Woody detritus decomposition in a shrub-invaded grassland: interactions among soil deposition, termites and radiant energy
Decomposition, a fundamental driver of biogeochemical cycling, is strongly influenced by land-cover. Changes in vegetation composition (e.g., shrub encroachment) alter the quantity and quality of organic matter inputs, while concurrently affecting microclimate, trophic structure, and soil transport. We sought to quantify how rates of woody detritus (WD) decomposition would vary in a shrub-encroached Sonoran Desert grassland as a function of spatial patterns of soil-litter mixing, termite activity, and vegetation cover. We hypothesized that the breakdown of WD would be highest under shrub canopies where surface termite activity is greatest. Alternatively, if WD decomposition is driven by abiotic factors, then mass loss would be greatest away from shrub canopies where soil-litter mixing, temperatures, and solar radiation levels are highest. We tested these hypotheses by placing WD [Quercus sp. dowels and air-dried green stems of local mesquite (Prosopis velutina)] of three diameter classes (<0.99 cm, 1-2.49 cm, 2.5-10 cm) under mature mesquite (shaded) and in paired intercanopy areas (full sun) in January 2012. A factorial soil-litter mixing treatment (none vs. covering of WD with 0.5 cm of patch-specific soil) assessed the role of soil-litter mixing. Samples were collected 0, 1, 3, 6, and 12 months after placement and analyzed for mass loss.
Soil moisture was comparable in subcanopy and intercanopy locations, whereas soil temperature and solar radiation were greatest in intercanopy areas (p<0.001). Mass loss in mesquite was significant in early harvests, but did not occur in oak until months 6 or 12. Pooled across dates and treatments, decomposition was slower for oak (k = 0.68 yr -1) than mesquite (k = 0.80 yr -1). Decomposition of oak was significantly influenced by vegetation cover, termite activity, and soil coverage, with highest rates occurring in subcanopy areas where termites were most active. Oak samples mixed with soil exhibited greater mass loss in subcanopy areas, whereas non-mixed samples lost more mass in intercanopy areas where radiation levels and soil temperatures were high and termite activity was low. In contrast to oak, termite activity was absent on mesquite. Mesquite decomposition was accelerated by soil mixing, but was not affected by vegetation cover. Results indicate that interactions between soil deposition, vegetation cover, and termite activity are spatially variable in their importance as drivers of WD decomposition in shrub-invaded grasslands. Vegetation structure and microsite conditions should be considered when modeling decomposition dynamics in systems with low plant cover and high rates of aeolian and fluvial soil movement.