The Role of Microbial Community Composition in Controlling Soil Respiration Responses to Temperature

Abstract

Rising global temperatures may increase the rates of soil organic matter decomposition by heterotrophic microorganisms, potentially accelerating climate change further by releasing additional carbon dioxide (CO2) to the atmosphere. However, the possibility that microbial community responses to prolonged warming may modify the temperature sensitivity of soil respiration creates large uncertainty in the strength of this positive feedback. Both compensatory responses (decreasing temperature sensitivity of soil respiration in the long-term) and enhancing responses (increasing temperature sensitivity) have been reported, but the mechanisms underlying these responses are poorly understood. In this study, microbial biomass, community structure and the activities of dehydrogenase and β-glucosidase enzymes were determined for 18 soils that had previously demonstrated either no response or varying magnitude of enhancing or compensatory responses of temperature sensitivity of heterotrophic microbial respiration to prolonged cooling. The soil cooling approach, in contrast to warming experiments, discriminates between microbial community responses and the consequences of substrate depletion, by minimising changes in substrate availability. The initial microbial community composition, determined by molecular analysis of soils showing contrasting respiration responses to cooling, provided evidence that the magnitude of enhancing responses was partly related to microbial community composition. There was also evidence that higher relative abundance of saprophytic Basidiomycota may explain the compensatory response observed in one soil, but neither microbial biomass nor enzymatic capacity were significantly affected by cooling. Our findings emphasise the key importance of soil microbial community responses for feedbacks to global change, but also highlight important areas where our understanding remains limited.

Fig 1. Linear regressions between a) RR_MT and CVA1 or b) CVA2.

Black diamonds represent each soil (n = 18).

Fig 2. Community composition based on pyrosequencing.

The relative abundance of each microbial phylum (Archaea, n = 1; Bacteria, n = 13; Fungi, n = 4) is indicated for precooling (pc), control and cooled treatments.

Fig 3. Number of OTUs changing (mean ± S.E) in response to time of incubation (comparing precooling vs. control) or cooling (comparing cooled vs. control soils) or both (comparing precooling vs. cooled).

a) in different respiration response groups, b) linear regression between the number of changing OTUs and site MAT with or without 1H. Grey bars represent each group with standard error bars. Error bars for compensatory group were not added (n = 1). Black diamonds represent each soil (n = 8).

Fig 4. Microbial biomass estimated using a) SIR, b) CFE and c) qPCR methods and corrected for carbon loss in the control treatment (interpolated between pre-cooling and control treatment end of experiment biomasses, so that the MAT plus 3°C soils were compared to the cooled (MAT minus 3°C) soils at a similar C loss, corresponding to the maximum C loss of cooled samples at the end of the experiment.

Each soil is identified by geographic/climatic area including 1 (subarctic), 2 (Scotland), 3 (England), 4 (Mediterranean) and 5 (tropical), and ecosystem type: A (arable), C (coniferous evergreen forest), D (deciduous broadleaf forest), G (grassland), H (ericaceous heath) and E (evergreen broadleaf forest) (21). Mean ± S.E. is presented (n = 3). Black bars (control treatment), grey bars (cooled treatment). Statistically significant differences (P<0.05) are marked with an asterisk.

Fig 5

a) R_mass at the end of the experiment for control (black bars) and cooled treatments (grey bars). Statistically significant differences (P<0.05) are marked with an asterisk. b) R_mass ratio (control/cooled) regression with RR_MT. Black diamonds represent each soil (n = 20). Code for soil origin provided in the legend to Fig 1.