Studying phenotypic evolution using multivariate quantitative genetics

Quantitative genetics provides a powerful framework for studying phenotypic evolution and the evolution of adaptive genetic variation. Central to the approach is G, the matrix of additive genetic variances and covariances. G summarizes the genetic basis of the traits and can be used to predict the phenotypic response to multivariate selection or to drift. Recent analytical and computational advances have improved both the power and the accessibility of the necessary multivariate statistics. It is now possible to study the relationships between G and other evolutionary parameters, such as those describing the mutational input, the shape and orientation of the adaptive landscape, and the phenotypic divergence among populations. At the same time, we are moving towards a greater understanding of how the genetic variation summarized by G evolves. Computer simulations of the evolution of G, innovations in matrix comparison methods, and rapid development of powerful molecular genetic tools have all opened the way for dissecting the interaction between allelic variation and evolutionary process. Here I discuss some current uses of G, problems with the application of these approaches, and identify avenues for future research.

The stability of P in coral reef fishes

The constancy of phenotypic variation and covariation is an assumption that underlies most recent investigations of past selective regimes and attempts to predict future responses to selection. Few studies have tested this assumption of constancy despite good reasons to expect that the pattern of phenotypic variation and covariation may vary in space and time. We compared phenotypic variance-covariance matrices (P) estimated for Populations of six species of distantly related coral reef fishes sampled at two locations on Australia's Great Barrier Reef separated by more than 1000 km. The intraspecific similarity between these matrices was estimated using two methods: matrix correlation and common principal component analysis. Although there was no evidence of equality between pairs of P, both statistical approaches indicated a high degree of similarity in morphology between the two populations for each species. In general, the hierarchical decomposition of the variance-covariance structure of these populations indicated that all principal components of phenotypic variance-covariance were shared but that they differed in the degree of variation associated with each of these components. The consistency of this pattern is remarkable given the diversity of morphologies and life histories encompassed by these species. Although some phenotypic instability was indicated, these results were consistent with a generally conserved pattern of multivariate selection between populations.

The genetic and environmental relationship between the interpersonal sensitivity measure (IPSM) and the personality dimensions of Eysenck and Cloninger

A shortened version of the Interpersonal Sensitivity Measure (IPSM) developed to predict depression prone personalities was administered in a self-report questionnaire to a community-based sample of 3269 Australian twin pairs aged 18-28 years, along with Eysenck's EPQ and Cloninger's TPQ. The IPSM included four sub-scales: Separation Anxiety (SEP); Interpersonal Sensitivity (INT); Fragile Inner-Self (FIS); and Timidity (TIM). Univariate analysis revealed that individual differences in the IPSM sub-scale scores were best explained by additive genetic and specific environmental effects. Confirming previous research findings, familial aggregation for the EPQ and TPQ personality dimensions was entirely due to additive genetic effects. In the multivariate case, a model comprising additive genetic and specific environmental effects best explained the covariation between the latent factors for male and female twin pairs alike. The EPQ and TPQ dimensions accounted for moderate to large proportions of the genetic variance (40-76%) in the IPSM sub-scales, while most of the non-shared environment variance was unique to the IPSM sub-scales. (C) 2001 Elsevier Science Ltd. All rights reserved.

Genetic covariation of neuroticism with monoamine oxidase activity and smoking

Variation in the personality trait of neuroticism is known to be affected by genetic influences, but despite a number of association studies, the genes involved have not yet been characterized. In a recent study of platelet monoamine oxidase in 1,551 twin subjects, we found a significant association between monoamine oxidase activity and scores on the Eysenck Personality Questionnaire neuroticism scale. Further analyses presented here indicate that both neuroticism and monoamine oxidase activity are associated with variation in smoking habits, and that adjusting for the effect of smoking strengthens the association between MAO and neuroticism. Analysis of the genetic and environmental sources of covariation between neuroticism, smoking, and monoamine oxidase activity show that approximately 8% of the genetic variance in neuroticism is due to the same additive genetic effects that contribute to variation in monoamine oxidase activity, suggesting that variation in neuroticism is associated in part with aspects of serotonin metabolism. (C) 2001 Wiley-Liss, Inc.

Some vocabulary and grammar for the analysis of multi-environment trials, as applied to the analysis of FPB and PPB trials

For the improvement of genetic material suitable for on farm use under low-input conditions, participatory and formal plant breeding strategies are frequently presented as competing options. A common frame of reference to phrase mechanisms and purposes related to breeding strategies will facilitate clearer descriptions of similarities and differences between participatory plant breeding and formal plant breeding. In this paper an attempt is made to develop such a common framework by means of a statistically inspired language that acknowledges the importance of both on farm trials and research centre trials as sources of information for on farm genetic improvement. Key concepts are the genetic correlation between environments, and the heterogeneity of phenotypic and genetic variance over environments. Classic selection response theory is taken as the starting point for the comparison of selection trials (on farm and research centre) with respect to the expected genetic improvement in a target environment (low-input farms). The variance-covariance parameters that form the input for selection response comparisons traditionally come from a mixed model fit to multi-environment trial data. In this paper we propose a recently developed class of mixed models, namely multiplicative mixed models, also called factor-analytic models, for modelling genetic variances and covariances (correlations). Mixed multiplicative models allow genetic variances and covariances to be dependent on quantitative descriptors of the environment, and confer a high flexibility in the choice of variance-covariance structure, without requiring the estimation of a prohibitively high number of parameters. As a result detailed considerations regarding selection response comparisons are facilitated. ne statistical machinery involved is illustrated on an example data set consisting of barley trials from the International Center for Agricultural Research in the Dry Areas (ICARDA). Analysis of the example data showed that participatory plant breeding and formal plant breeding are better interpreted as providing complementary rather than competing information.

Genetic Covariation between Serum {gamma}-Glutamyltransferase Activity and Cardiovascular Risk Factors

Background: Several studies have shown that variation in serum gamma-glutamyltransferase (GGT) in the population is associated with risk of death or development of cardiovascular disease, type 2 diabetes, stroke, or hypertension. This association is only partly explained by associations between GGT and recognized risk factors. Our aim was to estimate the relative importance of genetic and environmental sources of variation in GGT as well as genetic and environmental sources of covariation between GGT and other liver enzymes and markers of cardiovascular risk in adult twin pairs. Methods: We recruited 1134 men and 2241 women through the Australian Twin Registry. Data were collected through mailed questionnaires, telephone interviews, and by analysis of blood samples. Sources of variation in GGT, alanine aminotransferase (ALT), and aspartate aminotransferase (AST) and of covariation between GGT and cardiovascular risk factors were assessed by maximum-likelihood model-fitting. Results: Serum GGT, ALT, and AST were affected by additive genetic and nonshared environmental factors, with heritabilities estimated at 0.52, 0.48, and 0.32, respectively. One-half of the genetic variance in GGT was shared with ALT, AST, or both. There were highly significant correlations between GGT and body mass index; serum lipids, lipoproteins, glucose, and insulin; and blood pressure. These correlations were more attributable to genes that affect both GGT and known cardiovascular risk factors than to environmental factors. Conclusions: Variation in serum enzymes that reflect liver function showed significant genetic effects, and there was evidence that both genetic and environmental factors that affect these enzymes can also affect cardiovascular risk. (C) 2002 American Association for Clinical Chemistry.

Genetic and Environmental Influences on Migraine: A Twin Study Across Six Countries

Migraine is a common neurovascular brain disorder that is manifested in recurrent episodes of disabling headache. The aim of the present study was to compare the prevalence and heritability of migraine across six of the countries that participate in GenomEutwin project including a total number of 29,717 twin pairs. Migraine was assessed by questionnaires that differed between most countries. It was most prevalent in Danish and Dutch females (32% and 34%, respectively), whereas the lowest prevalence was found in the younger and older Finnish cohorts (13% and 10%, respectively). The estimated genetic variance (heritability) was significant and the same between sexes in all countries. Heritability ranged from 34% to 57%, with lowest estimates in Australia, and highest estimates in the older cohort of Finland, the Netherlands, and Denmark. There was some indication that part of the genetic variance was non-additive, but this was significant in Sweden only. In addition to genetic factors, environmental effects that are non-shared between members of a twin pair contributed to the liability of migraine. After migraine definitions are homogenized among the participating countries, the GenomEUtwin project will provide a powerful resource to identify the genes involved in migraine.

Genetic variation in foliar carbon isotope composition in relation to tree growth and foliar nitrogen concentration in clones of the F-1 hybrid between slash pine and Caribbean pine

The objectives of this study were: (1) to quantify the genetic variation in foliar carbon isotope composition (delta(13)C) of 122 clones of ca. 4-year-old F-1 hybrids between slash pine (Pinus elliottii Engelm var. elliottii) and Caribbean pine (Pinus caribaea var. hondurensis Barr.,et Golf.) grown at two field experimental sites with different water and nitrogen availability in southeast Queensland, Australia, in relation to tree growth and foliar nitrogen concentration (N-mass); and (2) to assess the potential of using delta(13)C measurements, in the foliage materials collected from the clone hedges at nursery and the 4-year-old tree canopies in the field, as an indirect index of tree water use efficiency for selecting elite F-1 hybrid pine clones with improved tree growth. There were significant differences in foliar delta(13)C between the nursery hedges and the 4-year-old tree canopies in the field, between the summer and winter seasons, between the two experimental sites, and between the upper outer and lower outer canopy positions sampled. This indicates that delta(13)C measurements in the foliage materials are significantly influenced by the sampling techniques and environmental conditions. Significant differences in foliar delta(13)C, at the upper outer canopy in both field experiments in summer and winter, were detected between the clones, and between the female parents of the clones. Clone means of tree height at age ca. 3 years were positively related to those of the upper outer canopy delta(13)C at both experimental sites in winter, but only for the wetter site in summer. There were positive, linear relationships between clone means of canopy delta(13)C and those of canopy N-mass, indicating that canopy photosynthetic capacity might be an important factor regulating the clonal variation in canopy delta(13)C. Significant correlations were found between clone means of canopy delta(13)C at both experimental sites in summer and winter, and between those at the upper outer and lower outer canopy positions. Mean clone delta(13)C for the nursery hedges was only positively related to mean clone stem diameter at 1.3 m height at age 3 years on the wetter site. The clone by site interaction for foliar delta(13)C at the upper outer canopy was significant only in summer. Overall, the relatively high genetic variance components for foliar delta(13)C and significant, positive correlations between clone means of foliar delta(13)C and tree growth have highlighted the potential of using foliar delta(13)C measurements for assisting in selection of the elite F-1 hybrid pine clones with improved tree growth. (C) 2002 Elsevier Science B.V. All rights reserved.

Divergent environments and population bottlenecks fail to generate premating isolation in Drosophila pseudoobscura

While the feasibility of bottleneck-induced speciation is in doubt, population bottlenecks may still affect the speciation process by interacting with divergent selection. To explore this possibility, I conducted a laboratory speciation experiment using Drosophila pseudoobscura involving 78 replicate populations assigned in a two-way factorial design to both bottleneck (present vs. absent) and environment (ancestral vs. novel) treatments. Populations independently evolved under these treatments and were then tested for assortative mating and male mating success against their common ancestor. Bottlenecks alone did not generate any premating isolation, despite an experimental design that was conducive to bottleneck-induced speciation. Premating isolation also did not evolve in the novel environment treatment, neither in the presence nor absence of bottlenecks. However, male mating success was significantly reduced in the novel environment treatment, both as a plastic response to this environment and as a result of environment-dependent inbreeding effects in the bottlenecked populations. Reduced mating success of derived males will hamper speciation by enhancing the mating success of immigrant, ancestral males. Novel environments are generally thought to promote ecological speciation by generating divergent natural selection. In the current experiment, however, the novel environment did not cause the evolution of any premating isolation and it reduced the likelihood of speciation through its effects on male mating success.

Genetic Simplex Modeling of Eysenck's Dimensions of Personality in a Sample of Young Australian Twins

The relative stability and magnitude of genetic and environmental effects underlying major dimensions of adolescent personality across time were investigated. The Junior Eysenck Personality Questionnaire was administered to over 540 twin pairs at ages 12, 14 and 16 years. Their personality scores were analyzed using genetic simplex modeling which explicitly took into account the longitudinal nature of the data. With the exception of the dimension lie, multivariate model fitting results revealed that familial aggregation was entirely explained by additive genetic effects. Results from simplex model fitting suggest that large proportions of the additive genetic variance observed at ages 14 and 16 years could be explained by genetic effects present at the age of 12 years. There was also evidence for smaller but significant genetic innovations at 14 and 16 years of age for male and female neuroticism, at 14 years for male extraversion, at 14 and 16 years for female psychoticism, and at 14 years for male psychoticism.

Proteomic analysis of k-casein micro-heterogeneity

The proteome of bovine milk is dominated by just six gene products that constitute approximately 95% of milk protein. Nonetheless, over 150 protein spots can be readily detected following two-dimensional electrophoresis of whole milk. Many of these represent isoforms of the major gene products produced through extensive posttranslational modification. Peptide mass fingerprinting of in-gel tryptic digests (using matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) in reflectron mode with alpha-cyano-4-hydroxycinnamic acid as the matrix) identified 10 forms of K-casein with isoelectric point (pl) values from 4.47 to 5.81, but could not distinguish between them. MALDI-TOF MS in linear mode, using sinapinic acid as the matrix, revealed a large tryptic peptide (mass > 5990 Da) derived from the C-terminus that contained all the known sites of genetic variance, phosphorylation and glycosylation. Two genetic variants present as singly or doubly phosphorylated forms could be distinguished using mass data alone. Glycoforms containing a single acidic tetrasaccharide were also identified. The differences in electrophoretic mobility of these isoforms were consistent with the addition of the acidic groups. While more extensively glycosylated forms were also observed, substantial loss of N-acetylneuraminic acid from the glycosyl group was evident in the MALDI spectra such that ions corresponding to the intact glycopeptide were not observed and assignment of the glycoforms was not possible. However, by analysing the pl shifts observed on the two-dimensional gels in conjunction with the MS data, the number of N-acetylneuraminic acid residues, and hence the glycoforms present, could be determined.

Multivariate genetic analysis of academic skills of the Queensland Core Skills Test and IQ highlight the importance of genetic g

This study examined the genetic and environmental relationships among 5 academic achievement skills of a standardized test of academic achievement, the Queensland Core Skills Test (QCST; Queensland Studies Authority, 2003a). QCST participants included 182 monozygotic pairs and 208 dizygotic pairs (mean 17 years +/- 0.4 standard deviation). IQ data were included in the analysis to correct for ascertainment bias. A genetic general factor explained virtually all genetic variance in the component academic skills scores, and accounted for 32% to 73% of their phenotypic variances. It also explained 56% and 42% of variation in Verbal IQ and Performance IQ respectively, suggesting that this factor is genetic g. Modest specific genetic effects were evident for achievement in mathematical problem solving and written expression. A single common factor adequately explained common environmental effects, which were also modest, and possibly due to assortative mating. The results suggest that general academic ability, derived from genetic influences and to a lesser extent common environmental influences, is the primary source of variation in component skills of the QCST.

Substantial changes in the genetic basis of tadpole morphology of Rana lessonae in the presence of predators

Predator-induced morphological plasticity is a model system for investigating phenotypic plasticity in an ecological context. We investigated the genetic basis of the predator-induced plasticity in Rana lessonae by determining the pattern of genetic covariation of three morphological traits that were found to be induced in a predatory environment. Body size decreased and tail dimensions increased when reared in the presence of preying dragonfly larvae. Genetic variance in body size increased by almost an order of magnitude in the predator environment, and the first genetic principal component was found to be highly significantly different between the two environments. The across environment genetic correlation for body size was significantly below 1 indicating that different genes contributed to this trait in the two environments. Body size may therefore be able to respond to selection independently in the two environments to some extent.

A twin study of psychometric intelligence and efficiency of information processing

Intelligence (IQ) can be seen as the efficiency of mental processes or cognition, as can basic information processing (IP) tasks like those used in our ongoing Memory, Attention and Problem Solving (MAPS) study. Measures of IQ and IP are correlated and both have a genetic component, so we are studying how the genetic variance in IQ is related to the genetic variance in IP. We measured intelligence with five subscales of the Multidimensional Aptitude Battery (MAB). The IP tasks included four variants of choice reaction time (CRT) and a visual inspection time (IT). The influence of genetic factors on the variances in each of the IQ, IP, and IT tasks was investigated in 250 identical and nonidentical twin pairs aged 16 years. For a subset of 50 pairs we have test–retest data that allow us to estimate the stability of the measures. MX was used for a multivariate genetic analysis that addresses whether the variance in IQ and IP measures is possibly mediated by common genetic factors. Analyses that show the modeled genetic and environmental influences on these measures of cognitive efficiency will be presented and their relevance to ideas on intelligence will be discussed.

High-breakdown linear discriminant analysis

The classification rules of linear discriminant analysis are defined by the true mean vectors and the common covariance matrix of the populations from which the data come. Because these true parameters are generally unknown, they are commonly estimated by the sample mean vector and covariance matrix of the data in a training sample randomly drawn from each population. However, these sample statistics are notoriously susceptible to contamination by outliers, a problem compounded by the fact that the outliers may be invisible to conventional diagnostics. High-breakdown estimation is a procedure designed to remove this cause for concern by producing estimates that are immune to serious distortion by a minority of outliers, regardless of their severity. In this article we motivate and develop a high-breakdown criterion for linear discriminant analysis and give an algorithm for its implementation. The procedure is intended to supplement rather than replace the usual sample-moment methodology of discriminant analysis either by providing indications that the dataset is not seriously affected by outliers (supporting the usual analysis) or by identifying apparently aberrant points and giving resistant estimators that are not affected by them.