Advertisement
Research| Volume 104, ISSUE 3, P3123-3143, March 2021

Economic losses due to Johne's disease (paratuberculosis) in dairy cattle

Open AccessPublished:January 14, 2021DOI:https://doi.org/10.3168/jds.2020-19381

      ABSTRACT

      Johne's disease (JD), or paratuberculosis, is an infectious inflammatory disorder of the intestines primarily associated with domestic and wild ruminants including dairy cattle. The disease, caused by an infection with Mycobacterium avium subspecies paratuberculosis (MAP) bacteria, burdens both animals and producers through reduced milk production, premature culling, and reduced salvage values among MAP-infected animals. The economic losses associated with these burdens have been measured before, but not across a comprehensive selection of major dairy-producing regions within a single methodological framework. This study uses a Markov chain Monte Carlo approach to estimate the annual losses per cow within MAP-infected herds and the total regional losses due to JD by simulating the spread and economic impact of the disease with region-specific economic variables. It was estimated that approximately 1% of gross milk revenue, equivalent to US$33 per cow, is lost annually in MAP-infected dairy herds, with those losses primarily driven by reduced production and being higher in regions characterized by above-average farm-gate milk prices and production per cow. An estimated US$198 million is lost due to JD in dairy cattle in the United States annually, US$75 million in Germany, US$56 million in France, US$54 million in New Zealand, and between US$17 million and US$28 million in Canada, one of the smallest dairy-producing regions modeled.

      Key words

      INTRODUCTION

      Johne's disease (JD), or paratuberculosis, is an infectious chronic inflammatory disorder of the intestines that can affect domestic and wild ruminants including dairy cattle (
      • Fecteau M.E.
      • Whitlock R.H.
      Paratuberculosis in cattle.
      ). The disease is caused by an infection with Mycobacterium avium ssp. paratuberculosis (MAP), a bacterial organism that is relatively resistant to environmental, physical, and chemical stressors (
      • Manning E.J.
      • Collins M.T.
      Mycobacterium avium ssp. paratuberculosis: Pathogen, pathogenesis and diagnosis.
      ;
      • Whittington R.J.
      • Marshall D.J.
      • Nicholls P.J.
      • Marsh I.B.
      • Reddacliff L.A.
      Survival and dormancy of Mycobacterium avium ssp. paratuberculosis in the environment.
      ;
      • Donaghy J.
      • Keyser M.
      • Johnston J.
      • Cilliers F.P.
      • Gouws P.A.
      • Rowe M.T.
      Inactivation of Mycobacterium avium ssp. paratuberculosis in milk by UV treatment.
      ) and that can persist for extended periods outside of its host environment, surviving up to 55 wk in a dry, fully shaded environment (
      • Whittington R.J.
      • Marshall D.J.
      • Nicholls P.J.
      • Marsh I.B.
      • Reddacliff L.A.
      Survival and dormancy of Mycobacterium avium ssp. paratuberculosis in the environment.
      ). Although MAP infection has also been observed in omnivores and carnivores such as wild rabbits (
      • Greig A.
      • Stevenson K.
      • Henderson D.
      • Perez V.
      • Hughes V.
      • Pavlik I.
      • Hines M.E.
      • McKendrick I.
      • Sharp J.M.
      Epidemiological study of paratuberculosis in wild rabbits in Scotland.
      ), foxes (
      • Beard P.M.
      • Daniels M.J.
      • Henderson D.
      • Pirie A.
      • Rudge K.
      • Buxton D.
      • Rhind S.
      • Greig A.
      • Hutchings M.R.
      • McKendrick I.
      • Stevenson K.
      • Sharp J.M.
      Paratuberculosis infection of nonruminant wildlife in Scotland.
      ), and nonhuman primates (
      • McClure H.M.
      • Chiodini R.J.
      • Anderson D.C.
      • Swenson R.B.
      • Thayer W.R.
      • Coutu J.A.
      Mycobacterium paratuberculosis Infection in a colony of stumptail macaques (Macaca arctoides).
      ;
      • Zwick L.S.
      • Walsh T.F.
      • Barbiers R.
      • Collins M.T.
      • Kinsel M.J.
      • Murnane R.D.
      Paratuberculosis in a Mandrill (Papio sphinx).
      ), JD is primarily associated with cattle and sheep. As the infection progresses in cattle through its 4 distinct stages, the clinical effects worsen in severity from diarrhea and reduced milk production to lethargy, hypoproteinemia, and severe emaciation (
      • Tiwari A.
      • VanLeeuwen J.A.
      • McKenna S.L.B.
      • Keefe G.P.
      • Barkema H.W.
      Johne's disease in Canada Part I: clinical symptoms, pathophysiology, diagnosis, and prevalence in dairy herds.
      ). These clinical effects can result in substantial economic losses for dairy producers (
      • Garcia A.B.
      • Shalloo L.
      Invited review: The economic impact and control of paratuberculosis in cattle.
      ), with decreased milk production (
      • Lombard J.E.
      • Garry F.B.
      • McCluskey B.J.
      • Wagner B.A.
      Risk of removal and effects on milk production associated with paratuberculosis status in dairy cows.
      ;
      • McAloon C.G.
      • Whyte P.
      • More S.J.
      • Green M.J.
      • O'Grady L.
      • Garcia A.
      • Doherty M.L.
      The effect of paratuberculosis on milk yield—A systematic review and meta-analysis.
      ), decreased slaughter value (
      • Benedictus G.
      • Dijkhuizen A.A.
      • Stelwagen J.
      (Economic losses to farms due to paratuberculosis in cattle).
      ;
      • Kudahl A.B.
      • Nielsen S.S.
      Effect of paratuberculosis on slaughter weight and slaughter value of dairy cows.
      ;
      • Raizman E.A.
      • Fetrow J.P.
      • Wells S.J.
      Loss of income from cows shedding Mycobacterium avium subspecies paratuberculosis prior to calving compared with cows not shedding the organism on two Minnesota dairy farms.
      ), and premature culling (
      • Ott S.L.
      • Wells S.J.
      • Wagner B.A.
      Herd-level economic losses associated with Johne's disease on US dairy operations.
      ;
      • Shephard R.W.
      • Williams S.H.
      • Beckett S.D.
      Farm economic impacts of bovine Johne's disease in endemically infected Australian dairy herds.
      ) among the primary sources. Annual losses per cow among MAP-infected dairy herds in the United States have been estimated at US$21 (
      • Ott S.L.
      • Wells S.J.
      • Wagner B.A.
      Herd-level economic losses associated with Johne's disease on US dairy operations.
      ), US$35 (
      • Groenendaal H.
      • Nielen M.
      • Jalvingh A.W.
      • Horst S.H.
      • Galligan D.T.
      • Hesselink J.W.
      A simulation of Johne's disease control.
      ), and up to US$79 (
      • Pillars R.B.
      • Grooms D.L.
      • Wolf C.A.
      • Kaneene J.B.
      Economic evaluation of Johne's disease control programs implemented on six Michigan dairy farms.
      ); annual losses among MAP-infected dairy herds in Canada have been estimated at CA$49 (approximately US$40) per cow (
      • Tiwari A.
      • VanLeeuwen J.A.
      • Dohoo I.R.
      • Keefe G.P.
      • Weersink A.
      Estimate of the direct production losses in Canadian dairy herds with subclinical Mycobacterium avium subspecies paratuberculosis infection.
      ).
      • Raizman E.A.
      • Fetrow J.P.
      • Wells S.J.
      Loss of income from cows shedding Mycobacterium avium subspecies paratuberculosis prior to calving compared with cows not shedding the organism on two Minnesota dairy farms.
      estimated income over feed cost losses of US$366 per MAP-shedding cow per lactation, and more recently,
      • Bhattarai B.
      • Fosgate G.T.
      • Osterstock J.B.
      • Fossler C.P.
      • Park S.C.
      • Roussel A.J.
      Perceptions of veterinarians in bovine practice and producers with beef cow-calf operations enrolled in the US Voluntary Bovine Johne's Disease Control Program concerning economic losses associated with Johne's disease.
      estimated annual losses of US$1,644 per 100 cows in a US herd with a true prevalence of 7%.
      With the herd-level prevalence of MAP infection likely exceeding 50% in most countries with significant dairy industries (
      • Barkema H.W.
      • Orsel K.
      • Nielsen S.S.
      • Koets A.P.
      • Rutten V.P.M.G.
      • Bannantine J.P.
      • Keefe G.P.
      • Kelton D.F.
      • Wells S.J.
      • Whittington R.J.
      • Mackintosh C.G.
      • Manning E.J.
      • Weber M.F.
      • Heuer C.
      • Forde T.L.
      • Ritter C.
      • Roche S.
      • Corbett C.S.
      • Wolf R.
      • Griebel P.J.
      • Kastelic J.P.
      • De Buck J.
      Knowledge gaps that hamper prevention and control of Mycobacterium avium subspecies paratuberculosis infection.
      ), there is a need to estimate the economic losses due to JD across major dairy-producing regions within a single methodological framework. This prospect is complicated by the heterogeneity of both economic characteristics across regions and methodologies across MAP prevalence studies. However, the former can be addressed by modeling with region-specific economic variables and the latter by modeling with a range of assumed prevalence scenarios across regions. Accordingly, this study uses Markov chain Monte Carlo methods to (1) model the spread of MAP infection within dairy herds, (2) estimate economic losses due to JD for regions with available prevalence estimates, and (3) estimate economic losses due to JD using a range of assumed prevalence scenarios across a more comprehensive group of major dairy-producing countries, states, and provinces.

      MATERIALS AND METHODS

      A MAP-negative Markov model was developed to establish baseline steady-state herd characteristics, followed by a MAP-positive Markov model where herd structure and the distribution of infected animals across age groups changed over time. Using a Monte Carlo simulation approach, MAP-positive herds from each region were compared with baseline MAP-negative herds from each region to generate 3 main sets of results: the general behavior of the model with assumptions of 10% initial within-herd prevalence and 50% herd-level prevalence for a generic herd with no region-specific economic variables; estimated losses due to MAP infection for a select group of dairy-producing regions with available prevalence estimates; and estimated losses for a more comprehensive group of dairy-producing regions over a range of assumed prevalence scenarios. Last, the Canadian estimates were reexamined with special consideration for the unique market conditions that arise due to supply management (e.g., fixed annual production levels and relatively high farm-gate prices).

      MAP-Negative Models

      A MAP-negative baseline Markov herd model was developed with purchased replacements coming from a separately modeled MAP-negative pool. The herd and replacement pools were modeled using the same constraints and structure: 15 age categories ranging from 0–3 mo to 8–9 yr of age, with each age category having its own matrix of transition probabilities, and the model spanning a 10-period horizon with each period representing 1 yr. In the MAP-negative models, animals either age or may be culled until they reach the 8–9 yr category when they must be culled. Natural replacements (aged 0 to 12 mo) and purchased replacements (aged 12 mo to 3 yr) are added to the herd each year to maintain the herd's age structure, with the former coming from within the herd and the latter coming from the replacement pool. After the initial age parameters were set, the herd and pool were modeled for fifty 1-yr periods stabilizing with an annual cow-culling rate of 27%, a young-stock percentage (including calves less than one year) of 48%, and for a 100-cow herd, 1.36 cows and 3.07 young-stock between 1 and 2 yr of age being brought in from the external replacement pool each year. These numbers are similar to those observed in Canadian dairy herds, which have an average cow-culling rate between 26% and 33% (), an average young-stock percentage of 48% (
      • STATCAN
      Table 32–10–0130–01 Number of cattle, by class and farm type (x 1,000).
      ), and purchase an average of 1.37 cows and 3.09 young-stock between 1 and 2 yr of age per 100 cows per year ().

      MAP-Positive Models

      Whereas the MAP-negative models had only 2 possible outcomes in each age group's transition matrix (i.e., continue aging or be culled), the MAP-positive models are more complex. Infection with MAP is characterized by a long incubation period (
      • Jubb T.F.
      • Sergeant E.S.G.
      • Callinan A.P.L.
      • Galvin J.W.
      Estimate of the sensitivity of an ELISA used to detect Johne's disease in Victorian dairy cattle herds.
      ;
      • Fecteau M.E.
      • Whitlock R.H.
      Paratuberculosis in cattle.
      ), with varying levels of susceptibility across age groups (
      • Windsor P.A.
      • Whittington R.J.
      Evidence for age susceptibility of cattle to Johne's disease.
      ;
      • Mortier R.A.R.
      • Barkema H.W.
      • Bystrom J.M.
      • Illanes O.
      • Orsel K.
      • Wolf R.
      • Atkins G.
      • De Buck J.
      Evaluation of age-dependent susceptibility in calves infected with two doses of Mycobacterium avium subspecies paratuberculosis using pathology and tissue culture.
      ) and shedding across stages of infection (
      • Weber M.F.
      • Kogut J.
      • de Bree J.
      • van Schaik G.
      • Nielen M.
      Age at which dairy cattle become Mycobacterium avium ssp. paratuberculosis faecal culture positive.
      ;
      • Mitchell R.M.
      • Medley G.F.
      • Collins M.T.
      • Schukken Y.H.
      A meta-analysis of the effect of dose and age at exposure on shedding of Mycobacterium avium subspecies paratuberculosis (MAP) in experimentally infected calves and cows.
      ;
      • Mortier R.A.R.
      • Barkema H.W.
      • Orsel K.
      • Wolf R.
      • De Buck J.
      Shedding patterns of dairy calves experimentally infected with Mycobacterium avium subspecies paratuberculosis..
      ;
      • Wolf R.
      • Orsel K.
      • De Buck J.
      • Barkema H.W.
      Calves shedding Mycobacterium avium subspecies paratuberculosis are common on infected dairy farms.
      ). The modeling implications of these characteristics will be expanded upon later in this section. However, in general terms, the MAP-positive models function in the following way: the animal can remain negative and continue aging, become infected and continue aging, or be culled. Once an animal is infected, it can either be culled or its infection can undergo progression, regression, or inertia (remain the same). Animals in each stage of infection are associated with a different risk of being culled, and within each stage, there are various nonshedding, lightly shedding, moderately shedding, and heavily shedding states. Infection pressure on animals in the herd ft is determined by the number and degree of shedding animals in the herd in period t, defined as
      infectionpressure=ft=MAPl×Lt+MAPm×Mt+MAPh×Htanimals,
      [1]


      where MAPl, MAPm, and MAPh equal the amount of MAP bacteria shed by lightly shedding, moderately shedding, and heavily shedding animals, respectively. Lt, Mt, and Ht equal the number of animals shedding at those levels in period t, and animals equals the total number of animals in the herd, which is fixed over the 10-period horizon.
      All other potential outcomes are functions of infection pressure ft and compete for the probability of remaining in the herd equal to (1 − x), where x is the probability of being culled. For MAP-negative animals, the probability of being culled remains the steady-state value according to their age category. For MAP-positive animals, the probability of being culled depends on the stage of their infection, with the probability increasing with the severity of infection. In their general forms across all time periods, the transition probability P for each outcome within each age-specific transition matrix can be defined as
      P(infection)=(1-e-i=0nfi×w)×fi=0nfi×(1-x),
      [2]


      P(progression)=(1-e-i=0n(fi+1fi)×w)×fi=0n(fi+1fi)×(1-x),
      [3]


      P(regression)=(1-e-i=0n(fi+1fi)×w)×1fi=0n(fi+1fi)×(1-x),
      [4]


      P(inertia)=(e-i=0n(fi+1fi)×w)×(1-x)a,
      [5]


      where n is the number of potential outcomes given an animal's age and stage of infection; P(infection) is the probability of an animal transitioning from a MAP-negative stage to the initial stage of infection (the probability of infection); P(progression) is the probability of transitioning from one stage of infection to a more severe stage (the probability of disease progression); P(regression) is the probability of transitioning from one stage of infection to a less severe stage (the probability of disease regression); and P(inertia) is the probability of transitioning to the same stage of infection (the probability of neither disease progression nor regression occurring). There is a positive relationship between infection pressure f and P(infection), and there is a positive relationship between infection pressure f and P(progression) and a negative relationship between infection pressure f and P(regression). This reflects the observed association between environmental MAP load and the rate of disease progression (
      • Weber M.F.
      • Kogut J.
      • de Bree J.
      • van Schaik G.
      • Nielen M.
      Age at which dairy cattle become Mycobacterium avium ssp. paratuberculosis faecal culture positive.
      ;
      • Windsor P.A.
      • Whittington R.J.
      Evidence for age susceptibility of cattle to Johne's disease.
      ;
      • Bolton M.W.
      • Pillars R.B.
      • Kaneene J.B.
      • Mauer W.A.
      • Grooms D.L.
      Detection of Mycobacterium avium subspecies paratuberculosis in naturally exposed dairy heifers and associated risk factors.
      ), including repeated exposure to MAP (
      • Fecteau M.-E.
      • Whitlock R.H.
      • Buergelt C.D.
      • Sweeney R.W.
      Exposure of young dairy cattle to Mycobacterium avium ssp. paratuberculosis (MAP) through intensive grazing of contaminated pastures in a herd positive for Johne's disease.
      ;
      • McGregor H.
      • Dhand N.K.
      • Dhungyel O.P.
      • Whittington R.J.
      Transmission of Mycobacterium avium ssp. paratuberculosis: Dose–response and age-based susceptibility in a sheep model.
      ;
      • Suwandi A.
      • Bargen I.
      • Pils M.C.
      • Krey M.
      • Zur Lage S.
      • Singh A.K.
      • Basler T.
      • Falk C.S.
      • Seidler U.
      • Hornef M.W.
      • Goethe R.
      • Weiss S.
      CD4 T Cell dependent colitis exacerbation following re-exposure of Mycobacterium avium ssp. paratuberculosis..
      ;
      • Marquetoux N.
      • Mitchell R.
      • Ridler A.
      • Heuer C.
      • Wilson P.
      A synthesis of the patho-physiology of Mycobacterium avium subspecies paratuberculosis infection in sheep to inform mathematical modelling of ovine paratuberculosis.
      ). Note that this differs from explicitly modeling initial infectious dose dependence, which would require complicated tracking of infected animals through time not readily accomplished within a Markov framework. However, the modeling approach employed here still captures the positive feedback between within-herd prevalence, infection pressure, and the rate of infection progression. Last, by definition, there is a negative relationship between the sum of all infection, progression, and regression probabilities and P(inertia) because the more likely the stage of infection is to change over one transition period, the less likely it is to remain unchanged. In Equation [5], a is the number of potential inertia outcomes where transitions would only occur to another state (level of shedding) within the same stage of infection (e.g., a transition from a nonshedding state of stage 3 infection to a moderately shedding state of stage 3 infection). In Equations [2] through [5], w is a weight that allows for adjustment in the rate of spread of MAP infection within the herd. For this study, w was set it at 0.075, which through trial and error was determined to result in an approximate doubling of within-herd prevalence over 10 yr. However, by adjusting this value, other within-herd prevalence scenarios could be modeled, from scenarios where prevalence remains relatively constant over 10 yr to explosive epidemics where prevalence increases dramatically.
      To accurately reflect MAP's infectivity-age relationship, its long incubation period, and its clinical progression, limits on the ranges of possible outcomes were set within each age group's transition matrix as well as the potential states of shedding across stages: infected animals up to 12 mo old can only be in stage 1 of infection (the initial stage of infection where animals are infected and may shed low levels of MAP bacteria); infected animals up to 24 mo old can be in stages 1 or 2 (stage 2 being the subclinical stage where animals appear healthy but may shed low and moderate levels of bacteria); from 2 to 5 yr old they can be in stages 1, 2, or 3 (stage 3 being the clinical stage of infection where animals may exhibit clinical symptoms such as intermittent diarrhea and weight loss and may begin to shed low, moderate, and high levels of bacteria); animals 5 to 9 yr old can be in all stages of infection, with the final stage of infection, stage 4 (the terminal stage characterized by severe emaciation, often mandibular edema, and high levels of shedding), being a de facto absorptive stage that most often results in culling within one period; all stages of infection have possible nonshedding states within them except stage 4; animals can only become infected in their first 12 mo; once an animal is MAP-positive they will remain positive for their lifetime; and progression or regression can only transition the animals to within one stage of their current stage of infection. For example, a “12–15 month” calf that is MAP-positive in stage 2 of infection can either progress to stage 3, regress to stage 1, remain in stage 2, or be culled within 1 yr. Similarly, a “5–6 year” cow in stage 3 of infection can either progress to stage 4, regress to stage 2, remain in stage 3, or be culled within 1 yr. Examples of the initial period 0 herd structure matrix, the period 0 transition matrices for animals aged 0–3 mo, 2–3 yr, and 8–9 yr, and the final period 1 herd structure matrix for the MAP-positive model, are available in Supplemental Tables S1–S5 (https://doi.org/10.5683/SP2/6OXE0R).

      Monte Carlo Simulations

      Monte Carlo simulations are a type of methodology that uses random sampling from a set of input variables, each with their own distributions, to determine a range of possible outcomes. In this case 10,000 iteration simulations were run using randomized variables in a Markov model according to the mean and distribution associated with each variable. For the MAP-specific variables in the model (Table 1), all variables were assumed to have normal distributions with standard deviations of 10% of their mean values. However, for the initial within-herd prevalence and initial regional cow-level prevalence, a set of herd conditions (disease distributions) was required for each potential initial value randomly selected in each of the 10,000 iterations. To obtain these initial conditions, MAP-positive replacements were introduced into the steady-state MAP-negative herd and fifty 1-yr periods were simulated using the MAP-positive model transition probabilities previously described. As prevalence increased over time in this 50-yr simulation, the distribution of infected animals in the herd changed, generating a range of disease distributions for every iteration within each subsequent 10-yr simulation. This model initialization allowed for within-herd and regional cow-level prevalence to be simulated with normal distributions and standard deviations equal to 20% of the mean. Although these assumed standard deviations may seem constrictive, data required to determine their true values were unavailable and the selected standard deviations capture a wide range of input values without destabilizing the simulations and their results.
      Table 1Input variables used in for the Monte Carlo simulations of the Mycobacterium avium ssp. paratuberculosis-positive Markov herd model
      All variables simulated with a normal distribution and a standard deviation of 10% of the mean.
      VariableMean valueUnitSource
      Effect of infection on production5.90%
      • McAloon C.G.
      • Whyte P.
      • More S.J.
      • Green M.J.
      • O'Grady L.
      • Garcia A.
      • Doherty M.L.
      The effect of paratuberculosis on milk yield—A systematic review and meta-analysis.
      Replacement pool5,000.00CowsAssumed
      Replacement cost (labor)2.00hAssumed
      Bacteria shed, light shedders
      Light and moderate shedding values based on median cfu count for the range. Heavy shedders at minimum cutoff.
      5.00cfu
      • Whitlock R.H.
      • Wells S.J.
      • Sweeney R.W.
      • Van Tiem J.
      ELISA and fecal culture for paratuberculosis (Johne's disease): Sensitivity and specificity of each method.
      and
      • Crossley B.M.
      • Zagmutt-Vergara F.J.
      • Fyock T.L.
      • Whitlock R.H.
      • Gardner I.A.
      Fecal shedding of Mycobacterium avium ssp. paratuberculosis by dairy cows.
      Bacteria shed, moderate shedders
      Light and moderate shedding values based on median cfu count for the range. Heavy shedders at minimum cutoff.
      25.00cfu
      • Whitlock R.H.
      • Wells S.J.
      • Sweeney R.W.
      • Van Tiem J.
      ELISA and fecal culture for paratuberculosis (Johne's disease): Sensitivity and specificity of each method.
      and
      • Crossley B.M.
      • Zagmutt-Vergara F.J.
      • Fyock T.L.
      • Whitlock R.H.
      • Gardner I.A.
      Fecal shedding of Mycobacterium avium ssp. paratuberculosis by dairy cows.
      Bacteria shed, heavy shedders
      Light and moderate shedding values based on median cfu count for the range. Heavy shedders at minimum cutoff.
      50.00cfu
      • Whitlock R.H.
      • Wells S.J.
      • Sweeney R.W.
      • Van Tiem J.
      ELISA and fecal culture for paratuberculosis (Johne's disease): Sensitivity and specificity of each method.
      and
      • Crossley B.M.
      • Zagmutt-Vergara F.J.
      • Fyock T.L.
      • Whitlock R.H.
      • Gardner I.A.
      Fecal shedding of Mycobacterium avium ssp. paratuberculosis by dairy cows.
      Weight at 0 to 3 mo74.24kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 3 to 6 mo141.82kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 6 to 9 mo214.70kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 9 to 12 mo286.52kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 12 to 15 mo354.86kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 15 to 18 mo425.32kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 18 to 21 mo477.63kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at 21 to 24 mo524.05kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Weight at maturity (2 to 9 yr)680.39kg
      • Jones C.M.
      • Heinrichs J.
      Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
      Value reduction, stage 4 animals31.0%
      • Kudahl A.B.
      • Nielsen S.S.
      Effect of paratuberculosis on slaughter weight and slaughter value of dairy cows.
      Value reduction, stage 3 animals
      Based on stage 4 value reduction observed in study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 0.308, α = 0.308, β = 0.031).
      29.0%Calculated
      Value reduction, stage 2 animals
      Based on stage 4 value reduction observed in study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 0.308, α = 0.308, β = 0.031).
      26.0%Calculated
      Value reduction, stage 1 animals
      Based on stage 4 value reduction observed in study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 0.308, α = 0.308, β = 0.031).
      15.0%Calculated
      Culling risk, stage 4 animals3.20Ratio
      • Hendrick S.H.
      • Kelton D.F.
      • Leslie K.E.
      • Lissemore K.D.
      • Archambault M.
      • Duffield T.F.
      Effect of paratuberculosis on culling, milk production, and milk quality in dairy herds.
      Culling risk, stage 3 animals
      Based on stage 4 hazard ratio observed in the study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 3.200, α = 3.200, β = 0.320). Stage 1 risk is based on a mean value of 1.00 with a normal distribution and standard deviation of 0.10, truncated with a minimum value of 1.00 to obtain a true mean of 1.08.
      2.98RatioCalculated
      Culling risk, stage 2 animals
      Based on stage 4 hazard ratio observed in the study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 3.200, α = 3.200, β = 0.320). Stage 1 risk is based on a mean value of 1.00 with a normal distribution and standard deviation of 0.10, truncated with a minimum value of 1.00 to obtain a true mean of 1.08.
      2.69RatioCalculated
      Culling risk, stage 1 animals
      Based on stage 4 hazard ratio observed in the study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 3.200, α = 3.200, β = 0.320). Stage 1 risk is based on a mean value of 1.00 with a normal distribution and standard deviation of 0.10, truncated with a minimum value of 1.00 to obtain a true mean of 1.08.
      1.08RatioCalculated
      1 All variables simulated with a normal distribution and a standard deviation of 10% of the mean.
      2 Light and moderate shedding values based on median cfu count for the range. Heavy shedders at minimum cutoff.
      3 Based on stage 4 value reduction observed in study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 0.308, α = 0.308, β = 0.031).
      4 Based on stage 4 hazard ratio observed in the study. Other values estimated by scaling the stage 4 value to a truncated cumulative logistic probability distribution (maximum = 3.200, α = 3.200, β = 0.320). Stage 1 risk is based on a mean value of 1.00 with a normal distribution and standard deviation of 0.10, truncated with a minimum value of 1.00 to obtain a true mean of 1.08.

      Economic Losses

      To estimate region-specific economic losses, region-specific economic input variables were incorporated into the simulations. Detailed values are available in Appendix Table A1., Table A2.. After each period, the MAP-positive herd was compared with the steady-state MAP-negative herd to estimate the economic losses per cow within infected herds due to MAP infection. Premature culling losses were estimated by tallying additional exits in the MAP-positive herd and assigning those exits a value according to their age-at-exit and associated replacement price. Additional aggregated labor costs of seeking out, purchasing, and introducing a replacement to the herd were also accounted for. Salvage losses were estimated by tallying MAP-positive exits and assigning them a reduced salvage value according to their stage of infection. Production losses were estimated by tallying the number of MAP-positive cows and determining the quantity of milk that would have been produced if those cows were negative instead, multiplied by the farm-gate price for milk. The sum of these 3 losses was divided by the number of cows in the herd to obtain an estimate of total losses per cow within MAP-infected herds. Total regional losses were estimated by the product of regional head count, herd-level prevalence, and total losses per cow in infected herds. Annual losses, both per cow and regional, were discounted over time at an assumed rate of 5% per annum and then averaged over the 10-yr horizon to obtain the reported annual loss estimates. This discount rate was selected because it is consistent with small private firm investment in a family enterprise; it falls between a public investment return rate of approximately 3% (
      • USDA
      National Dairy Comprehensive Report - Monthly - 02/29/2020.
      ) and a private investment return rate of approximately 10% (
      • Macrotrends
      S&P 500 Historical Annual Returns.
      ). Similarly, the Treasury Board of Canada selected a discount rate of 7% in its 2007 Cost-Benefit Analysis Guide but noted that it would likely be reduced in future years (
      • Treasury Board of Canada
      Canadian Cost-Benefit Analysis Guide: Regulatory Proposals.
      ). Finally, 2 main sets of simulations were run: one using available prevalence estimates (Table 2) for a limited number of regions, and a second using a range of assumed prevalence scenarios for a comprehensive selection of major dairy-producing regions.
      Table 2Available herd-level and regional cow-level Mycobacterium avium ssp. paratuberculosis (MAP) prevalence estimates for various dairy-producing regions
      Region
      USA = United States; WI = Wisconsin; MI = Michigan; DEU = Germany; GBR = Great Britain; NLD = the Netherlands; ITA = Italy; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; IRL = Ireland; ESP = Spain; BEL = Belgium; AUT = Austria.
      Herd-level prevalence (%)SourceRegional cow-level prevalence (%)Source
      USA21.6
      Apparent prevalence estimate.
      • Wells S.J.
      • Wagner B.A.
      Herd-level risk factors for infection with Mycobacterium paratuberculosis in US dairies and association between familiarity of the herd manager with the disease or prior diagnosis of the disease in that herd and use of preventive measures.
      3.4
      Apparent prevalence estimate.
      • Wells S.J.
      • Wagner B.A.
      Herd-level risk factors for infection with Mycobacterium paratuberculosis in US dairies and association between familiarity of the herd manager with the disease or prior diagnosis of the disease in that herd and use of preventive measures.
       WI34.0
      • Collins M.T.
      • Sockett D.C.
      • Goodger W.J.
      • Conrad T.A.
      • Thomas C.B.
      • Carr D.J.
      Herd prevalence and geographic distribution of, and risk factors for, bovine paratuberculosis in Wisconsin.
      4.8
      • Collins M.T.
      • Sockett D.C.
      • Goodger W.J.
      • Conrad T.A.
      • Thomas C.B.
      • Carr D.J.
      Herd prevalence and geographic distribution of, and risk factors for, bovine paratuberculosis in Wisconsin.
       MI66.0
      • Johnson-Ifearulundu Y.
      • Kaneene J.B.
      • Lloyd J.W.
      Herd-level economic analysis of the impact of paratuberculosis on dairy herds.
      6.9
      • Johnson-Ifearulundu Y.
      • Kaneene J.B.
      • Lloyd J.W.
      Herd-level economic analysis of the impact of paratuberculosis on dairy herds.
      DEU84.7
      Apparent prevalence estimate.
      • Hacker U.
      • Hüttner K.
      • Konow M.
      (Investigation of serological prevalence and risk factors of paratuberculosis in dairy farms in the state of Mecklenburg-Westpommerania, Germany).
      10.3
      Apparent prevalence estimate.
      Because the simulations include both a regional replacement pool and a MAP-positive herd, both regional cow-level prevalence estimates and within-herd prevalence estimates were required. Most prevalence studies referenced provided only herd-level and regional cow-level estimates. The following equality was used to convert prevalence estimates between the various types: pw = pr/ph, where pw, pr, and ph, equal within-herd, regional cow-level, and herd-level prevalence, respectively.
      • Hacker U.
      • Hüttner K.
      • Konow M.
      (Investigation of serological prevalence and risk factors of paratuberculosis in dairy farms in the state of Mecklenburg-Westpommerania, Germany).
      GBR76.1
      Apparent prevalence estimate.
      Herd-weighted average of results from first, second, and third tests.
      • Woodbine K.A.
      • Schukken Y.H.
      • Green L.E.
      • Ramirez-Villaescusa A.
      • Mason S.
      • Moore S.J.
      • Bilbao C.
      • Swann N.
      • Medley G.F.
      Seroprevalence and epidemiological characteristics of Mycobacterium avium ssp. paratuberculosis on 114 cattle farms in south west England.
      7.3
      • Woodbine K.A.
      • Schukken Y.H.
      • Green L.E.
      • Ramirez-Villaescusa A.
      • Mason S.
      • Moore S.J.
      • Bilbao C.
      • Swann N.
      • Medley G.F.
      Seroprevalence and epidemiological characteristics of Mycobacterium avium ssp. paratuberculosis on 114 cattle farms in south west England.
      NLD54.7
      Apparent prevalence estimate.
      • Muskens J.
      • Barkema H.W.
      • Russchen E.
      • van Maanen K.
      • Schukken Y.H.
      • Bakker D.
      Prevalence and regional distribution of paratuberculosis in dairy herds in the Netherlands.
      1.4
      Because the simulations include both a regional replacement pool and a MAP-positive herd, both regional cow-level prevalence estimates and within-herd prevalence estimates were required. Most prevalence studies referenced provided only herd-level and regional cow-level estimates. The following equality was used to convert prevalence estimates between the various types: pw = pr/ph, where pw, pr, and ph, equal within-herd, regional cow-level, and herd-level prevalence, respectively.
      • Muskens J.
      • Barkema H.W.
      • Russchen E.
      • van Maanen K.
      • Schukken Y.H.
      • Bakker D.
      Prevalence and regional distribution of paratuberculosis in dairy herds in the Netherlands.
      ITA70.5
      • Pozzato N.
      • Capello K.
      • Comin A.
      • Toft N.
      • Nielsen S.S.
      • Vicenzoni G.
      • Arrigoni N.
      Prevalence of paratuberculosis infection in dairy cattle in Northern Italy.
      Average of Lombardy and Veneto regions and weighted by herds sampled in study.
      6.9
      • Pozzato N.
      • Capello K.
      • Comin A.
      • Toft N.
      • Nielsen S.S.
      • Vicenzoni G.
      • Arrigoni N.
      Prevalence of paratuberculosis infection in dairy cattle in Northern Italy.
      Calculated from reported within-herd prevalence for Lombardy and Veneto regions, weighted by head of cattle sampled in each region.
      CAN42.0
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      3.4
      Apparent prevalence estimate.
      Weighted average
      Average of Canadian cow-level estimates weighted by head of cattle in each province.
       QC23.6
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      3.1
      Apparent prevalence estimate.
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
      Tiwari (2005) Canadian cow-level prevalence estimate.
       ON54.1
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      2.4
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Keefe G.P.
      • Tremblay R.
      • Power C.
      • Wichtel J.J.
      Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
       BC65.8
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian western provinces.
      3.1
      Apparent prevalence estimate.
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
      Tiwari (2005) Canadian cow-level prevalence estimate.
       AB65.8
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian western provinces.
      9.1
      Apparent prevalence estimate.
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
       MB65.8
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian western provinces.
      4.5
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Tiwari A.
      • Plaizier J.C.
      • Whiting T.L.
      Seroprevalences of antibodies against bovine leukemia virus, bovine viral diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in beef and dairy cattle in Manitoba.
       SK65.8
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian western provinces.
      2.7
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Forsythe L.
      • Tiwari A.
      • Chartier R.
      Seroprevalence of antibodies against bovine leukemia virus, bovine viral diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in dairy cattle in Saskatchewan.
       NS47.3
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian Atlantic provinces.
      3.3
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Keefe G.P.
      • Tremblay R.
      • Power C.
      • Wichtel J.J.
      Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
       NB47.3
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian Atlantic provinces.
      2.9
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Keefe G.P.
      • Tremblay R.
      • Power C.
      • Wichtel J.J.
      Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
       PE47.3
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian Atlantic provinces.
      1.3
      Apparent prevalence estimate.
      • VanLeeuwen J.A.
      • Keefe G.P.
      • Tremblay R.
      • Power C.
      • Wichtel J.J.
      Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
       NL47.3
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      Corbett et al. (2018) estimate for Canadian Atlantic provinces.
      3.1
      Apparent prevalence estimate.
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
      Tiwari (2005) Canadian cow-level prevalence estimate.
      AUS14.0
      • Kennedy D.J.
      • Citer L.
      Paratuberculosis control measures in Australia.
      Herd-level prevalence in control region (Victoria).
      1.8
      Apparent prevalence estimate.
      • Jubb T.F.
      • Galvin J.W.
      Effect of a test and control program for bovine Johne's disease in Victorian dairy herds 1992 - 2002.
      IRL20.6
      • Good M.
      • Clegg T.
      • Sheridan H.
      • Yearsely D.
      • O'Brien T.
      • Egan J.
      • Mullowney P.
      Prevalence and distribution of paratuberculosis (Johne's disease) in cattle herds in Ireland.
      2.7
      • Good M.
      • Clegg T.
      • Sheridan H.
      • Yearsely D.
      • O'Brien T.
      • Egan J.
      • Mullowney P.
      Prevalence and distribution of paratuberculosis (Johne's disease) in cattle herds in Ireland.
      ESP10.7
      • Diéguez F.J.
      • Arnaiz I.
      • Sanjuán M.L.
      • Vilar M.J.
      • López M.
      • Yus E.
      Prevalence of serum antibodies to Mycobacterium avium ssp. paratuberculosis in cattle in Galicia (northwest Spain).
      3.0
      • Diéguez F.J.
      • Arnaiz I.
      • Sanjuán M.L.
      • Vilar M.J.
      • López M.
      • Yus E.
      Prevalence of serum antibodies to Mycobacterium avium ssp. paratuberculosis in cattle in Galicia (northwest Spain).
      BEL36.0
      • Boelaert F.
      • Walravens K.
      • Biront P.
      • Vermeersch J.P.
      • Berkvens D.
      • Godfroid J.
      Prevalence of paratuberculosis (Johne's disease) in the Belgian cattle population.
      2.0
      • Boelaert F.
      • Walravens K.
      • Biront P.
      • Vermeersch J.P.
      • Berkvens D.
      • Godfroid J.
      Prevalence of paratuberculosis (Johne's disease) in the Belgian cattle population.
      AUT7.0
      Apparent prevalence estimate.
      • Gasteiner J.
      • Wenzl H.
      • Fuchs K.
      • Jark U.
      • Baumgartner W.
      Serological cross-sectional study of paratuberculosis in cattle in Austria.
      2.0
      Apparent prevalence estimate.
      • Gasteiner J.
      • Wenzl H.
      • Fuchs K.
      • Jark U.
      • Baumgartner W.
      Serological cross-sectional study of paratuberculosis in cattle in Austria.
      1 USA = United States; WI = Wisconsin; MI = Michigan; DEU = Germany; GBR = Great Britain; NLD = the Netherlands; ITA = Italy; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; IRL = Ireland; ESP = Spain; BEL = Belgium; AUT = Austria.
      2 Because the simulations include both a regional replacement pool and a MAP-positive herd, both regional cow-level prevalence estimates and within-herd prevalence estimates were required. Most prevalence studies referenced provided only herd-level and regional cow-level estimates. The following equality was used to convert prevalence estimates between the various types: pw = pr/ph, where pw, pr, and ph, equal within-herd, regional cow-level, and herd-level prevalence, respectively.
      3 Herd-weighted average of results from first, second, and third tests.
      4 Average of Lombardy and Veneto regions and weighted by herds sampled in study.
      5 Calculated from reported within-herd prevalence for Lombardy and Veneto regions, weighted by head of cattle sampled in each region.
      6 Average of Canadian cow-level estimates weighted by head of cattle in each province.
      7
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
      Canadian cow-level prevalence estimate.
      8
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      estimate for Canadian western provinces.
      9
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      estimate for Canadian Atlantic provinces.
      10 Herd-level prevalence in control region (Victoria).
      * Apparent prevalence estimate.

      RESULTS

      General Behavior

      With a mean initial within-herd prevalence of 10% and a mean initial regional herd-level prevalence of 50%, 90% of the 10,000 iterations resulted in proportional increases of within-herd prevalence ranging from 0.51 to 1.67, with a mean of 1.02, which is approximately equivalent to a doubling of within-herd prevalence from 10% to 20% over 10 yr (Figure 1). The percentage of infected animals that shed MAP increased from 6% to 12% over the 10-yr horizon, with the percentage of moderately and heavily shedding animals steadily increasing from 6% to over 10% of all shedding animals by yr 10 (Figure 2). As prevalence increased, so did infection pressure and the severity of infections, resulting in an increased cow-culling rate relative to the steady-state MAP-negative cow-culling rate (Figure 3A). Although this increased culling rate is indicative of worsening overall health in the herd, the precise sources of potential economic losses in the model (i.e., premature culling, reduced salvage value, and reduced production) can be seen changing over the 10-yr horizon in Figure 3B: the percentage of MAP-positive culls with varying degrees of reduced salvage value increased from just over 7% to 12% of all culls; the percentage of premature culls directly attributable to MAP infection doubled from 2% to 4% of all culls; forgone production as a percentage of total production increased from 0.63% to 1.16%.
      Figure thumbnail gr1
      Figure 1Simulated changes in within-herd Mycobacterium avium ssp. paratuberculosis (MAP) prevalence over time (10,000 iterations) assuming initial mean values of 10% within-herd prevalence and 50% herd-level prevalence. (A) Distribution of proportional changes in within-herd prevalence over a 10-yr horizon with distribution mean identified by µ. (B) Within-herd prevalence and its 90% confidence interval over time.
      Figure thumbnail gr2
      Figure 2Simulated changes in Mycobacterium avium ssp. paratuberculosis (MAP) shedding among animals over time (10,000 iterations) assuming initial mean values of 10% within-herd prevalence and 50% herd-level prevalence. (A) Animals in a MAP-shedding state as a percentage of total herd. (B) Degrees of shedding among MAP-shedding animals.
      Figure thumbnail gr3
      Figure 3Simulated changes in indicators of overall herd health over time (10,000 iterations) assuming initial mean values of 10% within-herd prevalence and 50% herd-level Mycobacterium avium ssp. paratuberculosis (MAP) prevalence. (A) Cow-culling rate compared with the MAP-negative (MAP-ve) herd over time. (B) Premature culls and MAP-positive (MAP+ve) culls as a percentage of total exits, and forgone production as a percentage of potential production.

      Economic Losses (Available Prevalence Estimates)

      On a national herd basis, estimated annual losses per cow within MAP-infected herds based on available prevalence estimates ranged from US$15.07 in the Netherlands to US$48.91 in Canada, and as a percentage of gross milk revenue on MAP-infected farms, from 0.39% in the Netherlands to 1.91% in Ireland. Total annual national losses ranged from US$1.91 million in Austria to US$150.19 million in Germany (Table 3).
      Table 3Estimated 10-yr average annual losses (US$) due to Mycobacterium avium ssp. paratuberculosis infection for various dairy-producing regions based on available prevalence estimates
      Region
      USA = United States; WI = Wisconsin; MI = Michigan; DEU = Germany; GBR = Great Britain; NLD = the Netherlands; ITA = Italy; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; IRL = Ireland; ESP = Spain; BEL = Belgium; AUT = Austria.
      Positive herds, losses per cow (90% CI)Positive herds, losses as % of milk revenue (90% CI)Regional losses in millions (90% CI)
      USA53.30(38.49–68.51)1.41(1.02–1.81)107.92(63.67–160.18)
       WI47.60(33.98–61.83)1.22(0.87–1.58)20.67(11.95–30.92)
       MI43.02(30.02–57.40)1.00(0.70–1.34)12.07(6.81–18.27)
      DEU43.08(30.46–56.25)1.32(0.93–1.72)150.19(85.82–225.82)
      GBR34.40(23.82–45.69)1.07(0.74–1.42)49.38(27.62–74.86)
      NLD15.07(9.52–21.40)0.39(0.25–0.55)12.85(6.82–20.35)
      ITA33.45(22.81–44.78)1.04(0.71–1.39)40.05(22.48–60.81)
      CAN48.91(33.41–65.85)0.86(0.59–1.16)20.02(11.28–30.63)
       QC60.29(42.91–78.80)1.09(0.78–1.43)5.05(2.94–7.56)
       ON30.77(19.73–42.86)0.56(0.36–0.78)5.41(2.86–8.47)
       BC37.52(24.56–51.92)0.59(0.39–0.82)2.08(1.13–3.24)
       AB84.04(59.59–110.12)1.37(0.97–1.79)4.39(2.53–6.61)
       MB44.46(29.48–60.85)0.75(0.50–1.03)1.21(0.67–1.85)
       SK33.17(21.61–46.03)0.55(0.36–0.76)0.64(0.34–1.00)
       NS37.09(23.49–51.93)0.66(0.42–0.92)0.38(0.20–0.60)
       NB36.21(24.29–49.73)0.68(0.45–0.93)0.33(0.18–0.51)
       PE19.56(12.47–27.81)0.35(0.22–0.49)0.13(0.07–0.21)
       NL57.59(38.78–78.26)0.76(0.51–1.03)0.16(0.09–0.25)
      AUS31.25(22.05–40.38)1.53(1.08–1.97)6.68(3.88–9.95)
      IRL42.81(30.21–55.52)1.91(1.35–2.48)12.09(7.02–17.96)
      ESP45.99(35.57–57.52)1.39(1.08–1.74)4.03(2.48–5.79)
      BEL21.72(14.28–29.62)0.72(0.48–0.99)4.15(2.26–6.43)
      AUT51.21(41.13–61.75)1.84(1.48–2.22)1.91(1.19–2.71)
      1 USA = United States; WI = Wisconsin; MI = Michigan; DEU = Germany; GBR = Great Britain; NLD = the Netherlands; ITA = Italy; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; IRL = Ireland; ESP = Spain; BEL = Belgium; AUT = Austria.

      Economic Losses (Assumed Prevalence)

      With an assumed within-herd prevalence of 10% in MAP-infected herds and a herd-level prevalence of 50% across major dairy-producing regions, annual losses per cow in infected herds ranged from US$8.31 in Brazil to US$81.53 in Japan, with a revenue-weighted average of US$32.84 per cow per year. Annual national herd losses ranged from US$5.84 million in Finland to US$198.42 million in the United States. Revenue-weighted average losses as a percentage of gross milk revenue were 1.07%, with significantly higher estimates for Ireland, New Zealand, and Australia (Table 4). Results for a wider range of prevalence scenarios are presented in Appendix Table A3. When annual losses are broken down into their component sources, premature culling losses ranged from 14% of total losses in Poland and Russia to 34% in New Zealand and Australia, reduced salvage value losses ranged from 6% in Poland to 19% in Ireland, and production losses ranged from 40% in Ireland to 80% in Poland. Revenue-weighted averages premature culling losses accounted for 24% of total losses, reduced salvage value losses accounted for 11%, and production losses accounted for 65%. Details of the breakdown of total losses into source components for all regions modeled are available in Appendix Table A4.
      Table 4Estimated 10-yr average annual losses (US$) due to Mycobacterium avium ssp. paratuberculosis infection for various dairy-producing regions assuming a mean herd-level prevalence of 50% and a mean within-herd prevalence of 10%
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Positive herds, losses per cow (90% CI)Positive herds, losses as % of milk revenue (90% CI)Regional losses in millions (90% CI)
      EU-2831.73(21.86–42.28)1.08(0.75–1.44)364.31(205.91–551.59)
       DEU36.40(25.19–47.83)1.11(0.77–1.46)74.86(42.13–113.37)
       FRA31.33(21.71–41.59)1.11(0.77–1.48)55.73(31.71–83.60)
       GBR34.27(23.66–45.46)1.06(0.73–1.41)32.28(18.24–48.74)
       POL20.77(14.27–27.88)0.86(0.59–1.15)23.05(12.98–35.39)
       NLD42.82(29.77–56.74)1.11(0.77–1.47)33.31(18.89–50.42)
       ITA33.06(23.09–43.92)1.03(0.72–1.37)28.08(15.94–42.38)
       IRL37.36(25.29–49.87)1.67(1.13–2.22)25.65(14.27–38.95)
       ESP30.76(21.13–41.14)0.93(0.64–1.25)12.61(7.10–19.03)
       DNK47.53(32.96–62.98)1.14(0.79–1.51)13.57(7.70–20.50)
       BEL33.64(23.31–44.70)1.12(0.78–1.49)8.92(5.03–13.56)
       AUT33.54(23.21–44.57)1.21(0.84–1.60)8.96(5.09–13.47)
       CZE30.33(20.80–40.86)0.88(0.60–1.18)5.46(3.10–8.33)
       SWE40.89(28.26–54.38)1.14(0.78–1.51)6.42(3.61–9.78)
       FIN44.10(30.85–58.60)1.09(0.76–1.44)5.84(3.33–8.84)
      USA42.26(29.09–56.00)1.12(0.77–1.48)198.42(112.42–300.44)
       CA41.73(29.00–55.33)1.10(0.76–1.46)36.29(20.51–55.30)
       WI39.42(27.19–52.44)1.01(0.70–1.34)25.18(14.26–38.06)
       ID37.23(25.50–50.04)0.92(0.63–1.24)11.37(6.40–17.54)
       NY43.61(30.35–57.84)1.12(0.78–1.49)13.62(7.83–20.61)
       TX40.69(28.27–54.12)1.05(0.73–1.39)10.97(6.19–16.69)
       MI40.88(28.06–54.44)0.95(0.65–1.27)8.69(4.91–13.14)
       PA36.40(25.20–48.13)1.09(0.75–1.44)9.47(5.38–14.23)
       MN38.56(26.68–51.19)1.09(0.75–1.44)8.76(4.95–13.26)
       NM38.82(26.77–51.81)0.95(0.66–1.27)6.42(3.60–9.67)
       WA42.57(29.54–56.43)1.08(0.75–1.43)5.91(3.36–8.96)
      BRA8.31(5.78–11.02)1.01(0.71–1.35)71.07(40.35–107.28)
      CHN13.29(9.17–17.83)0.96(0.66–1.29)79.59(45.14–120.88)
      RUS14.25(9.73–19.11)0.87(0.59–1.16)48.70(27.68–75.12)
      NZL21.75(15.05–28.85)1.37(0.95–1.82)53.95(30.52–81.69)
      TUR11.44(7.85–15.39)0.90(0.62–1.21)36.36(20.61–55.29)
      CAN56.99(39.35–75.72)1.00(0.69–1.33)27.78(15.60–42.04)
       QC53.30(36.93–70.94)0.97(0.67–1.29)9.47(5.35–14.38)
       ON55.35(38.36–73.56)1.01(0.70–1.34)8.98(5.09–13.6)
       BC62.94(43.57–83.46)0.99(0.69–1.31)2.65(1.48–4.00)
       AB67.16(46.36–89.14)1.09(0.76–1.45)2.66(1.51–4.04)
       MB56.64(38.94–75.90)0.96(0.66–1.29)1.17(0.65–1.79)
       SK62.32(43.15–82.92)1.03(0.71–1.37)0.91(0.52–1.39)
       NS52.26(36.20–70.16)0.93(0.64–1.25)0.56(0.32–0.85)
       NB51.05(35.08–68.36)0.95(0.66–1.28)0.49(0.28–0.75)
       PE52.63(36.33–70.82)0.93(0.64–1.25)0.38(0.21–0.58)
       NL77.01(53.31–102.24)1.01(0.70–1.34)0.23(0.13–0.35)
      AUS28.21(19.51–37.61)1.38(0.95–1.84)21.56(12.19–32.84)
      JPN81.53(56.37–108.24)1.02(0.70–1.35)34.63(19.45–52.64)
      1 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Annual losses due to MAP infection were also estimated with an assumed stable within-herd prevalence of 10% and the same stable herd-level prevalence of 50%. Similar results were observed (Table A5): annual losses per cow in infected herds ranged from US$7.30 in Brazil to US$71.57 in Japan, and annual national herd losses ranged from US$4.87 million in Czechia to $171.92 million in the United States. Losses per cow as a percentage of gross milk revenue ranged from 0.78% in Czechia to 1.40% in Ireland, with a revenue-weighted average of 0.93%. Compared with the estimated losses within the dynamic 10-yr prevalence models, the single-year stable prevalence (and therefore non-discounted) losses per cow were between 10 and 16% less, and annual regional losses were between 11 and 16% less, with revenue-weighted average differences of 13% for both per-cow and per-region annual losses.

      Economic Losses (Canada)

      This section reexamines the results for Canada with special consideration for the implications of supply management; production losses are no longer calculated as forgone production, but instead as the cost of having additional cows to maintain a fixed level of production over the 10-yr horizon. With assumed prevalence values of 10% within infected herds and 50% of herds being infected, annual losses per cow ranged from US$28.48 in Prince Edward Island to US$53.16 in Newfoundland, and annual provincial losses ranged from US$160,000 in Newfoundland to US$5.73 million in Québec (Table 5). Expanded results with fixed annual production are available in the Appendix; the breakdown of total losses per cow into their source components is available in Table A6, and results for a wider range of prevalence scenarios are available in Table A7.
      Table 5Estimated 10-yr average annual losses (US$) for Canadian dairy producers assuming a mean herd-level Mycobacterium avium ssp. paratuberculosis (MAP) prevalence of 50%, a mean within-herd prevalence of 10%, and with consideration for supply management (fixed output over time and production losses allocated as increased variable costs necessary to maintain production)
      Region
      CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.
      Variable costs per cow
      STATCAN–Table 32–10–0136–01, farm operating revenues and expenses, annual (STATCAN, 2019b). Sum of “Feed, supplements, straw, and bedding,” “Veterinary fees, medicine, and breeding fees,” and “Salaries and wages, including benefits related to employee salaries” for average dairy farms across all revenue levels in 2018. Total per farm divided by number of cows per farm. Number of cows per farm obtained by number of cattle divided by number of farms: CDIC–Number of farms with shipments of milk (CDIC, 2019b). Number of cattle: STATCAN–Table 32–10–0130–01, number of cattle, by class and farm type (STATCAN, 2020).
      (US$)
      Positive herds, losses per cow (90% CI)Positive herds, losses as % of milk revenue (90% CI)Regional losses in millions (90% CI)
      CAN2,47635.11(24.34–46.18)0.62(0.43–0.81)17.11(9.67–25.74)
       QC2,43032.23(22.54–42.34)0.59(0.41–0.77)5.73(3.23–8.58)
       ON2,25633.15(22.80–43.70)0.60(0.41–0.79)5.38(3.03–8.13)
       BC3,20441.39(28.87–53.94)0.65(0.45–0.85)1.74(0.98–2.62)
       AB3,10646.42(32.22–61.30)0.76(0.53–1.00)1.84(1.05–2.80)
       MB3,01436.94(25.82–48.50)0.63(0.44–0.82)0.76(0.44–1.15)
       SK2,78539.97(27.69–52.58)0.66(0.46–0.87)0.59(0.33–0.89)
       NS2,51531.00(21.64–40.45)0.55(0.39–0.72)0.33(0.19–0.50)
       NB2,46431.28(21.97–40.93)0.58(0.41–0.76)0.30(0.17–0.45)
       PE2,14428.48(19.84–37.38)0.50(0.35–0.66)0.21(0.12–0.31)
       NL4,11253.16(37.11–69.50)0.70(0.49–0.91)0.16(0.09–0.24)
      1 CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.
      2 STATCAN–Table 32–10–0136–01, farm operating revenues and expenses, annual (
      • STATCAN
      Table 32–10–0136–01 Farm operating revenues and expenses, annual.
      ). Sum of “Feed, supplements, straw, and bedding,” “Veterinary fees, medicine, and breeding fees,” and “Salaries and wages, including benefits related to employee salaries” for average dairy farms across all revenue levels in 2018. Total per farm divided by number of cows per farm. Number of cows per farm obtained by number of cattle divided by number of farms: CDIC–Number of farms with shipments of milk (
      • CDIC
      D056 - Number of Farms with Shipments of Milk by Province.
      ). Number of cattle: STATCAN–Table 32–10–0130–01, number of cattle, by class and farm type (
      • STATCAN
      Table 32–10–0130–01 Number of cattle, by class and farm type (x 1,000).
      ).

      Sensitivity Analyses

      For simplicity, an average Canadian herd with an assumed initial within-herd prevalence of 10% and an assumed herd-level prevalence of 50% in the region has been selected. Changes in estimated losses per cow were most sensitive to changes in within-herd prevalence, followed by annual production per cow, the farm-gate price, the effect of MAP infection on production, and the amount of bacteria shed by lightly shedding animals (Figure 4). The latter 4 variables had a similar effect on estimated losses per cow, which was significantly less than the effect of within-herd prevalence. Regional loss estimates were sensitive to similar variables, but most sensitive to herd-level prevalence. Premature culling losses were most sensitive to the culling risk associated with stage 1 MAP infection, whereas losses due to reduced salvage values and reduced production were most sensitive to within-herd prevalence. Annual production per cow, volume of bacteria shed by lightly shedding animals, effect of infection on production, and additional culling risk to stage 2 infected animals were also influential variables (Figure 5).
      Figure thumbnail gr4
      Figure 4Sensitivity of estimated losses due to Mycobacterium avium ssp. paratuberculosis (MAP) infection in Canadian dairy herds to a range of input variables (10,000 iterations) assuming initial mean values of 10% within-herd prevalence and 50% herd-level prevalence. The color of the sensitivity bars indicates the direction of the relationship between the variable and estimated losses (gray indicates the effect of variable values below their mean value, white indicates the effect of values above their mean, and black indicates that the effect is unclear). (A) Total annual losses per cow. (B) Annual regional losses.
      Figure thumbnail gr5
      Figure 5Sensitivity of estimated sources of losses due to Mycobacterium avium ssp. paratuberculosis (MAP) infection in Canadian dairy herds to a range of input variables (10,000 iterations) assuming initial mean values of 10% within-herd prevalence and 50% herd-level prevalence. The color of the sensitivity bars indicates the direction of the relationship between the variable and estimated losses (gray indicates the effect of variable values below their mean value, white indicates the effect of values above their mean, and black indicates that the effect is unclear). (A) Annual losses due to premature culling. (B) Annual losses due to reduced salvage value. (C) Annual losses due to reduced production.

      DISCUSSION

      There are 2 main sets of results from this research: first, results for the limited group of regions for which prevalence estimates are available, and second, results for the more comprehensive group of dairy-producing regions with assumed prevalence scenarios. The estimates from the first set may be useful on a region-by-region basis if the prevalence estimates they are based on reflect the true prevalence of MAP infection in those regions. However, relying on these prevalence estimates to valuate JD's economic impact across regions is problematic. Because estimated losses are highly dependent on prevalence and prevalence estimates are heterogeneously derived across studies with no centrally accepted quality assurance and quality control standards in widespread use, it is not possible to confidently compare prevalence estimates across regions. Therefore, variations in estimated losses across regions based on available prevalence estimates may reflect variations across prevalence studies as opposed to variations in economic losses due to JD.
      As an example, refer to the first set of results for Germany and the Netherlands using available prevalence estimates. The neighboring countries have many similar key dairy sector characteristics with above-average annual production per cow per year, similar farm-gate prices, aggregated wage rage rates, replacement costs, and other such production variables. Due to the geographic proximity, level of integration, and similarity of economic characteristics across the 2 dairy industries, it would be intuitively reasonable to expect Germany and the Netherlands to not only have comparable MAP prevalence, but also for MAP infection to have a comparable per cow economic impact on their dairy sectors. However, based on available prevalence estimates, Germany has a herd-level prevalence of 85% and a within-herd prevalence 12.2% (
      • Hacker U.
      • Hüttner K.
      • Konow M.
      (Investigation of serological prevalence and risk factors of paratuberculosis in dairy farms in the state of Mecklenburg-Westpommerania, Germany).
      ), whereas the Netherlands has a herd-level prevalence of 55% and a within-herd prevalence of 2.5% (
      • Muskens J.
      • Barkema H.W.
      • Russchen E.
      • van Maanen K.
      • Schukken Y.H.
      • Bakker D.
      Prevalence and regional distribution of paratuberculosis in dairy herds in the Netherlands.
      ). As a result, estimated losses per cow on infected farms across the 2 regions are significantly different: MAP-infected farms in Germany are estimated to lose US$43.08 per cow per year, or 1.41% of gross milk revenue, whereas infected farms in the Netherlands are estimated to lose only US$15.07 per cow per year, or 0.39% of gross milk revenue. However, when instead using the assumed prevalence scenario of 10% within MAP-infected herds and 50% of herds being infected, there are much closer estimates for the 2 countries: US$36.40 and US$42.82 per cow for Germany and the Netherlands, respectively, both accounting for 1.11% of gross milk revenue. There is a similar situation when comparing the 2 largest dairy-producing provinces in Canada: Québec and Ontario. The neighboring regions share many dairy characteristics, have relatively similar regional cow-level prevalence estimates of 3.1% (
      • Tiwari A.
      Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
      ) and 2.4% (
      • VanLeeuwen J.A.
      • Keefe G.P.
      • Tremblay R.
      • Power C.
      • Wichtel J.J.
      Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
      ), but have significantly different herd-level prevalence estimates of 24% and 54% (
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      ) for Québec and Ontario, respectively. Based on these herd-level and regional cow-level prevalence estimates, Québec would have a mean within-herd prevalence of 13% on infected farms, whereas Ontario would have a mean within-herd prevalence of 4%. As a result, infected herds in Québec are estimated to lose US$60.29 per cow per year or 1.09% of gross milk revenue, whereas infected farms in Ontario are estimated to lose just US$30.77 per cow per year or 0.56% of gross milk revenue. When prevalence is instead assumed to be uniform across the 2 regions, estimated losses are US$56.99 and US$55.35, or 0.97% and 1.01% of revenue for Québec and Ontario, respectively. By using assumed prevalence scenarios across regions, the differences in estimated losses can be wholly attributed to differences in the economic characteristics of the dairy sectors in the 2 regions.
      In the assumed prevalence simulations, the revenue-weighted average losses per cow in MAP-positive herds are approximately 1% of gross milk revenue. Production per cow and farm-gate price are positively related to production losses, which account for an average of 65% of total losses. However, for certain countries such as Ireland, New Zealand, and Australia, which have above-average losses as a percentage of milk revenue, there are interesting underlying structural factors that contribute to the variation across estimates; these 3 countries have average to below-average annual production per cow and farm-gate prices in conjunction with on-average replacement and salvage prices. This combination results in relatively lower milk revenue and therefore production losses, and the values of premature culling and salvage losses having disproportionate effects on total losses relative to other countries. On average for all countries, premature culling and salvage losses account for 24% and 11% of total losses, respectively, but in Ireland, New Zealand, and Australia, they account for 34% to 41% and 16% to 19%, respectively.
      The proportional increase in within-herd MAP prevalence over time is perhaps the most important parameter that is not directly identified by the sensitivity analyses; it is the driver of all estimated economic losses because of its effect on within-herd prevalence. As previously mentioned, it was assumed that within-herd MAP prevalence would, on average, double from its initial value over 10 yr. Because future losses are discounted over time, the greater losses resulting from the higher within-herd prevalence toward the end of the horizon had less of an effect on the 10-yr average annual losses than the relatively lesser losses in the early years. When within-herd prevalence was instead assumed to be stable at 10%, estimated losses decreased by a revenue-weighted average of 13% relative to the estimates obtained from the 10-yr dynamic prevalence models, but similar patterns emerged across regions and the magnitude of the losses were comparable. Other key input variables were revealed through the sensitivity analyses, aside from within-herd and herd-level prevalence, production per cow, and farm-gate price. Not surprisingly, the effect of MAP infection on production is a significant variable for estimated production losses, which account for most of losses per cow, and therefore regional losses due to MAP infection. The bacterial output of lightly MAP-shedding animals was also identified as a significant source of variation in both estimated total losses and sources of losses, particularly for premature culling losses and reduced salvage value losses. Another interesting input variable is the probability of culling associated with subclinical stage 1 MAP infections. This variable was negatively related to all components of total losses aside from premature culling losses. Although this may seem counterintuitive at first, it is logical. The most severe economic damages would occur at later stages of the disease once clinical signs emerge, but most infected animals in the herd are in the subclinical stage. As the risk of stage 1 animals being culled increases, likely for reasons not explicitly attributable to JD but for below-average weight gain, reproductive issues and infertility, or increased susceptibility to other diseases, 2 things happen in the model: the infection pressure, or amount of MAP bacteria present in the herd, decreases, resulting in less severe infections among other animals, and second, the animals are removed before their infections can progress to the more costly stages, from both the health and economic perspectives. However, for premature culling losses, these indirect effects are outweighed by the direct effect of less overall culling, so the overall relationship is positive. It is also important to recognize that the net costs associated with a higher culling rate may be overestimated in this model. Because only the economic impact of culling due to MAP infection was considered, this model ignores the potential benefits associated with having a greater proportion of younger animals in the herd. For example, age-related conditions such as reduced fertility, udder health, and foot health are all potential sources of economic losses that could be partially offset as a direct result of an increased cow-culling rate.
      Last, we discuss the estimates for Canadian herds with consideration for supply management. Although the general method described is appropriate for most dairy industries, the Canadian industry requires special attention. Canada's dairy sector operates with planned and controlled domestic production levels, administered cost-of-production-based pricing of fluid milk, and import controls which help to insulate producers from competitive forces both domestic and foreign. There are 2 consequences relevant to this model: (1) production losses can no longer be measured as forgone milk sales due to the production quota system, and (2) Canada has an above-average farm-gate price, the highest among countries modeled (aside from Japan, which subsidizes dairy production at particularly high rates) and much higher than the farm-gate price in the United States, Canada's most comparable counterpart. Apart from a higher level of total annual output in the United States, both countries have similar dairy sector characteristics in terms of genetics, marketing, consumer preferences, and annual production per cow, and assuming the same within-herd and herd-level MAP prevalence across the 2 countries, there should be similar estimated losses per cow due to MAP infection. However, the above-average farm-gate price in Canada results in a greater valuation of production losses and therefore total losses per cow in Canada, but those losses represent a lower percentage of gross milk revenue. Although these differences are attributable in part to varying technical and allocative efficiencies across dairy sectors, which are not addressed by this study, the effects of the differing market structures are addressed; to reflect the constraint of fixed production levels in Canada, production losses were subsequently re-estimated as the cost of having additional, less productive MAP-positive cows to maintain a fixed level of production over the 10-yr horizon. Once adjusted, estimated annual losses per cow within MAP-infected herds in Canada fell from US$56.99 to US$35.11. Although this is more in line with estimated losses of US$42.26 per cow in the United Sates, as a percentage of gross milk revenue, Canadian losses fell from 1.00% to 0.62% compared with 1.12% in the United States. This may be an overcorrection. Although overall production and farm-gate prices in Canada are set annually by the Canadian Dairy Commission and producers can only sell the amount of milk for which they have production quota, there is still competition among producers. The overall level of production generally increases year over year (
      • CDIC
      D037–3 - Average Production by Province.
      ) and producers trade quota among themselves; more profitable producers purchase quota from less profitable ones through a quota exchange market to increase the size of their operations. The number of dairy farms in Canada has steadily decreased over the last several decades while the size of herds has increased (
      • CDIC
      D056 - Number of Farms with Shipments of Milk by Province.
      ). In other words, producers operate in an environment somewhere in between fixed production and a purely competitive market, and therefore true losses due to MAP infection in infected Canadian dairy herds likely lie somewhere in between the 2 estimates, or between US$35.11 and US$56.99 per cow per year.
      Although the estimated losses due to JD in major dairy-producing regions are significant, they are likely much less than the losses associated with more easily addressable herd health issues such as mastitis. It may seem intuitive that healthier dairy herds generate more profits for producers, but dairy production is complex and producers are often forced to prioritize short-term, immediate losses from relatively acute diseases such as mastitis over long-term losses from chronic diseases such as JD. However, if left unchecked, JD will continue to spread, losses will continue to increase, and controlling the disease will become increasingly difficult. It is also important to recognize that research into potential interactions between MAP and other pathogens that affect dairy cows is still developing. For example, higher mastitis incidences (
      • Diéguez F.J.
      • Arnaiz I.
      • Sanjuán M.L.
      • Vilar M.J.
      • Yus E.
      Management practices associated with Mycobacterium avium subspecies paratuberculosis infection and the effects of the infection on dairy herds.
      ) and higher culling rates due to clinical mastitis (
      • Arrazuría R.
      • Arnaiz I.
      • Fouz R.
      • Calvo C.
      • Eiras C.
      • Diéguez F.J.
      Association between Mycobacterium avium ssp. paratuberculosis infection and culling in dairy cattle herds.
      ) have been observed in MAP-positive herds, and MAP-positive cows have been shown to have a higher incidence of clinical mastitis (
      • Rossi G.
      • Grohn Y.T.
      • Schukken Y.H.
      • Smith R.L.
      The effect of Mycobacterium avium ssp. paratuberculosis infection on clinical mastitis occurrence in dairy cows.
      ). National control programs have already been established in several countries including Australia, Ireland, Japan, the Netherlands, and the United States, and JD is listed by the World Organisation for Animal Health as a priority disease for international trade. The dairy herd-level MAP prevalence in Canada has recently been estimated to be 42% (
      • Corbett C.S.
      • Naqvi S.A.
      • Bauman C.A.
      • De Buck J.
      • Orsel K.
      • Uehlinger F.
      • Kelton D.F.
      • Barkema H.W.
      Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
      ), yet Canada has no mandatory control program in place despite JD's pervasiveness. MAP infection in dairy herds may also pose a public health threat; there is concern that MAP may be associated with Crohn's disease in humans (
      • El-Zaatari F.A.K.
      • Osato M.S.
      • Graham D.Y.
      Etiology of Crohn's disease: The role of Mycobacterium avium paratuberculosis..
      ;
      • Harris J.E.
      • Lammerding A.M.
      Crohn's disease and Mycobacterium avium ssp. paratuberculosis: Current issues.
      ;
      • Hermon-Taylor J.
      Mycobacterium avium subspecies paratuberculosis is a cause of Crohn's disease.
      ;
      • Chacon O.
      • Bermudez L.E.
      • Barletta R.G.
      Johne's disease, inflammatory bowel disease, and Mycobacterium paratuberculosis..
      ;
      • Naser S.A.
      • Ghobrial G.
      • Romero C.
      • Valentine J.F.
      Culture of Mycobacterium avium subspecies paratuberculosis from the blood of patients with Crohn's disease.
      ;
      • Feller M.
      • Huwiler K.
      • Stephan R.
      • Altpeter E.
      • Shang A.
      • Furrer H.
      • Pfyffer G.E.
      • Jemmi T.
      • Baumgartner A.
      • Egger M.
      Mycobacterium avium subspecies paratuberculosis and Crohn's disease: a systematic review and meta-analysis.
      ;
      • Abubakar I.
      • Myhill D.
      • Aliyu S.H.
      • Hunter P.R.
      Detection of Mycobacterium avium subspecies paratuberculosis from patients with Crohn's disease using nucleic acid-based techniques: A systematic review and meta-analysis.
      ;
      • Waddell L.A.
      • Rajić A.
      • Stärk K.D.
      • McEwen E.S.
      The zoonotic potential of Mycobacterium avium ssp. paratuberculosis: a systematic review and meta-analyses of the evidence.
      ), infected cows can shed live MAP in both their feces and milk, and live MAP have been recovered from pasteurized milk (
      • Ellingson J.L.E.
      • Anderson J.L.
      • Koziczkowski J.J.
      • Radcliff R.P.
      • Sloan S.J.
      • Allen S.E.
      • Sullivan N.M.
      Detection of viable Mycobacterium avium ssp. paratuberculosis in retail pasteurized whole milk by two culture methods and PCR.
      ;
      • Shankar H.
      • Singh S.V.
      • Singh P.K.
      • Singh A.V.
      • Sohal J.S.
      • Greenstein R.J.
      Presence, characterization, and genotype profiles of Mycobacterium avium subspecies paratuberculosis from unpasteurized individual and pooled milk, commercial pasteurized milk, and milk products in India by culture, PCR, and PCR-REA methods.
      ).

      CONCLUSIONS

      Although the economic impact of JD in dairy herds has been estimated before, this study is unique in 2 ways: first, it estimates economic losses due to JD across a comprehensive selection of major dairy-producing regions within one methodological framework, and second, it attempts to capture the relationship between economic losses due to JD and the market conditions that arise as a result of supply management in Canada. With assumptions of 10% within-herd prevalence, 50% herd-level prevalence, and a doubling of within-herd prevalence over 10 yr, an estimated 1% of gross milk revenue is lost annually in MAP-positive dairy herds. This translates to revenue-weighted average losses of US$33 per cow per year on infected farms, with greater losses in regions with higher farm-gate prices and production per cow. Twenty-four percent of those losses are attributable to premature culling, 11% are attributable to reduced salvage values, and 65% are attributable to reduced production. Each year, an estimated US$198 million is lost due to JD in the United States, US$75 million in Germany, US$56 million in France, US$54 million in New Zealand, and between US$17 million and US$28 million in Canada, one of the smallest dairy-producing regions modeled. As research into MAP infection in dairy herds continues to expand, the input values used in these models can be adjusted and updated, perhaps providing evidence of an even greater direct economic incentive for producers to control this disease and for the continued development of new testing methods and pharmaceutical interventions such as vaccines. This research is an important contribution to the policy discussion surrounding paratuberculosis control in Canada and internationally.

      ACKNOWLEDGMENTS

      This research benefited from discussions with Jeroen De Buck, Karin Orsel, and Paul Burden from the University of Calgary, Departments of Production Animal Health and Ecosystem and Public Health, and Joseph Rasmussen from the University of Lethbridge, Department of Biological Sciences. This research was supported by Genome Canada, Genome Prairie, and Genome British Columbia (225RVA). The authors have not stated any conflicts of interest.

      APPENDIX

      Table A1.Key dairy sector characteristics for various dairy-producing regions in order of decreasing 2018 annual production
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Annual production
      Canadian production values: CDIC–Average production based on official–Supervised records (CDIC, 2019a). US annual production values: USDA ERS–Dairy data–Milk cows and production by state and region (USDA ERS, 2019). All other regions: CLAL.it – Dairy by country (CLAL, 2019).
      (1,000 Mt)
      Annual production
      Canadian annual per cow production values: CDIC–Average production based on official–Supervised records (CDIC, 2019a). US production per cow values: USDA ERS–Milk cows and production by state and region (USDA ERS, 2019). All other regions: CLAL.it–Dairy by country (CLAL, 2019).
      (kg/cow)
      Dairy cattle
      Canadian cattle values: STATCAN–Table 32–10–0130–01–Number of cattle, by class and farm type (× 1,000) (STATCAN, 2020). US cattle values: USDA ERS–Dairy data–Milk cows and production by state and region (USDA ERS, 2019). All other regions: CLAL.it–Dairy by country (CLAL, 2019).
      (1,000 head)
      Farm-gate price
      Turkey: 2018 EU-28. Australian price: Australian Dairy Industry in Focus–2018 (Dairy Australia, 2019). Canadian values: CDIC–MI011–Canadian farm cash receipts from dairying (CDIC, 2019c). All other regions: CLAL.it–Dairy by country (CLAL, 2019). Converted to 2018 US$ using IRS.gov–Yearly average currency exchange rates (IRS, 2020).
      (US$/100 kg)
      EU-28166,7447,27922,90640.22
       DEU33,0878,0684,10140.53
       FRA25,0557,0583,55039.92
       GBR15,4888,2431,87939.07
       POL14,1716,4012,21437.71
       NLD14,0909,0791,55242.50
       ITA12,3407,2891,69344.08
       IRL7,8315,7201,36939.20
       ESP7,3368,96881836.83
       DNK5,6159,85157042.41
       BEL4,1787,89852937.96
       AUT3,8217,16953338.77
       CZE3,1628,80835939.33
       SWE2,7608,81831340.85
       FIN2,3989,08326444.72
      USA98,68810,5469,35835.86
       CA18,33110,5721,73435.86
       WI13,87010,8871,27435.86
       ID6,87111,28360935.86
       NY6,75010,83562335.86
       TX5,83010,85653735.86
       MI5,06611,94742435.86
       PA4,8389,32151935.86
       MN4,4769,88145335.86
       NM3,75811,38833035.86
       WA3,05511,03027735.86
      BRA33,4911,96317,06041.71
      CHN
      2017 values.
      30,6402,56311,95553.87
      RUS30,6114,4926,81536.58
      NZL21,9474,4374,94635.72
      TUR20,0373,1616,33840.22
      CAN10,22810,51997253.93
       QC3,67310,36935453.11
       ON3,37710,43232452.72
       BC90610,8038458.81
       AB86710,9657955.95
       MB44810,8804154.14
       SK32911,2562953.74
       NS22610,5662153.17
       NB1919,9851953.59
       PE15110,4941454.02
       NL6410,778670.74
      AUS9,1766,0171,52534.00
      JPN7,2908,60784793.15
      1 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      2 Canadian production values: CDIC–Average production based on official–Supervised records (
      • CDIC
      D037–3 - Average Production by Province.
      ). US annual production values: USDA ERS–Dairy data–Milk cows and production by state and region (
      • USDA ERS
      Dairy data – milk and cow production by State and region.
      ). All other regions: CLAL.it – Dairy by country (
      • CLAL
      Dairy Sector by Country.
      ).
      3 Canadian annual per cow production values: CDIC–Average production based on official–Supervised records (
      • CDIC
      D037–3 - Average Production by Province.
      ). US production per cow values: USDA ERS–Milk cows and production by state and region (
      • USDA ERS
      Dairy data – milk and cow production by State and region.
      ). All other regions: CLAL.it–Dairy by country (
      • CLAL
      Dairy Sector by Country.
      ).
      4 Canadian cattle values: STATCAN–Table 32–10–0130–01–Number of cattle, by class and farm type (× 1,000) (
      • STATCAN
      Table 32–10–0130–01 Number of cattle, by class and farm type (x 1,000).
      ). US cattle values: USDA ERS–Dairy data–Milk cows and production by state and region (
      • USDA ERS
      Dairy data – milk and cow production by State and region.
      ). All other regions: CLAL.it–Dairy by country (
      • CLAL
      Dairy Sector by Country.
      ).
      5 Turkey: 2018 EU-28. Australian price: Australian Dairy Industry in Focus–2018 (). Canadian values: CDIC–MI011–Canadian farm cash receipts from dairying (
      • CDIC
      MI011 - Canadian farm cash receipts from dairying.
      ). All other regions: CLAL.it–Dairy by country (
      • CLAL
      Dairy Sector by Country.
      ). Converted to 2018 US$ using IRS.gov–Yearly average currency exchange rates (
      • IRS
      Yearly average currency exchange rates.
      ).
      6 2017 values.
      Table A2.Region-specific aggregated input variables used in the Monte Carlo simulations of the Mycobacterium avium ssp. paratuberculosis-positive Markov herd model
      All variables simulated with a normal distribution and standard deviation of 10% of the mean.
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      gdppc
      Gross domestic product per capita (gdppc). US gdppc by state (BEA, 2019) and Canadian gross domestic product by province (STATCAN, 2019c). Converted to US$ (IRS, 2020). All other regions (World Bank, 2020).
      (US$)
      Wage
      Estimated aggregate dairy wage rate (wage). US 2018 (USDA NASS, 2019). All other regions (i through n) calculated using the following formula: wagei=wageUSA×gdppci×farm−gatepriceifarm−gatepriceUSA×gdppcUSA.
      (US$/h)
      r-c
      hd = head. Calf (c), open heifer (oh), bred heifer (bh), and mature cow (m) replacement (r) costs. US replacement prices (USDA, 2020). All other regions (i through n) calculated using the following formula: ri=rUSA×gdppci×farm−gatepriceifarm−gatepriceUSA×gdppcUSA.
      (US$/hd)
      r-oh
      hd = head. Calf (c), open heifer (oh), bred heifer (bh), and mature cow (m) replacement (r) costs. US replacement prices (USDA, 2020). All other regions (i through n) calculated using the following formula: ri=rUSA×gdppci×farm−gatepriceifarm−gatepriceUSA×gdppcUSA.
      (US$/hd)
      r-bh
      hd = head. Calf (c), open heifer (oh), bred heifer (bh), and mature cow (m) replacement (r) costs. US replacement prices (USDA, 2020). All other regions (i through n) calculated using the following formula: ri=rUSA×gdppci×farm−gatepriceifarm−gatepriceUSA×gdppcUSA.
      (US$/hd)
      r-m
      hd = head. Calf (c), open heifer (oh), bred heifer (bh), and mature cow (m) replacement (r) costs. US replacement prices (USDA, 2020). All other regions (i through n) calculated using the following formula: ri=rUSA×gdppci×farm−gatepriceifarm−gatepriceUSA×gdppcUSA.
      (US$/hd)
      s-12
      0–12-mo animals (12), 12–24-mo animals (24), and mature cow (m) salvage (s) prices. Canadian salvage prices (STATCAN, 2019a). Converted to kg at 50.8023 kg/cwt (hundredweight) and converted to US$ (IRS, 2020). All other regions (i through n) calculated using the following formula: si=sCAN×gdppci×farm−gatepriceifarm−gatepriceCAN×gdppcCAN.
      (US$/kg)
      s-24
      0–12-mo animals (12), 12–24-mo animals (24), and mature cow (m) salvage (s) prices. Canadian salvage prices (STATCAN, 2019a). Converted to kg at 50.8023 kg/cwt (hundredweight) and converted to US$ (IRS, 2020). All other regions (i through n) calculated using the following formula: si=sCAN×gdppci×farm−gatepriceifarm−gatepriceCAN×gdppcCAN.
      (US$/kg)
      s-m
      0–12-mo animals (12), 12–24-mo animals (24), and mature cow (m) salvage (s) prices. Canadian salvage prices (STATCAN, 2019a). Converted to kg at 50.8023 kg/cwt (hundredweight) and converted to US$ (IRS, 2020). All other regions (i through n) calculated using the following formula: si=sCAN×gdppci×farm−gatepriceifarm−gatepriceCAN×gdppcCAN.
      (US$/kg)
      EU-2839,92810.0859.94375.71650.80764.172.121.690.71
       DEU47,60312.1172.01451.35781.82918.022.552.030.85
       FRA41,46410.3961.78387.19670.69787.522.191.740.73
       GBR42,94410.5362.62392.48679.86798.292.221.760.74
       POL15,4213.6521.71136.05235.66276.710.770.610.26
       NLD53,02414.1584.11527.18913.171,072.252.982.370.99
       ITA34,4839.5456.73355.59615.94723.242.011.600.67
       IRL78,80619.39115.29722.641,251.751,469.814.083.251.36
       ESP30,3717.0241.75261.65453.24532.191.481.180.49
       DNK61,35016.3397.10608.601,054.221,237.873.442.731.15
       BEL47,51911.3267.32421.97730.94858.272.391.900.79
       AUT51,46212.5374.47466.79808.56949.412.642.100.88
       CZE23,0795.7033.88212.33367.79431.861.200.950.40
       SWE54,60814.0083.26521.84903.921,061.392.952.340.98
       FIN50,15214.0883.70524.64908.771,067.082.972.360.99
      USA62,79514.1484.05526.79912.501,071.462.982.370.99
       CA60,35913.5980.79506.36877.111,029.902.862.270.95
       WI48,66610.9665.14408.26707.19830.382.311.830.77
       ID36,4418.2048.77305.71529.54621.791.731.370.58
       NY65,22014.6887.29547.14947.741,112.843.092.461.03
       TX53,73712.1071.92450.80780.88916.912.552.020.85
       MI44,2019.9559.16370.81642.31754.202.101.670.70
       PA51,84111.6769.39434.90753.33884.562.461.950.82
       MN54,80512.3473.35459.76796.40935.132.602.060.87
       NM41,6199.3755.71349.15604.79710.141.971.570.66
       WA59,33313.3679.41497.75862.201,012.392.812.240.94
      BRA8,9212.3413.8987.04150.78177.040.490.390.16
      CHN9,7713.3019.64123.13213.28250.430.700.550.23
      RUS11,2892.5915.4196.60167.34196.490.550.430.18
      NZL41,9459.4155.92350.49607.12712.891.981.570.66
      TUR9,3702.3714.0788.17152.73179.340.500.400.17
      CAN46,26915.6693.13583.691,011.061,187.193.302.621.10
       QC40,38913.4780.06501.77869.161,020.572.842.250.95
       ON46,16715.2890.85569.43986.361,158.183.222.561.07
       BC45,54016.8199.96626.551,085.311,274.383.542.811.18
       AB61,81621.71129.08809.031,401.401,645.534.573.631.52
       MB41,40914.0883.68524.51908.551,066.822.962.360.99
       SK53,48718.04107.27672.371,164.681,367.573.803.021.27
       NS35,64111.9070.73443.34767.95901.732.511.990.84
       NB36,97012.4473.94463.44802.76942.602.622.080.87
       PE35,11111.9170.79443.67768.53902.412.511.990.84
       NL48,76121.57128.25803.861,392.441,635.014.543.611.51
      AUS57,37412.2572.81456.34790.46928.162.582.050.86
      JPN39,29022.98136.60856.161,483.021,741.374.843.841.61
      1 All variables simulated with a normal distribution and standard deviation of 10% of the mean.
      2 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      3 Gross domestic product per capita (gdppc). US gdppc by state (
      • BEA
      Regional Economic Accounts – SAGDP Tables: Annual GDP by State.
      ) and Canadian gross domestic product by province (
      • STATCAN
      Table 36–10–0222–01 Gross domestic product, expenditure-based, provincial and territorial, annual (x 1,000,000).
      ). Converted to US$ (
      • IRS
      Yearly average currency exchange rates.
      ). All other regions (
      • World Bank
      GDP per capita (current US$).
      ).
      4 Estimated aggregate dairy wage rate (wage). US 2018 (
      • USDA NASS
      Farm Labor – 05/30/2019.
      ). All other regions (i through n) calculated using the following formula: wagei=wageUSA×gdppci×farmgatepriceifarmgatepriceUSA×gdppcUSA.
      5 hd = head. Calf (c), open heifer (oh), bred heifer (bh), and mature cow (m) replacement (r) costs. US replacement prices (
      • USDA
      National Dairy Comprehensive Report - Monthly - 02/29/2020.
      ). All other regions (i through n) calculated using the following formula: ri=rUSA×gdppci×farmgatepriceifarmgatepriceUSA×gdppcUSA.
      6 0–12-mo animals (12), 12–24-mo animals (24), and mature cow (m) salvage (s) prices. Canadian salvage prices (
      • STATCAN
      Table 32–10–0077–01 Farm product prices, crops and livestock.
      ). Converted to kg at 50.8023 kg/cwt (hundredweight) and converted to US$ (
      • IRS
      Yearly average currency exchange rates.
      ). All other regions (i through n) calculated using the following formula: si=sCAN×gdppci×farmgatepriceifarmgatepriceCAN×gdppcCAN.
      Table A3.Estimated mean value 10-yr average annual losses (US$ per cow in Mycobacterium avium ssp. paratuberculosis-positive herds and millions of US$ per region) across a range of prevalence scenarios
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Prevalence scenarios (within-herd prevalence: herd-level prevalence)
      5%: 50%15%: 50%10%: 30%10%: 70%
      Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)
      EU-2818.77214.9240.82467.4930.40208.8932.41519.62
       DEU21.4844.0346.9396.2234.8942.9237.19106.77
       FRA18.3432.5539.9370.8730.8532.8532.5080.76
       GBR20.3019.0744.0241.3532.8218.5034.9946.02
       POL12.6313.9926.3129.1319.9213.2321.2332.90
       NLD25.2519.5955.1442.7941.0019.0943.7247.49
       ITA19.6716.6542.3935.8931.6816.0933.7740.02
       IRL21.1814.5049.0533.5835.7014.6638.0836.49
       ESP18.527.5739.1916.0329.497.2431.4318.00
       DNK27.967.9761.3017.4745.517.7848.5319.36
       BEL19.825.2443.3711.4732.235.1134.3612.72
       AUT19.615.2343.4011.5732.115.1434.2412.78
       CZE18.413.3038.486.9129.093.1331.017.79
       SWE24.063.7752.738.2539.163.6841.759.15
       FIN26.093.4456.777.4942.273.3545.068.33
      USA24.91116.5754.47254.8840.49113.6743.17282.77
       CA24.6221.3553.6946.5539.9520.7842.5951.69
       WI23.5114.9750.5132.1837.7914.4440.2835.92
       ID22.456.8347.3914.4335.696.5238.0416.22
       NY25.698.0056.2217.5141.777.8144.5419.42
       TX24.156.4952.2114.0238.976.2841.5515.62
       MI24.545.2052.1711.0639.184.9841.7712.40
       PA21.525.5846.8512.1634.885.4337.1813.51
       MN22.795.1649.6111.2436.945.0239.3812.49
       NM23.323.8549.548.1737.223.6939.689.16
       WA25.193.4954.737.5840.773.3943.478.43
      BRA4.9542.2510.6590.877.9740.788.49101.43
      CHN7.9747.6216.96101.4012.7345.6713.57113.59
      RUS8.6729.5318.0861.6113.6827.9714.5869.54
      NZL12.5631.0628.3170.0120.8130.8722.1976.82
      TUR6.9321.9514.5646.1410.9820.8811.7151.93
      CAN34.0016.5372.9935.4854.6215.9358.2339.63
       QC31.955.6668.1112.0651.115.4354.4813.51
       ON33.025.3470.9211.4853.075.1556.5712.82
       BC37.641.5880.603.3860.381.5264.363.78
       AB39.671.5786.443.4264.331.5368.593.80
       MB33.980.7072.351.4954.310.6757.891.67
       SK37.070.5479.941.1759.720.5263.671.30
       NS31.480.3466.600.7150.110.3253.410.80
       NB30.650.2965.160.6348.950.2852.170.70
       PE31.700.2367.040.4850.450.2253.780.54
       NL45.890.1498.670.3073.800.1378.670.33
      AUS16.2712.4136.7127.9926.9712.3428.7630.71
      JPN48.5620.57104.4744.2478.1219.8583.2849.38
      1 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Table A4.Estimated sources of losses per cow in various dairy-producing regions assuming a Mycobacterium avium ssp. paratuberculosis herd-level prevalence of 50% and a within-herd prevalence 10%
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Proportion of losses
      Premature cullingReduced salvage valueReduced production
      EU-280.250.120.63
       DEU0.260.120.61
       FRA0.260.120.62
       GBR0.240.110.64
       POL0.140.060.80
       NLD0.260.120.62
       ITA0.230.110.67
       IRL0.410.190.40
       ESP0.180.080.74
       DNK0.270.130.60
       BEL0.270.120.61
       AUT0.300.140.57
       CZE0.150.070.78
       SWE0.270.130.60
       FIN0.250.120.63
      USA0.270.120.61
       CA0.260.120.62
       WI0.220.100.68
       ID0.170.080.74
       NY0.270.120.61
       TX0.240.110.66
       MI0.190.090.72
       PA0.250.120.63
       MN0.250.120.63
       NM0.190.090.72
       WA0.250.110.64
      BRA0.220.100.67
      CHN0.200.090.71
      RUS0.140.070.79
      NZL0.340.160.50
      TUR0.160.080.76
      CAN0.220.100.68
       QC0.200.090.71
       ON0.220.100.68
       BC0.210.100.69
       AB0.260.120.63
       MB0.200.090.71
       SK0.230.110.66
       NS0.180.080.74
       NB0.190.090.72
       PE0.180.080.74
       NL0.220.100.68
      AUS0.340.160.50
      JPN0.220.100.67
      1 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Table A5.Estimated mean value annual losses (US$ per cow in Mycobacterium avium ssp. paratuberculosis-positive herds and millions US$ per region) with an assumed stable within-herd prevalence of 10% and a herd-level prevalence of 50%
      Region
      EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Annual losses (US$ per cow)Annual losses (% of milk revenue)Annual losses (millions of US$ per region)
      EU-2827.670.95316.89
       DEU31.670.9764.94
       FRA27.110.9648.12
       GBR29.930.9328.12
       POL18.600.7720.59
       NLD37.240.9728.90
       ITA28.990.9024.54
       IRL31.301.4021.43
       ESP27.280.8311.16
       DNK41.230.9911.75
       BEL29.230.987.73
       AUT28.941.047.71
       CZE27.100.784.87
       SWE35.480.995.55
       FIN38.460.955.08
      USA36.740.97171.92
       CA36.310.9631.48
       WI34.640.8922.07
       ID33.060.8210.07
       NY37.890.9811.80
       TX35.600.919.56
       MI36.150.847.66
       PA31.730.958.23
       MN33.610.957.61
       NM34.350.845.67
       WA37.140.945.14
      BRA7.300.8962.27
      CHN11.740.8570.15
      RUS12.760.7843.47
      NZL18.541.1745.86
      TUR10.200.8032.32
      CAN50.110.8824.36
       QC47.070.858.34
       ON48.670.887.88
       BC55.460.872.33
       AB58.500.952.31
       MB50.060.851.03
       SK54.640.900.80
       NS46.360.830.50
       NB45.150.840.43
       PE46.690.820.34
       NL67.630.890.20
      AUS24.031.1718.32
      JPN71.570.8930.31
      1 EU-28 = European Union; DEU = Germany; FRA = France; GBR = Great Britain; POL = Poland; NLD = the Netherlands; ITA = Italy; IRL = Ireland; ESP = Spain; DNK = Denmark; BEL = Belgium; AUT = Austria; CZE = Czechia; SWE = Sweden; FIN = Finland; USA = United States; CA = California; WI = Wisconsin; ID = Idaho; NY = New York; TX = Texas; MI = Michigan; PA = Pennsylvania; MN = Minnesota; NM = New Mexico; WA = Washington; BRA = Brazil; CHN = China; RUS = Russia; NZL = New Zealand; TUR = Turkey; CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland; AUS = Australia; JPN = Japan.
      Table A6.Estimated sources of losses per cow in Canadian regions assuming a Mycobacterium avium ssp. paratuberculosis herd-level prevalence of 50% and a within-herd prevalence 10%, and with consideration for supply management (fixed output over time and production losses allocated as increased variable costs necessary to maintain production)
      Region
      CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.
      Proportion of losses
      Premature cullingReduced salvage valueReduced production
      CAN0.350.160.48
       QC0.330.150.52
       ON0.370.170.47
       BC0.320.150.53
       AB0.370.170.46
       MB0.300.140.56
       SK0.360.170.48
       NS0.300.140.56
       NB0.310.150.54
       PE0.330.150.52
       NL0.320.150.53
      1 CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.
      Table A7.Estimated mean value 10-yr average annual losses (US$ per cow in Mycobacterium avium ssp. paratuberculosis-positive herds and millions US$ per region) for Canadian dairy across a range of prevalence scenarios, and with consideration for supply management (fixed output over time and production losses allocated as increased variable costs necessary to maintain production)
      Region
      CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.
      Prevalence scenarios (within-herd prevalence: herd-level prevalence)
      5%: 50%15%: 50%10%: 30%10%: 70%
      Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)Per cow (US$)Region (millions of US$)
      CAN20.219.8245.7522.2433.579.7935.8024.37
       QC18.683.3141.907.4230.853.2832.908.16
       ON19.033.0843.277.0031.703.0833.817.66
       BC24.051.0153.762.2639.631.0042.262.48
       AB26.611.0560.642.4044.391.0547.342.62
       MB21.560.4447.830.9935.360.4437.711.09
       SK22.990.3452.130.7638.230.3340.770.83
       NS18.080.1940.150.4329.670.1931.640.47
       NB18.200.1740.570.3929.940.1731.920.43
       PE16.490.1237.010.2727.240.1229.050.29
       NL30.860.0968.990.2150.860.0954.240.23
      1 CAN = Canada; QC = Québec; ON = Ontario; BC = British Columbia; AB = Alberta; MB = Manitoba; SK = Saskatchewan; NS = Nova Scotia; PE = Prince Edward Island; NL = Newfoundland.

      REFERENCES

        • Abubakar I.
        • Myhill D.
        • Aliyu S.H.
        • Hunter P.R.
        Detection of Mycobacterium avium subspecies paratuberculosis from patients with Crohn's disease using nucleic acid-based techniques: A systematic review and meta-analysis.
        Inflamm. Bowel Dis. 2008; 14 (17886288): 401-410
        • Arrazuría R.
        • Arnaiz I.
        • Fouz R.
        • Calvo C.
        • Eiras C.
        • Diéguez F.J.
        Association between Mycobacterium avium ssp. paratuberculosis infection and culling in dairy cattle herds.
        Arch. Med. Vet. 2014; 46: 39-44
        • Barkema H.W.
        • Orsel K.
        • Nielsen S.S.
        • Koets A.P.
        • Rutten V.P.M.G.
        • Bannantine J.P.
        • Keefe G.P.
        • Kelton D.F.
        • Wells S.J.
        • Whittington R.J.
        • Mackintosh C.G.
        • Manning E.J.
        • Weber M.F.
        • Heuer C.
        • Forde T.L.
        • Ritter C.
        • Roche S.
        • Corbett C.S.
        • Wolf R.
        • Griebel P.J.
        • Kastelic J.P.
        • De Buck J.
        Knowledge gaps that hamper prevention and control of Mycobacterium avium subspecies paratuberculosis infection.
        Transbound. Emerg. Dis. 2018; 65 (28941207): 125-148
        • BEA
        Regional Economic Accounts – SAGDP Tables: Annual GDP by State.
        https://www.bea.gov/data/gdp/gdp-state
        Date: 2019
        Date accessed: January 13, 2020
        • Beard P.M.
        • Daniels M.J.
        • Henderson D.
        • Pirie A.
        • Rudge K.
        • Buxton D.
        • Rhind S.
        • Greig A.
        • Hutchings M.R.
        • McKendrick I.
        • Stevenson K.
        • Sharp J.M.
        Paratuberculosis infection of nonruminant wildlife in Scotland.
        J. Clin. Microbiol. 2001; 39 (11283080): 1517-1521
        • Benedictus G.
        • Dijkhuizen A.A.
        • Stelwagen J.
        (Economic losses to farms due to paratuberculosis in cattle).
        Tijdschr. Diergeneeskd. 1985; 110 (4040661): 310-319
        • Bhattarai B.
        • Fosgate G.T.
        • Osterstock J.B.
        • Fossler C.P.
        • Park S.C.
        • Roussel A.J.
        Perceptions of veterinarians in bovine practice and producers with beef cow-calf operations enrolled in the US Voluntary Bovine Johne's Disease Control Program concerning economic losses associated with Johne's disease.
        Prev. Vet. Med. 2013; 112 (24034813): 330-337
        • Boelaert F.
        • Walravens K.
        • Biront P.
        • Vermeersch J.P.
        • Berkvens D.
        • Godfroid J.
        Prevalence of paratuberculosis (Johne's disease) in the Belgian cattle population.
        Vet. Microbiol. 2000; 77 (11118712): 269-281
        • Bolton M.W.
        • Pillars R.B.
        • Kaneene J.B.
        • Mauer W.A.
        • Grooms D.L.
        Detection of Mycobacterium avium subspecies paratuberculosis in naturally exposed dairy heifers and associated risk factors.
        J. Dairy Sci. 2011; 94 (21854939): 4669-4675
        • CDIC
        D037–3 - Average Production by Province.
        • CDIC
        D056 - Number of Farms with Shipments of Milk by Province.
        • CDIC
        MI011 - Canadian farm cash receipts from dairying.
        • Chacon O.
        • Bermudez L.E.
        • Barletta R.G.
        Johne's disease, inflammatory bowel disease, and Mycobacterium paratuberculosis..
        Annu. Rev. Microbiol. 2004; 58 (15487941): 329-363
        • CLAL
        Dairy Sector by Country.
        https://www.clal.it/en/
        Date: 2019
        Date accessed: September 15, 2020
        • Collins M.T.
        • Sockett D.C.
        • Goodger W.J.
        • Conrad T.A.
        • Thomas C.B.
        • Carr D.J.
        Herd prevalence and geographic distribution of, and risk factors for, bovine paratuberculosis in Wisconsin.
        J. Am. Vet. Med. Assoc. 1994; 204 (8163422): 636-641
        • Corbett C.S.
        • Naqvi S.A.
        • Bauman C.A.
        • De Buck J.
        • Orsel K.
        • Uehlinger F.
        • Kelton D.F.
        • Barkema H.W.
        Prevalence of Mycobacterium avium ssp. paratuberculosis infections in Canadian dairy herds.
        J. Dairy Sci. 2018; 101 (30316607): 11218-11228
        • Crossley B.M.
        • Zagmutt-Vergara F.J.
        • Fyock T.L.
        • Whitlock R.H.
        • Gardner I.A.
        Fecal shedding of Mycobacterium avium ssp. paratuberculosis by dairy cows.
        Vet. Microbiol. 2005; 107 (15863285): 257-263
        • Dairy Australia
        Dairy in Focus - 2018.
        • Diéguez F.J.
        • Arnaiz I.
        • Sanjuán M.L.
        • Vilar M.J.
        • López M.
        • Yus E.
        Prevalence of serum antibodies to Mycobacterium avium ssp. paratuberculosis in cattle in Galicia (northwest Spain).
        Prev. Vet. Med. 2007; 82 (17928078): 321-326
        • Diéguez F.J.
        • Arnaiz I.
        • Sanjuán M.L.
        • Vilar M.J.
        • Yus E.
        Management practices associated with Mycobacterium avium subspecies paratuberculosis infection and the effects of the infection on dairy herds.
        Vet. Rec. 2008; 162 (18480020): 614-617
        • Donaghy J.
        • Keyser M.
        • Johnston J.
        • Cilliers F.P.
        • Gouws P.A.
        • Rowe M.T.
        Inactivation of Mycobacterium avium ssp. paratuberculosis in milk by UV treatment.
        Lett. Appl. Microbiol. 2009; 49 (19486288): 217-221
        • El-Zaatari F.A.K.
        • Osato M.S.
        • Graham D.Y.
        Etiology of Crohn's disease: The role of Mycobacterium avium paratuberculosis..
        Trends Mol. Med. 2001; 7 (11378513): 247-252
        • Ellingson J.L.E.
        • Anderson J.L.
        • Koziczkowski J.J.
        • Radcliff R.P.
        • Sloan S.J.
        • Allen S.E.
        • Sullivan N.M.
        Detection of viable Mycobacterium avium ssp. paratuberculosis in retail pasteurized whole milk by two culture methods and PCR.
        J. Food Prot. 2005; 68 (15895728): 966-972
        • Fecteau M.E.
        • Whitlock R.H.
        Paratuberculosis in cattle.
        in: Behr M.A. Collins D.M. Paratuberculosis: Organism, Disease, Control. CABI, Wallingford, UK2010: 144-156
        • Fecteau M.-E.
        • Whitlock R.H.
        • Buergelt C.D.
        • Sweeney R.W.
        Exposure of young dairy cattle to Mycobacterium avium ssp. paratuberculosis (MAP) through intensive grazing of contaminated pastures in a herd positive for Johne's disease.
        Can. Vet. J. 2010; 51 (20436867): 198-200
        • Feller M.
        • Huwiler K.
        • Stephan R.
        • Altpeter E.
        • Shang A.
        • Furrer H.
        • Pfyffer G.E.
        • Jemmi T.
        • Baumgartner A.
        • Egger M.
        Mycobacterium avium subspecies paratuberculosis and Crohn's disease: a systematic review and meta-analysis.
        Lancet Infect. Dis. 2007; 7 (17714674): 607-613
        • Garcia A.B.
        • Shalloo L.
        Invited review: The economic impact and control of paratuberculosis in cattle.
        J. Dairy Sci. 2015; 98 (26074241): 5019-5039
        • Gasteiner J.
        • Wenzl H.
        • Fuchs K.
        • Jark U.
        • Baumgartner W.
        Serological cross-sectional study of paratuberculosis in cattle in Austria.
        J. Vet. Med. B Infect. Dis. Vet. Public Health. 1999; 46 (10528542): 457-466
        • Good M.
        • Clegg T.
        • Sheridan H.
        • Yearsely D.
        • O'Brien T.
        • Egan J.
        • Mullowney P.
        Prevalence and distribution of paratuberculosis (Johne's disease) in cattle herds in Ireland.
        Ir. Vet. J. 2009; 62 (21851740): 597-606
        • Greig A.
        • Stevenson K.
        • Henderson D.
        • Perez V.
        • Hughes V.
        • Pavlik I.
        • Hines M.E.
        • McKendrick I.
        • Sharp J.M.
        Epidemiological study of paratuberculosis in wild rabbits in Scotland.
        J. Clin. Microbiol. 1999; 37 (10325318): 1746-1751
        • Groenendaal H.
        • Nielen M.
        • Jalvingh A.W.
        • Horst S.H.
        • Galligan D.T.
        • Hesselink J.W.
        A simulation of Johne's disease control.
        Prev. Vet. Med. 2002; 54 (12114011): 225-245
        • Hacker U.
        • Hüttner K.
        • Konow M.
        (Investigation of serological prevalence and risk factors of paratuberculosis in dairy farms in the state of Mecklenburg-Westpommerania, Germany).
        Berl. Munch. Tierarztl. 2004; 117: 140-144
        • Harris J.E.
        • Lammerding A.M.
        Crohn's disease and Mycobacterium avium ssp. paratuberculosis: Current issues.
        J. Food Prot. 2001; 64 (11770646): 2103-2110
        • Hendrick S.H.
        • Kelton D.F.
        • Leslie K.E.
        • Lissemore K.D.
        • Archambault M.
        • Duffield T.F.
        Effect of paratuberculosis on culling, milk production, and milk quality in dairy herds.
        J. Am. Vet. Med. Assoc. 2005; 227 (16266021): 1302-1308
        • Hermon-Taylor J.
        Mycobacterium avium subspecies paratuberculosis is a cause of Crohn's disease.
        Gut. 2001; 49 (11709506): 755-757
        • IRS
        Yearly average currency exchange rates.
        • Johnson-Ifearulundu Y.
        • Kaneene J.B.
        • Lloyd J.W.
        Herd-level economic analysis of the impact of paratuberculosis on dairy herds.
        J. Am. Vet. Med. Assoc. 1999; 214 (10101415): 822-825
        • Jones C.M.
        • Heinrichs J.
        Growth Charts for Dairy Heifers: Comparing to a breed standard can indicate if heifer growth is progressing normally.
        https://extension.psu.edu/growth-charts-for-dairy-heifers
        Date: 2017
        Date accessed: January 17, 2020
        • Jubb T.F.
        • Galvin J.W.
        Effect of a test and control program for bovine Johne's disease in Victorian dairy herds 1992 - 2002.
        Aust. Vet. J. 2004; 82 (15149075): 228-232
        • Jubb T.F.
        • Sergeant E.S.G.
        • Callinan A.P.L.
        • Galvin J.W.
        Estimate of the sensitivity of an ELISA used to detect Johne's disease in Victorian dairy cattle herds.
        Aust. Vet. J. 2004; 82 (15478730): 569-573
        • Kennedy D.J.
        • Citer L.
        Paratuberculosis control measures in Australia.
        in: Behr M.A. Collins D.M. Paratuberculosis: Organism, Disease, Control. CABI, Wallingford, UK2010: 330-343
        • Kudahl A.B.
        • Nielsen S.S.
        Effect of paratuberculosis on slaughter weight and slaughter value of dairy cows.
        J. Dairy Sci. 2009; 92 (19700693): 4340-4346
        • Lombard J.E.
        • Garry F.B.
        • McCluskey B.J.
        • Wagner B.A.
        Risk of removal and effects on milk production associated with paratuberculosis status in dairy cows.
        J. Am. Vet. Med. Assoc. 2005; 227 (16379637): 1975-1981
        • Macrotrends
        S&P 500 Historical Annual Returns.
        • Manning E.J.
        • Collins M.T.
        Mycobacterium avium ssp. paratuberculosis: Pathogen, pathogenesis and diagnosis.
        Rev. Sci. Tech. 2001; 20 (11288509): 133-150
        • Marquetoux N.
        • Mitchell R.
        • Ridler A.
        • Heuer C.
        • Wilson P.
        A synthesis of the patho-physiology of Mycobacterium avium subspecies paratuberculosis infection in sheep to inform mathematical modelling of ovine paratuberculosis.
        Vet. Res. (Faisalabad). 2018; 49 (29514687): 27
        • McAloon C.G.
        • Whyte P.
        • More S.J.
        • Green M.J.
        • O'Grady L.
        • Garcia A.
        • Doherty M.L.
        The effect of paratuberculosis on milk yield—A systematic review and meta-analysis.
        J. Dairy Sci. 2016; 99 (26686704): 1449-1460
        • McClure H.M.
        • Chiodini R.J.
        • Anderson D.C.
        • Swenson R.B.
        • Thayer W.R.
        • Coutu J.A.
        Mycobacterium paratuberculosis Infection in a colony of stumptail macaques (Macaca arctoides).
        J. Infect. Dis. 1987; 155 (3559275): 1011-1019
        • McGregor H.
        • Dhand N.K.
        • Dhungyel O.P.
        • Whittington R.J.
        Transmission of Mycobacterium avium ssp. paratuberculosis: Dose–response and age-based susceptibility in a sheep model.
        Prev. Vet. Med. 2012; 107 (22762894): 76-84
        • Mitchell R.M.
        • Medley G.F.
        • Collins M.T.
        • Schukken Y.H.
        A meta-analysis of the effect of dose and age at exposure on shedding of Mycobacterium avium subspecies paratuberculosis (MAP) in experimentally infected calves and cows.
        Epidemiol. Infect. 2012; 140 (21524342): 231-246
        • Mortier R.A.R.
        • Barkema H.W.
        • Bystrom J.M.
        • Illanes O.
        • Orsel K.
        • Wolf R.
        • Atkins G.
        • De Buck J.
        Evaluation of age-dependent susceptibility in calves infected with two doses of Mycobacterium avium subspecies paratuberculosis using pathology and tissue culture.
        Vet. Res. (Faisalabad). 2013; 44 (24099491): 94
        • Mortier R.A.R.
        • Barkema H.W.
        • Orsel K.
        • Wolf R.
        • De Buck J.
        Shedding patterns of dairy calves experimentally infected with Mycobacterium avium subspecies paratuberculosis..
        Vet. Res. (Faisalabad). 2014; 45 (25224905): 71
        • Muskens J.
        • Barkema H.W.
        • Russchen E.
        • van Maanen K.
        • Schukken Y.H.
        • Bakker D.
        Prevalence and regional distribution of paratuberculosis in dairy herds in the Netherlands.
        Vet. Microbiol. 2000; 77 (11118710): 253-261
        • Naser S.A.
        • Ghobrial G.
        • Romero C.
        • Valentine J.F.
        Culture of Mycobacterium avium subspecies paratuberculosis from the blood of patients with Crohn's disease.
        Lancet. 2004; 364 (15380962): 1039-1044
        • OMAFRA
        Culling Decisions: Dairy Cows.
        • Ott S.L.
        • Wells S.J.
        • Wagner B.A.
        Herd-level economic losses associated with Johne's disease on US dairy operations.
        Prev. Vet. Med. 1999; 40 (10423773): 179-192
        • Pillars R.B.
        • Grooms D.L.
        • Wolf C.A.
        • Kaneene J.B.
        Economic evaluation of Johne's disease control programs implemented on six Michigan dairy farms.
        Prev. Vet. Med. 2009; 90 (19464741): 223-232
        • Pozzato N.
        • Capello K.
        • Comin A.
        • Toft N.
        • Nielsen S.S.
        • Vicenzoni G.
        • Arrigoni N.
        Prevalence of paratuberculosis infection in dairy cattle in Northern Italy.
        Prev. Vet. Med. 2011; 102 (21807432): 83-86
        • Raizman E.A.
        • Fetrow J.P.
        • Wells S.J.
        Loss of income from cows shedding Mycobacterium avium subspecies paratuberculosis prior to calving compared with cows not shedding the organism on two Minnesota dairy farms.
        J. Dairy Sci. 2009; 92 (19762809): 4929-4936
        • Rossi G.
        • Grohn Y.T.
        • Schukken Y.H.
        • Smith R.L.
        The effect of Mycobacterium avium ssp. paratuberculosis infection on clinical mastitis occurrence in dairy cows.
        J. Dairy Sci. 2017; 100 (28711261): 7446-7454
        • Shankar H.
        • Singh S.V.
        • Singh P.K.
        • Singh A.V.
        • Sohal J.S.
        • Greenstein R.J.
        Presence, characterization, and genotype profiles of Mycobacterium avium subspecies paratuberculosis from unpasteurized individual and pooled milk, commercial pasteurized milk, and milk products in India by culture, PCR, and PCR-REA methods.
        Int. J. Infect. Dis. 2010; 14 (19576834): e121-e126
        • Shephard R.W.
        • Williams S.H.
        • Beckett S.D.
        Farm economic impacts of bovine Johne's disease in endemically infected Australian dairy herds.
        Aust. Vet. J. 2016; 94 (27349883): 232-239
        • STATCAN
        Table 32–10–0077–01 Farm product prices, crops and livestock.
        • STATCAN
        Table 32–10–0136–01 Farm operating revenues and expenses, annual.
        https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3210013601
        Date: 2019
        Date accessed: September 16, 2020
        • STATCAN
        Table 36–10–0222–01 Gross domestic product, expenditure-based, provincial and territorial, annual (x 1,000,000).
        • STATCAN
        Table 32–10–0130–01 Number of cattle, by class and farm type (x 1,000).
        https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3210013001
        Date: 2020
        Date accessed: September 13, 2020
        • Suwandi A.
        • Bargen I.
        • Pils M.C.
        • Krey M.
        • Zur Lage S.
        • Singh A.K.
        • Basler T.
        • Falk C.S.
        • Seidler U.
        • Hornef M.W.
        • Goethe R.
        • Weiss S.
        CD4 T Cell dependent colitis exacerbation following re-exposure of Mycobacterium avium ssp. paratuberculosis..
        Front. Cell. Infect. Microbiol. 2017; 7 (28361039): 75
        • Tiwari A.
        Seroprevalence, production impacts, economics and risk factors of Mycobacterium avium subspecies paratuberculosis in Canadian dairy cattle.
        in: Health Management. Atlantic Veterinary College, University of Prince Edward Island, PE, Canada2005: 349
        • Tiwari A.
        • VanLeeuwen J.A.
        • Dohoo I.R.
        • Keefe G.P.
        • Weersink A.
        Estimate of the direct production losses in Canadian dairy herds with subclinical Mycobacterium avium subspecies paratuberculosis infection.
        Can. Vet. J. 2008; 49 (18624066): 569-576
        • Tiwari A.
        • VanLeeuwen J.A.
        • McKenna S.L.B.
        • Keefe G.P.
        • Barkema H.W.
        Johne's disease in Canada Part I: clinical symptoms, pathophysiology, diagnosis, and prevalence in dairy herds.
        Can. Vet. J. 2006; 47 (17017652): 874-882
        • Treasury Board of Canada
        Canadian Cost-Benefit Analysis Guide: Regulatory Proposals.
        • USDA
        National Dairy Comprehensive Report - Monthly - 02/29/2020.
        https://www.ams.usda.gov/mnreports/lsmdairycomp.pdf
        Date: 2020
        Date accessed: February 11, 2020
        • USDA ERS
        Dairy data – milk and cow production by State and region.
        • USDA NASS
        Farm Labor – 05/30/2019.
        • Van Biert P.
        Economics of Milk Production in Alberta, 2018: The Dairy Cost Study.
        • VanLeeuwen J.A.
        • Forsythe L.
        • Tiwari A.
        • Chartier R.
        Seroprevalence of antibodies against bovine leukemia virus, bovine viral diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in dairy cattle in Saskatchewan.
        Can. Vet. J. 2005; 46 (15759829): 56-58
        • VanLeeuwen J.A.
        • Keefe G.P.
        • Tremblay R.
        • Power C.
        • Wichtel J.J.
        Seroprevalence of infection with Mycobacterium avium subspecies paratuberculosis, bovine leukemia virus, and bovine viral diarrhea virus in Maritime Canada dairy cattle.
        Can. Vet. J. 2001; 42 (11265187): 193-198
        • VanLeeuwen J.A.
        • Tiwari A.
        • Plaizier J.C.
        • Whiting T.L.
        Seroprevalences of antibodies against bovine leukemia virus, bovine viral diarrhea virus, Mycobacterium avium subspecies paratuberculosis, and Neospora caninum in beef and dairy cattle in Manitoba.
        Can. Vet. J. 2006; 47 (16933557): 783-786
        • Waddell L.A.
        • Rajić A.
        • Stärk K.D.
        • McEwen E.S.
        The zoonotic potential of Mycobacterium avium ssp. paratuberculosis: a systematic review and meta-analyses of the evidence.
        Epidemiol. Infect. 2015; 143 (25989710): 3135-3157
        • Weber M.F.
        • Kogut J.
        • de Bree J.
        • van Schaik G.
        • Nielen M.
        Age at which dairy cattle become Mycobacterium avium ssp. paratuberculosis faecal culture positive.
        Prev. Vet. Med. 2010; 97 (20708807): 29-36
        • Wells S.J.
        • Wagner B.A.
        Herd-level risk factors for infection with Mycobacterium paratuberculosis in US dairies and association between familiarity of the herd manager with the disease or prior diagnosis of the disease in that herd and use of preventive measures.
        J. Am. Vet. Med. Assoc. 2000; 216 (10800519): 1450-1457
        • Whitlock R.H.
        • Wells S.J.
        • Sweeney R.W.
        • Van Tiem J.
        ELISA and fecal culture for paratuberculosis (Johne's disease): Sensitivity and specificity of each method.
        Vet. Microbiol. 2000; 77 (11118724): 387-398
        • Whittington R.J.
        • Marshall D.J.
        • Nicholls P.J.
        • Marsh I.B.
        • Reddacliff L.A.
        Survival and dormancy of Mycobacterium avium ssp. paratuberculosis in the environment.
        Appl. Environ. Microbiol. 2004; 70 (15128561): 2989-3004
        • Windsor P.A.
        • Whittington R.J.
        Evidence for age susceptibility of cattle to Johne's disease.
        Vet. J. 2010; 184 (19246220): 37-44
        • Wolf R.
        • Orsel K.
        • De Buck J.
        • Barkema H.W.
        Calves shedding Mycobacterium avium subspecies paratuberculosis are common on infected dairy farms.
        Vet. Res. (Faisalabad). 2015; 46 (26091858): 71
        • Woodbine K.A.
        • Schukken Y.H.
        • Green L.E.
        • Ramirez-Villaescusa A.
        • Mason S.
        • Moore S.J.
        • Bilbao C.
        • Swann N.
        • Medley G.F.
        Seroprevalence and epidemiological characteristics of Mycobacterium avium ssp. paratuberculosis on 114 cattle farms in south west England.
        Prev. Vet. Med. 2009; 89 (19327856): 102-109
        • World Bank
        GDP per capita (current US$).
        https://data.worldbank.org/indicator/NY.GDP.PCAP.CD
        Date: 2020
        Date accessed: January 4, 2020
        • Zwick L.S.
        • Walsh T.F.
        • Barbiers R.
        • Collins M.T.
        • Kinsel M.J.
        • Murnane R.D.
        Paratuberculosis in a Mandrill (Papio sphinx).
        J. Vet. Diagn. Invest. 2002; 14 (12152813): 326-328