Combining Salmonella Dublin genome information and contact-tracing to substantiate a new approach for improved detection of infectious transmission routes in cattle populations.

This study presents a new method for detection of between-herd livestock movements to facilitate disease tracing and more accurately describe network behaviour of relevance for spread of infectious diseases, including within-livestock business risk-carrying contacts that are not necessarily recorded anywhere. The study introduces and substantiates the concept of grouping livestock herds into business-units based on ownership and location in the tracing analysis of animal movement-based contact networks. To test the utility of this approach, whole core genome sequencing of 196 Salmonella Dublin isolates stored from previous surveillance and project activities was combined with information on cattle movements recorded in the Danish Cattle Database between 1997 and 2017. The aim was to investigate alternative explanations for S. Dublin circulation in groups of herds connected by ownership, but without complete records of livestock movements. The EpiContactTrace R-package was used to trace the contact networks between businesses and compare the network characteristics of businesses sharing strains of S. Dublin with different levels of genetic relatedness. The ownership-only definition proved to be an unreliable grouping approach for large businesses, which could have internal distances larger than 250 km and therefore do not represent useful epidemiological units. Therefore, the grouping was refined using spatial analysis. More than 90% of final business units formed were composed of one single cattle property, whereas multi-property businesses could reach up to eight properties in a given year, with up to 15 cattle herds having been part of the same business through the study period. Results showed markedly higher probabilities of introduction of infectious animals between proposed businesses from which the same clone of S. Dublin had been isolated, when compared to businesses with non-related strains, thus substantiating the business-unit as an important epidemiological feature to consider in contact network analysis and tracing of infection routes. However, this approach may overestimate real-life contacts between cattle properties and putatively overestimate the degree of risk-contacts within each business, since it is based solely on information about property ownership and location. This does not consider administrative and individual farmers behaviours that essentially keep two properties separated. Despite this, we conclude that defining epidemiological units based on businesses is a promising approach for future disease tracing tasks.

Cost-effectiveness of two different protocols for animal tracing investigations of bovine tuberculosis outbreaks in France.

In the French bovine tuberculosis (bTB) surveillance program, tracing-on and back investigations have a major importance as, in 2016, they represented about 21 % of the detected outbreaks. Building on our previous work on the other surveillance system components (Poirier et al., 2019), we evaluated for the first time the sensitivity and the cost of the two existing protocols of bTB's tracing-on investigations trough scenario tree modelling with a stochastic approach. We used French databases (national database for bovine identification and database recording all bTB surveillance and control results) and direct and indirect costs collected in a previous study. These assessments allowed us to calculate the cost-effectiveness index (cost/sensitivity) of each tracing-on protocol. In the first protocol (trace-and-cull protocol), the animal(s) linking the farm to an outbreak are systematically culled for bacteriology, PCR and histology testing. In the second protocol (trace-and-test protocol), the traced animal is culled only if it had non-negative result to an intradermal cervical comparative tuberculin test (ICCT). We estimated herd sensitivity of the two tracing-on protocols for 12 herd types defined by their production type, size and herd turnover. For the trace-and-cull protocol, mean herd sensitivity was estimated between 67.3 % [66.8-67.7]CI95 % and 89.2 % [88.7-89.7]CI95 % and between 51.2 % [50.8-51.5]CI95 % and 73.1 % [72.6-73.6]CI95 % for the trace-and-test protocol, depending on herd type. The trace-and-cull protocol was between 278 €/herd and 717 €/herd more expensive than the trace-and-test protocol, depending on herd type. Regardless of herd type, the trace-and-cull protocol had the smaller cost/sensitivity ratio and was therefore the most cost-effective protocol. That work showed that systematically culling traced animals to perform bacteriology and PCR on them (trace-and-cull protocol) is associated with a better herd sensitivity and is more cost-effective for all herd types. That is consistent with French veterinary authorities' recommendations but does not account for sociological aspects such as the bond between the farmer and his animals. Yet, cost-effectiveness difference was minor in small dairy and beef herds with a low turnover, suggesting the protocol could be chosen depending on the epidemiological context in those herds.

Emergence and Containment of Canine Influenza Virus A(H3N2), Ontario, Canada, 2017-2018.

Canine influenza virus (CIV) A(H3N2) was identified in 104 dogs in Ontario, Canada, during December 28, 2017-October 30, 2018, in distinct epidemiologic clusters. High morbidity rates occurred within groups of dogs, and kennels and a veterinary clinic were identified as foci of infection. Death attributable to CIV infection occurred in 2 (2%) of 104 diagnosed cases. A combination of testing of suspected cases, contact tracing and testing, and 28-day isolation of infected dogs was used, and CIV transmission was contained in each outbreak. Dogs recently imported from Asia were implicated as the source of infection. CIV H3N2 spread rapidly within groups in this immunologically naive population; however, containment measures were apparently effective, demonstrating the potential value of prompt diagnosis and implementation of CIV control measures.

Comparing the effects of non-homogenous mixing patterns on epidemiological outcomes in equine populations: A mathematical modelling study.

Disease transmission models often assume homogenous mixing. This assumption, however, has the potential to misrepresent the disease dynamics for populations in which contact patterns are non-random. A disease transmission model with an SEIR structure was used to compare the effect of weighted and unweighted empirical equine contact networks to weighted and unweighted theoretical networks generated using random mixing. Equine influenza was used as a case study. Incidence curves generated with the unweighted empirical networks were similar in epidemic duration (5-8 days) and peak incidence (30.8-46.4%). In contrast, the weighted empirical networks resulted in a more pronounced difference between the networks in terms of the epidemic duration (8-15 days) and the peak incidence (5-25%). The incidence curves for the empirical networks were bimodal, while the incidence curves for the theoretical networks were unimodal. The incorporation of vaccination and isolation in the model caused a decrease in the cumulative incidence for each network, however, this effect was only seen at high levels of vaccination and isolation for the complete network. This study highlights the importance of using empirical networks to describe contact patterns within populations that are unlikely to exhibit random mixing such as equine populations.

Integration of genetic and epidemiological data to infer H5N8 HPAI virus transmission dynamics during the 2016-2017 epidemic in Italy.

Between October 2016 and December 2017, several European Countries had been involved in a massive Highly Pathogenic Avian Influenza (HPAI) epidemic sustained by H5N8 subtype virus. Starting on December 2016, also Italy was affected by H5N8 HPAI virus, with cases occurring in two epidemic waves: the first between December 2016 and May 2017, and the second in July-December 2017. Eighty-three outbreaks were recorded in poultry, 67 of which (80.72%) occurring in the second wave. A total of 14 cases were reported in wild birds. Epidemiological information and genetic analyses were conjointly used to get insight on the spread dynamics. Analyses indicated multiple introductions from wild birds to the poultry sector in the first epidemic wave, and noteworthy lateral spread from October 2017 in a limited geographical area with high poultry densities. Turkeys, layers and backyards were the mainly affected types of poultry production. Two genetic sub-groups were detected in the second wave in non-overlapping geographical areas, leading to speculate on the involvement of different wild bird populations. The integration of epidemiological data and genetic analyses allowed to unravel the transmission dynamics of H5N8 virus in Italy, and could be exploited to timely support in implementing tailored control measures.

Cryptic connections illuminate pathogen transmission within community networks.

Understanding host interactions that lead to pathogen transmission is fundamental to the prediction and control of epidemics1-5. Although the majority of transmissions often occurs within social groups6-9, the contribution of connections that bridge groups and species to pathogen dynamics is poorly understood10-12. These cryptic connections-which are often indirect or infrequent-provide transmission routes between otherwise disconnected individuals and may have a key role in large-scale outbreaks that span multiple populations or species. Here we quantify the importance of cryptic connections in disease dynamics by simultaneously characterizing social networks and tracing transmission dynamics of surrogate-pathogen epidemics through eight communities of bats. We then compared these data to the invasion of the fungal pathogen that causes white-nose syndrome, a recently emerged disease that is devastating North American bat populations13-15. We found that cryptic connections increased links between individuals and between species by an order of magnitude. Individuals were connected, on average, to less than two per cent of the population through direct contact and to only six per cent through shared groups. However, tracing surrogate-pathogen dynamics showed that each individual was connected to nearly fifteen per cent of the population, and revealed widespread transmission between solitarily roosting individuals as well as extensive contacts among species. Connections estimated from surrogate-pathogen epidemics, which include cryptic connections, explained three times as much variation in the transmission of the fungus that causes white-nose syndrome as did connections based on shared groups. These findings show how cryptic connections facilitate the community-wide spread of pathogens and can lead to explosive epidemics.

Unraveling the contact patterns and network structure of pig shipments in the United States and its association with porcine reproductive and respiratory syndrome virus (PRRSV) outbreaks.

The analysis of the pork value chain is becoming key to understanding the risk of infectious disease dissemination in the swine industry. In this study, we used social network analysis to characterize the swine shipment network structure and properties in a typical multisite swine production system in the US. We also aimed to evaluate the association between network properties and porcine respiratory and reproductive syndrome virus (PRRSV) transmission between production sites. We analyzed the 109,868 swine shipments transporting over 93 million swine between more than 500 production sites from 2012 to 2014. A total of 248 PRRSV positive occurrences were reported from 79 production sites during those 3 years. The temporal dynamics of swine shipments was evaluated by computing network properties in one-month and three-month networks. The association of PRRS occurrence in sow farms with centrality properties from one-month and three-month networks was assessed by using the multilevel logistic regression. All monthly networks showed a scale-free network topology with positive degree assortativity. The regression model revealed that out-degree centrality had a negative association with PRRS occurrence in sow farms in both one-month and three-month networks [OR=0.79 (95% CI, 0.63-0.99) in one-month network and 0.56 (95% CI, 0.36, 0.88) in three-month network] and in-closeness centrality model was positively associated with PRRS occurrence in sow farms in the three-month network [OR=2.45 (95% CI, 1.14-5.26)]. We also describe how the occurrence of porcine epidemic diarrheac (PED) outbreaks severely affected the network structure as well as the PRRS occurrence reports and its association with centrality measures in sow farms. The structure of the swine shipment network and the connectivity between production sites influenced on the PRRSV transmission. The use of network topology and characteristics combining with spatial analysis based on fine scale geographical location of production sites will be useful to inform the design of more cost-efficient, risk-based surveillance and control measures for PRRSV as well as other diseases in the US swine industry.

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

Animals' exchanges are considered the most effective route of between-farm infectious disease transmission. However, despite being often overlooked, the infection spread due to contaminated equipment, vehicles, or personnel proved to be important for several livestock epidemics. This study investigated the role of indirect contacts in a potential infection spread in the dairy farm network of the Province of Parma (Northern Italy). We built between-farm contact networks using data on cattle exchange (direct contacts), and on-farm visits by veterinarians (indirect contacts). We compared the features of the contact structures by using measures on static and temporal networks. We assessed the disease spreading potential of the direct and indirect network structures in the farm system by using data on the infection state of farms by paratuberculosis. Direct and indirect networks showed non-trivial differences with respect to connectivity, contact distribution, and super-spreaders identification. Furthermore, our analyses on paratuberculosis data suggested that the contributions of direct and indirect contacts on diseases spread are apparent at different spatial scales. Our results highlighted the potential role of indirect contacts in between-farm disease spread and underlined the need for a deeper understanding of these contacts to develop better strategies for prevention of livestock epidemics.

Badgers prefer cattle pasture but avoid cattle: implications for bovine tuberculosis control.

Effective management of infectious disease relies upon understanding mechanisms of pathogen transmission. In particular, while models of disease dynamics usually assume transmission through direct contact, transmission through environmental contamination can cause different dynamics. We used Global Positioning System (GPS) collars and proximity-sensing contact-collars to explore opportunities for transmission of Mycobacterium bovis [causal agent of bovine tuberculosis] between cattle and badgers (Meles meles). Cattle pasture was badgers' most preferred habitat. Nevertheless, although collared cattle spent 2914 collar-nights in the home ranges of contact-collared badgers, and 5380 collar-nights in the home ranges of GPS-collared badgers, we detected no direct contacts between the two species. Simultaneous GPS-tracking revealed that badgers preferred land > 50 m from cattle. Very infrequent direct contact indicates that badger-to-cattle and cattle-to-badger M. bovis transmission may typically occur through contamination of the two species' shared environment. This information should help to inform tuberculosis control by guiding both modelling and farm management.

A nationwide database linking information on the hosts with sequence data of their virus strains: A useful tool for the eradication of bovine viral diarrhea (BVD) in Switzerland.

Pestiviruses infect a wide variety of animals of the order Artiodactyla, with bovine viral diarrhea virus (BVDV) being an economically important pathogen of livestock globally. BVDV is maintained in the cattle population by infecting fetuses early in gestation and, thus, by generating persistently infected (PI) animals that efficiently transmit the virus throughout their lifetime. In 2008, Switzerland started a national control campaign with the aim to eradicate BVDV from all bovines in the country by searching for and eliminating every PI cattle. Different from previous eradication programs, all animals of the entire population were tested for virus within one year, followed by testing each newborn calf in the subsequent four years. Overall, 3,855,814 animals were tested from 2008 through 2011, 20,553 of which returned an initial BVDV-positive result. We were able to obtain samples from at least 36% of all initially positive tested animals. We sequenced the 5' untranslated region (UTR) of more than 7400 pestiviral strains and compiled the sequence data in a database together with an array of information on the PI animals, among others, the location of the farm in which they were born, their dams, and the locations where the animals had lived. To our knowledge, this is the largest database combining viral sequences with animal data of an endemic viral disease. Using unique identification tags, the different datasets within the database were connected to run diverse molecular epidemiological analyses. The large sets of animal and sequence data made it possible to run analyses in both directions, i.e., starting from a likely epidemiological link, or starting from related sequences. We present the results of three epidemiological investigations in detail and a compilation of 122 individual investigations that show the usefulness of such a database in a country-wide BVD eradication program.

Multicriteria Evaluation of Classical Swine Fever Control Strategies Using the Choquet Integral.

An outbreak of the highly contagious animal disease classical swine fever (CSF) requires the selection of an optimal control strategy. The choice of a control strategy is a decision process depending on different aspects. Besides epidemiology, economic and ethical/social aspects must be taken into account. In this study, multicriteria decision-making (MCDM) was used to evaluate six control strategies for two regions with different farm densities. A strategy including only the minimum EU control measures and the traditional control strategy based on preventive culling were compared to alternative control strategies using emergency vaccination and/or rapid PCR testing ('emergency vaccination', 'test to slaughter', 'test to control' and 'vaccination in conjunction with rapid testing'). The MACBETH approach was used in order to assess the three main criteria (epidemiology, economics and ethical/social aspects). Subcriteria with both quantitative and qualitative performance levels were translated into a normalized scale. The Choquet integral approach was adopted to obtain a ranking of the six CSF control strategies based on the three main criteria, taking interactions into account. Three different rankings of the importance of the main criteria, which were to reflect the potential perceptions of stakeholders, were examined. Both the region under investigation and the ranking of the main criteria had an influence on the 'best' choice. Alternative control strategies were favourable to the minimum EU control and the traditional control measures independent of the farm density. Because the choice of the 'best' control strategy does not solely depend on the epidemiological efficiency, MCDM can help to find the best solution. Both MACBETH and the Choquet integral approach are feasible MCDM approaches. MACBETH only needs a qualitative evaluation and is therefore a comparatively intuitive approach. The Choquet integral does not only take the importance of the criteria into account but also the interaction between them.

Infectious disease transmission and contact networks in wildlife and livestock.

The use of social and contact networks to answer basic and applied questions about infectious disease transmission in wildlife and livestock is receiving increased attention. Through social network analysis, we understand that wild animal and livestock populations, including farmed fish and poultry, often have a heterogeneous contact structure owing to social structure or trade networks. Network modelling is a flexible tool used to capture the heterogeneous contacts of a population in order to test hypotheses about the mechanisms of disease transmission, simulate and predict disease spread, and test disease control strategies. This review highlights how to use animal contact data, including social networks, for network modelling, and emphasizes that researchers should have a pathogen of interest in mind before collecting or using contact data. This paper describes the rising popularity of network approaches for understanding transmission dynamics in wild animal and livestock populations; discusses the common mismatch between contact networks as measured in animal behaviour and relevant parasites to match those networks; and highlights knowledge gaps in how to collect and analyse contact data. Opportunities for the future include increased attention to experiments, pathogen genetic markers and novel computational tools.

Brown spider monkeys (Ateles hybridus): a model for differentiating the role of social networks and physical contact on parasite transmission dynamics.

Elevated risk of disease transmission is considered a major cost of sociality, although empirical evidence supporting this idea remains scant. Variation in spatial cohesion and the occurrence of social interactions may have profound implications for patterns of interindividual parasite transmission. We used a social network approach to shed light on the importance of different aspects of group-living (i.e. within-group associations versus physical contact) on patterns of parasitism in a neotropical primate, the brown spider monkey (Ateles hybridus), which exhibits a high degree of fission-fusion subgrouping. We used daily subgroup composition records to create a 'proximity' network, and built a separate 'contact' network using social interactions involving physical contact. In the proximity network, connectivity between individuals was homogeneous, whereas the contact network highlighted high between-individual variation in the extent to which animals had physical contact with others, which correlated with an individual's age and sex. The gastrointestinal parasite species richness of highly connected individuals was greater than that of less connected individuals in the contact network, but not in the proximity network. Our findings suggest that among brown spider monkeys, physical contact impacts the spread of several common parasites and supports the idea that pathogen transmission is one cost associated with social contact.

Complex system modelling for veterinary epidemiology.

The use of mathematical models has a long tradition in infectious disease epidemiology. The nonlinear dynamics and complexity of pathogen transmission pose challenges in understanding its key determinants, in identifying critical points, and designing effective mitigation strategies. Mathematical modelling provides tools to explicitly represent the variability, interconnectedness, and complexity of systems, and has contributed to numerous insights and theoretical advances in disease transmission, as well as to changes in public policy, health practice, and management. In recent years, our modelling toolbox has considerably expanded due to the advancements in computing power and the need to model novel data generated by technologies such as proximity loggers and global positioning systems. In this review, we discuss the principles, advantages, and challenges associated with the most recent modelling approaches used in systems science, the interdisciplinary study of complex systems, including agent-based, network and compartmental modelling. Agent-based modelling is a powerful simulation technique that considers the individual behaviours of system components by defining a set of rules that govern how individuals ("agents") within given populations interact with one another and the environment. Agent-based models have become a recent popular choice in epidemiology to model hierarchical systems and address complex spatio-temporal dynamics because of their ability to integrate multiple scales and datasets.

Characterization of contact structures for the spread of infectious diseases in a pork supply chain in northern Germany by dynamic network analysis of yearly and monthly networks.

A major risk factor in the spread of diseases between holdings is the transport of live animals. This study analysed the animal movements of the pork supply chain of a producer group in Northern Germany. The parameters in-degree and out-degree, ingoing and outgoing infection chain, betweenness and ingoing and outgoing closeness were measured using dynamic network analysis to identify holdings with central positions in the network and to characterize the overall network topology. The potential maximum epidemic size was also estimated. All parameters were calculated for three time periods: the 3-yearly network, the yearly and the monthly networks. The yearly and the monthly networks were more fragmented than the 3-yearly network. On average, one-third of the holdings were isolated in the yearly networks and almost three quarters in the monthly networks. This represented an immense reduction in the number of holdings participating in the trade of the monthly networks. The overall network topology showed right-skewed distributions for all calculated centrality parameters indicating that network resilience was high concerning the random removal of holdings. However, for a targeted removal of holdings according to their centrality, a rapid fragmentation of the trade network could be expected. Furthermore, to capture the real importance of holdings for disease transmission, indirect trade contacts (infection chain) should be considered. In contrast to the parameters regarding direct trade contacts (degree), the infection chain parameter did not underestimate the potential risk of disease transmission. This became more obvious, the longer the observed time period was. For all three time periods, the results for the estimation of the potential maximum epidemic size illustrated that the outgoing infection chain should be chosen. It considers the chronological order and the directed nature of the contacts and has no restrictions such as the strongly connected components of a cyclic network.

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.

EpiContactTrace: an R-package for contact tracing during livestock disease outbreaks and for risk-based surveillance.

BACKGROUND: During outbreak of livestock diseases, contact tracing can be an important part of disease control. Animal movements can also be of relevance for risk-based surveillance and sampling, i.e. both when assessing consequences of introduction or likelihood of introduction. In many countries, animal movement data are collected with one of the major objectives to enable contact tracing. However, often an analytical step is needed to retrieve appropriate information for contact tracing or surveillance. RESULTS: In this study, an open source tool was developed to structure livestock movement data to facilitate contact-tracing in real time during disease outbreaks and for input in risk-based surveillance and sampling. The tool, EpiContactTrace, was written in the R-language and uses the network parameters in-degree, out-degree, ingoing contact chain and outgoing contact chain (also called infection chain), which are relevant for forward and backward tracing respectively. The time-frames for backward and forward tracing can be specified independently and search can be done on one farm at a time or for all farms within the dataset. Different outputs are available; datasets with network measures, contacts visualised in a map and automatically generated reports for each farm either in HTML or PDF-format intended for the end-users, i.e. the veterinary authorities, regional disease control officers and field-veterinarians. EpiContactTrace is available as an R-package at the R-project website (http://cran.r-project.org/web/packages/EpiContactTrace/). CONCLUSIONS: We believe this tool can help in disease control since it rapidly can structure essential contact information from large datasets. The reproducible reports make this tool robust and independent of manual compilation of data. The open source makes it accessible and easily adaptable for different needs.

Control of classical swine fever epidemics under varying conditions--with special focus on emergency vaccination and rapid PCR testing.

SUMMARY: In case of a classical swine fever outbreak in the European Union (EU), its control is based upon the culling of swine on infected farms, movement restrictions in the protection and surveillance zones, and contact tracing. Additionally, preventive culling may be carried out. Emergency vaccination and rapid PCR testing are discussed as alternatives to avoid this measure. An outbreak of classical swine fever and the success of its control are influenced by different factors. Using a spatial and temporal Monte-Carlo simulation model the control strategies 'Restriction Zone', 'Traditional Control', 'Emergency Vaccination', 'Test To Slaughter', 'Test To Control' and 'Vaccination in conjunction with Rapid Testing' were compared under various conditions. Farm density, compliance with movement restrictions and delay in the establishment of an emergency vaccination were analysed as influencing factors. It was found that all these factors had a significant influence on the number of infected and culled farms. In a low-density region, the basic measures are sufficient to control an epidemic, provided strict compliance with movement restrictions is adhered to. In a high-density region, additional measures are necessary. They can compensate non-strict compliance with movement restriction to a certain extent. In the high-density region, 'Emergency Vaccination' and 'Vaccination in conjunction with Rapid Testing' reached the same level of infected farms as 'Traditional Control', independent of the value of compliance with movement restrictions. However, in the case of an emergency vaccination, an early start to the vaccination campaign is essential for successful disease control.

The performance of approximations of farm contiguity compared to contiguity defined using detailed geographical information in two sample areas in Scotland: implications for foot-and-mouth disease modelling.

BACKGROUND: When modelling infectious diseases, accurately capturing the pattern of dissemination through space is key to providing optimal recommendations for control. Mathematical models of disease spread in livestock, such as for foot-and-mouth disease (FMD), have done this by incorporating a transmission kernel which describes the decay in transmission rate with increasing Euclidean distance from an infected premises (IP). However, this assumes a homogenous landscape, and is based on the distance between point locations of farms. Indeed, underlying the spatial pattern of spread are the contact networks involved in transmission. Accordingly, area-weighted tessellation around farm point locations has been used to approximate field-contiguity and simulate the effect of contiguous premises (CP) culling for FMD. Here, geographic data were used to determine contiguity based on distance between premises' fields and presence of landscape features for two sample areas in Scotland. Sensitivity, positive predictive value, and the True Skill Statistic (TSS) were calculated to determine how point distance measures and area-weighted tessellation compared to the 'gold standard' of the map-based measures in identifying CPs. In addition, the mean degree and density of the different contact networks were calculated. RESULTS: Utilising point distances <1 km and <5 km as a measure for contiguity resulted in poor discrimination between map-based CPs/non-CPs (TSS 0.279-0.344 and 0.385-0.400, respectively). Point distance <1 km missed a high proportion of map-based CPs; <5 km point distance picked up a high proportion of map-based non-CPs as CPs. Area-weighted tessellation performed best, with reasonable discrimination between map-based CPs/non-CPs (TSS 0.617-0.737) and comparable mean degree and density. Landscape features altered network properties considerably when taken into account. CONCLUSION: The farming landscape is not homogeneous. Basing contiguity on geographic locations of field boundaries and including landscape features known to affect transmission into FMD models are likely to improve individual farm-level accuracy of spatial predictions in the event of future outbreaks. If a substantial proportion of FMD transmission events are by contiguous spread, and CPs should be assigned an elevated relative transmission rate, the shape of the kernel could be significantly altered since ability to discriminate between map-based CPs and non-CPs is different over different Euclidean distances.