Length and thickness of bamboo internodes: a beautiful curve
elegantexperiments.net
elegantexperiments.net
Bamboo is a grass, it doesn't layer bark. It's one and done. The internode distance is pretty much fixed once it hardens. The bamboo cells inflate and harden. So the graphs make a reasonable amount of sense.
The wall thickness is a function of time to harden, and time to add extra stuff to the wall. Segments close to the ground have simply longer time, and likely hardening begets hardening.
The internode distance is interesting in that there is a natural point at which the bamboo stops elongating and begins to start shortening (again in a negative exponential). My understanding is that bamboo does start to keep "leaves" that sprout from the nodes, to gather light and energy. I wonder if that is the point that the leaves start, and hence it is a mixture of both time (less time = shorter) and investment in energy (effort to grow sun-catching leaves).
Some interesting followups would be in the natural habitat, where does the typical canopy start?
Doesn’t have to be driven by “natural habitat”, just the current environment. You are right though, the stalk wants to grow tall quickly (but not any taller than necessary, hence the successively increasing node distance in its “search”) until there is sufficient light, then it begins to disfavor further height growth in favor of increased density of leaves (nodes, where the leaves come from)
This is good stuff.
It’s always growing up, even as it creates a node. The way it works is the hormones shoot down the stem from the tippy top (where the growth is happening) and tell a lower part of the plant to create a node there (even as the top continues to grow up)
https://en.wikipedia.org/wiki/D%27Arcy_Wentworth_Thompson#On...
edit: Maybe I misunderstood? I'm a bit confused on width/length here.
Internode length increases up to the middle node (not sure if it’s also middle height), then decreases from mid-node to top.
Then internodes at the top are shorter and lighter.
I don’t know if the length pattern has some evolutionary advantage.
y(t) = V * [ 1 - e^(-R * C * t) ]
to
y(t) = V * e^[-R * C * (t-T) ]
since the weight of the plant is cumulative, it's interesting to think about the integral of these functions as they hold up the plant.