Exploring the connection between emergent animal personality and fitness using a novel individual-based model and decision tree approach

Synergistic Integration Between Machine Learning and Agent-Based Modeling: A Multidisciplinary Review

Agent-based modeling (ABM) involves developing models in which agents make adaptive decisions in a changing environment. Machine-learning (ML) based inference models can improve sequential decision-making by learning agents' behavioral patterns. With the aid of ML, this emerging area can extend traditional agent-based schemes that hardcode agents' behavioral rules into an adaptive model. Even though there are plenty of studies that apply ML in ABMs, the generalized applicable scenarios, frameworks, and procedures for implementations are not well addressed. In this article, we provide a comprehensive review of applying ML in ABM based on four major scenarios, i.e., microagent-level situational awareness learning, microagent-level behavior intervention, macro-ABM-level emulator, and sequential decision-making. For these four scenarios, the related algorithms, frameworks, procedures of implementations, and multidisciplinary applications are thoroughly investigated. We also discuss how ML can improve prediction in ABMs by trading off the variance and bias and how ML can improve the sequential decision-making of microagent and macrolevel policymakers via a mechanism of reinforced behavioral intervention. At the end of this article, future perspectives of applying ML in ABMs are discussed with respect to data acquisition and quality issues, the possible solution of solving the convergence problem of reinforcement learning, interpretable ML applications, and bounded rationality of ABM.

Challenges in Linking Cognition and Survival: A Review

Linking cognitive performance with fitness outcomes, measured using both reproductive and survival metrics, of free-living animals is crucial for understanding the evolution of cognition. Although there is increasing evidence showing a link between cognitive traits and reproductive success metrics, studies specifically exploring the link between cognition and survival are scarce. We first explore which cognitive traits related to survival have been investigated in free-living animals. We also discuss the challenges associated with investigating the links between cognition and survival. We then review studies that specifically consider survival of animals of known cognitive abilities that are either free-living or in captivity and later released into the wild. We found nine studies exploring cognitive traits in wild populations. The relationships between these cognitive traits and survival were equivocal. We found a further nine studies in captive-reared populations trained to predator cues and later released into the wild. Training to recognize predator cues was correlated with increased survival in the majority of studies. Finally, different individual intrinsic characteristics (i.e., age, body condition, personality, sex) showed varied effects between studies. We argue that finding ecologically relevant cognitive traits is crucial for gaining a better understanding of how selection impacts certain cognitive traits, and how these might contribute to an individual’s survival. We also suggest possible standardized, easy to implement, cognitive tests that can be used in long-term studies, which would generate large sample sizes, take into account intrinsic characteristics, and provide an opportunity to understand the mechanisms, development and evolution of cognition.

The relationships between toad behaviour, antipredator defences, and spatial and sexual variation in predation pressure

BACKGROUND

Animal behaviour is under strong selection. Selection on behaviour, however, might not act in isolation from other fitness-related traits. Since predators represent outstanding selective forces, animal behaviour could covary with antipredator defences, such that individuals better suited against predators could afford facing the costs of riskier behaviours. Moreover, not all individuals undergo equivalent degrees of predation pressure, which can vary across sexes or habitats. Individuals under lower predation pressure might also exhibit riskier behaviours.

METHODS

In this work, I tested these hypotheses on natterjack toads (Epidalea calamita). Specifically, I gauged activity time, exploratory behaviour and boldness in standard laboratory conditions, and assessed whether they correlated with body size and antipredator strategies, namely sprint speed, parotoid gland area and parotoid gland colour contrast. Additionally, I compared these traits between sexes and individuals from an agrosystem and pine grove, since there is evidence that males and agrosystem individuals are subjected to greater predation pressure.

RESULTS

Sprint speed as well as parotoid gland contrast and size appeared unrelated to the behavioural traits studied. In turn, body mass was negatively related to activity time, boldness and exploration. This trend is consistent with the fact that larger toads could be more detectable to their predators, which are mostly gape unconstrained and could easily consume them. As predicted, females exhibited riskier behaviours. Nonetheless, agrosystem toads did not differ from pine grove toads in the behavioural traits measured, despite being under stronger predation pressure.

Parasites differentially impact crayfish personality in different contexts

The expression of an individual animal’s behaviour can be placed along many different personality spectra. Parasite load can alter animal behaviour and, thus, fitness. The personality traits of rusty crayfish, Faxonius rusticus, were analysed in three different behavioural contexts: foraging, exploration, and threatened. Each crayfish was tested in each context 3 times, giving a total of 9 assays per crayfish. After assays were completed, crayfish were dissected, and the hepatopancreas of each crayfish was photo analysed to determine the parasite load of the trematode, Microphallus spp. A composite personality score for each assay and parasite load was loaded into a PCA. The PCA model showed that as parasite load increased, crayfish became bolder in threatening contexts and less exploratory in novel environments, whether or not a food stimulus was present. Thus, parasite load alters the placement of crayfish on different personality spectra, but this change is context specific.

An efficient new assay for measuring zebrafish anxiety: Tall tanks that better characterize between-individual differences

BACKGROUND

Zebrafish (Danio rerio) are increasingly being used to model anxiety. A common behavioral assay employed for assessing anxiety-like behaviors in zebrafish is the "novel tank test". We hypothesized that using deeper tanks in this test would result in greater between-individual variation in behavioral responses and a more 'repeatable' assay.

NEW METHODS

After mapping the literature and identifying common behavioral parameters used in analysis, we performed novel tank anxiety tests in both custom-designed 'tall' tanks with increased depth and 'short' trapezoidal tanks. We compared the repeatability of the behavioral parameters between tall and short tanks and also investigated sex differences.

RESULTS

Overall, regardless of tank depth, almost all behavioral parameters associated with anxiety in zebrafish were significantly repeatable (R = 0.24 to 0.60). Importantly, our tall tanks better captured between-individual differences, resulting in higher repeatability estimates (average repeatability tall tanks: R = 0.46; average repeatability short tanks: R = 0.36) and clearer sex differences.

CONCLUSIONS

Our assay using tall tanks has advantages over tests based on short tanks which underestimate repeatability. We argue that use of deeper tanks will improve the reliability of behavioral data across studies using novel tank tests for zebrafish. Our results also call for increased attention in designing the most appropriate assay in biomedical and behavioral sciences as current methods may lack the sensitivity to detect subtle, yet important, information, such as between-individual variation, an important component in assessing the reliability of behavioral data.

Computational animal welfare: towards cognitive architecture models of animal sentience, emotion and wellbeing

To understand animal wellbeing, we need to consider subjective phenomena and sentience. This is challenging, since these properties are private and cannot be observed directly. Certain motivations, emotions and related internal states can be inferred in animals through experiments that involve choice, learning, generalization and decision-making. Yet, even though there is significant progress in elucidating the neurobiology of human consciousness, animal consciousness is still a mystery. We propose that computational animal welfare science emerges at the intersection of animal behaviour, welfare and computational cognition. By using ideas from cognitive science, we develop a functional and generic definition of subjective phenomena as any process or state of the organism that exists from the first-person perspective and cannot be isolated from the animal subject. We then outline a general cognitive architecture to model simple forms of subjective processes and sentience. This includes evolutionary adaptation which contains top-down attention modulation, predictive processing and subjective simulation by re-entrant (recursive) computations. Thereafter, we show how this approach uses major characteristics of the subjective experience: elementary self-awareness, global workspace and qualia with unity and continuity. This provides a formal framework for process-based modelling of animal needs, subjective states, sentience and wellbeing.

Paternal exposure to a common herbicide alters the behavior and serotonergic system of zebrafish offspring

Increasingly, studies are revealing that endocrine disrupting chemicals (EDCs) can alter animal behavior. Early life exposure to EDCs may permanently alter phenotypes through to adulthood. In addition, the effects of EDCs may not be isolated to a single generation − offspring may indirectly be impacted, via non-genetic processes. Here, we analyzed the effects of paternal atrazine exposure on behavioral traits (distance moved, exploration, bottom-dwelling time, latency to enter the top zone, and interaction with a mirror) and whole-brain mRNA of genes involved in the serotonergic system regulation (slc6a4a, slc6a4b, htr1Aa, htr1B, htr2B) of zebrafish (Danio rerio). F0 male zebraFIsh were exposed to atrazine at 0.3, 3 or 30 part per billion (ppb) during early juvenile development, the behavior of F1 progeny was tested at adulthood, and the effect of 0.3 ppb atrazine treatment on mRNA transcription was quantified. Paternal exposure to atrazine significantly reduced interactions with a mirror (a proxy for aggression) and altered the latency to enter the top zone of a tank in unexposed F1 offspring. Bottom-dwelling time (a proxy for anxiety) also appeared to be somewhat affected, and activity (distance moved) was reduced in the context of aggression. slc6a4a and htr1Aa mRNA transcript levels were found to correlate positively with anxiety levels in controls, but we found that this relationship was disrupted in the 0.3 ppb atrazine treatment group. Overall, paternal atrazine exposure resulted in alterations across a variety of behavioral traits and showed signs of serotonergic system dysregulation, demonstrating intergenerational effects. Further research is needed to explore transgenerational effects on behavior and possible mechanisms underpinning behavioral effects.