In this paper I present the results of an experimental study of the effects of genotype and density on the mean and variance of offspring numbers in both sexes of the flour beetle, Tribolium castaneum. From the observed variance in offspring numbers the effective population size at several different densities is estimated using the methods of Crow & Morton (1955).

I found that both the mean and variance of offspring numbers varied with genotype and density. In general, males were more variable in offspring numbers than females and this variability increased with density. Individuals homozygous for the black body colour mutant, b/b, were less variable in offspring numbers than + / + individuals, but the latter produced more offspring at most densities. As density increased, + / + individuals became more variable in offspring numbers whereas b/b individuals were less sensitive in this regard. These findings are discussed in relation to the ecology of selection at the black and closely linked loci.

Two linear models have been devised and applied to the study of the distribution of human male chromosomes on the metaphase plate in preparations from lymphocyte cultures not treated with spindle poisons. Using these it has been found that the chromosomes are approximately distributed around a centre of symmetry and that the lines joining the centromeres of homologous chromosomes (i.e. segments) have the centre of symmetry approximately at their mid-point. Thus each chromosome mirrors the position of its homologue relative to the centre of symmetry. The position of each chromosome in the metaphase plate was found to be approximately constant relative to the centre of symmetry and the other chromosomes. The significance of these findings is discussed in relation to the hypothesis on the distribution of the chromosomes in interphase nuclei and to data on acrocentric associations, acrocentric-non-acrocentric associations and the frequency of the most common translocations.

The effect of subdivision of a population on response to artificial directional selection for abdominal bristle number in Drosophila melanogaster was compared using large, replicated lines. Three different population structures were compared: (i) selection in an Undivided, large population with 50 pairs of parents (treatment U); (ii) selection in each of 10 sublines which were reconstituted every 6th generation by Crossing after Culling the 5 lowest sublines (treatment CC); and (iii) selection in each of 10 sublines which were reconstituted every 6th generation by Crossing after Retaining all 10 sublines (treatment CR). At the end of three cycles of selection and crossing, neither CR nor CC was superior to U; sublining did not increase response to selection. These results agree with the predictions arising from an entirely additive model and provide no evidence for the presence of epistasis.

A comparison of 50-pair lines (U) with several 5-pair lines was made over 31 generations. For the 50-pair lines, there was close agreement between response predicted from the base population (using ih2σp) and observed response throughout all 31 generations of selection. Although the best of the 5-pair lines exceeded the 50-pair lines in the early generations, average response to directional selection in the 5-pair lines soon fell behind that predicted from ih2σp, and soon reached a plateau.

Some yellow alleles which show full expression of the mutant phenotype have behavioural defects, including reduced locomotor activity and a low level of male competitive mating success. Other yellow alleles produce pattern mosaics in which some cuticular structures are yellow and others wild type within the same individual. The pattern mosaic expression of these alleles extends to behaviour. Mosaicism in respect of the different pleiotropic effects reveals something about the possible mode of action of this locus on behaviour.

Alcohol dehydrogenase (Adh) allozyme frequencies and tolerance of adult flies to ethanol were measured in population samples of D. melanogaster from a winery in Southern Ontario. Samples were also tested from a number of non-winery sites. Adhfast frequencies decrease as one moves away from the winery, but this drop in allozyme frequency is statistically significant only for those areas that are more than 3 km distant from the winery. Tolerance of adult flies to ethanol also differed between winery and non-winery populations, but these differences were not statistically significant. The data presented here may help to resolve the seeming conflict between the two previous studies of Adh allozyme frequencies in natural populations from highand low-alcohol environments.

The theory of path analysis is extended by considering a multivariate system of correlations from a dual perspective. Intrinsic factors exert a unidirectional influence on both the variance and covariance of dependent factors. In contrast, extrinsic factors have a bidirectional influence on the covariance structure of both antecedent and dependent factors and do not influence intrinsic variability. The mathematical model assumes a formally complete linear system of unitary factors. A coefficient is defined to quantify the influence of adventitious associations and is called a copath. Copaths are compared to path coefficients and to correlations due to common antecedents. The chain properties of these coefficients are derived along with a general formula and computational algorithm. The method is illustrated for multifactorial inheritance in extended pedigrees in the presence of different types of assortative mating.

A natural population of Atlantic salmon, Salmo salar L. was screened for genetic variation at 59 protein loci, using a sample of parr, the juvenile freshwater stage. The mean heterozygosity per locus, estimated at 0·033 ± 0·014, is similar to that described in other salmonid species but low for fish species in general. Variation was observed at AAT, IDH, MD–ME, GLO, ADA, and SDH loci. The methods described should prove useful in stock discrimination and in producing salmon for restocking and sea-rearing. The observed extent of gene duplication is discussed in relation to the evolution and systematics of the salmonid fishes.

The wild-type N group plasmid RN3, which is phenotypically Res+ Mod+(RII) is not eliminated by thymine starvation of its bacterial host. Derivatives of RN3 selected for the Res− phenotype are eliminated. The presence or absence of the RII modification specificity does not affect thymineless elimination of RN3 Res− plasmids. A Res−Mod(am) RN3 mutant is not eliminated by thymine starvation from either amber suppressing or non-suppressing hosts, suggesting that it carries a cryptic mutation in a novel genetic locus required for elimination. Thymineless elimination is shown to be recA+-dependent and the presence of the X group plasmid R6K significantly inhibits elimination of RN3 Res−Mod+. However, since R6K has no effect on two other plasmid-mediated functions of UV protection and UV-induced mutagenesis, which are also recA+-dependent, it would appear that elimination is determined by a separate plasmid gene than that encoding the UV functions. This is confirmed using derivatives of another N group plasmid R390, which eliminate but which have lost the ability to increase UV-induced mutagenesis in their host.

Geographic populations are not necessarily equivalent to genetic populations. Electrophoretic isozyme analysis shows that in the biennial, Verbascum virgatum, not only is spatial distribution a factor creating isolated populations, but time also is an isolating mechanism. Thus, in the biennial situation there is a possibility that selection can be rejected as the probable cause of variation, much as can random drift in cases involving spatial distributions. The magnitudeof difference infrequency of alleles between the odd and even year colonies studied was too large to be explained by random drift in populations of the size and short duration of those we observed. Similarly, it is unlikely that random fluctuations in selection intensity on one age class would produce a difference as large as that observed. It is possible, however, that variation was introduced (mutation or founder) when the population was much smaller and that the difference was trapped at a relatively high level when the population rapidly increased in size. Simulations and algebraic theory do not refute this idea. They also show that colony differentiation can occur with migration rates considerably greater than previously predicted.

The effect of migration and recombination on the equilibrium structure of populations subject to a common symmetric selection regime in all habitats is studied. Attention is restricted to a class of symmetric polymorphic equilibria which have been studied in two-deme systems by Bazykin (1972) and Karlin & McGregor (1972) for one locus and by Christiansen & Feldman (1975) for two loci. With increased migration and recombination the heterozygosity increases unless it is already at the maximum level. Although the populaton system as a whole is always at linkage equilibrium, the magnitude of linkage disequilibrium in the individual demes may either increase or decrease with more migration and recombination. In general, the less the migration and the less the recombination between interacting loci, the greater the possibilities of polymorphic equilibria.

Experiments have been performed to show that PGM polymorphism for the two common electrophoretic allozymes, PGMA and PGMB, in Drosophila melanogaster has adaptive value. Firstly, the allele frequencies converge to the same equilibrium value in six experimental populations. Secondly, density-dependent selection operates. Thirdly, the relative fitness of the three genotypes varies in modified culture media. PGM polymorphism is maintained by frequency-dependent selection and heterotic selection: the first mechanism operates to reach equilibrium frequency, the second cooperates to maintain it. The experiments performed with modified culture media favour the view that the two allozymes have different affinities for two components which are present in the nutritional environment. These components may be either substrates or other factors involved in the reaction catalyzed by PGM.

Mutations in the gatB gene as well as mutations in the putative structural gene gatA and the positive acting regulatory gene intA can affect γ-amino-n-butyrate (GABA) transaminase (EC 2.6.1.19) levels in the ascomycete fungus Aspergillus nidulans. Partial or complete loss of function mutations in gatA, gatB and ssuA, which specifies succinic semialdehyde dehydrogenase, can lead to accumulation of ω-amino acids resulting in pseudo-constitutivity and elevated expression of (retained) activities under intA control. These regulatory effects underlie selective methods for gatB−, ssuA− and leaky gatA− mutations. However, all three gatB− alleles which have been selected lead only to partial loss of GABA transaminase activity as judged by both in vivo and in vitro criteria. It has not been established whether the leakiness of these three gatB− mutations is an allele-specific or a locus-specific effect and whether or not the GABA transaminase present in gatB− strains differs from the wild type enzyme. Thus the role of the gatB product remains to be elucidated. The gatB gene is not closely linked to any other gene involved in ω-amino metabolism or related pathways.

The issue of incorporating the good alleles from two homozygous populations of a cross-fertilizing species into a single improved population was investigated assuming independent assortment, no epistasis, and either complete dominance (of the favourable or the unfavourable alleles) or additive genes. The selection limit in the foundation stock is a function of the effective population size (N), the proportion (x) contributed by the better source population (P1), the difference in relative fitness between single locus homozygotes and the proportion of loci (that will segregate in the foundation stock) fixed favourably in P1. In real life the last two of these are never known. We therefore focused on the response limits given x = 0·5, 0·75, 0·875 or the optimum value of x (which is a function of the other three parameters). Our general finding was that in situations where N is large enough so that a major portion of the potential can be achieved, the F2 population should be used as the foundation stock when the two source populations do not differ greatly in performance; but when one population performs considerably better than the other, the first backcross (but not second backcross) would be the choice.

Linkage disequilibrium between segregating amino acid sites in repeated genes which form a multigene family was investigated by using the population genetics theory. The degree of non-random association of amino acids is measured by the disequilibrium coefficient which is determined by the balance among various forces; unequal and equal crossing-over, mutation, random genetic drift and recombination which takes place between the two sites in question. Another measure of disequilibrium, ‘standardized identity excess’, represents excess probability of simultaneous identity at the two sites over that expected from random association of amino acids. Sequence data of variable region of immunoglobulins provide most interesting example of multigene family to apply the theory. Statistical analyses on identity excess for various groups and subgroups of variable region sequences of immunoglobulins suggest that a multigene family such as human κ or mouse κ gene family consists of several subfamilies between which recombination is limited. The analyses also indicate that the recombination may take place between any gene members in one subfamily.

In the course of an X-ray experiment, the normal allele of forked was transposed to the second chromosome, where it acts as a suppressor of forked. In this position, which is near the centromere, the duplication (Dp-f+) is subject to a variegated position effect. This was studied in its dependence on the hetero-euchromatin balance; the results agree with and extend those found for other position effects. In addition, we found regional preferences for variegation in the individual flies. The most interesting aspect of Dp-f+ is its tendency to transpose either to the homologous second chromosome or to Chromosome IV. In the latter position, Dp-f+ acts as a dominant near-lethal, so that the apparent selectivity of insertion sites is at least in part due to deleterious effects at insertion sites other than its original one. In a new, and presumably, centromere-far position of Dp-f+ on Chromosome II the variegated position effect disappeared and transposition was reduced in frequency or wholly abolished. The frequency of losses of Dp-f+ approximately equalled that of transpositions. Since there is good evidence that transpositions occurred pre-meiotically, the apparent losses of Dp-f+ may have been due to meiotic segregation separating the loss from the new insertion.

Twenty-four cold-sensitive (prototrophic) mutants were isolated after UV mutagenesis of protoplasts of the fungus Podospora anserina. Genetic analysis of these mutants was performed in order to detect those among them which were most likely to be impaired in translational fidelity. The 24 mutations belonged to 24 different genes. One half of the mutants were pleiotropic and displayed an altered phenotype: growth rate at the permissive temperature, germination of the spores, fertility and/or sporulation. Nine mutants differed from wild-type in their resistance levels to cycloheximide, trichodermin and/or paromomycin. Several mutations were linked to known ribosomal loci. Two mutations behaved like informational antisuppressors: one is allelic to the previously described As3 gene and the other one defines a new antisuppressor gene, AS6.

The hypothesis is advanced that the evolutionary stability of the unusual sex-determining mechanisms of the Wood Lemming (Myopus schisticolor) and of the Varying Lemming (Dicrostonyx torquatus) is a direct consequence of certain characteristic features of their population dynamics, and that these include phases of unrestrained population growth and of mass dispersal. Computer simulations confirm the feasibility of such an explanation. Predictions of this hypothesis are found to differ in a potentially testable manner from those of the ‘inbreeding’ hypothesis of Stenseth (1978). The demonstration of such a direct link between population ecology and evolutionary genetics would, if substantiated, be exceptional in mammals.