Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 rspb.royalsocietypublishing.org Research Cite this article: Rossetto M, Kooyman R, Yap J-YS, Laffan SW. 2015 From ratites to rats: the size of fleshy fruits shapes species’ distributions and continental rainforest assembly. Proc. R. Soc. B 282: 20151998. http://dx.doi.org/10.1098/rspb.2015.1998 Received: 19 August 2015 Accepted: 11 November 2015 Subject Areas: ecology, plant science, genomics Keywords: Australia, chloroplast genome, dispersal, fruit size, next generation sequencing, rainforest assembly Author for correspondence: Maurizio Rossetto e-mail: [email protected] Electronic supplementary material is available at http://dx.doi.org/10.1098/rspb.2015.1998 or via http://rspb.royalsocietypublishing.org. From ratites to rats: the size of fleshy fruits shapes species’ distributions and continental rainforest assembly Maurizio Rossetto1,2, Robert Kooyman1,3, Jia-Yee S. Yap1,2 and Shawn W. Laffan4 1 National Herbarium of NSW, The Royal Botanic Gardens and Domain Trust, Mrs Macquaries Road, Sydney, New South Wales 2000, Australia 2 QAAFI, The University of Queensland, St Lucia, Queensland 4072, Australia 3 Department of Biological Sciences, Macquarie University, Sydney, New South Wales 2109, Australia 4 Centre for Ecosystem Science, School of Biological, Earth and Environmental Science, University of New South Wales, Sydney 2052, Australia Seed dispersal is a key process in plant spatial dynamics. However, consistently applicable generalizations about dispersal across scales are mostly absent because of the constraints on measuring propagule dispersal distances for many species. Here, we focus on fleshy-fruited taxa, specifically taxa with large fleshy fruits and their dispersers across an entire continental rainforest biome. We compare species-level results of whole-chloroplast DNA analyses in sister taxa with large and small fruits, to regional plot-based samples (310 plots), and whole-continent patterns for the distribution of woody species with either large (more than 30 mm) or smaller fleshy fruits (1093 taxa). The pairwise genomic comparison found higher genetic distances between populations and between regions in the large-fruited species (Endiandra globosa), but higher overall diversity within the small-fruited species (Endiandra discolor). Floristic comparisons among plots confirmed lower numbers of large-fruited species in areas where more extreme rainforest contraction occurred, and recolonization by small-fruited species readily dispersed by the available fauna. Species’ distribution patterns showed that larger-fruited species had smaller geographical ranges than smaller-fruited species and locations with stable refugia (and high endemism) aligned with concentrations of large fleshy-fruited taxa, making them a potentially valuable conservation-planning indicator. 1. Introduction The source and type of propagules (fruit and seed) and their ability to move into available habitat are critical factors determining the abundance and distribution of plant species [1,2]. A range of functional and ecological factors can influence dispersal [3], and trade-offs occur between dispersal mode, dispersal distance, seed mass and seedling survival in different environmental conditions [4]. Here, we focus on how the size of fleshy fruits influences the dispersal potential of rainforest trees and vines, and consequently species’ distributions and continental assembly patterns. Many studies have contributed to the mechanistic understanding of temporal, spatial and phylogenetic variation in seed deposition patterns based on interactions between frugivore behaviour and the distribution of food resources (reviewed in [5]). While providing important insights into local assembly processes [6], it has been difficult for these approaches to generalize more broadly and across larger scales. Technical limitations related to tracking seed movements have constrained the interpretation of community-level dispersal patterns to a compilation of components examined individually across different scales. A more integrated picture of the role of dispersal in assembly patterns will only emerge if frugivore-mediated dispersal can be measured as part of a complex process that takes into account both temporal and spatial scales [3,5]. & 2015 The Author(s) Published by the Royal Society. All rights reserved. Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 type and size influence the dispersal and distribution of species provides important insights into temporal variations in floristic patterns, and can support the development of improved species- and community-level management strategies. 2. Material and methods (a) Fruit size and chloroplast genome variation (b) Fruit size and species’ distributions Although seed dry mass is generally used as a measure of dispersal because of the strong positive relationship to both reserve mass and diaspore size [21], here we are interested in how whole fleshy-fruit size can potentially shape seed dispersal and species’ distribution patterns. Since the quality and quantity of the pulp found on fleshy fruits can influence what dispersers they attract [22], we focused our investigations on fleshy fruits that fully enclose the seed and can potentially be ingested as a whole. A total of 1093 species with such fruits were identified in our original full database of 2306 free-standing and climbing woody rainforest plants [7]. We divided fleshy fruits into two main size groups: less than or equal to 30 mm and greater than 30 mm in width. This cut-off point was chosen because 30 mm is the size of fruits that can be ingested whole by the largest vertebrate dispersers remaining in southern subtropical rainforests, the wompoo fruit dove (Ptilinopus magnificus), topknot pigeon (Lopholaimus antarcticus) and channel-billed cuckoo (Scythrops novaehollandiae; [9]). In tropical Australia, fruits of up to 40 mm are ingested by other volant dispersers (such as the imperial pigeon; Ducula spilorrhoa) [23], and even larger fruits can be dispersed by the remaining Proc. R. Soc. B 282: 20151998 To analyse the influence of fruit size on genetic diversity and connectivity, we compared whole-cpDNA variation in two Endiandra (Lauraceae) species. Endiandra globosa is a tree of up to 30 m in height producing single-seeded black fleshy fruits 35– 60 mm wide. It has two major centres of distribution in the subtropics, from the Brunswick catchment in northern New South Wales (NNSW) to southeastern Queensland (SEQld), and in the tropics within rainforest areas from Ingham to Cairns (Qld). Endiandra discolor is a tree of up to 45 m producing single-seeded black fleshy fruits 10–13 mm wide with distribution ranging from Gosford in southern NSW to Mt Lewis (Qld) in the tropics. In order to test variations in genomic diversity within and between regions, each species was sampled from four sites (two in the tropics and two in the subtropics), targeting similar-sized populations located at similar geographical distances (figure 1). To obtain representations of within- and between-population diversities, total DNA was extracted using Qiagen DNeasy Plant Mini kits with DNA from eight distinct individuals collected for each species at each site. These eight DNA extracts were pooled to prepare genomic libraries for next generation sequencing, following the protocol described in McPherson et al. [16]. Library preparation was in accordance with standard Nextera protocols (Illumina Inc., San Diego, CA, USA), and paired-end (2 150 bp) shotgun sequencing was performed on an Illumina Genome Analyser (GAIIx) by the Ramaciotti Centre for Genomics (University of NSW, Australia). Chloroplast genome data were assembled from the total DNA shotgun libraries based on the approach described by van der Merwe et al. [17]. To investigate the influence of fruit size on overall diversity and genetic connectivity, single-nucleotide polymorphism (SNP) detection was conducted to obtain relative values of Nei’s genomic diversity and genomic distance as described in Rossetto et al. [20]. Further details on sample preparation, molecular analyses and sequencing outcomes are given in electronic supplementary material, tables S1–S3. 2 rspb.royalsocietypublishing.org Continental distributions and ecological interactions within and between the Australian rainforest flora and fauna have been unevenly affected by shifting environmental and climatic conditions [7,8]. Extensive contractions of rainforest vegetation during the Oligo-Miocene and on into the Pleistocene resulted in a considerable loss of vertebrate diversity. Only six functional classes of fruit dispersers remain in Australia with little overlap between birds and mammals [9]. South of the Australian Wet Tropics (AWT) the rainforest contractions resulted in the loss of critical dispersers such as the southern cassowary (Casuarius casuarius johnsonii), the musky rat kangaroo (Hypsiprymnodon moschatus), as well as a general decline in possum diversity [10]. The results of previous genetic and environmental studies on the Australian subtropical flora suggest a reduction in, or loss of, suitable dispersers can have a more significant impact on the local distribution of rainforest trees than habitat availability alone [11,12]. Investigating if and how constraints on dispersal can explain the discordance between available habitat and species’ distribution is an area of increasing global scientific interest, particularly in view of predicted climate change [13]. DNA-based analyses provide broadly applicable methods for quantifying plant dispersal across varying spatial scales [14]. The maternal inheritance and conserved nature of chloroplast DNA (cpDNA) make it particularly useful for exploring seed-mediated dispersal. However, when using standard analytical tools, these same attributes can also restrict the detection of interpretable genetic signatures [15]. The recent development of approaches that enable the analysis of wholecpDNA genomes provides new opportunities for detecting landscape-level patterns, even in non-model species with low diversity [16,17]. Here, we take advantage of a combination of new technologies and unique continent-wide datasets to assess if and how the size of fleshy fruits influences species’ distributions and rainforest assembly patterns at local, regional and continental scales. We focus on fleshy fruit size because this measure highlights ecological interactions with animals that can take and disperse whole fruits (including seeds). We first compare local (paired sites), regional (tropics versus subtropics) and continental chloroplast genome variation in two co-occurring Endiandra (Lauraceae) species that produce fleshy fruits and single seeds of different sizes. Our expectation is that if fruit size impacts on dispersal potential, greater among-site genomic distances should be measured in the large-fruited species. We also expect that genomic distance should increase with increasing geographical distance (regional versus continental) in both species. We then investigate the geographical distribution patterns of 1093 species representing the fleshy-fruited component of the Australian rainforest woody flora. We expect that if the size of fleshy fruits shapes dispersal potential, small-fruited species dispersed by multiple animals should have larger geographical ranges than large-fruited species. In addition, we expect larger-fruited species to be concentrated within temporally stable refugial areas [18,19] where distribution patterns are regulated by the persistence of a greater variety of vertebrate dispersers or through localized persistence and site-specific dispersal mechanisms [11]. To our knowledge, this is the first time that the associations between fruit type, dispersal and distributional patterns have been explored simultaneously at species, regional, wholebiome and whole-continent scales. Understanding how fruit Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 3 HC 0 T 50 km AWT CQ MS Blue rspb.royalsocietypublishing.org CY WA-NT NNSW 0 500 SNSW 1000 km VIC BS 0 25 km CB E. discolor TAS E. globosa Figure 1. Location of collection sites for E. globosa and E. discolor, plus ‘Regional map’ (as previously described in [7] and table 3). Endiandra globosa was sampled from Tchupala (T) and Harvey Creek (HC) in the AWT, and from Hogan’s Scrub (HS) and Coastal Byron (CB) in NNSW. Endiandra discolor was sampled from Bluewater Keelbottom Creek (Blue) and Mount Spec (MS) in the AWT, and from Big Scrub reserve (BS) and CB in NNSW. WA-NT, Western Australia and Northern Territory; CY, Cape York; AWT, Australian Wet Tropics; CQ, Central Queensland; NNSW, Northern New South Wales; SNSW, Southern New South Wales; VIC, Victoria; TAS, Tasmania. native megafauna representative, the southern cassowary, which has a gape of up to 62 mm [9,24]. We included counts of wind-dispersed species as a non-fleshy (outlier) comparison group only to provide additional insights into the continental rainforest flora and the relative proportion of distinct fruit types and dispersal modes. We investigated the relationship between fruit type and size, range size, and the distribution of tree and vine species across the Australian rainforests and within regional areas (figure 1) using the BIODIVERSE software, v. 1.0 (http://purl.org/biodiverse; [25]). The continent-wide distribution and trait data used were described by Kooyman et al. [7] and aggregated to square 50 50 km grid cells. The mean geographical range size for the small and large fruit size classes were calculated using the number of grid cells containing each species. We also used BIODIVERSE to investigate possible landscape-level correlations between the distribution of large-fruited species and patterns of endemism, where endemism was calculated using the weighted endemism index (WE; [26]). WE is a rangeweighted richness score where wide-ranging taxa are assigned a lower weight than narrow-ranging taxa, with the weight being calculated as the fraction of the taxon’s range that occurs in a sample region. As a range-weighted richness score, WE is expressed in the same units as the number of taxa in each fruit size class. Finally, we used plot data from subtropical NNSW to investigate the localized distribution of species with large fleshy fruits across local areas with different long-term disturbance histories. Previous floristic, phylogenetic and genetic studies in NNSW suggest that rainforests in the Washpool area have been recently re-colonized from more stable areas to the south (Dorrigo) and north (Nightcap/Border Ranges; [7,20]). As a result, we analysed plot data from a previous study (140 plots from Nightcap/ Border Ranges; 127 from Dorrigo; 43 from Washpool [27]) to explore the potential impact of long-term disturbance events on the local distribution of large-fruited species. We calculated the relative proportion of species with large fleshy fruits in relation to an expected regional average, looking for areas with a significantly larger than expected number using a randomization analysis in which the species were randomly assigned to sites, but where the species richness for each site and geographical range of each species were held constant [28]. 3. Results (a) Fruit size and chloroplast genome variation Shotgun sequencing of total DNA produced over 158 500 bp of cpDNA with sufficient coverage (average 139) to investigate within- and between-population diversities in the two related and co-occurring species. Pairwise alignment of the E. globosa and E. discolor cpDNA sequences identified a total of 200 indels and 943 fixed SNPs differentiating between species (see electronic supplementary material, tables S1 and S2). The genomic diversity measures obtained were highly informative and directly comparable at all scales within and between species and highlight the power and relative simplicity of this whole-genome approach. We found higher levels of Nei’s genomic diversity overall, regionally and at population level in the small-fruited E. discolor (table 1 and electronic supplementary material, table S3). Levels of genomic diversity were also more consistent throughout the distribution of E. discolor (table 1 and electronic supplementary material, table S3). Genomic distances between regional populations and between regions were higher in the large-fruited species E. globosa, while there were no fixed SNP variants differentiating subtropical sites or regions in E. discolor (table 1). At the regional level, genomic distances were higher between tropical sites than between subtropical sites for both species, while Nei’s genomic diversity was greater for E. globosa in the tropics and for E. discolor in the subtropics (table 1). (b) Fruit size and species’ distributions Of the total of 2306 woody Australian rainforest species in the dataset, 433 (18.8%) have wind-dispersed propagules, and 1093 (48.1%) have fleshy fruits that fully enclose the seed. Overall, the proportion of species with large fleshy fruits was low (table 2). However, while vines had a much higher percentage of wind-dispersed species than trees, trees Proc. R. Soc. B 282: 20151998 HS Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 Endiandra discolor chloroplast genome length (bp) 158 585 158 567 total number of SNPs detected 183 294 overall Nei’s genomic diversity 7.10 1025 1.87 1024 Nei’s genomic diversity in AWT Nei’s genomic diversity in NNSW 9.39 1025 2.04 1025 1.18 1024 2.41 1024 genomic distance between AWT sites 7.24 1025 1.55 1025 genomic distance between 9.05 1026 0 NNSW sites genomic distance between AWT 4.4 1024 0 and NNSW and vines had very similar proportions of fleshy-fruited species within each size class. It is unlikely that phylogenetic bias impacts on these results, as fleshy-fruited species were present in 249 genera and 100 families with only 23% of the families shared between trees and vines. Despite large-fruited species representing less than 20% of all fleshy-fruited species, they were present in 56% of all fleshy-fruited families. Finally, even large families with numerous large-fruited species (such as Apocynaceae, Fabaceae, Lauraceae and Myrtaceae) included a wide range of fruit sizes. Larger-fruited species had smaller geographical ranges than smaller-fruited ones (as represented by the number of cells across a continental grid, table 2), with the latter occupying similar ranges to wind-dispersed species. Differences in average geographical ranges spanned by the two size categories (less than or equal to 30 mm ¼ 25.3; more than 30 mm ¼ 12.1) were highly significant ( p , 0.001; t-test). Geographically, the proportion of large to small fruits varied from north to south (table 3) and from refugial to nonrefugial areas. The proportion of wind-dispersed and of fleshy-fruited species remained stable across regions (except for Tasmania where the wind-dispersed category increased and overall diversity was very low), and the frequency of species with larger fleshy fruits decreased with increasing latitude (table 3). The proportion of species with large fleshy fruits was particularly high in regions known to include multiple stable refugia: the AWT and Cape York in tropical Queensland, and NNSW in the subtropics (table 3). Mapped concentrations of species with large fleshy fruits corresponded to regions with high weighted endemism (WE; see electronic supplementary material, figure S1). At local scales, the number of species with large fleshy fruits within NNSW plots is significantly higher within areas that include multiple refugia (Nightcap–Border Ranges N ¼ 21; Dorrigo N ¼ 10), than in re-colonized areas (Washpool N ¼ 5). The proportion of plots containing a number of species with large fruits that is significantly higher than expected was total trees vines percentage fleshy percentage fleshy more than 30 mm 48.1 7.3 49.1 7.6 43.8 5.8 percentage wind-dispersed 18.8 13.1 43.3 100 80 43 number of families with fleshy fruits greater than 30 mm 39 32 14 number of wind-dispersed families 43 34 16 mean no. cells for fleshy less than or 25.3 25.5 26.6 mean no. cells for fleshy more than 30 mm 12.1 12.6 9.7 mean no. cells for wind-dispersed 22.1 21.3 23.2 number of fleshy families equal to 30 mm more than double in Nightcap–Border Ranges (7.8% of plots) than in Dorrigo (3.6% of plots), while Washpool had none (0% of plots; figure 2). 4. Discussion Seed dispersal by frugivorous species is a crucial component of plant population dynamics with consequences for spatial structure, maintenance of diversity and colonization of new habitats [3]. We found variation in species extent by wholefruit size and then determined the influence of this extent on community assembly across diverse spatial scales. We used an innovative approach that integrated whole-chloroplast genome variation in sister taxa with small and large fleshy fruits, with analyses of the distributional patterns of smalland large-fruited species across the whole of a continental biome. At a continental scale, we showed that species with large fleshy fruits occupied smaller areas than small-fruited species; across and within regions we showed that an Endiandra species with large fleshy fruit displayed greater betweenpopulation genomic divergence than one with smaller fruit; and across three regional refugia in NNSW, we showed that there were disproportionately low numbers of large fleshy-fruited species within more recently recolonized areas. The distinctive whole-cpDNA signatures detected across the two Endiandra species investigated corresponded to our expectations of differing dispersal potential, with high genomic distances between sites at regional and continental scales for the large-fruited species (E. globosa), and continent-wide cpDNA homogeneity for the small-fruited species (E. discolor; table 1). For E. globosa, high cpDNA heterogeneity at larger geographical scales (including greater genomic diversity in the tropics) suggests that while dispersal might be sufficient 4 Proc. R. Soc. B 282: 20151998 Endiandra globosa Table 2. Proportions of rainforest trees and vines with different fruit characteristics, including differently sized fleshy fruits and wind-dispersed fruits as non-fleshy comparison (out of a total of 2306 species). Representation of rainforest trees and vines with fleshy fruits and winddispersed across families (11.6% overlap between trees and vines with fleshy fruits, and 23% with wind-dispersed fruits) and mean distribution area (measured as number of geographical grid cells) of species within different fruit-size groups are also shown. rspb.royalsocietypublishing.org Table 1. A summary of chloroplast genome data for two Endiandra species: reported chloroplast genome length; total number of SNPs (including withinpopulation SNPs and fixed between-population SNPs); Nei’s regional genomic diversity (for AWT and NNSW); and genomic distances between regional sites and between regions (calculated following [17]). Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 5 rspb.royalsocietypublishing.org all plots significantly high proportion of fleshy fruits more than 30 mm Nightcap Border Ranges Proc. R. Soc. B 282: 20151998 significantly high proportion of any fruits more than 30 mm Washpool N Dorrigo 0 25 50 km Figure 2. Relative proportion of species with large fleshy fruits in relation to an expected regional average across plots located within Nightcap –Border Ranges, Dorrigo and Washpool, three important rainforest areas in NNSW. Table 3. Regional distributions (listed from north to south) as totals and percentages of fruit categories across trees and vines. WA-NT, Western Australia and Northern Territory; CY, Cape York; AWT, Australian Wet Tropics; CQ, Central Queensland; NNSW, Northern New South Wales; SNSW, Southern New South Wales; VIC, Victoria; TAS, Tasmania. total species tropics subtropics % fleshy % fleshy more than 30 mm % wind 477 46.1 5.2 18.4 CY AWT 946 1587 48.2 50.2 6.6 7.9 15.8 16.9 CQ NNSW 792 740 47.3 48.1 4.4 5.4 16.3 18.9 SNSW 218 51.8 2.3 21.6 98 50 50.0 32.0 2.0 0 24.5 46.0 WA-NT VIC TAS to maintain local population processes, longer distance connectivity is limited. For E. discolor, landscape-level genomic homogeneity confirms efficient range-wide dispersal, with high cpDNA diversity suggesting a scenario of long-term presence and continuing connectivity rather than one of more recent continental expansion [20,29]. The current distribution of the Australian rainforest flora is the product of longer-term geological and shorter-term climatic processes [30,31]. The fossil record suggests that in the Eocene to early Miocene, many species with large fleshy fruits had continent-wide distributions [32], as did the larger vertebrates capable of dispersing them [8]. During the glacial Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 Data accessibility. The E. globosa and E. discolor chloroplast genomes generated in this study have been lodged in the GenBank database at NCBI (KT588614-KT588615). All chloroplast sequences aligned to the genomes are available through Dryad: http://dx.doi.org/10. 5061/dryad.rm5dr. Authors’ contributions. M.R. conceived, designed and coordinated the study and led the writing of the manuscript; J.-Y.S.Y. carried out molecular analysis; R.K. contributed to the development of the project, writing of the manuscript and provided field data; S.W.L. carried out statistical analyses. All authors helped writing the manuscript and gave final approval for publication. Competing interests. We have no competing interests. Funding. We thank Bioplatforms Australia Ltd. and the Friends & Foundation of RBGS for financial support. Acknowledgements. We thank Andrew Ford for the collection of material in northern Queensland, M. van der Merwe for support with genomic analyses and the editor and reviewers for insightful comments. References 1. 2. Hanski IA, Gilpin ME. 1997 Metapopulation biology. San Diego, CA: Academic. Levine JM, Murrell DJ. 2003 The community-level consequences of seed dispersal patterns. Annu. Rev. Ecol. Evol. Syst. 34, 549–574. (doi:10.1146/ annurev.ecolsys.34.011802.132400) 3. 4. Cousens R, Dytham C, Law R. 2008 Dispersal in plants. A population perspective. Oxford, UK: Oxford University Press. Leishman MR, Westoby M, Jurado E. 1995 Correlates of seed size variation: a comparison among five temperate floras. J. Ecol. 83, 517–529. (doi:10.2307/2261604) 5. 6. Côrtes MC, Uriarte M. 2013 Integrating frugivory and animal movement: a review of the evidence and implications for scaling seed dispersal. Biol. Rev. 88, 255–272. (doi:10.1111/j.1469-185X.2012.00250.x) Muller-Landau HC, Wright SJ, Calderon O, Condit R, Hubbell SP. 2008 Interspecific variation in primary 6 Proc. R. Soc. B 282: 20151998 Endiandra, that connectivity can be relatively low even in areas where the highest numbers of dispersers persist (table 1). The genomic findings relate to the comparison of just two species within a single genus, and the assemblage-level data focus on within-region comparisons. However, despite those limitations, our results suggest that large-fruited species are an important target for conservation monitoring, and potentially important for the investigation of a range of evolutionary, biogeographic and ecological processes within rainforests. When viewed independently of current species richness, our results from refugia in NNSW suggest that the presence of multiple species with large fleshy fruits within fragmented habitats may provide an informative tool for identifying the location of environmentally stable sites, particularly within southern regions. Management strategies aimed at maintaining diversity across all fruit size groups need to be adapted according to regional contexts. For instance, if the Australian subtropical rainforests have survived without megafauna dispersers since the Oligocene or Miocene, persisting populations of large-fruited species will have attained some resilience to change by relying on available smaller dispersers to maintain localized connectivity and expansion potential. However, the fossil record suggests that extinctions of subtropical megafauna were more recent [10,34]. Within the latter megafauna extinction scenario, an increasing and continuing shift towards small-fruited species (including recent immigrants of Indo-Malesian origin; [20]) is conceivable. Because large-fruited species are unlikely to expand their range via natural dispersal mechanisms within a fragmented habitat matrix, specific management interventions targeting large-fruited taxa are likely to be necessary. These include population enhancements and assisted migration based on well-informed genetic guidelines [11]. Our results suggest that clearly defining the combined influence of environmental change on large fleshyfruited species and their dispersers will be central to improving interpretations of local evolutionary patterns, and the influence of isolation and vicariance on the Australian rainforest. rspb.royalsocietypublishing.org fluctuations of the Quaternary, an overall decline in moisture shaped the distribution of vegetation by increasing aridity, fire frequency and fire intensity, favouring Eucalyptus intrusions into tropical rainforests as recently as 8 Kya [33]. Climatic fluctuations were particularly intense during the last few glacial cycles [30] causing extreme rainforest contractions and extinctions, especially in subtropical areas [10]. Volant vertebrates that can ingest fleshy fruits less than or equal to 30 mm are widely distributed across Australian rainforests, while those that can potentially disperse larger fruits and seeds are restricted to more tropical latitudes [9]. In both the tropics and subtropics, richness of large-fruited species tends to track overall richness and high endemism (see electronic supplementary material, figure S1), the latter generally acknowledged as an important indicator of environmentally stable refugia [18,19]. These correlations suggest that these regions maintained high floristic diversity through the climatic extremes of the Quaternary, despite the likelihood that some southern refugia were just too small and/or locally dissected to retain vertebrate dispersers capable of supporting the re-expansion of species with large fruits into newly available habitats [12]. Areas that were affected by more extreme disturbance events through time are likely to depend on surviving or re-colonizing fauna to facilitate the re-establishment of fleshyfruited species. South of the tropics, where the number of frugivorous vertebrates is limited, local differences in the intensity of temporal climatic fluctuations may therefore be most apparent in the distribution of poorly dispersed (large-fruited) plant species. The floristic comparison among regional NNSW plots presented here substantiates the expectations of greater losses of large-fruited species within areas that have endured more long-term disturbances. The very low number of species with large fruits from Washpool (figure 2) is consistent with a scenario of extreme rainforest contractions followed by recent re-colonization by small-fruited species that are readily dispersed by the available fauna. It is important to note that the local disappearance of specialized large-fruit dispersers is unlikely to lead to an immediate loss of large-fruited species as other vectors such as bush rats (e.g. Rattus fuscipes and Melomys cervinipes), possums (e.g. Trichosurus spp.) and water can mediate localized dispersal. Previous studies have demonstrated that even within small, environmentally stable refugia, large-fruited species can survive and persist in small populations by relying on localized dispersal [11]. The low genomic divergence measured among subtropical populations of E. globosa (table 1) supports the concept that dispersal by small animals may be sufficient to maintain local connectivity across areas of uninterrupted habitat. Furthermore, although large-fruited species can have relatively large continental ranges that include the tropics and subtropics, we show that they generally occur in localized patchy distributions (table 2), and, in the case of a large-fruited Downloaded from http://rspb.royalsocietypublishing.org/ on June 15, 2017 8. 10. 11. 12. 13. 14. 15. 16. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. Australia. Biotropica 40, 338–343. (doi:10.1111/j. 1744-7429.2007.00372.x) Laffan SW, Lubarsky E, Rosauer DF. 2010 Biodiverse, a tool for the spatial analysis of biological and related diversity. Ecography 33, 643–647. (doi:10. 1111/j.1600-0587.2010.06237.x) Crisp MD, Laffan S, Linder HP, Monro A. 2001 Endemism in the Australian flora. J. Biogeogr. 28, 183–198. (doi:10.1046/j.1365-2699.2001.00524.x) Kooyman R, Rossetto M, Cornwell W, Westoby M. 2011 Phylogenetic tests of community assembly across regional to continental scales in tropical and subtropical rain forests. Glob. Ecol. Biogeogr. 20, 707–716. (doi:10.1111/j.1466-8238.2010.00641.x) Laffan SW, Crisp MD. 2003 Assessing endemism at multiple spatial scales, with an example from the Australian vascular flora. J. Biogeogr. 30, 511 –520. (doi:10.1046/j.1365-2699.2003.00875.x) Petit RJ et al. 2003 Glacial refugia: hotspots but not melting pots of genetic diversity. Science 300, 1563– 1565. (doi:10.1126/science.1083264) Kershaw AP, Bretherton SC, van der Kaars S. 2007 A complete pollen record of the last 230 ka from Lynch’s crater, north-eastern Australia. Palaeogeogr. Palaeoclimatol. Palaeoecol. 251, 23 –45. (doi:10. 1016/j.palaeo.2007.02.015) Kooyman RM et al. 2014 Paleo-Antarctic rainforest into the modern Old World tropics: the rich past and threatened future of the ‘southern wet forest survivors’. Am. J. Bot. 101, 2121 –2135. (doi:10. 3732/ajb.1400340) Greenwood DR, Christophel DC. 2005 The origins and tertiary history of Australian ‘tropical’ rainforests. In Tropical rainforests: past, present and future, 1st edn (eds E Bermingham, CW Dick, C Moritz), pp. 336–373. Chicago, IL: The University of Chicago Press. Hopkins MS, Ash J, Graham AW, Head J, Hewett RK. 1993 Charcoal evidence of the spatial extent of Eucalyptus woodland expansions and rainforest contractions in North Queensland during the Late Pleistocene. J. Biogeogr. 20, 357 –372. (doi:10. 2307/2845585) Miller AH. 1962 The history and significance of the fossil Casuarius lydekkeri. Rec. Aust. Museum 25, 235–238. (doi:10.3853/j.0067-1975.25.1962.662) 7 Proc. R. Soc. B 282: 20151998 9. 17. variation for molecular ecology studies: a simple next generation sequencing approach applied to a rainforest tree. BMC Ecol. 13, 1472 –6785. (doi:10. 1186/1472-6785-13-8) Van der Merwe M, McPherson H, Siow J, Rossetto M. 2014 Next-gen phylogeography of rainforest trees: exploring landscape-level cpDNA variation from whole-genome sequencing. Mol. Ecol. Resour. 14, 199– 208. (doi:10.1111/1755-0998.12176) Kooyman R, Rossetto M, Allen C, Cornwell W. 2012 Australian tropical and subtropical rain forest community assembly: phylogeny, functional biogeography, and environmental gradients. Biotropica 44, 668 –679. (doi:10.1111/j.1744-7429. 2012.00861.x) Weber LC, VanDerWal J, Schmidt S, McDonald WJF, Shoo LP. 2014 Patterns of rain forest plant endemism in subtropical Australia relate to stable mesic refugia and species dispersal limitations. J. Biogeogr. 41, 222–238. (doi:10.1111/jbi.12219) Rossetto M, McPherson H, Siow J, Kooyman R, van der Merwe M, Wilson PD. 2015 Where did all the trees come from? A novel multidisciplinary approach reveals the impacts of biogeographic history and functional diversity on rain forest assembly. J. Biogeogr. 42, 2172–2186. (doi:10.1111/jbi.12571) Leishman MR, Wright IJ, Moles AT, Westoby M. 2000 The evolutionary ecology of seed size. In Seeds: the ecology of regeneration in plant communities (ed. M Fenner), pp. 31– 57. Wallingford, CT: CAB International. Grubb PJ, Newbery DM, Prins HHT, Brown ND. 1998 Seeds and fruits in tropical rainforest plants: interpretation of the range in seed size, degree of defence and flesh/seed quotients. In Dynamics of tropical communities (eds DM Newbery, HHT Prins, ND Brown), pp. 1 –24. Oxford, UK: Blackwell Scientific. Westcott DA, Bentrupperbaumer J, Bradford MG, McKeow A. 2005 Incorporating patterns of disperser behaviour into models of seed dispersal and its effects on estimated dispersal curves. Oecologia 146, 57 –67. (doi:10.1007/s00442-005-0178-1) Bradford MG, Dennis AJ, Westcott DA. 2008 Diet and dietary preferences of the Southern Cassowary (Casuarius casuarius) in North Queensland, rspb.royalsocietypublishing.org 7. seed dispersal in tropical forest. J. Ecol. 96, 653–667. (doi:10.1111/j.1365-2745.2008.01399.x) Kooyman R, Rossetto M, Sauquet H, Laffan SW. 2013 Landscape patterns in rainforest phylogenetic signal: isolated islands of refugia or structured continental distributions? PLoS ONE 8, e80685. (doi:10.1371/journal.pone.0080685) Wroe S, Field JH, Archer M, Grayson DK, Price GJ, Louys J, Mooney SD. 2013 Climate change frames debate over the extinction of megafauna in Sahul (Pleistocene Australia-New Guinea). Proc. Natl Acad. Sci. USA 110, 8777 –8781. (doi:10.1073/pnas. 1302698110) Dennis AJ, Westcott DA. 2006 Reducing complexity when studying seed dispersal at community scales: a functional classification of vertebrate seed dispersers in tropical forests. Oecologia 149, 620–634. (doi:10.1007/s00442-006-0475-3) Hocknull SA, Zhao JX, Feng YX, Webb GE. 2007 Responses of Quaternary rainforest vertebrates to climate change in Australia. Earth Planet. Sci. Lett. 264, 317–331. (doi:10.1016/j.epsl.2007.10.004) Rossetto M, Kooyman R. 2005 The tension between dispersal and persistence regulates the current distribution of rare palaeo-endemic rain forest flora: a case study. J. Ecol. 93, 906– 917. (doi:10.1111/j. 1365-2745.2005.01046.x) Rossetto M, Kooyman R, Sherwin W, Jones R. 2008 Dispersal limitations rather than bottlenecks or habitat specificity can restrict the distribution of rare and endemic rainforest trees. Am. J. Bot. 95, 321–329. (doi:10.3732/ajb.95.3.321) Siefert A, Lesser MR, Fridley JD. 2015 How do climate and dispersal traits limit ranges of tree species along latitudinal and elevational gradients? Glob. Ecol. Biogeogr. 24, 581–593. (doi:10.1111/geb.12287) Cain ML, Milligan BG, Strand AE. 2000 Long distance seed dispersal in plant populations. Am. J. Bot. 87, 2117 –2227. (doi:10.2307/2656714) Schaal BA, Hayworth DA, Olsen KM, Rauscher JT, Smith WA. 1998 Phylogeographic studies in plants: problems and prospects. Mol. Ecol. 7, 465–474. (doi:10.1046/j.1365-294x.1998.00318.x) McPherson H, Van der Merwe M, Delaney SK, Edwards MA, Henry RJ, McIntosh E, Rymer PD, Milner ML, Siow J. 2013 Capturing chloroplast