Mangroves are intertidal ecosystems that have a pantropical distribution. The distributional range of species inhabiting these ecosystems is typically restricted to either the Indo-West Pacific (IWP) region or the Atlantic-East Pacific (AEP) region (Tomlinson, 1986). How this pattern of distribution formed is one of the main biogeographic questions in mangrove research. Phylogenetic studies have detected significant levels of divergence in several tree species across the IWP and AEP (Rhizophora L. in Duke et al., 2002 and Takayama et al., 2013; and Hibiscus L. in Takayama et al., 2008). However, the divergence history, at a global scale, of many other mangrove plants remains to be clarified. Acrostichum aureum L. (common name “mangrove fern”; Pteridaceae) is of particular interest because this species is the only mangrove plant that is distributed pantropically (i.e., in both the IWP and AEP regions). This species also differs from other mangrove plants in that it has wind-dispersed spores, while most other mangrove plants have sea-dispersed seeds, fruits, or propagules. This different dispersal system might have enabled this species to achieve its relatively wide distribution compared to other mangrove plants. To address this question, it is important to perform population genetic studies to analyze the genetic structure and demographic history of the species using highly polymorphic microsatellite markers. Therefore, we developed novel microsatellite markers for A. aureum using next-generation sequencing. We tested the markers on samples from across the pantropical distribution of the species to check their levels of polymorphism and to determine their usefulness as markers for future studies.

Table 1.

Characteristics of 27 microsatellite markers developed for Acrostichum aureum. ^a,b

METHODS AND RESULTS

One sample of A. aureum was collected from Sabah (Malaysia) (Appendix 1), and total DNA extracted using a DNeasy Plant Mini Kit (QIAGEN, Hilden, Germany). We then performed shotgun sequencing, using one-third of a run on a Roche 454 Genome Sequencer Junior (Roche Applied Science, Penzberg, Germany). The GS Junior Titanium Sequencing Kit (Roche Applied Science) and Multiplex Identifier (MID) adaptors (see Margulies et al., 2005) were used following the manufacturer's protocol. The run generated a total of 81,415 reads with an average length of 490 bp.

The program QDD version 2.1 (Meglécz et al., 2010) was used to identify di- to hexanucleotide motif microsatellites with at least five repeats. Sequence similarity and establishment contigs were detected following the procedure in Takayama et al. (2011). A total of 1452 perfect microsatellite sequences were obtained and 48 primer pairs designed using the following criteria: (1) PCR product size of 80–300 bp; (2) flanking region containing at least five repetitions of any di- to hexanucleotide motifs; and (3) primers with length 18–27 bp, annealing temperature 57–63°C, and GC content 20–80%. Forty-eight primer pairs with at least 12 repeats of various fragment sizes appropriate for multiplex PCR were selected. The detailed primer method (Schuelke, 2000) was used to label and visualize the PCR amplicons of the selected primers. The 19-bp U19 sequence (GGTTTTCCCAGTCACGACG) was added to the 5′-tail of forward primers, and the GTTT PIG-tail was added to the 5′ end of the reverse primer. This PIG-tail facilitates the addition of adenosine by Taq polymerase, thereby reducing stuttering (Brownstein et al., 1996). PCR amplification tests of each primer pair were performed in individual PCR reactions using two individuals, collected from Sabah (Malaysia) and Pará (Brazil), using the standard protocol of QIAGEN Type-it Microsatellite PCR Kit (QIAGEN), with a final volume of 5.0 µL and 1.0 µM of each primer. The PCR thermal conditions were as follows: initial denaturation at 95°C for 5 min; 30–32 cycles of denaturation at 95°C for 30 s, annealing at 57°C for 90 s, extension at 72°C for 30 s; and final extension at 60°C for 30 min. The PCR products were electrophoretically separated on 1.5% agarose gels stained with ethidium bromide. Thirty loci exhibited clear PCR amplification. Twenty-four individuals sampled from Sabah (Malaysia) were used to assess the quality of amplification and polymorphism of these 30 loci. Loci were amplified using QIAGEN Type-it Microsatellite PCR Kits (QIAGEN) in three tubes, each with 5.0-µL mixtures containing 0.5 µL of 1–10 ng of genomic DNA, 2.5 µL of multiplex PCR master mix buffer, 1.2 pL of primer mix (with the concentration of each primer pair adjusted from 1.0 µM), and 0.8 µL of U19 fluorescent dye–labeled primer (6-FAM, VIC, NED, or PET; 1.0 µM). We used the same PCR protocol as described above. Twenty-seven of the 30 loci showed clear fragment patterns using one singleplex and 11 multiplex PCR sets (two to three primer pairs per multiplex; Table 1). Samples from two more populations (16 individuals each from Piti [Guam] and Pará [Brazil]; Appendix 1) were then included to check the genetic diversity of these loci. Cross-species amplification of these loci was also assessed by testing in the other two species in the genus Acrostichum L.: four individuals of A. danaeifolium Langsd. & Fisch, collected in Pará (Brazil) and Colima (Mexico), and four individuals of A. speciosum Willd. from Sungei Buloh (Singapore) (Appendix 1).

Table 2.

Genetic variation of the 27 newly developed microsatellite markers in three Acrostichum aureum populations.^a

The amplified products were loaded into an ABI3500 automatic sequencer (Applied Biosystems, Waltham, Massachusetts, USA) with GeneScan 600 LIZ Size Standard (Applied Biosystems), and their sizes and genotypes were determined using GeneMarker (Holland and Parson, 2011). Expected heterozygosity (H_e) and fixation index (F_IS) were calculated to evaluate genetic diversity of the three populations using FSTAT version 2.9.3.2 (Goudet, 2001; hereafter, FSTAT). The significance of deviations of F_IS from zero, as evidenced by deviation from Hardy–Weinberg equilibrium, and genotypic disequilibrium for all locus pairs, were tested by randomization using FSTAT. The obtained P values (with a 0.05 significance threshold) were adjusted based on a sequential Bonferroni correction. The presence of null alleles and their bias on genetic diversity among the three populations (F_ST) (Weir and Cockerham, 1984) were evaluated using FreeNA (Chapuis and Estoup, 2007). In the Sabah population, the number of alleles detected and H_e ranged from one to 15 and 0.000 to 0.893, respectively, and 26 of the 27 loci were polymorphic (Table 2). A significant deviation in F_IS was found in only one locus (AA16). Although null alleles were detected and their frequencies estimated at each locus (Table 4), the F_ST value after the null allele correction was 0.619, the same as the original value without correction (= 0.619), suggesting that biases, due to null alleles, in genetic structure analysis would be limited. Although 19 of the 27 loci were amplified in samples from the other two populations, most were fixed for different alleles among populations. Seven and six loci were amplified in A. danaeifolium and A. speciosum, respectively (Table 3).

CONCLUSIONS

The 26 polymorphic microsatellite markers developed in this study will be useful to evaluate the genetic structure and to infer the past demographic history of A. aureum to study how this mangrove fern achieved the widest distributional range of all mangrove plants. Cross-species amplification also suggested that some markers could be used to evaluate genetic diversity in other species in the same genus.

Table 3.

Fragment sizes detected in cross-amplification tests of the 27 newly developed Acrostichum aureum microsatellite markers in two closely related species.^a

Table 4.

Null allele frequencies at each locus estimated by FreeNA software in three Acrostichum aureum populations.^a