Unreliable quantitation of species abundance based on high-throughput sequencing data of zooplankton communities

High-throughput sequencing (HTS) is rapidly becoming a popular and robust tool to characterize biodiversity of complex communities, especially for those dominated by microscopic species such as zooplankton. The popular use of HTS-based methods has prompted a possible method of inferring relative species abundance from sequencing data. However, these methods remain largely untested in many communities as to whether sequence data can reliably quantify relative species abundance. Here we tested the relationship between species abundance and sequence abundance in zooplankton using 2 methods: (1) spiking known amounts of indicator species into existing zooplankton communities, and (2) comparing results obtained from parallel replicates for the same natural zooplankton communities. Although we detected a general trend that low-abundance species usually corresponded to low-abundance sequence reads, further statistical analyses revealed that sequencing data could not reliably quantify relative species abundance, even for the same indicator species spiked into different zooplankton communities. The distribution of sequence reads statistically varied even between parallel replicates of the same natural zooplankton communities. Our study reveals that sequence abundance may generally qualitatively reflect species abundance as the general trend between these 2 variables exists; however, extra caution is required when using HTS-based approaches to make quantitative inferences regarding zooplankton communities.