Recordings Illustrate How Meerkats Communicate in Group Contexts

By Jillian Kunze

From flocking birds and schooling fish to cooperation among animal groups in order to band against common threats, coordination is widespread throughout nature. “All of these processes are underlain by communication,” Ariana Strandburg-Peshkin (University of Konstanz and Max Planck Institute of Animal Behavior) said. “This may be passive in a lot of cases, in the form of observing others’ behavior, and can emerge without any explicit signaling.” But active signaling is indeed present in many natural systems. During her minisymposium presentation at the 2021 SIAM Conference on Applications of Dynamical Systems, which took place virtually in May, Strandburg-Peshkin focused on the impact of vocal communication on coordination within animal groups.

Vocal signals can be used to help group-living animals maintain cohesion while on the move, to initiate a group’s departure, and to recruit others to join in a collective action. Strandburg-Peshkin aims to address the dynamics of signal exchange in animal groups and the interaction of that process with other kinds of behaviors, such as decisions regarding movement. To do so, she studies a spectrum of different species and collaborates with an international team that performs long-term studies of animal behavior.

During her presentation, Strandburg-Peshkin focused on communication among meerkats, a social mongoose found in the southern part of Africa. To stay together as they travel distances of up to several kilometers, meerkats must communicate with other members of their group. This is accomplished through their frequent exchange of calls. “They are very vocal, engaged in a constant back-and-forth exchange between individuals,” Strandburg-Peshkin said. “They use this to coordinate movement.” Meerkats exhibit over 30 different types of calls, which are well-characterized through the work of Marta Manser of the Kalahari Meerkat Project, a long-term study of meerkats in South Africa headed by Manser and Tim Clutton-Brock. That study has habituated multiple groups of meerkats to human presence, which enables scientists to study them at close range. 

Strandburg-Peshkin is interested in the interplay between signal exchange and collective movement by meerkats, which happen together on similar timescales. The signal exchange is essentially the dynamics on the network of meerkats, while the collective movement is the dynamics of the network. A meerkat is influenced by where it sees its group mates and what those other meerkats say, and in turn influences the network through its own vocalizations. 

A simple context in which to explore this idea is sunning, when meerkats stand stationary in the morning sun to warm up. Since there is no movement in this case, it is only necessary to consider the vocal aspect of the dynamics. Sunning meerkats often give short, soft calls, but only when other meerkats are around. The function of these calls, however, is unknown. 

Strandburg-Peshkin teamed up with collaborators Vlad Demartsev, Michaela Ruffner, and Marta Manser to understand the temporal dynamics of these exchanges and determine whether the meerkats were taking turns in vocalizing. Turn-taking may be a key part of effective communication; humans fundamentally take turns when speaking during a conversation in a process that is universal across all languages, including sign languages. While many animal species also vocalize in turns, this phenomenon has mainly been studied in pairs such as during mating displays. 

To investigate turn-taking by meerkats, Strandburg-Peshkin and her collaborators used recordings of meerkat vocalizations taken while the meerkats were sunning outside their burrow (see Figure 1). “We are basically able to interview the meerkats with a boom mic,” Strandburg-Peshkin said. The recordings that they collected contained both the focal calls made by the meerkat being recorded and background calls from other meerkats in the group. The researchers looked at the rate of overlap of these focal and background calls—i.e., how often the two types of calls happened at the same time—to determine whether the meerkats were indeed taking turns (see Figure 2).

To test whether the amount of overlap was more or less than should be expected, the researchers shifted the background calls in time such that they occurred either earlier or later than what was really recorded, then calculated the new overlap rates. “If the meerkats avoid overlap, this overlap rate should be minimized at a time shift of zero,” she said. And that is indeed what the data showed (see Figure 3) — the rate of overlap was lowest for the original data as compared to the time shifted data.

The data also revealed that when an individual meerkat called, the others became more likely to call in response. But not immediately — for a short time window after a call, the call rate of other individuals is suppressed. But after a slightly longer time window, the call rate becomes larger than it would be on average. The meerkats seem to be waiting to avoid overlapping with the initial vocalizer, then responding to the call. 

Studying a whole group of moving animals is not easy. Human observers can easily focus on one animal, but it becomes difficult to record the actions of multiple animals at once. Strandburg-Peshkin and her collaborators found a technical solution to this problem: a collar with an attached global position system and audio recorder. Since the meerkats at the Kalahari Meerkat Project are comfortable with humans, the researchers could put the collars on them while they were relaxed. The researchers were particularly interested in what are called close calls: vocalizations that meerkats produce while foraging to maintain cohesion with their group. Bringing together the movement and vocalization data from the collars created a more complete picture of these processes’ dynamical interactions. 

The team is now trying to understand how individuals respond to calls and how their responses vary with different configurations of the group. To start, they looked at the vocal interactions between only two meerkats and collected the average pattern of interactions and responses to calls (see Figure 4). Preliminary results suggest that at a short distance of zero to two meters between the meerkats, once an individual calls, the other becomes more likely to respond, demonstrating call-and-response dynamics at close range. However, this form of analysis did not show any evidence for turn-taking or overlap avoidance as had been seen in the sunning context.

At longer distances of two to five meters, there was still evidence for responses to calls, but not as strongly and slightly delayed. At even further distances, the response dropped to the background level, with no real response dynamics. These results align with a mechanism that was first proposed by Gabriella Gall and Marta Manser in 2017, in which calls by one meerkat stimulate further calls by other meerkats to create a local vocal hotspot. This attracts other meerkats, maintaining group cohesion. 

Moving forward, Strandburg-Peshkin and her collaborators hope to extend the analysis from interactions between two individuals to a more network-oriented approach. There are a number of additional interesting research directions, such as understanding how the vocal landscape and individual movement dynamics influence each other, what role leadership plays, and how calls coordinate transitions to different activities. Insights from dynamical systems and neuroscience will hopefully shed more light on these aspects of animal communication.

Jillian Kunze is the associate editor of SIAM News.