Phylogenetic patterns in telomere lengths and TROC
In analyses of associations of TL values and life-history axes, strong phylogenetic values indicate that correlations are not independent over the history of the species. Species that are closely related may show similar patterns. A good example was the correlation between TROC and the body size axis (PC1), for which the “phylogeny-free” correlation was actually different in sign from the unadjusted correlations (note the high phylogenetic correlation; Table 2B). Alternatively, different clades may show different patterns, such the Adult TL and pace of life (PC2) correlation in the Passeriiformes (positive) and in the Procellariiformes (negative) (note the high phylogenetic correlation; Table 2A). After adjustment for phylogeny, residual correlations reveal ahistorical associations, such as the fairly strong negative association of TROC and PC2 (see below and Table 2B).
Adult and Chick TL among species were poorly explained by phylogeny, while TROC showed the strongest phylogenetic signal (Figure 2). Life-history axes, however, were strongly associated with phylogenetic history. Our result from 53 avian species confirms and extends previous work done on 19 bird species (Tricola et al., 2018). These results are important to our understanding of how life histories and telomere length on one hand, and the proximate factors that modulate telomere erosion on the other hand, are inter-related. It suggests that similarity in telomere lengths between species is not primarily due to shared-ancestry. Moreover, there may be independent evolutionary processes that have operated within different avian orders and families, leading to the present diversity in telomere lengths. If the evolution of life history traits is strongly constrained by phylogeny, butAdult and Chick TL are not, it may not be surprising that PC1, 2 and 3 showed only weak associations with telomere lengths measured at both stages of life. However, even after controlling for phylogeny, we found that the remaining variance in PC axes andAdult or Chick TL were poorly related (Table 2). This supports the hypothesis that telomere length evolved short and long forms at both life stages, even in closely phylogenetically-related species that share comparable life history traits. Our results seem to accord with a low phylogenetic pattern of telomere lengths in mammals (Gomes et al., 2011; Seluanov et al., 2007). For TROC , the apparent high variability in telomerase activity in closely-related species (Gomes et al., 2011; Tian et al., 2018) suggests that there should be little phylogenetic pattern, as we found.
We found no relationship between PC axes and Chick TL . We expected slow-living species to exhibit more preserved and then longer telomere lengths at the end of growth, given that rapid growth rate may enhance telomere shortening (Monaghan & Ozanne, 2018; Vedder et al., 2018). However, it is too early for conclusions about the co-evolution of growth and telomere dynamics at the chick stage. Given that telomere lengths in early life are a complex mix of inherited traits and environmental conditions (Dugdale & Richardson, 2018; Eisenberg, 2019; Vedder et al., 2017), more data are needed on embryonic telomere lengths. Further comparative analyses should include ecological variables that may highlight how specific local conditions can influence Chick TL .
TROC (calculated over the chick to adult stages) appeared moderately (our data) or unrelated to mean adult telomere length, such that most species with long telomere lengths after growth were not showing high rates of bp loss per year (Tricola et al., 2018). This result did not support the hypothesis that long telomeres erode faster because they are more sensitive to damage, nor that species with short telomeres should exhibit lower TROC over adulthood (as suggested by intra-specific research; (Bize, Criscuolo, Metcalfe, Nasir, & Monaghan, 2009; Grasman et al., 2011; Salomons et al., 2009)). Whether this could be the case for TROC calculated from earlier development remains an open question, and the fact that we found a negative relationship between TROC and Chick TL suggests that, during growth, the mechanisms regulating telomere dynamics are different than those in play during adulthood. As for TROC within adults and chicks, these may reflect distinct and non-additive processes defining species and trade-offs of individual growth, short-term chick survival and adult lifespan (Boonekamp et al., 2014), through a tightly-regulated balance between energy investment in telomere maintenance and somatic growth. In addition, our results suggest a coevolution of low TROC with increasing body size along the bird phylogenetic tree. This conflict indicates that more detailed studies are needed on how body size, embryonic development and reproduction investments may be related to specific telomere dynamics. However, the strong phylogenetic correlation of TROC and PC1 left little evidence of a relationship between these variables under current environmental conditions. In short, the strong phylogenetic pattern rendered the apparent positive phenotypic correlations of TROC and PC1 non-significant and slightly negative.
Adult TROC was closer to zero in species that produced smaller clutches and exhibited low embryonic growth rates on the slow-fast continuum (more negative values on the PC2 axis). Of all the potential associations of life histories and telomere dynamics, the association of slow growth and low reproductive rate with less TROC was the only consistent and significant pattern that we found, whether phylogenetically adjusted or not, and univariate or multivariate. Little TROC in slow growing species with low clutch sizes might be maintained by genetic correlations among these traits. Such negative associations between life history traits that divert energy from body maintenance and TROC may reflect two mechanisms of selection. First, the selective disappearance of individuals that were submitted to rapid telomere loss early in life, either due to unfavourable growth trajectories or unbalanced reproductive investments. This has been extensively discussed previously (Dantzer & Fletcher, 2015; Tricola et al., 2018): within individuals, TROC values were significantly higher (i.e. a greater rate of loss) than between-individuals, suggesting the selective disappearance of individuals with short telomeres. However, comparisons of between- and within-individuals TROC relationships with lifespan were found to be consistent, suggesting that selective disappearance might have a weak effect at the interspecific level. The second possibility is that less TROC actually reflects the co-selection of physiological or cellular mechanisms that protect telomere ends from a rapid erosion. In our study, the TROC – PC2 relationship was independent of body size effects, suggesting that it is not specifically in the largest birds that a co-evolution of slow growth and low reproductive rates with less TROC has taken place. Because telomerase is expected to be inactive in somatic cells soon after birth in large-bodied species (Seluanov et al., 2007), the underlying mechanisms should be independent of telomerase.