A Note on Extending Taylor's Power Law for Characterizing Human Microbial Communities: Inspiration from Comparative Studies on the Distribution Patterns of Insects and Galaxies, and as a Case Study for Medical Ecology

Many natural patterns, such as the distributions of blood particles in a blood sample, proteins on cell surfaces, biological populations in their habitat, galaxies in the universe, the sequence of human genes, and the fitness in evolutionary computing, have been found to follow power law (e.g., Kendal 2004, Illian et al. 2008, Venter 2001, Ma 2012). Taylor’s power law (Taylor 1961: Nature, vol. 189:732–735.) is well recognized as one of the fundamental models in population ecology, thanks to its wide applicability in describing the spatial distribution patterns of biological populations. A fundamental property of biological populations, which Taylor’s power law reveals, is the near universal heterogeneity (also known as aggregation or nonrandomness) of population abundance distribution in habitat. Obviously, the heterogeneity also exists at the community level, where not only the distributions of population abundances but also the proportions of the species composition in the community are often heterogeneous or non-random. Indeed, community heterogeneity is simply a reciprocal term of community evenness, the major dimension what community diversity tries to measure. Nevertheless, existing community diversity indexes such as Shannon index and Simpson index can only measure “local” or “static” diversity in the sense that they are computed for each habitat at a specific time point, but the indexes alone do not reflect the diversity changes across habitats or over time. This inadequacy is particularly problematic if the research objective is to study the dynamics of community diversity, which is critical for understanding the possible mechanisms that maintain the community diversity and stability. In this note, I propose to extend the application scope of Taylor’s power law to the studies of human microbial communities, specifically, the community heterogeneity at both population (single species) and community (multiple species) levels. I further suggested that population dispersion models such as Taylor (1980, Nature: 286, 5355), which is known to generate population distribution patterns consistent with the power law, should also be very useful for analyzing the distribution patterns of human microbes within the human body. Finally, I suggest that the parameters of the power law model built at community level, especially when associated with community metadata (or environmental covariates) and when time series data from longitudinal studies are utilized, can reveal important dynamic properties of human microbiome such as community stability, which can be invaluable in investigating etiology of some diseases associated with human microbial communities. Overall, I hope that the approach to human microbial community with the power law offers an example that ecological theories can play an important role in the emerging medical ecology, which aims at studying the ecology of human microbiome and its implications to human diseases and health, as well as in personalized medicine.