Background/Question/Methods With drought frequency and severity likely to increase the future, understanding its effect on terrestrial carbon (C) and nitrogen (N) cycling has become essential for accurately modeling ecosystem response to climate change. Microbes respond to drought stress by accumulating internal solutes, or osmolytes, such as amino acids, betaines and polyols, to balance cell membrane water potential as the soil dries. When seasonal rains arrive, internal solutes are released and rapidly mineralized, and resources are returned to cell production. We have been studying these processes in a California grassland with a climate of cool wet winters and hot dry summers. Beginning in summer 2007, we made monthly measurements of soil moisture, individual amino acid concentration in total soil and in microbial biomass via HPLC, total dissolved organic carbon and nitrogen (DOC and DON), and microbial biomass carbon and nitrogen (MBC and MBN).

Results/Conclusions We expected the concentration of the known amino acid osmolytes glutamate (glu) and proline (pro) in microbial biomass to fluctuate inversely with soil moisture. However, pro was only recovered in Mar 2008 (0.30 μg pro-C g^-1 dry soil) and initially, glu-C concentration varied proportionally with soil moisture: lowest during summer 2007 (0.06 g H₂O g^-1 dry soil, 2.50 μg glu-C g^-1 dry soil) and highest in winter 2008 (0.27 g H₂O g^-1 dry soil, 5.14 μg glu-C g^-1 dry soil). The expected trend emerged following the dry down in summer 2008 (0.10 g H₂O g^-1 dry soil, 2.61 μg glu-C g^-1 dry soil) when glu-C increased (0.05 g H₂O g^-1 dry soil, 5.99 μg glu-C g^-1 dry soil). This investment in glu production accounted for 3% of the MBC, and 12% of MBN and represented a shift in organic matter composition and resource allocation in response to drought. The difference in summer microbial biomass glu concentration may be related to the rainfall in the preceding winter, as winter 2006 was one of the driest in the 50 yr record for this area (195 mm), while winter 2007 was average (398 mm). These data suggest that glu may be used for microbial osmoregulation in a CA grassland. The implications for terrestrial carbon and nitrogen cycles are considerable because as the frequency of drought increases in the future, the microbial response of osmolyte accumulation and release may alter N dynamics and serve to pump C out of the ecosystem.