Whirling disease, caused by the parasite Myxobolus cerebralis, has infected rainbow trout (Oncorhynchus mykiss) and other salmonid fish in the western United States, often with devastating results to native populations but without a discernible spatial pattern. The parasite develops in a complex 2-host system in which the aquatic oligochaete Tubifex tubifex is an obligate host. Because substantial differences in whirling disease severity in different areas of North America did not seem explainable by environmental factors or features of the parasite or its fish host, we sought to determine whether ecological or genetic variation within oligochaete host populations may be responsible. We found large differences in compatibility between the parasite and various laboratory strains of T. tubifex that were established from geographic regions with different whirling disease histories. Moreover, 2 closely related species of tubificids, Limnodrilus hoffmeisteri and Ilyodrilus templetoni, which occur naturally in mixed species assemblages with T. tubifex, were incompatible with M. cerebralis. Virulence of the parasite was directly correlated with the numbers of triactinomyxon spores that developed within each strain of T. tubifex. Thus, the level of virulence was directly related to the compatibility between the host strain and the parasite. Genetic analyses revealed relationships that were in agreement with the level of parasite production. Differences in compatibilities between oligochaetes and M. cerebralis may contribute to the spatial variance in the severity of the disease among salmonid populations.

Myxobolus cerebralis, the parasite that causes salmonid whirling disease, has had detrimental effects on several salmonid populations in the Intermountain West, including the rainbow trout in the Madison River, Montana, USA. The goal of this study was to examine relationships among characteristics of the environment, Tubifex tubifex (the alternate host) populations, and rainbow trout whirling disease risk in the Madison River. Environmental characteristics were measured in side channels of the Madison River, and differences were described with a principal components analysis. The density of T. tubifex, the prevalence of infection in T. tubifex, and the density of infected T. tubifex were determined for the side channels using benthic core samples and examination of live tubificids for infection. The site-specific contribution to whirling disease risk in the side channels was determined using in situ exposures of sentinel rainbow trout. Regression analyses were used to determine correlations among these characteristics. Side channels differed in site-specific contribution to rainbow trout whirling disease risk, which was positively correlated to the density of infected T. tubifex. Side channels with fine sediments and lower water temperatures made greater site-specific contribution to whirling disease risk and had higher densities of infected T. tubifex than side channels with coarser sediments and higher temperatures. The ability to characterize areas of high whirling disease risk is essential for improving our understanding of the dynamics of M. cerebralis such that appropriate management strategies can be implemented. In addition, this study provides a model of how the disease ecology of complex aquatic parasites can be examined when the influential processes operate on different spatial scales.

Abstract.-The exotic parasite Myxobolus cerebralis was first detected in native adult Yellowstone cutthroat trout Oncorhynchus clarkii bouvierii from Yellowstone Lake in 1998, seriously threatening the ecological integrity of this pristine, naturally functioning ecosystem. We immediately began to assess the prevalence and spatial extent of M. cerebralis infection in Yellowstone cutthroat trout within Yellowstone Lake and to determine the infection risk of age-0 Yellowstone cutthroat trout, the relative abundance and actinospore production of tubificid worms, and the basic environmental characteristics of tributaries. During 1999-2001, juvenile and adult Yellowstone cutthroat trout were infected throughout Yellowstone Lake; the highest prevalence (15.3-16.4%) occurred in the northern and central regions. Exposure studies in 13 streams indicated that Pelican and Clear creeks and the Yellowstone River were positive for M. cerebralis; the highest prevalence (100%) and severity was found in Pelican Creek during mid-July. Sexually mature individuals of the oligochaete Tubifex tubifex were most abundant in early summer, were genetically homogenous, and were members of a lineage known to produce moderate to high levels of M. cerebralis triactinomyxons. Only 20 of the 3,037 sampled tubificids produced actinospores after 7 d in culture, and none of the actinospores were M. cerebralis. However, one non-actinospore-producing T. tubifex from Pelican Creek tested positive for M. cerebralis by polymerase chain reaction. Stream temperatures at Pelican Creek, a fourth-order, low-gradient stream, were over 208C during the first exposure period, suggesting that T. tubifex were capable of producing triactinomyxons at elevated temperatures in the wild. Although the infection of otherwise healthy adult Yellowstone cutthroat trout within Yellowstone Lake suggests some resistance, our sentinel cage exposures indicated that this subspecies may be more susceptible to whirling disease than previous laboratory challenges have indicated, and M. cerebralis may be contributing to a significant recent decline in this population.

Laboratory populations of Tubifex tubifex from mitochondrial (mt)16S ribosomal DNA (rDNA) lineage III were generated from single cocoons of adult worms releasing the triactinomyxon stages (TAMs) of the myxozoan parasite, Myxobolus cerebralis. Subsequent worm populations from these cocoons, referred to as clonal lines, were tested for susceptibility to infection with the myxospore stages of M. cerebralis. Development and release of TAMs occurred in five clonal lines, while four clonal lines showed immature parasitic forms that were not expelled from the worm (non-TAM producers). Oligochaetes from TAM- and non-TAM-producing clonal lines were confirmed as lineage III based on mt16S rDNA and internal transcribed spacer region 1 (ITS1) sequences, but these genes did not differentiate these phenotypes. In contrast, random amplified polymorphic DNA analyses of genomic DNA demonstrated unique banding patterns that distinguished the phenotypes. Cohabitation of parasite-exposed TAM- and non-TAM-producing phenotypes showed an overall decrease in expected TAM production compared to the same exposure dose of the TAM-producing phenotype without cohabitation. These studies suggest that differences in susceptibility to parasite infection can occur in genetically similar T. tubifex populations, and their coexistence may affect overall M. cerebralis production, a factor that may influence the severity of whirling disease in wild trout populations.