Abstract:   The city of Brevard NC (Transylvania County) hosts an established colony of non-albino white squirrels with pigmented eyes, head patch and dorsal stripe.  There is much variation in both the white and gray morph and little is known of their developmental genetics, but they appear to segregate as separate Mendelian units. Although it is often assumed that the white morph is recessive, breeding experiments are sparse, uncontrolled, and inconclusive.   According to generally accepted chronologies, the white morph originated in Brevard in 1951 with the release of a pair of white squirrels trapped in Florida.  If squirrel abundance at that time was approximately what it is today, the frequency of this genetic predisposition (hereafter referred to as the white allele) would have been less than .001 regardless of the zygousity of the released squirrels.  Today the percent of the white morph in a 3 square mile study area centered in downtown Brevard is holding steady in the mid-to low 20's with an average of 23.2% over the last four years.
     Assuming that the white allele is indeed recessive, then its frequency q is approaching 0.5 (0.482).   This would be a rather dramatic change in gene frequency over half a century, comparable to that of the classic melanization of the peppered moth.  Eastern gray squirrels can, but do not necessarily, have two litters per year.   Thus, assuming the change in gene frequency is distributed over 100 generations it would still require a selective coefficient (s) of 0.85 or greater against the gray morph to account for such a /\q.  Not only is this difference in fitness between the two morphs not apparent, the continued progression toward fixation predicted by that model is not occurring.   This population genetics analysis thus would suggest that, contrary to public opinion, the morph is not a simple recessive.
     Assuming instead that the white allele is dominant would reduce its current allelic frequency to approximately p=.125.   Such a /\p over 100 generations could be explained by selective coefficients as low as 0.05 which could be accounted for by human preference/protection for the unusual white morph.   However, such a model still could not explain the current apparent equilibrium.
     The simplest models of equilibrium invoke heterozygote advantage with homozygote fitnesses being reduced by selective coefficients s and t, respectively.  This would require some sort of unseen, pleiotrophic effect of the allele that operates in the heterozygote independent of pelage color.   If the heterozygote was gray, the current phenotypic distribution could be explained if s=t with an  equilibrium at p=q=0.5; in order to  stabilize within 100 generations s and t must be >0.1.   As in the original recessive advantage model considered above, this would still represent a dramatic /\q over half a century, although unlike the former, it would not proceed toward fixation.  Note that under such conditions,  coat color, itself, would appear to segregate as a simple recessive in individual mating pairs.
     Alternatively, if the heterozygote were white, the current phenotypic distribution could be explained by a much smaller /\p=0.125.  This would require s=7t and in order to achieve equilibrium in less than 100 generations,  t=>0.05 and s=>0.35.  In this model s is the selective coefficient against the "dominant" homozygote white which can be expected to suffer greater predation than the cryptically colored "recessive" homozygote gray.  The unusual feature, here, is that the reduction in fitness due to coat color of the heterozygote must be balanced by some yet undetected physiological benefit of heterosis.
     Frequency dependent models where the benefits of human interference and energy savings from pigments not produced and the costs of predation vary in non-linear ways might also lead to equilibrium frequencies in the observed ranges.  With wild fluctuations in density as observed in the Annual Squirrel Count, a significant role for drift, also, cannot be discounted.  As expected, the inheritance of Brevard's pigmented white squirrel is proving to be more complicated that that of its albino cousins in Olney IL.  What is needed now are some carefully conducted breeding experiments.