Soil CO2 efflux in response to the addition of water and fertilizer in temperate semiarid grassland in northern China

Knowledge about the effects of water and fertilizer on soil CO2 efflux (SCE) and Q (10) is essential for understanding carbon (C) cycles and for evaluating future global C balance. A two-year field experiment was conducted to determine the effects of water, fertilizer, and temperature on SCE in semiarid grassland in northern China. SCE, as well as environmental factors was measured in two grasslands, one with bunge needlegrass (BNE, Stipa bungeana) and one with purple alfalfa (ALF, Medicago sativa), with four treatments: CK (unwatered and unfertilized); W (50 mm water addition yr(-1)); F (50 kg phosphorus (P) fertilizer ha(-1) yr(-1) for ALF, 100 kg nitrogen (N) + 50 kg P fertilizer ha(-1) yr(-1) for BNE); and W + F. During the 11-month experimental period from July 2010 to October 2011, the addition of water consistently stimulated mean SCE in BNE and ALF, and the positive effects were relatively stronger during dry seasons. P fertilization consistently enhanced SCE in ALF, and the positive effect was strongly dependent on the availability of soil water. The effects of N plus P fertilization on SCE in BNE varied seasonally from significant increases to small reductions to no response. Water addition increased the Q (10) of SCE in ALF by 11 % but had no effect in BNE. Fertilization, however, reduced the Q (10) of SCE by 21 % and 13 % for BNE and ALF, respectively. Models that rely only on Q (10) underestimated the emissions of soil CO2 by 8-15 % at the study site, which was mediated by species and treatment. Responses of SCE and its temperature sensitivity to water and fertilizer may vary with species and depend on the period of measurement. Models of SCE need to incorporate the availability of ecosystemic water and nutrients, as well as species, and incorporate different environmental factors when determining the impact of water, nutrients, and species on SCE.