POPULATION GROWTH IN DROSOPHILA CULTURES

Abstract

Summary .1. As Pearl (1926) recognized, the Drosophila culture can be taken as a model of population growth involving the interaction of a plant and an animal species. Early work suggested that the growth of the fly population could be adequately described by an S‐shaped curve of the kind previously applied by Verhulst (1839) to human populations. This curve, the logistic, shows that populations apparently possess: (a) a slow initial growth rate, which (b) increases until it reaches a maximum (the point of inflexion of the curve), and (c) then becomes progressively less as the population approaches the maximal size possible under the particular ecological conditions. This logistic law was subsequently widely applied to a variety of populations, from bacterial to human, developing under divers conditions. Part of the justification for this procedure derived from the apparently fundamental character of the Drosophila model, and this review is devoted to an examination of that assumption. 2. The mathematical derivation of the Pearl‐Verhulst logistic law is shown to rest on highly arbitrary postulates which are not supported even by the early experimental analysis of the circumstances influencing natality and mortality in a Drosophila culture. Pearl and his co‐workers generally failed to recognize the importance of the growth of the yeast population, and, by paying attention only to the adult flies in their cultures, took no account of the contribution of the larvae and pupae present to the total population. As a consequence, their formulation of the experimental analysis was in terms of the intraspecific relationships between the adult individuals of Drosophila , and little weight was attached to the competition between larvae for food and space. 3. Examination of the total population (eggs, larvae, pupae, adults and yeasts) present in a culture shows that the logistic has sometimes been applied to family growth and not to true population growth, and that it describes even this restricted phase very incompletely. The precise course of population growth is found to depend upon the changes of egg output of the females, pre‐adult mortality and the duration of the various instars. Each of these is affected to some extent by the availability of the yeasts, and the growth and distribution of the yeasts is modified by the activity of the feeding larvae in the culture. That is, the interspecific relationship is found to be the primary ecological regulator of population growth; but the data from a total count are not sufficiently precise to permit a detailed analysis of this relationship. 4. While the course of oviposition of normal females is fairly regular under optimal conditions, their potential fecundity is more than halved when they are introduced into a normal culture bottle. This reduction in egg output results from a change in the character of the yeasts, a qualitative change, induced by the activity of the larvae in the medium. Since the yeasts change quantitatively and qualitatively as the culture grows, the early experiments on the fecundity of adults crowded under optimal conditions have little relevance to the analysis of population growth. Indeed, it has been shown that this decline of fecundity with crowding is mainly a consequence of food shortage, and such a shortage is only likely to affect the second generation of adults in a culture. The viability of the eggs that are laid changes, in much the same way as the oviposition rate, as the females age, and it also is affected by the quality of the yeasts available to the flies. 5. The feeding activity of the larvae causes a progressive physical breakdown of the agar‐gel medium, and it is thought that this alteration causes the yeasts to change from a predominantly aerobic to an anaerobic form of respiration, and thus to change their qualitative characteristics. At the same time the yeasts grow, and are carried by the larvae, thus spreading through the medium. So the larvae hatching from the eggs find themselves under different conditions as the culture develops. These differences in the state of the medium, in availability and kind of yeasts and the numbers of other larvae of various ages present, result in a progressive, if irregular, slowing down of pre‐adult development rate and in an increase of pre‐adult mortality. Even during the phase of family growth this mortality may amount to 20–30%. Further, the flies hatching will have had their potential fertility modified by the conditions under which they developed, and this is likely to be of particular significance for the flies hatching out during the later stages of family growth. It is not surprising, then, that the number of adults found in a culture depends very much on the kind of medium used, and on the quality of yeast seeded on to it, or that the course of the growth of the adult population is also very variable and can only rarely be described as logistic. 6. The significance of these findings for ecological theory lies mainly in the emphasis it places on the inter‐specific relationship described, which is much more complex than the simple competition between adult flies previously taken as the main factor determining population growth. Part of this complexity arises from the qualitative alterations occurring in the yeast population, and it is shown that these may be important in work on selection and for other genetic experiments. However, the ecology of Drosophila in nature may involve even more complex reactions, some of which may be of greater significance than those described for laboratory populations. Field and laboratory experimental work could be profitably carried through together with this animal.