The role of wildlife in the epidemiology of mycobacterium avium subspecies paratuberculosis in domestic ruminants in Europe (PARA-TB TRANSMISSION)

We aimed to assess the potential role of wildlife hosts in the within herd dynamics of Map and develop a framework for the assessment of the risk posed by wildlife hosts in the within-farm dynamics of Map-infection in dairy herds, beef suckler herds and sheep breeding flocks. Two approaches have been followed - A quantitative model of within farm transmission of Map in cattle farms and - A qualitative assessment of the risk of a replacement animal becoming infected from infected wildlife in a dairy, beef suckler or sheep breeding farm. For the first modelling approach, the models allowed both point estimates of parameters and distributions to be used as input permitting both deterministic and stochastic modelling. A high level of detail in the specification of livestock population dynamics and animal life-histories was achieved by linking a considerable number of management input parameters in addition to those used to model Map transmission. By initialising the models with different combinations of these parameters the within-herd dynamics of Map in different production systems (such as the dominant production systems in the project partner countries), under different control strategies and with different level of exposure to wildlife infection can be simulated. In this first modelling approach two separate models were used - the difference being the way transmission was modelled - a dairy-herd model and a beef-cattle model. In the dairy-herd model transmission of Map was modelled using the S-L-I-R model, where the animals of the herd were divided into: susceptible, latently infected, infectious (shedding the bacterium) and resistant (escaped from the disease). Estimates of the transmission parameter (beta) were obtained using empirical data from a previous field study in The Netherlands. This transmission parameter was estimated for farms with different risk profiles, the risk profile being an index of the potential for within-farm transmission based on calving management and hygiene. The number of new infections per unit time due to Map contamination from wildlife was assumed to be proportional to - The contribution of wildlife to environmental contamination relative to that of an infected cow, - The transmission probability, due to environmental contamination, from an infected cow to a susceptible calf and - The number of susceptible calves. The contribution of wildlife to environmental contamination relative to that of an infected cow was estimated from - Faecal contamination by wildlife (g per hectare), - Prevalence of Map in wildlife and - Infectivity of faeces (cfu per g of faeces). In the beef-cattle model (due to the lack of suitable epidemiological data) the within farm transmission was modelled as one minus the joint probability of escaping the different contamination sources; these sources being not only clinically and subclinically infected animals but also wildlife hosts. Transmission probabilities from clinically and subclinically infected cows were obtained from recent literature and the number of new infections per unit time due to Map contamination from wildlife was modelled in the same way as described before for the dairy herd model. For the second modelling approach we applied a standard risk-assessment methodology to obtain a framework to assess the risk that wildlife pose to dairy herds, beef suckler herds and sheep breeding farms in a more qualitative fashion than the previous approach. The hazard was Map and the risk questions that we considered dealt with - The risk of a replacement heifer infected from wildlife being introduced in the production group on a dairy herd, - The risk of a replacement heifer infected from wildlife being introduced in the production group on a beef suckler herd and - The risk of a replacement ewe infected from wildlife being introduced in the production group on a sheep breeding flock. For each of the questions we formulated the risk pathways starting with a physical pathway, which included a postpartum period followed by time spent outdoors (on pasture) and/or indoors. The sources of the agent could be infected cows/ewes and wildlife hosts and from these sources, via different routes and through a series of dependencies a susceptible animal may become infected and may be introduced in the production group. By decomposing each route in a series of individual events and assessing the likelihood of its occurrence we formally assessed the risk.

A panel of 122 field isolates of Map from the Czech Republic (24), The Netherlands (38), Scotland (46) and Spain (14) was assembled. The isolates were from both domestic livestock and wildlife as follows: cow (Bos taurus)(39), sheep (Ovis aries) (11), goat (Capra hircus) (17), moufflon (Ovis musimon) (7), red deer (Cervus elaphus) (11), fallow deer (Dama dama) (4), badger (Meles meles) (1), fox (Vulpes vulpes) (3), stoat (Mustela erminea) (4), weasel (Mustela nivalis) (2), crow (Corvus corone)(1), rook (Corvus frugilegus) (1), jackdaw (Corvus monedula)(1), rabbit (Oryctolagus cuniculus) (14), hare (Lepus europaeus) (2), rat (Rattus norvegicaus) (1), wood mouse (Apodemus sylvaticus) (1), giraffe (Giraffa camelopardalis) (1) and cat (Felis domesticus) (1). The panel of isolates was typed by amplified fragment length polymorphism (AFLP), restriction fragment length polymorphism and hybridization to IS900 (IS900-RFLP) and pulsed-field gel electrophoresis (PFGE)although not all isolates were typed with all methods due to problems incurred when subculturing. The panel provides a valuable resource for the scientific community for further studies on paratuberculosis, particularly those relating to strain differentiation and epidemiology. No funds exist to maintain stocks of the panel but the individual partners may have DNA stocks available for further studies.

Partners in UK, Greece, Spain and the Czech Republic have a total of 19 refereed publications in international publications. These cover studies on: - The risk of interspecies disease transmission from rabbits to livestock. - Clustering of Mycobacterium avium subsp paratuberculosis in rabbits. - Routes of intraspecific transmission of Mycobacterium avium subsp paratuberculosis in rabbits. - Sensitivity and specificity of serum ELISA and faecal culture for the diagnosis of paratuberculosis is sheep and goats. - The association of subclinical paratuberculosis with fertility in dairy ewes and goats. - Mycobacterium avium subsp paratuberculosis in fallow deer and wild boar in Spain. - Paratuberculosis and avian tuberculosis infection in red deer. - Wild boar as a possible vector of mycobacterial infection. - Invertebrates as possible vectors of paratuberculosis.

Quantification of wildlife faecal contamination of the feeding environment. Two feeding environments were considered: - Farm stored livestock feed and - Livestock grazing pastures. In Greek farms with M.a.paratuberculosis infected rodents, livestock feed stored on the farms was found to be contaminated with up to 120 rodent faeces/kg of feed. In the UK, information from farms on stocking densities and known disease prevalence, combined with M.a.paratuberculosis excretion rates (for livestock and rabbits) and published literature were used to quantify the relative input of bacteria onto grazing pasture by rabbits and domestic livestock. It was estimated that cattle, sheep and rabbits contribute 2.1 x 10{10}, 6.2 x 10{8} and 1.7 x 10{8} cfu / ha respectively. Due to the limited data available, the figures for rabbits are likely to be conservative. Grazing livestock contact potentially infective doses of M.a.paratuberculosis from rabbits on a daily basis. Consequently the exposure of livestock to wildlife faeces and thus M.a.paratuberculosis can be high and the faecal oral route of wildlife to livestock transmission is a likely route of inter-species transmission.

The web page www.ucm.es/info/para-tb was established to disseminate the project findings to the scientific community or other interested audience. The webpage contains project information regarding objectives and achievements, description of work-packages, the consortium, and the management structure. During the lifespan of the project, the web page has been updated with the scientific papers, presentation to congresses (including the abstracts), and other dissemination activities that have been submitted by authors. The webpage will be maintained for a period of time after completion of the project.

The quantitative models were used to assess the potential impact of control practices in different EU-production systems, using transmission parameters obtained from the literature. For each production system simulations for time periods of 30 years following the introduction of 1 infected animal were done. The initial prevalence of Map in wildlife was set at 17% and the average time between the beginning of clinical signs and the culling of a clinically infected animal was 3 months. Using these specifications, the expected within herd prevalences after the 30 year period were for a dairy farm in East Scotland 25.7%, for a beef suckler herd in East Scotland 35.1%, for a dairy farm in the Czech Republic 46.2%, for a beef suckler herd in the Czech Republic 1.1%, for a dairy farm in Greece 4.7% and for a bullfighting farm in Spain 40.3%. No change in the prevalence was predicted in most countries if the only measure was total control of wildlife infection. The reduction of the average time between the beginning of clinical signs and the culling of a clinically infected animal by a month greatly reduced the expected prevalence in most countries. Although there was considerable heterogeneity across production systems with respect to the within-herd dynamics of Map infection, the largest benefits in terms of disease control were obtained by reducing the time a clinically infected animal remains in the production group. Therefore our results support the strategic use of diagnostic tests and culling decisions that minimize the time a highly infectious animal remains in the herd. The other group of strategies that should be prioritized were those aimed at reducing the risk profile of the farm by improving neonatal management and hygiene. Preventing livestock to livestock Map transmission must be the priority in any disease control strategy however, wildlife to livestock transmission should also be targeted especially where wildlife species represent a significant source of environmental contamination and the farming system results in high levels of livestock exposure to Map in wildlife. The combination of field studies and modelling suggest that the livestock systems most at risk from Map in wildlife are Scottish cattle systems where livestock have high levels of exposure to Map from rabbits and Greek small ruminant systems using common grazing that are exposed to Map from rodents contaminating farm stored feed. Individual holdings of livestock using the same grazing resource may in epidemiological terms be seen as a single larger herd as any transmission event from wildlife to livestock can then be transmitted to any other animal using the grazing resource. In contrast Dutch dairy systems operating zero-grazing management regimes are at least risk from Map in wildlife as livestock exposure to environmental sources of contamination are greatly restricted.

A list of wildlife species found to harbour Map in a range of European ecosystems. Previously in Scotland Map had been isolated from a range of non-ruminant wildlife species including, rabbit (Oryctolagus cuniculus), hare (Lepus europaeus), fox (Vulpes vulpes), badger (Meles meles), stoat (Mustela erminea), weasel (Mustela nivalis), Common rat (Rattus norvegicus), wood mouse (Apodemus sylvaticus), crow (Corvus corone), rook (Corvus frugilegus) and jackdaw (Corvus monedula). In Greece Map was isolated from fox (Vulpes vulpes), house mouse (Mus domesticus), ship rat (Rattus rattus) and hare. In Spain Map was isolated from wild boar (Sus scrofa) and fallow deer (Dama dama). In the Netherlands Map was isolated from roe deer (Capreolus capreolus), red deer (cervus elaphus) and buzzard (Buteo buteo). In the Czech Republic fallow deer, wild boar, brown hare and ship rat were found to harbour Map. The results suggest that Scotland currently has the largest range of wildlfe species harbouring Map. However, it should be noted that rabbit populations in Europe were at extremely low levels during the project.

Spatial and temporal distribution of Map in rabbit populations. Clustering of pathogens in the environment leads to hotspots of diseases at local, regional, national and international levels. Scotland contains regional hotspots of Map in rabbits. The spatial and temporal dynamics of paratuberculosis in rabbits were studied within a hotspot region with the overall aim of determining environmental patterns of infection and thus risk of inter-species transmission to livestock. Specifically, to determine if prevalence of paratuberculosis in rabbits varies temporally between seasons and whether the heterogeneous spatial environmental distribution of Map at a large scale (i.e. regional hotspots) is replicated at finer resolutions within a hotspot. The overall prevalence of Map in rabbits was 39.7%; temporal distribution of infection in rabbits followed a cyclical pattern with a peak in Spring of 55.4% and a low in Summer of 19.4%. Spatially, Map infected rabbits, and thus risk of inter-species transmission, were highly clustered in the environment. However, this is mostly due to the clustered distribution of rabbits. Understanding the spatial and temporal distributions of Map in rabbits enables better informed livestock management decisions to reduce risk of inter-species transmission.

A workshop devoted to discuss the role of wildlife reservoirs of M.a. paratuberculosis was held on 18th August as activity of the 8th International Colloquium on paratuberculosis (Copenhagen, Denmark). This workshop was organized as a dissemination task of the European Project �The Role of Wildlife in the epidemiology of Mycobacterium avium subspecies paratuberculosis in domestic ruminants in Europe� QLK2-CT-2001-00879. The objectives of the workshop were to present updated information on worldwide prevalence of paratuberculosis infection in wildlife species, and to present the results achieved in the EU project QLK2-CT-2001-00879 about the epidemiology of paratuberculosis in Europe. The workshop was attended by 49 scientists from 20 countries. The report of the workshop will be included in the Proceedings of the Colloquium. The agenda was as follows: 1.2.1. M.a. paratuberculosis infection in European wildlife The first part of the workshop was devoted to the situation of M.a. paratuberculosis infection in European wildlife and the information derived from the EU-funded project QLK2-CT-2001-00879. ParaTB in wildlife before the EU project. Mike Hutchings. Scottish Agricultural College, Edinburgh, Scotland (UK). This presentation was devoted to a description of the state of the art regarding wildlife and paratuberculosis before this project started. ParaTB in wildlife in EU countries. Lucía de Juan. Universidad Complutense de Madrid, Spain. Dr. Lucía de Juan, scientific secretary of the Co-ordination Action VENoMYC (www.ucm.es/info/venomyc), presented the data from paratuberculosis in wildlife in European countries kindly submitted by VENoMYC partners. Overview of the project. Alastair Greig. Scottish Agricultural College, Edinburgh, Scotland (UK). Co-ordinator of the project, presented the objectives, number of samples analysed, selection of strains for molecular characterization, results and general conclusions of the project QLK2-CT-2001-00879. Results for the three typing methods for M.a.paratuberculosis. Karen Stevenson. Moredun Research Institute, Edinburgh, Scotland (EU), presented the molecular characterization from domestic and wildlife M.a.paratuberculosis isolates by three different techniques (PFGE, AFLP and IS900-RFLP). Sensitivity of EU livestock systems to presence of a wildlife host of M.a.paratuberculosis. Dirk Pfeiffer & Javier Guitian. Royal Veterinary College, Hertfordshire, England (UK). This presentation was devoted to epidemiological modelling techniques to compare the efficacies of various paratuberculosis control strategies and to determine the role of wildlife reservoirs in the persistence of paratuberculosis in domestic livestock. 1.2.2. Situation of paratuberculosis in wildlife in other non-European countries. Invited speakers would present data of paratuberculosis in wildlife in different non-European countries. "Prevalence of Johnes disease (para Tb) in New Zealand wildlife: preliminary surveys and implications for management". Andrea Byrom. Tb epidemiology and management, Manaaki Whenua - Landcare Research Lincoln (New Zealand). �Presence of M. avium ss. paratuberculosis in North America: ruminants, non-ruminants and their habitats�. Becky Manning. Johne's Testing Center, School of Veterinary Medicine, University of Wisconsin (USA). �Map from free ranging deer and rabbits near Minnesota dairy farms". Eran Raizman. Department of Pathobiology, School of Veterinary Medicine, Purdue University, Indiana (USA).

Quantification of the intake of rabbit faeces and thus Map through grazing. Two experiments were conducted to quantify rabbit faecal pellet ingestion during grazing by calves and lambs, the most susceptible livestock. Calves: Overall the calves ingested the equivalent of one faecal pellet in every 54 bites. Level of faecal contamination had no effect on the proportion of faecal pellets ingested. Rate of ingestion of faecal pellets increased at lower sward heights. Lambs: Overall the lambs ingested the equivalent of one faecal pellet in every 857 bites. Higher levels of contamination were associated with lower proportions of pellets ingested. Sward height had no significant effect on the proportion of faecal pellets ingested. From these results it was estimated that calves and lambs could ingest between 140 - 1329 and lambs 53 - 498 Map infected faecal pellets per hectare grazed, respectively. Sheep are better able to avoid ingestion of non-food items than cattle that potentially ingest infective doses of Map from rabbits on a daily basis.

The models were used to identify key epidemiological parameters by exploring changing scenarios with respect to wildlife infection and characteristics of the farming systems. In the quantitative models and with respect to the level of wildlife infection, three scenarios were considered - Absence of wildlife host (prevalence of Map shedding in wildlife host = 0%), - Prevalence of 17% and - Prevalence of 50%. To illustrate the impact of control measures aimed at removing infectious cows, in the dairy herd model two options were considered for time between becoming infectious and culling, 13 and 15 months. In the beef-cattle herd model scenarios with 1, 2, 4 and 6 months between development of clinical signs and culling were considered. To illustrate the impact of the range of indoor/outdoor farming systems that are used across Europe three scenarios or combinations of indoor/outdoor calving were used for the beef-cattle herd model, 100% outdoor calving, 50% outdoor calving and 100% indoor calving. Several simulations were run investigating the impact of different combinations of the above on the within-herd prevalence of Map-infection. Given our set of assumptions, the results suggested that the potential impact of the contamination of the environment by wildlife was very limited when compared to the impact of factors such as calving management and hygiene, calving indoors vs. outdoors or time during which a cow remains infectious in the herd. These 3 factors emerged as key predictors of the within-herd prevalence of Map-infection. The impact of Map contamination of the environment by wildlife species was higher for larger farms. The highest impact was observed for a hypothetical farm with 100 cows, 100% outdoor calving in which the estimated prevalence after 30 years was 8% without contamination by wildlife and 12% when the highest level of wildlife infection (50%) was considered. This quantitative assessment and the qualitative assessment of the risk of a replacement animal infected from wildlife being introduced into the production group on a dairy farm, a beef suckler herd and a sheep breeding flock, agreed that further investigation of the potential impact of wildlife on the within-farm dynamics of Map should target specific production systems in which this component was less likely to be negligible, namely, relatively large farms with a relatively low within-herd prevalence of Map infection and where cows were kept outdoors.

A panel of 122 isolates of Mycobacterium avium subsp. paratuberculosis (Map) was typed by Amplified Fragment Length Polymorphism (AFLP), Restriction Fragment Length Polymorphism analysis followed by hybridization to IS900 (IS900-RFLP) and Pulsed-Field Gel Electrophoresis (PFGE). Eighty seven isolates were successfully typed by IS900-RFLP using both Bst EII and Pst I enzymes and 79 isolates by PFGE using both Sna BI and Spe I enzymes. Sixty-eight strains were typed by AFLP but due to the variability and reproducibility of the technique, further typing was considered pointless. Partial typing data was obtained for the remaining isolates. Of the 87 isolates typed by IS900-RFLP, 10 Bst EII - Pst I multiplex profiles were detected. The IS900-RFLP multiplex profile with the widest geographical distribution was B-C1 and the multiplex profile with the broadest host range was B-C17 in this cohort of isolates. The profile B-C17 was only identified in Scotland. Of the 79 isolates typed by PFGE, 18 Sna BI-Spe I multiplex PFGE profiles were detected. The PFGE multiplex profile with the widest geographical distribution was [2-1] and this multiplex profile also had the broadest host range. These results concurred with the typing results obtained in project QLRT-2000-01420. It was not possible to sub-group the Map isolates by AFLP due to the combination of low genotypic diversity and experimental variation inherent in this technique. Map was isolated from the following wildlife species: fallow deer, red deer, moufflon, badger, fox, stoat, weasel, crow, rook, jackdaw, hare, rabbit, rat and wood mouse. A number of profiles were associated with a single host species but the number of isolates with the profiles in question were too small to make any comment regarding the existence of host species specific strains. AFLP typing revealed what may be a goat-specific genotype polymorphism but this result requires further investigation. It was difficult to make any conclusions from the results regarding the transmission of Map between wildlife and domestic livestock. The same strain types could be found in both domestic ruminants and wildlife (IS900-RFLP types B-C1, B-C17 and B-C9; PFGE types [2-1], [1-1] and [2-30]), which were also the most prevalent. In the panel of isolates examined, there were few examples of wildlife and domestic ruminant isolates obtained from the same property. In the Czech Republic isolates from a hare and a cow were made on the same property but were shown to have different IS900-RFLP types. In Scotland, isolates from a cow, stoat and rat were found to have the same PFGE-type on property GE, and isolates from a cow, stoat, rook, rabbit and hare were found to have the same PFGE type on property EN. Analysis of the results showed that it was not uncommon for more than one strain of Map to be isolated from animals on a single property.

The Department of Knowledge Policy (formerly ECLNV) in the Netherlands and the National Beef Association (NBA) in the UK have been the organisations which have been active in the EU project (QLK2-CT-2001-00879) entitled �The role of wildlife in the epidemiology of Mycobacterium avium subspecies paratuberculosis in domestic ruminants� in disseminating the results and findings of this project to stakeholders involved with the paratuberculosis problem. The findings from the project when added to the results of previous studies indicate that preventing livestock to livestock transmission of Mycobacterium avium subspecies paratuberculosis (Map) must be the priority in any paratuberculosis control strategy but that wildlife to livestock transmission must also be targeted especially where wildlife species represent a significant source of environmental contamination with Map and the farming systems result in high levels of exposure of livestock to such contamination. The combination of field studies and modelling suggests that the livestock systems most at risk from Map in wildlife are Scottish cattle systems where rabbits represent a significant source of Map and Greek small ruminant systems where rodents can contaminate stored feed and common grazing of livestock allows between herd and flock transmission of Map. In contrast, in countries such as Czech Republic where almost zero grazing by dairy cattle is the norm, the risk from Map in wildlife is extremely limited. In addition, in the last two years the rabbit population in the Netherlands has become virtually extinct due to viral haemorraghic syndrome (VHS), further limiting the interaction between this wildlife species and domestic livestock. This information has been passed to the Department of Knowledge Policy of the Ministry of Agriculture, Nature and Food quality, in the Netherlands who have formulated a communication plan for disseminating this information to the livestock and wildlife sectors and to the policy making bodies. At the quarterly meetings of the scientific commission on paratuberculosis in the Netherlands, the stakeholders have been informed about the results. These recommendations are regularly distributed to all Dutch dairy farmers and available to all stakeholders in the form of a booklet, the so-called �Parawijzer�. In the UK, the NBA has held farmers meetings, published papers on the role of wildlife in the epidemiology of paratuberculosis and chaired an industry government partnership aimed at increasing the awareness of the livestock sector on the impact of paratuberculosis on herd health and measures to control the disease. These publications include 25,000 leaflets, 3,000 posters, a CD with technical information, frequently asked questions booklet and a powerpoint presentation for vets. This �literature pack� was sent to all project partners for consideration for use in their own countries in their paratuberculosis education and control programmes. The results from the project are also available to national farmers organisations on the project website www.ucm.es/info/para-tb and on the website of the International Association for Paratuberculosis http://www.paratuberculosis.org/ and will be fed into several workpackages and via the dissemination workpackage (including a dedicated website) to the stakeholders of the new EU funded STREP- project on paratuberculosis (ParaTBTools).

Exploitable results

Condividi questa pagina

Scarica