Threshold sensing yields optimal path formation in Physarum polycephalum – but the mould does not know

Daniele Proverbio, Giulia Giordano · 2025

Read Threshold sensing yields optimal path formation in Physarum polycephalum – but the mould does not know by Daniele Proverbio, Giulia Giordano (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-12347.txt.

Introduction

Daniele Proverbio and Giulia Giordano study the slime mould Physarum polycephalum, an organism with no nervous system that still solves mazes and finds good paths to food. Earlier work linked this skill to memory and adaptation, but the core mechanism stayed unclear.

The pair use a circuit-like network model of the mould's foraging. They prove that threshold sensing, where flow paths respond only past a set level, makes single best paths emerge on their own. They also show which settings allow other paths, giving the mould its flexibility. The results match lab evidence and may inspire threshold-based methods for computing.

Catalog record

CREATORDaniele Proverbio, Giulia Giordano
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-12347

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Threshold sensing yields optimal path formation in Physarum polycephalum – but the mould does not know Daniele Proverbio, Giulia Giordano Abstract A circuital network model for the foraging behaviour of Physarum polycephalum proves that threshold sensing yields the emergence of optimal paths that connect food sources and solve mazes. These findings are in agreement with the experimental evidence of emergent problem solving by P. polycephalum and provide insight into the evolution of primitive intelligence. Our results can also inspire the development of threshold-based algorithms for computing applications. Do primitive organisms follow optimal strategies? If so, which mechanisms enable their optimal behaviours? Here, we demonstrate that the networking and maze solving behaviour of the acellular slime mould Physarum polycephalum leads to optimal path formation, thus supporting previous conjectures based on empirical evidence. P. polycephalum , a member of the Myxogastria class of slime moulds, is a paradigmatic organism for the study of emergent problem-solving by brainless life forms. It is a leading specimen for investigations on motility, environmental sensing [ 1 ] and response to chemical and physical stimuli [ 2 ] . It can solve puzzles and mazes by connecting food sources via shortest paths [ 3 ] , link multiple food sources [ 4 ] while avoiding risky environments [ 5 ] , and optimise transport networks [ 6 , 7 ] . Understanding how a simple organism without neural system nor central brain-led coordination [ 8 , 9 ] can display such problem-solving capabilities would shed light on the origin of primitive intelligence and the evolution of computing in living beings. Moreover, it would provide efficient bio-inspired algorithms to solve problems e.g. in optimization, maze exploration, statistics [ 10 , 11 ] . Numerous models have been developed to capture various aspects of P. polycephalum , including its life cycle, point-of-interest and foraging behaviours, nutrient relay, and maze solving [ 12 ] . Some models offer insight into oscillations and peristalsis within the cellular endoplasm (the fluid inner layer of the cytoplasm in amoeboid cells) [ 13 , 14 ] , yielding memory effects and directional migration [ 15 , 16 ] . Other models capture network formation using Hagen-Poiseuille law on graphs [ 17 ] , reaction-diffusion [ 18 , 19 ] , cellular automata [ 7 , 20 , 21 ] or multi-agent approaches [ 22 , 23 ] . To various degrees, these models reproduce experimental findings, such as the reinforcement of the main veins that follow efficient paths via flows of endoplasm and nutrients [ 12 ] ; they help explain how P. polycephalum can efficiently connect food sources in different configurations, with high fault tolerance; and they reveal how simple laws and feedback mechanisms within the organism can lead to the emergence of space exploration and pruning of redundant branches. Alternative modelling approaches are needed [ 24 ] to demonstrate the problem-solving abilities of P. polycephalum , which require some form of memory and “learning” (within the context of a microbiological interpretation). Drawing on the parallelism between gel/sol fluxes and currents, Physarum systems have thus been modelled as bio-inspired memristors (circuit elements with state- and history-dependent resistance) [ 25 , 26 ] . In fact, in its plasmodium stage, the cytoplasm of P. polycephalum differentiates into two forms: endoplasm containing fluid sol, and viscous gel-like ectoplasm. Movement follows a process called shuttle-streaming: a pressure gradient is brought about by contracting fibres in response to environmental stimuli, and the pressure potential increases up to the point at which it causes the gel to break down into sol. This drives the formation of new low-viscosity channels after a certain threshold in pressure potential is reached [ 27 ] . Restoring initial conditions upon changes in the environment requires time to break the gel structures down again, and depends on the shape and number of the formed veins [ 28 ] . Sensing of food sources and chemotaxis (movement along chemical gradients) are also governed by threshold phenomena: the response of the slime mould to environmental stimuli is usually not gradual, but step-wise [ 2 , 19 ] . The phenomena of threshold responses, nonlinear dependence of sol evolution on environmental changes, and state-dependent restoration of conditions bear resemblance to those occurring in memristor devices [ 29 , 26 ] , see Fig. 1 a. In this modelling perspective, memristors represent the circuital counterpart of nonlinear gel/sol interactions, playing the role of memory elements, while the resistance of memristors encodes the behaviour of low-viscosity channels [ 25 ] . Models based on the memristor analogy successfully capture key elements of Physarum ’s emergent problem-solving capabilities (although they are not all-encompassing [ 11 , 30 ] , and neglect some physiological aspects of the cellular organism [ 31 , 32 ] in favour of a predominantly phenomenological representation of the shuttle-streaming process). In electronics, suitable parallel circuits were numerically shown to solve mazes [ 33 ] , and experiments employing circuit-level bio-inspired approaches to solve mazes effectively obtained consistent success [ 34 ] . Identifying the key elements that enable maze solving, and proving that the obtained solution is optimal, would thus cast light on the evolution of biological organisms and guide the development of bio-inspired technologies. In this work, we build upon the circuital models for P. polycephalum developed in [ 25 ] to prove that threshold effects in shuttle-streaming yield minimum-path formation in foraging and maze-solving. The nonlinear responses to chemical and pressure gradients allow optimal network formation, irrespective of the environment topology. We also show that the circuital dynamic model explains and qualitatively reproduces the final slime configurations observed empirically, where one or more branches effectively link food sources, after a transient period during which several probing branches are initially generated and subsequently dismissed, except for those directly connecting the food sources. Memory , learning and problem solving are thus global properties that emerge from nonlinear local decisions over a complex network. 1 The Model: Circuit Analogy and Asymptotically Stable Steady State Figure 1: Linking a maze to a model of P. polycephalum learning. a) Schematic representation of the circuit model for element k , with capacitance C_{k} and memristive element M_{k} , connecting nodes h and j . b) Example of maze with walls (grey) and with food sources (red circles) at its entry and exit. The maze is covered by a network of circuit elements, where white circles are associated with nodes and circuit elements with links. Encoding the maze topology means having open switches (hence, disconnection) in correspondence to a wall. In green, the shortest path connecting the two food sources; in blue, an alternative (longer) path. When it enters a maze with food sources, Physarum polycephalum tends to first explore most of the environment and connect food sources, and then refine its path by a backward flow of nutrients [ 3 , 4 ] . At each turn of the maze, Physarum has to decide which path to follow, by responding to environmental stimuli and pressure gradients as described above. Building on the memristive analogy, we can cover a given maze with a network of responsive elements (akin to those in [ 33 , 17 ] ), each corresponding to

Read the complete text