Putative Taxa Research Articles

A genetic determination of cryptic species (Copepoda: Calanoida) and their postglacial biogeography in North America

Allozyme electrophoresis of Hesperodiaptomus arcticus (Marsh) sensu lato at subarctic Churchill, Canada revealed significant deviations from random mating in several populations, as shown by the complete absence of certain heterozygous types. A broad geographic survey of 20 enzyme loci in 1077 individuals from 27 ponds at 11 arctic and alpine sites west of Hudson Bay demonstrated that there were three genetically distinct types. Morphological analysis of genetically classified populations uncovered consistent morphological differences that had been ignored previously. These morphological characteristics were used to reclassify preserved museum and private specimens from additional sites from the entire North American distribution of H. arcticus sensu lato. A detailed study of two of the distinct types co-occurring at Churchill showed reproduction timing and habitat differences consistent with reproductive isolation. I concluded that there are three species and each have different distributions. Most regions examined must have been repopulated by dispersal from glacial refugia after deglaciation. The nominal species is apparently restricted to western arctic localities in the vicinity of the Alaskan refugium while the other two putative taxa have dispersed widely into previously glaciated regions.

Systematic and evolutionary inferences from isozyme studies in the genus Eremaea (Myrtaceae)

Eremaea is a genus of woody shrubs endemic to the south-west of Australia. With the exception of E. pauciflora all species are restricted to the kwongan or sandplain regions of the coastal plain northwards from Perth. Forty four populations covering the seven recognised species and a further 12 putative taxa were examined for allozyme variation at 15 polymorphic loci. Gene diversity statistics indicated that a greater proportion of genetic variability in Eremaea species was due to within- rather than among-population differences. The allozyme data supported morphological studies which indicated five species groups or complexes. Differentiation among and within species complexes was examined using gene frequency and genetic distance data. The average genetic distance (Nei's D) among populations within species ranged from 0.011 (E. acutifolia) to 0.051 (E. aff. brevifolia 3). Within species complexes the average genetic distance between taxa ranged from 0.050 to 0.157 while between complexes it ranged from 0.164 to 0.558. Of the 159 possible pairwise combinations between the 19 morphologically distinct taxa, 15 showed little allozyme divergence. Lack of allozyme divergence was attributed to either rapid and recent speciation or to introgressive hybridisation. Based on allozyme data E. aff. brevifolia x violacea is either of hybrid origin or is a recently derived hybrid, and field observations indicated that interspecific hybridisation occurs within Eremaea. Rapid and recent speciation, combined with hybridisation have resulted in a reticulate pattern of evolution in sections of this genus. Phylogenetic analyses based on these allozyme data were generally consistent with those based on morphological data except for the placement of E. purpurea and E. aff. pauciflora 4.

Phenotypic Plasticity and the Selection of Taxonomic Characters in Puccinellia (Poaceae)

Phenotypic plasticity diminishes a character's taxonomic utility because the environmental contribution to phenotypic variation obscures genetically based relationships. To evaluate potential taxonomic characters of Puccinellia, divided clones representing several putative taxa were grown under a series of saltand drought-stress treatments representing naturally encountered environmental conditions. The resulting variation in 41 characters was partitioned by analysis of variance (ANOVA) into genetic and environmental components. Logarithmic transformation facilitates comparisons of variance levels among characters. Analogous nonparametric tests were used to analyze 40 characters not amenable to ANOVA. Levels of variance among genotypes and among treatments (phenotypic plasticity) continuous in distribution and highly variable in magnitude. Attributes of the spikelet dominate the list of characters best meeting the criteria of high genotypic variability and low plasticity. Ratios representing shapes and relative sizes tend to be more reliable than their component simple measurements. This analysis empirically confirms the importance of many characters emphasized in grass systematics and suggests characters for Puccinellia that previously have not been recognized as significant. While much recent effort has been directed at the theory and methodology of taxonomic and phylogenetic analysis, the choice of characters to be used in such analyses has received less attention. Yet the suite of characters underlying any systematic study profoundly influences the outcome. Different character sets derived from the same plants, and analyzed identically, may support different conclusions (Denton 1978; Hauser and Crovello 1982). The true relationships of natural groups never known with certainty. Faced with conflicting evidence, systematists must decide which characters reliably reflect past evolutionary events. This decision process is weighting. Two broad strategies of character weighting can be discerned, generalized correlation and a priori weighting (Cain 1959). Correlational methods weight individual characters by their tendencies to agree with others; weighting proceeds simultaneously with taxonomic analysis. These methods range from informal comparison of vxariation patterns to strictly defined relational criteria such as parsimony and character compatibility (cf. Felsenstein 1981). Systematic schemes based upon correlational methods often considered unweighted, because results determined by the aggregate number of equally weighted characters supporting a particular scheme. However, characters identified as reliable in one analysis often accorded greater weight in related taxa whose character combinations have not been analyzed. In contrast with correlational weighting, a priori methods assign weight to characters by their degrees of conformance to intrinsic criteria of good and bad characters. Some a priori criteria refer to general classes of characters, e.g., high weight to structurally complex organs (Mayr 1969), low weight to monogenic characters (Rollins 1958). Other a priori methods involve expectations of variation patterns to be exhibited by good characters. For example, Farris (1966) argued that good characters will vary little within biological populations. Davis and Heywood (1963, p. 119) stated the a priori criterion that good characters are not easily susceptible to environmental modification. In a general sense this principle is widely recognized. Leaf shape in Sagittaria, varying with the depth of water in which a plant grows, is well known to be taxonomically unreliable because it is environmentally modified. In less obvious cases, the environmental contribution to variation has been investigated for small numbers of characters (Jones and Newton 1970; Chapman 1975). In this paper, the environmental component of variation is analyzed for numerous characters of Puccinellia Parl. (Poaceae), in order to identify stable characters of potential taxonomic value.