Using F_ST to Detect Selection in the Periwinkle, Littorina saxatalis

The statistic F_ST measures the variance in allele frequency between populations relative to polymorphism within them and is expected to be the same across loci. This test has been used to locate selected regions of the genome in the periwinkle, Littorina saxatilis. Two distinct shell morphs have evolved (Fig. WN16.1A). The “H” form is found on the upper shore and has a wide, flat shell that can cling well to wave-lashed rocks. The “M” form is found on the lower shore and has a thicker, narrower shell that is better adapted to resist predation by crabs. Most genetic markers show rather little divergence both between populations of different morphs in the same place and between populations of the same morph in different places (F_ST ~ 0.025). However, about 5% of genetic markers showed substantial divergence between morphs in the same place, suggesting that they are closely linked to genes that are selected to maintain the adaptive differences between morphs (Fig. WN16.1B).

The same logic extends to morphological traits. If differences between populations are due to random drift, then the variance in trait mean between populations relative to the genetic variance within, measured by Q_ST, should be the same for all traits and should equal the analogous ratio F_ST for other kinds of genetic variations. In fact, morphological traits vary substantially more between populations than do genetic markers such as allozymes or microsatellites (i.e., Q_ST > F_ST; Fig. WN16.2). The most plausible explanation is that morphological traits experience stronger diversifying selection than do genetic markers. See Heller (1976), Hull et al. (1996), and Wilding et al. (2001).