Report Lifetime Fitness Costs of Inbreeding and Being Inbred in a Critically Endangered Bird Highlights d Inbreeding reduces lifetime fitness of critically endangered helmeted honeyeaters d Fitness costs are associated with both engaging in inbreeding and being inbred d Pairing with a genetically dissimilar mate reduces fitness costs of being inbred d Weak short-term effects of inbreeding can underlie stronger lifetime effects Authors Katherine A. Harrisson, Michael J.L. Magrath, Jian D.L. Yen, ..., Kimberly A. Miller, Karina Cartwright, Paul Sunnucks Correspondence [email protected] In Brief Harrisson et al. use genomic tools to quantify inbreeding depression in a critically endangered wild bird. They find lifetime fitness costs associated with both engaging in inbreeding and being inbred and show that pairing with a genetically dissimilar mate can reduce fitness costs associated with being an inbred individual. Harrisson et al., 2019, Current Biology 29, 2711–2717 August 19, 2019 ª 2019 Elsevier Ltd. https://doi.org/10.1016/j.cub.2019.06.064
Lifetime Fitness Costs of Inbreeding and Being Inbred in a Critically Endangered Bird
Feb 03, 2023
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Lifetime Fitness Costs of Inbreeding and Being Inbred in a Critically Endangered Birdnbreeding and Being Inbred in a Critically Endangered Bird
Highlights
helmeted honeyeaters
inbreeding and being inbred
costs of being inbred
lifetime effects
Harrisson et al., 2019, Current Biology 29, 2711–2717 August 19, 2019 ª 2019 Elsevier Ltd. https://doi.org/10.1016/j.cub.2019.06.064
Authors
Kimberly A. Miller, Karina Cartwright,
Paul Sunnucks
Correspondence [email protected]
In Brief
quantify inbreeding depression in a
critically endangered wild bird. They find
lifetime fitness costs associated with
both engaging in inbreeding and being
inbred and show that pairing with a
genetically dissimilar mate can reduce
fitness costs associated with being an
inbred individual.
Report
Lifetime Fitness Costs of Inbreeding and Being Inbred in a Critically Endangered Bird Katherine A. Harrisson,1,2,3,4,10,* Michael J.L. Magrath,5,6 Jian D.L. Yen,6,7 Alexandra Pavlova,4 Neil Murray,1 Bruce Quin,8
Peter Menkhorst,1,2 Kimberly A. Miller,9 Karina Cartwright,9 and Paul Sunnucks4 1Department of Ecology, Environment and Evolution, La Trobe University, Melbourne, VIC 3086, Australia 2Department of Environment, Land, Water and Planning, Arthur Rylah Institute for Environmental Research, Heidelberg, VIC 3084, Australia 3Research Centre for Future Landscapes, La Trobe University, Bundoora, VIC 3086, Australia 4School of Biological Sciences, Monash University, Clayton Campus, Clayton, VIC 3800, Australia 5Department of Wildlife Conservation and Science, Zoos Victoria, Parkville, VIC 3052, Australia 6School of BioSciences, The University of Melbourne, Parkville, VIC 3010, Australia 7ARC Centre of Excellence for Environmental Decisions, The University of Melbourne, Parkville, VIC 3010, Australia 8Department of Environment, Land, Water and Planning, Symes Road, Woori Yallock, VIC 3139, Australia 9Healesville Sanctuary, Zoos Victoria, Healesville, VIC 3777, Australia 10Lead Contact *Correspondence: [email protected]
https://doi.org/10.1016/j.cub.2019.06.064
SUMMARY
Reduced fitness as a result of inbreeding is a major threat facing many species of conservation concern [1–4]. However, few case studies for assessing the magnitudeof inbreedingdepression in thewildmeans that its relative importance as a risk factor for popula- tion persistence remains under-appreciated [5]. The increasing availability and affordability of genomic technologies provide new opportunities to address knowledge gaps around the magnitude and manifes- tation of inbreeding depression in wild populations [6–12]. Here, we combine over three decades of indi- vidual lifetime reproductive data and genomic data to estimate the relative lifetime and short-term fitness costs of bothbeing inbredandengaging in inbreeding in the lastwild population (<250 individuals remaining) of an iconicandcritically endangeredbird: thehelmet- ed honeyeater Lichenostomus melanops cassidix. The magnitude of inbreeding depression was sub- stantial: the mean predicted lifetime reproductive success of the most inbred (homozygosity = 0.82) in- dividuals was on average 87%–90% lower than that of the least inbred (homozygosity = 0.75). For individ- ual reproductive events and lifetime measures, we provide rare empirical evidence that pairing with a genetically dissimilar individual can reduce fitness costs associated with being an inbred individual. By comparing lifetime and short-term fitness measures, we demonstrate how short-term measures of repro- ductive success that are associated with only weak signatures of inbreeding depression can still underlie stronger lifetime effects. Our study represents a valu- able case study, highlighting the critical importance of inbreeding depression as a factor influencing the
Current
RESULTS AND DISCUSSION
Since 1983, the helmeted honeyeater has predominantly existed
as a single wild population of fewer than 250 birds at Yellingbo
Nature Conservation Reserve in Victoria, Australia. From 1995,
the Yellingbo population has been supplemented with captive-
reared offspring in most years (mean = 5.2 ± 6.1 SD per year;
range 0–18) [13]. A small number of captive-reared birds were
also released most years between 2001 and 2012 at additional
sites in Bunyip State Park near Tonimbuk. However, birds rein-
troduced into these areas have failed to establish colonies. The
wild Yellingbo population was estimated to consist of only 226
individuals (40 breeding pairs) in the 2018–2019 breeding
season, and genetic diversity has continued to decline in the
population over time [13]. Intensive monitoring of the helmeted
honeyeater since 1986 has involved banding of individual birds,
blood sampling, and routine monitoring of nesting attempts in
the wild. Here, we explore the extent of inbreeding depression
using reproductive and genomic data for 102 free-living helmet-
ed honeyeater individuals (45 females; 57 males) uniquely iden-
tified with colored leg bands and monitored between 1986 and
2018. Reproductive success was measured at the clutch level
(across 241 nests) and at the individual level across the lifetime
of individuals (n = 65). These data provide a rare case study
where high-quality individual lifetime reproductive success
data and associated genetic samples are available for a relatively
long-lived wild vertebrate; the longest lived individual is currently
21 years old.
Lifetime Fitness Costs Associated with Being Inbred Although many studies have demonstrated inbreeding depres-
sion acting on juvenile or fitness-related traits in wild vertebrates,
few have looked at lifetime fitness consequences [14–16], and
even fewer have done so using measures of inbreeding derived
Biology 29, 2711–2717, August 19, 2019 ª 2019 Elsevier Ltd. 2711
B
0
5
10
15
20
25
30
35
C
Figure 1. Lifetime Fitness Costs of Engaging in Inbreeding and Being Inbred
(A) Bayesian regression coefficients from the model of individual-level lifetime reproductive success (the number of fledglings produced in an individual’s lifetime
[LRS]) show the effects of genome-wide homozygosity (proportion of homozygous loci [HOM]) and mean genomic similarity (proportion of shared alleles [mean
GS]) of an individual to its social mates on LRS. The x axis is the magnitude of the increase or decrease in LRS per 1 SD increase in a given predictor variable. For
each coefficient, the middle line represents the median effect, the upper and lower bounds of the box are 80% Bayesian credible intervals, and the whiskers are
95% Bayesian credible intervals. The dotted vertical line indicates an effect size of zero (i.e., no effect). Descriptions of coefficients are in Table S1.
(B) Fitted relationships between HOM and LRS.
(C) Fitted relationships between mean GS and LRS.
For (B) and (C), separate curves were fitted for males (orange) and females (purple). Shaded regions encompass 80% Bayesian credible intervals, and solid lines
are mean fitted values.
from genome-wide data [10, 11]. Using a genomic proxy for
individual inbreeding (genome-wide homozygosity, based on
11,059 autosomal biallelic SNP loci) and Bayesian regression
modeling, we identified a negative effect of being inbred on the
lifetime reproductive success (LRS) (the number of fledglings
produced during an individual’s lifetime) of wild adult helmeted
honeyeater individuals (Figure 1). There was no evidence of an
interaction between genome-wide homozygosity and sex, sug-
gesting that the strength of inbreeding depression was similar
in males and females (Figure 1).
The model was fitted to observed data, accounting for varia-
tion among sites, bird origins (captive or wild origin), and hatch
seasons. If these effects are not controlled for statistically, the
impacts of inbreeding may not be accurately characterized.
Subsequently, we quantified the magnitude of inbreeding
depression by using the fitted model to predict the reproductive
success of individuals for a range of homozygosity values. For
example, a highly inbred (maximum observed homozygosity =
0.82) male or female of average genomic similarity to its social
mate (mean observed genomic similarity value = 0.78) was pre-
dicted to produce on average 87%–90% fewer fledglings over
its lifetime compared to a less inbred individual (minimum
observed homozygosity = 0.75; Table 1). In other words, on
average, a 9% increase in homozygosity resulted in an 87%–
90% decline in lifetime fitness. Although the majority of individ-
uals included in this study had more intermediate homozygosity
values (mean homozygosity = 0.78; Figure S1), even small
increases in degree of inbreeding resulted in fitness declines
(Table 1). Predictions of LRS based on themodel weremore pre-
cise at high values of homozygosity, suggesting that less inbred
birds can be associated with high or low LRS, whereas more
2712 Current Biology 29, 2711–2717, August 19, 2019
inbred birds have strong upper constraints on lifetime fitness.
The low-to-moderate model r2 values and high uncertainty in
model predictions are to be expected, given the unavoidably
small sample sizes and likely strong influence of environmental
effects and stochastic variation in offspring survival in wild pop-
ulations. Additional noise may also be present in the LRS data
due to the potential for a small number of offspring to be missed,
owing to some temporal variation in nest survey effort (see
Method Details). A benefit of a Bayesian approach to inference
is that uncertainty in parameter estimates is propagated through
to model predictions (Table 1).
Stronger inbreeding effects on LRS than on longevity (see
below) suggest that reduced LRS for very inbred individuals is
not solely the result of these individuals having shorter lifespans
and therefore reduced opportunities to reproduce. Given our
study was limited to adult birds associated with lifetime fitness
data, we have necessarily biased our study to birds that survived
to adulthood and that are likely to be the more reproductively
successful individuals in the population. Because of this, we
are likely to be underestimating the true magnitude of inbreeding
depression acting over the lifetime of individuals. Although the
majority of offspring attributed to an individual based on behav-
ioral observation would be expected to be true offspring of the
social pair, there is likely some degree of extra-pair paternity
(EPP) in helmeted honeyeaters as in many other passerines (in
the order of 10%–15%) [17], which means that the social father
may not always be the genetic father. However, any EPP would
be expected to affect male estimates of reproductive success
more than female estimates and either to introduce noise or, if fe-
males actively undertake extra-pair copulations with less inbred
or less related males to avoid inbreeding costs, weaken rather
Table 1. Mean Predicted LRS (Number of Fledglings Produced in an Individual’s Lifetime) for Male and Female Helmeted
Honeyeaters at Different Values of Genome-wide Homozygosity (HOM), Taking into Account the Mean Genomic Similarity of an
Individual to Its Social Mates (Mate GS)
Female LRS Male LRS
HOM Low Mate GS Mean Mate GS High Mate GS Low Mate GS Med Mate GS High Mate GS
0.75 16.2 [1, 25] 18.2 [2, 28] 35.1 [0, 59] 15.2 [1, 23] 17.5 [2, 27] 33.2 [2, 54]
0.77 11.7 [1, 18] 10.3 [1, 16] 7.2 [0, 12] 9.9 [1, 15] 8.7 [0, 14] 6.5 [0, 11]
0.79 8.6 [0, 13] 5.6 [0, 9] 1.8 [0, 3] 6.6 [0, 11] 4.5 [0, 7] 1.6 [0, 3]
0.82 5.6 [0, 9] 2.4 [0, 4] 0.3 [0, 1] 4.0 [0, 7] 1.7 [0, 3] 0.3 [0, 1]
Mean mate GS is calculated for all clutches over an individual’s lifetime; low GS is the minimum observed value (0.75), mean mate GS is the average
value (0.78), and high mate GS is the maximum observed value (0.89). Homozygosity values span the observed range. Predicted values were drawn
from the posterior predictive distributions of the fitted individual-level model. 10% and 90% quantiles are given in brackets. Related to Table S1.
than drive the observed effects. For example, if females prefer-
entially choose more outbred and more different males for
EPP, then LRS will be underestimated for those outbred males,
and overestimated for inbred males. Thus, potential biases in
our analyses resulting from EPP would most likely lead to an
overall downward bias in the magnitude of estimates of
inbreeding depression for males (for lifetime and clutch-level
reproductive success).
Pairing with a Genetically Dissimilar Mate Can Reduce Fitness Costs of Being Inbred Although inbreeding depression is typically expressed as the
reduction in an individual’s fitness resulting from its being inbred,
here, we also quantified the reduction in the LRS of breeding
adults resulting from pairing with genetically similar individuals
(i.e., engaging in inbreeding) [18]. Although most individuals
stayed with the same social partner for their entire lifetime,
36% of the individuals switched mates at least once (usually
following unsuccessful breeding attempts or the death of a
mate). We found that, in addition to negative lifetime fitness
costs associated with being inbred, individuals that had higher
genomic similarity to their social mates (averaged over their
lifetime and so weighted according to the number of years an
individual spent with different partners) had lower LRS than indi-
viduals more dissimilar to their social mates (Figure 1). The
magnitude of the negative effect of mean social mate similarity
was weaker than that of individual genome-wide homozygosity
(Figure 1). There was little evidence of an interaction between
mean social mate similarity and sex (Figure 1).
Evidence of an interaction between individual genome-wide
homozygosity and mean genomic similarity of social mates sug-
gests that the negative fitness effects of being inbred are greater
when an individual is paired with a genetically similar individual
(Table 1; Figures 1 and 2). Similarly, the negative fitness effects
of pairing with a genetically similar mate are greater when an in-
dividual is more inbred. Because of the negative interaction
between genomic similarity and homozygosity, the effect of
genomic similarity on LRS was positive at very low levels of
homozygosity, but this association was subject to high levels
of uncertainty (e.g., proportion of homozygous loci [HOM] =
0.75; Table 1). This association might reflect that more heterozy-
gous individuals in this population are likely to be more similar to
each other than tomore homozygous individuals. Overall, our re-
sults suggest that pairing with a genetically dissimilar mate can
reduce the lifetime fitness costs associated with being an inbred
individual.
Our study is unusual in quantifying the relative fitness costs
associated with an individual being inbred and an individual
engaging in inbreeding within the same analytical framework
[18, 19]. Studies that have tested for sex-specific fitness costs
associated with individuals engaging in inbreeding (as distinct
from sex-specific fitness costs associated with ‘‘being inbred’’)
in polygynous red deer [19] and in socially monogamous song
sparrows [18] have reported reproductive costs of engaging in
inbreeding for males, but not females. In the present study, we
found evidence of fitness costs associated with engaging in
inbreeding in both sexes, although it remains to be tested
whether helmeted honeyeaters are making active choices to
avoid breeding with genetically similar mates (i.e., exhibiting
inbreeding avoidance behavior). Although it is often assumed
that inbreeding depression will lead to selection against mating
between relatives and the evolution of inbreeding avoidance
mechanisms, this is not always borne out due to potential costs
of not inbreeding (e.g., lost mating opportunities) and potential
benefits of inbreeding (e.g., kin selection) [15, 18, 20–23].
Negative Consequences of Being Inbred for Longevity of Individuals Being inbred had a negative effect on the lifespan of helmeted
honeyeaters, although the magnitude of the inbreeding depres-
sion was less than that observed for LRS (Figure S2). There
was evidence of an interaction between genome-wide homozy-
gosity and sex, suggesting a stronger negative effect of being
inbred on lifespans of females than males (Figure S2). The
most inbred female (homozygosity = 0.82) was predicted by
themodel to live on average 1,979 days less than the least inbred
female (homozygosity = 0.75), representing a 49% decline in fe-
male lifespan due to inbreeding (Table S2). There was little evi-
dence of an effect of genome-wide homozygosity on the lifespan
of males (Figure S2; Table S2). Given that our study was
restricted to adult birds associated with lifetime reproductive
data, sex-specific effects of inbreeding on longevity should be
interpreted with some caution.
Short-Term Fitness Proxies for Measuring Inbreeding Depression As reported in other studies, effects of inbreeding were not as
strong when considering reproductive success at the clutch
Current Biology 29, 2711–2717, August 19, 2019 2713
0
20
40
60
80
Homozygosity
A − Low mate GS
B − Mean mate GS
C − High mate GSA B C
Figure 2. Engaging in Inbreeding Compounds Effects of Being Inbred
Illustration of the effects on an individual’s LRS (number of fledglings produced in an individual’s lifetime) of an interaction between individual inbreeding
(homozygosity) and mean genomic similarity of an individual to its social mates. Panels show fitted relationships between homozygosity and LRS for individuals
with (A) comparatively low (minimum observed = 0.75), (B) mean (mean observed = 0.78), and (C) comparatively high (maximum observed = 0.89) genomic
similarity to social mates. Separate curves were fitted for males (orange) and females (purple). Shaded regions encompass 80% Bayesian credible intervals, and
solid lines are mean fitted values. See also Table S1.
level, as opposed to LRS [10, 15, 24]. At the clutch level, we de-
tected inbreeding depression associated with some, but not all,
stages of reproduction. Helmeted honeyeaters produce up to
three fledglings per breeding attempt. There was no evidence
of a negative effect of a social parent’s degree of inbreeding,
parent age, or genomic similarity of socially bonded parent pair
on the number of eggs recorded per clutch (Figure 3; Tables
S3 and S4). However, the SD of clutch size (SD = 0.46) was lower
than that of number of hatchlings (SD = 0.88) and fledglings (SD =
0.91), and there may have been insufficient variation in the num-
ber of eggs recorded to detect inbreeding effects at the clutch
level (72% of nests had two eggs recorded). There was evidence
of inbreeding depression associated with hatching and fledging
stages of reproduction, with moderate-to-strong support for
negative effects of female social parent genome-wide homozy-
gosity (posterior probabilities 0.76–0.82), male social parent
genome-wide homozygosity (posterior probabilities 0.78–0.79),
and genomic similarity of social parents (posterior probabilities
0.78–0.92) on the number of hatchlings and fledglings produced
per clutch (Figures 3B, 3C, and S3; Tables S3 and S4). These re-
sults indicate reproductive costs associated with parents being
inbred and parents engaging in an inbred mating at the clutch
level.
effects on the number of fledglings produced per clutch (Fig-
ure 3C). Thus, as detected in lifetime models, for individual
breeding attempts, inbred individuals can reduce reproductive
costs associated with having high individual homozygosity by
pairing with a genetically dissimilar mate. Male parent homozy-
gosity had a stronger negative effect on clutch-level reproduc-
tive success than did female parent homozygosity, with a 10%
and 7% reduction in number of fledglings produced per clutch
2714 Current Biology 29, 2711–2717, August 19, 2019
per SD increase in male and female parent homozygosity,
respectively (Table S1). The effect of genomic similarity of a
pair on the number of fledglings produced per clutch was
similar in magnitude to that of male parent homozygosity: a
one SD increase in the proportion of shared alleles between
socially bonded parent birds was associated with a 17% reduc-
tion in the number of fledglings produced per clutch (Table S1).
Within the range of observed homozygosity and genomic
similarity values, inbreeding effects translated to a 53% decline
in the number of fledglings produced per clutch when
comparing the most similar pair with the least similar pair, a
42% decline in the number of fledglings produced per clutch
when comparing the most inbred male to the most outbred
male, and a 26% decline in the number of fledglings produced
per clutch when comparing the most inbred female with the
most outbred female (Table S3). Mating between genetically
similar individuals reduced the number of fledglings per clutch
more than it did the number of hatchlings, even after accounting
for the cumulative effects of inbreeding depression through
consecutive life stages (eggs to hatchlings to fledglings; Table
S1; Figure 3).
Although our results suggested that the negative lifetime con-
sequences of individual inbreeding were equally strong for both
sexes, male effects appeared to be stronger than those of fe-
males at the clutch level. More outbred females were predicted
to live longer than their more inbred counterparts and therefore
have more reproductive opportunities, which may explain why
strong effects of female inbreeding on LRS were not reflected
to the same degree in components of reproductive success
measured at the clutch level. Alternatively, because helmeted
honeyeaters can breed multiple…
Highlights
helmeted honeyeaters
inbreeding and being inbred
costs of being inbred
lifetime effects
Harrisson et al., 2019, Current Biology 29, 2711–2717 August 19, 2019 ª 2019 Elsevier Ltd. https://doi.org/10.1016/j.cub.2019.06.064
Authors
Kimberly A. Miller, Karina Cartwright,
Paul Sunnucks
Correspondence [email protected]
In Brief
quantify inbreeding depression in a
critically endangered wild bird. They find
lifetime fitness costs associated with
both engaging in inbreeding and being
inbred and show that pairing with a
genetically dissimilar mate can reduce
fitness costs associated with being an
inbred individual.
Report
Lifetime Fitness Costs of Inbreeding and Being Inbred in a Critically Endangered Bird Katherine A. Harrisson,1,2,3,4,10,* Michael J.L. Magrath,5,6 Jian D.L. Yen,6,7 Alexandra Pavlova,4 Neil Murray,1 Bruce Quin,8
Peter Menkhorst,1,2 Kimberly A. Miller,9 Karina Cartwright,9 and Paul Sunnucks4 1Department of Ecology, Environment and Evolution, La Trobe University, Melbourne, VIC 3086, Australia 2Department of Environment, Land, Water and Planning, Arthur Rylah Institute for Environmental Research, Heidelberg, VIC 3084, Australia 3Research Centre for Future Landscapes, La Trobe University, Bundoora, VIC 3086, Australia 4School of Biological Sciences, Monash University, Clayton Campus, Clayton, VIC 3800, Australia 5Department of Wildlife Conservation and Science, Zoos Victoria, Parkville, VIC 3052, Australia 6School of BioSciences, The University of Melbourne, Parkville, VIC 3010, Australia 7ARC Centre of Excellence for Environmental Decisions, The University of Melbourne, Parkville, VIC 3010, Australia 8Department of Environment, Land, Water and Planning, Symes Road, Woori Yallock, VIC 3139, Australia 9Healesville Sanctuary, Zoos Victoria, Healesville, VIC 3777, Australia 10Lead Contact *Correspondence: [email protected]
https://doi.org/10.1016/j.cub.2019.06.064
SUMMARY
Reduced fitness as a result of inbreeding is a major threat facing many species of conservation concern [1–4]. However, few case studies for assessing the magnitudeof inbreedingdepression in thewildmeans that its relative importance as a risk factor for popula- tion persistence remains under-appreciated [5]. The increasing availability and affordability of genomic technologies provide new opportunities to address knowledge gaps around the magnitude and manifes- tation of inbreeding depression in wild populations [6–12]. Here, we combine over three decades of indi- vidual lifetime reproductive data and genomic data to estimate the relative lifetime and short-term fitness costs of bothbeing inbredandengaging in inbreeding in the lastwild population (<250 individuals remaining) of an iconicandcritically endangeredbird: thehelmet- ed honeyeater Lichenostomus melanops cassidix. The magnitude of inbreeding depression was sub- stantial: the mean predicted lifetime reproductive success of the most inbred (homozygosity = 0.82) in- dividuals was on average 87%–90% lower than that of the least inbred (homozygosity = 0.75). For individ- ual reproductive events and lifetime measures, we provide rare empirical evidence that pairing with a genetically dissimilar individual can reduce fitness costs associated with being an inbred individual. By comparing lifetime and short-term fitness measures, we demonstrate how short-term measures of repro- ductive success that are associated with only weak signatures of inbreeding depression can still underlie stronger lifetime effects. Our study represents a valu- able case study, highlighting the critical importance of inbreeding depression as a factor influencing the
Current
RESULTS AND DISCUSSION
Since 1983, the helmeted honeyeater has predominantly existed
as a single wild population of fewer than 250 birds at Yellingbo
Nature Conservation Reserve in Victoria, Australia. From 1995,
the Yellingbo population has been supplemented with captive-
reared offspring in most years (mean = 5.2 ± 6.1 SD per year;
range 0–18) [13]. A small number of captive-reared birds were
also released most years between 2001 and 2012 at additional
sites in Bunyip State Park near Tonimbuk. However, birds rein-
troduced into these areas have failed to establish colonies. The
wild Yellingbo population was estimated to consist of only 226
individuals (40 breeding pairs) in the 2018–2019 breeding
season, and genetic diversity has continued to decline in the
population over time [13]. Intensive monitoring of the helmeted
honeyeater since 1986 has involved banding of individual birds,
blood sampling, and routine monitoring of nesting attempts in
the wild. Here, we explore the extent of inbreeding depression
using reproductive and genomic data for 102 free-living helmet-
ed honeyeater individuals (45 females; 57 males) uniquely iden-
tified with colored leg bands and monitored between 1986 and
2018. Reproductive success was measured at the clutch level
(across 241 nests) and at the individual level across the lifetime
of individuals (n = 65). These data provide a rare case study
where high-quality individual lifetime reproductive success
data and associated genetic samples are available for a relatively
long-lived wild vertebrate; the longest lived individual is currently
21 years old.
Lifetime Fitness Costs Associated with Being Inbred Although many studies have demonstrated inbreeding depres-
sion acting on juvenile or fitness-related traits in wild vertebrates,
few have looked at lifetime fitness consequences [14–16], and
even fewer have done so using measures of inbreeding derived
Biology 29, 2711–2717, August 19, 2019 ª 2019 Elsevier Ltd. 2711
B
0
5
10
15
20
25
30
35
C
Figure 1. Lifetime Fitness Costs of Engaging in Inbreeding and Being Inbred
(A) Bayesian regression coefficients from the model of individual-level lifetime reproductive success (the number of fledglings produced in an individual’s lifetime
[LRS]) show the effects of genome-wide homozygosity (proportion of homozygous loci [HOM]) and mean genomic similarity (proportion of shared alleles [mean
GS]) of an individual to its social mates on LRS. The x axis is the magnitude of the increase or decrease in LRS per 1 SD increase in a given predictor variable. For
each coefficient, the middle line represents the median effect, the upper and lower bounds of the box are 80% Bayesian credible intervals, and the whiskers are
95% Bayesian credible intervals. The dotted vertical line indicates an effect size of zero (i.e., no effect). Descriptions of coefficients are in Table S1.
(B) Fitted relationships between HOM and LRS.
(C) Fitted relationships between mean GS and LRS.
For (B) and (C), separate curves were fitted for males (orange) and females (purple). Shaded regions encompass 80% Bayesian credible intervals, and solid lines
are mean fitted values.
from genome-wide data [10, 11]. Using a genomic proxy for
individual inbreeding (genome-wide homozygosity, based on
11,059 autosomal biallelic SNP loci) and Bayesian regression
modeling, we identified a negative effect of being inbred on the
lifetime reproductive success (LRS) (the number of fledglings
produced during an individual’s lifetime) of wild adult helmeted
honeyeater individuals (Figure 1). There was no evidence of an
interaction between genome-wide homozygosity and sex, sug-
gesting that the strength of inbreeding depression was similar
in males and females (Figure 1).
The model was fitted to observed data, accounting for varia-
tion among sites, bird origins (captive or wild origin), and hatch
seasons. If these effects are not controlled for statistically, the
impacts of inbreeding may not be accurately characterized.
Subsequently, we quantified the magnitude of inbreeding
depression by using the fitted model to predict the reproductive
success of individuals for a range of homozygosity values. For
example, a highly inbred (maximum observed homozygosity =
0.82) male or female of average genomic similarity to its social
mate (mean observed genomic similarity value = 0.78) was pre-
dicted to produce on average 87%–90% fewer fledglings over
its lifetime compared to a less inbred individual (minimum
observed homozygosity = 0.75; Table 1). In other words, on
average, a 9% increase in homozygosity resulted in an 87%–
90% decline in lifetime fitness. Although the majority of individ-
uals included in this study had more intermediate homozygosity
values (mean homozygosity = 0.78; Figure S1), even small
increases in degree of inbreeding resulted in fitness declines
(Table 1). Predictions of LRS based on themodel weremore pre-
cise at high values of homozygosity, suggesting that less inbred
birds can be associated with high or low LRS, whereas more
2712 Current Biology 29, 2711–2717, August 19, 2019
inbred birds have strong upper constraints on lifetime fitness.
The low-to-moderate model r2 values and high uncertainty in
model predictions are to be expected, given the unavoidably
small sample sizes and likely strong influence of environmental
effects and stochastic variation in offspring survival in wild pop-
ulations. Additional noise may also be present in the LRS data
due to the potential for a small number of offspring to be missed,
owing to some temporal variation in nest survey effort (see
Method Details). A benefit of a Bayesian approach to inference
is that uncertainty in parameter estimates is propagated through
to model predictions (Table 1).
Stronger inbreeding effects on LRS than on longevity (see
below) suggest that reduced LRS for very inbred individuals is
not solely the result of these individuals having shorter lifespans
and therefore reduced opportunities to reproduce. Given our
study was limited to adult birds associated with lifetime fitness
data, we have necessarily biased our study to birds that survived
to adulthood and that are likely to be the more reproductively
successful individuals in the population. Because of this, we
are likely to be underestimating the true magnitude of inbreeding
depression acting over the lifetime of individuals. Although the
majority of offspring attributed to an individual based on behav-
ioral observation would be expected to be true offspring of the
social pair, there is likely some degree of extra-pair paternity
(EPP) in helmeted honeyeaters as in many other passerines (in
the order of 10%–15%) [17], which means that the social father
may not always be the genetic father. However, any EPP would
be expected to affect male estimates of reproductive success
more than female estimates and either to introduce noise or, if fe-
males actively undertake extra-pair copulations with less inbred
or less related males to avoid inbreeding costs, weaken rather
Table 1. Mean Predicted LRS (Number of Fledglings Produced in an Individual’s Lifetime) for Male and Female Helmeted
Honeyeaters at Different Values of Genome-wide Homozygosity (HOM), Taking into Account the Mean Genomic Similarity of an
Individual to Its Social Mates (Mate GS)
Female LRS Male LRS
HOM Low Mate GS Mean Mate GS High Mate GS Low Mate GS Med Mate GS High Mate GS
0.75 16.2 [1, 25] 18.2 [2, 28] 35.1 [0, 59] 15.2 [1, 23] 17.5 [2, 27] 33.2 [2, 54]
0.77 11.7 [1, 18] 10.3 [1, 16] 7.2 [0, 12] 9.9 [1, 15] 8.7 [0, 14] 6.5 [0, 11]
0.79 8.6 [0, 13] 5.6 [0, 9] 1.8 [0, 3] 6.6 [0, 11] 4.5 [0, 7] 1.6 [0, 3]
0.82 5.6 [0, 9] 2.4 [0, 4] 0.3 [0, 1] 4.0 [0, 7] 1.7 [0, 3] 0.3 [0, 1]
Mean mate GS is calculated for all clutches over an individual’s lifetime; low GS is the minimum observed value (0.75), mean mate GS is the average
value (0.78), and high mate GS is the maximum observed value (0.89). Homozygosity values span the observed range. Predicted values were drawn
from the posterior predictive distributions of the fitted individual-level model. 10% and 90% quantiles are given in brackets. Related to Table S1.
than drive the observed effects. For example, if females prefer-
entially choose more outbred and more different males for
EPP, then LRS will be underestimated for those outbred males,
and overestimated for inbred males. Thus, potential biases in
our analyses resulting from EPP would most likely lead to an
overall downward bias in the magnitude of estimates of
inbreeding depression for males (for lifetime and clutch-level
reproductive success).
Pairing with a Genetically Dissimilar Mate Can Reduce Fitness Costs of Being Inbred Although inbreeding depression is typically expressed as the
reduction in an individual’s fitness resulting from its being inbred,
here, we also quantified the reduction in the LRS of breeding
adults resulting from pairing with genetically similar individuals
(i.e., engaging in inbreeding) [18]. Although most individuals
stayed with the same social partner for their entire lifetime,
36% of the individuals switched mates at least once (usually
following unsuccessful breeding attempts or the death of a
mate). We found that, in addition to negative lifetime fitness
costs associated with being inbred, individuals that had higher
genomic similarity to their social mates (averaged over their
lifetime and so weighted according to the number of years an
individual spent with different partners) had lower LRS than indi-
viduals more dissimilar to their social mates (Figure 1). The
magnitude of the negative effect of mean social mate similarity
was weaker than that of individual genome-wide homozygosity
(Figure 1). There was little evidence of an interaction between
mean social mate similarity and sex (Figure 1).
Evidence of an interaction between individual genome-wide
homozygosity and mean genomic similarity of social mates sug-
gests that the negative fitness effects of being inbred are greater
when an individual is paired with a genetically similar individual
(Table 1; Figures 1 and 2). Similarly, the negative fitness effects
of pairing with a genetically similar mate are greater when an in-
dividual is more inbred. Because of the negative interaction
between genomic similarity and homozygosity, the effect of
genomic similarity on LRS was positive at very low levels of
homozygosity, but this association was subject to high levels
of uncertainty (e.g., proportion of homozygous loci [HOM] =
0.75; Table 1). This association might reflect that more heterozy-
gous individuals in this population are likely to be more similar to
each other than tomore homozygous individuals. Overall, our re-
sults suggest that pairing with a genetically dissimilar mate can
reduce the lifetime fitness costs associated with being an inbred
individual.
Our study is unusual in quantifying the relative fitness costs
associated with an individual being inbred and an individual
engaging in inbreeding within the same analytical framework
[18, 19]. Studies that have tested for sex-specific fitness costs
associated with individuals engaging in inbreeding (as distinct
from sex-specific fitness costs associated with ‘‘being inbred’’)
in polygynous red deer [19] and in socially monogamous song
sparrows [18] have reported reproductive costs of engaging in
inbreeding for males, but not females. In the present study, we
found evidence of fitness costs associated with engaging in
inbreeding in both sexes, although it remains to be tested
whether helmeted honeyeaters are making active choices to
avoid breeding with genetically similar mates (i.e., exhibiting
inbreeding avoidance behavior). Although it is often assumed
that inbreeding depression will lead to selection against mating
between relatives and the evolution of inbreeding avoidance
mechanisms, this is not always borne out due to potential costs
of not inbreeding (e.g., lost mating opportunities) and potential
benefits of inbreeding (e.g., kin selection) [15, 18, 20–23].
Negative Consequences of Being Inbred for Longevity of Individuals Being inbred had a negative effect on the lifespan of helmeted
honeyeaters, although the magnitude of the inbreeding depres-
sion was less than that observed for LRS (Figure S2). There
was evidence of an interaction between genome-wide homozy-
gosity and sex, suggesting a stronger negative effect of being
inbred on lifespans of females than males (Figure S2). The
most inbred female (homozygosity = 0.82) was predicted by
themodel to live on average 1,979 days less than the least inbred
female (homozygosity = 0.75), representing a 49% decline in fe-
male lifespan due to inbreeding (Table S2). There was little evi-
dence of an effect of genome-wide homozygosity on the lifespan
of males (Figure S2; Table S2). Given that our study was
restricted to adult birds associated with lifetime reproductive
data, sex-specific effects of inbreeding on longevity should be
interpreted with some caution.
Short-Term Fitness Proxies for Measuring Inbreeding Depression As reported in other studies, effects of inbreeding were not as
strong when considering reproductive success at the clutch
Current Biology 29, 2711–2717, August 19, 2019 2713
0
20
40
60
80
Homozygosity
A − Low mate GS
B − Mean mate GS
C − High mate GSA B C
Figure 2. Engaging in Inbreeding Compounds Effects of Being Inbred
Illustration of the effects on an individual’s LRS (number of fledglings produced in an individual’s lifetime) of an interaction between individual inbreeding
(homozygosity) and mean genomic similarity of an individual to its social mates. Panels show fitted relationships between homozygosity and LRS for individuals
with (A) comparatively low (minimum observed = 0.75), (B) mean (mean observed = 0.78), and (C) comparatively high (maximum observed = 0.89) genomic
similarity to social mates. Separate curves were fitted for males (orange) and females (purple). Shaded regions encompass 80% Bayesian credible intervals, and
solid lines are mean fitted values. See also Table S1.
level, as opposed to LRS [10, 15, 24]. At the clutch level, we de-
tected inbreeding depression associated with some, but not all,
stages of reproduction. Helmeted honeyeaters produce up to
three fledglings per breeding attempt. There was no evidence
of a negative effect of a social parent’s degree of inbreeding,
parent age, or genomic similarity of socially bonded parent pair
on the number of eggs recorded per clutch (Figure 3; Tables
S3 and S4). However, the SD of clutch size (SD = 0.46) was lower
than that of number of hatchlings (SD = 0.88) and fledglings (SD =
0.91), and there may have been insufficient variation in the num-
ber of eggs recorded to detect inbreeding effects at the clutch
level (72% of nests had two eggs recorded). There was evidence
of inbreeding depression associated with hatching and fledging
stages of reproduction, with moderate-to-strong support for
negative effects of female social parent genome-wide homozy-
gosity (posterior probabilities 0.76–0.82), male social parent
genome-wide homozygosity (posterior probabilities 0.78–0.79),
and genomic similarity of social parents (posterior probabilities
0.78–0.92) on the number of hatchlings and fledglings produced
per clutch (Figures 3B, 3C, and S3; Tables S3 and S4). These re-
sults indicate reproductive costs associated with parents being
inbred and parents engaging in an inbred mating at the clutch
level.
effects on the number of fledglings produced per clutch (Fig-
ure 3C). Thus, as detected in lifetime models, for individual
breeding attempts, inbred individuals can reduce reproductive
costs associated with having high individual homozygosity by
pairing with a genetically dissimilar mate. Male parent homozy-
gosity had a stronger negative effect on clutch-level reproduc-
tive success than did female parent homozygosity, with a 10%
and 7% reduction in number of fledglings produced per clutch
2714 Current Biology 29, 2711–2717, August 19, 2019
per SD increase in male and female parent homozygosity,
respectively (Table S1). The effect of genomic similarity of a
pair on the number of fledglings produced per clutch was
similar in magnitude to that of male parent homozygosity: a
one SD increase in the proportion of shared alleles between
socially bonded parent birds was associated with a 17% reduc-
tion in the number of fledglings produced per clutch (Table S1).
Within the range of observed homozygosity and genomic
similarity values, inbreeding effects translated to a 53% decline
in the number of fledglings produced per clutch when
comparing the most similar pair with the least similar pair, a
42% decline in the number of fledglings produced per clutch
when comparing the most inbred male to the most outbred
male, and a 26% decline in the number of fledglings produced
per clutch when comparing the most inbred female with the
most outbred female (Table S3). Mating between genetically
similar individuals reduced the number of fledglings per clutch
more than it did the number of hatchlings, even after accounting
for the cumulative effects of inbreeding depression through
consecutive life stages (eggs to hatchlings to fledglings; Table
S1; Figure 3).
Although our results suggested that the negative lifetime con-
sequences of individual inbreeding were equally strong for both
sexes, male effects appeared to be stronger than those of fe-
males at the clutch level. More outbred females were predicted
to live longer than their more inbred counterparts and therefore
have more reproductive opportunities, which may explain why
strong effects of female inbreeding on LRS were not reflected
to the same degree in components of reproductive success
measured at the clutch level. Alternatively, because helmeted
honeyeaters can breed multiple…
Related Documents
