Tag: システム行動学

  • What Exactly Is Systems Ethology?

    What Exactly Is Systems Ethology?

    Introduction

    Systems Biology was first proposed in 1998.
    In recent years, I feel that the concept of systems biology has gradually gained acceptance.

    For an excellent and very accessible overview of systems biology, see Tetsuya J. Kobayashi’s article “What exactly was systems biology?“.

    I apologize if I am mistaken, but I think systems biology can be summed up in a single phrase: “understanding life as a system.”

    My sense is that this introduced a new perspective across biology as a whole, and especially into a wide range of fields such as molecular biology, developmental biology, and chronobiology.

    My own research is in neuroethology, or more broadly ethology.
    In recent years, computational studies that treat animals as systems—neural simulations, connectome data analysis, and machine learning tools for analyzing animal behavior—have been widely published in neuroethology as well.

    Honestly, I do not know whether this kind of impact on neuroethology was anticipated when systems biology was first proposed, or whether the current situation falls within the scope of Systems Ethology.

    Still, it is a fact that this trend is advancing in neuroethology too, and in that sense I feel that the proposal of systems biology was remarkably prescient.

    The beginnings of Systems Ethology

    One question that naturally arises is: if the current trends in ethology are already being absorbed into systems biology, why bother to call it Systems Ethology?

    That is a fair point, and I do wonder about it myself. Even so, there were reasons that led me to deliberately propose Systems Ethology.

    This comes largely from a personal wish of mine: I wanted a place where researchers who study behavior could come together.

    Behavioral research today is pursued from many different perspectives. Alongside the central goal of neuroethology—understanding how neural circuits control behavior—there is ecology, which asks what function a behavior serves in its environment; evolutionary biology, which asks how a behavior evolved; and information science, which asks what strategies behavior implements and what kind of system it constitutes. Behavior is being studied across disciplinary boundaries.

    Even within neuroethology, the range is broad—from circuit-level work in model organisms such as fruit flies and mice to studies of unique behavioral mechanisms in non-model organisms—and it goes by different names, such as neuroethology or behavioral neurobiology.

    Adding to this, the approaches used to study behavior have also diversified: genetic manipulation of neural activity, recording of neural activity by electrophysiology and calcium imaging, behavioral quantification using machine learning, mathematical descriptions of behavior, exploration of behavioral strategies with reinforcement learning, and more classical detailed observation of behavior. Here too, methods cross disciplinary lines.

    As a result, my personal impression is that each of us attends whichever conference happens to be closest to the scope of our own research, and somewhere along the way I found myself wondering: where exactly is my home conference?

    Behavioral research is being pursued from ever more diverse perspectives, and the current reality is that we often want to learn about other groups’ methods but have no way to do so.
    It struck me as a great missed opportunity, and I imagined that if there were a venue where all of these came together, behavioral research might advance even further. That was the intent behind creating such a place.

    This is purely my own view, and I do not assume that those collaborating with me share the same motivation, nor that this rests on my own agenda alone—but the desire for a venue where everyone could gather was something I heard again and again in conversations with many people at conferences.

    And so, in order to create such a venue, I set out to propose Systems Ethology. In September 2024, at SWARM2024, I submitted a paper titled

    Toward Understanding the Principles of Animal Behaviors: Systems Ethology
    Hayato M Yamanouchi, Yusuke Notomi, Ryoya Tanaka, Shumpei Hisamoto, Shigeto Dobata

    and

    Systems Ethology: Toward Elucidating the Design Principles of Animal Behavior

    and organized the following Organized Session.

    The following is the abstract of this OS, quoted from the SWARM2024 website.

    The exploration of biology plays a crucial role in elucidating the behavioral mechanisms of individual agents and their collective behavior as swarms, owing to the complex and diverse nature of animal behavior. In particular, recent remarkable advances in information processing technology have helped to elucidate their complex behavioral patterns. Furthermore, these technological innovations also enable detailed investigations into non-model species where established research tools are lacking, thereby contributing to a broader understanding of various biological phenomena. 
     Currently, methods for animal behavior analysis are highly diversified, necessitating opportunities for integrated discussions where specialists with cutting-edge knowledge can share their techniques. We therefore propose a framework called “Systems Ethology” to elucidate the design principles of animal behavior. With the overarching goal of understanding animal behavior as a system, we aim to facilitate the exchange of information regarding various approaches to elucidating the mechanisms underlying each behavior. Such cross-disciplinary interactions among researchers can assist in performing more efficient and meaningful research.
     In this session, we aim to gather biological insights from various disciplines such as ecology, ethology, and neuroscience. Additionally, proposals for diverse behavioral analysis approaches, incorporating insights from information science and engineering, are also encouraged.

    For this OS we invited researchers studying behavior from a wide range of perspectives, and we were able to put together a very well-attended session.

    We are now planning what has long been a hope of mine: launching a Systems Ethology research meeting.

    The ideas behind Systems Ethology

    Next, I would like to discuss what Systems Ethology actually is and how we define it as a discipline.

    Ethology has long been centered on “Tinbergen’s four questions,” proposed by Niko Tinbergen.

    There are various ways of phrasing and framing them, but here we take the four perspectives to be “survival value,” “ontogeny,” “evolution,” and “causation.”

    Niko Tinbergen argued that behavior must be examined from these four perspectives, and they remain an important guiding framework for ethology today, and by extension for neuroethology.

    Today, however, each of these perspectives is pursued by research that focuses on it alone.

    To take a clear example, neuroethology focuses on how neurons control behavior from the standpoint of causation, ecology focuses on survival value, and evolutionary biology focuses on evolution.

    Answering each of these questions matters, but these perspectives have tended to exist separately, with anything outside a given focus treated as a black box.

    In other words, I feel there has never been a discipline that focuses on the causation and the survival value of behavior at the same time, covering both scopes. (Of course, at the level of individual studies, work that combines these views is becoming more common.)

    Truly understanding behavior requires all four perspectives—“survival value,” “ontogeny,” “evolution,” and “causation”—yet they have rarely been integrated. No one can say for certain why, but one factor is surely how difficult it is for these perspectives to understand one another.

    Put the other way around: what if we simply made mutual understanding possible?
    Here, the notion of a system may serve as the medium that solves this problem.

    That is, if the system becomes a kind of shared language linking these perspectives, each side can incorporate the others’ ideas.

    Here, I define this system as the behavioral system.

    There are many ways to conceive of behavior, and no single definition fits them all, but here I take a simpler view, defining behavior as:

    The internally coordinated responses (actions or inactions) of whole living organisms (individuals or groups) to internal and/or external stimuli, excluding responses that are more easily understood as developmental changes

    Japanese translation: 生物全体(個体または群れ)が内部および/または外部からの刺激に対して内部的に調整した反応(行動または不行動)であり、発達上の変化としてより容易に理解できる反応は除く。
    ~Source: Levitis et al. (2009)~

    That is the definition I adopt.

    Recast in terms of a system, behavior corresponds to the system’s output.
    The parts of the system responsible for input and processing are therefore treated as the behavioral system.

    A behavioral system can be viewed at multiple levels: with the individual as the unit, the neuron as the unit, or the molecule as the unit.
    Given this diversity of levels, understanding behavior requires working across them and making sense of behavior at several levels at once.

    Because this makes things so complex, however, fully understanding a single system within one study becomes impossible.
    That is precisely why we need a venue for sharing diverse approaches and insights across disciplinary boundaries.

    Returning to “Tinbergen’s four questions,” Systems Ethology proposes the following four guiding principles for answering them.

    1) System structure: Static structure represented by nodes and edges. It includes discovering the behavior and the pathways by which the behavior is formed from inputs. It also elucidates the system’s components and how they are connected. 

    2) System dynamics: The dynamic structure is represented by nodes and edges. This perspective includes dynamics of the system (structure changes of the systems) and dynamics on the system (inner-statements transition of the nodes in the system). Dynamic system structure changes as a result of learning, development, and other parameters. 

    3) The control method: Methods for controlling the system’s states and behaviors. This method includes behavioral changes caused by intervention on the system’s nodes or edges. 

    4) The design method: Meaning and principles of the system. This method includes understanding the system’s adaptive and evolutionary significance. 

    ~Source: Yamanouchi et al. SWARM2024, (2024)~

    These principles indicate which aspects of a system we set out to clarify, and they are based on the four principles of systems biology.

    No one is certain where Systems Ethology will go from here, and as things stand the details have yet to be unified.

    That is exactly why I want to build Systems Ethology as a venue for thinking about the future of the study of behavior.

    The future of Systems Ethology

    At present, Systems Ethology is a newly proposed framework, and one that exists only as a conference paper.

    Eventually, I hope to present it to the wider community in a form such as an opinion article.

    My more immediate hope is to realize the original goal: a place where people who study behavior can come together.

    To start, I would like to organize a Systems Ethology study group here in Japan.

    Closing remarks

    This article reflects my personal views and is not meant to define Systems Ethology definitively.

    Please keep in mind that others may see things differently, and that I may well be mistaken on some points.