If you spend enough time poking around bushes in California, Nevada or Arizona, you’ll find stick insects, long little guys that blend in with sticks or leaves. Sometimes you only notice them when they drop out of their camouflaged environment and onto your shirt. They’re funny looking, harmless and at the center of a recent high-impact study at Utah State University describing when and how you can predict evolution.
“In some cases you can make predictions and forecasts of evolution. It depends on various things like the timescale. Even when you can’t forecast into the future, you should understand why. You understand the mechanisms that are driving evolution still, but they’re inherently unpredictable,” said genetics expert Dr. Zachariah Gompert.
The stick insects he studies come in three different colors: brown, green and stripey, and just-plain-green. Gompert’s research associates tracked what percentage of the stick insect population was what color for 25 years. Gompert used the information to try to predict what the future color proportions of the population would be.
“For the green versus striped, you can explain on the order of 85-90 percent of the variation you see,” Gompert said. “We’re forecasting remarkably well. In the case of the melanistic allele frequency, there is some predictive ability but it’s much much poorer.”
Birds were choosing to eat the most common green colored stick insect, causing the rarer green stick insect to increase in the population. Because the factor driving evolution was easy to observe – the rarity of the stick insect color the previous year– it was easy to predict the color that would dominate in the future. But weather – which is very hard to predict from last year’s data – was the main factor determining how many brown stick insects were around.
Gompert thinks his study illustrates how complicated evolution can become. When one factor, like rarity, is affecting evolution, it’s easy to predict outcomes. But when more than one factor is present, like rain and heat, it becomes much more difficult to predict. Knowing this will help managers decide how much to trust predictions about the adaptability of species.