Access to this full-text is provided by Wiley.
Content available from Ecology and Evolution
This content is subject to copyright. Terms and conditions apply.
13476
|
Ecology and Evolution. 2020;10:13476–13487.www.ecolevol.org
Received: 28 June 2020
|
Revised: 14 September 2020
|
Accepted: 7 October 2020
DOI: 10.1002 /ece3.6952
ORIGINAL RESEARCH
Different landscape effects on the genetic structure of two
broadly distributed woody legumes, Acacia salicina and
A. stenophylla (Fabaceae)
Francisco Encinas-Viso1 | Christiana McDonald-Spicer1,2 | Nunzio Knerr1 |
Peter H. Thrall3 | Linda Broadhurst1
This is an op en access arti cle under the ter ms of the Creative Commons Attribution L icense, which pe rmits use, dis tribu tion and reprod uction in any med ium,
provide d the original wor k is properly cited.
© 2020 The Authors. Ecolog y and Evolution published by John Wiley & S ons Ltd.
1Centre for Australian National Biodiversity
Research, CSIRO, Canb erra , ACT, Austr alia
2The Australian National University,
Canbe rra, AC T, Australia
3CSIRO A griculture & Food, Canberra, ACT,
Australia
Correspondence
Francisco Encinas-Viso, Cent re for
Australian Nat ional Biodiver sity Research,
CSIRO, C anber ra, ACT, Australia.
Email: francisco.encinas-viso@csiro.au
Abstract
Restoring degraded landscapes has primarily focused on re-establishing native plant
communities. However, little is known with respect to the diversity and distribution
of most key revegetation species or the environmental and anthropogenic factors
that may affect their demography and genetic structure. In this study, we investi-
gated the genetic structure of two widespread Australian legume species (Acacia
salicina and Acacia stenophylla) in the Murray–Darling Basin (MDB), a large agricul-
turally utilized region in Australia, and assessed the impact of landscape structure
on genetic differentiation. We used AFLP genetic data and sampled a total of 28
A. salicina and 30 A. stenophylla sampling locations across southeastern Australia.
We specifically evaluated the importance of four landscape features: forest cover,
land cover, water stream cover, and elevation. We found that both species had high
genetic diversity (mean percentage of polymorphic loci, 55.1% for A. salicina ver-
sus. 64.3% for A. stenophylla) and differentiation among local sampling locations (A.
salicina: ΦPT = 0.301, 30%; A. stenophylla: ΦPT = 0.235, 23%). Population structure
analysis showed that both species had high levels of structure (6 clusters each) and
admixture in some sampling locations, particularly A. stenophylla. Although both spe-
cies have a similar geographic range, the drivers of genetic connectivity for each spe-
cies were very different. Genetic variation in A. salicina seems to be mainly driven by
geographic distance, while for A. stenophylla, land cover appears to be the most im-
portant factor. This suggests that for the latter species, gene flow among populations
is affected by habitat fragmentation. We conclude that these largely co-occurring
species require different management actions to maintain population connectivity.
We recommend active management of A. stenophylla in the MDB to improve gene
flow in the adversity of increasing disturbances (e.g., droughts) driven by climate
change and anthropogenic factors.
|
13477
ENCINA S-VISO Et Al .
1 | INTRODUCTION
Environmental changes through space and time can produce genetic
differentiation (Fenderson et al., 2020). However, determining the
role of specific environmental factors that cause genetic differenti-
ation is still challenging. Changes in the landscape produced by cli-
mate change and intensified land use can generate a severe decrease
of genetic connectivity and population viability in many plants and
animals (Frankham et al., 2010). Therefore, the per vasive effects of
global changes, and particularly habitat fragmentation, increase ex-
tinction risk of native species in urban and agriculturally intensified
areas even in apparently resilient plant species (Vranckx et al., 2012;
Young et al., 1996). Shrubby legumes belonging to the genus Acacia
are highly diverse and widespread across the Australian conti-
nent. Acacias form major components of many ecosystems across
the continent including many arid ecosystems with poor soils (Bui
et al., 2014; Maslin et al., 2003) and play an important role in eco-
system functioning including through the provision of resources and
habitat to a broad range of insect s and animals (Wandrag et al., 2015;
Ward & Branstetter, 2017; Young et al., 2008). It also helps rapid
colonization supporting ecosystem recovery following disturbance
(Spooner, 20 05). Consequently, acacias often play a critical role in
the restoration of highly degraded areas (Jeddi & Chaieb, 2012).
Restoration using Acacia species primarily occurs within re-
gions where fragmentation of native vegetation is extensive (Doi &
Ranamukhaarachchi, 2013; Jeddi & Chaieb, 2012). Fragment ation of
large and continuous vegetation results in smaller, more isolated popu-
lations, often with lower genetic diversity, an increased risk of further
genetic loss through drift and elevated inb reedin g (Aguilar et al., 20 06,
2008; Hamrick, 2004; Young et al., 1996). Consequently, there are
risks associated with using seed crops from small populations for res-
toration purposes. Our understanding, however, about how landscape
fragmentation and other environmental factors (e.g., elevation) shape
patterns of gene flow in Australian Acacia species remains unclear. A
recent meta-analysis of patterns of genetic diversity highlighted that
Australian species generally follow global expectations when factors
including range size, form, and abundance are considered (Broadhurst
et al., 2017). This study also found that genetic diversity is lower in
Australian shrubs (primarily acacias) when compared to trees or herbs
and that population genetic structure (Fst/Gst) in shrubs and trees was
estimated to be twice that observed in global studies. While observa-
tions such as these are useful for high-level comparisons, understand-
in g the ma j o r drive rs of am o n g- spec ies va r i a t ion in ge n etic di ver sit y and
structure may be more important for guiding conservation decisions.
Here, we compare genetic diversity, population genetic structure,
and landscape genetics using AFLP data in two functionally similar
shrubby legumes (Acacia salicina and Acacia stenophylla) to improve
our understanding of the main environmental factors shaping genetic
connectivity in these two species. Acacia salicina an d Acacia stenophylla
are both broadly distributed across the Murray Darling Basin (MDB)
(Figure 1) in eastern Australia, one of Australia's most large river sys-
tem that has been extensively used for agricultural production (Cai &
Cowan, 2008). Importantly, these two species however have partially
contrasting life-history and environmental requirements. A. salicina is
a perennial woody shrub that mainly occurs in semi-arid habitats and
it is a very successful colonizer of degraded areas with high tolerance
of bare soil (Gr ig g & Mulligan, 1999). Th is species has be en int ro du ced
successfully in different parts of the world to revegetate degraded
areas and restore soil conditions (Jeddi & Chaieb, 2010), and it is inva-
sive in some arid areas of Israel (Jeddi & Chaieb, 2012). Although not
much is known about seed dispersal mechanisms of A. salicina, some
evidence suggests that birds can disperse their seeds (O’Dowd & Gill,
1986). A. stenophylla is a small woody shrub that mainly occurs in ri-
parian ecosystems of Australian river dryland areas. This species pro-
vides nesting habitat for many birds in floodplains of inland Australia,
and its main seed dispersal mechanism is through hydrochory (Murray
et al., 2019). Thus, this species might have specific patterns of genetic
structure and diversity modulated by downstream unidirectional gene
flow through the MDB river system (Ritland, 1989).
Given their ecological and biological differences and to uncover
the effects of landscape fragmentation and other environmental
factors on genetic connec tivit y, we have formulated the following
hypotheses: We hypothesize that these two species potentially have
different gene flow connectivity patterns through the landscape.
We expect A. stenophylla to be more sensitive to habitat fragmen-
tation, historical changes in water fluctuations of the MDB (Cai &
Cowan, 2008), and being affected by hydrological connectivity di-
rectly shaping its genetic structure and diversity, while A. salicina
seems more resilient and can quickly (re)colonize degraded areas
(Grigg & Mulligan, 1999; Jeddi & Chaieb, 2012) and the river system
might ac t as geographi c barriers for gene flow. To test the se hypoth-
eses, we were interested in determining: (a) Did levels of genetic di-
versity differ between the two species? (b) Did population genetic
structure differ bet ween the two species? And 3. if differences be-
tween the two species were evident, could this be explained by en-
vironmental factors, such as elevation and/or habitat fragmentation?
2 | METHODS
2.1 | Site selection and collection of genetic
material
Location data from herbarium specimen records of the Australian
Virtual Herbarium were obtained to guide the selection of sites. A
survey was then conducted to multiple agricultural areas of the MDB
KEYWORDS
Australia, connectivity, gene flow, habitat fragmentation, landscape genetics, population
structure, resistance surfaces
13478
|
ENCINAS-VISO Et A l.
(New South Wales, Australia). We selected sites where the num-
ber of mature individuals of A. stenophylla and A. salicina exceeded
20–30 trees and distance between locations was greater than 30 km
(Thrall et al., 2007). A total of 28 A. salicina and 30 A. stenophylla
sampling locations were collected from across the MDB in south-
eastern Australia (Figure 1). 25 and 28 of those sampling locations of
A. stenophylla and A. salicina; respectively, were located close to riv-
erbanks or water streams. Phyllode material was collected from up
to 30 trees in each sampling location, kept cool during transpor t to
the laboratory, lyophilized for 2–5 days (Flexi-Dry MP FTS Systems,
USA), and stored for DNA analysis.
2.2 | DNA extractions and AFLP genotyping
DNA was extracted from ~10 mg of dried tissue ground to a fine
powder using 3-mm tungsten carbide beads in a Retsch MM300
mixer mill using the Qiagen 96-well DNEasy Ex traction Kit (Qiagen,
Melbourne) following the manufacturer's protocol. AFLP amplifica-
tion largely followed that of (Vos et al., 1995) with the exceptions
that 500ng of genomic DNA was digested for each sample using
EcoRI-MseI, the EcoI-A-MseI-C preamplification reaction was diluted
1:30 prior to selective amplification and selective amplification
primers were fluorescently labeled. Initial screening of 12 primer
combinations identified six with polymorphic and repeatable banding
patterns (Eco-AGC/Mse-CTC (FAM), Eco-AGT/Mse-CTC (PET), Eco-
ACC/Mse-CTC ( VIC), Eco-AGC/Mse-CTG (FAM), Eco-ACC/Mse-
CTG (VIC), Eco-AGA/Mse-CTG (NED)) (dye used shown in bracket s).
Amplicons were visualized on an ABI 3130XL sequencer using a LIZ
500-bp internal standard (Applied Biosystems) and scored using
GeneMapper Version 4.0 software (Applied Biosystems). A binary
matrix of present (1) and absent (0) bands was constructed for each
species. Ninety-five samples from 3 to 4 populations from across
the geographic range of both A. salicina and A. stenophylla were run
twice for each of the proposed primer pairs to test for reproduc-
ibility across a range of 120 to 450 bp. Markers with an error rate of
>5% were discarded with the error rate for markers selected for A.
salicina ranging from 0–3.4 and for A. stenophylla being 1.5–4.3. A
ne g ati ve co n tro l was ru n wit h eve r y set of sa mpl e s in a 96 -we ll blo c k.
2.3 | Genetic data analyses
The binar y data matr ix of each sp ecies was used to estimate th e pe r-
centage of polymorphic loci (%P) and expec ted heterozygosity (He)
for each sampling location using GenAlEx version 6.41 (Peakall &
FIGURE 1 Sampling locations for
A. salicina (orange filled circles) and A.
stenophylla (red filled circles). Gray shading
indicates extent of Murray Darling Basin.
Sampling location 5 (black filled circle)
indicates site where both species were
collected
24 27
1
13
19
18
9
5
24
29
15
20
8
6
17
3
11
25
12
16
21
23
28
26
14
22
10
57
34
56
43
47
52
32
58
55
30
50
48
37
51
38
42
35
46
31
33
39
45
36
40
49
41
44
54
53
0km150km 300km
N
|
13479
ENCINA S-VISO Et Al .
Smouse, 2006). We also checked for sample size effects for both
species (see Figure S12), and we did not find major effec ts on genetic
diversity (He); except for sampling location 48 (N = 7) of A. steno-
phylla, which showed significant differences of genetic diversity
(He) (Wilcoxon test: W = 322,230, p < .05) between all sample sizes.
GenAlEx was also used for an analysis of molecular variance (AMOVA
(Excoffier et al., 1992)) to determine the distribution of genetic varia-
tion among sampling loc ations for both species with the significance
of ΦPT (analogue of FST, population differentiation statistics) based
on 999 permutations. An exploratory analysis of population genetic
structure was undertaken using principal coordinate analyses (PCoA)
based on between-plant pairwise genetic distances (ΦPT) using the
GenAlEx covariance standardized method with the first two dimen-
sions then plotted. These analyses identified divergent sampling lo-
cations in each species, which were removed and the analyses rerun.
STRUCTURE version 2.3. 2 (Pritchard et al., 2000) was used to de-
termine population structure without prior knowledge of population
affinities based on the admix ture model, a 50,00 0 burn-in followed
by 500,0 00 MCMC repetitions, a uniform prior for alpha, an initial
alpha of 1 and allele frequencies correlated among sampling loca-
tions. The optimal number of K-clusters was determined with the
ad hoc statistic ΔK (Evanno et al., 2005) using Struc ture Har vester
v0.6.6 ( Earl & vonHoldt, 2012) from five runs for each K = 2–1 0 .
The popgraph R package (Dyer, 2009) was used to create pop-
ulation graphs that described the distribution of genetic variation
among sampling locations for each species. This graph–theory ap-
proach simultaneously identifies genetic covariance structures
among subpopulations, does not assume a priori hierarchical or
bifurcating statistical models of population arrangement, and is in-
depe ndent of evolut io nar y assum pt ions th at aim to mi nimize Hard y–
Weinberg and linkage disequilibrium within populations (Dyer &
Nason, 2004). Populations are represented as nodes with node
diameter representing the level of within-site heterozygosity, lines
connecting nodes show populations that are not significantly genet-
ically differentiated with line length representing among-site genetic
variation (Dyer & Nason, 200 4). Paths connec ting populations were
also examined for “extended edges,” designating long-distance dis-
persal, and “compressed edges,” indicating topological or ecological
sources of vicariance, both of which were identified by chi-square
tests at α = 0.05.
2.4 | Resistance surface analysis
To assess the effect of landscape structure on genetic differentia-
tion, we estimated four explanatory variables for each species: (a)
“forest cover” based on a continuous forest cover map for 2000
(University of Maryland: http://earth engin epart ners.appsp ot.com/
scien ce-2013-globa l-fores t/downl oad.html) where every pixel con-
tained a value of forest cover [0,100], (b) “land cover” based on a
categorical Global Land Cover Map for 2009 (GlobCover: http://due.
esrin.esa.int/globc over) where every pixel contained a land cover
class code, (c) “water stream cover” based on categorical “Surface
Hydrology Lines” map of Australia (http://pid.geosc ience.gov.au/
datas et/ga/83130), and (d) elevation base d on a GEODATA 9 Second
digital elevation model (DEM) version 3 2008 for the main study
area (https://data.gov.au/data/datas et/0fc33 57c-5852-4e6b-992c-
c78bd 10e9234) where every pixel contained an elevation value ex-
pre ssed in meters. To minimize border effec ts, we cropped al l rasters
to the extent of the study region for both Acacia species and masked
them using a shape file of Australia. We also standardized by repro-
jecting (EPSG:4326) and aggregating all rasters to similar size grid
cells. Spatial data were prepared using the R packages raster (Hijmans
& van Etten, 2012) and rgdal (Bivand, Keitt, & Rowlingson, 2020).
The “forest cover” was aggregated by a factor of 3, using the mean
function, “land cover” was similarly aggregate by a factor of 30 and
reclassified as outlined below. The DEM was resc aled (min 0.0 01 –
max 1) and also aggregated by a factor of 3 using the mean func tion.
We used circuit theor y (McRae et al., 2008) to estimate the re-
sistance to gene flow between sampling locations of both species
for each of the explanatory variables (forest cover, land cover, water
stream cover, elevation) using Circuitscape v4.0 (McRae, 2006) to
estimate pairwise resistance distances. Because we hypothesized
higher gene flow across intact forest remnants than bet ween re-
gions predominantly covered by agricultural areas, we created resis-
tance sur faces where agricultural area pixels had higher resistance
values. We created two separate resistance surfaces: one using land
cover and one using forest cover maps. We used raw values of forest
cover rasters (Figure S5) and transformed the categorical values of
land cover rasters to numerical resistance values ranging between 0
and 1 (Figure S4). More specifically, we assigned a minimal resistance
of 0.1 to all forested land cover classes, medium resistance values
(0.4–0. 5) to areas containing fragmente d hab itats , and a maximal re-
sistance of 0.9 to all other classes (agricultural and permanent snow
areas). We also tested two more the hypotheses: (a) that elevation
influenced genetic connectivity with mountain ranges being a po-
tential barrier to Acacia gene flow and (b) that water streams might
be positively af fecting gene flow between Acacia populations (par-
ticularly A. stenophylla). To do this, we created resistance surfaces
with pixels at higher elevations having higher resistance values using
the raw elevations from the DEMs as resistance values for each pixel
(Figure S6) and we created a “water stream cover” conductance sur-
faces only taking major perennial watercourses with conductance
values of 1 (Figure S7). Finally, to test for isolation by geographic
distance (IBD), we created null-model rasters by replacing all values
of the forest cover rasters with 0.5 and calculated resistance dis-
tances between sampling locations. Because Circuitscape does not
accept zero resistance values, we replaced zero values in all rasters
with 0.0 001.
2.5 | Landscape genetic analysis
We used conditional genetic distance (Dyer & Nason, 20 04) as our
response variable and it was calculated using the R package gstu-
dio (Dyer, 2009). This is an interindividual genetic distance, which
13480
|
ENCINAS-VISO Et A l.
considers genetic covariation among all studied sampling locations.
As ex planatory var iab les, we used the various dist ance mat rices de-
scribed above: geographic distance, forest cover, land cover, water
stream cover, and elevation. We tested correlations among genetic
and environmental distances according to the different scenarios
considered using: (a) Mantel (Mantel, 1967) and partial Mantel tests
(i.e., to control for spatial autocorrelation using the geographic
distance matrix); and (b) multiple regression on distance matrices
(MRM, (Lichstein, 2007)). Mantel tests were applied using the R
package vegan (vs. 2-0-10) (Oks anen et al., 2017) and estimated sig-
nificance based on 10 000 permutations. MRM were implemented
in the R package ecodist vs. 1. 2-9 (Gos lee & Urba n, 20 07) by bu il d in g
an initial model. For each covariate, we then estimated regression
coefficients and associated p-val ues ba se d on 10 ,000 per mut ati o ns .
To select the model that best explained genetic differentiation, we
used a backward elimination model selection approach as described
in Legendre and Legendre (2012). The first model included all vari-
ables: geographic distance, forest cover, land cover, water stream
cover, and elevation. The variable showing the highest nonsignifi-
cant p-value was removed and we repeated this procedure until all
variables included in the analysis showed p-values lower than 0.05.
We corrected p-values obtained for Mantel tests and MRM mod-
els for multiple testing using the Benjamini and Hochberg (1995)
method as implemented in the stats package in R .
3 | RESULTS
3.1 | Genetic diversity and population genetic
structure
Genetic diversity measures varied among sampling locations within
species as well as between A. salicina and A. stenophylla. The per-
centage of polymorphic loci ranged from 36%–76.4% in A. salicina
and 46.2%–83.9% in A. stenophylla while expected heterozygosity
for A. salicina was 0.117–0.216 and 0.151–0.280 in A. stenophylla
(Table 1). Although differences were also evident at the species level
for mean percentage of polymorphic loci (55.1% for A. salicina vs.
64.3% for A. stenophylla) and mean expected heterozygosity (0.162
for A. salicina vs. 0.214 for A. stenophylla), slight differences in the
primer pair com bi nations used to gen er ate th e data se ts sug ge st cau-
tion whe n co mpari ng these re su lts . The AMOVAs indica te d su bst an-
tial and significant differences in genetic variation among sampling
locations of both species (A. salicina: ΦPT = 0.301, 3 0.1%, p = .010;
A. stenophylla: ΦPT = 0.235, 23.5%, p = .010). Significant isolation by
distance was also detected in both species (p < .001).
The first two principal coordinates axes accounted for 56.9%
of the total variation in A. salicina and 49.3% of the variation for A .
stenophylla (Figure S1). Divergent sampling locations were evident
for both of these PCos, namely, A. salicina plants from sampling loca-
tions 13–16 were located in negative PCo1 space (Figure S1a) and a
group of A. stenophylla plants from sampling locations 37, 38, 42, 44,
and 46 were in negative PCo1 and PCo2 space (Figure S1c).
The analyses done by STRUCTURE showed that most likely
there are six genetic clusters (K = 6) for bot h spe c ie s (Fig ure s 2 an d
3) based on the statistic ∆K, although there was some evidence
that a smaller number of clusters (i.e., 3) might also be present
(Figure S11a,b). There was little evidence of admixture in many of
the A. salicina plants (Figure 2 and Figure S2). For example, some
northern sampling locations (23, 24, 26, 27, and 29) were strongly
ass ociate d with cl us te r 1 (dark blu e); ho wever, some pla nt s i n sam-
pling location 24 had associations with cluster 4 (yellow) and 6
(red). In contrast, other A. salicina sampling locations (e.g., 2, 5, 6,
22, 25, and 28) showed evidence of admixture or potentially immi-
gration from sampling locations belonging to other groups. The A .
salicina clusters were somewhat geographically partitioned across
the study region (Figure 2 and Figure S9) with cluster 1 (dark blue)
found to the n or theast, clus ter 2 (light blue) restr icted to the most
southerly edge while cluster 3 (purple) was broadly distributed.
Clusters 4 (yellow) and 5 (green) were distributed broadly in the
southern half while the single cluster 6 (red) sampling location is
found to the northeast. Two sampling locations (22 and 28) were
a mixture of several clusters and could not be assigned to a sin-
gle cluster at >70%. Results from K = 3, the second highest ∆K,
show a clear geographic pattern of northern and southern clus-
ters (Figure S11a). Unlike A. salicina, many A. stenophylla sampling
locations were dominated by cluster 1 (red, Figure 3d) and many
sampling locations showed evidence of admix ture or immigration
(Figure 3 and Figure S3). Many sampling locations could not also
be assigned to a single cluster at >70%. These results are also
clearly obser ved for K = 3 (Figure S11b). Geographically the ma-
jority of sampling locations assigned to cluster 1 were located in
headwaters with sampling locations not easily assigned or those
belonging to other clusters loc ated downstream.
Popgraph highlighted a complex web of connected sampling lo-
cations (Figure 4, Figures S9 and S10). Few compressed edges indi-
cating ecological or topographical barriers to dispersal were evident
in A. salicina except among several sampling locations at the south-
ern edge of the study area and between sampling loc ations 21 and
24 (Figure 4a). Extended edges indicative of long-distance dispersal
linked many of the A. salicina sampling locations across broa d spatial
scales up to 500 km (Figure 4b). More compressed edges were ob-
served in A. stenophylla, again at the southern end of the study area
(Figure 4c) while extended edges were found across the whole of the
region (Figure 4d).
3.2 | Landscape genetics analysis: Mantel tests and
multiple regression on distance matrices (MRM)
We used resistance surfaces of each environmental variable calcu-
lated by Circuitscape for our landscape genetic analysis on each spe-
cies. We did not find any significant relationship between genetic
clusters and the different resistance surfaces (LC, FC, and DEM) for
A. stenophylla and A. salicina, except for forest cover effects on ge-
netic clusters of A. salicina (Kruskal–Wallis χ2 = 12.657, p = .048).
|
13481
ENCINA S-VISO Et Al .
More specifically, the admixed cluster (white) consisting of two sam-
pling locations was mostly related to areas of high forest cover re-
sistance (Figure S7a).
Mantel tests showed that the genetic structure of A. salicina
was affected by both geographic distance and land cover resistance
(p < . 0 5), wh ile fo r A. stenophylla, o n ly la nd cove r re sis t a nce ap p eare d
to be important (Table 2). After correcting by spatial autocorrelation
(partial Mantel tests), none of these fac tors had a significant ef fect
on either of the Acacia species (Table 2). However, the MRM model
considering only geographic distance was the best model explain-
ing genetic variation between sampling locations in Acacia salicina
whereas the MRM model including only land cover and elevation
was the best model explaining sampling location genetic differentia-
tion for Acacia stenophylla (Table 3, Table S1).
4 | DISCUSSION
The MDB is Australia's longest river system (area of ~1.0 million
square kilometers) and a vital economic resource for the agricul-
tural industry (Cai & Cowan, 2008); however, it is also considered
one of Australia's most impacted ecosystems (Cole et al., 2016).
Disturbed landscapes, such as the MDB, alter spatial struc ture
and affect plant demography by decreasing population size and
increasing population isolation due to geographical distance or
barriers in the landscape (Kwak et al., 2009). While habitat frag-
mentation is expected to diminish gene flow between local popu-
lations, ultimately affecting patterns of genetic differentiation and
viabilit y (Ellstrand, 1992; Kwak et al., 2009; Young et al., 1996),
some evidence suggest s that a lack of inter vening vegetation may
A. salicina A. stenophylla
Site No. N%P He (SE)
Site
No. N%P He (SE)
110 40.0% 0.117 (0.010) 512 51.1 % 0.175 (0.013)
2 8 45.5% 0.151 (0.012) 30 10 52.0% 0.195 (0.014)
318 49. 5% 0.140 (0.011) 31 21 68.2% 0.228 (0.013)
423 42.9% 0.121 (0.010) 32 22 47.5 % 0.166 (0.013)
517 52.0 % 0.153 (0.011) 33 24 5 9.6% 0.190 (0.013)
614 42.2% 0.120 (0.011) 34 30 60.5% 0.188 (0.013)
821 65.8% 0.192 (0.011) 35 20 4 7.1 % 0.151 (0.013)
928 53.5% 0.150 (0.011) 36 25 73.1% 0.245 (0.013)
10 29 55.6% 0.154 (0.011) 37 20 71.3% 0.243 (0.013)
11 15 36.0% 0.118 (0.011) 38 21 73.5% 0.248 (0.013)
12 10 37. 8 % 0.126 (0.011) 39 29 64.6% 0.211 (0.013)
13 18 67. 3 % 0.216 (0.012) 40 26 61.0% 0.190 (0.013)
14 18 52.0 % 0.166 (0.012) 41 24 68.2% 0.233 (0.013)
15 30 68.0% 0.211 (0.012) 42 24 83.9% 0.275 (0.013)
16 17 62.9% 0.210 (0.012) 43 24 73.5% 0.241 (0.013)
17 23 76. 4% 0.203 (0.011) 44 16 54.3% 0.198 (0.014)
18 19 62.5% 0.174 (0.011) 45 18 67.7 % 0.230 (0.013)
19 21 56.0% 0.159 (0.011) 46 21 78.0% 0.280 (0.013)
20 21 52.7% 0.156 (0.011) 47 24 71.3% 0.215 (0.013)
21 26 42.5% 0.129 (0.011) 48 746.2% 0.168 (0.014)
22 25 62.5% 0.185 (0.012) 49 26 70.9% 0.224 (0.013)
23 28 60.4% 0.163 (0.011) 50 22 66.4% 0.215 (0.013)
24 19 57. 8 % 0.169 (0.011) 51 27 65.5% 0.217 (0.013)
25 24 66.9% 0.194 (0.011) 52 19 67. 7 % 0.232 (0.014)
26 26 56.7% 0.161 (0.011) 53 29 71.3% 0.216 (0.012)
27 19 49. 8% 0.152 (0.011) 54 22 69.1% 0.233 (0.014)
28 28 67. 3 % 0.194 (0.011) 55 19 59.2 % 0.202 (0.014)
29 30 59. 6% 0.155 (0.011) 56 24 63.7% 0.217 (0.013)
Mean 55.1% 0.162 (0.002) 57 29 65.5% 0.208 (0.013)
58 26 56.1% 0.181 (0.013)
Mean 64.3% 0.214 (0.001)
TABLE 1 Genetic diversity measures
for A. salicina and A. stenophylla. N,
number of plants; %P, percentage
of polymorphic loci; He, expected
heterozygosity. Bold values show mean %
P and He values per species
13482
|
ENCINAS-VISO Et A l.
FIGURE 2 STRUCTURE output of A.
salicina for K = 6. Geographical location
of A. salicina sampling locations colored
to match K = 6. Sampling location
numbers as per Table 1. Colors distinguish
the genetic clusters inferred from
STRUCTURE
FIGURE 3 STRUCTURE output of
A. stenophylla for K = 6. Geographical
location of A. stenophylla sampling
locations colored to match K = 6.
Sampling location numbers as per Table 1.
Colors distinguish the genetic clusters
inferred from STRUCTURE
|
13483
ENCINA S-VISO Et Al .
actually increase gene flow (Sork & Smouse, 2006). Our result s
show that A. salicina and A. stenophylla species had distinct pat-
terns of genetic differentiation among populations and that similar
element s of landscape structure had an impor tant influence on
the observed genetic variation.
Both STRUCTURE and genetic diversity analyses show that A.
salicina had high genetic differentiation with six distinct genetic clus-
ters. Population graph analysis and PCoA fur ther confirmed these
results, showing the effect on genetic differentiation by geographic
barriers that separate northern and southern sampling locations.
However, STRUCTURE analysis might be overestimating genetic
structure of A. salicina due to high levels of IBD (Frantz et al., 2009).
In contrast, A. stenophylla showed high levels of admixture in many
sampling locations dominated by a single cluster (1, red) from six
genetic clusters detected by STRUC TURE. This suggests that gene
flow of A . stenophylla is relatively high across many northern and
southern sampling locations within its geographical range. However,
few sampling locations were composed of distinc t genetic clusters
[e.g., cluster 3 (green) only present in the south and cluster 4 (black)
in the far southwest (see Figure 3). Our findings thus reveal that,
contrar y to A. salicina, A. stenophylla is able to maintain higher gene
flow across large distances (at least 300 km). Although we did not
find an effect of hydrological connectivity, this partially confirms
results from a previous study of population genetics of A. steno-
phylla, whic h sh ows high leve ls of po pu la ti on con ne ctivity acros s the
northwest region of the MDB mainly driven by hydrochory (Murray
et al., 2019). Seed dispersal through water streams can occur over
large distances in this species (>100 km) (Murray et al., 2019). The
high levels of genetic connectivity in A. stenophylla also provide
some evidence showing that gene flow could be maintained in trees
despite the landscape fragmentation caused by agricultural produc-
tion (Byrne et al., 20 07).
FIGURE 4 Visualization of A. salicina Popgraph compressed (A) (red dashed lines) and extended (B) (black dashed lines) edges for and
A. stenophylla compressed (C) (red dashed lines) and extended (D) (black dashed lines) edges. Geographical location of A. stenophylla and A.
salicina sampling locations colored to match K = 6. White nodes indicate < 70% assignment to a single cluster
13484
|
ENCINAS-VISO Et A l.
Our spatial analysis suggest s that distinc t environmental fac-
tors influence genetic differentiation in the two studied species.
Geographic distance is the main factor determining A. salicina ge-
netic structure, which confirms the findings by STRUCTURE and
the visualization of the population graph. Interestingly, this means
that habitat fragmentation measured by land cover and forest cover
in this area does not seem to have a strong effect on gene flow.
Geographical barriers, such as presence of major rivers (Darling
and Murray rivers) and geographical distance, seem to shape the
observed pattern of genetic variation. However, model fitting
(R2 = 0.03) was poor, suggesting that other environmental factors,
not considered here, might have an effect on the genetic connec-
tivity of A. salicina. For example, extensive drought s in the Murray
Darling Basin area are known to have an impact on Acacia demogra-
phy (Godfree et al., 2019). However, population graph analysis indi-
cates that gene flow between populations is relatively high based on
the number of compressed and extended edges across A. salicina's
range. Thus, the results overall suggest that gene flow is not affected
by habitat fragmentation and it is likely that this species, which is a
successful colonizer of disturbed areas (Grigg & Mulligan, 1999), is
fairly resilient to fragmentation because it has large seed banks, high
growth rate and tolerance to bare soil (Jeddi & Chaieb, 2012).
In the case of A. stenophylla, habitat fragmentation (predicted by
land cover) did have a significant effect on genetic structure. Thus,
genetic connectivity among A. stenophylla sampling locations does
not seem to be mainly driven by geographic distance (contrary to A.
salicina) and MRM showed that elevation might also act as a poten-
tial barrier for gene flow between sampling locations. Interestingly,
given the high levels of admixture in many sampling locations of A.
stenophylla, the effects of land cover and elevation are likely to ex-
plain the presence of distinct genetic clusters in southern sampling
locations (green, yellow, and light blue clusters; see Figure 4c,d)
where there is high habitat fragmentation produced by extensive
areas of agricultural land (Figure S4). Despite these factors, genetic
structure and diversity analysis shows that this species is largely
unconstrained across its range with long-distance seed movement
possibly helping to maintain homogeneity among sampling loca-
tions, especially within rivers and tributaries as it has been shown
in a previous study (Murray et al., 2019). Although we did not find
a significant effect of water stream cover, occasional one-direc-
tional long-distance seed dispersal may par tly explain high levels of
admixture in many sampling locations located at the river margins
(Figures 1 and 3).
Although high levels of gene flow have been found in the north-
ern MDB (Upper Murray River) in A. stenophylla (Mur ray et al., 2019),
we found that sampling locations in the southern MDB (Lower
Murray River) are more structured with lower levels of gene flow
(Figures 3 and 4). Evidence from a freshwater fauna study suggests
that population divergence and differentiation between Upper and
Lower Murray River are recent (~125 years) and likely induced by
anthropogenic disturbance (Cole et al., 2016). Water streams of the
MDB have been heavily managed for agricultural purposes since
European settlement and, in recent times, are known to suffer from
lower and more even flow volumes (Adamson et al., 2009; Cai &
Cowan, 2008). This suggests that the population connectivity of A.
stenophylla may now be partially affected by severe water flow fluc-
tuations and management, particular in the Lower Murray River, of
the MDB (Oliver & Merrick, 2006).
Several studies have pointed out the negative impacts of habi-
tat fragmentation on plant population viabilit y and genetic diversit y
(Millar et al., 2014; Young et al., 1996). Our re sults suggest that a de-
crease of forested areas can significantly alter genetic differentiation
for A. stenophylla, but our results do not support that for A. salicina.
Management actions to improve connectivity of these species (in-
cluding through water management) need to be tailored accordingly
based on our findings. For example, southern sampling locations of
A. stenophylla who continue to suffer severe water fluctuations that
TABLE 2 Result s of Mantel and Partial Mantel correlation tests
for A. salicina and A. stenophylla. For each test, rM is provided.
Significant results (p-value < .05) after p-value correction are
shown in bold and italicized. The variables considered are
geographic distance (GD), elevation (DEM), forest cover (FC), water
stream cover (WC), and land cover resistance (LC)
Species
A. salicina A. stenophylla
Mantel tests
GD 0.175 0.035
DEM −0.212 −0 .217
FC 0.077 −0.093
LC 0.205 0.152
WC −0.069 0.041
Partial Mantel tests
DEM −0.402 −0.26
FC 0.068 −0.158
LC 0.171 0.179
WC −0.062 0.043
Species R2FGD DEM FC LC WC
A. salicina 0.03 12.01 2.38e−05
(p = .004)
– – -– –
A. stenophylla 0.117 28.85 –275.996
(p = .019)
–39. 81
(p = .013)
–
TABLE 3 Best MRM models obtained
with back ward elimination for both A .
salicina and A. stenophylla. The variables
considered are geographic distance (GD),
elevation (DEM), forest cover resistance
(FC), water stream cover ( WC), and land
cover resist ance (LC). Bold text shows
significant values (p-value < .05)
|
13485
ENCINA S-VISO Et Al .
will decrease seed dispersal between populations might be bene-
fited by translocation of individuals where we can target populations
with relatively large effective sizes and that are relatively well con-
nected by gene flow to other large populations.
This study provides an important contribution to understanding
patterns of genetic differentiation for key plant restoration species
in Australia across an important agro-ecosystem region. Our results
show that (a) both species had relatively high levels of genetic diver-
sity and dif ferentiation; (b) both species also had high levels of ge-
netic structure across the MDB, although A. stenophylla also showed
high admixture levels in several sampling loc ations; and (c) habitat
fragmentation and elevation do not equally af fect the genetic con-
nectivity of these two woody legumes supporting our hypothesis.
While it seems that A. salicina genetic differentiation and connec-
tivity are mainly driven by geographic distance, anthropogenic dis-
turbances in the MDB do have an important impact on gene flow
in A. stenophylla and it is likely that it affects other less resilient
plant species in the region (for example, wetland specialists (Colloff
et al., 2014)). Previous studies show that severe impac t it is already
occurring in freshwater fauna in the MDB (Chessman, 2011; Cole
et al., 2016) augmented by the increasing effects of climate change
(Adamson et al., 20 09; Balcombe et al., 2011). We also suggest that
this work could ser ve as a reference for studies aiming to assess the
importance of their associated legume symbionts (nitrogen-fixing
rhizobial bac teria) (Thrall et al., 2005, 2007) to understand how their
composition and genetic variation across large geographic scales
might be associated with the survival and reproduction of Acacia
species.
ACKNOWLEDGMENTS
This research was jointly funded by CSIRO and the New South Wales
Environment Trust. The authors would like to thank Jo Slattery and
Jacqui McKinnon for field assistance and for bacterial isolation and
purification and Michelle Watt and Michael Grossman for construc-
tive comments on the original manuscript.
CONFLICT OF INTEREST
None declared.
AUTHOR CONTRIBUTION
Francisco Encinas-Viso: Conceptualization (lead); Formal analysis
(lead); Investigation (lead); Methodolog y (lead); Project administra-
tion (equal); Software (lead); Supervision (lead); Validation (equal);
Visualization (equal); Writing-original draft (lead); Writing-review &
editing (lead). Christiana McDonald-Spicer: Data curation (equal);
Formal analysis (equal); Investigation (equal); Methodology (equal);
Visualization (equal); Writing-review & editing (equal). Nunzio
Knerr: Data curation (equal); Formal analysis (equal); Investigation
(equal); Software (equal); Visualization (lead); Writing-review &
editing (equal). Peter Thrall: Conceptualization (equal); Data cu-
ration (equal); Funding acquisition (lead); Investigation (equal);
Methodology (equal); Project administration (equal); Resources
(lead); Super vision (equal); Validation (equal); Writing-review &
editing (equal). Linda Broadhurst: Conceptualization (lead); Data
curation (equal); Formal analysis (supporting); Funding acquisition
(lead); Investigation (equal); Methodolog y (equal); Project adminis-
tration (lead); Resources (lead); Software (supporting); Supervision
(equal); Validation (equal); Visualization (supporting); Writing-
original draft (supporting); Writing-review & editing (equal).
DATA AVAIL AB ILI T Y STAT EME N T
Genetic data of both Acacia species will be deposited at CSIRO Data
Access Portal.
ORCID
Francisco Encinas-Viso https://orcid.org/0000-0003-0426-2342
REFERENCES
Adamson , D., Mallawaarachchi, T., & Quiggin, J. (2009). Declining in-
flows and more frequent droughts in the Murray-Darling Basin:
Climate change, impacts and adaptation*. Australian Journal of
Agricultural and Resource Economic s, 53, 345–366. ht tps://doi.
org /10.1111/j .1467-84 89.2 009.0 0451.x
Aguilar, R., Ashwor th, L., Galetto, L., & A izen, M. A . (20 06). Plant repro-
duction susceptibility to habitat fragmentation: Review and synthe-
sis through a meta-analysis. Ecology Letters, 9, 968–980.
Aguilar, R., Quesada, M., Ashworth, L., Herrerias-Diego, Y., & Lobo, J.
(2008). Genetic consequences of habitat fragmentation in plant
populations: Susceptible signals in plant traits and methodolog-
ical approaches. Molecular Ecology, 17, 5177–5188 . http s://do i.
org /10.1111/j .136 5-294X. 20 0 8. 03971. x
Balcomb e, S. R., Sheldon, F., Capon, S. J., Bond, N. R., Hadwen, W. L.,
Marsh, N., & Bernays, S. J. (2011). Climate-change threat s to native
fish in degraded rivers and floodplains of the Murray-Darling Basin,
Australia. Marine and Freshwater Research, 62, 1099–1114. http s://
doi.org/10.1071 /MF11059
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery
rate: A practical and power ful approach to multiple testing. Journal
of the Royal Statistical Society: Series B (Methodological), 57, 289–300.
https://doi.org/10.1111/j.2517-6161.1995.tb020 31.x
Bivand, R., Keitt, T., & Rowlingson, B. (2020). rgdal: Bindings for the
'Geospatial' Data Abstraction Librar y. R package version 1.5-16.
https://CRAN.R-proje ct.org/packa ge=rgdal
Broadhurst, L., Breed, M., Lowe, A., Bragg, J., Catullo, R., Coates, D.,
Encinas-Viso, F., Gellie, N., James, E., Krauss, S., Pot ts, B., Rossetto,
M., Shepherd, M., & Byrne, M. (2017). Genetic diversity and str uc-
ture of the Aus tralian flora. Diversity and Distributions, 23(1), 41–52.
https ://doi.or g/10.1111/d di.125 05
Bui, E. N., Thornhill, A., & Miller, J. T. (2014). Salt- and alkaline-tolerance
are linked in Acacia. Evolutionary Biology, 10, 20140278.
Byrne, M., Elliott, C. P., Yates, C., & Coates, D. J. (2007). Extensive pollen
dispersal in a bird-pollinated shrub, Calothamnus quadrifidus, in a
fragmented landscape. Molecular Ecology, 16, 1303–1314.
Cai, W., & Cowan, T. (2008). Evidence of impact s from rising temperature
on inflows to the Murray-Darling Basin. Geophysical Research Letters,
35, L07701. htt ps://doi.org/10.1029/2008G L033390
Chessman, B. C. (2011). Declines of freshwater tur tles associated with
climatic drying in Australia’s Murray-Darling Basin. Wildlife Research,
38, 664–671. https://doi.org/10.1071/WR11108
Cole, T. L., Hammer, M. P., Unmack, P. J., Teske, P. R., Brauer, C. J., Adams,
M., & Beheregaray, L. B. (2016). Range-wide fragmentation in a
threatened fish associated with post-European set tlement modifica-
tion in the Murray-Darling Basin, Australia. Conservation Genetics, 17,
1377–1391. ht tps ://doi.org/10.1007/s1059 2-016- 08 68- 8
13486
|
ENCINAS-VISO Et A l.
Colloff, M. J., Ward, K. A., & Roberts, J. (2014). Ecology and conser va-
tion of gr assy wetlands dominated by spiny mud grass Pseudoraphis
spinescens in the southern Murray-Darling Basin, Australia. Aquatic
Conservation: Marine and Freshwater Ecosystems, 24, 238–255.
Doi, R., & Ranamukhaar achchi, S. L . (2013). Slow restor ation of soil mi-
crobial functions in an Acacia plantation established on degraded
land in Thailand. International Journal of Environmental Science
and Technology, 10, 623–634. https://doi.org/10.1007/s1376
2-012-0165-0
Dyer, R. J. (20 09). GeneticStudio: A suite of programs for spatial analysis
of genetic-marker. Molecular Ecology Resources, 9, 110–113.
Dyer, R. J., & Nason, J. D. (2004). Population graphs: The graph theoretic
shape of genetic structure. Molecular Ecology, 13, 1713–1727. https://
doi.org/10.1111/j.1365-294X.2004.02177.x
Earl, D. A . (2011). Structure harves ter v0.6.1.
Earl, D. A., & vonHoldt, B. M. (2012). S TRUCTURE HARVESTER: A web-
site and progr am for visualizing STRUC TURE output and implement-
ing the Evanno method. Conservation Genetics Resources, 4, 359–361.
htt ps://doi.org/10.10 07/s126 8 6- 011-9548-7
Ells t ra nd, N. C. (1992). Gene flow by po lle n: Impl ica tio ns for pl ant con ser-
vation genetics . Oikos, 63, 77–86. https://doi.org/10.2307/3545517
Evanno, G., Regnaut, S., & Goudet , J. (2005). Detecting the num-
ber of clusters of individuals using the soft ware struc ture: A
simulation study. Molecular Ecology, 14, 2611–2620. ht tps://d oi.
org/10.1111/j.1365-294X.2005.02553.x
Excoff ier, L., Smouse, P., & Quattro, J. (1992). Analysis of molecular
variance inferred from metric distances among DNA haplotypes:
Applic ation to human mitochondrial DNA restriction data. Genetics,
131, 479–491.
Fenderson, L. E., Kovach, A. I., & Llamas, B. (2020). Spatiotemporal land-
scape genetics: Investigating ecology and evolution through space
and time. Molecular Ecology, 29, 218–246. htt ps://doi. org/10 .1111/
mec.15315
Frankham, R., Ballou, J. D., & Briscoe, D. A. (2010). Introduction to conser-
vation genetics, 2nd ed. Cambridge Universit y Press.
Frantz, A. C., Cellina, S., Krier, A., Schley, L ., & Burke, T. (2009). Using
spatial Bayesian methods to determine the genetic structure of
a continuously distributed population: Clusters or isolation by
dis tance? Journal of Applied Ecology, 46, 493–505. https://doi.
org /10.1111/j .136 5-2664 .2 00 8.01606.x
Go df ree, R. C., Kn err, N., Go df ree , D. , Bu sby, J., Rob ert son, B. , & En ci nas-
Viso , F. (2019) . Hi st or ica l re co ns truc tion unveils the ris k of mas s mor-
tality and ecosystem collapse during pancontinental megadrought.
Proceedings of the National Academy of Sciences of the United States of
America, 116(31), 15580–15589.
Goslee, S., & Urban, D. (2007). The ecodist package for dissimilari-
ty-based analysis of ecologic al data. Journal of Statistical Software,
22, 1–19.
Grigg, A. H., & Mulligan, D. R. (1999). Biometric relationships for esti-
mating standing biomass, Litterfall and litter accumulation of Acacia
salicina on mined land in central Queensland. Australian Journal of
Botany, 47, 807–816. https://doi.org /10.1071/BT97122
Hamrick, J. L . (2004). Response of forest trees to global environment al
changes. Forest Ecology and Management, 197, 323–335. ht tps://doi.
org/10.1016/j.foreco.2004.05.023
Hijmans, R. J., & van Etten, J. (2012). raster: Geographic analysis and
modeling with raster dat a. R pack age version 2.0-12.
Jeddi, K ., & Chaieb, M. (2010). Changes in soil properties and vegetation
following livestock grazing exclusion in degraded arid environments
of South Tunisia. Flora - Morpho logy, Distributi on, Functional Ecol ogy of
Plants, 205(3), 184–189.
Jeddi, K., & Chaieb, M. (2012). Restoring degraded arid Mediterranean
areas with exotic t ree species: Influence of an age sequence of
Acacia salicina on soil and vegetation dynamics. Flora - Morpholog y,
Distribution, Functional Ecology of Plants, 207(9), 693–700.
Kwak, M. M., Velterop, O., & Andel, J. (2009). Pollen and gene flow in
fragmented habitats. Applied Vegetation Science, 1, 37–54. https://
doi .or g/10 .2307/147908 4
Legendre, P., & Legendre, L . (2012). Chapter 1 - Complex ecological data
sets. In P. Legendre, & L.- B.- T.-D. Legendre (Eds.), Numerical ecology
(1–57 ). Elsevier.
Lichstein, J. W. (2007). Multiple regression on distance matrices: A multi-
variate spatial analysis tool. Plant Ecology, 188, 117–131. https://doi.
org /10.1007/s1125 8-0 06 -9126 -3
Mantel, N. (1967). The detection of disease clustering and a generalized
regression approach. Cancer Research, 27, 209–220.
Maslin, B. R., Miller, J. T., & Seigler, D. S. (20 03). Overview of the generic
status of Acacia (Leguminosae: Mimosoideae). Australian Systematic
Botany, 16, 1–18.
McRae, B. H. (2006). Isolation by resistance. Evolution, 60, 1551–1561.
McRae, B. H., Dickson, B. G., Keitt, T. H., & Shah, V. B. (2008). Using cir-
cuit theory to model connectivity in ecology, evolution, and conser-
vation. Ecology, 89, 2712–2724. https://doi.org/10.1890/07-1861.1
Millar, M . A ., Coates, D. J., & Byrne, M. (2014). Extensive long-distance
pollen dispersal and highly outcrossed mating in historically small
and disjunct populations of Acacia woodmaniorum (Fabaceae), a
rare banded iron formation endemic. Annals of Botany, 114, 961–971.
https://doi.org/10.1093/aob/mcu167
Murray, B. F., Reid, M. A., Capon, S. J., Thoms, M., & Wu, S.-B. (2019).
Gene flow and genetic structure in Acacia stenophylla (Fabaceae):
Effects of hydrological connectivit y. Journal of Biogeography, 46,
1138–1151.
O’Dowd, D. J., & Gill, A . M. (1986). Seed dispersal syndromes of
Australian Acacia. In D. Murray (Ed.), Seed dispersal (pp. 87–121) . Ne w
York: Academic Press.
Oksanen, J., Blanchet, F. G., Friendly, M., Kindt , R., Legendre, P., McGlinn,
D., O’Har a, R. B., Leslie Simpson, G., Solymos, P., Stevenes, M. H. H.,
& Wagner, H. H. (2017). vegan: Communit y Ecolog y Package.
Oliver, R. L., & Merrick, C. J. (2006). Partitioning of river metabolism
identif ies phy toplankton as a major contributor in the regulated
Murray River (Australia). Freshwater Biology, 51 , 1131–1148. ht tps://
doi .org/10.1111/j.1365 -2427.200 6. 015 62.x
Peakall, R., & Smouse, P. E . (20 06). GENALEX 6: G enetic analy-
sis in Excel. Population genetic sof tware for teaching and re-
search. Molecular Ecology Notes, 6, 228–295. https://doi.
org /10.1111/j .1471-8286 .2 005. 01155 .x
Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of pop-
ulation structure using multilocus genotype data. Genetics, 155,
945–9 59.
Ritland, K. (1989). Genetic differentiation, diversity, and inbreeding in
the mountain monkeyflower (Mimulus caespitosus) of the Washington
Cascades. Canadian Journal of Botany, 67, 2017–2024.
Sork, V. L., & Smo use , P. E. (20 06). Genetic analysis of la nds cape connec-
tivity in tree populations. Landscape Ecology, 21, 821–836. https://
doi.org/10.10 07/s1098 0-0 05-5 415-9
Spooner, P. G. (2005). Response of Acacia species to disturbance
by roadworks in roadside environments in southern New South
Wales, Australia. Biological Conservation, 122, 231–242. https://doi.
org/10.1016/j.biocon.2004.07.012
Thrall, P. H., Millsom, D. A., Jeavons, A . C., Waayers, M., Harvey,
G. R., Bagnall, D. J., & Brock well, J. (2005). Seed inoculation
with effective root-nodule bacteria enhances revegetation
success. Journal of Applied Ecology, 42, 740–751. h t tps://doi .
org/10.1111/j.1365-2664.2005.01058.x
Thrall, P. H., Slat tery, J. F., Broadhurst, L. M., & Bick ford, S. (20 07).
Geographic patterns of symbiont abundance and adaptation in na-
tive Australian Acacia-rhizobia interactions. Journal of Ecology, 95,
1110–1122 . ht tps://doi.or g/10.1111/j.1365-2745. 20 07.01278 .x
Vos, P., Hogers, R., Bleeker, M., Reijans, M., Lee, T. V. D., Hornes, M.,
Friters, A., Pot, J., Paleman, J., Kuiper, M., & Zabeau, M. (1995). AFLP:
|
13487
ENCINA S-VISO Et Al .
A new technique for DNA fingerprinting. Nucleic Acids Research, 23,
4407–4414. https://doi.org/10.1093/nar/23.21.4407
Vranckx, G., Jacquemyn, H., Muys, B., & Honnay, O. (2012). Meta-
analysis of susceptibility of woody plants to loss of genetic diversity
through habitat fragmentation. Conservation Biology, 26, 228–237.
https ://doi.or g/10.1111/ j.1 523-1739.20 11.01778. x
Wandrag, E. M., Dunham, A. E., Miller, R. H., & Rogers, H. S. (2015).
Vertebrate seed dispersers maintain the composition of tropical for-
est seedbanks. AoB Plants, 7, plv130. https://doi.org/10.1093/aobpl
a/pl v13 0
Ward, P. S., & Branstetter, M. G. (2017). The acacia ants revisited: con-
vergent evolution and biogeographic context in an iconic ant/plant
mutualism. Proceedings of the Royal Society B: Biological Sciences, 284,
20 16 256 9.
Young, A. G., Boyle, T., & Brown, T. J. B. (1996). The population ge-
netic consequences of habitat fragmentation for plants. Trends in
Ecology & Evolution, 11 , 413–417. htt ps://doi.org/10.1016/0169-
5347(96)100 45 -8
Young, H., Fedigan, L. M., & Addicott, J. F. (2008). Look before leap-
ing: Foraging selec tivity of capuchin monkeys on acacia trees in
Costa Rica. Oecologia, 155, 85–92. https://doi.org /10.1007/s0044
2-007-0883-z
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section.
How to cite this article: Encinas-Viso F, McDonald-Spicer C,
Knerr N, Thrall PH, Broadhurst L. Different landscape effects
on the genetic structure of two broadly distributed woody
legumes, Acacia salicina and A. stenophylla (Fabaceae). Ecol Evol.
2020;10:13476–13487. https://doi.or g/10.1002/ece3.6952
Available via license: CC BY
Content may be subject to copyright.
