Long- and short-term soil storage methods other than freezing can be useful for DNA-based microbial community analysis

Abstract

Methodsfor sequencing microbial communities are constantly improving, necessitating storage of samples for future analyses. However, tests of different storage approaches have been limited, hindering our ability to use soil samples stored under various conditions to understand how microbial communities may be changing over time. Few studies have directly compared the effects of different storage methods on microbiome composition over month-to-decade time scales. Here, we present evidence from two experiments investigating how microbial community composition differs between soil samples collected at the same time and stored under different conditions. In a short-term experiment, we measured the effects of three months of refrigeration at 4 °C versus three months of freezing at −80 °C, and in a long-term experiment, we measured the effects of up to 8 years of air-drying at room temperature versus freezing at −80 °C. We used high throughput DNA sequencing (Illumina MiSeq) to analyze general fungal and arbuscular mycorrhizal fungal communities in both experiments and bacterial communities in the short-term experiment. We analyzed DNA sequencing efficiency, alpha and beta diversity, log fold change of microbial groups, and variation attributable to environmental variables. In the short-term experiment, we found no significant impacts of refrigeration versus frozen storage on bacterial or fungal community composition. In the long-term experiment, we found no significant differences in sequencing efficiency or alpha diversity between air-dried versus frozen samples, but significant differences in beta diversity between the groups. The differences in beta diversity generally did not alter the relationship among general fungal and AM fungal communities, sampling location, and aboveground plant communities. Overall, our results suggest that soil microbial communities are relatively robust to different storage methods compared to frozen storage. Nevertheless, downstream application and interpretation should still account for previous sample storage and for certain analyses, such as temporal tracking of specific taxa.