Academic journal article Genetics

The Adaptive Hypothesis of Clinal Variation Revisited: Single-Locus Clines as a Result of Spatially Restricted Gene Flow

Academic journal article Genetics

The Adaptive Hypothesis of Clinal Variation Revisited: Single-Locus Clines as a Result of Spatially Restricted Gene Flow

Article excerpt

AFTER the establishment of the field of ecological genetics more than 30 years ago (CLARKE 1975) rapid progress in molecular marker development and analysis technology has generated a surge of renewed interest in identification of selective footprints of natural selection in a wide range of species (e.g., SCHLÖTTERER 2002). Among numerous research strategies developed to infer the evidence of selection in natural populations at the molecular level (reviewed by NIELSEN 2005; VASEMÄGI and PRIMMER 2005), associations between environmental variables and molecular marker polymorphisms are commonly taken as strong support for the hypothesis that natural selection maintains singlelocus clinal variation (e.g., EANES 1999; BAINES et al. 2004). However, it has often been overlooked that single-locus clines can also be the result of various neutral evolutionary processes, such as hybridization of previously isolated populations, founder events, and migrational patterns, such as spatially restricted gene flow (but see also MANLY 1985; STORZ 2002). In Drosophila melanogaster, for example, extensive research over several decades has suggested the presence of latitudinal clines for a number of genetic polymorphisms, including cosmopolitan inversions (KNIBB et al. 1981), allozyme loci (OAKESHOTT et al. 1982), candidate genes (SCHMITD et al. 2000; DUVERNELL et al. 2003; FRYDENBERG et al. 2003), and microsatellite loci (GOCKEL et al. 2001). At the same time, widespread clinal variation has been demonstrated for several quantitative traits, including body size (JAMES et al. 1995), egg size, ovariole number (AZEVEDO et al. 1996), development time (JAMES and PARTRIDGE 1995), and a temperature resistance (HOFFMANN et al. 2002). Recently, SEZGIN et al. (2004) reported three new significant latitudinal clines in 14 metabolic genes from North America, which together with earlier studies reveal seven clines among 17 genes (41%). Similarly, an earlier study by GOCKEL et al. (2001) suggested that a considerable proportion of randomly selected microsatellite loci (5 loci out of 19, 26%) exhibited significant linear clines in a north-south transect along the east coast of Australia, providing further support of the presence of extensive single-locus clinal variation in D. melanogaster, both in the northern and the southern hemispheres. However, as argued by DUVERNELL et al. (2003), if ∼26% of all polymorphisms that show clinal variation are truly affected by selection, then on average every 10th-50th gene (80-400 kb apart) across the Drosophila genome is expected to exhibit an adaptive latitudinal cline along the environmental gradient. Here, I show that results like those of SEZGIN et al. (2004) and others can be alternatively explained by an isolation-bydistance (IBD) effect alone, without the need to invoke a widespread multilocus adaptive response, through the use of simple population genetic simulations and assessment of a recently published empirical data set.

The simulation considered a simple neutral population genetic model with 11 subpopulations of 1000 diploid individuals made up of equal numbers of males and females connected by a one-dimensional stepping stone model of migration using six different migration rates (m = 0.001, 0.005, 0.01, 0.02, 0.035, 0.05). This enabled evaluation of how often allele frequencies at 500 neutral loci (mutation rate 5 × 10^sup -4^, two-phase stepwise mutation model with 30% of double-step mutation events, maximum number of possible allelic states 25) exhibited statistically significant clinal variation in the presence of various levels of IBD. The simulations were run first for 2000 generations starting with minimal variability before reducing the migration rate to zero and simulating 0, 50, 150, 250, 350, and 450 generations in isolation. Such a two-step simulation regime is expected to generate a strong IBD pattern during the first migration-drift phase while random genetic drift is expected to reduce the strength of the IBD and increase the level of differentiation during the second isolation phase. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.