Hydrological and geochemical studies in the seasonal tropics of Panama over the past six years have revealed runoff behaviors not seen in temperate climates. This integrated international research program combines hydrological, geochemical, and geophysical measurements in the seasonal tropics including in situ data collection and analysis, enhanced by the use of natural and introduced tracers and modeling. The overall objective of this collaborative research program is to understand important runoff generating processes in the seasonal tropics and the roles of seasonal transitions and land-use changes on runoff mechanism thresholds. Hydrologic, geochemical and isotopic data collected over a range of scales will compare the virtually pristine 414 km2 old-growth Upper Rio Chagres and adjacent 330 km2 largely deforested Rio Pacora watersheds to examine land-use change effects. Different bedrock types underlay these basins, so geochemistry and isotopes will be used to assist in identifying flow paths and residence times and in model and hypothesis testing. The humid tropics cover 22% of the Earth's land surface and are home to 36% of humanity. The selected study area is representative of the humid tropics promoting transferability of the knowledge gained to a geographically large region. There are three very important questions that this research will help answer. What is the effect of deforestation or reforestation on the water yields from watersheds in the seasonal tropics? Why does runoff generation behave differently early in the rainy season than at other times of the year? How can we better predict these effects by improved hydrologic models? The Panama Canal watershed is extremely important for world commerce because the runoff from the watershed drives the Panama Canal, which is vitally important to the United States. These research results will be communicated to the Panama Canal Authority, Panamanian Universities, and broadly disseminated through peer-review articles. An international field course on tropical hydrology will be held each year of the project and will be open to qualified U.S. and Panamanian graduate and undergraduate students. The project compliments the Smithsonian Tropical Research Institute (STRI) Panama Canal Watershed Experiment, and STRI is a collaborator, as are the Panama Canal Authority and the Technological University of Panama.
This project studied land-use effects on water resources availability in the Panama Canal Watershed (PCW). PCW water resources face two extremes. They can be scarce during the pronounced December-April dry season yet excessive during the May-November rainy season. Panama Canal operations depend on reliable water yields from the PCW, particularly as the Panama Canal undergoes expansion that will increase water use. Water resources management in the PCW is critical to the United States economy as 15-19% of U.S. trade with Asia traverses the Panama Canal. Understanding land use effects on water resources availability is also important for countries in similar settings such as Caribbean island nations that depend on small drought and flood sensitive watersheds for their water supplies. Our tests using simulated rainfall showed that previous work studying the effects of land use on soil infiltration capacity was hampered by methodology. Previous investigators used only 8.9 cm diameter soil cores, while our rainfall simulator rained on a 2m x 6m area and measured much higher infiltration rates in both forest and pasture land uses compared to the results from the small soil cores. We visualized flow paths in the soil using Electrical Resistivity Imaging (ERI) while applying simulated rainwater with salt tracers. Applied simulated rain water that infiltrated in the pasture moved vertically to a depth of about 0.5 m then spread laterally. We observed surface runoff and flow through large decayed root channels (macropores) in the pasture. Chemical analysis of runoff and macropore water for the salt tracers in the applied water showed that this flow was mostly our applied water. In contrast, a similar test in a 30-year old forest showed that infiltrated rainfall moved downward about 1.5-2 m, and we observed no runoff and no macropore flow, even after applying 60 cm (23 inches) of rain in two hours. In 2013 and 2014, we deployed automatic water samplers for streams and rainwater. Stream water samplers were triggered to start sampling during storms as stream levels rose. We also sampled shallow groundwater, springs, and flowing streams between storms. These water samples allowed us to use naturally-occurring geochemical and stable water isotope tracers to help identify water sources and flow paths taken by rain water as it falls through vegetation canopies, travels in the soil, and mixes with some groundwater to form stream water. Pre-rainfall stream flow has a strong groundwater geochemical signal, while rainfall that quickly enters the stream as surface runoff exhibits a different geochemical signature similar to rainfall and the soil through which it flowed. The ratio of the geochemical and isotopic signals of groundwater and rainfall helped us identify flow paths. Natural geochemical and isotope tracer analyses suggest that larger rainfall events disproportionately contribute greater runoff by activating an additional flowpath through the soil that is not activated during smaller, more frequent rainfall events. Despite considerably smaller flood peaks in catchments covered by mature forest, our tracer analyses showed that the relative contributions of rainwater and groundwater were similar between a mature forest and subsistence agricultural land covers. The final results of our project indicate that grazing does not significantly reduce the infiltration capacity of the soil surface or the deeper depleted clay (saprolite) soils in the PCW. Rather, deforestation changes the flow path taken by rainfall on its trip to becoming streamflow. This is because deforestation promotes lateral downslope flow through soils whereas root-dense forests promotes vertical soil-water flow. We have published two peer-review papers as of the end of the project are preparing several additional papers for submittal. Furthermore we will use our results to inform process-based hydrological model developments aimed at improved predictions of land-use change effects on future water resources availability in the PCW and similar tropical settings as a function of land use change scenarios. This project engaged two Ph.D. students and two M.S. students in the use of sophisticated hydrologic technologies in rugged tropical conditions for extended field campaigns. These students interacted with project collaborators from the Smithsonian Tropical Research Institute and the Technological University of Panama. All graduate students are leading preparation of peer-review publications for submittal under the guidance of the project principal investigators.
