Several Australian arid-zone acacias are threatened by habitat loss, degradation, and fragmentation resulting from agricultural activities and exotic herbivores (Morton et al., 1995), although others, including Acacia ligulata A. Cunn. ex Benth., are thriving. Two long-lived and potentially clonal species facing a variety of potential threats are A. melvillei Pedley and A. pendula A. Cunn. ex G. Don. Both of these latter species likely suffer from infrequent seed production and chronic recruitment failure (Batty and Parsons, 1992). Moreover, there is some debate about the origin and taxonomy of stands of A. pendula found in the Hunter region of New South Wales (Bell et al., 2007), the extreme eastern range edge of its distribution and a notable anomaly for this species, given its predominate semi-arid/arid distribution in four Australian states. A clear understanding of the factors underlying the variation in the performance of these three species is hampered by a lack of genetic tools that allow assessment of the mating and dispersal and genetic diversity of remaining stands.

The three target species have partially overlapping ranges. “Acacia melvillei shrubland” endangered ecological community occurs in semiarid and arid eastern Australia. This community is considered threatened primarily because of senescence of the overstory (dominated by A. melvillei), infrequent seed set, and recruitment failure due to overgrazing (NSW Scientific Committee, 2008). Acacia pendula is more widespread, occurring throughout the eastern semiarid zone, but is considered threatened within the Hunter Valley (NSW Scientific Committee, 2008). In contrast, A. ligulata is one of the most widespread Acacia species, occurring throughout arid Australia. Seed set occurs annually in this species, recruits are common (personal observation), and most stands appear to be thriving (personal observation). For each of these species, we developed primers that amplify microsatellite loci. By comparing and contrasting the genetic structure of populations of these species with partially overlapping distributions and perceived variation in reproductive success, we aim to gain insights into the impact of anthropogenic disturbance on their genetic structure and diversity and, together with demographic assessments, will seek to use these data to predict the resilience of remaining stands.

METHODS AND RESULTS

We used GS FLX Titanium sequencing (Roche Diagnostics Corporation, Sydney, Australia) to generate databases of DNA sequences for A. melvillei and A. pendula. Specimens of each species were sourced from stands located in western New South Wales. Genomic DNA was extracted using a DNeasy Plant Mini Kit (QIAGEN, Melbourne, Australia). Multiple DNA extracts from the same individual were pooled to obtain 5 µg of high-molecular-weight DNA for library construction. The library was prepared in accordance with the manufacturer's instructions (Roche Diagnostics Corporation), and the sequencing was performed at the Otago Genomic Sequencing Unit, University of Otago, New Zealand, using the GS FLX system with the GS FLX Titanium Rapid Library Preparation Kit (catalog no. 05608228001; Roche Diagnostics Corporation). Upon receipt of the DNA sequence databases from the University of Otago, we used the program MSATCOMMANDER version 0.8.1 (Faircloth, 2008) to detect DNA sequences containing di-, tri-, and tetranucleotide repeats, and to design microsatellite primers for PCR assays.

Table 1.

Novel microsatellite loci for Acacia melvillei, A. ligulata, and A. pendula.^a

To PCR amplify loci of interest, we used Multiplex-Ready Technology. This method was developed by Hayden et al. (2008) and is briefly described below. For each species, 24 locus-specific primer sets were synthesized by Sigma-Aldrich (Sydney, Australia). We also made use of existing primers (obtained in the same way) that amplify microsatellite loci in A. carneorum Maiden and A. loderi Maiden (Roberts et al., 2013) to potentially increase the number of microsatellites available for use in A. melvillei, A. pendula, and A. ligulata. Each respective forward and reverse primer had the nucleotide sequence 5′-ACGACGTTGTAAAA-3′ and 5′-CATTAAGTTCCCATTA-3′ attached to its 5′-end. Tag primers, tagF (5′-ACGACGTTGTAAAA-3′) and tagR (5′-CATTAAGTTCCCATTA-3′), were also synthesized, with tagF 5′-end labeled with one of Applied Biosystems' (Carlsbad, California, USA) proprietary fluorescent dyes (VIC, FAM, NED, and PET). Each PCR assay contained 0.2 mM dNTP, 1× ImmoBuffer (Bioline, Alexandria, Australia), 1.5 mM MgCl₂, 100 ng/µL bovine serum albumin (BSA; Sigma-Aldrich), 75 nM each of dye-labeled tagF and unlabeled tagR primer, 0.15 units of Immolase DNA polymerase (Bioline), and 2 µL of genomic DNA (∼10 ng/µL). The optimal primer concentration of each forward and reverse locus-specific primer was determined in preliminary PCR assays varying the primer concentration between 5 and 120 nM (Table 1) and also was included within each 10 µL (total volume) assay. PCRs were conducted on either a Bio-Rad (Hercules, California, USA) or Eppendorf (Hamburg, Germany) thermocycler with a denaturing step at 95°C, primer annealing step of 63°C, and an extension step at 72°C repeated for 40 cycles. Genomic DNA was extracted from phyllodes from one individual from each of five stands across the range of each species using a standard cetyltrimethylammonium bromide (CTAB) method (Doyle and Doyle, 1987). For each species, we genotyped eight individuals separated by at least 10 m, from each of five stands separated by at least 30 km. This initial sampling allowed us to assess levels of polymorphism within and between stands, before primers were deemed sufficiently polymorphic to characterize population genetic structure.

We developed new polymorphic primers that had consistently clean profiles, six each for A. melvillei and A. ligulata, and five for A. pendula (Table 1). We were also able to cross-transfer 15 previously optimized loci, 11 of which are described in Roberts et al. (2013). Specifically, five of 11 primer sets amplified successfully and had equally clear profiles on electropherograms for A. melvillei (DCL0C, AO35A, DSGN5, BNQS6, and DZ709), A. ligulata (A4IKI, AQBUV, DCL0C, ARU19, and C03P6), and A. pendula (ACPU7, BAIR8, BBY8P, C5IMO, and DCLOC), respectively. This resulted in a total of 11 working primers each for A. melvillei and A. ligulata, and 10 for A. pendula. All other primers tested did not amplify consistently or were difficult to score because of complex stuttering of the amplified product. These primer sets were discontinued. Combinations of successful primers were trialed together in multiplex PCRs to look for repeatable and clean assays. Successful combinations of primers as multiplex PCRs, which were subsequently used for all further genotyping, are presented in Table 2.

Table 2.

Multiplex PCR combinations achieved and fluorescent dyes used. Primers listed in Table 1 but absent here were not successfully multiplexed.

Following our initial screening of loci described above, we preceded to genotype plants from two New South Wales populations of each species (A. melvillei: AMEL1, AMEL2; A. ligulata: ALIG1, ALIG2; A. pendula: APEN1, APEN2; Appendix 1) using 10 of the primer pairs developed for each plant species (Tables 3–5). All loci amplified consistently in duplicate PCR assays and were polymorphic with between three and 17 alleles per locus.

Because A. melvillei reproduces both sexually and asexually, we used Gen-Clone to estimate the probability that n (where n = 1, 2, 3…i) copies of a multilocus genotype were produced by distinct episodes of sexual reproduction, P_sex (Arnaud-Haond and Belkhir, 2007). Where P_sex is less than 0.05, it is improbable that n multilocus genotype copies were derived by sex alone.

All 30 plants in AMEL1 were identical, which far exceeds the maximum number of replicates of that genotype (n = 7) that is expected to result from sexual reproduction (P_sex = 0.073) with all replicates of n > 7 identical genotypes associated with P_sex values less than 0.05. In contrast, we detected 26 distinct genets in AMEL2, and it was improbable that the n = 4 replicated genotypes were produced by independent episodes of sexual reproduction (P_sex < 0.001), implying that while the vast majority of distinct genotypes in this stand were founded sexually, the replicate genotypes were produced by asexual reproduction. All A. pendula and A. ligulata plants were genetically distinct, with the exception of one pair in ALIG2. Levels of genetic diversity and expected genotypic diversity expressed as the average number of alleles per locus (A) and expected heterozygosity (H_e), respectively, were generally high for AMEL2, APEN1, APEN2, ALIG1, and ALIG2 (Table 2). However, average inbreeding within populations (F_IS) scores across all loci indicated significant deficits of heterozygotes in all five populations, suggesting inbreeding is a common phenomenon in these species (Tables 3–5). None of the pairwise tests for linkage equilibrium revealed significant associations between loci (P > 0.05).

Table 3.

Levels of genetic diversity and expected genotypic diversity for a nonclonal population of Acacia melvillei.

CONCLUSIONS

These polymorphic markers have proved effective in estimating levels of genetic diversity within populations of these three acacias (A. pendula, A. ligulata, and A. melvillei) and partitioning of variation within and among populations. Moreover, these primer sets can be used to compare levels of genetic diversity and structure within species as part of the process of investigating reproductive failure in A. melvillei and A. pendula. The amplification of DNA extracted from adult leaf material and the embryo of seeds enables estimation of mating system parameters and the assessment of the relative past contributions of sexual and asexual reproduction within and among populations and species. In this initial study, we found evidence of inbreeding in all three species, suggesting a history of isolation. We also identified a high degree of clonality in one population of A. melvillei, a phenomenon which, if widespread, may influence the choice of conservation actions. For the threatened A. melvillei, further landscape-level assessment of genetic diversity and structure, across a wider range of populations, will allow us to estimate historic levels of connectivity, identify populations containing novel genotypes, and assess the suitability of strategies such as genetic rescue. Ultimately, such strategies will inform management via translocation or augmentation. Our success in cross-amplifying markers among Acacia species implies that at least some of these primers will be transferable to other acacias. This study represents the first attempt to characterize the genetic structure of these three important coverstory Acacia species.

Table 4.

Levels of genetic diversity and expected genotypic diversity for two nonclonal populations of Acacia ligulata.

Table 5.

Levels of genetic diversity and expected genotypic diversity for two nonclonal populations of Acacia pendula.