Parthenogenetic populations can remain stable in spite of high mutation rate and random drift

Abstract

Genotypic models [1,2] imply that the accumulation of deleterious mutations is a mechanism which severely limits the evolutionary potential of parthenogenetic species. It is thought that high genomic mutation rate under weak selection will lead to genetic deterioration due to random fixation of deleterious mutations. This has been called "Muller's ratchet" [3]. For large populations where Muller's ratchet is unlikely to operate, the label of "evolutionary dead end" usually attached to reproduction by parthenogenesis has been shown to be untenable in phenotypic selection models [4]. Here we simulate weak stabilizing selection in small populations under the conditions for which Muller's ratchet is supposed to work, but we assume fitness to be determined by many quantitative characters. We demonstrate that, as expected, a high mutation rate may hinder the maintenance of an optimal genotype, but contrary to the "ratchet" analogy, we find that mean fitness reaches a stable stochastic equilibrium near the optimum. We simulate a small population of obligately parthenogenetic organisms with discrete and non-overlapping generations. We start with N e mothers. Each mother produces the same number of offspring, and each offspring carries a new mutation with probability /z. We assume that several (NDIM = dimension of the phenotypic space) quantitative characters z=(z~, ..., ZNDIM) contribute multiplicatively to total fitness W(z) of each individual. Defining fitness as survival probability, we apply viability selection by deleting an individual with probability (1-W(z)). After selection the next generation of N e mothers is chosen randomly from 204 geometric reasons. The probability of "back mutations" (which is ignored in the genotypic models [1, 2]), or mutations which decrease the length of the character vector, is given [6] approximately by the survivors. These processes are simulated by Monte Carlo methods: we do not evaluate any model equation numerically, but we simulate directly the stochastic events as occurring in natural populations.The decrease in fitness with increasing distance Izil of each character from its optimal value is modelled as a Gaussian function. By a proper scale transformation the fitness function can be treated with optimum value at zero and width We start with a population at the optimum that carry no mutations. After several generations the frequency of the class without mutations decreases and finally reaches zero by random drift: Muller's ratchet has clicked one notch (Fig. 1). The accumulation of mutations continues at its predicted

Colony structure and reproduction in the thelytokous parthenogenetic ant Platythyrea punctata (F. Smith) (Hymenoptera, Formicidae)

“…This is because parthenogens presumably lack the ability to rapidly generate genotypic diversity in reaction to changing environmental conditions (Maynard Smith, 1978;Itow et al, 1984). Several authors nevertheless pointed out that, based on modeling, rates of phenotypic evolution under unisexual and asexual conditions may be of similar magnitude and parthenogens may do as well as obligately bisexual species under similar environmental conditions assuming elevated mutation rates and depressed environmental sensitivity (Lynch and Gabriel, 1983;Gabriel and Wagner, 1988). The establishment of a parthenogenetic lineage may therefore be a dynamic evolutionary process depending on alterations in life history patterns that determine the successful establishment of an asexual population (Templeton, 1982).…”

Mutation Load and the Survival of Small Populations

“…Under these conditions, locus-specific back-mutation can do litde to stop the operation of Muller's ratchet, except in enormous populations. However, compensatory mutations of other loci influencing the same or different trait may lead to a substantial amount of effective back-mutation for fitness (Gabriel and Wagner, 1988). This is because selection sorts individuals on the basis ofthe composite properties of the entire phenotype regardless of the genotypie states at individualloci (Lynch and Gabriel, 1983).…”

Dispensable and indispensable genes in an ameiotic fish, the Amazon molly <i>Poecilia formosa</i>

“…These models, however, depend on population size. The effect of Muller's ratchet is weaker for larger populations, and, under certain circumstances, parthenogenetic lineages may remain stable (Gabriel and Wagner, 1988). Evidence is accumulating (Darnell and Abramoff, 1968;Avise et al, 1991;Schartl et al, 1995b) that the Amazon molly, as well as another ameiotic poeciliid fish, the hybridogenetic Poeciliopsis (Quattro et al, 1992), and some salamanders (genus Ambystoma) (Hedges et al, 1992;Spolsky et al, 1992) might have survived longer than predicted.…”