Many viruses of arthropods also infect other organisms including humans, sometimes with devastating consequences. Yet, for the vast diversity of arthropods, their associated viruses remain unexplored. Here, we mined meta-transcriptomes from 711 arthropod species, including insects, arachnids, myriapods, and crustaceans, and uncovered more than 1400 previously unknown RNA viruses, representing 822 novel evolutionary groups at a level between species and genus. These newly found viral groups fill major evolutionary gaps within the five branches of RNA viruses, bridging the evolution of viruses infecting early and later diverging eukaryotes. Additionally, co-phylogenetic analysis implies that RNA viruses of arthropods commonly co-evolved with their hosts. Our analyses indicate that arthropods have played a central role in the macroevolution of RNA viruses by serving as reservoirs in which viruses co-evolved with arthropods while being exchanged with a vast diversity of organisms.

Carbon stable isotope analysis is an important tool in studies of fish ecology. Studies using this tool must account for the effect of lipids present in fish tissue on carbon stable isotope (δ13C) values. Simple correction equations have been developed to correct for this effect. For taxa with high fat content, choosing an accurate correction equation can have a significant impact on results. Pacific salmon (Oncorhynchus spp.) are a lipid-rich genus. Their broad marine distribution and ecological, cultural, and commercial importance make them prime candidates for stable isotope analysis. To determine both an accurate lipid correction equation for Pacific salmon δ13C values, and the effect of lipid extraction on bulk nitrogen isotope (δ15N) values, we performed pairwise isotope analysis on lipid-extracted and untreated muscle samples from 68 Chinook salmon O. tshawytscha spanning a size range of 165-750 mm and C:N values of 2.96-14.25. We compared the fit of existing δ13C lipid correction equations from 3 previously published models to our data, and optimized the top performing model using a leave-one-out cross validation. The model of Kiljunen et al. (2006; https://doi.org/10.1111/j.1365-2664.2006.01224.x) performed the best (mean squared error: 0.22, r2: 0.91), while the optimized model only slightly improved on it (MSE: 0.20, r2: 0.93). For δ15N, we determined that Chinook δ15N values significantly increased by 0.6‰ following lipid extraction. Our results confirm a lipid normalization procedure that is broadly applicable to Pacific salmon, and supports streamlined analysis of δ13C and δ15N from a single untreated muscle tissue sample.

Infectious disease from domestic hosts, held for agriculture, can impact wild species that migrate in close proximity, potentially reversing selective advantages afforded by migration. For sockeye salmon in British Columbia, Canada, juveniles migrate past numerous Atlantic salmon farms from which they may acquire a number of infectious agents. We analyse patterns of molecular detection in juvenile sockeye salmon for one bacterial pathogen, Tenacibaculum maritimum, known to cause disease in fish species around the globe and to cause mouthrot disease in farmed Atlantic salmon in BC. Our data show a clear peak in T. maritimum detections in the Discovery Islands region of BC, where sockeye migrate close to salmon farms. Using well established differential-equation models to describe sockeye migration and T. maritimum infection spread, we fit models to our detection data to assess support for multiple hypotheses describing farm- and background-origin infection. Despite a data-constrained inability to resolve certain epidemiological features of the system, such as the relative roles of post-infection mortality and recovery, our models clearly support the role of Discovery-Islands salmon farms in producing the observed patterns. Our best models (with 99.8% empirical model support) describe relatively constant (background) infection pressure, except around Discovery-Islands salmon farms, where farm-origin infection pressure peaked at 12.7 (approximate 95% CI: 4.5 to 31) times background levels. Given the evidence for farm-origin transfer of T. maritimum to Fraser-River sockeye salmon, the severity of associated disease in related species, and the imperilled nature of Fraser River sockeye generally, our results suggest the need for a more precautionary approach to managing farm/wild interactions in sockeye salmon.