THE MEASUREMENT OF SELECTION ON QUANTITATIVE TRAITS: BIASES DUE TO ENVIRONMENTAL COVARIANCES BETWEEN TRAITS AND FITNESS

The use of regression techniques for estimating the direction and magnitude of selection from measurements on phenotypes has become widespread in field studies. A potential problem with these techniques is that environmental correlations between fitness and the traits examined may produce biased estimates of selection gradients. This report demonstrates that the phenotypic covariance between fitness and a trait, used as an estimate of the selection differential in estimating selection gradients, has two components: a component induced by selection itself and a component due to the effect of environmental factors on fitness. The second component is shown to be responsible for biases in estimates of selection gradients. The use of regressions involving genotypic and breeding values instead of phenotypic values can yield estimates of selection gradients that are not biased by environmental covariances. Statistical methods for estimating the coefficients of such regressions, and for testing for biases in regressions involving phenotypic values, are described.     

COMPONENTS OF PHENOTYPIC SELECTION: POLLEN EXPORT AND FLOWER COROLLA WIDTH IN IPOMOPSIS AGGREGATA

In the hummingbird-pollinated herb Ipomopsis aggregata, selection through male function during pollination favors wide corolla tubes. We explored the mechanisms behind this selection, using phenotypic selection analysis to compare effects of corolla width on two components of male pollination success, pollinator visit rate and pollen exported per visit. During single visits by captive hummingbirds, flowers with wider corollas exported more pollen, and more dye used as a pollen analogue, to stigmas of recipient flowers. Corolla width was less strongly related to visit rate in the field, and had no direct effect on visit rate after nectar production and corolla length were controlled for. Moreover, the phenotypic selection differential was 80% higher for the effect on pollen exported per visit, suggesting that this is the more important mechanism of selection.     

ON THE LOW HERITABILITY OF LIFE-HISTORY TRAITS

Life-history traits such as longevity and fecundity often show low heritability. This is usually interpreted in terms of Fisher's fundamental theorem to mean that populations are near evolutionary equilibrium and genetic variance in total fitness is low. We develop the causal relationship between metric traits and life-history traits to show that a life-history trait is expected to have a low heritability whether or not the population is at equilibrium. This is because it is subject to all the environmental variation in the metric traits that affect it plus additional environmental variation. There is no simple prediction regarding levels of additive genetic variance in life-history traits, which may be high at equilibrium. Several other patterns in the inheritance of life-history traits are readily predicted from the causal model. These include the strength of genetic correlations between life-history traits, levels of nonadditive genetic variance, and the inevitability of genotype-environment interaction.     

THE INTERACTION PHENOTYPE IN THE DROSOPHILA WILLISTONI–SPIROPLASMA SYMBIOSIS

Both the population and coevolutionary dynamics of hereditary male-lethal endosymbionts, found in a wide range of insect species, depend on host fitness and endosymbiont transmission rates. This paper reports on fitness effects and transmission rates in three lines of Drosophila willistoni infected with either male-lethal spiroplasmas or a spontaneous nonmale-lethal mutant. Overall fitness measures were reduced or unaffected by the infection; however, some infected females produced more offspring in early broods. Maternal transmission rates were high, but imperfect, and varied with a female's age, host line, and spiroplasma type. No evidence for paternal or horizontal transmission was found. If an altered temporal pattern of reproduction is not a factor in countering the loss of spiroplasma hosts through imperfect maternal transmission, persistence of this endoparasitism remains unexplained. Tolerance of the infection and ability to transmit bacteria varied with both host and spiroplasma line. Analysis of the interaction between the spontaneous nonmale-lethal mutant and its host suggests this symbiosis has undergone coevolution under laboratory culture.     

Fitness consequences of helping behavior in the western bluebird

We examined the fitness consequences of helping behavior inthe western bluebird (Sialia mexicana) at Hastings Reservationin Carmel Valley, California, USA, and tested hypotheses forhow helpers benefit from engaging in alloparental behavior.Both juvenile and adult western bluebirds occasionally helpat the nest During a 12 year period, all adult helpers and mostjuvenile helpers were male. Helpers usually fed at nests ofboth their parents and rarely helped when only one parent waspresent. The frequency of pairs with adult helpers was only7%, but nearly one-third of adult males helped among those withboth parents on the study area. At least 28% were breeders whosenests failed. The propensity to help appears to depend uponparental survival, male philopatry, and the breeding successof potential helpers. Feeding rates were not increased at nestswith juvenile helpers, apparendy because breeding males reducedtheir feeding rates. In contrast, adult helpers increased theoverall rates of food delivery to the nest in spite of a reductionin the number of feeding trips made by both male and femaleparents. Helpers did not derive any obvious direct fitness benefitsfrom helping, but they had greater indirect fitness than nonhelpersdue to increases in nestling growth rates and fledging successat their parents' nests. Helpers fledged fewer offspring intheir first nests than did nonhelpers, suggesting that theywere birds with reduced reproductive potential. Although wehave not yet measured the effect of extrapair fertilizationson the fitness benefits of helping, we calculated the differencein fitness between helpers and nonhelpers as a function of thepotential helper's paternity when breeding independently andhis father's paternity in the nest at which he might help. Inconjunction with constraints on breeding and indirect fitnessbenefits, we predict that relatedness of males to the youngin their own as well as their parents' nests will influencehelping behavior in western bluebirds.     

Fitness effects of sex ratio response to host quality and size in the parasitoid wasp Spalangia cameroni

The parasitoid wasp Spalangia cameroni oviposited a greaterproportion of daughters in stable fly pupae than in house flypupae, even when I controlled for stable flies being smallerthan house flies. Sex ratio manipulation in response to hostquality has been modeled as being adaptive through an effectof host quality on the size and hence offspring production ofdaughters. 5. cameronis response to host species may insteadbe adaptive through an effect on larval survivorship, the developmenttime of daughters, and the size of sons. There was greater survivalof daughters than sons on stable flies. Controlling for hostsize, I found that development time of daughters was about 2%less on stable flies than on house flies. The decrease in developmenttime corresponds to a 2% increase in fitness as estimated byr, the intrinsic rate of increase, and is equivalent to abouta 9% increase in offspring production. Sons were about 2% largerfrom house flies than stable flies, which may increase offspringproduction by up to 3%. Host species had no consistent effecton size of daughters or development time of sons. In additionto the response to host species, mothers oviposited a greaterproportion of daughters in larger stable fly hosts. Whetherthis behavior is adaptive is unclear. Although offspring werelarger when they developed on larger stable flies, the rateof increase was less for daughters dian for sons. Effects ofstable fly size on offspring development time were negligible.     

Rainfall and flowering synchrony in a tropical shrub: Variable selection on the flowering time ofErythroxylum havanense

Summary We tested the adaptive significance of flowering synchrony by means of a quantitative analysis of selection and by flowering induction experiments with the deciduous shrubErythroxylum havanense. Temporal schedules of flower and fruit production were determined for a local population (in three sites) in a Mexican seasonal forest for 2 years (1987–1988). The consequences of natural variation in flowering time (flowering initiation day) on maternal reproductive success (fecundity) were evaluated. We observed high levels of inter- and intraindividual flowering synchrony in 1987, but not in 1988 and this contrast was related to differences in rainfall patterns between the two years. A significant proportion (15.4%) of the phenotypic variation in flowering initiation day was accounted for by environmental variance. The expression of phenotypic variance of flowering time and, consequently, the opportunity for selection to act, are controlled by annual variation in rainfall. Despite the between-year difference in flowering synchrony, we detected a relatively intense directional selection on flowering initiation day in both years, but selection coefficients were of opposite sign (standardized directional gradients were –0.326 and 0.333 for 1987 and 1988, respectively). For both years there was a significant relationship between individual relative fitness and the number of neighbouring flowering plants in a given day, suggesting positive frequency-dependent selection.     

The effect of B chromosomes on fitness components inAegilops speltoides Tausch

The effect of B chromosomes (Bs) on various components of vigour and fertility inAegilops speltoides was studied. We designed three types of synthetic populations. In the first type we compared plants from the localities Haifa (which has Bs in nature) and Ashkelon (which has not). Non significant differences were found between the plants from both populations, and between B carriers and non carriers among Haifa individuals. In the second synthetic population we compared 0B, 1B, 2B, 3B and 5B plants from Haifa. We found that vigour variables were not affected by the presence of Bs, while fertility variables were significantly correlated with B number, in such a way that 1B plants produce more grains per plant than any other constitution, while 5Bs produce a remarkable deleterious effect. In the third type of synthetic populations we compared plants with 0B and 1B, 0B and 2B, 0B and 3B, and 0B and 5B. In this case we also found that 1B plants were the most fertile, but the variation of reproductive fitness was higher from plot to plot within the same B class than among B classes. We conclude that the effect of B chromosomes on the fitness variables that we have estimated is close to neutrality in this species. We discuss the relation between the strength of the mechanism of accumulation and the deleterious effects of B chromosomes.     

Assembly of local communities: consequences of an optimal body size for the organization of competitively structured communities

Much empirical evidence suggests that there is an optimal body size for mammals and that this optimum is in the vicinity of l00g. This presumably reflects an underlying fitness function that is greatest at this mass. Here, I combine such a fitness function with an equilibrium model of competitive character displacement to assess the potential influence of a globally optimal body size in structuring local ecological communities. The model accurately predicts the range of body sizes and the average difference in size for species in communities of varying species richness. The model also predicts a uniform spacing of body sizes, rather than the gaps and clumps in the sizes of coexisting species observed in real communities. Alternative explanations for this phenomenon are discussed. The allometric relationships that result in a body size optimum subsume a large number of characteristics associated with the physiological, behavioral, demographic, and evolutionary dynamics of the species. Further integration of the underlying dynamics (e.g. individual energetics) of these relationships into all hierarchical levels of ecology will have to incorporate multiple interactive sites, spatial heterogeneity, and phylogenetic structure, but it has the potential to provide important discoveries into the means by which natural selection operates.     

Plasmid RFLP profiling and DNA homology inThermus isolated from hot springs of different geographical areas

Thirty-four strains belonging to various species of the genus Thermus (T. aquaticus, "T. thermophilus," "T. brockianus," T. scotoductus, and genomic species 2) isolated from hot springs of different geographical areas were examined for plasmid content and restriction fragment length polymorphism (RFLP) of plasmid DNAs. The four strains of the numerical taxonomy cluster E of genomic species 2 did not harbor plasmid DNA. Overall examination of the HindIII-RFLP profiling of plasmid DNA showed considerable variability between and within genomic species, with the exception of presumed clonal isolates. In spite of this heterogeneity, HindIII plasmid digests within a numerical taxonomic cluster gave a subset of restriction fragments of similar or identical length. Strains belonging to genomic species 2 or unclassified isolates from S. Pedro do Sul that harbored plasmid DNA (7 of the 14 strains studied) exhibited strong DNA homology between plasmid regions. No homologous sequences to these plasmid regions were found in chromosomal DNA from strains isolated from S. Pedro do Sul in which no plasmids were detected. The strains belonging to T. scotoductus formed two plasmid DNA homology groups, as estimated by probing with a plasmid fragment that coincided with the two numerical taxonomy clusters proposed previously. Among the other species, homology of plasmid regions was also found between some strains. Strong homology was also found between plasmid regions from some strains of different taxonomic groups, isolated from the same and from different sources, suggesting that these sequences are highly conserved in plasmids present in Thermus. For plasmid-containing strains, results of plasmid RFLP profiling/DNA homology appear promising for the typing of Thermus at the level of biotypes or of individual strains, namely, for monitoring the diversity and frequency of isolates from a particular hot spring. Received: 24 October 1994 / Accepted: 6 March 1995