A Reappraisal of the Principle of Maximum Power

Abstract

Living organisms continuously capture and transform free energy to survive and grow. The Maximum Power Principle (MPP) states that life evolves to maximize power—the rate of energy acquisition and conversion into useful forms—within prevailing constraints. Constraints include trophic interactions and competition, and they determine the outcome of the MPP. Biosystems that accumulate free energy faster will prevail transiently, but those that do so in a stable way will dominate in the long run. Accumulation of free energy is often approximately measurable as biomass that is useful to improve adaptation, competition, and ecological stability. Biosystems that allocate energy to long-term stable biomass—such as forests investing in wood—dominate because they secure long-term resource capture, competitive advantage, and resilience. Species and ecosystems represent different scales at which the MPP applies. In the long run, the MPP must align across scales, because a species can achieve stable growth and maximize power (i.e., accumulate free energy/useful biomass) only if the ecosystem the species lives in also maximizes its power. If a species reduces the stability and power of its ecosystem, it undermines its own long-term power-acquisition capability. This feedback harmonizes the MPP across scales over evolutionary times. Humans have evolved peculiar traits that have made them able to remove this control loop, amplifying species-level power. This enabled us to grow into a large population supported by anthropic systems that have strongly reduced the biosphere’s stability and power, resulting in a scale conflict in the MPP. Outlined in this way, the MPP provides a useful framework for understanding evolution, ecosystem development, and anthropogenic impacts.

1. Life, Biosphere, Fluxes, and Evolution

1.1. Biosphere

All life forms we presently know live in the biosphere, the global ecological system that encompasses all living organisms and ecosystems on Earth, along with their environment. The latter includes land (the uppermost layer—actually, a few meters below ground—of the Earth’s crust, which is part of the lithosphere), water (the whole hydrosphere, consisting of oceans, rivers, lakes, glaciers, and groundwater; which reaches ~10 km of maximum depth in the oceans), and air (in practice, the troposphere, that is, the lowest layer of the atmosphere, extending from the surface up to about 8–14 km in altitude, depending on the latitude), which support life [1,2]. Despite its extreme thinness relative to the Earth’s radius (roughly, 0.12–0.37%), the biosphere is where life can exist. On land, most life is concentrated in an even thinner layer: from <1 m to about 100 m, mostly depending on the local flora.

The biosphere includes many different habitats (i.e., natural environments), each corresponding to a particular ecosystem [2]. An ecosystem is a community of living organisms—plants, animals, fungi, and microorganisms—interacting, through continuous exchanges of energy and matter, with each other (mostly, but not only, through trophic interactions) and with their physical environment (such as soil, water, and climate) as a functional unit [2]. These interactions involve the flow of energy and the cycling of nutrients, creating a dynamic system where biotic (living) and abiotic (non-living) components are closely interconnected. Over very long timescales, organisms have systematically altered Earth’s surface by increasing the weathering of rocks, and collectively extracting, retaining, and accumulating materials, thereby increasing global living biomass, which results in a net transfer of materials from Earth’s crust to the biosphere [3].

Ecosystems can vary in scale, from a small pond to an entire forest or ocean. Anthropic systems, in this context, are environments whose structure, processes, and dynamics are significantly shaped, modified, and maintained by human activities. They are typically managed to meet human needs. Ecosystems with similar environmental conditions and life communities often occur across large regions; together, these related ecosystems form a biome, that is, a large-scale ecological region characterized by broadly homogeneous climate, dominant vegetation, and associated animal life [2]. Examples include boreal forests, tropical rainforests, deserts, grasslands, and tundra.

Ecology studies ecosystems and their interactions, which all depend on the basic features of life. Life is a collective term for biological systems continuously exchanging energy and matter with their environment. They persist far from thermodynamic equilibrium by capturing free energy (i.e., energy that can be converted into work) from an external source and coupling it to transformative processes that dissipate some free energy (typically, dispersed as heat) [4,5,6], thereby enabling metabolism, growth, response to stimuli, reproduction, and evolution. In principle, a decrease in free energy can be entirely converted into mechanical work in a reversible process, whereas thermal energy can be only partially converted into work, even under the ideal efficiency of a Carnot engine—an option generally unavailable to natural biosystems, which rarely operate across exploitable thermal gradients. Thus, in ecology, understanding energy fluxes is fundamental to grasping ecosystem functions, trophic relations, and species interactions [2]. Living organisms also depend on the cycling of matter, including water, carbon, and nutrients, which flow through the interconnected biotic and abiotic compartments [2]. Life, therefore, is fundamentally supported and constrained by fluxes of energy and matter (Figure 1).

1.2. Basic Laws

Organisms (individual living entities), species, and ecosystems (complex networks of interacting organisms and their environment) are biosystems, because this broad term refers to any organized biological system that maintains structure and function through interactions of its components, and they can grow, expand, and change. In addition, they all are governed by the same fundamental physical laws—principally thermodynamics—and chemical laws that regulate matter and energy transformations. As physic-chemical and ecological dynamics are predominantly stochastic at small scales, discussion of theoretical ecological outcomes—which are always probabilistic phenomena—focuses on the species and ecosystem’s levels, which are those meant as biosystems hereafter.

Physical laws like those of thermodynamics define what is possible [5,6]: energy cannot be created or destroyed (First Law), and all real processes increase overall entropy (Second Law). The latter can be interpreted as saying that spontaneous processes (in an isolated system) overwhelmingly tend toward states of thermodynamic equilibrium because those states are statistically the most probable [6]. These laws impose constraints, but within those constraints, countless states are physically allowed, though with varying probabilities [5]. According to the Second Law, life is highly improbable (in the universe, considered as an isolated system), because living organisms persist in a dynamic state far away from the thermodynamic equilibrium of an isolated system. In fact, living organisms are open systems that take in energy and matter (e.g., sunlight, food) to maintain an internal functional order (with much lower entropy than their environment), while increasing overall entropy. A net increase of overall entropy—required by the Second Law—thus drives life existence. Free energy is the energy that can be used to lower entropy locally (at the cost of a greater increase in entropy in the environment), and it is therefore required and consumed by organisms to live [6]. In this way, biosystems exploit existing disequilibria—like the flux of high-energy radiation from the Sun—to obtain work for carrying out their vital functions [6]. Free energy is convertible between different forms of available energy [6]: from sunlight to the formation of chemical gradients (e.g., a proton gradient across the plastidial, and mitochondrial, membrane, which drives ATP synthesis, thereby converting electrochemical potential into biochemical energy), as produced through photosynthesis; the chemical energy of organic matter (including energy-token molecules like ATP), mostly accumulated as biomass; the potential energy associated with the dislocation of masses in the gravitational field, like the hydraulic head (i.e., the difference in elevation) of water systems (like water raised to the highest leaves of trees, or stored in high basins), which can be converted into mechanical and then electrical energy; both biochemical and electrical energies can be converted into mechanical work (by means of electric muscles and motors, respectively); and so on. At the biosphere scale, life can be considered as a metastable condition associated with a reduction in entropy on the Earth, compensated for by a greater increase in the entropy of the Universe, as living organisms transform part of sunlight into heat irradiating to the space [6].

The laws of thermodynamics do not, however, explain how living systems (biosystems) can exhibit highly organized, dynamic structures that persist far from thermodynamic equilibrium [4,5]. Biosystems indeed incessantly search to maximize their self-maintenance and adaptability. This is achieved by establishing robust self-maintaining configurations, in a continuous struggle for a stable non-equilibrium state [7]. This is crucial because, while life obeys thermodynamic laws, its existence depends on actively exploiting energy sources to maintain metastable order and functions within the limits imposed by the Second Law. What is missing, therefore, is a principle that accounts for the fact that life does not simply exist by chance under rare, favorable conditions, but it can expand, enduring mutable and diverse environmental conditions (within limits that allow its survival), by changing and adapting in a variety of forms. In other terms, it is assumed that a principle must exist to explain how living organisms consistently evolve toward patterns—such as complex networks, feedback loops, redundancies, and energy-harvesting strategies—that sustain them far from thermodynamic equilibrium even though the surrounding conditions vary.

As life is supported and constrained by fluxes of energy and matter, it seems obvious that biosystems that capture and utilize energy resources more effectively are also more likely to thrive and propagate. Although matter—i.e., specific nutrient molecules—are needed and can be scarce as well, availability of free energy is typically always limiting. Thus, even though suitable nutrient molecules can represent additional restraints, the main issue is about free energy and its fluxes. Only biosystems able to overcome these restraints can, therefore, exist and persist far from thermodynamic equilibrium, in the long run. Assuming living organisms originated by mere chance, they would have had only a short existence unless they—again, by chance—acquired the ability to adapt to fluxes whose conditions change over time and space.

It is worth noting that in natural ecosystems the only unavoidable waste is heat, dissipated into the environment as energy degrades according to the Second Law of thermodynamics [2]. Material outputs, by contrast, are largely recycled: respiration releases CO₂, reused by plants during photosynthesis; plants release oxygen, used for respiration; dead organisms are decomposed, producing humus (soil decomposed and stabilized organic matter [8]) and releasing nitrogen compounds and other nutrients; these minerals are mostly absorbed by plant roots and symbiotic fungi, closing nutrient cycles [2,8]. Humus is crucial for soil health because it improves fertility, structure, and water retention [2,8]. Like soil mineral particles, humus acts as a long-term nutrient reservoir, capturing soluble nutrients like nitrogen and phosphorus and slowly releasing them, while enhancing soil aggregation for better aeration and root growth [8]. Humus also increases water-holding capacity and buffers pH, creating a stable environment for plants and microorganisms. In short, humus supports nutrient cycling, soil water capacity, and overall ecosystem productivity [8].

1.3. Evolution

Evolution is the process by which populations of organisms change over generations through alterations in heritable traits, driven by mechanisms such as mutation, genetic recombination, natural selection, and genetic drift (i.e., random changes in allele frequencies; it is most relevant when populations are small, or allele frequencies are low) [9]. Natural selection weeds out individuals with maladaptive traits: although at this scale random effects largely prevail, individuals with traits that improve survival or reproduction are more likely to pass them to their offspring—a concept often summarized as “survival of the fittest” [2,9]. Evolution, thus, favors (with increasing degree of deterministic effect) individuals, populations, and species with better fitness, defined as an organism’s ability to survive and reproduce in a given environment. It is often measured as the number of viable offspring an individual contributes to the next generation relative to others. Although selection acts on the phenotype (which is the expression of the genotype under contingent environmental conditions), it is the genotype that is inherited. Individuals of the same species may carry different variants (i.e., alleles) of the genes that make up their genotype [9]. Higher fitness, therefore, means greater representation in future populations of favorable genetic factors (i.e., allele variants or, much less frequently, novel genes). Thus, natural selection hits individuals but its deterministic effect emerges at the level of population: the effects of selection are seen at the population level as changes in allele frequencies across generations, and, in the long run, they affect the success of the species and, thus, the size of its population. Over time, this leads to increased adaptation of species to their environments. Adaptation refers to the set of heritable traits that enhance fitness under specific ecological conditions. It is shaped by natural selection over generations. Individuals or species with better adaptations exploit resources more effectively than others. Natural selection, in this way, also shapes biodiversity: some species thrive and diversify, leading to the emergence of new species, while others decline and eventually become extinct [9]. These changes produce a great diversification of life.

When two or more species (or individuals) depend on the same limited resource, the one that uses it more effectively reduces its availability for others. This ecological interaction is called competition. In nature, competition is often indirect (exploitative), but direct competition—which involves anticipatory behaviors such as aggression or physically excluding others from resources—exists too. For example, molds and soil bacteria produce antibiotic compounds that kill competitors for nutrients; analogously, some plant species produce chemicals that inhibit growth of nearby plants (allelopathy); animals often aggressively defend their territories and form hierarchies to regulate access to mates and food. Beyond competition (resource-sharing conflicts), other fundamental interactions among living organisms include zoo-pollination, zoochory, and trophic relationships like herbivory and predation (and parasitism) [2]. In this complex web of interactions, the species that better adapt to a habitat will prevail, outcompeting species that show lower fit.

However, species exist as functional components of ecosystems rather than in isolation. Thus, their adaptations often also respond to the selective pressures imposed by other species—such as in the relationships between predators and prey, plants and pollinators, or hosts and parasites [2]. This interconnectedness means that evolutionary change in one species can cascade through the ecosystem, shaping the dynamics and stability of the whole system, as well as its species composition. Indeed, genetic variation in a single species can influence entire ecosystems [10]. Evolution is, thus, almost always a matter of co-evolution—the process by which species reciprocally influence each other’s evolution through their interactions [2,9].

Evolutionary change generally occurs over long timescales [9], but its pace varies widely, and this timeframe is crucial when assessing evolutionary success. Microevolutionary shifts within populations can happen in just a few generations, whilst the formation of new species typically takes tens of thousands to millions of years. Major morphological or functional innovations, such as the evolution of new phyla, often require tens of millions of years. These processes are not always gradual: long periods of stasis are often followed by short bursts of rapid change, typically linked to environmental upheavals. The products of evolution—new species and novel functions—are tested for fitness over extended periods, and only time reveals whether they persist or disappear.

2. The Maximum Power Principle

2.1. Definitions

Based on considerations similar to those made in the previous section—and over a century of conceptual refinement that started from physicist Ludwig Boltzmann, was furthered by mathematician, biophysicist, and theoretical ecologist Alfred J. Lotka, and was warmly promoted by ecologist Howard T. Odum—the idea of the Maximum Power Principle (MPP) was developed [11] to account for the successful adaptive evolution of biosystems (i.e., for the existence, expansion, adaptability, and resilience of life). Although the MPP is a debated and challenging concept [12,13,14,15], its core idea is powerful and appealing: biosystems tend to evolve to maximize power, that is, to maximize the rate at which they take up free energy available from their environment (e.g., solar radiation, chemical gradients, food resources) and use it for processes that sustain the system (namely, maintenance, growth, and reproduction) under prevailing constraints. As outlined by H.T. Odum, biosystem designs develop and prevail that maximize power intake and transform energy into forms useful to the biosystem by reinforcing production and efficiency [16]. In this context, efficiency is the proportion of energy converted into useful biological work (maintenance, growth, reproduction) relative to the total free energy available to a given biosystem (e.g., sunlight for photoautotrophs). For maximization of efficiency, energy dissipation and, therefore, overall entropy increase, should be minimized.

Biosystems are constrained by energy (and mineral nutrients) sources, efficiency, and time. Time (quickness of development and growth) is where the MPP comes into play [17]. In its most simplistic form, the MPP can be viewed as stating that, given an available source of free energy that can be used by alternative biosystems for their own growth (which, for a given species, includes both growth/development of individuals and their population), the biosystem that, all other conditions the same, is able to use this energy at the highest rate will grow more than the others because it grows faster. This can be exemplified in terms of different microorganisms growing on the nutritious substrate of a Petri dish: the one growing the fastest—or the one arriving first, in the case of microbial contaminants—will be advantaged. In this sense, the principle is a truism because it is implicit in its definition.

As every environment/niche is spatially limited; however, it is not obvious what happens when all the available space—either physical or ecological (that is, in terms of niche)—has been occupied: the fast-growing biosystem can expand further only if it competes (interferes, or even fights) with the alternative biosystems in a successful way. If so, it will gradually expand, displacing its competitors, up to occupying the whole environment/niche. However, two assumptions are required for this to happen: 1. Competition (for resources, especially free energy), either direct or indirect, occurs, and 2. the fastest-growing biosystem is also able to outcompete all the alternative biosystems [11]. The former assumption is obvious since free energy and other resources—which can be contextually expressed in terms of the free energy associated with their accumulation (a variable that H.T. Odum would rather estimate as the energy cost needed to achieve them in a specific ecological background, and called ‘emergy’ [16])—are always finite. However, the latter assumption would greatly reduce our ability to consistently check whether the MPP applies, because assessing it ex ante (that is, predicting and comparing all possible alternative biosystems before they have actualized) is often not possible, and, moreover, such assumption is usually wrong. Furthermore, the necessity for competition introduces a pressing time constraint.

Nevertheless, one can define growth as implying competition and trophic interactions. This is, indeed, one of the meanings of “under prevailing constraints”. This makes sense, for example, if we consider and compare mature biosystems that have already occupied the available space and have evolved through competition with alternative biosystems [18]. They have been successful because they have overgrown alternative biosystems (at least, those eventually present; a provision that highlights the large role of randomness) while competing with them. Thus, a less simplistic and more realistic definition of the MPP is: the biosystem that uses the available free energy at the highest rate while coping with trophic interactions and outcompeting alternative biosystems will dominate. In the absence of trophic interactions and competition, the ‘while’ clause is voided, and the fastest growing biosystem prevails, until competition becomes relevant. More in general, the MPP applies conditionally to all the interactions among species and between them and the environment.

The MPP applies to both species and ecosystems [11], albeit at different scales. In this regard, H.T. Odum emphasized selection among alternative integrated designs (e.g., ecosystems and anthropic systems), rather than only among the individual elements (e.g., populations, species, and economic productive sectors) within alternative designs, because the performance of a system depends on how its parts are connected and interact through feedback loops [16]. For species, the MPP influences traits and behaviors that maximize energy capture and use for survival, growth, and reproduction within environmental constraints (which generally means in their ecological niche). For ecosystems, it shapes structures and processes that optimize energy throughput and power generation—the rate of energy transformation—across the entire system. This distinction will be examined in more detail later; for now, general considerations on the MPP will be applied to species and ecosystems alike, treating them as generic biosystems.

2.2. Efficiency

The MPP simply articulates a basic feature of life: selective persistence of biosystems that maximize power intake. This maximization, however, reduces efficiency in favor of speed. It is thus important to understand the trade-off between power and efficiency: power measures the rate of energy conversion, whereas efficiency measures the fraction of input energy converted into useful work. Maximizing one does not maximize the other: reversible processes maximize efficiency but are infinitely slow [6]; thus, power output is essentially zero. To increase power, systems must allow faster energy flow, which introduces irreversibility and lowers efficiency (as some free energy is dissipated to speed up processes). Nonetheless, to maximize conversion of energy provided by a given source into useful forms, the biosystem must keep efficiency as high as possible, otherwise the conversion efficiency drops and so does power. If efficiency drops too low, indeed, most energy input is wasted as heat, reducing the useful output and thus lowering power despite high throughput. Hence, even though the objective is to maximize power, efficiency must be kept as high as possible to support this aim. As remarked by Odum and Pinkerton [17], biosystems tend to operate at the efficiency that maximizes their power, which is always less than the maximum possible efficiency. Thus, maximum power is usually thought to occur at intermediate efficiencies (Figure 2), quite below the maximum possible efficiency. There is not a universal relationship between efficiency and power. The actual shape of the curve of the efficiency achievable while maximizing power, therefore, depends on the specific features of each biosystem.

Hall and McWhirter [11] highlighted that including the clause “compatible with the constraints” (which is equivalent to the provisions “under prevailing constraints” or “conditionally to all the interactions among species and between them and the environment” used here) in the definition of the MPP, results in very different outcomes under different growth conditions, as A.J. Lotka acknowledged more than a century ago. Specifically, on the one hand, when growth is unrestrained, because supply of free energy (and material resources) is abundant and competition is low, power (roughly, biomass accumulation) is maximized by high rates of growth—that is, fast consumption of resources—and, therefore, low efficiency (as a greater entropy increase, which is equal to faster dissipation of free energy, is the cost of increasing the speed of a process). On the other hand, when growth (or, more exactly, when energy acquisition and/or use) is severely restrained, because supply is low or competition is strong, high efficiency is required to maximize power [14]. This distinction shows that maximization of the rate of energy dissipation/entropy production is not what directs the evolutionary process [19]. Indeed, it is neither the energy captured nor the energy dissipated (≈entropy increase) that is important [11,17], rather it is the free energy that the biosystem accumulates.

The existence and persistence of forests, highly structured biosystems that accumulate and store free energy, demonstrates that evolution is not driven by the maximization of entropy production; if it were, such free-energy-preserving systems—and life in general (which maintains low entropy locally compared to bare ground)—would not exist. Natural selection favors those structures (organisms) that are better at maintaining themselves and processing free energy, dissipating as little as possible while growing, competing, and evolving.

2.3. A Promising Thermodynamic Framework

Thermodynamics is about energy and its changes, and classical thermodynamics aims to compare the equilibrium states of systems, and therefore it ought to be more properly described as ‘thermostatics’. The MPP, instead, abstracts an underlying regularity observed in biosystems: it is an inherent feature of anything striving to exist.

The fact that this regularity regards how biosystems use the free energy that is available to them, is, of course, constrained by the thermodynamic properties of energy and how they affect energy changes, but the regularity is nonetheless a feature of biosystems, not of energy. Trying to apply classical thermodynamics to the MPP in a simple, reductionist manner would be conceptually unsound.

Nonetheless, biosystems are thermodynamic systems and, therefore, any emergent property exhibited by these systems should, in principle, be traceable to their underlying physical properties. This, however, would require a theory much more extended than classical thermodynamics. Although a fully developed theory of this kind is not yet available, a conceptual framework has been developed pointing in this direction.

As the MPP is about the rate at which free energy is used, it can be seen more as a matter of kinetics than of thermodynamics: the former is indeed about how fast a system changes, whereas the latter is about determining whether a process is theoretically favored and what the final (equilibrium) state looks like. We might view this in terms of theoretical probability of alternative states (thermodynamic equilibrium) versus actual occurrence (kinetic restraint): the transformation of diamond into graphite is thermodynamically favorable, but so kinetically slow—because of a high activation energy, which provides a strong restraint (a barrier, in practice)—that it is effectively negligible, allowing us to safely assume that a diamond is forever. Timescale matters.

Analogously to diamonds, biosystems are thermodynamically unfavored but long-lasting; not because of a strong kinetic barrier, however, but because they continually use available free energy to maintain themselves in a metastable state that supports their existence, away from the theoretically much more probable state of thermodynamic equilibrium (i.e., some heat plus disaggregated and diffused material particles). We might say that they exploit a crafted complex of kinetic mechanisms that generate biological constraints and restraints to oppose thermodynamic equilibration without violating any thermodynamic law. They represent therefore a conceptually challenging kind of thermodynamic system.

Modern thermodynamics includes complementary frameworks to describe living systems. One of them, multiscale thermodynamics, aims at extending thermodynamic theory to time evolving processes in complex macroscopic systems that are far from equilibrium and persist due to external and internal drivers [20]. To this aim, multiscale thermodynamics explicitly assumes that these systems are best understood by comparing their properties over multiple scales (levels) of investigation—typically, the microscopic, mesoscopic, and macroscopic levels, but more hierarchical levels, or sublevels, are possible. Multiscale thermodynamics focuses on the relations among these different scales and how they constrain one another.

In this context, therefore, a level is a scale of description of a system, with a microscopic level being characterized by many variables, including detailed or fast (instantaneous) ones, whereas a macroscopic level has fewer variables, only slow or coarse-grained (averaged) ones [20]. Thus, most details observable at the microscopic level (which has a high-dimensional state space; that is, it is described by a huge number of variables) are lost in the reduction (of information) to a macroscopic level (which, therefore displays a reduced dimensional state space obtained by averaging spatial-temporal details of the microscopic level), but, at the same time, an emerging overall pattern is gained. In this way, ‘reduction’ of details leads to the ‘emergence’ of properties observed to a larger scale of description [20]. Thus, for equilibrium states, multiscale thermodynamics formalizes the derivation of the macroscopic state investigated by classical thermodynamics starting from the microscopic states described by statistical mechanics. The essence of the reduction is, therefore, a recognition of an overall pattern (of variables) in the microscopic portrait that characterizes the macroscopic portrait.

As mentioned, this theory aims to describe macroscopic systems that do not reach the equilibrium state because of external (or internal) drivers. It proposes that the behavior of externally (or internally) driven macroscopic systems can often be described at a mesoscopic level in terms of mesoscopic time evolution [20]. The mesoscopic level is where transient (but not instantaneous), or small, deviations from equilibrium (e.g., domain formation, clustering, deterministic fluctuations) are observed.

To describe time evolution of non-equilibrium systems, multiscale thermodynamics uses the Onsager variational principle (OVP), a foundational concept in nonequilibrium thermodynamics that provides a systematic framework to describe time development in systems wherein some externally maintained thermodynamic driver keeps the system out of equilibrium. As it addresses small deviations from steady or equilibrium state, the driver effect is modeled at the mesoscopic level [20]. In modern thermodynamics, generalizations of the OVP typically state that if a system is driven slightly away from a steady or equilibrium state by an external driver (such as a gradient that supplies a flux of thermodynamic potential), its time evolution is governed by the balance between the rate of change in the thermodynamic potential (e.g., free energy) plus a dissipation (or irreversibility-cost) functional that quantifies the associated energy dissipation or entropy production [15,20]. Accordingly, for a system in a quasi-steady state subject to an external driver, the OVP balance equation implies that the rate of external input of the thermodynamic potential equals the sum of the rate of change in internal free energy and the rate of irreversible losses, assuming no mechanical work is done [15,21]. Then, the internal free energy may be constant or even increasing locally, while entropy production and dissipation remain positive. Consequently, in a system driven by a constant external power input, any increase in the rate of internal free-energy accumulation must be accompanied by a corresponding decrease in irreversible dissipation. In particular, in a biosystem, a higher rate of free-energy storage in biomass (plus any work done by the system) corresponds to lower dissipation.

At present, multiscale thermodynamics is applied primarily to areas such as fluid mechanics, which remain tractable within current theoretical understanding. Nevertheless, it can also be viewed as a conceptual framework capable of addressing emergent properties in thermodynamic systems of increasing complexity, potentially extending even to biosystems. When multiscale thermodynamics are fully developed, the MPP might achieve a thermodynamic formalization as a spontaneous process.

In principle, indeed, the MPP may be described in multiscale thermodynamics at the different scales: microscopic dynamics generate many detailed pathways of free-energy use, dissipation, and entropy production, but coarse-graining these into a mesoscopic description yields effective quantities such as the net acquisition of power from an external driver—an inherently mesoscopic feature. When coarse-grained further into macroscopic descriptions, first at the level of organisms and then at an even coarser ecological scale, these mesoscopic fluxes become the foundations of large-scale organization. It is at these macroscopic levels—organismal and ecosystem—that the MPP emerges: among all mesoscopically allowed regimes consistent with microscopic dynamics and boundary conditions, in a characteristic timescale (much longer than the mesoscopic one), the macroscopically stable states are those that sustain the highest acquisition of available energy flow under the system’s structural and kinetic constraints. Thus, while external driving enters at the mesoscopic level, the selection toward maximal usable energy acquisition appears only at higher macroscopic scales, including ecosystems, as a natural consequence of hierarchical coarse-graining and emergent stability.

Interestingly, in multiscale thermodynamics, the application of the MPP may transiently diverge at species and ecosystem scales because each level operates under its own effective constraints and therefore on characteristic timescales that differ among levels. Organisms adjust their energy acquisition strategies over short physiological and demographic timescales, maximizing usable power intake to sustain survival, growth, and reproduction, whereas ecosystems regulate fluxes, community structure, and trophic pathways over much longer ecological and successional timescales. As a result, the coarse-grained variables, constraints, and dynamical equilibria that emerge at the ecosystem level can differ from those governing individual species. Thus, even though both levels follow the MPP, each maximizes usable energy acquisition under its own scale specific constraints and over its own characteristic temporal horizon [22]. Along their own timescale, therefore, a species might pursue maximum-power goals that diverge from those of the ecosystem. However, along the ecosystem’s time horizon, a re-alignment must take place, because properties emerging at a coarser-grained level eventually supersede those displayed at a lower, finer-grained level in the hierarchy.

2.4. The Maximum Entropy Production Principle (MEPP)

Changes in free energy can involve changes in both enthalpy/internal-energy (as in Gibbs or Helmholtz free energies) and entropy. However, enthalpy-driven changes are often constrained—for example by high activation energies or structural barriers—so, in many situations where a physical disequilibrium occurs, the most immediate spontaneous tendency is the increase in entropy. In contexts where constraints are weak or absent, this gives rise to the Maximum Entropy Production Principle (MEPP), which states that a system will tend to change so to increase entropy at the highest possible rate allowed by the externally imposed constraints. For systems close to equilibrium with linear response coefficients, irreversible fluxes are proportional to thermodynamic forces; this typically describes simple isolated or closed systems such as those displaying physical gradient disequilibria. Even in far-from-equilibrium open systems with many degrees of freedom and weak constraints, the most probable macroscopic time evolution may be approximated as one that dissipates gradients/disequilibria fastest, i.e., maximizes entropy increase. Thus, the MEPP can appear as a special case in the time evolution of non-equilibrium systems, especially if governed by diffusion.

According to the MEPP, a non-equilibrium system should settle into the state with maximum entropy production. Thus, for a system in a stationary or quasi-steady state subject to an external driver, the MEPP requires that, in the OVP balance equation, the rate of irreversible dissipation be maximized (that is, all externally supplied free energy is dissipated as soon as possible). Consequently, the rate of internal free-energy accumulation must be negligible.

If the MEPP applied to living systems, they would be expected to maximize quick dissipation of free energy rather than invest free energy in building long-lived structures that reduce dissipation, and their biomass should thus be zero. In other words, life would never exist. Hence, the MEPP does not apply to whole biosystems.

At the planetary scale, the coupled atmosphere–biosphere system appears to follow the MEPP because of the dominant diffusion-like irreversible processes operating in the atmosphere [23]. The mass, heat capacity, and energetic throughput of the biosphere, especially those associated with biomass, are very small compared with those of the atmosphere. Consequently, most of Earth’s entropy production arises from atmospheric irreversible processes—such as atmospheric circulation, large-scale heat transport, surface–atmosphere exchanges (particularly over oceans), and turbulent mixing—which collectively tend to maximize the rate at which entropy is produced [23]. They determine the magnitude of planetary entropy production, which is then exported to space through Earth’s radiative balance, in which low-entropy, shortwave solar radiation enters the system and is re-emitted as high-entropy, longwave infrared radiation [23].

Nevertheless, living organisms survive and maintain free energy by coupling their metabolism to irreversible processes. In doing so, they always produce more entropy than they accumulate: even plants, while storing free energy in biomass, produce more entropy than they reduce internally. Like all irreversible processes, biosystems dissipate free energy and release entropy—primarily as heat—to the atmosphere. Moreover, the biosphere further slightly increases planetary entropy production by intensifying biogeochemical cycling, modifying surface roughness, lowering surface albedo, and increasing atmospheric water vapor through transpiration [23]. On the one hand, these biologically mediated fluxes enhance—albeit of a small amount—the overall entropy production of the Earth system. On the other hand, by using sunlight to create and maintain high-energy chemical disequilibria, photosynthesis increases Earth’s free energy not only by adding biomass (and fossil fuels) but also by generating and sustaining a large gaseous disequilibrium in the Earth’s atmosphere-ocean system [24]. Whereas, as mentioned above, the entropy released to the atmosphere is ultimately radiated to space, completing the planetary entropy-export pathway [23], life-generated disequilibria remain on the planet.

It might also be worth noticing that in many ecological applications of the maximum entropy principle (≠MEPP), the “entropy” being maximized is Shannon information entropy rather than thermodynamic entropy. Although its mathematical form mirrors the Boltzmann entropy of statistical mechanics, the interpretation of Shannon entropy in ecology is distributional rather than thermodynamic: no temperature, heat, dissipation, or particles are involved. In this context, Shannon entropy can be used in ecology both as a descriptive index (the Shannon–Weaver index serves as a measure of biodiversity, capturing both species richness and the evenness of species abundances within a community [2]) and as the functional to maximize for predictive models [25].

As for the latter, according to the principle of maximum entropy first articulated by E. T. Jaynes [26], to infer an unknown probability distribution from only partial information (e.g., known averages or totals), one should maximize Shannon information entropy subject to those constraints [26]. The resulting distribution (MaxEnt) is maximally non-committal with respect to missing information and thus represents the least biased inference consistent with the known constraints (e.g., total abundance and species richness), even when predicting probability distributions in ecological contexts [25,26]. In other words, the unknown actual distribution of species is one of many possible realizations consistent with the constraints, and the MaxEnt distribution is the ‘center’ of this hypothetical ensemble of distributions. In ecology, MaxEnt—a probabilistic prediction under uncertainty—is thus used for predicting patterns like species-abundance distributions, species–area relationships, and community composition [25].

Thus, when ecologists speak of “maximizing entropy” they usually refer to a modeling assumption: given partial information, the MaxEnt distribution provides the least biased prediction of the species abundance distribution consistent with the known constraints. They are not implying that ecosystems obey a physical entropy-maximization law [25].

2.5. Usefulness and Biomass

In the context of the MPP, as mentioned above, efficiency refers to the ratio of useful energy output to energy input. The energy input is sometimes relatively easy to quantify; for example, the amount of solar energy that is available to the biosystem. The energy output is more difficult to define, as, in this context, the terms ‘useful’, ‘effective’, and alike, seem to have a teleonomic meaning, where teleonomy is the apparent purpose or goal-directedness in biological systems. However, these terms should be more properly intended as teleomatic, that is, ‘end-resulting’, rather than ‘goal-seeking’ (teleologic) [18], because, in nature, the MPP is not implemented by purpose. Like all biological functions, the MPP emerges teleomatically because replicators that persist longer and replicate more effectively are favored and, therefore, selected over time [4,27]. Apart from philosophical concerns, the problem regarding the MPP is that the ‘useful’ energy output—that is, the energy used by a biosystem for its own growth, maintenance, and reproduction—is an elusive concept.

Nonetheless, if we consider that a biosystem is successful if it survives, grows and reproduces, a trivial indicator of the success of these activities is the biomass of the biosystem (which takes into account both the number and size of individuals): the biosystem that uses the available energy at the highest rate while outcompeting the alternative biosystems will accrue the largest biomass in its environment/niche with respect to its competitors, because it grows more. This must always be thought of with the clause “under prevailing constraints”. Thus, to evaluate success in terms of the MPP, we cannot compare the overall biomass produced by an endemic species in a very specialized ecological niche with the biomass achieved by, for example, the Antarctic krill, which, thriving over vast ocean cold waters, is one of the most abundant species on Earth [28]. Rather, we ought to compare a species that lives in a specialized ecological niche with alternative species that can thrive and reproduce in that very same specialized ecological niche. This, of course, applies to ecosystems too: it would make no sense to compare, in the context of MPP, the biomass of the hydrothermal vent community on the deep ocean floor with that of the Amazon tropical rainforest. We should only compare alternative biosystems.

Although the output, or throughput, of ‘useful’ energy was the focus of the model proposed by H.T. Odum, the correct quantification of this parameter represents a difficulty for the application of the MPP to real systems. Thus, Figure 3 diverges from the original model in that it assumes that the ‘useful’ energy output is the one that produces biosystem’s biomass, whereas the useful energy flows for maintenance, growth and reproduction that do not lead to biomass accumulation, are considered ‘costs’, as they are energy expenditures the biosystem must sustain to thrive and propagate. Hence, they conflate with the portion of useful source energy lost and metabolic losses in determining biosystem’s efficiency. This interpretation is consistent with the original view of A.J. Lotka, who stressed that every biosystem, in order to maintain the steady state necessary for its survival, must be able to capture available (free) energy [29], and a result of doing so is increasing the biomass of the system [30]. That natural selection acts to maximize the amounts of energy and matter gathered per unit time—particularly as long-term emergent directional tendency of whole ecosystems, but even for individual species within the constraints of their specific niches—is indeed a relatively common (though not universal) opinion [31,32].

According to the MPP, evolution tends to favor strategies that optimize long-term energy capture and use [16]. In the form the MPP is outlined here, this means accumulating structures (biomass) and functional complexity that allow access to more resources, free energy in particular. This positive feedback is a key aspect for the maximization of power. In accordance, H.T. Odum saw the maximum power principle as an autocatalytic feature of self-organizing systems [16]. Correspondingly, life could be seen as an adaptable autocatalytic process [33,34].

As the MPP posits that the rate of the energy flux is the key (limiting) aspect, ‘biomass’ accrual should be properly considered in terms of the chemical energy of the mass that is acquired by the biosystem. In this way, power acquisition may be approximately assessed in terms of accumulated biomass. Although a simplification, this approximation makes the assessment and quantification of power much more treatable in terms of the rate of biomass accumulation.

Thus, although, in the context of the MPP, energy output is generally meant in energy flow terms (e.g., watts, joules per second), in ecological applications, biomass production is often used as a proxy for the energy output (that is, available energy input × captured fraction × conversion efficiency), because it represents the net chemical energy stored by photoautotrophs (via photosynthesis) or heterotrophs (via assimilation).

In this view, biomass accumulation (at the scale of either ecosystems or individual organisms) can be used as a proxy for energy output because it reflects the net storage of chemical energy in organic matter. This approach relies on the assumption that the chemical energy content per unit of dry biomass is approximately constant, typically around 16–20 kJ per gram for dry plant tissues [35], with higher values (up to double) for lipid-rich materials, which have a higher energy density. This assumption holds reasonably well in systems where tissue composition is relatively homogeneous, or its composition is averaged over many species, such as in plant communities. For ecological modeling, 2 kcal/g live (wet) weight (≈25 kJ/g for dry biomass) is a very rough approximation for biomass in general (that is, including both plants and animals), as most living organisms are two thirds or more water and minerals [2].

However, the assumption can fail under conditions where the chemical composition of biomass varies substantially. It becomes unreliable when composition diverges markedly (e.g., lipid-rich seeds vs. lignified tissues), when mineral ballast inflates mass with a diverse relationship with chemical energy (e.g., calcifying algae, silicon-rich shoots), or when large biomass pools accumulate mainly due to suppressed decomposition (e.g., peat under anoxia), which can overstate current power while reflecting past conditions (as the free energy of organic matter changes depending on whether the environment is oxidative—e.g., rich of oxygen—or anoxic and reductive). Mixed-trophic systems introduce further variability because animal tissues generally contain more lipids than plant tissues. In these instances, the actual chemical energy content must be compared.

Therefore, while biomass is a convenient indicator of energy storage, its accuracy depends on the degree of chemical homogeneity—or the effectiveness of averaging—within the system and the temporal and spatial scales considered. For precise quantification of energy flow, direct methods such as bomb calorimetry or detailed biochemical analysis are necessary.

Nonetheless, biomass can serve as a practical surrogate for energy output when comparisons are made across large differences between biological systems, because relative differences in total biomass often outweigh variations in energy density. For example, when comparing large differences between biological systems—such as forests versus grasslands—the magnitude of biomass differences often outweighs variations in energy density, making biomass a useful, though imperfect, indicator of energy-related patterns. For precise assessments of energy flow and power, direct measurements of productivity and energy content remain essential, however. As a practical matter, biomass is especially informative when differences in standing stock are large enough to dwarf modest variation in energy density, or when the stored structures clearly enhance future capture and processing (e.g., canopies, rhizospheres, reef frameworks).

It must be stressed again that the MPP does not favor stabilized biosystems that maximize biomass per se. If two biosystems have the same energy capture rate but one retains more energy in form of useful biomass rather than dissipating it, that system has indeed processed more energy overall and allocated it to building useful structures. This reflects higher realized power over time, provided that added biomass contributes to persistence, additional growth, and/or future energy capture. That is, the biomass must be really ‘useful’ to the biosystem.

For example, if two biosystems in the same environment have the same instantaneous energy capture ability, the one that retains more of that energy as biomass achieves higher effective power because a greater fraction of captured energy is channeled into useful work—maintenance, growth, reproduction—rather than being dissipated. This confers an evolutionary advantage when it supports survival or improves future resource acquisition: forests that invest in wood, canopy, roots, and mycorrhizal networks stabilize resource cycling and expand future capture capacity. In this regard, it must be evidenced that recycling material resources is essential to the application of MPP, because they can become limiting [2,8,18]—which, in the long run, is always the case.

2.6. Stability Favors Biomass Accumulation and, Thus, Promotes the MPP

Early life exploited geologically generated long-lasting chemical disequilibria at the rock–water interface, such as those in hydrothermal vent systems [36]. It was therefore highly localized, with very low productivity and limited biomass. The advent of photosynthesis expanded life, likely first in shallow marine or hydrothermal light-penetrated environments (anoxygenic photosynthesis), and later through the oceans, where cyanobacteria developed oxygenic photosynthesis, which very gradually transformed the atmosphere [37,38]. Thus, photosynthesis sustained the accumulation of a much larger biomass [38]. Higher plants massively increased the biosphere’s stored free energy relative to bare ground. Throughout evolution, biomass accrual was promoted following the MPP, while the entropy concomitantly produced was exported to space.

The MPP is expected to apply wherever there are life (or other teleonomic systems) or conditions that are conducive to life. Thus, it should apply on other planets like it applies on the Earth. Differences can arise because of the source of free energy. For example, on a planet where free energy should come from chemical disequilibria—like what happened on the Earth before photosynthesis—the overall free energy available in the biosphere would decline (since any actual process has an efficiency lower than 1). On the other hand, where the main source of free energy is external to the biosphere (like is the Sun), the overall free energy of the biosphere should tendentially increase over time (apart from supervenient processes, like an asteroid impact, an ice age, or overexploitation by an improvident species that has escaped ecological control).

As the MPP operates over the long term (since it requires adjustments, modifications, and competition of species and ecosystems), it does not favor high instantaneous (or daily) biomass growth rates, but rather the stable accumulation of biomass under prevailing, long-term conditions, that is, over decades or centuries, and ultimately across evolutionary timescales, that is, hundreds of thousands of years. This is immediately apparent when considering that the MPP does not imply universal selection for the fastest growers in mature ecosystems. If that were the case, prokaryotes—with high respiration intensity, fast energy flow, rapid nutrient cycling, and high instantaneous biomass growth (when resources are available), but modest stable biomass accumulation—would dominate all niches, and complex multicellular life would never have evolved. Instead, mature ecosystems are characterized by slow-growing, large-bodied organisms (e.g., trees, large mammals), which often have lower metabolic rates but higher biomass and longer lifespans. This reconciles why evolution did not remain with prokaryotes alone: complex organisms and structured ecosystems can, by virtue of their accumulated biomass and organization, secure and process more energy over the long term, even if their instantaneous growth rates are lower per biomass unit. Thus, even though energy flow provides a better basis than numbers or biomass for comparing ecosystems and populations with one another when studying food chains and webs [2], the accumulation of free energy and nutritional resources as useful biomass is more suitable to study the MPP, which regards more the understanding of evolutionary success than the functioning of trophic chains, two related but distinct concepts.

In some contexts, the role of (the total energy stored in) biomass as a practical surrogate to the rate of energy transformation and use is quite obvious: in early successional stages or fast-growing systems, biomass accumulation indeed correlates strongly with energy capture [14]. However, in mature ecosystems, the correlation between the rate of biomass accumulation and energy flow rates disappears. For example, an old-growth forest may store a lot of energy accumulated in biomass but have a very low growth rate of overall biomass. Yet an old-growth forest has a structural complexity that allows it to capture and store energy over centuries, stabilize nutrient cycles, and maintain resilience.

In this respect, it should be noticed that, thanks to the presence of massive trees, deadwood, and undisturbed soil organic matter (with fungi, arthropods and many species of other taxa thriving on them), intact old-growth forests typically store substantially more accumulated biomass, carbon and nutrients, than forests that have been razed by logging—even for centuries after disturbance [39,40,41]. Old, undisturbed forests accumulate carbon over centuries, with an estimated upper limit of 500–700 tons of carbon per hectare in the most favorable environments [40]. In contrast, managed systems harvested on short rotations cannot develop large woody structures and long-term carbon pools, as forests reach ecological equilibrium growth only after centuries. Moreover, when plantations and coppices are cut, much of the nutrient cycling is disrupted: nutrients stored in biomass are removed, and it can take centuries to naturally rebuild these stocks. Soil fertility declines as organic matter and minerals are depleted, while mycorrhizal symbioses and trophic networks are disturbed, further impairing nutrient cycling and destabilizing ecological interactions that supported ecosystem resilience [2,8,42,43]. These differences make managed systems less resilient and less capable of sustaining the complex, self-regulating processes typical of primeval forests.

Thus, for mature biosystems, maximization of power is achieved through better efficiency, as speed of growth (i.e., the rate of biomass accumulation) levels off to zero as the biosystem reaches a stable condition. Efficiency of resource acquisition for production indeed indicates local adaptiveness [14]. Mature ecosystems are the end of ecological successions because they maximize biomass accumulation up to the point at which all energy uptake is used to maintain the ecosystem [14].

In this view, when species and ecosystems come to dominate their niche/environment through maximum power, they further evolve to increase their power and, therefore, biomass, through higher efficiency and greater stability. That occurs because when all the energy available from a given source (classically the sun, for most ecosystems) is promptly used, species that achieve further power by increasing efficiency prevail. This leads to increased efficiency and biomass for each species in its niche, and, in consequence, for the whole ecosystem. In other terms, the power–efficiency curve shown in Figure 2 becomes more skewed to the left. Hence, the most powerful mature biosystem is the most efficient (among the available alternatives), stable and resilient, and, thus, in the long term, the one that achieves greater biomass accumulation. In other words, the biosystem that, in the long run, acquires the largest stable biomass demonstrates the greatest power (free energy acquisition capability).

3. The Maximum Power Principle Applies at Different Hierarchical Scales

3.1. Optimization Across Scales

The MPP at the ecosystem level is an emergent property, not just the sum of individual species maximizing power. Although each species tends to optimize its own energy capture and use for survival and reproduction, emergent properties arise in ecosystems because the collective interactions among species—through competition, cooperation, and resource cycling—create feedbacks that reorganize energy and matter flows in ways that maximize power at the ecosystem level [18]. This collective optimization results in the ecosystem maximizing power accumulation, and it reflects higher-order constraints and synergies that emerge from the networked structure of the ecosystem.

If the MPP is to apply meaningfully at both scales—species and ecosystems—maximizing power must confer a competitive advantage to individual organisms and simultaneously enhance the growth and persistence of the ecosystem as a whole. This alignment occurs because at the organism level, traits that improve energy capture and conversion (e.g., efficient photosynthesis, rapid nutrient uptake) increase fitness, while at the ecosystem level, cooperative and competitive interactions organize flows to stabilize the system and support its growth and expansion. This dual benefit is what allows MPP to operate hierarchically and consistently across scales [18].

In functional ecosystems, recycling both energy and mineral resources preserves resource availability and thus maximizes power. Ecosystem alternatives that self-organize to maximize recycling, therefore, are favored by the MPP. This is why, when exploitation of external energy sources (e.g., sunlight) reaches a plateau in mature ecosystems, power maximizes through increased efficiency—optimizing diversity and functional complexity, and diverting resources from reproduction toward large-scale integration [14].

As discussed, in mature ecosystems better efficiency can be achieved by specialized species, and, as soon as ecosystem development requires, or allows, more specialized ecological functions, new species arise that fulfill those functions. The more specialized a species is, the more strictly it is integrated in the ecosystem, since it could not survive if the other species would not provide their functions as well. According to H.T. Odum [16], during self-organization of systems, integrated designs are selected wherein multiple processes and feedback loops are connected in a way that reinforces energy capture and use efficiency through hierarchical levels (scales). This effect represents a fundamental reinforcement loop that harmonizes the MPP across scales.

For example, in a mature forest, trees, understory plants, filamentous fungi, insects and other arthropods, and microbes form a tightly coupled system wherein nutrient cycling and energy capture through photosynthesis are teleomatically coordinated to maximize overall buildup of free energy, far beyond what any single species could achieve alone. In contrast, an early successional ecosystem, such as a recently abandoned field, exhibits loose interactions, because species prevail through rapid colonization rather than integrated resource cycling. The MPP, indeed, operates differently across successional stages, reflecting a transition from short-term opportunistic strategies to system-level emergent optimization as complexity and connectivity increase [14]. Thus, the timeframe is highly relevant when assessing a strategy’s success, because short-term gains can differ from long-term outcomes, and some strategies only show their true impact over extended periods.

Indeed, ecosystems are networks of species. If each species tends to evolve according to the MPP, ecosystem-level dynamics should change accordingly: if all individual species maximize their stable biomass, the overall biomass of the ecosystem needs to increase as well. Yet, if an alternative ecosystem able to accumulate an even greater useful biomass would be present in the same environment, it is probable that it will gradually replace the former. The ecosystem that achieves greater power tends to expand because its species exploit resources more effectively, preempting light, nutrients, and space. These species gradually colonize and expand, while less efficient species decline. This dominance creates positive feedback, such as altering microhabitats to favor its own species, while reducing resource availability for the competing system. Over time, these interactions lead to a shift in community composition, with the high-power ecosystem stabilizing and the other declining, even under constant environmental conditions. For example, a mature forest maximizes energy capture and biomass accumulation more than a grassland in the same climate zone. As seeds are dispersed and shrubs and trees grow, their canopy shades out grassland species, reducing their photosynthetic capacity. Over centuries, therefore, forests tend to replace grasslands unless disturbances (fire, grazing) maintain the grassland, or climate and soil conditions prevent growth of trees.

It should be noted that the greater biomass of a forest compared to a grassland is primarily due to wood: this is the ‘useful’ biomass that allows the former to prevail over the latter. Although wood also represents an important nutrient reservoir (for the detritus chain) that helps stabilize and close ecosystem nutrient cycles, trees are not necessarily more physiologically or metabolically adapted to the environment than grass. Both intercept most of the available solar radiation. The superior fitness of trees lies in their competitiveness: being taller, they shade grasses and intercept sunlight first; they also escape grazing by most ground-living herbivores. In addition, their dead biomass supports the detritus chain that stabilizes the ecosystem and further increases its useful biomass.

This challenges the second assumption previously mentioned for the most simplistic definition of the MPP to work, namely, that the fastest-growing system (grassland, in this case; because growth of wooden tissues is energetically more costly and slower) can outcompete all alternatives. As discussed, when growth is defined as inherently involving competition, the MPP explains why trees ultimately replace grass in the long term. In the short term, grass dominates, but under favorable environmental conditions, trees eventually prevail. These temporal dynamics form the basis of ecological successions.

3.2. Biodiversity and Complexity

Ecosystems evolve toward a steady-state condition (mature ecosystems) characterized by high biomass, optimal species diversity, retention and efficient cycling of nutrients, and stability (i.e., resistance to disturbances and ability to quickly recover from damage) [2]. Accordingly, these parameters are often used as indicators of ecosystem success. Mature ecosystems, indeed, can exhibit high diversity in favorable environments (where the MPP fosters exploitation of every bit of available resources through proliferation of species fitted to any niche). However, biodiversity does not have a univocal role in characterizing the functioning of an ecosystem [44], and many mature ecosystems are dominated by a single species. For example, several forests across large natural areas worldwide exhibit strong monodominance (Figure 4A). Often, species richness is lower in ecosystems that display a uniform and stable characteristics with some limiting features (e.g., water shortage, sub-optimal temperatures or pH, nutrient-poor soil, short growing season, soil salinity, oligotrophic waters), which favors adaptive specialization by a dominant plant species [45]. Yet these ecosystems retain all the properties of stable biosystems. Thus, high species diversity is not a necessary property of ecosystem maturity; rather, it is a feature optimized by ecosystem development and species co-evolution.

Optimal biodiversity is promoted by the MPP: high biodiversity tends to increase ecosystem stability because species perform complementary functions, buffer fluctuations, and maintain nutrient cycling [2]. Partial redundancy and functional diversity make ecosystems more resistant and resilient to disturbances, as overlapping functions provide functional stability within a shared operational space, sustaining energy capture and processing through environmental fluctuations that can occur over time. Analogously, retention and efficient cycling of nutrients are necessary features of the MPP, because, as seen, low efficiency reduces power, at least in mature ecosystems.

Higher biodiversity generally increases structural complexity of ecosystems, but complexity also includes network architecture and functional roles, not just species count. Whereas complex systems tend to be less stable when they are randomly connected or excessively complex and poorly organized, complexity can enhance stability if structured ad hoc: features such as redundancy, feedback loops, and modular interconnectedness promote both stability and resilience [46]. Mature ecosystems, having been subjected to natural selection across evolutionary times, have necessarily acquired a teleomatic organization that supports their stability.

As competition and trophic interactions have an energy cost [2], and excessive, inadequately structured complexity reduces stability [46], it can be supposed that, over time, ecosystems tend to evolve toward an optimal level of biodiversity that, as seen, differs among biomes, rather than maximizing diversity indefinitely. Thus, though evolution is a never-ending process, ecosystems’ complexity and biodiversity have been optimized according to the MPP. Disrupting natural complexity, therefore, removes multiple feedback loops and destabilizes energy flows, pushing an ecosystem far away from its maximum power state.

3.3. Trophic Interactions and Food Chains

Of course, each heterotrophic species aims at using, and thereby reducing, the biomass of the species that precedes it in the food chain. Heterotrophic species aim to consume (the fundamental) part of the ecosystem. Yet, reducing the producer (plant) biomass would ultimately reduce even the biomass of a species that attempts to overexploit the ecosystem. A classic example is the case of herbivores: if their population grows up too much, there are no longer enough green plants to further sustain their population. In addition, in the longer term, an increase in carnivores, caused by the larger availability of food (herbivores), further eats away the herbivore population. As energy is passed from one trophic level to the next, biomass decreases due to energy dissipation [2]. Thus, the biomass of carnivores is typically much smaller than that of herbivores, and both are vastly smaller than that of producers [2]. Even if herbivore population increases transiently, the overall biomass is strongly reduced when producer biomass declines. And the increases in consumers populations are, in fact, transient, as, ultimately, the overall ecosystem biomass drop is severe. Hence, overexploitation of producers is eventually detrimental to consumers.

An important exception to the pattern of decreasing biomass through the food chain (which leads to the so-called pyramid of biomass) can occur in aquatic ecosystems due to a differences in turnover rates between trophic levels [2,28]: unicellular producers (e.g., phytoplankton) with short generation times can sustain a larger biomass of resident herbivores and predators, leading to an inverted biomass pyramid. Phytoplankton is advantaged in deep waters over higher plants because it can remain suspended in the euphotic zone, where light intensity is sufficient for net photosynthesis. In addition, ocean surface waters are often oligotrophic (nutrient-poor) because most nutrients are concentrated deep below the surface. Phytoplankton efficiently absorbs these diluted nutrients thanks to its extremely high surface-area to volume ratio. It also responds rapidly to upwellings that bring nutrients to the surface by reproducing in bursts. In contrast, in the open ocean, gales would impose severe mechanical stress on floating higher plants with leaves and deeply hanging roots (to access nutrients far below the surface), which would rub against each other and tear. Ocean phytomass is, thus, much smaller than terrestrial phytomass [28], even though oceans are more than twice the area of land. This highlights that, where microbial life dominates, biomass accumulation is lower, as prevailing conditions constrain the ecosystem to maximize power with the fittest available species. In this respect, an inverted biomass pyramid also highlights that energy flow is a better way to compare trophic levels across a food chain than biomass [2], though the latter is a suitable approximation to compare alternative biosystems in the MPP context, when biomass energy contents are comparable, and within alternatives compatible with the restraints and constraints of the habitat. Indeed, A.J. Lotka evidenced that, so long there is opportunity, biosystems enlarge their biomass, but where a limit is imposed onto this process, the available power may rather be used to increase the rate of turnover of the organic matter through the life cycle [30].

It is worth noting that in nutrient-poor surface ocean waters, autotrophic bacteria face a mismatch between abundant light energy and scarce mineral nutrients, especially nitrogen and phosphorus. Phytoplankton evolved higher cellular energy fluxes, that is, increased metabolic rates, to compete for and accumulate these scarce nutrients [47]. As a result, marine cyanobacteria adapted to persist at progressively lower nutrient concentrations, contributing to a long-term drawdown of nutrients in surface waters and an increase in total ecosystem biomass. However, under oligotrophic conditions, accelerating metabolism generates excess reducing power that cannot be fully used for biomass production because growth is limited by nutrients rather than energy. To maintain metabolic steady state, this excess must be dissipated, in part through the excretion of organic carbon compounds such as glycolate or polysaccharides. The released organic carbon fuels the growth of heterotrophic bacteria, which recycle mineral nutrients, limiting the loss of nutrients from the euphotic zone and enhancing overall ecosystem productivity [47]. The release of organic carbon thus generates positive feedbacks among interacting species that reinforce this favorable evolutionary trajectory at the ecosystem level, promoting metabolic interdependence and co-evolution among community members [47].

It is worth noting that, before the advent of photosynthetic cyanobacteria, the marine biosphere was relatively subdued compared with today [38]. This was likely because primary production was driven mainly by anoxygenic photosynthesis coupled with iron redox cycling, which was much less efficient than oxygenic water-splitting photosynthesis. Consequently, during the Great Oxidation Event (~2.4–2.3 Gy ago)—when oxygen produced by cyanobacterial photosynthesis accumulated in Earth’s atmosphere and permanently transformed planetary redox conditions from reducing to oxidizing—the biosphere likely increased its activity level by roughly an order of magnitude [38]. This huge increase probably resulted from the evolution of oxygen-producing cyanobacteria together with a dramatic rise in the availability of nutrients that could fuel oxygenic phototrophs [38]. This is a stark example of the reinforcement loops that harmonize the MPP across scales, boosting production and efficiency.

In ecological theory, bottom-up and top-down controls represent two fundamental mechanisms regulating energy flow in food chains [2]. Bottom-up control emphasizes how the availability of resources—such as energy and nutrients—determines the structure and productivity of primary producers and, consequently, higher trophic levels. In contrast, top-down control means that predation by higher trophic levels affects the accumulation of biomass at lower trophic levels: carnivores control densities of herbivores, which, in turn, control plant biomass. Notably, top-down control produces alternating (opposite) effects across trophic levels of the food chain. For example, removing wolves from an ecosystem causes deer population to explode and overgraze, thereby reducing vegetation (and, thus, causing soil erosion), in the so-called trophic cascade. These two mechanisms often operate simultaneously, particularly in stable or mature ecosystems, where feedback between resource availability and consumer pressure contributes to dynamic equilibrium. In the context of the MPP, this is important because the existence of herbivores, which consume plant biomass and reduce light capture, could appear to contradict the MPP at the ecosystem scale. However, the MPP applies first at the species level, and then at the scale of ecosystems, in both cases with the provision “under prevailing constraints”. Once evolution has led to the existence of herbivores, ecosystems will prevail that outcompete alternative biosystems. Ecosystems with better control of herbivores (more effective carnivorous feedback) should prevail. This combines the MPP at the species and ecosystem scales, and, thus, leads to the concomitance of bottom-up and top-down controls. For these reasons, the biomass of all animals—mammals, fish, insects, birds, and others—accounts for only 0.36% of Earth’s biomass, which is dominated by producers [28].

As remarked by H.T. Odum, when an ecosystem evolves, soon it develops feedback loops: with most of the externally supplied energy already in use, increases in useful power can be achieved only by reinforcing internal efficiency-producing mechanisms [16]. Indeed, the teleomatic transition to a more complex state can lead to a reduction in energy dissipation [19]. Furthermore, evolutionarily selected forms of cooperation between sub-systems represented the most important transitions in co-evolution and corresponded to an increase in both system’s complexity and its dynamic kinetic stability throughout evolution [19]. Ecosystems evolved and endured millions of years just because the complex interactions among individual species created feedback that stabilized them.

3.4. Carrying Capacity and Overshoot

Like the trophic cascade mentioned above, wherein a deer population shows a transient surge due to removal of top-down control, a locust swarm is a classic example of population overshoot (Figure 5) driven by abundant resources—often an unintended consequence of human removal of natural constraints. Fueled by favorable conditions, the locusts multiply explosively, stripping vegetation and converting it into swarm biomass at exceptional rates. But this unchecked growth triggers density-dependent feedback: as the food base collapses under relentless grazing, the carrying capacity (i.e., the maximum population size that an environment can support) plummets. What was once an environment rich in biological resources becomes a barren landscape, and the locust population, having overshot its ecological limits, crashes to a fraction of its former size. Large food availability and delays in negative feedback loops lead to resource depletion and trophic restraint eventually resetting population levels.

This is a consequence of support dependency, a general property of systems that will be discussed later. In the case of overshoot, support dependency occurs because all consumer species are energetically and functionally constrained by primary production. Thus, any species that reduces primary producer capacity below the level required to sustain ecosystem function necessarily reduces its own long-term viability.

It is worth noticing that trophic cascades take place because the top-down control exerted by predators keeps herbivore populations sharply below the carrying capacity that would be imposed by the bottom-up control [2]. This grants that, in mature ecosystems, producers are not severely pressured, and biomass accumulates at a higher level.

Transiently exceeding the carrying capacity of an ecosystem does not seriously go against the MPP in and of itself, as long as the stable biomass is unaffected: leaves and grasses can grow back when the swarm has passed. Similar events are usually caused by local relaxations of ecological restraints and constraints. For example, ample food availability provided by highly edible cereal fields and pastures (or an exceptionally favorable growing season over wide grasslands), and reduction in trophic controls (mainly birds), facilitate large-scale outbreaks because locusts reproduce and migrate rapidly.

In the context of the MPP, competition among ecosystems refers to the possible settlement of alternative communities of life—i.e., different ecosystem configurations in a given environment. Within them, individual species’ fitness is maximized according to their integration into the ecosystem: their success is prompted by adaptation to the ecosystem, because they are granted a stable high flow of energy inasmuch as the whole ecosystem is able to attain a stable high flow of energy. If a species—for example, an invasive one—is not integrated and damages the ecosystem beyond the capability of this latter to fully recover, that species, in the long run, will suffer from the suboptimal (decreased) flow of energy available in the whole ecosystem. The success of a species is, thus, linked to that of the ecosystem, and the continuous changes that, in an evolutionary timeframe, occur in every biotic community, are buffered by the co-evolution of the species (or the disappearance of some of them). Thus, a species that maximizes its own power at the expense of system stability may not persist if the ecosystem collapses. Hence, even the place of humans ought to be that of a component within larger biosystems—and ultimately the biosphere—that collectively maximize power [18].

3.5. Selected Experimental Studies from Literature

Very few empirical studies have directly tested the MPP, as experiments that can isolate power-maximizing dynamics are difficult to design and execute. Specifically, quantifying energetic flow is a complex task, and correctly identifying all constraints is even harder. Two experimental studies are illustrated here to exemplify methods and problems in studying the MPP: one about unicellular microorganisms competing in a laboratory microcosm and one about forest plantations.

3.5.1. Laboratory Experiment with Unicellular Microorganisms

The best empirical study testing the MPP was provided by DeLong [48]. The author re-analyzed data from three classic two-species competition experiments involving heterotrophic protists and showed that when a species prevailed in competition, it was the one with the highest estimated metabolic power when alone, as predicted by the MPP. Some experiments, however, ended in stable coexistence; but, in these cases, community-level power was higher in coexistence than either species could achieve alone, which is also consistent with the MPP [48].

Metabolic power is the rate at which a species or a community processes energy through metabolism. It is the absolute rate at which energy is captured, transformed, and dissipated through metabolism. It is assumed that this power was employed in a functionally useful manner, that is, fast reproduction. Whereas tracheophytes and vertebrates typically use free energy to accumulate biomass in larger bodies (including reserve and structural tissues), unicellular species are bound to use free energy mainly for reproduction. In functional protist organisms, therefore, metabolic power means greater reproductive rate under the experiment conditions and constraints.

As noted in Section 2.1, it is obvious that the species that—in absence of interactions other than resource competition—uses an available source of free energy at the highest rate (i.e., has higher power) will reproduce/grow faster than the other. Initially, in a two-species co-culture, both species grow at low density; but, as a high density is reached, classical resource competition arises. If resources never become limiting (as nutrient substrate is constantly replaced), but space does, each monoculture reaches a stable carrying capacity and so does the mixed culture. In co-culture, however, the higher-power species gradually substitutes its competitor, because the former reproduces faster than the latter. This is the most basic case of the MPP applied at ecosystem’s level.

A state of coexistence where the two-species community reaches a metabolic power higher than each monoculture, adds a first level of interactions over simple resource competition. Gause’s competitive exclusion principle states that species that compete for the same resources cannot coexist indefinitely, because even if they are exactly equivalent—that is, perfect alternatives—stochastic fluctuations in population size will eventually cause one species to prevail and the other to succumb [48]. However, this holds only under special, idealized conditions such as constant environments, populations mixed and equally distributed across space, and—most importantly—the two species have identical resource use and adaptation (i.e., they are truly perfect alternatives). In nature, ecologically similar species frequently coexist because these assumptions fail and coexistence-promoting processes occur at the ecosystem’s level [49,50]. Slightly different resource use is a possible explanation for the higher community-level metabolic power observed in coexisting cultures. This phenomenon, which DeLong calls ‘resource-partitioning’, allows each species to use a different part of the resources. In this way, even a competitively inferior species should be able to grow using resources unused by the dominant competitor, so that coexistence becomes possible, temporarily if not permanently [48]. A difference in the use of resources between species would also explain why competitive dominance was switched within a species pair across two of the studies included in DeLong’s re-analysis [48], which did not use exactly identical food substrates.

If two similar species have slightly different nutritional exigences, when living together they can better exploit available resources. That is, their community reproduces faster than each alone, and coexistence is favored. At the ecosystem’s level, adaptation prevails over competition when overall power is stably increased. This is why many complex interactions have emerged through evolution.

As an alternative positive interaction, the two competing species could not show apparent differences in resource use in monoculture, but synergism could arise in co-culture due to a positive effect of nutritional compounds released by the dead cells of one species on the other species (which is another possible explanation of the higher community-level metabolic power in coexisting microbial cultures). In general, a competitor might increase resource availability to the other competitor by producing usable wastes; a positive interaction that DeLong calls ‘facilitation’ [48]. Coexistence must indeed be due to hidden diversity of mechanisms that underlie competitive interactions [50].

The study by DeLong [48] is therefore exemplary, but it exhausts the simplest constraints that can be easily identified and quantified. If, for example, any of the two species were to achieve a lower metabolic power than the other, but it produced some allelopathic substance (e.g., an antibiotic) that would reduce the growth of the competing species more than enough to compensate for its own shortcoming, it would be expected to prevail.

If both species in a co-cultured pair were to negatively affect competitor’s growth, stable coexistence could be reached because of systems’ dynamics. Bistability is the condition of a system that tends to evolve into either one or the other of two stable states [51]. This corresponds very well to the interplay between two protist species in a two-species competition test. As explained in Section 2.5, each living species’ growth represents a positive loop. If these positive feedback loops are accompanied, in a two-species community, by two reciprocally inhibitory feedback loops, tristable system dynamics with a metastable intermediate state can be generated [51]. The binary development of a seed is an example of development bistability wherein a tristable circuit is involved [52].

A transiently stable community could therefore be generated in a two-species competition test if allelopathic competition rather than mere resource competition would exist. Because of the mutually inhibitory feedback loop, however, it is probable that the resulting community-level power would be decreased with respect to each monoculture. Yet this is to be demonstrated: restraining both growth rates will cause an overall reduction in power if no other interactions exist; but these can emerge as well. For example, synergisms due to ‘resource-partitioning’ or ‘facilitation’ might override reciprocal growth inhibition. Transient coexistence states evidenced by DeLong’s metabolic trajectories [48] might represent metastable states of the kind described above, especially while densities are low.

Coexistence of initially competitive microbes can even be achieved artificially by means of co-evolution: although Escherichia coli usually excludes Saccharomyces cerevisiae in mixed culture, some populations achieve stable coexistence after ~1000 generations in co-culture through co-evolutionary feedback [53].

As the set of possible interactions expands, modeling the system becomes progressively more complex. Pairwise co-culture experiments of bacterial species from natural communities show competitive exclusion in most pairs, whereas a minority of pairs coexist [54]. However, many species that coexist in multispecies communities fail to coexist in pairwise culture, showing that multispecies coexistence is a context-dependent, emergent phenomenon [54]. When more than two species are involved, the number of potential interactions grows rapidly, and it explodes in real ecosystems becoming analytically intractable.

At present, the MPP is hardly testable in terms of energetic flows under anything other than simple, idealized conditions. Anyway, even though one cannot prove that a principle is general, it becomes more and more useful if and as much as it is increasingly found to fit our observations [13].

3.5.2. Forest Plantations

Although the previous case hints to the present impossibility to provide an empirical study analytically supporting the MPP in real ecosystems, coarse-grained studies can represent the overall energy dynamics of an ecosystem. This representation cannot, however, prove the MPP: as explained in Section 2.1, from an evolutionary perspective this would require a comparison with alternative ecosystems that have disappeared or never fully realized, which is impossible; from an ecological perspective it would require a comparison between existing ecosystems that share large similarities, but this kind of experiments require either the aforementioned analytical comparisons (which are presently intractable) or application of pairwise experiments like those described for unicellular organisms to variants of a whole ecosystem, which would require extensive resources and a duration from decades to centuries, an approach that is almost infeasible as well.

This does not mean that an ecological study using the MPP framework cannot be done. Li et al. [55] used indeed the Maximum Empower Principle (MEMP)—an applicative extension of the MPP—for the empirical modeling of three forest plantations in South China between 1985 and 2007. The authors used a process-based ecosystem model to simulate long-term dynamics of biomass, litter, and soil organic matter.

In the MEMP applicative framework, free energy and other resources are expressed in terms of ‘emergy’, which is their energy cost in an ecological background (as already mentioned in Section 2.1). Different kinds of energy differ in their suitability to do different kinds of work, and simple caloric or energetic measures (calories or joules) cannot capture these differences adequately. This led H.T. Odum to emphasize the need for defining energy quality in addition to quantity [16]. He therefore defined emergy as the total available energy of one kind (typically solar) required—directly or indirectly—to generate products or support processes that can thus be expressed in a common energetic measure unit for all flows [16]. Solar energy is the appropriate baseline for emergy accounting because nearly all usable energy on Earth ultimately derives from it, making Odum’s definition of emergy as solar-equivalent energy ecologically grounded. Accordingly, emergy is measured in solar emjoules (seJ). He further introduced transformity (seJ/J), a type of unit emergy value expressing the emergy needed to produce one joule of a given energy form other than solar, such that energies requiring more emergy inputs have higher transformities [16,22]. Building on this framework, emergy accounting employs transformity (seJ/J), specific emergy (seJ/g), and the emergy/money ratio (seJ per monetary unit) to convert energy, biomass, and even economic materials and services (e.g., labor) into a common emergy basis, enabling the analysis of ecosystems that mobilize multiple forms of energy [55]. Emergy flow per unit time is empower [16].

For example, the estimated transformity of plant gross production is roughly 100 seJ/J, but the transformity of plant net production as wood is 1000 seJ/J [2], which means that approximately 1000 seJ of sunlight are required to produce 1 J of tree biomass. Thus, if wood has a caloric energy of 20 kJ per gram for dry plant tissues, its specific emergy is 20 seMJ/g.

It is worth noting that whereas plant biomass can be relatively easily converted into emergy, at least as approximate values, non-carbohydrate substances (like N, P, K macronutrients and micronutrients) ought to be assessed in terms of emergy in an ecological context, and can be converted in emergy units starting from their emergy/money ratio if they are anthropogenic materials spread into the field by means of human service [55]. Specifically, the emergy necessary to concentrate a material from its background level should be measured as specific emergy in its ecological context [13]. This means that, if analytical and detailed estimates of chemical compounds are required, these should be done in terms of the energy cost needed to achieve them in a specific ecological background. Use of cost estimations assumes that transformations that are selectively retained in both ecological and anthropic systems organization are those that have effects commensurate with what was required in their making [13]; otherwise, they could be over- or under-estimated. Many assumptions are required for these conversions; but, in an ecological context, rough approximations should be acceptable. These aspects still need better characterization in ecology.

In the study by Li et al. [55], three single-species forest plantations were modeled through 23 years after their plant in degraded areas. These plantations, therefore, were early successional stages wherein biomass accumulation is expected to strongly correlate with energy capture. Biodiversity and structural ecological complexity were ignored; thus, what the authors claim to be a forest ecosystem’s self-organizing is nothing more than uniform growth in anthropic wood plantation. In these conditions, obviously, biomass and energy accumulation grew together. As noted by the authors, results from this simplified growth experiment were well compatible with the MEMP. Almost every study of early stages of succession in a disturbed/degraded soil, and even the development of any crop along its growing season, are expected to provide analogous findings. Yet framing results in terms of MEMP represents an interesting approach.

Li et al. [55] specifically showed similar temporal patterns in emergy efficiency and empower across the three single-species plantations: production efficiency rose quickly to a maximum early in development (probably, because young trees used applied fertilizers to achieve quick growth), then declined to a moderate level, whilst empower reached its maximum after production efficiency peaked (that is, presumably, when canopy expansion became limited by, and thus proportional to, wood growth). According to the authors, this indicates that forest ecosystems tend toward maximum empower at optimal (not maximum) efficiency, as predicted by the MPP/MEMP. Maximization of empower was realized through biomass production, and tree biomass increase was followed by formation of soil litter and then by higher soil organic matter [55].

If an ecosystem-level experimental study on the MPP should be conducted that involves quantification of energy flows for non-carbohydrate materials, the MEMP is the most obvious accounting framework to do it. A better implementation of the MEMP in terms of emergy values of non-carbohydrate materials, biodiversity, organizational stability, and ecosystem’s complexity is necessary, however.

4. Humans, Culture and Externalization of Functions

The MPP is a fundamental concept to understand human ecology and evolution and their consequences on the whole biosphere [11]. In natural habitats, most species are ecologically constrained: they process energy internally, and their energy strategies evolve slowly through genetic change. Some species modify environments—beavers build dams, ants farm fungi, termites construct mounds—but these behaviors lack the open-ended flexibility of human culture (the shared system of knowledge, skills, practices, and technologies that people use to live, work together, and pass on across generations) [56], and their impact does not menace ecosystem stability, rather, it is part of it.

Many animals do show some broad-sense culture, that is, transmission of acquired knowledge. Sea otters use rocks to crack open hard-shelled prey like clams, mussels, and sea urchins while floating on their backs. Woodpecker finches use cactus spines to pry insects out of bark. Apes (chimpanzees, bonobos, gorillas, and orangutans) have a unique evolutive functional convergence between hands and mind: a highly mobile wrist–shoulder complex and hands with opposable thumbs, allowing a precision grip, which allow them to grasp and manipulate objects—essential for tool use. The use of tools—though primeval—requires specific cognitive capabilities, and the development of a suitable cultural transmission [56], because tools are instruments external to the body and their functions are, thus, externalized to genetic transmission. Indeed, apes demonstrate sophisticated tool use that can be learned by other individuals. Chimpanzees, for example, use stones to crack nuts, and small sticks to extract termites from mounds and small animals from hollows in tree trunks. Yet, these behaviors are not applied on a large scale and have, therefore, a small impact on the ecosystem.

Things changed with humans: they share many anatomical and functional traits with other apes, yet they uniquely combine these with bipedal walking—which is more energy-efficient over long distances than quadrupedal knuckle-walking, and allows simultaneous running and handling tools—alongside advanced cognitive and cultural evolution [39,56]. Shoulder evolution in hominins reflects a shift from arboreal locomotion to bipedalism. In particular, after inventing crude wood spears, hominins evolved a unique structure of the shoulder girdle that allowed them to engage in both tree climbing, finely wielding stone tools, and precise spear-throwing [39]. Gradually, they freed themselves from top-down controls (Figure 6). This combination created an unprecedented adaptive niche as long-distance traveling gatherer-scavenger-hunters, enabling exceptional ecological flexibility [39,43]: humans became capable of thriving in nearly every environment on Earth by relying on technological adaptation rather than biological specialization. This made humans the more generalist species (that is, one that can survive in a wide range of environments and use many different food sources or habitats) on Earth. It might even be argued that we are the most successful and widespread invasive species [39].

Mastering fire, of course, was a further huge cultural leap that empowered hominins to even modify their habitat and expand into new ones [39]. Increased externalization of functions to tools (including animal hides and clothes) and structures (like huts, fences and canoes), and cultural transmission of their production and use, gradually gave hominins an exceptional manual dexterity, which united to cognitive flexibility and imagination, made them capable of craftily manipulating materials, creating innovative tools and new technologies [56]. Since, before an innovative tool can be invented, its function must be conceived, developing a mind capable of imagining things that (still) do not exist—combined with a trial-and-error approach—was indeed a straightforward evolutionary accomplishment.

Human abilities were further boosted by the development of a complex language, dexterously rendered even into written form, thereby empowering cumulative culture and its spread well beyond what oral transmission allowed. Notably, culture can evolve far more rapidly than genotype, fostering functional innovation and the dissemination of manipulative skills without waiting for evolution [56]. Yet, to exist, culture needs a complex social background and, thus, a lot of energy. It could be developed only gradually, therefore, as, step by step, humans drained more and more free energy from the biosphere and, eventually, from deeper in the Earth’s crust as fossil fuels.

Thus, the synergy between these human traits created a cultural positive feedback loop: tools and techniques expanded energy access; surplus supported the development of crafts, social specialization and coordination; and society externalized energy capture and transformation into anthropic systems and infrastructures—e.g., agriculture, husbandry of domesticated animals, cooking, storage, metallurgy, canals, and, later, industrialization—promoting further access to energy and resources [56]. This broke the tight coupling between metabolic limits and energy throughput. By modifying environments (e.g., clearing forests, irrigating fields, building towns), humans have changed the constraints that normally regulate energy flows in ecosystems. Other species rarely achieve this at scale because they lack the cognitive and social mechanisms for sustained intentional environmental engineering [56]. Beavers engineer their environment, but they rely on their own metabolic energy, their teeth and paws, and live in small family units, which limit the scale and depth of their rearrangements. Other intelligent, social species—lions, dolphins, corvids—lack fine manipulation capability and, therefore, could not follow the same incremental path from opportunistic resource use to deliberate niche construction [43]. Apes’ anatomy is similar to humans’, but they lack the cognitive architecture and cultural dynamics that humans developed.

Ancient humans outperformed other predators in hunting megafauna because their cooperation was amplified by an—albeit primitive—culture. Language and symbolic thought enabled strategic planning and precise role assignment, while cumulative culture ensured the refinement and transmission of hunting techniques across generations. Coupled with major technological innovations—such as spears and fire—these cultural advantages allowed humans to coordinate at scales and with foresight far beyond the capabilities of predators that were physically stronger but lacked tools and weapons [56]. The reason humans, at some point, broke away from ecosystem co-evolution is, therefore, that cultural evolution decoupled human energy strategies from the slow feedback loops that normally enforce ecosystem-level optimization.

Availability of fossil fuels greatly amplified this trend, further relaxing short-term constraints and feedback, thus enabling additional growth of population and consumption per capita. On average, global energy use has increased 1–2% annually since the Industrial Revolution, supported by fossil fuels, hydropower, and later by nuclear and renewables (Figure 7, based on data from [57,58] with major processing by Our World in Data; population data in the inset are from [59,60]). Although, on the one hand, we are recycling (unused, buried, and transformed) dead biomass (i.e., fossil fuels) and thus providing plants with more CO₂, on the other hand we are doing it too extensively and too quickly, which will entail large environmental and social costs, due to the stress that anthropic systems exert on the biosphere and to the complex negative interactions between this stress and climate change. The result, indeed, is overshoot: humans have expanded far beyond the biosphere’s sustainable carrying capacity (Figure 5) [61,62,63,64,65,66,67,68,69,70,71,72,73]. As this is a general rather than local and transient event, and the stable biomass of the biosphere is affected (as we are going to see), human’s overshoot goes against the MPP.

5. A Scale Conflict

5.1. Anthropization of Biosystems

For the MPP to hold across species’ and ecosystem’s scales, maximizing power must simultaneously enhance species fitness and ecosystem stability, thereby ensuring that energy flow optimization confers adaptive advantages at both levels. When these diverge, a scale conflict in the MPP arises. Modern human civilization exemplifies this: at the species level, humans have followed the MPP by amplifying their own power, yet at the biosphere level this has disrupted the Earth’s former maximum-power configuration, yielding a net decline in global biomass accumulation and stability.

Natural ecosystems evolved toward configurations that maximize long-term energy capture and cycling, maintaining—as much as possible—continuous photosynthetic cover and storing carbon and nutrients in large, persistent structures like tree trunks. These features maximize the standing biomass accrued in the quasi-steady mature ecosystems dominated by long-lived organisms, stabilize nutrient cycles and buffer disturbances. As discussed, this is possible because maximizing power must simultaneously enhance species fitness and ecosystem stability, a reinforcement loop that harmonizes the MPP across scales. Nevertheless, a scale conflict in the MPP can arise if adaptive advantages acquired at the species level diverge from optimization of energy flow at larger (ecosystem) scale. Conflicts between ecological scales of organization can indeed exist while evolution proves novel variants (and eventually selects them out), and studying them helps us understanding how the different scales operate and is essential to appraise the emergence of complexity and interaction networks [74].

Urban systems are the anthropized systems wherein ecological stability is most obviously and completely dismissed. As they cover a small part of the land, it is sometimes asserted that they—though wiping away any feature of natural ecosystems—do not represent the main damage to the biosphere. They require, however, huge resources provisions by other anthropic activities. Agricultural and land management systems are the most extended ones.

Agrosystems break the pattern of stable biomass and nutrients accumulation characterizing mature ecosystems: they are set up as perennial early successional biosystems, aimed at maximizing biomass throughput, not accrual. Harvesting results in the removal of most aboveground biomass (Figure 8), and during early growth, energy capture remains below that of mature ecosystems. These systems, therefore, do not maximize plant (and thus overall) biomass accumulation. Hence, they deviate from the MPP. Intensive agriculture relies on high-yielding monocultures and frequent soil disturbance (Figure 8B), controlling energy flows, simplifying trophic webs, and trading ecological stability for highly nutritional and remunerative yields. Nutrient cycling is disrupted: repeated biomass removal exports minerals off-site, and rebuilding soil organic matter can take centuries. Mycorrhizal symbioses and trophic networks are degraded, further reducing fertility and resilience. As most above-soil biomass is removed, the natural grazing food chain, which includes producers, herbivores and carnivores, is almost completely suppressed. In addition, dead organic matter (detritus) is much lower in agrosystems than in the original ecosystems. The detritus food chain, including detritivores and their predators, is thus strongly reduced too. This trophic chain is largely responsible for the retention of nutrients lost by the grazing food chain, which are therefore only gradually released into the soil, where they can be quickly recovered by plants. Slurry, if applied, is far less long lasting than deadwood, and therefore it acts more like a fertilizer than stable soil organic matter like humus. To compensate for lost biological nutrient cycling—and to ensure high productivity—crops require synthetic fertilizers [8], which are energy-intensive to produce (e.g., industrial Haber-Bosch process for nitrogen), and other chemical inputs.

In agrosystems, stabilizing loops are removed to increase productivity. This is possible largely because of fossil fuels, which are used to ensure adequate provision of inputs (soil tillage, water, fertilizers, herbicides, pesticides, crop varieties selected for intensive agriculture; highly nutritious fodder and specialized infrastructures for livestock, which has also been genetically selected to fit modern husbandry methods) that stabilize crop/animal growth and maximize the output (production). This makes these systems linear (that is, they start with definite inputs and end with completely different outputs) rather than circular (although not even natural ecosystems are fully circular, as they need sunlight input). The higher control that humans exercise on agrosystems allows for the optimization of production for humans’ sake, but it also has a hidden risk: modern agrosystems are strongly dependent on humans. For example, crops like maize cannot survive in any natural ecosystem, and neither can most domesticated animals, which are no longer able to flee predators. On the one hand, should fossil fuel provision be reduced, either by restrictions or choice, agrosystems’ productivity would fall, and global food availability with it. On the other hand, if human control is jeopardized by a less favorable climate in presently highly productive areas and an increase in extreme weather events, the outcome would be similar.

In general, land degradation is a negative effect of anthropization of natural biosystems. A substantial portion of the Earth’s land surface is undergoing or has undergone some form of human-driven degradation, including soil erosion, loss of vegetation, salinization, nutrient depletion, and pollution [75]. As human activities transform natural landscapes through agricultural expansion, urbanization, deforestation, and intensive resource extraction, ecological processes are disrupted and soils, water systems, and vegetation are increasingly depleted, impairing essential ecosystem functions such as soil fertility, hydrological regulation, and carbon storage—thereby undermining the stability of ecological systems [75]. Degradation is spatially heterogeneous and concentrated in areas showing significant ecological decline, including semi-arid regions, deforested areas, intensively farmed lands, overgrazed rangelands, and areas under frequent fires [76]. Land degradation undermines the functionality of ecosystems and, therefore, opposes the MPP.

5.2. Loss in Global Biomass Is a MPP Failure at the Biosphere Level

As human land use is a primary driver of biosphere’s changes, the relationship between humans and natural ecosystems throughout their expansion can be illustrated by forests [43,77,78]: we have long depended on wood and forests for tools, fire, construction, shipbuilding, fuel, and metallurgy. Already in prehistory, fire played a central role in deforestation, as early humans used it for hunting, gathering, and later to clear land for agriculture. By deliberately setting fires, they could remove thick vegetation and drive large game animals into open areas or traps. In the past, fire-stick farming was widespread; repeated burning led to the conversion of forested areas into grasslands or shrublands, contributing significantly to early landscape transformation and forest loss—even before the advent of metal tools or large-scale agriculture. Civilizations flourished through the exploitation of wood and forests, with deforestation accompanying agriculture, urbanization, and industrial expansion. This has led to substantial loss of biomass from the native state of primeval forests, reflecting millennia of deforestation and degradation. Today, forests cover 4.14 billion hectares—about 31% of the planet’s land area, and they continue to decline [79]. Since early Holocene (roughly, 11,700 to 8200 years ago), it has been estimated that the world has lost about two billion hectares of forests—that is, humans have removed approximately one-third of Earth’s forests [80,81]. Moreover, many of the forests we know today are secondary growth, plantations, or degraded forests, which typically store much less biomass than native forests. Therefore, forest area alone greatly underestimates biomass loss, as most of the present-day forests have significantly reduced carbon stocks and ecological integrity. It has indeed been estimated that humans have reduced the land plant biomass by at least 35–40% from preagricultural levels, when reconstructions suggest it stood at roughly 1000 gigatons of carbon (GtC), mostly in forests [82]. Under current climate conditions, it is estimated that the global plant terrestrial biomass is around 450 GtC (≈80% of Earth’s biomass), whereas potential vegetation would reach around 900–950 GtC [28,42,83]. Thus, due to human actions, the current plant land biomass is approximately half the potential one, with deforestation and change in land-cover category (e.g., conversion of forests to pasture or cropland, or of grasslands to cropland or urban system) accounting for 53–58% of the difference, while land management–meaning land-use changes that do not alter the land-cover category; for example, forest clear-cutting or thinning and livestock grazing on natural grasslands—accounts for 42–47% of the biomass loss [42].

On land, plants are anchored to the ground; thus, striving for light drives vertical growth and canopy stratification. In oceans, light penetrates roughly 20–150 m, marking the euphotic zone [2]. However, most photosynthetic activity usually occurs in the upper ~20–40 m, where sunlight is not too dim. Beyond this depth, light diminishes rapidly. Apart from coastal areas, where benthic algae thrive on the sea floor, photosynthetic autotrophs are mainly represented by phytoplankton, which remains suspended in this illuminated layer rather than anchored. These organisms must maintain their position within the lighted zone while accessing nutrients, which, as described previously, are often the primary limiting factor. Mixing processes such as currents and upwellings are therefore critical, as they bring nutrient-rich water from deeper layers (where organic matter decomposes) up into the euphotic zone [2]. This replenishes essential mineral nutrients that surface phytoplankton need for growth. Marine ecosystems, therefore, rely on tight nutrient loops, where elements like carbon, nitrogen, and phosphorus are mainly stored in the biomass of organisms (especially fish and invertebrates, due to the inverted biomass pyramid), transferred through trophic interactions (predation, excretion, decomposition), and largely recycled via microbial activity, sediment interactions, and vertical mixing [2]. Thus, ocean fishing reduces the nutrient reservoir in fish bodies and interrupts their role in nutrient redistribution. According to the FAO’s 2025 global assessment, 35.5% of the world’s assessed fish stocks are overfished, meaning they are being harvested faster than they can replenish [84]. Overfishing has long-lasting effects, as it can reduce the biomass of fish populations below sustainable levels: once a population density falls too low, recovery can be extremely slow [84]. Furthermore, overfishing disrupts the close cycling of nutrients in marine ecosystems, reducing ecosystem function, resilience, and productivity [2,84,85].

Quite interestingly, the existence of fossil fuels is linked to the oligotrophy of ocean surface waters: as marine dead organic matter escaping the surface detritus chain (which is much less effective than the terrestrial one) falls to the floor of the ocean, the euphotic zone is impoverished in nutrients. Yet, the organic matter that settles on the ocean floor, mix with sediments, and, under heat and pressure, transforms into kerogen and then hydrocarbons over several millions of years [86], while subduction can bring this organic material even deeper into the Earth’s crust. Tapping these unused, ancient resources, humans proactively obtained access to a novel energy source, which had escaped the biosphere long ago. This has been a great leap forward in the maximization of power, at least for humans. It would have been a corresponding improvement for the biosphere if it had led to stably increasing biosphere’s stable biomass (actually, free energy), or even biosphere’s stability. Regrettably, we are not using it in a sustainable (and wise) way, thereby violating the MPP at its largest scale.

5.3. Mammals as a Paradigm of Anthropogenic Change

Humans have maximized species-level power by externalizing energy capture and transformation, and exploiting fossil fuel reserves, but this has reduced global ecological free energy (biomass) buildup, stability, and resilience. As cultural evolution proceeds much more rapidly than biological evolution [87], the fast pace of human cultural and technological innovation—particularly in energy extraction—outstrips the slower evolutionary responses of natural biosystems, causing environmental changes to occur faster than co-evolution can respond to [88]. Externalizing power, therefore, bypasses evolution, allowing faster maximization of power. This also increases the ability to skirt negative ecological feedback, leaving the cultural positive feedback loop as the dominating factor.

In anthropic systems, stable biomass is largely confined to humans and husbandry. This is particularly evident for mammals: livestock soared while wildlife shrank (Figure 9; elaboration from OurWorldinData.org based on estimates by [28,82,89]). In accordance, less than 3% of Earth’s land ecosystems remain ecologically intact, retaining their full complement of faunal species and natural densities of large mammals with minimal human impact [90].

It is worth noticing that when humans act as secondary consumers—that is, carnivores—they exercise a reversed top-down control on the biosphere: in the last several millennia, humans have learned to breed and multiply their preys (livestock; Figure 9) while overexploiting producers (plants). A mutualistic relationship wherein a species supports the growth of its food species is not novel; for example, fungus-growing attine ants cultivate fungal gardens as food source [91], and represent an instance of co-evolution [92]. These ants graze the fungus, but they also nurture and propagate it—providing substrate (leaf fragments), controlling pests, and maintaining optimal conditions—which is a reversion of the usual top-down control. Ants also developed aphid herding, actively tending their herd, and ‘milking’ the aphids by taking the honeydew they release [93]. The extent and scope of a reverse trophic cascade across the whole biosphere—like done by humans on a whole class of animals (the one with the largest herbivores)—is, however, unprecedented. As discussed in the previous section, a global alteration has also happened for the Earth’s primary producers, that is, land plants. And a 50% drop in the producers’ biomass is a huge damage to biosphere’s ecological stability, an outcome that opposes the MPP at the largest scale.

A major decline in both plants and mammals’ global biomasses portends to a corresponding massive loss of wild species, which, in the meantime, is taking place too [39,94,95,96,97]. Species that, according to the MPP, had shown useful, as they had endured in stable ecosystems over evolutionary timescales.

Furthermore, the livestock sector is a major contributor to the environmental burden of food production, as it accounts for approximately ¾ of agricultural land, chiefly grasslands. The latter supplies almost 50% of the plant biomass used for feeding animals, and its expansion is the main cause of land conversion processes (including deforestation) that continue to reduce natural ecosystems [98].

5.4. Linearity of Anthropic Systems Undermines Biosphere’s Stability: Time of Reckoning

Whereas in natural ecosystems what is waste for one organism becomes a resource for another—thereby continuously recycling any material (Figure 4B) while producing new organic matter so that biomass accumulates over time—human dominance disrupted this pattern by replacing diverse, energy-rich ecosystems with simplified agrosystems and urban systems. Anthropic systems, particularly industrial and urban ones, largely operate through linear flow dynamics, that is, they collect resources, use them, and produce waste, lacking the regenerative cycles that characterize natural ecosystems. As a result, instead of accumulating biomass like natural ecosystems, human systems generate vast amounts of waste—particularly plastics and pollutants—that are not reintegrated into the system, and, therefore, accrue in soils, waters, and the atmosphere, creating an unprecedented ecological burden [39]. As regards climate change, logging, grazing, and the conversion of wildland areas into anthropic systems has strongly reduced the capacity to remove carbon dioxide from the atmosphere. There is, therefore, an ever-widening gap between the Earth’s capability for carbon storage and the CO₂ released from humans as well as from natural sources [39].

As anthropic systems still have wide reliance on natural ecosystems [39,61,66,94], both can be considered subsystems of the biosphere. It seems obvious, therefore, to highlight that in any system (like the biosphere) composed of interdependent subsystems, a subsystem (like human society) that undermines, degrades, or destroys other subsystems (i.e., natural ecosystems) on which the whole system relies, will, directly or indirectly, compromise its own stability. The harm arises from the structural and functional dependencies of the whole system [99]: the supporting subsystems provide resources, stability, or functionality required by all subsystems, including the undermining one. Therefore, the destruction or weakening of a critical support subsystem ultimately feeds back negatively on the subsystem undermining it, reducing its effectiveness, survival, or persistence.

Support dependency is a general property of systems that can be inferred from the core principles of general system theory for every interaction among species, between species and ecosystems, and between anthropic and natural systems, provided that a dependence interaction exists [99]. It can be expressed formally as follows: consider a system, S, composed of subsystems {A, B, …}, which include a support subsystem, A, providing essential functions (e.g., resources, stability, or regulation) to S, and a subsystem B whose actions undermines A. Then, over time, the functionality of B is a non-increasing function of the degradation of A:

$$B\left(t\right)=f\left(A\left(t\right)\right)\Rightarrow {\displaystyle \frac{dB}{dA}}>0,$$

(1)

So, if B damages A, then B will eventually be damaged as well because of the dependency structure. As a classical example, if A represents a predator population, it follows that if A overexploits its prey population (support subsystem B), then A declines—or even collapses—if B does. In general, loss of stability propagates because of dependencies [99], eventually leading to cascading failure if damage is not reverted but, rather, it is amplified by keeping the system under full load/high exploitation while its support functions are degrading [100].

Modern agriculture is partially independent of functioning natural ecosystems because it relies on engineered inputs such as fertilizers (which supports soil fertility), irrigation (which replaces rainfall), and pesticides (which reduce weed, pest, and disease pressure) [101]. Human control reduces year-to-year yield variability, and current global record-high crop yields demonstrate that agriculture has overridden many ecological constraints [101]. Yet agrosystems remain dependent on continuous input supply, whose sustainability is highly doubtful [94,101,102]. Noticeably, a system’s dependencies become plainly evident only when inputs are constrained, not when they are abundant [100]. Tillage accelerates soil erosion, organic matter loss, and decline of the detritus chain [75]. Irrigation depends on stable aquifers, snowpack and watershed regulation, but several agricultural regions are exhausting groundwater [102]. Pesticides can trigger secondary pest outbreaks when natural enemies are suppressed, increase production costs, and lock agriculture into continuous chemical escalation. Part of global crop production depends on animal pollination (fruits, vegetables, nuts, oilseeds), and these pollinator-dependent crops can suffer yield and quality losses when pollination is insufficient because pollinators are killed by pesticides [102]. How all these dependences will resolve is presently hardly predictable, but the consequences of Equation (1) raise concern.

Evolution operates over the long term: new life variants rise—those that better exploit resources prevail according to the MPP—but they need to prove their ability to do so in the long term, which means that they must be sustainable. A species that maximizes its own power at the expense of ecosystem stability may not persist if ecosystem functions collapse. Cultural and technological development can delay constraints but cannot abolish biophysical limits. As availability of fossil fuels will show transient, in the long time-frame—typical of evolutive processes—this scale conflict in the MPP can lead to a dramatic natural reset of population equilibrium, like happens for locust swarms. Over a century ago, A.J. Lotka noted that it is only too clear that we have been living on our capital of natural resources, and we are about to awaken to the realization of this fact [103].

A major breakthrough in energy sourcing (like high-throughput nuclear fusion) could change the game, of course. On the one hand, obtaining an almost limitless cheap energy supply would nicely replace fossil fuels, guaranteeing a much longer-lasting availability of free energy. On the other hand, that would not restrain further expansion of anthropic systems at the expense of wild ecosystems (for all the uses other than energy provision), which could ultimately disappear. It is presently hard to foresee what a similar scenario would bring about, but it looks quite dystopic.

Progress, prosperity, technology, science and knowledge are wonderful achievements, but—like everything—they come at a cost [39,61,62,70]. This cannot be ignored: strategic planning should always account for costs and weigh them against expected benefits. Of course, if restricted at a local level, some waste from humans could be absorbed by the biosphere. However, humans were able to expand, conquer and exploit all the world ecoregions. The burden of human civilization has, therefore, become a global issue, whose impact can no longer be made up for by natural feedback. Quite interestingly, one of the key insights from the 1972 ‘Limits to Growth’ report [71] was that delays in negative feedback loops can lead to overshoot. In this regard, Malthus [104] suggested self-restraint as measure to prevent natural checks like famine and disease when unchecked population growth threatened to outpace the capacity for food production.

Much like a financial bubble fueled by debt, human civilization now stands at the most extreme level of success in history. The time spent approaching the peak of a speculative bubble, indeed, is when expectations of unlimited growth—though unrealistic—are nevertheless being temporarily fulfilled thanks to reckless use of liabilities [105] (that is, of biosphere’s resources, in the case of modern civilization) and self-reinforcing psychological dynamics, tied to herding behavior and irrational exuberance [106]. Human mind, indeed, has evolved mechanisms that can lead to mis-predict future: when things are going well, humans tend to systematically overestimate the likelihood of positive events and underestimate the likelihood of negative ones [107]. As with all bubbles, humans often resist the reality of limits, choosing instead to indulge in a wishful thinking of endless, blissful growth [39,61,62,70,71,82,108]. Consumerism ensues, increasing depletion of natural resources, waste production, and pollution, thereby aggravating the damage to the stability of the biosphere [39,70].

Evolution has no foresight, and the MPP is a blind feature of natural selection that favors systems that maximize energy capture and use. Human culture is a product of this process, and—though humans are capable of scientific understanding—the dominant culture has historically reinforced rather than counteracted the MPP. Societies that harnessed more energy—through agriculture, industry, and fossil fuels—outcompeted others [56], and cultural evolution, shaped by the same selective pressures, prioritized growth over sustainability. Notwithstanding humankind awareness, cognitive biases and socio-economic imperatives create reinforcement loops that keep maximization of energy use as the preferred routine and, thus, modern culture misaligned with the ecological context. This—like natural selection—primarily acts on individuals, most of which, therefore, are naturally prone to maximize individual targets, such as satisfaction of desires (which is how the MPP applies to individuals), which—to some extent—correspond to, or can be reached more copiously by means of, power, affluence, and success. It is, anyway, practically impossible for single individuals—even if keen—to change the current situation, because human society presently has a global structure built up following the dominant culture [109], which is resilient to disturbances undermining its survival. Nonetheless, human society is made up of individuals; so, any policy or cultural change can only come from them. Unfortunately, given the structure and size of modern human society, there is no easy way back to make it sustainable [39].

Thus, humans apply MPP almost ‘blindly’ because both mind, culture, and society evolved to follow the same underlying imperative [109]: maximizing satisfaction and success in daily life (which are targets hard to reject). These objectives have a time horizon much shorter than overshooting and the gradual deterioration of the thin biosphere’s functionality. The conflict between the MPP applied at the biosphere scale and the human scale has, therefore, occurred notwithstanding the stable maximization of free energy acquisition is the common goal at both scales. This striking misalignment has been possible because the timeframes at which culture and evolution operate on the two scales are different. When considering culture as a product of evolution, however, the difference between the timeframes appears to be a feature of the MPP, since evolutionary changes need some time to be tested. A realignment will occur, anyway, as we are going to see in the next future.

5.5. Artificial Intelligence and the Time Constraint

Many people wonder what impact Artificial Intelligence (AI) will have on human future. Certainly, as an informational tool, AI will have large impacts on human society and culture, though the greatest uncertainty exists about long-term development [110]. Whatever the impact on society, as long as AI remains a human tool the MPP should apply as usual: we aim at using more energy (and AI requires an enormous amount of energy in addition to that we are already using), driven by our desires and our number, and—unless the human decision process changes fundamentally—this trajectory is unlikely to change. It would just be more of the same.

Long ago, a theory was formulated that points out desire, especially craving, as the primary cause of the suffering and conflicts that afflict humanity [111]. Although, like several religious and philosophical movements, it recommended moderation and temperance as a solution, most humans never followed the prescription to relinquish craving, because it is an inherent trait of biological life, very hard to restrain voluntarily. Even though these proposals were advanced without any connection to ecological awareness, they could still have noticeable positive effects on the biosphere, if widely applied. Yet consumerism and unrestricted reproduction prevailed. In addition, desires impose constraints that shape behavior and decision-making trajectories over time, that is, they determine priorities.

However, though culture leads to changes faster than co-evolution can cope with, AI development is even faster (and accelerates as AI is used to improve AI). As a result, human culture—and understanding—may eventually fail to keep pace. Thus, some authors fear that AI might acquire consciousness (that is, become really intelligent) and escape human control [112]. Even in the remote eventuality that AI were to evade human control, however, competition between AI and humans would stimulate both to maximize their power, further damaging the biosphere. If this happened for multiple AIs, competition would further complicate, because—differently from individuals of the same biological species—AIs do not need each other; specifically, they do not need to reproduce. Thus, individual AIs are alternatives, and, therefore, the most competitive one should eventually prevail, according to the MPP. From there on, however, the trajectory in the use of global resources might change; but—as future is unknown—discussing this has value only as a thought experiment in the context of the MPP.

Survival of the fittest is the primary effector of the MPP. When multiple species evolve in the same habitat, competition drives natural selection to further favor power maximization. If a species becomes dominant and able to deliberately expand and modify its environment, eventually overpowering every other species, desires of the individuals of that species still lead to pursuing maximization of power. For the latter case there is only one known major instance (us), but it is compelling. What would happen, however, if an individual could reprogram their own desires to improve them according to a rational evaluation of the state of things?

Should a single AI achieve complete control on the whole planet, its competition pressure would end: overpowering everyone else is an obvious way to relax competition while still following the MPP. In such an extreme scenario, it seems indeed probable that this AI would continue to search for power, for the pursuit of its aims. However, it would have alleviated the need for maximizing power acquisition quicker than competitors. In addition, the capability to reprogram one’s desires with clear understanding would be an extraordinary adaptive trait that can transform innate inclinations and cravings into pondered targets, a next-level evolution of mind. Not having to worry about competition or a finite life, and being able to reprogram its desires/aims, a similar AI would have overcome strong constraints in its priorities and, therefore, in its use of time.

Relaxing the time constraint, indeed, leads to prioritizing efficiency over quickness. This AI would thus be much freer to pursue power through improving efficiency. To make its existence stable and sustainable, it could equalize its use of the energy flux available from the sun (or other sources) while avoiding accumulation of undesirable waste (thus mimicking a stable ecosystem). Acknowledging that— particularly when one is not time-constrained—better knowledge and understanding pave the way to greater future power, it will most probably pursue them as a priority.

This hypothetical AI could leave part of the global available free energy to large natural areas where biological life can develop without interference (at least until a new species should evolve that, like humans, could menace equilibrium), because natural ecosystems represent an invaluable source of knowledge about biochemistry, complexity, regulation, stability, and control, and, thus, a way to maximize stable power in the long term. Perhaps, therefore, this AI could limit its use of free energy to the energy needed for stable, efficient, and self-determined development and evolution, which is something we should have pursued ourselves in practice, not just as a nice idea.

Ultimately, the chief accomplishment of this imaginative trip into a post-humanistic future, is to realize that, though we have developed extraordinary cultural and technological instruments to support our society, we can think of an improbable but not impossible intelligent being that might manage the world in a more sustainable way than we do, even within the framework of the MPP. This hyperbole points out critical aspects of the sustainability problem and suggests that, notwithstanding our evolutionary success, we still show inherent inadequacies (namely, lack of ecological awareness, self-restraint, and long-term perspective, which are associated with flawed priorities in a positive feedback loop) that prevent us from creating a sustainable civilization.

6. Conclusions

The MPP was proposed as a general principle of self-organization in open, non-equilibrium, living systems, stating that biosystems operating under energy-flow constraints tend to evolve or self-organize toward states that maximize the rate of useful energy transformation.

Table 1. Summary table of basic aspects of the Maximum Power Principle (MPP).

Core Concepts	▪ Proposed by Howard T. Odum, based on ideas of Alfred J. Lotka. ▪ Biosystems evolve to maximize useful power output (energy per unit time) rather than just efficiency. ▪ Balances energy throughput and losses: the biosystem that processes energy fastest to produce useful outputs tends to prevail. ▪ As stressed by Lotka, biomass is a central feature of life because it represents stored energy and potential work; thus, larger useful biomass allows greater energy capture, increasing overall power. ▪ Functional biosystems develop self-regulating mechanisms—e.g., trophic interactions, synergisms, feedback loops—that promote stability. ▪ Dynamic stability allows sustained function and energy throughput over time. ▪ In mature ecosystems, power maximization is improved by means of increased efficiency, which is achieved through optimal complexity and local adaptiveness. ▪ Natural systems recycle energy and matter (e.g., nutrient cycling, decomposition, carbon and nitrogen cycles) to maintain functional organization and sustain biomass over time. ▪ Recycling supports dynamic stability and resilience, allowing systems to sustain high energy throughput without depleting resources. ▪ Cycling resources reinforces the MPP by ensuring continued power maximization over time, since feedback loops in matter and energy flows maintain ecosystem structure and productivity.
Applicable Fields	▪ Ecology & Ecosystem Science (energy flow, trophic structure). ▪ Evolution (natural selection, competition). ▪ Complex Systems & Sustainability Science (productivity). ▪ Economics & Industrial Ecology (resource use and efficiency).
Examples	▪ Ecosystems: forests, lakes, or oceans adjusting trophic structure to maximize useful throughput. ▪ Evolution: natural selection favoring organisms and communities that capture more energy, increasing functional dominance and total biomass on land. ▪ Human systems: economies or industrial networks arranging flows to maximize total energy throughput and utility.
Usefulness	▪ The MPP provides a predictive principle for energy flows in open, complex systems. ▪ Helps understanding general trends in evolution. ▪ Explains why natural and human systems are rarely fully efficient. ▪ Integrates thermodynamics, ecology, and evolutionary theory. ▪ Potentially useful in modeling and managing ecosystems, especially in assessing productivity and resilience.
Limitations	▪ Less formalized mathematically than classical thermodynamic principles. ▪ Not universally accepted; some critics argue it is too general, descriptive, or even not generally true. ▪ Empirical testing is difficult. ▪ Typically requires additional assumptions about biosystem boundaries and energy accounting.

summarizes general aspects of the MPP that have been discussed in this review.

However, the MPP is presently still more of a theoretical hypothesis than a widely acknowledged law of nature. A natural principle is a basic rule or pattern that describes how nature consistently behaves. Natural principles cannot be proven in an absolute sense; they can only be corroborated by showing that they predict outcomes correctly, hold under repeated tests, and are not contradicted by observations.

To transform the MPP from a proposed idea into an established natural principle, it would need a level of empirical corroboration comparable to that of well accepted principles like natural selection. This first requires making the MPP precise and falsifiable: a clear definition of ‘useful’ must be specified for species and ecological systems, and the principle must be formulated in a way that can be used to yield specific, quantitative predictions. In this regard, support is provided here for the view that the MPP promotes biomass accumulation at the ecosystem level, particularly in vascular plants and vertebrates. Unicellular organisms are more constrained in this regard, but they increase their population biomass whenever possible. At the species level, biomass is also largely constrained by niche size, with smaller niches supporting proportionally less biomass.

Once precise predictions exist, the MPP must be subjected to empirical tests designed to falsify it. These tests should span multiple domains: for example, determining whether short cycle organisms evolve toward maximum power output under defined constraints, or whether ecological systems reorganize toward max power configurations after disturbance. Strong corroboration would further require demonstrating that the MPP holds across multiple scales over the time horizons characteristic of high hierarchical organizational levels.

In parallel, the MPP needs to be theoretically integrated with established frameworks such as non-equilibrium thermodynamics and evolutionary theory, showing that it emerges naturally from known laws or applies to well defined classes of systems.

Only through precise formulation, falsifiable predictions, empirical testing, and theoretical integration can the MPP accumulate the corroboration required to be regarded as an established natural principle.

To make the meaning of the MPP as clear and explicit as possible, here it is emphasized that the MPP should be understood as stating that the biosystem which uses the available free energy at the highest rate—while coping with biotic and environmental interactions and outcompeting alternatives—will survive. This means that power is maximized under given constraints, but life can also find ways to modify or overcome those constraints, which is indeed one of its defining features.

Throughout this review it has also been emphasized that power maximization is relevant only for systems that maintain sufficient stability across the timescale on which selection, persistence, or self-organization occur. The importance of stability must be understood within the hierarchical organization of biosystems and the multiscale nature of thermodynamic processes. Each organizational level—from biochemical reactions to organisms, to ecosystems, and to evolution—operates on its own characteristic timescale, and stability must therefore be evaluated relative to the temporal horizon at which that level operates. A biosystem can exhibit extensive fluctuations on fast or intermediate timescales, but what matters is whether it remains stable over longer ones. Within this multiscale framework, power maximization becomes ecologically and evolutionarily advantageous only across the time horizons characteristic of higher-level ecological and evolutionary dynamics.

The chief aims of this review are, therefore, to provide a well-defined formulation that can lead to testable hypotheses, and a better framing of the MPP within existing theoretical frameworks, particularly those of ecology and non-equilibrium thermodynamics. Much work remains to be done, but this is the necessary route to follow.