Theor. Appl. Genet. 56,183-185 (1980) 9 by Springer-Verlag 1980 Heterozygosity Levels in a Mammal: A Determination Derived from Allele Frequencies at Loci Controlling Gross Morphology B. Blumenberg Faculty of Sciences, Lesley College, Cambridge, Massachusetts (USA) Summary. Estimates of heterozygosity derived from elec- trophoretic data may seriously underestimate the true lev- els of genetic variability. An estimate derived from allele frequencies in Felis that relate to macroscopic morpho- logical traits appears to be congruent with several theoreti- cal predictions of heterozygosity levels. Key words: Allele frequency - Felis - Gross morphology - Heterozygosity Introduction During the last decade, it has become routine and estab- lished procedure to study genetic variation in natural pop- ulations by means of the technique of gel electrophoresis (Lewontin and Hubby 1966; Harris 1966). Such analysis reveals electrophoretic allozymic variation. In a recent and comprehensive review, Nevo (1978) surveyed the litera- ture and computed a mean heterozygosity level (H) for mammals as 0.0359. The H levels in the tabulation range from 0.0 for Mirounga anqustirostris (and others) to 0.106 for Mus rnusculus. To a minority of workers, a fundamental question re- mains. Are such determinations truly representative of the genome or do they merely reflect the limitations of avail- able technique? Several authors, after considering both theoretical and practical aspects of such methodology, have aligned themselves with the latter view. King and Wilson (1975), Marshall and Brown (1975) and Nixon (1977) estimated that perhaps as few as 25% of all muta- tions are electrophoretically detectable. Bonhomme and Selander (1978) in a survey of variation in both electro- phoretic charge and thermostability at 14 structural loci in Mus rnusculus estimated that electrophoresis detects about one-third of the variants which is equivalent to 30% of the alleles at such loci. An important factor certain to complicate this problem further stems from the observa- tion that some isozymes show patterns of inheritance that cannot be interpreted by simple Mendelian models (Bow- en and Yang 1978). Nei (1975: pp. 134-135) estimated that real heterozygosity levels in outbreeding organisms may be on the order of 30-40%. Observations of H for individual loci suggest a wide range of values. Johnson (1974) provides a summary of such data. In small vertebrates, variable substrate enzymes range in H levels to 30% (esterases); regulatory enzymes to 24% (phosphoglucomutase 2,3) and non-regulatory en- zymes to 13% (isocitrate dehydrogenase). In a recent re- view, Ward (1977) gives a mean vertebrate heterozygosity level for adenosine deaminase of 28%. In man, Johnson (1974) presents H levels for variable substrate enzymes that range upwards to 47% (pepsinogen), for regulatory enzymes that range to 50% (glutamate pyruvate amino- transferase) and for non-regulatory enzymes that range to 9% (amylase). Hedrick and Murray (1978) report H levels for 12 red blood cell loci in man that exceed 30%. Those loci showing heterozygosity levels in excess of 20% in- clude representatives of both the fast and slow evolving groups (Sarich 1977). It is worth noting that a) loci pro- ducing regulatory enzymes do show lower levels of hetero- zygosity than all others and b) the proportion of any ge- nome upon which heterozygosity (or indeed any other) determinations rest is well under 1%. Methods Is there any approach to heterozygosity conceptualization and calculation that results in mean H levels in excess of 20%? There appears to be one set of studies that provides data that lead to such a determination. Free ranging populations of the domestic cat (F. catus) have long been the subject of investigations in popu- lation genetics. These studies have focused upon an examination of several polymorphic loci the alleles of which determine gross morphological traits such as coat color, hair length, presence or absence of tail, manifestations of polydactyly, etc. A discussion of 0040-5752/80/0056/0183/$ 1.00