The problem is if you're not consuming extra trees to make science packs, you're going to make too many tree seeds, which clog up your system. The only way to get rid of tree seeds is to harvest them, which gives cellulose (easily disposed of) and oil, which is not easily disposed of. The only way I can see to deal with it in a reliable fashion is to use the oil on tree progenitors which are then consumed to make trees in the inefficient way. Oil cracking can be used to easily void oil by turning the oil to fatty acids and back.

On a separate note, making tree science is very annoying since you get waaaaaaaay too much wood and cellulose in the process, which isn't used for a lot of things.
...why does generating arthropods consume guns?

Woodworking can use wood to replace other resources. Cellulose can make graphite. What you don't need you can just burn. If your factory is still making graphite the old fashioned way it's just a logistics problem to switch. The tech for oil cracking comes immediately after the tech for trees. You're not likely to fill up a large tank with tree seed oil alone in the meantime, unless you have an extremely high tech multiplier.

Arthropods are not very cooperative and need stern discipline. Despite our best intentions, there are disputes.

The quantity of wood involved in making tree research is just too much, is the problem. I have my factory set up to make 0.26 tree science per second. (so that I make enough for the 10k Mecha 2 research) It produces over 40 excess BOXES of wood per second, most of which has to be disposed of. I've been disposing of over 50 boxes of cellulose per second, by turning it into graphite, which is eaten by a carbon sink. Using the graphite is kindof pointless since I can just make it from air.

I could understand needing a gun for that reason, but the gun is consumed. It's like the arthropods are eating the guns.

Spitters are particularly hard to manage. Equipment does need to be recycled on occasion. There is more than 1 gun, the 1 gun is just the maintenance.

The alternative to using byproducts productively is to add extra effort to dispose of them some other way. 0.26 bio science per second is comparable to 1500 SPM. A base that large uses a lot of graphite naturally. No reason to waste a bunch of space and energy in one spot making it from electrolysis and air separation and then use extra effort in some other spot disposing of unwanted graphite.

50 boxes of cellulose with the burning recipe instead of pyrolysis turns into 12.5 boxes of graphite, which your factory probably needs for that much science. Getting it there is a logistics problem. Then any carbon sink for graphite can be much smaller. Once you get woodworking, which you should have before you start making science based on the capstone tech requirement, then even less excess. That's assuming you use it though.

Bio science is definitely not 100 times more expensive. It's slower to produce, and is consumed slower, but it's not that expensive. Though, considering at this point I'm post-scarcity, expense is merely in how much room the infrastructure takes up. As far as that's concerned, it seems to be about 5 times as expensive, with my inorganic-science base taking up about twice as much room as my bio science base.
I went and checked and while consuming/producing about 344 red science per minute researching Braking 10 (and thus half that for the others) my best estimate at graphite consumption is 360 boxes/minute. Tree research was generating around 20/s aka 1200/m so I would still have to dump the vast majority. The other science packs consumed during 0.26/s tree research would be around 1/s of the relevant packs, which is 60 SPM not 1500.

But this is talking around the real issue here-- I DO NOT NEED THE GRAPHITE. I am making it already! Keep in mind that to even reach bio science you need to do carbon dumping. That means you have a lot of infrastructure lying around making mass amounts of graphite already. If you stop dumping carbon, then you have a lot of excess graphite production that doesn't require any logistical challenges to deliver to the rest of your base. Secondly, since you're making it from air and the associated structures are relatively compact, you can just copypaste it near where you're consuming it. Graphite is a trivial resource at this point, making the excess created from tree science purely an annoyance, and even if it could be used to supply my inorganic science costs, I'm generally not running my inorganic science at the same time as my bio science! In fact, my tree science pack production is done. I've stockpiled about 8k of it, after running it passively while I did other things. I won't be able to do the same thing with inorganic science unfortunately, since even boxed I'd need like a hundred large containers for each type to stockpile the millions of packs needed for the most expensive researches.

The amount of storage required to buffer conventional tech at megabase scales is an interesting point. It might make sense in the end game tech tier to unlock data compression for non-bio science, to compress boxes of science even further.

I'll look into the rate of wood production for science. But you're talking about one of the 2 most expensive techs in the game. And you're saying on the one hand that it's too cheap, because it's nowhere near as expensive as I'm saying it's balanced to be. And on the other hand you're saying that it's too expensive because you need a whole big subsystem for byproduct management. That's part of the cost too, moreso with biology than industry. It's not meant to be cheap and easy to unlock the optional post-victory tech goals, especially the very last ones. They're like infinite research in vanilla, just a bit more discrete.

The rate of conventional science consumed by bio science is not a good way to estimate the relative cost. The regular science component is meant to be a fraction of the overall cost, not to require your regular science to be running at full capacity. So you can buffer some regular science while working on bio science (which is why the scaling of storage requirements is concerning). So the equivalent SPM of biology is higher than the SPM it directly consumes. And yes, some of the cost is in the space the equipment takes up and the time it takes. A lot of biology comes from just air and water like other parts of your factory. Furthermore, your biology factory is meant to be smaller than your regular factory. There aren't as many biology researches so it's not meant to take up as much of your production time. You only need it part time, so even though the biology packs are considered expensive, the research costs aren't tuned to be crippling and require immense investment in science production that you won't be using most of the time. Instead of aiming for a high SPM target for biology, you can make it at a more gradual pace and switch it with non-biology research, which has a lot more tech options, including the infinite ones.

You're making graphite already using energy, space, and UPS. You can sunset parts of your factory that biology replaces. For example oil is more efficient for hydrocarbons, even though you can make hydrocarbons from nothing. Yes you're already making them the old way, but you can make your factory smaller and more efficient by doing it the new way, which may ultimately let you scale up your SPM and scale down your UPS. Biology is also making stuff from nothing, but faster, smaller, with less energy. If you don't want to do that you can, but it's not as efficient to keep making something that you're also making an effort to dispose of.

The goal of Nullius is to create planets similar to vanilla. The vanilla planet has coal deposits. So there has been some consideration of changing carbon sinks a bit so you can keep using them to automate disposal, but you can also making proper coal and get a coal deposit drone to create coal deposits required as a mission objective.

Okay, to be clear, the annoying part of the wood/cellulose byproduct production from tree science is that it's difficult to fit it onto belts, even on what seemed to me to be a compact factory block, only feeding two assembly machines at the end making tree science. Over 50 wood box/s means that I have to use the top tier belt to move it around. Green belts are extremely expensive even compared to blue belts due to the pipe 4 ingredient, so to save on costs I immediately split off all but 45/s onto a second belt that is always disposed of. The infrastructure needed to dispose of ALL the wood is actually not too bad, it's like 8 combustion chamber 3, which can run off the oxygen produced by the trees.
Now that I'm looking at it, my other bio science types don't fill up nearly as much of the square, so perhaps I oversized tree science. (by which I mean I should've made it smaller and then duplicated it in another square to ensure I had enough production)

Here's the tree science factory block: https://i.imgur.com/VmDhj3j.png (modified to run at half speed at most) It only has 6 combustion chambers so some of the load is handled elsewhere. I remove as much cellulose as I can via train instead of immediately burning it, to be put in an overfill storage area which itself has four combustion chambers to handle whatever I can't store/use. I think I'm using more cellulose than graphite...

Thinking about it, I came up with several things to potentially change, any one of which would reduce the wood produced during tree science production:
1, less wood and/or cellulose from the tree harvesting "wood box" recipe
2, less trees required per tree science pack (and maybe more tree progenitors per pack?)
3, less tree seeds required per tree, or more tree seeds per tree harvested (this would mean you have to harvest less trees during science pack production)

An alternate approach would be to make the tree science pack recipe much slower, to communicate it's not meant to produce as much per assembling machine as other packs, and resulting in a smaller support structure per assembly machine (since they work slower, and thus consume less materials over time)

Yes, I'll probably look into it and do something along those lines. I want wood to be quite cheap so it's a clear improvement to use the woodworking recipes to replace metal, even though with asteroid mining everything is "free". But some things could change around a bit, particularly the science recipe, to result in a bit less byproduct. At some point I'm also still planning to add a coal patch and oil well terraforming mission goals, which will greatly increase the demand for graphite and oil, but that doesn't preclude tweaking the research.

I love that guns are consumed.

The ratios of tree science are changed in the latest update, so there should be less wood byproduct. I think it's already pretty evident from the research counts that you're not meant to make as many SPM of biology as regular science, so I'm not changing anything about that. But managing surplus wood won't be quite as big a part of dendrology. Wood is meant to be cheap enough to make it worth fitting woodworking into your supply chain to replace some basic uses of metal, but it doesn't need to have negative value.