The Journal of Experimental Biology

The ability of animals to adjust their heading, i.e. to turn, is essential for all walking animals. While several studies have addressed how leg movement or inter-leg coordination may change during turning, relatively little is known about how turning-related changes scale with turn magnitude. Here, we used spontaneous and visually induced turns of unrestrained walking stick insects to test (i) how high-level parameters of unrestrained turning scale with low-level parameters of leg movement, and (ii) the effect of visual guidance on turning parameters. To this end, we used a step change in stationary landmark position in an open-field arena to constrain timing and magnitude of target-directed turns. These visually guided turns were compared with spontaneous turns in an all-white condition. We show that visually induced turns were walked at a larger forward velocity and had fewer short steps than spontaneous turns. The scaling of turning responses was dominated by an increase in turning duration (factor 1.87) rather than turning speed (factor 1.32). Increased rotational velocity correlated with reduced forward velocity, though with flexible timing of both effects. These changes were accompanied by larger shifts in step direction, as well as an increased asymmetry of step types between inner and outer legs, suggesting a mix of distinct turning strategies, that depend on overall turn angle. Future models on six-legged locomotion should thus consider the incorporation of more than one mechanism to govern turning.

Although muscle spindle sensory signals have been extensively studied, little is known about how and why muscle spindle firing is modulated by the central nervous system during movement. Specialized motor neurons to the muscle spindle, i.e. gamma motor neurons, can profoundly alter spindle firing during behavior, but technological limitations hinder our ability to record gamma motor and muscle spindle sensory signals during most behaviors. We used a biophysical model of a muscle spindle within a muscle-tendon unit to simulate how gamma drive may modulate muscle spindle Ia firing during locomotion. Based on a few available recordings from decerebrate animals, we demonstrate that our model, tuned to passive stretch conditions, can reproduce profound changes in muscle spindle firing in response to identical joint motions in locomotor vs. relaxed stretch conditions. Our model can discover phasic patterns of two types of gamma motor neuron drive based on recorded muscle spindle Ia firing and joint motion. By simulating perturbations, we conclude that: 1) sinusoidal activation of static gamma motor neurons during locomotion, encoding intended movement, modulates muscle spindle signals such that they act as sensorimotor feedback signals based on errors from the intended muscle fascicle length; 2) phasic on/off activation of dynamic gamma motor neurons during locomotion acts as an event detector, heightening muscle spindle Ia responses to discrete perturbations. As such, their muscle-within-muscle structure allows the muscle spindle to act as a highly tunable physical internal model of muscle state to guide movement. Our model supports proposed but as-yet-untested theories of muscle spindle function and offers a framework for extending the testing of muscle spindle function to active, behavioral conditions.

Animals fight for resources to obtain fitness benefits; most contests are intrasexual, and males tend to fight more than females. Although the genetic basis of male aggression is well studied, we know little about the genetic variation of female aggression. Female aggression varies with reproductive status and is potentially influenced not only by her genotype, but also by the genotype of her mate. Here we measured both male and female aggression in a set of Drosophila melanogaster inbred lines by competing each line against a standard competitor. Aggression varied among lines for both sexes, but male and female aggression were not correlated. Female aggression for many lines increased with mating, as expected, but not all lines changed aggression. However, when females were mated to males of different lines, male genotype did not affect the post-mating change in aggression, suggesting that ejaculate-mediated effects do not vary across these lines. The aggression level of the standard opponent was positively correlated with that of focal individuals indicating that individuals modulate their behavior according to the genotype of their opponent.

Anoxia is a major stress for most vertebrates and frequently accompanies harsh winter conditions, particularly in species that spend much of the season frozen solid. North American freeze-tolerant wood frogs (Rana sylvatica) can survive several months without oxygen and endure whole-body freezing for up to eight months of the year, with [~]70% of total body water frozen as extracellular ice, yet revive when temperatures rise in spring. Survival depends on multiple adaptations, including tolerance of prolonged oxygen deprivation while frozen, when breathing and circulation are halted. A key strategy involves hepatic glycogen mobilization, producing large amounts of glucose that are distributed to tissues where it functions both as a cryoprotectant and as a substrate for anaerobic ATP production. The present study examines the role of histone lysine methylation and demethylation in regulating liver proteins under anoxic conditions. Relative protein expression of seven histone methyltransferases (ASH2L-S, ASH2L-L, RBBP5, SETD8, SMYD2, ESET, SETD1), six lysine demethylases (KDM1A, KDM3B, KDM4A, KDM4B, KDM5A, KDM5C), and eight histone marks (H3K4me1, H3K4me2, H3K9me3, H3K27me3, H3K36me3, H3K79me3, H4K20me1, H4K20me3) were evaluated in wood frog liver under control, 4-hour, and 24-hour anoxia exposures. The data indicate that histone lysine methylation and demethylation contribute significantly to transcriptional regulation under anoxia. Specifically, H3K4, H3K36, and H3K79 methylation were associated with transcriptional activation, whereas H3K9, H3K27, and H4K20 methylation correlated with transcriptional repression. These findings highlight the dynamic role of epigenetic regulation in supporting hypometabolism and stress adaptation in freeze-tolerant wood frogs.

Crown-of-thorns sea star (CoTS), Acanthaster cf. solaris, outbreaks are a major cause of hard coral cover decline across the west Pacific, threatening coral reefs. Coral taxa vary in susceptibility to CoTS predation from preferred (Acropora spp.) to non-preferred (Porites spp.), yet the mechanisms underlying these differences are poorly understood. We investigated coral defenses during an ongoing CoTS outbreak in Mo'orea, French Polynesia by examining gene expression (including putative toxin genes) in healthy and actively predated colonies of a preferred (Acropora hyacinthus) and a non-preferred (Porites sp.) coral prey species. During predation, A. hyacinthus exhibited molecular signatures of cellular stress responses involving oxidative stress signalling, inflammation, and tissue proteolysis. In contrast, Porites sp. showed enrichment of genes involved in mitochondrial metabolic adjustment and aerobic metabolism, suggesting metabolic compensation to maintain cellular function. Furthermore, A. hyacinthus demonstrated a reactive defense behaviour by differentially expressing toxins (e.g., kunitz-type neurotoxins) while Porites sp. employed constitutive expression of all putative toxins regardless of active predation, suggesting a proactive defense strategy. Together, these findings suggest that preferred and non-preferred coral prey exhibit fundamentally different molecular and defensive strategies during CoTS predation, shedding light on the evolutionary arms race between corals and their predators.

In a wide variety of animals, developmental crowding results in adults with smaller bodies. The crowding effect on body size in Drosophila melanogaster is canonically attributed to heightened competition for nutrition. However, whether other consequences of crowding also contribute to its effect on size remains an open question. We tested the relative contributions of nutritional competition, oxygen availability, and larval-generated metabolites to the crowding effect on size. We found that while nutrition explains most of the variation in body size due to crowding, oxygen also contributes in a sex- and nutrition-dependent manner. We found no evidence that larval-generated chemicals affect body size. These data confirm a widely suspected but untested role of nutrition in producing the crowding effect on size in D. melanogaster, while revealing an unexpected role of oxygen, and raise the possibility that behavior may be a mediator of density-dependent plasticity. Research HighlightsWe found that both nutrition and oxygen mediate the crowding effect on size in Drosophila melanogaster.

Many insects enter diapause, a programmed state of developmental arrest that enables survival under adverse environmental conditions. In Drosophila melanogaster Meigen, 1830, diapause is characterized by reproductive arrest and reduced intestinal growth, accompanied by suppressed intestinal stem cell (ISC) activity. Juvenile Hormone (JH) promotes ISC proliferation under favorable conditions, but its capacity to modulate stem cell dynamics during cold-induced diapause remains unclear. Here, we investigated whether JH signaling can reactivate midgut remodeling in adult females maintained at 11. At this temperature, flies exhibited pronounced gut atrophy and elevated Phospho-histone H3 (PH3+) cell abundance, consistent with temperature-dependent G2/M phase arrest JH treatment significantly increased the proportion of Delta-positive progenitor cells in the anterior (R2) and posterior (R5) midgut regions at both 11 and 25, demonstrating that JH acts as a conserved mitogen for the ISC pool irrespective of thermal environment. A trend toward reduced PH3+ accumulation in the posterior midgut following JH treatment (p = 0.061) suggests possible facilitation of mitotic exit, though this effect did not reach statistical significance. Despite cellular-level changes, JH treatment did not restore overall gut size, indicating that the 72-84 hour exposure window was insufficient for subsequent tissue hypertrophy. Additionally, we identified a recurrent cold-induced pathology of gut distension, provisionally termed Lumen Obstruction Syndrome (LOS), which was independent of JH signaling. These findings reveal an uncoupling of JH-driven stem cell expansion from gross organ growth under diapause conditions, highlighting the selective sensitivity of the ISC compartment to endocrine signaling during environmental stress.

Sleep is a ubiquitous, yet highly variable, behavior across species. The duration and timing of sleep are influenced by ecological demands and dietary context. In the mosquito Aedes aegypti, a blood-feeding insect with specialized nutritional requirements, the relationship between feeding and sleep remains poorly understood. Here, we investigated how blood-derived dietary protein influences sleep regulation. Using postural analysis, videography, and arousal-threshold assays, we established that immobility bouts of greater than 10 minutes reliably define sleep in Ae. aegypti. Mosquitoes lacking the circadian clock gene cycle still maintained daily sleep rhythms but exhibited reduced sleep duration and heightened overall activity. Infrared activity monitoring revealed that blood-fed females showed a marked increase in sleep beginning immediately after feeding and persisting for several days, accompanied by reduced locomotor activity. Notably, this sleep elevation lasted well beyond the cessation of previously reported host-seeking phases, raising the possibility of distinct phases of opportunistic versus targeted host pursuit. To determine the dietary basis of this effect, we tested mosquitoes fed a bovine serum albumin (BSA)-based diet. BSA feeding alone was sufficient to mimic the sleep-promoting and activity-reducing effects of blood, suggesting dietary protein is a major nutritional regulator. Moreover, RNAi-mediated knockdown of the leucokinin receptor (Lkr), which has previously been associated with fluid homeostasis and feeding behavior, resulted in enhanced sleep and reduced activity, implicating mosquito LK signaling in the modulation of postprandial sleep. Together, these findings demonstrate that blood-derived proteins drive sustained increases in sleep and reductions in locomotor activity in Ae. aegypti. This work positions Ae. aegypti as a model for dissecting nutrient-specific regulation of sleep and highlights potential adaptive functions of protein-induced quiescence, such as energy conservation and predator avoidance. More broadly, it provides insight into how specialized diets shape the neural and behavioral architecture of sleep.

Energy expenditure (EE) during running depends on the interplay between active muscle work and elastic energy storage and return, yet the relative contribution of mechanical power to EE remains debated. Quantifying the relative contributions of segment-level mechanical power can provide a way to address this debate. In this study, we aimed to quantify how segment-level mechanical power contributes to EE during running and to demonstrate that these mechanistic insights support wearable-based EE estimation. Joint dynamics and respiratory gas-based EE were collected from healthy young adults running at multiple speeds. Scale factors were derived to quantitatively link efficiency-weighted segment power to measured EE. The stance leg consistently showed the strongest correlation with EE, and this dominance was preserved across speeds. Including swing-leg hip power further improved accuracy. Scale factors were approximately 0.45, suggesting that active muscle work and elastic energy return contribute comparably to the mechanical power associated with EE. Using a lightweight machine learning model, stance-leg and swing-leg hip joint power were reconstructed from a single sacral IMU, enabling accurate EE prediction. These findings demonstrate that lower-limb mechanical power is a robust predictor of running EE, supporting both the extensibility of biomechanically-informed frameworks and wearable-based EE monitoring.

Perception and action form a loop as an organism interacts with its environment. Here, we probe the interdependence between the two in the context of choice-history effects, the tendency of past choices and their outcomes to bias current choices. We investigated if the modulation by previous choices depended on the coupling between perceptual choices and the motor responses reporting these choices. We analyzed data from non-human primates (Macaca mulatta; three females and five males) performing two different perceptual decision-making tasks. Each task had a coupled variant with a fixed mapping between perceptual choices and motor plans used to report the choice, and an uncoupled variant, in which this mapping varied randomly across trials. We found that, in both tasks, the animals in the coupled variant had larger choice-history effects compared to the animals in the uncoupled variant. Decoupling choices from motor responses was further associated with an inability by the animals to learn experimentally induced stimulus sequence regularities. Together, these findings identify choice-response coupling as a factor shaping animals ability to use recent history to guide behavior and highlight the tight link between cognitive and motor processes.

Quantitative gait analysis could enhance personalized treatment for many movement-related conditions; however, it is not routinely integrated into clinical care. Advances in mobile sensing, such as smartphone-based motion capture, enable rapid clinical gait assessment, but extracting actionable insights remains challenging. Although machine learning models can support clinical decisions from gait data, they typically require costly task- and condition-specific datasets, which limits progress across various gait-related conditions. Here we present a generative foundation model of walking kinematics that enables various downstream clinical tasks across diverse patient populations using clinically accessible smartphone video-based gait analysis. We aggregated eight gait datasets comprising 657 individuals across seven unique pathologies. Using weakly-supervised learning, we trained a variational autoencoder to distill high-dimensional gait kinematics into a 16-dimensional learned latent representation. We demonstrate generalizability across four downstream clinical tasks spanning pathologies both seen and unseen during training, with and without model fine-tuning, including: 1) classification of neuromuscular disorders unseen during training, 2) predicting clinical severity scores for individuals with Parkinson's disease, 3) tracking of subacute recovery post-stroke, and 4) generating patient-specific kinematic changes following total hip arthroplasty. Our model also computes a deviation from mean unimpaired (DMU) score, an interpretable scalar metric that captures an individual's deviation from typical unimpaired gait, providing rapid, holistic quantification of impairment. This generalizable model provides a foundation for clinically actionable tools that translate mobile sensing-derived gait data into precise biomechanical insights for clinical research and decision-making. The open-source model is deployed in the cloud for automated smartphone video-based gait analysis on our freely available OpenCap platform.

Some blind individuals use echolocation, a skill that allows them to better navigate their environment using echoes from self-generated mouth clicks reflected off surrounding surfaces. Echolocation involves a complex interplay of sensory accumulation, information processing, dynamic prediction, motor planning and execution in real-time. Computational modeling offers a valuable approach to understanding the cognitive and neural mechanisms underlying echolocation performance, in particular the temporal dynamics of the process. We present a computational model of human echolocation behavior based on a Kalman filter, where we treat the echolocator as an active sensor that maintains an internal belief about the target's location and continuously refines it via echo feedback. The model, based on observations of echolocation in blind human experts, simulates the use of mouth clicks and returning echoes to localize and orient toward a target under varying conditions. In the experiment, the target is placed at a random azimuth in the frontal plane. An echolocator aims a series of mouth clicks in various directions and infers the target azimuth using acoustic information received from the click echoes. The system integrates three major components: (1) a simulation of echoacoustic interaural time differences (ITD) to estimate the relative head-target angle; (2) a Kalman filter that processes these ITDs to iteratively update probabilistic beliefs about target location and associated uncertainty; and (3) a motor control system that modulates head movements with the current belief state. The Kalman filter serves as a representation of the internal state of the observer, where its beliefs drive the direction of head rotation, and its uncertainty estimates drive head velocity adjustments. Model performance demonstrates that simple predictive computational approaches can reproduce key aspects of echo-guided sensorimotor learning, providing a framework that may be leveraged to develop biologically plausible models, advance understanding of best practices, and potentially improve intervention strategies.

Low-intensity vibration (LIV) is a non-invasive mechanical stimulus capable of regulating skeletal adaptation and cellular signaling pathways involved in bone remodeling. Despite growing interest in LIV, substantial methodological heterogeneity persists in the selection of experimental vibration parameters such as frequency, expressed in Hertz (Hz) and intensity, defined as earth's gravitational field (g) (9.81 m/s2). Focusing on micro-computed tomography (CT) derived trabecular bone volume fraction (BV/TV) as the main outcome measure, this study sought to synthesize the effects of different LIV frequency and intensity on BV/TV in small rodents (mice and rats) as they remain as the most studied pre-clinical model. To accomplish this, we performed a systematic review searching for publications in English on PubMed, Web of Science, CINAHL, and Embase databases. Two independent investigators followed inclusion criteria to select only peer-reviewed studies with mature mice, using whole-body vibration experiments without other co-variables. We further restricted to include studies that analyzed non-fractured bones and compared pre- and post-intervention or control values. In addition to these core criteria, a detailed hierarchical screening framework was applied during full-text review. The two independent investigators extracted data independently and considered the characteristics of the study, animals' characteristics, intervention characteristics, and results. For this study we considered load-bearing hindlimbs, femur and tibia, separately but did not include vertebrae in the analysis. A Bayesian network meta-analysis and a revised SYRCLE risk of bias (RoB) tool were used to evaluate the risk of bias across included studies. Seven studies met the inclusion criteria. Results showed that an LIV regime applied at 45Hz at 2g presented higher chances to increase trabecular BV/TV of the mouse tibia (estimated effect 3.22 [CrI 1.98, 4.45]), while LIV regimes applied to the femur at 90Hz and 1.4g (estimated effect 3.08 [CrI -1.99, 7.97]) present better chances to increase trabecular BV/TV results compared to other interventions but with no significant differences. Finally, we applied 45Hz at 0.2g LIV to 5 month old male C57BL/6 for 5 weeks (n=10/group) which showed significantly increased Trabecular Thickness (Tb.Th) for both the tibia (10%, p<0.01) and femur (17%, p<0.001), with the femur showing further increases in trabecular BV/TV (32%, p<0.05) compared to non-LIV controls. We conclude that changes in the microarchitectures of the tibia and femur respond differently to the same application of LIV (45Hz, 0.2g) in mice and rats.

The heterotrimeric G-protein transducin (Gt) is among the key proteins mediating phototransduction in rod and cone cells of the vertebrate retina. Even though this protein has been extensively characterized in mammals, little is known about its expression patterns in migratory songbirds. Here we characterised Gt expression in the European robin, a night-migratory songbird known for its light-dependent magnetoreception. The mechanism underlying magnetoreception is not fully understood, but one well-supported hypothesis involves a radical-pair formation in the blue light receptor cryptochrome type 4a. The - and {gamma}-subunits of cone specific transducin have been identified as possible interaction partners of cryptochrome 4a. Therefore, we analysed the expression patterns of various G-protein subunits in bird photoreceptors. Specifically, we combined single cell RNA sequencing and immunohistochemistry, and tested for protein interaction by pulldown, co-immunoprecipitation, and NanoBiT luminescence assays. We show that genes for G-protein subunits GNB1 and GNB3 (coding for Gt{beta}1 and Gt{beta}3, respectively) are predominantly expressed in rods and cones. Among {gamma}-subunits, GNGT2 (coding for Gt{gamma}T2) was the principal isoform in cones, whereas GNG11 (coding for Gt{gamma}11) was associated with rods. In contrast, we did not detect GNG10 (coding for Gt{gamma}10) expression in either photoreceptor type. Interaction assays demonstrated that all three {beta}{gamma} combinations; {beta}{gamma}T2, {beta}{gamma}10, and {beta}{gamma}11, can associate in vitro. These findings indicate that {beta}{gamma} dimer formation in vivo is likely constrained by the photoreceptor-specific expression of the respective subunits. Furthermore, the absence of GNG10 expression in rods and cones does not support a role of this {gamma}-subunit in photoreceptor-based magnetoreception.

Collective movement requires coordination between individuals, yet how this emerges during early interactions remains poorly understood. We investigated how partner familiarity influences coordination, leader-follower dynamics, and learning in homing pigeon pairs navigating from novel sites. Birds were released repeatedly with either familiar or unfamiliar partners, followed by solo releases to assess learning. By quantifying bidirectional information flow, we found familiarity influenced information-transfer dynamics during the first release: familiar pairs exhibited more asymmetric information transfer, likely reflecting established leader-follower relationships, whereas unfamiliar pairs showed more symmetric exchange. These differences disappeared after one release. Conversely, familiarity had little effect on cohesion or navigational performance. There was some evidence for an influence on learning: birds from familiar pairings had higher homing efficiency on a subsequent solo release. Finally, across partnerships, followership was more predictable than leadership with respect to individual identity and flight speed, indicating stable variation in individuals' tendency to follow rather than lead. This suggests that a shift in emphasis from leadership to followership might enhance our understanding of collective decision-making dynamics. Our results demonstrate how flight partners rapidly coordinate, producing limited downstream effects on navigation and learning, with implications for many animals that travel in fission-fusion transitory collectives.

Animals balance staying in a favorable environment with exploring new ones. In C. elegans chemotaxis, the process by which worms migrate toward an attractant has been extensively studied. However, what happens after they reach it remains largely unexplored, partly because conventional assays immobilize worms at the point of arrival. Here, we quantitatively analyzed chemotactic behavior upon reaching an attractive odor source using an immobilization-free chemotaxis assay. We observed that 62% animals left the isoamyl alcohol region after initially approaching it, a behavior we termed "leaving behavior." Quantitative analysis revealed that leaving behavior represents a distinct locomotor state compared with free-moving, high-concentration odor avoidance, and approach behavior. To test whether leaving behavior is related to olfactory adaptation, we analyzed mutants in adaptation-related genes. The proportion of leaving behavior was significantly increased in egl-4 loss-of-function mutants compared with wild-type animals, whereas arr-1 mutants showed no significant difference. These results suggest that egl-4 negatively regulates leaving behavior, suggesting a role for this kinase in stabilizing post-arrival behavioral states beyond its known function in olfactory adaptation. Our findings indicate that chemotaxis involves dynamic behavioral transitions even after reaching an attractant, consistent with an exploration-exploitation trade-off framework.

Understanding the genetic structure of keystone species within river networks is essential for effective conservation and management. While population differentiation of anadromous species often occurs between river systems, less research has been conducted on differentiation within rivers with smaller catchment areas. In this study, we investigated the population genetic structure of wild Atlantic salmon (Salmo salar) across the small-scale river Ronne [a] system in southernmost Sweden using Restriction-site Associated DNA sequencing (RADseq). Although the Admixture analysis did not detect clearly defined genetic clusters, significant pairwise FST values and DAPC revealed emerging population differentiation among the Ronne [a] tributaries. The observed patterns are consistent with a system characterized by connectivity, where genetic flow is present but can be reduced by behavioral and ecological factors such as spawning homing behavior and selective movements. These findings suggest that, despite overall connectivity, Atlantic salmon populations in the Ronne [a] catchment area may function as partially independent sub-populations. This highlights the importance of conservation and management strategies in fragmented river systems to consider population genetic structure to support resilient salmon populations under ongoing anthropogenic pressures.

To make adaptive foraging decisions, naive individuals need to gather information about the local prey community. Besides sampling prey personally, the young could gather information about prey profitability by observing the foraging behaviour of other individuals, and parental provisioning provides the first opportunity to acquire this social information. Still, previous research on vertical transmission of prey preferences from parents has provided mixed results that are often confounded with other information sources, such as siblings and peers. It is also not known whether information from parents can change potential innate biases against certain prey types, such as avoidance of warningly coloured insects. Here, we tested whether social information acquired by offspring during parental provisioning influences the development of prey preferences in a generalist predator, the Great Tit (Parus major). We brought 15 great tit broods and their parents into captivity at late nestling stage (14 days old) and divided them into three social information treatments where parents were provided with either brown, red or yellow palatable maggots to feed to their dependent young for 8 days. Once foraging independently from parents, we conducted a preference test where juveniles were offered the full array of coloured maggots. Regardless of palatable exposure to typical warning-coloured maggots (i.e. red and yellow), juveniles consistently preferred yellow over red, and preferred brown maggots the most (i.e. lacking warning coloration). This supports the existence of innate biases against typical warning colours, and that social information from parents is unlikely to override these, at least when alternative prey is easily available.

Capturing naturalistic behavioral dynamics is essential for understanding social interaction in ecologically valid settings. Existing investigations of naturalistic social interaction rely on time-aggregated analysis methods better suited for task-based experiments, which lose the complex, moment-to-moment dynamics exhibited in naturalistic settings. The emerging field of topological data analysis (TDA) provides new tools to characterize fine-grained dynamics in time-series data that cannot be captured by time-averaged methods. The present work utilizes Temporal Mapper, a recently developed TDA specifically tailored to analyzing dynamical systems. Temporal Mapper characterizes complex temporal dynamics as transition networks, where nodes are stable states and edges are transitions between states. Originally designed for human neural time series analysis, here we demonstrate the utility of Temporal Mapper to capture rich animal postural dynamics during naturalistic social interaction. We utilized an existing dataset with 12 video recording sessions of two domestic ferrets (Mustela putorius furo) during naturalistic interaction and tracked the postures of animals during social interaction. Ferrets were chosen due to their strong social-cognitive skills and rich postural dynamics for investigating social behavior via posture estimation. Temporal Mapper was then used to represent the postural dynamics as transition networks for each recording session. Here, we found that posture states are significantly smaller and more widespread during active social interaction compared to non-social activities. Additionally, the number of sequential postural states before transitioning to new behaviors is more consistent during active social interaction than non-social activities. Together, our findings suggest that social activity has a broad range of unstable postural states arranged in consistent sequences. Our method, Temporal Mapper, allows for network structure analysis of complex naturalistic data, applicable for characterizing rich dynamics in different species, scales, and paradigms.

The evolution of flippers for wing-powered diving in penguins is a striking example of tetrapod limb specialization. The modern penguin flipper is structurally reinforced by a characteristic dorsoventral flattening of the long bones accompanied by a reduction in distal forelimb musculature, features which emerged convergently in flightless diving birds and aquatic mammals. While an extensive fossil record informs the morphological sequence through which these changes occurred, the evolutionary pressures and developmental mechanisms underlying these modifications are unknown. We find that in avian and mammalian forelimbs, a flattened bone morphology only emerged in aquatic lineages that lost ancestral modes of locomotion, including in flightless diving birds, pinnipeds, and cetaceans. Using penguin embryos as an accessible model for investigating flipper development, we demonstrate that early patterning of forelimb musculoskeletal morphology is similar to that seen in forelimbs of non-aquatic birds. Instead, later modifications of gene expression and cell and tissue behaviors underlie flipper phenotypes. Thus, we find that in the early penguin forelimb, the initial cues that pattern the muscle do not differ from other avian species, however late embryonic changes in proliferation result in dramatic reduction of muscle. Likewise, forelimb bones in penguins initially have similar cross-sectional proportions to those in flighted birds. The shape of these bones is, however, remodeled late in embryonic development through a process that shares molecular hallmarks with bone ridge formation at tendon attachment sites. In these bones, ridge-forming tissue initiates at the ends of the bones (the epiphyses) and extends into tendon-like connective tissue along the lateral edges of the bone, widening the long bones along the anterior-posterior axis and producing a flattened bone. Using spatial transcriptomics and comparative genomic tools we determine that differentially expressed genes between the ridge-forming tissues and long bone cartilage are significantly enriched for signals of selection in the penguin lineage and that these genes may also be convergently evolving in marine mammals. Together, these data show that the evolution of musculoskeletal morphology in the penguin flipper occurred through expansion or novel deployment of molecular programs typically associated with tendon-attachment sites during late embryonic development.