It is often assumed that evolutionary specialization by an organism must lead to losses of performance and functionality in other environments, but there is little evidence to support this theory. Part of this uncertainty stems from a problem of perception: how does one operationally define a specialist? Also, a paucity of direct experimentation has limited the consideration of alternative population genetic processes that can lead to such losses of function. We describe here a long-term experiment in which twelve populations of the bacterium, Escherichia coli, have evolved in the laboratory for 20,000 generations. During this time, all of the populations became progressively better adapted to growth in a simple glucose-limited environment. We will examine the consequences of this evolution for the bacteria's capacity to utilize a variety of substrates other than glucose, and thereby quantify the extent to which specialization led to losses of function. We will also evaluate the relative contributions of two population processes, tradeoffs (antagonistic pleiotropy) and drift (mutation accumulation), to the observed changes in function.
