The dispersal and colonization of plant populations allow species to occupy novel habitats, migrate, and undergo range shifts in response to changing environmental factors and, as such, are fundamental ecological processes for ensuring the long-term persistence of species. Natural landscape disturbance often generates habitats available for colonization. Patterns of colonization and population expansion can be inferred from the levels and partitioning of genetic variation of plant populations with known disturbance histories, such as recent volcanic eruptions. We sampled and mapped 496 individuals from two populations of the colonizing terrestrial orchid, Sobralia chrysostoma, on the 1992 lava flow of Volcán Arenal in central Costa Rica. We used neutral co-dominant markers to genotype individuals and estimate population genetic statistics.  Both populations had high mean levels of genetic diversity (P = 100%; AP = 3.31; H_e = 0.259) suggesting that the lava flow was colonized by numerous individuals that likely originated from multiple source populations. However, significant spatial genetic structure (SGS) was only present in one population at the smallest distance class (≤ 2 m) and was low (r = 0.032). That these large and genetically diverse populations had such low SGS and an absence of SGS respectively is contrary to expectations and differs significantly from the pattern in Epidendrum radicans (Orchidaceae), with which S. chrysostoma is growing sympatrically. Our results suggest that these two populations either consist primarily of immigrant individuals or that seeds produced in situ dispersed over longer distances, thereby producing larger seed shadows and greater overlap of seed shadows.

In April 2004 leaves were collected from 288 S. chrysostoma individuals at the first Arenal Slope site (AS-1; 10.4504° N and 84.7250° W). In April 2010, leaves were collected from 208 S. chrysostoma individuals at the second site (AS-2). The populations are separated by ~90 m and the collections were separated by six years. From a central point in each population, a transect line was placed successively in a radial pattern and a leaf was collected from every individual, with its distance and angle relative to the central point recorded.

Leaf samples were also collected from a mean of 49 individuals (range 46 - 50) from each of three regional populations of S. chrysostoma within 0.3 to 2.9 km (mean = 1.9 km) of the lava flow populations. Because of challenges posed by the terrain, sampled individuals were not mapped. While we did not survey the surrounding landscape comprehensively, we collected from every population found in the vicinity. 

Leaves were snap frozen in liquid nitrogen within a few hours of collection, and transported to the University of Georgia in an ultra-cold dry shipper.  Leaves were crushed in chilled mortars with a pestle, liquid nitrogen, and a pinch of sea sand to disrupt cellular compartmentalization.  Enzymes were extracted from leaf tissue with the Wendel and Parks (1982) extraction buffer.  The resulting slurry containing crude protein extract was absorbed onto 4x6 mm wicks punched from Whatman 3 mm chromatography paper.  Wicks were stored in microtest plates at -70°C until used for electrophoresis.  For electrophoresis, wicks were placed in horizontal starch gels (10%). 

Seven enzyme stains in four buffer systems resolved one monomorphic and eight polymorphic loci. Enzymes stained and polymorphic loci (in parentheses) for each buffer system were: (1) system 4, malate dehydrogenase (Mdh-2, Mdh-3), and 6-phosphogluconate dehydrogenase (6-Pgd); (2) system 7, aspartate aminotransferase (Aat), and menadione reductase (Mnr); (3) system 8-, triosephosphate isomerase (Tpi-1); (4) system 11, isocitrate dehydrogenase (Idh), and shikimic dehydrogenase (Skdh). All buffer and stain recipes were adapted from Soltis et al. (1983) except Aat (Cheliak & Pitel, 1984). Buffer system 8- is a modification of buffer system 8 described by Soltis et al. (1983). Banding patterns were consistent with Mendelian inheritance patterns expected for each enzyme system (Weeden & Wendel, 1989) and were consistent with disomic inheritance expected for diploid individuals.