The productivity of four inbred strains of Drosophila melanogaster

Four inbred strains of flies have been maintained by brother × sister mating since November, 1963. Counts of production per vial were made at 14 days, and the flies from the most productive vial were used as the parents of the next generation. For analysis the data were separated into three groups: vials with no production (blanks), vials with less than 20 flies (low production), and those with over 20 flies. The number of blanks did not change significantly in strain A but did show significant decreases in strains B and D. Until generation 60 the number of blanks per generation was not distributed in a Poisson series, but after this generation it was. The number of vials with low production was high in all strains for the first 20 generations and then decreased in widely different generations for each strain.Changes took place in all four strains but by different mechanisms. In strain A there was no increase in range, but the mode moved from 40 to 150 flies per vial and the proportion of low productive vials fell from 15% to 2% per generation. In strain D the low production vials fell from 22% to 6% and there was an extension of the range. Each of the strains shows definite trends of increase and decrease which last from 3 to 30 generations, and vary in slope from 0.3 to 13.8 flies per vial per generation. While we cannot determine how much or how many of these slopes are environmental, it can be argued that since many of the slopes of the four strains are not correlated, part of the cause is genetic. The changes in the direction of the trends show that phenotypic selection in small populations was not effective in accumulating and exploiting mutations for greater productivity.

FIVE GENES LOCATED ON CHROMOSOME 4 OF THE TOMATO

The gene for broadleaf has been the only gene known to be on chromosome 4 of the tomato. Four other genes are shown to be linked with broadleaf. These are wiry (w1), which distorts the leaves and flowers; clausa (clau), which causes excessive proliferation of the pinnae; divergens (di) which affects plant growth and stem color; and stamenless (sl1), which causes the stamens to be absent or greatly reduced. The genes w1, clau, and di give mongenic ratios which are deficient in number of mutants. The order of the genes on the basis of F2 data is w1clau 39 e 21 di 2 sl1. The genes tf, ao, rv, and ah are not in this group.

PERIODICITIES IN THE ANNUAL MUSKRAT POPULATION FIGURES FOR THE PROVINCE OF SASKATCHEWAN

The yearly production of muskrat pelts in Saskatchewan for the period 1915–60 exhibits a 6-year periodicity instead of the expected 10-year cycle. To check the synchrony of these periodicities, the province is divided into thirds: the southern prairie section, the middle aspen grove section, and the northern Canadian Shield section. In the prairie section there are two main peaks, one in 1923 and the other in 1955. The remainder of the peaks, which are at 6-year intervals, vary greatly in magnitude. The aspen grove section has peaks which are also 6 years apart but are of similar magnitude. The northern section has peaks of equal magnitude which are 30 years apart. The environmental control exercised on the conservation blocks in the Cumberland House area tends to disrupt the cycle.Some of the population changes are correlated with precipitation data. In the prairie section, the highest catches occur after 2 wet years. In the other sections, the muskrat population is not correlated with the precipitation. In the aspen grove section there is a slight correlation between snowfall and the muskrat catch the next spring.

THE INHERITANCE OF LITTER SIZE, BODY WEIGHT, AND VARIABILITY, IN A CROSS BETWEEN TWO INBRED STRAINS OF MICE

Crosses were made between the two inbred lines, C57BL and BALB, and the effects of maternity and genotype on litter size, survival, 30- and 60-day weights, and their variances were assessed. The maternal effect of BALB reduced litter size by.78 to 1.91 when compared with reciprocal crosses of identical genotype. No maternal effect was found for C57; neither did the genotype of the young have any effect on the litter size borne by mothers of this strain. There was a significant increase (2.26 ±.67) in litter size when BALB females produced hybrid instead of inbred young. This was not the result of hybrid males being more prolific, because BALB males mated to non-BALB females produced the largest litters in this experiment. The cross F1 × F1produced the largest mean litter size. A change in ration affected the litter size when BALB was the mother, but not when C57 was the mother. Fox chow plus linseed produced more young than any of the other three diets.There were no significant differences in the weights of males of any of the generations at 30 days, but the F1females borne in C57 mothers were larger than the females of other generations, and it appears that the C57 female exerts a maternal effect which causes the female young to mature earlier. The data indicate that the mother exerts the greatest influence on 30-day weights, the genotype of the young having little or no effect. The F1female has the greatest maternal effect: she increases the 30-day weight by 10–20%.At 60 days the F1mouse was 15% larger than the corresponding inbred borne in the same female. The mean weights of the F1and F2were the same, but the backcrosses were 10% lighter.The variance at 30 days was larger than the corresponding variance at 60 days, and 90% of the variance was in the between-litters component. At 30 days the largest variances were found in the P1, F1, and B.C. generations, while the smallest was found in the F2indicating that much of the variability was the result of the maternal environment. At 60 days the variances for all generations were essentially the same.MacArthur's large and small strains, which had been produced by 30 generations of selection with only moderate inbreeding, were carried through 20 generations of brother × sister mating. This inbreeding brought about a decrease in the weight of the large line, and an increase in the weights of the small strain. The former may be attributed to the loss of the heterozygotic effect on size, while the latter may be the result of natural counterselection; the smallest mice had few or no young.

GENETICS AND PHYSIOLOGY OF THE XANTHOPHYLLOUS MUTANT OF THE TOMATO

Xanthophyllous is a dominant yellow-leafed mutant; the genotype Xa Xa is lethal in either the seed or the very early seedling stage, while the genotype Xa xa is viable but reacts differently to light intensity than does the normal xa xa plant. The 1: 2: 1 ratio is never realized because less than half the lethals appear as seedlings. The 2: 1 ratio is affected by germination so that under good conditions for germination a 2: 1 ratio is obtained, but under poor conditions the ratio may approach 1: 1. Maximum likelihood formulae are given for calculating linkage when one gene is a dominant with recessive lethal effects, and it is shown that the simple product moment method gives similar results for these data. Xa is in linkage group VII (chromosome 10), and its lethal effect disturbs the monogenic ratios of all genes in this group. The following crossover values between these genes and Xa are found: H 44.5%, pe 50%, t 35.5%, tv31.7%, and ag 46.5%.Under standard growing conditions the pigment concentrations of xanthophyllous contrasted with green is: chlorophyll, 790 instead of 3240 μg./g.; xanthophyll, 54 instead of 211 μg./g.; and carotene, 60 instead of 130 μg./g. Both green and yellow plants, when grown under various intensities and spectral compositions of light, have different responses. Green plants make only limited response to changes in light intensity, whereas Xa xa plants become green and cannot be distinguished from xa xa plants. From 600 to 8000 ft-c. the Xa xa plants show a linear trend of decreasing pigment (increasing yellowness) with increasing light intensity. The hypothesis is advanced that Xa xa plants make as much pigment as do green ones, but a light-dependent reaction breaks down the pigments faster than they are being formed.

THE NATURE OF CYCLES IN POPULATIONS OF CANADIAN MAMMALS

The number of pelts of the various fur animals collected each year since 1915, are examined critically to determine whether the peak or high collections are cyclic, or whether they could arise by chance in a series of random numbers as suggested by Cole. Cole's criterion of a peak is rejected as impractical and misleading. By splitting the country into 53 sections, it is shown that peaks do not occur at random as would be expected if these are chance occurrences. Nor do the peaks occur in the same year all over the country, but rather they are spread over three to four 'good' years which are followed by five or six 'bad' years, when no peaks occur in any section. Peak collections for the whole of Canada are the resultant of simultaneous high collections in most parts of the country. Rarely are all the sections of the country in phase; whenever a high-producing section gets out of phase it may produce a secondary peak, or even a split peak in the total for the whole country. It is shown that there is a relationship between the number of sections experiencing a peak, and the peak for the whole country. In general, the peak collection of most furs is reached first in the central provinces, then in the western, and finally in the eastern provinces. Data from annual questionnaires confirm this sequence. They show that the increase phase for snowshoe rabbit, mink, and fox begins as an island of increase in the prairie provinces surrounded by an area of decrease or no change. The reports of increase on subsequent years can be plotted by isophasal lines which spread out like waves from the original area of increase. The easiest interpretatation for these waves is migration from the areas where increase began into the more sparsely populated surrounding sections. These migrations would have a synchronizing effect on the population changes over the whole area. They would also give hybrid vigor to the local population by mixing gene pools which have been isolated for several generations. It is suggested that the causal agent of these cycles is the accumulation of favorable factors in good years, acting over most of the country on populations of different initial densities. Migration from the more favored areas will cause an upswing of the cycle over most of the range. Decrease, being largely density-dependent, will follow across the country in the same order.

A STUDY OF SIZE INHERITANCE IN THE HOUSE MOUSE: II. ANALYSIS OF FIVE PRELIMINARY CROSSES

Five crosses were made between mice of different body size, and over 2000 mice were raised in the F1, F2, and backcross generations. The body weight at 60 days after birth was used as the criterion of size. The male means were always larger than the corresponding female means and the difference in weight between the two sexes increased progressively with body size. Litter size and sequence had no effect on body size. The adequacy of the gram scale was tested with inconclusive results which indicated that in at least two crosses some other scale should be used. Log-grams were substituted for grams and gave a good fit in cross No. 3 but not in crosses Nos. 1 and 2. The evidence from selection experiments, environmental variability, and sex differences in size indicate that on a gram scale at least part of the factors which affect body size are proportionate rather than additive in nature. In all five crosses the F1and F2means are intermediate between the parents. The backcross means are halfway between the F1and the respective parent. Only one cross showed increased size in the F1which might be interpreted as due to heterosis. Reciprocal crosses gave significantly different results and the dissimilarity was carried over into the next generation. This difference was attributed to the environmental effects of female body size. As expected, the variances of the P1's and the F1were similar but, contrary to expectation, the F2variance was no larger than that of the F1. Litter size showed a different type of inheritance. One cross between P1's with mean litter sizes of 5.1 and 10.2 gave an F1mean of 13.2 young. This was tentatively interpreted as dominance of large litter size and hybrid vigor allowing more embryoes to reach parturition.

POPULATION CYCLES AND COLOR PHASE GENETICS OF THE COLORED FOX IN QUEBEC

The pelt collection figures for colored fox in the province of Quebec were examined. The figures for the central and southern part of the province show a typical nine-year cycle. In the northern sections the data show that until 1930 there was a nine-year cycle in colored fox coexisting with a four-year cycle in white fox. After 1930 the four-year and nine-year cycles exist simultaneously with the four-year gradually dominating the scene. In the Upper James Bay region the typical nine-year cycle shows a supplementary peak corresponding with the four-year peak observed in the regions to the north. The coat color phase ratios cannot be explained by monohybrid equilibrium but they are consistent with the view that the population consists of isolates. A partial breakdown of isolate barriers would account for the ratios observed without the necessity of the large unexplained gene frequency changes which occur if panmixia is postulated. Migration causes the breakdown of isolate barriers and this accounts for the sudden shift in gene frequency and explains the long term trends which have resulted in a lower percentage of the silver phase. The degree of isolation changes with the population pressures. The cause of the cycles appears to be resident in the respective area in which the animal breeds. Both the cycle and the color phase data indicate that northward migrations have taken place.

A STUDY OF SIZE INHERITANCE IN THE HOUSE MOUSE: I. THE EFFECT OF MILK SOURCE

Three strains of mice were used to study the effect of fostering on the growth pattern of the mouse. The strains used breed true for size and have been designated as "Large", "Small", and "Intermediate". The 14-day mean weight of mice that received milk from "Large" strain mothers is significantly different from those that received milk from either the "Small" or the "Intermediate" strain mothers. Although these differences tend to remain, they are not statistically significant at 140 days. The significance of these results are discussed in relation to the arithmetic and geometric concepts of polygenic growth.

THE GENETICS OF THE COLOUR PHASES OF THE RED FOX IN THE MACKENZIE RIVER LOCALITY

The red fox (Vulpes fulva) exists in the wild in three coat colour phases; red, cross, and silver or black. These three phases result from the action of one pair of alleles, the homozygotes being silver and red and the heterozygote being cross. At least two different mutations have occurred giving rise to the Canadian gene in eastern Canada and the Alaskan gene in western Canada. The mixing of these two mutant genes complicates the gene frequency analysis.The proportions of the three colour phases are shown to vary with (1) locality, (2) state of population cycle, (3) population trend, (4) migration pressure. Of these factors the variations with locality and population trend are fairly satisfactorily explained by selection but the fluctuation of colour phase proportions with the population cycle is not. On the other hand all the facts can be explained by a migration theory, with or without selection. By migration, a mixture of native and migrant populations with different gene frequencies is obtained. Such migrations tend to be rhythmic since they are connected with the population cycle. In the year that migration took place the pelt returns reveal aberrant gene frequencies or colour phase ratios. In the following years the gene frequencies quickly approach equilibrium that may be at the premigration level or at a new one depending upon the success of the migrants in establishing themselves in the breeding population.The marked diminution in the percentage silver and cross is due to the rapid population increase in an area of low frequency of the silver producing gene and the migration of this type into areas of higher frequency.

THE INHERITANCE OF FRUIT SIZE IN THE TOMATO

It is pointed out that size data from over 50 tomato crosses are explained by the assumption of the geometric action of size factors but not by a simple additive theory.The fact that the F1 results fitted such a theory was pointed out in a previous paper when the theory was proposed. The analysis is here extended to the F2 generation and to cell size measurements.The use of the geometric scale introduces regularity into the otherwise unpredictable F2 segregations, and they become amenable to a simple logarithmic scheme. Analysis by such a scheme indicates that differences in cell number or ovary size are caused by the segregation of three to five pairs of major genes, whereas mature cell size differences seem to be brought about by the segregation of at least twice as many factors.Final weight is thus the resultant of the proportionate action of the following factors:—1. The number of mitotic divisions in the pre-anthesis period and therefore the number of cells at anthesis.2. The cell expansion after anthesis.3. Fruit shape, locule number, and other size-modifying effects.