Animals learn and integrate information in constantly changing environments to enhance their survival and fitness. Cognitive abilities constitute adaptive traits that allow organisms to learn and integrate information to perform tasks, such as foraging, nesting, selecting breeding sites, avoiding predators, acquiring mates, or communicating optimally. The cognitive abilities of primates, including humans, and other vertebrates have been extensively explored (Kummer and Goodall, 1985; Marler, 1991; Bateson, 2002; Sol et al., 2002; Sol et al., 2005; Reader and Laland, 2003; Amiel et al., 2011; Benson-Amram and Holekamp, 2012; Cole and Quinn, 2012; Pritchard et al., 2016; Preiszner et al., 2017; Huebner et al., 2018; Chow et al., 2021; Santana and Garcia-Mijares, 2022).

The cognitive capacities of insects, which comprise the largest group of invertebrates, have, however, been explored to a much lesser extent, possibly due to their simple nervous systems and miniature brains (Chittka and Rossi, 2022). Nevertheless, recent findings suggest that the complexity of insect societies is not solely based on innate behavior, as they exhibit some features of social intelligence, like vertebrates, including recognition of individuals, object manipulation learning, cultural traditions, and emotion-like states. This highlights the need to explore the cognitive abilities of social insects in unusual situations, which could even be experimentally generated but possibly never examined earlier (Giurfa et al., 2001; Giurfa, 2013; Giurfa, 2019; Reznikova, 2018; Chittka and Rossi, 2022).

Ants are eusocial insects and all 12,000 known species of ants exhibit the hallmarks of eusociality, including reproductive division of labor, cooperative alloparental brood care, and overlapping generations (Hölldobler and Wilson, 1990). Ants, living in complex societies, are expected to have faced many challenges, but they appear to have been successful over their 150 million years on Earth, despite the changing environments that they must have confronted (Cole, 1985; Anderson and McShea, 2001). Some of the insights into the cognitive abilities of Hymenoptera have been obtained, for example, the cognitive differences between ants performing mass foraging and solitary foraging. Solitary hunters, such as members of the genus Myrmecia and subgenus Serviformica, are much more flexible than mass-foraging ants, as they independently search for food and make important decisions (Reznikova, 2018). Honeybee foragers can count, optimize paths, innovate, and even self-evaluate their chances of solving a task (Giurfa et al., 2001; Lihoreau et al., 2012; Perry and Barron, 2013; Loukola et al., 2017; Howard et al., 2018). Experiments have also been performed on insects under laboratory conditions to understand their ability to learn and retain their memory of coping with various environmental challenges (Papaj and Lewis, 1992; Collett et al., 2013; Giurfa, 2013; , 2019; Feinerman and Korman, 2017; Perry et al., 2017).

However, many laboratory-based mechanistic studies fail to consider the ecological contexts, within which animals have to solve specific problems. This limitation reduces the scope of insect-based cognition research in several ways, as it may hinder the ability of insects to perform cognitive tasks, as observed in Drosophila, honeybees, and bumblebees (Bateson et al., 2011; Yang et al., 2013). The current opinion in insect cognition research thus emphasizes prioritizing ecologically inspired experiments, which can be generated by closely observing insects in their natural habitat and understanding their real-life problems. Such experiments should be relevant to the umwelt of the organisms and should pose meaningful challenges (Lihoreau et al., 2019). At the collective level, for instance, social insects, such as bees and ants, generally make precise decisions for selecting their nesting site, sources of food, and solving mazes or transporting large food items as a group, rather than alone (Goss et al., 1989; Beckers et al., 1990; Czaczkes and Ratnicks, 2013; Sasaki et al., 2013; Mukhopadhyay et al., 2019). Hence, there is a need to understand the mechanisms that enhance performance in animal groups, when performing such tasks; some factors, such as learning from other group members or increased complementarity among group members, may contribute to such improvement in task performance (Collet et al., 2023).

The current study draws inspiration from observing Diacamma indicum colonies in their natural habitat. D. indicum is a metallic black, one-centimeter-long ant, found in India, Sri Lanka, and possibly Japan. They reside in colonies of approximately 85 adult ants, brood, and a single reproductively active female, known as the gamergate (Annagiri, 2021). When there is a disturbance at their nest or a better nesting option available, they relocate to a new nest, using tandem running recruitment. In tandem running, an informed ant (“leader”) recruits only one nestmate (“follower”) and walks her to the destination while maintaining tactile contact throughout the journey. This form of recruitment is used to bring nest mates to the new nest, to food sources or to slave colonies to raid them (Wilson, 1959; Franklin, 2014; Kaur et al., 2017).

During our earlier observations on colony relocation, we observed several tandem leaders and their followers accumulating at a narrow concrete crevasse, forming the new nest’s entrance, as it was a challenge to enter this narrow opening with their pupa (MP, pers. obs.). Generally, when ants relocate into a subterranean nest, the entrance dimension and other features of the nest, such as the tunnels and chambers, are often enlarged by excavating the surrounding soil. However, this may not always be possible, as we have seen in D. indicum colonies, residing at other locations, such as crevices inside tree trunks, bamboo stems, brick walls, or concrete slabs, which may pose various degrees of challenges to the ants in their efforts to redesign their new nest (Bhattacharyya et al., 2021). Observing the colony’s struggle to insert the pupa, we were inspired to design an experiment that utilized a narrow nest entrance, thus posing a cognitive challenge to the relocating colonies.

Our experimental design was unique in several ways. First, relocation is a goal-oriented and self-motivated process that forms an essential component in the life history of several social insects. It is often an adaptive response to factors, such as competition, nest deterioration, or colony expansion (McGlynn, 2012). Delaying the relocation process would expose the colony to both biotic and abiotic stress for relatively longer periods of time. Therefore, it becomes crucial to execute the relocation process efficiently to minimize the colony’s exposure to stress factors. Second, unlike other learning experiments, no training sessions were conducted. Colonies experienced the challenge at the start of relocation and had an opportunity to perform insertion as many times as the number of pupae within their colony. They were then tested with additional pupae in the control and manipulation experiments. Third, instead of using food, the most commonly used reward in cognition experiments, we utilized pupae, the most valuable item among the brood, as the source of motivation. Pupae require urgent and careful attention, as any damage to them during transportation would result in the loss of the investment in time, social care, and resources for the colony. Unlike food, which can be carried back piecemeal, pupae have to be kept intact. Pupae eventually eclose to forming the next generation of workers or reproductives for the colony and are thus linked to the colony’s fitness, making it unlikely that they would be abandoned. We, therefore, hypothesized that pupa transporters are likely to face much higher selection pressures than would food transporters because more is at stake for the former.

The combination of these features, we believe, creates a new paradigm to explore the cognitive abilities of these superorganisms, in general. More specifically, we examined how individual D. indicum ants and colonies were able to address this cognitive challenge and the manner in which they solved the problem.

Behavioral innovations are critical for organisms to explore, access, and utilize new resources, enabling them to survive in changing environments and occupy new ecological niches (Reader and Laland, 2003; Sol et al., 2005; Liker and Bókony, 2009; Benson et al., 2012; Thornton and Samson, 2012; Griffin and Guez, 2014; Reader et al., 2016). Although extensive research has been conducted on behavioral innovation and problem-solving in primates and birds, relatively few studies have been conducted on social insects (Reader and Laland, 2003; Sol et al., 2005). Jean-Henri Faber, a pioneering scientist, thus wrote, “What happens in this little brain of Hymenoptera? Are their abilities similar to ours? Is there a thought? What problem, if we could solve it; what chapter of psychology, if we could write it!” (Translated from Fabre, 1882).

To explore the cognitive abilities of Hymenopteran insects, we therefore investigated whether D. indicum ants could overcome a cognitive challenge that confronted them. We used a novel paradigm that did not involve separate training or rewards and found that these ants could cooperate and successfully deciphered how to insert pupae through narrow nest entrances. Furthermore, their efficiency improved with experience, at both the individual and colony levels. It is true that not all animals can overcome challenges posed to them, as even motivated individuals may face difficulties in problem-solving, due to a lack of suitable motor action or physical abilities (Griffin and Guez, 2014). However, D. indicum ants overcame the challenge by performing a novel emergent behavior – a cooperative insertion of pupa, with one ant pushing and the other ant(s) pulling it from inside the nest. This behavior is distinctively novel and innovative, as D. indicum is known to be a solitary forager and transporter of food, and these ants typically transport brood individually (Kaur and Sumana, 2014).

Typically, in the case of a new nest entrance that is narrower than optimum, ants try to widen the entrance (SA, pers. obs.). When the substrate is soft, as for soil or decaying wood, this behavior can successfully enlarge the nest entrance by mandibular action. However, in the current set-up, the nest entrance, constructed from rigid plastic, did not allow the ants’ mandibles to modify it and thus, this became a hindrance to brood transportation. However, ants continued to bite the edges of the nest entrance, probably trying to solve the new problem with an existing solution, as is the case with higher primates that use already existing behavioral patterns to solve new problems (Kummer and Goodall, 1985). In the manipulation experiments, the ants took four times longer to relocate the colony, indicating that the narrow nest entrance indeed posed a challenge. Transportation from the old nest to the roof of the new nest was, however, qualitatively similar to what occurred in the control experiments. Ants were seen crowding the entrance and accumulating on the roof, with pupae held in their mandibles ( Supplementary Figure 4 ), as they repeatedly tried to enter through the narrow entrance, still holding the pupae. Occasionally, after multiple unsuccessful trials, they would leave the pupae on the roof, enter the new nest, explore it, and then return to attempt the pupal insertion again. As ants neither had prior exposure to this challenge nor received any reward from experimenters for successfully inserting the pupae, their self-motivation and innovation were examined.

Interestingly, although individual ants attempted to introduce a pupa into the narrow nest entrance, they often could not complete the task. Due to the unsuitable alignment, the pupa got stuck at the entrance. Even when the ant aligned a pupa correctly and inserted more than 60% of the pupae into the entrance, their inability to let go of the pupae from their mandibles prevented them from successfully posting the pupa. Thus, only a small percentage of pupal insertions were through individual transport (ITPP mode). In the control experiments, all pupal transport was performed by individual transporters holding a pupa each (ITHP) but, in the manipulation experiments, the majority of the insertions were through cooperative transport, in which one ant pushed the pupa through the nest entrance while another ant pulled that pupa from inside the nest (COIP). Such a shift in behavior can be truly termed “innovation”. Cooperative transport was, however, not seen in this species for pupae, as the control experiments did not show a single cooperative insert. On facing a challenge, however, the ants showed “a new or modified learned behavior not previously found in a population”, which is the definition of innovation, as recorded in other social insects like bumblebees and Aphaenogaster ants (Fellers and Fellers, 1976; Reader and Laland, 2003; Kaufman and Kaufman, 2015; Chittka and Rossi, 2022). To understand the possible steps involved in this innovation, we thus constructed a flow chart, inspired by Kaufman and Kaufman (2015) ( Figure 8 ).

During pupal transportation, only one ant acted as the pusher most of the time while one to three ants participated in pulling the pupa from inside the new nest. The most efficient method of pupa transportation involved a single puller, as the time taken to insert a pupa was lowest with a combination of one pusher and one puller ( Supplementary Figure S3 ). Previous research on Novomessor cockerelli ants, transporting food, indicated that they predominantly pulled the load individually and also cooperatively (Buffin and Pratt, 2016). However, in the current experiment with D. indicum ants, both pushing and pulling were performed equally by individual ants, possibly because the pupae were relatively light but required careful handling to avoid damage during transportation. During the experiments, in fact, no pupa was observed to be damaged or cannibalized by nestmates. Even though minor injuries may not have been detected by unaided human eyes, ants typically notice any such damage or infection and cannibalize the injured brood (Maák et al., 2020). It was observed, nevertheless, that most ants specialized either as pushers or pullers, with only about 40% of the ants exhibiting behavioral flexibility by performing both tasks.

An increase in latency has been noted when birds or primates perform problem-solving tasks (Griffin and Guez, 2014). Similarly, in the current experiment, our study ants showed a relatively higher latency for pupal transport through the narrow entrance of the nest. During the relocation phase, the crowding, seen at the entrance, with multiple workers entering and exiting the entrance for a tandem run, may have increased the latency for pupal transportation. However, the main factor, influencing latency, appeared to be the inability of the ants to transport pupae through the narrow entrance, as multiple individuals kept trying to do so, leading to several failed attempts. To quantify this, we analyzed the number of failed attempts before the focal ant successfully inserted a given pupa. At the individual level, we found that ants, in their third pupal insertion, took 19% fewer attempts, indicating that they improved their performance in this task with time. However, a significant improvement was not seen when the first six attempts by individual ants were examined, possibly due to its relatively small sample size or its complex dynamics while these insertions were performed, as staged in a natural context.

Further, at the colony level, we analyzed the number of attempts required to insert the pupae across the sequence of successful insertions and found a significantly negative correlation during the test phase. A comparable improvement was, however, not seen during the relocation phase, possibly because of all the other activities that simultaneously occurred at the nest entrance during this time. The colonies had thus clearly learned to perform this task better with experience.

In conclusion, the current study aimed to investigate the problem-solving abilities of an ant society, using a nature-inspired design that utilized its brood as motivation. Our findings suggest that these innovative ants could indeed solve the given problems and even improve their performance with experience. Generalizing further, the ants’ capabilities to innovate and adapt to changing environments could be expected to significantly impact their survival, particularly in circumstances, such as relocation, which threaten the whole colony, including the brood. Further studies involving other social insects and ecologically-inspired designs that consider the organisms’ umwelt may uncover more insights into the intricate workings of these superorganism brains.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

This study was conducted under the prevalent animal care regulations in India. The model organism was not endangered and was not collected from any legally protected area. After experimentation, the ant colonies were released back into their natural habitat.

The idea for the experiment emerged from an observation made by MP. Conceptualization was by SA. Fund acquisition was by SA. The experimental design was by SA, MP, and SH. Data acquisition was conducted by MP and SH. Interpretation of data was by SA, SM, MP, and SH. Statistical analysis was done by SM and MP. Visualization was done by SH, MP, and SA. The original draft of the manuscript was prepared by MP, and it was reviewed and edited by SA. Supervision was by SA. All authors contributed to the article and approved the submitted version.