Object play and problem solving in infancy: Insights into tool use.

Tool use is primarily, but not exclusively, present in species with otherwise advanced cognitive traits. However, the interaction between such traits and conspecific inter-individual variation in the presence, complexity, or intensity of tool use is far from being established. We addressed this matter among human infants, seeking factors that relate to differences in tool use. We examined, both correlationally and experimentally, whether the propensity to engage in object combinations predicts performance in means-end problem-solving tasks involving or not involving the use of a tool. We tested 71 infants aged 15, 18, 21, and 24 months, dividing them into two subgroups: one exposed to an adult demonstrating object-object combinations (i.e., "prompting" infants to combine objects together) and another with comparable social exposure but where the adult demonstrated single-object manipulations. We found a correlation between the combined level of spontaneous and prompted object combinations and problem-solving performance regardless of the involvement of tools in the problem. However, we did not find differences in tool-use performance between the two demonstration subgroups. The correlational analysis suggests that complexity of play, as measured by the frequency of combining objects, is linked to infants' problem-solving skills rather than being specifically associated with tool use, as previously suggested in the literature.

Serial reversal learning in nectar-feeding bats.

We explored the behavioral flexibility of Commissaris's long-tongued bats through a spatial serial reversal foraging task. Bats kept in captivity for short periods were trained to obtain nectar rewards from two artificial flowers. At any given time, only one of the flowers provided rewards and these reward contingencies reversed in successive blocks of 50 flower visits. All bats detected and responded to reversals by making most of their visits to the currently active flower. As the bats experienced repeated reversals, their preference re-adjusted faster. Although the flower state reversals were theoretically predictable, we did not detect anticipatory behavior, that is, frequency of visits to the alternative flower did not increase within each block as the programmed reversal approached. The net balance of these changes was a progressive improvement in performance in terms of the total proportion of visits allocated to the active flower. The results are compatible with, but do not depend on, the bats displaying an ability to 'learn to learn' and show that the dynamics of allocation of effort between food sources can change flexibly according to circumstances.

GoFish: A low-cost, open-source platform for closed-loop behavioural experiments on fish.

Fish are the most species-rich vertebrate group, displaying vast ecological, anatomical and behavioural diversity, and therefore are of major interest for the study of behaviour and its evolution. However, with respect to other vertebrates, fish are relatively underrepresented in psychological and cognitive research. A greater availability of easily accessible, flexible, open-source experimental platforms that facilitate the automation of task control and data acquisition may help to reduce this bias and improve the scalability and refinement of behavioural experiments in a range of different fish species. Here we present GoFish, a fully automated platform for behavioural experiments in aquatic species. GoFish includes real-time video tracking of subjects, presentation of stimuli in a computer screen, an automatic feeder device, and closed-loop control of task contingencies and data acquisition. The design and software components of the platform are freely available, while the hardware is open-source and relatively inexpensive. The control software, Bonsai, is designed to facilitate rapid development of task workflows and is supported by a growing community of users. As an illustration and test of its use, we present the results of two experiments on discrimination learning, reversal, and choice in goldfish (Carassius auratus). GoFish facilitates the automation of high-throughput protocols and the acquisition of rich behavioural data. Our platform has the potential to become a widely used tool that facilitates complex behavioural experiments in aquatic species.

Testing cognitive models of decision-making: selected studies with starlings.

The behavioural sciences are home to controversies that have survived for centuries, notably about the relation between observable behaviour and theoretical constructs addressing out-of-sight processes in the agents' brains. There is no shared definition for cognition, but the very existence of a thriving journal called Animal Cognition proves that such controversies are still live and help to (a) promote research on the complexity of processes leading to action, and (b) nudge scholars to restrict their cognitive models to those that can be falsified experimentally. Here, we illustrate some of these issues in a limited arena, focusing on the construction and expression of subjective value and choice. Using mainly work from our own laboratory, we show that valuation of alternatives is sensitive to options' properties, to subject's state, and to background alternatives. These factors exert their influence at the time the subject learns about individual options, rather than at choice time. We also show that valuation can be experimentally dissociated from the cognitive representation of options' metrics and argue that experimental animals process options independently at the time of choice, without elaborated comparisons along different dimensions. The findings we report are not consistent with the hypothesis that preference is constructed at the time of choice, a prevalent view in human decision-making research. We argue that animal cognition, viewed as a research program at the crossroads of different behavioural sciences rather than as a debate about properties of mental life, is inspiring and solid, and a progressive and progressing paradigm.

Paradoxical choice and the reinforcing value of information.

Signals that reduce uncertainty can be valuable because well-informed decision-makers can better align their preferences to opportunities. However, some birds and mammals display an appetite for informative signals that cannot be used to increase returns. We explore the role that reward-predictive stimuli have in fostering such preferences, aiming at distinguishing between two putative underlying mechanisms. The 'information hypothesis' proposes that reducing uncertainty is reinforcing per se, somewhat consistently with the concept of curiosity: a motivation to know in the absence of tractable extrinsic benefits. In contrast, the 'conditioned reinforcement hypothesis', an associative account, proposes asymmetries in secondarily acquired reinforcement: post-choice stimuli announcing forthcoming rewards (S+) reinforce responses more than stimuli signalling no rewards (S-) inhibit responses. In three treatments, rats faced two equally profitable options delivering food probabilistically after a fixed delay. In the informative option (Info), food or no food was signalled immediately after choice, whereas in the non-informative option (NoInfo) outcomes were uncertain until the delay lapsed. Subjects preferred Info when (1) both outcomes were explicitly signalled by salient auditory cues, (2) only forthcoming food delivery was explicitly signalled, and (3) only the absence of forthcoming reward was explicitly signalled. Acquisition was slower in (3), when food was not explicitly signalled, showing that signals for positive outcomes have a greater influence on the development of preference than signals for negative ones. Our results are consistent with an elaborated conditioned reinforcement account, and with the conjecture that both uncertainty reduction and conditioned reinforcement jointly act to generate preference.

Imprinting on time-structured acoustic stimuli in ducklings.

Filial imprinting is a dedicated learning process that lacks explicit reinforcement. The phenomenon itself is narrowly heritably canalized, but its content, the representation of the parental object, reflects the circumstances of the newborn. Imprinting has recently been shown to be even more subtle and complex than previously envisaged, since ducklings and chicks are now known to select and represent for later generalization abstract conceptual properties of the objects they perceive as neonates, including movement pattern, heterogeneity and inter-component relationships of same or different. Here, we investigate day-old Mallard (Anas platyrhynchos) ducklings' bias towards imprinting on acoustic stimuli made from mallards' vocalizations as opposed to white noise, whether they imprint on the temporal structure of brief acoustic stimuli of either kind, and whether they generalize timing information across the two sounds. Our data are consistent with a strong innate preference for natural sounds, but do not reliably establish sensitivity to temporal relations. This fits with the view that imprinting includes the establishment of representations of both primary percepts and selective abstract properties of their early perceptual input, meshing together genetically transmitted prior pre-dispositions with active selection and processing of the perceptual input.

Sex differences in learning flexibility in an avian brood parasite, the shiny cowbird.

Females of brood parasitic shiny cowbirds, Molothrus bonariensis, search and prospect host nests, synchronizing parasitism with host laying. This behavior is sex-specific, as females perform this task without male's assistance. Host nests must be removed from the female's memory "library" after being parasitized, to avoid repeated parasitism, or when they become unavailable because of predation. Thus, females must adjust their stored information about host nest status more dynamically than males, possibly leading to differences in learning flexibility. We tested for sex differences in a visual (local cues) and a spatial discrimination reversal learning task, expecting females to outperform males as an expression of greater behavioral flexibility. Both sexes learned faster the spatial than the visual task during both acquisition and reversal. In the visual task there were no sex differences in acquisition, but females reversed faster than males. In the spatial task there were no sex differences during either acquisition or reversal, possibly because of a ceiling effect: both sexes learned too fast for differences in performance to be detectable. Faster female reversal in a visual but not spatial task indicates that the greater behavioral flexibility in females may only be detectable above some level of task difficulty.

Sex differences in the use of spatial cues in two avian brood parasites.

Shiny and screaming cowbirds are avian interspecific brood parasites that locate and prospect host nests in daylight and return from one to several days later to lay an egg during the pre-dawn twilight. Thus, during nest location and prospecting, both location information and visual features are available, but the latter become less salient in the low-light conditions when the nests are visited for laying. This raises the question of how these different sources of information interact, and whether this reflects different behavioural specializations across sexes. Differences are expected, because in shiny cowbirds, females act alone, but in screaming cowbirds, both sexes make exploratory and laying nest visits together. We trained females and males of shiny and screaming cowbird to locate a food source signalled by both colour and position (cues associated), and evaluated performance after displacing the colour cue to make it misleading (cues dissociated). There were no sex or species differences in acquisition performance while the cues were associated. When the colour cue was relocated, individuals of both sexes and species located the food source making fewer visits to non-baited wells than expected by chance, indicating that they all retained the position as an informative cue. In this phase, however, shiny cowbird females, but not screaming, outperformed conspecific males, visiting fewer non-baited wells before finding the food location and making straighter paths in the search. These results are consistent with a greater reliance on spatial memory, as expected from the shiny cowbird female's specialization on nest location behaviour.

Choosing fast and simply: Construction of preferences by starlings through parallel option valuation.

The integration of normative and descriptive analyses of decision processes in humans struggles with the fact that measuring preferences by different procedures yields different rankings and that humans appear irrationally impulsive (namely, show maladaptive preference for immediacy). Failure of procedure invariance has led to the widespread hypothesis that preferences are constructed "on the spot" by cognitive evaluations performed at choice time, implying that choices should take extra time in order to perform the necessary comparisons. We examine this issue in experiments with starlings (Sturnus vulgaris) and show that integrating normative and descriptive arguments is possible and may help reinterpreting human decision results. Our main findings are that (1) ranking alternatives through direct rating (response time) accurately predicts preference in choice, overcoming failures of procedure invariance; (2) preference is not constructed at choice time nor does it involve extra time (we show that the opposite is true); and (3) starlings' choices are not irrationally impulsive but are instead directly interpretable in terms of profitability ranking. Like all nonhuman research, our protocols examine decisions by experience rather than by description, and hence support the conjecture that irrationalities that prevail in research with humans may not be observed in decisions by experience protocols.

Ducklings imprint on chromatic heterogeneity.

Avian filial imprinting is a rapid form of learning occurring just after hatching in precocial bird species. The acquired imprint on either or both parents goes on to affect the young bird's survival and social behaviour later in life (Bateson in Biol Rev 41:177-217, 1966). The imprinting mechanism is specialized but flexible, and causes the hatchling to develop high-fidelity recognition and attraction to any moving stimulus of suitable size seen during a predefined sensitive period. It has been observed (Martinho and Kacelnik in Science 353:286-288, 2016; Versace et al. in Anim Cogn 20:521-529, 2017) that in addition to visual and acoustic sensory inputs, imprinting may incorporate informational rules or abstract concepts. Here we report a study of mallard ducklings (Anas platyrhynchos domesticus) undergoing imprinting on the chromatic heterogeneity of stimuli, with a focus on how this may be transferred to novel objects. Ducklings were exposed to a series of chromatically heterogeneous or homogeneous stimuli and tested for preference between two novel stimuli, one heterogeneous and the other homogeneous. Exposure to heterogeneity significantly enhanced preference for novel heterogeneous stimuli, relative to ducklings exposed to homogeneous stimuli or unexposed controls. These findings support the view that imprinting does not rely solely on exemplars, or snapshot-like representations of visual input, but that instead young precocial animals form complex multidimensional representations of the target object, involving abstract properties, either at the time of learning, or later, through generalization from the learnt exemplars.

Maintaining performance in searching dogs: Evidence from a rat model that training to detect a second (irrelevant) stimulus can maintain search and detection responding.

Scent-detecting dogs perform a sequence, or chain, of behaviors that, at minimum, includes searching followed by a detection behavior that signals the presence of a target stimulus to the handler. However, when working, dogs often engage in prolonged periods of searching without encountering a target. It is therefore important for trainers to use methods that promote persistent search behavior and target detection accuracy. Laboratory models can provide insights to the important variables that influence search persistence and accuracy. The present experiments examined a rat model of detection dog behavior. Two experiments assessed the use of practice with a single target stimulus to maintain search and detection of another previously-trained target. In Experiment 1, after learning a search→detection chain with two auditory targets, rats received either brief or extended training with only one of the targets before being tested for detection of both targets in extinction. The results suggest that single-target training strengthened the ability of the other target to control the detection behavior. Experiment 2 found that even infrequent target encounters were still effective at maintaining detection behavior to the other trained target. Importantly, the treatment was effective when the target stimuli were from different sensory modalities. Overall, the results support the utility of the rat model of search-dog behavior for evaluating novel training methods. We suggest several useful procedures for enhancing search persistence and accuracy in detection dogs that can be implemented in training protocols.

Priors in Animal and Artificial Intelligence: Where Does Learning Begin?

A major goal for the next generation of artificial intelligence (AI) is to build machines that are able to reason and cope with novel tasks, environments, and situations in a manner that approaches the abilities of animals. Evidence from precocial species suggests that driving learning through suitable priors can help to successfully face this challenge.

Parallel vs. comparative evaluation of alternative options by colonies and individuals of the ant Temnothorax rugatulus.

Both a single ant and the colony to which it belongs can make decisions, but the underlying mechanisms may differ. Colonies are known to be less susceptible than lone ants to "choice overload", whereby decision quality deteriorates with increasing number of options. We probed the basis of this difference, using the model system of nest-site selection by the ant Temnothorax rugatulus. We tested the applicability of two competing models originally developed to explain information-processing mechanisms in vertebrates. The Tug of War model states that concurrent alternatives are directly compared, so that choosing between two alternatives takes longer than accepting a single one. In contrast, the Sequential Choice Model assumes that options are examined in parallel, and action takes place once any option reaches a decision criterion, so that adding more options shortens time to act. We found that single ants matched the Tug of War model while colonies fitted the Sequential Choice model. Our study shows that algorithmic models for decision-making can serve to investigate vastly different domains, from vertebrate individuals to both individuals and colonies of social insects.

Midsession reversal task with pigeons: Parallel processing of alternatives explains choices.

Most models of choice assume a "tug of war" (ToW) between options present at the time of the choice, arguing that preferences are built on this process, and implying that adding options increases delay to act. In contrast, the sequential choice model (SCM) proposes that choices are driven by parallel expression of the mechanisms that control action in sequential encounters, without comparative deliberation at choice time. Only the SCM predicts choice preferences based on latencies to respond in single-option encounters. SCM further predicts that latencies to choose should either be the same or shorter than those in sequential encounters. We contrasted these models using a midsession reversal task with pigeons. Responses to one alternative (S1) were rewarded in the first half of each session and those to the other (S2) in the second half. Single-option (sequential) and two-option (choice) trials were intermingled. In choice trials subjects strongly preferred S1 early in the session, showed intermediate preferences toward the midsession, and preferred S2 late. These preferences were all predicted by changes in latency toward the presently negative alternative (S2 early and S1 late) in single-option trials. Latency toward presently positive stimuli were minimal throughout, in both single and two option trials, with no evidence of an evaluation time cost of choice. The ability to predict choice preference from latencies in sequential encounters and the absence of a choice delay support the SCM against ToW models, consistently with results from other protocols and species. (PsycINFO Database Record

Paradoxical choice in rats: Subjective valuation and mechanism of choice.

Decision-makers benefit from information only when they can use it to guide behavior. However, recent experiments found that pigeons and starlings value information that they cannot use. Here we show that this paradox is also present in rats, and explore the underlying decision process. Subjects chose between two options that delivered food probabilistically after a fixed delay. In one option ("info"), outcomes (food/no-food) were signaled immediately after choice, whereas in the alternative ("non-info") the outcome was uncertain until the delay lapsed. Rats sacrificed up to 20% potential rewards by preferring the info option, but reversed preference when the cost was 60%. This reversal contrasts with the results found with pigeons and starlings and may reflect species' differences worth of further investigation. Results are consistent with predictions of the Sequential Choice Model (SCM), that proposes that choices are driven by the mechanisms that control action in sequential encounters. As expected from the SCM, latencies to respond in single-option trials predicted preferences in choice trials, and latencies in choice trials were the same or shorter than in single-option trials. We argue that the congruence of results in distant vertebrates probably reflects evolved adaptations to shared fundamental challenges in nature, and that the apparently paradoxical overvaluing of information is not sub-optimal as has been claimed, even though its functional significance is not yet understood.

Adaptations to different habitats in sexual and asexual populations of parasitoid wasps: a meta-analysis.

BACKGROUND: Coexistence of sexual and asexual populations remains a key question in evolutionary ecology. We address the question how an asexual and a sexual form of the parasitoid Venturia canescens can coexist in southern Europe. We test the hypothesis that both forms are adapted to different habitats within their area of distribution. Sexuals inhabit natural environments that are highly unpredictable, and where density of wasps and their hosts is low and patchily distributed. Asexuals instead are common in anthropic environments (e.g., grain stores) where host outbreaks offer periods when egg-load is the main constraint on reproductive output. METHODS: We present a meta-analysis of known adaptations to these habitats. Differences in behavior, physiology and life-history traits between sexual and asexual wasps were standardized in term of effect size (Cohen's d value; Cohen, 1988). RESULTS: Seeking consilience from the differences between multiple traits, we found that sexuals invest more in longevity at the expense of egg-load, are more mobile, and display higher plasticity in response to thermal variability than asexual counterparts. DISCUSSION: Thus, each form has consistent multiple adaptations to the ecological circumstances in the contrasting environments.

Response to Comments on "Ducklings imprint on the relational concept of 'same or different'".

Two Comments by Hupé and by Langbein and Puppe address our choice of statistical analysis in assigning preference between sets of stimuli to individual ducklings in our paper. We believe that our analysis remains the most appropriate approach for our data and experimental design.

Goffin's cockatoos make the same tool type from different materials.

Innovative tool manufacture is rare and hard to isolate in animals. We show that an Indonesian generalist parrot, the Goffin's cockatoo, can flexibly and spontaneously transfer the manufacture of stick-type tools across three different materials. Each material required different manipulation patterns, including substrates that required active sculpting for achieving a functional, elongated shape.

Ducklings imprint on the relational concept of "same or different".

The ability to identify and retain logical relations between stimuli and apply them to novel stimuli is known as relational concept learning. This has been demonstrated in a few animal species after extensive reinforcement training, and it reveals the brain's ability to deal with abstract properties. Here we describe relational concept learning in newborn ducklings without reinforced training. Newly hatched domesticated mallards that were briefly exposed to a pair of objects that were either the same or different in shape or color later preferred to follow pairs of new objects exhibiting the imprinted relation. Thus, even in a seemingly rigid and very rapid form of learning such as filial imprinting, the brain operates with abstract conceptual reasoning, a faculty often assumed to be reserved to highly intelligent organisms.

Pea Plants Show Risk Sensitivity.

Sensitivity to variability in resources has been documented in humans, primates, birds, and social insects, but the fit between empirical results and the predictions of risk sensitivity theory (RST), which aims to explain this sensitivity in adaptive terms, is weak [1]. RST predicts that agents should switch between risk proneness and risk aversion depending on state and circumstances, especially according to the richness of the least variable option [2]. Unrealistic assumptions about agents' information processing mechanisms and poor knowledge of the extent to which variability imposes specific selection in nature are strong candidates to explain the gap between theory and data. RST's rationale also applies to plants, where it has not hitherto been tested. Given the differences between animals' and plants' information processing mechanisms, such tests should help unravel the conflicts between theory and data. Measuring root growth allocation by split-root pea plants, we show that they favor variability when mean nutrient levels are low and the opposite when they are high, supporting the most widespread RST prediction. However, the combination of non-linear effects of nitrogen availability at local and systemic levels may explain some of these effects as a consequence of mechanisms not necessarily evolved to cope with variance [3, 4]. This resembles animal examples in which properties of perception and learning cause risk sensitivity even though they are not risk adaptations [5].