Why do species have restricted geographic distributions?
Classic ecological perspectives tell us distribution limits occur where ecological parameters coincide with the boundaries of ecological niches. Evolutionary perspectives, on the other hand, surmise distribution boundaries reflect a failure of niche evolution. Though small population size and limited genetic variation in ecologically relevant traits likely limits a species’ distribution, few studies have investigated multivariate phenotypes.
Paccard et al. (2016) did just that in The American Naturalist. They investigated whether populations at latitudinal limits had different quantitative genetic architecture of ecologically relevant traits compared to those populations found in the middle of the range.
They estimated genetic variance-covariance (G-) matrices for 10 traits related to size, development and water balance.
Populations at the southern and northern range limits had reduced levels of genetic variation across many traits. Though genetic variance declined at the edges, there were weaker genetic correlations as compared to midlatitude populations.
The consequence of declining [genetic variance] but weaker correlations at the edge was that the predicted short-term response to randomly generated selection gradients was not worse or, if anything, slightly better at the southern and northern range margins.
Basically, their results contradict the expectations from univariate perspectives on phenotype. Moreover, Paccard et al. suggest the patterns they uncovered in A. lyrata may in fact be more general and could impact our understanding of species range limits.
Drift is important at range limits, but the consequences of drift may result in pretty high evolvability at range limits.
How general is this model? Only time will tell.