Can hook-bending be let off the hook? Bending/unbending of pliant tools by cockatoos

Hooded crows (Corvus cornix) manufacture objects relative to a mental template

It was recently found that not only tool-specialized New Caledonian crows, but also Goffin cockatoos can manufacture physical objects in accordance with a mental template. That is, they can emulate features of existing objects when they manufacture new items. Both species spontaneously ripped pieces of card into large strips if they had previously learned that a large template was rewarded, and small strips when they previously learned that a small template was rewarded. Among New Caledonian crows, this cognitive ability was suggested as a potential mechanism underlying the transmission of natural tool designs. Here, we tested for the same ability in another non-specialised tool user-Hooded crows (Corvus cornix). Crows were exposed to pre-made template objects, varying first in colour and then in size, and were rewarded only if they chose pre-made objects that matched the template. In subsequent tests, birds were given the opportunity to manufacture versions of these objects. All three crows ripped paper pieces from the same colour material as the rewarded template, and, crucially, also manufactured objects that were more similar in size to previously rewarded, than unrewarded, templates, despite the birds being rewarded at random in both tests. Therefore, we found the ability to manufacture physical objects relative to a mental template in yet another bird species not specialized in using or making foraging tools in the wild, but with a high level of brain and cognitive development.

Dunking rusk: innovative food soaking behaviour in Goffin's cockatoos (Cacatua goffiniana)

Foraging innovations in animals involving the processing of resources that are already edible in an unprocessed state, yet of improved quality in a processed state, are rare but important to study the evolution of food preparation. Here, we present the first scientific report of food dunking behaviours in parrots by Goffin's cockatoos, a model species for innovative problem solving. Observations during lunch showed seven out of 18 cockatoos placing their food into water and soaking it prior to consumption. This was largely done with dry rusk which was eaten almost exclusively when dunked. Furthermore, their transport effort and waiting times before retrieving food from the water indicate their willingness to invest considerable time to prepare a soaked rusk piece of a higher texture quality. Our present results suggest that the function of this behaviour is to soak the food. Because only some individuals dunked food and dunking has not been observed in the wild, we believe this to be a spontaneous foraging innovation either by one or multiple individuals.

Ratcheting up tool innovation in Goffin's cockatoos (Cacatua goffiniana): The effect of contextually diverse prior experience

The ability to gain information from one situation, acquire new skills and/or perfect existing ones, and subsequently apply them to a new situation is a key element in behavioural flexibility and a hallmark of innovation. A flexible agent is expected to store these skills and apply them to contexts different from that in which learning occurred. Goffin's cockatoos (Cacatua goffiniana) are highly innovative parrots renowned for their problem-solving and tool-using skills and are thus excellent candidates to study this phenomenon. We hypothesized that birds allowed to use a tool in a larger variety of contingencies would acquire a broader expertise in handling it, facilitating its transfer to new tasks. In our study, we compared the performance of two groups of captive Goffin's cockatoos (N = 13): A test group received more diverse learning and motor experiences on multiple applications of a hook-type tool, while a control group received intensive, total trial-matched, experience with a single application of the same tool. Then, both groups were tested on two novel tasks to determine whether experience with the tool in multiple contexts would facilitate performance during transfer. While both groups transferred to both novel tasks, group differences in performance were apparent, particularly in the second transfer task, where test birds achieved a higher success rate and reached criteria within fewer trials than control birds. These results provide support for the prediction that experiencing a diverse range of contingencies with a tool appears to allow birds to acquire generalizable knowledge and transferrable skills to tackle an untrained problem more efficiently. In contrast, intensive experience with the tool in a single context might have made control birds less flexible and more fixated on previously learned tool-dependent instances.

Innovative composite tool use by Goffin's cockatoos (Cacatua goffiniana)

Composite tool use (using more than one tool simultaneously to achieve an end) has played a significant role in the development of human technology. Typically, it depends on a number of specific and often complex spatial relations and there are thus very few reported cases in non-human animals (e.g., specific nut-cracking techniques in chimpanzees and capuchin monkeys). The innovative strategies underlying the innovation and spread of tool manufacture and associative tool use (using > 1 tools) across tool using animals is an important milestone towards a better understanding of the evolution of human technology. We tested Goffin's cockatoos on a composite tool problem, the 'Golf Club Task', that requires the use of two objects in combination (one used to control the free movement of a second) to get a reward. We demonstrate that these parrots can innovate composite tool use by actively controlling the position of the end effector and movement of both objects involved in a goal directed manner. The consistent use of different techniques by different subjects highlights the innovative nature of the individual solutions. To test whether the solution could be socially transmitted, we conducted a second study, which provided only tentative evidence for emulative learning. To our knowledge, this indicates that the cognitive preconditions for composite tool use have also evolved outside the primate lineage.

Current Understanding of the "Insight" Phenomenon Across Disciplines

Despite countless anecdotes and the historical significance of insight as a problem solving mechanism, its nature has long remained elusive. The conscious experience of insight is notoriously difficult to trace in non-verbal animals. Although studying insight has presented a significant challenge even to neurobiology and psychology, human neuroimaging studies have cleared the theoretical landscape, as they have begun to reveal the underlying mechanisms. The study of insight in non-human animals has, in contrast, remained limited to innovative adjustments to experimental designs within the classical approach of judging cognitive processes in animals, based on task performance. This leaves no apparent possibility of ending debates from different interpretations emerging from conflicting schools of thought. We believe that comparative cognition has thus much to gain by embracing advances from neuroscience and human cognitive psychology. We will review literature on insight (mainly human) and discuss the consequences of these findings to comparative cognition.

Goffin's cockatoos discriminate objects based on weight alone

Paying attention to weight is important when deciding upon an object's efficacy or value in various contexts (e.g. tool use, foraging). Proprioceptive discrimination learning, with objects that differ only in weight, has so far been investigated almost exclusively in primate species. Here, we show that while Goffin's cockatoos learn faster when additional colour cues are used, they can also quickly learn to discriminate between objects on the basis of their weight alone. Ultimately, the birds learned to discriminate between visually identical objects on the basis of weight much faster than primates, although methodological differences between tasks should be considered.

Captive great apes tend to innovate simple tool behaviors quickly

Recent studies have highlighted the important role that individual learning mechanisms and different forms of enhancenment play in the acquisition of novel behaviors by naïve individuals. A considerable subset of these studies has focused on tool innovation by our closest living relatives, the great apes, to better undestand the evolution of technology in our own lineage. To be able to isolate the role that individual learning plays in great ape tool innovation, researchers usually employ what are known as baseline tests. Although these baselines are commonly used in behavioral studies in captivity, the length of these tests in terms of number of trials and duration remains unstandarized across studies. To address this methodological issue, we conducted a literature review of great ape tool innovation studies conducted in zoological institutions and compiled various methodological data including the timing of innovation. Our literature review revealed an early innovation tendency in great apes, which was particularly pronounced when simple forms of tool use were investigated. In the majority of experiments where tool innovation took place, this occurred within the first trial and/or the first hour of testing. We discuss different possible sources of variation in the latency to innovate such as testing setup, species and task. We hope that our literature review helps researchers design more data-informed, resource-efficient experiments on tool innovation in our closest living relatives.

Individual Goffin´s cockatoos (Cacatua goffiniana) show flexible targeted helping in a tool transfer task

Flexible targeted helping is considered an advanced form of prosocial behavior in hominoids, as it requires the actor to assess different situations that a conspecific may be in, and to subsequently flexibly satisfy different needs of that partner depending on the nature of those situations. So far, apart from humans such behaviour has only been experimentally shown in chimpanzees and in Eurasian jays. Recent studies highlight the prosocial tendencies of several bird species, yet flexible targeted helping remained untested, largely due to methodological issues as such tasks are generally designed around tool-use, and very few bird species are capable of tool-use. Here, we tested Goffin's cockatoos, which proved to be skilled tool innovators in captivity, in a tool transfer task in which an actor had access to four different objects/tools and a partner to one of two different apparatuses that each required one of these tools to retrieve a reward. As expected from this species, we recorded playful object transfers across all conditions. Yet, importantly and similar to apes, three out of eight birds transferred the correct tool more often in the test condition than in a condition that also featured an apparatus but no partner. Furthermore, one of these birds transferred that correct tool first more often before transferring any other object in the test condition than in the no-partner condition, while the other two cockatoos were marginally non-significantly more likely to do so. Additionally, there was no difference in the likelihood of the correct tool being transferred first for either of the two apparatuses, suggesting that these birds flexibly adjusted what to transfer based on their partner´s need. Future studies should focus on explanations for the intra-specific variation of this behaviour, and should test other parrots and other large-brained birds to see how this can be generalized across the class and to investigate the evolutionary history of this trait.

Goffin's Cockatoos (Cacatua goffiniana) Can Solve a Novel Problem After Conflicting Past Experiences

Novel problems often partially overlap with familiar ones. Some features match the qualities of previous situations stored in long-term memory and therefore trigger their retrieval. Using relevant, while inhibiting irrelevant, memories to solve novel problems is a hallmark of behavioral flexibility in humans and has recently been demonstrated in great apes. This capacity has been proposed to promote technical innovativeness and thus warrants investigations of such a mechanism in other innovative species. Here, we show that proficient tool-users among Goffin's cockatoos-an innovative tool-using species-could use a relevant previous experience to solve a novel, partially overlapping problem, even despite a conflicting, potentially misleading, experience. This suggests that selecting relevant experiences over irrelevant experiences guides problem solving at least in some Goffin's cockatoos. Our result supports the hypothesis that flexible memory functions may promote technical innovations.

Extending the Reach of Tooling Theory: A Neurocognitive and Phylogenetic Perspective

Tool use research has suffered from a lack of consistent theoretical frameworks. There is a plethora of tool use definitions and the most widespread ones are so inclusive that the behaviors that fall under them arguably do not have much in common. The situation is aggravated by the prevalence of anecdotes, which have played an undue role in the literature. In order to provide a more rigorous foundation for research and to advance our understanding of the interrelation between tool use and cognition, we suggest the adoption of Fragaszy and Mangalam's (2018) tooling framework, which is characterized by the creation of a body-plus-object system that manages a mechanical interface between tool and surface. Tooling is limited to a narrower suite of behaviors than tool use, which might facilitate its neurocognitive investigation. Indeed, evidence in the literature indicates that tooling has distinct neurocognitive underpinnings not shared by other activities typically classified as tool use, at least in primates. In order to understand the extent of tooling incidences in previous research, we systematically surveyed the comprehensive tool use catalog by Shumaker et al. (2011). We identified 201 tool use submodes, of which only 81 could be classified as tooling, and the majority of the tool use examples across species were poorly supported by evidence. Furthermore, tooling appears to be phylogenetically less widespread than tool use, with the greatest variability found in the primate order. However, in order to confirm these findings and to understand the evolution and neurocognitive mechanisms of tooling, more systematic research will be required in the future, particularly with currently underrepresented taxa.

Problem Solving in Animals: Proposal for an Ontogenetic Perspective

Simple Summary

Animals must be able to solve problems to access food and avoid predators. Problem solving is not a complicated process, often relying only on animals exploring their surroundings, and being able to learn and remember information. However, not all species, populations, or even individuals, can solve problems, or can solve problems in the same way. Differences in problem-solving ability could be due to differences in how animals develop and grow, including differences in their genetics, hormones, age, and/or environmental conditions. Here, we consider how an animal’s problem-solving ability could be impacted by its development, and what future work needs to be done to understand the development of problem solving. We argue that, considering how many different factors are involved, focusing on individual animals, and individual variation, is the best way to study the development of problem solving.

Abstract

Problem solving, the act of overcoming an obstacle to obtain an incentive, has been studied in a wide variety of taxa, and is often based on simple strategies such as trial-and-error learning, instead of higher-order cognitive processes, such as insight. There are large variations in problem solving abilities between species, populations and individuals, and this variation could arise due to differences in development, and other intrinsic (genetic, neuroendocrine and aging) and extrinsic (environmental) factors. However, experimental studies investigating the ontogeny of problem solving are lacking. Here, we provide a comprehensive review of problem solving from an ontogenetic perspective. The focus is to highlight aspects of problem solving that have been overlooked in the current literature, and highlight why developmental influences of problem-solving ability are particularly important avenues for future investigation. We argue that the ultimate outcome of solving a problem is underpinned by interacting cognitive, physiological and behavioural components, all of which are affected by ontogenetic factors. We emphasise that, due to the large number of confounding ontogenetic influences, an individual-centric approach is important for a full understanding of the development of problem solving.

Convergent evolution of complex cognition: Insights from the field of avian cognition into the study of self-awareness

Pioneering research on avian behaviour and cognitive neuroscience have highlighted that avian species, mainly corvids and parrots, have a cognitive tool kit comparable with apes and other large-brained mammals, despite conspicuous differences in their neuroarchitecture. This cognitive tool kit is driven by convergent evolution, and consists of complex processes such as casual reasoning, behavioural flexibility, imagination, and prospection. Here, we review experimental studies in corvids and parrots that tested complex cognitive processes within this tool kit. We then provide experimental examples for the potential involvement of metacognitive skills in the expression of the cognitive tool kit. We further expand the discussion of cognitive and metacognitive abilities in avian species, suggesting that an integrated assessment of these processes, together with revised and multiple tasks of mirror self-recognition, might shed light on one of the most highly debated topics in the literature-self-awareness in animals. Comparing the use of multiple assessments of self-awareness within species and across taxa will provide a more informative, richer picture of the level of consciousness in different organisms.

String-pulling in the Goffin's cockatoo (Cacatua goffiniana)

Goffin’s cockatoos, a parrot species endemic to the Tanimbar Islands in Indonesia, demonstrate remarkable cognitive skills across various technical tasks. These neophilic extractive foragers explore objects with their beak and feet, and are skilled in several modes of tool use. In this study, we confronted the animals for the first time with a vertical string-pulling setup, including a set of classic and novel controls. Nine of the 12 subjects, two of which were subadults, immediately interacted with the single-string task, with seven individuals successfully obtaining the reward on their very first attempt. Four different double string discrimination tests with varying spatial relations were used to assess the Goffin’s cockatoos’ apprehension of basic physical task properties. We found significant differences in performance between the respective experimental conditions, as well as the development of side biases. The results suggest that while the birds seem to consider simple cause–effect relationships, there is no evidence for a mental representation of the causal mechanisms underlying the string-pulling tasks, as subjects failed the crossed strings condition out of immediate sight. Finally, we provide suggestions on improving the methodology, and discuss our findings in regard to the Goffin’s cockatoo’s ecology.

Innovative problem solving in macaws

Behavioural innovations with tool-like objects in non-habitually tool-using species are thought to require complex physical understanding, but the underlying cognitive processes remain poorly understood. A few parrot species are capable of innovating tool-use and borderline tool-use behaviours. We tested this capacity in two species of macaw (Ara ambiguus, n = 9; Ara glaucogularis, n = 8) to investigate if they could solve a problem-solving task through manufacture of a multi-stone construction. Specifically, after having functional experience with a pre-inserted stick tool to push a reward out of a horizontal tube, the subjects were required to insert five stones consecutively from one side to perform the same function as the stick tool with the resulting multi-component construction. One Ara glaucogularis solved the task and innovated the stone construction after the experience with the stick tool. Two more subjects (one of each species) did so after having further functional experience of a single stone pushing a reward out of a shortened tube. These subjects were able to consistently solve the task, but often made errors, for example counter-productive stone insertions from the opposing end, even in some of the successful trials. Conversely, multiple trials without errors also suggested a strong goal direction. Their performance in the follow-up tasks was inconclusive since they sometimes inserted stones into un-baited or blocked ‘dummy tubes’, but this could have been an attention-deficit behaviour as subjects had not encountered these ‘dummy tubes’ before. Overall, the successful subjects’ performance was so erratic that it proved difficult to conclude whether they had functional understanding of their multi-stone constructions.

Object manufacture based on a memorized template: Goffin's cockatoos attend to different model features

Although several nonhuman animals have the ability to recognize and match templates in computerized tasks, we know little about their ability to recall and then physically manufacture specific features of mental templates. Across three experiments, Goffin cockatoos (Cacatua goffiniana), a species that can use tools in captivity, were exposed to two pre-made template objects, varying in either colour, size (long or short) or shape (I or L-shaped), where only one template was rewarded. Birds were then given the opportunity to manufacture versions of these objects themselves. We found that all birds carved paper strips from the same colour material as the rewarded template, and half were also able to match the size of a template (long and short). This occurred despite the template being absent at test and birds being rewarded at random. However, we found no evidence that cockatoos could carve L-shaped pieces after learning that L-shaped templates were rewarded, though their manufactured strips were wider than in previous tests. Overall, our results show that Goffin cockatoos possess the ability to physically adjust at least the size dimension of manufactured objects relative to a mental template. This ability has previously only been shown in New Caledonian crows, where template matching was suggested as a potential mechanism allowing for the cumulative cultural transmission of tool designs. Our results show that within avian tool users, the ability to recreate a physical template from memory does not seem to be restricted to species that have cumulative tool cultures.

Exploring the role of individual learning in animal tool-use

The notion that tool-use is unique to humans has long been refuted by the growing number of observations of animals using tools across various contexts. Yet, the mechanisms behind the emergence and sustenance of these tool-use repertoires are still heavily debated. We argue that the current animal behaviour literature is biased towards a social learning approach, in which animal, and in particular primate, tool-use repertoires are thought to require social learning mechanisms (copying variants of social learning are most often invoked). However, concrete evidence for a widespread dependency on social learning is still lacking. On the other hand, a growing body of observational and experimental data demonstrates that various animal species are capable of acquiring the forms of their tool-use behaviours via individual learning, with (non-copying) social learning regulating the frequencies of the behavioural forms within (and, indirectly, between) groups. As a first outline of the extent of the role of individual learning in animal tool-use, a literature review of reports of the spontaneous acquisition of animal tool-use behaviours was carried out across observational and experimental studies. The results of this review suggest that perhaps due to the pervasive focus on social learning in the literature, accounts of the individual learning of tool-use forms by naïve animals may have been largely overlooked, and their importance under-examined.

Old and New Approaches to Animal Cognition: There Is Not "One Cognition"

Using the comparative approach, researchers draw inferences about the evolution of cognition. Psychologists have postulated several hypotheses to explain why certain species are cognitively more flexible than others, and these hypotheses assume that certain cognitive skills are linked together to create a generally "smart" species. However, empirical findings suggest that several animal species are highly specialized, showing exceptional skills in single cognitive domains while performing poorly in others. Although some cognitive skills may indeed overlap, we cannot a priori assume that they do across species. We argue that the term "cognition" has often been used by applying an anthropocentric viewpoint rather than a biocentric one. As a result, researchers tend to overrate cognitive skills that are human-like and assume that certain skills cluster together in other animals as they do in our own species. In this paper, we emphasize that specific physical and social environments create selection pressures that lead to the evolution of certain cognitive adaptations. Skills such as following the pointing gesture, tool-use, perspective-taking, or the ability to cooperate evolve independently from each other as a concrete result of specific selection pressures, and thus have appeared in distantly related species. Thus, there is not "one cognition". Our argument is founded upon traditional Darwinian thinking, which-although always at the forefront of biology-has sometimes been neglected in animal cognition research. In accordance with the biocentric approach, we advocate a broader empirical perspective as we are convinced that to better understand animal minds, comparative researchers should focus much more on questions and experiments that are ecologically valid. We should investigate nonhuman cognition for its own sake, not only in comparison to the human model.

Using an Innovation Arena to compare wild-caught and laboratory Goffin's cockatoos

The ability to innovate, i.e., to exhibit new or modified learned behaviours, can facilitate adaptation to environmental changes or exploiting novel resources. We hereby introduce a comparative approach for studying innovation rate, the ‘Innovation Arena’ (IA), featuring the simultaneous presentation of 20 interchangeable tasks, which subjects encounter repeatedly. The new design allows for the experimental study of innovation per time unit and for uncovering group-specific problem-solving abilities – an important feature for comparing animals with different predispositions and life histories. We applied the IA for the first time to investigate how long-term captivity affects innovative capacities in the Goffin’s cockatoo, an avian model species for animal innovation. We found that fewer temporarily-captive wild birds are inclined to consistently interact with the apparatus in comparison to laboratory-raised birds. However, those that are interested solve a similar number of tasks at a similar rate, indicating no difference in the cognitive ability to solve technical problems. Our findings thus provide a contrast to previous literature, which suggested enhanced cognitive abilities and technical problem-solving skills in long-term captive animals. We discuss the impact and discrepancy between motivation and cognitive ability on innovation rate. Our findings contribute to the debate on how captivity affects innovation in animals.

Innovative problem solving in great apes: the role of visual feedback in the floating peanut task

Nonhuman great apes show remarkable behavioural flexibility. Some individuals are even able to use water as a tool: They spit water into a vertical tube to make a peanut float upwards until it comes into reach (floating peanut task; FPT). In the current study, we used the FPT to investigate how visual feedback, an end-state demonstration and a social demonstration affect task performance in nonhuman great apes in three experiments. Our results indicate that apes who had acquired the solution with a clear tube maintained it with an opaque one. However, apes starting with an opaque tube failed to solve the task. Additionally, facing the peanut floating on a water-filled tube (i.e., an end-state demonstration) promoted success independent on the availability of visual feedback. Moreover, experiencing how water was poured into the tube either by a human demonstrator or by a water tap that had been opened either by the ape or a human did not seem to be of further assistance. First, this study suggests that great apes require visual feedback for solving the FPT, which is no longer required after the initial acquisition. Second, some subjects benefit from encountering the end-state, a finding corroborating previous studies.

Extraction without tooling around — The first comprehensive description of the foraging- and socio-ecology of wild Goffin’s cockatoos (Cacatua goffiniana)

When tested under laboratory conditions, Goffin’s cockatoos (Cacatua goffiniana) demonstrate numerous sophisticated cognitive skills. Most importantly, this species has shown the ability to manufacture and use tools. However, little is known about the ecology of these cockatoos, endemic to the Tanimbar Islands in Indonesia. Here we provide first insights into the feeding- and socio-ecology of the wild Goffin’s cockatoos and propose potential links between their behaviour in natural settings and their advanced problem-solving capacities shown in captivity. Observational data suggests that Goffin’s cockatoos rely on a large variety of partially seasonal resources. Furthermore, several food types require different extraction techniques. These ecological and behavioural characteristics fall in line with current hypotheses regarding the evolution of complex cognition and innovativeness. We discuss how the efficiency of (extractive) foraging may preclude tool use in wild Goffin’s cockatoos, even though the corresponding cognitive and ecological prerequisites seem to be present.

Spontaneous innovation of hook-bending and unbending in orangutans (Pongo abelii)

Betty the crow astonished the scientific world as she spontaneously crafted hook-tools from straight wire in order to lift a basket out of vertical tubes. Recently it was suggested that this species’ solution was strongly influenced by predispositions from behavioural routines from habitual hook-tool manufacture. Nevertheless, the task became a paradigm to investigate tool innovation. Considering that young humans had surprising difficulties with the task, it was yet unclear whether the innovation of a hooked tool would be feasible to primates that lacked habitual hook making. We thus tested five captive orangutans in a hook bending and unbending task. Orangutans are habitually tool-using primates that have been reported to use but not craft hooked tools for locomotion in the wild. Two orangutans spontaneously innovated hook tools and four unbent the wire from their first trial on. Pre-experience with ready-made hooks had some effect but did not lead to continuous success. Further subjects improved the hook-design feature when the task required the subjects to bent the hook at a steeper angle. Our results indicate that the ability to represent and manufacture tools according to a current need does not require stereotyped behavioural routines, but can indeed arise innovatively. Furthermore, the present study shows that the capacity for hook tool innovation is not limited to large brained birds within non-human animals.

Tool making cockatoos adjust the lengths but not the widths of their tools to function

The ability to innovatively use or even manufacture different tools depending on a current situation can be silhouetted against examples of stereotyped, inborn tool use/manufacture and is thus often associated to advanced cognitive processing. In this study we confronted non-specialized, yet innovative tool making birds, Goffin's cockatoos (Cacatua goffiniana), with an apparatus featuring an out-of-reach food reward that could be placed at different distances from a tool opening. Alternatively, the food stayed at a constant distance but the tool opening in the front of the apparatus had different diameters. We used a novel material for tool manufacture (cardboard) that demanded an incrementally increased manufacturing effort from the actor, depending on the length of the tool required. We found that our subjects used two strategies to succeed in this tasks: either by making carboard-stripe tools using the full length of the material sheets originally offered or by adjusting the lengths of their tools to different goal distances. Subjects also discarded cardboard stripes that were too short to reach the goal prior to use and discarded longer pieces when the goal was further away than when it was close. Nevertheless, likely due to morphological constraints, the birds failed to adjust the widths of their tools depending on the diameter of the tool opening.

Habitual tool use innovated by free-living New Zealand kea

The emergence of flexible tool use is rare in the animal kingdom and thought to be largely constrained by either cognitive ability or ecological factors. That mostly birds with a high level of intelligence innovate tool use in captivity is consistent with the former hypothesis. We report here the first documented case of habitual tool use innovated in the wild by a bird species only known to have used tools in captivity. Trap-boxes containing food-bait and snap-trap(s) were installed in the remote Murchison Mountains, New Zealand, from 2002 to catch introduced stoats. Kea tampered with the trap-boxes in various ways. Over approximately 2.5 years, sticks were found inserted into at least 227 different trap-boxes that were up to 27 km apart. Video footage confirmed that the stick insertion was kea tool use. Trap-boxes are unlikely to have provided the only possibility for kea tool use in their habitat given their extractive foraging and skilled object manipulation. We argue that they instead greatly facilitated the opportunity for tool use, thus increasing the chance that kea would invent the behaviour. The innovation of tool use by kea in response to facilitation provides rare field support for the cognitive constraints hypothesis.