Recent research in Costa Rica has shown that the growth of tropical moist forest trees is strongly related to the strength of the dry season across a range of rainfall that by all standard measures would be considered more than sufficient for growth. Understanding the basis for this finding is crucial because tropical moist forests play important roles in the global carbon and water cycles and therefore regional and global climate. Predicted future warming and drying of tropical moist forests could have strong feedback effects on global climate. However, recent satellite studies from the Amazon report that trees have their greatest leaf cover during the dry season, suggesting that the dry season does not affect production. These two findings suggest a serious gap in our understanding of role of the dry season on tropical tree growth and carbon sequestration. Dry-season controls on tree growth could result from both direct water limitation or via other indirect mechanisms. Direct limitation via dry soil and dry air seems unlikely given the high rainfall, soil moisture, and humidity that occur during the dry season. Potential indirect mechanisms include excessive leaf temperatures lowering photosynthesis, premature leaf loss in response to dry air, low light-use efficiency and canopy photosynthesis during periods of direct sun, and changes in allocation of resources to root biomass or reproduction instead of aboveground growth. This project will test hypotheses of direct water limitation and indirect mechanisms to determine the basis for the dry season rainfall correlation with growth. The study will use both observational and experimental approaches to test for direct water limitation, combining measurement of whole forest photosynthesis and the form of energy losses, leaf-and whole-tree water use, monthly tree growth and climate measurements, and a dry-season water-addition experiment. Indirect limitations will be evaluated by measurements of canopy leaf area, leaf temperature, and measurements of root production and metabolism. Measurements of growth and ecosystem carbon, water and energy balance will be used to evaluate the importance of dry season growth reduction for regional climate.
This project will support outreach and education components in both South Florida and Costa Rica. In South Florida a partnership has been developed with a local private grade 1-8 school for environmental outreach activities. Project personnel will continue to support the Fairchild Challenge school environmental awareness program at Fairchild Tropical Botanical Garden in Miami. The project will also support the environmental education programs of the Organization for Tropical Studies at the La Selva Biological Station, the location of the study. The project will support one postdoctoral fellow, several undergraduates, and provide training for two Costa Rican technicians. Finally, understanding the basis of the sensitivity of dry season limitation on aboveground biomass production is critical to forecasting the responses of tropical wet forests to future climate regimes.
Tropical rain forests (Figure 1) play a key role in the global carbon cycle contributing ~60% of forest biomass and ~27% of the soil carbon. Rain forests are thought to contribute 20-40% of the global terrestrial photosynthesis. Tropical forests are key drivers of global climate both through carbon and energy balances. The current and future carbon and water balances of rain forests are of considerable interest to scientists and policymakers alike. Correct parameterization of the responses of tropical forest in global system models and their validation is critical for future climate projections of the Earth System. The controls on tropical tree growth and their sequestration of carbon is one of the great uncertainties of earth system science and have tremendous implications for the response of the earth system to change. Despite the critical need for information on the topic, data on the controls of tropical tree growth and ecosystem productivity are few and largely restricted to correlations of adult tree growth with temperature and growth experiments with seedlings. Long-term plot-level data of rain forest trees in Central America show independent growth responses to two factors, nighttime minimum temperature (negative) and dry season rainfall (positive), even though the site has relatively high dry season rainfall. That the relationship extends up to very high rainfall amounts suggests that tree growth is not driven by direct water limitation, but rather some other factor(s) correlated with dry season rainfall. In this project we experimentally tested the role of direct soil water limitation on tropical tree growth using a dry season irrigation experiment (Figure 2). We also investigated other potential controlling factors that correlate with dry season rainfall. Increased soil moisture did NOT increase diameter growth of experimentally irrigated trees; both irrigated and control trees continued to show a reduction in growth during the dry season. Estimates of leaf litterfall (leaf loss) during the dry season showed similar patterns between irrigated and control trees, as did plot leaf area coverage as estimated by canopy photography (Figure 3). Our irrigation experiment directly tested whether tree growth is controlled by soil moisture limitation. The experimental treatments were successful in maintaining high levels of soil moisture on the water addition plots. The results from the 3 years of treatment were consistent among years and conclusive; tree growth declines during the dry season regardless of the soil moisture availability. This is not to say that soil moisture limitation never reduces tree growth in the Central American Wet Tropics. However, the strong reduction in tree growth during the dry season seen each year is not a generally a result of soil moisture limitation. By eliminating soil moisture limitation as a primary controlling factor for tree growth, we are left with two likely causes of lower growth during the dry season. Tree sensitivity to high vapor pressure deficits (VPD, the dryness of the air compared to how much water vapor the air can hold) during the dry season and differences in direct and diffuse light (cloudiness) in the dry season compared to the remainder of the year. Evidence from this project and previous work points to a strong role of VPD controlling tree growth. Tree responses were most closely aligned to control by high vapor pressure deficits that occur during the dry season, that is, the arrival of dry air masses reducing photosynthesis. Our diameter measurements indicate a negative relationship between tree growth and VPD of the air. Our leaf level work has shown strong reductions in photosynthesis with higher VPD. The diameter measurements also hint at a possible secondary limiting role of quantity of diffuse light (as influenced by cloudiness) on canopy photosynthesis. However, these relationships between growth and VPD and diffuse light are correlations among factors that are themselves strongly correlated and point to the strong need for experimental tests of the potential limiting factors. Nevertheless, these findings greatly clarify the growth response of tropical trees to climate and provide clear directions for modeling the control of carbon storage in these forests. This project has provided direct opportunities for research to undergraduate assistants hired on the FIU campus, as well as exposure to the research for undergraduate and graduate students in four different courses taught at FIU. The project postdoc has had tremendous opportunities to increase his knowledge base, teaching and mentoring experience, and networking. K 6-8 grade students and their science teacher were directly exposed to tropical ecology and the project, as well as nearly all the projects ongoing at the La Selva Biological Station in Fall 2012, when their teacher conducted a research expedition to the site. Our research was presented to many researchers and a substantial number of courses and education groups, as well as reporters who visited the site. An annual open house at the station exposed hundreds of local citizens to our research.
