Selection in nature: experimental manipulations of natural populations

Numerous studies have documented evolution by natural selectionin natural populations, but few are genuine selection experimentsthat are designed and then executed in nature. We will focuson these few cases to illustrate what can be learned from fieldselection experiments alone or field and laboratory selectionexperiments together that cannot be learned from laboratoryexperiments alone. Both types of study allow us to evaluatecause and effect relationships because a planned experimentcan be accompanied by a more direct evaluation of the factorsthat cause evolution. A unique benefit of field experimentsis that they give us the opportunity to measure the rate andmagnitude of selection in nature. We have found that this rateis far greater than one might imagine based on observationsof the fossil record. A combination of field and laboratoryselection experiments has revealed the importance of populationsize and structure in shaping the genetics of adaptation. Forexample, laboratory selection experiments on insecticide resistancetend to attain resistance though polygenic inheritance. Theevolution of insecticide resistance in nature often eventuallyyields to single genes of large effect that are rare but, oncethey arise, represent a higher fitness solution to resistanceand spread among populations. Finally, field studies enableus to test evolutionary theory in a context in which all ofthe tradeoffs associated with a trait are realized; in the laboratory,organisms may be shielded from the fitness tradeoffs associatedwith the evolution of a trait. For example, we have comparedthe patterns of senescence in guppies from high and low mortalityrate environments in the laboratory and in the field. In thelaboratory, guppies from high predation environments had delayedsenescence relative to those from low predation environments.In the field the apparent relationship is the opposite. Onehypothesis for this difference is that a tradeoff associatedwith the evolution of the high predation life history is a decreasein the investment in the immune system. Such a sacrifice wouldbe evident in nature where there is exposure to disease andparasites but less so in the laboratory, which is relativelydisease and parasite free.