Systematic Biology

Vogl, C.; Wagner, G. P.

doi: 10.2307/2992449pmid: N/A

AbstractComputer simulations of randomly evolving lineages were performed using the model of Raup and Gould (1974). The mean, variance, and distribution of quantitative traits were evaluated. These results were compared with the predictions of diffusion theory (Slatkin, 1981). It is a common feature of all simulations that the predictions fit better in rapidly expanding clades than in slowly expanding clades. Hence, high probabilities of extinction compared to the probability of speciation cause a high degree of stochasticity in character dispersion. We therefore recommend restricting the statistical analysis of interspecific variation to recent and rapidly evolving clades. The results also show that the within clade mean has exceedingly high interclade variance, which calls into question the significance of a direct comparison of within clade means. In contrast the distribution of variances and variance ratios of two traits is more predictable and these statistical parameters might be useful in comparative biology: variances increase linearly with time according to the diffusion approximation, variance ratios of two independent traits follow approximately an F- (variance-ratio) distribution, and the distribution of the variance ratio is independent of speciation rate and the age of the clade. On the basis of these results it is possible to design two kinds of statistical tests: 1) Comparing the variances of a set of homologous traits in two monophyletic clades allows one to test whether the traits show cladespecific evolutionary flexibilities. 2) The ratio of variances of two different traits within the same clade provides a measure of the relative evolutionary flexibility of two traits which is independent of the exact phylogeny and the age of the clade. To illustrate the application of these tests the interspecific variation of seven meristic characters from three clades of blennioid fishes is analysed.

Bailey, Robert C.; Byrnes, Janice

doi: 10.2307/2992450pmid: N/A

AbstractA new approach to assessing percent measurement error (%ME), using an old statistical technique (Model II Analyses of Variance and Covariance: ANOVA/ANCOVA), was developed using both weight and linear measurements made three times each on 87 freshwater snails (Cipangopaludina chinensis; Vivipariidae). Variability of each measurement, as well as covariability between pairs of measurements, was partitioned into among- and within-snail (i.e., measurement error) components. The %ME varied by two orders of magnitude across the ten measurement variables considered. Shell weight had the lowest %ME (0.059%), while body whorl height had the highest (3.88%). There was a low correlation between %ME and the among-snail coefficient of variation for each variable (r = −0.28; P > 0.20). Within-snail correlations between pairs of measurement variables were uniformly low (rwithin ≤ 0.20), while among-snail correlations were generally very high (r = 0.74 to 0.99) due to the large size range of snails in the sample. In any morphometric study, if %ME is found to be high for a particular variable, it should either be deleted from the study or remeasured a number of times on every individual to be included in the dataset.

Porter, Adam H.

doi: 10.2307/2992451pmid: N/A

AbstractGene flow estimates between groups of populations (NmGT) derived from Wright's hierarchical F-statistics can be used to test taxonomic hypotheses about the location of biological species boundaries. The interpretation of NmGT is limited by the assumptions of the island model: balancing selection, time to equilibrium, and the limits to detectability of alleles are biases acting in the same direction as gene flow to cause similarity among populations. However, confidence limits of 0.01 ≤ NmGT ≤ 0.15 between sympatric sibling species suggest that these biases are weak relative to gene flow in causing genetic similarity. NmGT also compares favorably to gene flow estimated from a mark-recapture study, and NmGT values across a well-studied grasshopper hybrid zone suggest that gene flow is limited but not stopped by the barrier.The nominal species Limenitis lorquini and L. weidemeyerii hybridize along their contact areas in western North America. Their hybrid zone was mapped using wing pattern characters, and the same populations were analyzed electrophoretically at 18 enzyme loci. Gene flow is strong within these taxa (1.56 ≤ Nm ≤ 15.23). Across the hybrid zone, 0.57 ≤ NmGT ≤ 1.97, suggesting that although there is a partial barrier to genetic exchange, these taxa share significant portions of their gene pools. Small population sizes and other circumstantial evidence argue that much of the similarity observed among populations is actually due to gene flow, rather than balancing selection, although undetected allelic differences may also be partially responsible. L. weidemeyerii and its “subspecies” are best considered as univoltine subspecies of lorquini, and all should be treated as a single genetic unit when cladistic analyses are performed. The ease with which these results were obtained suggests that systematists should pay more attention to the genetic population structure of taxonomic boundaries.

Zink, Robert M.; Avise, John C.

doi: 10.2307/2992452pmid: N/A

AbstractAnalyses of mitochondrial DNA (mtDNA) and allozymes were used to estimate phylogenetic patterns in the avian genus Ammodramus. Levels of interspecific genetic differentiation were greater than most previous estimates for other congeneric avian taxa. Phenetic and phylogenetic patterns were highly concordant for these two genetically independent data sets, suggesting a robust estimate of the evolutionary history of these sparrows. However, the genetic pattern was not concordant with an estimate of variation in skeletal morphometrics produced by Robins and Schnell (1971); we suggest that ecological pressures effect convergence in skeletal morphology. Independent calibrations of mtDNA and allozyme distances suggest times of divergence that differ by a factor of two among the species assayed.

Downes, Barbara J.

doi: 10.2307/2992453pmid: N/A

AbstractCoevolution of parasites and their hosts is presumed if the phylogeny of a group of parasites matches that of their hosts, producing a high value of a consistency index. Such comparisons are done using relatively host specific parasite species that are usually separated only by morphological characters. Parasitic, unionicolan mites in the subgenus Unionicolides are described, from sampling data, as highly host specific and are thought to have co-speciated with their freshwater mussel hosts. However, experimental observations of Unionicola Unionicolides poundsi and U. U. lasellai were inconsistent with these mites' supposed high host specificity. A transplant experiment, in which nymphs metamorphosed in the “wrong” host, demonstrated that the morphological characters used to differentiate these mite species are plastic and hostdetermined. Hence, when taxonomists use only morphological characters and host-induced morphological variation is present, morphological variants can be elevated to the status of species. Consequently, parasites can appear to be highly host specific. If morphs are then grouped, as species, into the same taxon, phylogenetic comparisons can produce high consistency indices independently of any coevolutionary processes. The conditions under which this bias can occur are commonly true of phylogenetic comparisons. Hence, host-induced morphology is a viable, alternative explanation for high consistency indices.

Archie, James W.

doi: 10.2307/2992454pmid: N/A

Maddison, Wayne; Slatkin, Montgomery

doi: 10.2307/2992455pmid: N/A

Whitehead, Peter

doi: 10.2307/2992456pmid: N/A

doi: 10.1093/syszoo/39.2.185pmid: N/A