The influence of empirical contact networks on modelling diseases in cattle

Analysis of within-pen and between-pen fenceline temporal contact networks in confined feedlot cattle

Though contact networks are important for describing the dynamics for disease transmission and intervention applications, individual animal contact and barriers between animal populations, such as fences, are not often utilized in the construction of these models. The objective of this study was to use contact network analysis to quantify contacts within two confined pens of feedlot cattle and the shared "fenceline" area between the pens at varying temporal resolutions and contact duration to better inform the construction of network-based disease transmission models for cattle within confined-housing systems. Two neighboring pens of feedlot steers were tagged with Real-Time Location System (RTLS) tags. Within-pen contacts were defined with a spatial threshold (SpTh) of 0.71 m and a minimum contact duration (MCD) of either 10 seconds (10 s), 30 seconds (30 s), or 60 seconds (60 s). For the fenceline network location readings were included within an area extending from 1 m on either side of the shared fence. "Fenceline" contacts could only occur between a steer from each pen. Static, undirected, weighted contact networks for within-pen networks and the between-pen network were generated for the full study duration and for daily (24-h), 6-h period, and hourly networks to better assess network heterogeneity. For the full study duration network, the two within-pen networks were densely homogenous. The within-pen networks showed more heterogeneity when smaller timescales (6-h period and hourly) were applied. When contacts were defined with a MCD of 30 s or 60 s, the total number of contacts seen in each network decreased, indicating that most of the contacts observed in our networks may have been transient passing contacts. Cosine similarity was moderate and stable across days for within pen networks. Of the 90 total tagged steers between the two pens, 86 steers (46 steers from Pen 2 and 40 steers from Pen 3) produced at least one contact across the shared fenceline. The total network density for the network created across the shared fenceline between the two pens was 17%, with few contacts at shorter timescales and for MCD of 30 s or 60 s. Overall, the contact networks created here from high-resolution spatial and temporal contact observation data provide estimates for a contact network within commercial US feedlot pens and the contact network created between two neighboring pens of cattle. These networks can be used to better inform pathogen transmission models on social contact networks.

Quantifying the relationship between within-host dynamics and transmission for viral diseases of livestock

Understanding the population dynamics of an infectious disease requires linking within-host dynamics and between-host transmission in a quantitative manner, but this is seldom done in practice. Here a simple phenomenological model for viral dynamics within a host is linked to between-host transmission by assuming that the probability of transmission is related to log viral titre. Data from transmission experiments for two viral diseases of livestock, foot-and-mouth disease virus in cattle and swine influenza virus in pigs, are used to parametrize the model and, importantly, test the underlying assumptions. The model allows the relationship between within-host parameters and transmission to be determined explicitly through their influence on the reproduction number and generation time. Furthermore, these critical within-host parameters (time and level of peak titre, viral growth and clearance rates) can be computed from more complex within-host models, raising the possibility of assessing the impact of within-host processes on between-host transmission in a more detailed quantitative manner.

Bovine tuberculosis in youngstock cattle: A narrative review

Bovine tuberculosis (bTB), caused by Mycobacterium bovis, remains a high-priority global pathogen of concern. The role of youngstock animals in the epidemiology of bTB has not been a focus of contemporary research. Here we have aimed to collate and summarize what is known about the susceptibility, diagnosis, transmission (infectiousness), and epidemiology to M. bovis in youngstock (up to 1-year of age). Youngstock are susceptible to M. bovis infection when exposed, with the capacity to develop typical bTB lesions. Calves can be exposed through similar routes as adults, via residual infection, contiguous neighborhood spread, wildlife spillback infection, and the buying-in of infected but undetected cattle. Dairy systems may lead to greater exposure risk to calves relative to other production systems, for example, via pooled milk. Given their young age, calves tend to have shorter bTB at-risk exposure periods than older cohorts. The detection of bTB varies with age when using a wide range of ante-mortem diagnostics, also with post-mortem examination and confirmation (histological and bacteriological) of infection. When recorded as positive by ante-mortem test, youngstock appear to have the highest probabilities of any age cohort for confirmation of infection post-mortem. They also appear to have the lowest false negative bTB detection risk. In some countries, many calves are moved to other herds for rearing, potentially increasing inter-herd transmission risk. Mathematical models suggest that calves may also experience lower force of infection (the rate that susceptible animals become infected). There are few modeling studies investigating the role of calves in the spread and maintenance of infection across herd networks. One study found that calves, without operating testing and control measures, can help to maintain infection and lengthen the time to outbreak eradication. Policies to reduce testing for youngstock could lead to infected calves remaining undetected and increasing onwards transmission. Further studies are required to assess the risk associated with changes to testing policy for youngstock in terms of the impact for within-herd disease control, and how this may affect the transmission and persistence of infection across a network of linked herds.

Quantifying changes in the British cattle movement network

The modelling of disease spread is crucial to the farming industry and policy makers. In some of these industries, excellent data exist on animal movements, along with the networks that these movements create, and allow researchers to model spread of disease (both epidemic and endemic). The Cattle Tracing System is an online recording system for cattle births, deaths and between-herd movements in the United Kingdom and is an excellent resource for any researchers interested in networks or modelling infectious disease spread through the UK cattle system. Data exist that cover many years, and it can be useful to know how much change is occurring in a network, to help judge the merit of using historical data within a modelling context. This article uses the data to construct weighted directed monthly movement networks for two distinct periods of time, 2004-2006 and 2015-2017, to quantify by how much the underlying structure of the network has changed. Substantial changes in network structure may influence policy-makers directly or may influence models built upon the network data, and these in turn could impact policy-makers and their assessment of risk. We examined 13 network metrics, ranging from general descriptive metrics such as total number of nodes with movements and total movements, through to metrics to describe the network (e.g., Giant weakly and strongly connected components) and metrics calculated per node (betweenness, degree and strength). Mixed effect models show that there is a statistically significant effect of the period (2004-2006 vs 2015-2017) in the values of nine of the 13 network metrics. For example median total degree decreased by 19%. In addition to examining networks for two time periods, two updates of the data were examined to determine by how much the movement data stored for 2004-2006 had been cleansed between updates. Examination of these updates shows that there are small decreases in problem movements (such as animals leaving slaughterhouses) and therefore evidence of historical data being improved between updates. In combination with the significant effect of period on many of the network metrics, the modification of data between updates provides further evidence that the most recent available data should be used for network modelling. This will ensure that the most representative descriptions of the network are available to provide accurate modelling results to best inform policy makers.

Spatial and temporal variation in proximity networks of commercial dairy cattle in Great Britain

The nature of contacts between hosts can be important in facilitating or impeding the spread of pathogens within a population. Networks constructed from contacts between hosts allow examination of how individual variation might influence the spread of infections. Studying the contact networks of livestock species managed under different conditions can additionally provide insight into their influence on these contact structures. We collected high-resolution proximity and GPS location data from nine groups of domestic cattle (mean group size = 85) in seven dairy herds employing a range of grazing and housing regimes. Networks were constructed from cattle contacts (defined by proximity) aggregated by different temporal windows (2 h, 24 h, and approximately 1 week) and by location within the farm. Networks of contacts aggregated over the whole study were highly saturated but dividing contacts by space and time revealed substantial variation in cattle interactions. Cows showed statistically significant variation in the frequency of their contacts and in the number of cows with which they were in contact. When cows were in buildings, compared to being on pasture, contact durations were longer and cows contacted more other cows. A small number of cows showed evidence of consistent relationships but the majority of cattle did not. In one group where management allowed free access to all farm areas, cows showed asynchronous space use and, while at pasture, contacted fewer other cows and showed substantially greater between-individual variation in contacts than other groups. We highlight the degree to which variations in management (e.g. grazing access, milking routine) substantially alter cattle contact patterns, with potentially major implications for infection transmission and social interactions. In particular, where individual cows have free choice of their environment, the resulting contact networks may have a less-risky structure that could reduce the likelihood of direct transmission of infections.

Spatial and network characteristics of Irish cattle movements

Our aim was to examine, for the first time, the spatial and network characteristics of cattle movements between herds in the Republic of Ireland (ROI), to inform policy and research of relevance to the surveillance and management of disease in Irish cattle. We analysed movements in 2016 as discrete herd to herd pairings (degree), herd to herd pairings by date of move (contacts) and herd to herd pairings by date and individual animal (transfers), and looked at each of these as movements out of a herd (out degree, out contacts, out transfers) and into a herd (in degree, in contacts, in transfers). We found that the frequency distributions, by herd, of these six move types were all heavily right skewed but in the case of the 'out' data types more closely followed a log-normal than the scale free distribution often reported for livestock movement data. For each distinct herd to herd contact in a given direction, over 90 % occurred only once, whereas the maximum number of occurrences was 62. Herd-level Spearman rank correlations between inward moves (represented as in degree, in contacts, in transfers) and outward moves (out degree, out contacts, out transfers) were weak or even negative whereas correlations between different measures of outward moves or inward moves (e.g. out degree vs. out contacts, in transfers vs. in degree) were stronger. Correlations between these variables and the network measure betweenness varied between r = 0.513 and r = 0.587. Some herds took part in a relatively large number of movements whilst also retaining their cattle for long periods (> 100 days) between moves. In and out degree, contacts and transfers were mapped across Ireland on a 5 km grid, and additionally normalized per 1000 animals and per herd. We found considerable variation in the number of movements by county. Approximately half of transfers were conducted within a single county, but the number and distance of between county movements varied considerably by county of origin and county of destination, with the proportion of moves completed within a single county correlated with its size. Herds exchanging cattle via a market were generally further apart than when moves were made directly herd to herd. For contacts, the distances moved away from the herd were on average greater for origin herds in the west of ROI whereas distances moved to a herd were generally greater for destination herds in the centre-east and the north-west.

Considering appropriate replication in the design of animal social network studies

Social network analysis has increasingly been considered a useful tool to interpret the complexity of animal social relationships. However, group composition can affect the contact structure of the network resulting in variation between networks. Replication in contact network studies is rarely done but enables determination of possible variation in response across networks. Here we explore the importance of between-group variability in social behaviour and the impact of replication on hypothesis testing. We use an exemplar study of social contact data collected from six replicated networks of cattle before and after the application of a social disturbance treatment. In this replicated study, subtle but consistent changes in animal contact patterns were detected after the application of a social disturbance treatment. We then quantify both within- and between-group variation in this study and explore the importance of varying the number of replicates and the number of individuals within each network, on the precision of the differences in treatment effects for the contact behaviour of the resident cattle. The analysis demonstrates that reducing the number of networks observed in the study would reduce the probability of detecting treatment differences for social behaviours even if the total number of animals was kept the same.

Transmission on empirical dynamic contact networks is influenced by data processing decisions

Dynamic contact data can be used to inform disease transmission models, providing insight into the dynamics of infectious diseases. Such data often requires extensive processing for use in models or analysis. Therefore, processing decisions can potentially influence the topology of the contact network and the simulated disease transmission dynamics on the network. In this study, we examine how four processing decisions, including temporal sampling window (TSW), spatial threshold of contact (SpTh), minimum contact duration (MCD), and temporal aggregation (daily or hourly) influence the information content of contact data (indicated by changes in entropy) as well as disease transmission model dynamics. We found that changes made to information content by processing decisions translated to significant impacts to the transmission dynamics of disease models using the contact data. In particular, we found that SpTh had the largest independent influence on information content, and that some output metrics (R₀, time to peak infection) were more sensitive to changes in information than others (epidemic extent). These findings suggest that insights gained from transmission modeling using dynamic contact data can be influenced by processing decisions alone, emphasizing the need to carefully consideration them prior to using contact-based models to conduct analyses, compare different datasets, or inform policy decisions.

Development of a network based model to simulate the between-farm transmission of the porcine reproductive and respiratory syndrome virus

Contact structure within a population can significantly affect the outcomes of infectious disease spread models. The objective of this study was to develop a network based simulation model for the between-farm spread of porcine reproductive and respiratory syndrome virus to assess the impact of contact structure on between-farm transmission of PRRS virus. For these farm level models, a hypothetical population of 500 swine farms following a multistage production system was used. The contact rates between farms were based on a study analyzing movement of pigs in Canada, while disease spread parameters were extracted from published literature. Eighteen distinct scenarios were designed and simulated by varying the mode of transmission (direct versus direct and indirect contact), type of index herd (farrowing, nursery and finishing), and the presumed network structures among swine farms (random, scale-free and small-world). PRRS virus was seeded in a randomly selected farm and 500 iterations of each scenario were simulated for 52 weeks. The median epidemic size by the end of the simulated period and percentage die-out for each scenario, were the key outcomes captured. Scenarios with scale-free network models resulted in the largest epidemic sizes, while scenarios with random and small-world network models resulted in smaller and similar epidemic sizes. Similarly, stochastic die-out percentage was least for scenarios with scale-free networks followed by random and small-world networks. Findings of the study indicated that incorporating network structures among the swine farms had a considerable impact on the spread of PRRS virus, highlighting the importance of understanding and incorporating realistic contact structures when developing infectious disease spread models for similar populations.

Replicating disease spread in empirical cattle networks by adjusting the probability of infection in random networks

Comparisons between mass-action or "random" network models and empirical networks have produced mixed results. Here we seek to discover whether a simulated disease spread through randomly constructed networks can be coerced to model the spread in empirical networks by altering a single disease parameter - the probability of infection. A stochastic model for disease spread through herds of cattle is utilised to model the passage of an SEIR (susceptible-latent-infected-resistant) through five networks. The first network is an empirical network of recorded contacts, from four datasets available, and the other four networks are constructed from randomly distributed contacts based on increasing amounts of information from the recorded network. A numerical study on adjusting the value of the probability of infection was conducted for the four random network models. We found that relative percentage reductions in the probability of infection, between 5.6% and 39.4% in the random network models, produced results that most closely mirrored the results from the empirical contact networks. In all cases tested, to reduce the differences between the two models, required a reduction in the probability of infection in the random network.

Bovine tuberculosis: within-herd transmission models to support and direct the decision-making process

Use of mathematical models to study the transmission dynamics of infectious diseases is becoming increasingly common in veterinary sciences. However, modeling chronic infectious diseases such as bovine tuberculosis (bTB) is particularly challenging due to the substantial uncertainty associated with the epidemiology of the disease. Here, the methodological approaches used to model bTB and published in the peer-reviewed literature in the last decades were reviewed with a focus on the impact that the models' assumptions may have had on their results, such as the assumption of density vs. frequency-dependent transmission, the existence of non-infectious and non-detectable stages, and the effect of extrinsic sources of infection (usually associated with wildlife reservoirs). Although all studies suggested a relatively low rate of within-herd transmission of bTB when test-and-cull programs are in place, differences in the estimated length of the infection stages, sensitivity and specificity of the tests used and probable type of transmission (density or frequency dependent) were observed. Additional improvements, such as exploring the usefulness of contact-networks instead of assuming homogeneous mixing of animals, may help to build better models that can help to design, evaluate and monitor control and eradication strategies against bTB.

The risk of disease to great apes: simulating disease spread in orang-utan (Pongo pygmaeus wurmbii) and chimpanzee (Pan troglodytes schweinfurthii) association networks

All great ape species are endangered, and infectious diseases are thought to pose a particular threat to their survival. As great ape species vary substantially in social organisation and gregariousness, there are likely to be differences in susceptibility to disease types and spread. Understanding the relation between social variables and disease is therefore crucial for implementing effective conservation measures. Here, we simulate the transmission of a range of diseases in a population of orang-utans in Sabangau Forest (Central Kalimantan) and a community of chimpanzees in Budongo Forest (Uganda), by systematically varying transmission likelihood and probability of subsequent recovery. Both species have fission-fusion social systems, but differ considerably in their level of gregariousness. We used long-term behavioural data to create networks of association patterns on which the spread of different diseases was simulated. We found that chimpanzees were generally far more susceptible to the spread of diseases than orang-utans. When simulating different diseases that varied widely in their probability of transmission and recovery, it was found that the chimpanzee community was widely and strongly affected, while in orang-utans even highly infectious diseases had limited spread. Furthermore, when comparing the observed association network with a mean-field network (equal contact probability between group members), we found no major difference in simulated disease spread, suggesting that patterns of social bonding in orang-utans are not an important determinant of susceptibility to disease. In chimpanzees, the predicted size of the epidemic was smaller on the actual association network than on the mean-field network, indicating that patterns of social bonding have important effects on susceptibility to disease. We conclude that social networks are a potentially powerful tool to model the risk of disease transmission in great apes, and that chimpanzees are particularly threatened by infectious disease outbreaks as a result of their social structure.

Highly dynamic animal contact network and implications on disease transmission

Contact patterns among hosts are considered as one of the most critical factors contributing to unequal pathogen transmission. Consequently, networks have been widely applied in infectious disease modeling. However most studies assume static network structure due to lack of accurate observation and appropriate analytic tools. In this study we used high temporal and spatial resolution animal position data to construct a high-resolution contact network relevant to infectious disease transmission. The animal contact network aggregated at hourly level was highly variable and dynamic within and between days, for both network structure (network degree distribution) and individual rank of degree distribution in the network (degree order). We integrated network degree distribution and degree order heterogeneities with a commonly used contact-based, directly transmitted disease model to quantify the effect of these two sources of heterogeneity on the infectious disease dynamics. Four conditions were simulated based on the combination of these two heterogeneities. Simulation results indicated that disease dynamics and individual contribution to new infections varied substantially among these four conditions under both parameter settings. Changes in the contact network had a greater effect on disease dynamics for pathogens with smaller basic reproduction number (i.e. R₀ < 2).

An integrated web system to support veterinary activities in Italy for the management of information in epidemic emergencies

The management of public health emergencies is improved by quick, exhaustive and standardized flow of data on disease outbreaks, by using specific tools for data collection, registration and analysis. In this context, the National Information System for the Notification of Outbreaks of Animal Diseases (SIMAN) has been developed in Italy to collect and share data on the notifications of outbreaks of animal diseases. SIMAN is connected through web services to the national database of animals and holdings (BDN) and has been integrated with tools for the management of epidemic emergencies. The website has been updated with a section dedicated to the contingency planning in case of epidemic emergency. EpiTrace is one such useful tool also integrated in the BDN and based on the Social Network Analysis (SNA) and on network epidemiological models. This tool gives the possibility of assessing the risk associated to holdings and animals on the basis of their trade, in order to support the veterinary services in tracing back and forward the animals in case of outbreaks of infectious diseases.

Temporal-spatial heterogeneity in animal-environment contact: implications for the exposure and transmission of pathogens

Contact structure, a critical driver of infectious disease transmission, is not completely understood and characterized for environmentally transmitted pathogens. In this study, we assessed the effects of temporal and spatial heterogeneity in animal contact structures on the dynamics of environmentally transmitted pathogens. We used real-time animal position data to describe contact between animals and specific environmental areas used for feeding and watering calves. The generated contact structure varied across days and among animals. We integrated animal and environmental heterogeneity into an agent-based simulation model for Escherichia coli O157 environmental transmission in cattle to simulate four different scenarios with different environmental bacteria concentrations at different areas. The simulation results suggest heterogeneity in environmental contact structure among cattle influences pathogen prevalence and exposure associated with each environment. Our findings suggest that interventions that target environmental areas, even relatively small areas, with high bacterial concentration can result in effective mitigation of environmentally transmitted pathogens.