Modelling the spread of infectious salmon anaemia among salmon farms based on seaway distances between farms and genetic relationships between infectious salmon anaemia virus isolates

Effects of regional coordination of salmon louse control in reducing negative impacts of salmonid aquaculture on wild salmonids

Parasitic salmon lice (Lepeophtheirus salmonis) are a constraint to the sustainable growth of salmonids in open net pens, and this issue has caused production to level off in recent years in the most aquaculture-intensive areas of Norway. The maximum allowed biomass at a regional level is regulated by using the so-called "traffic light" system, where salmon louse-induced mortality of migrating wild salmon post-smolts is evaluated against set targets. As a case study, we have investigated how a specific aquaculture-intensive area can reduce its louse levels sufficiently to achieve a low impact on wild salmon. Analyses of the output from a virtual post-smolt model that uses data on the reported number of salmon lice in fish farms as key input data and estimates the salmon louse-induced mortality of wild out-migrating Atlantic salmon post-smolts, suggested that female louse abundance on the local farms must be halved in spring to reach the goal implied by the traffic light system. The outcome of a modelling scenario simulating a proposed new plan for coordinated production and fallowing proved beneficial, with an overall reduction in louse infestations and treatment efforts. The interannual variability in louse abundance in spring, however, increased for this scenario, implying unacceptably high louse abundance when many farms were in their second production year. We then combined the scenario with coordinated production with other louse control measures. Only measures that reduced the density of farmed salmonids in open cages in the study area resulted in reductions in salmon louse infestations to acceptable levels. This could be achieved either by stocking with larger fish to reduce exposure time or by reducing fish numbers, e.g. by producing in closed units.

Development and application of a whole genome amplicon sequencing method for infectious salmon anemia virus (ISAV)

Infectious salmon anemia (ISA) is an infectious disease primarily affecting farmed Atlantic salmon, Salmo salar, which is caused by the ISA virus (ISAV). ISAV belongs to the Orthomyxoviridae family. The disease is a serious condition resulting in reduced fish welfare and high mortality. In this study, we designed an amplicon-based sequencing protocol for whole genome sequencing of ISAV. The method consists of 80 ISAV-specific primers that cover 92% of the virus genome and was designed to be used on an Illumina MiSeq platform. The sequencing accuracy was investigated by comparing sequences with previously published Sanger sequences. The sequences obtained were nearly identical to those obtained by Sanger sequencing, thus demonstrating that sequences produced by this amplicon sequencing protocol had an acceptable accuracy. The amplicon-based sequencing method was used to obtain the whole genome sequence of 12 different ISAV isolates from a small local epidemic in the northern part of Norway. Analysis of the whole genome sequences revealed that segment reassortment took place between some of the isolates and could identify which segments that had been reassorted.

Description of spatiotemporal patterns of infectious salmon anemia virus (ISAV) detections in marine Atlantic Salmon farms in Newfoundland and Labrador

OBJECTIVE

The objectives of this study were to describe spatiotemporal patterns of infectious salmon anemia virus (ISAV) detections in marine salmonid production sites in the province of Newfoundland and Labrador in Canada.

METHODS

Infectious salmon anemia virus surveillance data between 2012 and 2020 from the province of Newfoundland and Labrador were used. Data comprised a total of 94 sampling events from 20 Atlantic Salmon Salmo salar production sites in which ISAV was detected. Using linear regression models, factors influencing time to detection (days from stocking to first ISAV detection) and time to depopulation (days from first detection to production site depopulation) were investigated.

RESULT

Based on 28 unique cases, site-level annual incidence risk of ISAV detection ranged from 3% to 29%. The proportion of ISAV detection by PCR in fish samples ranged from 2% to 45% annually. Overall, ISAV variants from the European clade were more common than variants from the North American clade. The type of ISAV clade, detections of ISAV in nearest production sites based on seaway distances, and year of infectious salmon anemia cases were not associated with time to first ISAV detection. Time to depopulation for sites infected with the ISAV-HPRΔ variant was not associated with ISAV North American or European clades.

CONCLUSION

Our results contribute to the further understanding of the changing dynamics of infectious salmon anemia detections in Newfoundland and Labrador since its first detection in 2012 and will likely assist in the design of improved disease surveillance and control programs in the province.

A Scoping Review on GIS Technologies Applied to Farmed Fish Health Management

Finfish aquaculture, one of the fastest growing intensive sectors worldwide, is threatened by numerous transmissible diseases that may have devastating impacts on its economic sustainability. This review (2010-2022) used a PRISMA extension for scoping reviews and a text mining approach to explore the extent to which geographical information systems (GIS) are used in farmed fish health management and to unveil the main GIS technologies, databases, and functions used to update the spatiotemporal data underpinning risk and predictive models in aquatic surveillance programmes. After filtering for eligibility criteria, the literature search provided 54 records, highlighting the limited use of GIS technologies for disease prevention and control, as well as the prevalence of GIS application in marine salmonid farming, especially for viruses and parasitic diseases typically associated with these species. The text mining generated five main research areas, underlining a limited range of investigated species, rearing environments, and diseases, as well as highlighting the lack of GIS-based methodologies at the core of such publications. This scoping review provides a source of information for future more detailed literature analyses and outcomes to support the development of geospatial disease spread models and expand in-field GIS technologies for the prevention and mitigation of fish disease epidemics.

Graph-Based Reconstruction and Analysis of Disease Transmission Networks Using Viral Genomic Data

Understanding the patterns of viral disease transmissions helps establish public health policies and aids in controlling and ending a disease outbreak. Classical methods for studying disease transmission dynamics that rely on epidemiological data, such as times of sample collection and duration of exposure intervals, struggle to provide desired insight due to limited informativeness of such data. A more precise characterization of disease transmissions may be acquired from sequencing data that reveal genetic distance between viral genomes in patient samples. Indeed, genetic distance between viral strains present in hosts contains valuable information about transmission history, thus motivating the design of methods that rely on genomic data to reconstruct a directed disease transmission network, detect transmission clusters, and identify significant network nodes (e.g., super-spreaders). In this article, we present a novel end-to-end framework for the analysis of viral transmissions utilizing viral genomic (sequencing) data. The proposed framework groups infected hosts into transmission clusters based on the reconstructed viral strains infecting them; the genetic distance between a pair of hosts is calculated using Earth Mover's Distance, and further used to infer transmission direction between the hosts. To quantify the significance of a host in the transmission network, the importance score is calculated by a graph convolutional autoencoder. The viral transmission network is represented by a directed minimum spanning tree utilizing the Edmond's algorithm modified to incorporate constraints on the importance scores of the hosts. The proposed framework outperforms state-of-the-art techniques for the analysis of viral transmission dynamics in several experiments on semiexperimental as well as experimental data.

Methods Combining Genomic and Epidemiological Data in the Reconstruction of Transmission Trees: A Systematic Review

In order to better understand transmission dynamics and appropriately target control and preventive measures, studies have aimed to identify who-infected-whom in actual outbreaks. Numerous reconstruction methods exist, each with their own assumptions, types of data, and inference strategy. Thus, selecting a method can be difficult. Following PRISMA guidelines, we systematically reviewed the literature for methods combing epidemiological and genomic data in transmission tree reconstruction. We identified 22 methods from the 41 selected articles. We defined three families according to how genomic data was handled: a non-phylogenetic family, a sequential phylogenetic family, and a simultaneous phylogenetic family. We discussed methods according to the data needed as well as the underlying sequence mutation, within-host evolution, transmission, and case observation. In the non-phylogenetic family consisting of eight methods, pairwise genetic distances were estimated. In the phylogenetic families, transmission trees were inferred from phylogenetic trees either simultaneously (nine methods) or sequentially (five methods). While a majority of methods (17/22) modeled the transmission process, few (8/22) took into account imperfect case detection. Within-host evolution was generally (7/8) modeled as a coalescent process. These practical and theoretical considerations were highlighted in order to help select the appropriate method for an outbreak.

Infectious Salmon Anemia Virus Shedding from Infected Atlantic Salmon (Salmo salar L.)-Application of a Droplet Digital PCR Assay for Virus Quantification in Seawater

Infectious salmon anemia virus (ISAV) infection is currently detected by fish sampling for PCR and immunohistochemistry analysis. As an alternative to sampling fish, we evaluated two different membrane filters in combination with four buffers for elution, concentration, and detection of ISAV in seawater, during a bath challenge of Atlantic salmon (Salmo salar L.) post-smolts with high and low concentrations of ISAV. Transmission of ISAV in the bath challenge was confirmed by a high mortality, clinical signs associated with ISA disease, and detection of ISAV RNA in organ tissues and seawater samples. The electronegatively charged filter, combined with lysis buffer, gave significantly higher ISAV RNA detection by droplet digital PCR from seawater (5.6 × 10⁴ ISAV RNA copies/L; p < 0.001). Viral shedding in seawater was first detected at two days post-challenge and peaked on day 11 post-challenge, one day before mortalities started in fish challenged with high dose ISAV, demonstrating that a large viral shedding event occurs before death. These data provide important information for ISAV shedding that is relevant for the development of improved surveillance tools based on water samples, transmission models, and management of ISA.

DTU-DADS-Aqua: A simulation framework for modelling waterborne spread of highly infectious pathogens in marine aquaculture

Simulation models are useful tools to predict and elucidate the effects of factors influencing the occurrence and spread of epidemics in animal populations, evaluate the effectiveness of different control strategies and ultimately inform decision-makers about mitigations to reduce risk. There is a paucity of simulation models to study waterborne transmission of viral and bacterial pathogens in marine environments. We developed a stochastic, spatiotemporal hybrid simulation model (DTU-DADS-Aqua) that incorporates a compartmental model for infection spread within net-pens, an agent-based model for infection spread between net-pens within and between sites and uses seaway distance to inform farm-site hydroconnectivity. The model includes processes to simulate infection transmission and control over surveillance, detection and depopulation measures. Different what-if scenarios can be explored according to the input data provided and user-defined parameter values, such as daily surveillance and depopulation capacities or increased animal mortality that triggers diagnostic testing to detect infection. The latter can be easily defined in a software application, in which results are summarized after each simulation. To demonstrate capabilities of the model, we simulated the spread of infectious salmon anaemia virus (ISAv) for realistic scenarios in a transboundary population of farmed Atlantic salmon (Salmo salar L.) in New Brunswick, Canada and Maine, United States. We assessed the progression of infection in the different simulated outbreak scenarios, allowing for variation in the control strategies adopted for ISAv. Model results showed that improved disease detection, coupled with increasing surveillance visits to farm-sites and increased culling capacity for depopulation of infected net-pens reduced the number of infected net-pens and outbreak duration but the number of ISA-infected farm sites was minimally affected. DTU-DADS-Aqua is a flexible modelling framework, which can be applied to study different infectious diseases in the aquatic environment, allowing the incorporation of alternative transmission and control dynamics. The framework is open-source and available at https://github.com/upei-aqua/DTU-DADS-Aqua.

Short communication: Evaluation of charged membrane filters and buffers for concentration and recovery of infectious salmon anaemia virus in seawater

Infectious salmon anaemia virus (ISAV) is the cause of an important waterborne disease of farmed Atlantic salmon. Detection of virus in water samples may constitute an alternative method to sacrificing fish for surveillance of fish populations for the presence of ISA-virus. We aimed to evaluate different membrane filters and buffers for concentration and recovery of ISAV in seawater, prior to molecular detection. One litre each of artificial and natural seawater was spiked with ISAV, followed by concentration with different filters and subsequent elution with different buffers. The negatively charged MF hydrophilic membrane filter, combined with NucliSENS® lysis buffer, presented the highest ISAV recovery percentages with 12.5 ± 1.3% by RT-qPCR and 31.7 ± 10.7% by RT-ddPCR. For the positively charged 1 MDS Zeta Plus® Virosorb® membrane filter, combined with NucliSENS® lysis buffer, the ISAV recovery percentages were 3.4 ± 0.1% by RT-qPCR and 10.8 ± 14.2% by RT-ddPCR. The limits of quantification (LOQ) were estimated to be 2.2 x 103 ISAV copies/L of natural seawater for both RT-qPCR and RT-ddPCR. The ISAV concentration method was more efficient in natural seawater.

Challenges and Solutions to Viral Diseases of Finfish in Marine Aquaculture

Aquaculture is the fastest food-producing sector in the world, accounting for one-third of global food production. As is the case with all intensive farming systems, increase in infectious diseases has adversely impacted the growth of marine fish farming worldwide. Viral diseases cause high economic losses in marine aquaculture. We provide an overview of the major challenges limiting the control and prevention of viral diseases in marine fish farming, as well as highlight potential solutions. The major challenges include increase in the number of emerging viral diseases, wild reservoirs, migratory species, anthropogenic activities, limitations in diagnostic tools and expertise, transportation of virus contaminated ballast water, and international trade. The proposed solutions to these problems include developing biosecurity policies at global and national levels, implementation of biosecurity measures, vaccine development, use of antiviral drugs and probiotics to combat viral infections, selective breeding of disease-resistant fish, use of improved diagnostic tools, disease surveillance, as well as promoting the use of good husbandry and management practices. A multifaceted approach combining several control strategies would provide more effective long-lasting solutions to reduction in viral infections in marine aquaculture than using a single disease control approach like vaccination alone.

Data-Driven Network Modeling as a Framework to Evaluate the Transmission of Piscine Myocarditis Virus (PMCV) in the Irish Farmed Atlantic Salmon Population and the Impact of Different Mitigation Measures

Cardiomyopathy syndrome (CMS) is a severe cardiac disease of Atlantic salmon caused by the piscine myocarditis virus (PMCV), which was first reported in Ireland in 2012. In this paper, we describe the use of data-driven network modeling as a framework to evaluate the transmission of PMCV in the Irish farmed Atlantic salmon population and the impact of different mitigation measures. Input data included live fish movement data from 2009 to 2017, population dynamics events and the spatial location of the farms. With these inputs, we fitted a network-based stochastic infection spread model. After assumed initial introduction of the agent in 2009, our results indicate that it took 5 years to reach a between-farm prevalence of 100% in late 2014, with older fish being most affected. Local spread accounted for only a small proportion of new infections, being more important for sustained infection in a given area. Spread via movement of subclinically infected fish was most important for explaining the observed countrywide spread of the agent. Of the targeted intervention strategies evaluated, the most effective were those that target those fish farms in Ireland that can be considered the most connected, based on the number of farm-to-farm linkages in a specific time period through outward fish movements. The application of these interventions in a proactive way (before the first reported outbreak of the disease in 2012), assuming an active testing of fish consignments to and from the top 8 ranked farms in terms of outward fish movement, would have yielded the most protection for the Irish salmon farming industry. Using this approach, the between-farm PMCV prevalence never exceeded 20% throughout the simulation time (as opposed to the simulated 100% when no interventions are applied). We argue that the Irish salmon farming industry would benefit from this approach in the future, as it would help in early detection and prevention of the spread of viral agents currently exotic to the country.

Infectious Salmon Anemia and Farm-Level Culling Strategies

Infectious salmon anemia (ISA) is an infectious disease, and outbreaks must be handled to avoid spread between salmon sea farms. Intensive culling at infected farms is an important biosecurity measure to avoid further spread but is also a costly intervention that farmers try to avoid. A lack of action, however, may lead to new outbreaks in nearby salmon sea farms, with severe impacts on both economy and animal welfare. Here, we aim to explore how a time delay between a detected outbreak and the culling of both infected cages and entire farms affects the further spread of the disease. We use a previously published model to calculate how many salmon sea farms were directly infected in each outbreak. To investigate the effect of culling on the further spread of disease, we use the number of months elapsed from the detected outbreak to (a) the first cage being depopulated, and (b) to the entire salmon sea farm being depopulated as predictors of how many new farms the virus was transmitted to, after controlling for contact between the farms. We show that the lapse in time before the first cage is depopulated correlates positively with how many new salmon sea farms are infected, indicating that infected cages should be culled with as little time delay as possible. The model does not have sufficient power to separate between culling of only cages assumed to be infected and the entire farm, and, consequently, provides no direct empirical evidence for the latter. Lack of evidence is not evidence, however, and we argue that a high probability of spread between cages in infected salmon sea farms still supports the depopulation of entire farms as the safest option.

Estimating risk factors for the daily risk of developing clinical cardiomyopathy syndrome (CMS) on a fishgroup level

Cardiomyopathy syndrome (CMS) is a viral disease, causing significant mortality and decreased welfare in farmed salmon in the North Atlantic Ocean. In Norway, it has become the most important disease in animal husbandry, affecting more than 100 farms each year. Control of CMS is based on mitigation of risk factors, since no treatment or vaccine is available. However, little is known about how the disease spreads and develops, thus rendering disease control difficult for farmers and competent authorities. The objective of the present study was to identify risk factors leading to the development of clinical CMS, using data provided from the salmon producers. Daily production data from individual fishgroups in more than 120 salmon farms along the coast of Norway from fish put to sea in 2012-2014 was collected. The data included cause-specific mortalities, which was used to identify outbreaks of CMS and risk factors for disease. A model for describing the daily probability of outbreak of CMS in each fishgroup was developed. The model was run to find the most likely value for each of the parameters, given the observed outbreak data. From the data, we found that fish in the southern region of Norway have a much higher risk of developing CMS than fish in mid and west (parameter estimates (PE) 4.43 (CI: 2.54-7.04) vs. 3.27 and 2.58 (CI: 2.45-4.37 and 2.01-3.57). Further, across all regions, fish put to sea in the late fall develop CMS twice as often as fish put to sea in the early spring (PE 2.18-2.59; CI:1.54-4.6). Previous outbreaks of pancreas disease increased the risk of getting CMS with 3.36 (CI:2.97-3.78) in the west and 1.41 (CI: 1.24-1.63) in the mid regions and decreased the risk with 0.519 (CI: 0.456-0.611) in the south. Previous outbreaks of heart- and skeletal muscle inflammation increased the risk of CMS with 1.56-1.73 (CI:1.34-2.11) in the mid and south regions, and had no effect in the west. In addition, we found that fish groups originating from certain hatcheries had a higher risk of CMS than other fishgroups, independent on which farm they were farmed on. The risk of developing CMS also increased with the number of days at sea. The use of production data in the study gave the possibility to study disease development on a fish group level, and on a daily basis. Thus, the identification of risk factors provides new possibilities for control of disease.

Transmission Trees on a Known Pathogen Phylogeny: Enumeration and Sampling

One approach to the reconstruction of infectious disease transmission trees from pathogen genomic data has been to use a phylogenetic tree, reconstructed from pathogen sequences, and annotate its internal nodes to provide a reconstruction of which host each lineage was in at each point in time. If only one pathogen lineage can be transmitted to a new host (i.e., the transmission bottleneck is complete), this corresponds to partitioning the nodes of the phylogeny into connected regions, each of which represents evolution in an individual host. These partitions define the possible transmission trees that are consistent with a given phylogenetic tree. However, the mathematical properties of the transmission trees given a phylogeny remain largely unexplored. Here, we describe a procedure to calculate the number of possible transmission trees for a given phylogeny, and we then show how to uniformly sample from these transmission trees. The procedure is outlined for situations where one sample is available from each host and trees do not have branch lengths, and we also provide extensions for incomplete sampling, multiple sampling, and the application to time trees in a situation where limits on the period during which each host could have been infected and infectious are known. The sampling algorithm is available as an R package (STraTUS).

Reconstructing foot-and-mouth disease outbreaks: a methods comparison of transmission network models

A number of transmission network models are available that combine genomic and epidemiological data to reconstruct networks of who infected whom during infectious disease outbreaks. For such models to reliably inform decision-making they must be transparently validated, robust, and capable of producing accurate predictions within the short data collection and inference timeframes typical of outbreak responses. A lack of transparent multi-model comparisons reduces confidence in the accuracy of transmission network model outputs, negatively impacting on their more widespread use as decision-support tools. We undertook a formal comparison of the performance of nine published transmission network models based on a set of foot-and-mouth disease outbreaks simulated in a previously free country, with corresponding simulated phylogenies and genomic samples from animals on infected premises. Of the transmission network models tested, Lau’s systematic Bayesian integration framework was found to be the most accurate for inferring the transmission network and timing of exposures, correctly identifying the source of 73% of the infected premises (with 91% accuracy for sources with model support >0.80). The Structured COalescent Transmission Tree Inference provided the most accurate inference of molecular clock rates. This validation study points to which models might be reliably used to reconstruct similar future outbreaks and how to interpret the outputs to inform control. Further research could involve extending the best-performing models to explicitly represent within-host diversity so they can handle next-generation sequencing data, incorporating additional animal and farm-level covariates and combining predictions using Ensemble methods and other approaches.

Risk Factors Associated With Outbreaks of Infectious Salmon Anemia (ISA) With Unknown Source of Infection in Norway

The occurrence of infectious salmon anemia (ISA) outbreaks in marine farmed Atlantic salmon constitutes a recurring challenge in Norway. Here, we aim to identify risk factors associated with ISA outbreaks with an unknown source of infection (referred to as primary ISA outbreaks). Primary ISA outbreaks are here defined by an earlier published transmission model. We explored a wide range of possible risk factors with logistic regression analysis, trying to explain occurrence of primary ISA with available data from all Norwegian farm sites from 2004 to June 2017. Explanatory variables included site latitude and a range of production and disease data. The mean annual risk of having a primary outbreak of ISA in Norway was 0.7% during this study period. We identified the occurrence of infectious pancreatic necrosis (IPN), having a stocking period longer than 2 months, having the site located at high latitude and high fish density (biomass per cage volume) in the first six months after transfer to sea site as significant risk factors (p < 0.05). We have identified factors related to management routines, other disease problems, and latitude that may help to understand the hitherto unidentified drivers behind the emergence of primary ISA outbreaks. Based on our findings, we also provide management advice that may reduce the incidence of primary ISA outbreaks.

outbreaker2: a modular platform for outbreak reconstruction

BACKGROUND

Reconstructing individual transmission events in an infectious disease outbreak can provide valuable information and help inform infection control policy. Recent years have seen considerable progress in the development of methodologies for reconstructing transmission chains using both epidemiological and genetic data. However, only a few of these methods have been implemented in software packages, and with little consideration for customisability and interoperability. Users are therefore limited to a small number of alternatives, incompatible tools with fixed functionality, or forced to develop their own algorithms at considerable personal effort.

RESULTS

Here we present outbreaker2, a flexible framework for outbreak reconstruction. This R package re-implements and extends the original model introduced with outbreaker, but most importantly also provides a modular platform allowing users to specify custom models within an optimised inferential framework. As a proof of concept, we implement the within-host evolutionary model introduced with TransPhylo, which is very distinct from the original genetic model in outbreaker, and demonstrate how even complex model results can be successfully included with minimal effort.

CONCLUSIONS

outbreaker2 provides a valuable starting point for future outbreak reconstruction tools, and represents a unifying platform that promotes customisability and interoperability. Implemented in the R software, outbreaker2 joins a growing body of tools for outbreak analysis.

QUENTIN: reconstruction of disease transmissions from viral quasispecies genomic data

Motivation

Genomic analysis has become one of the major tools for disease outbreak investigations. However, existing computational frameworks for inference of transmission history from viral genomic data often do not consider intra-host diversity of pathogens and heavily rely on additional epidemiological data, such as sampling times and exposure intervals. This impedes genomic analysis of outbreaks of highly mutable viruses associated with chronic infections, such as human immunodeficiency virus and hepatitis C virus, whose transmissions are often carried out through minor intra-host variants, while the additional epidemiological information often is either unavailable or has a limited use.

Results

The proposed framework QUasispecies Evolution, Network-based Transmission INference (QUENTIN) addresses the above challenges by evolutionary analysis of intra-host viral populations sampled by deep sequencing and Bayesian inference using general properties of social networks relevant to infection dissemination. This method allows inference of transmission direction even without the supporting case-specific epidemiological information, identify transmission clusters and reconstruct transmission history. QUENTIN was validated on experimental and simulated data, and applied to investigate HCV transmission within a community of hosts with high-risk behavior. It is available at https://github.com/skumsp/QUENTIN.

Contact

pskums@gsu.edu or alexz@cs.gsu.edu or rahul@sfsu.edu or yek0@cdc.gov.

Supplementary information

Supplementary data are available at Bioinformatics online.

Connectivity-based risk ranking of infectious salmon anaemia virus (ISAv) outbreaks for targeted surveillance planning in Canada and the USA

Infectious salmon anaemia (ISA) can be a serious viral disease of farmed Atlantic salmon (Salmo salar). A tool to rank susceptible farms based on the risk of ISA virus (ISAv) infection spread from infectious farms after initial incursion or re-occurrence in an endemic area, can help guide monitoring and surveillance activities. Such a tool could also support the response strategy to contain virus spread, given available resources. We developed a tool to rank ISAv infection risks using seaway distance and hydrodynamic information separately and combined. The models were validated using 2002-2004 ISAv outbreak data for 30 farms (24 in New Brunswick, Canada and 6 in Maine, United States). Time sequence of infection spread was determined from the outbreak data that included monthly infection status of the cages on these farms. The first infected farm was considered as the index site for potential spread of ISAv to all other farms. To assess the risk of ISAv spreading to susceptible farms, the second and subsequent infected farms were identified using the farm status in the given time period and all infected farms from the previous time periods. Using the three models (hydrodynamic only, seaway-distance, and combined hydrodynamic-seaway-distance based models), we ranked susceptible farms within each time interval by adding the transmission risks from surrounding infected farms and sorting them from highest to lowest. To explore the potential efficiency of targeted sampling, we converted rankings to percentiles and assessed the model's predictive performance by comparing farms identified as high risk based on the rank with those that were infected during the next time interval as observed in the outbreak data. The overall predictive ability of the models was compared using area under the ROC curve (AUC). Farms that become infected in the next period were always within the top 65% of the rank predicted by our models. The overall predictive ability of the combined (hydrodynamic-seaway-distance based model) model (AUC = 0.833) was similar to the model that only used seaway distance (AUC = 0.827). Such models can aid in effective surveillance planning by balancing coverage (number of farms included in surveillance) against the desired level of confidence of including all farms that become infected in the next time period. Our results suggest that 100% of the farms that become infected in the next time period could be targeted in a surveillance program, although at a significant cost of including many false positives.

Bayesian reconstruction of transmission within outbreaks using genomic variants

Pathogen genome sequencing can reveal details of transmission histories and is a powerful tool in the fight against infectious disease. In particular, within-host pathogen genomic variants identified through heterozygous nucleotide base calls are a potential source of information to identify linked cases and infer direction and time of transmission. However, using such data effectively to model disease transmission presents a number of challenges, including differentiating genuine variants from those observed due to sequencing error, as well as the specification of a realistic model for within-host pathogen population dynamics. Here we propose a new Bayesian approach to transmission inference, BadTrIP (BAyesian epiDemiological TRansmission Inference from Polymorphisms), that explicitly models evolution of pathogen populations in an outbreak, transmission (including transmission bottlenecks), and sequencing error. BadTrIP enables the inference of host-to-host transmission from pathogen sequencing data and epidemiological data. By assuming that genomic variants are unlinked, our method does not require the computationally intensive and unreliable reconstruction of individual haplotypes. Using simulations we show that BadTrIP is robust in most scenarios and can accurately infer transmission events by efficiently combining information from genetic and epidemiological sources; thanks to its realistic model of pathogen evolution and the inclusion of epidemiological data, BadTrIP is also more accurate than existing approaches. BadTrIP is distributed as an open source package (https://bitbucket.org/nicofmay/badtrip) for the phylogenetic software BEAST2. We apply our method to reconstruct transmission history at the early stages of the 2014 Ebola outbreak, showcasing the power of within-host genomic variants to reconstruct transmission events.

We present a new tool to reconstruct transmission events within outbreaks. Our approach makes use of pathogen genetic information, notably genetic variants at low frequency within host that are usually discarded, and combines it with epidemiological information of host exposure to infection. This leads to accurate reconstruction of transmission even in cases where abundant within-host pathogen genetic variation and weak transmission bottlenecks (multiple pathogen units colonising a new host at transmission) would otherwise make inference difficult due to the transmission history differing from the pathogen evolution history inferred from pathogen isolets. Also, the use of within-host pathogen genomic variants increases the resolution of the reconstruction of the transmission tree even in scenarios with limited within-outbreak pathogen genetic diversity: within-host pathogen populations that appear identical at the level of consensus sequences can be discriminated using within-host variants. Our Bayesian approach provides a measure of the confidence in different possible transmission histories, and is published as open source software. We show with simulations and with an analysis of the beginning of the 2014 Ebola outbreak that our approach is applicable in many scenarios, improves our understanding of transmission dynamics, and will contribute to finding and limiting sources and routes of transmission, and therefore preventing the spread of infectious disease.

Can biosecurity and local network properties predict pathogen species richness in the salmonid industry?

Salmonid farming in Ireland is mostly organic, which implies limited disease treatment options. This highlights the importance of biosecurity for preventing the introduction and spread of infectious agents. Similarly, the effect of local network properties on infection spread processes has rarely been evaluated. In this paper, we characterized the biosecurity of salmonid farms in Ireland using a survey, and then developed a score for benchmarking the disease risk of salmonid farms. The usefulness and validity of this score, together with farm indegree (dichotomized as ≤ 1 or > 1), were assessed through generalized Poisson regression models, in which the modeled outcome was pathogen richness, defined here as the number of different diseases affecting a farm during a year. Seawater salmon (SW salmon) farms had the highest biosecurity scores with a median (interquartile range) of 82.3 (5.4), followed by freshwater salmon (FW salmon) with 75.2 (8.2), and freshwater trout (FW trout) farms with 74.8 (4.5). For FW salmon and trout farms, the top ranked model (in terms of leave-one-out information criteria, looic) was the null model (looic = 46.1). For SW salmon farms, the best ranking model was the full model with both predictors and their interaction (looic = 33.3). Farms with a higher biosecurity score were associated with lower pathogen richness, and farms with indegree > 1 (i.e. more than one fish supplier) were associated with increased pathogen richness. The effect of the interaction between these variables was also important, showing an antagonistic effect. This would indicate that biosecurity effectiveness is achieved through a broader perspective on the subject, which includes a minimization in the number of suppliers and hence in the possibilities for infection to enter a farm. The work presented here could be used to elaborate indicators of a farm’s disease risk based on its biosecurity score and indegree, to inform risk-based disease surveillance and control strategies for private and public stakeholders.

SCOTTI: Efficient Reconstruction of Transmission within Outbreaks with the Structured Coalescent

Exploiting pathogen genomes to reconstruct transmission represents a powerful tool in the fight against infectious disease. However, their interpretation rests on a number of simplifying assumptions that regularly ignore important complexities of real data, in particular within-host evolution and non-sampled patients. Here we propose a new approach to transmission inference called SCOTTI (Structured COalescent Transmission Tree Inference). This method is based on a statistical framework that models each host as a distinct population, and transmissions between hosts as migration events. Our computationally efficient implementation of this model enables the inference of host-to-host transmission while accommodating within-host evolution and non-sampled hosts. SCOTTI is distributed as an open source package for the phylogenetic software BEAST2. We show that SCOTTI can generally infer transmission events even in the presence of considerable within-host variation, can account for the uncertainty associated with the possible presence of non-sampled hosts, and can efficiently use data from multiple samples of the same host, although there is some reduction in accuracy when samples are collected very close to the infection time. We illustrate the features of our approach by investigating transmission from genetic and epidemiological data in a Foot and Mouth Disease Virus (FMDV) veterinary outbreak in England and a Klebsiella pneumoniae outbreak in a Nepali neonatal unit. Transmission histories inferred with SCOTTI will be important in devising effective measures to prevent and halt transmission.

We present a new tool, SCOTTI, to efficiently reconstruct transmission events within outbreaks. Our approach combines genetic information from infection samples with epidemiological information of patient exposure to infection. While epidemiological information has been traditionally used to understand who infected whom in an outbreak, detailed genetic information is increasingly becoming available with the steady progress of sequencing technologies. However, many complications, if unaccounted for, can affect the accuracy with which the transmission history is reconstructed. SCOTTI efficiently accounts for several complications, in particular within-patient genetic variation of the infectious organism, and non-sampled patients (such as asymptomatic patients). Thanks to these features, SCOTTI provides accurate reconstructions of transmission in complex scenarios, which will be important in finding and limiting the sources and routes of transmission, preventing the spread of infectious disease.

Using genomics data to reconstruct transmission trees during disease outbreaks

Genetic sequence data from pathogens present a novel means to investigate the spread of infectious disease between infected hosts or infected premises, complementing traditional contact-tracing approaches, and much recent work has gone into developing methods for this purpose. The objective is to recover the epidemic transmission tree, which identifies who infected whom. This paper reviews the various approaches that have been taken. The first step is to define a measure of difference between sequences, which must be done while taking into account such factors as recombination and convergent evolution. Three broad categories of method exist, of increasing complexity: those that assume no withinhost genetic diversity or mutation, those that assume no within-host diversity but allow mutation, and those that allow both. Until recently, the assumption was usually made that every host in the epidemic could be identified, but this is now being relaxed, and some methods are intended for sparsely sampled data, concentrating on the identification of pairs of sequences that are likely to be the result of direct transmission rather than inferring the complete transmission tree. Many of the procedures described here are available to researchers as free software.

Surveillance of the Sensitivity towards Antiparasitic Bath-Treatments in the Salmon Louse (Lepeophtheirus salmonis)

The evolution of drug resistant parasitic sea lice is of major concern to the salmon farming industry worldwide and challenges sustainable growth of this enterprise. To assess current status and development of L. salmonis sensitivity towards different pesticides used for parasite control in Norwegian salmon farming, a national surveillance programme was implemented in 2013. The programme aims to summarize data on the use of different pesticides applied to control L. salmonis and to test L. salmonis sensitivity to different pesticides in farms along the Norwegian coast. Here we analyse two years of test-data from biological assays designed to detect sensitivity-levels towards the pesticides azamethiphos and deltamethrin, both among the most common pesticides used in bath-treatments of farmed salmon in Norway in later years. The focus of the analysis is on how different variables predict the binomial outcome of the bioassay tests, being whether L. salmonis are immobilized/die or survive pesticide exposure. We found that local kernel densities of bath treatments, along with a spatial geographic index of test-farm locations, were significant predictors of the binomial outcome of the tests. Furthermore, the probability of L. salmonis being immobilized/dead after test-exposure was reduced by odds-ratios of 0.60 (95% CI: 0.42–0.86) for 2014 compared to 2013 and 0.39 (95% CI: 0.36–0.42) for low concentration compared to high concentration exposure. There were also significant but more marginal effects of parasite gender and developmental stage, and a relatively large random effect of test-farm. We conclude that the present data support an association between local intensities of bath treatments along the coast and the outcome of bioassay tests where salmon lice are exposed to azamethiphos or deltamethrin. Furthermore, there is a predictable structure of L. salmonis phenotypes along the coast in the data, characterized by high susceptibility to pesticides in the far north and far south, but low susceptibility in mid Norway. The study emphasizes the need to address local susceptibility to pesticides and the need for restrictive use of pesticides to preserve treatment efficacy.

Orthomyxoviruses of Fish

The family Orthomyxoviridae is well known for containing influenza viruses with a segmented RNA genome that is prone to gene reassortment in mixed infections (known as antigenic shift) resulting in new virus subtypes that cause pandemics, and cumulative mutations (known as antigenic drift), resulting in new virus strains that cause epidemics. This family also contains infectious salmon anemia virus (ISAV) and tilapia lake virus (TiLV), which are a unique orthomyxoviruses that infect fish and is unable to replicate above room temperature (24°C). This chapter describes the comparative virology of members in the family Orthomyxoviridae in general, helping to understand the emergent teleost orthomyxoviruses, ISAV and TiLV. The most current information on virus–host interactions of the fish orthomyxoviruses, particularly ISAV, as they relate to variations in virus structure, virulence, persistence, host range and immunological aspects is presented in detail.

Epidemic Reconstruction in a Phylogenetics Framework: Transmission Trees as Partitions of the Node Set

The use of genetic data to reconstruct the transmission tree of infectious disease epidemics and outbreaks has been the subject of an increasing number of studies, but previous approaches have usually either made assumptions that are not fully compatible with phylogenetic inference, or, where they have based inference on a phylogeny, have employed a procedure that requires this tree to be fixed. At the same time, the coalescent-based models of the pathogen population that are employed in the methods usually used for time-resolved phylogeny reconstruction are a considerable simplification of epidemic process, as they assume that pathogen lineages mix freely. Here, we contribute a new method that is simultaneously a phylogeny reconstruction method for isolates taken from an epidemic, and a procedure for transmission tree reconstruction. We observe that, if one or more samples is taken from each host in an epidemic or outbreak and these are used to build a phylogeny, a transmission tree is equivalent to a partition of the set of nodes of this phylogeny, such that each partition element is a set of nodes that is connected in the full tree and contains all the tips corresponding to samples taken from one and only one host. We then implement a Monte Carlo Markov Chain (MCMC) procedure for simultaneous sampling from the spaces of both trees, utilising a newly-designed set of phylogenetic tree proposals that also respect node partitions. We calculate the posterior probability of these partitioned trees based on a model that acknowledges the population structure of an epidemic by employing an individual-based disease transmission model and a coalescent process taking place within each host. We demonstrate our method, first using simulated data, and then with sequences taken from the H7N7 avian influenza outbreak that occurred in the Netherlands in 2003. We show that it is superior to established coalescent methods for reconstructing the topology and node heights of the phylogeny and performs well for transmission tree reconstruction when the phylogeny is well-resolved by the genetic data, but caution that this will often not be the case in practice and that existing genetic and epidemiological data should be used to configure such analyses whenever possible. This method is available for use by the research community as part of BEAST, one of the most widely-used packages for reconstruction of dated phylogenies.

With sequence data becoming available in increasing high volumes and at decreasing costs, there has been substantial recent interest in the possibility of using pathogen genome sequences as a means to retrace the spread of disease amongst the infected hosts in an epidemic. While several such methods exist, many of them are not fully compatible with phylogenetic inference, which is the most commonly-used methodology for exploring the ancestry of the isolates represented by a set of sequences. Procedures using phylogenetics as a basis have either taken a single, fixed phylogenetic tree as input, or have been quite narrow in scope and not available in any current package for general use. For their part, standard phylogenetic methods usually assume a model of the pathogen population that is overly simplistic for the situation in an epidemic. Here, we bridge the gap by introducing a new, highly flexible method, implemented in the publicly-available BEAST package, which simultaneously reconstructs the transmission history of an epidemic and the phylogeny for samples taken from it. We apply the procedure to simulated data and to sequences from the 2003 H7N7 avian influenza outbreak in the Netherlands.

Space-time modelling of the spread of pancreas disease (PD) within and between Norwegian marine salmonid farms

Infectious diseases are a constant threat to industrialised farming, which is characterised by high densities of farms and farm animals. Several mathematical and statistical models on spatio-temporal dynamics of infectious diseases in various farmed host populations have been developed during the last decades. Here we present a spatio-temporal stochastic model for the spread of a disease between and within aquaculture farms. The spread between farms is divided into several transmission pathways, including (i) distance related spread and (ii) other types of contagious contacts. The within-farm infection dynamics is modelled by a susceptible-infected-recovered (SIR) model. We apply this framework to model the spread of pancreas disease (PD) in salmon farming, using data covering all farms producing salmonids over 9 years in Norway. The motivation for the study was partly to unravel the spatio-temporal dynamics of PD in salmon farming and partly to use the model for scenario simulation of PD control strategies. We find, for example, that within-farm infection dynamics vary with season and we provide estimates of the timing from unobserved infection events to disease outbreaks on farms are detected. The simulations suggest that if a strategy involving culling of infectious cohorts is implemented, the number of detected disease outbreaks per year may be reduced by 57% after the full effect has been reached. We argue that the high detail and coverage of data on salmonid production and disease occurrence should encourage the use of simulation modelling as a means of testing effects of extensive control measures before they are implemented in the salmon farming industry.

Infectious salmon anemia (ISA) virus: infectivity in seawater under different physical conditions

Infectious salmon anemia (ISA) virus (genus Isavirus, family Orthomyxoviridae), present in all major salmon producing countries, is the causative agent for a serious and commercially important disease affecting Atlantic Salmon Salmo salar. Nearly all ISA outbreaks occur in the marine production phase and knowledge about survival time for ISA virions in seawater is crucial for an adequate strategy to combat the disease. To acquire knowledge about this important factor, a study of ISA virus exposed to four different physical conditions was carried out. The virions' survival was tested in sterile seawater, sterile seawater with normal ultraviolet light radiation (UVR), natural seawater, and natural seawater with UVR. During the 72-h experiment both presence of ISA virus RNA and the infectivity of ISA virions were monitored. The result of this study showed that the infectivity of ISA virions is lost within 3 h of exposure to natural seawater or sterile seawater with UVR. However, it was possible to detect ISA virus RNA throughout the experimental period. This indicates that the effect of both UVR and biological activity of natural seawater limits the survival time of ISA virions under normal conditions. The survival time of ISA virions in sterile seawater was less than 24 h. Based on the available literature and the present study it is not very likely that passive horizontal transmission in seawater over long distances can occur. This is due to the following factors: (1) the effect of UVR and biological activity on ISA virions infectivity found in the present study, (2) the speed and dilution effect in seawater currents in salmon farming areas, (3) the temperature during the major outbreak periods, and (4) the need for an infective dose of ISA virions to reach naive Atlantic Salmon.

Within-farm spread of infectious salmon anemia virus (ISAV) in Atlantic salmon Salmo salar farms in Chile

Spread of infectious salmon anemia virus (ISAV) at the cage level was quantified using a subset of data from 23 Atlantic salmon Salmo salar farms located in southern Chile. Data collected from official surveillance activities were systematically organized to obtain detailed information on infectious salmon anemia (ISA) outbreaks. Descriptive statistics for outbreak duration, proportion of infected fish, and time to secondary infection were calculated to quantify the magnitude of ISAV incursions. Linear and multiple failure time (MFT) regression models were used to determine factors associated with the cage-level reproduction number (Rc) and hazard rate (HR) for recurrent events, respectively. In addition, the Knox test was used to assess if cage-to-cage transmissions were clustered in space and time. Findings suggest that within farms, ISA outbreaks, on average, lasted 30 wk (median = 26 wk, 95% CI = 24 to 37 wk) and affected 57.3% (95% CI = 47.7 to 67.0%) of susceptible cages. The median time to secondarily diagnosed cages was 23 d. Occurrence of clinical ISAV outbreaks was significantly associated with increased Rc, whereas increased HR was significantly associated with clinical outbreaks and with a large number of fish. Spatio-temporal analysis failed to identify clustering of cage cases, suggesting that within-farm ISAV spread is independent of the spatial location of the cages. Results presented here will help to better understand ISAV transmission, to improve the design of surveillance programs in Chile and other regions in which salmon are intensively farmed, and to examine the economic impact of ISAV and related management strategies on various cost and demand shifting factors.

Space-time modelling of the spread of salmon lice between and within Norwegian marine salmon farms

Parasitic salmon lice are potentially harmful to salmonid hosts and farm produced lice pose a threat to wild salmonids. To control salmon lice infections in Norwegian salmonid farming, numbers of lice are regularly counted and lice abundance is reported from all salmonid farms every month. We have developed a stochastic space-time model where monthly lice abundance is modelled simultaneously for all farms. The set of farms is regarded as a network where the degree of contact between farms depends on their seaway distance. The expected lice abundance at each farm is modelled as a function of i) lice abundance in previous months at the same farm, ii) at neighbourhood farms, and iii) other, unspecified sources. In addition, the model includes explanatory variables such as seawater temperature and farm-numbers of fish. The model gives insight into factors that affect salmon lice abundance and contributing sources of infection. New findings in this study were that 66% of the expected salmon lice abundance was attributed to infection within farms, 28% was attributed to infection from neighbourhood farms and 6% to non-specified sources of infection. Furthermore, we present the relative risk of infection between neighbourhood farms as a function of seaway distance, which can be viewed as a between farm transmission kernel for salmon lice. The present modelling framework lays the foundation for development of future scenario simulation tools for examining the spread and abundance of salmon lice on farmed salmonids under different control regimes.

A comparison of modelling approaches to assess the transmission of pathogens between Scottish fish farms: the role of hydrodynamics and site biomass

Scotland is the largest Atlantic salmon (Salmo salar) producer in the EU with an output of over 150,000 t, contributing over £500 million annually towards the economy. Production continues to increase, predominantly due to the increase in output per farm and reduction in losses due to infectious diseases. Farms are grouped within disease management areas whose boundaries are defined by where the closest pair of farms is separated by more than twice the tidal excursion distance (TE) Tidal excursion is defined as 7.2 km in mainland Scotland, or 3.6 km in the Shetland Islands). The majority of salmon farms are located within relatively sheltered inshore areas where non-tidal advective current speed is minimal. However there is an aspiration for offshore production where it might be possible to increase stocking levels and where current speeds will be greater so TE models could break down. Separation distances whereby farms would avoid infection risk were obtained using an analytical, discrete-time Susceptible-Exposed-Infectious-Recovered (SEIR) model coupled with a hydrodynamic transport expression representing transmission of pathogenic agents between fish farms. The model incorporated transmission, expression and recovery parameters as well as pathogen shedding and decay. The simplified hydrodynamic model incorporated residual advection, tidal advection and turbulent diffusion elements. The obtained separation distances were compared to a computationally intensive, numerical model and were demonstrated to be comparable, although the analytical model underestimated the variation within the transmission distances. Applying characteristics for a robust pathogen, infectious pancreatic necrosis virus type (IPNV-type), and less robust pathogens such as infectious salmon anaemia virus type (ISAV-type) and Aeromonas salmonicida type (AS-type) pathogens, it was possible to obtain separation distances whereby farms avoided infection. Simulation outputs indicated that separation distances should increase to avoid disease as farm size and current speed increase. The more conserved IPNV-type pathogen required separation distances of hundreds of kilometres, AS-type required tens of kilometres, whilst the distances for ISAV-type were within the scale of the current DMAs, that were developed for ISAV control. However, should production be moved to areas of faster moving currents and increased farm production the current disease management area principles might need readdressing.

Low virulent infectious salmon anaemia virus (ISAV-HPR0) is prevalent and geographically structured in Norwegian salmon farming

Infectious salmon anaemia (ISA) is a severe disease in farmed Atlantic salmon Salmo salar that has caused epidemic outbreaks in most salmon-producing countries worldwide. The disease is caused by virulent ISA virus (ISAV). Low virulent variants of the virus, characterised by a full-length sequence in the highly polymorphic region of segment 6 in the virus genome, have been reported with increasing frequencies. These variants of the virus, termed HPR0, have been proposed to be ancestors of virulent ISAV. We examined this idea through studies of the phylogeographic and environmental distribution of ISAV-HPR0, as well as phylogeographic associations between virulent ISAV and ISAV-HPR0. Samples from 232 fish groups were screened for ISAV. Real-time RT-PCR was used for detection of ISAV, and the ISAV haemagglutinin esterase (HE) gene was characterised for positive samples. A Mantel test was used to test phylogeographic associations between pairs of ISAV-HPR0 HE gene sequences. A rank test was used to test associations between HE gene sequences from virulent ISAV and ISAV-HPR0. ISAV-HPR0 was detected in fish groups both in freshwater and marine environments, and in juveniles, on-grown marine salmon and broodstock salmon. Genetic and geographic distances between pairs of ISAV-HPR0 HE gene sequences were positively correlated, suggesting that the population of ISAV-HPR0 is geographically structured. Finally, we found a spatial association between fish groups with virulent ISAV (n = 21) and fish groups with ISAV-HPR0 (n = 27), supporting the hypothesis that ISAV-HPR0 may undergo a transition to virulent ISAV.

Risk mapping of heart and skeletal muscle inflammation in salmon farming

Heart and skeletal muscle inflammation (HSMI) is an infectious disease causing losses to the Norwegian salmon farming industry due to increased mortality and high morbidity in infected salmon. The disease is listed as a notifiable disease on list 3 (national list) by the Norwegian Food Safety Authority. HSMI is believed to be a viral disease, but the association to the recently discovered Piscine reovirus (PRV) remains unclear. Undoubtedly, other factors interact to determine whether PRV-infected fish develop disease or not. In this study, logistic regression was used to model the risk of an outbreak of HSMI at the cohort level, by including spatio-temporal risk factors. The data consisted of fish cohorts grown on geo-referenced farms from 2002 to 2010. The risk factors included were: infection pressure, cohort size (maximum number of fish), cohort index (smolt characteristics), cohort lifespan (months in sea) and a geo-index calculated as the position along a local polynomial regression line based on the longitude and latitude of each farm included in the study. The results showed that the risk of developing HSMI increased with increasing cohort lifespan, increasing infection pressure and increasing cohort size, and was mostly low for cohorts grown on farms in Southern-Norway, high for farms in Mid-Norway and variable for farms in Northern-Norway (based on the geo-index). The final model was used to explore three different scenarios with regards to the risk of developing HSMI, and to calculate the probability for each cohort of developing HSMI, independent of their actual disease-status. The model suggested that the probability of developing HSMI was much higher in Mid-Norway than in the rest of the country. Even though PRV seems to be widely distributed in the environment, the finding that infection pressure has a large influence on the probability of developing HSMI, suggests that it might be possible to reduce the number of clinical outbreaks, if measures are taken to reduce infection pressure. However, the prospects of controlling the spread of HSMI and reducing clinical outbreaks might be difficult because of indications of large distance spread of the disease.

Sea lice as a density-dependent constraint to salmonid farming

Fisheries catches worldwide have shown no increase over the last two decades, while aquaculture has been booming. To cover the demand for fish in the growing human population, continued high growth rates in aquaculture are needed. A potential constraint to such growth is infectious diseases, as disease transmission rates are expected to increase with increasing densities of farmed fish. Using an extensive dataset from all farms growing salmonids along the Norwegian coast, we document that densities of farmed salmonids surrounding individual farms have a strong effect on farm levels of parasitic sea lice and efforts to control sea lice infections. Furthermore, increased intervention efforts have been unsuccessful in controlling elevated infection levels in high salmonid density areas in 2009-2010. Our results emphasize host density effects of farmed salmonids on the population dynamics of sea lice and suggest that parasitic sea lice represent a potent negative feedback mechanism that may limit sustainable spatial densities of farmed salmonids.