Bovine tuberculosis (bTB) caused by Mycobacterium bovis, is a chronic debilitating disease of livestock, wildlife and humans [1, 2]. Cattle may serve as the main host for M. bovis worldwide [3], while many or most other species such as possums, pigs, cats, dogs, horses and sheep are considered to be spill-over hosts [4]. Aerosol is considered to be the main route of infection in animals [3, 4]. Other routes of infection such as ingestion of contaminated feeds, water and fomites have been identified [3]. Information concerning routes of transmission and different potential sources of infection in Africa is scarce [5].

Bovine tuberculosis is a disease with potential public health and economic importance [6] since it can affect international trade of animals and animal products [7]. The presence of bTB in domesticated and wild animals in synergy with the HIV pandemic in developing countries makes zoonotic tuberculosis a potential threat to human health [1, 8]. In developed countries, bTB has been controlled through ‘a test-and slaughter policy’. Nevertheless, bTB remains a problem in most developing countries where surveillance and control activities are often inadequate or unavailable [6] possibly due to lack of funds to support the whole exercise and compensate for tested and slaughtered animals in these countries.

Previously published information indicates that bTB is endemic in Tanzania’s cattle, with regional prevalences ranging from 0.2% to 13.2% [5, 9–15] suggesting the presence of foci of infection [13]. This could be underestimated if not confirmed by currently available bacteriological or molecular techniques [16]. A high prevalence (13.2%) of bTB was reported in pastoral cattle in the southern highlands of Tanzania and was associated with high numbers of indigenous cattle kept under intensive husbandry practice [17]. In Tanzania, M. bovis has been isolated from human lymph biopsies [18, 19], cow’s milk and tissue samples from slaughter houses [12, 19] and from a range of wildlife species including migratory wildebeest, topi and lesser kudu [20]. In their study, Renwick et al. [21] revealed bTB to be established in domestic stock and recently native wild bovids particularly African buffalo has been infected. In Tanzania, up-to-date information concerning risk factors for transmission of bTB at the livestock-wildlife interface is lacking. Studies conducted in the Southern Highlands and northern parts of the country reported an association of bTB infection with age, sex, breed, lactation and variation in climate [17] and proximity to wildlife [5]. Likewise, studies conducted in other African countries and elsewhere reported an association of bTB infection with age, sex, physiological status and husbandry practices such as cattle movement and contact with wildlife [22–26].

The Serengeti ecosystem comprises an ecosystem defined by annual movement of herds of ungulates [27] interacting with populations of livestock. The husbandry practice of most of the pastoral communities in the region is pastoralism, based on transhumance, which refers to a pattern of seasonal movement between dry season and wet season pastures. Given the husbandry practices and proximity to wildlife, studies are needed to explore the disease status, dynamics in the ecosystem and risk factors for bTB infection in animals and ultimately also humans. Here we present findings of a study conducted to determine the prevalence of bTB and risk factors associated with bTB infection in pastoral and agro-pastoral communities at the livestock-wildlife interface in the Serengeti ecosystem in Tanzania.

This study has shown an overall bTB apparent prevalence of 2.4% in cattle around the Serengeti ecosystem. Generally the prevalence in this study is consistent with studies conducted in other regions of Tanzania [5, 11, 13–15] and like for other areas previously studied, bTB is endemic. Results from this study show low prevalence of bTB as compared to a previous study by Kazwala et al. [17] in Southern Highlands of Tanzania. This discrepancy could be attributed to difference in sample sizes and study design. The herd prevalence of 50% (16 out of 32) might reflect that bTB infection varies considerably between herds and is widespread in pastoral and agro-pastoral communities in the Serengeti ecosystem and surrounding areas.

The study has also shown that most of the cattle owners interviewed in villages had poor knowledge on bTB. Moreover, the majority of the respondents had never heard of bovine tuberculosis and were not aware of bTB presence in wildlife. The infection of pastoral and agro-pastoral cattle with bTB, poor community knowledge on the mode of transmission for bTB among people poses serious risks of infection with zoonotic diseases including bTB to people in the Serengeti ecosystem. Therefore, there is a great need for awareness creation and community involvement in planning and implementation of disease control programmes especially zoonotic diseases in the Serengeti ecosystem.

The prevalence of bTB among the three districts varied considerably with significant differences between the three districts (Table 2). Although our point estimates of prevalence in the three districts varied five-fold, the sample size was not large enough to be sure that these are significant differences. The relatively high proportion of bTB infection in Ngorongoro district compared to other two districts might be due to husbandry practices of semi-nomadic system where the Maasai pastoralists move their cattle during the dry season searching for water and pasture [40], and multiple land use system in Ngorongoro where human, livestock and wildlife live in the same area [31]. These practices predispose cattle to bTB infection due to higher chances of coming in contact with infected animals. Nevertheless, it is proposed that bTB occurs in hotspots [41]. The difference in bTB prevalence among hotspots are influenced by ecological factors, cultural factors, livestock trade patterns, host populations, timing of contact and reproductive rate of pathogens [41–43].

Previous studies conducted in wildlife in Serengeti have confirmed the presence of M. bovis in buffalo, wildebeest and lion [5]. The buffalo is considered as the natural host of M. bovis and its social behaviour of living in large herds provides favourable conditions for aerosol transmission of M. bovis to other members of the same herd [44]. The migration of wild animals traversing the Serengeti ecosystem heading northwards to Maasai Mara in Kenya could contribute to the spread of bTB to the nearby surroundings’ villages. For example, 88.9% of the cattle owners interviewed in this study reported contact of cattle with wildlife at water sources. In this study, there was a significant association between age and intradermal skin positivity with more positive animals in medium aged groups (2–4 years). These are weaners which are likely to get exposed to infection with time and remain sensitised to bovine tuberculin for the rest of their life [40]. While young animals are unlikely to get infected with M. bovis, it is said that age alone is not enough to account for susceptibility to infection but continued opportunities to exposure to the bacterium. Previously published studies conducted in the Southern Highlands and northern regions of Tanzania [5, 17] indicated a trend-wise increase in positive reactivity as age increases which was not the case for our study. Existing belief is that, in endemic situation, the duration of exposure to bTB infection increases with age [5, 17, 24]. Similarly, the size of the households correlated to intradermal skin positivity with more reactors in medium sized households. Despite the lack of a clear explanation for this association, other risk factors can play a role as pointed out in a stratified classification of worldwide bovine tuberculosis risk factors in cattle [2]. For example animals regardless of size of the household herdsmen could move their animals to longer distances in search of pasture and water sources. In so doing chances of meeting infected herds including wildlife may occur thus exposing to infection.

Findings from this study have indicated no significant association between sex and intrademal skin positivity (Table 1). The finding concurs with previous studies conducted in northern regions of Tanzania and elsewhere, which reported similar findings for bTB positivity between male and female cattle [3, 5]. Contrary to our findings, other studies have reported an association between sex and intradermal skin positivity [17, 24, 45]. For example Inangolet et al. [45] and Cadmus et al. [24] reported female cattle being at greater risk of testing positive than males contrary to findings by Kazwala et al. [17], where male cattle were more affected than female cattle as male cattle particularly castrates are kept longer and hence more chances of contracting a disease than female cattle. In controlled studies no difference in susceptibility to bTB in relation to gender/sex has been found (Anita Michel, personal communication). It is probably more the likelihood of production type (e.g. dairy cows) and exposure that can cause a difference, e.g. breeding bulls which are shared between farmers.

Our results have shown that about 64.8% of respondents reported presence of very thin and emaciated animals in their herds. However, through physical observation, most of thin and emaciated animals in those herds tested negative for bTB as compared to positive reactors whose body condition was good. This supports previous findings by Amen et al. [22] and Munyeme et al. [46] in which most of positive reactors cattle had good body condition as compared to negative reactors. With exception of animals in advanced stages of bTB when emaciation and laboured breathing are more prominent clinical signs, body condition has a direct association with individual health condition and it is probably not a reliable determinant for bTB infection. During this study, it was interesting to note that some very healthy animals in Bunda district, although had less bTB overall as compared to other districts, tested positive to tuberculin skin test. The good body condition score (even for infected cattle) in Bunda district might be due to quality and plentiful pastures and water for cattle, making animals look healthier than in Serengeti and Ngorongoro districts. Poor body condition to animal herds in Ngorongoro districts could possibly be attributed by drought condition in Maasai steep lands and long trekking distance where animals move long distances every day searching for pastures and water sources. Experience in human TB, the negative tuberculin skin testing (TST) (the equivalent of bTB infection) is often seen in people with overwhelming infection or disease, or with measles, HIV or malnutrition [47, 48]. In our study site bTB is not routinely tested in cattle at slaughter houses for association of TB lesions with emaciation. Available reports do not indicate whether the same is true of cows and whether some of these emaciated animals have false negative reactions.

The study has shown a high prevalence of atypical mycobacteria in cattle, which signifies environmental contamination. Cattle could have acquired the infection through contaminated environment during grazing or at water sources. The high prevalence of atypical mycobacteria in this study concurs with previous findings in Tanzania whereby Shirima et al. [13] and Durnez et al. [15], reported a prevalence of 6% and 10.1% of atypical mycobacterioses respectively. In their study Mdegella et al. [14] and Durnez et al. [15] found a prevalence of 14% and 19% of atypical mycobacteria in milk samples that could expose milk consumers at great risk of contracting milkborne zoonotic infections. Consumption of undercooked meat and unpasturalized milk is a common practice in most pastoral communities in Tanzania [17]. The prevalence of 10.6% of atypical mycobacteria obtained in this study may alert for a possibility of immune compromised individuals such as HIV/AIDS patients and puts them into a risk of being infected with opportunistic infections. This is particularly critical if at all the atypical mycobacteria such as M. avium and M. fortuitum are shed in meat or milk from cattle as it happens in case of M. bovis. Atypical mycobacteria which are commonly environmental have been reported in cattle farms [49] with wide distribution in nature (soil, water, animals and humans).

Available reports revealed atypical mycobacteria to include M. gordonae, M. smegmatis, M. fortuitum, M. phlei, M. flavescenes and M. avium intracellular to be shed in cattle milk samples in Morogoro region and Kibaha, Tanzania [14]. M. avium is said to be clearly a human pathogen in infants and also in HIV-infected adults; but the others are arguably less virulent, with M. fortuitum very rarely causing disease in individuals with genetic immune system problems. Others are regarded to be commensals. However, the non-specific reactions are not always caused by atypical mycobacteria. Sometimes, other closely related bacteria species (Nocardia, Corynebacterium, Trueperella etc.), as well as some other factors can cause the unspecific reactions (Anonymus, 2013).

The study has shown that older cattle (>4 years) had a high response to PPD and doubtful reaction as compared to young cattle, with more doubtful reaction in females as compared to males (Figure 4). It is believed that doubtful reaction in older cattle might be due to immune suppression owing to old age [45]. Furthermore, it is suggested that, stressful conditions such as drought conditions, return of animals from transhumance, long trekking, clinical ill-health due to trypanosomiasis, tick borne diseases, ectoparasites, heavy burden of endoparasites and other environmental stressors could contribute to higher rates of doubtful reactions [50]. However, in this study the doubtful reactors were not re-tested after 60 days as per International Office of Epizootics (OIE) guidelines due to time limitation and field logistics.

In conclusion, findings from this study add useful epidemiological data/information regarding bTB infection at the livestock-wildlife interface in Tanzania. It was revealed that agro-pastoral and pastoral communities in the study area have low knowledge on bTB and its presence in other animals including wildlife. Moreover, majority of the community engage in husbandry practises such as free movement of cattle, sharing of water sources and pasture which increase chances of diseases transmission to livestock. This information is useful and could be used by respective authorities in designing appropriate control measures against bTB. Therefore, we propose that education programs should be implemented in pastoral communities to raise awareness on preventive measures against bTB and other infectious diseases. The education programs at livestock-wildlife interface areas should primarily focus on minimizing contact between cattle and wildlife. However, in order to be more certain whether strains are circulating between cattle and wild animals, we need to cultivate mycobacteria and use genetic strain typing methods to characterise them accurately.