Limited Lifetime Costs of Inbreeding in a Large Wild Population of Superb Fairy‐Wrens

ABSTRACT

Inbreeding depression can have major negative effects on both individual fitness and population performance. The impact of inbreeding depression in wild populations remains poorly understood, and our knowledge is primarily based on estimations in insular populations with limited dispersal, which has been predicted to enhance the capacity for outcrossing. Limited migration and small population sizes, which make insular and endangered populations tractable models for fieldwork, often experience low genetic variation and may be subject to stronger inbreeding effects. Estimating inbreeding effects is difficult in many populations because pedigrees of migrants are unknown. Genomic data offer methods to estimate realised inbreeding even in migrants. Here, we apply genomic methods to a superb fairy‐wren ( Malurus cyaneus ) population that has been subject to detailed study of fitness variation over three decades. The population has sex‐biased dispersal, resulting in many immigrant females. We use genomic inbreeding estimates to explicitly test for inbreeding depression across several key components of individual fitness, including nestling condition, juvenile survival and adult traits. Genomic methods were able to recover previous pedigree‐based findings of inbreeding depression in early nestling condition and additionally identified some evidence for inbreeding depression in fledgling but not juvenile mortality. The only adult fitness trait affected, male extra‐pair paternity, is highly variable and under strong sexual selection. Our findings suggest that natural populations experiencing high effective population sizes and regular dispersal might be less affected by inbreeding depression than previously assumed based on studies of often insular and smaller populations.

1. Introduction

Inbreeding can negatively impact both individual fitness and population performance via inbreeding depression, and consequently contribute to population decline (Frankham 2005; Saccheri et al. 1998). Inbreeding occurs when related individuals mate, resulting in offspring with genomic regions that are identical‐by‐descent (IBD) as they originate from a common ancestor (Charlesworth and Charlesworth 1987; Keller and Waller 2002). The associated increased homozygosity of IBD regions in inbred offspring can have significant negative fitness consequences, through either expression of deleterious but recessive alleles or loss of heterozygosity at overdominant loci (Charlesworth and Charlesworth 1987; Charlesworth and Willis 2009). Such costs may be countered by benefits of inbreeding, such as inclusive fitness, or reduction of costs of inbreeding avoidance (Duthie and Reid 2016; Kokko and Ots 2006; Szulkin et al. 2013). A reduction of fitness in inbred individuals relative to outbred individuals, called inbreeding depression, is common across many species and fitness traits (Angeloni et al. 2011; Doekes et al. 2021; Keller and Waller 2002). While there is growing evidence for inbreeding depression in wild populations (Kardos et al. 2016; Keller and Waller 2002), advancing such studies in natural settings has been limited due to three major challenges: estimating both inbreeding and lifetime fitness in natural populations, and complex condition‐dependence of inbreeding effects.

Quantifying individual pedigree inbreeding has previously required intensive multigenerational studies to estimate the expected proportion of the genome that is IBD. Such methods are unable to capture stochastic variation associated with Mendelian sampling variation and linkage, and are limited by the availability, depth and accuracy of pedigrees. Genomic estimates of realised inbreeding are increasingly accessible and potentially more sensitive for detecting inbreeding depression (Kardos et al. 2015, 2016), and, even in populations subject to long‐term study, can estimate the inbreeding history of both founders and immigrants (Chen et al. 2016). However, using genomic data to quantify inbreeding is complicated by population structure (Lavanchy et al. 2024), genotyping methods (Duntsch et al. 2021; Lavanchy and Goudet 2023) and differences between inbreeding metrics (Caballero et al. 2021; Duntsch et al. 2021; M. C. Keller et al. 2011). Regardless, genomic methods offer powerful approaches to quantify inbreeding depression in wild populations.

Quantifying inbreeding depression, however, still requires some estimate of fitness which, for lifetime measures such as lifespan and lifetime reproductive success, often still requires intensive multi‐generational individual‐level study. Consequently, while inbreeding depression in juvenile traits may be more commonly quantified (Keller and Waller 2002), studies assessing the effect of inbreeding across the full lifespan of individuals in natural populations remain relatively rare (Huisman et al. 2016; Keller et al. 2008; Kruuk et al. 2002; Stoffel et al. 2021; Szulkin et al. 2007). Theory predicts purging will be inefficient against late‐acting or cumulative effects of deleterious alleles (including those contributing to inbreeding load), leading to their accumulation and high fitness costs in late‐life and senescence (the mutation accumulation hypothesis: Medawar 1953). However, lifetime effects of inbreeding remain understudied in wild populations (Grueber et al. 2010; Keller et al. 2008; Szulkin et al. 2007; Trask et al. 2021), and results remain conflicting. For example, inbreeding depression was stronger in early life survival in island populations of red deer ( Cervus elaphus ) and Soay sheep ( Ovis aries ) (Huisman et al. 2016; Stoffel et al. 2021), but inbreeding depression in survival increased with age in song sparrows ( Melospiza melodia ) (Keller et al. 2008), and appeared to affect most life‐history stages including reproductive traits in great tits ( Parus major ) and collared flycatchers ( Ficedula albicollis ) (Kruuk et al. 2002; Szulkin et al. 2007). As early life traits are easier to quantify and predicted to be less affected by accumulated inbreeding load, it is possible that the effect of inbreeding depression in lifetime and adult components of fitness in the wild has been underestimated. While challenging, quantifying inbreeding depression in late‐life and lifetime fitness is vital to understanding its full impact on both individual fitness and population performance.

Lastly, condition‐dependence of inbreeding effects may serve to mask or exacerbate inbreeding depression in the wild (Armbruster and Reed 2005; Keller et al. 1994), such as environmental conditions (Bebbington et al. 2016; Cheptou and Donohue 2011; Fox and Reed 2011; Gooley et al. 2017; Keller et al. 2002; Springer et al. 2020), social and population dynamics (Nielsen et al. 2012; Yun and Agrawal 2014), and sex differences (Bilde et al. 2009; Brekke et al. 2010; Saccheri et al. 2005; Vega‐Trejo et al. 2022). For example, decreased food availability and interspecific competition increased inbreeding depression in both juvenile and adult survival of cactus finches ( Geospiza scandens ) (Keller et al. 2002). Conversely, favourable environments may mask inbreeding depression, as found in meerkats ( Suricata suricatta ) where the presence of subordinates who provision and protect young within their social group buffered inbreeding effects on pup traits (Nielsen et al. 2012). Natural populations therefore may be subject to multiple complex interactions of inbreeding and environmental condition‐dependence.

Consequently, ecological and genetic data are required over sufficient temporal and environmental scales to quantify inbreeding depression and understand its effects in the wild. Indeed, the majority of studies with requisite data to quantify inbreeding depression in lifetime fitness are limited to highly endangered or managed populations (Bilski et al. 2013; Brekke et al. 2010; Duntsch et al. 2023; Foster et al. 2021; Grueber et al. 2010; Harrisson et al. 2019; Trask et al. 2021), or island populations (Bérénos et al. 2016; Huisman et al. 2016; Keller et al. 2008; Kruuk et al. 2002). The frequency of inbreeding in such populations is generally strongly increased by small population sizes (Caballero et al. 2017; Frankham 2005), historical bottlenecks (Brekke et al. 2010; Keller et al. 1994) or limited migration (Chen et al. 2016; Keller et al. 2001), which in combination strongly affect purging of deleterious variation. Small population size and frequent inbreeding may permit effective purging of deleterious inbreeding load (Eyer et al. 2018; reviewed in Hedrick and Garcia‐Dorado 2016; Robinson et al. 2018), or lead to fixation of deleterious inbreeding load and strong inbreeding depression (Frankham 2005) depending on the specific demographic circumstances and population history. Further, dispersal may prevent fixation of inbreeding load, or introduce inbreeding load that is consequently expressed after initial heterosis (Keller and Waller 2002; Smeds and Ellegren 2023). Thus, many populations studied to date are not necessarily representative of populations with substantial migration, and relatively large and stable population sizes. Consequently, our understanding of inbreeding depression magnitude in the wild is generally biased towards estimates in insular populations with limited dispersal.

Here, we leverage long‐term data from a wild population to assess inbreeding depression across the complete life‐history of superb fairy‐wrens ( Malurus cyaneus ), a common and iconic passerine distributed broadly across south‐east Australia. Our study population has been the subject of intensive individual‐level monitoring for over three decades, with complete life‐history data for over 10,000 individuals, making it one of the most comprehensive long‐term studies of a natural population (Cockburn et al. 2003, 2021). The species and study population present a unique opportunity to address several aspects of inbreeding depression in the wild.

Firstly, the long‐term study population has been subject to extensive monitoring that quantified demography and fitness to a high level in an open population. The species is distributed broadly across south‐east Australia, inhabiting Eucalyptus woodlands and common even in urbanised environments throughout its range (Crates et al. 2012). The species is a cooperative breeder, with young males remaining on their natal territory for several years as subordinate helpers who assist a dominant pair‐bonded male and female to provision young (Cockburn, Osmond, Mulder, et al. 2008; Dunn et al. 1995) (Figure 1). Males generally remain highly philopatric, and gain dominant breeding positions via inheritance or dispersal to an immediately neighbouring territory, whereas females must disperse to gain a dominant breeding position (Cockburn et al. 2003; Cockburn, Osmond, Mulder, et al. 2008). In our study population, this is reflected in frequent dispersal of females both to and from the surrounding metapopulation, with approximately 60% of recruiting females being immigrants (Cockburn, unpublished). The population is also exposed to environmental factors, such as predation, overwinter climatic effects (Lv et al. 2023), and fluctuations in population size and social structure (Cockburn, Sims, Osmond, et al. 2008). Consequently, we are able to estimate inbreeding depression in an open population subject to the natural contexts of regular dispersal, and assess condition‐dependence of inbreeding effects in association with both social and environmental factors.

FIGURE 1.

Superb fairy‐wren (A) social structure, (B) sex‐specific life‐history and reproductive pathways and (C) inbreeding depression in fitness traits. Territorial social groups are composed of a dominant breeding pair and 0–5 subordinate male helpers, usually philopatric young from previous broods. Males will generally remain on their natal territory forming orderly age‐based queues for dominant breeding positions gained via inheritance, fissioning or dispersing to immediately neighbouring territories. Male reproductive success can be gained via two pathways, either within‐pair offspring produced after gaining a dominant breeding position, or offspring produced via competition for extra‐pair copulations. In contrast, females must disperse from their natal territories prior to recruiting into the breeding population, and will only produce offspring as a dominant breeder. We tested for inbreeding depression in several critical life‐history stages, including (i) nestling condition and survival from fledging to foraging independence (Hajduk et al. 2018), (ii) juvenile survival through annual mortality peaks in winters (Lv et al. 2023) and (iii) the survival and reproductive success of adults recruited into the breeding population. We found evidence of inbreeding depression only in nestling condition, and male extra‐group reproductive success, and some evidence of inbreeding depression in survival from fledgling to nutritional independence.

Secondly, although disentangling genetic and environmental influences on performance can be difficult in highly philopatric species, the unusual mating system of fairy‐wrens can be leveraged to decouple such correlations, and assess the contribution of sexual selection to inbreeding depression. The breeding female consistently cuckolds her mate, with approximately two‐thirds of offspring sired by extra‐pair paternity (EPP) (Cockburn et al. 2003; Double et al. 2005). Extra‐group parentage is dominated by attractive and often older males that display to their neighbours for prolonged periods before the breeding season, as they acquire the nuptial plumage that allows courtship to commence much earlier (Cockburn, Osmond, and Double 2008; Double and Cockburn 2000). Less attractive males that have not acquired breeding plumage early can gain some success pretending to be the dominant when females seek fertilisations shortly before dawn (Double and Cockburn 2003; Double and Cockburn 2000). Females will generally copulate with their social partner in order to secure the social relationship and paternal care from her social partner, allowing the social males some paternity (Dunn and Cockburn 1998). However, extra‐group paternity is most common regardless of social circumstances (Hajduk et al. 2018; Mulder et al. 1994). Females also never copulate with their social partner or helpers if he is her son (Cockburn et al. 2003; Hajduk et al. 2018). Male lifetime reproductive success is consequently a combination of both extra‐pair and within‐pair reproductive success. As gaining dominant breeding positions in male superb fairy‐wrens is primarily via inheritance through an age queue of subordinate males rather than competition (Cockburn, Osmond, Mulder, et al. 2008), we would generally not expect inbreeding depression to impact within‐pair reproductive success. The exception to this would be if fertilisation success is affected (e.g., through infertility or poor sperm performance related to inbreeding) (Calhim et al. 2011; Losdat et al. 2018; Opatová et al. 2016). Alternatively, if inbreeding impacts male quality and associated sexual selection (van Oosterhout et al. 2003) in gaining extra‐pair parentage, we would expect inbreeding depression to be more extreme in reproductive success gained via EPP.

Lastly, obligate female dispersal in this species reduces the risk of mating with nuclear family relatives, and we believe that the strong sex differences in this case are driven by the advantages of inbreeding avoidance (Cockburn et al. 2003; Hajduk et al. 2018). Importantly, despite extremely common social circumstances in which close inbreeding could occur—mothers often have sons as helpers and in the immediate area surrounding her breeding territory—both mother and sons never copulate (Cockburn et al. 2003). Males pay a high cost by forgoing copulations, as the young males living with their mother are also disadvantaged in obtaining extra‐group fertilisations because they are unable to display the plumage characteristics that are attractive to females (Cockburn, Osmond, and Double 2008). This complete avoidance of close inbreeding is known from other cooperatively breeding birds (Koenig and Haydock 2004; Leedale et al. 2020). Although most bird species show female‐biased dispersal (Greenwood 1980), populations of some species within the genus Malurus have substantial rates of female philopatry and helping at the nest. In M. elegans , there is no bias and there is a substantially greater rate of inbreeding among nuclear family members, with some evidence of costs (Brouwer et al. 2011). We suspect that these differences arise because some species are much more long‐lived (up to 17 years), and therefore any reproductive restraint in a queue for dominance becomes extremely costly (Kokko and Ekman 2002). In superb fairy‐wrens, quantification of both early and lifetime consequences of inbreeding may allow us to better understand whether inbreeding depression is a cost of close inbreeding.

Here, we leverage 30 years of data from a long‐term study population to precisely estimate survival and fitness. We combine these with genomic inbreeding estimates to explicitly test for inbreeding depression across several components of individual fitness, including social and environmental condition‐dependence in early‐life and different pathways of male reproductive success. Thus, we are able to quantify inbreeding depression in lifetime fitness of a large wild population.

2. Materials and Methods

2.1. Study Population and Field Methods

The superb fairy‐wren study population is located in the Australian National Botanic Gardens (ANBG) and surrounding bushland in Canberra, Australia (35°16 S, 149°06 E) on the traditional lands of the Ngunnawal and Ngambri people. A long‐term project has colour‐banded and regularly censused all individuals within this ~60 ha area since 1992, with some prior sampling of a smaller internal area since 1988 (Cockburn et al. 2003). Social groups maintain territories year‐round, and defend them vigorously preceding and during the breeding season (September—March). Although pairs can breed without assistance, they are often helped to defend nests and provision young by up to five subordinate male helpers (Figure 1A). As discussed in the introduction, despite social partnerships between dominant breeding pairs, the species is infamous for high rates of EPP, with approximately two‐thirds of offspring sired by males outside of social partnerships (Cockburn et al. 2003). Subordinates are often sons from previous broods who remain philopatric throughout their life, either inheriting dominant breeding vacancies on their natal territory or filling vacancies when the last male on an immediately neighbouring territory dies (Cockburn, Osmond, Mulder, et al. 2008). In contrast, young females must disperse in their first year in order to gain a breeding vacancy (Figure 1B). As the breeding season approaches, any supernumerary females must compete to gain a vacancy by either fissioning an existing territory or dispersing to a vacancy. We call this period the scramble phase, and it lasts approximately from 1 August to 15 November. Once breeding positions are gained, adult dispersal is both infrequent and usually limited to short distances (see Cockburn et al. 2003; M. C. Double et al. 2005).

We were able to summarise individual fitness through measures of body condition, survival and fecundity across different life stages. The majority of birds in the project were banded and weighed as nestlings (~80%). Survival and social status of individuals in the study area were confirmed by regular censuses aiming to sight all individuals every 2–3 days during the breeding season, and once a week otherwise. If individuals were not sighted, groups were revisited often on the same day or every few days until disappearance was validated after two weeks without sighting. Estimates of survival based on disappearance are consequently highly reliable, except for juvenile females who disappear during known dispersal windows as described above.

2.2. Inbreeding and Relatedness Estimation

To estimate genomic inbreeding for birds captured between 1988 and 2020 we sequenced samples using a reduced representation sequencing approach. Blood samples were collected when birds were captured for banding using a hypodermic needle to puncture the brachial vein, sampling blood with a heparinised capillary tube, and stored by freezing or in ethanol (see Supporting Information). Blood and DNA samples were sent to Diversity Arrays Technology (DArT Pty Ltd., Canberra, ACT, Australia) for reduced‐representation sequencing (DArTseq; Kilian et al. 2012). DNA extractions from blood samples were completed by DArT using the Macherey‐Nagel NucleoMag Blood extraction kit, following the protocol using proteinase K. For sequenced samples (n = 7164), SNPs were called from allele count data provided by DArT with our own custom genotype calling scripts applying additional coverage thresholds (see Supporting Information). Further quality filtering to remove samples and loci with poor call‐rate (individual call‐rate < 0.5, SNP call‐rate < 0.85), remove monomorphic loci, and retain only loci mapping to one location in the Malurus cyaneus genome (Peñalba et al. 2020) was completed with dartR (Gruber et al. 2018). We also further removed loci that may be unreliable due to poor detectability, allelic dropout or the presence of multilocus contigs by removing loci clearly out of Hardy–Weinberg Equilibrium (p < 1e‐20), loci with low minor allele frequency (MAF < 0.01), and loci that had a high rate of Mendelian inheritance errors (per‐locus Mendelian error rate > 10%) using PLINK (Purcell et al. 2007). After filtering, 5667 individuals were scored across 9675 autosomal loci, with a mean call‐rate of 92.6% across all individuals. Further information on quality control measures and filtering of individuals is available in Supporting Information.

We used a genomic estimate of inbreeding, F _ROH, which is the proportion of an individual's genome contained in runs of homozygosity (ROH). This measure offers a direct estimate of the proportion of the genome that may be identical‐by‐descent (Caballero et al. 2021, 2022). Alternate measures of genomic inbreeding such as F _GRM were considered; however, many relatedness‐based inbreeding metrics assume no population substructure is present (Lavanchy et al. 2024; Zilko et al. 2020). This raised issues in our dataset due to complex population substructure associated with male philopatry, which has previously been found to affect genetic parameterisation in our population (Cockburn et al. 2021; Double et al. 2005). Consequently, we used PLINK to identify ROH in the population (Purcell et al. 2007), with parameters based on recommendations from Meyermans et al. (2020) and analyses with similar SNP density (Duntsch et al. 2021; Hewett et al. 2023) (Table S1). In short, we used a sliding window of 30 SNPs and identified ROH that passed the following criteria: ROH had to include a minimum of 30 SNPs, have a minimum length of 2000 kb, and a minimum density of 1 SNP per 100 kb. ROH were excluded if they had an internal gap larger than 500 kb, or more than 5 missing SNP calls or any heterozygote calls. To calculate F _ROH for each individual, we summed the total length of identified ROH and divided this by the total length of assembled autosomal chromosomal contigs of 905.95 Mb (Peñalba et al. 2020). While our dataset SNP density and relatively conservative ROH parameterisation means we may underestimate the true distribution of ROH, F _ROH provides a good relative measure of genomic inbreeding (Lavanchy and Goudet 2023).

The population pedigree is based on individual parentage assignment using established targeted genotyping methods (Cockburn et al. 2021; DArTcap; Kilian et al. 2012). Using the pedigree, we calculated individual estimates of pedigree inbreeding (F _PED) for comparison against genomic inbreeding estimates, and calculated an additive relatedness matrix with the pedigreemm R package (R Core Team 2022; Vazquez et al. 2010) to include in animal models as specified below, accounting for trait variation that may be due to heritable components (Kruuk 2004; Wilson et al. 2010). An additive matrix was used because genomic relatedness matrices may also be affected by complex population substructure in our study population. Within the modelling framework described below, we calculated heritability from posterior distributions of variance associated with random effects to allow calculation of associated error (Martin and Videlier 2024).

2.3. Statistical Methods—Early‐Life Fitness

2.3.1. Nestling Weight and Survival to Independence

We assessed the effect of inbreeding on nestling mass and survival from fledging to foraging independence. Nestling weight is an indicator of early body condition and can have consequent impacts on early life survival (Hajduk et al. 2018). We define foraging independence as being 28 days after fledging, as the first records of dispersal occur in the fifth week out of the nest. Mortality during this period from fledging to independence is high (~40% of fledged nestlings), the majority of which occurs in the first days of fledging when predators actively hunt fledglings that beg incessantly and have only weak flight. We used only individuals who were banded and weighed as 5–8 day old nestlings from 1991 to 2020 to assess nestling mass (n = 5243) and a subset of these who successfully fledged to evaluate survival from fledging to foraging independence (n = 3931). We fitted a Gaussian general linear‐mixed effects animal model for nestling mass and a Bernoulli general linear‐mixed effects animal model for binary survival in brms in R (Bürkner 2017; R Core Team 2022) (see Table S2 and Supporting Information for further description of models and associated parameters, including priors).

2.3.1.1. Fixed Effects

In the model of nestling weight, genomic inbreeding (F _ROH) was included to estimate inbreeding depression. Variability in number of helpers in this cooperatively breeding species is also important for maternal investment via load‐lightening and group defence, and consequently may impact early life fitness (Langmore et al. 2016; Russell et al. 2007). Here we include number of helpers as a three‐level factor (0, 1, ≥ 2) as instances of two or more helpers are relatively rare (median number of helpers = 0, range = 0–5; 15% of nestlings had ≥ 2 helpers) (Hajduk et al. 2018). We also include brood size (median brood size = 3, range = 1–5). Increasing brood size may have a negative effect on nestling mass due to within‐brood resource competition as noted previously in this species (Hajduk et al. 2018), though as superb fairy‐wrens do not exhibit hatching asynchrony, such effects are expected to be distributed evenly across the brood. Sex was also included to account for differences between males and females. Sex was established using a subset of targeted genotyping markers (Cockburn et al. 2021), validated using sexually dimorphic adult plumage characteristics where observed. In the nestling mass model, we included nestling age in days at banding as a quadratic expression to account for growth during this key developmental phase. We also include a factor to account for a change in weighing protocols from 1992 onwards (pre‐1992, 1992+). We also fitted interactions of inbreeding with sex and number of helpers to account for possible sex‐bias or social environment buffering inbreeding; however, these were removed from the final model as their effects were not significant in exploratory models (Table S3).

In the model of survival from fledging to independence, we include all fixed effects and interactions as described for the model of nestling weight, except for weighing protocols and nestling age, and similarly tested for interactions of inbreeding which were excluded from the final model as their effects were not significant (Table S3). We ran two models, with and without nestling weight adjusted for nestling age and weighing protocol to account for the influence of weight variation on early life survival (Hajduk et al. 2018).

2.3.1.2. Random Effects

We fitted nest ID (within‐brood variation), maternal ID, cohort (multilevel factor 1991–2020 accounting for interannual variation; for example, the 2020 cohort being birds hatched September 2020 through March 2021), hatch fortnight (multilevel factor 1–14 accounting for intra‐annual variation, splitting the breeding season into fortnight intervals from 9 September to 15 March), and pedigree‐based additive genetic effect (Kruuk 2004) as random effects in both models to account for shared and heritable variation due to these factors that may otherwise bias estimates of inbreeding depression (Becker et al. 2016). Paternal ID was omitted from these models as it does not fully and truly account for male parental effects, due to the prevalence of extra‐pair paternity in the species, the presence of helpers, and the change in social father contribution in the presence of helpers.

2.3.2. Overwinter Mortality

After the very heavy mortality of fledglings, mortality is limited until a major peak that occurs over the first winter of life. This peak is correlated with adverse environmental conditions and affects all sex and age classes (Lv et al. 2023). Consequently, we investigated inbreeding depression in juvenile survival over their first winter nonbreeding period from 1 April to 1 August following their hatch season (as in Eastwood et al. 2023), a period in which dispersal is negligible. We fitted separate models for two groups: individuals from the core study area (n = 1887; including males and both females who remain on their natal territory overwinter and any early disperser females who internally dispersed within the core study area), and female immigrants who arrive during the early dispersal window prior to 1 April (n = 196). We fitted a Bernoulli general linear‐mixed effects animal model of binary survival from 1 April through to 1 August for each group.

2.3.2.1. Fixed Effects

As for nestling fitness, we fitted genomic inbreeding, sex, and an interaction term of sex and inbreeding. Sex was not included in the model including only immigrant females. Overwinter survival is known to be influenced by climatic variables. Consequently, we fitted mean maximum and mean minimum temperature over the period of 1 April to 1 August, based on the Australian Bureau of Meteorology records for Canberra Airport (approximately 8 km from the study site; http://www.bom.gov.au/climate/data). We also fitted interaction terms of each of these with inbreeding to assess whether environmental conditions may enhance or mask inbreeding depression.

2.3.2.2. Random Effects

We accounted for cohort, hatch fortnight, nest ID, maternal ID and additive genetic effect. As nest, maternal ID, and hatch fortnight are unknown for immigrant females, these effects were excluded from that model.

2.4. Statistical Methods—Adult & Lifetime Fitness

For individuals surviving to recruitment at the end of the scramble phase (i.e., on 15 November in the year following their hatch season), we assessed the effect of inbreeding depression on lifespan and lifetime reproductive success. Due to female‐biased dispersal, we estimate lifespan and lifetime reproductive success only in recruiting females who internally disperse within the core study area or immigrate; we cannot distinguish between mortality and dispersal of females who disappear prior to recruitment, and so estimates of lifetime fitness are unknown. We separately assessed the effect of inbreeding on lifetime reproductive success for all males (i.e., not just those who successfully recruit) as a measure of ‘whole’ fitness. For all three traits, we only considered cohorts where all individuals had died and so had complete measures of lifespan and lifetime reproductive success. As a result, cohorts from 2016 onwards (where some birds were still living) were excluded as these would include only short‐lived birds and bias the fitness of these cohorts towards short lifespan and smaller numbers of offspring. For both adult traits, we fit log‐normal linear‐mixed effect animal models (see Table S2 and Supporting Information for further description of models and associated parameters, including priors).

2.4.1. Lifespan

We estimated lifespan based on hatch date and disappearance from the population following regular censusing described in the Field Methods section. In order to ensure accuracy and completeness of lifespan measures, only individuals banded as nestlings who survived to recruitment were included in modelling inbreeding depression in lifespan (n = 920). We included genomic inbreeding, sex and their interaction as fixed effects, and cohort and additive genetic effect as random effects.

2.4.2. Lifetime Reproductive Success in Adults

We quantified lifetime reproductive success (LRS) as the number of banded offspring who survive to nutritional independence (41 days). We use this metric to reduce variation associated with high offspring mortality that does not necessarily reflect parent fitness, particularly from whole‐of‐brood predation and high rates of predation post‐fledging. We quantified LRS for individuals who survived to recruitment, including those banded as nestlings, fledglings and juveniles. As the pathways to reproductive success differ substantially between males and females (Cockburn et al. 2003), we fitted separate models for each sex. In both sexes, the number of breeding seasons in which individuals are alive is highly correlated with reproductive success (Figure S2; in females, Pearson correlation r = 0.79, df = 378, p < 0.001; in males, Pearson correlation r = 0.69, df = 770, p < 0.001). To account for this, we included breeding span, calculated as the number of breeding seasons during which an individual was alive as a potential breeder.

2.4.2.1. Female LRS

For females (n = 380), we included genomic inbreeding and breeding span as fixed effects, and cohort and additive genetic effect as random effects.

2.4.2.2. Male LRS

We explicitly tested whether inbreeding affected parentage pathways available to males via within‐pair paternity and extra‐pair paternity. Extra‐pair paternity can either be extra‐group (i.e., with a breeding female outside of a males own social group), or within‐group as a subordinate helper with the breeding females in their social group (Hajduk et al. 2021). Cases of extra‐pair within‐group offspring are rare (approximately 2.5% of nestlings), as many subordinate males are philopatric offspring of the breeding female, and the mother never mates with her son. We therefore included offspring produced by this mechanism within the extra‐pair classification. We tested for inbreeding depression in male LRS by fitting two separate models, for extra‐pair offspring of all recruited males (n = 772) and within‐pair offspring of males who gained dominance (n = 515). We included genomic inbreeding as a fixed effect, and cohort and additive genetic effect as random effects. We also fitted a fixed effect of breeding span when modelling extra‐pair paternity, and social span (the number of breeding seasons holding a dominant breeding position) when modelling within‐pair paternity. As dominant breeding positions in male superb fairy‐wrens are gained primarily via inheritance through an age queue of subordinate males (Cockburn, Osmond, Mulder, et al. 2008), we accounted for variation in opportunity for within‐pair paternity by fitting social span, but do not expect dominance acquisition itself to be a trait in which inbreeding depression is expressed. This is because dominance acquisition is a trait determined by the death of existing dominant males, and the age of other subordinate males; dominant positions are always acquired by the oldest subordinate male, or in the cases where all males on a territory have died or there are no subordinate helpers on that territory, a neighbouring subordinate male.

2.4.3. Lifetime Reproductive Success as a Measure of Whole Fitness

In addition to our specific models addressing reproductive success in adults that survived to recruitment, we tested whether inbreeding affected total lifetime reproductive success in all individuals including those who did not survive to recruitment. As with other adult LRS models, we quantified lifetime reproductive success (LRS) as the number of banded offspring who survive to nutritional independence (41 days). We fitted zero‐inflated Poisson models separately for each sex, and included genomic inbreeding as a fixed effect for both components of the model, and cohort and additive genetic effect as random effects. In these models, we did not include breeding span and social span. We included individuals banded as nestlings, fledglings and juveniles (males n = 2584). For females, we specifically excluded females where we cannot distinguish between mortality and dispersal in key windows (see Supporting Information for more details) (n = 1033).

2.5. Immigrant Genetic Contributions

Although immigrants arrive in the population each year, they are only genetically significant if they make lasting contributions to the population. In order to assess this, we quantified immigrant genetic contributions using pedigree‐based methods described by Hunter et al. (2019) and similarly to Chen et al. (2019). For this purpose, we described the population bi‐annually, taking a snapshot of birds alive on 15 November (post scramble‐phase) and 1 April (end of breeding season) each year between 1992 and 2021, similarly to yearly post‐breeding census snapshots used to capture individual genetic contributions to extant populations in Reid et al. (2019). We described each focal cohort of immigrants based on the year they arrive to the population (e.g., any individual appearing between 15 November 1992 and 15 November 1993 would be classified within the 1992 immigrant cohort). For each subsequent cohort of immigrant females, we calculated the expected genetic contribution of that cohort to each bi‐annual census population based on an additive genetic relatedness matrix, created using a trimmed pedigree of the focal set of immigrants and their descendants. The relatedness of the focal cohort to all descendants alive at a given bi‐annual census point were summed and divided by the total population size at that time point, to give the expected genetic contribution of the focal cohort to the population at that point (i.e., the expected proportion of genomes in the current population that originate from immigrant females). Further details and methods used to apply the Hunter et al. (2019) approach to our population are provided in Supporting Information.

3. Results

3.1. Inbreeding Estimates

We identified 33,464 runs of homozygosity in the individuals included in our analyses (n = 5667). The average ROH length was 2988 kb (95% percentiles = 2022–6430 kb). The longest ROH identified was 26 Mb and occurred in an individual with known pedigree inbreeding (F _PED = 0.25), though we note that due to SNP density in our dataset, longer ROH may not have been identifiable in some genomic regions, or broken into multiple short ROH due to internal gaps between SNP locations. The proportion of the genome in ROH (F _ROH) ranged from 0% to 20% (mean F _ROH = 1.9%, 95% percentiles = 0.4%–4.6%). Individual F _ROH was well correlated with pedigree inbreeding (Figure 2; for individuals with at least three grandparents known, Pearson correlation r = 0.68, df = 1597, p < 0.001). A number of individuals with no known pedigree‐based inbreeding had elevated values of F _ROH, suggesting that there may be inbreeding that pedigree methods have previously not detected in this population. This is likely attributable to immigrant females, who make up approximately 60% of recruiting females in our adult fitness analyses, and likely disperse from the immediate broader population surrounding the study site. Immigrant females are important contributors to the study population, with the expected genetic contributions of immigrant cohorts to the population in 2021 totalling approximately 55% (Figure S3). While a few immigrants had high genomic inbreeding estimates (12/314 genotyped immigrant females had F _ROH ≥ 0.04; Figure S3), all except two of these relatively inbred immigrants did recruit and breed (10/12; 83%).

FIGURE 2.

Pairwise correlation between pedigree and genomic estimate of inbreeding for individuals included in modelling with at least three grandparents known (n = 1599; Pearson correlation r = 0.68, df = 1597, p < 0.001). Boxplots show F _ROH (genomic inbreeding estimated based on runs of homozygosity) distribution. Because F _PED (pedigree‐based inbreeding) estimates are not continuously distributed and low values are disproportionally represented in this dataset, they are binned into categories (F _PED = 0, 0–0.025, 0.025–0.05, 0.05–0.10, 0.10–0.15, 0.15–0.2, and F _PED = 0.25, upper bound inclusive). Numbers above boxplots indicate sample size within each bin. F _ROH estimates can be interpreted as the proportion of the genome contained in runs of homozygosity, and so can take values ranging from 0 to 1, and the range extending here to similar bounds as pedigree‐based inbreeding is coincidental. Genomic methods of inbreeding recover the pattern of known inbreeding based on genealogy well.

3.2. Inbreeding Depression in Early Life

There was evidence for negative effects of inbreeding on nestling weight which appear to partially mediate inbreeding depression in survival to independence subsequent to the nestling phase (Table 1; Figure 3). Given the real distribution of inbreeding in our nestlings, an average inbred individual (F _ROH = 0.019) would weigh approximately 0.47 g more than the most inbred individual (F _ROH = 0.204). Beyond inbreeding effects, nestling weight was on average 7 g, was significantly higher in males, showed a quadratic increase with age at banding, increased with the presence of subordinate helper(s), increased if measured with a previous weighing protocol (associated with chicks being weighed later in the day), and decreased with brood size. Probability of survival to nutritional independence increased with weight (adjusted for age and weighing protocol), such that a 1 g increase in weight on average increased probability of survival by approximately 28% (95% CI = 5%–57%). However, the effect of inbreeding on survival to nutritional independence was marginally non‐significant when including weight (F _ROH posterior mean = −0.131, 95% CI = −0.269 to 0.004) and marginally significant when excluding weight (F _ROH posterior mean = −0.133, 95% CI = −0.268 to −0.003) (Table 1). In both weight and fledgling survival, we found no evidence that inbreeding depression was sex‐specific or mediated by social environment. The heritability of nestling weight was estimated as 12.8% (95% CI = 8.1%–17.8%). Together, these results are largely consistent with a previous study of inbreeding depression in these traits using pedigree‐based estimates of inbreeding (Hajduk et al. 2018).

TABLE 1.

Effect of genomic inbreeding on (A) nestling weight and (B, C) survival from fledging to nutritional independence. Estimates presented are the posterior means from Bayesian models, and associated 95% credible values. Bolding indicates where the 95% credible intervals do not span zero.

Fixed effects	(A) Nestling mass		(B) Survival from fledging to 41 days (with weight)		(C) Survival from fledging to 41 days (no weight)
	Posterior mean (95% CI)		Posterior mean (95% CI)		Posterior mean (95% CI)
Intercept	6.912	(6.814, 7.011)	0.652	(−0.321, 1.564)	0.623	(−0.348, 1.543)
Inbreeding coefficient (F ROH)	−0.035	(−0.054, −0.016)	−0.131	(−0.269, 0.004)	−0.133	(−0.268, −0.003)
Sex (male)	0.159	(0.126, 0.192)	0.128	(−0.107, 0.369)	0.161	(−0.069, 0.393)
Helpers (0, 1, 2+)
1	0.071	(0.006, 0.137)	0.244	(−0.149, 0.653)	0.256	(−0.137, 0.659)
2+	0.095	(0.013, 0.178)	0.266	(−0.235, 0.773)	0.289	(−0.196, 0.792)
Brood size	−0.042	(−0.069, −0.015)	−0.123	(−0.300, 0.048)	−0.134	(−0.306, 0.031)
Nestling age (days)	2.117	(1.841, 2.389)
Nestling age squared	−1.418	(−1.689, −1.143)
Weight protocol (pre‐1992)	0.549	(0.126, 0.973)
Weight—adjusted			0.180	(0.041, 0.330)

Random effects	Posterior mean (95% CI)		Posterior mean (95% CI)		Posterior mean (95% CI)
Cohort	0.156	(0.106, 0.223)	1.344	(0.869, 2.023)	1.325	(0.867, 1.955)
Hatch fortnight	0.091	(0.043, 0.169)	1.226	(0.719, 2.051)	1.225	(0.717, 2.065)
Additive genetic effect	0.272	(0.217, 0.323)	1.220	(0.407, 2.060)	1.155	(0.355, 1.880)
Maternal ID	0.048	(0.002, 0.121)	0.506	(0.036, 0.958)	0.501	(0.037, 0.955)
Nest ID	0.511	(0.487, 0.537)	2.220	(1.849, 2.724)	2.185	(1.828, 2.619)
Residual variance	0.457	(0.434, 0.478)	N/A		N/A
Sample size	5243		3931			3931

FIGURE 3.

Negative effects of genomic inbreeding on (A) nestling weight, and (B) effect of nestling weight on survival from fledging to nutritional independence. Black circles, error bars and numbers represent mean, standard error and sample size of weight and survival data for groups of individuals binned based on genomic inbreeding values (0.025 intervals from 0 to 0.175, until > 0.175), or nestling weight (2–4 then every 0.5 until 9–11). Black lines and grey ribbons indicate mean conditional effects and 95% credible intervals for (A) inbreeding coefficient averaged across sexes, and (B) nestling weight. Weight data presented in both panels has been adjusted for nestling age and weighing protocol. Model fit and large credible intervals in (B) resulted from plotting mean conditional effects and binary trait distribution. Genomic inbreeding estimates for fitted models and raw data have been projected back to their original scale. Inbreeding depression affects nestling weight, which may have some consequences for survival to independence via the effect of weight.

We found no evidence of inbreeding depression in survival in first overwinter (Table S4). We found a significant negative effect of mean maximum temperature on survival in the overwinter period consistent with prior research (Lv et al. 2023, 2026), however, there was no support for effects of inbreeding being worse in adverse environmental conditions (F _ROH × T _MAX posterior mean = 0.02, 95% CI = −0.14 to 0.18; Table S4).

3.3. Inbreeding Depression in Adult & Lifetime Fitness

There was no evidence that inbreeding depression affected total lifespan in either sex (F _ROH posterior mean = −0.01, 95% CI = −0.05 to 0.03; Table S5). In both sexes, lifetime reproductive success unsurprisingly increased with lifespan (Table 2; Figure S2). In males there was a negative effect of F _ROH inbreeding on the number of extra‐pair offspring sired, but not within‐pair offspring (Figure 4A,B). Given the real distribution of inbreeding in our recruited males (mean F _ROH = 1.9%, 95% percentile = 0.3%–4.1%), this would result in a reduction of male extra‐pair LRS from approximately 1 offspring for the most outbred individual (F _ROH = 0) to no offspring for the most inbred individual (F _ROH = 0.11). However, there was no support for inbreeding depression in female reproductive success (Table 2; Figure 4C).

TABLE 2.

Sex‐specific effects of genomic inbreeding on lifetime reproductive success in recruited adults, via (A) male extra‐pair paternity; (B) male within‐pair paternity and (C) female lifetime reproductive success. While female breeding success and male extra‐pair success generally commence at 1 year of age, male within‐pair success only becomes available once they inherit a dominant breeding position (i.e., through orderly age‐based queues, and given that their own mother has died or left the territory). Estimates presented are the posterior means from Bayesian models, and associated 95% credible values. Bolding indicates where the 95% credible intervals do not span zero.

Fixed effects	Males						Females
	(A) Extra‐pair offspring			(B) Within‐pair offspring			(C) Offspring
	Posterior mean (95% CI)			Posterior mean (95% CI)			Posterior mean (95% CI)
Intercept	0.567	(0.493, 0.638)		0.688	(0.621, 0.752)		1.161	(1.049, 1.271)
Inbreeding coefficient (F ROH)	−0.066	(−0.112, −0.021)		−0.025	(−0.081, 0.029)		0.005	(−0.063, 0.074)
Breeding span (log‐years)	0.517	(0.47, 0.563)					0.694	(0.626, 0.762)
Social span (log‐years)				0.380	(0.326, 0.433)

Random effects	Posterior mean (95% CI)			Posterior mean (95% CI)			Posterior mean (95% CI)
Cohort	0.051	(0.003, 0.126)		0.062	(0.003, 0.155)		0.187	(0.075, 0.310)
Additive genetic effect	0.313	(0.216, 0.404)		0.121	(0.006, 0.279)		0.275	(0.017, 0.564)
Residual variance	0.568	(0.516, 0.618)		0.602	(0.550, 0.648)		0.576	(0.379, 0.683)
Sample size	772			515			380

FIGURE 4.

Sex‐specific effects of genomic inbreeding on lifetime reproductive success, with negative effects in males (blue) only via (A) extra‐pair paternity but not (B) within‐pair paternity and (C) no effect on female (red) lifetime reproductive success. Lifetime reproductive success is defined as the number of offspring surviving to nutritional independence. Axes have been projected back to their original scale. Circles represent raw data, and the lines and shading indicate the fitted linear mixed effects model for each trait and associated 95% credible intervals in each sex. Inbreeding depression only affects male extra‐pair reproductive success, which may be under strong sexual selection in this species.

We additionally tested for and found no effect of inbreeding on total lifetime reproductive success in all individuals including those who did not survive to recruitment in either sex. Genomic inbreeding did not significantly affect either whether individuals had offspring or the number of offspring (Table S6).

4. Discussion

The impact of inbreeding depression in large wild populations remains poorly understood. Although conservation biologists have been concerned with inbreeding in very small populations (Frankham 2005; Keller and Waller 2002; Spielman et al. 2004), estimates from healthier populations have often been conducted on island populations where migration is rare. Limited migration and small population sizes generally have strong effects on inbreeding dynamics and purging of deleterious variation, and are thus not necessarily representative of inbreeding depression in populations with more immigration and emigration (Boakes et al. 2007; Caballero et al. 2017; Hedrick and Garcia‐Dorado 2016). However, studies in these contexts have previously been restricted to pedigree‐based methods unable to estimate inbreeding in both immigrants and founders (Dickel et al. 2021; Kardos et al. 2016; Keller and Waller 2002). Genomic data offer methods to directly estimate realised inbreeding without intensive multigenerational study, and importantly, for both founders and immigrants.

Here, we apply genomic methods to a superb fairy‐wren population which has been subject to multidecadal study of fitness variation. The population has sex‐biased dispersal with male philopatry, but obligate female natal dispersal, often over considerable distances (Cockburn et al. 2003; Cockburn, Osmond, Mulder, et al. 2008), resulting in many immigrant females of unknown ancestry. In general, our results support the suitability of genomic methods for characterising inbreeding in large wild populations, as we recovered the pattern of inbreeding estimates and inbreeding depression in early‐life traits based on genealogy previously established in this population (Hajduk et al. 2018), and were able to identify inbred individuals among the annual recruitment of immigrant females and founders. We obtained measures of inbreeding depression for traits across three life‐history phases, capturing both early‐life condition and survival through key mortality peaks, and adult fitness in recruiting adults and lifetime fitness: (a) nestling condition and survival from fledging to foraging independence, (b) juvenile survival through annual mortality peaks in winters, (c) the survival and reproductive success of adults recruited into the breeding population and (d) lifetime reproductive success. In general, inbreeding depression was low or negligible for many of these traits, including lifetime fitness.

4.1. Inbreeding Depression in Early‐Life

We found evidence for inbreeding depression in nestling weight, and some effect of inbreeding on survival from fledging to independence when not accounting for weight. Nestling weight is an important indicator of body condition in early life and can have downstream effects on early‐life survival (Hajduk et al. 2018), as we found in our analyses; an increase in weight of 0.73 g on average results in an approximately 20% increase in probability of survival from fledging to independence. Given that our analyses suggest inbreeding reduces nestling weight and survival to independence when not accounting for nestling weight, it appears that inbreeding depression in survival to independence may be partly due to effects of inbreeding on weight, or longer‐term costs of growth compensation (Birkhead et al. 1999; Hewett et al. 2026; Johnsson and Bohlin 2006). Here, our results are largely in line with inbreeding effects on nestling body condition which have been documented in several bird species, though consequent survival impacts vary (Chen et al. 2016; Hewett et al. 2026; Kruuk et al. 2002; Lichtenauer et al. 2019). While we are unable to explicitly test for earlier inbreeding effects in embryo or early development (Bensch et al. 1994; Daniels and Walters 2000; Foster et al. 2021; Kempenaers et al. 1996; Kingma et al. 2013; Kruuk et al. 2002), hatching failure is not more common in this population than in other songbirds and is not prevalent in known incestuous pairings (Cockburn pers. comm.). Importantly, applying genomic methods also largely recovered the pattern of inbreeding depression in these early‐life traits based on genealogy previously established in this population (Hajduk et al. 2018).

We also explicitly tested for condition‐dependence of inbreeding depression via social environment in fledgling traits, and environmental conditions in juvenile overwinter survival, but found no evidence to support condition‐dependent inbreeding effects. While subordinate helpers in this species may help with provisioning and group defence (Dunn et al. 1995), dominant breeders load‐lighten such that increasing the number of helpers does not lead to a related increase in corporate provisioning (Cockburn, Sims, Osmond, et al. 2008; Russell et al. 2007). To our knowledge, cooperatively breeding meerkats are the only species where there is evidence that social environment can buffer inbreeding depression, as the number of subordinate helpers improves various pup traits including growth and survival and to some extent negate the effect of inbreeding on pup mass (Nielsen et al. 2012; Russell et al. 2002). Consequently, load‐lightening behaviour in superb fairy‐wrens may explain why social environment does not buffer inbreeding depression in nestling weight in our population, though we do find that the presence of helpers improves nestling weight. However, such social environment effects may also be confounded with other factors such as habitat quality, as higher quality territories are more likely to be productive and thus have more helpers (Backhouse et al. 2023; Cockburn, Sims, Osmond, et al. 2008). We also found no evidence of inbreeding depression in first overwinter survival, including no significant environmental interactions despite strong environmental mortality effects, which are linked to higher maximum temperatures in winter and associated metabolic and resource costs (Lv et al. 2023, 2026). Overall, our results do not suggest overall or condition‐dependent inbreeding affecting fledgling or first overwinter survival in our population. These findings are in line with prior studies finding inbreeding depression by environment interactions in wild populations were rare compared to experimental contexts (Pemberton et al. 2017).

4.2. Inbreeding Depression in Adult & Lifetime Fitness

Inbreeding did not affect adult fitness in most cases. Female adult performance in both survival and reproductive success did not exhibit significant inbreeding depression. Survival and within‐group reproductive success of adult males was similarly unaffected by inbreeding depression. In both cases, variation in reproductive success was largely explained by the number of breeding seasons an individual was alive as a potential breeder. As we found no effect of inbreeding on lifespan, there is also no support for indirect effects of inbreeding on lifetime reproductive success via reduced lifespan as seen in another Malurus species. We additionally found no support for inbreeding depression in lifetime reproductive success in either sex even when including individuals that did not survive to recruitment. Our results are also in contrast to theory and studies predicting accumulation of inbreeding effects over individual lifetime resulting in reduced lifespan or accelerated senescence processes (Beccardi et al. 2024; Charlesworth and Hughes 1996; de Boer et al. 2018; Keller et al. 2008).

Although male survival and within‐pair reproductive success were not affected by inbreeding, the success of males at obtaining extra‐group parentage did exhibit inbreeding depression. The absence of inbreeding depression in male within‐pair success is not entirely unexpected, given that usually dominance acquisition is determined by stochastic and external processes, not inherent male quality—males usually gain dominance via inheritance in ordered queue systems (Cockburn, Osmond, Mulder, et al. 2008), fission, or dispersal to neighbouring territories (Double et al. 2005; Mulder 1995). In contrast, male success in gaining extra‐pair parentage is highly variable, and the dynamics of this trait are influenced by strong sexual selection. Once higher quality males become sufficiently old, they are capable of moulting into nuptial plumage months before the breeding season and immediately commence displays to females in their neighbourhood. Early moulting males are much more likely to attract extra‐group females (Cockburn, Osmond, and Double 2008; Dunn and Cockburn 1999). However, moulting and maintaining early nuptial plumage appears to be costly and can be impossible during cold, dry winters (Cockburn, Osmond, and Double 2008). Even when conditions are good, most males can achieve early moult only three times before they die. Despite these strong environmental effects, males can first attempt early moult as young as three years of age, while some never moult early, and others defer until late life. Such variations mean that male prowess does not depend entirely on lifespan, and the individual differences are likely to have both genetic and environmental causes (Cockburn et al. 2016; Cooper et al. 2021). Consequently, if male quality signals are affected by inbreeding as has been found in some other species (Bolund et al. 2010; Joron and Brakefield 2003; van Oosterhout et al. 2003), sexual selection via female choice of extra‐pair partners may lead to inbreeding depression in this trait. We found that inbreeding depression does have a small but significant negative effect on male extra‐pair reproductive success in our population, which could suggest genetic components underlying sexual selection in gaining extra‐group reproductive success.

In addition to male‐specific processes, it is possible that the lack of inbreeding depression in female reproductive success may be due to differences in life‐history between the two sexes. In general, females may be more sensitive to inbreeding depression (Vega‐Trejo et al. 2022), particularly when investment and cost of survival or reproduction are imbalanced (Kokko and Jennions 2003). Dispersal is female‐biased in superb fairy‐wrens (Cockburn et al. 2003; Mulder 1995). It is possible that females who disperse and successfully gain breeding vacancies have already passed some selection hurdle imposed by dispersal costs (Hansson et al. 2004) or intra‐sexual competition for vacancies (Fan et al. 2018). While we found no significant differences in the distribution of inbreeding between recruited immigrant females and the base population (Figure 2B), it is challenging to quantify long‐term dispersal costs directly. In our system, purging of inbred individuals during dispersal phases may be difficult to detect due to the porous nature of our study boundary and unknown dispersal distances; fates of emigrant females leaving within dispersal windows are unknown, and immigrants have unknown dispersal distances. Further assessment of inbreeding depression in dispersal success in both immigrant females and internal dispersers within the study area may elucidate whether there is any evidence for inbreeding depression in such traits that could explain the lack of inbreeding effects in female reproductive success.

4.3. Limited Inbreeding Effects in a Large Open Population

Recent studies that use genomic methods to quantify inbreeding depression in wild vertebrate populations vary in their findings of inbreeding effects over various life stages (Table 3). However, almost all studies reported inbreeding depression in some trait and frequently in lifetime fitness measures. In part, this is likely a result of publication bias in both the context of positive results and characteristics of populations that have sufficient data to address these questions (L. F. Keller and Waller 2002); notably, all studies in Table 3 leveraged at least 5 years of population monitoring or sampling, and were often part of larger long‐term projects that likely benefit from conservation interests or monitoring small manageable populations. However, populations with a history of decline, bottlenecks, isolation or ongoing small effective population sizes, will experience population dynamics such as limited migration and small population sizes, which generally have strong effects on inbreeding dynamics and purging of deleterious variation (Boakes et al. 2007; Caballero et al. 2017; Hedrick and Garcia‐Dorado 2016). Based on studies to date, these population history characteristics appear to often lead to a high degree of inbreeding load, the effects of which may be more easily detectable due to both strong inbreeding depression and a higher frequency of inbreeding events in those populations.

TABLE 3.

Examples of recent studies using genomic methods to quantify inbreeding depression in wild vertebrate populations. Studies included used genomic inbreeding estimation (excluding studies using microsatellite data) to quantify inbreeding depression across life‐history stages, including components of adult fitness, in free‐living populations. Studies are separated into subsections based on population characteristics, including (A) large, open populations, (B) insular populations and (C) endangered or small populations.

Species	Population & data characteristics	Genomic data type	Genomic inbreeding a	Trait(s) assessed	Findings	References
(A) Large, open populations
Caribou (Rangifer tarandus)	Migratory species experiencing ongoing declines, one of focal populations estimated ~8900 in 2016 (99% decline since 1990s) Long‐term study (1996–2016)	ddRADSeq n ind = 400 n loc = 22,073	sMLH & F GRM	Body mass Annual adult survival	No significant inbreeding depression with either metric in any trait after model selection	(Gagnon et al. 2019)
Leach's storm‐petrel (Hydrobates leucorhous)	Large outbreeding species. (2010–2015)	ddRADseq n ind = 560 n loc = 2514	IR	Breeding pair success		(Sin et al. 2021)
				Adult morphology
				Chick morphology	* [maternal IR only]
				Chick growth rate	* [maternal IR only]
(B) Insular or structured populations
Red deer (Cervus elaphus)	Island population, with some migration across broader population on Isle of Rum Long‐term study (1981–2013)	Cervine SNP chip n ind = 2254 n loc = 37,410	F GRM	Birth date		(Huisman et al. 2016)
				Birth weight	*
				First summer survival
				First winter survival	*
				Age 1–2 years survival	*
				Age 0–2 years survival	*
				Female age at first reproduction
				Adult annual survival
				Adult annual breeding success	*
				Female LRS	*
Soay sheep (Ovis aries)	Isolated island population on St Kilda, historical bottlenecks Long‐term study (1985–2016)	Ovine SNP chip n ind = 5805 n loc = 37,037	F HOM, F GRM, F ROH	Birth weight	*[only maternal]	(Bérénos et al. 2016)
				Juvenile hindleg	*[only ind]
				Juvenile weight	*[ind & maternal]
				First‐year survival	*[male only, FGRM & FROH]
				Annual survival	*[male all, female FHOM & FGRM]
				ABS	*[male only, FGRM]
Harbour seals (Phoca vitulina)	Part of broader metapopulation with existing strong substructure. (1997–2003)	ddRADseq n ind = 60 n loc = 14,585	sMLH	Longevity [binary young/old]	* [see paper for outlier effect]	(Hoffman et al. 2014)
				Parasite infection	*
House sparrows (Passer domesticus)	Island metapopulation, local Ne ranges 10–100 (1998–2013)	Custom SNP array n ind = 3116 n loc = 181,529	F GRM & F ROH	Adult morphology	* [not in all traits assessed]	(Niskanen et al. 2020)
				Survival	*
				ARS	*
				LRS	*
(C) Endangered or threatened populations
Leadbeaters possum (Gymnobelideus leadbeateri)	Critically Endangered, historical and ongoing declines. Focal pop < 40 individuals in 2014. (1996–2001 & 2010–2017)	ddRADseq (DArTseq) n ind = 287 n loc = 1575	F H & F GRM	Survival to maturity	*	(Zilko et al. 2020)
				Breeding success	*
				Lifespan	*
				LRS	* [FGRM only]
Florida scrub‐jays (Aphelocoma coerulescens)	Threatened, ongoing declines and decreased immigration. Study population ~263 families in 2016, migration to regional population. (1995–2013)	Custom SNP array n ind = 3583 n loc = 7834	Pairwise IBD & F GRM	Hatching success	* [parental relatedness]	(Chen et al. 2016)
				Nestling weight	*
				Juvenile survival (indep to 1yr)	*
				Breeder lifespan	*
				LRS	* [females only]
Hihi (Notiomystis cincta)	Threatened, remnant islands population with largest population ~192, some managed migration (2001–2015)	Custom SNP array n ind = 363 n loc = 41,195	F ROH‐MOD	Juvenile survival	*[worse in females]	(Duntsch et al. 2023)
				Adult annual survival	No adult ID
				ARS
				LRS
Kākāpō (Strigops habroptilus)	Critically Endangered, < 250 individuals remaining, island populations with severe bottlenecks	Both WGS and RRS n ind = 148 n loc‐WGS = 1,552,757 n loc‐RRS = 19,341		Clutch size		(Foster et al. 2026)
				Egg volume
				Hatching success of females	*
Helmeted honeyeater (Lichenostomus melanops cassidix)	Critically Endangered, < 250 individuals remaining Long‐term study (1986–2018)	ddRADSeq (DArTseq) n ind = 123 n loc = 11,059	F HOM	LRS	*	(Harrisson et al. 2019)
				Lifespan	* [stronger in females]
Purple‐crowned fairy‐wren (Malurus coronatus coronatus)	Endangered subspecies with relatively fragmented but connected subpopulations. Long‐term study (2004–2022)	ddRADSeq (DArTseq) n ind = 1970 n loc = 5848	F GRM	Survival from fledging to 6 months		(Evans et al. 2026)
				Survival from 6 months to 1 year
				Lifespan	*
				LRS	* male‐specific effects

^a F _HOM = proportion of individual genome in homozygous state. F _H = PLINK method of moments based on observed and expected homozygosity (Purcell et al. 2007). F _GRM = GCTA allelic similarity between gametes, giving more weight to homozygous rare alleles (Yang et al. 2011). F _ROH = PLINK or other rule‐based ROH classification (Purcell et al. 2007). F _ROH‐MOD = ROH classification using model‐based approach often based on genotype‐likelihoods, such as RZooRoH (Bertrand et al. 2019). sMLH = standardised multi‐locus heterozygosity. MLH = multi‐locus heterozygosity. Pairwise IBD = pairwise coancestry via R package related (Wang 2011; Pew et al. 2015). IR = internal relatedness (Amos et al. 2001). F _IS = inbreeding estimation via dartR R package (Gruber et al. 2018).

The absence of strong inbreeding depression over life history through juvenile stages and in adult survival and reproduction—excluding male extra‐pair parentage—in our population suggests that natural populations experiencing high effective population sizes and regular dispersal might be affected by inbreeding depression less than commonly studied populations characterised by small population sizes and limited dispersal. In our population, this may be coupled with behavioural mechanisms such as sex‐biased dispersal and kin recognition reducing incidents of inbreeding (Cockburn et al. 2003; Hajduk et al. 2018); indeed, individuals resulting from close inbreeding are rare in our population. While relatively low rates of inbreeding may impact our ability to detect very small effects of inbreeding depression (only overcome by having larger sample sizes, for example, human studies leveraging over one million samples; Clark et al. 2019), we do have sufficient power to detect small but significant inbreeding depression in adult male extra‐pair paternity. As a result, we expect that our study is able to capture meaningful fitness variation associated with inbreeding depression.

In particular, the connectivity with the broader surrounding meta‐population may explain why our results differ from other bird studies in isolated, endangered or small populations (Ewing et al. 2008; Grueber et al. 2010; Harrisson et al. 2019; L. F. Keller et al. 2008; Kruuk et al. 2002). In our study, regular female immigration and emigration may lead to sufficient gene flow to assist in preventing fixation of deleterious variation via drift (L. F. Keller and Waller 2002). In perhaps the most demographically similar population to our own, a long‐term study of Florida scrub‐jays which experiences regular immigration from a broader metapopulation (Chen et al. 2016, 2019), ongoing population declines and decreasing immigration rates likely contribute to inbreeding depression pervasive across multiple life stages (Chen et al. 2016). Notably, inbreeding effects were absent or limited in the largest populations documented in table 3, the migratory caribou (Gagnon et al. 2019) and Leach's storm petrels (Sin et al. 2021). In this context, our finding of limited inbreeding effects on adult fitness emphasises that while inbreeding depression has previously thought to be pervasive in the wild (L. F. Keller and Waller 2002), it may have relatively limited individual effects in large and open populations which remain understudied.

5. Conclusions

In conclusion, we found low or negligible effects of inbreeding over several life‐history phases including key mortality peaks and adult fitness in a large open population of superb fairy‐wrens. While genomic methods were able to recover previous results on inbreeding depression in early nestling condition (Hajduk et al. 2018), early inbreeding effects were not carried through fledgling or juvenile mortality. The only adult fitness trait affected, male extra‐pair paternity, is highly variable and under strong sexual selection (Cockburn, Osmond, and Double 2008). Our findings, in line with few other studies with similar demographic contexts, suggest that natural populations experiencing high effective population sizes and regular dispersal might be affected by inbreeding depression less than previously assumed based on studies in small or insular populations. Further assessment of immigrant genetic contributions and possible dynamics contributing to purging of inbreeding load in our population would be valuable in explicitly disentangling the evolutionary mechanisms underlying the minimal effects of inbreeding in our population, and allow us to reframe how inbreeding depression affects similar free‐living large populations.

Author Contributions

J.E., L.K., A.C. and A.M. formulated the research question. A.C. and H.O. contributed to sample collection in the field. J.E. led laboratory work and data analysis, with contributions from L.K. and A.C. on analysis. J.E. wrote the manuscript, with input from all co‐authors.

Funding

The long‐term research project funding was generously provided by the Australian Research Council to A.C. and L.K. (most recently DP150100298 to A.C. and L.K., FT110100453 and FL200100068 to L.K.). We are grateful to the Holsworth Wildlife Research Endowment ‐ for their continuing support of the long‐term superb fairy‐wren project through multiple grants in support of PhD research.

Ethics Statement

Research was approved by the Australian Bird and Bat Banding Scheme, and the Animal Experimentation Ethics Committee of the Australian National University, who work under the guidelines of the Australian National Health and Medical Research Council.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Distributions for all fitness traits, including (A) nestling weight, (B) survival to nutritional independence (41 days), survival to first overwinter for natal males and females (C) and immigrant females (D), (E) lifespan for recruited adults, and reproductive success in recruited adults including females (F), male extra‐pair paternity (G) and male within‐pair paternity for males who gained a dominant breeding position (H) and total lifetime reproductive success in all females (I) and males (J).

Figure S2: Breeding span is positively correlated with lifetime reproductive success in both (A) males (blue) and (B) females (red). Lifetime reproductive success is defined as the number of offspring surviving to nutritional independence. Here, we include both within‐pair and extra‐pair offspring for males. We include only individuals in LRS inbreeding depression models (male n = 772, female n = 380). Colour intensity indicates more individuals with these values. LRS and breeding span were correlated both in males (Pearson correlation r = 0.69, p < 0.001) and females (Pearson correlation r = 0.79, p < 0.001).

Figure S3: (A) The expected genetic contribution of female immigrant cohorts based on pedigree‐relatedness to extant populations and (B) the distribution of genomic inbreeding in immigrant females.

Table S1: Parameters used to define runs of homozygosity via PLINK.

Table S2: Summary of fitness traits, model types and samples.

Table S3: Effect of inbreeding on (A and B) nestling weight and (C and D) survival from fledging to independence when including sex‐specific inbreeding effects and interactions with social environment.

Table S4: Effect of inbreeding and environment on first overwinter survival in (A) birds hatched in the study area, and (B) immigrant early disperser females.

Table S5: Effect of inbreeding on lifespan, including (A) interactions and (B) dropping insignificant interactions.

Table S6: Effect of inbreeding on total lifetime reproductive success in all (A) females and (B) males, including those that did not survive to recruitment.

Data Availability Statement

The data and scripts that support the findings of this study are openly available in a Dryad repository at https://doi.org/10.5061/dryad.9cnp5hqwb.