Maintaining a pair bond year after year (perennial monogamy) often enhances reproductive success, but what familiar pairs are doing differently to improve success is unknown. We tested the hypothesis that endocrine changes mediate improvements in parental attendance in known-age Cassin’s auklets Ptychoramphus aleuticus, which we found limited evidence for. Instead, we found sex-specific parental roles in familiar pairs. Males modulated their nest attendance depending on the attendance of their mate, but the direction depended on mate familiarity.  We suggest his flexibility may be mediated by prolactin. In a historical dataset, females with a familiar mate laid larger eggs that hatched into more robust chicks, but larger eggs correlated with lower female body condition. In study birds, attendance by males and females in good condition predicted chick weight, but attendance by females in poor condition did not, suggesting female-specific energetic constraint. Our findings suggest that males and females contribute differently to their joint reproductive fortunes, and that improvements in their respective roles may result in the benefits of mate familiarity. Since improved reproductive success is presumed to be a main benefit of maintaining a long-term pair bond, these results suggest a new avenue of research in the evolution of monogamy.

Fieldwork and sample collection: We conducted our study on Southeast Farallon Island (37.6989° N, 123.0034° W; Farallon Islands National Wildlife Refuge, California, USA) in 2019 and 2021. These two years represented a positive El Niño Southern Oscillation (ENSO) event where on average only 8% of followed pairs fledged a chick (2019), and a negative ENSO event (2021) where 65% of pairs fledged a chick [1]. Many individuals of the Southeast Farallon Island Cassin’s auklet population are of known age and have known reproductive histories [2], allowing us to disentangle the confounding effects of age and mate familiarity. Nest boxes were checked every 15 days until an incubating bird was found. If the individual was of known age, it was measured and evaluated for inclusion in the study. All morphometrics are therefore from birds in their first two weeks of incubation. We weighed known-age incubating adults between 13:00 and 15:00 with a Pesola spring scale (+/- 1g), measured wing cord with a wing rule (+/- 1mm) and measured bill depth with a digital calliper (+/- 0.1mm). The following day, we measured the mate. We checked nests every 5 days to determine egg fate, and after hatching, chicks were weighed every 5 days until fledging to measure the maximum weight attained.

We acknowledge that sex is a nonbinary phenotype and that treating it as such can obscure important and relevant biological information, and inadvertently support social agendas that cause real harm [3]. For brevity and clarity, we refer to the egg-laying member of the pair as “female” and the non-egg laying member of the pair as “male”. Males have deeper bill depths compared to females, but there is significant overlap [4]. To sex birds accurately, we used bill-depth based sex estimations from multiple previous years for all individuals that had been measured before with a different mate (63 of 76 individuals in 2019 and 66 of 80 individuals in 2021), assuming a pair was one male and one female. Sex-specific body condition was estimated using the residual of the model mass~bill depth+wing cord. This model was built using morphometrics for this colony from 1973 to 2021 (N = 7964).

Pairs were selected for the study if there was at least one known-age mate. We attempted even representation of individuals in their first or second breeding attempt (i.e., 2-3 years old; “young”; in this population many individuals breed for the first time at 2 years) or had more years of previous breeding (i.e., 5+ years old; “old”.) If the pair contained an individual of unknown age, the pair was classified based on the known-age mate, but because ages of mates are only weakly correlated (coefficient = 0.39, SE = 0.051, p < 0.001, R² = 0.14, N = 353), we used the categorical classification of pairs as “old” or “young” for pair selection purposes only. We used individual age for 97 known-age birds and band year-based age estimation for 51 non-known-age mates of known-age birds in our models. Although there were 9 individuals mistakenly classified because they were not thought to be known-age and were later discovered to have been banded as chicks, there was no difference in the mean age between experience groups (student’s t test, p = 0.3). When we compared newly formed pairs (“inexperienced”) to pairs that had been observed breeding together at least once before (“experienced”) in our analyses, we used a categorical variable to describe mate familiarity (i.e., “experienced” or “inexperienced”; “pair experience”). When exploring changes over time, we used the number of years the pair had been observed breeding together (“bond duration”). In this population, known-age individuals are usually recaptured yearly from age 2 or 3 until senescence, so it is unusual to have gaps in our knowledge of their reproductive history. We followed 79 auklet pairs over two years, 19 inexperienced pairs and 20 experienced pairs in 2019 and 19 experienced and 21 inexperienced pairs in 2021. A total of 145 individuals were used in this study, 23 of which contributed to the study in both years.

To measure nest attendance, nest boxes of were fitted with radio frequency identification (RFID) readers and both individuals were fitted with a passive integrated transponder (PIT) tag on the tarsus, a method that is reliable for determining nest attendance and causes little disturbance [5,6]. Our RFID readers were programmed to continuously record from 2 circular antennas deployed at the nest entrance to record tagged birds entering or exiting the nest box. Tags and readers were deployed 1-2 days after birds were weighed and measured. We measured nest attendance rate as the proportion of nights an individual was present at the nest. During incubation, presence at night means the bird was either finishing an incubation bout or relieving its mate. During chick rearing, presence at the nest at night was assumed to be prey delivery. Every effort was made to equalize the number of nights of recordings across nests, but it was not always possible as readers occasionally ran out of battery or malfunctioned. We accounted for this variation in our statistical analysis by weighting models by the number of nights of RFID recording.

Physiological sampling was conducted 10 days after the egg was first observed during incubation (May 13-16 in 2019 and April 5-17 in 2021) between 10:00 and 19:00 PST and when the chick was 10 days old (June 17-24 in 2019 and April 28-May 19 in 2021) between 22:00 and 0:00, as Cassin’s auklets are only present on the island at night during chick rearing. A one-time measurement will not fully capture the range of hormones across a breeding period, which can be inconsistent within an individual [7]. However, repeated sampling would cause levels of disturbance that nesting auklets may not tolerate and we were unwilling to risk triggering nest abandonment, given that this population has significantly declined since monitoring began in the 1970’s [2]. We therefore limited our disturbance of the colony, at the cost of a lower likelihood of finding correlations between behaviour and hormone levels. We measured PRL in plasma and CORT in faeces, since faecal CORT is less susceptible to momentary stressors from investigator disturbance [8]. An analysis of faecal CORT data collected in this population in 2021 show no correlation between CORT and time of collection (generalized linear mixed model with faecal sample emulsification as a random effect, p = 0.61). To our knowledge, there is no existing assay for measuring PRL in avian faeces, but it takes much longer than CORT to respond to handling stress [9]. Other studies have found little evidence of circadian rhythm in plasma PRL[10]. We collected a faecal sample by holding the bird over a Pyrex container for no more than 1 minute, then collected a 150µL blood sample from the brachial vein into heparinized capillary tubes to measure PRL. Plasma and faecal pellets were frozen at -10°C in the field. After the field season, plasma samples were frozen at -80°C and faecal samples at -15°C.

Laboratory methods: We dried faecal samples in a fume hood, then weighed and homogenized them before rehydrating with 2mL dH₂O. To validate our radioimmunoassay protocol, we made a pool from three samples, added 1 and 0.25 ng CORT and performed serial dilution to confirm parallelism to the standard curve. We measured 1.05 ng and 0.15 ng CORT in these pooled samples. Steroids were extracted in duplicate following Krause et al. [11] across two assays. Mean percent recovery was 61.5% (SD = 20.6%). Intra-assay CV was 3.66% and inter-assay CV was 7.16%. 26 out of 194 faecal samples were emulsified by dichloromethane during steroid extraction, which decreased the resulting CORT measurement by 6.8 ng/mg (SE = 1.2, t-value = 8.70, p < 0.001). There was no evidence that emulsified samples were nonrandomly distributed among any variables of interest (Supplementary table 1).

We determined plasma concentrations of PRL via a heterologous radioimmunoassay as detailed in Cherel et al. [12] in 20µL duplicates across two assays. The detection limit was 0.99 ng/mL. The intra- and inter-assay CVs were 11.20 and 12.56%, respectively. Pooled plasma samples of Cassin’s auklet produced a dose–response curve that paralleled chicken prolactin standard curves, validating our PRL radioimmunoassay for this species.

Historical Data: We used historical data of known-age Cassin’s auklet egg length and diameter from Southeast Farallon Island (the same population as our study birds) from 1992-2017 (N = 783). Egg length and diameter of the widest point were measured using digital callipers (+/- 0.1mm) at the same time as adults were measured in early incubation. Egg measurements of the study birds were not available because objective 3 was added post-hoc, in response to interesting patterns we observed in our results for the first two objectives. Accordingly, egg measurements were not in our original study design and regular egg measurement was discontinued from the monitoring protocol in 2018. We calculated egg volume as 0.5202LD²-0.4065 [13], where L = length and D = diameter. These data were paired with information on pair experience, female age, wing cord (a proxy of body size), body condition, hatching success, and the maximum weight of the chick. Adult and chick morphometrics were collected using the same protocol as for our study birds, and body condition was calculated using the same model. We omitted one female with an anomalously low residual body condition of -58, which was likely due to measurement error (population mean = -1, SD = 10).

1.         Johns ME, Warzybok P, Jahncke J, Doak P, Lindberg M, Breed GA. 2022 Episodes of high recruitment buffer against climate-driven mass mortality events in a North Pacific seabird population. J. Anim. Ecol. 91, 345–355. (doi:10.1111/1365-2656.13630)

2.         Pyle P, Sydeman WJ, Hester M. 2001 Effects of age, breeding experience, mate fidelity and site fidelity on breeding performance in a declining population of Cassin’s auklets. J. Anim. Ecol. 70, 1088–1097. (doi:10.1046/j.0021-8790.2001.00567.x)

3.         Massa MG, Aghi K, Hill M. 2023 Deconstructing sex: Strategies for undoing binary thinking in neuroendocrinology and behavior. Horm. Behav. 156, 105441. (doi:10.1016/j.yhbeh.2023.105441)

4.         Nelson DA. 1981 Sexual differences in measurements of Cassin’s Auklet. J. F. Ornithol. 52, 233–234. (doi:10.2307/4512659)

5.         Marcouillier L, Miranda E, Whelan S, Fifield D, Hatch S, Shoji A, Nakajima C, Elliott KH. 2024 PIT tagging does not measurably reduce reproductive success in sensitive burrow-nesting seabirds. Mar. Biol. 171, 1–13. (doi:10.1007/S00227-023-04387-X/METRICS)

6.         Bonter DN, Bridge ES. 2011 Applications of radio frequency identification (RFID) in ornithological research: a review. J. F. Ornithol. 82, 1–10. (doi:10.1111/j.1557-9263.2010.00302.x)

7.         Grant AR, Baldan D, Kimball MG, Malisch JL, Ouyang JQ. 2020 Across time and space: Hormonal variation across temporal and spatial scales in relation to nesting success. Gen. Comp. Endocrinol. 292, 113462. (doi:10.1016/j.ygcen.2020.113462)

8.         Ludders JW, Langenberg JA, Czekala NM, Erb HN. 2001 Fecal corticosterone reflects serum corticosterone in Florida sandhill cranes. J. Wildl. Dis. 37, 646–652. (doi:10.7589/0090-3558-37.3.646)

9.         Chastel O, Lacroix A, Weimerskirch H, Gabrielsen GW. 2005 Modulation of prolactin but not corticosterone responses to stress in relation to parental effort in a long-lived bird. Horm. Behav. 47, 459–466. (doi:10.1016/j.yhbeh.2004.10.009)

10.       Lea RW, Sharp PJ, Chadwick A. 1982 Daily variations in the concentrations of plasma prolactin in broody bantams. Gen. Comp. Endocrinol. 48, 275–284. (doi:10.1016/0016-6480(82)90027-2)

11.       Krause JS, Mcguigan MA, Bishop VR, Wingfield JC, Meddle SL. 2015 Decreases in Mineralocorticoid but not Glucocorticoid Receptor mRNA Expression During the Short Arctic Breeding Season in Free-Living Gambel’s White-Crowned Sparrow (Zonotrichia leucophrys gambelii). J. Neuroendocrinol. 27, 66–75. (doi:10.1111/JNE.12237)

12.       Cherel Y, Mauget R, Lacroix A, Gilles J. 1994 Seasonal and Fasting-Related Changes in Circulating Gonadal Steroids and Prolactin in King Penguins, Aptenodytes patagonicus. https://doi.org/10.1086/physzool.67.5.30163887 67, 1154–1173. (doi:10.1086/PHYSZOOL.67.5.30163887)

13.       Narushin VG. 2005 Egg geometry calculation using the measurements of length and breadth. Poult. Sci. 84, 482–484. (doi:10.1093/ps/84.3.482)