Phenotypic plasticity in mass loss during chick-rearing in the European starling (Sturnus vulgaris)

It has long been recognized that mass loss during breeding could be adaptive (e.g., by ameliorating the costs of increased parental activity). However, many studies still commonly interpret mass loss as evidence of “stress” or a cost of reproduction (a negative effect of high workload during chick provisioning). Despite several studies reporting evidence in support of both hypotheses, the ecological and/or life-history contexts under which mass loss may be viewed as a “cost” or an adaptive strategy are still unclear. Here, we used a long-term dataset from a breeding population of European starlings to investigate the natural annual and individual variation in body mass and mass loss and to test whether mass loss during chick-rearing represents a phenotypically plastic trait that varies predictably in relation to ecological context and individual quality. While there was significant annual variation in incubation mass, chick-rearing mass, and mass change, there were no systematic relationships between mass loss and (1) current breeding success or (2) future fecundity and survival. In addition, we found moderate repeatability of mass loss (R = 0.59) suggesting there might be additive genetic variation for this trait, though with considerable residual environmental variation. However, we found no covariation between this residual, intra-individual variation and other reproductive or life-history traits. We therefore found no support for the idea that mass loss reflects “reproductive stress” in our system: there were no negative relationships between mass loss and either current or future reproduction and survival (local return rate). Our results are consistent with mass loss being an individually plastic trait, potentially with moderate additive genetic variance, with individuals using mass loss to “level the playing field” and individually optimize reproductive effort and fitness.

Methods
Fieldwork was conducted at the Davistead Farm in Langley, British Columbia, Canada (49°10’N, 122°50’W) between April and July 2012–2022 using a free-living population of European starlings (Sturnus vulgaris). The field site is comprised of approximately 150 nest boxes mounted on fence posts around open pastures and on farm structures. In each year, a constant basic field protocol was followed: nest boxes were checked daily from April 1 to determine the laying date, clutch size, hatch date (day 0), and brood size at hatching, day 6, and fledging (day 21). Morphometric measures including mass, wing chord, and tarsus length were collected from adult females at day 6 incubation (6.3 ± 1.3 d) and day 10 chick-rearing (9.6 ± 1.3 d). Here, “mass loss” is defined as the difference between day 6 incubation and day 10 chick-rearing mass measures (approx. 15 days apart). Using tripod-mounted video cameras (Canon VIXIA HF R800), 30–60 min behavioral observations of chick-rearing parents were conducted at individual nest boxes from 09:00–15:00 h when chicks were 6–10 days old. Data were obtained on the provisioning rate (number of visits/30 min) and the sex of the visiting adult. Chick quality metrics are derived from brood mean morphometric measures (mass, wing chord, tarsus length) that were obtained from all chicks on day 17. Upon capture, all females were fitted with individually numbered metal bands (Environment Canada Permit # 10646). After the fledging of the first broods, data collection was repeated for the second broods. In some years, we conducted experiments that potentially affected natural mass change via wing-clipping and/or attachment of radio-transmitters (e.g., Fowler and Williams, 2017; Serota and Williams, 2019), so in this paper, we excluded these and restricted analysis to non-manipulated birds and controls. All protocols were approved by the Simon Fraser University Animal Care Committee (permit # 1172B-96).