Skip to main content

Higher sociability leads to lower reproductive success in female kangaroos

Cite this dataset

Carter, Alecia et al. (2020). Higher sociability leads to lower reproductive success in female kangaroos [Dataset]. Dryad.


In social mammals, social integration is generally assumed to improve females’ reproductive success. Most species demonstrating this relationship exhibit complex forms of social bonds and interactions. However, female eastern grey kangaroos (Macropus giganteus) exhibit social preferences, yet do not appear to cooperate directly. It is unclear what the fitness consequences of sociability could be in species that do not exhibit obvious forms of cooperation. Using four years of life history, spatial, and social data from a wild population of approximately 200 individually recognizable female eastern grey kangaroos, we tested whether higher levels of sociability are associated with greater reproductive success. Contrary to expectations, we found that the size of a female’s social network, her numbers of preferential associations with other females, and her group sizes all negatively influenced her reproductive success. These factors influenced the survival of dependent young that had left the pouch rather than those that were still in the pouch. We also show that primiparous females were less likely to have surviving young. Our findings suggest that social bonds are not always beneficial for reproductive success in group-living species, and that female kangaroos may experience trade-offs between successfully rearing young and maintaining affiliative relationships.


Social network data:

Collection: Field observations were conducted between January 2010 and December 2013 at Sundown National Park in Queensland, Australia (28°55’03’’S, 151°34’46’’E). Each month between 10 and 16 surveys were conducted within the study site in the two hours following sunrise or the two hours prior to sunset. From January 2010 until September 2012, surveys were conducted once a day, alternating between mornings and evenings, when the kangaroos were feeding. Between October 2012 and December 2013, surveys were conducted either once or twice a day (once in the morning and once in the evening). The 15 m chain rule was used to assign group membership; any individual who was within 15 m of at least one other group member was included in the group. Female kangaroos at the study site had been previously observed to maintain social and spatial cohesion at this distance. All members of a group were considered to be associating. Estimates of group size included all adult and sub-adult females within a group but did not include males of any age. This was done because we were primarily interested in females’ association patterns with each other. We recorded the location of the center of the group (or, where possible, the approximate location of each group member) using a Garmin eTrex H GPS (Garmin International Inc., Olathe, KS, USA). During field sessions in which surveys were not being conducted, GPS locations were taken for females ad libitum from 2010 until mid-2012. 

Processing: Only females of reproductive age that were recorded on association surveys at least 10 times in a year were included in the analyses. We restricted our analysis only to females whose reproductive state had been recorded at least once every two months because of the possibility that we may have missed reproductive milestones for less frequently seen animals. For Approach 2, association survey data were grouped into three overlapping two-year periods: Jan 2010–Dec 2011 = time period A, Jan 2011–Dec 2012 = time period B, and Jan 2012–Dec 2013 = time period C.

Reproductive state data: 

Collection: The reproductive state of female kangaroos was visually assessed based on the presence and developmental stage of their young. Kangaroos give birth after 36 days of gestation to a small (approximately 800 mg), blind and hairless young that then continues its development inside the mother’s pouch. Young are permanently excluded from the pouch (PEP) at approximately 46 weeks after birth, although they begin to leave the pouch on temporary excursions approximately 1-2 months prior to PEP. After PEP they continue to nurse for approximately another seven months before they are weaned. The reproductive state of each female we encountered during association surveys was recorded in one of six categories: females with no visible pouch-young (NPY), females with a small pouch-young (SPY), females with a medium pouch-young (MPY), females with a large pouch-young (LPY), females with a young-at-foot (YAF) (young that had reach PEP but continued to nurse and remain close to its mother), and females who were caring simultaneously for both a small pouch-young and a young-at-foot (SPY, YAF).

Processing: For approach 1 that investigated annual reproductive success, for each female for a given year, we analyzed (1) whether she produced a young that survived to the LPY stage, (2) to PEP, and (3) to weaning. For approach 2, a longer-term dataset was used to assess the relationship between individuals’ sociability and offspring survival, incorporating a measure of females’ association preferences. We analyzed (1) whether she produced a young that survived to the LPY stage, (2) to PEP, and (3) to weaning in the second year of the two-year time period. This ensured that the same young were not counted in two time periods.

Usage notes

Social network data: There are four social network files, one for each year of the study. These data list the individuals and the groups in which they were seen.

Reproductive state data: These tables indicate whether each female produced a young that survived to the LPY stage (Yng.LPY), to PEP (Yng.PEP), and to weaning (Yng.Wean) for the given year (binary 0 = no, 1 = yes). It also includes data on: 

Numrec.Yr: Number of recorded sightings in a given year

Year / Period: The year or period (A, B, or C) of the study

MedianGS: The median group size in which the female was found

Prim.mum: A binary variable indicating whether the female was primiparous (1) or not (0)


Australian Research Council, Award: DP120102693