A line of rainbow trout with improved resistance to Flavobacterium psychrophilum, the causative agent of bacterial cold water disease (BCWD) , was developed over five generations of selective breeding. As a result of selection , mean survival following laboratory F. psychrophilum challenge in the resistant (ARS-Fp-R) and contemporary control (ARS-Fp-C) and susceptible (ARS-Fp-S) lines was 84.9, 50.5, and 22.0%, respectively, and the g enetic trend for survival in the ARS-Fp-R line was +10.1 percentage points per generation. Importantly, improved survival in the ARS-Fp -R line based on laboratory challenges translated to improved resistance to BCWD in production-scale field challenges where fish were naturally exposed to the pathogen.
Preliminary laboratory challenges of families from the three genetic lines with F. columnare , the causative agent of columnaris disease, suggested that: 1) survival following F. columnare challenge is heritable; and 2) resistance to F. psychrophilum and F. columnare has a positive (favorable) genetic correlation in this population. As a result, selection to improve survival following F. columnare challenge became the sole breeding objective. Fifth-generation families (n = 100) from the ARS-Fp -R line served as the base population, and selection is currently being practiced to develop double-resistant (ARS-Fp/Fc-R), randomly-mated control (ARS-Fp-R), and susceptible (ARS-Fc-S) lines over five subsequent generations of selection.
Contrary to expectations, no upward selection response and only modest downward selection response was observed through the first two generations; mean survival following F. columnare challenge of second-generation families was 19.2, 19.7, and 6.2% for the ARS-Fp/Fc-R, ARS-Fp-R, and ARS-Fc-S lines, respectively. Whereas typical family variation in survival was observed each generation (i.e., ranges of 27 – 98%, 9 – 89%, and 1 – 62% survival in the base, first, and second generations, respectively) , mid-parent breeding values were not predictive of progeny survival in the subsequent generation (within-line correlations ≤ 0.06).
Compared to the BCWD challenge (injection-based, flow-through challenge model), increased variation (~1.4-fold) in survival across replicates (challenge tanks) within a family was observed for the columnaris challenge (immersion-based, flow-through challenge model ), suggesting reduced accuracy of the survival phenotype due to unknown non-genetic effects. Efforts are currently underway to develop and evaluate the utility of a mixed-family (common garden) recirculation-based columnaris challenge model to improve accuracy of the survival phenotype and resulting breeding value estimates.