Balancing experimental control with ecological validity remains a central challenge for studying brain function. Here, we developed the Tower Foraging Park (TFP), a self-paced behavioral paradigm that emulates patch foraging and in which mice collect rewards by performing directional quarter-turns around square towers (exploit) and switching between them as they become depleted (explore). Mice rapidly learned the rewarded turning direction, increased movement speed, reduced trajectory variability, and typically abandoned towers before depletion. Reversal of the rewarded direction triggered rapid adaptation, accompanied by a dissociation between movement variability and speed. Repeated reversals progressively improved flexibility. Increasing the difficulty of locating rewarding towers prolonged exploitation, revealing adaptive regulation of patch-leaving decisions. Finally, mice trained under flexible contingencies ultimately outperformed those trained under stable contingencies in a challenging context. Altogether, the TFP reveals mechanisms underlying flexible foraging and provides a versatile, ecologically grounded platform for investigating the neural bases of adaptive behavior.

With practice, animals perform reward-oriented actions faster and with less variability. The dorsal striatum (DS) plays a key role in this process, potentially through opposing changes in cortical input strength to the two main striatal projection neurons (D1- and D2-SPNs). To test this hypothesis, we trained mice in a foraging task requiring them to perform quarter-turns (QTs) in a single direction (counterclockwise, CCW) along the walls of square towers to collect drops of water. As training progressed, the number and speed of CCW QTs increased while the variability of their trajectory decreased. Whisker trimming in well-trained mice altered QTs kinematics, highlighting the role of tactile inputs in guiding these actions. Combining ex vivo patch-clamp recordings in the DS with glutamate uncaging in the barrel cortex, we mapped the cortical neurons monosynaptically connected to D1- or D2-SPNs and measured the strength of these connections in the contralateral hemispheres, relative to the turn direction. In well-trained mice, compared to naive controls, the contralateral hemisphere showed no major changes in D1- or D2-SPNs connectivity, despite increased excitation of cortical pyramidal neurons. In contrast, the ipsiversive hemisphere exhibited no change in cortical excitability but showed increased cortical connectivity and input strength to SPNs. These findings reveal an unexpected hemispheric asymmetry in corticostriatal connectivity during lateralized foraging, which may reflect a homeostatic process normalizing striatal activity across hemispheres despite unbalanced cortical input. ### Competing Interest Statement The authors have declared no competing interest. Agence Nationale de la Recherche, https://ror.org/00rbzpz17, ANR-20-CE16-0002

The density and overlap of cortical axons in the dorsolateral striatum (DLS) have suggested that striatal neurons integrate widespread information from cortical regions that are functionally related. However, in vivo, DLS neuronal responses to sensory stimuli have shown unexpectedly high selectivity, raising questions about the actual degree of convergence of functional corticostriatal projections on individual striatal cells. Here, we investigated this question by focusing on the projections from different whisker cortical columns in mice, as they overlap in the striatum and are co-active during behavior. Using ex vivo patch-clamp recordings in the DLS and glutamate uncaging for focal stimulations in the barrel cortex, we were able to map the location of presynaptic neurons to individual striatal projection neurons (SPNs). We found that each SPN was innervated by cells located in a small number of whisker cortical columns scattered across the barrel field in the slice. Connectivity of single SPNs with cortical neurons was thus highly discontinuous horizontally, despite the presence of more potential connections. Moreover, connectivity patterns were specific to each cell, with neighboring SPNs sharing few common clusters of presynaptic cells in the cortex. Despite this sparse and distinct innervation of individual SPNs, the projection was topographically organized at the population level. Finally, we found similar innervation patterns for D1- and D2-type SPNs, but observed differences in synaptic strength in their connections with certain cortical layers, notably the associative layer 2/3. Our results suggest that the high convergence of somatosensory inputs to the striatum, enabled by diffuse and overlapping cortical innervation, is accomplished through sparse yet complementary connectivity to individual SPNs.

Adaptive behavior depends on a variety of brain functions, such as learning, decision-making, spatial navigation and motor control, which have been studied using two main strategies. Trial-based tasks allow their mechanistic dissection but tend to generate highly stereotypical behavior, whereas open-field investigations capture naturalistic dynamics with less experimental control. To leverage the strengths of both approaches, we developed a behavioral framework which recreates dilemmas faced by animals during patch foraging. In the Tower Foraging Park (TFP), mice harvest rewards along square towers (patches) by making quarter-turns around them in a single direction (exploit) and alternating between towers (explore) as patches eventually deplete. Within a couple of sessions, naïve mice performed quarter-turns in the rewarded direction with increasing vigor and reduced variability, and switched towers after short exploitation bouts. When the harvest direction was reversed, mice rapidly adapted their turning direction, with quarter-turn trajectory variability and speed becoming decoupled. Mice subjected to daily reversals adapted progressively faster, revealing meta-learning. When the next rewarding tower became harder to locate, all trained mice increased exploitation duration, although metalearners outperformed animals trained under stable contingencies. Altogether, the TFP produced behavior consistent with foraging theory and revealed new processes facilitating flexible foraging: meta-learning and the decoupling of movement variability and speed. Moreover, because the TFP accommodates diverse protocol variants, adheres to FAIR principles, and is fully compatible with modern neurophysiological techniques, it provides a promising platform for mechanistic investigations of brain functions underlying adaptive behavior while maintaining ethological validity. ### Competing Interest Statement The authors have declared no competing interest. Agence Nationale de la Recherche, ANR-20-CE16-0002, ANR-17-EURE-0029, ANR-16-CONV-0001 Aix-Marseille Université, https://ror.org/035xkbk20, AMX-22-RE-AB-007, AMX-19-IET-004 Fondation pour la Recherche Médicale, https://ror.org/04w6kn183, FDT202204014828

During reward-oriented behaviors, animals –including humans– spontaneously adjust the speeds of their decisions and movements based on dynamically changing costs and benefits. The mechanisms constraining these adaptive modulations remain unclear, especially in freely moving animals. Here, we developed a naturalistic foraging task in which rats decided when and how fast to run across a motorized treadmill to collect rewards. Model-based analyses explained why decision and movement speeds were coupled or decoupled as rats adapted to changes in reward value or motor cost, respectively. Moreover, lesions of the dorsal striatum increased the animals’ sensitivity to motor cost, limiting their running speed in the most effortful conditions while sparing reward-related behavioral modulations. Altogether, our study describes how neuroeconomic constraints influence decision and movement speeds in foraging rats, and paves the way for a refined understanding of the role of the basal ganglia in motor control and decision-making. ### Competing Interest Statement The authors have declared no competing interest.

Monteiro and colleagues used temperature manipulation to bidirectionally alter the speed of neuronal dynamics in the dorsal striatum of anesthetized rats. This manipulation selectively slowed down or sped up time perception, providing insights into the mechanisms of time-based decisions.

It is well-accepted in neuroscience that animals process time internally to estimate the duration of intervals lasting between one and several seconds. More than 100 years ago, Henri Bergson nevertheless remarked that, because animals have memory, their inner experience of time is ever-changing, making duration impossible to measure internally and time a source of change. Bergson proposed that quantifying the inner experience of time requires its externalization in movements (observed or self-generated), as their unfolding leaves measurable traces in space. Here, studies across species are reviewed and collectively suggest that, in line with Bergson's ideas, animals spontaneously solve time estimation tasks through a movement-based spatialization of time. Moreover, the well-known scalable anticipatory responses of animals to regularly spaced rewards can be explained by the variable pressure of time on reward-oriented actions. Finally, the brain regions linked with time perception overlap with those implicated in motor control, spatial navigation and motivation. Thus, instead of considering time as static information processed by the brain, it might be fruitful to conceptualize it as a kind of force to which animals are more or less sensitive depending on their internal state and environment.

It is commonly accepted in neuroscience that animals process time internally to estimate the duration of intervals lasting between one and several seconds, an ability called prospective time perception. More than 100 years ago, Henri Bergson nevertheless remarked that, because animals have memory, their inner experience of time is ever-changing (two lived moments cannot repeat). This should make duration impossible to measure internally and time a source of change. Bergson therefore proposed that quantifying the inner experience of elapsing time requires its externalization in observed or self-generated movements, as their unfolding leaves discretizable, hence measurable, traces in space. Here, studies across species are reviewed and collectively suggest that, in line with Bergson's counterintuitive ideas, animals spontaneously solve time estimation tasks through a movement-based spatialization of time and struggle to do that while being immobile, or independently of their movements. Moreover, the well-known scalable anticipatory responses of animals to regularly spaced rewards can be explained by the variable pressure of time on reward-oriented actions, rather than explicit reliance on internal representation of time. Finally, the brain regions linked with prospective time perception overlap with those implicated in motor control, spatial navigation and motivation. Thus, instead of considering time as static information processed in the brain (i.e., like space), it might be fruitful to conceptualize it as a kind of force to which animals are more or less sensitive depending on their internal state and environment.

The dorsal striatum (dS) has been implicated in storing and retrieving procedural memories and controlling movement kinematics (e.g., speed). Since procedural memories are expressed through movements, the exact nature of the dS function has proven difficult to delineate. Here we challenged rats in complementary tasks designed to alleviate this performance confound. Surprisingly, dS lesions spared task-specific procedural memories but altered the kinematics of their expression in motor routines. Further behavioral analyses combined with simulations in the optimal control framework indicated that these alterations reflected an increased sensitivity to effort with preserved reward-seeking and ability to modulate movement speed. By setting the sensitivity to effort, the dS contributes to the optimization of the energy invested into voluntary movements. Such an elementary function of the dS might explain its implication in both procedural decisions and the control of movement speed.

The dorsal striatum (dS) has been implicated in storing procedural memories and controlling movement kinematics. Since procedural memories are expressed through movements, the exact nature of the dS function has proven difficult to delineate. Here, we challenged rats in complementary locomotion-based tasks designed to alleviate this confound. Surprisingly, dS lesions did not impair the rats' ability to remember the procedure for the successful completion of motor routines. However, the speed and initiation of the reward-oriented phase of the routines were irreversibly altered by the dS lesion. Further behavioral analyses, combined with modeling in the optimal control framework, indicated that these kinematic alterations were well explained by an increased sensitivity to effort. Our work provides evidence supporting a primary role of the dS in modulating the kinematics of reward-oriented actions, a function that may be related to the optimization of the energetic costs of moving.

How animals adapt their behavior according to regular time intervals between events is not well understood, especially when intervals last several seconds. One possibility is that animals use disembodied internal neuronal representations of time to decide when to initiate a given action at the end of an interval. However, animals rarely remain immobile during time intervals but tend to perform stereotyped behaviors, raising the possibility that motor routines improve timing accuracy. To test this possibility, we used a task in which rats, freely moving on a motorized treadmill, could obtain a reward if they approached it after a fixed interval. Most animals took advantage of the treadmill length and its moving direction to develop, by trial-and-error, the same motor routine whose execution resulted in the precise timing of their reward approaches. Noticeably, when proficient animals did not follow this routine, their temporal accuracy decreased. Then, naïve animals were trained in modified versions of the task designed to prevent the development of this routine. Compared to rats trained in the first protocol, these animals didn’t reach a comparable level of timing accuracy. Altogether, our results indicate that timing accuracy in rats is improved when the environment affords cues that animals can incorporate into motor routines.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.