Are microsatellite fragment lengths useful for population-level studies? The case of Polygala lewtonii (Polygalaceae)(1)

PREMISE OF THE STUDY: Microsatellites, despite being commonly used population-level markers, contain biases because scoring relies solely on fragment length. Their complexity can lead to homoplasy, the effects of which are poorly understood. We investigated the impact of using fragment lengths, repeats, or flanking region sequences on common population-level analyses. METHODS: Five polymorphic microsatellite markers amplified across the central Florida scrub endemic Polygala lewtonii (Polygalaceae) and its close, widespread congener P. polygama. We genotyped 147 individuals of P. lewtonii and 156 of P. polygama, and sequenced the amplicons of four markers across all individuals. We ran basic statistics, spatial clustering analysis, historical demographics, and migration tests. RESULTS: One population of intermediate morphology was genetically clearly identified as P. polygama, making it the southernmost population of this species. Statistics were comparable between the fragment length and repeat numbers, with some notable differences. Flanking regions exhibited surprisingly high polymorphism between species, and between geographically distant conspecific populations. DISCUSSION: The increasing use of markers developed in one species and amplified in another is only a good practice if precautions are taken, notably the sequencing of the fragments between species and populations. Flanking region sequences are a useful marker at the interspecific level.