Article II. Historical land use cover and change and its effects on hydrological processes in the upper Reventazón River, Costa Rica
Across the tropics, and the world, land use change has been a primary result of the development with major impacts on hydrological processes. In this study, I combined land use parameterization (e.g., forest, coffee, sugar and pasture) derived from field observations to investigate the interplay of antecedent moisture content (AMC), infiltration, surface runoff and percolation using a one-cell model. To look for similar patterns at the watershed scale, I performed a supervised land use classification of two satellite images (1986 and 1996) of the upper Reventazón watershed in Costa Rica. While minimal land use change occurred from 1986 to 1996, seven nested watersheds were classified as either forested (>85% forested) or mixed-use watersheds (<50% forested). In the one-cell modeling, at AMC of field capacity or greater, sugar and pasture initiated percolation quicker and with less cumulative rainfall than the forest and coffee at lower precipitation intensities (<3.5 cm/hr). At the wetter moisture contents in the forest and coffee, increased lateral flow (matrix and macropore) delayed the initiation of percolation. Also at the greater AMC, sugar and pasture initiated surface runoff more quickly and with less cumulative precipitation due to their lower conductivities (0.7 and 1.3 cm/hr respectively). With high precipitation intensity (10 cm/hr), surface runoff was greater in the sugar and pasture than in the forest and coffee. At the same intensity, no percolation occurred within the sugar and pasture. At the drier AMC, much more distinct differences were observed between the forest, coffee, sugar and pasture. At the drier AMC, hydrological connectivity throughout the entire soil profile was disrupted in the sugar and pasture. While the sugar and pasture initiated surface runoff quickly and with less cumulative precipitation than in the forest and pasture, percolation was delayed in the sugar and pasture due to the lower conductivities resulting in a smaller vertical flow to the deeper, soil horizons. In the forest and coffee, greater conductivities allowed for the rapid transfer of water throughout the soil profile. In general, the one-cell modeling showed that differences in surface runoff and percolation between land uses diminished with increasing moisture content and precipitation intensity. Basic hydrological analyses (Flow Duration Curves) and runoff coefficients) showed that the forested watersheds had much greater water yields with less variability than mixed-land use watersheds. However, on average, the forested watersheds also received more precipitation due to elevation and regional microclimates. Due to a good regional cover of data, but a poor temporal resolution, model results were compared to monthly discharge values within the seven watersheds. The majority of the simulations predicted discharge better than using the mean of the data. Similar results to the one-cell modeling were observed at the watershed scale with greater amounts of percolation occurring at the transition from the dry season to the wet season. Also, during the transition to the wet season, surface runoff occurred much earlier in the mixed-land use watersheds. Finally, greater percolation within the forested watersheds led to greater baseflow contribution to river discharge. Therefore, while bedrock topography may provide a quantitative depth threshold for subsurface connectivity and ultimately lateral flow, in these deep tropical soil systems, the soil and its conductivity may control the rate at which this depth is exceeded. This connection may be particularly important during the transition from dry to wet season.