I agree with you on all points, but I do think this still supports the point I was initially trying to make:
Classification of “patches” into biomes (in the true ecosystem sense) is in my opinion necessary but by itself insufficient for Thrive’s purposes. In order play a mountain theme in mountain “patches”, we do need to categorise which patches are mountains. This means we need a terrain classification either baked into biomes (as is the case in microbe stage) or a new separate system entirely parallel to the real ecosystem-based biome. So that would respectively be “Biome: Desert (mountains)“ or “Biome: Desert + Terrain: Mountains”.
The same goes for hills, plateaus, presence of rivers maybe, etc.
It’s the thing that keeps making me wonder if it would be better to have a microbe-stage-style patch classification determined purely by terrain + climate conditions as a “base layer”. The actual biome classification would then be a layer on top decided by the actual species present and ecosystem.
That would be in place of the multi-tier biome system where you define then “end biome”, with lower tier biomes replacing them if the right type of plants are not available. So this is also in response to HH:
The end result will probably be similar, since as @HyperbolicHadron mentioned, we’d be using the generator’s basic outputs to determine either the “potential biome” or the “non-biological patch classification“. And in the end a “tropical high-temperature high-precipitation” patch is still most likely going to end up with a “Rainforest” biome built on top of it, same as if you defined it as “potential biome: Rainforest”.
But mechanically, I feel more comfortable with a “bottom-up” definition than a “top-down” definition. It seems inherently more flexible to me. The top-down system of potential biomes with conditional downgrades seems a bit too rigid and linear. For example:
I agree designing this to be likely is probably a nice thing to do. But this is Thrive, so likely does not mean guaranteed. It’ll likely be at least possible for something grass-like to evolve before something tree-like does. What happens to the succession tier pattern when grasses exist but trees do not?
And do we really want to make it impossible for there to be a possibility of a “tree” species to evolve a way to compete more effectively against the “grasses” (for example, poisoning the soil with something the grasses have not evolved resistance to), thereby having forests displace some of the grasslands?
Another possible factor affecting biomes could be the fauna available. For example, I know that the presence of grazing animals is a major factor in maintaining grass/scrubland by preventing the growth of forests.
So again: I think your list of biomes is so far a very good one, but I don’t think it’s a good idea to define patches by their end-point “potential biome”, with a linear set of lower tier biomes leading up to it. I can imagine some potential problems already with just having earth-like life evolving in a different order, or with different competitive effectiveness. But what concerns me more is the problems we are not foreseeing right now.
So I would greatly prefer a system that at the base level defines patches by their geography and climate. Then let auto-evo decide based on those characteristics and the inter-play between species what type of vegetation actually ends up dominating. Then based on those results, we can select from your list of biomes. This seems more flexible to me, since I expect each “climate-geographical zone” to have more of a “tree”, “field” or “list“ of potential biomes, rather than a linear list leading up to a final ultimate biome, based purely on available plant species.
Seems like here you’re primarily talking about the climate-side (which makes sense since that is also what your (very interesting!) source is talking about! But I’m still thinking more about the biological side. Since ultimately, biomes to me seem defined by the ecology. Which is very strongly influenced, but not entirely determined, by climate.
Some things that still stand out to me:
-
You seem to assign all pre-shrub scrubland as (cold) desert. Now, shrubs are actually defined as still pretty decently sized woody plants. And as far as I can tell from a quick wiki search, those are a middle-Devonian development. But in the early Devonian (and later Silurian?), there were still a lot of small herbaceous plants. Rather than desert, I think this would be visually more similar to (alpine) tundra, rather than desert? (Please ignore the Prototaxites, they’re… weird)
-
Going even earlier into the Silurian and late Ordovician, we’re probably moving into seeing only moss and lichen (and algal/bacterial mats). Still tundra? Or is that Barrens? To us maybe, but to the arthropods we know were crawling around, probably not.
-
Speaking of algal/bacterial mats, which make for example stromatolites, they’re also around during some of the timeframes above, and before then. Afterwards, the evolution of other flora and grazing fauna prevented them from defining the ecosystem anymore. How would you classify a biome where just this stuff is covering everything?
-
Another problem could be that the plant-analogs may not be super plant-like? With the way auto-evo is going, you could just as easily end up with a main photosynthetic phylum that is built on soft-body cells or cells individually encased in silicate or carbonate rocks instead of the the cellulose-walled plants we all know and love.
I would expect that will change the calculus on for example growing tall, and how different climates affect different types of growth.
Which still all brings me back to “categorise areas based on geographic-climatic conditions,define biome from dominant species“.