The Potential Role of Direct and Indirect Contacts on Infection Spread in Dairy Farm Networks

Analyzing the intrastate and interstate swine movement network in the United States

Identifying and restricting animal movements is a common approach used to mitigate the spread of diseases between premises in livestock systems. Therefore, it is essential to uncover between-premises movement dynamics, including shipment distances and network-based control strategies. Here, we analyzed three years of between-premises pig movements, which include 197,022 unique animal shipments, 3973 premises, and 391,625,374 pigs shipped across 20 U.S. states. We constructed unweighted, directed, temporal networks at 180-day intervals to calculate premises-to-premises movement distances, the size of connected components, network loyalty, and degree distributions, and, based on the out-going contact chains, identified network-based control actions. Our results show that the median distance between premises pig movements was 74.37 km, with median intrastate and interstate movements of 52.71 km and 328.76 km, respectively. On average, 2842 premises were connected via 6705 edges, resulting in a weak giant connected component that included 91 % of the premises. The premises-level network exhibited loyalty, with a median of 0.65 (IQR: 0.45 - 0.77). Results highlight the effectiveness of node targeting to reduce the risk of disease spread; we demonstrated that targeting 25 % of farms with the highest degree or betweenness limited spread to 1.23 % and 1.7 % of premises, respectively. While there is no complete shipment data for the entire U.S., our multi-state movement analysis demonstrated the value and the needs of such data for enhancing the design and implementation of proactive- disease control tactics.

Descriptive network analysis and the influence of timescale on centrality and cohesion metrics from a system of between-herd dairy cow movements in Ontario, Canada

Previous research has demonstrated that static monthly networks of between-herd dairy cow movements in Ontario, Canada were highly fragmented, reducing potential for large-scale outbreaks. Extrapolating results from static networks can become problematic for diseases with an incubation period that exceeds the timescale of the network. The objectives of this research were to: 1) describe the networks of dairy cow movements in Ontario, and 2) describe the changes that occur among network analysis metrics when conducted at seven different timescales. Networks of dairy cow movements were created using Lactanet Canada milk recording data collected in Ontario between 2009 and 2018. Centrality and cohesion metrics were calculated after aggregating the data at seven timescales: weekly, monthly, semi-annual, annual, biennial, quinquennial, and decennial. There were 50,598 individual cows moved between Lactanet-enrolled farms, representing approximately 75% of provincially registered dairy herds. Most movements occurred over short distances (median = 39.18 km), with fewer long-range movements (maximum = 1150.80 km). The number of arcs increased marginally relative to the number of nodes with longer network timescales. Both mean out-degree, and mean clustering coefficients increased disproportionately with increasing timescale. Conversely, mean network density decreased with increasing timescale. The largest weak and strong components at the monthly timescale were small relative to the full network (267 and 4 nodes), whereas yearly networks had much higher values (2213 and 111 nodes). Higher relative connectivity in networks with longer timescales suggests pathogens with long incubation periods and animals with subclinical infection present increased potential for wide-spread disease transmission among dairy farms in Ontario. Careful consideration of disease-specific dynamics should be made when using static networks to model disease transmission among dairy cow populations.

Multilayer network analysis of FMD transmission and containment among beef cattle farms

As a highly contagious livestock viral disease, foot-and-mouth disease poses a great threat to the beef-cattle industry. Direct animal movement is always considered as a major route for between-farm transmission of FMD virus. Sharing contaminated equipment and vehicles have also attracted increasing interests as an indirect but considerable route for FMD virus transmission. With the rapid development of communication technologies, information-sharing techniques have been used to control epidemics. In this paper, we built farm-level time-series three-layer networks to simulate the between-farm FMD virus transmission in southwest Kansas by cattle movements (direct-contact layer) and truck visits (indirect-contact layer) and evaluate the impact of information-sharing techniques (information-sharing layer) on mitigating the epidemic. Here, the information-sharing network is defined as the structure that enables the quarantine of farms that are connected with infected farms. When a farm is infected, its infection status is shared with the neighboring farms in the information-sharing network, which in turn become quarantined. The results show that truck visits can enlarge the epidemic size and prolong the epidemic duration of the FMD outbreak by cattle movements, and that the information-sharing technique is able to mitigate the epidemic. The mitigation effect of the information-sharing network varies with the information-sharing network topology and different participation levels. In general, an increased participation leads to a decreased epidemic size and an increased quarantine size. We compared the mitigation performance of three different information-sharing networks (random network, contact-based network, and distance-based network) and found the outbreak on the network with contact-based information-sharing layer has the smallest epidemic size under almost any participation level and smallest quarantine size with high participation. Furthermore, we explored the potential economic loss from the infection and the quarantine. By varying the ratio of the average loss of quarantine to the loss of infection, we found high participation results in reduced economic losses under the realistic assumption that culling costs are much greater than quarantine costs.

Direct and indirect pathways for the spread of African swine fever and other porcine infectious diseases: An application of the mental models approach

In this study, we investigated the occurrence of direct and indirect infectious disease transmission pathways among pig farms in Switzerland, as well as their specific relevance for the spread of African swine fever, porcine reproductive and respiratory syndrome (PRRS), and enzootic pneumonia. Data were collected using an adapted mental models approach, involving initial interviews with experts in the field of pig health and logistics, semi-structured interviews with pig farmers, and a final expert workshop, during which all identified pathways were graded by their predicted frequency of occurrence, their likelihood of spread of the three diseases of interest, and their overall relevance considering both parameters. As many as 24 disease pathways were identified in four areas: pig trade, farmer encounters, external collaborators, and environmental or other pathways. Two thirds of the pathways were expected to occur with moderate-to-high frequency. While both direct and indirect pig trade transmission routes were highly relevant for the spread of the three pathogens, pathways from the remaining areas were especially important for PRRS due to higher spread potential via aerosols and fomites. In addition, we identified factors modifying the relevance of disease pathways, such as farm production type and affiliation with trader companies. During the interviews, we found varying levels of risk perception among farmers concerning some of the pathways, which affected adherence to biosecurity measures and were often linked to the degree of trust that farmers had towards their colleagues and external collaborators. Our findings highlight the importance of integrating indirect disease pathways into existing surveillance and control strategies and in disease modelling efforts. We also propose that biosecurity training aimed at professionals and risk communication campaigns targeting farmers should be considered to mitigate the risk of disease spread through the identified pathways.

Modelling transmission of Mycobacterium avium subspecies paratuberculosis between Irish dairy cattle herds

Bovine paratuberculosis is an endemic disease caused by Mycobacterium avium subspecies paratuberculosis (Map). Map is mainly transmitted between herds through movement of infected but undetected animals. Our objective was to investigate the effect of observed herd characteristics on Map spread on a national scale in Ireland. Herd characteristics included herd size, number of breeding bulls introduced, number of animals purchased and sold, and number of herds the focal herd purchases from and sells to. We used these characteristics to classify herds in accordance with their probability of becoming infected and of spreading infection to other herds. A stochastic individual-based model was used to represent herd demography and Map infection dynamics of each dairy cattle herd in Ireland. Data on herd size and composition, as well as birth, death, and culling events were used to characterize herd demography. Herds were connected with each other through observed animal trade movements. Data consisted of 13 353 herds, with 4 494 768 dairy female animals, and 72 991 breeding bulls. We showed that the probability of an infected animal being introduced into the herd increases both with an increasing number of animals that enter a herd via trade and number of herds from which animals are sourced. Herds that both buy and sell a lot of animals pose the highest infection risk to other herds and could therefore play an important role in Map spread between herds.

Asymmetric host movement reshapes local disease dynamics in metapopulations

Understanding how the movement of individuals affects disease dynamics is critical to accurately predicting and responding to the spread of disease in an increasingly interconnected world. In particular, it is not yet known how movement between patches affects local disease dynamics (e.g., whether pathogen prevalence remains steady or oscillates through time). Considering a set of small, archetypal metapopulations, we find three surprisingly simple patterns emerge in local disease dynamics following the introduction of movement between patches: (1) movement between identical patches with cyclical pathogen prevalence dampens oscillations in the destination while increasing synchrony between patches; (2) when patches differ from one another in the absence of movement, adding movement allows dynamics to propagate between patches, alternatively stabilizing or destabilizing dynamics in the destination based on the dynamics at the origin; and (3) it is easier for movement to induce cyclical dynamics than to induce a steady-state. Considering these archetypal networks (and the patterns they exemplify) as building blocks of larger, more realistically complex metapopulations provides an avenue for novel insights into the role of host movement on disease dynamics. Moreover, this work demonstrates a framework for future predictive modelling of disease spread in real populations.

Multiple species animal movements: network properties, disease dynamics and the impact of targeted control actions

Infectious diseases in livestock are well-known to infect multiple hosts and persist through a combination of within- and between-host transmission pathways. Uncertainty remains about the epidemic dynamics of diseases being introduced on farms with more than one susceptible host species. Here, we describe multi-host contact networks and elucidate the potential of disease spread through farms with multiple hosts. Four years of between-farm animal movement among all farms of a Brazilian state were described through a static and monthly snapshot of network representations. We developed a stochastic multilevel model to simulate scenarios in which infection was seeded into single host and multi-host farms to quantify disease spread potential, and simulate network-based control actions used to evaluate the reduction of secondarily infected farms. We showed that the swine network was more connected than cattle and small ruminants in both the static and monthly snapshots. The small ruminant network was highly fragmented, however, contributed to interconnecting farms, with other hosts acting as intermediaries throughout the networks. When a single host was initially infected, secondary infections were observed across farms with all other species. Our stochastic multi-host model demonstrated that targeting the top 3.25% of the farms ranked by degree reduced the number of secondarily infected farms. The results of the simulation highlight the importance of considering multi-host dynamics and contact networks while designing surveillance and preparedness control strategies against pathogens known to infect multiple species.

Unravelling direct and indirect contact patterns between duck farms in France and their association with the 2016-2017 epidemic of Highly Pathogenic Avian Influenza (H5N8)

Live animal movements generate direct contacts (via the exchange of live animals) and indirect contacts (via the transit of transport vehicles) between farms, which can contribute to the spread of pathogens. However, most analyses focus solely on direct contacts and can therefore underestimate the contribution of live animal movements in the spread of infectious diseases. Here, we used French live duck movement data (2016-2018) from one of the largest transport companies to compare direct and indirect contact patterns between duck farms and evaluate how these patterns were associated with the French 2016-2017 epidemic of highly pathogenic avian influenza H5N8. A total number of 614 farms were included in the study, and two directed networks were generated: the animal introduction network (exchange of live ducks) and the transit network (transit of transport vehicles). Following descriptive analyses, these two networks were scrutinized in relation to farm infection status during the epidemic. Results showed that farms were substantially more connected in the transit network than in the animal introduction network and that the transit of transport vehicles generated more opportunities for transmission than the exchange of live animals. We also showed that animal introduction and transit networks' statistics decreased substantially during the epidemic (January-March 2017) compared to non-epidemic periods (January-March 2016 and January-March 2018). We estimated a probability of 33.3 % that a farm exposed to the infection through either of the two live duck movement networks (i.e. that was in direct or indirect contact with a farm that was reported as infected in the following seven days) becomes infected within seven days after the contact. However, we also demonstrated that the level of exposure of farms by these two contact patterns was low, leading only to a handful of transmission events through these routes. As a consequence, we showed that live animal movement patterns are efficient transmission routes for HPAI but have been efficiently reduced to limit the spread during the French 2020-2021 epidemic. These results underpin the relevance of studying indirect contacts resulting from the movement of animals to understand their transmission potential and the importance of accounting for both routes when designing disease control strategies.

Integrating diverse data sources to predict disease risk in dairy cattle-a machine learning approach

Livestock farming is currently undergoing a digital revolution and becoming increasingly data-driven. Yet, such data often reside in disconnected silos making them impossible to leverage their full potential to improve animal well-being. Here, we introduce a precision livestock farming approach, bringing together information streams from a variety of life domains of dairy cattle to study whether including more and diverse data sources improves the quality of predictions for eight diseases and whether using more complex prediction algorithms can, to some extent, compensate for less diverse data. Using three machine learning approaches of varying complexity (from logistic regression to gradient boosted trees) trained on data from 5,828 animals in 165 herds in Austria, we show that the prediction of lameness, acute and chronic mastitis, anestrus, ovarian cysts, metritis, ketosis (hyperketonemia), and periparturient hypocalcemia (milk fever) from routinely available data gives encouraging results. For example, we can predict lameness with high sensitivity and specificity (F1 = 0.74). An analysis of the importance of individual variables to prediction performance shows that disease in dairy cattle is a product of the complex interplay between a multitude of life domains, such as housing, nutrition, or climate, that including more and diverse data sources increases prediction performance, and that the reuse of existing data can create actionable information for preventive interventions. Our findings pave the way toward data-driven point-of-care interventions and demonstrate the added value of integrating all available data in the dairy industry to improve animal well-being and reduce disease risk.

Survey of emergency response plans for managing the movement of cattle during a foot-and-mouth disease outbreak in North America

OBJECTIVE

To collect information from US state animal health officials (SAHOs) and beef feedlot managers and veterinarians regarding emergency response plans for movement of cattle in the event of a foot-and-mouth disease (FMD) outbreak in North America.

SAMPLE

36 SAHOs, 26 feedlot veterinarians, and 7 feedlot managers.

PROCEDURES

3 versions of an electronic questionnaire were created and distributed to SAHOs and US feedlot veterinarians and managers to gather information about planned or expected responses to an FMD outbreak that originated at 1 of 3 geographic locations (Mexico or Canada, a bordering state, or a nonbordering state). Descriptive data were reported.

RESULTS

All respondents recognized that the risk of FMD transmission to livestock in their area or care increased as the outbreak got closer in proximity to their location. Most SAHOs indicated that they would immediately close their state's borders to livestock movement at the beginning of an FMD outbreak, particularly if the disease was identified in a bordering state. During an extended FMD outbreak, 29 of 36 (80.6%) SAHOs reported they would resume interstate movement of cattle under some conditions, including enhanced permitting, whereas feedlot veterinarians and managers commonly reported they would be willing to receive cattle from states where no FMD-infected animals were identified, regardless of permit requirements.

CONCLUSIONS AND CLINICAL RELEVANCE

Information gained from this survey can be used to inform disease modeling and preparedness efforts to facilitate business continuity of US beef feedlots in the event of an FMD outbreak in North America.

Assessing potential routes of Streptococcus agalactiae transmission between dairy herds using national surveillance, animal movement data and molecular typing

Streptococcus agalactiae, also known as group B Streptococcus (GBS), is a pathogen of humans and animals. It is an important cause of mastitis in dairy cattle, causing decreased milk quality and quantity. Denmark is the only country to have implemented a national surveillance and control campaign for GBS in dairy cattle. After a significant decline in the 20th century, prevalence has increased in the 21st century. Using a unique combination of national surveillance, cattle movement data and molecular typing, we tested the hypothesis that transmission mechanisms differ between GBS strains that are almost exclusive to cattle and those that affect humans as well as cattle, which would have implications for control recommendations. Three types of S. agalactiae, sequence type (ST) 1, ST23 and ST103 were consistently the most frequent strains among isolates obtained through the national surveillance programme from 2009 to 2011. Herds infected with ST103, which is common in cattle but rarely found in people in Europe, were spatially clustered throughout the study period and across spatial scales. By contrast, strains that are also commonly found in humans, ST1 and ST23, showed no spatial clustering in most or any years of the study, respectively. Introduction of cattle from a positive herd was associated with increased risk of infection by S. agalactiae in the next year (risk ratio of 2.9 and 4.7 for 2009-2010 and 2010-2011, respectively). Moreover, mean exposure to infection was significantly higher for newly infected herds and significantly lower for persistently susceptible herds, as compared to random simulated networks with the same properties, which suggests strong association between the cattle movement network and new infections. At strain-level, new infections with ST1 between 2009 and 2010 were significantly associated with cattle movements, while other strains showed only some degree of association. Sharing of veterinary services, which may serve as proxy for local or regional contacts at a range of scales, was not significantly associated with increased risk of introduction of S. agalactiae or one of the three predominant strains on a farm. Our findings support the reinstatement of restrictions on cattle movements from S. agalactiae positive herds, which came into effect in 2018, but provide insufficient evidence to support strain-specific control recommendations.

Quantitative risk assessment of introduction of BVDV and BoHV-1 through indirect contacts based on implemented biosecurity measures in dairy farms of Spain

A stochastic quantitative risk assessment model was developed to estimate the annual probability of introduction of bovine viral diarrhea virus (BVDV) and bovine herpesvirus 1 (BoHV-1) on 127 dairy farms through indirect contacts. Vehicles transporting calves, cattle to slaughterhouse, dead animals, and mixture of feed, as well as visits by veterinarians and hoof trimmers, farm workers and contacts with neighbors were considered in the model. Data from biosecurity questionnaires of each farm, scientific literature and expert opinion from field veterinarians, animal vehicle drivers, hoof trimmers and personnel from rendering transport companies were used to estimate values for input parameters. Results showed that the annual probability of introducing BVDV or BoHV-1 through indirect contacts was very heterogeneous. The overall distribution of median values for each farm ranged from 0.5 to 14.6% and from 1.0 to 24.9% for BVDV and BoHV-1, respectively. The model identified that providing protective clothing and boots to visits, not allowing the animal vehicle driver to come into contact with animals present on the farm and ensuring that calf vehicles arrived empty, were the measures with the highest impact on the probability of infection for most farms. This model could be a useful tool to show the impact of the measures to farmers and veterinarians, thus increasing their awareness on biosecurity. In addition, it could support decision making on which measures should be prioritized in dairy cattle herds to reduce the probability of introduction of diseases.

Impact of truck contamination and information sharing on foot-and-mouth disease spreading in beef cattle production systems

As cattle movement data in the United States are scarce due to the absence of mandatory traceability programs, previous epidemic models for U.S. cattle production systems heavily rely on contact rates estimated based on expert opinions and survey data. These models are often based on static networks and ignore the sequence of movement, possibly overestimating the epidemic sizes. In this research, we adapt and employ an agent-based model that simulates beef cattle production and transportation in southwest Kansas to analyze the between-premises transmission of a highly contagious disease, foot-and-mouth disease. First, we assess the impact of truck contamination on the disease transmission with the truck agent following an independent clean-infected-clean cycle. Second, we add an information-sharing functionality such that producers/packers can trace back and forward their trade records to inform their trade partners during outbreaks. Scenario analysis results show that including indirect contact routes between premises via truck movements can significantly increase the amplitude of disease spread, compared with equivalent scenarios that only consider animal movement. Mitigation strategies informed by information sharing can effectively mitigate epidemics, highlighting the benefit of promoting information sharing in the cattle industry. In addition, we identify salient characteristics that must be considered when designing an information-sharing strategy, including the number of days to trace back and forward in the trade records and the role of different cattle supply chain stakeholders. Sensitivity analysis results show that epidemic sizes are sensitive to variations in parameters of the contamination period for a truck or a loading/unloading area of premises, and indirect contact transmission probability and future studies can focus on a more accurate estimation of these parameters.

Antimicrobial & antiparasitic use and resistance in British sheep and cattle: a systematic review

A variety of antimicrobials and antiparasitics are used to treat British cattle and sheep to ensure animal welfare, a safe food supply, and maintain farm incomes. However, with increasing global concern about antimicrobial resistance in human and animal populations, there is increased scrutiny of the use of antimicrobials in food-producing animals. This systematic review sought to identify and describe peer and non-peer reviewed sources, published over the last ten years, detailing the usage of, and resistance to, antimicrobials and antiparasitics in sheep and cattle farming systems in Britain as well as identify knowledge gaps. Applying the PRISMA review protocol and guidelines for including grey literature; Scopus, Web of Science, Medline, and government repositories were searched for relevant articles and reports. Seven hundred and seventy titles and abstracts and 126 full-text records were assessed, of which 40 scholarly articles and five government reports were included for data extraction. Antibiotic usage in sheep and cattle in Britain appear to be below the UK average for all livestock and tetracyclines and beta-lactam antibiotics were found to be the most commonly used. However, the poor level of coverage afforded to these species compared to other livestock reduced the certainty of these findings. Although resistance to some antibiotics (using Escherichia coli as a marker) appeared to have decreased in sheep and cattle in England and Wales over a five-year period (2013-2018), levels of resistance remain high to commonly used antibiotics. The small number and fragmented nature of studies identified by this review describing anthelmintic usage, and the lack of available national sales data, prevented the identification of trends in either sheep or cattle. We recommend that additional efforts are taken to collect farm or veterinary level data and argue that extraction of this data is imperative to the development of antimicrobial and antiparasitic resistance strategies in Britain, both of which are needed to reduce usage of these anti-infective agents, curb the development of resistance, and safeguard national agricultural production. Finally, metrics produced by this data should be generated in a way to allow for maximum comparability across species, sectors, and countries.

Review: Application and Prospective Discussion of Machine Learning for the Management of Dairy Farms

Dairy farmers use herd management systems, behavioral sensors, feeding lists, breeding schedules, and health records to document herd characteristics. Consequently, large amounts of dairy data are becoming available. However, a lack of data integration makes it difficult for farmers to analyze the data on their dairy farm, which indicates that these data are currently not being used to their full potential. Hence, multiple issues in dairy farming such as low longevity, poor performance, and health issues remain. We aimed to evaluate whether machine learning (ML) methods can solve some of these existing issues in dairy farming. This review summarizes peer-reviewed ML papers published in the dairy sector between 2015 and 2020. Ultimately, 97 papers from the subdomains of management, physiology, reproduction, behavior analysis, and feeding were considered in this review. The results confirm that ML algorithms have become common tools in most areas of dairy research, particularly to predict data. Despite the quantity of research available, most tested algorithms have not performed sufficiently for a reliable implementation in practice. This may be due to poor training data. The availability of data resources from multiple farms covering longer periods would be useful to improve prediction accuracies. In conclusion, ML is a promising tool in dairy research, which could be used to develop and improve decision support for farmers. As the cow is a multifactorial system, ML algorithms could analyze integrated data sources that describe and ultimately allow managing cows according to all relevant influencing factors. However, both the integration of multiple data sources and the obtainability of public data currently remain challenging.

Multilayer and Multiplex Networks: An Introduction to Their Use in Veterinary Epidemiology

Contact network analysis has become a vital tool for conceptualizing the spread of pathogens in animal populations and is particularly useful for understanding the implications of heterogeneity in contact patterns for transmission. However, the transmission of most pathogens cannot be simplified to a single mode of transmission and, thus, a single definition of contact. In addition, host-pathogen interactions occur in a community context, with many pathogens infecting multiple host species and most hosts being infected by multiple pathogens. Multilayer networks provide a formal framework for researching host-pathogen systems in which multiple types of transmission-relevant interactions, defined as network layers, can be analyzed jointly. Here, we provide an overview of multilayer network analysis and review applications of this novel method to epidemiological research questions. We then demonstrate the use of this technique to analyze heterogeneity in direct and indirect contact patterns amongst swine farms in the United States. When contact among nodes can be defined in multiple ways, a multilayer approach can advance our ability to use networks in epidemiological research by providing an improved approach for defining epidemiologically relevant groups of interacting nodes and changing the way we identify epidemiologically important individuals such as superspreaders.

Illustration of Different Disease Transmission Routes in a Pig Trade Network by Monopartite and Bipartite Representation

Simple Summary

Besides direct animal movements between farms; indirect transmission routes of pathogens can have an immense impact on network structure and disease spread in animal trade networks. This study integrated these indirect transmission routes between farms via transport companies or feed supply as bipartite networks; which were compared to the monopartite animal movements network representing the direct transmission route. Both bipartite networks were projected on farm level to enable a comparison to the monopartite network. The number of edges increased immensely from the monopartite animal movements network to both projected networks. Thus, farms can be highly connected over indirect connections, although they are not directly trading animals. The ranking of the animals according to their centrality parameters, indicating their importance for the network, showed moderate correlations only between the animal movements and the transportation network. The epidemiological models based on the different network representations revealed significantly more infected farms for the networks including indirect transmission routes compared to the direct animal movements. Indirect transmission routes had an immense impact on the outcome of centrality parameters, as well as on the spreading process within the network. This knowledge is needed to understand disease spread and to establish reliable prevention and control measurements.

Abstract

Besides the direct transport of animals, also indirect transmission routes, e.g., contact via contaminated vehicles, have to be considered. In this study, the transmission routes of a German pig trade network were illustrated as a monopartite animal movements network and two bipartite networks including information of the transport company and the feed producer which were projected on farm level (n = 866) to enable a comparison. The networks were investigated with the help of network analysis and formed the basis for epidemiological models to evaluate the impact of different transmission routes on network structure as well as on potential epidemic sizes. The number of edges increased immensely from the monopartite animal movements network to both projected networks. The median centrality parameters revealed clear differences between the three representations. Furthermore, moderate correlation coefficients ranging from 0.55 to 0.68 between the centrality values of the animal movements network and the projected transportation network were obtained. The epidemiological models revealed significantly more infected farms for both projected networks (70% to 100%) compared to the animal movements network (1%). The inclusion of indirect transmission routes had an immense impact on the outcome of centrality parameters as well as on the results of the epidemiological models.

Perception of farmers about endometritis prevention and control measures for zero-grazed dairy cows on smallholder farms in Rwanda

BACKGROUND

Endometritis is a prevalent uterine disease in postpartum cows. The disease reduces fertility performance and milk yield, and subsequently, productivity and profitability of dairy farms. The reduction in performance is associated with considerable economic losses on dairy farms. Smallholder farmers are likely to incur considerable economic losses from the disease where they lack knowledge of effective prevention and control measures for the disease. This study used farmer's perspectives to determine the effectiveness of different management interventions (MIs) for endometritis prevention and control on smallholder farms in Rwanda practicing dairy zero-grazing. The best-worst scaling (BWS) choice method was applied that relied on past 1 year recall data obtained from 154 farmers. These farmers were identified through snowball sampling in a cross-sectional study.

RESULTS

Of the 20 MIs evaluated, 12 scored highly for effectiveness. The top four most effective are: avoiding sharing equipment with neighbouring farms (45.5%), consulting animal health service provider about disease treatment (31.8%), keeping cows in a clean and dry shed (26.7%), and selecting sires based on calving ease (26.6%). The MIs considered least effective were: maintaining clean transition cow housing (35.1%), removal of fetal membrane immediately after passing (33.1%), disinfecting the equipment used in calving assistance before and after use (32.5%), and selecting sires with low percent stillbirths (29.2%).

CONCLUSION

This study has demonstrated the application of BWS object case method in understanding the MIs that farmers consider are most effective in the prevention and control of endometritis disease in the dairy herds. The MIs are on-farm biosecurity and hygiene, seeking veterinary services for disease treatment and selecting sires for ease of calving. These MIs should be considered for prioritization in extension services and research to continuously improve and enhance their practical application on smallholder dairy farms.

Implications of Cattle Trade for the Spread and Control of Infectious Diseases in Slovenia

The objectives of this study were to gain insight into the structure of the cattle trade network in Slovenia and to evaluate the potential for infectious disease spread through movements. The study considered cattle movements between different types of premises that occurred between August 1, 2011 and July 31, 2016 with the exclusion of the movements to the end nodes (e.g., slaughterhouses). In the first part, we performed a static network analysis on monthly and yearly snapshots of the network. These time scales reflect our interest in slowly spreading pathogens; namely Mycobacterium avium subsp. paratuberculosis (MAP), which causes paratuberculosis, a worldwide economically important disease. The results showed consistency in the network measures over time; nevertheless, it was evident that year to year contacts between premises were changing. The importance of individual premises for the network connectedness was highly heterogeneous and the most influential premises in the network were collection centers, mountain pastures, and pastures. Compared to random node removal, targeted removal informed by ranking based on local network measures from previous years was substantially more effective in network disassociation. Inclusion of the latest movement data improved the results. In the second part, we simulated disease spread using a Susceptible-Infectious (SI) model on the temporal network. The SI model was based on the empirically estimated true prevalence of paratuberculosis in Slovenia and four scenarios for probabilities of transmission. Different probabilities were realized by the generation of new networks with the corresponding proportion of contacts which were randomly selected from the original network. These diluted networks served as substrates for simulation of MAP spread. The probability of transmission had a significant influence on the velocity of disease spread through the network. The peaks in daily incidence rates of infected herds were observed at the end of the grazing period. Our results suggest that network analysis may provide support in the optimization of paratuberculosis surveillance and intervention in Slovenia. The approach of simulating disease spread on a diluted network may also be used to model other transmission pathways between herds.

Pig trade networks through live pig markets in Guangdong Province, China

This study used social network analysis to investigate the indirect contact network between counties through the movement of live pigs through four wholesale live pig markets in Guangdong Province, China. All 14,118 trade records for January and June 2016 were collected from the markets and the patterns of pig trade in these markets analysed. Maps were developed to show the movement pathways. Evaluating the network between source counties was the primary objective of this study. A 1‐mode network was developed. Characteristics of the trading network were explored, and the degree, betweenness and closeness were calculated for each source county. Models were developed to compare the impacts of different disease control strategies on the potential magnitude of an epidemic spreading through this network. The results show that pigs from 151 counties were delivered to the four wholesale live pig markets in January and/or June 2016. More batches (truckloads of pigs sourced from one or more piggeries) were traded in these markets in January (8,001) than in June 2016 (6,117). The pigs were predominantly sourced from counties inside Guangdong Province (90%), along with counties in Hunan, Guangxi, Jiangxi, Fujian and Henan provinces. The major source counties (46 in total) contributed 94% of the total batches during the two‐month study period. Pigs were sourced from piggeries located 10 to 1,417 km from the markets. The distribution of the nodes' degrees in both January and June indicates a free‐scale network property, and the network in January had a higher clustering coefficient (0.54 vs. 0.39) and a shorter average pathway length (1.91 vs. 2.06) than that in June. The most connected counties of the network were in the central, northern and western regions of Guangdong Province. Compared with randomly removing counties from the network, eliminating counties with higher betweenness, degree or closeness resulted in a greater reduction of the magnitude of a potential epidemic. The findings of this study can be used to inform targeted control interventions for disease spread through this live pig market trade network in south China.

When resolution does matter: Modelling indirect contacts in dairy farms at different levels of detail

Animal exchanges are considered the major pathway for between-farm transmission of many livestock infectious diseases. Yet, vehicles and operators visiting several farms during routine activities can also contribute to disease spread. Indeed, if contaminated, they can act as mechanical vectors of fomites, generating indirect contacts between visited farms. While data on animal exchanges is often available in national databases, information about the daily itineraries of trucks and operators is rare because difficult to obtain. Thus, some unavoidable approximations have been frequently introduced in the description of indirect contacts in epidemic models. Here, we showed that the level of detail in such description can significantly affect the predictions on disease dynamics. Our analyses focused on the potential spread of a disease in a dairy farm system subject of a comprehensive data collection campaign on calf transportations. We developed two temporal multilayer networks to model between-farm contacts generated by either animal exchanges (direct contacts) and connections operated by trucks moving calves (indirect contacts). The complete model used the full knowledge of the daily trucks' itineraries, while the partial informed one used only a subset of such available information. To account for various conditions of pathogen survival ability and effectiveness of cleaning operations, we performed a sensitivity analysis on trucks' contamination period. An accurate description of indirect contacts was crucial both to correctly predict the final size of epidemics and to identify the seed farms responsible for generating the most severe outbreaks. The importance of detailed information emerged even more clearly in the case of short contamination periods. Our conclusions could be extended to between-farm contacts generated by other vehicles and operators. Overcoming these information gaps would be decisive for a deeper understanding of epidemic spread in livestock and to develop effective control plans.

Analysing livestock network data for infectious disease control: an argument for routine data collection in emerging economies

Livestock movements are an important mechanism of infectious disease transmission. Where these are well recorded, network analysis tools have been used to successfully identify system properties, highlight vulnerabilities to transmission, and inform targeted surveillance and control. Here we highlight the main uses of network properties in understanding livestock disease epidemiology and discuss statistical approaches to infer network characteristics from biased or fragmented datasets. We use a 'hurdle model' approach that predicts (i) the probability of movement and (ii) the number of livestock moved to generate synthetic 'complete' networks of movements between administrative wards, exploiting routinely collected government movement permit data from northern Tanzania. We demonstrate that this model captures a significant amount of the observed variation. Combining the cattle movement network with a spatial between-ward contact layer, we create a multiplex, over which we simulated the spread of 'fast' ( R0 = 3) and 'slow' ( R0 = 1.5) pathogens, and assess the effects of random versus targeted disease control interventions (vaccination and movement ban). The targeted interventions substantially outperform those randomly implemented for both fast and slow pathogens. Our findings provide motivation to encourage routine collection and centralization of movement data to construct representative networks. This article is part of the theme issue 'Modelling infectious disease outbreaks in humans, animals and plants: epidemic forecasting and control'. This theme issue is linked with the earlier issue 'Modelling infectious disease outbreaks in humans, animals and plants: approaches and important themes'.

Contact chains of cattle farms in Great Britain

Network analyses can assist in predicting the course of epidemics. Time-directed paths or 'contact chains' provide a measure of host-connectedness across specified timeframes, and so represent potential pathways for spread of infections with different epidemiological characteristics. We analysed networks and contact chains of cattle farms in Great Britain using Cattle Tracing System data from 2001 to 2015. We focused on the potential for between-farm transmission of bovine tuberculosis, a chronic infection with potential for hidden spread through the network. Networks were characterized by scale-free type properties, where individual farms were found to be influential 'hubs' in the network. We found a markedly bimodal distribution of farms with either small or very large ingoing and outgoing contact chains (ICCs and OCCs). As a result of their cattle purchases within 12-month periods, 47% of British farms were connected by ICCs to more than 1000 other farms and 16% were connected to more than 10 000 other farms. As a result of their cattle sales within 12-month periods, 66% of farms had OCCs that reached more than 1000 other farms and 15% reached more than 10 000 other farms. Over 19 000 farms had both ICCs and OCCs reaching more than 10 000 farms for two or more years. While farms with more contacts in their ICCs or OCCs might play an important role in disease spread, farms with extensive ICCs and OCCs might be particularly important by being at higher risk of both acquiring and disseminating infections.

Identifying outbreaks of Porcine Epidemic Diarrhea virus through animal movements and spatial neighborhoods

The spread of pathogens in swine populations is in part determined by movements of animals between farms. However, understanding additional characteristics that predict disease outbreaks and uncovering landscape factors related to between-farm spread are crucial steps toward risk mitigation. This study integrates animal movements with environmental risk factors to identify the occurrence of porcine epidemic diarrhea virus (PEDV) outbreaks. Using weekly farm-level incidence data from 332 sow farms, we applied machine-learning algorithms to quantify associations between risk factors and PEDV outbreaks with the ultimate goal of training predictive models and to identify the most important factors associated with PEDV occurrence. Our best algorithm was able to correctly predict whether an outbreak occurred during one-week periods with >80% accuracy. The most important predictors included pig movements into neighboring farms. Other important neighborhood attributes included hog density, environmental and weather factors such as vegetation, wind speed, temperature, and precipitation, and topographical features such as slope. Our neighborhood-based approach allowed us to simultaneously capture disease risks associated with long-distance animal movement as well as local spatial dynamics. The model presented here forms the foundation for near real-time disease mapping and will advance disease surveillance and control for endemic swine pathogens in the United States.

Characterization of swine movements in the United States and implications for disease control

Understanding between-farm movement patterns is an essential component in developing effective surveillance and control programs in livestock populations. Quantitative knowledge on movement patterns is particularly important for the commercial swine industry, in which large numbers of pigs are frequently moved between farms. Here, we described the annual movement patterns between swine farms in three production systems of the United States and identified farms that may be targeted to increase the efficacy of infectious disease control strategies. Research results revealed a high amount of variability in movement patterns across production systems, indicating that quantities captured from one production system and applied to another may lead to invalid estimations of disease spread. Furthermore, we showed that targeting farms based on their mean infection potential, a metric that captured the temporal sequence of movements, substantially reduced the potential for transmission of an infectious pathogen in the contact network and performed consistently well across production systems. Specifically, we found that by targeting farms based on their mean infection potential, we could reduce the potential spread of an infectious pathogen by 80% when removing approximately 25% of farms in each of the production systems. Whereas other metrics, such as degree, required 26-35% of farms to be removed in two of the production systems to reach the same outcome; this outcome was not achievable in one of the production systems. Our results demonstrate the importance of fine-scale temporal movement data and the need for in-depth understanding of the contact structure in developing more efficient disease surveillance and response strategies in swine production systems.

The spread of bovine tuberculosis in Canadian shared pastures: Data, model, and simulations

Bovine tuberculosis (bTB), caused by Mycobacterium bovis, is a chronic disease typical of cattle. Nonetheless, it can affect many mammals including humans, making it one of the most widespread zoonotic diseases worldwide. In industrialized countries, the main pathways of introduction of bTB into a herd are animal trade and contact with infected wildlife. In addition, for slow-spreading diseases with a long latent period such as bTB, shared seasonal pastures might be a between-herd transmission pathway, indeed farmers might unknowingly send infected animals to the pasture, since clinical signs are rarely evident in early infection. In this study, we developed a dynamic stochastic model to represent the spread of bTB in pastures. This was tailored to Canadian cow-calf herds, as we calibrated the model with data sourced from a recent bTB outbreak in Western Canada. We built a model for a herd with seasonal management, characterized by its partition into a group staying in the main facility and the remaining group(s) moving to summer pastures. We used this model to estimate the time of the first introduction of bTB into the herd. Furthermore, we expanded the model to include herds categorized as high-risk contacts with the index herd, in order to estimate the potential for disease spread on shared pastures. Finally, we explored two control scenarios to be applied to high-risk farms after the outbreak detection. Our results showed that the first introduction likely happened 3 to 5 years prior to the detection of the index herd, and the probability of bTB spreading in pastures was low, but not negligible. Nevertheless, the surveillance system currently in place was effective to detect potential outbreaks.

A descriptive analysis of swine movements in Ontario (Canada) as a contributor to disease spread

In recent times, considerable efforts have been made to develop infrastructure and processes of tracing livestock movements. One of common use of this type of data is to assess the potential for spread of infections in source populations. The objectives of this research were to describe Ontario pig movements in 2015, and to understand the potential for disease transmission through animal movement on a weekly and yearly basis. Swine shipments from January to December 2015 represented 224 production facilities and a total of 5398 unique animal movements. This one-mode directed network of animal movements was then analyzed using common descriptive network measures. The maximum yearly (_y) weak component (WC_y) size and maximum weekly (_w) weak component size (WC_w) was 224 facilities, and 83 facilities, respectively. The maximum WC_w did not change significantly (p > 0.05) over time. The maximum strong component (SC) consisted of two facilities both on a weekly, and on a yearly basis. The size of the maximum ingoing contact chain on a yearly basis (ICC_y) was 173 nodes with one abattoir as the end point, and the maximum ICC_w consisted of 53 nodes. The size of the maximum outgoing contact chain (OCC_y) contained 79 nodes, with one sow herd as a starting point. The maximum OCC_w was 6 nodes. Regression models resulted in significant quadratic associations between weekly count of finisher facilities with betweenness >0 (p = 0.02) and weekly count of finisher facilities with in-degree and out-degree >0 (p = 0.01) and week number. Higher weekly counts of nursery and finisher facilities with betweenness >0 and in-degree and out-degree both >0 values occurred during summer months. All study facilities were connected when direction of animal movement was not taken into consideration in the yearly network. As such, yearly networks are potentially representative of infections with long incubation periods, subclinical infections, or endemic infections for which active control measures have not being taken. When the direction of animal movement was considered, such infection could still spread substantially and affect 35% of the study population (79/224). In the study population, finisher sites were proportionally and consistently most represented in WC_w (min = 51%, max = 78%), which reflects current Ontario herd demographics. However, abattoirs were over-represented when the number of facilities in the study population was taken into consideration. This, and the size of the maximum ICC_w both suggest that abattoirs could be, at least for some infectious diseases, suitable establishments for targeted sampling.

Modelling farm-to-farm disease transmission through personnel movements: from visits to contacts, and back

Infectious diseases in livestock can be transmitted through fomites: objects able to convey infectious agents. Between-farm spread of infections through fomites is mostly due to indirect contacts generated by on-farm visits of personnel that can carry pathogens on their clothes, equipment, or vehicles. However, data on farm visitors are often difficult to obtain because of the heterogeneity of their nature and privacy issues. Thus, models simulating disease spread between farms usually rely on strong assumptions about the contribution of indirect contacts on infection spread. By using data on veterinarian on-farm visits in a dairy farm system, we built a simple simulation model to assess the role of indirect contacts on epidemic dynamics compared to cattle movements (i.e. direct contacts). We showed that including in the simulation model only specific subsets of the information available on indirect contacts could lead to outputs widely different from those obtained with the full-information model. Then, we provided a simple preferential attachment algorithm based on the probability to observe consecutive on-farm visits from the same operator that allows overcoming the information gaps. Our results suggest the importance of detailed data and a deeper understanding of visit dynamics for the prevention and control of livestock diseases.