Abstract:
Marine microbes are the most abundant cellular life forms on Earth yet we know viruses to be even more numerous than microbes. Marine microbes are estimated to total about 10^30 while marine viruses have long been assumed to be ten-times more abundant, thus roughly 10^31 viruses are estimated to be present in the global ocean. In this thesis, the numerical relationship between viruses and hosts is examined through the perspective provided by global marine datasets. Fixed models which describe virus densities in terms of microbe densities such as the 10:1 ratio are examined for their predictive capacity, ultimately dispelling their use in favor of a non-linear model referred to here as the power-law model. Not only is the ability to predict virus densities from microbe densities improved by the power-law model but further analyses describe how the virus densities vary across across microbial densities, highlighting that the power-law model is most reliable when predicting virus densities of low-microbe density samples. Potential causes of the variability of virus densities are examined, ultimately determining that the studies which collected the samples are themselves a non-trivial source of virus-to-microbe ratio (VMR) variability. The data examined in this thesis comes from 22 different studies, totaling more than 5,500 records, and presents an opportunity to test the current knowledge of virus to microbe ratios with empirical data. Finally, the methods and data used in this analyses are described in detail and provided freely to the public to use to so that the findings presented in this thesis can be easily expanded upon by less quantitative but otherwise dedicated researchers in the marine microbiology community.