At this point, Elizabeth, Nathan, and their team had been monitoring the beetle populations for more than a decade. To connect another dot in their chain of evidence, they compared data on PGI frequency among beetles in Bishop Creek from back when Nathan started working on them in 1988 to beetles collected in 1996. Before 1988, the weather in the area had been particularly dry, but between 1988 and 1996 it was cooler and wetter. As we would expect if PGI-1 were a cold-adapted protein undergoing natural selection based on temperature, its frequency increased between the two samples. Other enzyme polymorphisms that were not associated with temperature adaptation didn’t show any notable change.

This was another line of evidence suggesting that the PGI frequency differences Nathan had observed so long ago were caused by selection based on temperature — but they were still missing some key steps. The team had investigated molecular interactions and population dynamics, but they still needed to understand how PGI version affects the performance and fitness of actual beetles.