INTRODUCTION
Telomeres, the nucleoprotein complexes involving tandem repeats of a non-coding DNA sequence, prevent the ends of linear chromosomes from inappropriately activating the DNA repair machinery (Blackburn, 1991). In the absence of restoration, telomeres shorten with each cell division due to incomplete replication of DNA at the chromosome ends, and their eventual dysfunction limits cell replicative potential (Hayflick, 1965). Telomeres may be further eroded by other processes including oxidative damage (von Zglinicki, 2002; Reichert & Stier, 2017). Telomere length (TL) changes might therefore reflect the cumulative costs associated with acquiring and maintaining a particular body size, since this is linked to cell replication levels (Monaghan & Ozanne, 2018), as outlined in Fig. 1. The functional relationships between size, growth and telomere dynamics might thus play an important role in shaping the optimal body size in wild species under natural selection (Ringsby et al. 2015; Erten & Kokko, 2020). Body size is a fundamental species characteristic, which is intertwined with many aspects of species ecology and evolution (Haldane, 1928; Peters, 1983; Woodward et al., 2005; Sibly & Brown, 2007), and is under directional selection in many species (Kingsolver & Pfennig, 2004). Both across and within species, body size has been shown to correlate with important fitness-related traits including survival, lifespan, fecundity and metabolic rate (Bumpus, 1899; Gaillard et al., 1989; White et al., 2019). Life-history theory predicts that organisms need to allocate their limited available energy among different components of fitness, which leads to trade-offs and selection for optimal solutions (Stearns, 1989). Trade-offs between life-history traits may also occur due to antagonistic interactions; for instance, if allocation of energy into developmental growth has negative consequences due to long-term effects of telomere shortening (Fig. 1, Monaghan et al., 2009; Young, 2018). For instance, Heidinger et al. (2012) demonstrated a negative relationship between early-life TL and lifespan in captive zebra finches (Taeniopygia guttata ). Negative correlations between fitness-related traits may conform to the hypothesized life-history trade-offs (Futuyma, 2010), but they are difficult to observe in the wild, for instance due to variation among individuals in resource acquisition (van Noordwijk and de Jong 1986) and random environmental variation (Pujol et al., 2018). Nonetheless, physiological or genetic constraints generating life-history trade-offs may be detected by comparing different phenotypes or genotypes (Reznick, 1985), or through experimental manipulations involving for example natural selection in a controlled environment or artificial selection in the traditional breeder’s approach (Connor, 2003; Postma et al., 2007; Pick et al., 2020). Correlated responses to artificial selection then suggest additive genetic covariance between a trait and the selected trait (Connor, 2003).
Several studies have investigated potential for telomere dynamics to underpin individual variation in life-history strategies (Monaghan, 2010; 2014; Selman et al., 2012; Vedder et al., 2017). In the wild, many of these studies, including long-lived bird species, mammals and reptiles, suggest that most telomere loss occur during early life (Hall et al., 2004; Spurgin et al., 2018) and that TL may be negatively correlated with various developmental stress factors at this stage (Fig. 1), including brood competition (Boonekamp et al., 2014; Nettle et al., 2016), poor nutrition and catch-up growth (Jennings et al., 1999; Geiger et al., 2012). Such factors may result in the release of stress hormones, which have been shown experimentally to increase early-life telomere loss in the wild (Herborn et al., 2014), and oxidative stress (Reichert & Stier, 2017) that may directly increase the shortening of telomeres. In addition, body size has been shown to negatively correlate with TL within different tetrapod species (Scott et al., 2006; Pauliny et al., 2006; Debes et al., 2016; Caprioli et al., 2013; Ringsby et al., 2015; Spurgin et al., 2018), which is thought to be due to the additional number of cell divisions required for acquiring larger size, and the increased oxidative stress associated with maintaining larger size (Monaghan & Ozanne, 2018). The emerging field of telomere ecology aims to identify factors that influence variation in individual TL and their potential fitness consequences in free-living animals (Spurgin et al., 2018). Whether there is a causal relationship between telomere dynamics and individual variation in fitness in populations in the wild and if TL is an indicator of individual quality are not yet fully understood (Simons, 2015; Angelier et al., 2019). Experimental approaches in both the field and the laboratory play an important role in increasing our understanding of TL and life-history evolution.
In a large scale experimental study conducted in the wild, an artificial directional selection regime on body size, as indicated by tarsus length, was imposed annually and in opposite directions during four consecutive years in two island populations of wild house sparrows (Passer domesticus ) in northern Norway (Fig. 2). Each winter, adult sparrows with tarsus lengths smaller or larger than given thresholds were removed from each population to produce a significant bidirectional change in mean tarsus length across the adult breeding populations (Kvalnes et al., 2017). Relatively high heritabilities for tarsus length have been found in these house sparrow populations (h 2=0.3-0.4, Kvalnes et al., 2017), thus we expected, based on the Breeder’s equation (Lande, 1979), the artificial selection to result in significant responses in offspring tarsus lengths and growth rates. Here, we initially show how the artificial selection on parental tarsus length affected the size of their offspring measured during the nestling stage. Tarsus length is commonly used as a proxy for structural body size in house sparrows (Rising & Somers, 1989; Araya-Ajoy et al. 2019). In a previous study based on a subsample of chicks from the population undergoing artificial selection for larger body size, Ringsby et al. (2015) showed that the selection regime had indeed extended the range of chick body size at its upper end, and that this was associated with a reduction in TL in red blood cells. In this study, we examined whether the reciprocal effect, leading to longer TL, occurred in the population in which body size was reduced, and also examined the results in chicks whose parents were not subjected to the selection regime. We constructed genetic pedigrees to identify nestlings with parents that were subjected to artificial selection. We then investigated how individual TL in nestling cohorts changed under the different size selection regimes: We expected that increasing body size through artificial selection led to shorter TLs through increased number of cell divisions (Falconer et al., 1978) and oxidative stress associated with increased energy expenditure (Geiger et al., 2012; Pauliny et al. 2015; Smith et al., 2016; Monaghan & Ozanne, 2018). Similarly, we expected that smaller body size and thus decreased nestling growth led to a slower rate of TL reduction (Vedder et al., 2018). Since TL was measured in early-life (fledgling stage) we also examined environmental factors previously shown to influence telomere loss during this period (Chatelain et al., 2020), specifically brood competition (Boonekamp et al., 2014) and weather conditions (Graham et al., 2019).
The artificial selection was expected to shift the populations away from their optimal body size and increase the phenotypic variance across the populations (Kvalnes et al., 2017). Here, we investigated whether changes in TL following the artificial size selection might mechanistically underpin fitness effects due to the deviation from the optimal body size. Thus, the survival and reproduction of all individuals on both islands were monitored during and after the selection events. Hypothesizing that shorter TL relative to body size would be associated with lower survival and reduced lifespan (Heidinger et al., 2012), because short telomeres reflect adverse early-life conditions (Wilbourn et al., 2018; Eastwood et al, 2019), we tested the effect of TL on short-term (first-year) survival under the two selection regimes, and on long-term survival (lifespan) after the artificial selection events. Similarly, we tested if longer early-life TL predicted higher future reproductive success (e.g. Pauliny et al., 2006; Olsson et al., 2011; Heidinger et al., 2012; Bauch et al. 2013) and if any of the potential trade-offs (Fig. 1) were affected by the artificial selection.