Response of Rhizospheric and Basal Soil Respiration to Changes in Soil Moisture and Temperature under Different Irrigation Phases in Arid Cotton Field in Northwestern China

Understanding the response of rhizospheric respiration (from roots, mycorrhizal hyphae and associated microbes) (Rr) and basal soil respiration(from decomposers) (Rb) to environmental variables (soil temperature (Ts) and soil moisture (Ms)) is critical for ecosystem C budget research, as these two components represent the major source of soil respiration. To study the different effects of Ts and Ms on Rr and Rb, six plots, three under the crown of cotton and three on the bare land were built in one of cotton fields at the Aksu National Experimental Station of Oasis Farmland Ecosystem on April 5, 2012. Based on the difference of soil respiration between cotton plots and bare plots, the components of soil respiration, Rr and Rb were divided. Then, the responses of Rr and Rb to soil temperature and moisture were analyzed, respectively. The results showed that. (1) During the whole experimental period, the variation in Rr was more strongly influenced by Ms (R2 = 0.6065) than by Ts (R2 = 0.0023), whereas Rb exhibited higher temperature sensitivity (Q10 = 2.291) than Rr(Q10 = 0.844). (2) During active irrigation phase, both Rr and Rb decreased with increasing Ms but showed only a weak relationship with Ts. (3) During the post-irrigation phase, the temperature responses of both Rr and Rb were well characterized by Q10 models, whereas their relationships with Ms followed parabolic trends. Our results demonstrate three key findings: (1) Rb and Rr shown significantly divergent responses to Ts and Ms; (2) while Rb displayed higher temperature sensitivity (Q10) than Rr, both were strongly regulated by Ms, with supra-optimal soil moisture levels (e.g., from irrigation or heavy precipitation) disrupting thermal control of respiration; (3) Ms exerted threshold-dependent effects on both Rr and, with Rb’s threshold (approximated field capacity) being substantially higher than Rr’s. These findings underscore the necessity of integrating the differential responses of Rr and Rb to Ts and Ms into ecosystem models to enhance projections of soil carbon cycling under future climate scenarios.