In tallgrass prairie, belowground processes are even more important than in forested systems because aboveground biomass and standing dead litter are periodically removed by frequent fires or grazers. Thus, studies that address factors regulating belowground processes are especially relevant for tallgrass prairie. We predicted that effects of grazing and burning differ belowground and that changes in root productivity caused by burning or grazing provide feedback that affects ecosystem fluxes of C and N. These differences in belowground response should be driven largely by changes in N dynamics and the degree to which burning and grazing affect the pathway and magnitude of N loss and the degree of N limitation in these systems. Fire, the major pathway of N loss in ungrazed tallgrass prairie, should result in reduced net N mineralization and N availability. We expected plants to compensate for increased N limitation by increasing their allocation to roots, as manifested in increased soil respiration and C cycling belowground. In contrast, grazing conserves N in the ecosystem by redistributing the N once contained in grass to labile forms in urine and dung. Thus, we predicted that grazing should increase N cycling rates and N availability to plants. Consequently, grazed plants should be less N limited and should allocate less C to roots and more to shoots. This, in turn, should decrease belowground C cycling, manifested as reduced soil CO₂ flux.

We explored the roles of grazing and burning on root growth in experimental watersheds at Konza Prairie, Kansas, USA. To assess effects of fire on root productivity, we installed root ingrowth cores in two watersheds without grazers that differ in fire frequency: annually vs. infrequently burned (four years since the last fire). To assess effects of grazing, we installed root ingrowth cores in an annually burned watershed grazed by bison and in fenced controls (exclosures). Within bison “grazing lawns,” root ingrowth cores were installed in lightly and heavily grazed patches. Concurrently, we measured in situ rates of net N mineralization and soil respiration as indices of soil N and C cycling.

Annual burning resulted in a 25% increase in root growth compared to the unburned watershed (four years since last fire), as plants compensated for N limitation by increasing allocation to roots. Grazing had the opposite effect: it decreased root growth, especially in heavily grazed patches (∼30% less than in fenced controls). Grazing by ungulates increased N cycling and availability. Therefore, grazed plants, instead of being N limited, experienced C limitation as shoots regrew and plants allocated less C to roots. Interestingly, root ingrowth on the long-term unburned watershed was as low as in lightly grazed patches in the grazed watershed. Thus, seemingly disparate treatments such as infrequent burning (characterized by accumulation of detritus aboveground) and grazing (periodic biomass removal) both had higher levels of N availability than annually burned prairie in the absence of grazers. Root growth in unburned and grazed watersheds must be limited by resources other than N (e.g., C in grazing lawns or light in infrequently burned prairie).

Burning and grazing also altered root tissue chemistry in contrasting ways that further accentuated the root growth differences caused by these treatments. Frequent fires lowered substrate quality of roots (C:N = 60), thus increasing N limitation. In contrast, grazing and infrequent burning improved root tissue quality (C:N = 40), promoting faster cycling of N. These large differences in root growth and tissue chemistry can result in profound ecosystem-level changes. Grazing increased net N mineralization rates from 87% to 617% compared to watersheds without grazers, whereas annual burning decreased it by ∼50% compared to unburned prairie. Although grazing speeded up N cycling, it reduced soil respiration by 50% compared to fenced controls, presumably because of reduced root mass. On the other hand, annual burning increased soil respiration, presumably because of increased root biomass. Ultimately, differences in the quantity and quality of roots provide feedback to affect C and N cycling and help to maintain and even promote the fundamental differences in N cycling between burning and grazing in tallgrass prairie.