I conducted field studies to determine the effect of insect herbivory and intraspecific plant competition on ragwort Senecio jacobaea. The objectives were to determine the patterns and causes in the distribution of the ragwort flea beetle Longitarsus jacobaeae foraging among varying densities of ragwort, to measure the behavioral and numerical responses of the beetle to changes in ragwort density, and to estimate the impact of insect herbivory and intraspecific competition on ragwort performance. Host density was manipulated by planting 1, 4, 8, or 16 plants per 0.5 x 0.5 m patch. Beetles were counted in each patch to assess the effect of host density on the beetle population. I measured four components of reproductive success represented by growth rate, development rate, reproduction, and annual survivorship to assess the effect of herbivory and intraspecific plant competition on ragwort performance In the first experiment, beetle populations were manipulated by establishing equal numbers of beetles in patches with unequal number of hosts (1, 4, 8, 16 plants per patch), which were then subsequently allowed to move freely about. Beetles rapidly re-distributed themselves, such that the number of beetles was strongly and positively correlated with the number of hosts. This indicates that ragwort flea beetles are highly sensitive to local distribution of their food plants. In the second experiment, host density was manipulated by planting ragwort in densities of 1, 4, 8, 16 plants per patch, and beetles were then allowed to colonize the experimental patches. Beetle behavioral response to a change in host density was dependent on host population size: the numbers of colonizing beetles increased asymptotically with increasing plant density. The number of beetle-days ranged from 261 for 1-plant patches to 1822 for 16-plant patches. In contrast, the numerical response (represented as observed multiplication rate per capita per generation per year) appears to be inexplicably low in the single plant population and levels off in the 4, 8, and 16 plant patches (grand mean for multiplication rate 1 was 5 and for multiplication rate 2 was 10.4 progeny per individual per generation). Combining these results, the beetles apparently respond to spatial variation in the density of hosts primarily by changes in their movement behavior rather than by changes in their per capita reproductive rates. These results highlight the importance of a natural enemy's colonizing behavior for controlling a sudden upsurge in pest abundance. Both insect herbivory and intraspecific competition had an effect on ragwort performance. For example, over approximately one year, ragwort's rate of biomass accumulation was 48% lower, and seed-head production was 18% lower in exposed compared to protected plots, while intraspecific competition reduced ragwort's rate of biomass accumulation and seed-head production, such that a 16-fold increase in host density (in protected patches) led to a 12-fold decrease in biomass per plant and a 11-fold decrease in the number of seed-heads per plant. Herbivore effects were independent of host density: variation in plant density from 1 to 16 plants led to no detectable change in magnitude of the herbivore effect. This suggests there is no density-dependent refuge for host plants operating at these local scales of observation. Keywords: Host density effects, behavioral response, reproductive response, biological control agent, Longitarsus jacobaeae, Senecio jacobaea. / Graduation date: 1995

Ragwort (Jacobaea vulgaris syn. Senecio jacobaea), a common weed of pasture and poorly managed land worldwide, is toxic to livestock and horses. There is no fully satisfactory control option available. The aim of the work described in this thesis was to assess the potential of root infecting pathogens to cause disease in ragwort, and to examine the possibility of using them as biological control agents against this weed. Thirty-six root infecting isolates were obtained from ragwort roots, collected from a nationwide postal survey where Pony Club adult leaders were asked to provide samples. Twenty-one of these were identified as Fusarium spp. by morphological identification and tested for the ability to cause disease on aseptically raised ragwort seedlings. Twelve isolates demonstrated virulence towards ragwort when measured by disease score over a 14 day period. These isolates were identified by molecular means using the internal transcribed spacer and translation elongation factor genes, as either Fusarium avenaceum or F. acuminatum while isolates of F. solani, F. redolens, F. cerealis or F. culmorum did not exhibit virulence except one isolate of F. culmorum which was weakly virulent. To investigate the biological control potential of virulent isolates, a representative of F. avenaceum and F. acuminatum plus the one weakly virulent F. culmorum isolate, were tested against plants of agricultural importance in grasslands. Six grasses (crested dogs tail, Cynosurus cristatus; Timothy, Phleum pratense; red fescue Festuca rubra; Italian ryegrass, Lolium multiflorum; and two varieties of perennial ryegrass, L. perenne) were unaffected by inoculation with any of the isolates. There was no significant difference between the symptoms caused by F. avenaceum and F. acuminatum on ragwort, red clover (Trifolium pratense) and white clover (T. repens). F. culmorum was more virulent towards white clover than either red clover or ragwort. This suggests that although high concentrations of pure PA may inhibit fungal growth, the range of other nutrients in the plants can counteract this negative effect.

Bacterial cultures enriched from sheep rumen fluid have demonstrated the ability to detoxify pyrrolizidine alkaloids (seneciphylline and jacobine) in tansy ragwort (Senecio jacobaea). The microbes are difficult to isolate using classical anaerobic techniques, therefore, microbes from two different enrichment cultures demonstrating similar degradation activity were identified using their 16S ribosomal RNA genes. Gene sequences from a rich medium enrichment were matched to Clostridium bifermentans, Prevotella ruminicola, Escherichia coif, and from a minimal medium enrichment to, C. clostridiiforme, C. aminophilum, Streptococcus bovis, and Butyrivibrio fibrosolvens. There were no identical organisms between the two libraries, but the common genus was Clostridium. / Graduation date: 1998

The cinnabar moth, Tyria jacobaeae (L.) (Lepidoptera:Arctiidae), was released in 1959 to control the grassland weed tansy ragwort, Senecio jacobaea L. (Asteraceae), despite evidence that caterpillars of this species can feed on native plants within the genera Senecio and Packera. Previous studies confirmed the moth's ability to develop on the native Senecio triangularis Hook., although no systematic study has been conducted to determine the extent of non-target impact on all potential host species. To address the lack of systematic studies we conducted a regional survey to determine the consequences of exposure of non-target plants to cinnabar moth caterpillars. We also conducted a local field experiment to determine the influence of habitat on the patterns of association of the moth and non-target plants. In the regional survey, we mapped the potential distribution of the cinnabar moth in Oregon to determine the extent of exposure of native Senecio and Packera species, and systematically sampled exposed species to assess the frequency and severity of feeding on these plants. We found that nine of the 20 native non-target species in Oregon were exposed to the cinnabar moth, three of the 10 native Senecio and six of the 10 native Packera. Ten of the native species escaped exposure because they occur east of the Cascade Mountain Range where the cinnabar moth does not occur. We found feeding damage on three of the nine exposed species: Packera cymbalarioides, P. pseudaurea, and S. triangularis were attacked at one of three (33%), two of six (33%), and seven of 15 (47%) sites that supported populations of each species, respectively. Within sites, attack frequency of stems was 33% (of six total stems sampled) for P. cymbalarioides, and ranged from 53% to 56% (of 20 to 108 total stems sampled) for P. pseudaurea and 7% to 64.5% (of 32 to 458 total stems sampled) for S. triangularis. Conditional median damage per site (median of attacked stems only) was 10% in P. cymbalarioides, 5% to 17.5% in P. pseudaurea, and 5% to 37.5% in S. triangularis. The attack rate on non-target plants (7.1 to 64.5 percent of stems attacked at a singe site) was equal to or greater than on the target weed (8.3 to 50.0 percent of stems attacked at a single site). At three sites, caterpillars attacked non-target plants but the target weed was absent, and at one site, the target was present but caterpillars fed on non-target plants only. We conclude that attack frequency and severity on the three species is not high, but equaled or exceeded the level of attack on the target weed. We also conducted a mark-release-recapture experiment to relate habitat preference to patterns of non-target host use in the field. We compared adult moth dispersal patterns and larval development between a meadow habitat and a forest habitat. We found that long-term dispersal distance (spanning days) was similar in both habitats but we recaptured a higher percentage of moths from the meadow (47%) compared to the forest (10%). Short-term displacements, based on direct observations of flights immediately after release, differed between habitats: moths in the meadow flew short distances (8.5m ± 1.5, n=13) at or below the herbaceous canopy (0.8 m ± 0.2, n=13) while moths in the forest flew longer horizontal (22.8 m ± 2.8, n=15) and vertical distances (5.9 m ± 0.9, n=15). We recovered seven fifth instar larvae (of 278 eggs) from the meadow habitat but no larvae beyond the second instar (of 119 eggs) were recovered from the forest habitat. We conclude that the cinnabar moth is limited to meadow habitats because adult moths display movement patterns that remove them from forest habitats (possibly due to disorientation) and larvae are unable to survive on plants growing in the forest. Taken together, the regional survey and the local field-experiment indicate that the cinnabar moth uses only a small proportion of available non-target host plant species. Other species are likely unused because of geographic isolation from the moth, habitat selection by the moth, or phenological differences between the moth and non-target plants. / Graduation date: 2003

Graduation date: 1998

Thesis (Ph. D.)--Oregon State University, 2008. / Printout. Includes bibliographical references. Also available on the World Wide Web.