Not all of the beanstalks, though, had the same relationship with the fungi. In each mesocosm, one plant was surrounded by a mesh penetrated by holes half a micron across. Gaps that size are too small for either roots or hyphae to penetrate, but they do permit the passage of water and dissolved chemicals. Two plants were surrounded with a 40-micron mesh. This can be penetrated by hyphae but not by roots. The two remaining plants, one of which was at the centre of the array, were left to grow unimpeded.
Five weeks after the experiment began, all the plants were covered by bags that allowed carbon dioxide, oxygen and water vapour in and out, but stopped the passage of larger molecules, of the sort a beanstalk might use for signalling. Then, four days from the end, one of the 40-micron meshes in each mesocosm was rotated to sever any hyphae that had penetrated it, and the central plant was then infested with aphids.
At the end of the experiment Dr Johnson and his team collected the air inside the bags, extracted any volatile chemicals in it by absorbing them into a special porous polymer, and tested those chemicals on both aphids (using the winged, rather than the wingless morphs) and wasps. Each insect was placed for five minutes in an apparatus that had two chambers, one of which contained a sample of the volatiles and the other an odourless control.
The researchers found, as they expected from their previous work, that when the volatiles came from an infested plant, wasps spent an average of 3½ minutes in the chamber containing them and 1½ in the other chamber. Aphids, conversely, spent 1¾ minutes in the volatiles’ chamber and 3¼ in the control. In other words, the volatiles from an infested plant attract wasps and repel aphids.
Crucially, the team got the same result in the case of uninfested plants that had been in uninterrupted hyphal contact with the infested one, but had had root contact blocked. If both hyphae and roots had been blocked throughout the experiment, though, the volatiles from uninfested plants actually attracted aphids (they spent 3½ minutes in the volatiles’ chamber), while the wasps were indifferent. The same pertained for the odour of uninfested plants whose hyphal connections had been allowed to develop, and then severed by the rotation of the mesh.
Broad beans, then, really do seem to be using their fungal symbionts as a communications network, warning their neighbours to take evasive action. Such a general response no doubt helps the plant first attacked by attracting yet more wasps to the area, and it helps the fungal messengers by preserving their leguminous hosts.
Plant-fungus symbiosis is a surprisingly underexplored area of biology. The limited data available suggest most plants go in for it in one form or another, but its role is only slowly being illuminated. Work like Dr Johnson’s suggests this is a serious omission, not least for the understanding of how crops like beans actually grow. The underground world, though invisible to the human eye, should not for that reason be ignored or underestimated.