Tuesday, August 3, 2010

PS 34-87: Spatial and temporal variations of soil respiration in an upland tropical rainforest

Scott Pitz, Katalin Szlavecz, Lijun Xia, Doug Carlson, and Jayant Gupchup. Johns Hopkins University

Background/Question/Methods

Soil respiration is one of the largest fluxes of carbon dioxide to the atmosphere. But little is known how biotic and abiotic factors affect soil respiration. And a compounding issue is that soil respiration can vary greatly on both spatial and temporal scales. Also, human disturbance can have dramatic effects on nutrient cycling and soil respiration.  Wireless sensor networks could help address some of the problems by collecting high-resolution data over multiple scales.

We deployed a wireless sensor network to monitor soil conditions in an upland tropical rainforest in the Ecuadorian Amazon. The network consisted of 12 well method CO2 sampling locations at four sites, two in an old forest and two in a young forest. The young forest was cleared 20 to 25 years ago for a helicopter pad. At each site we collected five-minute readings for soil CO2 concentrations, soil moisture, and soil and air temperatures. Next to each sampling location were ring used for daily chamber flux measurements.  The sampling locations were arranged on two scales.  The spacing at each site was 3 meters and between sites was 25 meters.

Results/Conclusions

The chamber method demonstrated that soil respiration was 30 percent higher, on average in the old forest than the young forest.  The well method’s estimated soil efflux did not match the measured values collected via the chamber method. Often the well method overestimated measured fluxes. This elucidates the need for refinement of well method models so that the predicated values are closer to actual fluxes. The average chamber fluxes measured in the tropical forest (3.07 umol m-2 s-1) were lower than the average flux that we have observed in a temperate forest under similar soil conditions (3.36).  Also, the network was deployed at the end of the dry season, a time of year, which may not be representative of the average annual flux. To fully characterize the soil flux the network needs to be deployed different a different time of the year, preferably the wet season.