Feral horses keep their distance — Neither too close nor too far

A research group consisting of Special Postdoctoral Researcher Tamao Maeda at the Research Center for Integrative Evolutionary Science, the Graduate University for Advanced Studies (SOKENDAI), and researchers from Kyoto University have announced findings showing that feral horses in Portugal live in large groups made up of multiple non-territorial groups (units), and that these units never mix with one another, maintaining a distance that is neither too close nor too far from other units. The researchers used drones to observe 25 units and also found that when units are about to come into contact, they dynamically adjust their spatial arrangement by changing the shape of the unit depending on the distance. The results were published in Proceedings of the Royal Society B on January 21.

Many animals establish "territories" or use similar strategies to minimize direct conflict, but how do "non-territorial" species respond when different groups occupy the same space? This has largely remained unclear.

Feral horses in Portugal do not hold clear territories and form a multilevel society—similar to human society—in which multiple family groups (units) consisting of males, females, and foals live together. In the case of horses, each unit typically comprises one to two males and multiple females.

Gathering in large multi-unit groups is thought to offer benefits such as protection from predators. However, males from different units are rivals, which means the horses face conflicting demands.

How horses maintain an appropriate distance from other units under these circumstances, and the mechanism behind this behavior, had not been understood.

In this study, the researchers used drones to observe 25 units and statistically analyzed changes in inter-unit distances and unit shapes.

The results showed that units never mixed with one another (with the exception of two special units).

To test whether this behavior was intentional, the team conducted a randomization analysis. The analysis showed that real horses statistically and significantly "avoided being near other units" and that, even when approached, "actively moved so as not to mix."

Further examination of unit shapes revealed that when another unit was nearby, units became more "circular" and members clustered closer together. When the distance narrowed further and units were on the verge of contact, the unit shape shifted to become "elongated," suggesting that the units were maneuvering to prevent their boundaries from overlapping.

The two special units were a unit led by a male named "Kobe" and a unit led by a male named "Uzumasa." These two units alone showed exceptional behavior: over 59 observations, members of the two units crossed boundaries and intermixed 21 times. Mixing between these two units was observed on multiple occasions during the 2016-2017 observation period. No similar behavior was observed in any other units at the same site. Why this particular pair was exceptional remains to be clarified, including factors such as kinship and long-term social bonds.

Maeda commented: "Most of the time, the horses we observed were either eating or sleeping, and at first glance they appeared to be leading a very relaxed life. However, this study suggests that even so, they are finding ways to ensure that other groups (units) do not come too close or mix with them. Going forward, I would like to investigate what information horses use and how they make decisions in carrying out this kind of spatial adjustment."

Journal Information
Publication: Proceedings of the Royal Society B
Title: Spatial strategies in non-territorial societies: how feral horses maintain boundaries with other groups
DOI: 10.1098/rspb.2025.2468