The Keeping of the World

What the Marsh Remembers

There is a place I return to each autumn, when the light has thinned to something the colour of old linen and the reed beds have begun their long bronze conversation with the wind. It lies in the borderlands between Carmarthenshire and Pembrokeshire, where the land has never quite decided whether it belongs to the terrestrial or the aquatic, and where that indecision has produced, over millennia, one of the most quietly extraordinary ecosystems these islands possess. I go there to listen, which is not quite the same as going there to hear. Listening requires a particular quality of attendance, a willingness to be instructed by what presents itself rather than by what one has anticipated finding. The marsh, in my experience, is always several steps ahead of whatever understanding I bring to its edges.

What strikes me first, arriving each year, is how completely the place refuses the human appetite for clarity. A marsh is neither land nor water, neither fixed nor fluid, neither silent nor loud in any register we easily interpret. It operates in gradients and thresholds. The ground beneath your feet has a memory encoded in its very texture: press down and the peat releases water that fell as rain decades, sometimes centuries, before your birth. That water carries within it chemical signatures, pollen grains, spores of mosses that no longer grow in this part of Wales. The marsh holds all of this with an equanimity that feels, to the attentive visitor, less like indifference than like a very deep patience. It has been keeping records far longer than we have been asking questions.

Peat bogs and coastal marshes are, in the strictest ecological sense, among the most faithful archivists the planet possesses. Their anaerobic conditions — the low-oxygen, high-acid environment created by saturated soils and the slow decomposition of sphagnum and sedge — preserve organic material with extraordinary fidelity. Pollen analysts working on British bog cores have reconstructed the entire vegetational history of the post-glacial period, watching forests of pine and birch colonise bare tundra, giving way to the great mixed woodlands of the Atlantic period, then retreating as Neolithic farmers cleared, burned, and grazed. The marsh watched all of this. It recorded the arrival of plantain — Plantago lanceolata, the weed that follows human disturbance like a shadow — whose pollen appears in bog profiles precisely at the moment human agriculture takes hold, a floral breadcrumb trail left by cultures that kept no written word. There is something almost unbearable in this, the idea that the land itself has been witnessing our entire tenure here, noting our interventions with the same dispassion it applies to volcanic eruptions and shifts in oceanic current.

But I do not want to speak of marshes only in the register of the scientific, though science is the instrument by which we have learned to translate what they say. What interests me here is something both simpler and more elusive: the quality of memory that a living wetland embodies, not as metaphor but as ecological fact. When I stand on the boardwalk at the edge of the fen and watch a bittern move through the reeds — that impossible creature, striped like weathered timber, which tilts its head skyward and becomes, in an instant, invisible against the vertical stems — I am watching the product of evolutionary time so vast it defeats imagination. The bittern's camouflage is memory of a kind: the accumulated experience of every ancestor that survived by standing still in exactly this way, in exactly this kind of light, among exactly this kind of vegetation. The reed bed remembers the bittern. The bittern remembers the reed bed. They have shaped one another across spans of time that make human history look like a brief interruption.

This is one of the things conservationism asks us to sit with, and it is not comfortable. It asks us to apprehend a timescale in which we are genuinely peripheral, and yet to accept that our peripheral status does not excuse us from moral responsibility. Quite the opposite. The very brevity and recency of our dominance over the natural world — and the speed with which that dominance has translated into ecological devastation — means that we carry a weight of obligation disproportionate to our tenure. We arrived late. We disrupted comprehensively. And we are only now, in the geological equivalent of the last few seconds, beginning to understand what it is we have been dismantling.

The marsh, in this respect, is an uncomfortable teacher. It shows you abundance and loss simultaneously, often in the same square metre of ground. Where I walk each autumn, there are plants that were once common across the lowland wetlands of Britain — marsh helleborine, grass-of-Parnassus, the pale lavender spires of water mint — that now survive in fragments, in pockets, in the margins of places that have themselves become marginal. The greater part of the wetland that once stretched across this valley was drained for agriculture in the mid-twentieth century, a process so rapid and so total that most people living nearby have no memory of what the land looked like before. Drainage is among the most violent acts we perform upon a landscape, precisely because it is invisible in its operation. You do not see the water leaving. You see only the changed surface, the tidied fields, the absence of the particular unruliness that water on flat land produces. But underneath, in the disrupted peat, the absence is profound and structural. The archive has been interrupted. Pages have been torn out.

And yet. The marsh that remains — that stubborn, peripheral, ecologically impoverished remnant — is still doing the work of marshes. It is still sequestering carbon, still filtering the water that moves through it toward the estuary, still providing the structural complexity of stem and root and mud that a hundred specialist species require. The reed warblers still come in May, threading their nests between the stems with an artisanal precision that has not changed since reed beds first colonised these latitudes after the last ice. The otters still move along the ditch margins at dusk, their sinuous passage leaving a chevron of ripples that widens and smooths long after they have gone. Life, in a marsh, is extraordinarily persistent. It presses into every available niche with what I can only describe as insistence — not the triumphalist insistence of conquest, but the quiet, non-negotiable insistence of things that have been doing this for a very long time and intend to continue.

I think often, when I am there, about what it means to call a place degraded. The word carries a moral charge that we apply uneasily, because to call a place degraded is implicitly to name the agent of degradation, and that agent is always, in the end, ourselves. But degraded does not mean dead, and it does not mean finished. A marsh that has lost sixty percent of its species richness over a century is impoverished, certainly — but it is also evidence. It carries, in its surviving communities, in the chemistry of its remaining peat, in the behavioural adjustments of the creatures that persist within it, a detailed account of what has happened and, crucially, of what still functions. The remnant is not merely a shadow of the whole. It is also a seed of it, if we can learn to read what it is telling us and respond accordingly.

There is a particular quality of light that comes in the late afternoon on the marsh in October, when the sun is low and raking and everything throws a shadow ten times its own length. The seed heads of the common reed catch it and become momentarily luminous, a moving silver just above the dull bronze of the main stems, and the whole landscape seems to breathe in a way that makes the distinction between animate and inanimate feel not merely arbitrary but wrong. The reed bed is breathing. It is transpiring water into air that will condense tonight as mist over the lower fields. It is fixing carbon drawn down from the atmosphere into the tissue of new stems that will, in autumn, fall and join the slow accretion of the peat. It is providing thermal insulation for the roost of starlings that arrives each dusk in a murmuration that makes the sky above it look temporarily alive with a new form of intelligence. All of this is happening because the marsh is still here, still doing what marshes do, still remembering in the only way a living system can: by continuing.

Conservationism, at its most fundamental, begins with this: the recognition that the living world has a memory older, deeper, and more consequential than our own, and that the destruction of habitat is not merely an environmental loss but an act of erasure. When a marsh is drained, the archive is not simply interrupted. It is, in many of its most complex and irreplaceable dimensions, destroyed. The pollen record ends. The specialist communities built up over centuries are dispersed or extinguished. The hydrological relationships between soil and atmosphere and plant and creature — relationships that took ecological time to establish and that we are only beginning to understand — are severed. What remains is the memory of what was there, carried in the minds of those who knew it, growing fainter with each generation until it, too, is gone. The marsh does not forgive or condemn this. It simply becomes something else, or becomes nothing. It is we who must live with what that means.

The Weight of Witness

There is a particular quality of grief that arrives not with noise but with the slow recognition of absence — the way you might notice, only after some weeks of mild bewilderment, that a familiar sound has ceased. The reed bunting that once threaded its thin, reedy song through the morning air above the fen; the lapwing that tumbled and keened above the water meadow in April; the small tortoiseshell that made a blazing point of colour against the old stone wall in September. These are not merely sentimental losses, though sentiment is not the dishonour we have been taught to believe it is. They are ecological facts, measurable and documented, catalogued in population indices and breeding bird surveys with the cold authority of numbers. And yet between the number and the lived experience of loss lies an enormous and largely unacknowledged distance, a space in which something essential about what conservation actually demands of us goes consistently unspoken.

To witness is to be present to what is happening without the consolation of turning away. It is a harder discipline than it first appears. The naturalist, the ecologist, the farmer who has walked the same field boundary for forty years — all of them carry a knowledge that is simultaneously privilege and burden: the knowledge of what was, held against the evidence of what remains. This temporal layering, this capacity to hold past abundance as a living standard against present diminishment, is among the most ethically demanding aspects of caring for the natural world. It is not passive. It is not merely the accumulation of loss. It is a form of active moral attention, and it asks things of us that our culture has not yet found adequate language to describe.

The ecologist Peter Kareiva, writing with uncomfortable candour about the limitations of conservation science, argued that the movement had too often retreated into a posture of mourning, cataloguing decline in terms that left the general public feeling implicated but helpless. There is something in this critique worth sitting with, even for those of us who find his broader conclusions about so-called "new conservation" troubling. The witness to ecological loss can become, if not carefully tended, a form of paralysis — a witnessing so saturated with grief that it forecloses rather than enables action. And yet the corrective cannot be to look away, or to reframe every loss as an opportunity, every degraded landscape as a blank canvas for human reinvention. The corrective must be something more difficult: a witnessing that holds grief without being consumed by it, that transforms the weight of what has been seen into the discipline of sustained attention and, ultimately, into responsibility.

I think often of the ornithologist and writer W. H. Hudson, who spent long years in Patagonia and the English countryside developing what can only be called a practice of witnessing. His observations in Birds and Man and Afoot in England, written in the early years of the twentieth century, carry in them the ambient anxiety of someone watching a world begin to accelerate away from itself. Hudson understood, even then, that the naturalist's role was not simply to describe but to remember on behalf of those who did not know what to remember. He bore witness to the diversity of bird life in the English countryside with the urgency of someone who sensed that he was writing a record that would one day read as elegy. That it now does so is a fact that should discomfort us more than it typically seems to.

This question of what witnessing requires is not merely philosophical. It has direct consequences for how conservation is practised, funded, communicated and imagined. Consider the so-called "shifting baseline syndrome," a concept introduced by the fisheries scientist Daniel Pauly in 1995, which has since become one of the more illuminating and troubling frameworks in ecological thinking. Pauly observed that each generation of fisheries scientists tends to calibrate their baseline of what constitutes a healthy fish population against the conditions they first encountered at the start of their careers — conditions that were already depleted relative to those encountered by the generation before. The result is a slow, generational amnesia in which catastrophic long-term decline is systematically invisible at any given moment because each new observer compares present conditions not to some objective historical baseline but to the recent past as they personally experienced it. What once seemed abundance becomes, within a generation or two, a remembered anomaly rather than a standard. And what was depleted becomes normal.

The implications of this are quietly devastating. They mean that our capacity to understand the true scale of ecological loss is itself subject to erosion — that the very consciousness needed to motivate and sustain conservation effort degrades over time in direct proportion to the losses it ought to be responding to. This is the cruellest possible mechanism: a grief that unmakes its own witness. To set it in motion requires nothing more than the ordinary passage of time and the replacement of one generation of observers by the next. No malice is needed. Only forgetting.

The writer and naturalist Robert Macfarlane has spoken movingly about the loss of what he calls a "language for landscape" — the attrition of the specific, locally rooted words by which particular ecological features and phenomena were once distinguished and named. When the words go, something of the perceiving goes with them. The child who has no word for "loaning" — the old Scots term for a lane through which cattle were driven — cannot quite see what the adult who grew up with that word sees when she looks at a certain kind of field boundary. This linguistic erosion is not merely a cultural curiosity. It is a form of perceptual loss, a narrowing of the aperture through which the natural world can be witnessed and therefore known and therefore loved and therefore defended. Language and witness are inseparable, and both require constant renewal if they are to remain instruments adequate to ecological attention.

What I am reaching towards here is something about the nature of conservation as a practice rooted not only in science and policy — though it requires both, urgently and rigorously — but in a specific quality of human attention that must be cultivated, sustained and transmitted across time. The ecologist is only part of the story. Equally necessary are the writers, the artists, the farmers who have watched the seasons turn across the same ground for decades, the beachcombers who notice when the wrack-line changes, the hill walkers who remember the blanket of bilberry that once covered a slope now stripped to heather monoculture or bare peat. All of these people are witnesses in the sense I mean, and all of them carry knowledge that the technical apparatus of conservation science cannot fully capture or replace.

This is not an argument against science. I want to be clear on that point. The longitudinal population studies, the satellite tracking of migratory species, the soil carbon analyses, the acoustic monitoring arrays that now allow us to hear a forest and assess its biodiversity from the signature of sound — these are extraordinary achievements of human ingenuity placed at the service of the living world, and they deserve both celebration and vastly increased investment. But science records what witnessing makes visible to begin with, and witnessing requires a certain disposition, a certain willingness to look and to be affected by what is seen, that must be separately cultivated. The two are complementary, and neither is sufficient without the other.

There is also a social and political dimension to witness that demands consideration. Whose witnessing counts? Whose testimony about the state of a landscape, a river, a hedgerow is accorded authority? The history of conservation is not innocent of power — has often, in fact, reproduced and even amplified existing inequalities in its decisions about which landscapes merit protection, which communities are included in the design of that protection, and whose knowledge of local ecological conditions is treated as credible evidence. The Indigenous land managers of Australia, the small-scale fisher communities of the North Atlantic seaboard, the crofters of the Scottish Highlands who have maintained traditional practices of low-intensity land use for generations — all of these groups carry forms of witnessing, accumulated across lifetimes and generations, that conservation institutions have historically struggled to accommodate within their frameworks of expertise. This is changing, slowly, in some places, as the field of conservation biology develops more sophisticated understandings of local and traditional ecological knowledge. But the change is not moving at the pace the ecological emergency demands.

To witness well, then, is to witness widely. It is to extend the circle of testimony, to cultivate the humility to recognise that knowledge of a landscape may arrive through many different instruments of attention — through childhood memory and agricultural practice and Indigenous protocol and scientific method alike — and that no single mode of knowing contains the whole of what a place holds. The weight of witness, taken seriously, asks not only that we look but that we listen, across whatever distances of culture and class and geography separate us from others who are also attending to what the world is becoming.

And it asks something more intimate still. It asks that we allow ourselves to be changed by what we see. The naturalist who has watched a fen drain, a meadow ploughed, a hedgerow grubbed out to accommodate a larger turning circle for agricultural machinery — this person has witnessed something that, if received fully, cannot leave them unaltered. The question is what form that alteration takes. It can curdle into helpless rage or retreat into protective numbness. Or it can be refined, with difficulty, into a quality of attention so precise and so committed that it becomes its own form of devotion: a refusal to look away, a refusal to accept the normal erosion of standards, a refusal, above all, to stop knowing what the marsh remembered and the meadow once held. This is the weight that witness asks us to carry, and it is, I believe, among the most necessary burdens of our time.

Tending What Remains

There is a word in Welsh — cynefin — that resists clean translation, as the best words always do. It means, approximately, habitat, but the approximation fails almost immediately. Cynefin carries within it the accumulated weight of belonging: the place where a creature knows itself, where a sheep knows which hill is hers across generations, where a person's bones understand, without instruction, the particular slant of light in October, the precise temperature of a certain river, the smell of a hillside after the first rain of autumn. It is not merely a location but a covenant — the ongoing, largely unspoken agreement between a living thing and the ground that made it. To lose cynefin is to lose not only a place but the self that was calibrated by that place. And it is this loss — intimate, structural, often irreversible — that those who tend what remains are truly working against, even when the work is framed in the cooler language of ecology or carbon budgets or biodiversity net gain.

To tend is a quieter word than to save. It lacks the drama of rescue, the clarity of conquest. And yet it is, I think, the more honest verb for what conservation, at its most serious, actually demands. Saving suggests a single act: a hand extended, a thing pulled back from the edge. Tending implies duration, repetition, the willingness to return and return again without guarantee of reward. It implies a relationship rather than a transaction. When I think of the people I have known who have devoted their lives to this work — the warden who has walked the same coastal headland every morning for thirty years counting choughs, the botanist who crawls on her hands and knees through Caledonian remnants looking for a lichen that may or may not persist, the farmer in Carmarthenshire who has sacrificed yield to keep his field margins unsprayed and his hedgerows broad — I recognise in all of them this quality of faithful return. They are not heroes of a single dramatic hour. They are practitioners of an extraordinarily patient love.

And the world they tend is not pristine. This is perhaps the central discomfort that conservationism in our era must finally reckon with honestly: there is no original, uncorrupted nature to restore. The ecosystems we call ancient are themselves the products of long human entanglement. The wildflower meadows that ecologists rightly cherish as havens of invertebrate life were, from a deep-time perspective, created by clearance, by centuries of low-intensity grazing, by the particular poverty of peasant farming that lacked the capital for improvement. The upland blanket bogs of Wales and Scotland are in part artefacts of Neolithic and Bronze Age deforestation. The English wood pasture tradition, which produced the great veteran oaks now draped in rare lichens and home to stag beetles and lesser spotted woodpeckers, grew from medieval systems of common land management. Nature and culture are not opposite terms. They are, and perhaps always were, a braid.

This does not diminish the urgency of protection. It deepens it, because it means that what we are conserving is not a world apart from human meaning but a world saturated with it — a world that carries within its living tissue the memory of how people once moved through it, used it, understood themselves in relation to it. When a traditional hay meadow is ploughed out and reseeded with rye grass, something is lost that cannot be measured only in species counts. Gone with the hay rattle and the green-winged orchid and the ox-eye daisy is a particular set of human relationships to that ground, a vocabulary for it, a set of practices that encoded a kind of knowledge that took generations to develop and cannot be reconstructed from a textbook. The conservationist who fights to maintain such places is also, whether they articulate it this way or not, fighting to keep alive certain possibilities of human understanding — certain ways of knowing and being known by the land.

The organisations that do this work operate under formidable constraint. In Britain alone, the landscape of conservation is vast and complicated: statutory bodies whose mandates shift with political weather, wildlife trusts working county by county with chronic underfunding, private landowners whose motivations range from genuine ecological commitment to what can only be called reputation management, rewilding projects that attract both genuine excitement and fierce controversy, farmers caught between the economics of survival and the genuine desire to steward their land well. The policy environment has changed significantly since Britain's departure from European agricultural frameworks, with new schemes attempting to reward public goods — clean water, biodiversity, carbon sequestration, flood mitigation — rather than mere production. Whether these schemes will deliver at the necessary scale remains, at this early stage, genuinely uncertain. What is not uncertain is that the need is vast.

Consider the numbers plainly, as they deserve to be considered. The UK is among the most nature-depleted nations on Earth — a striking and sobering fact given that it contains some of the world's most passionate and long-established conservation traditions. Roughly half of its native species have declined since 1970; one in six are threatened with extinction. Ancient woodland, which took centuries to develop its particular complexity and which cannot, despite what some consultancy documents suggest, be meaningfully offset by new planting, has been reduced to a small fraction of its former extent. Upland rivers run with agricultural sediment. Chalk streams — a globally rare habitat, found almost exclusively in England — are over-abstracted and damaged. Seagrass meadows, which support juvenile fish, cycle nutrients, and store carbon with extraordinary efficiency, have declined by over ninety per cent in the past century. Against this catalogue, the response, though real and in places genuinely encouraging, can seem achingly inadequate.

And yet the inadequacy must not become an excuse for despair, because despair, in this context, is simply another form of abandonment. The moth recorder who has watched his local site shrink and the species list thin over four decades does not stop recording. He adjusts his methodology, publishes his data, writes letters, turns up again in August with his trap. The data he accumulates is not only a record of loss; it is a resource for recovery, a map of what was here and therefore of what might, under different conditions, return. This is the quiet heroism of the citizen naturalist: not the expectation of triumph but the refusal to stop bearing witness, and the understanding that witness itself is a form of tending.

There is something important, too, in the recent shift — still contested, still working itself out — towards thinking about restoration not only of individual sites but of functional landscapes, of connected corridors through which species can move, adapt, recolonise. The logic is compelling and well-supported by ecological science: isolated fragments of habitat are inherently fragile, vulnerable to local extinction events, to the gradual erosion of genetic diversity, to the accumulation of edge effects. Connectivity — the ability of hedgehogs to move between gardens, of dormice to travel through hedgerow systems, of freshwater invertebrates to recolonise rivers from intact source populations — is not a luxury but a necessity for long-term viability. The ambition of projects like the Nature Recovery Network, or the Lawton Report's memorably plain instruction to make habitats bigger, better and more joined up, is essentially an ambition to think at the scale of ecology rather than the scale of ownership or administrative convenience.

This requires a transformation not only in land management but in how we understand the relationship between private ownership and public obligation. Land in Britain has always been held within a web of rights and responsibilities, though the balance has shifted dramatically in the centuries since enclosure towards an understanding of ownership as near-absolute. The argument for nature recovery is, among other things, an argument for a different settlement — one that recognises that a landowner's decisions about drainage, pesticide use, tree cover, and field management have consequences that extend far beyond their title deeds, downstream into the catchment, outwards into the atmosphere, forward into the futures of people not yet born. This is not a comfortable argument for everyone. But discomfort is sometimes where the necessary conversations begin.

I think often about what it means to tend something you may not live to see flourish. The oak planted now for the veteran tree of 2300. The river restored by removing a weir, knowing that the full return of migratory fish to its upper reaches may take decades. The bog rewetted by blocking drainage channels cut in the 1960s, its sphagnum mosses recolonising millimetre by millimetre across a surface that will not reach its former depth of peat for centuries, if ever. These are acts of what one might call deep time generosity — gestures towards a future one cannot inhabit, made on the basis of a belief that that future matters and that the world will be better with these things in it than without them. There is something in this quality that resembles, more than almost any other human endeavour, the spirit of genuine love: attentive, undemanding of immediate return, willing to be faithful to something whose full value may never be visible to the one who tends it.

My grandmother kept a garden on the edge of a coal-scarred valley in south Wales, in ground that had been contaminated and cleared and left and slowly, painfully recolonised. She grew vegetables, yes, but she also kept corners deliberately wild — a damp patch she never drained where marsh marigolds came every spring, a dry-stone wall left loose and unmortared so that the lizards could bask in its crevices on summer mornings. She did not have a word for what she was doing. She would have been puzzled by the idea that it required one. She was tending what remained because what remained deserved tending, and because the act of tending it kept her connected to something she needed to stay connected to — something that had nothing to do with the economy of the valley or the noise of the century, and everything to do with what she called, simply, being in the world properly.

That phrase has stayed with me across all the years since, through all the reading and the walking and the long hours in hides and on headlands and beside rivers, watching, noting, trying to articulate what it is that the living world asks of us. Being in the world properly. It is not a scientific formulation. It does not map easily onto policy frameworks or funding streams or species recovery targets, though it can, I believe, quietly animate all of them. It is an ethical orientation, a posture of attention and responsibility that recognises the world not as a resource to be managed but as a community to be served — a community of which we are members, whose other members have claims upon our care that exist prior to their utility to us. To tend what remains, in this spirit, is not to concede defeat, not to be content with fragments and remnants, not to stop arguing for more and pushing for better. It is to begin, each day, where one actually is: with what is here, still alive, still holding, still asking, in its patient and extraordinary way, to be known.