When it comes to flocking together, homing pigeons use a simple strategy to find better ways home, according to a recent report.
The study, published previously as a Reviewed Preprint in eLife and appearing today as the final Version of Record, suggests that homing pigeons use simple route averaging when navigating as a group. eLife’s editors say the work addresses an important question, and provides compelling evidence based on multiple models and data on how homing pigeons can generate social routes from solitary ones.
The findings open avenues for future research to investigate the evolution of the mechanisms used by homing pigeons and other social animals when deciding on the best route to travel.
How animals navigate complex environments depends on their cognitive abilities. When travelling in groups, some animals pool individual information to improve their navigation. This can be achieved by following experienced leaders, which requires recognising the experts of the group, or by using simpler mechanisms, such as the ‘wisdom of crowds’ principle, which average the routes of all individuals. These strategies therefore range from cognitively complex to simple, but their prevalence or interplay in nature remains unexplored.
“This is where the homing pigeon comes in: as a social species that have been studied extensively for their ability to develop and recall routes, these birds are an ideal model organism for studying navigational strategies,” says author Shoubhik Banerjee, a PhD student in senior author Albert Kao’s lab at the University of Massachusetts Boston (UMass Boston), US. Banerjee and Kao conducted the study with Postdoctoral Researcher Fritz Francisco, also a member of the Kao Lab.
A previous study published in 2017* showed that pairs of experienced and naïve homing pigeons could continuously improve their homing routes over the course of the experiment. The study proposed the key driver to be cumulative cultural evolution (CCE), where chains of birds improve their routes by exploring different options and choosing better ones. However, a detailed mechanistic understanding of how these route improvements emerge is still lacking.
“Building on that study, we aimed to investigate the mechanisms that pigeons use to improve their route efficiency and whether those mechanisms fall under the criteria required for CCE,” Banerjee adds.
The previous work involved creating ‘chains’ of birds, similar to a game of telephone, and allowing them to fly back home repeatedly from a release site 8.4km away. Each chain was composed of five ‘generations’ and included an experienced bird that knew about the homing task from the previous generation, paired with a naïve bird that lacked this information. At the end of the generation (12 flights), the experienced pigeon was replaced with a new, naïve pigeon that travelled with the remaining, now-experienced bird. Working as control groups, solo and fixed pairs of birds carried out the same amount of flights as the experimental group (a total of 60 flights). The study found that the experimental chains of birds significantly outperformed both the solo and fixed pair controls by the end of the fifth generation – a result attributed to CCE.
Banerjee, Francisco and Kao set out to explore which navigation mechanism is necessary and sufficient to replicate those experimental results. They developed seven plausible learning mechanisms, categorised into three types with increasing cognitive complexity. The first type represents the simplest process, where birds have no knowledge of their partner’s level of experience or performance, and includes only the averaging strategy. The second type assumes that birds can recognise the more experienced individual in the pair and maximise their performance using this knowledge. And the third type introduces the highest level of cognitive complexity, which requires birds to actively evaluate their individual or paired performance, aligning with the mechanistic criteria required for CCE.
The team then compared the results of the seven mechanisms with the experimental data to identify which strategies are most likely to be used by real birds. In particular, they explored whether the cognitive requirements of CCE are necessary for the observed improvement in navigation ability.
They found that all of the strategies resulted in route improvements, regardless of their underlying complexity, which suggests that a wide range of decision-making mechanisms can lead to navigational improvements – not just the ones compatible with the definition of CCE. However, when they combined the results with those from a social weight analysis, they found that the experimental data aligned best overall with the simplest strategy: averaging individual routes.
“We show that an improvement in route efficiency alone is not sufficient evidence for cultural transmission, as the experimental birds did not demonstrate some of the criteria of CCE,” says author Fritz Francisco. “This could be due to the wisdom of crowds improving routes ‘for free’ without placing additional cognitive load on the birds. On average, birds in this experiment influenced each other’s routes equally, disregarding any differences in experience, which raises broader questions about which social learning mechanisms truly align with the requirements for CCE.”
The team further observed that mixed strategies, while not supported by the experimental data, theoretically combined advantages from both averaging and active selection of better routes, resulting in even greater performance.
“Our results therefore pave the way for future studies to investigate the evolution of social learning and trade-offs among the different decision mechanisms that may be available to animals in the wild,” concludes senior author Albert Kao, Assistant Professor and Principal Investigator at UMass Boston.
“For this navigation task, simple averaging is sufficient to explain the experimental results in homing pigeons, but other tasks may be less amenable to the wisdom of crowds – there’s a lot of complexity in this area. It would be interesting to explore how collective navigation strategies evolve in different contexts, taking into account, for example, typical group sizes, error rates, and how many times a task is repeated, to better understand social decision-making in homing pigeons and other animals.”
*Sasaki T., Biro D.,. 2017. Cumulative culture can emerge from collective intelligence in animal groups. Nature Communications 8:15049. DOI: https://doi.org/10.1038/ncomms15049 .
##
Media contacts
Emily Packer, Media Relations Manager
eLife
+44 (0)1223 855373
About eLife
eLife is a non-profit organisation advancing open science by transforming how research is communicated, reviewed, and assessed. By developing open tools and collaborating with research communities, institutions, and funders, we are building a fairer, more effective global research ecosystem. In support of our mission, we introduced our publish-review-curate (PRC) model in which every article we review is published as a Reviewed Preprint that includes the article, feedback from the reviewers, and an eLife Assessment summarising the significance of the findings and the strength of the evidence. eLife is supported by the Howard Hughes Medical Institute , Knut and Alice Wallenberg Foundation , the Max Planck Society and Wellcome . Learn more at https://elifesciences.org/about .
To read the latest Ecology research in eLife, visit https://elifesciences.org/subjects/ecology .
And for the latest in Physics of Living Systems, see https://elifesciences.org/subjects/physics-living-systems .
Computational simulation/modeling
Animals
Cognitive simplicity drives collective route improvements in homing pigeons
26-May-2026