The Epidemiology of Mycobacterium bovis Infection in Cattle 53
24 months of clearing restrictions (Karolemeas
et al., 2011). Heuristically, recurrence can be
explained either by the presence of infection
missed by testing during the original breakdown
or the re-introduction of infection through cat-
tle movements, residual infection within the
farm’s environment or cross-species transmis-
sion from wildlife.
We used herd-level models with a structure
similar to our conceptual model (see Fig. 4.1) to
estimate the extent to which these patterns of
persistence were consistent with re-introduction
of infection or a hidden burden of infection
missed by testing (Conlan et al., 2012). Our
results suggested that at most 21% (12–33%) of
breakdowns harbour a median of one infected
animal (95% CI: 1–4) and at most 50% (95% CI:
33–67) of recurrent breakdowns could be
explained by infection missed by testing.
4.3.3 Between-herd transmission of
bovine tuberculosis and the role of
cryptic infection in wildlifeThe introduction of statutory reporting of cattle
movements in the wake of the bovine spongi-
form encephalopathy epidemic has provided
exquisite cattle movement data for the UK
(Mitchell et al., 2005), Ireland and other Euro-
pean countries. In the UK, the so-called cattle
tracing system has allowed sophisticated net-
work analyses to be carried out that attempt to
quantify the contribution of cattle movements
to between-herd transmission.
Second only to a past history of infection,
cattle movements from high-risk herds are per-
haps the greatest, and most consistent, risk fac-
tor for bovine TB breakdowns outperforming
environmental and other geographic factors
(Gilbert et al., 2005). The importance of cattle
movements to transmission was most dramati-
cally demonstrated by the geographic expansion
of disease that followed re-stocking of farms
after the 2001 Foot and Mouth Epidemic
( Carrique-Mas et al., 2008). However, the com-
bination of insensitive cattle tests and slow
transmission means that the relative contribu-
tion of environmental sources and cattle move-
ments to between-herd transmission of bovine
tuberculosis is surprisingly hard to quantify. The
situation in Scotland is perhaps the exception to
this, where all, or nearly all, recent breakdowns
have been attributed to cattle movement (Bessell
et al., 2013), even though at least one was
detected late after substantial within-herd trans-
mission that cattle movement allowed to be
seeded into several other herds.
The low prevalence of infection in Scotland
limits the utility in estimating rates of between-
herd transmission. In order to progress, national
level network models have had to make relatively
strong assumptions about the role of wildlife in
transmission that colour their findings. Green
et al. (2008) developed a relatively simple frame-
work for weighting the contribution of cattle
movements to other sources where the risk of
infection from environmental sources was mod-
elled as a constant risk.
This pragmatic assumption, necessary due
to our lack of systematic surveillance of wildlife
populations, neglects the feedback loop that
exists when a disease is freely transmitted
between two host populations. Amid all of the
controversy over the role of badgers, and culling
efforts on the epidemiology of M. bovis in cattle,
what is clear is that the disease can be trans-
mitted between both species and interference
with one host affects the prevalence of disease in
the other (Godfray et al., 2013). This is most
clearly demonstrated by the close geographical
association in the genotypes of M. bovis isolated
from cattle and badgers (Goodchild et al., 2012),
but the relative rates of transmission between
populations are not identifiable given available
methods and data, with considerable uncer-
tainty in estimates even with strong assump-
tions about directionality (Donnelly and
Nouvellet, 2013).
The importance of considering the full
dynamics of this two-host ecological system
was highlighted by Brooks-Pollock and Wood
(2015), who demonstrated how even relatively
low levels of between-species infection can allow
infection to persist in one host, be it cattle or
badger, even when the inherent within-species
reproductive ratio is below the threshold value
of 1. In this situation, which on the balance of
evidence is likely in Great Britain, marginal
gains from intervention measures could be suffi-
cient to tip the two-host system below threshold- particularly if the measures target inter-
specific spread.