Revising classification of focal vegetation communities of Mona Island Reserve, a Caribbean subtropical dry forest ecosystem

Background/Question/Methods

Previous studies of the vegetation of Mona Island described a topographic-related  type of vegetation (depressions or “bajuras”), characterized by taller and denser canopies.  A study conducted using remote sensing discriminates 15 types of vegetation cover classes, and revealed an East-West gradient of increased productivity.  In this context, we asked: are “bajuras” a class of vegetation community?, and are soil variables drivers for vegetation composition and structure?   Vegetation data was gathered at three depression and three plateau forest sites.  At each site we set up one  240 m permanent transect with 17 equidistant (15 m) sample points.  Each sample point represents the center of a 3 m radius circular plot within which we also established two 1 m² quadrats.   We counted and identify all established vegetation within each circular plot (> 1m height), and each quadrat pair (< 1 m).  Soil samples were collected in each circular plot for chemical analyses of nutrients and physical properties.  The effects of forest type and location on vegetation and soil variables were tested using MANOVA and ANOVA (for each variable).  To test for differences in vegetation composition among locations and the contribution of soil variables to these differences, we conducted a Redundancy Analysis using CANOCO.

Results/Conclusions

Forest type had a significant effect on species density variation at the overstory and understory strata, but so did the forest*site interactions.  On average, plateau sites had higher species densities in the overstory but this pattern was inconsistent across replicates. The density of species in the understory was higher at depression sites compared to plateau sites but replicates within forest types were significantly different from each other.  Herbs and grasses were the dominant growth forms of the understory at depression sites while non-woody growth forms were dominant in the understory of plateau sites.  MANOVA showed significant differences among forest types, replicates within forest types as well as significant interaction effects between forest type and replicates in the soil variables. The RDA analyses showed that the separation between sites was better explained by the variance in the amount of data explained by environmental variables.  Our results did not show that depression forests have a clear distinction from the plateau forests in terms of their species compositions and plant densities, neither a special soil characterization.  The high spatial heterogeneity of species distribution, microtopography and soil variables contribute to overcome the differences associated to the vegetation among particular locations within Mona Islands.