The DIRT on Q10: Differential temperature response of soils depleted of labile inputs

The decomposition of soil organic carbon (SOC) is expected to increase with global warming.  SOC decomposition has been commonly described by kinetic models with at least two pools with slow and fast turnover times.  The fast pool is thought to consist of labile substrates, while the slow pool consists of more chemically complex or resistant compounds which decompose more slowly and represent more stable SOC.  If the slow pool is indeed more chemically complex, then it’s necessarily higher activation energy should lead to a higher sensitivity (Q₁₀) to temperature and a proportionally larger response to warming than the fast, or labile, pool.  Experimental tests of the relative Q₁₀ of the fast and slow pools have been inconclusive and frequently contradictory in part due to two major challenges:

1. All pools are decomposing simultaneously, making it difficult to distinguish between the response of the fast and slow pools
2. Maintaining soils under differing conditions (such as temperature) over long periods of time may cause divergence in more than the Q₁₀ response (such as carbon quality, microbial communities, etc.), making the treatments less comparable over time.  We present here a test of the temperature response on soils from a 20 yr litter manipulation experiment incubated under an experimental regime that minimizes divergence among the soils over time.

We hypothesized that

1. If exclusion of inputs has depleted the fast/labile pool
2. The slow pool is more chemically complex, then the input exclusion treatments should show a higher Q₁₀ compared to the ambient or increased input treatments. 

The soils are taken from the Detritus Input and Removal Treatment (DIRT) plots in the Bousson Forest, Pennsylvania, US.  The DIRT treatments consist of litter and root exclusion (no inputs = NI), no roots (NR), no litter (NL), double litter (DL), and ambient conditions (C).  In the laboratory soils were incubated at 25^oC for 142 days.  Periodically, replicate sets were rotated into 15^oC, 35^oC or remained at 25^oC for 24 hr.  The headspace CO₂ concentration was measured before and after the 24 hr temperature treatments, and then all replicate sets were returned to 25^oC.   

Input exclusion decreased respiration rate, with NI < NR = NL < C = DL throughout the duration of the incubation.  Input exclusion also similarly decreased total carbon content.  The respiration rate at 25^oC remained consistent across all replicate sets throughout, indicating that the temperature treatment rotations do not cause the soils to diverge from each other.  Our preliminary data indicates the Q₁₀ of soil respiration was similar among the DIRT treatments, despite the clear differences in their carbon pools.  Other studies that have investigated this phenomenon have often examined the temperature effects of depleting labile soil carbon pools through time in laboratory incubations, instead of beginning with a single soil with very different initial labile soil carbon pools.  While we continue to incubate these samples to further deplete their soil carbon, our results to date strongly suggest that different soil carbon pools have the same relative temperature responses.  Recent studies have suggested that classically conceived chemically recalcitrant carbon does not exist in soils, which would explain our results relative to the predictions of kinetic theory.