Microsatellite markers for an endemic Atlantic Forest tree, Manilkara multifida (Sapotaceae)

We developed microsatellites for Manilkara multifida for future conservation genetics studies. M. multifida is a tropical tree that is endemic to Brazil which is currently restricted to fragmented landscapes. Our analysis indicated that all eight microsatellites are promising for assessing population genetics questions in this species.


Introduction
The Atlantic Forest is the second largest tropical forest in South America. Its original area measured 1 500 000 km 2 ranging along the Brazilian Atlantic coast, with additional patches in Argentina and Paraguay (Ribeiro et al. 2009). Currently, only 11.4-16% of its original cover remains, and it is distributed in a fragmented landscape (Ribeiro et al. 2009) as a consequence of land use and occupation. Most Brazilians live in areas that were originally covered by the Atlantic Forest, which increases the pressure on native forest species. Southern Bahia, Brazil, is a particularly diverse region of the Atlantic Forest, especially in terms of tree species, and for many years, inadequate forest management and agriculture intensification reduced the amount of forest cover as well as the abundance of tree species. During the 1970s, the timber industry alone caused major environmental damage in this region (Mesquita 1997), and several species, including Manilkara multifida Penn., suffered a significant reduction in that period due to the high commercial value of their timber.
Manilkara multifida is a rare tree species endemic to the Atlantic Forest of southern Bahia, and it is included on the IUCN Red List as an endangered species (O'Brien 1998). Members of the genus Manilkara are known to provide important food resources for primates (Oliveira et al. 2009), but there are no data about the functional and evolutionary role of M. multifida in its habitat. Trees of this species are lactescent and have flowers with red sepals and white petals (Pennington 1990). They can also be identified by their wide and discolorous leaves, which are covered by an adpressed indumentum on the abaxial face, and by their numbers of petals, stamens and staminoids (8-9 altogether) (Pennington 1990). In the present study, we report the development and characterization of the first microsatellite markers for M. multifida. This resource will help researchers obtain experimental data for the elaboration of genetic inferences that will contribute to conservation programmes for this species.

Genomic library and primer design
A genomic library was constructed from template DNA (250 ng mL 21 ) extracted via the cetyl trimethylammonium bromide method (Doyle and Doyle 1990). Isolated DNA was digested with the restriction enzyme RsaI (Invitrogen, Carlsbad, CA, USA). Fragments of between 200 and 800 bp were captured and linked to the adapters Rsa21 and Rsa25 (Edwards et al. 1996) using T4 DNA ligase. The DNA fragments containing microsatellites were captured by hybridization with biotinylated probes (CT) 8 and (GT) 8 , and recovered using magnetic streptavidin beads (Promega, Madison, WI, USA). These fragments were then amplified by polymerase chain reaction (PCR) according to the following conditions: initial denaturation at 958C for 1 min; 25 cycles of denaturation at 948C for 40 s, annealing at 60 8C for 40 s and extension at 72 8C for 2 min; and final extension at 72 8C for 5 min. Amplicons were cloned into the pGEM-T Easy vector (Promega), which was then transformed into Escherichia coli XL1-Blue (Promega). A set of 96 positive clones was amplified using M13 primers and Big Dye w Terminator reagents (version 3.1; Applied Biosystems, Foster City, CA, USA) and sequenced using a DYEnamic TM kit (GE Healthcare). Among these, 22 clones were discarded due to low-quality sequencing data (chromatograms). We analysed the remaining 74 sequences using SSRLocator software (Maia et al. 2008) and detected 67 containing microsatellites. A total of 14 pairs of primers were designed using Primer3 software (Rozen and Skaletsky 2000).

Analyses performed
We assessed paternity exclusion Q, diversity indices H E and H O , and allelic identity (I) using Cervus 3.0.3 (Marshall et al. 1998). We also computed the fixation index (F ) in FSTAT 2.9.3.2 (Goudet 2002) and examined for null alleles using Micro-Checker (Van Oosterhout et al. 2004).

Results
Eight polymorphic SSR loci for M. multifida were identified (Table 1). The average of 11.9 alleles per locus indicated high levels of polymorphism. The expected heterozygosity ranged from 0.622 to 0.911, and the observed heterozygosity ranged from 0.563 to 0.969. The MM03 and MM86 loci showed negative values of F (indicating excess heterozygotes). The combined exclusion probability of Q indicates that this set of microsatellites will be a powerful tool for future paternity studies. The I combined estimation confirms the potential use of this set of markers from M. multifida for identity analyses (Table 2). Null alleles were not observed at any locus, according to the Brookfield 1 method (Brookfield 1996). In many cases, the presence of null alleles, which usually occur due to a mutation in the primer annealing region, inflates the frequency of homozygotes. The Micro-Checker software found no genotype scoring errors due to stutter bands or allelic dropout.

Conclusions and forward look
The eight polymorphic microsatellite markers developed in this study display important discrimination capacities. These loci therefore have excellent potential to be used in genetic studies of M. multifida aimed at the conservation of this species.

Accession numbers
The sequences were deposited in the GenBank database with the following accession numbers: JX183540   Table 1 Description of the eight nuclear SSR markers developed for M. multifida in the present study, with forward (F) and reverse (R) sequences, repeat motif, annealing temperature and size range.