Problem IdentificationZCharacterization Triad

7.3.1 Agent

Mycobacteria are grouped in the order Actinomycetales, along with other genera such as Corynebacterium, Nocardia, and Rhodococcus. These bacteria contain unique mycolic acids in their cell wall that confer distinct immune-stimulatory properties and resistance to drying, acidity/alkalinity, and many antibiotics (Sakamoto 2012).

Theobald Smith (1898) initially classified the mammalian tubercle bacilli into various types based primarily on the differences in virulence of cattle-derived (M. bovis) and human-derived (M. tuberculosis) strains. Although the differences in susceptibility are by no means absolute (Lewis and Sanderson 1927), calves and rabbits inoculated with minimal amounts of cultures of bovine-type bacilli devel­oped progressive, usually fatal TB, whereas equivalent amounts of cultures of the human type only caused local, non-progressing lesions.

7.3.1.1 Transmission

An understanding of the routes of transmission is critical to effectively control BTB. Determining the source of infection is often difficult, if not impossible (Brush 1898) because after becoming infected, it can take a year or even up to a decade for the animal to become visibly diseased. Although certain physiological characteristics are common to all of the susceptible hosts, the course of infection and the rate of the development of the disease and clinical signs can vary substantially in the various host species (Biet et al. 2005). The locality of the primary lesions observed in necropsied animals provides the best indication of the route of infection (Phillips et al. 2003). Tuberculous lesions are frequently found in more than one anatomical site in the same animal. However, both direct inoculation and in-contact infection in cattle resulted in lesions largely confined to the respiratory tract, confirming this location as the primary site of infection in cattle (Cassidy et al. 1998).

Respiratory Route (Aerosol Transmission) In the respiratory route of infection, mycobacteria, from an infected animal (or human) with open pulmonary lesions, are aerosolized in the respiratory tract and shed when exhaling. This creates a contam­inated air space, putting any susceptible human or animal sharing that air space at risk (Renwick et al. 2007). That cattle serve as sources of M. bovis is supported by the bacteria consistently being isolated from the upper respiratory tract of 30-50% of tuberculin-test-positive reactors (Costello et al. 1998).

Tuberculous lesions in cattle are much more commonly found in the lungs and associated lymph nodes than in the abdominal organs or in the rest of the carcass (Table 7.1). That tuberculous lesions occur predominantly in the respiratory tract and associated lymph nodes of naturally (Langmuir 1961; Wilesmith et al. 1982; Collins and Grange 1983; Morris et al. 1994; O'Reilly and Daborn 1995; Whipple et al. 1996; Cleaveland et al. 2007) and experimentally infected cattle (Cassidy et al. 1998) have been accepted as sufficient evidence that the respiratory route is the primary portal of entry in this species. The minimum dose required to infect calves via the respiratory tract was up to 1000 times less than that required to infect animals via the oral route (Collins and Grange 1983). That few mycobacteria are required to establish an infection is also supported by mathematical models indicating that an M. bovis infection could be established in cattle by the inhalation of a single bacillus within an aerosol droplet (Neill et al. 1991).

The locality in the body of tuberculous lesions in livestock other than cattle and in wildlife species throughout Africa indicates that the respiratory route is also the primary route of infection in most of them (see Chap. 5 for a more detailed discussion about the situation in the various wildlife species).

Respiratory Excretion of M. bovis The results of several epidemiological investi­gations suggest that naturally infected cattle in the early stages of the disease do not readily infect other cattle.

Table 7.1 Distribution of visible lesions of BTB in naturally infected animals in Africa

NR not reported ^aLung lesions ^bLymph nodes ^cNumber of cases

organisms excreted, the duration and closeness of contact, and the production of aerosols of small particle size that contain viable mycobacteria (O’Reilly and Daborn 1995). Even when these criteria are satisfied, transmission of the disease is not rapid. This assumption is supported by the findings of an experiment where only 4 of 10 in-contact animals housed with 20 tuberculin-test-positive cattle for a year contracted the disease (Costello et al. 1998). A model simulating transmission within New Zealand cattle herds estimated that each M. bovis-infected animal would infect an average of 2.6 cattle during the course of a year (Barlow et al. 1997).

That cattle-to-cattle transmission does not occur rapidly benefits the test-and- slaughter campaigns in countries in the absence of wildlife or other maintenance hosts (Anon 1994). The problem in most of Africa is aggravated by the practice of not immediately culling infected animals after they tested positive for BTB. Under these circumstances, those with advanced pulmonary lesions continue to shed large numbers of organisms, and they remain an ongoing source of infection.

Oral Route of Infection The distribution of lesions in tuberculous cattle, goats, and camels indicates that the oral route of infection may be significant in certain areas where traditional African livestock husbandry is practiced (Table 7.1). In Tanzania, for instance, the majority of visible lesions in cattle in one study were found in the gastrointestinal rather than in the respiratory tract (Cleaveland et al. 2007). Though oral transmission may occur in several ways, such as by the consumption of infected feed and water, during predator/prey interaction, pseudo-vertically, and percutane- ously (Renwick et al. 2007), there are indications that, with some exceptions, this mode of transmission is uncommon and not important from an epidemiological perspective.

Feed and Water Environmental contamination by fecal and nasal excretion of M. bovis appears to be a less important mode of transmission as the bacteria are irregularly excreted (Neill et al. 1988). As an example, in an extensive study assessing the likelihood of transmission from infected pastures on BTB-positive farms, only mycobacteria other than M. bovis were cultured from the environmental substrates (Fine et al. 2011).

Mycobacteria can readily survive in water, and it is a major source of environ­mental mycobacteria that cause some of the non-tuberculous mycobacterioses in humans (Dailloux et al. 1999). Animals excreting M. bovis in fecal and nasal discharges contaminate running water (Humblet et al. 2009), and seasonal flooding in the environs of a village is a significant risk factor in the transmission of M. bovis in cattle (Cleaveland et al. 2007). Transmission by drinking M. bovis-contaminated water, however, is considered of little importance because the dilution of the numbers of mycobacteria in running water reduces the population density of M. bovis to a level where too few organisms occur to cause infection (Phillips et al. 2003). Stagnant water, however, can become heavily contaminated with animal waste containing M. bovis, and it constitutes a significant risk factor in smallholder/ pastoral production settings (Oloya et al. 2007).

In terms of wildlife in South Africa, it appears that tuberculous buffaloes, some with advanced pulmonary lesions making them active shedders of M. bovis, do not commonly shed M. bovis in sufficient numbers in nasal or oral discharges to contaminate water and forage, and this potential source of infection appears also not to play an important role in the transmission of BTB in free-ranging ecosystems (Michel et al. 2007). Here too, non-tuberculous Mycobacterium species were cul­tured and appear to thrive under these conditions.

Contaminated feed and pastures may be a source of M. bovis but appear to play a negligible role in the transmission of BTB because M. bovis on fomites does not remain infective for long (Morris et al. 1994). Repeated environmental contamina­tion does appear to increase the chances of oral infection (Maddock 1934) (refer to the section for further information). However, recognizing the normally rapid rate of dispersal of fecal matter and the resulting low bacterial numbers, oral transmission is unlikely to pose a major risk. There are examples, though not in Africa, of trans­mission through the sharing of feed and exposing cattle to pens that contained deer infected with M. bovis (Palmer et al. 2004) and the speculated role of latrines in the transmission of BTB from badgers to cattle (Humblet et al. 2009).

Other Sources of Oral Transmission Another type of oral transmission occurs during predator/prey encounters, which involves the consumption of infected tissues (lesions, blood, internal organs, etc.) by a susceptible predator. This type of trans­mission is a cause of concern in conservation areas (Michel et al. 2006), but dogs and cats can also fall victim to this type of transmission. Consumption of infected offal (and perhaps meat) poses a moderate risk because of the high oral dose that is required to establish an infection (Drewe et al. 2014).

Pseudo-vertical transmission can occur through consumption of infected milk or simply from close contact between lactating dams and their offspring. According to Michel (2008), social behavior, including regular and close physical contact such as licking, grooming, and suckling between members of the same herd over a protracted period of time, provides favorable conditions for pseudo-vertical trans­mission of M. bovis. Evangelista and De Anda (1996) examined the TB status of two groups of calves in M. bovis-infected dairy herds in Mexico. From 91 calves fed with pooled colostrum and raw milk from the bulk tank, 27% tested positive to the caudal fold test (CFT) at 6 months of age, whereas only 15% (n = 279) of calves fed with colostrum derived from tuberculin-negative cows and powdered milk were positive. Furthermore, the risk of skin-test-positive reactors increased from 11% in calves born from skin-test-negative cows to 33% in calves born from skin-test-positive cows. However, Francis (1950) noted that since only about 1% of tuberculous cows have TB of the udder, the risk of calves being infected by milk is considered low, unless they were fed pooled milk from the entire herd. Similarly, Gallagher et al. (1972) found no cases of TB in immature (under 18 months) lechwe (Kobus leche kafuensis) in Zambia, suggesting that neither congenital nor the consumption of infected milk was common. The offspring of infected cattle also do not appear to be at a significantly increased risk to become infected with M. bovis (Drewe et al. 2014).

Percutaneous Transmission A unique mode of transmission involving the percu­taneous route has been documented in greater kudus (Tragelaphus strepsiceros) (Renwick et al. 2007), where transmission is caused by contact with thorns contam­inated with purulent exudates discharged from fistulated parotid and submandibular lymph nodes containing tuberculous lesions.

7.3.2 Host-Related Risk Factors

Cowie et al. (2014) define risk factors as aspects of the system that influence the introduction of a disease and persistence of the disease in a population. Several studies have been conducted in Africa to provide information about the role of farm- and animal-level characteristics in the epidemiology of BTB. Most of the factors identified as consistently contributing to the occurrence of BTB were locality­specific (Awah-Ndukum et al. 2012) and related to the type of farm management (Berrada 1993). Specifically, the prevalence of BTB was high in peri-urban and urban areas where large numbers of dairy farms keep exotic cattle breeds (Tschopp et al. 2010). In contrast, low prevalence was recorded in pastoral settings, where indigenous cattle breeds are kept under a traditional management system (Gumi et al. 2012). At the livestock-wildlife interface (Munyeme et al. 2008), and in smallholder settings (Bernard et al. 2005), where animal movement between farms is uncontrolled (Firdessa et al. 2012), BTB was seen to spread very quickly. Many of these risk factors have been identified elsewhere (Marangon et al. 1998), and improving farm biosecurity is known to achieve effective control (Skuce et al. 2012; Cowie et al. 2014).

Animal-level risk factors also were extensively explored in Africa. It is, however, important to emphasize that BTB is a herd problem, and attempts to control the infection should be directed at herd/farm-level risk factors. In the following section, herd- and animal-level risk factors of BTB, with reference to African livestock production, are discussed.

7.3.2.1 Herd-Level Risk Factors

Herd Size A large herd size in association with a high cattle density and the housing of cattle, often exacerbated by between-herd movement, facilitates cattle-to-cattle transmission of BTB. In Africa (Cook et al. 1996; Asseged et al. 2000; Cleaveland et al. 2007) and on other continents (Griffin et al. 1996; Munroe et al. 1999; Karolemeas et al. 2010), herd size has been determined as one of the major risk factors for the transmission of BTB. However, in many regions of the African continent where animals constantly live in the open and under extensive conditions, BTB, irrespective of the herd size, is rare (Alhaji 1976; el Sanousi and Omer 1985). It is possible that herd size per se is not that important but that it is a proxy variable for other risk factors, such as a high animal density (Griffin et al. 1993). The detection of BTB reactors too is a problem in large herds. Since the tuberculin skin test is not perfect, the probability of a false-positive reactor would be greater in large herds (Humblet et al. 2009). Furthermore, the more animals that are skin-tested, the higher the probability to have a reactor. However, in a survey of cattle belonging to Somali and Guji pastoralists of southeastern Ethiopia (Gumi et al. 2012), herd size was not associated with the outcome of tuberculin tests. By analogy, the general practice of tuberculin skin testing of cattle and early removal of those infected with M. bovis has led to a progressively lower prevalence of BTB in all countries that have BTB control programs. In those countries, herd prevalence has effectively been reduced to less than the critical community size (CCS), and hence cattle are then no longer maintenance populations for BTB. The objective is to prevent spillover to cattle from other main­tenance or spillover hosts. In Africa, where cattle are still the main source of M. bovis infection for all other species, including humans, this situation will prevail until such time that effective control programs have been implemented in those countries with infected livestock and wildlife populations.

Transmission and Host Dynamics Two categories of hosts are distinguished in BTB: maintenance and spillover hosts (Humblet et al. 2009). The maintenance of an M. bovis infection in a given species (a maintenance host) can be achieved only when sufficient intra-specific transmission occurs to sustain the disease at and beyond the threshold level. This level is determined by the basic reproductive number (R₀), which is defined as the expected number of new cases produced in a population by a single infected animal (Renwick et al. 2007). The threshold for disease mainte­nance by intra-specific transmission alone is met when R₀ ≥ 1, meaning that each infected animal transmits the disease to one or more individuals (Palmer 2013); where R₀ < 1.0, the disease will disappear from the population. In addition to other factors, R₀ is determined by the abundance of the host species (Nugent 2011), referred to as the “critical community size” (CCS), which is the minimum size of a closed popu­lation within which a pathogen can persist indefinitely (Palmer 2013). In populations smaller than the CCS, the number and/or density of infected host levels are low enough for random extinction of the pathogen to occur.

Threshold values such as R₀ and CCS are not clear-cut, and they are also difficult to measure (Palmer 2013). The situation is also much more complex in multi-host ecosystems, such as those in some of the South African and Tanzanian national parks where these species-specific thresholds can be perturbed by the interspecies trans­mission of BTB from either maintenance or spillover hosts (Nugent 2011). In Africa, maintenance hosts include cattle (Cousins 2001) and wildlife species such as African buffaloes (Renwick et al. 2007), greater kudus (Tragelaphus strepsiceros), and Kafue lechwe (Gallagher et al. 1972), while several domesticated species, a number of wild ungulates and a few wild carnivores are spillover hosts (O’Reilly and Daborn 1995; Renwick et al. 2007). Since the common elements in maintenance hosts appear to be close family groupings and an aerogenous route of excretion of M. bovis (Fitzgerald and Kaneene 2012), it is possible that some of the other species may be able to become maintenance hosts of M. bovis under particular circum­stances. Only a small proportion of these spillover species, however, have the required attributes to become maintenance hosts (Biet et al. 2005).

Infection persists by intra-specific transmission alone in maintenance hosts, and they are a source of infection for other species. The presence of a single maintenance host in an ecosystem is likely to cause the spread of TB to other susceptible species within that system. Conversely, in a spillover host, infection will not persist indef­initely unless there is regular reinfection from other species. It is important to keep in mind that transmission from spillover to maintenance hosts may occur and that both maintenance and spillover hosts may act as disease vectors in susceptible multi­species populations (Corner 2006).

The characteristics of an effective maintenance host include susceptibility to M. bovis, prolonged host survival after infection, the ability of infected females to reproduce, and the shedding of M. bovis via multiple routes (Drewe et al. 2014). Additionally, ethological factors such as gregarious behavior, and ecological char­acteristics such as eating behavior, population density, and interaction with other species are factors determining whether a particular species is likely to become a reservoir for M. bovis (Biet et al. 2005). Mycobacterium bovis-induced lesions such as extensive pulmonary lesions that drain into bronchi and bronchioles increase the likelihood of generating infectious respiratory aerosols that are required for the transmission and maintenance of the infection in a particular herd (Palmer 2013).

To cope with the complexity of BTB in a multi-host system, the threshold density theory is replaced by a threshold community configuration approach (Renwick et al.

2007). Within these complex systems, the rate of interspecies transmission is dependent on the contact rate between various host species (which is density­dependent) and on the way in which M. bovis is excreted. In a multiple host­species environment, the direction of flow of M. bovis varies according to the epidemiological role of the host species involved in the transmission of the infection.

Although cattle are the original and most common reservoir host (Cousins 2001), M. bovis also infects other domesticated mammalian species (O’Reilly and Daborn 1995). All these species, however, are not equally susceptible to M. bovis. In Africa, studies on animal TB focused mainly on cattle, and data about other domesticated species are scarce (Cosivi et al. 1998; Asseged et al. 2014).

Mixed Rearing of Domestic Stock Both pastoralists and smallholder crop­livestock farming communities in Africa employ a mixed livestock rearing system, in which small ruminants are important components and are herded together with cattle during the day (Boukary et al. 2012). At night, they are usually kept indoors in poorly ventilated farmers’ houses for protection against theft and predators. Such multi-host husbandry practices and the sharing of air spaces are epidemiologically important for potential lateral spread of BTB within and between small and large ruminants and for potential zoonotic transmission (Tschopp et al. 2011).

Goats Tuberculosis in goats is widespread in certain regions of Africa where they co-graze with BTB-infected cattle (Table 7.2). Goats are susceptible to infection by M. caprae, M. bovis (Jenkins et al. 2011; Napp et al. 2013), and M. tuberculosis (Hiko and Agga 2011), and it appears that their resistance to become infected with M. bovis is not high (Anon. 2013; Napp et al. 2013). When exposed to a high infection pressure, large numbers of goats can be infected, and rapid spread of the disease within a flock can be expected. The involvement of the respiratory tract in tuberculous goats renders them suitable to act as a maintenance host for BTB and to disseminate the disease (Napp et al. 2013; Pesciaroli et al. 2014). TB caused by M. caprae is known to spread from infected flocks to in-contact cattle herds (Napp et al. 2013).

Table 7.2 Prevalence of BTB in different domestic animals across Africa

PM, necropsy; CIT, comparative intradermal TST; -, no info ^aData for 1910 ^bData for 1916

Most BTB cases in goats in Africa occur as a result of contact with infected cattle (Cousins 2001). It appears, however, that mixing different species of domestic animals in the traditional, extensive farming practices does not play a major role in the transmission of BTB to goats. Exposure to M. bovis infection mostly occurs when sharing pastures, watering points, and night shelters with BTB-infected cattle (Deresa et al. 2013). Under those circumstances, the number of infected flocks is low, not exceeding 2.9%, and with an equally low herd prevalence of 0.2% (Tschopp etal. 2011; Gumietal. 2012).

Sheep There is common belief that sheep are innately resistant to tuberculosis (Anon 2013). They do, on the contrary, appear to be quite susceptible to the infection and are at times infected with M. bovis or M. caprae with the development of lesions particularly in the respiratory tract, which are both macroscopically and histologi­cally very similar to the tubercles seen in cattle with BTB (Cordes et al. 1981; Davidson et al. 1981; Malone et al. 2003). TB is infrequently reported in sheep in Africa (Tag el Din and el Nour Gamaan 1982; Fullerton 1902), and its prevalence in this species is usually very low (Fullerton 1902). Tuberculous sheep have been detected in a number of countries (that reported its presence) including Egypt (Moustafa et al. 1964), Sudan (Tag el Din and el Nour Gamaan 1982), Niger (Boukary et al. 2012), (Kassa et al. 2012), and Nigeria (Bala et al. 2011) (Table 7.2). The reasons for the low prevalence of BTB in sheep probably are multiple (Allen 1988). Sheep appear not often exposed to high infection pressures of TB because of specific management and behavioral factors. They are usually extensively managed, and they mostly graze during daylight hours. These extensive conditions limit the spread of TB within flocks (McFadyean 1900). Furthermore, sheep tend to flock together both when grazing and when resting, thus evading contact with other species and limiting interaction with potentially tuberculous species to the minimum (Malone et al. 2003). They are more likely to acquire the infection when sharing pastures (Davidson et al. 1981), during communal housing with infected cattle (Malone et al. 2003), or during close contact with infected wildlife species (Deresa et al. 2013).

In sheep, lesions occur in the lungs, bronchial and mediastinal lymph nodes, and in the liver and spleen (Moustafa et al. 1964; Marianelli et al. 2010). In some instances, lesions are caseous, calcified, and well encapsulated, creating the impres­sion that sheep have a degree of resistance to the infection causing it to remain localized.

The perception is that the low prevalence of BTB in sheep is likely to be the consequence of inadequate diagnostic facilities on most of the continent (Tag el Din and el Nour Gamaan 1982), and the prevalence of the disease in sheep may be substantially higher in those countries where it has been detected. Because tubercu­lous lesions develop in the respiratory tract, sheep are considered to be a potential maintenance host of M. bovis, particularly when they share grazing with a number of species (Malone et al. 2003).

Camels Tuberculosis in dromedary (Arabian) camels (Camelus dromedarius) was first reported in Egypt in 1888, and M. bovis was confirmed as the cause of the disease in 1911 (Mason 1912). A prevalence of BTB of 3.2% (n = 1579) was recorded in camels slaughtered at Cairo abattoir during the course of 1916 (Mason 1917).

While they are susceptible to both experimental and natural infection with M. bovis and M. tuberculosis, M. bovis is more commonly isolated from tuberculous camels. In two reports, 97.4% (n = 76) and 2.6% and 93.4% and 6.6%, respectively, of bovine and human bacilli were isolated from tuberculous lesions (Elmossalami et al. 1971; Kinne et al. 2006). The disease is more frequently seen in camels kept in close contact with cattle than in those living under extensive conditions (Mason 1917; Elmossalami et al. 1971). Accordingly, camels of the Bedouin in Egypt were seldom affected compared to those owned by farmers, the Fellaheen, who custom­arily at night keep their camels indoors in close contact with cattle. It is most likely that BTB-infected camels could be a source of infection when introduced into a herd of healthy camels (Mason 1917).

Camels are also found in sub-Saharan Africa where owing to their drought tolerance and multiple economic values, traditional cattle-rearing communities are increasingly replacing their cattle with camels. During the transitional period of phasing-in and phasing-out, herds of camels and cattle are reared together during the day and rounded in a confined space at night (Beyi et al. 2014), thus increasing the risk of camels contracting M. bovis. Currently, the prevalence of BTB in camels in these areas appears to remain low as at the Niamey abattoir in Niger, only 0.3% of 2604 slaughtered camels contained lesions consistent with BTB (Boukary et al. 2012). A similar low prevalence of 0.3% was recorded in Nigeria (Bala et al. 2011), but a prevalence of 8.3% was reported in Eastern Ethiopia (Beyi et al. 2014). The low prevalence of BTB in camels in many African countries is probably because most of the camels remain in the open and harsh environment where it is difficult for the pathogen to spread. The prevalence of BTB in camels is higher than that of pigs and small ruminants in all the African countries where they occur.

In camels with BTB, the lungs and the bronchial lymph nodes are frequently affected, and lesions are limited to these organs in 60% of cases (Cousins 2001). In a study conducted in Eastern Ethiopia (Mamo et al. 2011), the lungs were the most frequently affected organ (40.0%) followed by the mesenteric (38.0%), mediastinal (12.5%), and retropharyngeal lymph nodes (8.3%). Additionally, generalized BTB, with numerous lesions in the organs and lymph nodes throughout the carcass, was fairly common (Table 7.1). The predominant occurrence of lesions in the respiratory tract of camels suggests that airborne transmission is the principal route of infection in this species.

A unique spoligotype (SB 1433) was isolated from camels in Niger (Boukary et al. 2012), where they are customarily reared separately from other species, and this unique strain appears to be maintained in camels in this particular region. Because of their relatively longer life expectancy, camels are an ideal maintenance host, once the infection has established itself in a fairly large population.

Pigs Tuberculosis is highly prevalent in pigs in Africa owing to their indoor management, frequent contact with infected cattle, and access to the contaminated material of animal origin. A study conducted in Uganda (Muwonge et al. 2010) indicated that 9.3% (n = 997) of pigs had lesions compatible with those of TB, and 33.3% (n = 93) of the carcasses with lesions yielded mycobacteria on culture.

Pigs usually acquire M. bovis from shared grazing or by the intake of M. bovis- contaminated dairy products, especially milk. On rare occasions, M. tuberculosis has also been isolated from pigs in Nigeria (Jenkins et al., 2011). Pigs are considered to be spillover hosts (Cousins 2001), and the disease does not seem to spread in a herd, and it often disappears from a herd when the source of infection, such as contam­inated milk, is removed (Anon 2013). Tuberculosis is now rare in domestic pigs in countries that successfully apply BTB control programs. In some countries, TB in pigs is re-emerging following the widespread adoption of outdoor pig farming systems (Pesciaroli et al. 2014).

Pigs appear to complicate the epidemiology of M. bovis, as they may be a source of infection for multiple species. Thus M. bovis strains isolated from the mesenteric lymph nodes of tuberculous pigs and from humans and cattle in the Northeastern part of Uganda had identical spoligotype patterns (SB 1469) (Muwonge et al. 2012).

The oral route appears to be the primary route of infection in pigs. Tuberculous lesions are thus common in the lymph nodes of the head, neck, and abdomen, but they may also be found in the lungs and abdominal viscera and in the mesenteric lymph nodes (Anon 1994). Tubercles often have an intense yellowish discoloration with a gritty texture. Although a rapidly progressive, disseminated disease with caseation and liquefaction of lesions has been reported in pigs, the lesions that usually develop tend to be localized and small and disappear or become inactive over time (Anon 2013).

After investigating TB in Sicilian black pigs, an endemic semi-free-ranging domestic pig breed, Di Marco et al. (2012) suggested that this species might act as a BTB reservoir in this ecosystem. The authors listed several facts to support their argument: that the pigs are infected, that the disease is active and self-maintaining in the pig population, and that conditions are favorable for black pigs to transmit M. bovis to cattle and humans. Furthermore, a spoligotype that was not previously described in the international spoligotyping database was isolated from pigs, suggesting that M. bovis with the potential of infecting cattle and humans circulates among pigs.

Warthogs (Phacochoerus africanus) may sustain the infection and similarly appear to be infected via the oral route and manifest lesions that are similar to those in domestic pigs (Woodford 1982).

7.3.2.2 Animal-Level Risk Factors

Age Although there appears to be some inconsistencies in the data, age is one of the most important animal-level risk factors identified in numerous studies in Africa (Berrada 1993; Katale et al. 2009; Ibrahim et al. 2012; Moiane et al. 2014). Animals of 6 years of age and older consistently had a higher prevalence (Vekemans et al. 1999; Tarnagda et al. 2014; Asseged et al. 2000; El-Olemy et al. 1985). When the average age of the herd was reduced and young stock was reared separately from older animals, the prevalence of BTB was much reduced (Awad 1962). It thus appears that the older the animal, the bigger the chances of being infected because of the prolonged and repetitive exposure to M. bovis that increases the likelihood of them becoming infected (de Vos et al. 2001; Cleaveland et al. 2007; Gumi et al. 2012).

A few reports exist where age was not significantly associated with BTB (Bedard et al. 1993). In a traditionally managed Moroccan herd, the number of tuberculin skin-test-positive animals did not differ between the age groups (Berrada 1993), and in Algeria, cattle aged between 2 and 5 years and goats the prevalence of BTB in older animals is significantly higher. These trends are consistent with those in Great Britain, where, on average, the incidence of BTB in cattle increases by 7.5% for every year of life, reaching 40% in the 5-6-year-old age group (O’Reilly and Daborn 1995). In intensive management systems, cows, heifers, and bullocks are, respectively, 10.3, 14.8, and 8.0 times more likely to fail a tuberculin test than calves (Griffin et al. 1996).

Studies involving other species revealed that age similarly is the most important risk factor. A logistical regression analysis (Rodwell et al. 2001) demonstrated that older African buffaloes were at greater risk for acquiring BTB than younger buffa­loes but that the risk increased for all age groups as the prevalence in a herd increased. Older camels also have a higher prevalence of TB than young animals (Elmossalami et al. 1971).

Sex Sex is one of a few animal-level risk factors that has only been mentioned in studies conducted in Africa (Humblet et al. 2009), but other countries are also now assessing this issue (Phillips et al. 2002). In Africa, the findings of various studies about the role of sex as a risk factor for BTB are inconsistent. More females than males were positive in countries such as Ethiopia (Elias et al. 2008), Uganda (Inangolet et al. 2008), Ghana (Bonsu et al. 2000), Chad (Ngandolo et al. 2009), Morocco (Berrada 1993), and Nigeria (Damina et al. 2011). In other studies, the opposite was true for cattle in Algeria (Sahraoui et al. 2011), Ethiopia (Tschopp et al. 2010), Nigeria (Ibrahim et al. 2012), and Tanzania (Tanner et al. 2014) where males were significantly more affected than females, while in a number of studies there were no statistically significant association between the prevalence of BTB and sex (Cook et al. 1996; Oloya et al. 2006; Awah-Ndukum et al. 2012; Okeke et al. 2014).

In beef cattle, similar inconsistencies exist. The majority of beef cattle are slaughtered young, giving them a much reduced lifetime exposure compared to dairy cattle, and the prevalence of the disease in them is usually low (Driscoll et al. 2011). In smallholder production systems, higher prevalence in males is seen. Although farmers usually take good care of their oxen, in these systems oxen shoulder the burden of farm activities because they are essential for ploughing, threshing, and harvesting (Tschopp et al. 2010). They are therefore kept in the herd for longer, thereby increasing the chance of being exposed to infection compared to females (Ibrahim et al. 2012). As sharing of oxen is commonplace in smallholder, crop-livestock production systems (Laval and Ameni 2004), there is also a greater probability that they will be in contact with an M. bovis-infected herd.

The inconsistencies in the prevalence based on sex appear to be the consequence of the management system followed rather than sex and the number of animals tested in each of the systems. The higher prevalence of BTB in dairy cows, also in pastoral settings, may predominantly be due to production stress during lactation and a longer life expectancy than that of males thus increasing the time that they are exposed to cattle with BTB. Their usual close confinement generally plays a bigger role in causing the higher prevalence of the disease in dairy cows (Francis 1950; Humblet et al. 2009).

Breed Initially, during the 1930s, the perceived lower and higher prevalence of BTB in various indigenous and exotic European breeds, respectively, were ascribed to an innate resistance of the African zebu breeds to BTB. The effect of this perceived resistance was probably enhanced by the extensive farming practices employed in Africa that tended to reduce the prevailing prevalence and spread of BTB in the indigenous breeds (Du Toit 1936; Carmichael 1938).

There also appeared to be differences in susceptibility between some of the indigenous breeds, such as the short-horned zebu and the long-horned Ankole cattle sharing the same environmental conditions and management practices. However, no observable differences in susceptibility to BTB between sub-Saharan White Fulani (3.3%), Sokoto Gudali (3.7), N’dama (3.6%), and Bunaji (3.3%) breeds were detected. Similarly, indigenous Moroccan cattle breeds were also thought to be less susceptible to BTB than the exotic dairy breeds, but they too were found to be equally susceptible (Bedard et al. 1993). In Tanzania, a high prevalence (13.2%) of BTB has been reported in indigenous cattle kept under intensive husbandry practices (Katale et al. 2009).

The perception of resistance may have been strengthened by the impact of the limited exposure of indigenous cattle to the African strains of M. bovis prior to colonialism (Brosch et al. 2002), because of the prevailing management practices at the time that limited contact between cattle of different owners and tribes and the spread of the disease. It is known that autochthonous cattle breeds in Africa (zebus, N’dama, and Mtui) have several important positive production traits, including resistance to certain diseases, but according to current knowledge, it is clear that resistance to BTB is not one of them.

The results of more recent studies suggest that the effect of breed on BTB in Ethiopia (and elsewhere in Africa) on the difference in prevalence of BTB in imported, high-value, European breeds compared to local zebus was markedly confounded by differences in herd management (Asseged et al. 2014), and that when these are taken into consideration, breed itself played a minor role in deter­mining the prevalence of BTB. There was, in addition, no significant difference in susceptibility between indigenous and exotic breeds after adjusting the data for age and sex (Berrada 1993; Asseged et al. 2000).

An example of the impact of management practices is seen in short-horned zebu and long-horned Ankole, which share the same environment and management conditions; however, Ankole cattle have a higher prevalence of tuberculin test positive reactors (54.9% compared to 4.6%) and gross lesions at slaughter (16% compared to 0.93%) (Carmichael 1937). Certain investigators reported no differ­ences between the indigenous breeds (Oloya et al. 2007), while others reported lesion prevalence in abattoirs in the Ankole longhorn cattle of 12.5-51.1% compared to the 0.1% in autochthonous zebus (Opuda-Asibo 1995). Various factors appear to influence the prevalence of BTB in the indigenous breeds. The variation of preva­lence in Ankole cattle exemplifies the concurrent influence of environmental and other risk factors. The high prevalence in Ankole cattle in Uganda was ascribed to a particular herd management style. In order to ward off biting insects, a smudge fire of dried cow dung is lit at the center of a kraal, and animals cluster around it, head-to- head, relishing the warmth and thus increasing the close contact between them. Under these circumstances, in infected herds, the coughing elicited by the smoke would facilitate the spread of BTB (Carmichael 1938). Elsewhere, Munroe et al. (1999) found no significant difference in risk based on breed between dairy and beef herds in BTB outbreaks in Canada. Francis (1950) concluded that “there is no reason to think that the various European breeds of cattle differ in their susceptibility to BTB.” In the final analysis, it is clear that genetic variation in susceptibility to M. bovis infection exists between mammalian families and species, but not in breeds (Phillips et al. 2002).

There is also a lack of association with breed or strains and susceptibility to BTB in other species. There was similarly a perception that the susceptibility of various human races differed, but it became clear that the differences are primarily deter­mined by the levels of disease in specific communities and levels of poverty prevailing in the different racial and ethnic groups (Opie 1931). No variation in susceptibility too was found in various mouse strains, and it was concluded that natural strain resistance simply does not exist (Gray et al. 1960).

A number of factors appear to influence the differences in prevalence that have been ascribed to the indigenous and exotic breeds, including the following:

1. The common practice, in Ethiopia, for example, to keep high-value Holstein cows in roofed cowsheds rather than in the open kraals (Elias et al. 2008).

2. The practice in the central highlands of Ethiopia for smallholders to keep dairy cows indoors, separately from other local zebus and crossbreds, which usually graze on communal pastures (Tschopp et al. 2013).

3. That genetically improved dairy animals are routinely kept under intensive conditions (Elias et al. 2008).

4. Different farms vary greatly in stocking density, extent of stock movement, the presence of M. bovis, and other risk factors (Driscoll et al. 2011).

Body Condition Score The body condition of cattle, assessed by using the body condition score (BCS), is another factor that has been used to subjectively assess the presence of BTB in cattle (Nicholson and Butterworth 1986). The scoring is based on visual inspection and palpation to assess the sharpness of the backbones and lumbar processes, and on occasion, the tail head, brisket, ribs, and hips, and the amount of muscle and fat covering these areas. The body condition score is typically categorized into poor (score 1-3), medium (score 4-6), and good (score 7-9) groups. The problem with using the body score as an assessment for the presence of BTB is that cattle in Africa often suffer from nutritional stress, evidenced by a low BCS, and that those with advanced clinical BTB characteristically also lose condition and manifest low BCSs.

It appears, based on available data, that BCS is not significantly associated with tuberculin test outcomes, and contradictory results have been recorded. In a study in Nigeria, the likelihood of detecting AFB in cattle increased as the BCS declined (Okeke et al. 2014). Similar results were obtained in Addis Ababa, Ethiopia (Elias et al. 2008), where, respectively, 36.1%, 30.9%, and 15.5% of 1869 of the cattle tested had poor, medium, and good BCSs. Similarly, in Zambia, a low BCS was associated with higher numbers of BTB-positive skin test cattle (Cook et al. 1996; Munyeme et al. 2008). The results using this parameter, however, are inconsistent as elsewhere a prevalence of 11.8%, 10.0%, and 13.1%, respectively, was recorded in cattle with poor, medium, and good scores (Asseged et al. 2000).

Immune Status Numerous factors impair the immune status of livestock, including poor hygienic conditions, nutritional deficiencies and imbalances, climatic stress, crowding, and poor ventilation, as well as those concomitant diseases that cause immune suppression (Berrada 1993). There appears to be no substantial evidence that the strain of milk production and calf bearing lower resistance to TB (Francis 1950; Ibrahim et al. 2012).

The rearing of cattle in Africa is mainly dependent on access to natural pastures, the availability of which varies seasonally thus causing a marked fluctuation in the nutritional status of animals during the dry periods of the year when they appear to become immunosuppressed (Ejeh et al. 2014). Malnutrition and nutritional stress are considered as important predisposing factors determining the susceptibility of cattle to BTB and are likely to play a significant role.

In certain African countries, co-infection with M. bovis and liver flukes appears to confound the interpretation of tuberculin test results, and it may also have an impact on the susceptibility of cattle to M. bovis (Munyeme et al. 2012) as the co-infection appears to influence the immune response to certain antigens. The extensive tissue damage caused by migrating metacercariae cause elevated immunoglobulin E levels, eosinophilia, and a T-lymphocyte-helper (Th) cell, type-2 immune response that has been shown to suppress the protective, M. bovis-specific, Th1 response (Phillips et al. 2002).

7.3.3 Environmental Factors

7.3.3.1 Geographic Distribution

It is generally accepted that the incidence of TB increased in industrialized countries in the seventeenth and eighteenth centuries, peaking at different times in different places. Looking at the trends, Lonnroth et al. (2009) suggested a temporal associa­tion between increased TB incidence and periods of rapid industrialization and urbanization. In a similar manner, many argue that BTB became widely dissemi­nated in countries where cattle were housed for long periods and in large herds. For instance, the British Economic Advisory Council in 1934 reported that at least 40% of the cows in dairy herds in Great Britain were infected with M. bovis (Alhaji 1976; Collins and Grange 1983). A prevalence of 50% also was reported in Germany in the mid-twentieth century (Pittler and Steel 1995). The progress made toward the eradication of BTB in the developed countries has changed its global distribution. Countries that at the turn of the twentieth century had the highest prevalence of BTB (United States, Canada, Western Europe, Australia, and Japan) now have the lowest prevalence, while African countries, which probably acquired the bulk of the infection during European colonization (Alhaji 1976; El Sanousi and Omer 1985; Phillips et al. 2002), are today the problem areas. Accordingly, prevalences of up to 50% in some areas in Malawi (Bernard et al. 2005), 25% in parts of Northern Nigeria (Alhaji 1976), 7.4% in Zambia (Cook et al. 1996), and 4.0% in Algeria (Sahraoui et al. 2009) have been reported.

Contrary to several opinions that BTB was absent in Africa before colonization (Du Toit 1936; Carmichael 1938), there is ample evidence, however, that BTB existed in the local cattle population before the introduction of European cattle breeds. Molecular typing of M. bovis isolates from Tanzania (Kazwala et al. 1998) showed two lineages of M. bovis: an aboriginal lineage with atypical properties and a lineage imported from Europe displaying a classical spoligotype profile. Muller et al. (2009) and Berg et al. (2011) later refined this observation. Accordingly, two clonal complexes Af1 and Af2 of M. bovis are geographically localized to Central-West and Eastern Africa. The East African clonal complex (Af2) is uniquely confined to Ethiopia, Kenya, Tanzania, and Uganda, although other non-Af2 strains were present at reasonably high frequencies (between 5% and 33%). The Central-West African clonal complex (Af1) is confined to Burkina Faso, Chad, Cameroon, and Nigeria. The introduction of improved European breeds of cattle and the advent of intensive dairying rapidly changed the distribution of BTB, and it is likely that some of the European cattle were infected with M. bovis European 1 clonal complex, which has a worldwide distribution, when imported, thereby amplifying the African burden of BTB (Smith 2012).

Mason (1912) paid particular attention to the origin of camels slaughtered in Cairo abattoir, Egypt. Although up to 40,000 camels were imported annually from Syria and Arabia, TB was found only in Egyptian camels and as far as it had been possible to ascertain, never in Syrian or Arabian camels.

7.3.3.2 Spatial Distribution

The pattern of BTB occurrence is not uniform: it can be classified as sporadic, persistent, or recurrent, while some herds may remain free of infection (Skuce et al. 2012). Biological and social mechanistic models have identified several risk factors, on a case-by-case basis, although some sporadic episodes offer few clues with respect to cause-and-effect relationships.

The prevalence of BTB varies widely in different areas within a country and in different herds within a given area. In South America (Cosivi et al. 1998), the highest prevalence of BTB was reported from areas surrounding major cities where intensive dairy production with large herd size was most common. In Spain, where about 97% of herds were officially free of BTB, bovine TB control programs that initially focused on dairy herds resulted in a dramatic reduction of the disease in that sector (Anon 2007a), and positive herds are now mainly located in the dry rural regions where there are more beef herds.

In the United States, 69% of the tuberculin reactors during 1917-1957 came from nine states: these nine states yielded 75% of the total reactors during 1955-1957 (Ranney 1958). In the United Kingdom, South West England and South West Wales have been consistently affected over the years (Anon 2007b). In New Zealand, which is divided into endemic and non-endemic areas based on difficulties in eradicating BTB, a total of 90% of tuberculous cattle comes from endemic areas (Collins et al. 1994). The common denominator in all of these reports is a localized, and possibly shared, external risk of the spread of BTB between neighboring herds, probably because of the presence of an additional maintenance host in the ecosystem (Johnston et al. 2011). In support of this argument, Costello et al. (1999) reported that isolates from cattle, badgers, and deer shared the same geographic range in Ireland, suggesting interspecies transmission. According to Skuce et al. (2012), historical incidence was a robust predictor of future breakdowns in UK and Irish herds, suggesting that the source was not totally removed. In intensive systems, this effect is expected to be less because of reduced exposure created by biological barriers to wildlife (Edwards et al. 1997).

The African Situation The relation between humans and their domestic animals in sub-Saharan Africa (SSA) is much closer than in the other regions. The human­animal interaction is growing in intensity due to the shift from extensive production system in the rural areas to intensified livestock husbandry in the peri-urban and urban centers, where even the most primitive of hygienic precautions are conspic­uously absent (Carmichael 1938). Under these circumstances, it is likely that BTB spreads rapidly once an M. bovis infection is established on a farm. Furthermore, in rural and small country towns across Africa, the home is shared with goats, sheep, fowls, and often cattle, posing a significant risk for zoonotic TB transmission. Therefore, large variations in BTB occurrence between regions have been reported in Africa.

The prevalence of BTB in Nigeria ranges from 0.5% in Oyo (low livestock population) to 12.3% in Gombe (northeastern state), where the livestock concentra­tion is high (Ejeh et al. 2014). In Ghana, BTB was seen to be highly prevalent (50%) in Ningo, a low-lying wetland (Bonsu et al. 2000), and in Tanzania, where the distribution of BTB is uneven, a significantly higher prevalence was seen in the southern highlands (14%) and southern regions (11.8%) compared to western (2.9%), northern (0.77%), lake (0.21%), and central (0.19%) regions (Daborn and Grange 1993). Consistent with the notion that the BTB prevalence is higher in humid and intensified conditions than in extensive, dryer areas, Bernard et al. (2005) reported a lower BTB prevalence in a pastoral area, which is dryer than the agropastoral zone where intensified conditions occurred. Much higher BTB preva­lences have been reported from peri-urban areas where intensified dairy production is practiced (Bonsu et al. 2000). In South Africa, where the incidence of infection was about 6%, a stark difference in prevalence was noted: it was about 40% in the larger towns, 10% in the small country towns, and 2% in rural areas (du Toit 1936).

Geographic variation in the prevalence of BTB within a country may suggest the existence of foci of M. bovis (or hotspots), such as communal pastures, watering points, and auction markets (Katale et al. 2012; Shirima et al. 2003). Since high infection rates were recorded in rangelands traversed by major rivers (Mwakapuja et al. 2013), areas with good pasture and water sources can attract more stock resulting in overcrowding (Brahmbhatt et al. 2012) as seen in the Mediterranean habitats of southern Spain (Acevedo et al. 2007). Humblet et al. (2009) propounded the view that water points are a potential risk factor, because areas around ponds are generally moist, with greater amounts of shade, two favorable conditions for the survival of M. bovis in the environment.

Munyeme et al. (2008) suggested that area could be a proxy variable for contact with wildlife. To give some insight, Munyeme et al. (2011) assessed whether black lechwe (Kobus leche smithemani) on the Bangweulu swamps, an area occupied by wildlife, were infected with TB. For comparison, Kafue lechwe (K. leche kafuensis) dwelling in the Kafue basin also were included in the study. The difference between the Kafue basin and the Bangweulu swamp was the absence of a livestock/wildlife interface in the latter. As expected, M. bovis was cultured from 4 of 11 Kafue lechwe, whereas none of the 30 black lechwe showed gross lesions suggestive of TB. The grazing range of Kafue lechwe and cattle extensively overlapped, particularly during the dry season, thus increasing the frequency of intermingling at watering points and on ranges. According to the authors, this scenario, which is considered a day-to-day phenomenon in all livestock/wildlife interface areas in Africa (Guilbride et al. 1963; Munyeme et al. 2008; Brahmbhatt et al. 2012; Katale et al. 2012), increases the opportunity for interspecies transmission of TB. Another perspective provided by the authors was the fact that early settlers from South Africa (Gallagher et al. 1972) may have introduced TB into the Kafue basin with their cattle when they settled there. This scenario is similar to the situation in Kruger National Park (KNP), South Africa, where it is hypothesized that comingling with M. bovis-infected cattle herds along the southern border of the Park introduced BTB into the buffalo population (Bengis et al. 1996; de Vos et al. 2001). A subsequent follow-up study based on the presence of identical molecular fingerprints (IS6110 and PGRS) of M. bovis isolates clearly established an epidemiological link between BTB outbreaks in the buffaloes of KNP and a neighboring cattle herd (Michel et al. 2008). Based on the disease timeline and its geographical expansion northward (from its entry point in the south), de Vos et al. (2001) calculated that BTB would spread at about 6 km/ year in a northerly direction in the KNP’s buffalo population.

7.3.3.3 Climatic Factors

Following fecal contamination of pastures, M. bovis withstands nutrient deprivation and osmotic shock; however, the weather will influence its survival. According to Phillips et al. (2003), exposure to direct sunlight destroys the bacilli within about 12 h. The survival time of M. bovis was investigated in the KNP (Tanner and Michel 1999). Tuberculous lungs and/or lymph nodes of African buffaloes (Syncerus caffer) and spiked fecal specimens placed at seven different sites in the habitat of free- ranging wildlife were analyzed over a 1-year period. Mycobacterium bovis could only be isolated for a period of 6 (for tissues) and 4 (for feces) weeks although the assessment was done during winter and under moist conditions. The survival time of M. bovis was greatly reduced (to a maximum of 5 days) when specimens were buried underground.

In the United Kingdom, with its markedly different climatic conditions, Duffield and Young (1985) previously failed to re-isolate M. bovis from fecal substrates at 4 weeks following artificial inoculation into bovine feces held under various envi­ronmental conditions. Maddock (1934) cited studies conducted in 1931 and 1932 where virulent M. bovis could be recovered from infected materials exposed in the open for 178 and 152 days, respectively. However, a follow-up study conducted during the hot summer of 1933 failed to recover tubercle bacilli from artificially infected grass 21 days after infection. Nevertheless, tubercle bacilli were shown to survive for 63 days after infection of the grass during the autumn and winter. Adams et al. (2013) experimentally inoculated environmental substrates (hay, soil, corn, water) and then exposed them to natural weather conditions in Michigan, USA. In 128 samples tested monthly during the 12-month period, M. bovis was not detectable by culture after 2 months although its DNA was still detectable by PCR for at least 7 months.

7.3.3.4 Herd Management Factors

In Africa, both commercial dairy farming with specialized dairy breeds, often of European origin, and traditional milk and beef production with extensively managed herds are widely practiced. The pattern of BTB prevalence in recorded surveys also reflects the prevailing livestock husbandry systems. According to Humblet et al. (2009), the management system will define the extent of contact between cattle, and between cattle and environmental sources. It is therefore important to assess the linkage between the prevailing livestock husbandry system and the occurrence of BTB.

From the point of view of BTB risk, African livestock production can be grouped into three broad categories (Katale et al. 2012; Unger and Munstermann 2004):

1. Pastoral (also called transhumance or nomadic)

2. Peri-urban dairy

3. Agropastoral, involving the integration of crop and livestock subsistence farming

In the following section, a detailed review of each system, as it relates to BTB risk, is provided.

Pastoral Production system Pastoral (transhumance or nomadic) production is an extensive, traditional, and common husbandry practice in many parts of Africa. It is practiced in every corner of the continent, usually forced by unevenly distributed precipitation with marked spatial and temporal variation in scarce feed and water resources. Herders and their livestock in Arid and Semi-Arid Land (ASAL) areas are forced to migrate periodically in search of surface water and green pasture, which are often shared by entire communities. Bovine TB seldom reaches a high incidence in cattle kept in the open; it is only when they are housed for intensive milk production that the incidence increases (Francis 1950). Extensive husbandry, with no housing as in pastoral production, presents a low risk for within-herd BTB transmission (Morris et al. 1994; Costello et al. 1998; Menzies and Neill 2000).

Both transhumant and nomadic systems of livestock production involve annual movement of livestock over variable distances, in some cases many hundreds of kilometers. Herders practicing transhumance have a regular and pre-defined move­ment pattern. The Fulani herdsmen, for example, migrate with their cattle to the southern parts of Nigeria during the rainy season and return north when the rain starts there (Opara 2005). In Zambia, cattle owners take their herds to the Kafue plains during the dryer months (May to October) and return to their villages during the rainy season (November to April) (Cook et al. 1996). Not infrequently, several discrete herds are rounded up at a given site (for safety) before the annual migration starts, forming a pool of herds, referred to as a “super herd.” Upward of eight large herds could be brought together for this purpose (Munyeme et al. 2008). This type of herd management involving “super herds” has also been described in Uganda (Oloya et al. 2007) and elsewhere (Ryan et al. 2006). A particular type of “super herd” was described in Botswana, where herds sharing the same range in the north congregate regularly (twice a month) at crush pens for health monitoring. In the process, cattle herds brought to the crush pens cohabit, share the same grazing spaces, and are managed under the same grazing strategy (Jori et al. 2013). Analogous to what has been described in African buffaloes (Michel and Bengis 2012), and dictated by seasonal availability of pasture and surface water, these alternating fission- and fusion-like events constitute a powerful driving force for pathogen dispersal among herds and across geographic niches as well as for bringing the source of infections and susceptible populations together.

In Northeastern Kenya, Southern Ethiopia, and parts of Somalia, two types of pastoral production are recognized: resident and migratory. The resident type incor­porates a permanent base where a few milking cows are kept to provide milk for part of the family of herders, including children, women, and the elderly. The cows usually graze far away from home in a communal grazing land, whereas calves graze around the homestead and are rounded up in open bush kraals at night. In the migratory type of production, a few herds belonging to close relatives are organized into a mobile herding group (analogous to a super herd). Migration distances, which may be in any direction, are determined by tribal interfaces and often do not exceed a 150-mile radius (Sheik-Mohamed and Velema 1999). During the wet season, when surface water and pasture are abundant, pastoralists disperse over large areas, while in the dry season, they tend to remain within a day’s reach of a water source. Individual herds are kept separate during the day (although several herds share ranges and water sources) and in separate bush kraals at night. Most of the cattle kept in this production setting are local zebus (Katale et al. 2012) or other indigenous breeds such as the N’dama (Unger and Munstermann 2004).

While it is true that animals provide milk for home consumption, draft power for cultivation, and are the main source of meat in many countries (Mwakapuja et al. 2013), livestock keeping in rural areas of SSA transcends purely economic reasons. In particular, cattle are an integral part of human social life, including the generation and accumulation of wealth and in maintaining a complex, deeply interwoven social system of mutual obligation (Michel et al. 2006). This is accomplished in part by the exchange of cattle within and between families and other social groups during traditional events such as marriage (dowries) and for other types of goods and services. As a result of this value system, animals are seldom sold or slaughtered (unless for funerals or other cultural ceremonies), and it is difficult to enforce culling, even for disease control purposes.

Bovine TB Risk Under extensive farming conditions, cattle live entirely in the open and are only rounded up at night in an open kraal made of thorny bushes, as a protection from theft and predators. In such areas, the prevalence of BTB is normally very low. For example, Paine and Martinaglia (1929) detected only 2 cases out of 17,263 cattle slaughtered at the Grahamstown, South Africa, abattoirs over 5.5 years. The authors listed several factors that might have led to the observed low BTB prevalence: (a) the cattle were mostly local grade stock, (b) they were not housed, and (c) the area was very lightly stocked. Carmichael (1938) similarly attributed the rarity of BTB in the tropics to the open-air existence of local livestock, whereas du Toit (1936) espoused the view that BTB is relatively unimportant in tropical Africa. However, a different perspective was brought to light when it was established that the long-horned Ankole breed were infected with BTB to an extent far in excess of any known in the worst infected regions of Europe (Carmichael 1938). According to the author, these animals constantly roamed over open range and were never housed; yet, 80% were tuberculin-test-positive, while abattoir inspection yielded a preva­lence of more than 40%.

To better understand the high prevalence of BTB among southern cattle in Sudan, Awad (1962) assessed the method of husbandry and herd management. During the wet season, in accordance with transhumant practice, herders drove their cattle to the highlands where they had built huts for them. The animals were crowded into the humid, unhygienic, and unventilated huts full of wood smoke to repel insects. The situation is exacerbated by the general practice to keep animals indefinitely, in most cases, until they die. Since younger and older animals were not separated, older tuberculous cows would transmit M. bovis to the younger stock. Daniel et al. (2009) similarly noted the transmission of M. bovis associated with introduction of animals reared in the adult female group.

A number of factors might result in the unexpected high prevalence of M. bovis infection in cattle in the extensive production sector. Animal movement, either through trading or communal grazing, including exposure to a wildlife maintenance host, was the most frequently reported variable determining herd-to-herd spread of BTB (Anon 1994; Johnston et al. 2011). The unrestricted movement of animals in the extensive husbandry system involves mixing of herds on communally owned ranges and around watering points during the dry season, leading to overcrowding. Studies conducted in Zambia provided compelling evidence in this regard. Cattle herds from the Monze district in Zambia move to the Kafue plains during the dry season in search of surface water and pasture. In the process, they mix with large populations of the Kafue lechwe, an endemic antelope and a maintenance host of BTB (Gallagher et al. 1972). Furthermore, cattle from different locations around the basin congregate on the plains, thus increasing the risk of contact with other herds (Cook et al. 1996). de Vos et al. (2001) suggested that a short period of such comingling is enough for M. bovis transmission. As a result, a high prevalence of BTB was recorded in cattle originating from the Monze area. Ryan et al. (2006) stated that movement of animals in and out of super herds is a plausible cause of the spread of BTB between herds. Consistent with this assumption, in Zambia, the herd­level BTB prevalence in transhumant herds (TH) was comparably higher than in the village-resident herds (Munyeme et al. 2008). In a similar study conducted in Tanzania, significantly more farms in the extensive sector had tuberculin reactors compared to farms in the intensive sector (Durnez et al. 2009). The authors con­cluded that because of the higher turnover of animals (buying, selling, gift, and mixing) in the extensive sector, animals were at a greater risk of exposure to and contracting the disease. Added to this is the problem of social unrest due to political instability and skirmishes along tribal interface areas, resulting in displacement of human and animal populations, and stock rustling.

In a nutshell, transhumance/nomadism and the consequent intermingling of animals are a way of pastoral life in many parts of Africa, and it is not likely to change any time soon. Even in the developed nations, in modern livestock produc­tion, the concept of discrete herds does not seem to be practiced as extensively as in the past. A study conducted in Australia found that only 10% of cattle herds were “closed” and that 13% of animals were moved between farms over a year (Ryan et al. 2006). In such open herds, enhanced BTB surveillance was necessary to reduce the number of infected herds (Barlow et al. 1998).

Surveillance for BTB in Africa is generally inadequate. For example, despite the reported high prevalence of BTB in Chad, Cameroon, and Nigeria, there are no facilities to quarantine and/or test animals for BTB before they cross national borders. Similarly, many African veterinary authorities at state or local levels do not actively apply measures for the control of BTB (Aliyu et al. 2009). As a consequence of this, two conditions for the continued persistence of the disease are being met: uncontrolled animal movement and poor disease surveillance. Hence, the stage is set for the further rapid dissemination of BTB throughout the pastoral systems of SSA, with the risk of creating a large number of BTB-positive herds, albeit with relatively low within-herd prevalences (Bernard et al. 2005; Gumi et al. 2012).

Intensive (Urban/Peri-urban) Dairy Production As part of a poverty reduction strategy, several African governments have been encouraging increased milk pro­duction through the introduction and/or expansion of peri-urban and other dairy development schemes. The impetus for transitioning to market-oriented production has been the rising consumer demand due to increasing population growth, urban­ization, and affluence (Unger and Munstermann 2004). Dairy production in Africa is largely characterized by unrestricted livestock keeping in towns and cities, and involves increasing intensification of large numbers of small production units. For instance, in Addis Ababa, Ethiopia, there were more than 5200 dairy farms in 1994, with an average herd size of 12 (Asseged et al. 2000). Importation of European breeds of cattle such as Holstein-Friesian was the centerpiece of these operations. Therefore, the majority of animals are exotic breeds or have a high proportion (>75%) of exotic blood. This was particularly evident in the peri-urban and urban areas (Shirima et al. 2003) owing to market access and availability of resources and services (Katale et al. 2012). Occasionally, peri-urban dairy cattle development projects in small towns may promote the use of milk and dairy products from indigenous cows (Bonsu et al. 2000). Commercial dairy farms tend to remain few in number (500 tested in central Ethiopia, the reactors were purchased from areas around Addis Ababa in the previous 5 years. Since BTB persists on a farm unless some mitigation is put in place (Marangon et al. 1998), cattle movement, without preliminary BTB testing and certification, could spread the infection from the infected farms to the newly established ones. Two recent studies in the United States suggested that the sources of infection in BTB in California (McCluskey et al.

2014) and Minnesota (Ribeiro-Lima et al. 2015) dairies were the purchase of BTB-infected animals and their movement onto new premises. Similarly, a compar­ative study of open (in which mixing with animals from other herds is possible) and closed (isolated) herds conducted in Addis Ababa (Asseged et al. 2000) corroborates these facts. According to the study, open herds were significantly more affected (63.4% BTB prevalence) compared to closed herds (36.8% BTB prevalence).

Global trade agreements can contribute to significantly increasing the risk of the inter-regional spread of BTB. For example, BTB in Danish cattle first appeared after tuberculous cows were imported from Germany (Magnus 1966). A recent assess­ment of BTB infections in the United States, in areas without a wildlife reservoir, identified a variety of risk factors, including the importation and comingling of steers from Mexico (Ribeiro-Lima et al. 2015). Similarly, it was postulated that cattle imported from France (in 1935) and the United Kingdom (in 1945) were responsible for the introduction of European strains of M. bovis into Iran (Tadayon et al. 2013). A review by Karolemeas et al. (2011) underscored the problem associated with trading of infected, TST-negative animals: this happened in 2008 when infected calves were exported from the United Kingdom into the Netherlands.

In summary, there is every reason to consider that the main cause of the high prevalence of BTB in dairy cattle is not their greater natural susceptibility to infection, but rather their crowded indoor existence, which increases their chances of exposure to infection. Francis (1950) emphasized the significance of inadequate ventilation and confining of livestock to the same house: “calves, which were housed with cows were exposed to a constant risk of infection by the aerogenous route, but when calves were not housed with cows, about 90% of them reached maturity without being infected, even if they were pastured with infected cows.” Opie (1931) reached a similar conclusion with regard to the transmission M. tuberculosis in humans: “the prevalence of TB in Jamaica was in large part referable to the crowded dwellings of the poorer people, where whole families, consisting of as many as six children, lived in small huts or yard rooms.” Therefore, although BTB occurs in both intensive and pastoral farming systems, the expanding dairy sector, along with intensive livestock management, will alter BTB dynamics in Africa in the future (Shirima et al. 2003).

To meet the rising demand for animal products in an increasingly growing population, there seems no better alternative than keeping improved stocks and changing the husbandry practices. There is, however, sufficient evidence to suggest that the application of broad principles of biosecurity will reduce the risk of cattle becoming infected by other animals, including wildlife (Skuce et al. 2012). Most of these emerging dairy producers are organized into associations (Asseged et al. 2000) or discrete bulk milk collection centers (Tschopp et al. 2013). These associations and centers coordinate requests for supplies and other services and also facilitate the acquisition of biotechnological inputs provided by livestock development agents. The silver lining is that the collected bulk milk is delivered to processing plants where it is pasteurized or processed at temperatures known to destroy mycobacteria. The risk of zoonotic TB through consumption of infected milk is therefore greatly reduced.

Smallholder Crop-Livestock Production System

Smallholder, crop-livestock production is a low input/low output production system that operates at the subsistence or semi-subsistence level, relying on family labor for growing crops. Livestock are kept largely as a secondary activity to support work on crop fields, with multiple other objectives including sales, inputs to agriculture (traction and manure), and subsistence (milk and milk products). Herd sizes are characteristically small (up to 15 head of cattle), and animals consist mainly of autochthonous breeds, such as the N’dama and short-horned zebu, which are typically adapted to inferior feeding and health regimens and to tropical climatic stressors that place serious limitations on their productivity levels. The animals are reared around the homestead and fed on natural pasture, cut grass, farm leftovers, and other by-products; only selected animals are nutritionally supplemented during the dry season (Unger and Munstermann 2004). Since infrastructures, such as road networks, are poorly developed, severely restricting access to market and other farm inputs, smallholder farmers have no incentives to invest in innovative production technologies. As of late, smallholders have been increasingly responding to new challenges such as an increasing demand for milk and other livestock products, by moving to more intensified and integrated mixed farming systems (Bonsu et al. 2000). As the availability of land is the main constraint, comingling of individual herds on communally owned grazing land is a common practice. Therefore, live­stock owned by several owners living in the same village or neighborhood consti­tutes an epidemiological unit (Moiane et al. 2014).

Moiane et al. (2014) suggested that a high prevalence of BTB was observed in almost all livestock areas where small-scale farming was practiced. Bonsu et al. (2000) examined 400 heifers and 747 cows in the Dagme District of Ghana. The prevalence of BTB in heifers was 9.8% and 16.7% in cows. Similarly, Bernard et al. (2005) conducted a BTB survey in the dairy-producing areas of Mbarara District, Uganda. In total, 252 of 340 herds had at least 1 BTB-positive cow. Furthermore, the BTB prevalence was higher in herds (79.3%) and animals (8.3%) from the agropastoral zone, compared to the pastoral areas. In Tanzania, 5.7% (n = 105) of smallholder dairy farms contained tuberculin reactors compared to none from traditionally managed herds (Swai and Schoonman 2012). According to Moiane et al. (2014), trading of animals among smallholders is frequently performed without regard to BTB status. The management system in the smallholder sector, which involves sharing water source and grazing areas, and mixing of animals from different herds, increases the likelihood of the spread of BTB. A study conducted in the central Ethiopian highlands indicated that 94% of owners kept their animals, predominantly zebu and low-grade crosses, on communal grazing land (Tschopp et al. 2013). Additionally, during vaccination campaigns and other animal health service activities such as deworming and tick control, animals from different herds use the same dip tanks, thereby comingling in holding areas. Cattle regularly attending a dip tank had a higher prevalence of BTB than cattle in bulking groups who more commonly practiced on-farm disease control measures (Bedard et al. 1993).

7.4