Gapped text
Mapping a network that was never one
In the oldest part of Renwick Wood, ecologist Tomas Vale stopped beside a fallen beech and pointed at what looked like a line of pale dust beneath the bark. He was leading a team mapping underground fungal networks, a phrase that encouraged visitors to imagine luminous threads connecting every tree in a cooperative system. Tomas disliked the image. Fungal filaments do connect roots, but connection does not by itself reveal what moves, who benefits or how long a relationship lasts.
37The project began with soil cores taken along a grid. Laboratory analysis identified fungal DNA, allowing the team to map where related sequences appeared. The first map was spectacular: coloured zones crossed paths, streams and the boundaries between old and recently planted woodland. Yet Tomas warned that a sequence detected in two places did not prove a continuous living network between them.
38To investigate continuity, researchers inserted narrow mesh bags containing material that fungi could colonise but roots could not enter. Months later, some bags showed matching chemical markers on opposite sides of a path. Others did not, although DNA maps suggested the same fungal group was present. The difference forced the team to separate presence from active connection.
39Weather complicated the picture further. A dry summer reduced visible fruiting bodies, leading volunteers to report that several species had disappeared. Soil samples, however, still contained their genetic traces. At the same time, some traces persisted after the associated filaments were no longer metabolically active.
40The team therefore added measurements of respiration and nutrient movement. In one area, carbon labelled in the leaves of young birch appeared later in fungal tissue near older oaks. The result attracted attention as evidence that trees were “sharing”. Tomas resisted the word. Transfer had occurred, but the experiment did not show intention, generosity or even net benefit to the birch.
41This caution did not make the findings less interesting. It redirected attention from moral stories about cooperation to the conditions under which exchange became possible. Soil moisture, root damage, fungal species and the age of surrounding trees all altered the pattern. A connection that redistributed nutrients in one season might carry disease or impose a cost in another.
42The practical test came when the estate planned to widen a path for fire access. The DNA map suggested a dense fungal zone beneath the proposed route, while the activity measures indicated that only part of it was currently linking roots. Moving the path would protect some active connections but cut through a different area likely to become important during wetter years.
43The final recommendation did not identify a single invisible network to preserve. It proposed a narrower path, seasonal restrictions on heavy machinery and monitoring after construction. The map became a record of different claims—presence, activity, transfer and future possibility—rather than a picture of one organism-like community. Tomas considered that less satisfying as an image and more useful as ecology. Sampling disturbed the system it described. Removing a core severed filaments and changed moisture around the opening, while mesh bags introduced an unusually concentrated resource. The team therefore treated each experiment as an intervention with a limited observational window rather than a transparent view of ordinary exchange. Repeated measurements at untouched comparison sites did not eliminate the effect, but they made it possible to ask whether a striking pattern belonged to the woodland or to the method used to reveal it. Scale further unsettled the network image. At the width of a root tip, exchange could be intense and short-lived; across the sampling grid, related genetic traces produced broad continuity; across decades, tree succession changed which partners occupied the same soil. A diagram could represent any one of these scales coherently and become misleading when its coherence was carried into another. The team therefore attached spatial and temporal scope to every network claim, preventing a local transfer from being narrated as the stable architecture of the entire woodland.