Field work and blood sampling
Fieldwork was conducted at Fowlers Gap Arid Research Station, in far-western New South Wales (31o05’S, ‘141o42’E) during the main part of the Austral breeding season (August–December) in 2016. Data were collected from zebra finch nestlings in nest boxes in their natural habitat at ‘Fowlers Gap’ (details regarding field site characteristics can be found in Griffith, Pryke, & Mariette, 2008). Given the mobility of the species in the wild, we were working with a population in which most adults were not banded, and we were unable to ascertain parents’ age or reproductive history. Nest boxes were monitored periodically (every 2 days) during the nest-building stage, and after the first egg was laid nest boxes were monitored each morning until the entire clutch was laid, enabling us to ascertain clutch size and projected hatch date. In the wild, we have previously shown that parent zebra finches only initiate incubation on the day that the last egg is laid (Gilby, Mainwaring, & Griffith, 2013). Consequently, we used the day of the last laid egg to represent the first day of embryonic development and the usual incubation period of ~12 days after the onset of incubation, to predict clutch hatch date. Two days prior to a clutch’s estimated hatch date, we monitored the nest three times per day between 06.00 and 17.00; our first nest check was at ~06.00, our second nest check was at ~11.30 and our third nest check was at ~17.00, which allowed us to identify the hatch date of each nestling. The hatch date was used to calculate the post-hatch age of each nestling (in days). Synchronous hatching meant that we could not tie nestlings to specific eggs, we were unable to account for the order in which eggs/nestlings were produced, something that has been shown to impact telomere dynamics in the zebra finch (Noguera, Metcalfe, Reichert & Monaghan, 2016).
To measure ambient developmental temperature, we obtained hourly atmospheric temperature data in the shade (i.e. air temperature) from the Australian Bureau of Meteorology’s automated weather station at Fowlers Gap, located within 20km of the study site. We calculated the mean temperature during each day of zebra finch development using hourly data from 7am to 7pm. These data represent the period of the day when the adults are periodically away from the nest foraging together (Mariette & Griffith 2015), and the nestlings (or embryos) are therefore subject to the greatest exposure to ambient (rather than brooding/ incubation) temperatures; the duration of nestling exposure to ambient temperature increases as the nestlings grow, and parents reduce brooding. The daily temperature values were then averaged from day 3 to day 11 of post-hatch development (prior to day 3, the ectothermic nestlings are typically brooded or incubated for long periods during the day, thus are not exposed to ambient temperatures, day 3 was also the age at which the earliest blood sample was taken, see below). During our sampling period (Aug-Nov), average temperatures across post-hatch development ranged from 14.6- 27.6oC (average 20.3oC, SEM: 0.295). At day 3 a small patch of down feathers was trimmed from a different area on each nestling allowing us to identify individual nestlings throughout development. Trimming the small patch of hair (approximately 0.5cm2) from each nestling was unlikely to impose thermoregulatory costs.
At day 3 and day 11, we extracted blood (<20µl) from the metatarsal vein of younger (day 3) nestlings, and the brachial vein of older (day 11) nestlings with a hypodermic needle and capillary tube. Blood was preserved in 0.5ml of 95% ethanol and stored at room temperature in an Eppendorf tube that was labelled with a unique ID. Growth rate (a measure of cell proliferation) can affect telomere dynamics (Boonekamp et al., 2020; Ringsby et al., 2015). To account for variation in growth rate between individual nestlings and broods we measured tarsus length (a reliable metric of body size) at day 3 and day 11 using digital calipers. The associated effects of clutch and brood size (i.e. parental care and resource acquisition) can also affect telomere length (Costanzo et al., 2016; Boonekamp et al., 2014) and DNAm changes (Sheldon et al., 2018), thus, we included clutch and brood size in our model. We collected data on the clutch size of each nest daily from the day of first lay to the onset of incubation, and on the two days prior to the estimated day of hatch (clutch size ranged from 2 to 11 eggs). We collected data on the brood size of each nest on the day the entire clutch hatched, day 3, and day 11 of post-hatch development (brood size ranged from 2 to 8 nestlings). Due to egg/nestling mortality, clutch/brood size often varied across development, thus we averaged the clutch/brood size of each nest across the two time-points.