Abstract

Equine piroplasmosis, caused by Theileria equi and Babesia caballi, have been identified as presumptive causes of mortality in newly reintroduced Przewalski’s horses (Equus ferus przewalskii) in the Dzungarian Gobi (Mongolia). To quantify the risk of infection for these horses, the parasites’ epidemiology was investigated in the local domestic horse population. The prevalence of the two pathogens was determined in 510 domestic horses in the reintroduction area by indirect immunofluorescence antibody test (IFAT) and multiplex polymerase chain reaction (PCR). The data was analyzed by fitting four nonlinear mathematic transmission models to the field data with the maximum likelihood method to determine the transmission parameters and selecting the best fitting model by conducting a Monte Carlo simulation study. Each of the models used different assumptions:

1.  Passive, maternal immunity or acquired immunity to the protozoa did not influence transmission dynamics.

2.  Maternal immunity had a different effect on transmission than acquired immunity, which was again different to transmission in naïve animals.

3.  Maternal and acquired immunity had the same effect, which was different in naïve animals.

4.  Maternally protected animals were infected equally to naïve animals, but acquired immunity affected transmission differently.

For the data for T. equi, model one fit best, which indicated that antibodies had no significant influence on the infection dynamics of T. equi. The maximum likelihood estimates of the parameters indicated that infection was expected 2¼ years after first exposure and then persisted lifelong. For B. caballi, model two performed best, and the infection rates suggested that foals from immune dams were expected to be infected at 4 months of age, naïve animals were expected to be infected 12 years after exposure, whereas immune animals were unlikely to be reinfected. The parasites were eliminated after approximately 2 years. Because geographic causes could be excluded, the association of maternal and filial infections suggested that the risk for infection had a strong behavioral component. Thus, under certain circumstances, Przewalski’s horses are expected to become infected within 4 months after arrival, whereas others will not acquire the infection for as long as 12 years. Because the outbreak of clinical babesiosis can be related to the general immune status of a horse, which is suppressed by stress, a preventive treatment with imidocarb-dipropionate (1.5–3 mg/kg IM delivered by remote injection; Carbesia^®, Pitman-Moore, Meaux, France) is being administered upon their arrival at the reintroduction site. Assuming similar transmission dynamics in Przewalski’s horses as in domestic horses, this will protect the animals with high-risk behavior during the stressful reintroduction period for approximately 3 weeks. However, when the drug level wears off, the animals with low-risk behavior will not have been infected with B. caballi, and only a few animals will have been infected with T. equi. Thus, further infections and clinical cases of piroplasmosis are to be expected after the animals have been released into the field.