The constancy of the G matrix through species divergence and the effects of quantitative genetic constraints on phenotypic evolution: a case study in crickets

Long-term phenotypic evolution can be modeled using the response-to-selection equation of quantitative genetics, which incorporates information about genetic constraints (the G matrix). However, little is known about the evolution of G and about its long-term importance in constraining phenotypic evolution. We first investigated the degree of conservation of the G matrix across three species of crickets and qualitatively compared the pattern of variation of G to the phylogeny of the group. Second, we investigated the effect of G on phenotypic evolution by comparing the direction of greatest quantitative genetic variation within species (g(max)) to the direction of phenotypic divergence between species (Delta(z)). Each species, Gryllus veletis, G. firmus, and G. pennsylvanicus, was reared in the laboratory using a full-sib breeding design to extract quantitative genetic information. Five morphological traits related to size were measured. G matrices were compared using three statistical approaches: the T method, the Flury hierarchy, and the MANOVA method. Results revealed that the differences between matrices were small and mostly caused by differences in the magnitude of the genetic variation, not by differences in principal component structure. This suggested that the G matrix structure of this group of species was preserved, despite significant phenotypic divergence across species. The small observed differences in G matrices across species were qualitatively consistent with genetic distances, whereas ecological information did not provide a good prediction of G matrix variation. The comparison of g(max) and Delta(z) revealed that the angle between these two vectors was small in two of three species comparisons, whereas the larger angle corresponding to the third species comparison was caused in large part by one of the five traits. This suggests that multivariate phenotypic divergence occurred mostly in a direction predicted by the direction of greatest genetic variation, although it was not possible to demonstrate the causal relationship from G to Delta(z). Overall, this study provided some support for the validity of the predictive power of quantitative genetics over evolutionary time scales.     

Comparing G matrices: a MANOVA approach

There is considerable interest in comparing genetic variance-covariances matrices (G matrix). However, present methods are difficult to implement and cannot readily be extended to incorporate effects of other variables such as habitat, sex, or location. In this paper I present a method based on MANOVA that can be done using only standard statistical packages (coding for the method using SPLUS is available from the author). The crux of the approach is to use the jackknife method to estimate the pseudovalues of the estimates; these estimates can then be used as datapoints in a MANOVA. I illustrate the method using two published datasets: (1) variation in G matrices resulting from differences in rearing condition, species, and sex in the crickets Gryllus firmus and G. pennsylvanicus; and (2) variation in G matrices associated with habitat and history in the amphipod Gammarus minus.     

Geographic variation in the G matrices of wild populations of the barn swallow

In this paper, we present an analysis of genetic variation in three wild populations of the barn swallow, Hirundo rustica. We estimated the P, E, and G matrices for six linear morphological measurements and tested for variation among populations using the Flury hierarchical method and the jackknife followed by MANOVA method. Because of nonpositive-definite matrices, we had to employ 'bending' to analyse the G and E matrices with the Flury method. Both statistical methods agree in finding that the P and G matrices are significantly different but comparison between the analysis of the P matrices and pairwise analyses of the P, E, and G matrices suggests caution in interpreting the Flury results concerning differences in matrix structure. The significant variation among the populations in the G matrices appears to be due in large measure to the most geographically distant population.     

Comparison of genetic (co)variance matrices within and between Scabiosa canescens and S. columbaria

In the current study, we used bootstrap analyses and the common principal component (CPC) method of Flury (1988) to estimate and compare the G ‐matrix of Scabiosa columbaria and S. canescens populations. We found three major patterns in the G ‐matrices: (i) the magnitude of the (co)variances was more variable among characters than among populations, (ii) different populations showed high (co)variance for different characters, and (iii) there was a tendency for S. canescens to have higher genetic (co)variances than S. columbaria. The hypothesis of equal G ‐matrices was rejected in all comparisons and there was no evidence that the matrices differed by a proportional constant in any of the analyses. The two ‘species matrices’ were found to be unrelated, both for raw data and data standardized over populations, and there was significant between‐population variation in the G ‐matrix in both species. Populations of S. canescens showed conservation of structure (principal components) in their G ‐matrices, contrasting with the lack of common structure among the S. columbaria matrices. Given these observations and the results from previous studies, we propose that selection may be responsible for some of the variation between the G ‐matrices, at least in S. columbaria and at the between‐species level.     

From micro- to macroevolution through quantitative genetic variation: positive evidence from field crickets

Quantitative genetics has been introduced to evolutionary biologists with the suggestion that microevolution could be directly linked to macroevolutionary patterns using, among other parameters, the additive genetic variance/ covariance matrix (G) which is a statistical representation of genetic constraints to evolution. However, little is known concerning the rate and pattern of evolution of G in nature, and it is uncertain whether the constraining effect of G is important over evolutionary time scales. To address these issues, seven species of field crickets from the genera Gryllus and Teleogryllus were reared in the laboratory, and quantitative genetic parameters for morphological traits were estimated from each of them using a nested full-sibling family design. We used three statistical approaches (T method, Flury hierarchy, and Mantel test) to compare G matrices or genetic correlation matrices in a phylogenetic framework. Results showed that G matrices were generally similar across species, with occasional differences between some species. We suggest that G has evolved at a low rate, a conclusion strengthened by the consideration that part of the observed across-species variation in G can be explained by the effect of a genotype by environment interaction. The observed pattern of G matrix variation between species could not be predicted by either morphological trait values or phylogeny. The constraint hypothesis was tested by comparing the multivariate orientation of the reconstructed ancestral G matrix to the orientation of the across-species divergence matrix (D matrix, based on mean trait values). The D matrix mainly revealed divergence in size and, to a much smaller extent, in a shape component related to the ovipositor length. This pattern of species divergence was found to be predictable from the ancestral G matrix in agreement with the expectation of the constraint hypothesis. Overall, these results suggest that the G matrix seems to have an influence on species divergence, and that macroevolution can be predicted, at least qualitatively, from quantitative genetic theory. Alternative explanations are discussed.     

VARIATION IN GENETIC ARCHITECTURE OF CALLING SONG AMONG POPULATIONS OF ALLONEMOBIUS SOCIUS,A. FASCIATUS,AND A HYBRID POPULATION: DRIFT OR SELECTION?

Predictions using quantitative genetic models generally assume that the variance-covariance matrices remain constant over time. This assumption is based on the supposition that selection is generally weak and hence variation lost through selection can be replaced by new mutations. Whether this is generally true can only be ascertained from empirical studies. Ideally for such a study we should be able to make a prediction concerning the relative strength of selection versus genetic drift. If the latter force is prevalent then the variance-covariances matrices should be proportional to each other. Previous studies have indicated that females in the two sibling cricket species Allonemobius socius and A. fasciatus do not discriminate between males of the two species by their calling song. Therefore, differences between the calling song of the two males most likely result from drift rather than sexual selection. We test this hypothesis by comparing the genetic architecture of calling song of three populations of A. fasciatus with two populations of A. socius. We found no differences among populations within species, but significant differences in the G (genetic) and P (phenotypic) matrices between species, with the matrices being proportional as predicted under the hypothesis of genetic drift. Because of the proportional change in the (co)variances no differences between species are evident in the heritabilities or genetic correlations. Comparison of the two species with a hybrid population from a zone of overlap showed highly significant nonproportional variation in genetic architecture. This variation is consistent with a general mixture of two separate genomes or selection. Qualitative conclusions reached using the phenotypic matrices are the same as those reached using the genetic matrices supporting the hypothesis that the former may be used as surrogate measures of the latter.     

QUANTITATIVE GENETICS OF THE WING COLOR PATTERN IN THE BUCKEYE BUTTERFLY (PRECIS COENIA AND PRECIS EVARETE): EVIDENCE AGAINST THE CONSTANCY OF G

Models for the evolution of continuously varying traits use heritabilities, genetic correlations, and the G -matrix to quantify the genetic variation upon which selection acts. Given estimates of these parameters, it is possible to predict the long-term effects of selection, infer past selective forces responsible for observed differences between populations or species, and distinguish the effects of drift from selection. Application of these methods, however, requires the unproven assumption that the G -matrix remains constant from one generation to the next. This study examines the assumption of constancy for the wing pattern characteristics of two sibling species of butterflies, Precis coenia and P. evarete (Lepidoptera: Nymphalidae). Quantitative genetic parameters were estimated from parent-offspring regression. Two approaches were taken to test the null hypothesis of equality between species. First, pairwise tests between corresponding elements of G and between heritabilities and genetic correlations for the two species were constructed. Second, a modification of Bartlett's modified likelihood-ratio test was used to test for equality between the G -matrices. The matrix test failed to detect any between species differences. In contrast, pairwise comparision revealed significant differences. Thus, it appears that constancy cannot be assumed at the species level in quantitative genetic studies. In particular, the assumption of constancy was violated for the trait with the greatest difference in mean phenotype.     

EMPIRICAL COMPARISON OF G MATRIX TEST STATISTICS: FINDING BIOLOGICALLY RELEVANT CHANGE

A central assumption of quantitative genetic theory is that the breeder's equation ( R = GP ⁻¹ S ) accurately predicts the evolutionary response to selection. Recent studies highlight the fact that the additive genetic variance–covariance matrix ( G ) may change over time, rendering the breeder's equation incapable of predicting evolutionary change over more than a few generations. Although some consensus on whether G changes over time has been reached, multiple, often-incompatible methods for comparing G matrices are currently used. A major challenge of G matrix comparison is determining the biological relevance of observed change. Here, we develop a "selection skewers" G matrix comparison statistic that uses the breeder's equation to compare the response to selection given two G matrices while holding selection intensity constant. We present a bootstrap algorithm that determines the significance of G matrix differences using the selection skewers method, random skewers, Mantel's and Bartlett's tests, and eigenanalysis. We then compare these methods by applying the bootstrap to a dataset of laboratory populations of Tribolium castaneum . We find that the results of matrix comparison statistics are inconsistent based on differing a priori goals of each test, and that the selection skewers method is useful for identifying biologically relevant G matrix differences.     

Similarity in G matrix structure among natural populations of Arabidopsis lyrata

Understanding the stability of the G matrix in natural populations is fundamental for predicting evolutionary trajectories; yet, the extent of its spatial variation and how this impacts responses to selection remain open questions. With a nested paternal half‐sib crossing design and plants grown in a field experiment, we examined differences in the genetic architecture of flowering time, floral display, and plant size among four Scandinavian populations of Arabidopsis lyrata. Using a multivariate Bayesian framework, we compared the size, shape, and orientation of G matrices and assessed their potential to facilitate or constrain trait evolution. Flowering time, floral display and rosette size varied among populations and significant additive genetic variation within populations indicated potential to evolve in response to selection. Yet, some characters, including flowering start and number of flowers, may not evolve independently because of genetic correlations. Using a multivariate framework, we found few differences in the genetic architecture of traits among populations. G matrices varied mostly in size rather than shape or orientation. Differences in multivariate responses to selection predicted from differences in G were small, suggesting overall matrix similarity and shared constraints to trait evolution among populations.     

Population divergence of genetic (co)variance matrices in a subdivided plant species,Brassica cretica

Abstract The present study of Brassica cretica had two objectives. First, we compared estimates of population structure (Q_st) for seven phenotypic characters with the corresponding measures for allozyme markers (F_st) to evaluate the supposition that genetic drift is a major determinant of the evolutionary history of this species. Secondly, we compared the genetic (co)variance ( G ) matrices of five populations to examine whether a long history of population isolation is associated with large, consistent differences in the genetic (co)variance structure. Differences between estimates of F_st and Q_st were too small to be declared significant, indicating that stochastic processes have played a major role in the structuring of quantitative variation in this species. Comparison of populations using the common principal component (CPC) method rejected the hypothesis that the G matrices differed by a simple constant of proportionality: most of the variation involved principal component structure rather than the eigenvalues. However, there was strong evidence for proportionality in comparisons using the method of percentage reduction in mean‐square error (MSE), at least when characters with unusually high (co)variance estimates were included in the analyses. Although the CPC and MSE methods provide different, but complementary, views of G matrix variation, we urge caution in the use of proportionality as an indicator of whether genetic drift is responsible for divergence in the G matrix.     

The alignment between phenotypic plasticity,the major axis of genetic variation and the response to selection

Phenotypic plasticity is the ability of a genotype to produce more than one phenotype in order to match the environment. Recent theory proposes that the major axis of genetic variation in a phenotypically plastic population can align with the direction of selection. Therefore, theory predicts that plasticity directly aids adaptation by increasing genetic variation in the direction favoured by selection and reflected in plasticity. We evaluated this theory in the freshwater crustacean Daphnia pulex, facing predation risk from two contrasting size-selective predators. We estimated plasticity in several life-history traits, the G matrix of these traits, the selection gradients on reproduction and survival, and the predicted responses to selection. Using these data, we tested whether the genetic lines of least resistance and the predicted response to selection aligned with plasticity. We found predator environment-specific G matrices, but shared genetic architecture across environments resulted in more constraint in the G matrix than in the plasticity of the traits, sometimes preventing alignment of the two. However, as the importance of survival selection increased, the difference between environments in their predicted response to selection increased and resulted in closer alignment between the plasticity and the predicted selection response. Therefore, plasticity may indeed aid adaptation to new environments.     

Environmentally triggered variability in the genetic variance–covariance of herbivory resistance of an exotic plant Solidago altissima

The variability in the genetic variance–covariance (G‐matrix) in plant resistance and its role in the evolution of invasive plants have been long overlooked. We conducted an additional analysis of the data of a reciprocal transplant experiment with tall goldenrod, Solidago altissima, in multiple garden sites within its native range (USA) and introduced range (Japan). We explored the differences in G‐matrix of resistance to two types of foliar herbivores: (a) a lace bug that is native to the USA and recently introduced to Japan, (b) and other herbivorous insects in response to plant origins and environments. A negative genetic covariance was found between plant resistances to lace bugs and other herbivorous insects, in all combinations of garden locations and plant origins except for US plants planted in US gardens. The G‐matrix of the resistance indices did not differ between US and Japanese plants either in US or Japanese gardens, while it differed between US and Japanese gardens in both US and Japanese plants. Our results suggested that the G‐matrix of the plant resistance may have changed in response to novel environmental differences including herbivore communities and/or other biotic and abiotic factors in the introduced range. This may have revealed a hidden trade‐off between resistances, masked by the environmental factors in the origin range. These results suggest that the stability of the genetic covariance during invasion, and the environmentally triggered variability in the G‐matrices of plant resistance may help to protect the plant against multiple herbivore species without changing its genetic architecture and that this may lead to a rapid adaptation of resistance in exotic plants. Local environments of the plant also have a critical effect on plant resistance and should be considered in order to understand trait evolution in exotic plants.     

Sex‐specific responses of phenotypic diversity to environmental variation

Identifying the factors generating ecomorphological diversity within species can provide a window into the nascent stages of ecological radiation. Sexual dimorphism is an obvious axis of intraspecific morphological diversity that could affect how environmental variation leads to ecological divergence among populations. In this paper we test for sex‐specific responses in how environmental variation generates phenotypic diversity within species, using the generalist lizard Gallotia galloti on Tenerife (Canary Islands). We evaluate two hypotheses: the first proposes that different environments have different phenotypic optima, leading to shifts in the positions of populations in morphospace between environments; the second posits that the strength of trait‐filtering differs between environments, predicting changes in the volume of morphospace occupied by populations in different environments. We found that intraspecific morphological diversity, provided it is adaptive, arises from both shifts in populations’ position in morphospace and differences in the strength of environmental filtering among environments, especially at high elevations. However, effects were found only in males; morphological diversity of females responded little to environmental variation. These results within G. galloti suggest natural selection is not the sole source of phenotypic diversity across environments, but rather that variation in the strength of, or response to, sexual selection may play an important role in generating morphological diversity in environmentally diverse settings. More generally, disparities in trait–environment relationships among males and females also suggest that ignoring sex differences in studies of trait dispersion and clustering may produce misleading inferences.     

Genotype?��?environment interaction is weaker in genitalia than in mating signals and body traits in Enchenopa treehoppers (Hemiptera: Membracidae)

Theory predicts that selection acting across environments should erode genetic variation in reaction norms; i.e., selection should weaken genotype × environment interaction (G × E). In spite of this expectation, G × E is often detected in fitness-related traits. It thus appears that G × E is at least sometimes sustained under selection, a possibility that highlights the need for theory that can account for variation in the presence and strength of G × E. We tested the hypothesis that trait differences in developmental architecture contribute to variation in the expression of G × E. Specifically, we assessed the influence of canalization (robustness to genetic or environmental perturbations) and condition-dependence (association between trait expression and prior resource acquisition or vital cellular processes). We compared G × E across three trait types expected to differ in canalization and condition-dependence: mating signals, body size-related traits, and genitalia. Because genitalia are expected to show the least condition-dependence and the most canalization, they should express weaker G × E than the other trait types. Our study species was a member of the Enchenopa binotata species complex of treehoppers. We found significant G × E in most traits; G × E was strongest in signals and body traits, and weakest in genitalia. These results support the hypothesis that trait differences in developmental architecture (canalization and condition-dependence) contribute to variation in the expression of G × E. We discuss implications for the dynamics of sexual selection on different trait types.     

Strategies for the genetic modification of the medicinal plant Catharanthus roseus (L.) G. Don

This review describes the different plant transformation techniques, including guided infection with Agrobacterium tumefaciens and A. rhizogenes, particle bombardment and protoplast fusion, that have been attempted to create transgenic Catharanthus roseus (L.) G. Don cell cultures, hairy roots and whole plants. The review also focuses on the different approaches used to manipulate and improve secondary metabolite yields in various culture systems, with special attention to the most relevant results achieved. Finally, under future perspectives, the authors propose several approaches which would likely be implemented with this species, to try to boost the accumulation of the anti-tumour agents, vinblastine and vincristine. Some comments on how the future of the genetic manipulation of medicinal plants may proceed aiming at achieving higher secondary metabolite yields are also given.     

QUANTITATIVE GENETICS OF MORPHOLOGICAL DIFFERENTIATION IN PEROMYSCUS. I. TESTS OF THE HOMOGENEITY OF GENETIC COVARIANCE STRUCTURE AMONG SPECIES AND SUBSPECIES

Phenotypic and additive genetic covariance matrices were estimated for 15 morphometric characters in three species and subspecies of Peromyscus. Univariate and multivariate ANOVAs indicate these groups are highly diverged in all characters, P. leucopus having the largest body size, P. maniculatus bairdii the smallest, and P. maniculatus nebrascensis being intermediate. Comparing the structure of P and G within each taxon revealed significant similarities in all three cases. This proportionality was strong enough to justify using P in the place of G to analyze evolutionary processes using quantitative genetic models when G can not be estimated, as in fossil material. However, the similarity between genetic and phenotypic covariance structures is sufficiently low that estimates of the genetic parameters should be used when possible. The additive genetic covariance matrices were compared to examine the assumption that they remain constant during evolution, an assumption which underlies many applications of quantitative-genetic models. While matrix permutation tests indicated statistically significant proportionality between the genetic covariance structures of the two P. maniculatus subspecies, there is no evidence of significant genetic structural similarity between species. This result suggests that the assumption of constant genetic covariance structure may be valid only within species. (It does not, however, necessarily imply a causal relationship between speciation and heterogeneity of genetic covariance structures.) The low matrix correlation for the two P. maniculatus subspecies' genetic covariance matrices indicates G may not be functionally constant, even within species. The lack of similarity observed here may be due partly to sampling variation.     

A note on simulating null distributions for G matrix comparisons

Genetic variances and covariances, summarized in G matrices, are key determinants of the course of adaptive evolution. Consequently, understanding how G matrices vary among populations is critical to answering a variety of questions in evolutionary biology. A method has recently been proposed for generating null distributions of statistics pertaining to differences in G matrices among populations. The general approach facilitated by this method is likely to prove to be very important in studies of the evolution of G . We have identified an issue in the method that will cause it to create null distributions of differences in G matrices that are likely to be far too narrow. The issue arises from the fact that the method as currently used generates null distributions of statistics pertaining to differences in G matrices across populations by simulating breeding value vectors based on G matrices estimated from data, randomizing these vectors across populations, and then calculating null values of statistics from G matrices that are calculated directly from the variances and covariances among randomized vectors. This calculation treats breeding values as quantities that are directly measurable, instead of predicted from G matrices that are themselves estimated from patterns of covariance among kin. The existing method thus neglects a major source of uncertainty in G matrices, which renders it anti‐conservative. We first suggest a correction to the method. We then apply the original and modified methods to a very simple instructive scenario. Finally, we demonstrate the use of both methods in the analysis of a real data set.     

Comparing macroecological patterns across continents: evolution of climatic niche breadth in varanid lizards

Macroecological analyses often test hypotheses at the global scale, or among more closely related species in a single region (e.g. continent). Here, we test several hypotheses about climatic niche widths among relatively closely related species that occur across multiple continents, and compare patterns within and across continents to see if they differ. We focus on the lizard genus Varanus (monitor lizards), which occurs in diverse environments in Africa, Asia, and Australia. We address three main questions. 1) How do climatic niche breadths of species on a given niche axis change based on the position of species along that niche axis? (E.g. are species that occur in more extreme environments more narrowly specialized for those conditions?) 2) Are there trade‐offs in niche breadths on temperature and precipitation axes among species, or are niche widths on different axes positively related? 3) Is variation in niche breadths among species explained primarily by within‐locality seasonal variation, or by differences in climatic conditions among localities across the species range? We generate a new time‐calibrated phylogeny for Varanus and test these hypotheses within and between continents using climatic data and phylogenetic methods. Our results show that patterns on each continent often parallel each other and global patterns. However, in many other cases, the strength of relationships can change dramatically among closely related species on different continents. Overall, we found that: 1) species in warmer environments have narrower temperature niche breadths, but there is no relationship between precipitation niche breadth and niche position; 2) temperature and precipitation niche breadths tend to be positively related among species, rather than showing trade‐offs; and 3) within‐locality seasonal variation explains most variation in climatic niche breadths. Some of these results are concordant with previous studies (in amphibians and North American lizards), and might represent general macroecological patterns.     

Transient sensitivities of non-indigenous shrub species indicate complicated invasion dynamics

As biological invasions increasingly affect natural systems, the need for methods that can quantify the processes responsible for invasion success has increased. Further, methods should be geared to the formulation of management strategies. Demographic analyses are designed to explore the causes and properties of population change. Matrix population models, a commonly used technique for demographic analysis, have been applied to the analysis of stage-structured populations. However, most commonly, analyses have focused on long-term outcomes dynamics (ergodic dynamics). The methods available for analysis of matrix population models have recently been extended to facilitate analysis of the transient dynamics most important to invasion analysis. In this paper we analyze the transient population dynamics of three invasive shrubs and compare them to ergodic dynamics. Cytisus scoparius, Clidemia hirta, and Ardisia elliptica come from different parts of the world and are all now found in the United States of America. They also have published transition matrices that measure the probabilities that any one life-history stage will transition to another over an annual time step. These matrices have been estimated from multi-year data collected from plots in various environments. Our comparative study of transient and ergodic dynamics of invasive shrubs shows that, for all the considered shrub species, there was a clear difference between the sensitivities drawn from these two approaches. The transient sensitivities of earlier life-history transitions showed magnified importance relative to ergodic sensitivities. This was especially true of A. elliptica for which the stable population structure was most different from the starting structure analyzed in detail here. For other species, as stable population structures were heavily weighted towards early stages, the differences in the importance of early transitions transiently and ergodically were less dramatic. Late life transitions showed magnified importance in areas towards the center of the invasion or in older invasion areas. Finally, populations with shorter estimated generation times show greater transient sensitivity to early life-history stages; but the pattern was complex and varied according to species, and was also observed across other life-history transitions. Overall, the ambiguity and complexity of the results highlight the power of considering transient population dynamics for invading species, as well as the importance of specific biological and ecological knowledge of the invading species. Although there may be commonalities across invasions, important decisions on control or inference on population dynamics should treat invasions as individual, unique events.     

Structure and stability of genetic variance–covariance matrices: A Bayesian sparse factor analysis of transcriptional variation in the three‐spined stickleback

The genetic variance–covariance matrix ( G ) is a quantity of central importance in evolutionary biology due to its influence on the rate and direction of multivariate evolution. However, the predictive power of empirically estimated G ‐matrices is limited for two reasons. First, phenotypes are high‐dimensional, whereas traditional statistical methods are tuned to estimate and analyse low‐dimensional matrices. Second, the stability of G to environmental effects and over time remains poorly understood. Using Bayesian sparse factor analysis (BSFG) designed to estimate high‐dimensional G ‐matrices, we analysed levels variation and covariation in 10,527 expressed genes in a large (n = 563) half‐sib breeding design of three‐spined sticklebacks subject to two temperature treatments. We found significant differences in the structure of G between the treatments: heritabilities and evolvabilities were higher in the warm than in the low‐temperature treatment, suggesting more and faster opportunity to evolve in warm (stressful) conditions. Furthermore, comparison of G and its phenotypic equivalent P revealed the latter is a poor substitute of the former. Most strikingly, the results suggest that the expected impact of G on evolvability—as well as the similarity among G ‐matrices—may depend strongly on the number of traits included into analyses. In our results, the inclusion of only few traits in the analyses leads to underestimation in the differences between the G ‐matrices and their predicted impacts on evolution. While the results highlight the challenges involved in estimating G , they also illustrate that by enabling the estimation of large G ‐matrices, the BSFG method can improve predicted evolutionary responses to selection.