Gross margin losses due to Salmonella Dublin infection in Danish dairy cattle herds estimated by simulation modelling.

Salmonella Dublin affects production and animal health in cattle herds. The objective of this study was to quantify the gross margin (GM) losses following introduction and spread of S. Dublin within dairy herds. The GM losses were estimated using an age-structured stochastic, mechanistic and dynamic simulation model. The model incorporated six age groups (neonatal, pre-weaned calves, weaned calves, growing heifers, breeding heifers and cows) and five infection stages (susceptible, acutely infected, carrier, super shedder and resistant). The effects of introducing one S. Dublin infectious heifer were estimated through 1000 simulation iterations for 12 scenarios. These 12 scenarios were combinations of three herd sizes (85, 200 and 400 cows) and four management levels (very good, good, poor and very poor). Input parameters for effects of S. Dublin on production and animal health were based on literature and calibrations to mimic real life observations. Mean annual GMs per cow stall were compared between herds experiencing within-herd spread of S. Dublin and non-infected reference herds over a 10-year period. The estimated GM losses were largest in the first year after infection, and increased with poorer management and herd size, e.g. average annual GM losses were estimated to 49 euros per stall for the first year after infection, and to 8 euros per stall annually averaged over the 10 years after herd infection for a 200 cow stall herd with very good management. In contrast, a 200 cow stall herd with very poor management lost on average 326 euros per stall during the first year, and 188 euros per stall annually averaged over the 10-year period following introduction of infection. The GM losses arose from both direct losses such as reduced milk yield, dead animals, treatment costs and abortions as well as indirect losses such as reduced income from sold heifers and calves, and lower milk yield of replacement animals. Through sensitivity analyses it was found that the assumptions about milk yield losses for cows in the resistant or carrier stages had the greatest influence on the estimated GM losses. This was more influential in the poorer management scenarios due to increased number of infected cows. The results can be used to inform dairy farmers of the benefits of preventing introduction and controlling spread of S. Dublin. Furthermore, they can be used in cost-benefit analyses of control actions for S. Dublin both at herd and sector level.

Strategies for time of culling in control of paratuberculosis in dairy herds.

Effect of time for culling cows infected with Mycobacterium avium ssp. paratuberculosis on prevalence and profitability was identified through simulations. Seven test-and-cull strategies with different culling criteria and no attempts to close infection routes were compared with strategies with (1) no control and (2) closure of infection routes and no culling. The effects on true prevalence and gross margin were evaluated in a herd with typical reproduction management (heat detection rate of 38%). This was repeated in a herd with poor reproduction management (heat detection rate of 28%), because poor reproduction leads to lack of replacement animals, which was hypothesized to affect the economic effects of culling. Effects of varying prices of milk, replacement heifers, and hourly wages were also evaluated. The simulated results predicted that immediate culling after the first positive antibody ELISA test would be the most effective culling strategy to reduce prevalence. However, closing transmission routes was even more effective in reducing the prevalence. In the first 3 to 6 yr, all test-and-cull strategies reduced gross margin by US$5 to 55/stall per year. These losses were fully compensated by increased gross margin in yr 6 to 19. In the short run (7 yr with typical reproduction and 10 yr with poor reproduction), it was most profitable to cull test-positive cows when their milk yield decreased below 85% of that expected according to their parity and lactation stage, especially in herds with poor reproduction management. However, this strategy only stabilized the prevalence and did not reduce it. In the long term (>7 yr from implementation of a strategy), it was most profitable to cull cows immediately or as soon as possible after testing positive the first time. Varying milk prices did not affect the ranking between the different culling strategies. Increased market price (20%) of replacement heifers made all culling strategies less profitable and made culling based on a milk yield criterion the most profitable culling strategy for a longer period (11 to 13 yr). A 20% reduction in heifer price made immediate culling after a positive test the most profitable strategy overall in herds with typical reproduction, and after 9 yr in herd with poor reproduction. To conclude, the ideal culling strategy depends on the aim of intervention, the time horizon, and the reproductive capabilities combined with prices of replacement animals.

Economy, efficacy, and feasibility of a risk-based control program against paratuberculosis.

Long-term effects of paratuberculosis on within-herd prevalence and on-farm economy of implementing risk-based control strategies were compared with alternative strategies by using a herd-simulation model. Closing transmission routes is essential for effective control of paratuberculosis. However, many farmers lack the resources to carry out these procedures for all cows in the herd. When using risk-based control strategies 1) all cows are tested quarterly with a milk ELISA, 2) specific cows with a high risk of being infectious are identified, and 3) the farmer can focus only on these infectious animals to close infection routes. In this way the workload can be reduced, making these control strategies more feasible. This study evaluates potential long-term effects of the risk-based approach compared with non-risk-based strategies by simulations conducted with the herd-simulation model PTB-Simherd. Seven control strategies were simulated in herds with initial true herd prevalences of 5, 25, and 50%, respectively. The results predicted the risk-based control strategies to be very efficient and comparable to the best whole-herd strategies in reducing the within-herd prevalence of paratuberculosis with considerably less labor. If infection routes are closed efficiently, prevalence can be reduced to 10% of initial prevalence within 5 to 7 yr. Test-and-cull strategies without closing infection routes were found, by simulation, to be ineffective in reducing prevalence and were not cost-effective methods. The profitability of the various control strategies depends on hourly wages and time spent per cow/calving. Furthermore, simulations show that immediate culling of highly infectious cows is only necessary and cost-effective if infection routes from these cows are not efficiently closed. The risk-based control strategies are recommended in the Danish voluntary control program "Operation Paratuberculosis," which was initiated in February 2006 and now includes 1,220 dairy farmers in Denmark.