Partitioning soil respiration: examining the artifacts of the trenching method.

Thursday, 18 December 2014
Kathleen E Savage1, Eric A Davidson1, Adrien Finzi2, Marc-Andre Giasson2 and Richard A Wehr3, (1)Woods Hole Research Center, Falmouth, MA, United States, (2)Boston University, Boston, MA, United States, (3)University of Arizona, Tucson, AZ, United States
Soil respiration (Rs) is a combination of autotrophic (Ra) and heterotrophic respiration (Rh). Several methods have been developed to tease out the components of Rs, such as isotopic analyses, and removing Ra input through tree girdling and root exclusion experiments. Trenching involves severing the rooting system surrounding a plot to remove the Ra component within the plot. This method has some potential limitations. Reduced water uptake in trenched plots could change soil water content, which is one of the environmental controllers of Rs in many ecosystems. Eliminating root inputs could reduce heterotrophic decomposition of SOM via lack of priming. On the other hand, the severed dead roots may temporarily increase available carbon substrate for Rh.

At the Harvard Forest, MA, we used the trenching method to partition Rs into its components Ra and Rh. Pre-trenched Rs was measured from spring to fall of 2012. In late fall of 2012, a trench was excavated to 1m depth around a 5x5m area, severing all roots. Plastic tarp was placed along the trench walls and then backfilled. Four automated Rs chambers were placed in the trenched plot and four in an un-trenched plot. Respiration was measured hourly for each chamber along with soil temperature and moisture from spring through fall of 2013 and 2014. Eighty root decomposition bags were placed in the organic soil horizon of the trenched (40) and un-trenched (40) plots at the time of trenching in 2012 and measured in 2013 and 2014. These data are being used to estimate the size and duration of any artifact due to root death. As expected, Rs was lower in the trenched plot (Rh only) than in the un-trenched plot (Rh + Ra) in 2013, but the reverse was unexpectedly observed during a period of low precipitation in 2014. High rates of ET combined with below-average precipitation dried the un-trenched plot to a point where Rh was inhibited, whereas less ET allowed the un-trenched plots to remain measurably wetter.