Estimating shares of various sources of variability in determining human phenotypic variance

Abstract

The authors have applied a method of repeated at specified time intervals ' observations of same individuals (repeatability) to a set of anthropometric characters (see table 1, where characters not denoted by standard anthropometric coding are — from top of the table down — : body weight, waist circumference, max, arm circumference and vital capacity) and three parameters of the acid-base status of the blood (see table 2). The method allows one to separate a “fluctuative” ‘portion of total observed variance and thus to estimate a “durable” portion of variance. The method is outlined in the text with formulas. Symbols of variables used in them, and consequently in headings of tables, are in majority the same as in D. S. Falconer's Introduction to Quantitative Genetics, (Edinburgh 1960). Exceptions are: V_er that denotes variance of random measurement error, P_g, D_eg, D_e(s+r) and D_es. These last four symbols are used for indices of genetic polymorphism and ecosensitivity described in Henneberg & Lewicki [1978]. Anthropometnic data were collected on a group of 40 physical laborers aged 20-60 yrs measured twice or thrice during summers of 1974, 1975, 1976. Physiological observations were carried on 181 individuals subdivided into five groups (boys aged 8 yrs, draft solidiers, male athletes, girls aged 8 yrs, and female athletes). Measurements were repeated 2-15 times on each individual and taken over periods ranging from one week to several years. It has been found, as would be theoretically expected, that anthropametric trails contain a large share of genetic and “durable” environmental variance, whereas variance resulting from these sources is practically absent in physiological characteristics observed here. Thus total variance of examined blood properties is attributable to reversible short-term fluctuations in physiological states of organisms plus observational errors. Furthermore the authors discuss relevance of partitioning of variance for arriving at more precise estimates of heritability and ecosensitivity of various characters.