Contrary to the general trend in the tropics, Puerto Rico underwent a process of agriculture abandonment during the second half of the 20th century as a consequence of socioeconomic changes toward urbanization and industrialization. Using data on landuse change, biomass accumulation in secondary forests, and ratios between gross domestic product (GDP) and carbon emissions, we developed a model of the carbon budget for Puerto Rico between 1936 and 2060. As a consequence of land abandonment, forests have expanded rapidly since 1950, achieving the highest sequestration rates between 1980 and 1990. Regardless of future scenarios of demography and land use, sequestration rates will decrease in the future because biomass accumulation decreases with forest age and there is little agricultural land remaining to be abandoned. Due to high per-capita consumption and population density, carbon emissions of Puerto Rico have increased dramatically and exceeded carbon sequestration during the second half of the 20th century. Although Puerto Rico had the highest percent of reforestation for a tropical country, emissions during the period 1950–2000 were approximately 3.5 times higher than sequestration, and current annual emission is almost nine times the rate of sequestration. Additionally, while sequestration will decrease over the next six decades, current socioeconomic trends suggest increasing emissions unless there are significant changes in energy technology or consumption patterns. In conclusion, socioeconomic changes leading to urbanization and industrialization in tropical countries may promote high rates of carbon sequestration during the decades following land abandonment. Initial high rates of carbon sequestration can balance emissions of developing countries with low emission/GDP ratio. In Puerto Rico, the socioeconomic changes that promoted reforestation also promoted high-energy consumption, and resulted in a net increase in carbon emissions.

Secondary forests are becoming increasingly widespread in the tropics, but our understanding of how secondary succession affects carbon (C) cycling and C sequestration in these ecosystems is limited. We used a well-replicated 80-year pasture to forest successional chronosequence and primary forest in Puerto Rico to explore the relationships among litterfall, litter quality, decomposition, and soil C pools. Litterfall rates recovered rapidly during early secondary succession and averaged 10.5 (± 0.1 SE) Mg/ha/y among all sites over a 2-year period. Although forest plant community composition and plant life form dominance changed during succession, litter chemistry as evaluated by sequential C fractions and by 13C-nuclear magnetic resonance spectroscopy did not change significantly with forest age, nor did leaf decomposition rates. Root decomposition was slower than leaves and was fastest in the 60-year-old sites and slowest in the 10- and 30-year-old sites. Common litter and common site experiments suggested that site conditions were more important controls than litter quality in this chronosequence. Bulk soil C content was positively correlated with hydrophobic leaf compounds, suggesting that there is greater soil C accumulation if leaf litter contains more tannins and waxy compounds relative to more labile compounds. Our results suggest that most key C fluxes associated with litter production and decomposition re-establish rapidly—within a decade or two—during tropical secondary succession. Therefore, recovery of leaf litter C cycling processes after pasture use are faster than aboveground woody biomass and species accumulation, indicating that these young secondary forests have the potential to recover litter cycling functions and provide some of the same ecosystem services of primary forests.

This paper compares aboveground forest structure and macronutrient stoichiometry over 5 15 years of hurricane induced secondary succession by species, life history groups, community 6 species composition, and geomorphic setting. Stem density continually increased after the 7 impact of the Category 4 hurricane Hugo and 15 years later, it was greater than pre-hurricane. 8 There were significant spatial and temporal differences in the number of species, the diversity 9 index per plot, forest structure, and biomass. The greatest compositional differences occurred 10 between the post-Hugo and the 15-year census. Prior to hurricane Hugo most plots had very 11 similar species composition and abundances, and thus occupied a small area in non-metric 12 multidimensional species space. Following the hurricane new species combinations occurred 13 and the location of plots was spread in multidimensional space. Diversity indices were 14 significantly different among geomorphic settings before and immediately after hurricane Hugo. 15 However, these differences were not observed again until the 15-year census where they returned 16 to pre-hurricane levels. Plant associations based on abundance, life history traits, and landscape 17 position had measurable differences in their structure, composition, aboveground nutrient 18 storage, and stoichiometry. However, these differences were reflected in a variety of ways at 19 different spatial scales. At the species level differences in macronutrient tissue concentrations 20 were apparent when comparing co-existing primary forests dominants, early successional 21 dominants, high-light and low-light species, and species whose stem densities are negatively 22 correlated. Community level differences were greater for forest structure and total nutrient 23 storage compared to the mass weighted concentrations of macronutrients. The largest differences observed were in Mg and can be attributed to the succession of pioneer species 2 following the hurricane. Over the entire 15-year period, the watershed average aboveground 3 stoichiometry was relatively consistent and this is linked to the biomass dominance of a few 4 species. The successional history recorded here also suggests that community level differences 5 in species composition, structure, and stoichiometry were well established after 10 to 15 years of 6 secondary succession.