Networks Allocate by Demand

01The evidence

• The functionality of arbuscular mycorrhizal networks across scales has been reviewed, with nutrient transfer depending on the physiological state of both partners rather than on gradients alone.Functional Ecology
• Narrative reviews of mycorrhizal networks and plant community dynamics have catalogued the evidence for directed transfer, including cases where movement runs against the gradient.Mycorrhizal networks review

02The tension

Nutrient movement is modelled with source-sink physics: material flows from high concentration to low. That explains some transfers and fails on the ones that matter most, such as phosphorus moving to a plant that already has moderate levels while a more depleted neighbour receives less.

03The theory

The fungal partner is not a passive conduit. It has its own physiology, and it responds to signals from the host, including sugar status and phosphorus demand. Allocation is therefore an active process that can be directed.

If allocation is demand-driven, then the pattern should follow the host's need rather than the concentration gradient, and it should change within hours when demand changes. That is a testable difference from source-sink physics, which predicts a fixed pattern set by concentrations.

The prediction is that transfer should run against the gradient in specific conditions, and that the pattern should be reversible when the demanding partner is removed. It also predicts that networks should be able to favour kin, which is the claim that has been hardest to test.

The applied consequence is that inoculation and network management should be designed around demand signalling rather than around abundance, and that the carbon cost of the symbiosis should be treated as a price rather than a leak.

04What this is not

05Predictions

• P1Nutrient transfer should run against the concentration gradient in defined conditions.
• P2Removing the demanding partner should reverse the allocation pattern within hours.
• P3Allocation should respond to the host's demand signal rather than to its size.

06What would refute it

Transfer following gradients in all well-measured cases would leave the network as a pipe.

07What it would change

It would make mycorrhizal management a signalling problem, and reframe the symbiosis as a market.