Taxonomic boundaries and systematic relationships in the grass subspecific complex Dichanthelium acuminatum were investigated with both morphological and molecular methods. Circumscription of subspecific taxa comprising the complex has been difficult due to a continuum of morphological character variation among taxa and possibly due to infraspecific and interspecific hybridization. Qualitative and quantitative morphological character data was collected from herbarium specimens and field-collected specimens and analyzed using multivariate statistical techniques. Representative specimens were selected for molecular phylogenetic analysis of DNA sequences from the GBSSI (waxy) nuclear gene. Subspecific boundaries as circumscribed in the most recent taxonomic treatment (10 subspecies) were tested from: 1) a morphological perspective with results of the multivariate statistical analysis to determine if the study specimens formed natural groupings that corresponded to the recent treatment; and, 2) with molecular phylogenetic analysis to estimate the evolutionarily significant lineages present and to determine if such lineages supported the natural groupings revealed from the multivariate morphological analysis. A separate investigation was conducted using a molecular technique to screen for putative hybrid specimens from DNA obtained from field-collected specimens. Multivariate statistical analysis of the morphological data provided support for four of the 10 taxa tested and additional support for two taxa considered as a single unit. Further research is needed to determine the appropriate status of the remaining six taxa of the ten taxa tested. Molecular phylogenetic analysis provided support for recognizing four evolutionarily significant units and provided parallel support for four of the five taxa recognized from the morphological analysis. The hybridization investigation identified two putative hybrid specimens, which were confirmed as hybrids with GBSSI sequence data and also with multivariate statistical analysis of morphological data to provide provisional evidence for the role of hybridization in producing specimens with intermediate morphological phenotypes. A taxonomic treatment and dichotomous key was produced for the 10 subspecific taxa of the Dichanthelium acuminatum complex.

Chenopodium L. is a relatively under-studied genus that includes the cultivated seed crop quinoa (Chenopodium quinoa Willd.). Quinoa is an allotetraploid (2n=4x=36, AABB genomes) that is cultivated by subsistence farmers and commercial growers in the Andean regions of South America. Approximately 60% of a quinoa seed is starch, a glucose polymer that is an important carbohydrate energy source in the human diet. Seed starch is normally comprised of amylose and amylopectin in a 1:3 ratio, but starches with different amylose:amylopectin ratios have different properties and potential uses. The accumulation of the amylose fraction of starch is controlled by a single dominant gene in quinoa, GBSSI. We report the sequencing and characterization of the GBSSI gene in 18 accessions of Chenopodium, including Andean quinoa and the related Mesoamerican chenopod grain species, C. berlandieri subsp. nuttalliae Saff. Two distinct homoeologs (GBSSIa and GBSSIb) were identified in the tetraploid accessions, and 19 different alleles were identified, including three null mutants – one in an accession of quinoa and two in a waxy landrace of C. berlandieri subsp. nuttalliae, referred to as ‘H02’. Expression analysis of the null mutants revealed that GBSSIa and GBSSIb were both strongly expressed late in quinoa and C. berlandieri subsp. nuttalliae seed development. Starch phenotyping indicated that ‘H02’ produced no amylose, likely due to its having two mutated copies of GBSSI. GBSSI sequences were used to analyze the phylogenetic relationships between quinoa and other members of the Chenopodium genus. This study and the discovery of Chenopodium GBSSI null-mutants will assist in the development of new Chenopodium crops with novel starches.

Eight of fifteen species in genus Anthoxanthum (Poaceae) can be found in Europe. Five of them are perennials forming A. odoratum complex, the remaining three are annual, more or less mediterranean taxa (A. aristatum, A. ovatum and A. gracile). Within the A. odoratum s. l. complex the following taxa are distinguished: widely spread A. odoratum s. str. (4x; 2n = 20), arcto-alpine A. alpinum (2x a 4x; 2n = 10 and 20), Madeiran endemic species A. maderense (2x; 2n = 10), endemic species of Balkan mountains A. pauciflorum (2x; 2n = 10) and the Iberian peninsula endemic A. amarum (?x; 2n ~ 90). The aim of this thesis is to clearify the unknown evolutionary relationships between the taxa, between the annuals and perennials, diploids and polyploids. The following questions should be answered in this study: 1) What is the origin and distribution of the rediscovered diploid perennial taxon and what is its relationship to the other members of the group; 2) What is the distribution pattern of the perennial taxa of the genus Anthoxanthum in Europe and what is their haplotype differenciation (overall distribution of the taxa and haplotypes and the existence of their sympatric occurence); 3) What evolutionary ties exist among the species and what is the origin of allotetraploid taxon A. odoratum s.str. The...

The Malacothamnus alliance consists of three genera, Iliamna, Malacothamnus, and Phymosia. The genera are considered taxonomically complex since hybridization freely occurs, polyploidy levels vary, and there is a lack of distinct morphological characters to delineate taxa. Several taxonomic treatments have been prepared for each genus, but relationships within the genera and the relationship of the Malacothamnus alliance to others in the Malvaceae remains unknown. This multifaceted study aimed to (a) examine the monophyly of the Malacothamnus alliance and its position in the Malvaceae, (b) determine the relationships between genera in the alliance, (c) compare variation of nuclear and chloroplast genes in the alliance, (d) prepare revised taxonomic treatments for Iliamna and Malacothamnus, and (e) examine the probability of successful hybridization in Iliamna. The monophyly of the Malacothamnus alliance was not confirmed using DNA sequences of both nuclear and chloroplast regions. In Iliamna, little sequence variation was detected among taxa in the Rocky Mountains; however, the nuclear and chloroplast regions conflicted with regard to the relationships of the western and eastern taxa. An ancestral copy of the chloroplast genome is shared between the two eastern U.S. Iliamna species and Phymosia (Bahamas and Mexico). The nuclear ITS sequences indicated the western U.S. Iliamna species were more closely related to Phymosia and Malacothamnus than to other species in Iliamna. Neither data set provided sufficient variation to resolve relationships of species in Malacothamnus. Genetic variation and the feasibility of hybridization in Iliamna supported the results of the broader phylogenetic studies. Iliamna corei and I. remota are recently derived from I. rivularis. Hybrid offspring of I. corei and I. remota had higher viability and fecundity than did hybrids between crosses of either species and I. rivularis. The Virginia populations of I. corei and I. remota are more genetically similar than either is to Illinois populations of I. remota. However, the species are morphologically distinct and can easily be distinguished from others in the genus. Revised taxonomic treatments for Iliamna and Malacothamnus based on surveys of herbarium material are presented. Taxonomic revisions include the new combinations of Iliamna grandiflora subsp. grandiflora and I. grandiflora supsp. crandallii and the resurrection of Malacothamnus hallii and M. orbiculatus. / Ph. D.