Biology subjects

Hein, A. M.

Publications and source records attributed to Hein, A. M..

7 recordsLinked to original sources

bioRxiv2026

Brinkmanship and the sensory-motor limits of pursuit predation

Predators routinely fail to capture prey despite being larger, faster, and more powerful. Half a century ago, theory predicted that prey could evade biomechanically superior predators using brinkmanship, delaying evasive maneuvers until the last possible instant, but this hypothesis has never been tested in the wild. Here, using more than 2,000 hours of high-speed stereo-camera recordings of predator and prey coral reef fishes, we show that prey successfully evade 97% of attacks. However, contrary to the standard explanation that prey evade through superior maneuverability, we find that prey evade by exploiting their predator's sensory-motor delay - the interval between a stimulus reaching the predator and the onset of the predator's motor response. Although predators partially compensate for this by forecasting prey position during interception steering, prey consistently time evasive maneuvers within a brief window just tens of milliseconds before collision, within which no viable predator trajectory can intercept them. Prey do not, however, turn at the time that maximizes miss distance, but slightly later, a bias that makes escape robust to variability in their own response timing. These findings reveal how temporal delays in perception and action place fundamental limits on the outcome of predator-prey interactions, and show that prey can exploit delays in a way that approaches the theoretical limit of escape performance.

animal behavior and cognition↗

bioRxiv2026

Natural enemies mediate the impact of plant microbiota on insect-borne virus transmission

Individual hosts can be protected from disease vectors through their microbiota, but how those benefits scale up to the host population level is rarely obvious. Disease spread may be inhibited through reduced vector visitation or amplified as vectors redirect their attention to unprotected hosts. The outcome--suppression or escalation of disease outbreaks--is likely to be influenced by natural enemies that redistribute or suppress vector populations. The challenge of determining how these processes interact across scales is exemplified by pea aphids, major virus vectors in pea crops that are commonly managed using parasitoid wasps. Recent evidence suggests that epiphytic bacteria in the genus Pseudomonas can repel or kill pea aphids, yet whether Pseudomonas-mediated protection from aphids complements or undermines parasitoid-based vector control remains unknown. We developed a multi-scale mathematical model to investigate why Pseudomonas complements versus undermines biocontrol of aphid-vectored virus outbreaks. Because Pseudomonas-induced mortality leaves aphids patchily distributed, the effect of Pseudomonas on virus outbreaks depends most strongly on how well parasitoids track aphid densities: When parasitoids efficiently locate aphids, Pseudomonas inhibits virus outbreaks by reducing aphid densities. However, when parasitoids are less efficient, they often search unproductively for aphids on Pseudomonas-inhabited plants with few aphids, slowing parasitoid population growth. Reduced parasitism intensifies aphid crowding on Pseudomonas-free plants and boosts the production of winged aphids that accelerate viral spread. Counterintuitively, the more effective Pseudomonas is at killing aphids, the more strongly it generates spatial variability and promotes virus spread. Our results show how community context dictates whether individual microbiota protection slows or accelerates viral outbreaks at the population scale.

ecology↗

bioRxiv2026

Collective behavior drives diversification across the tree of ray-finned fishes

Trait evolution and lineage diversification are thought to be driven by selection on the phenotypes of individuals. However, many important phenotypes are only manifested when individuals interact. Collective behavior is one such phenotype. During collective behavior, interactions among individuals lead to emergent group-level phenotypes that can govern individual fitness. Yet, whether collective behaviors exhibit similar patterns of trait evolution to individual traits and whether these group phenotypes have influenced macroevolutionary diversification is unknown. In ray-finned fishes, the most species-rich of all vertebrate classes, many species exhibit striking displays of collective behavior in the form of collective movement. Here, we use state-of-the-art machine learning methods for text and image analysis to quantify collective movement phenotypes for 2,839 Actinopterygian fishes, comprising nearly 10% of all extant species. We find that collective movement is highly evolutionarily stable in some clades and extremely labile in others, exhibiting over 300 evolutionary transitions. Consistent with theoretical predictions, collective movement is associated with more patchy diets and higher predation pressure. Finally, we find that the evolution of collective movement coincides with an increase in net macroevolutionary diversification rate. Our analyses reveal that collective behavior is deeply intertwined with ecological and evolutionary dynamics across a vast phylogenetic scale, unlocking new questions about the role of collective phenotypes in shaping the diversity of life on Earth. SignificanceCollective behavior is thought to contribute to the fitness of individuals that live in groups, but we still do not understand why some species exhibit collective behavior and others do not. We developed machine learning methods to quantify collective phenotypes for thousands of species spanning the phylogeny of ray-finned fishes and found collective behavior to be widespread and consistently correlated with experiencing higher predation pressure and foraging for grouping prey. Lineages that evolve collective movement show a higher net diversification rate. Our results illustrate that collective behavior is foundational to the broad-scale patterns of evolution and diversification in the largest vertebrate class.

evolutionary biology↗

bioRxiv2026

Collective learning and manifold behaviors in predator groups

Collective foraging in animal groups often relies on behavioral diversity, with individuals adopting different, sometimes complementary roles during shared tasks. However, most theoretical models predict that individuals responding to similar information cues in a shared environment should converge on a single optimal behavioral strategy. Using a spatially explicit multi-agent deep reinforcement learning model embedded in a three-species food chain, we show that stable behavioral diversity can emerge spontaneously among initially naive agents. Rather than converging on a single optimum, agents differentiate along a low-dimensional manifold of sensorimotor control, reflecting tradeoffs in speed regulation, spatial exploration, and deterministic turning rules. While multiple strategies yield comparable individual energetic returns, they are not interchangeable; group performance depends on how specific strategies combine to produce spatial resource partitioning and distributed directional influence. Replacing co-learned individuals with similarly competent agents trained in other groups disrupts these interaction structures and strongly reduces total energy acquisition. These results demonstrate that coordinated collective behavior and diverse, compatible strategies can arise endogenously from shared learning histories, but that this form of collective performance is path dependent and may be fragile to changes in group composition.

ecology↗

bioRxiv2024

Behavioral plasticity and the valence of indirect interactions

Behavioral plasticity in animals influences direct species interactions, but its effects can also spread unpredictably through ecological networks, creating indirect interactions that are difficult to anticipate. We use coarse-grained models to investigate how changes in species behavior shape these indirect interactions and influence the broader dynamics of ecological networks. As an illustrative example, we examine predators that feed on two types of prey, each of which temporarily reduces activity after evading an attack, thereby lowering vulnerability at the expense of growth. We demonstrate that this routine behavior can shift the indirect interaction between prey species from apparent competition to mutualism or parasitism. These shifts occur when predator capture efficiency drops below a critical threshold, causing frequent hunting failures. As a result, one prey species indirectly promotes the growth of the other by relaxing its density dependence through a cascade of network effects, paradoxically increasing predator biomass despite decreased hunting success. Empirical capture probabilities often fall within the range where such dynamics are predicted. We characterize these shifts in the qualitative nature of species interactions as changes in interaction valence, highlighting how routine animal behaviors reshape community structure through cascading changes within ecological networks.

ecology↗

bioRxiv2022

Demystifying image-based machine learning: A practical guide to automated analysis of field imagery using modern machine learning tools

Image-based machine learning methods are quickly becoming among the most widely-used forms of data analysis across science, technology, and engineering. These methods are powerful because they can rapidly and automatically extract rich contextual and spatial information from images, a process that has historically required a large amount of manual labor. The potential of image-based machine learning methods to change how researchers study the ocean has been demonstrated through a diverse range of recent applications. However, despite their promise, machine learning tools are still under-exploited in many domains including species and environmental monitoring, biodiversity surveys, fisheries abundance and size estimation, rare event and species detection, the study of wild animal behavior, and citizen science. Our objective in this article is to provide an approachable, application-oriented guide to help researchers apply image-based machine learning methods effectively to their own research problems. Using a case study, we describe how to prepare data, train and deploy models, and avoid common pitfalls that can cause models to underperform. Importantly, we discuss how to diagnose problems that can cause poor model performance on new imagery to build robust tools that can vastly accelerate data acquisition in the marine realm. Code to perform our analyses is provided at https://github.com/heinsense2/AIO_CaseStudy

ecology↗

bioRxiv2022

Wild animals suppress the spread of socially-transmitted misinformation

Understanding the mechanisms by which information and misinformation spread through groups of individual actors is essential to the prediction of phenomena ranging from coordinated group behaviours [1-3] to global misinformation epidemics [4-7]. Transmission of information through groups depends on the decision-making strategies individuals use to transform the perceived actions of others into their own behavioural actions [8-10]. Because it is often not possible to directly infer these strategies in situ, most studies of behavioural spread in groups assume individuals make decisions by pooling [7, 8, 10, 11] or averaging [8, 9] the actions or behavioural states of neighbours. Whether individuals adopt more sophisticated strategies that exploit socially-transmitted information, while remaining robust to misinformation exposure, is unknown. Here we uncover the impacts of individual decision-making on misinformation spread in natural groups of wild coral reef fish, where misinformation occurs in the form of false alarms that can spread contagiously. Using automated tracking and visual field reconstruction, we infer the precise sequences of socially-transmitted stimuli perceived by each individual during decision-making. Our analysis reveals a novel feature of decision-making essential for controlling misinformation spread: dynamic adjustments in sensitivity to socially-transmitted cues. We find that this property can be achieved by a simple and biologically widespread decision-making circuit. This form of dynamic gain control makes individual behaviour robust to natural fluctuations in misinformation exposure, and radically alters misinformation spread relative to predictions of widely-used models of social contagion.

ecology↗