Sociability, i.e. the tendency to interact with other individuals, varies significantly within populations, with some individuals being consistently more sociable than others. This variation may be maintained because more sociable individuals can thrive better in certain conditions, in which social interactions facilitate information exchange or cooperation, but not so if social encounters increase aggressive disputes or infection risk. At the proximate level, apart from genes, mothers transfer non-genetic compounds to their offspring that can influence the development of social skills. In this context, they may adjust their offspring’s sociability to match the social environment they will experience after birth, for example, via prenatal hormones. To test this, we experimentally manipulated the social density as perceived by blue tit females before egg laying. We subsequently measured yolk testosterone concentrations and social interactions among family members post-hatching. Females that were exposed to a simulated high social density transferred less testosterone to their eggs than control females. Network degree (i.e., the number of social interactions of the brood) was not affected by the social density treatment, but broods with lower yolk testosterone concentration showed a higher network degree. This suggests that mothers experiencing an environment with high social density (but not increased resource competition) deposit less yolk testosterone to produce offspring that are likely less aggressive but more sociable.

General methods

The study was carried out in Miraflores de la Sierra, Madrid, central Spain (40° 48′ N, 03° 47′ W) during the spring of 2021. We studied a wild blue tit population breeding in nest-boxes located in a deciduous forest, mainly dominated by Pyrenean oak (Quercus pyrenaica), at an elevation of 1250 m. The experimental manipulation took place from mid-March to the beginning of May, i.e., until most clutches were completed. The study area was divided into four plots, two high-density and two control, which included, respectively, 83 and 87 nest-boxes distributed across a similar-sized area. From mid-March onwards, almost three weeks prior to the laying of the first egg in the population, we exposed blue tit adults daily to either a recording of social stimuli simulating high-social density or to control playback. For the high-social density treatment, we broadcasted vocalizations of male and female blue tits. These included male songs and female calls (excluding alarm calls). For the control plots, we broadcasted vocalizations of male and female common chaffinches (Fringilla coelebs), which are common in the study area, but do not breed in nest-boxes and do not compete for food with blue tits. All recordings were obtained from Xeno-Canto (www.xeno-canto.org).

Once laying started, i.e., when the first egg (or the two first eggs) was (were) detected, nests were visited daily to collect the third egg of each clutch on the day it was laid. Collected eggs were kept cool during transport and stored at -80 °C within the same day of collection until analysis. 

In the second week after hatching (days 9-12), we trapped both adults when entering the nest-box for feeding the nestlings, and weighed them with a digital scale to the nearest 0.01 g. Their sex was determined based on plumage characteristics and the presence of an incubation patch. Thus, we could distinguish male and female parents during video observations. On day 12, nestlings were ringed, weighed with a digital scale to the nearest 0.01 g, and we marked individually on the head or wings with a white permanent marker (Edding 751). We also collected a blood sample from the brachial vein for molecular sexing.

Behavioural observations

On day 12, we replaced the original nest-box of each nest with a recording nest-box to familiarize blue tit breeding adults with it prior to video recordings. On day 13, we placed a night-vision video camera (wide-angle 8-LED IR DVR camcorder, DX, China) on the recording nest-box, approximately 10 cm above the nest. We recorded the behaviour of all family members during 30 min, excluding the first 30 and the last 10 min of the video to avoid possible interferences due to the presence of researchers when placing and removing the camera. A single observer analysed all video recordings and was unaware of the social density manipulation. For each feeding event, we registered whether the adult was marked or not and the nestlings’ begging intensity.  Begging intensity was rated on a 5-point scale following Kölliker et al. (1998): 0 = calm, 1 = weak gaping, 2 = gaping and neck stretched, 3 = gaping, neck stretched, and standing, and 4 = gaping, neck stretched, standing, and wing flapping. For each nestling, we calculated the average overall begging intensity across all feeding events directed to both parents together. Additionally, we obtained sex specific scores by averaging begging intensities for each parent separately. Also, for each parent, we extracted the number of feeding events and the average time spent in the nest. 

We scored the social behaviour of the nestlings in 10 events between the parental visits, that is, when no parent was present (hereafter: parent-absent events). We considered that a parent-absent event started at least 5 seconds after the parents had left the nest, and as soon as all chicks stayed quiet and lasted until the parents returned. Nestlings change positions mainly during or soon after feeding events (personal observation), and thus we recorded nestling positions at a single time point immediately after each parent-absent event had started. To this end, the video was paused and each nestling was assigned an individual number. We considered a social interaction as the direct physical contact between two nestlings. With this information, we created an N x N matrix where N is the number of individuals, and each cell indicates the existence of an interaction between two individuals (1) or the absence of interaction (0). As a proxy of sociability, we calculated network degree, that is, the average number of physical interactions among all nestlings within a nest.

Yolk hormone analysis

On the day of hormone analyses, each egg was slowly defrosted to isolate the yolk, which was weighed with a high-precision scale to the nearest 0.001 g. Subsequently, each yolk was diluted (1:2 w/v) in double distilled water. We took half of the sample and extracted steroids by adding 3 ml of a mixture of petroleum and diethyl ether (30:70) to the diluted yolk sample, vortexed it for 10 min, and then centrifuged the suspension for 10 min (4 °C, 1500 rpm). The ether phase was decanted after snap-freezing the tube in an ethanol bath with dry ice. This procedure was repeated a second time. Both ether phases were combined in a single tube and evaporated to dryness in a heating block at 50 °C under a stream of nitrogen. To remove protein residues, the extract was resuspended in 0.5 ml of 90% ethanol, vigorously mixed, and kept at -20 °C overnight. On the following day, the extract was centrifuged for 10 min (4 °C, 1500 rpm), the supernatant was then decanted to a new clean tube, dried again in a heating block at 50 °C under a stream of nitrogen and subsequently resuspended in 0.5 ml of steroid-free serum (DRG Labs, Germany). After a 10-min vortexing, the samples were transferred to Eppendorf tubes and frozen at -20 °C until analysis. To measure testosterone concentrations, we used an enzyme immunoassay (EIA) according to protocol (Testosterone ELISA; ref. EIA-1559; DRG, Germany). A Synergy HT Multi-Mode Microplate Reader (Biotek, USA) at 450 nm was used to measure changes in absorbance according to EIA procedure. We multiplied the obtained concentrations by the yolk weight of each sample to calculate the absolute values contained in each egg.

Molecular sexing of nestlings

DNA was extracted from blood samples using the Qiagen DNeasy Blood and Tissue kit (Qiagen Inc, Valencia, CA, USA). Sex identification was performed by polymerase chain reaction (PCR) amplification of the CHD-W and CHD-Z genes with primers P2 and P8. An initial denaturizing step at 94 °C for 4.5 min was followed by 40 cycles of 94 °C during 30 s, 49 °C during 45 s and 72 °C during 45 s. A final run of 72 °C during 10 min completed the program. Amplification was carried out in a total volume of 10 µl. Each PCR sample contained 2 µl DNA, 0.08 µl Taq polymerase (TaKaRa BIO Inc, Japan), 0.8 µl dNTP 2.5 mM, 0.5 µl of each primer 10 µM, 1 µl of 10X PCR buffer, and 5 µl of sterilized distilled water. The sex of 2 chicks from 2 nests could not be determined due to unsuccessful DNA extraction, and were hence excluded from analyses that included nestling sex.