Vulnerability of tropical forest ecosystems and forest dependent communities to droughts.

Energy captured by and flowing through a forest ecosystem can be indexed by its total Net Primary Productivity (NPP). This forest NPP can also be a reflection of its sensitivity to, and its ability to adapt to, any climate change while also being harvested by humans. However detecting and identifying the vulnerability of forest and human ecosystems to climate change requires information on whether these coupled social and ecological systems are able to maintain functionality while responding to environmental variability. To better understand what parameters might be representative of environmental variability, we compiled a metadata analysis of 96 tropical forest sites. We found that three soil textural classes (i.e., sand, sandy loam and clay) had significant but different relationships between NPP and precipitation levels. Therefore, assessing the vulnerability of forests and forest dependent communities to drought was carried out using data from those sites that had one of those three soil textural classes. For example, forests growing on soil textures of sand and clay had NPP levels decreasing as precipitation levels increased, in contrast to those forest sites that had sandy loam soils where NPP levels increased. Also, forests growing on sandy loam soil textures appeared better adapted to grow at lower precipitation levels compared to the sand and clay textured soils. In fact in our tropical database the lowest precipitation level found for the sandy loam soils was 821 mm yr(-1) compared to sand at 1739 mm yr(-1) and clay at 1771 mm yr(-1). Soil texture also determined the level of NPP reached by a forest, i.e., forest growing on sandy loam and clay reached low-medium NPP levels while higher NPP levels (i.e., medium, high) were found on sand-textured soils. Intermediate precipitation levels (>1800-3000 mm yr(-1)) were needed to grow forests at the medium and high NPP levels. Low thresholds of NPP were identified at both low (∼750 mm) and high precipitation (>3500 mm) levels. By combining data on the ratios of precipitation to the amount of biomass produced in a year with how much less precipitation input occurs during a drought year, it is possible to estimate whether productivity levels are sufficient to support forest growth and forest dependent communities following a drought. In this study, the ratios of annual precipitation inputs required to produce 1 Mg ha(-1) yr(-1) biomass by soil texture class varied across the three soil textural classes. By using a conservative estimate of 20% of productivity collected or harvested by people and 30% precipitation reduction level as triggering a drought, it was possible to estimate a potential loss of annual productivity due to a drought. In this study, the total NPP unavailable due to drought and harvest by forest dependent communities per year was 10.2 Mg ha(-1) yr(-1) for the sandy textured soils (64% of NPP still available), 8.4 Mg ha(-1) yr(-1) for the sandy loam textured soils (60% available) and 12.7 Mg ha(-1) yr(-1) for the clay textured soils (29% available). Forests growing on clay textured soils would be most vulnerable to drought triggered reductions in productivity so NPP levels would be inadequate to maintain ecosystem functions and would potentially cause a forest-to-savanna shift. Further, these forests would not be able to provide sufficient NPP to satisfy the requirements of forest dependent communities. By predicting the productivity responses of different tropical forest ecosystems to changes in precipitation patterns coupled with edaphic data, it could be possible to spatially identify where tropical forests are most vulnerable to climate change impacts and where mitigation efforts should be concentrated.