Impact of common sample pre-treatments on key soil microbial properties

Pre-treatment of soil samples prior to analysis is acknowledged to affect microbial activity and community parameters, but recommendations are diverging and there are no best-practice or standardised laboratory protocols. For the microbial carbon use efficiency (CUE) as a ratio parameter of C allocated to anabolism over total metabolised C, the effects of sample pre-treatment are unpredictable because the underlying parameters of microbial metabolism may be affected differently. In treatment or land use comparisons, it is important to know whether soils are affected in a similar way by pre-treatment. Based on five different paired sites (each with cropland and forest), the effects of six common pre-treatments were evaluated on 18O-CUE and associated parameters, i.e. respiration rate, soil microbial biomass C, total DNA extracted, and estimated abundances of fungi, bacteria and archaea. The six common pre-treatments were: (i) immediate analysis of field-fresh soil samples, and analysis following 14 d pre-incubation of soil samples that were (ii) field-fresh, (iii) air-dried, (iv) oven-dried (40 ◦C), (v) frozen at -20 ◦C and (vi) frozen in-situ (dry ice and subsequently liquid N2). The experiment revealed two main findings. The first was that pre-incubation of 14 d had the most pronounced effect and reduced microbial respiration rate, growth rate and microbial biomass C by between 28% and 63% on average compared with field-fresh samples, suggesting an impact of the reduction in readily available substrates on microbial activity. The second finding was that even after 14 days of pre-incubation, drying & rewetting caused an increase in the microbial respiration rate in forest soil samples of 64 ±53% (air-drying) and 86 ±65% (oven-drying), known as the Birch effect. However, CUE as a ratio parameter was found to be unaffected by sample pre-treatment, with only minor changes after 14 d of pre-incubation. Pre-treatment (including pre-incubation) affected cropland and forest samples similarly, with one exception: the estimated abundance of fungi increased only with drying & rewetting in forest soils. This suggests that dried cropland soils could potentially yield appropriate community data, while forest soils respond to soil drying in a more complex way. It was concluded that CUE values can be derived from soil samples regardless of whether they were dried & rewetted, frozen & thawed, or stored fresh.