Download From ratites to rats: the size of fleshy fruits shapes species

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