Parental effects occur when the phenotype of one or both parent(s) affects the phenotype of offspring beyond direct effects of genetic inheritance (e.g. Mousseau & Fox 1998). Such parental effects are of evolutionary importance because they can influence the direction and rate of phenotypic and genetic change (Badyaev & Uller 2009). Parental age is one aspect of the parental phenotype known to have consequences for aspects of offspring phenotype and fitness. Offspring from old parents have, for example, been found to have a reduced probability to recruit to the breeding population (e.g. Bouwhuis et al. 2009) or an altered (e.g. Bouwhuis et al. 2010) or reduced annual reproductive success or lifespan (Bouwhuis et al. 2015, Schroeder et al. 2015) after recruitment.
From an evolutionary point of view, parental age effects should arise from a resource-based trade-off between survival and reproduction, i.e. from disposable soma (Kirkwood 2017). Individuals shift resources from self-maintenance and repair to investment in germline-maintenance and reproduction. This increases an individual’s fitness since maintaining a healthy soma beyond the expected lifespan due to unavoidable extrinsic mortality would be less beneficial than investing the resources available into the production of offspring. Limited self-maintenance and repair then allow for the accumulation of damage with age, underlying the late-life performance declines defining senescence. Recent theoretical work has suggested that the disposable soma process can indeed account for parental senescence as well as transgenerational parental age effects observed in offspring (van den Heuvel et al. 2016), and when including trans-generational parental age effects, the evidence for disposable soma is convincing (Maklakov & Immler 2016).
The inheritance of epigenetic alterations to gene expression is a potential mechanistic explanation underlying parental effects in general (e.g. Burton & Metcalfe 2014), and of specific interest especially for species in which negative parental age effects on offspring have been found, despite parental care improving with age due to increased experience (Bouwhuis et al. 2015). Recent meta-analytical work has shown that DNA methylation patterns assessed in blood are predictive of age, but better predictive of time to death than age itself, and can therefore be viewed as accurate measures of biological age (Chen et al. 2016). In addition, empirical studies have shown that the extent to which somatic repair genes are upregulated with age, perhaps in response to DNA methylation levels (Razin & Cedar 1991), can explain variation in lifespan (e.g. Lucas et al. 2016). It therefore seems reasonable to hypothesise that a change in parental DNA methylation with age leads to changes in somatic and germline state, which then leads to offspring inheriting different epigenetic marks based on whether they are produced early or late in the life of their parents. These epigenetic marks could then affect how offspring balance their allocation of resources between self- and germline-maintenance and therefore be responsible for the observed variation in life-history trajectories of offspring of young and old parents (e.g. Bouwhuis et al. 2015, Schroeder et al. 2015).
The common tern (Sterna hirundo) is a relatively long-lived migratory seabird that is easily marked and shows very limited breeding dispersal, such that marked individuals can be followed throughout their lifespan (Zhang et al. 2015a,b). We study common terns breeding at the Banter See in Wilhelmshaven on the German North Sea coast. Birds from this study population have been ringed and marked with individually numbered subcutaneously injected transponders since 1992 and their individual life histories have been assessed since. Recent work has shown that many phenological and reproductive traits improve with age, before leveling off (Zhang et al. 2015a), but that there are sex-specific pathways of negative parental age effects on offspring fitness (Bouwhuis et al. 2015). Specifically, recruited daughters from older mothers suffer from reduced annual reproductive success and therefore obtain a reduced lifetime reproductive success. In contrast, recruited sons from older fathers suffer from a reduced lifespan, and this effect also translates to a reduced lifetime reproductive success. Because parental care, however, improves with parental age in our population (Limmer & Becker 2009), we expect the mechanism underlying the observed pattern to occur on the (epi)genetic level.
In the breeding seasons of 2013, 2014 and 2017, we obtained repeated blood samples from 17 breeding pairs and their offspring. These samples have undergone reduced representation bisulfite sequencing (RRBS) as an efficient and high-throughput sequencing technique to characterise their genome-wide DNA methylation profiles on the single nucleotide level. In addition, we will sequence, de novo assemble and annotate a draft reference genome for the so far genomically not characterised common tern, using DNA from an adult male with well characterised life history data (strong pedigree links, geolocator tracks of migratory journeys, known behavioural strategy and reproductive success) from within the focal population.
With these data, we will be able to answer the following questions:
* do adult DNA methylation patterns change with age within individual common terns and does any within-individual change in adult DNA methylation pattern depend on age-at-first-sampling or sex?
* do offspring DNA methylation patterns resemble those of (one of) their parent(s) and do (sex-specific) offspring DNA methylation patterns predict offspring development or survival?
In addition, we will be able to also assess whether adult DNA methylation status is predictive of remaining lifespan (e.g. Chen et al. 2016) and whether offspring DNA methylation status is predictive of the probability to recruit to the breeding population. Follow-up studies of recruits throughout their subsequent reproductive lifespan are within reach as well.