66 R.A. Skuce et al.
geographical ‘home range’, often confirmed by
cattle movement tracings. Due to the striking
biogeography and geographical localization of
molecular types, isolates have a genetic signa-
ture characteristic of their geographic origin.
This is now an important, consistent and exploit-
able finding in several countries (Smith et al.,
2003; Skuce et al., 2010; Robbe-Austerman and
Turcotte, 2014). When deployed in the outbreak
investigation setting, and in conjunction with
cattle movement databases and wildlife surveil-
lance, this is a powerful means of investigating
source, maintenance and spread of bovine TB.
Depending on their movement history, infected
cattle either ‘qualified’ or were ‘excluded’ from
particular outbreaks or clusters (Skuce et al.,
2010). Where M. bovis molecular types were
recovered at some distance from even a coarse
definition of their normal home range, this dem-
onstrated a role for cattle movement in dispersal
of bovine TB. Hence, it will be important to
investigate the extent to which recorded local
and longer-range cattle movements and cattle
social (contact/movement) networks could
explain the observed geographical localization
of M. bovis molecular types. TB molecular types
were also found outside their normal home
range in home-bred cattle, indicating previous
contact and transmission involving cattle from a
distant home range.
The existence of home ranges for molecular
types from molecular surveillance provides for a
distinction between ‘purchased’ infection and
infection ‘acquired on arrival’ in most cases, i.e.
local versus non-local source(s). In Northern
Ireland, significant numbers of specific MLVA
types were found outside their normal home
range. Many such animals had ear tags that
linked them directly with their proposed MLVA
type home range. The remainder may have had
more complex links (secondary, tertiary, etc.) to
the home range (Skuce and others, unpublished
data).
Whether geographical localization of
M. bovis molecular types in cattle reflects the
underlying spatial segregation of the disease in
wildlife remains to be determined in Northern
Ireland. Considering the detailed livestock, land-
scape and wildlife data sources that are now
being collected in various ecological studies,
there is an opportunity to investigate the extent
to which the geographical localization of
M. bovis molecular types in cattle reflects vari-
ables such as cattle contact and movement net-
works, landscape features, badger genetic
structure, etc. (Biek and Real, 2010).
Several local studies confirm that the num-
ber of reactors by herd is highly skewed and con-
sequently a relatively small number of herds
provide a disproportionate number of reactors.
In Northern Ireland, herd- and animal-level
molecular surveillance has identified >300
M. bovis MLVA types in >50,000 isolates (Skuce
et al., 2010; Trewby et al., 2016). Herd-level
MLVA surveillance disclosed on average 73
MLVA types each year, with 29 MLVA types pres-
ent in all years sampled, implying that a rela-
tively small number of MLVA types provides a
disproportionate number of culture-confirmed
cases. In human TB epidemiology, such data are
consistent with the presence of the ‘super-
spreader’ phenotype (Ypma et al., 2013), TB
cases that contribute disproportionately to
transmission. For M. bovis in Northern Ireland,
such ‘super-spreaders’ could reside at herd and
animal levels. For example, certain herds may
act as key nodes and may be highly connected in
transmission networks underpinned by animal
movement through trade. For M. bovis, such
data are also consistent with the clonal expan-
sion of molecular types identified in British data
(Smith et al., 2003).
The most parsimonious explanation for
observing the same, or very similar, genotypes in
large multi-reactor herds would tend to be that
extensive cattle–cattle transmission (amplifica-
tion, regardless of initial source) is occurring.
Smaller herds with multiple reactors tended to
contain multiple MLVA types, the most plausible
explanation being multiple introductions from
several external sources. However, it remains
important to index and understand micro-
evolutionary events within-herd (Navarro et al.,
2016). M. bovis MLVA types were observed to
expand and contract, at least in their frequency
(Skuce et al., 2010). Hence, the local population
seems to have remained largely, but not entirely
static, which probably reflects the expansion and
contraction of molecular types in different loca-
tions (Smith et al., 2003) in response to the effi-
cacy of local control measures, the transmission
dynamics of the various genotypes and other
undefined factors, such as potential competition
between types and constraints imposed by host