Showing papers by "Andy Purvis published in 1995"

Comparative analysis by independent contrasts (CAIC): an Apple Macintosh application for analysing comparative data.

Abstract: CAIC is an application for the Apple Macintosh which allows the valid analysis of comparative (multi-species) data sets that include continuous variables. Comparison among species is the most common technique for testing hypotheses of how organisms are adapted to their environments, but standard statistical tests like regression should not be used with species data. Such tests assume independence of data points, but related species often share traits by common descent rather than through independent adaptation. CAIC uses a phylogeny of the species in the data set to partition the variance among species into independent comparisons (technically, linear contrasts), each comparison being made at a different node in the phylogeny. There are two partitioning procedures--one used when all variables are continuous, the other when one variable is discrete. The resulting comparisons can be analysed validly in standard statistical packages to test hypotheses about correlated evolution among traits, to estimate parameters such as allometric exponents, and to compare rates of evolution. Previous versions of the package have already been used widely; this version is simpler to use and works on a wider range of machines. The package and manual are freely available by anonymous ftp or from the authors.

A composite estimate of primate phylogeny

Abstract: This paper presents an estimate of the phylogeny of all 203 species of primate. The composite tree is derived by applying a parsimony algorithm to over a hundred previous estimates, and is well resolved, containing 160 nodes. The ages of over half the clades in the tree have been estimated from information in the literature. Bootstrapping has been used to indicate the degree of certainty associated with each clade. The tree will be a useful framework for comparative biologists and shows which areas of primate phylogeny are still only sketchily known.

Evolutionary radiation of visual and olfactory brain systems in primates, bats and insectivores

Abstract: How brains have evolved in response to particular selection pressures is illuminated by ecological correlates of differences in brain structure among contemporary species. The focus of most comparative studies has been on the overall size of brains relative to body size, hence ignoring the ways in which selection operates on specific neural systems. Here we investigate evolutionary radiations in the size of visual and olfactory brain structures within three orders of mammals: primates, bats and insectivores. The comparative relationships within these three orders show both similarities and differences. After removal of the allometric effect of overall brain size, the sizes of different structures within each sensory modality are positively correlated in all three orders. Correlations between visual and olfactory structures, however, are negative in primates, negative but non-significant in insectivores, and positive in bats. In both primates and insectivores, nocturnal lineages tend to have larger olfactory structures than do diurnal or partly diurnal lineages, and among the primates diurnal lineages have larger striate visual cortexes. Hence the apparent trade-off between vision and olfaction in primates seems to be related to the divergence of nocturnal and diurnal forms. However, negative correlations between visual and olfactory structures were also found when nocturnal strepsirhines and diurnal haplorhines were analysed separately, suggesting that ecological variables in addition to activity timing may be significant. Indeed, there were also associations with diet: frugivory was associated with enlargements of the geniculostriate visual system in diurnal primates, enlargements of olfactory structures in nocturnal primates, and possibly enlargements of both in bats. Further ecological associations were found within insectivores: aquatic lineages had smaller olfactory structures than in their non-aquatic counterparts, and fossorial lineages had smaller optic nerves than in non-fossorial forms. We conclude that activity timing, diet and habitat have each played a role in the evolutionary radiation of mammalian sensory systems, but with varying effects in the different taxa. Some of the associations between ecology and sensory systems suggest alternative explanations for correlates of overall brain size, which have in the past commonly been interpreted in terms of selection on intelligence.

Sperm competition: mating system, not breeding season, affects testes size of primates

Abstract: In taxa in which several males mate with a fertile female, males have larger testes relative to their body size than do males of taxa where only one male normally mates. However, breeding seasonality could confound the association: some species with breeding seasons have unusually large testes; seasonal breeding could concentrate fertile females in time, so making them difficult to defend and thus promoting multi-male groups. Using data from 58 species of primates, and carefully accounting for phylogeny, we investigate the potential for breeding seasonality to confound the relation between testes size and mating system. No confounding effect exists. Multi-male taxa have very significantly larger testes for their body size than do single-male taxa, independently of seasonality of breeding. Seasonality has no effect whatsoever in our sample, although few primates have very short breeding seasons. While multi-male genera are equally likely to be seasonal as non-seasonal breeders, the single-male taxa of our sample are rarely seasonal, apparently supporting the difficulty of defence hypothesis but the sample by no means allows a firm conclusion.

Macroevolutionary Inferences from Primate Phylogeny

Abstract: We apply new statistical methods to a recent estimate of the phylogeny of all living primate species to test a range of models of cladogenesis. Null models in which probabilities of speciation and extinction do not differ among contemporaneous lineages are not consistent with the phylogeny. We present evidence that the net rate of cladogenesis (speciation rate minus extinction rate) increased in the lineage leading to the Cercopithecidae (Old World monkeys), and that there have been further increases in some lineages within that family. Such increases suggest the occurrence of clade selection, although we have not identified the selected trait or traits. There is no evidence that the net rate of cladogenesis is a function either of how many primate lineages are already present or of time. Intriguingly, three other major clades-Strepsirhini, Platyrrhini and Hominoidea-appear to have had very similar rates of clade growth, in spite of their great biological differences.

Mammal life-history evolution: a comparative test of Charnov's model

Abstract: We present a comparative test of Charnov's recent theoretical model of mammalian life-history evolution. Phylogenetic analysis of life-table data from 64 species, ranging across nine orders. supports all of Charnov's assumptions and most of his predictions. The allometries of time from independence to maturity (a), annual fecundity, and adult and juvenile mortality rates are in agreement with previous work and with the theory, as are the signs of the relationships among these traits when body size is controlled for. As predicted, the non-dimensional products of a and each of the other three traits are independent of adult body size, as is survivorship to maturity. However, we find that the ratio of weaning weight to adult weight (δ) is correlated with adult weight, in contradiction with the theory, and we do not find the predicted relationships between δ and the three non-dimensional products. The discrepancies could be because we have equated independence with weaning, or because the model assumes determinate growth: they could arise if large mammals have relatively longer periods of post-weaning care, or continue to grow after starting to reproduce. There is some evidence that δ is influenced by the nature of mortality around independence (density-dependent or density-independent), and we suggest this as a possible area for further work. In general, the areas of agreement between Charnov's theory and the data are more impressive than the differences, indicating that it could be a major breakthrough in understanding the evolution of life histories in placental mammals.

Phylogenetic Noise Leads to Unbalanced Cladistic Tree Reconstructions

Abstract: ?Cladistic tree balance is the extent to which internal nodes on a cladistic tree define clades of equal size. More robust maximum-parsimony trees taken from the literature are more balanced. Simulation studies suggest that a methodological bias is responsible for this correlation because incorrect reconstructions are also likely to be less balanced than the true trees they esti? mate. Misinformative cladistic characters can be expected to make trees more unbalanced if there is marked variation in their rates of change. This bias may contribute to the excess of unbalanced phylogenies reported in the literature. [Cladistic tree balance; tree robustness; retention index; homoplasy; Markov models.] Guyer and Slowinski (1991) presented evidence that a subset of well-corroborated phylogenies having five tips was more bal? anced than a larger sample of published trees, where tree balance denotes the de? gree to which branch points define sub? groups of equal size (Heard, 1992). There are three types of explanation for the ob? served difference in the balance between well and poorly corroborated trees. First, poorer phylogenetic data may lead to more unbalanced reconstructions. Such an inter? action would affect the work of those sys? tematists and evolutionary biologists who assume that the shapes of trees are unbi? ased. A second possibility is that the to? pology of the true tree influences the qual? ity of the phylogenetic data measured at the tips. More unbalanced true trees may produce more phylogenetic noise; thus, true unbalanced trees might be more dif? ficult to reconstruct correctly. Finally, there may be some third factor that influences both tree balance and data quality inde? pendently (see Arnold, 1990, for an ex? ample). The empirical study of potential bias is difficult because true and false trees can almost never be distinguished in prac? tice (see Hillis et al., 1992, for a rare excep? tion). Without the true tree one cannot quantify how incorrect a given reconstruc? tion is or whether it is more or less bal1 Present address: Department of Zoology, Univer? sity of British Columbia, Vancouver, British Columbia V6T 1Z4, Canada. E-mail: mooers@bcu.ubc.ca. anc than the true tree. However, if Guyer and Slowinski's well-corroborated trees are closer to the true tree, the prediction is that unbalanced reconstructed trees are more likely to be (more) incorrect than are bal? anced reconstructed trees. Here, we inves? tigate the relationship between data quali? ty and tree shape using published trees and simple computer simulation.