Discussion
Using longitudinal telomere measurements from the Lundy Island house sparrow population, where precise ages, death status and reproductive success are known, we estimated the relationships between telomere dynamics and fitness measures, including survival and reproductive success. We found that in post-fledging birds, independent of age, longer telomeres were associated with higher chance to survive to the next year. This finding was consistent with existing literature on adult telomere length (e.g. Angelier et al., 2013; Barrett et al., 2013) and meta-analytic results (Wilbourn et al., 2018). It also agrees with the speculation of the selective disappearance of older birds with short telomeres in the Lundy sparrows (Chik et al., 2023). The link between telomere length and survival/mortality could be explained by two mechanisms: Telomeres could play a causal and active role, by inducing cell senescence and cell death at a critically short length. The accumulation of senescent cells could hinder tissue functions, lead to organ failure, and eventual death (Barrett et al., 2013; Monaghan, 2010; Sahin et al., 2011). Alternatively, telomere length could also not participate directly in causing death, but serve as an indicator of the accumulative damage received by the body, or as a measure of ‘frailty’, the capacity of the body to withstand and/or recover from damage (Monaghan, 2010). Regardless of causality, our finding supports that telomere length could serve as a biomarker of immediate survival.
Nevertheless, the demonstrated association between adult telomere length and survival in our study contradicts others. In another insular house sparrow study in Norway, authors found no correlation between early-life telomere length and adult survival (Pepke et al., 2022). This could be a result of habitat differences – in the Norwegian population, some sparrows resided on islands with limited food and shelter, leading to higher competition and increased juvenile mortality, ultimately the decoupling of early-life telomere length and adult survival (Pepke et al., 2022); whereas in the Lundy population, food and shelter is available to sparrows year round, and mortality was less dependent on resources availability and population density (Simons et al., 2019), thus revealing a stronger effect of telomere length. As telomere dynamics are influenced by environmentally-induced oxidative stress (Monaghan & Ozanne, 2018), it is perhaps not surprising that the telomere-mortality link would be context-dependant, necessitating further studies using different ages, populations, and taxa (Wilbourn et al., 2018).
Compared with survival, the link between telomere dynamics and lifespan was much weaker, though still in the expected direction. This weaker link could be the result of the more removed nature of lifespan as an indicator of survival, or extrinsic factors. Independent of telomere length, age was linked with mortality: the youngest and oldest birds had a higher probability of dying. This could mean that other age-specific factors, such as predation, became the main cause of death in the shortest and longest living birds. This would weaken the link between lifespan and telomere length at the extreme ages, and drive down sample sizes, especially of long-lived birds, such that we could no longer detect an effect of telomere length on lifespan. In the Lundy sparrows, predation pressure was stronger in adults than in juveniles (Simons et al., 2019), but we do not know the main cause of death in each age class, nor have we tested for age dependency in TL-mortality association. Further studies should address these topics. Nevertheless, the effect found here agreed with the positive link we found between telomere length and immediate survival.
If telomere length acts as an indicator of somatic redundancy/frailty, then the TL-mortality link would be weaker at older ages, and the rate of telomere shortening could emerge as a better predictor of lifespan (Boonekamp et al., 2013; Monaghan, 2010). However, we did not find such association here, as covariance between the rate of RTL change and lifespan was not statistically significant, despite finding individual variation in the rate of telomere shortening (Chik et al., 2023). This could be a result of not having enough statistical power: In our dataset, only 270 birds were sampled three times or more, and few individuals lived to old ages of 9 and above.
In addition to survival, we also found a link between telomere length and reproductive success, such that individuals with longer telomeres on average, produce more genetic recruits over their lifetime, which in our population, predicts expected genetic contribution and fitness (Alif et al., 2022). In contrast, there was no evidence of any relationship between annual telomere length and reproductive output. Our results indicated that the link between telomere length and fitness is primarily through higher survival, where individuals with longer telomeres survive longer and as a result reproduced more, similar to the finding by Heidinger et al. (2021), and consistent with the ‘individual quality hypothesis’, i.e. individuals with a higher quality will have better body conditions, and hence survival and/or reproductive prospects, than poorer quality individuals, a trend found also in classical brood size manipulation studies testing for survival-reproduction trade-offs (Winder et al., 2022). One important contributor to variation in individual quality is parental age at conception – previously we detected such Lansing effect in the Lundy sparrows, where birds whose biological parents were older when they hatched, produced fewer recruits annually and over a lifetime, suggesting epigenetic detrimental effects that were carried down generations (Schroeder et al., 2015). Further studies should test for a similar Lansing effect in telomere dynamics to better elucidate the intrinsic and extrinsic contributors to variation in individual quality (e.g. Drake & Simons, 2023), and how telomere dynamics is mechanistically linked to quality and reproduction.
In contrast, there was no relationship between annual telomere length and reproductive output among individuals. This did not align with the ‘pace-of-life hypothesis’, under which individuals with a faster pace-of-life are expected to sacrifice somatic maintenance for reproduction, trading higher output for shorter telomeres. This could mean that there is little variation in the pace-of-life in the Lundy population, which is in line with another finding by Heidinger et al. (2021). Alternatively, our result could also indicate that the physiological costs of reproduction was not reflected on telomere dynamics, or that the trade-off between reproduction and ageing is not as strong within-species as previously considered, and masked by quality effects (Winder et al., 2022). Indeed, we did not find an association between the rate of telomere shortening and lifetime reproductive output, nor a trade-off between telomere length and reproductive output within an individual, suggesting that the lack of association could not be attributed solely to differences in individual quality, e.g. in resource acquisition or stress resistance. Note, however, that reproductive success need not equate to reproductive effort. For example, previous experiments have shown parents of enlarged broods had shorter telomeres and faster shortening than those with unmanipulated or reduced broods (Reichert et al., 2014; Sudyka et al., 2014). Further studies should therefore examine the effect of e.g. parental care on telomere dynamics, to determine whether the latter is an indicator of the costs of reproduction in the Lundy sparrows.
In conclusion, in this study we examined the fitness consequences of telomere dynamics in a longitudinal, closed house sparrow population, and found evidence that indeed telomere length was correlated with fitness. Our results provide additional support that telomere length is linked with survival and therefore in turn with lifetime reproductive success, but also add to the debate of the role of telomere shortening as an indicator of senescence, somatic resilience, and fitness. It is important as a next step to determine whether the associations we found are only at the phenotypic level, or occur also at the genetic level, which coupled with heritable variation in telomere dynamics (Chik et al., 2023), would inform how telomere dynamics evolve in the wild.