Toxoplasmosis was not an important contributor to foetal and lamb mortality between pregnancy scanning and marking for the primiparous ewe flocks in this study. Whilst T. gondii seropositivity was detected by indirect ELISA on 57% of farms suggesting widespread exposure at farm level, seroconversion after mating was evident for only 1% of primiparous ewes that were confirmed to be pregnant and subsequently failed to raise a lamb. Low incidence of T. gondii seroconversion in primiparous ewes was consistent with the absence of detection of T. gondii using qPCR on aborted and stillborn lambs from a subset of farms. These findings are in accord with a recent review of submissions to Australian veterinary laboratories that reported T. gondii was implicated in 5% of sheep abortion investigations, and suggests that toxoplasmosis is a sporadic cause of abortion in Australian sheep [13]. In contrast, T. gondii is considered a significant economic and production-limiting pathogen in New Zealand and the United Kingdom where vaccination is utilised to reduce associated losses [10, 25]. Our findings indicate that routine vaccination for toxoplasmosis is unlikely to be economically justified for many Australian sheep producers unless there is evidence indicting high risk of exposure to T. gondii and toxoplasmosis contributing to reproductive disease for sheep in the specific region.
This study used serial serology to assess the timing of seroconversion relative to the timing of abortion or lamb death for primiparous ewes that were seropositive at lamb marking. Reproductive disease associated with toxoplasmosis is most likely to occur when primary infection of a naïve ewe occurs during pregnancy, with the outcome of infection being largely dependent on the timing of exposure [7, 26]. Detection of IgG antibodies is reported to occur from 2-3 weeks post-infection [27], and abortions may occur within 1-2 weeks of infection [26, 28–30], or approximately 3-8 weeks post-infection, depending on stage of gestation when infection occurred [26, 29, 31, 32]. Toxoplasma gondii IgG antibodies are detectable by P30 ELISA between 3 and 10 weeks after infection [33] and persist for several years [34]. So, it is unlikely that antibodies would have failed to rise to detectable levels or waned to below detectable levels by lamb marking if ewes had become infected with T. gondii during pregnancy and foetal or lamb mortality was related to toxoplasmosis. Based on these assumptions, toxoplasmosis was a plausible aetiology for abortions in only two ewes (12114 and 13091; Table 2). The relationship for timing of seroconversion relative to abortion or lamb death was less clear for other ewes (Table 2), consistent with the notion that a spectrum of outcomes may occur for pregnant ewes that become infected with T. gondii (and particularly during late gestation).
An unclear association between timing of T. gondii seroconversion and abortion in some ewes was also consistent with observation that other factors were likely to be contributing to abortion or perinatal lamb deaths in these flocks. Abortions, stillbirths and polyarthritis associated with Chlamydia pecorum were detected in primiparous ewe flocks from Western Australia, and non-infectious causes of death (including dystocia and starvation-mismothering) were important contributors to cause of lamb death identified at necropsy [3, 35]. There was no evidence neosporosis was contributing to foetal or lamb mortality in these flocks [24].
The selection of primiparous ewe for serological surveillance was biased towards ewes that aborted or failed to rear lambs. Therefore, it was not possible to determine the risk for foetal and lamb mortality associated with seroconversion during gestation. Notwithstanding this, even if seroconversion during pregnancy was associated with a higher relative risk of failing to rear a lamb, the low incidence of seroconversions during pregnancy in primiparous ewes that failed to raise lambs suggests that toxoplasmosis was unlikely to be an important contributor to reproductive losses between scanning and marking on these farms.
Attempting to relate maternal antibody titre to the occurrence of abortion or lamb mortality can be unreliable [32]. Seropositivity may be detected in ewes who successfully rear lambs either due to the enduring nature of IgG antibodies from infections prior to gestation or the spectrum of outcomes that may occur when naïve ewes become infected during gestation. While extensive laboratory testing was not performed across all sites, the low incidence of T. gondii seroconversion for primiparous ewes that aborted or failed to raise lambs was supported by the absence of detection of T. gondii via molecular methods on a subset of farms in Western Australia [3]. Association between T. gondii seroconversion in ewes and foetal or lamb mortality in this study could be confirmed using foetal serology [32], molecular diagnostics [36], and histopathology [37] for tissues from a larger subset of lambs that were aborted or died during perinatal period. Additionally, quantitative serology to evaluate changes in titre for ewes with suspected toxoplasmosis would have allowed a more accurate interpretation of serial serological results. Nevertheless, even if foetal and lamb mortality for all ewes that were seropositive at lamb marking was due to toxoplasmosis, this still represents a very small contributor to overall foetal and lamb mortality in these flocks.
This study reported seroprevalence for ewes that were determined to be pregnant by transabdominal ultrasound and subsequently aborted or failed to raise lambs. However, toxoplasmosis can also have impacts on early pregnancy before pregnancy scanning, including embryonic death, resorption and early foetal mortality [38]. This study did not determine association between T. gondii seroconversion and pregnancy status at Scan 1. It is possible that infections occurring early in pregnancy could result in primiparous ewes being not pregnant at scanning. Low seroprevalence reported in this study for primiparous ewes that were pregnant at scan 1 suggests that toxoplasmosis was unlikely to be an important contributor to early reproductive losses on these farms. However, further investigation is required to determine if infection early in pregnancy is an important contributor to early pregnancy loss evident as ewes that are not pregnant at scanning.
Toxoplasma gondii seropositivity was detected for 57% of farms in this study, providing further evidence for the parasite’s endemicity and suggesting that exposure to T. gondii occurs on many Australian sheep farms. This was consistent with other Australian studies reporting farm-level seroprevalence ranging 41 – 97% [15, 16, 18, 20]. Seroprevalence in mature ewes (8.1%) was within the 7% – 57% range reported for other Australian studies in the last 15 years [17–21], and similar to national seroprevalence 11.5% (46/401) reported for mutton (mature sheep) in a recent Australian abattoir survey [21].
Individual-animal T. gondii seroprevalence for primiparous ewes in our study (1.1%) was lower than two previous studies reporting seroprevalence of 15% and 17% for Australian slaughter-age lambs in abattoir surveys [14, 17]. The most recent abattoir survey reporting seroprevalence in lambs used a commercial modified latex agglutination test (Eiken Indirect Latex Agglutination Test) [17]. Our study used a commercial indirect ELISA that has good sensitivity and specificity relative to latex agglutination test for sheep sera reported by the manufacturer [39], suggesting that differences in testing methodology are unlikely to explain the difference in seroprevalence between the studies. It is therefore more likely that the low seroprevalence in primiparous ewes reflects the sporadic nature of T. gondii infection in Australian sheep rather than differences in serological methods.
Flock and animal seroprevalence in mature ewes indicated variable exposure to T. gondii on Australian sheep farms. Most seropositive mature ewes were concentrated on five farms where within-farm seroprevalence ranged 25-50%. Variable seroprevalence in Australian sheep contrasts to the more consistent exposure reported for sheep farms in New Zealand, United States and United Kingdom [11, 12, 40], and suggests sporadic point source exposure to oocysts, likely via contaminated drinking water or feed source, are the major source of T. gondii exposure on Australian sheep farms. The higher (50%) seroprevalence observed for ewes on a farm on Kangaroo Island (Additional file 2) was consistent with previous studies reporting high T. gondii seroprevalence on Kangaroo Island [19, 20]. Interestingly, there was no evidence of seropositivity in the primiparous ewes from the farm on Kangaroo Island, consistent with sporadic exposure to T. gondii. The risk of T. gondii exposure is associated with a range of factors including abundance of cats, access to surface water and rainfall [41, 42]. The number of farms included in this study was too small to compare seroprevalence across regions, and further work is warranted to determine regional differences in T. gondii exposure and incidence of toxoplasmosis. Improved understanding of regional variation and risk factors for T. gondii exposure on Australian sheep farms would inform cost-benefit analyses for interventions to reduce the risk of toxoplasmosis.
Sampling primiparous ewes was biased towards ewes that were determined to be pregnant and subsequently failed to rear a lamb. If toxoplasmosis was an important contributor to foetal and lamb mortality on these farms, then this sampling bias could result in overestimation of seroprevalence relative to the general population. However, there was no difference observed in flock-level seroprevalence for primiparous ewes compared to multiparous ewes.
The commercial ELISA used in our study has been utilised for other seroprevalence surveys in sheep [18, 43–45], but there is a lack of published data validating the test against other ‘gold standard’ tests such as microagglutination tests or PCR. A commercial modified agglutination test for T. gondii was validated for Australian sheep [21], but this test is no longer available. The indirect ELISA used in this study was found to have good agreement with modified agglutination test on a subset of sheep sera collected from another study (Additional file 3). Further validation of commercial indirect ELISA for natural T. gondii infections in Australian sheep will inform improved estimation of true prevalence for field studies.
Whilst the observations from this study do not support the need for widespread routine vaccination to reduce foetal and lamb mortality for primiparous ewes in Australia, T. gondii is endemic on Australian farms and associated with sporadic reproductive losses on some farms [13]. The low rates of seropositivity in primiparous and multiparous ewes suggests a lack of protective immunity in a large proportion of ewes [10]. These ewes remain susceptible to reproductive disease if a toxoplasmosis outbreak was to occur during gestation, most likely through point source exposure to oocysts via contamination of drinking water or feed source. Control of feral and domestic cat populations on sheep farms and measures to prevent contamination of feed and water sources from cat faeces can reduce the risk of toxoplasmosis outbreaks in susceptible sheep. Vaccination could be warranted in some regions where high farm and individual animal seroprevalence is identified [20], and high incidence of reproductive losses due to T. gondii are confirmed using foetal and lamb necropsy and laboratory investigation. The interpretation of T. gondii serology for the purpose of diagnosing toxoplasmosis is challenging, particularly for field investigations of reproductive loss in extensive sheep production systems where abortions or unusually high incidence of perinatal mortality are challenging to detect at the time losses are occurring. Diagnosis of toxoplasmosis should be supported with detection of T. gondii in tissues where possible, as well as the exclusion of other endemic pathogens.