A comparison of substitution and addition designs for the analysis of competitive interactions in Drosophila melanogaster

Abstract

Two distinct experimental designs are commonly employed for the analysis of competitive interactions in genetically heterogeneous mixtures. These are referred to as substitution and addition designs respectively, and either may be used quantitatively to separate the effects of intra-and inter-genotypic competition. The results presented in this paper relate to two characters which measure competitive success in mixed cultures of Drosophila melanogaster, namely the proportion of developing individuals which survive to emerge as adults and the mean weight of those adults. For each of these characters the estimates of intra-and inter-genotypic competitive values obtained from substitution and addition design experiments were equivalent. The implications of this result with respect to the choice of a suitable experimental design and its application to competition in various plant and animal systems are discussed. INTRODUCTIONThe importance of competition as an environmental and selective force has long been appreciated (Andrewartha and Birch, 1954;Bakker, 1961).However, detailed investigation of the nature of competitive interactions has been hampered by elusive definitions and the absence of a suitable form of analysis. Recent developments in the analysis of data from competition experiments include the introduction of nearest neighbour and spacing models (Kempton, 1982; Cannel et al., 1984;Benjamin, 1982) which are used primarily in agriculture and forestry related experiments and descriptive, density related regression models (Mather and Caligari, 1981; Spitters 1983a), where the competitive success of a genotype is related to the densities of the various competitors. This latter approach has also been found suitable for the analysis of density related forms of competition in animal systems, as exemplified by the investigation of competition between larvae of Drosophila melanogaster for a constant and limiting amount of food Caligari, 1981, 1983;Eggleston, 1985). Mather and Caligari (1981) A characteristic, therefore, of the substitution design is that the independent variable is negative.Department of Genetics, University of Liverpool, P.O. Box 147, Liverpool L69 3BX, U.K. J. RODRIGUES DE MIRANDA AND P. EGGLESTONThe slope obtained from the monoculture regression (bm) again measures intra-genotypic competition. The slope of the duoculture regression (bd) measures both intra-genotypic competition among the indicator genotype competitors and the inter-genotypic competitive effect that the associate complement has on the performance of the indicator genotype. If the associate has no effect, then bd would be the same as bm. The inter-genotypic competitive value is therefore estimated by the difference between the two slopes (bdbm). The intra-genotypic competitive value is quantified similarly as (b0_bm), b0 being the slope obtained if the indicator competitors had not been substituted. One of the useful features of this design is that with the inclusion of a monoculture density series for the secondary genotype th...

Competitive interactions in Drosophila melanogaster: recurrent selection for aggression and response

“…All of these experiments have identified considerable amounts of genetical variation with respect to competitive interactions (Eggleston, 1985;de Miranda and Eggleston, 1987) and this supports earlier evidence for the heritability of competitive ability in genetical mixtures of Drosophila (Mather and Cooke, 1962;Gale, 1964), Hordeum (Sakai and Gotoh, 1955) and Tribolium (Lerner and Ho, 1961). In addition, the results of competition experiments between D. melanogaster and D. simulans (Moore, 1952) and those between wild type and dumpy wing mutants of D.…”

“…Subsequently, this model for the analysis of competitive interactions (although developed using Drosophila melanogaster as a model system) has been applied to a range of experimental and commercial material including Lolium perenne (Mather, Hill and Caligari, 1982) and Hordeum vulgare (Powell, Caligari and Thomas, 1985). In addition, the flexibility of this multiple linear regression approach has been demonstrated through a variety of experimental designs, including both substitution and addition techniques (de Miranda and Eggleston, 1987) and with a variety of competitive characters (Eggleston, 1987;Powell et a!., 1985).…”

Larval competition in Drosophila melanogaster. II. Comparing biological and competitive parameters

“…The experimental procedures used are identical to those of de Miranda and Eggleston (1987a). In short, competition took place in glass vials each with 50 ml 2 per cent bacto agar as a non nutritive base, and 55O mg live yeast (Sigma YSC-2) as the food source.…”

“…p was transformed to p using the angular transformation (Bartlett 1947) and ii' was transformed to 1/ for regression purposes (de Miranda and Eggleston, 1987a). The results were analysed using the multiple linear regression procedure of Mather and Caligari (1981).…”

The competition diallel and the exploitation and interference components of larval competition in Drosophila melanogaster

“…The diallel is formed by competing each strain with every other strain, including itself, and recording the performance of the primary (indicator) genotype in the presence of various associate competitors. However, with notable exceptions (McGilchrist & Trenbath, 1971) the analysis of such diallels has been largely empirical (Williams, 1962;McGilchrist, 1965;Norrington-Davies, 1967, 1968Breese & Hill, 1973;Mather & Caligari, 1983;Eggleston, 1985;de Miranda & Eggleston, 1987, 1988c with little regard for the biological determinants underlying the diallel. Hence, in this paper we present an analytical model for the competition diallel based on three linear biological parameters, representing both the exploitation (acquisition and utilization of a common resource) and interference (unique resources and mixture benefits) components of competition (Birch, 1957).…”