A Sampling Method Considering Body Size for Detecting the Associated Microbes in Plankton Populations: A Case Study, Using the Bloom-Forming Cyanobacteria, Microcystis

Abstract

It is challenging to accurately measure the microbes living in close association with tiny aquatic organisms. This is because many laboratory methods can only detect the relative amounts of microbes compared to their host, not the absolute number. If the host’s size affects this relative amount, simply testing random individuals may yield misleading results about the whole population. In this study, we simulated a dataset based on a globally widespread phytoplankton to test a different approach called group analysis. This method tests individuals in groups rather than individually. We found that this approach greatly improves accuracy and reduces errors, especially when analyzing groups with many individuals per group. This method also facilitates comparison of different populations and makes it easier to detect microbes present in very small numbers. Our findings offer a practical way to better understand the relationships between small aquatic organisms and their associated microbes, which is crucial for protecting water quality and ecosystem health.

Accurately quantifying associated microbes is essential to understand the interactions between microplankton and their associated microbes. Most DNA-based methods, such as high-throughput sequencing, primarily assess the ratio of target objects to references in microplankton samples. However, simple random sampling (SRS) of individuals may lead to deviations in quantifying these ratios at the population level if these characteristics are associated with the reference content of individuals. This study considered group analysis, which involves detecting k groups with n individuals in each group, as an alternative approach and used simulated data based on the detection of Microcystis populations to evaluate the accuracy of different sampling plans. Our results indicate that increasing the number of individuals in each group could reduce sampling bias and improve the accuracy of comparisons between populations. Group analysis could also minimize the impact of the detection limit. This study demonstrated that, when detection methods only provide the ratio of target objects to references, group analysis is more appropriate than SRS for characterizing microplankton populations. Group analysis can be used not only for detecting associated microbes but also for identifying ingested organisms or the biochemical composition of microplankton. Our results also demonstrate how in situ individual-level studies support ecological investigations.