Soil carbon dynamics in primary and secondary tropical forests in Colombia

Soils play a central role in the dynamics of biospheric carbon and in climate change. They contain the largest carbon stock of terrestrial ecosystems and return to the atmosphere a significant proportion of carbon fixed by photosynthesis. Soils of tropical forests are tremendously important in the carbon cycle because they receive the largest organic matter inputs, they have the largest respiration rates, and they are among the largest carbon reservoirs among world soils. This research assesses the main components of the soil carbon dynamics in primary (PF) and secondary (SF) tropical forests in Colombia. I evaluated the production, stocks, and decomposition rates of aboveground detritus as well as the stocks, growth, mortality, and decomposition of fine roots in these two forest types. Soil carbon outputs were evaluated as total soil, heterotrophic, and root respiration. The stocks of soil organic carbon down to 4 m deep in these two cover types and in degraded pastures (PAS) were also evaluated. ^ Soil inputs of organic carbon from above and belowground sources were lower in SF than in PF. Litterfall in SF was 58% and production of fine root detritus was 60% of that in PF. When production of woody detritus and palm fronds was considered, the difference between these forest types was even larger. However, outputs of mineral carbon through heterotrophic soil respiration were similar; in SF they equaled 97% of those in PF. As a result, soil carbon balance was positive in PF and negative in SF. Despite that soil carbon balances suggest that soils of SF are losing carbon, soil carbon stocks of SF were higher than of degraded pastures, suggesting that they have already started to recover soil carbon stocks lost under degraded pastures. This discrepancy can be partially explained by the effect of drier conditions on heterotrophic soil respiration as a consequence of a moderate El Niño event during the period of soil respiration measurements. The positive carbon balance in soils of PF despite the El Niño event, suggests that soils of PF accumulated about 664 Kg C ha−1 yr−1. Therefore, soil carbon dynamics mainly depended on successional status of vegetation and on climatic conditions. ^

Connectivity of fragmented amphibian populations in a neotropical landscape

A high proportion of amphibian species are threatened with extinction globally, and habitat loss and degradation are the most frequently implicated causes. Rapid deforestation for the establishment of agricultural production is a primary driver of habitat loss in tropical zones where amphibian diversity is highest. Land-cover change affects native assemblages, in part, through the reduction of habitat area and the reduction of movement among remnant populations. Decreased gene flow contributes to loss of genetic diversity, which limits the ability of local populations to respond to further environmental changes. The focus of this dissertation is on the degree to which common land uses in Sarapiquí, Costa Rica impede the movement of two common amphibian species. First, I used field experiments, including displacement trials, and a behavioral landscape ecology framework to investigate the resistance of pastures to movement of Oophaga pumilio. Results from experiments demonstrate that pastures do impede movement of O. pumilio relative to forest. Microclimatic effects on movement performance as well as limited perceptual ranges likely contribute to reduced return rates through pastures. Next, I linked local processes to landscape scale estimates of resistance. I conducted experiments to measure habitat-specific costs to movement for O. pumilio and Craugastor bransfodrii, and then used experimental results to parameterize connectivity models. Model validation indicated highest support for resistance estimates generated from responses to land-use specific microclimates for both species and to predator encounters for O. pumilio. Finally, I used abundance and experiment-derived resistance estimates to analyze the effects of prevalent land uses on population genetic structure of the two focal species. While O. pumilio did not exhibit a strong response to landscape heterogeneity and was primarily structured by distances among sites, C. bransfordii genetic variation was explained by resistance estimates from abundance and experiment data. Collectivity, this work demonstrates that common land uses can offer different levels of resistance to amphibian movements in Sarapiquí and illustrates the value of investigating local scales processes to inform interpretation of landscape-scale patterns.^

Connectivity of fragmented amphibian populations in a Neotropical landscape

A high proportion of amphibian species are threatened with extinction globally, and habitat loss and degradation are the most frequently implicated causes. Rapid deforestation for the establishment of agricultural production is a primary driver of habitat loss in tropical zones where amphibian diversity is highest. Land-cover change affects native assemblages, in part, through the reduction of habitat area and the reduction of movement among remnant populations. Decreased gene flow contributes to loss of genetic diversity, which limits the ability of local populations to respond to further environmental changes. The focus of this dissertation is on the degree to which common land uses in Sarapiquí, Costa Rica impede the movement of two common amphibian species. First, I used field experiments, including displacement trials, and a behavioral landscape ecology framework to investigate the resistance of pastures to movement of Oophaga pumilio. Results from experiments demonstrate that pastures do impede movement of O. pumilio relative to forest. Microclimatic effects on movement performance as well as limited perceptual ranges likely contribute to reduced return rates through pastures. Next, I linked local processes to landscape scale estimates of resistance. I conducted experiments to measure habitat-specific costs to movement for O. pumilio and Craugastor bransfodrii, and then used experimental results to parameterize connectivity models. Model validation indicated highest support for resistance estimates generated from responses to land-use specific microclimates for both species and to predator encounters for O. pumilio. Finally, I used abundance and experiment-derived resistance estimates to analyze the effects of prevalent land uses on population genetic structure of the two focal species. While O. pumilio did not exhibit a strong response to landscape heterogeneity and was primarily structured by distances among sites, C. bransfordii genetic variation was explained by resistance estimates from abundance and experiment data. Collectivity, this work demonstrates that common land uses can offer different levels of resistance to amphibian movements in Sarapiquí and illustrates the value of investigating local scales processes to inform interpretation of landscape-scale patterns.