The third in the behind-the-scenes series for the Journal of Animal Ecology’s Animal Social Network Special Issue, this blog post is provided by Beth Preston and tells the #StoryBehindThePaper for the article “Network‐level consequences of outgroup threats in banded mongooses: Grooming and aggression between the sexes“, which was recently published in the Journal of Animal Ecology Animal Social Networks Special Issue.

Social networks are a great tool to use to answer loads of different science questions. Who doesn’t want to delve down into the friendships, frictions and general social drama of their favourite animal species!

My study species the banded mongoose are notoriously sociable. They live in tight-knit social groups with multiple males and females. The females synchronise their pregnancies and give birth on the same day (!) and then all of the adults club together to look after the new pups, babysitting them at the den, and then escorting them during foraging as they grow.

This all sounds very cuddly and lovely, but the mongooses are also very violent! When social groups meet whilst foraging or moving around their territories, violent battles break out. Groups often manoeuvre themselves into battle lines opposite each other to assess their enemies, but then chaos breaks out when they all pile in and start biting and scratching each other! Adults and pups alike can be gravely injured or killed in these fights, and mortality rates are comparable to chimpanzee and human violence.

In light of this I wanted to investigate the impact of these fights (which we call intergroup conflicts or encounters) on the social lives of the groups involved. Social network analysis seemed like a great tool to get into the nitty-gritty of how intergroup conflict can affect groups and their relationships.

Managing to find fights in the wild is tricky, let alone managing to track behaviour before and after natural fights. So to tackle these problems I simulated these intergroup conflicts by showing various stimuli to focal mongoose groups. I wanted to really capture the drama of encounters, and ensure that my simulations covered as many senses as possible, following the patterns of real fights.

First, I presented faeces, urine and scent marks of a rival group. Luckily, mongooses really love to pee, poo and rub themselves all over plastic! That meant that by carrying around sheets of plastic and popping them in front of our target mongooses we could collect these scents very easily!

Next – sound! I recorded “war cries” a very distinctive call type that mongooses make when they encounter other mongoose groups. I then played these back from a tiny speaker that I hid in bushes, to simulate the sound of a rival group. The mongooses went crazy for this, and despite my best efforts often found the speaker and scent-marked all over it!

For the finale, I presented the mongoose group with four adult male mongooses, safely kept in traps, for five minutes. This was to provide the full experience of an invasion from a group of rival mongooses – but kept to only five minutes to avoid any unnecessary stress. The trapped mongooses held their own war-crying and snapping and rushing at the mongooses outside the traps rather than cowering in the middle out of reach of their rivals.

This was just one small part of the experiment though, each of the trials of the experiment took five days in the field. For the first two days I simply observed the behaviour of the mongooses. For this study the focus was on social networks – so I recorded every interaction between individuals under two categories – grooming and aggression. When two mongooses groomed each other I recorded their identities, and whether the grooming was from one mongoose to the other or reciprocal, so both ways. You might think – wow that sounds tough, how do you tell them apart? And you’d be right, the data collection was very intensive and it was a scramble to record all of the interactions, as often mongooses gather in a group to relax and groom each other, switching partners rapidly. Luckily Solomon (a co-author on the paper) has been working with the mongooses for 20 years and is great at identifying them very quickly. The mongooses also have handy shave marks on their backs to help us to identify them, and although it’s a steep learning curve you soon learn to tell them apart! Similarly when mongooses were aggressive to each other I recorded the individuals and the direction of the interaction.

Putting all of these interactions together I could build a network of relationships between all the individuals in a group. This gave a baseline network for each group. Then on the third day, the simulated intrusion with all the scents, calls and live intruders were presented. Next on the fourth and fifth days the behavioural observations were repeated and a second “post-conflict” network was made from all these interactions.

Each trial of the experiment created two networks to compare, leading to over 80 social networks for the mongooses in my study! I looked at various network measures to try and assess changes in relationships after exposure to a conflict simulation. I looked at eigenvector centrality – a measure that is often used to represent social cohesion – as many other studies have suggested that when there is an external threat groups should pull together and become more tight-knit and cohesive. I also looked at strength of the relationships – which is what it sounds like – how strong relationships between individuals are.

For each individual we calculated the change in their network measures from before to after a conflict, and explored whether this was linked to their sex or age. We found some interesting results. Eigenvector centrality, and therefore social cohesion, didn’t change for either males or females, but we did find that males reduced grooming towards other males and females, and females reduced their grooming towards males. Females kept grooming females at the same rate though. Aggression within the group was also affected by exposure to external threat, with males reducing their aggression towards females. Other aggression rates stayed the same.

Mongooses clearly react differently to each other when exposed to another group, and it seems that males and females react quite differently to each other. We have a number of suggestions as to why this might be. Males are much more likely to die in these fights, whereas females can actually benefit through sneaky mating with males in the other group, this changes the costs and benefits associated with fights which might well impact their within-group behaviour. There’s still a lot more to find out, and you can follow the Banded Mongoose Research Project here to follow the latest research.

Read the paper

Read the full paper here: Preston, EFR, Thompson, FJ, Ellis, S, Kyambulima, S, Croft, DP, Cant, MA. Network‐level consequences of outgroup threats in banded mongooses: Grooming and aggression between the sexes. J Anim Ecol. 2021; 90: 153– 167. https://doi.org/10.1111/1365-2656.13323