About 10,000 years ago, light-colored quartz began depositing in what would become modern-day Nebraska, eventually forming sand dunes covering an area of over 50,000 square kilometers. The Nebraska Sand Hills region is designated as a distinct grassland ecoregion and is home to a diversity of plants and animals including charismatic bison and pronghorn. This relatively recent geologic feature has also served as a natural evolutionary laboratory for one of its smallest residents – the deer mouse (Peromyscus maniculatus).
The coat color of deer mice varies dramatically from dark grey to light tan. Interestingly, dark mice are commonly found on the surrounding dark grey soil, whereas mice living on the geologically recent light-colored Sand Hills habitat are typically lighter, suggesting that lighter coat colors of Sand Hills mice evolved recently. It has been hypothesized that the evolution of this color variation is driven by natural selection for coat colors that blend in with their habitat against avian predators like owls and hawks that mainly hunt by sight. But to really understand how coat color in deer mice evolved, we need to dig deeper – to the level of genes that provide the heritable information for advantageous coat colors to be passed from one generation to the next. A long-standing challenge in biology has been to connect the mutations within genes to variation in traits that are subject to natural selection.
In this study, we make this connection using a combination of field and laboratory experiments. We first conducted a large-scale field experiment with wild populations of deer mice to directly estimate natural selection on coat color. We used 14,000 kg of steel, some heavy machinery, and plenty of manual labor, blood, sweat, and tears to build multiple large enclosures both on the Sand Hills and on soil. After weeks of rising before dawn to set mouse traps and a few rattlesnake scares, we managed to catch hundreds of wild mice. We then filled each enclosure with ~100 individually radio tagged mice caught on and off the Sand Hills and waited to see which ones would survive an attack by birds of prey from above.
After three months, we compared coat color and genetic differences between the original mice and the survivors. We first found that mice with matching coat colors to the habitat of their enclosures were more likely to survive. On average, surviving mice on dark soil had darker coat colors than founding individuals, while surviving mice on light sand had lighter coat colors. In addition, we found large changes in frequencies of mutations in the gene Agouti, which is known from previous research to affect the production of pigment cells in many vertebrates.
Showing that these mutations were important in nature was one thing, but we also wanted to demonstrate that they directly affected coat color in mice, so took off our field boots and donned our white lab coats. In the lab, we focused on one of these mutations, a single amino acid deletion. By genetically engineering it into laboratory mice, we discovered that those mice carrying the deletion had light coat colors because of significantly less yellow-brown pigment (pheomelanin) in their hairs. Biochemical tests confirmed that the deletion causes less pigment by reducing interactions between agouti proteins and a specific receptor (attractin) in the membrane of pigment-producing cells. Finally, we returned to the wild deer mice in our experimental enclosures back in Nebraska to show that this deletion did in fact decrease in the dark enclosures as expected.
With these experiments, we show how a single mutation (an amino acid deletion) within a single gene (Agouti) alters a trait (coat color) that is under natural selection (affects survival) due to ecological pressures (avian predation). The differential survival of light and dark mice feeds back at the genetic level causing certain mutations to become more or less common over time. Demonstrating these causal links between genes, traits, and survival provides a complete picture of evolution in action, which has rarely been observed outside of microbes in a lab.
These exciting results provide evidence that evolution can be predicted in the wild for certain traits under certain conditions. Whether this predictability holds true for other traits in other species will require more experiments that link genetics to traits to survival. Such research is becoming more important as we begin to realize that evolution can happen relatively quickly, that humans are increasingly the cause of these changes, and that evolution can have significant ecological, medical, and economic consequences.