Print PageForest fragmentation can result in increased genetic differentiation among populations and overall erosion of genetic diversity at large scales. Little is known, however, about the specific mechanisms affecting the genetic make-up of individual populations across modified landscapes. Fragmentation can alter pollen and seed dispersal, effective population size, and post-dispersal mortality, all of which are reflected in the magnitude and extent of fine-scale spatial genetic structure (i.e., non-random distribution on genotypes in the environment). We use the Amazonian understory herb Heliconia acuminata, an outcrossed plant, pollinated and seed-dispersed by birds, to investigate how the fine-scale spatial genetic structure (SGS) varies across a fragmented landscape. A combination of demographic and genetic data was collected from all plants in 50 x 100m plots in two continuous forest sites and three 1-ha fragments of the Biological Dynamics of Forest Fragments Project, Amazon, Brazil. Leaf tissue from all plants was collected in 2009 and genetic analysis was conducted using ten polymorphic microsatellite markers. We measured inbreeding and characterized SGS by assessing genetic relatedness between pairs of individuals in function of distance in each population.
Results/Conclusions
Continuous forest plots presented the lowest and highest values of abundance (122 and 765 plants) and proportion of flowering plants (2 and 6%), whereas fragments exhibited lower variation in plant abundance (198, 210, and 229) and proportion of flowering plants (2, 5 and 3%). SGS and inbreeding were significant but relatively low in all five populations. The highly dense continuous forest population presented the weakest SGS, followed closely by the least dense fragmented population. Plant density may be playing different roles on SGS of Heliconia acuminata in each plot. In the continuous forest site, high effective population size may weaken SGS by promoting parental diversity, decreasing bi-parental inbreeding and increasing overlap among seed shadows. In the fragment, isolated populations with low plant density may experience enhanced pollen and seed dispersal from other forested areas, which in turn decrease SGS. SGS of the three remaining populations, including one continuous forest plot, differed from the dense continuous forest population. Results suggest that different factors, such as effective population size, patch size (1-ha fragments vs. continuous forest) and possibly other features of the system, may interplay in a complex way to generate the observed pattern of SGS across this fragmented landscape.
