Mycelium – the underground fungal network that connects trees in a forest – is nature’s own DLT 1. When designing machine ecosystems, mimicking the principles of these natural networks give us a good head start in creating resilient autonomous assets. We’ve looked at aspects of natural ecosystems as an inspiration before, but this time we dive deep into the inner workings of interaction in ecosystems.
This post looks at two things; how and why these natural networks work. Looking at how they work are for example resources and information shared between individuals and species, and how does it react when circumstances change. We will also consider why it works: the system is characterized by communication, coöperation, altruism and long term self interest. But most importantly it is designed for adaptability; not for a predicted, optimized demand/supply scenario 2. Nature’s own DLT makes it possible for slow-growing, immobile trees to adapt to a constantly changing environment.
Forest networks – How they work
A forest is bigger than the trees and plants you see growing around you. Underground lies a vast network of strain-like fungal cells. A fungal network called mycelium 3. According to fungal scientist Paul Stamets 4 every step in a forest influences about 300 miles of mycelium5.
Resource trading between trees and fungus (1)
Trees transfer sugar6 to the fungus and receive water and nutrients such as phosphate and nitrogen in return 7. It is unclear to what extend this exchange is a trade with a set price and you only get as much as you pay for, or if the system is built on long term reciprocity and benefits for the ecosystem. It is more likely based on abundance, giving what you can to the system and receiving back what you need when you need it. This keeps the system healthy and balanced.
Big trees share energy with smaller trees (2)
Bigger trees that get more sun transfer sugar to smaller, shaded trees. There is a preference for the tree’s own offspring, but the whole system get a piece of the pie; even other species. There’s a big incentive to do this. The alternative would be competing for the sunshine. If trees didn’t share, each individual tree would try to grow as tall as possible as fast as possible to get the most sunshine. The problem is that this makes them less sturdy and thus more vulnerable to strong winds as well as sickness and bug attacks. In optimizing one factor, the trade-off is usually losing quality in other aspects. In the longer term this could be hurtful, or even devastating.
Seasonal energy sharing between tree species (3)
Seasonal resource transfers also take place between leafy trees8 and conifers9. Leafy trees receive energy in the winter and autumn from evergreens like conifers. In the summer leafy trees are more efficient at photosynthesis and transfer energy back to the conifers.
Altruism – nutrients to all (4)
Mycelium transfers nutrients to all trees in the forest. What I understand of the current research it does not seem like every tree is bargaining for its own resources, negotiating a sugar for nutrients price. The mycelium gets the sugar it needs from the trees that have an excess and transfers nutrients to all trees. It seem this is not done as purely bartering, but that a more resilient forest is better for its individual agents, regardless of their individual capacity to generate energy for themselves.
Trees share information (5)
Trees exchange information in addition to sharing resources. Trees under attack can release airborne chemicals10 in the form of acids and alcohol that work as a defense against the attacker, e.g. larvae, insect. These chemicals can a deterrent for the attacker or an attractor for the attackers natural enemy. The chemical also serves as communication to other trees. It both tells them an attacker is present and how to deal with it. Trees further use enzymes as part of their defense system. These defense signals are shared with other trees when triggered. This way younger trees learn from the experience of older trees on what defense to use.
Dying trees can rapidly transfer their legacy across generations and species. Research shows that an injured tree dumps its energy (carbohydrates) and wisdom (defense signaling knowledge) in the network to the benefit of other trees. It even does this to benefit other species 11. If conditions in the forest change due to for example climate change, a dying tree will even give a boost to a new species – one that is better adapted to these new conditions.
Hub trees give access to a bigger network (7)
Hub trees are crucially important for the network. A hub tree is an older, bigger tree that is connected to a lot of other trees through mycelium12. They are important for the connectivity in the network. Without them, many smaller trees would not be connected to each other. Having access to the network gives access to more resource and information transfer. By chopping down the hub trees the whole forest suffers from the loss of knowledge about defense strategies and loss of connectivity. It seems a forest runs like a resilient distributed network, combined with an efficient more centralized network on top for longterm knowledge and efficient routing.
Ready to adapt to changing circumstances (8)
Holes in canopy are a huge problem, because even a sturdy tree can only take so much wind. Like herd immunity, a closed canopy protects all trees – strong an weak. When a tree falls ill and dies, the other trees rush to help a smaller tree grow and close the hole. It seems one of the reasons big trees share resources is to have candidates for such unfortunate events ready. Smaller trees are not necessarily young, but can be “backup trees”. They can patiently wait for decades to get their chance to grow and claim their place in the forest canopy13. A healthy ecosystem seems to have a lot seemingly wasteful resources lying around in case of emergency.
Compatibility between networks (9)
The mycelial network is comprised of many species of fungus15. So it’s not just about selfishness from the individual trees.
Or is it just about power?
Could the fungus be the mastermind behind it all? Both the trees and the fungus benefit from the exchange of resources. It is possible that the fungal network distributes resources and information to keep the forest as strong and healthy as possible for selfish reasons: a healthy forest is better for the fungus. But this contradicts that hub-trees send more resources to seedlings of its own DNA. The whole concept of a mastermind seems more like an anthropomorphism 16 if anything.
Are we destined to share?
Could the driving force be altruism instead of selfishness? They do share information and knowledge. This is a good example of a positive sum economy: more trees gain knowledge, but the tree sharing does not lose its knowledge.17 In technology this can be compared with open source sharing.
Or are we born selfish?
Even the altruism can to some extent be explained by self interest – by sharing knowledge on defense signals the whole forest is healthier. Not sharing the information could give a relative advantage. But that could be penny wise and pound foolish. A healthier forest is more resilient against big threads like storms, floods, draught, fire, and countless other attacks when its inhabitants look beyond short term self interest.
Or is it all of the above?
Or it could be something in-between. Even if there is no short term gain in sharing, in cooperation there is usually a long term benefit for both parties. The forest ecosystem has characteristics of all of the above. Regardless of whether plants have feelings or can make plans, what we see in nature is that life prevails. Ecologists agree that the guiding force is a higher underlying principle that is common in nature: complex adaptive systems.
Next: complex adaptive systems
Complex Adaptive Systems are found throughout nature, on every scale. From bacteria to planets, and everything in between – complexity is found everywhere and explained with this theory. While CAS have various definitions, they basically consist of a large numbers of participants called agents, that interact and adapt or learn. Over many iterations they form complex patterns and tend to find a balance, resulting in a resilient ecosystem.
We see CAS as a perfect inspiration to create resilient machine ecosystems in the near future and hope to expand on this soon™.