The dynamics and carbon implications of fires in the Andes

Fire in natural ecosystems is a major agent of biome structure formation, biome degradation, atmospheric pollution and carbon emission. In this proposal we will focus attention on fire dynamics on a potentially important but barely studied ecosystem: the tropical montane cloud forests (TMCFs) of the Andes. In the Andes, humid TMCFs sit immediately below highly flammable, high altitude dry grasslands (the puna) that have suffered from recurrent anthropogenic fires for millennia, with the treeline sitting at approximately 3000 m. This treeline is a zone of ecological and climatic tension: on the one hand, rising temperatures and cloud heights may have a tendency to push the ecotone upwards, encouraging forest expansion into the puna. On the other hand, increased aridity in the puna (driven by rising temperates and evapotranspiration, and possibly by reducing precipitation), coupled with intensified human pressure, is increasing fire occurrence and penetration into the cloud forest. There is plentiful evidence that such fires are now widespread in Andean TMCFs However, there have been few systematic studies of these fire events because of the remoteness of this study system. Our analysis of satellite data has shown that Andean fires may be increasing in abundance, and display a mysterious five year cycle. The tendency to increased fires is particularly marked at elevations above 2000m, and may be linked to increasing water stress as potential transpiration rates rise in response to rising temperatures. Given the increasing occurrence of fire, we conjecture than an increasing number of fires are therefore moving down from the puna region towards the surrounding TCMFs, which are being degraded into grasslands or shrub communities. TMCFs contain a large stock of carbon, and their loss through fire is a potential large source of carbon dioxide and other combustion compounds to the atmosphere. Both the carbon stock and the loss on ignition are very comparable to those from a lowland Amazonian rainforest, but in the mountains almost all of this loss comes from soil organic layers. In this proposal we will conduct the first detailed evaluation of spatial and temporal dynamics of fire at the Andean treeline, and attempt to quantify their implications for carbon emissions. We will do this by focussing on an intensive study region in the southern Peruvian Andes (in and around Manu National Park), where we will determine the environmental controls on fire ignition and spread, quantify the carbon emissions and recovery times associated with specific fires, and evaluate the potential of remote sensing data to scale up fire emissions to the wider Andean region. Our project will involve the following componets: 1. Construction of the fire detailed fire history and fire map of a region of the Andean treeline, covering the period 1990-2011; 2. Conducting intensive tracking and evaluation of all fires in our study region in the period 2009-2011. 3. Measuring the biomass and carbon stocks in unburned and burned forests 4. Installing some carbon plots in forests prior to burn, and visiting them afterwards 5. Conducting drying and ignition experiments to determine thresholds of flammability at the treeline 6. Using our detailed field data to evaluate and calibrate satellite-based estimates for fire occurrence, extent and intensity 7. Using calibrated satellite data to estimate carbon emissions from fires across the tropical Andes