where d k 1 is the mean pairwise phylogenetic distance between species in community k 1 and d jk 1 is the mean pairwise phylogenetic distance between species in community k 2 . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. In the above I describe the eight metrics as relatively ‘terminal’ or ‘basal’ metrics. To demonstrate this property I have calculated each of the presence-absence weighted metrics using the four simplified scenarios showed in Figure 1. I performed the calculations on the original tree in Figure 1 and on four transformed versions of that tree using a lambda transformation [24]. The last of these transformations generated a star phylogeny, which allowed for the comparison of the metrics when all species are equally related. The results in Table 1 provide initial insights into the similarity of some of the metrics and their ability to detect terminal versus basal phylogenetic turnover. In general the nearest neighbor metrics of Dnn, PhyloSor and UniFrac were able to detect terminal turnover in Scenarios C and D and contrast them with the basal turnover in Scenarios A and B. The pairwise metrics of Dpw, Rao’s D and Rao’s H were able to do the same except the magnitude of the beta diversity measured was the inverse of that for the terminal metrics. This suggests that these two classes of metrics are complementary, rather than redundant, and may be utilized to differentiate patterns such as Scenario C versus ...
Context 3
... data is needed for phylogenetic beta diversity metrics to reflect the functional beta diversity and how does this vary from metric to metric?; and ( iii ) are any of the eight phylogenetic beta diversity metrics used in this study redundant and which provide novel insights? The second and third questions are largely of a methodological nature, but answering these questions is critical for one to appropriately address the first question posed. That is, without addressing the statistical underpinnings and relationships of the large number of phylogenetic beta diversity metrics that are accumulating it is difficult, if not irresponsible, to address the biological questions of interest with these metrics. Thus, the work will primarily focus on the key methodological questions while trying to provide some biological insights along the way. Several metrics of phylogenetic beta diversity have been produced in recent years. In Figure 1 I present a simplified picture of different types of phylogenetic beta diversity or turnover where phylogenetic beta diversity is relatively ‘basal’ or ‘terminal’. In this hypothetical set of scenarios the species turnover between the communities being compared is complete or in other words species beta diversity is the maximum possible. In contrast the phylogenetic beta diversity is more variable. The present work seeks to analyze eight of the most commonly implemented metrics. There are undoubtedly alternative metrics that have been developed or that will be developed, but for the time being the manuscript will be constrained to the follow set of eight. The first metric I used is phylogenetic analog of Sorensen’s Index termed PhyloSor [15]: where BL k1k2 is the total length of the branches shared between community k 1 and k 2 , BL k1 and BL k2 are the total branch lengths found in communities k 1 and k 2 respectively. This metric may be considered a ‘basal’ metric upon initial inspection, but in reality most of the variability in values necessarily comes from the terminal aspects of the phylogeny unless communities turnover over almost entirely between very basal clades, but this is likely never occurring. The second metric used is a presence-absence weighted dissimilarity metric representing the unique fraction ( UniFrac ) of the phylogeny represented between two communities [12]: where n is the number of branches in the phylogeny, BL i is the length of branch l , k 1l and k 2l are the numbers of species descendent from branch l in communities k and k . Lastly k and k 2T are the total numbers of species in communities k 1 and k 2 respectively. Similar, to the PhyloSor metric this metric primarily will detect ‘terminal’ phylogenetic beta diversity. The third metric used is presence-absence weighted and calculates the mean nearest phylogenetic neighbor between two communities [21]: where min d ik 2 is the nearest phylogenetic neighbor to species i in community k 1 in community k 2 and min d ik 1 is the nearest phylogenetic neighbor to species j in community k 2 in community k 1 . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fourth metric is similar to the above nearest neighbor metric except that it is abundance weighted [21,21]: where f i and f j are the relative abundance of species i and species j . This metric like those above is a ‘terminal’ metric of phylogenetic beta diversity. The fifth metric is a presence-absence weighted pairwise phylogenetic dissimilarity metric [21]: where d ik 2 is the mean pairwise phylogenetic distance between species i in community k 1 to all species in community k 2 and d ik 1 is the mean pairwise phylogenetic distance between species j in community k 2 to all species in community k 1 . This metric unlike those above is a ‘basal’ metric of phylogenetic beta diversity. The sixth metric is an abundance weighted version of the above pairwise phylogenetic dissimilarity [21,22]: where f i and f j are the relative abundance of species i and species j . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. The seventh metric is derived from Rao’s quadratic entropy [13,23]: where the variables are the same as those used the above nearest neighbor and pairwise metrics. This metric can be considered a ‘basal metric of phylogenetic beta diversity. The final metric standardizes Rao’s D based upon differences in alpha diversity between the two communities: where d k 1 is the mean pairwise phylogenetic distance between species in community k 1 and d jk 1 is the mean pairwise phylogenetic distance between species in community k 2 . This metric can be considered a ‘basal’ metric of phylogenetic beta diversity. In the above I describe the eight metrics as relatively ‘terminal’ or ‘basal’ metrics. To demonstrate this property I have calculated each of the presence-absence weighted metrics using the four simplified scenarios showed in Figure 1. I performed the calculations on the original tree in Figure 1 and on four transformed versions of that tree using a lambda transformation [24]. The last of these transformations generated a star phylogeny, which allowed for the comparison of the metrics when all species are equally related. The results in Table 1 provide initial insights into the similarity of some of the metrics and their ability to detect terminal versus basal phylogenetic turnover. In general the nearest neighbor metrics of Dnn, PhyloSor and UniFrac were able to detect terminal turnover in Scenarios C and D and contrast them with the basal turnover in Scenarios A and B. The pairwise metrics of Dpw, Rao’s D and Rao’s H were able to do the same except the magnitude of the beta diversity measured was the inverse of that for the terminal metrics. This suggests that these two classes of metrics are complementary, rather than redundant, and may be utilized to differentiate patterns such as Scenario C versus D. The results in Table 1 also show the behavior of the metrics when the phylogeny becomes more ‘star-like’. In particular, each metric converged on a single value across all four scenarios when a star phylogeny was utilized. In other words the phylogenetic metric could not tell the scenarios apart because all species are equally related and every scenario demonstrates maximum phylogenetic turnover. This is intuitive as phylogenetic relatedness is equal between all species and no additional information regarding similarity can be gleaned from this phylogeny. This elucidates the fact that the phylogenetic beta diversity metrics when utilized on a star phylogeny are essentially the same as most species beta diversity metrics. For example, a presence-absence metric like PhyloSor will converge on a traditional Sorensen’s Index and an abundance-weighted metric like Dpw will converge on a traditional Bray-Curtis Distance when a star phylogeny is used. Thus nearly all information is lost when a star phylogeny is utilized and this is particularly so for metrics scales between zero and one such as UniFrac and PhyloSor. Metrics that are not scaled between zero and one do provide additional information above and beyond what can be gleaned from a traditional species beta diversity metric in that they still relay branch length information in the form of the distances from the root to the tips of the tree. Whether this information is actually useful for inferences regarding community structure and assembly is another question. Lastly, it should be noted that when the phylogeny was very ‘tippy’ (I.E. lambda = 0.25) signifying an early burst of speciation followed by stasis, the metrics were still able to differentiate between the scenarios. Thus in the unlikely scenario of a star phylogeny the present metrics of phylogenetic beta diversity may convey little additional useful information to the ecologist, but even in scenarios where there was a rapid radiation followed by little net diversification the metrics still can differentiate between the patterns of importance to the ecologist. The first goal of this study was to quantify the relationship between the phylogenetic beta diversity of tropical tree communities and their spatial distance or climatic difference. The results of the Mantel tests show that species and phylogenetic beta diversity was generally more correlated with differences in annual precipitation rather than changes in altitude or spatial distance (Table 2). When comparing the phylogenetic metrics, pairwise metrics Dpw , Dpw ’, Rao’s D, and Rao’s H generally had weaker correlations with annual precipitation differences than did PhyloSor , UniFrac , Dnn , and Dnn ’ (Table 2). These results were consistent whether randomly resolved or the original less well resolved phylogeny was utilized (Table 2). The second goal of this study was to examine the relationship between different patterns of trait evolution and the ability to predict functional beta diversity from patterns of phylogenetic beta diversity. The prediction was that a high degree of phylogenetic signal in trait data should strengthen the correlation between phylogenetic and functional beta diversity values. This prediction was supported when using the PhyloSor , UniFrac and nearest neighbor metrics where the stronger the phylogenetic signal in trait data (i.e. a higher K value) the stronger the correlation between the phylogenetic and functional beta diversity patterns (Figure 2). Conversely the pairwise and Rao metrics were less likely to accurately predict the pattern of functional beta diversity even when there was moderate to high phylogenetic signal in the trait data. A final goal of the present study was to examine the statistical relationships between the eight phylogenetic beta diversity metrics. This was done by calculating Pearson’s correlations between the outputs from all metrics and using a principal components analysis. The correlation analyses show strong correlations between many pairs of metrics with some ...
Citations
... This difference is expected due to changes in the landscape, whether natural (i.e., different river basins) or generated by anthropogenic activity. Phylogenetic dissimilarity is strongly correlated with environmental heterogeneity (Swenson, 2011;Swenson et al., 2011), contributing to the findings of our study. ...
Amazon river floodplains sustain distinct kinds of seasonally flooded habitats along with their specialized biota. River sediment load and geomorphology determine vegetation physiognomy and the occurrence of specialized taxa. Hydropower dams disrupt the annual flooding cycle, affecting the floodplain habitats and associated avifauna. Our goal was to understand how permanent flooding caused by large dams affects taxonomic, functional, and phylogenetic diversity of bird communities in two Amazonian rivers with distinct characteristics. We sampled 35 sites, including undisturbed sites and sites impacted by the permanent flooding caused by large dams operating on the Madeira and Xingu rivers since 2012 and 2016, respectively. We recorded 202 bird species through passive acoustic monitoring. We did not find differences in mean species richness between rivers or between impacted vs. undisturbed sites. However, we found species turnover between distinct river basins and between sites. Undisturbed sites were characterized by floodplain specialists and were more phylogenetically clustered and functionally more similar than impacted sites, which were occupied by more generalist species from different phylogenetic clades, with distinct functional traits. By assessing multiple dimensions of bird diversity, we show that permanent flooding leads to extensive changes in the floodplain avifauna, favoring generalist over specialist species and thus reducing the uniqueness of the affected communities. We demonstrate that alpha diversity metrics alone are insufficient to characterize the impacts of river damming on Amazonian floodplain avifauna. To properly monitor dams' environmental impacts, additional measures of community change should be standardized, with emphasis on species replacements in seasonally flooded environments. Abstract in Portuguese is available with online material.
... We calculated the standardised effect size of phylogenetic diversity sensu stricto (hereafter ses.PD). This metric represents the deviation, in branch length, of the phylogeny of all species (tips) occurring in each patch (Faith 1992), from branch length obtained from random shuffling phylogeny tree permuted 999 times (Swenson 2011). This deviation was standardised and then expressed in standard deviations by using the following equation: ...
Context. Most aquatic macrophytes are ecozone-endemic species, and approximately two-thirds of them have rare occurrence at global scale. These small-range plants are seriously under-studied at macroecological scale, despite their marked vulnerability to extinction through habitat loss and climate change. Aims. To identify global hotspots of endemism and rarity of aquatic macrophytes and examine the factors that resulted in speciation hotspots of macrophytes in some areas of the planet. Methods. We analysed a database of 3499 macrophyte species to locate speciation hotspots and assess the biogeographic and environmental drivers that maintain ecozone-endemic, and globally rare species within their current limited global areas of occupancy. Key results. Ecozone-endemic and globally rare macrophyte species hotspots across the planet showed similar occurrence patterns and drivers among ecozones. Ecozone environmental conditions, particularly harsh environments, influenced macrophyte phylogenetic diversity and structure. Most macrophyte species diversification is recent (<10 million years ago). A negative association with bird-mediated zoochory was seen for endemicity and rarity hotspots. Conclusions. This study identified hotspots of endemicity and rarity, and potential cradle and museum speciation areas. Implications. Our findings could inform global action to conserve the macrophyte diversity of wetlands, and other inland aquatic habitats, across the world.
... Furthermore, the relationship between beta diversity and environmental factors may indicate specific drivers of differences in community composition (Dobrovolski et al., 2012). Ultimately, the integration of the patterns of beta diversity with phylogenetic and functional information of their species allows the investigation of how evolutionary and ecological factors interact to influence compositional differences among communities, such as geographic isolation, phylogenetic restrictions, biotic interactions, geographic distance, and environmental filtering (Graham & Fine, 2008;Magurran & McGill, 2011;Siefert et al., 2013;Swenson, 2011;Yang et al., 2015). ...
Aim
Patterns of beta diversity reflect the formation dynamics of ecological communities. Here, we integrated geographic, phylogenetic, and phenotypic information of coastal woody vegetation to investigate (1) whether the observed dissimilarity between assemblages differs from that expected by chance, examining the roles of spatial and deterministic processes; (2) the relative contribution of beta‐diversity components (turnover and nestedness) for taxonomic, phylogenetic, and functional beta diversity; and (3) what environmental factors drive the differences in composition between assemblages for all these dimensions.
Location
Brazil.
Taxon
Angiosperm trees.
Methods
We built dissimilarity matrices and partitioned the taxonomic, phylogenetic, and functional beta diversity from an incidence matrix, a phylogeny including the region's plants, and a matrix expressing functional distances. Using linear regressions, we tested the effects of different environmental predictors representative of the effects of water availability, thermal energy, habitat heterogeneity, edaphic constraints, climatic stability, and human influence on beta‐diversity patterns.
Results
Taxonomic, phylogenetic, and functional dissimilarities exhibited a typical pattern of greater dissimilarity with distance (i.e., as expected by chance). However, these patterns showed different contributions of beta‐diversity components, predominating turnover in taxonomic and phylogenetic dissimilarity, and nestedness in functional dissimilarity. Water availability had a slight effect on patterns of taxonomic and phylogenetic dissimilarities.
Main conclusions
The Brazilian coastal woody vegetation appears to have emerged through a dynamic of colonisation of evolutionarily distinct but functionally similar lineages that originated from adjacent phytogeographic domains, proportional to their diversity. This is consistent with a combination of both neutral and non‐neutral processes. Our findings underscore the complementary roles of different dimensions of beta diversity in explaining the dynamics of these vegetation communities.
... Phylogenetic diversity is a useful metric for the identification of areas of conservation priority (Faith, 2021;González-Orozco and Parra-Quijano, 2023), because it is a robust biodiversity statistical tool that takes into account the species relationships rather than the traditional species richness (Martin, 2002;Swenson, 2011;González-Orozco and Parra Quijano, 2023). Furthermore, PD has been identified as one of the major indicators of 'maintenance options' in conservation strategies (Laity et al., 2015;Perino et al., 2021), with the International Union for the Conservation of Nature (IUCN) being charged with implementing the use of PD in conservation planning (Gumbs et al., 2021). ...
Background and aims
The quartz fields of the Greater Cape Floristic Region (GCFR) are arid and island-like special habitats, hosting about 142 habitat-specialized plant species of which 81% are local endemics, characterized by a rapid turnover of species between and among sites. We use several phylogenetic community metrics to i) examine species diversity and phylogenetic structure within and among quartz fields; ii) investigate whether quartz field specialists are evolutionarily drawn from local species pools, while the alternative hypothesis posits that there is no significant evolutionary connection between quartz field specialists and the local species pools; and iii) determine whether there is an association between certain traits and the presence of species in quartz fields.
Methods
We sampled and developed dated phylogenies for six species-rich angiosperm families (Aizoaceae, Asteraceae, Crassulaceae, Cyperaceae, Fabaceae and Santalaceae) represented in the quartz field floras of southern Africa. Specifically, we focused on the flora of three quartz field regions in South Africa (Knersvlakte, Little Karoo and Overberg) and their surrounding species pools to address our research questions, scoring traits associated with harsh environments.
Key results
We found that the Overberg and Little Karoo had the highest level of species overlap for families Aizoaceae and Fabaceae, while the Knersvlakte and the Overberg had the highest species overlap for families Asteraceae, Crassulaceae, and Santalaceae. Although our phylogenetic community structure and trait analyses showed no clear patterns, relatively low pairwise phylogenetic distances between specialists and their local species pools for Aizoaceae and Fabaceae suggest that quartz species could be evolutionarily drawn from their surrounding areas, with phylogenetic overdispersion in Knersvlakte and Little Karoo for Aizoaceae and Crassulaceae.
Conclusions
Despite their proximity to one another within the GCFR, the studied areas differ in their species pools and the phylogenetic structure of their specialists. Our work provides further justification for increased conservation focus of these unique habitats under future scenarios of global change.
... Third, whether functional and phylogenetic structure show consistent patterns depends on the extent to which traits are conserved phylogenetically (e.g. Blomberg's K greater than two) (Swenson, 2011). ...
Aim
Understanding the patterns and drivers of biodiversity across space and time is commonly based on species diversity, which may ignore species' functional role and evolutionary history and result in an incomplete understanding of community assembly. It is suggested that integrating species, functional, and phylogenetic diversity could provide a more holistic assessment of community assembly in natural ecosystems. This study aimed to explore the elevational patterns and environmental drivers of multiple facets of fish diversity and community structure in a subtropical river during the wet and dry seasons.
Location
The Chishui River basin, China.
Methods
We investigated the responses of fish species richness, functional richness, and phylogenetic diversity to elevation in different seasons. Moreover, we compared functional dispersion and mean pairwise distance with those obtained from null models to infer assembly mechanisms shaping community structure. Additionally, we examined the environmental drivers (e.g. water chemistry, temperature, and river size) of fish diversity and community structure.
Results
Fish species richness, functional richness, and phylogenetic diversity showed a negative relationship with elevation in the Chishui River basin. Fish communities tended to be on average functionally random but phylogenetically clustered. Furthermore, phylogenetic structure exhibited a decreasing pattern along the elevational gradient. Despite no significant seasonal changes for fish diversity (except for phylogenetic diversity), fish communities became more phylogenetically overdispersed and clustered at low and high elevations in the dry season. Additionally, the responses of fish diversity and community structure to environmental variables were not synchronous.
Conclusions
At the basin scale, environmental filtering was prevalent in shaping fish phylogenetic structure, whereas stochasticity was likely more important for functional structure. Moreover, the ecological mechanisms shaping individual fish communities switched from limiting similarity to environmental filtering as elevation increased, and the underlying forces at two ends of the elevational gradient became more prominent in the dry season.
... For functional and phylogenetic β-diversity, we used the mean pairwise functional or phylogenetic distance (FD_Dpw or PD_Dpw) between two communities and the nearest neighbor trait or phylogenetic distance (FD_Dnn or PD_Dnn) between two communities (Swenson, 2011;Wang et al., 2023). FD/PD_Dpw and FD/PD_Dnn represent the two main mathematically independent classes (basal and terminal metrics) of trait or phylogeny metrics, with FD_Dpw or PD_Dpw assessing deeper turnover in the trait or phylogeny, and FD_Dnn or PD_Dnn being sensitive to turnover near the tips of trees (Jin et al., 2015). ...
Understanding long‐term changes in fish diversity and community assembly rules is crucial for freshwater conservation. Growing evidence indicates that studying functional and phylogenetic diversity beyond purely taxonomic considerations can provide different but complementary information on community assembly. Here, the taxonomic, functional, and phylogenetic β‐diversity of fish communities, as well as the community assembly mechanisms, were explored in five impounded lakes of the China's South‐to‐North Water Diversion Project (SNWDP) from the 1980s to the 2010s. We found that (1) there was an obvious trend of species homogenization in the five impounded lakes, but the long‐term transformations of different dimensional β‐diversity were divergent; (2) water quality and land use variables have greater impacts on multidimensional β‐diversity; and (3) community assembly process in taxonomic and functional dimensions were dominated by random process in both periods, while shifting from limiting similarity to habitat filtering in the phylogenetic dimension. These results highlight that functional and phylogenetic diversity are important additional ecological indices for assessing the patterns of fish diversity in lakes.
... For both phylogeny and functional traits, we used the mean pairwise (D pw ) and the mean nearest neighbor distance (D nn ) metrics to measure the phylogenetic and functional dissimilarity of each pairwise root neighborhood. We used both these two metrics, as D pw represents the "basal" metrics of phylogenetic or functional diversity, reflecting the differentiation of habitat preferences among species, whereas D nn represents the "terminal" metrics of phylogenetic or functional diversity, reflecting the differentiation in the utilization of specific resources among species (Swenson, 2011). Particularly, phylogenetic β diversity for example between pairwise root neighborhoods (e.g., A and B) was determined as the sum of phylogenetic distances between all intercommunity pairwise species among A and B divided by the total number of intercommunity species pairs (Yang et al., 2015). ...
Root‐centric studies have revealed fast taxonomic turnover across root neighborhoods, but how such turnover is accompanied by changes in species functions and phylogeny (i.e., β diversity) remains largely unknown. As β diversity can reflect the degree of community‐wide biotic homogenization, such information is crucial for better inference of below‐ground assembly rules, community structuring, and ecosystem processes. We collected 2480 root segments from 625 0–30 cm soil profiles in a subtropical forest in China. Root segments were identified into 138 species with DNA‐barcoding with six root morphological and architectural traits measured per species. By using the mean pairwise (Dpw) and mean nearest neighbor distance (Dnn) to quantify species ecological differences, we first tested the non‐random functional and phylogenetic turnover of root neighborhoods that would lend more support to deterministic over stochastic community assembly processes. Additionally, we examined the distance‐decay pattern of β diversity, and finally partitioned β diversity into geographical and environmental components to infer their potential drivers of environmental filtering, dispersal limitation, and biotic interactions. We found that functional turnover was often lower than expected given the taxonomic turnover, whereas phylogenetic turnover was often higher than expected. Phylogenetic Dpw (e.g., interfamily species) turnover exhibited a distance‐decay pattern, likely reflecting limited dispersal or abiotic filtering that leads to the spatial aggregation of specific plant lineages. Conversely, both functional and phylogenetic Dnn (e.g., intrageneric species) exhibited an inverted distance‐decay pattern, likely reflecting strong biotic interactions among spatially and phylogenetically close species leading to phylogenetic and functional divergence. While the spatial distance was generally a better predictor of β diversity than environmental distance, the joint effect of environmental and spatial distance usually overrode their respective pure effects. These findings suggest that root neighborhood functional homogeneity may somewhat increase forest resilience after disturbance by exhibiting an insurance effect. Likewise, root neighborhood phylogenetic heterogeneity may enhance plant fitness by hindering the transmission of host‐specific pathogens through root networks or by promoting interspecific niche complementarity not captured by species functions. Our study highlights the potential role of root‐centric β diversity in mediating community structures and functions largely ignored in previous studies.
... Species and functional diversity are, therefore, not interchangeable. Indeed, one could imagine two communities comprised of entirely different species (high species beta diversity) that carry out the same functions (low functional beta diversity; Swenson, 2011;Swenson et al., 2012;Siefert et al., 2013). Because functional diversity quantifies ecological impacts and species interactions, studies including the functional diversity of organisms provide more robust data for conservation and restoration than studies on species presence and abundance alone (Cadotte et al., 2009). ...
Arthropod diversity is often linked to variation in resource use, dispersal ability, habitat connectivity, and climate factors that differ across spatial scales. The aim of this research was to examine how species richness, functional diversity, and community composition of two taxa differing in functional roles and dispersal ability are structured across spatial scales and to identify the importance of vegetation, climate, and landscape in explaining these patterns at different scales. Organisms were collected from tree canopies using insecticidal fogging in the summer of 2000 from 96 trees in 24 stands of 6 deciduous forest sites in 2 ecoregions of the eastern United States. Taxonomic and functional beta diversity of ants (Hymenoptera: Formicidae) and spiders (Araneae) were partitioned across four hierarchical spatial scales (individual tree, forest stand, site, and ecoregion). The contributions of climatic, landscape, and vegetation variables were determined using model selection. Ant and spider species richness, functional diversity, and community composition differed between taxa and across spatial scales. Alpha diversity (within trees) was lower than expected for both taxa and types of diversity, with host tree species supporting different species of ants and spiders. While the beta components of species diversity among trees and forest stands were greater than expected for both taxa, spiders also showed significant levels of beta diversity among sites. Functional beta diversity was less scale‐dependent than taxonomic beta diversity. Stand‐level patterns of beta diversity were significantly predicted by variation in climate and landscape connectivity. The effects of climate and landscape fragmentation on the diversity and community structure of both taxa indicate that anthropogenic climate change and land use change will alter canopy arthropod communities. Results also suggest that patterns of diversity among fragmentation metrics are influenced by differences in dispersal ability.
... Given that biological communities result from a complexity of interactions between organisms, environment and space, the use of an integrative approach where the different dimensions of diversity are taken into account has been widely encouraged (Ouchi-Melo et al., 2018). While taxonomic diversity measures treat all species as functionally equivalent and evolutionarily independent (Swenson 2011), phylogenetic diversity considers shared ancestry relationships between species (Magurran, 2004), thus reflecting the evolutionary history of communities. Indeed, phylogenetic diversity has been progressively recognized as an essential dimension for biodiversity conservation (Pollock et al., 2017). ...
Fire is a natural disturbance that has shaped Earth's biodiversity for millions of years. In a fire-prone ecosystems, fire acts as an important environmental filter, selecting species presenting tolerant traits to fire events and post-fire environmental conditions. It is expected that intense fire activity selects for closely-related species, thus promoting a phylogenetic and functional clustering of communities. In view of the severe changes in natural fire regimes observed in different regions of the planet, an increasing body of research has been dedicated to exploring their impact on frogs, a particularly susceptible group in the context of wildfires. However, it is notable that most of the research focuses on the traditional dimension of diversity, the taxonomic diversity, leaving a significant gap in our understanding of how fire disturbances affect the phylogenetic diversity of these communities. Here, we tested the effects of three fire regime parameters (i.e., total burned area, time since the last fire and fire count at the landscape scale) on the taxonomic and phylogenetic diversity of frog communities in 26 sites within a fire-prone Brazilian protected area. We used Hill numbers to characterize the taxonomic (Species richness, Shannon and Simpson's diversity) and phylogenetic diversity (Phylogenetic richness, Mean phyloge-netic diversity of common lineages and Mean phylogenetic diversity of dominant lineages). We found that the fire regime did not explain patterns of the taxonomic diversity. Nonetheless, there was a positive correlation between phylogenetic richness and the frequency of fire occurrences, while a slight negative correlation was observed with the percentage of burned area. In addition, moderate fire activity seems to be an important driver of phylogenetic diversity. Therefore, management practices toward a mosaic of areas with different fire histories are fundamental in this protected area. We finally emphasize that all diversity facets of anurans should be assessed and considered in management decisions to guaranteeing anuran conservation in this region.
... The need for explicitly accounting for between-species relatedness generated a wave of methodological improvements that introduced new methods for calculating diversity. Next to a lively scientific discussion on how functional alpha diversity can be appropriately quantified (Mason et al. 2005, Petchey and Gaston 2006, Villéger et al. 2008, Mouchet et al. 2010, some authors developed on the concept of functional beta diversity, too (Swenson 2011, Botta-Dukát 2018, Chao et al. 2019). Among them, various indices for calculating pairwise functional dissimilarity between communities have been proposed (Ricotta and Burrascano 2008, Cardoso et al. 2014, Ricotta and Pavoine 2015. ...
... It is also notable that Swenson et al. (2011) and Swenson (2011) use the quantity Q(p, q) as a standalone index of pairwise beta diversity and call it D pw or 'Rao's D'. The latter name is misleading since Rao (1982) himself noted with D ij the DISC (or D Q ) index. ...
One of the effective tools to study the variation between communities is the use of pairwise dissimilarity indices. Besides species as variables, the involvement of trait information provides valuable insight into the functioning of ecosystems. In recent years, a variety of indices have been proposed to quantify functional dissimilarity between communities. These indices follow different approaches to account for between‐species similarities in calculating community dissimilarity, yet they all have been proposed as straightforward tools. In this paper, we review the trait‐based dissimilarity indices available in the literature and identify the most important conceptual and technical properties that differentiate among them, and that must be considered before their application. We identify two primary aspects that need to be considered before choosing a functional dissimilarity index. The first one is the way communities are represented in the trait space. The three main types of representations are the typical values, the discrete sets using the combination of species × sites and species × traits matrices, and the hypervolumes. The second decision is the concept of dissimilarity to follow, including two options: distances and disagreements. We use the above scheme to discuss the available functional dissimilarity indices and evaluate their relations to each other, their capabilities, and accessibility.
