Particulate soil carbon loss through priming in response to woody encroachment of a tallgrass prairie ecosystem

Background/Question/Methods

Approximately 40% of the earth’s land surface is covered by grassy ecosystems, and shrub encroachment on grasslands is a widespread phenomenon. Woody encroachment has been shown to affect soil carbon (C) dynamics, however, C-cycling responses in soil horizons below 20-cm depth are not well understood. Since most plant biomass in grasslands is belowground and most soil C is at depth, understanding subsoil C response is critical. In this study, we asked how shrub encroachment affects soil C cycling to 1-m depth at replicated dogwood-invaded or native grass-dominated locations. Because woody invaders have deeper roots, we expected higher C inputs and C-cycling activity, particularly deeper in the soil profile, under dogwoods than grasses. To evaluate this, we collected 6 replicate 1-m soil cores under dogwood islands and open grass areas in both heavily and weakly invaded areas, measured plant and soil C pools at 10-cm increments, and used isotopic C signatures to differentiate C3-woody-plant-derived from C4-grass-derived pools and fluxes. Also, to learn whether increased dissolved C inputs would stimulate or repress decomposition of standing soil C stocks, we conducted laboratory priming experiments in soils collected from non-invaded and heavily invaded areas.

Results/Conclusions

We found that the d13C‰ of soil microbial respiration was lower under woody plants than grasses (-19.9 (1.0) < -17.1 (1.1), mean (SE), p=10-16) over the whole soil profile, and this difference was greater at both 0-10cm and 50-100cm depths (p=0.0002). This shows that soil microbes use more C3-plant-derived dissolved C, particularly at depth, under encroaching dogwood plants, which supports our predictions. Surprisingly, however, root stocks, soil microbial respiration, 30-day mineralizable C, C-cycling enzyme activities, and total soil C concentrations were lower under dogwoods than grasses, despite a higher proportion of woody-plant-derived C under dogwoods. There was also a strong positive priming response to sugar addition in soils collected from intact grasslands (40-110%), but no priming under dogwoods. After soil C stocks were summed through the 1-m soil profile (g C ha-1), woody encroachment had no significant effect: Grassland soils hold large amounts of C, so changes in stocks may not be significant until many decades pass. However, our results clearly show that woody encroachment should not be expected to increase soil C, and that there is a significant pool of particulate C in native tallgrass prairies that is vulnerable to loss as a consequence of woody plant encroachment.