Zoonotic Pathogen Seroprevalence in Cattle in a Wildlife–Livestock Interface, Kenya

Brucellosis and leptospirosis are neglected bacterial zoonotic diseases of veterinary and public health importance worldwide (Seleem et al. 2010; de Vries et al. 2014). In livestock-dependent households, these diseases cause direct economic losses due to the reduction in animal’s milk yields, abortion and infertility (Adler and de la Peña Moctezuma 2010; Franc et al. 2018), significantly affecting the well-being of communities whose livelihood depends on livestock. Bovine brucellosis is caused by facultative intracellular gram-negative coccobacilli of the genus Brucella (Seleem et al. 2010). Whereas Brucella abortus is the main causative agent of bovine brucellosis, Brucella melitensis, the species that primarily affects sheep and goats, can occasionally infect cattle (Seleem et al. 2010). Bovine leptospirosis is caused by pathogenic spirochetes of the genus Leptospira (de Vries et al. 2014).

Knowledge on the epidemiology of these pathogens is limited in livestock, wildlife and human populations in the Maasai Mara ecosystem (in Kenya) and indeed in many resource-poor areas due to lack of prioritization, poor surveillance systems and diagnostic capacities (Allan et al. 2015; Ducrotoy et al. 2017). The Maasai Mara ecosystem has a rich biodiversity of wildlife and a thriving tourism industry that provides additional livelihoods to the local people (Bedelian and Ogutu 2017). In recent years, the area has undergone major land use changes due to increased human populations, infrastructure development (e.g., roads and fencing) and land privatization (Ogutu et al. 2009; Løvschal et al. 2019). An example of these changes is the establishment of wildlife conservancies in areas adjacent to Mara reserve and increased mixed farming (livestock production and crop cultivation) in areas further away from the reserve (Nthiwa et al. 2019). Whereas the establishment of wildlife conservancies provides a sustainable way of integrating wildlife conservation alongside livestock production (Løvschal et al. 2019), it also intensifies livestock–wildlife interactions which may increase infectious disease transmission (Nthiwa et al. 2019).

This study investigated how different land use types affect disease exposure among cattle herds raised in the Mara ecosystem, using Leptospira spp. and Brucella spp. as case study pathogens. Specifically, we determined the seroprevalence of these pathogens in cattle across three zones with varying levels of wildlife–livestock interactions and identified risk factors associated with exposure. This study provides information on the current epidemiological situation of these pathogens in the area. It will also provide additional data to inform discussions on the linkages between host diversity and infectious disease risk.

Blood samples were obtained from 1170 cattle (21.4% and 78.6% males and females, respectively) belonging to 390 herds. The median cattle herd size was 50 (range 4–570).

The overall apparent animal-level seroprevalences of Brucella spp. and Leptospira spp. were 36.9% (95% CI 34.1–39.8) and 23.5% (95% CI 21.1–26.0), respectively. Animal-level seroprevalence of both diseases differed between zones; Brucella spp. seroprevalence was higher in zone 1 (high interface area) than in zones 2 and 3 (χ² = 25.1, df = 2, P < 0.001) (Table 1). Zones 1 and 2 had significantly higher Leptospira spp. seroprevalence than zone 3 (χ² = 7.0, df = 2, P = 0.029) (Table 1). Overall, the level of Brucella spp. and Leptospira spp. co-exposure in animals was estimated at 8.8% (95% CI 7.3–10.4) and differed significantly by sex (χ² = 9.9, df = 1, P = 0.001) with females having higher levels of co-exposure (10.2%; 95% CI 8.4–12.1) than males (3.6%; 95% CI 1.6–5.6). There were no differences in the levels of co-exposure between zones (P > 0.05) (Table 1).

At herd level, 68.7% (95% CI 66.1–71.5) of the herds had at least one seropositive animal for Brucella spp. and 52.7% (95% 49.7–55.6) had at least one animal positive for Leptospira spp. (Table 1). The herd-level seroprevalence of both diseases varied significantly by zones (P < 0.001), following a similar pattern as that of animal-level seroprevalence mentioned above. Herd-level seroprevalence of brucellosis was higher in zone 1 than other zones, while for leptospirosis, zones 1 and 2 had a significantly higher seroprevalence than zone 3. The spatial distributions of Brucella spp. and Leptospira spp. seropositive herds are presented in Fig. 2.

To our knowledge, this is the first survey to determine the association between Brucella spp. and Leptospira spp. seroprevalences with land use patterns in the Mara ecosystem, Kenya. We found both Brucella spp. and Leptospira spp. to be prevalent in the area as evidenced by the high levels of exposure at animal and herd levels. The animal-level seroprevalence of Brucella spp. in our study was higher compared to 12.44% (n = 225) previously reported in the area (Enström et al. 2017), but was aligned with the findings of Nina et al. (2017) in Uganda (44%) and Madut et al. (2018) in Sudan (31%). Leptospira spp. animal-level seroprevalence was also consistent with earlier reports in resource-poor areas, for instance, 25–34% in Kenya (de Vries et al. 2014) and 22.2% in Lao People’s Democratic Republic (Olmo et al. 2018). The high seroprevalences of these pathogens are worrying as both diseases are well known threats to animal productivity. Both diseases cause abortion and reduced productivity. They are also highly infectious and hard to eradicate from a herd or area without active interventions to identify, treat and/or remove infected animals. Vaccination of livestock against brucellosis (Njeru et al. 2016) and leptospirosis is rarely done in the region and in Kenya, and therefore, the high seroprevalences found for these pathogens are likely due to natural exposure. The ELISA kits used are also not perfect as the specificity of both kits have been estimated to range from 98 to 100% for Brucella spp. (Tschopp et al. 2015) and to be about 85% for Leptospira spp. (Barrett et al. 2018). The lack of 100% specificity in both ELISA tests may have yielded low rates of false positives. Nevertheless, these high seroprevalences are aligned with previous studies in the region: for example, 35.6% for Leptospira spp. in Kenya (Rajeev et al. 2017) and 7.5 to 40% for Brucella spp. in various pastoral production systems across Africa (McDermott and Arimi 2002). While our study did not assess humans’ exposure to Brucella spp. or Leptospira spp., the communities in the surveyed zones may be at risk of zoonotic infection through food (e.g., raw milk consumption) and occupational exposure. Indeed, an earlier study conducted in the area reported a seroprevalence of 21.2% (range 13.8–35.9) among hospital patients with flu-like symptoms (Muriuki et al. 1997). A recent study conducted in the area also reported clinical signs that are compatible with brucellosis in cattle and among animals handlers (Enström et al. 2017).

Our analyses accounted for clustering of cattle within herds and villages using herd and village IDs as random effects, respectively. We found the intra-cluster (intra-herd) correlation coefficients (ICCs) for both diseases to be moderate. This is likely due to cattle within herds sharing a common environment (i.e., common grazing and shared water sources) and similar management practices (Segura-Correa et al. 2010). With respect to villages, there was a substantial clustering of herd-level brucellosis while for Leptospira spp., the low ICC indicated lack of village-level clustering (i.e., herd-level exposure was independent of villages).

The acquisition of new animals through purchase was an important risk factor for herd-level brucellosis, in agreement with earlier reports in Uganda (Bugeza et al. 2018). This finding could be due to the likelihood of introducing infected animals into a healthy herd if the health status of sourced animals is not determined or temporally quarantine is not enforced. Further studies should clarify how herd dynamics due to livestock offtake or purchase can influence the prevalence of Brucella spp. in the area. Brucella spp. seroprevalence was also higher as the sites got closer to the Mara reserve. This finding may be associated with the different land use types adopted in the surveyed zones. Land use changes are thought to modify the interactions between host species and thus can directly or indirectly influence the level of pathogen transmission between hosts (Patz et al. 2004; Gottdenker et al. 2014). In zone 1, for example, cattle are raised in extensive systems as farmers utilize wildlife conservancies and MMNR compared to zones 2 and 3 with sedentary and crop-livestock mixed agriculture, respectively. Extensive livestock production systems (e.g., pastoralism) allow multiple herds to share common grazing and watering points which may increase chances of naive cattle encountering infected or carrier state animals including wildlife (McDermott and Arimi 2002). Indeed, this study identified pastoral husbandry practice as a significant predictor of both brucellosis and leptospirosis seropositivity in cattle. For Leptospira spp., seroprevalence in zones 1 and 2 differed significantly with zone 3 (low interface area), but not between zones 1 and 2. The lack of significant differences in seroprevalence between zones 1 and 2 indicated that variations in land use patterns between the two zones alone may be inadequate to show exposure difference for this pathogen in cattle.

The higher seroprevalence of Brucella spp. in zone 1 compared to zones 2 and 3 could also be partly due to the likelihood of high interactions between wildlife and livestock, given that these animals graze within the MMNR. Although the biological sampling of wildlife was not conducted in the study, the interactions between wildlife and livestock in the area are a possible factor that could also account for the differences in seroprevalence of this pathogen in the zones. Zone 1, for example, has a higher diversity of wildlife species (i.e., wildlife host species richness) than zone 3, which could increase infectious disease transmission as it may create a large pool of pathogen reservoirs (Daszak et al. 2000; Keesing et al. 2010) including Brucella spp. which is shared with cattle (Godfroid 2018). Information on brucellosis (Njeru et al. 2016) and leptospirosis in wildlife species is very limited in the area and indeed in Kenya, but Brucella spp. exposure in various wildlife species including the African buffalo (Syncerus caffer) and blue wildebeest (Connochaetes taurinus) has been documented in the Mara ecosystem (Waghela and Karstad 1986). Besides wildlife, rodents are also important sources of various Leptospira species (Allan et al. 2015) and can contaminate grazing areas or watering resources utilized by livestock. Both Leptospira spp. and Brucella spp. are known to persist in the environment with survival duration being affected by factors such as ultraviolet (UV) light, pH, salinity, soil moisture and temperature (Estrada-Peña et al. 2014). The persistence of Brucella spp. in water and soil may range between 21 and 81 days (Aune et al. 2012), while for Leptospira spp., it can vary from hours to 193 days (Casanovas-Massana et al. 2018). The ability of these pathogens to persist in water or grazing areas can influence the indirect transmission processes of zoonotic diseases if these resources are contaminated with infected excreta or urine (Mwachui et al. 2015). Sharing of these ecological resources by different livestock herds may also promote direct transmission of zoonotic diseases through increased intra- and inter-herd interactions (Rajeev et al. 2017). Indeed, our study found common utilization of watering points, grazing areas or mixing of cattle herds at these key resources as important predictors of brucellosis and leptospirosis seropositivity in cattle. Whereas the role played by small ruminants (sheep and goats) in the epidemiology of Brucella spp. and Leptospira spp. in the area is largely unknown, the interactions between cattle and small ruminants may also increase the interspecies transmission levels of these pathogens. Small ruminants are increasingly becoming important sources of household livelihoods in the area (Løvschal et al. 2019), and their population densities are estimated to have increased by 235.6% between 1977 and 2014 compared to cattle populations by 0.8% between same period (Bedelian and Ogutu 2017). The high population densities of small ruminants in the area may also create a large pool of maintenance hosts for these pathogens.

This study found higher seroprevalences of Brucella spp. and Leptospira spp. among female cattle than males. In general, cows have lower offtake rates than bulls in Maasai Mara ecosystem as they are raised to provide milk, an important diet for the locals (Nthiwa et al. 2019), and also for breeding purposes to replace animals that may die due to recurrent droughts (Huho et al. 2011). As cows stay in herds longer than bulls, they could have high chances of repeated exposure to these pathogens over time. The high proportion of exposed females also presents a major risk of transmission to male populations through natural breeding which is predominant in the surveyed zones.

The finding that positive history of abortions among surveyed herds was associated with animal-level leptospirosis could be due to poor husbandry practices such as improper disposal of aborted fetuses and placenta, resulting to environmental contamination (Mwachui et al. 2015). Aborting animals retained in the herds may also act as sources of infections with subsequent parturitions through uterine discharges (Loureiro et al. 2017). Although abortions in cattle are caused by many diseases (e.g., foot and mouth disease, bovine trypanosomiasis and contagious bovine pleuropneumonia), our results suggest that Leptospira spp. could be one of the major causes in the area and further studies should clarify this finding. The positive association between large herd sizes (≥ 50 animals) and Leptospira spp. exposure of animals may be due to greater animal contacts within larger herds (Barrett et al. 2018). The management of large herds also involves frequent movements in search of water and pasture, more so in dry season. This practice may contribute to the spread of infectious diseases but may also expose herds to diseases that may be limited to an area (Alhaji et al. 2016).

This study provides data on the current epidemiological situation of Brucella spp. and Leptospira spp. exposure in cattle herds raised in the Mara ecosystem. Our findings demonstrated that both diseases are prevalent in the area and had a considerable level of co-exposure in animals. Seroprevalence of Brucella spp. was higher in areas near Mara reserve (zone 1) compared to other zones. For Leptospira spp., zones 1 and 2 had significantly higher seroprevalence than zone 3. The seropositivity of both diseases was also significantly associated with grazing cattle in wildlife reserves. As these pathogens could spill over from wildlife reservoirs into livestock in areas with close interactions, further studies are needed to establish exposure levels in wildlife, sheep and goats and humans. Furthermore, mapping the transmission routes of these pathogens and quantifying their impacts on cattle production will help in the development of appropriate control strategies.