Development of an SSR-based DNA fingerprinting method for black wattle (Acacia

Background: The most commonly used method for extracting DNA from plant leaf tissue involves cetyl trimethylammonium bromide but some species, such as Acacia mearnsii, contain high levels of secondary metabolites and polysaccharides that interfere with this process. Various modifications have been proposed for effective removal of these biomolecules but these methods can be time consuming. Therefore, this study was initiated to optimise the cetyl-trimethylammonium bromide protocol for the extraction of high-quality genomic DNA and to develop a fingerprinting tool using cross species transferable simple sequence repeat markers for genetic diversity studies in A. mearnsii. Methods: Five CTAB-based modification were examined and 49 cross-species microsatellite markers, developed for several Acacia species, were tested in four multiplex panels of A. mearnsii populations. Results: The modified protocol yields high quantity and quality DNA from A. mearnsii leaves using high concentration of NaCl to remove polysaccharides and polyvinylpolypyrrolidone (PVPP) to eliminate polyphenols during DNA purification. In addition, omitting the selective precipitation and NaCl gradient steps in the extraction protocol, enabled us to extract DNA 10–20 min faster than the normal protocol. Of the tested microsatellite loci, 11 were successful in amplifying sharp and high-intensity bands in all the four multiplex panels and were polymorphic. The level of polymorphism ranged from 0.115 to 0.794, with a mean 0.50 and mean number of alleles varied from 2 to 10, with overall mean of 6 alleles per locus. The mean observed and expected heterozygosity ranged from 0.058 to 0.970 and 0.102 to 0.796, respectively. The 11 microsatellite loci that were effectively amplified from A. mearnsii DNA were adequate in detecting genetic variation among the tested populations. Conclusions: These PCR-based, multi-allelic, co-dominant microsatellite markers provide a powerful tool for genetic, breeding and conservation studies in A. mearnsii. New Zealand Journal of Forestry Science Bairu et al. New Zealand Journal of Forestry Science (2020) 50:6 https://doi.org/10.33494/nzjfs502020x56x


Introduction
Acacia mearnsii is a well-known commercially grown tree species in South Africa. This species is cultivated mainly for its timber, timber products, pulp and its tannin-rich bark for the leather tanning industry. Acacia mearnsii is indigenous to Australia but is currently grown intensively in India, Japan, South Africa, Kenya, Tanzania, Uganda, Brazil, Uruguay and Argentina (ILDIS Keywords: Acacia mearnsii, cross-species microsatellite markers, DNA extraction, PCR amplification and Water Affair 2003).
Acacia mearnsii was first introduced into South Africa via a bag of seeds with no prior information on the genetic variability. Consequently, A. mearnsii plantations were presumed to have a narrow genetic base. The high out crossing nature and the low self-incompatibility of Acacia species and their broad geographic adaptability enabled the species to maintain high genetic diversity (Duminil et al. 2009). However, the out crossing rate is highly dependent on flower fecundity, pattern and synchronization (Butcher et al. 2000). These factors are highly influenced by the environment. Intensive breeding was carried out over the years and populations were established in geographically isolated regions depending on the objective of breeding (frost tolerance, disease resistance and timber and bark quality). These populations were exposed to specific environmental factors that led to populations with distinct phenotypes. It was suggested that fragmentation of populations in a given region might result in an increase in the level of inbreeding and a decrease in allelic richness (Butcher et al. 2000). Natural selection, mutation, genetic drift and mating systems all affect the patterns of genetic variation among populations (Millar et al. 2008). To investigate the influence of the above genetic forces on the genetic diversity and patterns of variation, it is helpful to develop appropriate genetic analysis tools.
Population genetic diversity and conservation studies using molecular techniques are important for population or species survival; and the extraction of high quality genomic DNA and polymerase chain reaction (PCR) amplification protocols are essential pre-requisites (Bonin et al. 2004;Tan & Yiap 2009). Extraction of highquality DNA from plant tissue is challenging because plants have variable levels of metabolites and structural biomolecules that interfere with existing DNA extraction protocols (Salblok et al. 2009;Sahu et al. 2012). Secondary metabolites, polysaccharides and polyphenols, are plant biomolecules that often interfere with DNA isolation, enzymatic digestion and PCR (Weising et al. 2005). In the case where species are characterized by their high secondary metabolite content, such as A. mearnsii, DNA extraction procedures require intensive testing and adaptation of different protocols.
The most commonly used basic DNA extraction methods from plant leaf tissue are derived from the original cetyl trimethylammonium bromide (CTAB) based protocol (Doyle & Doyle 1990). For species like A. mearnsii, which contain high levels of secondary metabolites and polysaccharides, ensuring extraction of high-quality DNA is important for downstream genetic analysis (Sahu et al. 2012;Healey et al. 2014). To contend with the problems associated with secondary metabolites various modifications have been proposed for effective removal of these biomolecules (Weising et al. 2005). The modified protocols include the addition of polyvinylpyrrolidone (PVP), insoluble PVPP, bovine serum albumin (BSA), reducing agents like β-mercaptoethanol or high salt solutions to the extraction buffer (Allen et al. 2006;Varma et al. 2007;Lade et al. 2014;Arruda et al. 2017). However, these methods can be time consuming, relying on long incubation steps, or requiring multiple DNA washes and precipitations that decrease overall yield (Healey et al. 2014). Therefore, this study was initiated to optimise the CTAB protocol for the extraction of high-quality genomic DNA and to develop a fingerprinting tool using cross species SSR markers for genetic diversity studies in A. mearnsii.

Sample preparation
The development of the DNA extraction and SSR-PCR amplification protocols were part of the genetic diversity study to support the Institute for Commercial Forestry Research's A. mearnsii breeding and seed production programmes (https://www.icfr.ukzn.ac.za). The four breeding populations used in this study were derived from a base population constituted with 1081 genotypes collected from different progeny trials resulting from controlled crosses and targeted selfing (Dunlop et al. 2003). The four breeding population were reconstituted based on similarity in disease resistance, bark yield and quality, stem form and vigour. Two hundred and twenty-eight individuals were selected from the four breeding populations (TP1 = 78, TP2 = 34, TP3 = 64, and TP4 = 52). At sampling time, three of these breeding populations were seven years old and one was a 15-yearold population (TP1). Young, fresh, and healthy leaves free from visible contamination were sampled, wrapped in moist paper and placed in a cooler box. Upon arrival to the laboratory, each sample was ground into fine powder with mortar and pestle following the addition of liquid nitrogen. The powder was transferred to duplicate 2 ml reaction tubes and immediately stored in a -80 o C freezer. One set of samples were used for DNA extraction and the duplicates were stored for future reference.

DNA Extraction
Various CTAB DNA extraction protocols were tested for their suitability with A. mearnsii leaf material. The initial protocol used here was developed by Centro Internacional de Mejoramiento de Maíz y Trigo (CIMMYT 2005) and modified by Bairu et al. (2006) and Moyo et al. (2008). The protocol was amended to increase DNA quantity and quality and to shorten the time spent on extractions. Briefly, some of the modifications included an altered NaCl concentration, to adding PVPP to the CTAB extraction buffer, and the collapse of the chloroform and CTAB/NaCl steps into a single step as described below and summarised in Table 1.
The final protocol contained the following steps: about 150 mg ground leaf material was added to 800 µL preheated at 65 °C CTAB extraction buffer (2% w/v CTAB, 100 mM Tris-HCl (PH 8.0) 20 mM EDTA, 1.4 M NaCl), with 8 µL 2-mercaptoethanol and 2% PVPP in a clean 2 mL tube. The mixture was incubated in a water bath at 65 °C for 1 hour, while mixing every 10 min by inversion. Immediately following incubation 800 µL chloroform: isoamyl alcohol (24:1) and 160 µL preheated at 65 °C CTAB/NaCl (10% CTAB, 0.7 M NaCl) was added and mixed for 10 min by inversion. The mixture was centrifuged for 10 min at 10000 rpm followed by recovery of the top aqueous layer into a 1.5 ml clean tube. Precipitation of DNA was achieved by adding 800 µL ice-cold isopropanol to the supernatant and mixed gently by inversion. The mixture was then centrifuged for 15 min at 10 000 rpm followed by careful decanting of the isopropanol, leaving the pellet in the tube. The pellet was washed in two steps, first with 70% ice cold ethanol and then 95% ethanol by inversion. After removing the ethanol by decanting, the pellet was allowed to air dry in a laminar flow bench. Finally, the dried pellet was re-suspended in 100 µL TE buffer (10 mM Tris-HCl, 1 mM EDTA, PH 8.0) and stored at -20 °C. Following DNA extraction, the concentration and purity of DNA was estimated using the Jenway Genova Nano spectrophotometer (Bibby Scientific Ltd., Staffordshire, UK).

SSR marker screening
Due to the lack of species specific SSR markers for A. mearnsii, 49 simple sequence repeat (SSR) markers were used from other Acacia species (Additonal File). The markers were chosen based on their ability to amplify across species (Adamski et al. 2013;Aggarwal et al. 2011;Miller 2009;Ng et al. 2005;Butcher et al. 2000). Initially, unlabelled primer sets were tested using PCR amplification and the PCR products were visualization on a 2% MetaPhor™ agarose gel (Lonza Rockland Inc., Rockland, USA) with TAE buffer. Only those markers with visible bands were selected for further evaluation. The forward primer of each selected primer set was then labelled with a fluorescent dye and used for downstream amplification. The PCR products were sent to the Central Analytical Facility at Stellenbosch University, South Africa for fragment analysis. The software program, Gene Marker® v2.4.0 (Soft Genetics) was used to score the electropherograms. After this step, only polymorphic markers that amplified more than two alleles across eight A. mearnsii samples of known genetic variation in the breeding programme were selected for further analyses.

PCR amplification
The initial PCR reactions were performed using the DreamTaq master mix, ready-to-use solution containing DreamTaq DNA Polymerase, optimized DreamTaq buffer, MgCl 2 , and dNTPs, from Thermo Scientific. Each reaction mixture contained; ~150 ng DNA, 5 µL DreamTaq PCR master mix (Thermo Scientific), 0.4 µM of each primer, 0.8 µL BSA (1mg/ml) and dH 2 O to make up the final volume of 10 µL. A gradient PCR was performed to determine the optimal annealing temperature (Ta) for each primer set tested. The temperature range for the gradient PCR was set from 48-60 °C, while the remaining parameters set were according to the manufacturer's protocols. The PCR products were viewed on a 2% MetaPhor™ agarose gel. The optimum Ta was established for each primer set and the touchdown PCR method was used for all subsequent PCR amplifications using the DreamTaq master mix. The touchdown protocol had the following conditions: initial denaturation step at 95 °C for 3 min, 10 cycles of 95 °C for 20 seconds, Ta 10 °C higher than the optimum and decreasing every cycle by 1 °C for 20 seconds and an elongation step of 72 °C for 30 seconds, followed by 25 cycles of 95 °C for 20 seconds, optimum Ta for 20 seconds, 72 °C for 30 seconds, with a final elongation step of 72 °C for 5 minutes. Additional magnesium chloride (MgCl 2 ; 1 mM) was added to PCR mixtures for primer sets with weak amplicons. The various annealing temperatures for each primer pair (Ta = 48-52 °C) and the different levels of amplicons across the primer sets, made multiplexing difficult. Each primer was, therefore, amplified in single-plex and added to a single tube for fragment analysis as a 'multiplex'. Four multiplex sets (Multiplex A, B, C and D) were established ( Table 2). All amplified products were sent to the Central Analytical Facility at Stellenbosch University, South Africa for fragment analysis. Gene Marker® v2.4.0 was used for scoring all genotypes. This process was time consuming and an alternative was needed.
The KAPA2G Fast Multiplex kit (KAPA Biosystems, Cape Town, South Africa) was tested for its multiplexing utility, using the selected Acacia SSR primer sets. The KAPA2G Fast Multiplex kit contains KAPA2G Fast HotStart DNA Polymerase, a buffer optimised for multiplex PCR, with 0.2 mM of each dNTP and 3 mM MgCl 2 (at 1X). This kit is pre-optimized for multiplex PCR reactions to use with primer interactions, primer concentrations, DNA quality and quantity, with marginal changes to annealing temperatures. Each PCR reaction mixture contained: ~150 µL DNA, 5 μL KAPA2G Fast Multiplex mix, 0.6 µL BSA (1mg/mL), 0.1-0.4 μM of each primer and dH 2 O to make up the final volume of 10 µL. The PCR cycle Bairu et al. New Zealand Journal of Forestry Science (2020)

Statistical analyses
Estimates of null allele frequencies were performed using the software program FreeNA (Chapuis and Estoup 2007) using the Expectation Maximization Algorithm (EM) (Dempster et al. 1977). The influence of null alleles on genetic diversity estimates was assessed with a Wilcoxon signed rank test using R (R Core Team 2015) and F st per locus values for corrected uncorrected null alleles were made using the excluding null alleles (ENA) method (Chapuis & Estoup 2007). Genetic diversity per locus was assessed using the mean number of alleles (N a ), observed (H o ) and unbiased expected (uH e ) heterozygosity using GenAlEx v6.5 (Peakall & Smouse 2012), and the inbreeding coefficient (F is ) was determined using Genepop v4.3 (Rousset 2008). Polymorphic information content (PIC) was estimated using Cervus v3.0 (Kalinowski et al. 2007), probability of identity (P ID ) for unrelated individuals and probability of identity for full siblings (P IDSibs ) were estimated using GeneAlEx V6.5, and allelic richness (Ar), using the rarefaction method, as implemented in FSTAT (Goudet 2001). Deviations from Hardy-Weinberg equilibrium were calculated using Genepop v4.3. P ID and P IDSibs estimate the probability that two randomly chosen full-sibs within a given population that have the same genotype on a set of makers. The P ID assumes that there is no linkage disequilibrium and population substructure. When such assumptions do not hold, the P IDSibs is often used as a conservative upper bound of the "real" probability.

DNA extraction, quality and quantity
In this study, the CTAB DNA isolation technique was employed to extract DNA from Acacia mearnsii leaf samples with slight modifications. The mean concentration and purity of DNA samples extracted from the leaves of Acacia mearnsii based on the five protocols tested are presented in Table 1. The mean DNA concentration varied significantly ranging from 44 to 159 ng/µL, at low concentration of NaCl and at high concentration of polyvinyl polypyrrolidone (PVPP), respectively. The addition of a high concentration NaCl and PVPP provided the best results, with high DNA concentrations (>100 ng/µL) and near optimum wavelength ratios (mean 260/280 = 1.887; mean 260/230 = 1.754). Efficient DNA extraction was achieved with the optimised CTAB protocol, with less than 5% reextraction. However, the 260/230 ratio values from the spectrophotometer assessments indicated that some residual phenols and/or carbohydrates might still be present in the DNA extracts. Subsequently, Bovine Serum Albumin (BSA) was used during PCR amplification to eliminate the effect of the co-extracted compounds.

SSR screening and summary statistics
Of the 49 assessed microsatellite makers developed for other Acacia species, only 11 (22.5%) loci amplified clear reproducible high-quality DNA bands in A. mearnsii (Table 2). These markers were polymorphic and consistent amplification was achieved among the four populations of A. mearnsii tested. The proportion of missing data ranged from 0% to 3.5% per locus. Missing data of up to ~4% of the scored genotypes is, however, acceptable for population genetic studies (Putman and Carbone 2014) and no genotype was eliminated from the analysis. The proportion of null alleles calculated per locus over all samples ranged from 0.00 (Ak89) to 25.8 (AH3-10), with a mean of 10.8 over all loci. The null allele frequencies (NAF) for locus AH2-13 and AH3-10 were higher than 20% (Table 3). The observed null allele frequencies per population for AH2-13 and AH3-10 ranged from 15.5-20.6% and from 19.4-32.1% per population, respectively (data not shown). Although the values are high, they are still within the range of values often reported in other studies using SSR loci develop in other species (Dakin and Avise 2004). There was no significant difference detected between Excluding Null Allele (ENA) corrected and uncorrected F ST values (p-value > 0.05; Bonferroni corrected) and therefore, it was decided to keep all loci for the subsequent analyses. The genotypes showed a wide range of allelic diversity from 2 to 10 alleles per locus. The highest allele number (N a = 10) was observed at markers AH2-1 and AH3-10 and the lowest was for AH16, with an overall mean N a of 5.3 (Table 2). Based on alleles detected among four multiplex panels in all four A. mearnsii populations, dinucleotide repeat SSRs were relatively more polymorphic than those with tri-nucleotide repeats.
The unbiased expected heterozygosity (uH e ) ranged from 0.103 to 0.796 per locus, while the observed heterozygosity (H O ) varied from 0.058 to 0.97 (Table 3). Only three loci (AH3-1, AH3-18 and Am465) did not show significant deviation from Hardy-Weinberg equilibrium (Table 3). A high level of variation was observed for the inbreeding coefficient (F IS ) estimates among loci (F IS = -0.567 to 0.653). Negative F IS values were only observed at two loci (AH3-18, Ak89), indicating a heterozygote excess for Ak89. Two loci were found to be highly polymorphic (AH2-1 and AH3-10). Both loci had PIC values greater than 0.7, with high allelic richness (Ar) values. More than 50% of the markers had PIC values greater than 0.50. The combined P ID value for the SSR panel was 7.2 x 10 -8 . Therefore 1 in about 14 million trees will have the same genotype. The P IDSibs value of 1 x 10 -3 provides a lower bound for the number of loci required for the successful identification of individuals, with 1 in about 1000 individuals sharing the same genotype if all individuals are full siblings. This indicates that these markers had a high discriminatory power and were found to be highly suitable for genetic diversity analysis. The allelic richness (A r ) estimates were based on a minimum sample size of 30 individuals, the values ranged from 2 (AH16 and AH3-1) to 9 (AH3-1), with a mean of 4.8 alleles per locus.

Discussion
Exploiting the differences in solubility of polysaccharides and DNA in the CTAB buffer by adjusting the concentration of sodium chloride can aid the removal of polysaccharides (Weising et al. 2005). It has been described that a high salt concentration in the extraction buffer assists in eliminating polysaccharides by increasing their solubility in ethanol (Fang et al. 1992;Lodhi et al. 1994;Varma et al. 2007). The addition of NaCl at concentrations higher than 0.5 M, along with CTAB, successfully removes polysaccharides during DNA extraction (Moreira & Oliveira 2011;Lucas et al. 2019). In the present study, a concentration of 1.4 M NaCl was used in the extraction buffer that further improved the quality of the extracted DNA. Endogenous DNases can degrade the extracted DNA unless EDTA is added (Weising et al. 2005). EDTA has an inhibitory effect on magnesium-dependant DNases by binding to magnesium ions through chelation (Weising et al. 2005). In this study, the addition of PVPP, which have strong H-receptor for binding and removal of polyphenolics into the CTAB buffer was helpful in removing the polyphenols and polysaccharides from leaf samples of Acacia mearnsii (Kolosova et al. 2004).
This protocol resulted in a mean total DNA amount of 15.95 μg extracted from 150 mg leaf tissue. Similar results were obtained with modified CTAB methods in cotton (15-30 μg from 100 mg plant tissue; Ali et al. 2019), as well as Arabidopsis thaliana, Zea mays and Nicotiana sp. (5-30 μg from 200 mg plant tissue; Allen et al. 2006). However, the values obtained in this study are much lower than those obtained from other species characterized by high levels of secondary metabolites (Sahu et al. 2012;Arruda et al. 2017). Tiwari et al. (2012), obtained a DNA concentration ranged from 179 to 833 ng/µL using the modified CTAB protocol in selected medicinal plants. Similarly, Sahu et al. (2012) achieved DNA concentration ranged from 8.8 to 9.9 μg/μL that was amenable of RAPD markers analysis. The difference in DNA concentration could be attributed to the various modifications such as high salt concentration, PVP, PVPP and avoiding the use of liquid nitrogen and selective precipitation and washing steps in addition to species differences.
The 260/280nm and 260/230nm wavelength ratios are well-known measures of nucleic acid quality. The recommended values for the 260/280 ratio ranged from 1.8 to 2.0 and the optimised CTAB protocol in this study resulted in an absorbance value of 1.89, which is within the accepted range indicating the protocol is efficient in obtaining high-quality DNA samples. The quality of the DNA obtained using the optimized protocol is sufficient since ratios in the range of 1.6-1.8 are acceptable for PCR reactions. Similarly, in a pure DNA sample, the reference interval for 260/230 ratio is 2.0-2.2, however, the ratio obtained in this study was 1.75 which falls below the recommended level. This value was similar to those reported by Murray and Thompson (1980) but lower than values reported by Arruda et al. (2017). This indicates that the quality of the DNA is low due to possible contamination by residual polyphenolic compounds and/or carbohydrates in the extract (Moncada et al. 2013). The high quality DNA obtained by Arruda et al. (2017) might be attributed to the higher concentration of CTAB (3%) and NaCl (2.5 M) compared to CTAB (2%) and NaCl (1.4 M) used in this study. This could aid to more efficient elimination of polysaccharides since the composition and concentration of reagents can interfere with the quality and quantity of extracted DNA (Borges et al. 2012). In addition, the low 260/280 and 260/230 ratios obtained in this study might be attributed to high quantities of tannin in A. mearnsii leaves (Elgailani & Ishak 2014).
The systematic exploration of microsatellite markers across species should be the first step in developing SSRs Bairu et al. New Zealand Journal of Forestry Science (2020) (Yosodha et al. 2005;Ravishankar et al. 2015). It is, therefore, more feasible to source SSR primers developed for other related species. The use of SSR markers across species from the same genus (Omondi et al. 2010;Aggarwal et al. 2011;Adamski et al. 2013;Le Roux et al. 2013;Roberts et al. 2013) and across different genera (Peakall et al. 1998) have been reported for Leguminosae. In this study, the possibility of crossspecies transferability of 49 microsatellites, derived from A. mangium, A. koa and A. auriculiformis, into A. mearnsii was investigated. Of the tested 49 cross-species markers, only 11 were effective in amplifying sharp and high intensity bands. The transfer rate was 22.5%, which is moderately low. This result was in agreement with Butcher et al. (2000), who reported low levels of crossspecies SSR amplification within Acacia species. The low transferability might be attributed to the genetic divergence among the Acacia species due to exposure to different bottlenecks and geographic isolation (Varshney et al. 2005;Barbará et al. 2007). Similarly, several studies indicated the low transferability rate of SSRs among plant species (Luro et al. 2008;Koppolu et al. 2010;Lee et al. 2011). Although transferability was in general very low, the mean number of alleles amplified per locus was relatively higher (6.0) in this study compared to the 2.9 allele per locus reported by (Adamski et al. 2013) using 16 Acacia koa SSR markers on seven Acacia species. The high levels of polymorphisms of SSR loci indicated that the application of these markers for genetic diversity studies in A. mearnsii was useful and cost-effective compared to developing new SSR markers specific to the species. This study provided an important insight in the development of microsatellite markers suitable for genetic studies in A. mearnsii. In the current study, we only used SSRs to study their cross-species transferability and on the genetic diversity of four A. mearnsii breeding population. However, their genomic distribution, biological function and other uses should be further investigated (Vieira et al. 2016).
Various factors such as genotyping errors of loci should be considered in choosing a sufficient SSR panel for population genetic and parentage studies. A wellknown problem with SSR is the occurrence of null alleles and over estimation of alleles due to polymerase strand-slippage in DNA replication (Vieira et al. 2016). The analysis of null allele frequencies per genotype conducted using the maximum likelihood estimates in this study confirmed the presence of null alleles in all loci but Ak89. For these loci, null-allele frequencies estimates ranged between 0% and 25.8%, with 64% of the loci showing null-allele frequencies above 5%. In this case, the null alleles occur when an allele fails to amplify via PCR and heterozygotes might falsely be scored as homozygote. Null alleles can possible be either caused by mutations in the flanking regions of a microsatellite sequence (Chapuis & Estoup 2007;Dakin & Avise 2004), or associated with shorter alleles outperforming the longer alleles, usually due to lower DNA quality or quantity (Gagneux et al. 1997;Wattier et al. 1998).
Consequently, null alleles may inflate genetic diversity parameter estimates (Chapuis & Estoup 2007;Carlsson 2008) and result in inaccurate parentage assessments (Dakin & Avise 2004). Carlsson (2008) reported that in datasets containing null alleles, the genetic differentiation estimate F ST were slightly overestimated. If the frequency of a null allele is <0.1, some estimators can be used directly without adjustment; if it is >0.5, the potency of estimation is too low and such a locus should be excluded (Huang et al. 2016). In this study, the estimated null allele frequencies were generally below the threshold levels.
All SSR loci were used for the downstream diversity analyses. It was reported by Carlsson (2008) that loci affected by null alleles can still be useful for population level studies. However, breeders should interpret the results with caution as the effect of null alleles on parentage analysis is more substantial, as it could lead to false parentage assignments (Pemberton et al. 1995). Since these breeding populations were derived from the same base population, it is therefore, advisable to use parentage assignment methods that consider the presence of null alleles in the dataset (Dakin & Avise 2004) during further selection and breeding planning.
The mean genetic diversity value observed in A. mearnsii (H e = 0.524) in this study was much higher than the study conducted on A. longifolia in Portugal (H e = 0.190) (Vicente et al. 2018) using SSRs. However, the value is more comparable with a study of A. senegal (H e = 0.519) (Djibo et al. 2017), A. auriculiformis (H e = 0.60) (Son et al. 2016) and A. mangium (H e = 0.56) (Son et al. 2016) using SSRs. The inbreeding coefficient (F IS ) measures the deviation of heterozygosity from expected values under the assumptions of Hardy Weinberg. A negative and positive F IS value indicates an excess or deficiency of heterozygosity in a population relative to the ratio under HWE. All the markers except Ak89 revealed a positive F IS value suggesting a deficiency of heterozygotes. The deficiency of heterozygotes can be attributed to either inbreeding or existence of null alleles or the presence of subpopulations within populations (Wahlund effect) (Jordana 2003). Of the 11 microsatellite loci analysed, 9 loci deviated from HWE in A. mearnsii (P < 0.001). This might be due to the small sample size used or the existence of null alleles (Wang 2008). Linkage disequilibrium observed among pairs of loci in this study appeared to be associated with the presence of null alleles in some of the analysed loci. All the microsatellite loci, except Ak89 and AH01, which were found to depart from the Hardy-Weinberg equilibrium (HWE), had null allele frequencies exceeding 10%.
The probability of identity (P ID ) is also a widely used individual identification estimator (Peakall & Sydes 1996;Reed et al. 1997;Waits et al. 2001;Ferrie et al. 2013), which provides the probability that two randomly drawn individuals within a given population will have identical genotypes at multiple loci. For P ID , it is assumed that there is neither linkage disequilibrium nor population substructure. In the case of such assumptions not holding true, the P ID might drastically underestimate the above-mentioned probability. Hence, probability of identity for siblings (P IDsibling ) is often used as a conservative upper bound of the "real" probability (Waits et al. 2001). Low P ID values are required to accurately estimate individual identity and not falsely classify individuals as identical (Waits et al. 2001;Vidya & Sukumar 2005). Locus P ID values in the range of 0.01 -0.0001 are generally used in wildlife forensics (Waits et al. 2001), suggesting the values in the current study are slightly higher than ideal. It is important to take the population size and genetic history of the species under study into account in choosing an appropriate marker panel using P ID estimates (Schwartz & Monfort 2008). For example, if the study population comprise of 10000 individuals, a P ID < 0.0001 (1 in about 10000 individuals) should be sufficient for individual identification.

Conclusions
In this study, we managed to get high quantity and quality DNA from A. mearnsii leaves by modifying the CTAB-based DNA extraction procedure using high concentration of NaCl to remove polysaccharides and PVPP to eliminate polyphenols during DNA purification. In addition, by combining the CTAB and NaCl steps in the extraction protocol, we succeeded in the extraction of DNA 10-20 min faster than the normal protocol. According to our knowledge, this is the first assessment of cross-species microsatellite marker transferability and the use of these markers to study the genetic variation in A. mearnsi. However, it was also demonstrated that the majority of cross-species microsatellite markers were not transferable across species in the genus Acacia. Considering the high polymorphism demonstrated by the studied microsatellite markers, the cross-species amplification is an interesting alternative to the development of new microsatellites in A. mearnsii. These multi-allelic, PCR-based, co-dominant microsatellite loci provides a powerful tool for genetic, breeding and conservation genetic studies of A. mearnsii.