Research and version note This is a version 0.1 research draft in Who Maintains the World? It examines the conceptual boundaries of ecosystem maintenance, with Australian fire, invasive-species management and Caring for Country as part of the context. Ecological objectives require place-, species- and law-specific research. This article cannot replace a local ecological assessment, and it does not treat Indigenous knowledge as a general technique detachable from its knowledge holders and rights-bearing communities.
Who Maintains the World? · Article 10
Quick read
For a machine, good maintenance usually means continued operation within a designed range. Nature does not have one comparable specification. Forests renew, rivers move, species distributions respond to disturbance and climate, and ecological succession changes the visible landscape. Even without direct human intervention, an ecosystem does not remain in one historical year. Treating all change as failure can suppress the fire, flood, movement and renewal on which ecological continuity depends.
It does not follow that nature needs no maintenance. Many environments remain under pressure from land clearing, greenhouse emissions, pollution, altered hydrology and introduced species. Non-intervention begins from those inherited conditions; it does not remove them. Habitat repair, control of particular invasive processes, and restoration of suitable water or fire relationships can therefore be necessary.
The difficult question is what is being maintained. Species composition, ecological processes, cultural relationships, land use and adaptive capacity may not all be maximised together. A baseline is not a neutral fact either. Choosing pre-colonial reconstruction, the first systematic survey, a statutory listing date or the present distributes responsibility differently and relies on different knowledge.
Australian accounts of Caring for Country show another limit in the machine analogy. For many Aboriginal and Torres Strait Islander peoples, Country involves interconnected land, waters, life, law, ancestors, people and obligations. Relevant practices must remain connected to their knowledge holders, authority and Country. Extracting cultural burning as a generic fuel-reduction technique removes governance and relationship from the knowledge.
Measurement remains necessary, but a hectare treated or a seedling planted is an activity rather than proof of ecological continuity. Monitoring must stay long enough to observe survival, function, unintended harm and response to disturbance. Baselines and success measures should disclose whose values they represent, especially where a numerical target can hide a conflict over Country, species or future land use.
This draft argues that ecological maintenance should sustain the relationships through which living communities can continue to form and adjust. Objectives must be explicit, contestable and revisable. Monitoring must search for unintended effects, while governance recognises Indigenous rights and local knowledge. Sometimes maintenance requires intervention; sometimes restraint. Neither is responsible until it identifies the harm, time horizon and authority against which it is judged.
Nature is not an object waiting for a service schedule
The word maintenance commonly implies return to an intended state. A road crack is filled, an instrument is calibrated and a software defect is corrected. Those objects normally have a designer, a defined function and some tolerance for deviation. An ecosystem has no equally simple specification.
A woodland is continually produced through relations among organisms, soil, water, fire, climate and human activity. Trees mature and die. Species enter, retreat or change in abundance. A storm may destroy one structure while creating habitat for another. Change is not always a fault in this setting. It is often part of how the system continues.
Environmental management therefore cannot classify every departure from a familiar view as damage. Excluding fire or flood indefinitely to preserve one visible state may weaken renewal and accumulate risk. The appearance has been kept while the relations that support it have been diminished.
This also distinguishes ecological maintenance from simple preservation. Preservation can be an appropriate objective for a particular heritage feature, specimen or highly vulnerable refuge. At ecosystem scale, however, continuity often depends on controlled variation. The relevant question is whether change remains within conditions that sustain future life, not whether every component stays where an observer first saw it.
Leaving nature alone is not a neutral starting point
Recognition of ecological change is sometimes converted too quickly into a rule of non-intervention. Contemporary ecosystems operate within altered climates, fragmented habitat, dammed rivers, contaminated sites and global transport networks. An introduced predator does not leave when a management programme stops, and the hydrological consequences of past clearing do not disappear because current managers abstain.
Non-intervention always leaves some already formed conditions in operation. It can be the right choice when an intervention is speculative, damaging or unnecessary. It can also disguise accumulated human influence as a natural baseline. A decision must compare what action and inaction would each continue.
Repair and maintenance overlap but are not identical. Repair responds to identified degradation and attempts to recover a function or relation. Maintenance includes the repeated observation, care and adjustment through which recovery persists. A planting event without later attention to survival, weeds, water, grazing and fire has completed an activity; it has not necessarily restored an ecosystem.
A baseline selects a past
Ecological assessment needs comparison. Without it, improvement and deterioration lose meaning. Yet a baseline may come from the earliest systematic survey, a statutory list, an estimate of pre-colonial conditions or a deliberately chosen future reference. Each has evidential limits, and each frames responsibility differently.
A recent baseline can create a shifting-baseline problem: every generation treats the already diminished world it first encountered as normal. A deeper historical baseline makes long loss visible, but complete return may be impossible under present climate, land occupation and extinction. History supplies evidence and direction without automatically specifying a feasible destination.
The selection also determines whose history counts. Describing pre-colonial Australia as an untouched wilderness erases long-standing Aboriginal and Torres Strait Islander governance and relationships involving fire, water and living communities. Maintenance built on an imagined human absence begins with the wrong account of how the landscape was formed.
A more accountable programme states why its baseline was chosen, what evidence is missing, and whether it aims to sustain a species, a process, a cultural relation or a risk threshold. Where those aims conflict, a composite score should not be allowed to conceal the underlying choice.
Fire demonstrates that change can be a means of continuity
Australian fire ecology varies greatly by place. Some communities depend on fire within particular ranges, while others are damaged by frequent burning. Intensity, season, spatial pattern and time since earlier fire can all change the outcome. Neither “more burning” nor permanent suppression is a universal ecological rule.
Climate change further alters fire weather and operational windows. A regime that was workable in the past may not remain safe. Managers must consider human life and property alongside species and ecological processes. These purposes sometimes align, but not in every location or at every time scale. A treatment that lowers risk immediately around one asset may have a different landscape-scale ecological effect.
Cultural burning should not be treated as a portable tool extracted from Indigenous knowledge. The Australian State of the Environment material on Caring for Country describes Country through interconnections among land, waters, skies, life, culture and people, and emphasises Indigenous-led management and knowledge transmission. Particular practices belong to particular Country, law, communities and knowledge holders.
An agency that purchases an isolated “burning service” while withholding decision-making authority, protection of knowledge and a continuing relationship has obtained only a fragment. In this case the structure through which knowledge is maintained cannot be separated from the governance of its use.
Invasive-species control creates a continuing obligation
Invasive-species management makes the ethical boundary of maintenance especially visible. Some introduced species impose severe predation, competition or disease pressure on native species. Continuing control may be necessary to prevent extinction. Control may also kill animals, affect non-target species and consume resources indefinitely.
The labels native and introduced are not sufficient by themselves. Decisions need evidence about actual impact, feasibility, animal welfare, non-target effects and alternatives. A programme that produces only a temporary reduction before rapid rebound creates a recurring commitment. Its sponsors should identify who will carry that commitment, how outcomes will be evaluated and when the method will change or stop.
Climate change complicates older categories. Species may shift their ranges in search of suitable conditions. Such movement is not automatically equivalent to a damaging human introduction. Managers need to distinguish demonstrated harmful processes from ecological responses to a changing climate. Otherwise, a classification designed for an earlier distribution can punish adaptation itself.
These decisions also illustrate why maintenance is not necessarily benign. Sustaining one valued relation may require coercive intervention in another. Ethical scrutiny does not disappear when a programme has conservation aims. It becomes part of the definition of what responsible maintenance entails.
Global goals still require local judgment
The Kunming–Montreal Global Biodiversity Framework establishes global goals and 2030 targets concerning effective conservation and management, restoration of degraded ecosystems, invasive alien species, and the rights and participation of Indigenous peoples and local communities. It helps make the scale of loss and shared obligation visible. It does not determine how water should be allocated in a particular wetland or when a particular fire-adapted area should be burned.
Between a global percentage and a local action sit ecological evidence, legal authority, funding and negotiation. An area measure can count a line on a protected-area map while saying little about effective management. A restoration target can report hectares while hiding poor survival or an unsuitable species mix. Numbers are valuable for coordination, but they do not replace maintained relationships on the ground.
Local discretion, in turn, must not become immunity from evaluation. A project should record its objective, intervention, expected effects and uncertainties so that failure can be detected. Adaptive management is not the retrospective redefinition of success whenever results disappoint. It sets testable expectations before action and preserves the capacity to revise practice when evidence changes.
Who may decide what an ecosystem should become?
Environmental management can appear technical while carrying substantial value choices. Protecting a threatened species may constrain development. Restoring river processes can alter agricultural extraction. A measure intended to reduce local fire exposure can have ecological and cultural consequences elsewhere. Expertise can estimate some effects, but it does not by itself decide whose interests should bear the cost.
Affected communities, responsible public authorities, scientific workers and field maintainers require roles appropriate to the decision. Where Indigenous Country is involved, participation cannot end with an invitation to comment. Land rights, cultural authority, free, prior and informed consent, and governance of knowledge must be worked out through the relevant legal and living relationships.
Future generations and non-human life cannot participate as present-day stakeholders do. Environmental law, independent science and long-term monitoring partly represent those interests, but each can be weakened by budgets and political cycles. Maintenance institutions need ways to keep long consequences present in short decisions.
There is no single formula for allocating authority. Immediate operational choices may belong close to Country and field conditions, while irreversible public decisions require broader legal legitimacy. The important point is to disclose the allocation rather than presenting a political settlement as a purely ecological calculation.
Metrics can redirect maintenance away from its purpose
Environmental programmes need reportable measures: hectares treated, seedlings planted, animals removed or protected area declared. These figures help with budgeting and comparison. They also make easily counted activity more attractive than slow ecological outcome. Seedling numbers arrive sooner than evidence of a functioning community ten years later. Consultation events are easier to total than actual transfer of authority.
A measure should stay close to the result it claims. If the aim is ecological resilience, evaluation may need species and functions, landscape connection, recovery after disturbance and exposure to new pressures. If the aim includes cultural relationships, outside agencies cannot unilaterally define success.
Monitoring infrastructure itself requires maintenance. Sensors fail, survey methods change and funding interruptions break time series. People on Country may recognise changes in a waterway or season that a central model cannot interpret. The data system supports learning only if it can receive, compare and respond to that situated knowledge without appropriating it.
This is a recurring visibility problem. Work that produces a reportable event can dominate work that preserves observation, trust and response capacity. A funding model that rewards projects but not the conditions of long follow-up may repeatedly create restoration without maintenance.
A Sustenesis reading: continuity is not static identity
This article uses Sustenesis only at the applied level. It does not redefine the philosophy from an ecological example. It takes up a central question: how can something remain recognisably continuous while its components and surroundings change?
An ecosystem makes the problem unusually concrete. The continuity of a Country cannot depend on every tree, channel and population remaining numerically identical. It lies partly in relations, constraints, regenerative capacities and continuing practices. A structure that refuses all change may lose its ability to respond. Yet if every transformation is accepted as natural adjustment, maintenance loses all critical content.
Ecological maintenance must therefore identify which differences remain within tolerable transformation and which destroy the conditions through which the system can continue to form. Philosophy cannot supply that boundary alone. Species evidence, local history, cultural authority and public judgment constrain the answer. New climate conditions and better knowledge may require it to be revised.
The result is not relativism. It is a more demanding form of accountability. A programme has to state the continuity it values and the evidence by which it will recognise loss, instead of assuming that a familiar appearance settles both questions.
Current judgment: maintain the conditions of continued formation
Nature cannot be maintained by freezing a historical picture. Responsible practice should sustain ecological processes, relationships, diversity and the capacity to adjust under disturbance, while assigning responsibility for continuing human pressures.
Every programme should identify its object, baseline, time horizon and rights-bearing parties. Intervention requires evidence and conditions for correction or withdrawal. Inaction also requires an account of the historical forces it leaves in place. Where Indigenous knowledge is involved, governance and continuing relationships are conditions of valid practice, not ethical decorations added after technical design.
Maintenance may include restoring water flows, controlling demonstrated invasive harm, protecting habitat and re-establishing suitable fire relationships. It may also require stopping an engineering intervention and allowing room for ecological processes. No single posture is permanently correct.
The version 0.1 conclusion is that ecological continuity consists not in immobility but in the capacity of living relations to keep forming under constraint and disturbance. Good maintenance does not promise a return to an uncontested past. It makes choices visible, observes their consequences, and can change direction when it discovers that it has been maintaining the wrong state.
Primary sources and further reading
- Australian Government, Australia State of the Environment 2021: Caring for Country.
- Australian Government, Australia State of the Environment 2021: Indigenous outlook.
- Convention on Biological Diversity, Kunming–Montreal Global Biodiversity Framework.
- Australian Government, Australia State of the Environment 2021: Management approaches.
Series navigation: Who Maintains the World? series overview
Discover more from Geoffrey Chen
Subscribe to get the latest posts sent to your email.