McDonald, D.B., W.K. Potts, J.W. Fitzpatrick, and G.E. Woolfenden. 1999. Proc. Royal Soc. London B. 266: 1117-1125.

Threatened Florida scrub-jays (Aphelocoma coerulescens) breed communally, are restricted to xeric sandy scrub habitat, generally disperse fewer than three territory diameters. Closely related Western scrub-jays (A. californica) do not breed communally, have a broader habitat range, disperse greater distances, and are not threatened. We compared the genetic structure of 445 individuals in 11 populations in Florida with 157 individuals in eight populations of Western scrub-jays. At
ten microsatellite loci, Florida had 24 out of 47 total alleles, while Western scrub-jays had 44. The Florida populations were more differentiated (G_ST = 0.048) than were a set of five California populations (G_ST = 0.015). A randomization extension of a Mantel test showed a stronger correlation between geographic and Cavalli-Sforza genetic distances among Western than Florida populations. Neighbour-joining trees clustered Florida populations from the same sandy ridge systems, suggesting that habitat continuity is more important than geographic proximity in allowing gene flow and preventing differentiation. For Western populations, isolation by distance appears to be the major determinant of genetic structure. Our results suggest that contrasting genetic structures may arise between closely related species, as a result of differences in ecology and social system. Conserving extant genetic variation in Florida jays will require maintaining viable populations in each of the major sandy ridge systems.

Macintosh (and IBM) software is available on request.         dbmcd@uwyo.edu

The software will do both a conventional Mantel test and a randomization extension
(allowing one to compare the strength of correlation between sets of matrices,
whereas the Mantel test asks only whether two matrices are correlated. That is, we can ask "Are Matrix A and Matrix B more correlated than are Matrix C and Matrix D?")

The tests should be applicable to a wide array of problems in ecology and evolution,
not just the genetic case shown here.