In this wildly divergent Earth timeline, a group of diatoms evolved to be up to a few inches across (rather than 2mm or less). With this group presumably developing some similar traits to organisms like macroscopic unicellular algae and slime mold.
The diatoms in this group all share the following traits:

  • Frustules are impermeable glass, with small openings which can be sealed shut with a trapdoor like mechanism. This means a giant diatom should theoretically be able to survive passing intact through the digestive tract of a whale (since glass can resist hydrochloric acid).
  • Have anaerobic internal chambers which let them efficiently fix nitrogen from the air or water.
  • Most of the diatoms volume is made of vacuules for storage and regulating bouyancy.

Diatoms can fix nitrogen, and get other nutrients by regularly descending into deeper waters. The combination of traits described means giant diatoms aren't limited by the availability of nutrients in the surface water like other pelagic photosynthesizers, so they should have a massive pelagic biomass.

Am I missing some obvious reason why large floating glass photosynthesizers like this are implausible?

Given the surface area issue any living parts of the cell will be no more than a few mm thick. Many will grow into connected clonal colonies, like real life diatoms.

diatom colony

If giant diatoms are feasible, then could fish, marine mammals, and reptiles plausibly evolve to prey on them?


3 Answers 3


Probably, but you might have to make some compromises.

Macroscopic unicellular life has been well-documented, especially in marine environments, with examples including the foraminiferan Xenophyophora and the giant amoeba Gromia, as well as the algal Caulerpa you mentioned. However, these all kinda cheat by being coenocytic organisms; consisting of multiple nuclei enclosed in a single cell membrane. They get around the issue of nutrient absorption by either having a highly wrinkled and porous surface, or with a massive central vacuole that would take the function of a coelom in multicellular organisms.


To my knowledge multinucleate diatoms have not been observed — but given that this property has arisen multiple different times I don't think it's a stretch that this could happen in your alternate history. They might look quite different from the diatoms we're used too though, and require completely different cellular architecture. I imagine they might look something like a sphere or ovoid surrounded by a glass shell, with a large central cavity and multiple pores to allow movement of fluid into the interior.


I think you will face the problem given by the surface growing less than the volume with increasing the size.

The surface grows with $R^2$ and is what you have to exchange with the environment, while the volume, which grows with $R^3$, is what you need both to fill and clean by means of those exchanges.

While multicellular organisms can use tricks to increase the exchange area, a unicellular organism is pretty much limited in what can achieve.

I guess there is a limit after which increasing the size will make the organism less efficient at interacting with the environment.

  • $\begingroup$ Unicellular Caulerpa already gets big enough here, so does that mean you can get enough surface area if ensure no living part of the organism is more than a few mm from a pore? $\endgroup$ Apr 5, 2022 at 19:25
  • 1
    $\begingroup$ I don't know what a diatom is. But presumably if they have little hatches they can stick little cilia out of the hatches to increase surface area. How much bigger this gets you is a mystery. $\endgroup$
    – Daron
    Apr 5, 2022 at 20:17

Frame challenge (minor): You don't get crystalline silica or silica glass from biological sources. They both require high temperatures. There are enormous kinetic barriers to forming glass and/or quartz.

That said, silica gel and even plain old amorphous silica also resist virtually all acids. The latter is what diatoms basically are, so they could just grow a slightly more acid-impervious outer shell.

  • 1
    $\begingroup$ I suppose if glass sponges are anything to go by the composite glass structure will probably perform better on every metric than inorganically occuring glass. $\endgroup$ Apr 7, 2022 at 14:57
  • $\begingroup$ Glass sponges don't actually use glass, they use silica, an ordered form of amorphous silica. Yes, it negates the need for actual glass. No, it won't be more chemically resistant than actual glass; caustic soda will dissolve any amorphous silica relatively quickly, while silica glass and quartz will only dissolve very slowly. $\endgroup$
    – user86462
    Apr 7, 2022 at 20:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .