Can prescribed fires restore C4 grasslands invaded by a C3 woody species and a co-dominant C3 grass species?

Prescribed fire is used to reduce woody plant invasion and restore herbaceous production and diversity in grasslands and savannas worldwide. Here, we determined whether a concentrated series of repeated-winter, repeated-summer, or alternate-season (winter and summer) fires in a short timeframe ("transition fires") could catalyze the restoration of C-4 perennial grasses in Southern Great Plains, USA grasslands that had become dominated by a fire-tolerant C-3 woody N-2-fixer (honey mesquite, Prosopis glandulosa) and a C-3 perennial bunchgrass (Texas wintergrass, Nassella leucotricha). We applied transition fires over a 5-yr span and maintenance fires on a portion of each plot 7 or 8 yr later. We measured herbaceous standing biomass and cover and soil variables (soil organic C, N, delta C-13, and delta N-15) in unburned, transition-burned, and maintenance-burned treatments. Greater delta C-13 at 10-20 (-17 parts per thousand) than 0-10 (-20 parts per thousand) cm depth increment confirmed that vegetation was historically mostly C-4 grassland that shifted toward C-3 dominance. Transition treatments with summer fire were most effective at top-killing mesquite, but no treatments root-killed >3%. Regrowth of top-killed mesquite was similar in all treatments and reached pre-fire height by 9-10 yr post-fire. Herbaceous production and cover responses showed that: (1) Alternate-season transition fires increased C-4 mid-grass, but did not change Texas wintergrass; (2) repeated-summer fires reduced Texas wintergrass, but did not change C-4 mid-grass; and (3) repeated-winter fires did not change C-4 mid-grass or Texas wintergrass compared with the unburned control. All maintenance fires stimulated Texas wintergrass biomass and cover, thus eliminating the reduction of Texas wintergrass caused by repeated-summer transition fires. There were no long-term effects of transition fires on soil C, N, delta C-13, or delta N-15. Results advance our understanding of the expectations and limitations of prescribed fire in shifting a woodland alternate state toward what was historically a fire supported C-4 grassland/savanna.