Skip to main content


The Agua Salud Project utilizes the Panama Canal’s central role in world commerce to focus global attention on the ecosystem services provided by tropical forests when compared to other types of land cover.  The hydrology portion of the Agua Salud project focuses on studies of the roles of de- and aforestation on water related ecosystem services such as flood and drought mitigation and water purification.

In today’s parlance, the Canal is a “green” operation, powered largely by water (Table 1).  The locks, three pairs on each end with a net lift of 27 meters, are gravity fed.  For each tonne of cargo that is transferred from ocean to ocean, about 13 tonnes of water (m3) are used.   The Canal watershed is also the source of drinking water over 1 million residents of Panama City and Colon City, at either end of the Canal, and numerous towns in between.  Water from the watershed also generates hydroelectricity.  In the Canal watershed, the use of lake water during the dry season is supplemented by groundwater baseflow derived from infiltration of water during the wet season.

About half of the watershed has been deforested, and the official policy in the Canal watershed (Law 21) is to reforest in anticipation of regaining ecosystem services. Land use decisions often necessitate tradeoffs.  Forests consume water, typically losing 150 to 600 mm per year through evapotranspiration compared to other types of vegetation [Jackson and others, 2005]. Where water is a precious commodity, this dichotomy can lead to difficult policy choices.  A recent United Nations report [Canadell and others, 2009] ranks systems research into the hydrological effects of reforestation as especially important in our changing world.  And indeed, the Canal has been a center of attention for the forest-versus-water controversy.  Reports commissioned by the World Bank [see, Calder, 2001] assert that reforestation could be detrimental to the Canal.  This conflict was featured by articles in the Economist [2005-04-21], the New York Times [2005-05-25].

Agua Salud Project

The Agua Salud Project examines this tradeoff of water for trees through an experimental design that focuses on the careful characterization of water movement under different land covers and land uses in the central Canal region.  Figure 1 shows 2003 land cover for the Panama Canal watershed.  Our principal field site includes the Agua Salud Watershed and the headwaters of several adjacent rivers. The region encompasses both protected mature forests and a wide variety of land uses that are typical of rural Panama. Experiments at the scale of entire catchments, typically 10 to 150 ha in area, will permit complete water and carbon inventories and exchanges for different land uses.  We maintain three types of control catchments: mature forest, pasture, and covered with an invasive grass (Saccharum spontaneum – canal grass). The experimental catchments consist of three treatments following pasture: native-species plantation, native second growth, and teak plantation, and a native species plantation following S. spontaneum.  The controls serve to distinguish between effects of tree growth and interannual-climate variation. About 700 ha of degraded land have been purchased to ensure the long-term stability of the experiment. Properties under control of the Agua Salud project are shown in Figure 2.

The project is currently funded through the HSBC climate Partnership and is operated by the Smithsonian Tropical Research Institute (STRI) in close cooperation with the Panama Canal Authority (ACP), which has funded much of the reforestation and some of the meteorological work, and the Panama Environmental Authority (ANAM).  The HSBC Climate Partnership funding ends on 31 December, 2011.  Private donors bought the research land and have leased it to STRI for 20 years, extendable to 40.  The University of Wyoming, The University of Potsdam, STRI, the USGS, and US Army Research Office provided additional funding for personnel and major equipment.

In our research we ask:  How do landscape treatments and management approaches affect ecosystem services and costs such as carbon storage, water quality and quantity, dry-season water supply, and biodiversity? (2) Can management techniques be designed to optimize forest production along with ecosystem services and costs during reforestation? (3) Do different tree-planting treatments and landscape management approaches influence groundwater storage, thought to be critical to maintaining dry-season flow, thus insuring the full operation of the Canal during droughts? (4) Fire is a major component in the landscaping tool kit of the rural agriculturalist that also cements the dominance of the invasive wild sugar cane, S. spontaneum; what are the biological, hydrological, and biogeochemical roles of fire in this landscape? (5) Can we identify treatment approaches, beyond conventional management, that can enhance desired ecosystem services?  Additional work will also address biodiversity, social, and economic values of these forests.

Syntheses and modeling will be employed to up-scale our findings to the entire Panama Canal Watershed.  At the larger scale we ask:   (6) How are water and carbon processed for different major land covers, water covers, and land uses of the entire basin?  (7) What are the operative ecosystem services and costs at the scale of the entire Canal watershed, and how do these relate to the population within the Canal Basin, the operation of the Canal locally and globally, and to the regional and national economy of Panama?

Our study is advanced thanks to prior work by others.   Most areas of the tropics are data poor.   Thanks to the presence of the Panama Canal, our situation is quite different.  The plant and animal biodiversity of the Panama Canal Basin has been characterized by various environmental organizations, including STRI.  The US AID funded STRI, in 1996, to assist the Panamanian government in setting up an environmental monitoring program in the Canal Basin, Canal Basin Monitoring Project (PMCC), with a focus on monitoring the effects of deforestation and urbanization.  These data underpin all aspects of our study at the Canal Basin scale.  Principal results are summarized in Heckadon Moreno, et al. [1999], Condit et al. [2001], and Ibáñez et al. [2001].

Agua Salud Hydrology

Our catchment-level work is highly detailed with the objective of describing processes in sufficient detail that they may be rigorously modeled for current and future conditions. This objective requires the use of “physics-based” hydrologic models (as opposed to empirical models) that are compartmentalized and connected in a realistic way. Each compartment endeavors to rigorously represent its physics, biology, and chemistry. This complexity requires that full physics-based models run on parallel-processing super computers, available to us in the United States. Because models explicitly include the important hydrologic processes, they can make predictions about novel states of a system or about new settings.

Land-use change modeling requires the use of models that are capable of resolving and accurately predicting activation of different water flow paths. Co-Investigator Fred L. Ogden is one of the main developers of the U.S. Army Corps of Engineers, Gridded Surface / Subsurface Hydrologic Analysis (GSSHA) model, which is capable of simulating a range of runoff generation processes from first principles. The GSSHA model was developed specifically for assessing the effects of land-use change on hydrology, erosion, and water quality.

A huge challenge in developing physics-based models is that of characterizing hydrologic processes in the invisible parts of a landscape including deep-soil environments and transient soil states such as the transition from profoundly dry to very wet at the beginning of the wet season. Our field program endeavors to do this characterization.

Dr. Helmut Elsenbeer, of the University of Potsdam, Germany, is focusing on hydrologic processes involving the plant canopy and shallow soil. The approach includes continuous throughfall monitoring using a stratified random design on 1-ha plots, along with a model-based component to characterize interception loss and to describe spatial patterns of throughfall. Along side the throughfall monitoring, overland flow is being described on select hillslopes. Soil matrix hydrology is being characterized by annual surveys of saturated hydraulic conductivity using a stratified random design extending over each focus watershed. We anticipate the first indications of changes in soil hydrology will be evident in this data set.

Small-catchment studies provide an ideal setting for the exploration of subsurface hydrology through a variety of indirect approaches that typically couple modeling to observation. Stream discharge throughout the Canal watershed shows distinct changes that appear to be associated with the initial wetting of soils and the activation of deeper low paths. The characterization of steady-state properties, such as saturation conductivity and porosity cannot be used to describe how soils wet up during the dry season as cracks close, leaf litter decays, and soils are affected by biological actors. These processes are being examined through the use of natural (end-member-mixing analysis or EMMA) and applied tracers (such as LiBr). The modeling of these processes is the responsibility of Dr. Ogden, who collaborates in designing applied tracer experiments with Dr. Robert Stallard. In this type of work even subtle nuances in stream-discharge response to rainfall become significant. Accordingly, most of our study catchments are instrumented with two-stage, V-notched weirs designed to catch such subtleties. The interpretation of deeper flowpaths will be supplemented by data from deep wells for most treatments.

Dr. Stallard of the US Geological Survey and STRI is overseeing the assembly of geographic and hydrologic data for the Agua Salud Region and the Panama Canal Basin as a whole.  Data sets will not only encompass the current state of these watersheds but also historic data documenting changes in land cover and hydrology. The historic data will provide strong model constraints especially when we begin to consider the effects of land-use and climate change for the entire region. These data are being archived by the STRI Bioinformatics Program.

Cooperators and Collaborators

The Agua Salud Project is operated by the Applied Forestry Program in the Center for Tropical Forest Science (CTFS) of the Smithsonian Tropical Research Institute (STRI) in close cooperation with the Panama Canal Authority (ACP) and the Panama Environmental Authority (ANAM). Research collaborators include the National Research Program (NRP) of the Water Resources Discipline (WRD) of the U.S. Geological Survey (USGS), Institute of Geoecology University of Potsdam, Germany, Department of Civil & Architectural Engineering at the University of Wyoming, and the Technological University of Panama.  The project is also part of the Panama Canal Watershed Experiment, which is, in turn, part of the Smithsonian Institution Global Earth Observatory (SIGEO) network.

Results to Date

From November 2007 to present, the Agua Salud project has planted over 150,000 trees.  Our forestry collaborators have mapped and described 108 0.1 ha plots in a 25-year chronosequence of secondary succession plots on our lands.  Over 400 soil samples taken across the project catchments show little variation in texture (over 95 % are classified as clays) and color.  For hydrology, we have established a network of 12 gaged catchments with weirs.  We also have a rain-gage network with nine three-gage clusters, two eddy-flux towers, one in forest and the other in a native-species plantation, one surface-energy-balance station, two complete meteorological stations, and ten shallow groundwater monitoring wells.  For hydrochemistry, we are sampling streamflow using automated samplers, rain, throughfall, overland flow, and groundwater.  In 2009, core staff hiring was completed.  With 2010, the hydrochemistry lab is processing samples, and the Agua Salud research program is fully operational.  Our project is attracting other research groups that are studying  carbon modeling, biodiversity monitoring, microbial biology, and ecosystem-services.

We are also educating large numbers of undergraduate, graduate, and post-doctoral students in the important field of tropical hydrology in collaboration with Panamanian universities.


Calder, I. R., Young, D., and Sheffield, J., 2001, Scoping Study to Indicate the Direction and Magnitude of the Hydrological Impacts Resulting from Land Use Change on the Panama Canal Watershed: Newcastle upon Tyne, U.K., Centre for Land Use and Water Resources Research.  45 p.

Canadell J. G., Ciais P., Dhakal S., Le Quéré C., Patwardhan A., and Raupach M.R., 2009, The Global Carbon Cycle – 2, UNESCO-SCOPE-UNEP Policy Briefs Series, n. 10, Persic, A., editor, ITC Grigny, France, 6 p..

Condit, R., Robinson, W.D., Ibáñez D., R., Aguilar, S., Sanjur, A., Martínez, R., Stallard, R.F., García, T., Angehr, G.R., Petit, L., Wright, S.J., Robinson, T.R., and Heckadon Moreno, S., , 2001, The status of the Panama Canal Watershed and its biodiversity at the beginning of the 21st Century: Bioscience, v. 51, no. 5, p. 389-398.

Heckadon Moreno, S., Ibáñez D., R., and Condit, R., 1999, La Cuenca Del Canal:  Deforestación, Contaminación y Urbanización: Panama, Republic of Panama, Smithsonian Tropical Research Institute.

Ibáñez D., R., Condit, R., Angehr, G.R., Aguilar, S., García, T., Martínez, R., Sanjur, A., Stallard R.F., Wright SJ, Rand A.S., Heckadon Moreno S., 2002, An ecosystem report on the Panama Canal:  Monitoring the status of the forest communities and the watershed: Environmental Monitoring and Assessment, v. 80, p. 65-95.

Jackson, R. B., E. G. Jobbágy, R. Avissar, S. B. Roy, D. J. Barrett, C. W. Cook, K. A. Farley, D. C. le Maitre, B. A. McCarl, and B. C. Murray, 2005, Trading water for carbon with biological carbon sequestration. Science, v. 310, 1944-1947.


Table 1. Panama Canal Average Water Budget
Annual discharge km3 yr-1 Average discharge m3 sec-1
Total Runoff: 4.4 139
Lockages1: 2.6 82
Hydroelectricity2 1.2 39
Drinking: 0.27 9
L+H+D = 4.1 130
Total, 1982: 3.3 105
1. 37 per day at 191,000 cubic meters per lockage.
2. Power generation is a buffer. When water is in short supply, electricity is either purchased from the power grid or generated on site using fossil fuels.

Figure 1
2003 land cover map for the Panama Canal Watershed (ANAM/ACP).

Figure 2
Location of the STRI Agua Salud Project Properties. The grid is UTM, NAD27 Canal Zone, with 1 km grid spacing.

Figure 3
Site map.  Left panel – Panama, gray rectangle indicates right panel.  Right panel –sites used in the Agua Salud Project.

  1. Barro Colorado – A long-term research site of STRI, with two research catchments.
  2. Agua Salud – gray rectangle, the primary research site for this project, a typical rural landscape with 650 ha of project lands on which 150,000 trees have been planted
  3. Gamboa – site of STRI labs accessing the Agua Salud through Soberania National Park.
  4. Sardinilla – Site of the McGill University Carbon Project, a biodiversity plantation of over 10,000 trees on former pasture.
  5. Ciudad del Arbol – site of an ACP – University of Panama native-species reforestation following Saccharum spontaneum.
  6. Prorena – site of a STRI native species reforestation following S. spontaneum.