iiinova — where your digital self keeps evolving

Diet profile is a crucial and broad survival skill, encompassing know-what, know-how, know-when, and know-where knowledge types. Our results therefore bear significance not just for orangutans, but also for other species with similar knowledge requirements. 7/11

The circular dendrogram shows all the recorded unique food items eaten by a sample of 3 focal immatures. Each vertex (grey line) indicates a food item. Each node (colored circle) represents a food species. Node size represents the recorded number of times each food item was consumed by these immatures during the study period. Below the circular dendrogram, each infant's social and individual learning propensities along with the respective population average propensities are shown. The images of the immatures are shown in the middle of the dendrogram.

170

Social and individual learning worked in tandem in diet profile development. However, when some infants had low individual learning propensities, the associated reduction in diet profile breadth was partially offset by their heightened social learning propensities. 6/11

Predicted effects of social and individual learning propensities on the development of a broad ecological competence are shown. The graph has two panels: (A) Estimated development of the breadths of the diet profiles of immatures between birth and the onset of independence and (B) Estimated breadths of the diet profiles at independence, as a function of immatures’ estimated average social and individually initiated learning propensities. Low, Average, High refer to mean-1SD, mean, and mean+1SD. Each circle represents a posterior prediction. The yellow circle represents the posterior mean, and the vertical lines represent 95% CI.

140

We therefore considered only independently initiated explorations when quantifying individual learning. Infants differed from one another in their social/individual learning propensities (accounting for confounds), but social learning was 5-6 times more frequent than individual learning. 5/11

Individual variation in learning propensities and within-individual relationships between propensities of different forms of learning are shown. Graphs on the first row show posterior distributions of the estimated average (A) social, (B) socially initiated and (C) individually initiated learning propensities for 21 immatures over their 8.5-year developmental period. The dashed black vertical line represents the population average, and the yellow diamond represents the posterior mean of the estimate for each immature. Values to the left of the dashed line represent lower, and those to the right higher, than population average propensities. The bottom row shows the scatterplots of individuals’ posterior mean for (D) social and individually initiated, (E) social and socially initiated, and (F) socially and individually initiated learning propensities for the 21 immatures. Circle size represents the total observation duration for each immature between birth and 8.5 years of age (N = 2-360 hours). Correlation r and posterior probabilities for the hypothesis ‘r > 0’ are shown.

161

After peering, infants engaged in exploration behavior. Peering triggered subsequent explorations for 2.5 hours! The key takeaway here is that engaging in social learning has a knock-on effect; not all exploration events are truly indicative of individual learning. 4/11

172

Using the data collected over 20 years on 21 infants between birth & independence, we find that social learning does the heavy lifting in diet profile development. How, you ask? 3/11

1104

Using ~7000 records of behavioral indicators of social (peering) & individual (exploration) learning, we find that heightened social & individual learning propensities led to the development of the broadest diet profile at independence. But which is more important: social/individual learning? 2/11

191

Tying all these results together, we conclude that the presence of allomaternal, or more generally, alloparental care (APC) can lead to the development of differentiated social relationships from a young age in social species. It would be interesting to investigate this in other species with APC!

We found that despite calves & other females (OF: females that do not show coordinated movement with the calf beyond that required to remain in the group) spending little time in proximity, OF initiated the greatest number of aggressive interactions with the calf. Calves rarely interacted with OF.

100

Similarly, we found that the calf behaved in a similar manner with its mother and the escort: responsibility in changes to proximity, duration of proximity, no. & class of interactions initiated (feeding, resting, social).

100

Using 1332 interactions between calves and conspecific females, we found that the escort initiated the greatest number of, specifically positive, interactions with the calf even more than the calf’s mother, despite the calf spending a similar proportion of their time near the mother & the escort.

100

T Revathe @revathe.bsky.social · 14/05/2025

We studied calf-conspecific female interactions by first categorizing females as mothers, escorts & other females, as there was coordinated movement between calves & only certain females in the group. This includes the calf’s mother & certain female(s) w/o a dependent young, whom we call ‘escort’.

100

We then fit 2 sets of models - one where we controlled for all the known confounding effects and one where we didn't. Not controlling for them led to false +ve or false -ve individual variation. There are a lot of factors that can affect the estimation of individual variation or "personality"!

Graph showing the estimated variation among Sumatran orangutan mothers when confounding effects are controlled vs not controlled for six maternal behaviours.

T Revathe @revathe.bsky.social · 14/05/2025

So, on average, mothers are different from one another, but what about their behavioural trajectories - or plasticity - during infant development? Well, we found that mothers consistently differed from one another in how they adjusted their behaviour as their infants grew!

Graphs showing maternal behavioural trajectory over infant age for two infants of the same mother and two infants of different mothers in Sumatran orangutans (behaviour: close proximity termination).

T Revathe @revathe.bsky.social · 14/05/2025

We show that over the developmental period (~8 years), mothers differ from one another in four maternal behaviours. They remained consistent in these differences across their different infants and even when we accounted for the various predictors of maternal behaviour.

Graphs showing variation among mothers and infants in four maternal behaviours in Sumatran orangutans: contact termination, close proximity termination, carrying, and feeding in close proximity