This guide is intended to serve as a reference for players stumped by the breadth of botany (Greenhouse and Arboretum) within the game. It has a number of pre-made production chains that can be used as a quick reference to identify immediate construction and production goals when this recipes first become available. It will note what production chains will be important later and thus worth properly setting up and which are functional dead-ends (as of Hot and Cold patch). It will also cover strategies for priming up relevant chains and early substitutes for chains which depend on later closed loops.

Because botany can become significantly sophisticated and requires more finesse in setting up logistics this guide also covers logistic management strategies and serves as a mid-level introduction to some of the logistics ideas that broadly apply to the whole of the game. Some strategies such as the use of spark launchers for bootleg botany with Loamy sparks, path-crate chains for Hauling sparks and Manifold design are covered.

This guide will, for hopefully obvious reasons, contain spoilers related to production facilities and recipes. It is inevitable that I talk about later recipes when introducing earlier ones as sometimes seemingly useless recipes are critical for later chains and its a good idea to build buffers earlier for ease of priming later production chains.

A buffer is any storage medium with input-output capabilities (both explicit storage such as crates but also implicit buffers such as buildings and trains). Buffers are not an in-game term and they are a concept in the sense of logistics setups. A crate whose inputs aren't being consumed is pure storage instead of a buffer. As astute readers will note, all buffers technically double as storage, but not all storage can function as a buffer, case in point, barrels.




Oversaturation refer to the state of production lines. An oversaturated production line has higher resource input flow than consumption. When a production line segment oversaturates its input buffer (either a building or intermediate storage) will start filling up as resources cannot be removed faster than they come in. Oversaturation can be symptomatic of production inefficiency (the resource is not being consumed as fast as its produced so one can make more buildings to consume said resource), they can be intended (such as in Manifold designs), or they can be indicative of production imbalances (a recipe using 2 resources is missing the other resource). Oversaturation is visible as resources "backing up".


Balancer setups (in contrast to Manifold setups) explicitly divide traffic flow (and thus resource flow) amongst a number of lanes. Balancer setups allow granular control of traffic at the cost of significant space requirements. Balancers do not have a warm-up period but without saturation (or sufficient buffering) they pose risks of seizing up on self-feeding setups because of incomplete stacks in all machines that, if concentrated would have enabled an additional cycle that would complete the feeding proccess. Balancer setups are usually implemented ingame using the Splitter Path Add-on



Because balancers have shared path areas, traffic volume is capped by the common area. Exceeding the traffic volume of the common area risks deadlock on the merger lanes. This means that high-balancer ratio setups have a mandatory traffic flow inefficiency in that the individual lanes cannot be fully utilised because the common path area would collapse.



Balancer setups with Hauling Sparks are possible, but they require even more area and buffers.


A manifold is a logistics setup where a group of machines are all fed in immediate succession with no rationing or filtering. Manifold setups are more space efficient than balancer setups and approximate peak efficiency better than balancers, on the downside manifolds have a warm-up period before which they are less efficient (for some cases, 0% efficiency until warm-up depending on the setting).




Manifolds can operate at max traffic without risk of self-gridlock and are generally easier to manage, for large production setups with a high number of machines manifolds are practically mandatory.


Manifolds are fed sequentially, this has a particularly important implication. Machines earlier on the sequence are more likely to be fed (in fact the first machine is guaranteed), this is of particular importance for botany where self-feeding loops require such a guarantee for their continued functioning. However its not always possible to build a manifold such that a specific machine sequence is fed. Because of the right-hand rule of traffic the rightmost machines are natural candidates to be the high priority targets. Conversely leftmost machine priority is hard to manage with manifolds and sometimes require path self-intersection.


As mentioned before manifolds have a warm-up period. Because a manifold feeds machines sequentially, and machines themselves have buffers (and you might have additional buffers before said machines), follow-up machines don't get fed until the first saturates. Because saturation does not increase efficiency past the point the machine starts operating this means that a manifold needs to "fill up" before successive machines come online. This might be a positive if you are progressively building the production chain, but in many cases it means that you won't get the efficiency you'd expect from the number of machines you have built. For Greenhouse and Arboretum setups, because of how self-feeding operates saturation will take a VERY long time. Because of priority management this means that you won't see outputs AT ALL for a long while until the manifold saturates the self-feeding segment of the mechanism.

A way around this is priming. Priming is simply the act of bringing items out of storage to intentionally oversaturate the machines in a manifold. This might seem pointless or counterintuitive, after all, it'd involve reducing production elsewhere simply to obtain equivalent production on the line of interest. However note that priming new lines often involves resources one has had access for a long while, while the new line might be a recent unlock. Under that lens, it makes sense to have storage piles for items you know you will want for priming new lines down the line.

Because Plant Extractors have multiple outputs, and production stops if EITHER resource backs up, manifolds have problems dealing with plant extractors even with careful planning. In most cases the output of an extractor will have a priority resource needed for self-feeding and one that is a "bonus". The first immediate solution is the use of compressors for the non-priority resource.


More complicated setups include using this resource on secondary production lines and storage. But its always a good idea for there to be escapement systems for excess of non-priority outputs (compressor or otherwise) even with secondary production lines.

Stone paths are 150% faster than dirt paths. You want to automate this ASAP. A buffered setup with a storage chest at the end is particularly efficient method of having it up and running early since Wooden Log byproducts are cheap but Storage Chests aren't.



Hauling sparks are 125% faster than any other sparks and they carry whole stacks. You will want to automate their creation and use them liberally wherever possible. The ability to carry whole stacks is particularly important when dealing with some aspects of Botany.





Hauling sparks requirements of closed road loops might seem problematic at first but can easily be assayed through the use of crate-broken segments. Keep in mind that this naturally induces buffers so manifold setups where sections of items being fed back might require priming as a result of the use of this pattern.


Much can be said about trains, they are alluded to in my dedicated guide to trains. You will want them for botany starting with the fact that the Arboretum is stuck on the Tundra Plateau. Once you start dealing with temperature (which you very much want to do because of Greenhouse production rates) it becomes even more vital to your production chains.

When dealing with the volcanic area, which does not allow trains, the shared path pattern becomes a quasi-necessity if dealing with significant shipping inputs into the area. By deploying exactly as many hauling sparks as the capacity of a large crate, its possible to ensure deadlock free shared pathing through the use of sensors. Because of space and wiring considerations the limit for resources that can share pathing is 8 per large loop. Each resource is given a dedicated large crate for input and output. The input crates are gated behind a one-way filter, the output crates are observed by a capacity sensor. The capacity sensor is set to only activate when the crate is not empty (Absolute, >= 1). Conversely the one-way filters for an input crate are set to activate to the output of the crate sensor. This ensures that even if all sparks where carrying an item, the output cannot overflow. The exact number of hauling sparks on this loops is 20 (2 x 10 item stacks).



Fundamentally if you have a pure loop it risks one resource clogging it because of deadlock. If you have resources ABCD but B and C are only consumed when there's enough of A, and there's too much B then there will never be A to be consumed with B, and thus you will be stuck with B filling your loop into eternity.



To avoid any resource from overflowing, we use logic to manage the ability of sparks to pull or push resources into the loop. For this we need to use path filters which require additional space on the loop.



We can differentiate each crate as a different resource through path addons (Left side is input, right side are the receivers). The reason we use one way on the inputs is because sparks will reach that point without a resource at hand, so we have to control what crates are accessible based on logic



The receivers (right side) have a capacity sensor that writes to a ribbon (through a Ribbon Writer) if they are not completely empty. That ribbon writer is set to the "corresponding bit". You pick what bit each resource is. In this case Coal is 0, Leaves are 1, Glowshrooms are 2 and Stellar Leaves are 3.
The Ribbon Readers then read the corresponding bit on the other end of the system
in this case, the leftmost crate is reading bit 0 (coal), followed by bit 1 (leaves) followed by bit 2 (glowshroom) and bit 3 (stellar leaves).

This means that when the signal is set, that crate cannot be accessed by sparks on the loop,
stopping its ability to push resources into the loop.



The setup shown isn't logistically optimal, its better to leave a space between crates so that you can have loops pushing and pulling into the crates, but this was necessary to make it fit in a single screenshot.

Notice that we only used one of the path lanes in our loop, that's why the limit is 8 resources
you can have a ribbon on each lane of the loop going through the cave entry. Its also why we use ribbons, because they let us cram 4 logic wires into one tile.

To achieve a 24 feeder Arboretum the resulting layout ends up with very stringent requirements on 8 of the feeders, namely that they have to be fed by a single path (and one in turn ALSO has to extract the outputs). The solution to this is the compact shared path pattern. It avoids deadlock by having dedicated sparks for each leg (a single spark, even if not using Hauling Sparks) and using logic to control the alternation. The counters are set to tick over at 2 (the minimum). To activate it, you run into the pad that will activate the side where the arrow is pointing (in the 2-leg case the one opposite to the arrow's facing). One of the major downsides of this pattern is that sparks following you (or yourself) can break it by triggering the pressure plate when moving around, so caution is advised.



The triple (and more, though at that point your throughput is so abysmal its not really usable since you have 1/4th the volume the path would otherwise have ... with a single Hauling Spark) setup is simply the setup for 2 but instead of connecting to each other each counter connects to the next leg (and its own reset) and the last leg connects back to the first.



Each leg contains a pressure plate, a counter and a one-way path add-on. Its counter is ticked by the pressure plate of the previous leg and reset by its own pressure plate, and the counter output connects to a NOT Gate that connects to the one-way path add-on for the leg.

The design is relatively simple and shouldn't require wire bridges. Something important to note is that while the pressure pad and the one way arrow should (ideally) be contiguous they DO NOT connect to each other. Additionally its ABSOLUTELY IMPERATIVE that the whole thing goes counter-clockwise otherwise it will break because the path splitter add-on will not match the order of gates opening and misroute sparks coming back.

Broadly speaking when setting up Greenhouses for self-feeding you will want to set the priority buildings on the right side because of how pathing works.


This isn't always possible and there is a workaround with an inefficiency induced by the non-self-fed output following the non-priority order.



Another useful pattern to know is that a single temperature control building can affect up to 16 greenhouses simultaneously. The pathing and input feeding is somewhat sensitive to travel times, in many cases improper setup might limit it to 13-14 functional greenhouses.


Arboretums are entirely dependent on their geographical surroundings. In general a ring of feeders is going to be as close as you can get to maximization (without some very exotic logic setups). Pathing is an open problem. Input flow rings are one option if you have sufficiently flat terrain.


Input rings into crates for segments is a compromise for when the feeders are in awkward spots.



Sometimes there is just not enough perimeter so one can build "broken rings". Somethign to keep in mind is that ideally, when breaking a ring, one should actually have two halves flowing in clockwise and counterclockwise direction and meeting in the opposite end. This is because on crate-linked broken rings the travel time for a segment might exceed the consumption rate and thus leave a section of the ring unsupplied.



Something to note about production in an Arboretum is that the Arboretum itself is irrelevant, what really matters is the number of feeders. The actual production building is the number of Arboretum Feeders when calculating production ratios. As a rule of thumb you will normally be able to achieve between 12 and 16 Arboretum Feeders.



Given sufficiently flat terrain, dual pathed feeding can go up to 16 feeders. In theory 18 is the "normal" dual path upper limit and requires VERY specific geography to be plausible (read: Cannons), is is possible given sufficiently flat geometry and a very specific loop pattern and 24 is the absolute upper limit and requiring some utterly absurd level of planning due to path sharing being required for 8 inputs (and one of those combines with the output) on top of having to supply the 16 other feeders at the same time.



One such example inside the game (non-functional because of the lack of space to properly feed the northern segment).



Manifold Spark[github.com] is a tool I developed ENTIRELY because of Botany. It is a GUI Graph Node editor for planning out production chains in Oddsparks. Visuals for production chains are all taken from this (freely available) tool and I (biased as I am) highly recommended using it later on when recipe complexity goes up and you have a number of dead-end production lines that can be repurposed for botany.




Manifold Spark operates on the principle of substitution. You select a final output product and then can replace inputs by using right click to substitute the input production chain for any alternative.





Its possible to save production chains and export them to a file.





Saving a production chain allows it to be substituted in other production chain.




Additionally, when substituting a user made production chain in, its possible to do so in Closure. The closure of a production chain pipes in outputs back as inputs (self-feeding mode). When imported as a Closure the production chain is treated as if it was a single machine.




Serializing allows having the production chain listed in a text file.




Its possible to serialize the closure. It will identify self-feeding situations and deduct them from the final output calculations, as well as indicate instances of self-feeding.




To be able to re-use saved production chains its necessary to export them.




And they can then be reimported.





An important feature for later production setups is Mergers. A Merger indicates that the inputs for a machine come from two flows. Right-click merger will produce a Merger node, which can then have both of its inputs substituted. The Merger node controls the ratio between net inputs from both production chains and adjustst them accordingly.






A number of premade recipes can be found here, all the non-trivial recipes in this guide included.[github.com]

The following are the self-feeding ratios for Greenhouse recipes. Ratios are read as "How many of this same kind of building could this building feed if its outputs were piped back in". So for instance a ratio of 1:3 means that 1 instance of the building could feed 2 additional ones on top of itself (3 buildings fed total), meaning that in self-feeding mode its output can be adjusted by a factor of 2/3, or conversely, if you plop 3 of the building, one can be dedicated to self-feeding while the two others are used elsewhere. Priming amounts are the minimum to finish a single cycle given a single greenhouse in self-feeding mode. This is simply 10x the amount to finish a greenhouse cycle. Its recommended that one actually have x1.2 the priming amount so that no time is lost while the greenhouse outputs are converted back into usable inputs as outputs aren't always compatible inputs. When priming larger setups its advisable to even prime the entire machine setup and not just the priority feeders.

Coral Seeds are listed twice, once without the use of Fluted Coral as an intermediate step and once when it is added. There are caveats to this production chain that are discussed in more detail on its relevant section which explains why this ratio is not a straight improvement. First the recipe WITHOUT using Fluted Coral as a magnification step is listed, then the recipe implementing the intermediate Fluted Coral Step.

Stellar Leaves are listed twice, once when the Stellar Ice is also fed back and once when its used as an output instead. There are reasons to want to store or use said Stellar Ice elsewhere and conversely reasons not to. First the recipe WITHOUT consuming the Stellar Ice is listed, then the recipe consuming the Stellar Ice.


This are the actual recipes with intermediate building counts, set so that the self-feeding ratio ends up with integer ratios on the greenhouse count, to get the number of dedicated self-feeders, multiply by the ratio:

One of the things you will want to do as soon as possible is jam Ore Miners on every column you can, and path them all into some colossal storage area. While copper mining will have pitiful numbers after the drill exhausts the deposit (it will never fully deplete), a lot of the botany early and midgame requires copper and its better if its already extracted and sitting in barrels rather than having to wait for it to be mined. Ore miners are relatively cheap and you can ignore them after setting them up. When I say "colossal" I truly do mean as many large barrels as you can physically fit, having 5k+ Copper Ore sitting around on barrels might seem absolutely absurd but it will save you a lot of time and effort in several quests.

Additionally, when the recipe becomes avilable you are going to want to pipe copper ore to furnaces for copper seeds. Ideally making 500 to 1000, while the greenhouse copper seed recipe will not produce excess seeds or ores, the item that DOES produce excess (the copper cutting) requires 500 copper seeds and its good to have excess for priming (since you will want a few loops of self-feeding so that it can properly "explode" into a bonanza of copper (to clear some very annoying late game tasks) and ideally have at least two copper sapplings going on. The ideal recipe is converting the Copper Sap to Copper Seeds to then convert it to Copper Ingot rather than going directly from Copper Sap to Copper Ingot (only do this after you feed the Copper Seeds back to the Copper Sappling though).


The humble Leaf Knot might seem catastrophically wasteful. It makes leaves, takes leaves and takes a lot of time to basically produce 1/16th of what a logger would produce in the same amount of time (when in self-feeding mode).

However looks are deceiving. Having a self-feeding source of leaves you can plop anywhere is a way of activating teleporters far away from distribution lines. Additionally its one of the inputs of Copper Sapplings (so you will want a healthy supply for priming that production line). It can be used together with Aether Flowers to produce Loamy Sparks locally anywhere on the map. Finally on late and post-game builds transporting leaves is inordinate because of volume, leaf knots are an entirely leaf-input-based product that can in turn be converted back to leaves after being shipped around.

Leaving a single logger feeding 2 cutters and 4 greenhouses running leaf knots early on to build a supply to then prime later chains is not a bad idea.



For feeding Copper Sapplings you will want Logger based supply chains simply because of how inordinate self-fed setups become at those volumes. In this case a total of 4 trees feeding 16 greenhouses.


Conversely for more conservative uses of leaves including Rope for Artsy Sparks, Fabric for Puffy Sparks, and Fertiliser for Loamy Sparks, self-fed setups are perfectly functional.


Another seed that comes SIGNIFICANTLY EARLIER than its intended use. Coral Seeds main use is not to make coral. They are a primary input in the Geode Cluster production chain. They are (incidentally) the more volumetrically demanding of the input pair for Geode Clusters in fact. Coral Seeds are particularly well suited for being grown on the west side of the map, in the border between the deep forest on the outer ring past the mountains and the start of the tundra area. You will always want to operate Coral Seeds in self-feeding mode.

You will want to have a disgusting supply of Coral Seeds before you need them. Geode Clusters might look like a colossal pain to deal with (requiring filters to clean up Geode Breaking) but being the sole source of machine automated Volcanic Soil makes it a VERY VERY desirable product. And for that you will need Coral Seeds. Having a priming (and quest clearing) supply early on is near mandatory specially because breaking Geode Clusters requires relinquishing the first you get to the Mason Sisters, ideally you want to have 800 Coral Seeds sitting around at that point 400 for the quest and then 400 for actual production, though if you (wisely) hook up Coral Seed production to the Arboretum you can forgo this initial supply (though no reason not to build it early and let it accumulate before its needed).

Keep in mind that the volumes for Geode Cluster production are pretty aggresive, Coral Seeds might seem an efficient recipe in self-feeding mode but the raw volumes demanded are nothing to scoff at.



Though it is of note that a byproduct of this production chain is sandstone, if you are making significant volumes of it, its not a bad idea to have Glass Vials coming out on the side.



Outside of Geode Clusters, Coral Seeds can also be used to make Fluted Coral which can be sawed down to more Coral. The astute reader will wonder if its more efficient to directly self feed the Coral Seeds back to the Greenhouses or to go through Fluted Coral. And the answer is the second.

Fluted Coral doesn't significantly reduce the number of Loggers or Greenhouses required but it does astronomically shrink the number of Seed Extractors required. A 20 Arboretum Feeder setup requires 30 Plant Extractors operating on a Pure Coral Setup, in contrast, if using Fluted Coral this number shrinks to 1.3.

The chain does gain marginal complexity but it's worth the effort simply because of the savings on the astronomical number of Plant Extractors. The added number of Stone Cutters is minuscule but their temperature requires might prove onerous, in practice it is more convenient to simply build more of them (since the number required is sufficiently small) then bothering to fully temperature manage them to full efficiency (it is fine to use a Heater but a Temperature Regulator is wasted on them).

The true final boss of Oddsparks. At first Stellar Leaves looks like a relatively amenable recipe, its byproduct Stellar Ice multiplies fertiliser and its self-feeding ratio looks pretty good. But do not be fooled, this single ingredient is going to eventually become the source of any and all your production related misery.



Stellar Leaves look innocuous at first sight, they require Stellar Seeds which are annoying to gather, but then its self-feeding ratio ensures a healthy supply in perpetuity. Except that Stellar Seeds AREN'T the self-feeding hazard here, its Volcanic Soil. Stellar Leaves forms part of both of the most flexible and complex production chains in the game. First for Geode Clusters (which are the only machine based supply chain that produces Volcanic Soil) and later for Aetheric Apples. Because Geode Clusters also produce Glowshrooms, Geode Clusters feed BOTH of its inputs but not with clean (or deterministic for that matter) ratios.

You will want to make a safety storage of Stellar Seeds in case your production lines ever go awry and dry up because scavenging for stellar seeds is an exercise in frustration. The alternate way of rebuilding a supply is using Stellar Ice sources (Frozen Rocks and Logs into Stellar Ice and then Ice into seeds).

This is a production line that you will have to jumpstart with gathered Volcanic Soil (either from gathering it yourself or from the spark thrower setup recommended above, the earlier you get it going and building a Volcanic Soil buffer, the better). And which later on you will have to significantly expand. Geode Clusters are not particularly demanding since they only require Stellar Leaves tangentially but Aetheric Apples have strong demands on Stellar Leaves because they require both Stellar Leaves and Fireshroom Clusters (which require Stellar Leaves for production).



Stellar Leaves HAVE to be self-fed, there are no other automated means of producing Stellar Seeds, you will likely want to use the Stellar Ice for even more seeds. Stellar Ice is needed for Freezing Sparks, but you will very rarely ever need more than 30 built (perhaps a few more if you lose them to the lava) and the Stellar Ice for them can be swiftly gathered from alternate gatherable inputs.


Fireshroom Clusters are the reason Stellar Leaves are so inordinately complicated to factor into production chains. They require Glowshrooms and Stellar Fertiliser. Stellar Fertiliser requires Stellar Leaves. In a twist of production nightmares, Fireshroom Clusters in self-feeding mode are EXTREMELY inefficient (4 buildings feed 5, meaning self-feeding requires 5 buildings to match 1 that is being externally fed), but in a cursed twist, their self-feeding mode has a byproduct of Lavacaps, which go into Geode Cluster production, which in turn produces Glowshrooms and Volcanic Soil. As a result Fireshroom Cluster production scales on itself in very weird ways, particularly when you are readying for the input onslaught that is Aetheric Apples in the endgame.

Fireshroom Clusters CAN be externally fed from the outputs of both Loamy Spark Thrower farming and from Geode Clusters, this two feeding sources SHOULD NOT BE DISCOUNTED. Because of the inefficiency of self-feeding, Aetheric Apples will have inordinate infrastructure needs if you don't include this two into your potential inputs.

However your first setup will have to be self-fed. It's a good idea to keep the Fireshroom Clusters that are the excess of this self-feeding in significantly large storage for priming Aetheric Apple production much later into the game. In Fireshroom Cluster state they also can be converted quickly into Glowshroom and Lavacaps if needed but the reverse is not true.




Later on you will have much more complicated hybrid setups where you have SOME self-fed Fireshroom Clusters and some externally fed from the Geode Clusters the first is already feeding. Manifold prioritization becomes a key issue if you don't have dedicated Fireshroom Cluster farms for the Omnigeode separate from those that are in Hybrid setups.



My personal favorite production chain. Friendly self-feeding ratio, obsoletes Miasma Vials for spark production (until you need Aether Crystals), can be plopped near literally anywhere, both of its outputs are extremely valuable, very easy temperature management, gotten pretty early in botany. What's not to love?

You will likely start building this all around on your own. Aetheric Seeds are vital for Aetheric Apples, so do keep them in storage for priming later on, but they are also extremely fast to produce so its not as critical as all the other inputs I've underlined needing priming before.

Self-fed mode is the only way this is going to operate (unless you are feeding it from a Rock Teron or Beelephant Farm, which is unreliable as of the Hot'n'Cold patch because of Boss Monster respawn rules), but that's ok, the ratio is decent.



Not much more to say, the best production chain in the game, I'd make a gigantic LED display for it if I could.

It makes Volcanic Soil. This will feel VERY underwhelming at the time you get it. It has some pretty stringent inputs, the ratios are a PITA because of the Fireshroom Clusters, and it produces geodes which you then have to break down, twice, and has random outputs (but the volume is so absurd that it will, statistically, approach the averages). But then you will unlock Aether Apples (and what they can be used for) and you will DESPERATELY want to have Geode Clusters going because you NEED Geode Clusters to sustain Aether Apple production, Spark Throwers do not suffice.



Geode Clusters require Geode Breakers, you will need at least 4 because of how Geode Breakers operate. One for the Cluster itself, which then has to be fed to 3 Geode Breakers, one per Geode type, ideally 3 per geode type (so 9). You will then have to separate the outputs (Glowshrooms, Coal, Copper Ore, Volcanic Soil) before shipping them where relevant. Of note is that the separator need not be close to the Geode Cluster production. Shipping Geode Clusters is significantly more efficient, and if you are using the majority of the outputs in a single location its better to break it near where its actually needed.



Something to be careful about is that you will have to escape Geode Breaker outputs, otherwise a single output can jam up the entire production chain, Geode Breakers cannot work while their tray has an item so your logistics chains should account for this.

The Copper Ores are quite useful to complete priming Copper Cutting production if you didn't start with miners early. And if you have the space, consider also storing them for the final quest.

Fireshroom Cluster Closure and Coral Seed Closure numbers are the number of Greenhouses in self-feeding mode inside a self-feeding setup.



This only exists to prime the Copper Cutting production chain. Self-Feeding has a ratio of 1:1 meaning they do not actually multiply your resources. Do not bother with this after you've already gotten enough Copper Seeds to fully prime your first or second Copper Sappling.

The source of efficient Copper. Copper Sapplings more then supply their own production. Be EXTREMELY careful about ensuring that the output loop back is prioritizing the sappling otherwise you will have to re-prime it and anything except another Copper Sappling will be an efficiency loss.



Since Copper Ingots are a prerequisite for Handy and Drilly, odds are you don't need me to sell you on why you want a Copper Sappling. Its inputs are not hard to sort, Leaf Knots (which I recommended you start stockpiling back when they were introduced) and Copper Seeds (which the Copper Sappling chain supplies after the initial priming).

It lets you create Ancient Bases. You already have the ability to place teleporters in Ancient Bases. Therefore, this lets you create teleporters anywhere. It lets you make sparks anywhere. The things you can do with this apple are basically what you have wanted to do since the very beggining of the game, and likely went to the developers discord or email to demand it be added to the game.

Well, here it is. And the price is your soul. The inputs are Liquid Fertiliser and Aether Seeds. Aether Seeds are generally simple to procure. It comes as a byproduct of the Aether Flower production process. You will more than trivially be able to satisfy this half. The issue is THE OTHER HALF.

Liquid Fertiliser requires Fireshroom Clusters and Stellar Leaves. Fireshroom Clusters requires Stellar Leaves. The volumes of Fireshroom Clusters involved are titanic. You will need a SIGNIFICANT number of Greenhouses producing Fireshroom Clusters, ideally a hybrid between pure and self-feeding. Because self-feeding requires MORE Greenhouses, and more Stellar Fertiliser, this means that self-feeding isn't always a better choice, if you can, use the Glowshrooms produced by the Geode Cluster producing Volcanic Soil. But have self-feeding as a secondary "emergency source" in case Glowshrooms dry out from said input stream. As for how to balance both input streams, Manifold Spark is your best bet, the previously linked guide covers using the pre-made recipe as well as how to introduce self-feeding cycles into larger recipes and how to use the splitter node.

Liquid Fertiliser production is, in itself another hassle. The temperature management (which you will DESPERATELY WANT at this point in the game) requires input shipping all across the map for optimal production. Stellar Leaves normally require importing into the volcano for Fireshroom Clusters and if you are also importing Geode Clusters, you will notice that the real estate in the mouth of the cave will start becoming an issue. Additionally shipping it out is a problem because of how space inefficient it is.

The reward for all of that is basically the panacea of great things, the Aether Apple. By the point you have this automated (which you actually can complete the game without fully automating if you do the final mile by manually combining existing storage by yourself) you have basically completed the game. And hopefully you listened to my advice about saving up on Copper Ore for the final quest.




This is the monstrosity I used, I do not advise copying this exactly but tailoring it to your existing production chains.