Introduction
Senescence affects most living organisms (Gaillard, Lemaître, & Fox, 2020). With advancing age, individual performance in reproduction, physical activities, or cognition diminish, while concomitantly the rate of actuarial senescence (i.e. mortality rate with age) increases (Nussey, Froy, Lemaitre, Gaillard, & Austad, 2013). The age at which senescence occurs varies among species and individuals (Gaillard et al., 2020), while energy gathered from the environment may be differently allocated to promote self-maintenance, and thus delay ageing (Bouwhuis, Charmantier, Verhulst, & Sheldon, 2010; Jones et al., 2008; Ricklefs, 2008). The view of senescence as an outcome of energy trade-offs that underlie life history trajectories has shaped our understanding of the evolution of animal longevities. Fast living species, with rapid growth, early sexual maturity, and high reproduction rates experience senescence earlier (Jones et al., 2008). However, whether trade-offs are the only determinants of the rate of senescence has been debated, and environmental and biological constraints may also be influential, particularly in comparisons among species (Cohen et al., 2019). In addition, senescence might differently affect phenotypic traits in relation to their impact on fitness (Hayward et al., 2015), with senescence in reproduction beginning at different points in the lifecycles of different species, compared to senescence in other traits that are directly linked to survival (Gaillard & Lemaître, 2017). Studying ageing among species will reveal how longevities co-vary with other organismal traits, and under which trade-offs among patterns of growth, reproduction, and lifespan (Boonekamp, Bauch, & Verhulst, 2020; Cohen et al., 2019). Further, studies among species may also reveal underlying cellular and physiological mechanisms of interest that sustained the diversity of animal longevities.
When comparing species, there is a strongly supported pattern of larger body size associated with longer life expectancy (Lindstedt & Calder, 1976; LIndstedt & Calder, 1981). This association should reflect physiological traits of large species, like lower mass-specific metabolic rates, which have not been fully characterized (Speakman, 2005a). Metabolism produces damages, like radical oxygen species (ROS) that are by-products of aerobic respiration. Radical oxygen species were proposed as the main underlying mechanism of ageing, thus linking together the co-evolution of body size and lifespan via the higher ROS production of fast-living, and generally small-sized species, with higher rates of oxygen consumption (Beckman & Ames, 1998; Buttemer, Abele, & Costantini, 2010; Delhaye et al., 2016). However, there are caveats to the so-called free-radical damage theory of ageing (e.g. lower ROS production at high metabolism due to uncoupling of mitochondrial respiration; (Speakman et al., 2015). Cohen et al. (2019) called for additional study of age-related mechanistic processes that may be integral to the important trade-offs in play. Thereby, such research might yield a more complete explanation of inter-specific variation in lifespan (Hulbert, Pamplona, Buffenstein, & Buttemer, 2007).
Shortening of telomere length is well known for its role in cell senescence in vitro (Harley, Futcher, & Greider, 1990). Within species, an increasing number of studies have examined the link between telomere length or its rate of erosion (i.e. telomere rate of change, TROC, expressed asbp lost per year, a mostly negative variable) with ageing and lifespan (e.g. a seminal paper by (Haussmann et al., 2003)). For instance, in recent comparative analyses, short telomeres have been associated with higher risks of mortality in non-model and wild animals (Wilbourn et al., 2018) and TROC was found inversely related to maximum lifespan in birds and mammals (Sudyka, Arct, Drobniak, Gustafsson, & Cichoń, 2015; Tricola et al., 2018; Whittemore, Vera, Martínez-Nevado, Sanpera, & Blasco, 2019). However, the sample size in the latter study was small, and a negative relationship between TROC and lifespan was lacking in intraspecific studies (Beaulieu, Reichert, Le Maho, Ancel, & Criscuolo, 2011; Sudyka et al., 2014). In addition, experimental evidence is lacking for a causal link between telomere length or TROC and lifespan ((Simons, 2015; Young, 2018); but see (Varela, Muñoz-Lorente, Tejera, Ortega, & Blasco, 2016)). Telomere length and TROC need not be highly correlated (Tricola et al., 2018), since telomere length an individual starts in life with may have a strong heritable basis ((Vedder et al., 2021), but see (Becker et al., 2015; Viblanc et al., 2020)), and both variables may be differently regulated by up-stream cell mechanisms (i.e. oxidative stress or telomerase activity, Grasman, Salomons, & Verhulst, 2011). As a consequence, they may not display similar associations with lifespan when comparing species.
Among species, TROC seems more strongly and negatively associated with mean longevity compared to a lack of significant association of mean telomere length and longevity (Dantzer & Fletcher, 2015; Haussmann et al., 2003; Sudyka et al., 2015; Tricola et al., 2018). Those studies also showed that TROC is phylogenetically conserved within bird families, thereby suggesting that mechanisms modulating telomere loss co-evolved with lifespan similarly among closely-related species (Tricola et al., 2018). However, Dantzer and Fletcher (2015) extended analyses to associations of TROC and life history traits, finding that among the 14 bird species, those with a more rapid pace of life also had greater TROC (Dantzer & Fletcher, 2015). The pace of life was calculated using a principal component analysis (PCA) that merged 9 variables on a single PCA axis, including adult body mass. However, body size effects must be removed from pace of life analysis, because the slow-fast continuum in life histories is statistically independent of body size (Gaillard et al. 1989; Read and Harvey 1989; reviewed by Harvey and Purvis 1999; Dobson and Oli 2007). The results of Dantzer and Fletcher (2015), therefore, could not differentiate associations of body size and the slow-fast pace of life with TROC. Body mass was further used as covariate in comparative analysis of interspecific link between TROC and lifespan among 19 species of birds, and this analysis likely adjusted for variation among species in body size (via body mass vs lifespan and TROC; (Tricola et al., 2018). Body size may also modulate the inter-specific variation of telomere length through its interaction with telomere maintenance enzyme, i.e.telomerase, which appears inoperative in larger rodent species; (Gorbunova & Seluanov, 2009), perhaps with an associated risk of cell immortalization (Risques & Promislov, 2018; Seluanov et al., 2007; Tian et al., 2018). In addition, body mass may be a key determinant of survival in animals (Miller, Harper, Galecki, & Burke, 2002), including bird species (e.g. (Briga, 2016; Stier et al., 2014); but see (Briga, Jimeno, & Verhulst, 2019) and has a significant association with senescence rate among species (Jones et al., 2008). Thus, one needs to statistically account for differences in body mass among species, to avoid spurious correlative conclusions about relationships between TROC or telomere length and longevity (Speakman, 2005b).
Telomere length at the end of growth seems to predict individual lifespan (Asghar et al., 2015; Heidinger et al., 2012; Lieshout et al., 2019). Early growth is the period when the greatest amount of telomere loss occurs, due either to rapid cell proliferation or high metabolism during early growth (Monaghan & Ozanne, 2018). Alternatively, longer telomeres early in life may inevitably erode faster (Grasman, Salomons, & Verhulst, 2011). Early growth is also the period during which physiological maturation of tissues occurs (Cornell, Gibson, Williams, & Portugal, 2017; Ricklefs, Shea, & Choi, 1994; Starck & Ricklefs, 1998), a phenomenon influenced by both genetic and environmental factors. From such cells and physiological mechanisms, ontogeny produces the adult phenotype. Rapid growth or stressful early-life conditions may increase telomere loss and shorten individual lifespan (Boonekamp, Mulder, Salomons, Dijkstra, & Verhulst, 2014; Herborn et al., 2014; Metcalfe & Monaghan, 2003; Tarry-Adkins et al., 2009). One hypothesis is that growth and longevity are connected via the processes that shape telomere maintenance during early-life. Still, only a tenuous and indirect relationship presently relates short telomeres to reduced long-term survival (Quque et al., 2021) and more rapid growth to increased telomere shortening (Vedder, Verhulst, Zuidersma, & Bouwhuis, 2018), both in wild birds. However, in a related study, inter-individual differences in telomere loss were only moderately sensitive to early life conditions (i.e. hatching order and resource acquisition ability), termed “canalization of resource allocation” towards telomere maintenance (Vedder, Verhulst, Bauch, & Bouwhuis, 2017). If we hypothesized that this extends to the inter-specific level, it suggests selection for reduced erosion of telomere length during growth in long-lived species. Since those species generally exhibited extended developmental periods, slow growth patterns might have been associated with reduced shortening of telomere length. This hypothesis needs to be tested at the phylogenetic level, as well as within long-lived avian groups of species like procellariforms, were body size effects can be statistically controlled due to a broad range in body mass (about 30 g to nearly 10 kg; (Dobson & Jouventin, 2007). Further, the relatively small procellariform species Oceanodroma leucorhroaexhibits an intriguing apparent telomere lengthening over life, and maintains telomerase activity in adult somatic cells (Haussmann, Winkler, Huntington, Nisbet, & Vleck, 2004; Haussmann et al., 2003; Haussmann, Winkler, & Vleck, 2005; Tricola et al., 2018).
In the present study, we tested the association of telomere length and TROC with lifespan, reproduction, and growth patterns among 53 bird species. Our study described the correlative relationships between life history traits and telomere variables at the interspecific level. We applied a principal components analysis to typify major elements of life histories, including body size, the pace of life syndrome, and parental care (after, e.g. (Bennett & Owens, 2002; Dobson & Jouventin, 2007; Dobson & Oli, 2007). This approach has the advantage of separating body size effects from influences of the pace of life on telomeres, and will nicely complement previous study on evolution of telomere length and dynamics in birds (Tricola et al., 2018). We then examined the association of these life-history indices and telomere variables, both without and with statistical adjustments for phylogenetic influence. Because lifespan is more closely associated with TROC than with telomere length (i.e. TROC being estimated using between-individual data, (Tricola et al., 2018), we predicted that after statistically controlling for body size, species with a fast pace of life should be characterized by higher TROC (i.e. greater telomere loss per time unit) and little pattern of change in adult telomere length (Adult TL ). In addition, we tested the predicted negative association of body mass and telomere length, previously found in rodents (Seluanov et al., 2007). We expected thatAdult TL would be negatively associated with body size, and TROC would be positively associated with body size (TROC is a negative value, and large-bodied species lose less TL). Predictions for Chick TLare more difficult to make. Since early growth is when telomeres are shortened the most and long-lived species have slower developmental rates, other things being equal, fledgling telomere lengths (Chick TL ), once controlled for body size, might be relatively longer in species with a slow pace of life because telomere shortening is drawn out over a longer period (based on low growth rates and reduced telomere shortening). But Chick TL should also be longer in those with a rapid pace of life (based on greater TROC, so long Chick TL and short Adult TL ). Following results of Tricola et al.(Tricola et al., 2018), we did not expect phylogeny to have a strong influence on Adult TL or Chick TL , even after controlling for the possible influence of body size (via body mass). On the other hand, phylogeny was found to explain a significant part of interspecific variation of TROC (Tricola et al., 2018). We hypothesized that extracting the influence of body size from the pace of life would lead to a comparable result, if closely related species with identical paces of life have evolved similar TROC independently of their body masses. This expectation seemed reasonable because some orders, like the Procellariiformes, have evolved long lifespans along with considerable variation in body masses among species (Dobson & Jouventin, 2007), so they should present different telomere lengths (in relation to body size) but similar TROC (in relation to lifespan).