Corticosterone sampling and measurement
We sampled 75 µl blood from each subject’s brachial vein within 3 min of capture (mean = 157.9 sec, min = 85, max = 216, SD = 30.877; for two birds their blood collection ended after the 3 min mark). The time to end blood collection was not correlated with natural log-transformed baseline corticosterone levels (Pearson’s r = 0.03, df = 41, p = 0.87) and we did not account for it in the statistical models.
Blood was kept on ice until centrifugation 2-8 hrs later at +4 °C for 10 min at 8000 rpm. Following centrifugation, blood plasma was removed and stored at -80 °C until analysis. Plasma corticosterone was measured using enzyme immunoassay (Cayman Chemical, catalogue number 501320; Ann Arbor, MI, USA) which has been validated and optimized for American robins (Abolins-Abols & Hauber 2020b). We extracted the non-polar components of plasma using diethyl ether extraction, described in (Abolins-Abols & Hauber 2020b). Briefly, 10 µl plasma was suspended in 200 µl ultrapure water and mixed with 1 ml diethyl ether. After passive phase separation, the mixture was flash-frozen, and the ether phase was decanted. Ether was then evaporated using nitrogen gas at 40 °C. Ether was added to the aqueous portion and decanted two more times. The extract was suspended in 600 µl assay buffer overnight at 4 °C. The concentration of corticosterone in the extract was measured according to the manufacturer’s instructions. Each extracted sample was analyzed in triplicate. We measured absorbance at 405 nm using Biotek 800TS plate reader (Winooski, VT, USA) and analyzed the data using an 8-point logistic curve using the Cayman Chemical analysis spreadsheet. The average intraplate coefficient of variation (CV) was 5.78%, while inter-plate CV was 6.19%; we therefore averaged values from the three replicates for each sample and standardized concentrations across places using the mean concentration of a standard control sample, for use in the statistical models.