Effect of habitat fragmentation on the growth, reproduction, and population dynamics of an Amazonian understory herb (Heliconia acuminata, Heliconiaceae)

Understanding the physical and biological consequences of habitat fragmentation remains one of the central areas of research in applied ecology. While numerous studies have demonstrated that the local extinction of plant species from habitat fragments is common, the mechanisms responsible for these extinctions are usually unknown. The disruption of reproduction and recruitment by plant populations in habitat fragments is frequently hypothesized to be a primary mechanism driving these local extinctions. I tested this hypothesis using as a model system the Amazonian understory herb Heliconia acuminata (Heliconiaceae), found in both the experimentally isolated forest fragments and the continuous forest reserves of Brazil's Biological Dynamics of Forest Fragments Project. I did this by: (1) determining via a series of manipulative experiments how and why H. acuminata growth and reproduction vary fragments and continuous forest; (2) establishing a series of large, permanent demography plots in both continuous forest and fragments; and (3) incorporating these demographic and empirical data into matrix-based demographic models. I found that H. acuminata individuals in habitat fragments are significantly smaller and have fewer vegetative shoots than those in continuous forest. Reductions in size may be the result of exacerbated dry-season abiotic conditions in fragments, a conclusion borne out by experimental transplants. Since reproduction in H. acuminata is size dependent, reductions in individual plant size lower the numbers of reproductive plants in forest fragments relative to areas of continuous forest. While the levels of fruit production by plants in fragments and continuous forest are similar, seed germination rates are significantly lower in forest fragments. Collectively, these results suggest that changes in abiotic conditions have a large impact of H. acuminata growth, reproduction, and recruitment, which could substantially impact population dynamics. Demographic models incorporating the observed levels of seed germination did result in projections of negative population growth in fragments, however only minimal increases in seedling establishment were necessary for population growth rates to become positive. Comparisons of projected and observed numbers of seedlings suggest there is seed dispersal into forest fragments maintaining positive population growth rates; Heliconia acuminata populations in this fragmented Amazonian landscape may therefore exhibit source-sink dynamics.