Behavioural data and R code for: Naked mole-rats (Heterocephalus glaber) do not specialise on cooperative tasks

The dataset includes behavioural scan data from naked mole-rats and their respective individual characteristics and group characteristics. The code is used to generate the analyses and figures presented in the manuscript:https://www.biorxiv.org/content/10.1101/2021.03.22.436002v1 Further information: Animals and housing The study includes data from five groups of naked mole-rats housed at the Vienna Zoo (Tiergarten Schönbrunn) in Austria, and six groups housed at the University of Pretoria in South Africa, with group sizes ranging from 12 to 45 individuals. All animals were born and raised in captivity and housed in tunnel systems made of either transparent PVC or glass. Each group occupied a self-contained tunnel system (3.20–7 m) including at least one nest box and one toilet area. Temperatures in the housing facilities were maintained close to natural burrow conditions at 28°–30°C. The animals were fed ad libitum daily on a diet of sweet potatoes, carrots, beetroot, apples and cucumber, and provided with wood wool (Vienna) or paper towel shreds (Pretoria) as nesting material. The boxes (toilet chamber) were cleaned once a day and the food container once a week. During observations, a standardised amount of digging substrate (1 x 200 ml wood shavings) was inserted into the tunnel system every 2 h to provide substrate for burrowing activity. All individuals were identified via passive integrated transponder tags, and prior to observations, the individuals received unique colour marks applied with permanent markers. Sex was determined from the external genitalia (Pretoria) or via molecular sexing using buccal mucosa samples (Vienna). The breeding females were identified by their characteristic genital morphology. In Vienna, we were unable to identify the breeding males in the groups morphologically, and no sexual behaviour was observed during the study. Therefore, we included all individuals except for queens in the behavioural analysis as non-breeders. Data collection Data from 169 non-breeding individuals (67 females, 102 males) were included in this study. In Vienna, data were collected from 72 animals between July 2018 and July 2019. Body mass was recorded a mean of 7.0 ± 1.5 times from every animal whenever the group was removed from the tunnel system (e.g. before observation sessions or when taking mucosa samples) by placing them on an electronic scale (accurate to the nearest gram). In Pretoria, data were collected from 97 animals in August 2020. Body mass measurements were taken once for each animal before their first observation. The mean body mass for all non-breeders was 38.3 ± 12.3 g (range 16-74 g), with 36.1 ± 11.6 g for females and 39.8 ± 12.5 g for males. Ages were known for 91 non-breeders in Pretoria and 8 non-breeders in Vienna. The mean age at the time of observation was 415.4 d, ranging from 140-1254 d. Behavioural data were collected using instantaneous scan sampling. The behaviour of every animal in a group was recorded in 6-10-min intervals, depending on group size. In larger groups, 20 animals were arbitrarily chosen for observation, whereas in groups smaller than 20 individuals, all animals were included in the observation. The ethogram included 16 behaviours (Supplementary Table S1), and the observations were recorded on a handheld device using software Animal Behaviour Pro version 1.2 (University of Kent, UK). In Vienna, groups remained in their usual tunnel systems, whereas in Pretoria, they were transferred one day before observation to a tunnel system better suited for observations. The animals were allowed 24 h to habituate to the observational tunnels. Observation sessions lasted 6 h and were carried out between 08:00 and 16:00 by the same one or two observers that alternated every 30 min. In Vienna, each group was observed five times over a mean period of 216 ± 61 d, with a mean time of 54 ± 41 d between sessions. In Pretoria, each group was observed on three consecutive days. Over all 43 sessions, a mean of 161 ± 69 sampling events was recorded per individual (range 78-300, see Table 1).