Thesis (M.S.)--University of Wyoming, 2007. / Title from PDF title page (viewed on Nov. 13, 2008). Includes bibliographical references (p. 173-183).

Thesis (Ph. D.)--Oregon State University, 2009. / Printout. Includes bibliographical references (leaves 176-194). Also available on the World Wide Web.

Ecomorphological specialization within Catostomidae in several large western North American lakes has produced populations including typical benthic suckers (Catostomus) and lakesuckers (Chasmistes), mid-water planktivores, with a continuum of morphologies existing between them. All extant lakesuckers are endangered, and population declines have been attributed in part to hybridization with sympatric Catostomus spp. Chapter 2 describes assessment for concordance of morphological and genetic variation in suckers in Utah Lake, Utah (June sucker, Chasmistes liorus; Utah sucker, Catostomus ardens; and suckers of intermediate morphology) by comparing a morphological analysis with amplified fragment length polymorphism and microsatellite analyses. Suckers were differentiated using characters associated with presumed feeding strategies: zooplanktivory (June sucker) and benthivory (Utah sucker). No molecular evidence was found for deep genetic divergence between morphs or for hybridization among ancient lineages. Slight population structuring accompanied substantial morphological variation. Chapter 3 describes the investigation of distribution and movement, spawning behavior, and diet of suckers in Utah Lake and their growth at different densities in a laboratory experiment. Acoustic / radio telemetry revealed little difference in movement and distribution of June sucker and Utah sucker or in timing of spawning runs. Stable isotopes analysis revealed that Utah sucker were enriched in 13C relative to June sucker as presumed diets would predict. Intermediate morphs were intermediate for δ13C and δ15N. Neither species nor density was a significant predictor of growth rate of June sucker or Utah sucker reared at different conspecific densities. Chapter 4 examines morphology, genetics, and diet of the sucker population inJackson Lake, Wyoming, once home of the extinct Snake River sucker, Chasmistes muriei, a lakesucker known from a single specimen. Currently, suckers in Jackson Lake are identified as Utah sucker; however, recently sampled individuals resemble lakesucker. No molecular evidence was found for deep genetic divergence between lakesucker and benthic morphs or for hybridization among ancient lineages. The benthic morph was significantly enriched in 13C relative to the lakesucker morph, consistent with presumed diets. Morphologically, the lone Snake River sucker holotype specimen grouped strongly with extant lakesucker morphs, suggesting that the status of the Snake River sucker be updated accordingly.

Pantosteus, a subgenus of Catostomus, includes the mountain sucker (Catostomus playthyrnchus), whose speculated older origins in the Miocene/Pliocene can provide insight into the ancient geographical events of western North America. We believe that major geologic events influencing the diversification of mountain suckers include the rise of the Colorado Plateau, the connections between the ancient Snake River system and the Lahontan system and subsequently the connection of the Snake River system to the Columbia Basin, dispersal of mountain suckers across the continental divide, as well as the Pleistocene Bonneville flood. If this is true, we should see evidence of geologic separation and timing through studying the phylogenetics of the mountain sucker. In order to clarify relationships of the mountain sucker with respect to other Pantosteus species, we examined cytochrome b (cyt b) sequences for 144 mountain suckers, 24 other Pantosteus species, and ten outgroup species. Phylogenetic relationships among haplotypes were constructed based on maximum likelihood and Bayesian criterion. In an effort to provide better resolution at some nodes, we also sequenced additional mitochondrial genes (ND1, ND2, ATPase, ND4L, ND4, ND5, ND6, and cyt b) for a subset of 44 individuals taken from the major clades obtained from the cyt b phylogentic analyses. Trees from this data set were also constructed under maximum likelihood and Bayesian criterion. All phylogenetic analysis revealed that mountain sucker are paraphyletic, with two major clades of mountain suckers separated by other members of the subgenus Pantosteus. One clade included two sub-clades, one from the upper Snake River drainage/northern Bonneville/Green River drainage Basins and the other from the southern Bonneville Basin. The other major clade included sub-clades from the Lahontan Basin, Columbia River Basin, and Upper Missouri River Basin. Molecular clock analysis revealed that Pantosteus likely split from Catostomus during the Miocene and that major speciation events within Pantosteus occurred during the Pliocene and Pleistocene. Genetic structuring and gene flow estimates for mountain sucker populations, with groupings based on major drainage basins, were calculated with AMOVA and Fst estimates in Arlequin and revealed that most of the genetic structuring was explained by variation among drainage basins with limited gene flow occurring between drainage basins. Based on this study, the role of the Colorado Plateau's geologic history in the evolution of the mountain sucker remains unclear. However, all other geologic events as discussed in this study seem to have played a significant role in the evolution of the mountain sucker.

Several hypotheses from community ecology were tested using habitat and diet patterns for six catostomid fishes of the Assiniboine River, Manitoba. Specifically, I examined expected assemblage patterns based on the equilibrium-nonequilibrium continuum that are based on competition as a structuring mechanism. The catostomid assemblage showed characteristics of both equilibrium and nonequilibrium assemblages. Habitat utilization was not proportional to habitat availability for depth, velocity, and substrate indicating habitat selection occurs. These patterns are influenced by the distributions of soft Lake Agassiz deposits and harder glacial till-plain habitats and therefore localized. Species co-occurred in relation to abundance indicating negative associations assumed in competition-based theory were not apparent. In addition, species from the same subfamily co-occurred more often than expected by chance, indicating positive within-group comparisons, except for silver redhorse, which did not co-occur differently than random. Species from the same feeding group co-occur most frequently with conspecifics, indicating species have specific habitat patterns. Benthic invertebrate distributions and fish distributions were positively correlated with fish diet. Within-feeding-group comparisons indicated species that co-occurred frequently consumed the same food items, but showed subtle differences in abundance of diet items. The most common diet items varied among white sucker, silver redhorse, golden redhorse, and shorthead redhorse, indicating that, although these species co-occur, subtle differences in feeding behavior may account for differences in relative abundance and frequency of diet items. Quillback and bigmouth buffalo shared a few core food items, but co-occurred infrequently. Inter-specific interactions showed silver redhorse diets were a subset of white sucker diets, white sucker diets were a subset of golden redhorse diets, shorthead redhorse diets were a subset of silver redhorse and golden redhorse diets, while bigmouth buffalo diets were a subset of quillback diets. Lower richness and prevalence of organisms in quillback and bigmouth buffalo diets were due to highly aggregated prey items. Benthic invertebrate distributions were aggregated and dependent on the predictable patterns of substrates, based on hydraulic sorting in the meandering along the river. Large-scale redundancy of species-habitat associations was correlated with historical (phylogenetic) or adaptive (morphological) constraints on habitat selection. Stream habitats are constrained by regional factors of slope and sediments, but also determined by stable, repetitive and predictable local processes of erosion, transport, and deposition (meandering). The fish-habitat associations of catostomids in the Assiniboine River provide an example of interaction between the abundance of the component species, the phylogenetic constraints on the niche, and the deterministic nature of the spatial distribution of habitats. Within-feeding-group pairs showed that local habitat overlap is positively correlated with geographic overlap, while local diet overlap is negatively correlated with geographic overlap. Two species pairs (silver redhorse and golden redhorse and bigmouth buffalo and quillback) were concluded to have symmetric overlaps at the geographic scale and stable interactions. Using the functional niche concept ecological theory provides a link between ecology and biogeography of sympatric species. The multi-analytical approaches in this study provide insights into the structuring of north temperate prairie river fish communities, through hypothesis testing and correlation that have application beyond Prairie Rivers. / October 2005

Several hypotheses from community ecology were tested using habitat and diet patterns for six catostomid fishes of the Assiniboine River, Manitoba. Specifically, I examined expected assemblage patterns based on the equilibrium-nonequilibrium continuum that are based on competition as a structuring mechanism. The catostomid assemblage showed characteristics of both equilibrium and nonequilibrium assemblages. Habitat utilization was not proportional to habitat availability for depth, velocity, and substrate indicating habitat selection occurs. These patterns are influenced by the distributions of soft Lake Agassiz deposits and harder glacial till-plain habitats and therefore localized. Species co-occurred in relation to abundance indicating negative associations assumed in competition-based theory were not apparent. In addition, species from the same subfamily co-occurred more often than expected by chance, indicating positive within-group comparisons, except for silver redhorse, which did not co-occur differently than random. Species from the same feeding group co-occur most frequently with conspecifics, indicating species have specific habitat patterns. Benthic invertebrate distributions and fish distributions were positively correlated with fish diet. Within-feeding-group comparisons indicated species that co-occurred frequently consumed the same food items, but showed subtle differences in abundance of diet items. The most common diet items varied among white sucker, silver redhorse, golden redhorse, and shorthead redhorse, indicating that, although these species co-occur, subtle differences in feeding behavior may account for differences in relative abundance and frequency of diet items. Quillback and bigmouth buffalo shared a few core food items, but co-occurred infrequently. Inter-specific interactions showed silver redhorse diets were a subset of white sucker diets, white sucker diets were a subset of golden redhorse diets, shorthead redhorse diets were a subset of silver redhorse and golden redhorse diets, while bigmouth buffalo diets were a subset of quillback diets. Lower richness and prevalence of organisms in quillback and bigmouth buffalo diets were due to highly aggregated prey items. Benthic invertebrate distributions were aggregated and dependent on the predictable patterns of substrates, based on hydraulic sorting in the meandering along the river. Large-scale redundancy of species-habitat associations was correlated with historical (phylogenetic) or adaptive (morphological) constraints on habitat selection. Stream habitats are constrained by regional factors of slope and sediments, but also determined by stable, repetitive and predictable local processes of erosion, transport, and deposition (meandering). The fish-habitat associations of catostomids in the Assiniboine River provide an example of interaction between the abundance of the component species, the phylogenetic constraints on the niche, and the deterministic nature of the spatial distribution of habitats. Within-feeding-group pairs showed that local habitat overlap is positively correlated with geographic overlap, while local diet overlap is negatively correlated with geographic overlap. Two species pairs (silver redhorse and golden redhorse and bigmouth buffalo and quillback) were concluded to have symmetric overlaps at the geographic scale and stable interactions. Using the functional niche concept ecological theory provides a link between ecology and biogeography of sympatric species. The multi-analytical approaches in this study provide insights into the structuring of north temperate prairie river fish communities, through hypothesis testing and correlation that have application beyond Prairie Rivers.

Several hypotheses from community ecology were tested using habitat and diet patterns for six catostomid fishes of the Assiniboine River, Manitoba. Specifically, I examined expected assemblage patterns based on the equilibrium-nonequilibrium continuum that are based on competition as a structuring mechanism. The catostomid assemblage showed characteristics of both equilibrium and nonequilibrium assemblages. Habitat utilization was not proportional to habitat availability for depth, velocity, and substrate indicating habitat selection occurs. These patterns are influenced by the distributions of soft Lake Agassiz deposits and harder glacial till-plain habitats and therefore localized. Species co-occurred in relation to abundance indicating negative associations assumed in competition-based theory were not apparent. In addition, species from the same subfamily co-occurred more often than expected by chance, indicating positive within-group comparisons, except for silver redhorse, which did not co-occur differently than random. Species from the same feeding group co-occur most frequently with conspecifics, indicating species have specific habitat patterns. Benthic invertebrate distributions and fish distributions were positively correlated with fish diet. Within-feeding-group comparisons indicated species that co-occurred frequently consumed the same food items, but showed subtle differences in abundance of diet items. The most common diet items varied among white sucker, silver redhorse, golden redhorse, and shorthead redhorse, indicating that, although these species co-occur, subtle differences in feeding behavior may account for differences in relative abundance and frequency of diet items. Quillback and bigmouth buffalo shared a few core food items, but co-occurred infrequently. Inter-specific interactions showed silver redhorse diets were a subset of white sucker diets, white sucker diets were a subset of golden redhorse diets, shorthead redhorse diets were a subset of silver redhorse and golden redhorse diets, while bigmouth buffalo diets were a subset of quillback diets. Lower richness and prevalence of organisms in quillback and bigmouth buffalo diets were due to highly aggregated prey items. Benthic invertebrate distributions were aggregated and dependent on the predictable patterns of substrates, based on hydraulic sorting in the meandering along the river. Large-scale redundancy of species-habitat associations was correlated with historical (phylogenetic) or adaptive (morphological) constraints on habitat selection. Stream habitats are constrained by regional factors of slope and sediments, but also determined by stable, repetitive and predictable local processes of erosion, transport, and deposition (meandering). The fish-habitat associations of catostomids in the Assiniboine River provide an example of interaction between the abundance of the component species, the phylogenetic constraints on the niche, and the deterministic nature of the spatial distribution of habitats. Within-feeding-group pairs showed that local habitat overlap is positively correlated with geographic overlap, while local diet overlap is negatively correlated with geographic overlap. Two species pairs (silver redhorse and golden redhorse and bigmouth buffalo and quillback) were concluded to have symmetric overlaps at the geographic scale and stable interactions. Using the functional niche concept ecological theory provides a link between ecology and biogeography of sympatric species. The multi-analytical approaches in this study provide insights into the structuring of north temperate prairie river fish communities, through hypothesis testing and correlation that have application beyond Prairie Rivers.

Low copy number anonymous nuclear loci were used to search for species markers in four species of Klamath Basin suckers. We sequenced 28 randomly chosen loci representing 10,421 bp; 21 loci were similar to sequences in GenBank. Eight fixed sequence differences were found among Klamath species. Locus 120 contained rare but diagnostic markers for Deltistes luxatus and for Catostomus rimiculus. Locus 4 also contained three rare but unique sites in Catostomus rimiculus. No sequence differences were found between Chasmistes brevirostris and Catostomus snyderi. Loci 4 and 120 exhibited allele frequency differences between Rogue River C. rimiculus and all Klamath Basin suckers. Genotype BB of locus 4 was a fixed diagnostic marker and genotype BB of locus 120 was a frequency dependent marker for Rogue C. rimiculus. Although Klamath suckers represent three genera, very limited variation was found among 10,431 base pairs. We examined phylogenetic patterns of five loci in eleven catostomid genera and 25 species to determine if the homogeneity in the Upper Klamath Basin was due to massive hybridization and introgression or to retention of ancestral sequences. Two loci with no similarity to GenBank sequences (non-coding loci) and three loci with substantial similarity to GenBank sequences (coding loci) gave similar results, providing support for various subfamilies and tribes, more support for eastern genera and little support for western genera. Each locus was a mosaic of species or population markers, sometimes providing discriminatory power for allopatric populations of a species, such as C. macrocheilus, while not discriminating other species. Upper Klamath Basin species were noteworthy in their lack of autapomorphies, but had similar numbers of derived informative sites as other catostomins. Upper Klamath Basin species consistently shared ancestral or equivocal informative sites either with moxostomatins or a variable group of western species and shared derived sites with other western species, especially C. occidentalis. The data suggest that Upper Klamath Basin species have retained a largely ancestral genome at these loci. Thus, the failure of this technique to uncover significant variation in Upper Klamath Basin species may be a reflection of their plesiomorphic genome at these loci and not necessarily hybridization. / Graduation date: 2004