The Hidden Conversations Of Trees

AThe idea that a forest might be a communicating whole rather than a crowd of competitors has had an unusually rapid career. First proposed in the 1980s on the strength of a few glasshouse experiments, it reached popular books within a decade, and by the turn of the century a phrase describing the soil fungi as a kind of web had entered ordinary speech. The picture it conjures is a companionable one: older trees feeding sugars to shaded seedlings, damaged plants warning their neighbours of insects on the way, a wood behaving less like a market than like a household. Almost every element of that picture rests on a real observation, and almost every element travels further than the observation permits. Sugars do move between root systems, and volatile chemicals given off by a chewed leaf do alter the chemistry of leaves on plants nearby. The dispute is therefore not about whether anything travels, but about how the traffic should be described. What remains contested is who is doing the moving, who gains by it, and whether either process is usefully described as a message sent from one tree to another.

BMarta Ilves, a plant physiologist at the Tarnow Institute of Forest Science, argues that the airborne half of the story has been read backwards. A plant, she points out, has no nervous system, so a leaf attacked on one branch cannot alert the rest of the plant except through the air; the volatile compounds are, on this account, a message the plant sends to itself, and any neighbour that responds is eavesdropping rather than being addressed. Between 2011 and 2017 her group worked with 240 shrubs, wrapping the crown of each in a sleeve that let light through but kept air from moving between branches. The sleeves were taken off for two hours each week so that the plants could be inspected. Wrapped plants failed to raise their defences on undamaged branches after one branch had been cut, while unwrapped controls did so within a day. In the field the reaction of neighbours followed the wind: defensive enzymes rose by 38 per cent in plants downwind of a damaged shrub and by 4 per cent in those upwind of it. A signal aimed at a neighbour, Ilves observes, would not depend so heavily on the weather.

CThe underground half is harder to dismiss, since labelled carbon does travel from one tree to another, but Ilves regards the quantities as the crucial and neglected point. In a study of paired birch and fir saplings, carbon that had demonstrably crossed the fungal connection accounted for 2.6 per cent of the carbon budget of the receiving plant over a season, and rather more than half of the labelled material was still in fungal tissue when the trees were harvested. The saplings in that study were four years old and grew under a closed canopy. A fungus that draws sugar from one host and passes a fraction of it to another is not a postal service; it is an organism managing its own supply. Transfer of any measurable size, moreover, was detected only where the two trees were linked by the same fungal species and only where the receiving seedling stood in deep shade. Ilves is careful here. She does not deny that the carbon moves, nor that a shaded seedling may benefit from it; she denies that the movement was arranged by the donor.

DEwan Traill, who directs the Kinloch Forest Laboratory, maintains that experiments of this kind answer a question nobody in forestry is asking. Pots, sleeves and single seasons, he contends, cannot capture a relationship that develops over the life of a stand, and the only test that matters is what happens to regeneration when mature trees are taken out. His own trial, begun in 2009 and still running, compared plots in which selected old trees were retained with plots cleared entirely: after nine years, seedling survival stood at 41 per cent in the retained plots against 23 per cent in the cleared ones. Kinloch has since opened two further sites, though neither has yet yielded usable figures. Traill concedes, however, that his design cannot separate the fungal network from everything else a standing tree provides, including shade, wind shelter, leaf litter and a soil that machinery has not disturbed. He also accepts that no plot in the trial was given a fungicide treatment, which is the one manipulation that might have told the network apart from its surroundings.

EDecisions are meanwhile being taken on the strength of whichever version is believed. Retention rules in several forestry codes now require that a proportion of mature trees be left standing, a policy that Traill's figures support whether or not the network explains them, and a market has grown up in fungal inoculants sold to growers on the promise of connection. Products of that kind are not cheap. The wording matters here in a way that is easy to underestimate: a wood in which resources are traded by fungi and a wood in which trees care for their offspring imply different laws, different products and different reasons for public support. Most specialists now accept that both processes occur and that the fungus, rather than the tree, is the party with something to gain, while conceding that neither position has been tested at the scale at which forests are actually managed. The conversation, if that is the word for it, may turn out to be one the trees are not having.