Comparison of composition and diversity of protected and unprotected seasonally dry tropical forests in the cattle ranching landscape of Liberia county, Guanacaste, Costa Rica
MetadataShow full item record
Seasonally Dry Tropical Forests (SDTFs) are growing back in Costa Rica after a heavy deforestation period, but only a small part of them is under formal conservation status. Most secondary forests in the Guanacaste province are located in cattle ranches, and are regularly disturbed by fire, as well as the local practice of using forests as a dry season supplementation for cattle, where cows browse in those forests and find shelter from the harsh dry season conditions. We sampled four secondary forests of the same age in farms under different levels of fire and browse pressure, and compared them to a forest in a protected area in terms of taxonomical composition and ecological diversity, at three different forest strata, using principally a cluster analysis, indicator species and a Non-Metric Multidimensional Scaling (NMDS). Canopy composition and diversity did not differ between protected and unprotected forests, although fire was an influencing factor of canopy composition. Fire was also a discriminating factor for shrubs and saplings, showing different responses in composition to this disturbance, while browsing was the only discriminating factor between floor vegetation clusters. A Mantel correlation analysis returned a significant correlation between all forest strata, while the associations between clusters of the different strata of the forest were variable and not always significant, despite some strong trends. We conclude that, although fire and cattle browsing might have an effect on forests, these are not the main factors that influence forest composition and diversity in SDTFs of this landscape, which could be abiotic factors. Moreover, this study suggests that SDTFs are more resilient to cattle browsing than to fire. We suggest that this practice does not negatively impact forest integrity and can even preserve it, while maintaining local livelihoods. However, a more thorough experiment with a tighter control on all covariables is recommended to confirm these findings.