The increasing appreciation this last decade that associated microorganisms (`microbiota') are fundamentally
tied to the basic physiological and biological processes of virtually all organisms on the planet necessitates we
revise many established models to account for the influences of these microorganisms. One such model is our
understanding of the impact of spatially varying selection on animal populations, which has traditionally
described how the pressures of distinct environmental characters on animal genotype lead to favorable
phenotypic outcomes in distinct geographic locations. Recent evidence, including our preliminary work, shows
that the abundance of different members of the microbiota also varies predictably with geography, suggesting
the abundance and identity of associated microorganisms may be related to these locally adaptive processes. A
major gap is that virtually all the current work in this area is descriptive, attributing geography-dependent
variation in host-associated microbiota to environmental or host genetic variation without defining the
underlying mechanisms. Thus, we propose an experimental design to explicitly define how variation in a model
animal's microbiota is established. Our approach is based on defining specific influences for two of the most
obvious candidate processes: the environment, including diet, and host genotype. To understand if the
environment determines geographic variation in the microbiota we will compare the microbiota of wild flies in
the eastern United States, a broad location across which the microbiota is known to vary predictably with
geography, with the microbiota of their diet and abiotic environment. We will also rear flies in the laboratory
and the wild under different environmental conditions (temperature, photoperiod, and humidity), to define
causal deterministic roles for these characters on the fly microbiota. Then, to reveal how host genotype
influences the fly microbiota we will focus on one host process, feeding preference, and how this process
influences the composition of the fly microbiota. Since the sequences of many fly feeding preference genes vary
in the eastern United States, we will also create fly mutants that swap alleles between flies from different
geographies, to determine if these alleles are responsible for variation in fly feeding preference and microbiota
composition with fly geography. Together, the results will define the relative contributions of the environment
and host genotype in determining geographic variation in the microbiota of the fruit fly, the organism upon
which many of our current models of animal evolution are based. Thus, our findings will be directly relevant to
current models, and will facilitate the incorporation of the microbiota and its predictable geographic variation
into improved definitions of animal evolution.