Rhizosphere Priming Effects Greater in Deep than Shallow Soils under Miscanthus x giganteus

High-yielding perennial grasses, such as Miscanthus x giganteus (Mxg), can contribute to a sustainable bioeconomy through net ecosystem carbon uptake to mitigate climate change. However, soil carbon sequestration by Mxg is inconsistent across sites and concentrated largely in surface soils. We hypothesized that the release of root exudates, rhizodeposits, and other labile carbon by Mxg can accelerate the decomposition of soil organic matter (known as the rhizosphere priming effect, RPE) more in deep compared to shallow soil because microbes are more carbon-limited in deep soil. To test this hypothesis, we incubated shallow (0-15 cm depth) and deep (50-100 cm depth) soil from mature Mxg stands (n = 5) to determine if RPE effects are greater in deep soils compared to surface soil. At the initiation of the incubation, we either added an artificial cocktail of sugars, organic acids, and amino acids to simulate root exudation as an experimental treatment or deionized water as a control treatment. We estimated RPE from the difference in carbon dioxide production and a suite of hydrolytic extracellular enzyme activities between the experimental and control treatments. We found that only in the deep soil did the artificial root exudates stimulate both α-glucosidase (AG) and acid phosphatase (AP) activity in comparison to the control (P = 0.0188 and P < 0.0005, respectively), with no effect of experimental treatment on hydrolytic extracellular enzyme activity in the shallow soil. This reflects the potential forncrease decomposition of complex soil organic matter in deep soil for both organic phosphorus (as indicated by a fivefold increase in AP activity over the control) and complex starches and celluloses (evidenced by a tenfold increase in AG activity over the control). Our results suggest that deep-rooted perennial grasses could induce RPE in deep soils to offset soil carbon sequestration, highlighting the need to understand controls on RPE to predict its importance across sites.