Spatiotemporal cluster and incidence analysis of cattle mortality caused by bovine babesiosis in Styria, Austria, between 1998 and 2016

Reported fatal cases of bovine babesiosis (syn.: piroplasmosis, red water fever) in cattle were analyzed to identify spatial and temporal clusters of their incidence in the Austrian province of Styria. Data were collected within a governmental babesiosis compensation program. Diagnosis was performed using a standardized necropsy protocol. Between 1998 and 2016, a total of 1257 cases of fatal babesiosis were registered and compensated. Within the study interval, annual numbers of fatal babesiosis differed significantly among municipalities. Spatiotemporal analysis covering the entire study period revealed one high-risk cluster in the western and central northern region of Styria and a low-risk cluster in the southeastern part of Styria. Annual temporal analysis demonstrated that cases accumulated in June. Annual spatial analysis revealed consistently that cases mainly occurred in the western and central northern regions, whereas they occurred rarely in the southeastern regions. These results should increase awareness and facilitate protective actions against ticks during certain time periods and geographic areas.

In Europe, cattle are mainly infected with Babesia divergens (M'Fadyean and Stockman 1911), transmitted by ixodid ticks (Edelhofer et al. 2004;Krampitz et al. 1986). Apparently, only one clinical case of Babesia bovis (Babes 1888) in cattle has been reported in Austria, suggesting it is of only minor clinical importance in this geographic region (Edelhofer et al. 2004). In contrast, Babesia divergens not only affects cattle but can also affect immunosuppressed, especially splenectomized, humans (Schuster 2002;Zintl et al. 2003).
Subclinical, acute, and chronic courses of bovine babesiosis have been described (Gray and Murphy 1985). After an incubation period of a few days after the tick bite, acute cases are characterized by high fever (up to 42°C), red urine, and, in longer-surviving animals, ischemic changes in the skeletal and heart muscle (Radostits et al. 2000). If not treated during the acute phase, babesiosis is often lethal. In the chronic state, anorexia, jaundice, hepatomegaly, and splenomegaly are common, whereas retinal detachment has been described (Taylor and Andrews 1992;Naucke 2008). Differential diagnoses include anaplasmosis, eperythrozoonosis, leptospirosis, postparturient and bacillary hemoglobinuria (Williams and Andrews 1992;Taylor and Andrews 1992), and intoxication with bracken fern, lead, or copper.
To avoid infection, various preventive measures can be applied against vector and pathogen. For example, vaccines for babesiosis (Holzheu et al. 2016) and topical treatments containing deltamethrin are used to prevent tick bites. Targeted vaccination and antiparasitic treatment may prevent cattle from being infected. However, there is a lack of information regarding high-or low-risk areas in the province of Styria, and no specific high-risk period has been defined. Thus, the objective has been to investigate and identify spatial and temporal clusters of the incidence of bovine babesiosis.

Data collection
Data were collected in the province of Styria, Austria, where fatal babesiosis is reported to the local official veterinarian to be considered for compensation. For all reported cases, the local official veterinarian requests a necropsy (standardized protocol) done in the rendering plant by trained official veterinarians. To qualify as an official case that will be funded, the three following pathological findings must present (1) red urine in the urinary bladder, (2) raspberry-red spleen, and (3) anemia. The present study contains all fatal cases of babesiosis in Styria that have met these three criteria and occurred between 1 January 1998 and 31 December 2016. Location (municipality), age, gender, and date of reporting to the official veterinarian were recorded.

Descriptive analysis
Data analysis was conducted in Microsoft Excel 2016® (Microsoft Corp., Redmond, Washington, USA) and in R® (R Core Team 2017). Individual cases were stratified by location, age, gender, and date of necropsy. The number of cases was aggregated by year and normalized based on the number of cattle in a municipality (population data obtained from the Austrian cattle database). The mean incidence was calculated, and a heat map that included all reported fatal cases was prepared to indicate when most piroplasmosis cases had been reported.

Statistical analysis
Identification of spatial and temporal clusters was conducted using a discrete Poisson model in SaTScan (Kulldorf 1997). This analysis provides the annual number of fatal cases in relation to the number of animals at risk within and outside the cluster, the ratio of observed cases versus expected cases, and the relative risk for cattle dying of babesiosis within the cluster versus the risk for cattle dying of babesiosis outside the cluster.
Cases were aggregated at a municipality level and a likelihood ratio test and Monte Carlo simulation done to estimate significance. The time interval was set to 1 day, and the maximum temporal and spatial window size was set to 50% of the population. Clusters were considered significant if the p value was < 0.05. A relative risk > 1 defined a high-risk cluster. A relative risk < 1 defined a low-risk cluster. A first spatiotemporal analysis was carried out covering the entire study period and region. Additionally, an analysis of each, the spatial and temporal clustering, was performed for every year to identify changes in the annual clustering. A relative risk was reported for cases that occurred within vs. those that occurred outside the identified cluster. If all cases observed in a year were included in a single cluster, no relative risk could be reported.  Table 1).

Results
Cases mainly occurred during the summer months, with significant temporal clusters between May and October  2001, 2005, and 2009, no relative risk was reported, because all cases occurred within the cluster. All clusters were highly significant (p value < 0.001). The annual number of fatal babesiosis cases varied among municipalities and years. Spatiotemporal analysis covering the entire study period revealed one high-risk cluster in the western and central northern part of Styria. In addition, a low-risk cluster was detected in the southeastern part of Styria (Fig. 2). Annually, the spatial distribution of high-and lowrisk clusters changed throughout the study period: high-risk clusters were always observed in the western and the central northern regions (districts Murau, Murtal, Leoben, Bruck-Mürzzuschlag, and Voitsberg and parts of Graz-Umgebung), whereas low-risk clusters were observed in the southeastern region in 50% to 90% of the years studied (districts Leibnitz, Graz, Südoststeiermark, and Hartberg-Fürstenfeld and parts of Weiz). No clustering was observed in more than 50% of the years of the study period in the northern region of Styria (district: Liezen). Over the entire study period, the highest mean incidence (fatal cases per 100,000 cattle) was observed in some municipalities in the central and western region of Styria, mainly in the districts Murau, Bruck-Mürzzuschlag, and Graz-Umgebung (Fig. 3).

Discussion
There are no recent reports regarding incidence of diseases caused by hematotropic parasites such as Babesia spp. in Austria. Furthermore, there are limited data on the density of Ixodes ricinus, the vector of Babesia spp. Based on recent studies in Germany, forest or forest-like habitats offer better survival conditions for ticks than meadows or clear areas, where lower population densities are observed (Böhnke et al. 2015). In addition, there were also lower densities reported in higher altitudes with coniferous woods or an alpine climate. Although most parts of Styria allow ticks to thrive, this study clearly identified high-risk clusters for babesiosis mortality in the central parts and low-risk clusters in the southeastern parts of Styria. Locations of spatial and temporal clusters may be linked to topology, temperature, and differences in pasture management. In northern and central Styria, young stock and calf herds are frequently kept on alpine pastures throughout the summer months when most babesiosis cases occur. Therefore, increased incidence of babesiosis in central Styria might also be due to an increased exposure on alpine pastures.
I. ricinus, the most common tick present on cattle and a known vector for B. divergens, is a key factor in transmission of Babesia spp. (Holzheu et al. 2016). Minimum temperature for I. ricinus to develop is 8.4°C, and the mean duration of larval development is 51 days at 20.0°C (Dautel 2010). I. ricinus needs habitats with a relative humidity of over 80% for a prolonged time (Kahl 1982). As ticks reproduce faster with increasing temperature (Dautel 2010), risk of babesiosis infection is highest during summer months, which explains disease clusters between May and October. While the presented heatmap (Fig. 1) includes all babesiosis cases, cluster dates in Table 1 mark a period of aggregated cases (elevated risk). In years with a short temporal cluster duration, the estimated number of annual cases per 10,000 within a cluster was high, while the relative risk stayed within a normal range (e.g., 1999 with 116.4 annual fatal cases within the cluster and a relative risk of 13.7, see Tab. 1). This suggests that a high estimated number of annual cases within a cluster were rather driven by a short cluster duration than by a high relative risk within the cluster.
The number of fatal babesiosis cases varied between years, which may be linked to weather conditions that favor tick development in certain years. However, an investigation of a The study was based on reported cases of mortality caused by Babesia spp. Consequently, it excludes cases not reported by farmers, which could have introduced a bias into the analysis. However, reporting of suspect cases is free of charge, and all farmers are obliged to financially contribute to the Styrian animal disease fund which motivates them to seek financial reimbursement in case of animal losses. Therefore, authors considered the impact of underreporting as low.
Vaccinated cattle are unlikely to develop clinical signs of babesiosis (Holzheu et al. 2016). As vaccinated cattle are reported to the government of Styria, their numbers are well documented. By far the highest vaccination rate was reported in 2006 with 2173 cattle, which amounts to 0.7% of the Styrian cattle population (Wagner et al. 2006). Thus, vaccination coverage was very low and likely had a negligible impact on the results of the study.
The present study shows that the risk of cattle dying of babesiosis as evaluated based on reported fatal cases was non-randomly distributed across the province of Styria. We conclude that cattle farmers in high-risk areas should be motivated to implement preventive control measures for their herds. To ensure development of protective antibodies before pasture season, the best time to vaccinate cattle is between March and April. According to Holzheu et al. (2016), a recombinant Babesia vaccine used in 55 animals resulted in a significant increase in the development of specific antibodies compared to a control group of 57 animals. No signs of general or local reaction after vaccination could be noticed, approving vaccine safety (Holzheu et al. 2016). Another study by Edelhofer et al. (1998) successfully used an inactivated vaccine against babesiosis without causing severe clinical symptoms in splenectomized calves and afforded resistance to disease after challenge with B. divergens. In Austria, only veterinarians are allowed to administer this vaccination, which might be seen as a drawback for farmers. However, farmers could apply spot-on or pour-on suspensions of Deltamethrin to cattle before grazing. The disadvantage of this strategy is that animals have to be treated twice or three times per year to reduce the number of tick bites throughout the whole pasture season. Nevertheless, cattle should frequently be checked for signs of babesiosis in early summer when most cases occur, as that would enable treatment in early stages, with high chances of success.

Conclusion
The present study showed that babesiosis is not evenly distributed in the studied area, which may be due to dissimilarities of farm management strategies and geographic and climate conditions. Areas where cattle are mainly kept on alpine pastures are at higher risk, especially in summer.
Funding information Open access funding provided by University of Veterinary Medicine Vienna. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflicts of interest.
Ethical approval KSS, JAH, and RW collaboratively designed the study. KSS performed the statistical analyses and prepared the manuscript for submission. JAH performed data extraction from various databases. IK prepared all figures and reviewed statistical methods. PW and AD gave further input and JK was involved with manuscript preparation. All authors approved the final version of the manuscript for publication.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.