Age-specific infectious period shapes 
dynamics of pneumonia in bighorn sheep
Authors: Raina K. Plowright, Kezia R. Manlove, Thomas E. 
Besser, David J. Paez, Kimberly R. Andrews, Patrick E. 
Matthews, Lisette P. Waits, Peter J. Hudson, and Frances 
E. Cassirer
This is the peer reviewed version of the following article: see full citation below, which has been 
published in final form at https://dx.doi.org/10.1111/ele.12829. This article may be used for non-
commercial purposes in accordance with Wiley Terms and Conditions for Self-Archiving.
Plowright, Raina K., Kezia R. Manlove, Thomas E. Besser, David J. Paez, Kimberly R. Andrews, 
Patrick E. Matthews, Lisette P. Waits, Peter J. Hudson, and Frances E. Cassirer. "Age-specific 
infectious period shapes dynamics of pneumonia in bighorn sheep." Ecology Letters 20, no. 10 
(September 2017): 1325-1336. DOI: 10.1111/ele.12829.
Made available through Montana State University’s ScholarWorks 
scholarworks.montana.edu 
Age-specific infectious period shapes dynamics of pneumonia
in bighorn sheep
Raina K. Plowright,1*
Kezia R. Manlove,2
Thomas E. Besser,2 David J. P
aez,1 
Kimberly R. Andrews,3
Patrick E. Matthews,4
Lisette P. Waits,3 Peter J. Hudson5 
and E. Frances Cassirer6
Abstract
Superspreading, the phenomenon where a small proportion of individuals contribute dispropor-
tionately to new infections, has profound effects on disease dynamics. Superspreading can arise
through variation in contacts, infectiousness or infectious periods. The latter has received little
attention, yet it drives the dynamics of many diseases of critical public health, livestock health
and conservation concern. Here, we present rare evidence of variation in infectious periods under-
lying a superspreading phenomenon in a free-ranging wildlife system. We detected persistent infec-
tions of Mycoplasma ovipneumoniae, the primary causative agent of pneumonia in bighorn sheep
(Ovis canadensis), in a small number of older individuals that were homozygous at an immunolog-
ically relevant genetic locus. Interactions among age-structure, genetic composition and infectious
periods may drive feedbacks in disease dynamics that determine the magnitude of population
response to infection. Accordingly, variation in initial conditions may explain divergent popula-
tion responses to infection that range from recovery to catastrophic decline and extirpation.
INTRODUCTION
A small proportion of the host population can be responsible
for the majority of pathogen transmission (Woolhouse et al.
1997; Lloyd-Smith et al. 2005; Buhnerkempe et al. 2017).
These individuals produce a disproportionate number of
secondary infections by engaging in behaviour that increases
transmission, shedding more pathogen per unit time, or
remaining infected for longer than the average host (Lloyd-
Smith et al. 2005; Streicker et al. 2013; VanderWaal & Ezenwa
2016). The consequences of superspreading were exemplified
by the pandemics of severe acute respiratory syndrome (SARS)
and human immunodeficiency virus (HIV) that were spread
worldwide by a small number of highly infectious or connected
individuals (May & Anderson 1987; Lloyd-Smith et al. 2005).
Since those landmark pandemics, the dynamic consequences
of variation in infectiousness and social contact have been
studied in detail (e.g., Ferrari et al. 2006; Bansal et al. 2007;
Chase-Topping et al. 2008). By contrast, relatively little atten-
tion has focused on variation in infectious periods among
individuals and the dynamic consequences of this heterogeneity
(Woolhouse et al. 1997; Hawley & Altizer 2011).
This knowledge gap is particularly relevant to management
of many common pathogens of humans and animals that are
persistently carried by some hosts. Persistently infected
individuals help to maintain Salmonella, Brucella, Staphylo-
coccus, Streptococcus, Mycobacterium, Haemophilus and Lep-
tospira species, Hepatitis B and C virus, and Epstein–Barr
virus within host populations (Medley et al. 2001; Young
et al. 2002; Virgin et al. 2009; Prager et al. 2013; Buhn-
erkempe et al. 2017). Yet, for many if not most of these
pathogens, the mechanisms driving persistence and prevalence
are not understood (Wertheim et al. 2005). Such understand-
ing comes from resampling individuals over time, but longitu-
dinal empirical studies are expensive, difficult and rare
(VanderWaal & Ezenwa 2016). Moreover, inferences derived
from cross-sectional studies can mislead control efforts. For
example, longitudinal studies of Escherichia coli O157 in cat-
tle, thought to be maintained by persistently infected super-
shedders, revealed that the pathogen may be serially shed by
many individuals – in other words, different individuals take
turns at supershedding (Spencer et al. 2015). If supershedding
is not a stable individual-host characteristic, then targeted
interventions based on specific individuals will fail.
Here, we investigate heterogeneity in infectious periods
among hosts, and mechanisms that may give rise to hetero-
geneity, in a system where the pathogen is limiting host popu-
lations: Mycoplasma ovipneumoniae in bighorn sheep (Ovis
canadensis) (Fig. 1). M. ovipneumoniae is the primary causal
agent of bacterial pneumonia in bighorn sheep (Besser et al.
1Department of Microbiology and Immunology, Montana State University,
109 Lewis Hall, Bozeman, MT 59717, USA
2Department of Veterinary Microbiology and Pathology, Washington State
University, Pullman, WA 99164, USA
3Department of Fish and Wildlife Sciences, University of Idaho, 875 
Perimeter Drive MS 1136, Moscow, ID 83844, USA
4Oregon Department of Fish and Wildlife, 65495 Alder Slope Road, Enter-
prise, OR 97828, USA
5Center for Infectious Disease Dynamics, 201, Life Sciences Building,
Pennsylvania State University, University Park, PA 16802, USA
6Idaho Department of Fish and Game, 3316 16th Street, Lewiston, ID 
83501, USA
2008, 2012a,b, 2013, 2014). This pathogen, likely introduced
with domestic sheep during European settlement, contributed
to the historic decline of bighorn sheep to less than 20 000
individuals across the United States (Buechner 1960). Domes-
tic sheep tolerate M. ovipneumoniae in their upper respiratory
mucosa with varying morbidity but generally low mortality
(Alley et al. 1999; Besser et al. 2012a). Introduction of M.
ovipneumoniae into na€ıve bighorn sheep populations can result
in severe mortality ranging from 10 to 100% (Cassirer et al.
2013, 2017). Animals that survive acute infection and asso-
ciated disease can clear infection or can remain subclinically
infected in the upper respiratory tract. If previously exposed
animals are reinfected, they are relatively immune to lower
respiratory tract disease but can develop a subclinical upper
respiratory tract infection (Cassirer et al. 2013; Plowright
et al. 2013). Infected populations often remain stagnant, or
decline – sometimes to extirpation – in the years to decades
after M. ovipneumoniae introduction because most lambs die
of pneumonia by 3 months of age (Manlove et al. 2016). Sus-
tained lack of recruitment is the primary impediment to big-
horn sheep recovery.
We evaluated the support for two competing hypotheses
about population-level persistence of M. ovipneumoniae in big-
horn sheep. First, M. ovipneumoniae could be maintained by
persistent infections within a few specific individuals. Second,
M. ovipneumoniae could be maintained through serial
(sequential) infections transmitted among many individuals.
For the first hypothesis, the prediction is that infection status
across sampling events would be positively correlated within
individuals. In the second, the prediction is that an animal’s
infection status during one test would be independent of its
status during previous tests, provided the intervals between
testing are longer than the infectious period and that acquired
immunity does not affect infection status. Moreover, the
probability that an animal tested positive would reflect current
population-level prevalence. Additionally, we investigated
individual characteristics that might drive persistent carriage
including age, genetic diversity and coinfection with lung-
worms. Finally, we predicted that lambs born to carrier ewes
would be more likely to die than lambs from non-carriers
because of early transmission directly from mother to off-
spring. To examine these hypotheses, we repeatedly sampled
free-ranging bighorn sheep in the Lostine population (Wal-
lowa Mountains, Eastern Oregon, USA) over a 4-year period.
METHODS
Study population
The Lostine bighorn sheep comprise a single, well-mixed pop-
ulation that has been surveyed annually since it was estab-
lished in 1971 (see Supporting Information; Coggins 1988;
Figure 1 Epidemiology of Mycoplasma ovipneumoniae invasion and persistence in bighorn sheep populations.
1326 R. K. Plowright et al. Letter
Fig. 2). In 1986, a pneumonia epidemic killed two-thirds of
the population; however, the population rebounded to 92%
of its pre-epidemic size within 12 years (Coggins 1988; Cog-
gins & Matthews 1992; Cassirer et al. 2013). In 2000, a year
before a novel M. ovipneumoniae strain was detected in the
population (Cassirer et al. 2016), monitoring was expanded to
include survival and productivity of individuals tracked with
radio collars (Cassirer & Sinclair 2007; Cassirer et al. 2013).
Population estimates during this study ranged from 75 to 90
sheep (lambs and adults; ODFW, unpublished data; Fig. 2b).
Sampling methods and diagnostic testing
Bighorn sheep in the Lostine population migrate seasonally
between low and high elevations (Coggins 1988) and are
habituated to humans through supplemental feeding on their
winter range. We used a baited corral trap and ground dart-
ing to capture animals on their winter range (Supporting
Information). We placed unique markers on each animal and
recaptured individuals within and between seasons with a bias
towards sampling females because they transmit pathogens to
lambs. We monitored 41 radio-collared animals during the
study (34 females, 7 males) of which seven had been collared
prior to the study.
We collected nasal samples with dacron swabs and placed
these samples into mycoplasma broth (Hardy Diagnostics,
Santa Maria, CA) for culture enrichment or into sterile tubes
(dry swabs). We conducted PCR and RT-PCR (realtime
PCR) detection of M. ovipneumoniae with the methods
described in McAuliffe et al. (2003) and Ziegler et al. (2014),
respectively. Both tests were used to ascribe infection status to
a sample, and RT-PCR was also used to estimate relative
pathogen load. We collected blood for detection of anti-M.
ovipneumoniae antibodies by competitive enzyme-linked
immunosorbent assay (ELISA) on serum (Besser et al. 2014).
We collected faecal samples and measured lungworm larvae
intensity with the Baermann technique (Forrester & Lankester
1997). We also scored body condition by palpating the sacral
crest (Cook et al. 2007). Diagnostic testing was conducted at
the Washington Animal Disease Diagnostic Laboratory and
in the Besser laboratory at Washington State University.
Data analyses
Persistent carriage analysis
We used a suite of non-parametric resampling-based methods to
evaluate our hypotheses (Supporting Information). We com-
pared the observed frequency of changes in animal infection sta-
tus through time to that expected by chance. We repeatedly
randomly permuted infection status for all individuals across all
capture events. We calculated the average duration of an ani-
mal’s infection status for each permutation and used those aver-
ages to build a null distribution reflecting the average time
infected or not infected under the serial transmission hypothesis.
We then compared the quantile of the null distribution to the
observed average duration in each infection state to obtain a P-
value for departure from the serial transmission hypothesis.
Moreover, we considered multiple tests of the same animal
assuming a Hardy–Weinberg-like equilibrium (Wigginton
et al. 2005). We considered a situation in which two tests of
1970 1980 1990 2000 2010
0
20
40
60
80
100
120
140
Year
N
um
be
r o
f b
ig
ho
rn
 s
he
ep
178
5
9
20
10
10
1516
22
18
15
15
6
3
2
Infected or suspected infected
Suspected healthy
Lambs recruited
No health status data
01 02 03 04 05
Kilometers
Lostine population
Other Hells Canyon populations
(a) (b)
Figure 2 (a) Map of the Lostine Population. (b) Population trajectory and lamb recruitment of the Lostine population from 1970 to 2014. Red dots
indicate that Mycoplasma ovipneumoniae was detected by polymerase chain reaction in one or more individuals, or M. ovipneumoniae infection was
suspected because pneumonic animals were observed, or juvenile survival through summer was less than 50%. Black dots indicate that recruitment was
high and we suspect that M. ovipneumoniae was not transmitted to lambs. Open circles indicate that no health data were available. Digits indicate the
number of animals that were translocated out of the population and the length of the arrow corresponds to this number.
the same individual were treated as analogous to two alleles,
and we assumed the transmission and recovery processes were
sufficiently rapid that consecutive test results for the same ani-
mal were independent (Supporting Information). Departures
from an equilibrium pattern suggest that infection concen-
trates in a few hosts.
We created a frequency distribution of the proportion of pos-
itive test results for all animals tested three or more times. Each
individual contributed one proportion to this distribution, and
proportions were weighted according to the number of times
animals were sampled, such that individuals with more samples
received greater weight. Under the serial transmission hypothe-
sis, the distribution would be unimodal and centred at the pop-
ulation prevalence; under the persistent carrier hypothesis it
would be bimodal with individuals repeatedly testing either
positive or negative. We fit a kernel density estimator over the
range of this distribution (from 0 to 1), identified a natural
breakpoint at the proportion corresponding to the minimum of
the kernel density estimator and generated bootstrapped confi-
dence intervals for the breakpoint (Supporting Information).
To explore the consistency of an individual’s infection status
across years, we estimated the probability that an animal had
a positive test result in year t + 1 given two or more positive
test results in year t. We repeated this analysis for animals
that always tested negative in the first year. We used binomial
confidence intervals to quantify uncertainty in our estimates.
To account for the temporal pattern of infection status, we
based our classification of individual animals as persistent car-
riers, intermittent carriers or negatives on the following defini-
tion: persistent carriers were individuals that tested positive in
two or more consecutive sampling events across at least
2 years (even if they tested as negative at other times) and
intermittent carriers tested both positive and negative in con-
secutive years. Animals that were only sampled once were not
assigned a carriage status.
To explore whether anti-M. ovipneumoniae antibody levels
and pathogen load (measured in RT-PCR cycle thresholds
[Cts] on DNA extracted directly from nasal swab samples)
differed among carriage classifications, we used Welch’s two-
sample t-test.
Age-specific infection status
We assigned ages and birth years to bighorn sheep younger
than 4 years old at initial capture (n = 67) on the basis of
incisor eruption (Dimmick & Pelton 1994). We assigned the
ages of five animals 4 years or older when first captured on
the basis of cementum analysis of incisors obtained at
necropsy (Turner 1977; Mattson Laboratory, Milltown, MT).
For other animals that were 4 years of age or older at first
capture, definitive age estimation was not possible (n = 20).
We used a Monte Carlo simulation to estimate their ages. We
defined a window of possible ages for each animal such that
the minimum possible age at first capture was four and the
maximum possible age at final capture was 20 (Nussey et al.
2008). Each time the animal was recaptured, we narrowed the
lower and upper bound on this window. We drew the individ-
ual’s estimated age at capture from this window with proba-
bilities proportionate to age-specific survival estimates from
Jorgenson et al. (1997) (Supporting Information).
We repeated the age estimation process for animals with
unknown ages 10 000 times and concatenated these ages to a
fixed dataset of animals with known ages. At each iteration,
we fit a mixed-effects logistic regression model (Bates et al.
2014) on the presence or absence of M. ovipneumoniae. The
linear and quadratic effects of sheep age were included as
fixed effects, whereas the animal’s identity was modelled as a
random effect. We regressed infection status on median age
estimates, and used Akaike’s information criteria (AIC) to
compare models with linear and quadratic age. We also evalu-
ated infection in cohorts of animals reconstructed from longi-
tudinal demographic data (Supporting Information).
Genetic diversity
We extracted DNA from blood samples and then PCR-ampli-
fied and genotyped 15 autosomal DNA microsatellite loci and
one sex chromosome locus for sex identification (Supporting
Information and Table S1). Eleven of the autosomal loci were
expected to be adaptively neutral given their position in the
genome. The remaining four autosomal loci were located
within (or close to) genes associated with immune system
function (Supporting Information) and were putatively under
disease-related adaptive pressures (e.g. Paterson et al. 1998b;
Coltman et al. 2001; Luikart et al. 2008). We calculated neu-
tral genetic diversity for each individual as the proportion of
the 11 neutral loci that were heterozygous, and used Student’s
t-tests to evaluate whether carrier classes differed in the mean
proportion of heterozygous genotypes across loci. For each
putatively adaptive locus, we used Fisher’s exact tests to eval-
uate whether the proportion of heterozygous individuals
across carrier classes differed. We also incorporated heterozy-
gosity at adaptive loci into the logistic regression model
described above with each locus, age, and quadratic age
included as covariates. We used AIC to compare models with
and without heterozygosity (Supporting Information).
Parentage, lamb survival and lamb infection status
In 2012–2015, we assessed survival of offspring by observing
marked ewes and lambs monthly in their remote habitat
through summer. We classified a female as having a lamb if
she was observed nursing a lamb, alone with a lamb or if she
was pregnant the previous winter. We assumed she had lost
her lamb if she was observed in a group with no lambs over
several consecutive observations. In 2014 and 2015, we added
genetic maternity analyses on blood and tissue samples from
marked ewes and lambs and faecal samples from unmarked
lambs, yearlings and ewes.
For all samples, we PCR-amplified the same 15 autosomal
microsatellite and sex chromosome loci as described above,
but with extra (3–6) independent PCR amplifications for fae-
cal samples to account for potential allele dropout (Waits &
Paetkau 2005; Supporting Information). Duplicate samples
from the same individual were genetically identified and
removed as described in the Supporting Information.
We identified potential mothers as females ≥ 2 years at the
birth of the individual of interest. We used CERVUS 3.0.7
(Marshall et al. 1998; Kalinowski et al. 2007) to conduct like-
lihood assessment of maternity. We estimated that 90% of
potential mothers in the population had been sampled, and
required an 80% confidence level to assign maternities. We
submitted serum for pregnancy testing (Drew et al. 2001) if a
female sampled over the winter was never seen with a lamb in
spring to determine whether she lost her lamb or was not
pregnant.
We used a generalised mixed model with binomial errors
and a logit link function to estimate whether offspring sur-
vival differed as a function of ewe carrier classification, or the
results of the two closest tests preceding the lamb’s birth (both
positive, one positive or both negative). We included infection
status and linear and quadratic ewe age as covariates and ewe
identity as a random effect. We used AIC-based approaches
for model selection.
Lungworms
Lungworm larvae counts were overdispersed and zeroinflated,
hence we applied negative binomial models to these data with
age, carriage status, infection status and heterozygosity at four
adaptive loci as fixed effects. We also fit models with random
effects for individual (Supporting Information).
RESULTS
Infection status
From April 2012 through March 2016, we captured and sam-
pled 86 animals for a total of 227 samples. Forty percent
(N = 34) of animals were sampled two to nine times
(Table 1).
All adults, including infected individuals, were in good body
condition throughout the study and most animals appeared
healthy, although we occasionally saw signs of respiratory
tract disease and associated weakness. Lamb-ewe ratios in late
winter (when lambs are 
 9 months old), and adult female
survival, averaged 0.43 (0.37–0.50) and 0.93 (0.9–1), respec-
tively (Supporting Information).
M. ovipneumoniae was detected in 47% of adult females
(95% bootstrapped confidence interval [35%, 65%]), 44% of
yearlings (11%, 78%) and 89% of lambs (78%, 100%)
(Fig. 3a). Antibodies to M. ovipneumoniae (at ≥ 50% inhibi-
tion) were detected at least once in 95% (41/43) of adult
females, 85% of yearlings (17/20) and 97% of lambs (31/32).
Antibody levels from animals that tested PCR-positive were
significantly higher than those from animals that tested
negative on PCR on the same sampling event (Kruskal–Wallis
test, P < 0.001, N = 114 PCR positives and 111 PCR nega-
tives; Fig. S1). Antibody levels and pathogen loads in inter-
mittent carriers (when infected) did not differ significantly
from those in infected persistent carriers (Wilcoxon rank-sum
W = 534.5, P = 0.13; Wilcoxon rank-sum W = 32, P = 0.96,
respectively, Supporting Information).
Persistent carriage analysis
During our 4-year sampling period, the longest observation of
consistently positive and negative test results from an individ-
ual animal were 3.2 and 3.1 years, respectively. However, the
true duration of the same infection status could be longer; 44
of 50 infection events detected in adults in this study were
either interval or right-censored (Fig. 4a).
Both the permutation analysis and the Hardy–Weinberg
equilibrium analysis suggested that the average duration of
positive or negative states was greater than expected by
chance (H-W p2 = 0.14; q2 = 0.39; 2pq = 0.47; P-value <
0.001; observed average exceeded all 5000 permutations for
average duration in either state and for duration of the nega-
tive state; observed average exceeded all but 2 of 5000 permu-
tations for the infected state; Fig. 3e–g; Fig. S2).
The kernel density approach suggested that animals sepa-
rated into two groups when the proportion of positive tests
was above or below 65.7% (95% bootstrapped CI for the sep-
aration threshold = [38%, 85%]) (Fig. 3c). Eight of 11 (73%)
animals that tested positive twice in year t and were sampled
at least once in year t + 1 tested positive at least once in t + 1
(95% CI from 39 to 93%). Seven of 10 animals that tested
negative at least twice in year t (70%), did not test positive in
year t + 1 (95% CI from 35 to 92%).
The final classifications of individuals that were sampled
more than once are presented in Fig. 4a. Thirty-seven percent
of animals were consistently positive over two or more years
and were classified as persistent carriers. Forty-five percent of
animals were both positive and negative, and were classified
as intermittent carriers. The remaining 18% of animals had
no PCR-positive test results, and were classified as negative
(Figs 3b and 4a).
Age-specific infection status
We found a significant, U-shaped quadratic relation between
age and M. ovipneumoniae prevalence such that prevalence of
infection was higher in relatively young (< 2 years) and old
(> 15 years) individuals than in intermediate-aged animals
(Fig. 5a). The AIC of a model that included both quadratic
and linear terms was less than that of a model that included
only a linear term (DAIC = 32; Table S2). Results from the
cohort assignment ageing method were similar (Fig. 5b,
Fig. S3).
Genetic diversity and infection status
Diversity at a putatively adaptive locus differed across carrier
classes, whereas neutral genetic diversity did not (Fig. S4). A
locus within the Major Histocompatibility I gene complex
Table 1 Number of bighorn sheep sampled between 2012 and 2016, and
number of times sampled, by sex and age*
Number of times sampled 1 2 3 4 5 6 7 8 9 Total
Total individuals 52 6 3 6 5 11 2 0 1 86
Adult ewes 12 5 3 6 5 11 2 0 1 45
Adult rams 5 1 0 0 0 0 0 0 0 6
Yearlings* 14 2 0 0 0 0 0 0 0 16
Lambs* 31 1 0 0 0 0 0 0 0 32
Total samples 62 18 9 24 25 66 14 0 9 227
*Animals 2 years or older are classified as adults. Three lambs and six
yearlings were sampled again as adults, and are included in the table
twice. Two animals were sampled as lambs, yearlings and adults and are
included in the table three times.
(MHCI) had lower heterozygosity for persistent carriers than
intermittent carriers (P = 0.005) or negative animals (P = 0.04)
(Fig. 3d). However, diversity at the other three putatively
adaptive loci was not correlated with carrier status. In a model
with mean age, but no random effect for individual identity
(mixed models did not converge), animals that were
0.0 0.2 0.4 0.6 0.8 1.0
0
0.25
0.50
0.75
1.00
D
ensity
0
2
4
6
8
10
Fr
eq
ue
nc
y
Density over all tested animals
Animals with 3 or more tests
Proportion positive
Pr
ev
al
en
ce
1.0
0.8
0.6
0.4
0.2
0.0
Adult ewes Yearlings Lambs Persistent Intermittent Negative
Pr
op
or
tio
n 
of
 a
du
lt 
ew
es
1.0
0.8
0.6
0.4
0.2
0.0
+ in consec. years
+ twice in one year
Mean length of time
same status
Fr
eq
ue
nc
y
100 300 500
0
500
1000
1500
2000
2500
Bootstrapped values
Observed value
Mean length of time
PCR-negative
Fr
eq
ue
nc
y
100 300 500
0
500
1000
1500
Mean length of time
PCR-positive
Fr
eq
ue
nc
y
100 300 500
0
500
1000
1500
(e) (f) (g)
(c)
(a) (b)
Bootstrapped values
Observed value
Bootstrapped values
Observed value
Persistent Intermittent Negative
Fr
eq
ue
nc
y
Homozygous
Heterozygous
(d)
0
2
4
6
8
10
heterozygous at MHCI were less likely to be infected than ani-
mals that were homozygous at this locus (P = 0.016); age and
quadratic age were both significant (P < 0.0001; Table S3).
However, models excluding and including all four adaptive loci
had equivalent AIC scores (Supporting Information).
Parentage, lamb survival and lamb infection status
After removing duplicate samples, we identified 15 individuals
from 97 faecal samples with final consensus genotypes.
Genetic maternity analyses assigned 36 of 37 lambs to a dam,
with 25 assignments at the 95% confidence level and 11
assignments at the 80% confidence level.
A greater proportion of lambs died from ewes classified as
persistent carriers compared to lambs from ewes classified as
intermittent carriers or negative (Fig. S5). However, this result
was not statistically significant (P = 0.45). When defining the
carrier status of ewes based on the two tests prior to lambing,
the difference in the survival of offspring of persistent and
intermittent carrier ewes vs. negative ewes was non-significant
(P = 0.09) (Fig. S5). Ewe age (linear or quadratic) was not
associated with offspring survival (Supporting Information).
0 1 3 4
Age (years)
An
im
al
s 
sa
m
pl
ed
 a
s 
juv
en
ile
s
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
(a) (b)
An
im
al
s 
sa
m
pl
ed
 o
nl
y 
as
 a
du
lts
Persistent carriers
Interm
ittent carriers
N
egatives
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
2013
Biological year Years since first positive
22014 2015 2016 –3 –2 –1 0 1 2 3
Figure 4 (a) Infection status of adult bighorn sheep over time. Red dots represent a positive test, black dots represent a negative test, based on PCR. Three
classes were detected for animals captured two or more times: persistent carriers, intermittent carriers and negative. The star denotes an animal that was
classified in part on samples (and PCR tests) collected prior to this study (Animal 13). All animals except 61, 41 and 6 were adult females. Biological year
is 1 May to 31 April. (b) Infection status of bighorn sheep first sampled as juveniles (lambs or yearlings).
Figure 3 (a) Mean prevalence (with 95% confidence intervals) of Mycoplasma ovipneumoniae shedding in adult females, yearlings and lambs. (b) Proportion
of adult ewes characterised as persistent carriers, intermittent carriers or negative, given that M. ovipneumoniae was detected in consecutive years (light
grey) or was detected at least twice in 1 year (dark grey). (c) The proportion of positive tests for 30 adult females tested three or more times in the Lostine
bighorn sheep population, 2012–2016. The density curve is weighted by the number of tests. (d) Distribution of homozygosity and heterozygosity at a locus
within the Major Histocompatibility I gene complex among three classes of M. ovipneumoniae carriage. (e, f, g) The expected and observed distributions of
consecutive positive and negative tests. The average duration of both infected and uninfected states was greater than expected by chance by both a
permutation analysis (e, f, g) and an equilibrium analysis (Fig. S2).
Of the 12 lambs or yearlings sampled over multiple years,
four were negative and remained negative the following year;
eight were positive and six of these lambs or yearlings became
negative within 2 years (Fig. 4b).
Both ageing and carriage status (but not genetics, or infec-
tion status) competed to explain variation in lungworm counts,
however, evidence of confounding made it difficult to establish
causal links. Age probably underlies both carriage and lung-
worm status (Table S4, Fig. S6, Supporting Information).
DISCUSSION
Mycoplasma ovipneumoniae prevalence was highest in rela-
tively young and old animals. However, the majority of
infected young animals recovered within 1 to 2 years, whereas
the majority of infected old animals remained infected for the
duration of the study. The presence of older persistent carriers
supports the hypothesis that M. ovipneumoniae is maintained
by persistent infections within a few individuals, although, a
high proportion of animals were intermittently infected and
their contribution to pathogen persistence at the population
level is unknown. If intermittent carriers sustain serial trans-
mission without reinfection from persistent carriers, or if
intermittent carriers are persistently infected with low patho-
gen loads that are only intermittently detected, removal of
both intermittent and persistent carriers may be necessary to
eliminate M. ovipneumoniae from populations.
Age-related responses to microparasitic and macroparasitic
infections are well documented and are usually attributed to
dynamic interactions among exposure, parasite-induced host
mortality, parasite accumulation, acquired immunity, and age-
related variation in immune function (Hudson & Dobson
1995; Wilson et al. 2002). The early peak in pathogen preva-
lence in bighorn sheep lambs may reflect early exposure of
lambs to infected ewes, lack of protection by maternal immu-
nity, and intense transmission among lambs concentrated in
nursery groups. The high prevalence in older animals may be
influenced by seasonal re-exposure during epidemics of pneu-
monia among lambs. However, given that evidence of expo-
sure (by antibody detection) was also high in middle-aged
animals that tested negative, low resistance to infection in
young and old animals seems a more parsimonious explana-
tion for the U-shaped relations between age and infection in
bighorn sheep.
High risk of infection or mortality in the young and the old
has been attributed to immune system immaturity and senes-
cence, respectively (Lloyd 1995; Vallejo 2007). For example,
children have the highest rates of infection with the persistent,
upper respiratory tract pathogens Staphylococcus aureus and
Kingella kingae; then, as the immune system matures, children
either clear the pathogens or become intermittently infected
(Wertheim et al. 2005; Yagupsky 2015). Higher risk in older
animals is often attributed to immunosenescence and waning
of immune memory. As individuals age, telomere lengths in
immune cells erode, triggering replicative senescence of those
cells (Vallejo 2007). The result is reduced activation, memory
and responsiveness of the adaptive and innate immune sys-
tems (Tarazona et al. 2002; Solana et al. 2006). Declines in
immunity as age increases explain increasing intensities of gas-
trointestinal helminths in Soay sheep (Ovis aries) (Hayward
et al. 2009) and progression of bovine tuberculosis in badgers
(Meles meles) (Beirne et al. 2016). Moreover, as exemplified
by avian malaria (Asghar et al. 2015), long-term persistent
infections can exacerbate telomere degradation and
immunosenescence.
A combination of the mechanisms outlined above may
explain the U-shaped pattern observed in this study. Similar
combinations of mechanisms were suggested to explain the U-
shaped relation between age and prevalence of human papillo-
mavirus (HPV) infection and between age and mortality from
influenza and varicella-zoster virus (Castle et al. 2005; Heinin-
ger & Seward 2006; Greer et al. 2010). U-shaped age-infection
curves are infrequently reported; age-intensity or age-preva-
lence curves that are convex or rise to an asymptote are more
Age (years)
Pr
op
or
tio
n 
po
sit
ive
0 2 4 6 8 10 12 14 16 18 20
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Most likely age
80
60
40
20
0
0 5 10 15
Positive test result
Negative test result
(a)
(b)
In
di
vid
ua
l s
he
ep
Figure 5 Relations between Mycoplasma ovipneumoniae prevalence and
bighorn sheep age. (a) Error bars represent 95% confidence intervals. The
solid line is the fitted result from a generalised mixed-effects model with a
logit link function and binomial errors, whereas the dotted lines are the
95% confidence intervals estimated by bootstrapping. (b) Age estimated
by assigning animals to cohorts on the basis of sampling history, age-
specific ewe survival estimates and annual recruitment data.
common and are a consequence of immunity or parasite-
induced mortality. Examples include lungworm larvae burdens
observed in this study (Table S4) and Trichostrongylus retor-
taeformis in wild European rabbits (Oryctolagus cuniculus)
(Cornell et al. 2008).
Despite many studies of age-varying risk of infection, few
studies report age-varying duration of infection, probably
because resampling individuals with known ages is seldom
possible. However, there are rare exceptions. In humans, the
probability of persistent Hepatitis B infection decreased with
age (Ott et al. 2012), whereas, in some populations, the prob-
ability of persistent HPV infection increased with age (Castle
et al. 2005). In cattle, the probability of persistent foot-and-
mouth disease virus infection decreased with age (Bronsvoort
et al. 2016), however in wild rodents, no association was
observed between the duration of cowpox infection and age
(Hazel et al. 2000).
The interaction between age-structured carriage and lack of
recruitment may explain why bighorn sheep population
responses to M. ovipneumoniae infection range from recovery
to catastrophic decline and extirpation. One population-level
consequence of low recruitment is a shift towards an older
age structure (Plowright et al. 2013; Manlove et al. 2016). If
older animals are more likely to carry M. ovipneumoniae, then
positive feedbacks may determine the prevalence in a popula-
tion and the probability of the population’s long-term persis-
tence. For example, increasing mean age may lead to
increasing prevalence of M. ovipneumoniae, increasing proba-
bility and severity of lamb pneumonia outbreaks and acceler-
ating long-term population declines. Conversely, decreasing
mean age may facilitate pathogen extinction and population
recovery. Therefore, the age-structure during and immediately
after pathogen-invasion could be important in determining the
long-term population response to infection. Moreover, inter-
ventions to remove carriers may be most efficiently conducted
after the initial outbreak and before the mean age increases as
a consequence of lack of recruitment. These dynamic aspects
of carriage are relatively unexplored. Nevertheless, disentan-
gling these dynamic interactions is a challenging and necessary
step towards understanding control options and long-term
sequelae of infection.
In addition to age, genetic diversity at a locus within the
MHCI gene complex was lower in persistent carriers than in
negative or intermittent carriers. These results remained signif-
icant in a model including age. However, there was no rela-
tion between neutral genetic diversity and carrier status,
suggesting that an individual’s genotype at certain loci may
influence infection status, whereas an individual’s relative level
of inbreeding does not. MHC class I alleles usually respond
to intracellular pathogens and the mechanisms by which they
would affect the presumably extracellular M. ovipneumoniae
are unknown. However, MHC class I and II genes were
linked in Soay sheep; if also true for this population, selective
pressures at any MHC gene could generate genetic signatures
across all MHC genes, including class II genes that respond
to extracellular pathogens (Paterson 1998). MHC alleles in
Soay and Scottish Blackface sheep were associated with sur-
vival and resistance to macroparasites (Schwaiger et al. 1995;
Buitkamp et al. 1996; Paterson et al. 1998a).
The abundance of MHCI homozygous genotypes in older
animals (Fig. S7) may reflect recent introduction of new geno-
types through immigration, and may confound the age-car-
riage relations. Alternatively, homozygous and heterozygous
individuals may have a survival advantage, or disadvantage,
respectively. Counterintuitively, homozygotes could live longer
than heterozygotes if homozygotes are tolerant (maintaining
health despite their pathogen load) and heterozygotes are
resistant but at a cost (e.g. energetic cost of immunity or
immunopathology) (Raberg et al. 2009). In theory, selection
for tolerant homozygotes could lead to a stable association
between host and pathogen (Roy & Kirchner 2000). However,
for bighorn sheep, the phenomenon could be maladaptive
because lambs would be susceptible to infections from tolerant
adults. Thus, relations among genetic diversity, survival, toler-
ance and resistance, could drive another feedback loop that
results in population decline. More data are needed to exam-
ine this hypothesis.
We found little evidence that a lamb’s survival is directly
related to its mother’s infection status. Instead, the offspring
of both carriers and non-carriers died, probably because
lamb-to-lamb transmission dominates when lambs are concen-
trated in nursery groups, concealing the origin of the epidemic
(Cassirer et al. 2013; Plowright et al. 2013; Manlove et al.
2014). Some offspring of persistent carriers survived, despite
their putative early exposure, and offspring of non-carriers
were exposed to other infected lambs or ewes. Lambs and
adults that were PCR negative had serological evidence of
exposure to M. ovipneumoniae, indicating that their status was
not due to lack of exposure.
Priorities for future research
Our work highlights some critical knowledge gaps in under-
standing the population dynamics of persistently carried
pathogens. First, there is a dearth of age-structured longitudi-
nal data on infections in natural systems. Without data on
age, potential mechanisms and feedbacks driving disease
dynamics may be overlooked or misinterpreted (Medley et al.
2001). For example, one might mistakenly conclude that M.
ovipneumoniae status drives lungworm burdens, whereas age
probably drives both lungworm burden and M. ovipneumoniae
status.
Second, there is little understanding of the role of intermit-
tent carriers in maintaining and transmitting infections (Plo-
wright et al. 2016). Experiments in the field and on captive
animals, supported by modelling studies, may be necessary to
clarify the role of intermittent carriers. Ultimately, replicated
field experiments that monitor the outcome of removal of per-
sistent or intermittent carriers may be needed to elucidate
mechanisms of persistence at the population level.
Third, there is no robust statistical definition of persistent
carriage, even for common human pathogens such as S. aur-
eus (Wertheim et al. 2005). Our definition of persistent car-
riage within an individual as two consecutive years of
infection was subjective given the constraints imposed by cen-
soring, lack of data and limited knowledge of the relevant
mechanisms. Longer studies are needed to empirically estab-
lish this definition in this system and others. Ideally, a
definition informed by the number of secondary infections
produced by individuals, as was developed for superspreaders
(Lloyd-Smith et al. 2005), could be developed for persistent
carriage. However, quantifying transmission of mucosal bacte-
rial pathogens remains very challenging.
For bighorn sheep, we suggest that replicate studies be con-
ducted in other locations to determine if the Lostine popula-
tion – with its winter aggregation and particular strain of
M. ovipneumoniae – is representative of other infected bighorn
sheep populations. In addition, longer time series would allow
stronger inference about age-prevalence relations and explo-
ration of alternative explanations such as cohort effects (Sup-
porting Information). The discovery of age-structured
persistent carriage and variation in diversity at an MHC locus
provides a basis for simulations and field experiments that
explore options for eradication of pneumonia in bighorn
sheep populations.
ACKNOWLEDGEMENTS
Thanks to Vic Coggins for his years of dedicated bighorn
sheep management that made this study possible. Thanks to
Holly E. Tuers-Lance and Chad Dotson for their assistance in
the field and Nathan Justice, Alan Swanson, Jordan Schup-
bach and Robyn Egloff for guidance on analyses or figures.
Funding was provided by Morris Animal Foundation grant
D13ZO-081, National Institutes of Health IDeA Program
grants P20GM103474 and P30GM110732, P. Thye, Montana
University System Research Initiative: 51040-MUSRI2015-03,
Oregon Chapter of the Foundation for North American Wild
Sheep, Idaho Department of Fish and Game, Washington
Department of Fish and Wildlife, Oregon Department of Fish
and Wildlife, Federal Aid to Wildlife Restoration, and
University of Idaho. KRM was supported on a Pennsylvania
State University Academic Computing Fellowship.
AUTHORSHIP
RKP and EFC designed the study; PEM, EFC and RKP sup-
ported field data collection; TEB contributed to diagnostic
testing; KRA and LPW performed genetic analyses; KM,
RKP, DP and EFC analysed data; RKP, EFC and KM wrote
the first draft of the manuscript and all authors contributed
to revisions.
DATA ACCESSIBILITY STATEMENT
Data are available from the Dryad Digital Repository: http://
dx.doi.org/10.5061/dryad.s86sd.
REFERENCES
Alley, M., Ionas, G. & Clarke, J. (1999). Chronic non-progressive
pneumonia of sheep in New Zealand–a review of the role of
Mycoplasma ovipneumoniae. N. Z. Vet. J., 47, 155–160.
Asghar, M., Hasselquist, D., Hansson, B., Zehtindjiev, P., Westerdahl, H.
& Bensch, S. (2015). Hidden costs of infection: chronic malaria
accelerates telomere degradation and senescence in wild birds. Science,
347, 436–438.
Bansal, S., Grenfell, B.T. & Meyers, L.A. (2007). When individual
behaviour matters: homogeneous and network models in epidemiology.
J. R. Soc. Interface, 4, 879–891.
Bates, D., M€achler, M., Bolker, B. & Walker, S. (2014). Fitting linear
mixed-effects models using lme4. arXiv preprint arXiv:1406.5823.
Beirne, C., Waring, L., McDonald, R.A., Delahay, R. & Young, A.
(2016). Age-related declines in immune response in a wild mammal are
unrelated to immune cell telomere length. Proc. R. Soc. London B:
Biol. Sci., 283.
Besser, T.E., Cassirer, E.F., Potter, K.A., VanderSchalie, J., Fischer, A.,
Knowles, D.P. et al. (2008). Association of Mycoplasma ovipneumoniae
infection with population-limiting respiratory disease in free-ranging
Rocky Mountain bighorn sheep (Ovis canadensis canadensis). J. Clin.
Microbiol., 46, 423–430.
Besser, T.E., Cassirer, E.F., Yamada, C., Potter, K.A., Herndon, C.N.,
Foreyt, W.J. et al. (2012a). Survival of bighorn sheep (Ovis canadensis)
commingled with domestic sheep (Ovis aries) in the absence of
Mycoplasma ovipneumoniae. J. Wildl. Dis., 48, 168–172.
Besser, T.E., Highland, M.A., Baker, K., Cassirer, E.F., Anderson, N.J.,
Ramsey, J.M. et al. (2012b). Causes of pneumonia epizootics among
Bighorn Sheep, Western United States, 2008–2010. Emerg. Infect. Dis.,
18, 406–414.
Besser, T.E., Frances Cassirer, E., Highland, M.A., Wolff, P., Justice-
Allen, A., Mansfield, K. et al. (2013). Bighorn sheep pneumonia:
sorting out the cause of a polymicrobial disease. Prev. Vet. Med., 108,
85–93.
Besser, T.E., Cassirer, E.F., Potter, K.A., Lahmers, K., Oaks, J.L.,
Shanthalingam, S. et al. (2014). Epizootic pneumonia of bighorn sheep
following experimental exposure to Mycoplasma ovipneumoniae. PLoS
ONE, 9, e110039.
Bronsvoort, B.M.d., Handel, I.G., Nfon, C.K., Sørensen, K.-J., Malirat,
V., Bergmann, I. et al. (2016). Redefining the ‘carrier’ state for foot-
and-mouth disease from the dynamics of virus persistence in
endemically affected cattle populations. Sci. Rep., 6, 29059.
Buechner, H.K. (1960). The bighorn sheep in the United States, its past,
present, and future. Wildl. Monogr., 4, 3–174.
Buhnerkempe, M.G., Prager, K.C., Strelioff, C.C., Greig, D.J., Laake,
J.L., Melin, S.R., et al. (2017). Detecting signals of chronic shedding to
explain pathogen persistence: Leptospira interrogans in California sea
lions. J Anim Ecol., 86, 460–472.
Buitkamp, J., Filmether, P., Stear, M.J. & Epplen, J.T. (1996). Class I
and class II major histocompatibility complex alleles are associated
with faecal egg counts following natural, predominantly Ostertagia
circumcincta infection. Parasitol. Res., 82, 693–696.
Cassirer, E.F. & Sinclair, A.R.E. (2007). Dynamics of pneumonia in a
bighorn sheep metapopulation. J. Wildl. Manage., 71, 1080–1088.
Cassirer, E.F., Plowright, R.K., Manlove, K.R., Cross, P.C., Dobson,
A.P., Potter, K.A. et al. (2013). Spatio-temporal dynamics of
pneumonia in bighorn sheep. J. Anim. Ecol., 82, 518–528.
Cassirer, E.F., Manlove, K.R., Plowright, R.K. & Besser, T.E. (2016).
Evidence for strain-specific immunity to pneumonia in bighorn sheep.
J. Wildl. Manage., 81, 133–143.
Cassirer, E. F., Manlove, K. R., Almberg, E. S., Kamath, P. L., Cox, M.,
Wolff, P., et al. (2017), Pneumonia in bighorn sheep: Risk and
resilience. J. Wild. Manage. https://doi.org/10.1002/jwmg.21309
Castle, P.E., Schiffman, M., Herrero, R., Hildesheim, A., Rodriguez,
A.C., Bratti, M.C. et al. (2005). A prospective study of age trends in
cervical human papillomavirus acquisition and persistence in
Guanacaste, Costa Rica. J. Infect. Dis., 191, 1808–1816.
Chase-Topping, M., Gally, D., Low, C., Matthews, L. & Woolhouse, M.
(2008). Super-shedding and the link between human infection and
livestock carriage of Escherichia coli O157. Nat. Rev. Microbiol., 6,
904–912.
Coggins, V.L. 1988. The Lostine Rocky Mountain bighorn sheep die-off
and domestic sheep. In Proceedings of the 6th Biennial Symposium of
the Northern Wild Sheep and Goat Council (ed Baniff W. M.). Banff,
Alta., 11–15 April 1988. pp.57–64.
Coggins, V. L. & Matthews, P. E. (1992). Lamb survival and herd status
of the Lostine bighorn herd following a Pasteurella die-off. In
Proceedings of the Biennial Symposium of the Northern Wild Sheep and
Goat Council, (eds Emmerich J., Hepworth, W.G.). Cody, WY, 8:147–
154.
Coltman, D., Wilson, K., Pilkington, J., Stear, M. & Pemberton, J.
(2001). A microsatellite polymorphism in the gamma interferon gene is
associated with resistance to gastrointestinal nematodes in a naturally-
parasitized population of Soay sheep. Parasitology, 122, 571–582.
Cook, R.C., Stephenson, T.R., Myers, W.L., Cook, J.G. & Shipley, L.A.
(2007). Validating predictive models of nutritional condition for mule
deer. J. Wildl. Manage., 71, 1934–1943.
Cornell, S.J., Bjornstad, O.N., Cattadori, I.M., Boag, B. & Hudson, P.J.
(2008). Seasonality, cohort-dependence and the development of
immunity in a natural host–nematode system. Proc. Biol. Sci., 275,
511–518.
Dimmick, R.W. & Pelton, M.R. (1994). Criteria of sex and age. In:
Research and Management Techniques for Wildlife and Habitats, 5th
edn. The Wildlife Society, Bethesda, MA, pp. 169–214.
Drew, M.L., Bleich, V.C., Torres, S.G. & Sasser, R.G. (2001). Early
pregnancy detection in mountain sheep using a pregnancy-specific
protein B assay. Wildl. Soc. Bull., 29, 1182–1185.
Ferrari, M.J., Bansal, S., Meyers, L.A. & Bjørnstad, O.N. (2006).
Network frailty and the geometry of herd immunity. Proc. Biol. Sci.,
273, 2743–2748.
Forrester, S.G. & Lankester, M.W. (1997). Extracting Protostrongylus
spp. larvae from bighorn sheep feces. J. Wildl. Dis., 33, 868–872.
Greer, A., Tuite, A. & Fisman, D. (2010). Age, influenza pandemics and
disease dynamics. Epidemiol. Infect., 138, 1542–1549.
Hawley, D. & Altizer, S.M. (2011). Disease ecology meets ecological
immunology: understanding the links between organismal immunity
and infection dynamics in natural populations. Funct. Ecol., 25, 48–60.
Hayward, A.D., Wilson, A.J., Pilkington, J.G., Pemberton, J.M. &
Kruuk, L.E. (2009). Ageing in a variable habitat: environmental stress
affects senescence in parasite resistance in St Kilda Soay sheep. Proc.
Biol. Sci., 276, 3477–3485.
Hazel, S., Bennett, M., Chantrey, J., Bown, K., Cavanagh, R., Jones,
T. et al. (2000). A longitudinal study of an endemic disease in its
wildlife reservoir: cowpox and wild rodents. Epidemiol. Infect., 124,
551–562.
Heininger, U. & Seward, J.F. (2006). Varicella. Lancet, 368, 1365–1376.
Hudson, P. & Dobson, A. (1995). Macroparasites: observed patterns in
naturally fluctuating animal populations. In: Ecology of Infectious
Diseases in Natural Populations (eds. Grenfell, B.T. and Dobson, A.P.).
Cambridge University Press, Cambridge, pp. 144–176.
Jorgenson, J.T., Festa-Bianchet, M., Gaillard, J.-M. & Wishart, W.D.
(1997). Effects of age, sex, disease, and density on survival of bighorn
sheep. Ecology, 78, 1019–1032.
Kalinowski, S.T., Taper, M.L. & Marshall, T.C. (2007). Revising how the
computer program CERVUS accommodates genotyping error increases
success in paternity assignment. Mol. Ecol., 16, 1099–1106.
Lloyd, S. (1995). Environmental Influences on Host Immunity. Ecology of
Infectious Diseases in Natural Pop-ulations. Cambridge University
Press, Cambridge, UK, pp. 327–361.
Lloyd-Smith, J.O., Schreiber, S.J., Kopp, P.E. & Getz, W.M. (2005).
Superspreading and the effect of individual variation on disease
emergence. Nature, 438, 355–359.
Luikart, G., Pilgrim, K., Visty, J., Ezenwa, V.O. & Schwartz, M.K.
(2008). Candidate gene microsatellite variation is associated with
parasitism in wild bighorn sheep. Biol. Let., 4, 228–231.
Manlove, K.R., Cassirer, E.F., Cross, P.C., Plowright, R.K. & Hudson,
P.J. (2014). Costs and benefits of group living with disease: a case study
of pneumonia in bighorn lambs (Ovis canadensis). Proc. Biol. Sci., 281,
20142331.
Manlove, K., Cassirer, E.F., Cross, P.C., Plowright, R.K. & Hudson, P.J.
(2016). Disease introduction is associated with a phase transition in
bighorn sheep demographics. Ecology, 97, 2593–2602.
Marshall, T., Slate, J., Kruuk, L. & Pemberton, J. (1998). Statistical
confidence for likelihood-based paternity inference in natural
populations. Mol. Ecol., 7, 639–655.
May, R.M. & Anderson, R.M. (1987). COMMENTARY~ Transmission
dynamics of HIV infection. Nature, 326, 137.
McAuliffe, L., Hatchell, F. & Ayling, R. (2003). Detection
ofMycoplasma. Vet. Rec., 153, 687–688.
Medley, G.F., Lindop, N.A., Edmunds, W.J. & Nokes, D.J. (2001).
Hepatitis-B virus endemicity: heterogeneity, catastrophic dynamics and
control. Nat. Med., 7, 619–624.
Nussey, D., Coulson, T., Festa-Bianchet, M. & Gaillard, J.M. (2008).
Measuring senescence in wild animal populations: towards a
longitudinal approach. Funct. Ecol., 22, 393–406.
Ott, J., Stevens, G., Groeger, J. & Wiersma, S. (2012). Global
epidemiology of hepatitis B virus infection: new estimates of age-
specific HBsAg seroprevalence and endemicity. Vaccine, 30, 2212–
2219.
Paterson, S. (1998). Evidence for balancing selection at the major
histocompatibility complex in a free-living ruminant. J. Hered., 89,
289–294.
Paterson, S., Wilson, K. & Pemberton, J. (1998a). Major
histocompatibility complex variation associated with juvenile survival
and parasite resistance in a large unmanaged ungulate population (Ovis
aries L.). Proc. Natl Acad. Sci., 95, 3714–3719.
Paterson, S., Wilson, K. & Pemberton, J.M. (1998b). Major
histocompatibility complex variation associated with juvenile survival
and parasite resistance in a large unmanaged ungulate population (Ovis
aries L.). Proc. Natl Acad. Sci. USA, 95, 3714–3719.
Plowright, R.K., Manlove, K., Cassirer, E.F., Cross, P.C., Besser, T.E. &
Hudson, P.J. (2013). Use of exposure history to identify patterns of
immunity to pneumonia in bighorn sheep (Ovis canadensis). PLoS
ONE, 8, e61919.
Plowright, R.K., Peel, A.J., Streicker, D.G., Gilbert, A.T., McCallum, H.,
Wood, J. et al. (2016). Transmission or within-host dynamics driving
pulses of zoonotic viruses in reservoir-host populations. PLoS Negl.
Trop. Dis., 10, e0004796.
Prager, K., Greig, D.J., Alt, D.P., Galloway, R.L., Hornsby, R.L.,
Palmer, L.J. et al. (2013). Asymptomatic and chronic carriage of
Leptospira interrogans serovar Pomona in California sea lions
(Zalophus californianus). Vet. Microbiol., 164, 177–183.
Raberg, L., Graham, A.L. & Read, A.F. (2009). Decomposing health:
tolerance and resistance to parasites in animals. Philos. Trans. R. Soc.
Lond. B Biol. Sci., 364, 37–49.
Roy, B. & Kirchner, J. (2000). Evolutionary dynamics of pathogen
resistance and tolerance. Evolution, 54, 51–63.
Schwaiger, F.-W., Gostomski, D., Stear, M., Duncan, J., McKellar, Q.,
Epplen, J. et al. (1995). An ovine major histocompatibility complex
DRB1 allele is associated with low faecal egg counts following natural,
predominantly Ostertagia circumcincta infection. Int. J. Parasitol., 25,
815–822.
Solana, R., Pawelec, G. & Tarazona, R. (2006). Aging and innate
immunity. Immunity, 24, 491–494.
Spencer, S.E., Besser, T.E., Cobbold, R.N. & French, N.P. (2015).
‘Super’or just ‘above average’? Supershedders and the transmission of
Escherichia coli O157: H7 among feedlot cattle. J. R. Soc. Interface, 12,
20150446.
Streicker, D.G., Fenton, A. & Pedersen, A.B. (2013). Differential sources
of host species heterogeneity influence the transmission and control of
multihost parasites. Ecol. Lett., 16, 975–984.
Tarazona, R., Solana, R., Ouyang, Q. & Pawelec, G. (2002). Basic
biology and clinical impact of immunosenescence. Exp. Gerontol., 37,
183–189.
Turner, J.C. (1977). Cemental annulations as an age criterion in North
American sheep. J. Wildl. Manag., 41, 211–217.
Vallejo, A.N. (2007). Immune remodeling: lessons from repertoire
alterations during chronological aging and in immune-mediated disease.
Trends Mol. Med., 13, 94–102.
VanderWaal, K.L. & Ezenwa, V.O. (2016). Heterogeneity in pathogen
transmission: mechanisms and methodology. Funct. Ecol., 30, 1606–
1622.
Virgin, H.W., Wherry, E.J. & Ahmed, R. (2009). Redefining chronic viral
infection. Cell, 138, 30–50.
Waits, L.P. & Paetkau, D. (2005). Noninvasive genetic sampling tools for
wildlife biologists: a review of applications and recommendations for
accurate data collection. J. Wildl. Manage., 69, 1419–1433.
Wertheim, H.F., Melles, D.C., Vos, M.C., van Leeuwen, W., van Belkum,
A., Verbrugh, H.A. et al. (2005). The role of nasal carriage in
Staphylococcus aureus infections. Lancet Infect. Dis., 5, 751–762.
Wigginton, J.E., Cutler, D.J. & Abecasis, G.R. (2005). A note on exact
tests of Hardy-Weinberg equilibrium. Am. J. Hum. Genet., 76, 887–893.
Wilson, K., Bjørnstad, O., Dobson, A., Merler, S., Poglayen, G.,
Randolph, S. et al. (2002). Heterogeneities in macroparasite infections:
patterns and processes. Ecol. Wildl. Dis., 44, 6–44.
Woolhouse, M.E., Dye, C., Etard, J.-F., Smith, T., Charlwood, J. &
Garnett, G., et al. (1997). Heterogeneities in the transmission of
infectious agents: implications for the design of control programs. Proc.
Natl Acad. Sci., 94, 338–342.
Yagupsky, P. (2015). Kingella kingae: carriage, transmission, and disease.
Clin. Microbiol. Rev., 28, 54–79.
Young, D., Hussell, T. & Dougan, G. (2002). Chronic bacterial
infections: living with unwanted guests. Nat. Immunol., 3, 1026–
1032.
Ziegler, J.C., Lahmers, K.K., Barrington, G.M., Parish, S.M., Kilzer, K.,
Baker, K. et al. (2014). Safety and immunogenicity of a Mycoplasma
ovipneumoniae Bacterin for domestic sheep (Ovis aries). PLoS ONE, 9,
e95698.
SUPPORTING INFORMATION
Additional Supporting Information may be found online in
the supporting information tab for this article.
Editor, Giulio De Leo
Manuscript received 16 March 2017
First decision made 23 April 2017
Second decision made 14 July 2017
Manuscript accepted 23 July 2017