Figures
Abstract
Amazonian rainforests sustain some of the richest tree communities on Earth, but their ecological and evolutionary responses to human threats remain poorly known. We used one of the largest experimental datasets currently available on tree dynamics in fragmented tropical forests and a recent phylogeny of angiosperms to test whether tree communities have lost phylogenetic diversity since their isolation about two decades previously. Our findings revealed an overall trend toward phylogenetic impoverishment across the experimentally fragmented landscape, irrespective of whether tree communities were in 1-ha, 10-ha, or 100-ha forest fragments, near forest edges, or in continuous forest. The magnitude of the phylogenetic diversity loss was low (<2% relative to before-fragmentation values) but widespread throughout the study landscape, occurring in 32 of 40 1-ha plots. Consistent with this loss in phylogenetic diversity, we observed a significant decrease of 50% in phylogenetic dispersion since forest isolation, irrespective of plot location. Analyses based on tree genera that have significantly increased (28 genera) or declined (31 genera) in abundance and basal area in the landscape revealed that increasing genera are more phylogenetically related than decreasing ones. Also, the loss of phylogenetic diversity was greater in tree communities where increasing genera proliferated and decreasing genera reduced their importance values, suggesting that this taxonomic replacement is partially underlying the phylogenetic impoverishment at the landscape scale. This finding has clear implications for the current debate about the role human-modified landscapes play in sustaining biodiversity persistence and key ecosystem services, such as carbon storage. Although the generalization of our findings to other fragmented tropical forests is uncertain, it could negatively affect ecosystem productivity and stability and have broader impacts on coevolved organisms.
Citation: Santos BA, Tabarelli M, Melo FPL, Camargo JLC, Andrade A, Laurance SG, et al. (2014) Phylogenetic Impoverishment of Amazonian Tree Communities in an Experimentally Fragmented Forest Landscape. PLoS ONE 9(11): e113109. https://doi.org/10.1371/journal.pone.0113109
Editor: Xiao-Yong Chen, East China Normal University, China
Received: July 9, 2014; Accepted: October 17, 2014; Published: November 19, 2014
Copyright: © 2014 Santos et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. All relevant data are within the paper and its Supporting Information files.
Funding: The Instituto Nacional de Pesquisas da Amazônia (INPA), Smithsonian Institution, US National Science Foundation, Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), NASA-LBA program, US-AID, Mellon Foundation, Blue Moon Fund, and Marisla Foundation generously supported the Biological Dynamics of Forest Fragment Project. BAS, FPLM and MT are thankful to CNPq for postdoctoral fellowship and research grants. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
The current biodiversity crisis driven by human-induced species loss is likely to drastically alter the tree of life [1], [2], but this is still poorly understood due to the historical gap between ecology and phylogenetics [3]. Fortunately, this gap is shrinking as evolutionary relationships among species can be incorporated into ecological studies [4]–[6]. Phylogenetic information, besides providing an additional measure of biodiversity beyond traditional measures of species richness and diversity, is useful to infer community assembly processes and ecosystem stability [7]–[9].
In the context of conservation, community phylogenetics is a powerful tool to understand how biological communities respond to human disturbances of varying type, intensity, and frequency [10]. Important advances in this field are helping us to understand the effects of fires on community assembly [11]–[13], the processes involved in community organization during forest regeneration [14], [15], and the outcome of species loss and gain on community phylogenetic diversity [16], [17]. However, we are still far from drawing general conclusions about the organization and persistence of biodiversity in our rapidly changing world [18].
Several studies have revealed shifts in the taxonomic and functional profile of tropical tree communities in response to habitat loss and fragmentation [19]–[22]. Compared to continuous forest, tree communities near forest edges and in small forest fragments (hereafter edge-affected habitats), may support only a small fraction of (1) emergent and understory tree species; (2) heavy-wooded, shade-tolerant and slow-growing trees; (3) large-seeded trees dispersed by medium to large-bodied frugivores; and (4) tree species pollinated by specialized vectors and those bearing supra-annual reproduction (see Tabarelli et al. [23] and references therein). Long-term tree monitoring in Central Amazonia reveals that edge-affected habitats are hyperdynamic systems characterized by increased rates of tree recruitment and mortality, especially in the first few decades following edge creation [24]. This hyperdynamism is guided by the remarkable recruitment of fast-growing, early-to-mid successional trees that replace slow-growing late-successional ones [19], [25], [26], which are most probably killed by increased wind turbulence or desiccation near fragment edges [27], [28]. Similarly, in the Atlantic forest of northeast Brazil, a small number of native, fast-growing successional species proliferate from forest edges to entire landscapes [21], [29], leading to a sort of regional biotic homogenization [23], [30].
However, important questions remain before we can properly predict the destiny of tropical tree communities in human-modified landscapes. Two of them are addressed here: (1) how much evolutionary diversity is being lost in edge-affected habitats along with the taxonomic and functional impoverishment? (2) what is the role of human-induced biotic homogenization on the phylogenetic diversity of remaining communities? A recent study on secondary forests of Costa Rica [31] demonstrated that disturbance-favored early-successional tree species are more phylogenetically related than expected by chance, whereas late-successional species tend to be less related than random expectations drawn from the same regional phylogeny. If these trends hold for Amazonian tree species, we could expect an increase in phylogenetic clustering or a decrease in phylogenetic evenness in the edge-affected habitats, where fast-growing trees are thriving and replacing disturbance-sensitive ones.
In this paper we used one of the largest datasets currently available on tree dynamics in a 1000-km2 experimentally fragmented Amazonian forest landscape and a recent phylogeny of angiosperms to assess whether tree communities have lost phylogenetic diversity and become phylogenetically clustered since their isolation two decades previously. First, we examined the effects of forest site (1-ha, 10-ha, and 100-ha fragments and continuous forest), time since forest isolation, and distance to nearest forest edge on richness of tree genera, phylogenetic community diversity and structure. Then, we evaluated potential relationships between changes in phylogenetic diversity and the proliferation or declining of 59 tree genera that have significantly changed in importance values since forest isolation. Finally, we estimated the phylogenetic distribution of declining and increasing tree genera and discuss the implications of our results for the ecology and conservation of tropical tree communities in human-modified landscapes.
Methods
Study landscape
The Biological Dynamics of Forest Fragments Project is a 1,000-km2 experimental landscape in Central Amazonia. Within this landscape, nine forest fragments ranging from 1 to 100 ha in area were isolated from nearby intact forest during the early 1980s by clearing and burning the intervening vegetation to create cattle pastures. Some of the pastures have been abandoned and now support secondary forests. Detailed descriptions of the project, including its study design, fragment histories, the matrices of modified vegetation surrounding fragments, and the methods used for censusing and identifying trees are provided elsewhere [24].
Before fragment isolation, permanent 1-ha forest plots were established within each fragment and in eight comparable sites in nearby intact forest. The present study incorporates tree demography data from 40 1-ha plots distributed throughout the entire landscape, either along forest edges (plot center <100 m from the nearest edge) or in ‘intact’ forest interiors (170–3,000 m from the edge). After an initial, exhaustive inventory of tree communities, each plot was resampled after fragmentation at typical intervals of 4–6 years to assess tree mortality, growth, and the recruitment of new trees. Altogether, the fates of nearly 32,000 trees (dbh>10 cm) were followed for periods of up to 18 years (mean 14.7 years) [19]. All the tree specimens from the different censuses were identified together by collecting a voucher specimen for virtually every single tree. In the herbarium, all the voucher specimens were pooled into families (the initial family identification was made in the field) and then each specimen was identified to species or genus/morphospecies level.
In this study we used the same genus-level dataset presented in Laurance et al. [19], [32], which were mainly interested in describing broader taxonomic and ecological responses of tree communities to human disturbance in 40 1-ha BDFFP plots. We are aware that examining phylogenetic community structure at the genus level has shortcomings when compared with species-level analyses, as we cannot assume that shifts in tree genera are a proxy for shifts in tree species. However, in the absence of robust species-level phylogeny, the genus-level approach provides reliable information about lineages and can shed light on the structure of hyperdiverse communities such as ours. This approach has been used in other studies, including ant [33] and bacteria communities [34], and contributed considerably for the development of phylogenetic community ecology.
Shifts in phylogenetic community structure and diversity
To evaluate shifts in tree communities we first listed all genera observed in the 40 1-ha plots. Altogether, 267 genera belonging to 62 families were recorded. This genera list was then assembled into a phylogeny (i.e. the regional phylogeny) using the phylomatic function of Phylocom 4.2 [35] and the maximally resolved supertree of angiosperms R20100701, available for free in the software. The regional phylogeny had branch lengths estimated with the bladj algorithm. For this, we used the node ages provided by Bell et al. [36] and further corrected for inconsistencies in syntax and nomenclature of internal nodes and the regional phylogeny following procedures described in Gastauer and Meira-Neto [37]. The resulting time-calibrated regional phylogeny was used in the subsequent analyses (see Appendix S1).
To examine shifts in phylogenetic community structure and diversity within the landscape, we calculated two intra-sample phylogenetic metrics for each community: mean phylogenetic distance (MPD) and net related index (NRI) (see Webb et al. [35] for a complete description of the metrics and Table S1 for raw data of tree plots). We used the argument ‘-a’ of comstruct function in Phylocom 4.2 to take into account the importance value of each taxon, estimated by the combination of basal area and relative abundance as described in Laurance et al. [19]. To be conservative, we avoided using metrics more sensitive to phylogenetic resolution at the tips of the phylogeny, such as MNTD and NTI, because the evolutionary relationships of many tropical genera within families are not resolved yet.
We treated MPD as the metric of phylogenetic diversity and NRI as the metric of phylogenetic structure. MPD was expressed in million years and represented the observed phylogenetic diversity of a given community. NRI, the measure of ‘standarzided effect size’ of phylogenetic community structure, was calculated comparing the observed MPD to 999 null communities generated by null model 0 of Phylocom 4.2. This model shuffles species labels across the entire phylogeny [35] and is more appropriate for temporal analyses than other null models (e.g. Phylocom model 2) because it maintains plot abundance distribution, plot species richness, occupancy rates, and conserves the spatial contagion of species (see Norden et al. [31], p.S74, for more details on the use of this null model for temporal analyses).
We then made before vs. after comparisons to assess whether average levels of tree-wide phylogenetic structure changed over time across forest sites. Positive trends of NRI indicate decrease in phylogenetic evenness or increase in phylogenetic clustering since forest isolation, whereas negative trends indicate the opposite. Because communities with similar values of NRI can differ in terms of absolute evolutionary diversity, we made similar comparisons using the observed values of MPD to estimate potential shifts in community phylogenetic diversity.
Increasing and decreasing tree genera
Two previous studies [19], [32] identified 28 tree genera that are significantly increasing and 31 that are significantly decreasing in importance value throughout the landscape, totaling 59 genera that respond positively or negatively to human disturbance in demographic terms (see Table S2). The bulk of increasing genera is composed by fast-growing, small-seeded species typical of early to mid successional stages; 15 are thriving in edge-affected habitats [19] and 13 in the continuous forest [32]. The group of decreasing genera is mostly represented by slow-growing, large-seeded species typical of old-growth forests; 18 are declining in edge-affected habitats, six are reducing in the continuous forests and seven are declining in both habitats [19], [32].
To assess the relationship between the shifts in phylogenetic diversity and the proliferation or declining of these particular genera, we calculated for each tree community the relative change in MPD and mean importance value of the 59 tree genera that have significantly responded to human disturbance. The relative changes in MPD and mean importance value of genera were calculated as the difference between the final and the initial value, divided by the initial value. Negative relative changes in MPD indicate that tree community loses phylogenetic diversity over time, while positive changes indicate the opposite. If shifts in phylogenetic diversity were influenced by the proliferation of increasing genera, we would expect positive relationship between the gain of increasing genera and the loss of phylogenetic diversity. A similar positive relationship could arise if the negative shifts in MPD were affected by the loss of decreasing genera.
Phylogenetic distribution of increasing and decreasing genera
To assess the phylogenetic distribution of declining and increasing tree genera, we estimated the degree of phylogenetic relationships for each group using NRI, as adopted elsewhere [31]. Both groups of genera, one containing 28 increasing genera and other 31 decreasing genera, were assumed to be a community drawn from the 267-genera regional pool, resulting in a value of NRI for group. According to the findings of Norden et al. [31], we expected that the group of increasing genera was more phylogenetically related than the group of decreasing ones (i.e. NRIincreasing>NRIdecreasing).
Statistical analyses
We used linear mixed models to test for the effects of forest site, time since forest isolation and distance to the nearest forest edge on the richness of tree genera and phylogenetic community metrics. We set forest site (four levels: 1-ha [n = 4], 10-ha [n = 9], 100-ha forest fragments [n = 10] and continuous forest [n = 17]), time since forest isolation (first and last census), edge distance and the interaction between forest site×time as fixed effects. We included this interaction term to assess whether the magnitude of the shifts in phylogenetic metrics was greater in the smaller forest fragments (1-ha and 10-ha), where tree communities have changed suddenly since forest isolation [19]. Forest plot (N = 40), the subject in which tree censuses were done, was set as the random effect. We adopted the residual maximum likelihood method to separate the variance of fixed and random effects [38]. We used Pearson product-moment correlations to assess the relationship between the relative changes in MPD and mean importance value of decreasing and increasing genera. All analyses were performed in JMP 8 (SAS Institute Inc.).
Results
Genera richness decreased significantly from 116.2±1.2 genera/ha (mean ± SE) before fragmentation to 113.5±1.4 genera/ha after fragmentation (time effect in Table 1), but this reduction occurred regardless of whether plot was in forest fragments or continuous forest (interaction term not significant; Table 1). We observed a similar trend for MPD, which decreased slightly (<2%) but significantly over time and irrespective to plot location (Table 1; Fig. 1). Surprisingly, shifts in phylogenetic structure and diversity were not related with plot distance to the nearest edge (Table 1).
Initial tree census refers to the first tree inventory carried out before forest isolation in early 1980's; final tree census refers to the last census available in our dataset (on average 14 years after the first census, see Methods). Red lines indicate loss of tree genera and phylogenetic diversity over time.
Thirty-two out of 40 communities (80%) experienced a decrease in MPD over the period of tree monitoring, as indicated by negative rates of change in MPD (Fig. 2). In these impoverishing tree communities, the loss of phylogenetic diversity was positively correlated with the increase in importance value of increasing genera (r = 0.38; P<0.05) and marginally correlated with the reduction in importance value of decreasing genera (r = 0.34; P = 0.059), suggesting that both the proliferation of increasing genera and the reduction of decreasing genera are partially underlying the loss of phylogenetic diversity at the study landscape.
Negative values indicate loss of phylogenetic diversity over time.
Consistent with the reduction in genera richness and MPD, NRI increased after forest isolation, resulting in a 50%-decrease in tree-wide phylogenetic evenness of tree communities throughout the landscape (NRIinitial = −0.33 vs. NRIfinal = −0.17). Furthermore, as expected, the group of 28 tree genera thriving in the landscape was more phylogenetically clustered than that of 31 declining genera (NRIincreasing = 1.56 vs. NRIdecreasing = 0.09).
Discussion
In the last decades, the scientific community has invested immense efforts to anticipate the fate of tropical biodiversity as it becomes increasingly threatened by human-disturbances, such as habitat fragmentation, agricultural expansion and global warming [39], [40]. Much has been learnt from either theoretical models or empirical studies, at least at the species and community level, but the scarcity of long-term data and lack of reliable evolutionary information still limit our ability to understand the extent to which human-modified landscapes can operate as a repository for the tropical evolutionary heritage (see references [41]–[45] for contrasting views).
The present study bridges part of this knowledge gap by documenting shifts in the phylogenetic structure and diversity of tree communities inhabiting multiple sites in a human-modified landscape (see also Arroyo-Rodríguez et al. [46]). Despite the genus-level analyses and the great variation in the phylogenetic metrics across plots and sites, tree communities experienced significant reduction in tree genera richness, loss of phylogenetic diversity and reduction in phylogenetic evenness over time throughout the entire landscape. Such impoverishment was correlated with demographic shifts of particular genera, suggesting that the replacement of less related, slow-growing tree taxa by more related, fast-growing ones is partially underlying the erosion of evolutionary heritage at the landscape scale. Considering the long lifespan usually exhibited by tropical tree species, potentially exceeding several centuries [47], it is alarming that such shifts have already emerged within an experimental landscape exposed to habitat loss and fragmentation in the early 1980s.
Identifying the causes for the demographic and phylogenetic shifts we documented in this human-modified landscape is complex, especially across the supposed ‘intact’ continuous forest plots located from 170–3000 m from the nearest forest edge. A possibility is that these forest plots could be in a state of disequilibrium recovering from a large-scale past disturbance, leading to time-dependent shifts in community membership and phylogenetic community structure [14], [15], [31]. Major forest fires has been recognized as the only natural disturbance likely to operate at this scale [32], but charcoal and phytolith data suggest that the study area has been continuously forested for at least 4,500 years [48]. Another possibility is that old-growth forests would be responding to multi-decadal changes in rainfall, which affects forest productivity, species composition and possibly phylogenetic diversity. However, no trends in Central Amazonia rainfall were evident for the 1984–1999 period [32] or across preceding decades [49]. What has been documented so far is that our continuous forest plots are experiencing elevated mortality and turnover rates [50], leading to a significant decline in plot species richness [51] in parallel with the proliferation of fast-growing species [32], which also occur across all forest habitats [19], [26]. It has been hypothesized that the elevated mortality and recruitment are a consequence of rising atmospheric CO2 concentrations, which might stimulate plant growth and competition [32].
Independent findings in the Amazon and Brazilian Atlantic forests have provided evidence for a nonrandom taxonomic and functional impoverishment of tree communities inhabiting edge-affect habitats from local to regional scales [19]–[21], [25], largely due to shifts in physical conditions imposed by the creation of forest edges [25], [52]. Such a floristic/functional drift towards communities dominated by few disturbance-adapted fast-growing species ultimately results in increasing levels of biotic homogenization [23], [30] and loss of phylogenetic diversity along forest edges [16]. Although the phylogenetic shifts we documented in the study landscape were not related with plot distance to the nearest forest edge, it is not surprising that (1) tree communities experience phylogenetic diversity loss, clustering and possibly homogenization across multiple habitats and spatial scales, and (2) to some extent the evolutionary heritage of sensitive tropical forest biota will not be retained across human-modified landscapes [24].
Nonetheless, we cannot definitively discard the hypothesis that our findings are simply the outcome of stochastic variation in forest dynamics within the study landscape. This possibility is supported by the 17 reference plots in the continuous forest, which also showed significant temporal changes in genera richness and phylogenetic structure and diversity. We would need even longer tree monitoring, possibly over a century, to properly identify the causes of the taxonomic and ecological shifts that ultimately lead to the phylogenetic shifts at that spatial scale. Also, the long-term magnitude of tree community responses to habitat loss and fragmentation remains uncertain, precluding efforts to scale-up conclusions for other tropical forests. It has been proposed that taxonomic, functional and phylogenetic impoverishment of tree communities across multiple spatial scales are more likely to occur among those biota that have evolved and diversified in the presence of less intensive natural or human-related disturbance regimes, such as in Central Amazonia and the Brazilian Atlantic forests [53]–[55]. In contrast, forests that evolved under intense volcanic activity, frequent hurricanes, and extreme climates, such as some Mexican and Costa Rican forests, support a higher proportion of disturbance-tolerant species than do South America forests, and thus may be less susceptible to human disturbances [56]–[58]. Although the generality of the phylogenetic impoverishment we detected is unclear, it could negatively affect ecosystem productivity and stability and have broader impacts on coevolved organisms [8], [9]. Further studies should examine the generality of both patterns and processes proposed here.
Supporting Information
Table S1.
Raw data of 40 1-ha tree plots, before and after fragmentation.
https://doi.org/10.1371/journal.pone.0113109.s001
(XLSX)
Table S2.
Increasing and decreasing tree genera at the study landscape in Central Amazonia, Brazil.
https://doi.org/10.1371/journal.pone.0113109.s002
(XLSX)
Appendix S1.
Regional phylogeny of the study landscape in Central Amazonia, Brazil.
https://doi.org/10.1371/journal.pone.0113109.s003
(TXT)
Acknowledgments
We thank the many taxonomic specialists and field technicians who have collected and identified plant material during the study. We are grateful to Víctor Arroyo-Rodríguez and an anonymous reviewer for valuable comments. This is publication number 652 in the BDFFP technical series.
Author Contributions
Conceived and designed the experiments: BAS WFL. Performed the experiments: JLCC AA SGL. Analyzed the data: BAS MT FPLM. Contributed reagents/materials/analysis tools: BAS. Wrote the paper: BAS MT FPLM JLCC SGL WFL.
References
- 1. Purvis A, Agapow P, Gittleman JL, Mace GM (2000) Nonrandom extinction and the loss of evolutionary history. Science 288: 328–330.
- 2. Vamosi JC, Wilson JRU (2008) Nonrandom extinction leads to elevated loss of angiosperm evolutionary history. Ecol Lett 11: 1047–1053.
- 3. Webb CO, Ackerly DD, McPeek MA, Donoghue MJ (2002) Phylogenies and community ecology. Annu Rev Ecol Syst 33: 475–505.
- 4. Vamosi SM, Heard SB, Vamosi C, Webb CO (2009) Emerging patterns in the comparative analysis of phylogenetic community structure. Mol Ecol 18: 572–592.
- 5. Pausas JG, Verdu M (2010) The jungle of methods for evaluating phenotypic and phylogenetic structure of communities. Bioscience 60: 614–625.
- 6. Cavender-Bares J, Ackerly DD, Kozak KH (2012) Integrating ecology and phylogenetics: the footprint of history in modern-day communities. Ecology 93(Supplement): S1–S3.
- 7.
Magurran AE (2004) Measuring biological diversity. Oxford: Blackwell Science. 264 p.
- 8. Cavender-Bares J, Kozak KH, Fine PVA, Kembel SW (2009) The merging of community ecology and phylogenetic biology. Ecol Lett 12: 693–715.
- 9. Cadotte MW, Dinnage R, Tilman D (2012) Phylogenetic diversity promotes ecosystem stability. Ecology 93(Supplement): S223–S233.
- 10. Helmus MR, Keller W, Paterson MJ, Yan ND, Cannon CH, et al. (2010) Communities contain closely related species during ecosystem perturbance. Ecol Lett 13: 162–174.
- 11. Verdu M, Pausas JG (2007) Fire drives phylogenetic clustering in Mediterranean Basin woody plant communities. J Ecol 95: 1316–1323.
- 12. Cavender-Barres J, Reich PB (2012) Shocks to the system: community assembly of the oak savanna in a 40-yearfire frequency experiment. Ecology 93(Supplement): S52–S69.
- 13. Cianciaruso MV, Silva IA, Batalha MA, Gaston KJ, Petchey OW (2012) The influence of fire on phylogenetic and functional structure of woody savannas: Moving from species to individuals. Perspect Plant Ecol Evol Syst 14: 205–216.
- 14. Letcher SG (2010) Phylogenetic structure of angiosperm communities during tropical forest succession. Proc R Soc B 277: 97–104.
- 15. Letcher SG, Chazdon RL, Andrade ACS, Bongers F, Breugelf M, et al. (2012) Phylogenetic community structure during succession: Evidence from three Neotropical forest sites. Perspect Plant Ecol Evol Syst 14: 79–87.
- 16. Santos BA, Arroyo-Rodríguez V, Moreno CE, Tabarelli M (2010) Edge-related loss of tree phylogenetic diversity in the severely fragmented Brazilian Atlantic forest. Plos One 5: e12625
- 17. Cadotte MW, Strauss SY (2011) Phylogenetic patterns of colonization and extinction in experimentally assembled plant communities. Plos One 6: e19363
- 18. Winter M, Devictor V, Schweiger O (2013) Phylogenetic diversity and nature conservation: where are we? Trends Ecol Evol 28: 199–204.
- 19. Laurance WF, Nascimento HEM, Laurance SGW, Andrade A, Ribeiro JELS, et al. (2006) Rapid decay of tree-community composition in Amazonian forest fragments. Proc Natl Acad Sci USA 103: 19010–19014.
- 20. Michalski F, Nishi I, Peres CA (2007) Disturbance-mediated drift in tree functional groups in Amazonian forest fragments. Biotropica 39: 691–701.
- 21. Santos BA, Peres CA, Oliveira MA, Grillo A, Alves-Costa CP, et al. (2008) Drastic erosion in functional attributes of tree communities in Atlantic forest fragments of northeastern Brazil. Biol Conserv 141: 249–260.
- 22. Lopes AV, Girão LC, Santos BA, Peres CA, Tabarelli M (2009) Long-term erosion of tree reproductive trait diversity in edge-dominated Atlantic forest fragments. Biol Conserv 142: 1154–1165.
- 23. Tabarelli M, Peres CA, Melo FPL (2012) The ‘few winners and many losers’ paradigm revisited: Emerging prospects for tropical forest biodiversity. Biol Conserv 155: 136–140.
- 24. Laurance WF, Camargo JLC, Luizão RCC, Laurance SGW, Pimm SL, et al. (2011) The fate of Amazonian forest fragments: A 32-year investigation. Biol Conserv 144: 56–67.
- 25. Laurance WF, Nascimento HEM, Laurance SGW, Andrade A, Fearnside PM, et al. (2006) Rain forest fragmentation and the proliferation of successional trees. Ecology 87: 469–482.
- 26. Santos GGA, Santos BA, Nascimento HEM, Tabarelli M (2012) Contrasting demographic structure of short- and long-lived pioneer tree species on Amazonian forest edges. Biotropica 44: 771–778.
- 27. D'Ângelo SA, Andrade ACS, Laurance SGW, Laurance WF, Mesquita RCG (2004) Inferred causes of tree mortality in fragmented and intact Amazonian forests. J Trop Ecol 20: 243–246.
- 28. Briant G, Gond V, Laurance SGW (2010) Habitat fragmentation and the desiccation of forest canopies: A case study from eastern Amazonia. Biol Conserv 143: 2763–2769.
- 29. Tabarelli M, Aguiar AV, Girão LC, Peres CA, Lopes AV (2010) Effects of pioneer tree species hyperabundance on forest fragments in northeastern Brazil. Conserv Biol 24: 1654–1663.
- 30. Lôbo D, Leão T, Melo FPL, Santos AMM, Tabarelli M (2011) Forest fragmentation drives Atlantic Forest of northeastern Brazil to biotic homogenization. Divers Distrib 17: 287–296.
- 31. Norden N, Letcher SG, Boukili V, Swenson NG, Chazdon R (2012) Demographic drivers of successional changes in phylogenetic structure across life-history stages in plant communities. Ecology 93(Supplement): S70–S82.
- 32. Laurance WF, Oliveira AA, Laurance SGW, Condit R, Nascimento HEM, et al. (2004) Pervasive alteration of tree communities in undisturbed Amazonian forests. Nature 428: 171–175.
- 33. Lessard J-P, Fordyce JA, Gotelli NJ, Sanders NJ (2009) Invasive ants alter the phylogenetic structure of ant communities. Ecology 90: 2664–2669.
- 34. Youssef NH, Ashlock-Savage KN, Elshahed MS (2012) Phylogenetic diversities and community structure of members of the extremely halophilic Archaea (order Halobacteriales) in multiple saline sediment habitats. Appl Environ Microbiol 78: 1332–1344.
- 35. Webb CO, Ackerly DD, Kembel SW (2008) Phylocom: software for the analysis of phylogenetic community structure and trait evolution. Bioinformatics 24: 2098–2100.
- 36. Bell CD, Soltis DE, Soltis PS (2010) The age and diversification of the angiosperms re-revisited. Am J Bot 97: 1296–1303.
- 37. Gastauer M, Meira-Neto JAA (2013) Avoiding inaccuracies in tree calibration and phylogenetic community analysis using Phylocom 4.2. Ecol Inform 15: 85–90.
- 38.
Grafen A, Hails R (2002) Modern Statistics for the Life Sciences. Oxford: Oxford University Press. 368 p.
- 39.
Laurance WF, Peres CA (2006) Emerging threats to tropical forests. Chicago: University of Chicago Press. 520 p.
- 40. Laurance WF, Sayer J, Cassman KG (2014) Agricultural expansion and its impacts on tropical nature. Trends Ecol Evol 29: 107–116.
- 41. Wright SJ, Muller-Landau N (2006) The future of tropical forest species. Biotropica 38: 287–301.
- 42. Laurance WF (2007) Have we overstated the tropical biodiversity crisis? Trends Ecol Evol 22: 65–70.
- 43. Tabarelli M, Lopes AV, Peres CA (2008) Edge-effects drive tropical forest fragments towards an early-successional system. Biotropica 40: 657–661.
- 44. Chazdon RL, Peres CA, Dent D, Sheil D, Lugo AE, et al. (2009) The potential for species conservation in tropical secondary forests. Conserv Biol 23: 1406–1417.
- 45. Melo FPL, Arroyo-Rodríguez V, Fahrig L, Martínez-Ramos, Tabarelli M (2013) On the hope for biodiversity-friendly tropical landscapes. Trends Ecol Evol 28: 461–468.
- 46. Arroyo-Rodríguez V, Cavender-Bares J, Escobar F, Melo FPL, Tabarelli M, et al. (2012) Maintenance of tree phylogenetic diversity in a highly fragmented rain forest. J Ecol 100: 702–711.
- 47. Martínez-Ramos M, Alvarez-Buylla ER (1998) How old are tropical rain forest trees? Trends Plant Sci 3: 400–405.
- 48. Piperno DR, Becker P (1996) Vegetation history of a site in the Central Amazon Basin derived from phytolith and charcoal records from natural soils. Quat Int 45: 202–209.
- 49. Mahli Y, Wright J (2004) Spatial patterns and recent trends in the climate of tropical rainforest regions. Philos Trans R Soc Lond B Biol Sci 359: 311–329.
- 50. Laurance SGW, Nascimento HEM, Laurance WF, Andrade A, Fearnside P, et al. (2009) Rapid change in Amazonian forest dynamics: effects of climate change? J Veg Sci 20: 323–333.
- 51. Laurance SGW, Andrade A, Laurance WF (2010) Unanticipated effects of stand dynamism on Amazonian tree diversity. Biotropica 42: 429–434.
- 52. Oliveira MA, Santos AMM, Tabarelli M (2008) Profound impoverishment of the large-tree stand in a hyper-fragmented landscape of the Atlantic forest. For Ecol Manage 256: 1910–1917.
- 53. Barlow J, Gardner TA, Lees AC, Parry L, Peres CA (2012) How pristine are tropical forests? An ecological perspective on the pre-Columbian human footprint in Amazonia and implications for contemporary conservation. Biol Conserv 151: 45–49.
- 54. Arroyo-Rodríguez V, Rös M, Escobar F, Melo FPL, Santos BA, et al. (2013) Plant β-diversity in fragmented rain forests: testing floristic homogenization and differentiation hypotheses. J Ecol 101: 1449–1458.
- 55. Arroyo-Rodríguez V, Pineda E, Escobar F, Benítez-Malvido J (2009) Value of small patches in the conservation of plant-species diversity in highly fragmented rainforest. Conserv Biol 23: 729–739.
- 56. Borouncle C, Finegan B (2011) Tree regeneration and understory woody plants show diverse responses to forest–pasture edges in Costa Rica. Biotropica 43: 562–571.
- 57. Rös M, Escobar F, Halffter G (2011) How dung beetles respond to a human-modified variegated landscape in Mexican cloud forest: a study of biodiversity integrating ecological and biogeographical perspectives. Divers Distrib 18: 377–389.
- 58. Hernández-Ruedas MA, Arroyo-Rodríguez V, Meave JA, Martínez-Ramos M, Ibarra-Manríquez G, et al. (2014) Conserving Tropical Tree Diversity and Forest Structure: The Value of Small Rainforest Patches in Moderately-Managed Landscapes. Plos One 9: e98931