“What are the environmental determinants of phenotypic selection? A meta-analysis of experimental studies”

Posted on

Christina M. Caruso, Ryan A. Martin, Nina Sletvold, Michael B. Morrissey, Michael J. Wade, Kate E. Augustine, Stephanie M. Carlson, Andrew D. C. MacColl, Adam M. Siepielski, and Joel G. Kingsolver

Across study systems, natural selection is stronger in environments where mean fitness is low

Some of the species included in the database of selection estimates compiled by Caruso et al. Top left, ambush bug (Phymata americana); bottom left, great blue lobelia (Lobelia siphilitica); right, Lapland march orchid (Dactylorhiza lapponica).
(Photo credits: top left by David Punzalan; bottom left by Brian Husband, right by Nina Sletvold)

Natural selection on plant and animal populations is not uniform, but varies in space and time. Is this variation in natural selection more likely to be caused by some aspects of the environment than by others?

To answer this question, a team of researchers from universities in Canada, the UK, the US, and Sweden compiled data from more than 90 studies that experimentally manipulated aspects of the environment and estimated natural selection on plant and animal populations. The researchers used this dataset to test two predictions about which aspects of the environment are more likely to cause natural selection to vary in space and time.

The researchers did not find support for the prediction that aspects of the biotic environment (such as predation and competition) are more likely to cause selection to vary in space and time than aspects of the abiotic environment (such as temperature and precipitation). Instead, they found that manipulating the biotic environment had a similar effect on selection estimates as manipulating the abiotic environment.

In contrast, the researchers found support for the prediction that aspects of the environment that have a large effect on the average fitness (survival and reproduction) of plant and animal populations are more likely to cause selection to vary in space and time. Specifically, they found that selection estimates were larger in experimental treatments where manipulating the environment reduced average fitness.

These results indicate that the aspects of the environment that cause selection to vary in space and time are predictable. Specifically, the aspect of the environment that has the largest effect on mean fitness should be the most important cause of variation in selection. Read the Article