Mauritia flexuosa L.f. (Arecaceae) is a dioecious, diploid palm (Röser et al. 1997) covering millions of hectares in Amazonia in monodominant stands (Peters et al. 1989). M. flexuosa populations are naturally isolated from one another as they generally dominate swamps along meandering rivers that originated as ox-bow lakes (Kalliola et al. 1991). The fruit of M. flexuosa has a high wild harvest potential, but harvesters often fell females for easier access to the fruit (Holm et al. 2008). This harvesting strategy, combined with artisanal gold mining that often occurs along meandering rivers in Amazonia in both Peru and Brazil, could restrict gene flow in naturally fragmented populations. To investigate questions regarding habitat fragmentation, impacts of wild-harvest, and natural isolation, we developed 13 novel polymorphic dinucleotide microsatellite loci for M. flexuosa.

Microsatellite library development followed a modified version of Glenn and Schable (2005). DNA was extracted from the leaf tissue of one M. flexuosa individual from Pará, Brazil, donated by the Montgomery Botanical Center. Dry tissue was homogenized in a Qiagen Tissue Lyser (Qiagen, Inc.) and DNA extracted following the Qiagen DNeasy Plant Kit (Qiagen, Inc.) protocol with modified Buffer AP1 to which we added 40 μL of 10 mg/mL PVPP (polyvinylpolypyrrolidone). DNA was digested with the restriction enzyme Sfo I (New England Biolabs) and ligated to double-stranded Super-SNX linkers. Following ligation, the restricted DNA was denatured, hybridized to biotin-labeled oligonucleotides [(GA)12, (CA)12] and captured on magnetic streptavidin coated beads (Invitrogen). Retrieved microsatellite-enriched DNA was amplified via PCR and sent for commercial Rolling Circle Amplification and sequencing to Sequetech (Mountain View, CA). Sequences were searched for the presence of microsatellites using MSATCOMMANDER 0.8.1 (Faircloth 2008) and primers were designed using Primer3 (Rozen and Skaletsky 2000).

We tested 36 primer-pairs on ten individuals of M. flexuosa from Estrada do Amapá, Acre, Brazil, collected by collaborators at the University of Acre. Loci were amplified via PCR on an Eppendorf Mastercycler (Eppendorf, Westbury, NY). The M13-tailed primer method (Boutin-Ganache et al. 2001) was used for genotyping individuals. Reactions involved forward primers 5′-tailed with a 15-mer M13 sequence (5′-TCCCAGTCACGACGT-3′), unmodified reverse primers and fluorescently labeled (6-FAM, VIC, NED) 15-mer M13 primers (Applied Biosystems). Reactions were performed in 10 μL volumes using 1× Qiagen Type-It master mix (Qiagen, Inc.), 0.05 μM Forward Primer, 0.5 μM Reverse Primer, 0.5 μM M13 primer and ~20 ng template DNA. The thermal cycling conditions used for PCR amplification included an initial denaturation step at 94°C for 8 m followed by 35 cycles of 30 s at 94°C, 30 s at 54 or 57°C (Table 1), 30 s at 72°C, and a final extension step at 72°C for 5 m. PCR products were run through an Applied Biosystems 3730xl DNA Analyzer at the DNA Analysis Facility on Science Hill at Yale University (http://www.dna-analysis.research.yale.edu) and the data were analyzed in GENEMARKER v1.91 (SoftGenetics, State College, PA).

Table 1 13 polymorphic microsatellite loci for M. flexuosa
Full size table

Observed and expected heterozygosity were calculated using GENALEX 6.41 (Peakall and Smouse 2006). We used GENEPOP 4.0.10 (Rousset 2008) to test for departures from Hardy–Weinberg equilibrium (HWE) and to determine possible linkage disequilibrium (LD) between loci using a burn-in of 100,000 and 1,000 batches with 10,000 iterations per batch. We performed 10,000 permutations to estimate null allele frequencies via two different methods: Dempster et al. (1977), as implemented in FREENA (Chapuis and Estoup 2007), and Oosterhout et al. (2006), as implemented in MICROCHECKER (Oosterhout et al. 2004).

Table 1 shows the results of the 13 loci that were polymorphic and easily scored out of the original 36 tested. Those 13 loci were tested on 25 individuals from a single population in Estrada do Amapá, Acre, Brazil. Loci ranged in size from 180 to 291 bp. Number of alleles ranged from 6 to 15 and expected heterozygosity varied from 0.582 to 0.918. All loci were in HWE (P > 0.004) after sequential Bonferroni correction (Holm 1979). Additionally, no loci were in LD after sequential Bonferroni correction. Nine loci were in HWE (P > 0.05) before Bonferroni correction, eight of which had null allele frequencies below 5%. The higher null allele frequencies and possible deviation from HWE in these loci are likely to be fixed by lowering the annealing temperature and should still be useful for future genetic studies.

These polymorphic loci provide an opportunity for the study of gene flow in a largely unstudied system with high potential for conservation initiatives and sustainable wild harvest management programs (Peters et al. 1989; Holm et al. 2008). In future studies, we will apply these microsatellite loci to multiple populations of M. flexuosa in Brazil that encompass a variety of land-use types to inform conservation and management initiatives.