Growth rates and other life history measurements
As the expected hatch date was approaching, we intensified nest monitoring to multiple times a day in order to accurately identify what is conventionally defined as day 0 of the nestling stage. Once the nestlings emerged from the egg we painted their tarsi using an individual combination of non-toxic markers to allow measurement of individual growth trajectories. On the same day we also started weighing each nestling using a portable electronic scale (Nuweigh; ±0.01 g). Measurements proceeded every second day at the same time (16:00 ± 1h) until fledging. We estimated the nestling growth rate constant (K ) using logistic regressions (Sofaer et al., 2013; Remeŝ and Martin, 2002; Ricklefs, 1968). This method allows us to obtain asymptotic body mass at fledge date, and to calculate the slope of the line tangent to the growth curve at inflection point (K ). This is a dimensionless and mass-independent metric of growth speed that is not confounded by variation among individuals in size at hatch. These measurements allowed us to test whether fast rates of somatic growth (Vedder et al. 2017; Monaghan & Ozanne 2018), and offspring mass (Ringsby et al. 2015) affect telomere length and attrition.
We also recorded brood size, and kept track of hatching order, which are both known to influence telomere dynamics (Reichert et al. 2014; Costanzo et al. 2016; Noguera et al. 2016). Finally, when the offspring produced in our experiment reached 60 days ,of age we recorded their sex, on the basis of plumage, to account for gender differences in telomeres (Noguera et al. 2015).