« Annoyed | Main | Still on the road, sort-of »

On the road yet again (and again) ...

This Thursday (the day after tomorrow) I'm off to sunny Oslo, Norway, for The Oslo SF Festival, where I'm one of the guests of honour over the weekend.

I'm back Monday night, but off again on Thursday (that's next Thursday) to Novacon 39 in Nottingham for the following weekend. Back mid-week.

When I'm not bouncing around Europe like a demented flea, I'll be working on the next novel ("Rule 34") and hopefully on the copy edits to "The Fuller Memorandum".

So that's two conventions, a CEM to check, and I'm writing a novel. If you want a reason why I'm not going to be strangling kittens in public posting provocative essays on my blog for the next couple of weeks, those should be sufficient.

Meanwhile, though, a question arises from the previous two discussions, and it is this: what is the minimum number of organisms one would need to create a biosphere capable of sustaining human life (for example, aboard a deep space mission too prolonged to subsist solely on canned goods)? Just to scope it out a bit further: I'm assuming you want a stable minimal ecosystem drive by sunlight-equivalent illumination. I'm not concerned with its size, so much as its complexity. What aspects do we need to consider? (For the sake of simplification, let's hand-wave cosmic radiation and solar flares out of the picture for now.)




Where's the sunlight coming from in a "deep space mission"?

You want a biosphere not a colony-pod type thing, right? Algae + humans + energy (sunlight). Algae process sunlight > humans eat algaeburgers (yum!). Water + air are assumed also. Recycle human waste + waste humans. Work for you? Can't see why you would need anything else.


Charly: Broken link to OSFF.


Kevin @1: Algae doesn't have all the nutrients you need. It's lacking a whole bunch of vitamins.

So unless you're going to assume some advanced genetic engineering, you'd need some other plants. A citrus plant, for example. Some legumes as well, maybe.


kevin: "sunlight equivalent illumination" is not automatically sunlight. It could be radiance supplied by daylight-spectrum lamps powered by a nuclear reactor; or it could be something else. It's just a convenient way of putting energy into an ecosystem.

You need a metric buttload of extra stuff beyond just algae. (In fact, "just algae" is the first-order simplistic approximation I was kind of expecting to hear first. It's a common misconception among the space cadet types who seem to think that biology is an irritating detail of space flight, rather than the elephant in the littlest room.) There are a couple of orders of magnitude more prokaryotic passengers (and some eukaryotes) riding around in your gut than there are cells in your body, in numerous species (which have interesting effects on how well you digest your food); how are you nurturing and recycling them? How are you ensuring that none of these strains of microscopic passengers run riot elsewhere in your food chain? What are you using in your anaerobic digesters to deal with faecal waste? Where are the vitamins coming from? What do you do about the roughly 5 million (dead) skin particles each human passenger sloughs per 24 hours period?

"A citrus plant ... some legumes as well" is barely beginning to approach the actual complexity of the problem.


Well, the only thing even close to trying to set up a self-contained biosphere which we have done yet is Biosphere 2, which of course was a rather dismal failure. I tend to think that they went quite overboard on the complexity. (I mean, trying to build 6 ecosystems, some fairly incompatible with the others, in less than 10000 sq. m?? Really?) Better to build one ecosystem really, really well, to my mind. But I'm waiting for heteromeles or one of the other biology people to show up and talk about it with far more knowledge than I could ever muster.


Biosphere 2's initial failure lay in the concrete -- it acted as a huge CO2 sink, forming calcium carbonate as it cured.

Hint: it's not the only experiment along these lines (although in scale alone it was the biggest so far).


Wasn't Biosphere 2 an attempt to do just that?

And even with that size of biosphere they had problems maintaining oxygen levels on the 1st mission. - And then the project disintegrated in a management mess - good job it wasn't somewhere really hostile, like mars or LEO.

I have a feeling a closed ecosystem capable of maintaining a long term crew is going to be a) quite big and b) more complex than we imagine.


@6: Well, I was following the other thread, so I knew that the concrete had been a problem (I even read the wikipedia article you just linked before I posted). I was just expressing my view that it might have been a better idea to try to implement one ecosystem on a large-ish scale and really well than 6 with very limited space and poorly.

I knew there had been other biosphere experiments, but I didn't know of any that had the same focus that Biosphere seemed to have (that is, supporting human life in an artificial habitat).


I'm on the panel with you at http://www.osff.no/ this friday - The "Disasters on a galactic scale" gig. Looking forward to seeing you there. As for sunny? Well, yeah... ;-)

By the way: I get a "Comment Submission Error - invalid request" when trying post without first previewing.


Darn, you guys type faster than my interweb connection takes to post a response :(


It probably looks like Karl Schroeder's Permanence.

...Otherwise, I'm not at all certain of the answer, but I suspect it involves a dynamic interior cladding that mimics coral or other living structures. I hate to say, "Cheese caves...in spaaaaaaace!" but the problem is wicked enough that the architecture itself must be pointed toward the solution.


I know that this is probably an extremely naive response but here goes. Considering we're already on the first step of re-engineering bacteria to produce chemicals that we need, couldn't the biosphere that you need be very simple?

I understand that the interaction of all the bacteria would be difficult to model, but wouldn't this be much simpler than creating a ship with a volume of a cubic mile and the ecosystem to fill it?


It's not just nutrients and calories. There has to be variety, too. I'd get mighty sick of nothing but algaeburgers after a few months.


Potatoes, corn, peas, beans, some fruit trees, etc.

A metric buttload of soil to plant them in, plus the microbes, fungi, and earthworms necessary. Yeah, we could do hydroponics, but I'm trying to keep the human maintenance to a minimum.

Maybe some domestic animals to eat the parts of the plants that we can't. Goats? They also provide meat and milk. Could probably skip this one.

And bees for pollination/honey.

This is starting to look like it would take some serious acreage.

"A citrus plant ... some legumes as well" is barely beginning to approach the actual complexity of the problem.
Yes, sorry -- I was just trying to point out that the minimum is more than just algae.

I didn't even try to provide a list, 'cause bio's no longer my field, and hasn't been for decades now.

Charlie didn't mention aesthetic concerns, so I wouldn't try to consider tastes. And you don't need bees for pollination -- unless you're talking about a very large group, people can pollinate flowers manually. It's boring, tedious work, but it's already done.


I suspect the minimum number is actually quite low, and involves a fairly large amount of brute force processing of various wastes back into fertilizer, your ecosystem does not have to be able to deal with everything if your basic philosophy is that you are going to feed anything problematic, or hell, just all organic waste, into an oven, reduce it to basic chemical compounds, and dissolve those into the hydropondic nutrient fluid your food plants are growing in. After all, any space ship designer is going to find energy and miniature chemical plant a heck of a lot easier to fit aboard than termites, worms, and so on and so forth.

Which means the question is going to boil down to "how many species of plant do you need to cover all the vitamins/amino acids/ect needed for human life and avoiding morale problems due to an excessively boring diet?
If you are willing to cheat (IE: bring supplements), and do not care about morale, the answer is going to be "two: Humans and potatoes". problem being that your crew might lynch any designer that actually made anyone fly with that life support system.


As if Biosphere 2 was the first, or even the most successful, attempt; but I guess it was American, shiny, and on TV.

The Soviet Union did a lot of work in this field. Their first attempt was indeed basically a bloke in a hermetically sealed portapotty with a bucket of algae, but BIOS-3 sounds pretty close to Charlie's requirements; that was a 315 cubic metre hab for three human occupants, a whole lot of Chlorella, some complicated air filters, wheat, and 400kW of electricity. There's more information here although the sourcing may be flaky. Some poor sods spent 180 days locked in it in 1972-1973.

There's a paper in Bioscience, and there's a book available on Google Books here. (I notice from the book that NASA tried some algal species in the 1960s that basically caused epic diahorrea, vomiting, and flatulence. FAIL.)

I've seen a rather poor photo somewhere; the inside of it looked like a small caravan...a windowless, Soviet caravan you can't leave.

Also, you might want to ask Kathryn Cramer's sister, Karen Cramer Shea, who I believe is an expert.


A family of back to the land types living my city have managed to produce about 6,000 pounds (2700 kg) of food off 1/10 of an acre (~400 square meters). It looks like the USDA estimates an average person eats about 4.7 lb (2.2kg) per day. I think they're a family of 6, so they're only producing about half of their needed food. Also I don't know how labor intensive their method is, but in theory they could feed 3 people.

Also NASA had a project BioHome that found a set of plants good at producing oxygen, removing VOCs, and filtering sewage in something the size of a mobile home.

Though the for the ISS they chose to use a mechanical system, that is unfortunately not 100% efficient.


Rather than an answer, I'll try to outline a path to figuring out an answer. 1) Need to know the various cycles-- oxygen and nitrogen are the big ones, but there are many more. 2) Cycles need to be robust, i.e., fault-tolerant-- no critical paths, every cycle has to have a backup, even if it's only temporary. 3) Everybody has to be compatible-- if the byproducts of cycle C poison the inputs to cycle Q, that's a bad thing. 4) Don't forget cycling and sensing the physical variables-- energy, force, volume, momentum, power, temperature.


Cool problem.

I think the answer is going to depend on where you want the life-support system to fall between "fire and forget" and "constant monitoring and tweaking by humans."

For one thing, I think that having just one thing, or just a handful of things, to eat becomes more than just a morale problem after a while even if all nutritional needs are met. I'm pretty sure I remember reading that at some point people lose their appetite and can't maintain their body weight (fine as long as they are losing fat; bad when they start losing muscle). I just can't remember where, drat it, and a little googling is not turning anything up for me, so maybe this isn't an issue.

But if you want an ecosystem that doesn't need a lot of tweaking I think it's going to involve a lot of microbes--they are an important component of soil and some of them are absolutely required if you don't want to be having to add nitrogen fertilizer all the time. Keeping any one organism from overgrowing and unbalancing the system will, I think, be a non-trivial problem; I remember reading that some of the space stations have had serious problems with mold, for example. You can't avoid bringing prokaryotes in with you, either, (maybe if we're talking some sort of "bubble-boy" scenario for the astronauts?) because as Charlie points out only a fraction of the cells in a human body are actually that human's. That being the case I think that dust mites will take care of the shed human skin problem.

If we want the astronauts to have more to eat than algae and bacteria, even supposing they're willing to be vegans, we're looking at a number of varieties of plants with their attendant pollinators (unless the astronauts are planning on making the rounds with little brushes...) The pollinators may have attendant symbiotes that need to be included and kept healthy. The pollinators will have predators that we may want to introduce to keep their numbers under control. The predators will have symbionts, predators and parasites of their own.

Also, do we want to take precautions against accidentally introduced predators and parasites on those plants (by bringing along organisms that prey on or parasitize, for example, aphids) or do we want to assume we'll get it perfectly right and include only those organisms we actually want? Do we want to handle all nutrient cycling (nitrogen, phosphorus, sulfur, carbon dioxide/oxygen iron, manganese etc) biologically or are we willing to do some chemistry?

I do not feel competent to guess what number of species you might need for a self-sustaining ecology that can support human beings. I feel reasonably safe in saying it will be pretty large.


We should probably start looking at the problem from the perspective of how much stuff a person needs per day and then we can work backwards to figure out what inputs we need (plants, animals etc) to produce this.

An adult requires about 2,000 calories per day made up from all this stuff:
Which suggests that an adult requires about 500g of food per day to take in the minimum recommended.

This is broken down into a simple food pyramid like so:
(I assume this takes vitamins and minerals already in the food into account.)

I'll also assume that given the right amount of air, water and food the necessary bacteria/microbes/etc will turn up.

On top of the food you'll also need a giant water tank containing somewhere between 2 and 3 litres of water per person/day(not to mention the extra water you'll need for crops or animals) + air tanks. Much of which can be recycled (although not indefinitely).

It's possible if you were keeping animals in a zero (or micro) gravity environment they would not produce as much muscle for meat. Which is another point, as the people on board begin to succumb to muscle weakening and bone density loss they'll probably require less food (since they'll be doing less work) and this would need to be figured in for a very long journey (although you'd probably have a rotating section of the ship if, at the end of that journey, the passengers were going to end up on a planet).

So, who knows the most resource efficient way to produce the above?


This is an entire science discipline! The problem is not just in figuring out how many separate species are required for a viable ecology, but also the number of members of each species needed to create further (healthy) generations to avoid inbreeding.

The last issue is why creatures like the bigfoot and yeti do not make sense. There needs to be a community of such creatures for breeding. I've read that around 200 individuals is the minimum for a breeding population (there goes the last man and woman on Earth scenario, not to mention Adam and Eve), but its more complex than this, and depends on species behaviour too - breeding couples have to roam, and find each other; so if a rat and its potential mate are separated by a mountain range, then its goodbye to that rat species. Cheetas are one of the most inbred species - they went thru a population bottleneck during the last iceage, when only am estimated couple of hundred individuals were alive at the one time to keep the species going, and the species is vulnerable as a result, being less adaptive and more fragile than other species. The smaller the breeding population during a bottleneck, the more fragile the species becomes, until more time passes and more genetic variations arise again.

This matters is you want to set up a biome that survives over time, but not so much if its just for a generation or two. A gene bank is one solution, but artificial insemination is difficult and only works for some species. And try doing it with beetles! You need a lot of genetic variation for a healthy ecosystem, and so you need a lot of room, especially for higher organisms...

Look at the time it took evolution to create complex life on Earth. Most of the time went on creating single cell organisms, and the ecosystem to support those, with a food supply and so on, and complex life exploded (relatively) once the microorganism ecosystem was in place.


I've a feeling that for a "fire and forget" system you are going to need to be pretty big. There's a lot of cyclic and generally chaotic stuff going on - remember population dynamics was an early area for chaos research. On a planet you can cope, but it's very easy to imagine a vital species going extinct very early on when the entire ecosystem is the size of a school, say.

So I think that for a "fake world" you are going to need a lot of organisms, a lot of non-living systems (weather, lightning for nitrogen, oceans for buffering) and generally a really big space.

For a controlled system I think that with some careful pre-work to produce a strain of germfree humans (but look at the pictures at http://www.narl.org.tw/en/topic/topic.php?topic_id=1 and consider if you'd fancy being a germfree human), and a suitable genetically engineered photosynthetic organism (the infamous "algae") that produced all nutrients and micronutrients needed (actually, a bit of genetic engineering of your humans - stick vitamin C synthesis back in for a start - would help) you could get away with just the two.


An interesting aspect of this question is the similarity to the discourse about different strategies for sustainability - do you opt for a sufficience strategy, i.e. adaptation of life-styles to (here) a minimal, but "natural" ecosystem, or do you aggressively replace "nature capital" with technology (e.g. waste scrubbers, genentech algae etc), thus creating a rigorosly managed "artificial" eco-system.


Re #12: Nope. When you say that modeling the system would be difficult, you're implying that it's possible to get bacteria to do everything we want, and nothing we don't want. Bacteria mutate, in readily observable time. Under industrial conditions, you can work with that -- but if you turn 'em loose, they'll stop working for you and start doing their own thing in short order.

Some things to think about generally:

* Hobby aquaculture setups are simple-ish model systems that you could look at -- plants grow, fish eat 'em, fish droppings fertilize plants. In theory.

* Stability is difficult on a small scale -- any leakage out of the system is more significant.

* Monocultures are generally fragile systems.

* Stuff you didn't want always gets into the system.

* What's doing your thermal regulation?

* Where's the clean water coming from?

* Processing your waste products may be the most complicated part of a minimal system.


I friend of mine is working in the Dutch R&D for Greenhouse based agriculture. For the last few years they have been toying with the idea of self-sufficient agricultural centres capable of delivering a full scale of staple foods, from a single point a full meal could be sourced. Fully integrated and containing many recycling and reuse loops and centred around the concept of levelled working floors, (top greens, then cattle/poultry/pork, then fish, then mushrooms, all more or less growing on the reject from the level above) a agricultural complex of a 100 ha was expected to deserve a community of half a million people.

The lighting doesn't even have to be tweaked to "natural" sunlight, more or less everything works, from fluorescent tubes to leds and in some cases even lasers(why bother but they researched it). And only some relatively minor tweaks have to be considered to bring greens to shape,color or taste(like adding blue light to certain tomatoes to finish the ripening process in the latter stages).

Anyway translate a 100 ha to an orbital or spacefaring capacity (virtually no soil, but some height, say 10 meters) feeding half a million people (in a very rough pronosis ofcourse). It would be a cube 215 meters to the sides. Which isn't all that big. Or in a rotating cylinder configuration as to generate gravity(the cube would then be on an accelerating starship) a double levelled 200 meter wide torus with a 800 meter diameter.

This would be one of the possible ways you could come to palatable food in space with more or less readily available technology and knowledge if you had to build it on a very short timescale.


VictorS et al: I'm going to stick a finger in the air and speculate that given where the implicit requirement for such a system would come from (space journey too long to take canned goods), an implicit corollary is that there need to be at least two, or more, redundant food chains in the biosphere. (Otherwise, if your carefully managed potato monoculture succumbs to a blight halfway to Mars, you're going to be in Donner Party territory.)

Till: a technology-supported biosphere is entirely possible ... until you factor in the need for humans to ride herd on the tech. If the maintenance crew are down with diarrhoea or Martian hyper-scabies or something because the system's gone out of envelope, it's not going to fix itself. (All die: oh, the embarrassment.)

Random rule change: I'm going to allow genetic modification of any organism in the system except humans. (For the pragmatic reason that the lead time on GM humans it as least an order of magnitude longer than for anything else they're likely to want on the voyage. Never mind the ethical constraints of modifying humans for a mission they might not want to embark on when they grow up ...)


Genetic modifications are now go?

In that case, we revert back to algae as a solution - only, we have something in the order of hundreds, if not thousands, of various interdependent strains, with a primary one engineered for photosynthesis or heat/radiation processing. Maintenance will be done by siphoning off a sample or waste products from one hermetically sealed and maintained vat to another - it'll be the primary form of employment for most of the seed colony.

Breakfast, lunch and dinner for the next few hundred years will be algae soup, algae pudding, algae biscuits...


There's a key design question here - where to draw the line between what your biosystem recycles and what you bring along?

Yes, you're going to want to produce and recycle all your macronutrients, but it makes no sense to treat trace elements in the same way. For example, adult humans need about 50 micrograms of selenium per day. One gram of supplements will do you for decades (ignoring what's already in the food). There's just no point trying to set up an additional cycle for selenium, with the attendant weight, power and management requirements when one gram of pills in a bottle will provide a lifetime supply.

But consider:
Calcium, 1 gram per day
Magnesium, 0.5 g/day
Zinc, 0.01 g/day
Copper, 0.001 g/day

Which of these will be cycled versus dragged along with you? And if they are to be cycled, then what's the effect of this cycle upon all the other cycles? (Hint: 50 ug/day of selenium is essential, 400 ug/day is toxic, so the problem here is not supplying selenium, but keeping it from being concentrated by the rest of your nutrient cycles.)


VictorS: Let them mutate then, you've changed them once you can do it again. Unfortunately Charlie has already pre-empted my next suggestion about how to control it: an immune system would have been a pretty good way to control things.

On a different note, an important point to consider is that as well as food for the journey, you're probably going to want to eat at the other end. So unless you're on a roundtrip, your ecology is going to have to be fairly adaptable as well.


@Tetrarch (25): The idea of the "self-sufficient agroskyscraper" is one of the binding threads in the new & fascinating short story collection "Metatropolis".


Cat @19: I think you might be thinking of some of the British studies during WWII? I believe that specifically, it's reduced or unfamiliar meal options, coupled with great stress that caused the catastrophic lack of appetite. Subtract the stress, and even the pickiest eater will usually eat something bland or mildly disgusting if they get hungry enough.

Which leads me to another thought that I don't think anyone's considered yet in this -- spices. You can do amazing things with bland staple foods if you have enough fresh spices, although spicy enough food would become its own problem in a confined environment with recirculated air; hence the millions of won expended to create space-safe kimchi:


@26: Charlie, genetic modification would be a waste of time, I suspect. The only thing it would definitely be useful for is a black-box "cure all the problems free fall causes on physiology so it acts like it would on Earth." Since we don't know how to do that, out comes the GM magic wand to make it happen.

Cat answered about the size scale I would answer. In general, every large organism is going to have more prokaryotic cells than eukaryotic cells, and that's going to be most of the complexity right there. Food: assuming a mostly vegan diet, along with mushrooms to provide the B vitamins that said diet needs, you can get from less than 1 hectare per person. You can minimize the size by maximizing productivity (i.e. maize instead of wheat), but then you get a system that cycles quickly, and if something goes wrong, it goes wrong faster, and you're SOL. I tend towards the bet-hedging strategy of planting multiple varieties of many different species, so that if something fails, something else succeeds. That requires more room, but it's less prone to failure. Another thing that requires more room is if you want multiple generations of seed crops, since you're going to need to set some of your seed aside to produce future generations.

The big problem here is the decomposition cycle. You can run it all with bacteria and fungi (i.e. compost pile) but composting is generally slow. Worm bins are a bit faster (and should be part of the system), and hot composting requires you to get your shit together in cubic meter increments. Still, the fastest ways to get fertilizer fast are a) to recycle your own feces into the soil, and b) to get something like a goat, pig, chicken, or in a pinch, crickets and/or termites to eat inedible crop wastes.

If the decomposition cycle is too slow, your system will have pulse productivity, where there's a spurt of growth, followed by a lot of decay that cranks up the CO2. A fluctuating biosphere is doable, but scary. If it's too fast, you also get into pulses, because many of the decomposers starve waiting for their wastes to build up again.

A final thing I would suggest for a space mission is that they have at least two biosphere-type greenhouses separated from each other by airlocks. That way, if something goes wrong in one, it doesn't doom the mission immediately, and it can be rebuilt using material from the other.

Anyway, numbers of eukaryotes: I would take dozens of species of food plants: something like a Mesoamerican milpa (maize, beans melons, tomatoes, chilis, sweet potato jicama, amaranth, peanuts, posibly avocadoes, etc), stingless bees as pollinators, red worms in a worm bin, the organisms to run both slow and fast compost bins (mostly microbes: I'd use the sourdough approach and just take half a pile from some system that works well with the same inputs I'd give them), probably some microbes to help break down human and other feces, and either a couple of cages of crickets and termites (mmmm, termites!), or a flock of chickens and a few rabbits. Oh yes, and probably a dozen or two fungal species to work on decomposition, condition the soil, keep the plants happy, and feed me. Plus the hundreds of microbial symbionts that each of these big ones have.

So, we're looking at, on order, 50-100 macroscopic species and 500-1000 microscopic species.

That would be my happy spot. Now, if you're into crazy survival stories, you can figure out how the Independence I culture of the high Arctic (ca. 2500-2000 BCE) survived. They seem to have subsisted largely on musk oxen. The archeologists can't really figure out how they got through the winters, because the evidence says they spent the winters in tents burning bundles of willow twigs (i.e, one or two cubic meters of twigs for a six month winter in a tent). One archeologist (McGhee, Ancient People of the Arctic) postulates they spent most of the winter sleeping, because they can't figure out how they could have kept a fire going for even a short part of the winter. AFAIK, they were the first people to populate the high Arctic, and somehow, they made a living in a way that would probably kill expert survivalists. The only reason to bring this up is that I'm looking at space survival in relative comfort and stability. It's quite possible that there's a crazy way (or twelve) to get by on resources that would starve a goat, and my guess is that crazy is going to be first out of the starting gate and into deep space.


Why have macro organisms at all? In the end we're trying for a self-sustaining chemical reactor. Perhaps you can get some sort of simple light harvesting organism as your base (yes, the sorely beaten algae horse), and then have an ecosystem of tumors at the top. Tumors are immortal, evolve quickly, nearly every organism has them, and it would be fun to have a slab of ecosystem pâté to eat every day.


Does the answer involve throwing in a few of those food printer doodahs for sustenance?


Backyard chicken coop in space?


On the surface it seems like an interesting question, but the real answer is that we haven't got the slightest idea. Postulating a space faring human race without any genetic engineering is absurd on its face, but allowing unbound GE just gives you magic super algae defined to produce everything we need.

We just don't know enough to even set the bounds of the question, let alone start answering it.


What is "number of organisms"? Individuals? Species?

For individuals, it's the prokaryotes plus some noise, let's call it somewhere in the 10^15 to 10^18 range. We have no idea how much of that is "necessary", beyond "some". For species, the term is more-or-less meaningless for prokaryotes. I don't think we have any good measure of biodiversity at the prokaryotic level. The idea of "gene" still holds together, more-or-less, so I guess you could have a theoretical measure of the number of different gene expressions. No way of actually measuring that for a sample of slop, but maybe it's a step in the right direction.

Do you want us to count viruses? How about endogenous retroviruses?

Or is this only a question about eukaryotes? Multi-cellular eukaryotes? Macroscopic eukaryotes? Are we allowed to genetically engineer stuff? Using how much magic? (enough and the answer is surely 1). Give us some guidance here.


heteromeles @3: I had in mind specifically the GM soya that Monsanto's getting ready to launch commercially that produce omega-3 oils. Or similar. It's not a magic wand, but if it saves you from having to pack a fish tank as well as the soya beans you're growing for tofu it may be worth it. Also: golden rice.

A vegan diet is probably the way to go, unless you've got a specific need for animal husbandry (with all the waste handling, reproductive and veterinary issues it implies -- a whole extra layer of complexity on top of growing crops). If you've got to have animals, chickens, tilapia, and rabbits or guinea pigs will probably work best (and I'm really not enthusiastic about the idea of RODENTS IN SPAAAAACE!! escaping and chewing on the wiring ...).

(Late here now, I'm going to bed. More tomorrow.)


Off the cuff, I'd say the more the better. Cat mentions "fire and forget", but there are some additional benefits to going for increased numbers of species.

For one, we simply don't understand the problems and processes of biosphere-type arrangements very well. The larger set you take with you (the more you try to mimic an open, working biome) the more likely you are to get some things incidentally right. After all, who knew you needed thalium to make yams have the appropriate nutrient effect?

Another is that there don't seem to be any one-to-one niche-to-organism situations. Take multiple organisms for a given niche. If one fails due to some unexpected lossage, the others still fill the niche. I'm assuming, of course, that we only care about a given niche because it's important. Lose the niche, get ecological collapse, lose the mission. Redundancy is your friend, especially when so many of the risks are unknown.

I strongly suspect the number of organisms in the human body are going to be a complicating factor. We need something in the environment that's going to break down every byproduct of human life. If the mission is sufficiently long, that breakdown process must also return the byproducts to the cycle rather than simply sequester them. That may radically increase the number of species needed.

What type of mission length are we talking here? A 3-4 year trip to Mars and back? If so, we might be able to get by quite simply. For a lot of critical vitamins and minerals, we might be able to get by with frozen supplements. Then we'd "only" have to worry about the biosphere for calories and oxygen. It may be monotonous, but we'd get there and back.

Heck, even O2 might not be an issue. Assuming sufficient power and tech, you can convert CO2 back to O2+C and plow the C back into the system. Now all we have to worry about is generating those 1200 (my deliberately lower number) calories a day. At that point we're back to simple hydroponics. There are still a lot of species when you count the bacteria, etc, but if it only has to last five years, you're looking at a much smaller problem.


@39: 1200 calories a day? That's nearly a starvation diet. Your astronauts are not going to be doing *anything* when they get to Mars (or wherever they're going). Budgeting for something closer to 2000-2500 calories per day with peaks of up to 3-4000 is a much smarter approach in my mind.

I'm inclined to go with the 'more' option here. Like I said earlier, try to set up a complete mini-ecosystem, based around probably two main crop packages, from different parts of the world. Something like the old Native American standby crops (maize, beans, chilis, etc.) and the Mesopotamian package (wheat, lentils, etc.). Animals, at least insects, will be necessary, which probably means that you'll need to bring at least a *few* larger animals for pest control and the like. More details are necessary for a real answer, but at the moment I'm assuming the goal is simple indefinite existence.

This, by the way, is another reason why I think predictions of vast, flourishing self-sufficient deep-space colonies are delusional. A planet is just so huge that an accident or crash in one part of your biosphere can be absorbed by the rest. If you make a habitat big enough for that, you might as well have just built on a planet in the first place. That's aside from the whole zero-g thing, which may cause birth defects. I know, I know, spin the dang thing, but that itself has certain problems. Going to a planet is just easier.


@Nick@37: I'm glad someone brought this up. So far we've been discussing humans within the system as though they are a single organism. They're not. They're a composite of multiple organisms, some of which are very nasty. (I'm looking at you, E. coli.) If I recall correctly, the whole meat probe needs that entire personal ecosystem to survive -- without it, there's no food digestion, and the vagina turns into a desert wasteland that spawns even more virulent bacteria. Assuming that's true, a thorough count means we have to count those species as well. But maybe I'm wrong. Again, we need a biologist. Get me a Stross/Watts mashup, stat!


Going for silly-fun, rather than serious (which I lack the knowledge-base for):

What's the technology base that you are assuming to have available? Can we have engineered plants that emit whatever specific algae we need to have the set of inputs and outputs of chemical compounds that we desire (while preserving atoms, of course; so needing the ability to route some chemicals away from general immediate release-to-air).

How about: for every chemical which needs to be in balance in the environment, there needs to be something which can increase the level and something which can decrease it, plus some general scrubbers for noxiousness. So the question is one of how intercorrelated those chemical levels are and how many levels is optimum to adjust per tailored organism which consumes some and emits others as waste-product. For resiliency, you probably want two organisms per increase and per decrease in a chemical, so that if you have too much A and organism O1 reduces A,B,C but increases D,E and you already have too much D and using O2 to decrease D will increase A, then you want a different option, O3.

Assuming tailored organisms can be engineered well enough to choose any combination of increasing and decreasing, independently of how the chemicals react in nature, it's then a matter of math, solving to find the minimum.

Of course, anyone who goes into space depending upon the minimum is even more optimistic than those who go into space in craft built by lowest bidders. If you have the ability to manufacture the organisms on demand, then you're really concerned with what's the maximum rate you'd ever need to change the level of any chemical in the biosphere (processing smoke from raging fires?) and what volume and physical design you need to interact with the environment best (fractal branching of a tree-like base), to scrub it, with organisms returning for reprocessing. Some kind of gengineered plant which can make the other plant organisms on demand. But you'll want three different root plants, for redundancy in case cosmic rays turn out to mess up plant X, even if only in preventing it from emitting organism O1 and X works for everything else.

Once you have the scrubbers, you then add in food sources for variety to suit the humans.

If you use a computer to control the master plants and which algae they spawn, and you turn it into a biological computer and embed it in the plants, can you end up with sentient "trees" as a necessity for space travel? :-) Three of them, voting!


@Phil:42, three trees voting - have to love Iain Banks.
@VictorS:24, I would imagine that some frozen bacteria of the optimal/engineered type would be periodically recultured to ensure stability (at least if it's the tank-bound stuff you're talking about in the first place, obviously re-innoculating the soil with the Ur-bacteria would be significantly less effective.)

Regarding ruminants, I would tend to think that cows would present a CO2 problem unless you had an extremely vigorous growth rate in your little bubble. Victor's model of the aquaculture set-up would seem to present less of an atmospheric hazard. (no dirt!)

I would tend to think that you would want to take a lot more topsoil than you'd hope to need, if going the Space Pastoral route; a rash of bad crops or a malfunctioning bioprocessor (dead cow?) could turn into a downward spiral of productivity.

Hippie thinking: find an ecosystem that you can simplify. Plains model: herd animals eat simple photosynthetic organism, shit. Birds follow herd, picking through debris, also shit. (Additionally: insects eat shit, are eaten by birds, herd animals.) Shit feeds topsoil. Agricultural hack: use buffalo or cow as herd animals, rotate pasture using fencing, rotate flock of chickens over just-grazed pasture, then leave fallow or plant crops. Eat most eggs, occasionally eat chicken or buffalo. (Hope insect populations stable.) It's a hell of a lot of area, but it seems relatively simple in terms of numbers of elements. Now whether you'd run into serious problems with herd sizes, etc. I'm not sure. Optimally we would select animals very, very distantly related to each other but with proven sound traits, otherwise we might have some kind of frozen embryo stash of novel genetic stock?


I've heard it said that with good soil, subsistence farming could be done at a rate of a quarter-acre per person [citation needed].

If you did a Three Sisters kind of approach, then you'd want to trace a square a meter to a side, then dig down about a meter ((sorry to be mixing units...quarter-acre is a squidge over 1000 square meters?)) then have a lab list every single one of the components it found in the soil while the 'sisters' were in full form.


Oh, and be sure to bring some worms. ((sorry for multiple post.))


(i.e. maize instead of wheat)

You're probably better off with wheat (or some other cereal), because maize requires a lot of water as well as light. And warm weather: don't plant it before the soil temp hits 15C.


@38: Charlie, the traditional role for animals is to eat the things we can't eat, and to convert them into things we can eat. The additional, ecosystem role for animals is as fast decomposers (24 hours through a goat, rabbit, or chicken vs. 16 days through hot compost, up to a month through worms, and months through a composter, or years in soil.

In a small system, getting your nutrients tied up in decomposing plant matter and unprocessed fecal material isn't good. Additionally, you get to convert some inedible plant material and waste into edible animal products (eggs, milk, or meat).

Insects would work better (mostly because you can start with a few and grow many, rather than needing a bunch of large animals), but many people have cultural problems with eating them. Fish work as well, but then you have the potential issues of water in low gravity.

I agree that escapee animals are a problem, but then again, so is condensation.

I realize that a vegan diet is possible. However, most traditional societies, while largely vegetarian (not arguing that) also ate a certain amount of animal products, often as a recycling step or to obtain needed nutrients. I think there's a good reason for people to keep animals as well as plants.


What supports you?

All known elements would need to be present for minimal survivability and sustainability, anything else threatens a suicide mission beyond reasonable odds that can likely include all manner of unexpected consequences and surprises that might involve some previously unknown or undiscovered elements.
It's getting the optimal mix of all known ingredients, the main goal will be to understand the value or not of so-called nuisance species (i.e. pests), to an extent that some may be a vital catalyst that will somehow might 'save the day', when all else fails to overcome an intractable problem. High risk demands high hopes and a well stocked larder. Miniaturization will go a long way to help solve the storage problem, one key question is what are we not taking that remains a hidden variable.


I'm going to take all the upper-limit suggestions made by others for complexity, and add in the suggestion that you'd want a very high ratio of biomass to top-level consumers (i.e. the crew). At least 100:1, probably more like 1000:1 or even higher. Any less and you're looking at the kind of high-turnover ecosystem that's notoriously unstable, and I think stability would be rather a pressing criteria for your life-support system.

That adds up to rather a lot of stuff: tons of living matter, plus substrate. Forget starvation diets, part of the crew's job will be to eat as much of the excess production as they can.

An interesting model to look at is the way people do aquaculture (and traditionally, rice paddies) in South-East Asia: a big pond full of algae and multiple species of herbivorous fish. Human and animal wastes go in, and food comes out. You could add a few species of burrowing crustaceans to clean up dead bodies and turn over the mud. The rice paddies would run much the same way, sans fish. I can't see how having a large amount of water would be a bad thing on a spaceship: at the very least the thermal mass might come in handy. You'd probably want to keep the rice fields in rotation, to avoid peaks of CO2 and methane. As heteromeles pointed out, having your ecosystem pulse gases would probably be a bad thing.


Whoops, meant to say: you wouldn't be adding nitrogen fertilisers to a closed system, either. Nor any other fertilisers, with the possible exception of a trace elements.


Forget about land crops. Go purely aquaculture (salt water, algae, seaweed, fish, crustaceans. It's restricted to where the water is so you minimise contamination of dry areas, and you can make it quite complex without needing a huge area. As far as I know, everything you need can be produced in an ocean environment and it's not affected by weightlessness.
You don't need land animals, insects, birds or crops. since we are assuming that energy isn't a problem, then the purified water can be used for drinking, bathing and probably as fuel and backup oxygen after a little electrolysis.


Honestly, the whole "it needs to be really big/diverse" argument makes me break out in hives - the more things you bring, the more unpredictable outcomes are possible, and it does not actually protect you from outcomes of the "Eh, shit, we forgot termites" or "Oh shit the termites are eating everything!" types, and then you are suddenly trying to fix a really complex system by hand. Nothing that actually fits on a spaceship smaller than a Culture vessel is going to be fire and forget, so brute force chemistry, hydropondics, and as much labor saving automation as you can get away with is going to be key. Adding redundancy in the form of overcapacity and segmentation is going to weigh less than trying to approximate a natural ecosystem. - This means you are, indeed, relying on constant vigilance, but really? If the entire crew is incapacitated, that means they are going to be dead anyways. We are unlikely to build space craft that are very forgiving if the crew stops paying attention to the controls any time soon.


Madeline@41: I used to be a biologist, and your points are just why I referred to "germfree". It appears possible to rear, at least, mice with no intestinal (or skin or other) flora, and for these to reproduce within an aseptic environment. They aren't the happiest looking beasts, but they live and breed.

I can't see any technical problem why we couldn't do the same for people. I can see lots of other problems, but not technical. And this passes Charlie's "no genetic engineering of humans" clause.


I'm going to go for an uneducated orders of magnitude guess:

The Isle of Wight in a can

I feel that the ecosystem is at least an order of magnitude more complicated than we currently comprehend it being and that nothing less than the equivalent of a decent sized island is going to be able to be sustainable.

To offer some support to my hypothesis we could investigate various sizes of island around the globe and see what is the smallest size capable of supporting the required number of humans we want.

This has been a similar problem for humans for a while: if I settle on an island how big a family can I have?


Nick@53: I have a feeling rodents are much better at making their own vitamins than humans are. If you were going to make an aseptic human (if you could) they'd need an awfully big list of dietary supplements. That's assuming they didn't develop a charming array of auto-immune disorders.


Well, assuming we're allowed a lot of raw energy from mysterious source X - can't we grow sugar cane? Apart from fats and fibres, it'll get us most of the nutrition we need - let them eat Mint Cake!

We can, to a certain extent, forgo proteins (man can, almost, live on carbs alone, through protein sparing). So, we'll need fats, we could go for grapes, or rapeseed, or other oil producing plants to supplement the sugary evil I'm proposing. A few herbs for variety. All we're missing now is fibre...


Oh, obvious when you think about it - make room for some cacao trees, and grow cocoa - fibre sorted, and opens up for a chocolate diet. We'll need plenty of CO2 on board, so that means sugary fizzy drinks are on order as well.

For a savoury, maybe marmite yeast?


Let's assume I'm the budget director for the starship project. If we ever get to the moment where the ecosystem team says that we need a 100:1 ratio of non-human to human biomass - perhaps more - I intend to throw up my hands and _then and there_, _in the meeting_, ring up all the tinned and freeze-dried food reps I know, to start discussing an RFP.

In other words, if you make your ecosystem too large, it begins to look very bad alongside brute-force solutions to the problem of human survival. All this shit (literally) needs accelerating, after all. Unless, that is, the main raison for the mission is to move the ecosystem, and the humans are just pursers to the cargo, rather than the payload themselves.


The point is that in fact you are feeding energy, either from solar panels or some fusion or fission plant to the spacefarers. Since placing electrodes directly on the test subjects has proved to add no nourishment so far (except by producing lots and lots of cooked and charred meat) The determining factor for any food production system would be the turnover rate from poo and expelled air(and hair, and skin follicles,etc) back into hamburgers.

The more time is needed to make that happen, the more mass and volume you need, to allow for the transformation process to happen and to continually ensure food availability. By using electrochemical and enzyme based processes you would have the fastest turnover rate, but probably have some really unappetizing goo after a few cycles. Which incidentally corresponds with a lot of fast food and baby food, so that might be a possibility.

I would expect then to see something with continuously producing plants (like tomatoes, that can be mechanically forced to continuous production) and maybe an algae to crustaceans cycle. (better have to have shrimp for dinner everynight than some disgusting insect)

Anyway, as always it is a matter of energy concentration, yet tempered by the human nature, as in how many time can you be forced to eat your own fecal matter (even with generous amount of ketchup) before you commit suicide.


An Interesting twist on @58:

You are Ecological systems director of the Mars project.
Fully-enclosed-ecosystems aren't reliably working, so most of the food for the say 2-year Mars trip is tinned, and you've been reduced to backup / experiment. So you're really mass-bound.

Now what choices do you make, given that the biomass (shit) is going to Mars anyway ? You're concentrating on creating a soil / composting infrastructure, and brought some seeds, rather than trying to feed the crew.


There's an obvious solution to the poop problem: use it as propellant.

Quite simple steps:

1. Fill spaceship with canned food and humans
2. Humans open cans, eat food, and produce poop
3. Seal poop into cans
4. Use a big ole electromag rail to accelerate cans to high velocity (free leccy is available right) and shoot them out in a useful direction for navigation purposes
5. Laugh in the face of all those eejits in their Silent Running biodome spaceships as you arrive at the New World light and poop free!


Suddenly, Im getting this vison of a spaceship where the one remaining crew member that didnt go down with the flu is spriting around 24/7 higher than a kite on sleep antagonists and stimulants trying to do nurse duty on top of keeping the waste treatment plant, the reactor, the shrimp tanks, the hydropondic baths, and so on and so forth all running smoothly and hoping desperately somebody gets back on their feet before the medical cabinet runs out of amphetamines..


Interesting comments here.
Thought I might join in.

The Biosphere experiments were interesting but flawed in the sense that they were conducted in earth's gravity well. While it makes a good model for self sustainable living here on earth, its (in my mind) the wrong model for space habitats.

Earth is a Water planet, our atmosphere and biology are aquatic by genesis, everything that happens on land started due to the effects that the oceans have on our biosphere.

Life on the land (homeostasis) could not exist without the oceans.

So, if your going to start a Space Biosphere you would need to start with a substantial body of water. Probably all water. (see amazing Don Petit experiments on ISS:)

http://www.youtube.com/watch?v=jXYlrw2JQwo&feature=related )

Spherical Fishbowl in space,? An ocean on earth is rarely just water, its a whole organic soup, this soup is what (I think) one would need to begin a self sustaining Biosphere.

How to start one? If we ever get the space elevator we can just plonk a long drinking straw down to the ocean. :) Or find some water rich asteroids to mine.

Start with this 'base' to build upon.
Fit an artificial magnetosphere (fudge) at the core. (tether extending out right from the start). I would bet my imaginative brain that if you started with a balloon and eventually got a functioning ecology producing gases you would end up with an atmosphere. Aquaculture from that point is likely to be far more productive for food, oxygen and energy then conventional (earthbound) thinking of land farming.

The questions are how much mass is critical to self sustainability? Also as mass grows there will be increasing pressure at the core, this offers tremendous possibilities for energy generation assuming you (engineered) the core for the purpose.

The Don Petit Experiments on the ISS illustrate to me that the behavior of water in Zero Gravity is a whole new outlook on the space habitat theme, sprinkled with a bit of imagination of course.

I can see the illustration now, Dolphins in space... :)

Damian Kuczynski


If you're allowing hand-waving for energy input to drive (say) photosynthesis, you also have to allow hand-waving for digesting turds and skin cells - who knows: supercritical wet oxidation or high intensity UV. Once you have the water, nitrogen, carbon dioxide, and macro/micronutrients you just feed that back in as raw materials for the bio system.

I'd claim you also get to hand-wave on vitamins and assume genetically engineered microbes, or at least specify where the limits to hand-waving go. In one horrific, yet imaginable scenario, you could have a supplemented glucose drip without any foodstuffs whatever, all produced by microbes (bacteria, yeast, algae).

Conceivably this could even be done with a variety of strains of nitrogen fixing photosynthesizers, with no eukaryotes whatever, each tailored to poop a particular nutrient.

This comment is along the lines of Charlie's intent to attempt to specify a minimum; I'm not saying I would want to live in such a situation. However, once you have a lower limit on what you'd have to put in place, you can then build from there to counter any seeming deal-breakers (e.g. postulate some kind of cool psychoactive drugs that program you to enjoy that glucose drip... also produced by microbes in my scenario, of course!)


Put the humans in VR and bodily in some food fluid/bioreactor?


Two species; Humans and algae Chronically and perilously unstable, what about the several trillion fellow travellers in the average gut? Not to mention the fact that absence of cosmic/solar radiation doesn't stop mutations?

Earth normal mutation rates are in the range one in ten to the six to one in ten to the eight. The vast majority are either silent -no immediate effect, or deleterious and so sewlected against. Apply even a slight selective advantage and watch the mutation spread. Monsanto seemed surprised when it's 'Roundup ready' crop resulted in herbicide resistance to Roundup in weed species inside five years. It was inevitable and didn't require genes 'leaking' from the the crop genome. Ten years earlier European biologists were watching the spread of Paraaquat resistance in multiple weeds and grasses across central Europe - years before before genetic engineering was practised on plants in quarantined labs.

Starting point
Biosphere 2 had one point two hectares in five biomes and included zero point two seven hectares of agricultural system. The agricultural system barely supported the eight occupants, The ecology. Ants took over and half trashed the rainforest biome -probable cause lack of controlling agents - be they disease or Aarvarks or the unnaturally still weather conditions ~(drowning nests?)

Crop yields suffered from various factors, but one no body considered was that without breezes and rain to fight against the plants grew spindly and weak, The forest Biome suffered similarly.

So lessons: minimum 350m2 per adult of agricultural area. This can be reduced by companion planting/permaculture systems the already mentioned three sisters scheme is an example but there are better. The downside is that they are labour intensive so if you don't want a cast of farmers it requires either intelligent mechs or monoculture cropping and larger areas.

Given the lifting of the genetic modification ban how about carniculture, whether plants or vat grown to replace animal protein

Life support ecology Basically we won't live long enough to grow a new one from scratch unless we know what we're doing, and we don't currently. The UK's Ecology is less that 10,000 years old and compared to almost anywhere else is terribly impoverished with a mere 1200 native plant species several thousands of arthropods, multi thousands of fungi, prokaryotes, eurkaryotes etc. All this on a land which has virtually no wilderness areas untended by humans or their agents (eg sheep grazing the fells)

Actual size of support ecology that would be 'fire and forget'? well I've seen 'authoritative estimates' (aka WAGs from people prominent in a related field - no one in the field would dare as they know the depths of our ignorance) suggesting four hectares per human.

Personal take is somewhere closer to Biosphere 2's 1600m2 per human, but simplified so that you run it as three biomes 1 agriculture, 1 aquaculture - fish, water reserve,CO2 sink algae farm and one 1 mixed woodland/heath environment. Do duplicate the support Ecology using different models. Different agriculture, maybe coral reef and tropical rainforest. and as a Last ditch backup CO2 scrubbers and hydroponic greenhouse.

Please note this is not a fire and forget, the ecologies are too small, they will need management but the large an individual biome the less work it should take per unit of area

I'm trying to prevent one point of failure killing the mission whatever it may be. So long as both Ecologies tick along life should be good and the hydroponics/scrubber system in storage.

The more you look at this the harder it becomes!


The issue with aquaculture on Earth is that CO2 diffuses slowly in water and is the major limiting factor on plant growth. Those fancy aquariums full of aquatic plants are fed bottled CO2 so that the plants can grow.

Conversely, if you have a very humid chamber, you can actually grow plants aeroponically, and that (or Forward's levoponics) might actually make more sense than a soil based system, *PROVIDED* that you have a way of breaking wastes down into a liquid that you can spray on the roots without poisoning yourself.

The sticky part is actually the solar power driving the system. These big, complex Silent Running biospheres are designed around the idea that it's running on about 150 to 300 w/m2. If you've got a lot of surplus energy to play with, you can, of course, burn the wastes to break wastes down to their mineral constituents.

That's also the reason why a lot of human labor is involved. Human max work output is on order of 70 watts (that's peddling hard), and a caloric intake of 2000 Cal works out to about 100 W input over 24 hours. These numbers can be quibbled with, but the important point is that energetically, human labor is a great way to do a lot of complex tasks with relatively little energy input. If the system is energy-limited, one can make a good argument for minimizing automation.


Nick@53: So, could germfree deep space humans become Wells' invaders, brought down by the common cold? Wouldn't they be toast the moment they a) encountered foreign bacteria from a core sample, b) touched the surface of an oft-forgotten panel that an engineer had once sneezed on pre-launch, or c) rendezvoused with another station? (How do we vaccinate for an environment we've never visited?)

I don't dispute the possibility of the germfree human, but it's sounding more like a vulnerability than an asset at the moment. I think maintaining the quarantine would add a layer of complexity that this experiment is striving to avoid. If we've got to count the flora just to make sure the probes remain functional, then we should count them.

This brings up another question, though: are we hatching probes in the hab, or sending some from home? I think the layers of complexity necessary for sustaining life might be different for those two groups. Native probes could be cultured to need less, and might be completely content to pluck raw mushrooms off their walls and ceilings when they're feeling peckish.


I think there's a great deal of automation going on in topsoil already. A lot of root structure evolved to deal with wind loading...a little root tending and you might not have to use a big block of soil. What's the performance per pound of various and sundry techniques?


@69: Kelson, I agree with you about the "automation" in the soil. Plants shed a lot of photosynthate (i.e. stuff synthesized through photosynthesis) into the soil, on the order of 10-20%. If you think of plants as the big energy companies, they power a lot of small businesses (i.e. fungi and bacteria) in the soil through that photosynthate. Energy-wise, that's around, say 200 w/m2 hitting them *20% efficiency*20% to soil or 8 w/m2 into the soil (yes, feel free to quibble with the numbers. It's right order of magnitude). So, energy and resource-efficient breakdown of wastes into plant material that takes time to work? This is why? It takes less wattage than, say, an incinerator, and as a consequence, it's slower. That's why getting the decomposer cycling rates correct is so critical.

As for roots as support...not so much. The major support material is wood, which is some variation on the theme of CH2O. Wood is great stuff, because it takes surplus photosynthate and turns it into a mostly-dead support material that has a low maintenance cost. The plant also needs to have properly sized vessels to move sap around (some of which is powered active transport), and a lot of metabolically active surface area to take up nutrients and sunlight from whatever you're growing them in. On balance, you can get by with a smaller root system in free fall, I think, but it's not huge savings in weight or size.



The only correct answer to your original question is: not enough data.

But if you allow me to (briefly) depart from the original question, I can say with some confidence that you need a lot less complexity if you allow for solutions that aren't perfect. As in being ready to add some oxygen and scrub some CO2 from the atmosphere by means other than biological. Or adding phosphorus and nitrogen from stockpiles to make up for not perfectly closed cycles.

The complexity you would end up with would largely depend on how much of a stockpile you have and how much power you can put into stuff like scrubbing and splitting CO2 and finally, how long the mission is supposed to last. 10 years or 1000 years ...

So, think about your problem the other way around. Given a certain complexity, how close can we get to perfection?

We can have an interplanetary mission of, say, 10 years with not much of a biosphere to speak of (think ISS + huge load of cargo). The mission could last 20 years if the biosphere had an efficiency of 50%, 100 years if it were 90% closed.


An actual starship ecosystem would look like a Chinese village with the mix of rice paddies and animals. That would require a stable gravity for the system to work.

Look up the Wild China series by the BBC for examples of what a complex ecosystem really looks like. There are many downloads on YouTube.

This is a highlight clip that shows the complexity.

A true system would have to match that complexity with plants, animals, fish, birds, etc... That's the only way to handle food production, air, and waste handling; unless of course you have a magic recycle system like Niven had in his Known Space series.

Nature is complex, not simple.


heteromeles, No quibble necessary on the flux :) I'm just amazed at how much gets done on the available light as is.

I was thinking that roots could be trained, or GM'd, to be shallow and numerous in the cozy confines of the habitat. In a manner that a sequoia stands up fine for 200+ years on a root ball, instead of taproot.


Madeline@68: Foreign bacteria - unless you believe in panspermia the chances of alien bacteria harming you are minute. The ship will have to be completely sterile before we start (and see more on this below). Yes, re-integrating with human society will need to be handled with care.

John@66: You solve the mutation problem by sterilising your algae vats every however often is necessary (say by boiling them with that copious energy we've got floating around - and a bit of UV while we're at it)and re-seeding from one of a huge number of small packets of freeze dried spores you carry.

I remain convinced that you are either on a mini-earth (and so so mini at that - I disagree that an island is a model - the oceanic systems (including the deep convection) and the planetary atmosphere are a huge factor) or you want the absolute bare minimum of species. And that bare minimum is germfree humans plus algae in vats. The algae are extracted, cooked, and sterilised before they come into contact with our germfree humans. The human waste products pass through some sort of airlock/steriliser combination.

I'm not volunteering for this - I'm in full thought-experiment mode here.


@73: That's easy. Snip off the tap root, and the roots will grow laterally. Actually, most monocots (i.e., corn and other grasses) don't particularly go in for taproots anyway. Training the roots into a shallow space is fairly straightforward, and hydroponicists and bonsai artists do it pretty routinely.

The thing is, a tree grows much like a coral: it's a skin of very active tissue over a metabolically inert core (wood for trees, coral for reefs). Would is a very stable CHO sink, which might be good or bad, depending on what you're short of in this spaceship of the mind. Conversely, the green parts of plants have a lot more elemental diversity and are a lot more metabolically active. This might be good or bad, again depending on whether your system can handle that much activity without running short of something.

One of fields of information we're critically short on is a tissue-by-tissue elemental breakdown of the kinds of plants that we'd use in constructing a biosphere. FOr example, I'm only aware of a few studies that looked at C:N ratios (for example) in different tissues in a tree, and that ignores the other 14 or so elements that plants normally need. If you're going to try to have 17 closed biotechnochemical loops going in your space habitat, knowing where the stuff is needed and when on a per species basis would be quite useful. If someone's looking for a "find 'em and grind 'em" postdoc, this would be one subject to look at.


Seems to me that you might grow the macro-nutrients with specialized gengineered aqueous cultures, and bring along the micro-nutrients. You can use the tanks for rad shielding in solar storms, and reseed them from cold storage when mutants cause trouble.

Move the macro-micro slider according to how long your journey is.

Keeping the cultures working properly long term would require professional tending, and significant RnD to get gengineered up and working in the first place.

Air recycling is probably best done via rugged brute force machines if energy is cheap.

Human waste recycling with a sterilization step inserted will hopefully allow the system to be fairly closed.

I would recommend developing several working culture systems and packing multiple starter cultures for them all, if at all economically feasible.

Also consider starter cultures for required human commensal organisms. Some RnD on what is required would be useful here.

There will be significant problems making something like this last really long term (decades plus), but on that scale you have the same problems with your machinery, electronics, and crew.


Let's do it the other way.
Of the species in our observable working (so far) biome, which ones do we know we can manage without.



Well, OK, for a starter, I suspect we could manage without the Blue Whales, although this remains an incompletely tested assumption, and various things do seem to have gone awry since they became scarce.


So we're dealing with a two part problem, supply and demand.

So lets begin with the first, we take a number of volunteers, and close them into a absolutely sealed building, large enough to put them mentally at ease-- this will be the habitat portion of the experiment. The habitat by default is extremely bare, furnished, but no food or supplies (music, video, clothing, toothbrushes, paste, etc).

They have an airlock which leads down to the 'factory' where the engineers and experimenters work and is used for deliveries--

Each time the volunteers require something they bring up a terminal and ask: food, drugs, entertainment anything they require (within reason..).

For each of those requests, the experimenters grant the request in the most simplistic way they can from off-the-shelf goods, then apply themselves to working out the production line to actually generate a working ongoing solution that keeps the product coming at the interval required by the volunteers ("We really need a litre of shower-gel, but only once a month").

The production lines are then built in the factory portion of the experiment.

So essentials and not-so essential elements are slowly ticked off the list, each time the production line fails, it gets tweaked and rebuilt, the volunteers don't suffer for the failures, the experimenters fall back to off-the-shelf until they refactor the lines and move back to using them again.

Every week the volunteers get an intense medical check and the results again, fed into the factory portion if needed.

After running the test for a generation or two, you should be fairly confident in your supply lines and redundancies, and indeed how much 'off-the-shelf' supplies you need to cope with unforeseen failures.

I suspect by the end of the experiment the habitat portion will be absolutely tiny compared to the factory element, but if the world is our factory, surface area/population means we're running at 70,000 square metres per person, though that numbers shrinking fast.


You can also get some very crude mass estimates from the good ol' trophic pyramid.

Basically, for every 1 kg of herbivore, you need at least 10 kg of plant material, and there's roughly 90% loss of biomass between trophic levels. That loss goes to the decomposers, so double the mass to include the decomposer cycle if you're going to run decomposition biologically.

If you eat animals, you get 1% of the plant mass (10% efficiency across two steps), assuming you eat the critter.

The complexity here is that humans typically don't eat all of the plant, so trophic efficiency is much lower, and animals will eat parts of the plants that humans won't eat.

Still, for every 1 kg of human, you need at a minimum 21 kg of plant and decomposer biomass for biological processing. What that 21 kg is composed of is the real question.

Here's a simple recipe from 19th Century Ireland: Potatoes and goat milk, anyone? That's complete nutrition. If you engineer the potatoes so that the goat can eat the potato greens, vigorously sterilize the system so that no potato blight gets in, and engineer some sort of fast composting system, it might work. Of course, the goats don't always give milk, and the potatoes take time to grow.... And then, there is the small example of what happened to the Irish when they tried this particular strategy.


h@80: Potatoes are funny things. They're the most efficient carbohydrate crop, but they need deep soil, without too many rocks if you plan to till it mechanically. Is it even possible to grow them hydroponically?


h@80: Marvelous stuff. The trophic pyramid scheme is interesting. Is it rated out? By which I mean, is that 1 kg from 10 kg per day/month/instant?

And for goodness sakes, throw in some dandelion greens with that potato and milk diet...:D


Tomatoes have deep root systems and can be grown hydroponically, although I don't know how successful that is. They aren't good container plants, though.


Erm... ecosystems tend to be self organising and self stabilising. Trying to design and build one and have it not be a spectacular clusterfail is the wrong approach. How about trying evolve one, using natural selection.

Get a great big rotating habitat filled with dirt, stuff a huge variety of organisms in there seal it up and give it a year or ten with sunlight input. Once the ecosystem has sorted itself out, introduce a human burden to the system *slowly* and sustainably. The CO2 a human cycles could be buffered while the ecosystem catches up. Likewise other compounds elements. But the secret is to allow for adapation.

Kind of like what happend to the earth, and what didn't happen with the advent of hairless overclocked primates that have eaten most of the biosphere.

(Ultra long term: Cosmic radiation, even with good shield would accelerate evolution due to increase mutation).


These guys seem to do pretty well...


@85: Yeah, ecospheres are fun. We used to do that as a lab experiment every semester, except we used aquarium snails instead of shrimp, and a 200 ml screwtop jar. The best systems lasted a few months, which is well short of the ecosphere, but pretty damn good considering how tiny they were.

As for the trophic pyramid, that's basic ecosystem science, and the 10% is what they teach in college. It's right within an order of magnitude. The real lesson is that there are ~90% energy losses when something eats something else. That's why Charlie was arguing for a vegan diet, and he's got a point.

This 90% loss doesn't hold for elements, though, nor for necessary amino acids, vitamins, omega 3 fatty acids, or similar necessities. That's why I think a few, well-chosen animals can be a good thing.


Will@43: I thought my three trees voting was completely original; haven't read any Iain Banks. While I'm waiting for the next book from Charlie, can you please point me to the book you're thinking of? My search-engine-fu is apparently lacking and it's time to try a new author.

How closely does it relate to the topic which Charlie's asking about? :)


I'm surprised some prior UK work on a related subject hasn't come up, by one Captain
J. Cook. Long sea voyages, of course, had similar challenges - and, in fact, they WERE suicidal for 80-90% of their crew in the early days.

Survivability of sea voyages kept slowly improving, but only grew to acceptable modern standards during Cook's voyage, and on the ships that copied what he did. So, that should give a good profile of one successful diet low on variety. Of course, he had tons of external inputs - like salt pork,water, ship's biscuit, limes, and some other things.

I, and plenty of other people need meat to thrive, and lots of meat when I exercise; don't think the salt pork was so optional (yes, I've almost certainly tried whatever you're thinking might work as a replacement, short of vat-grown meat...).


Kelson@82: strictly speaking, it's turnover that matters. There are ecosystems where the consumers out-mass the producers. They're prone to sudden and massive collapse -- some of the small commercial fish species like anchovies fit this category, and whenever there's an El Nino year fishermen on the west coast of South America come home empty-handed because the fish stock has collapsed.


Ironically, the idiots from the previous thread who vaguely wished for "ecology equations" had a tiny shred of a valid point.

Almost everyone who is NOT a Population Biologist makes back-of-envelope guesses that are orders of magnitude too small.

For EVERY species onboard, you need to have enough organisms to avoid genetic bottlenecks (Founder effect, or isolation and genetic drift). You have to use the equations of population biology. Even with panmixia, you need order of magnitude a million people, for example, to avoid losing genes over time.

Some of these equations, by the way, were repurposed from Astrophysics while I was in grad school (1973-1977). One again, Physicists moved in and picked the low lying fruit in someone else's field, as they had strong Equation Fu.

My genetics advisor lived with Hutterites to study what was previously seen with Amish.


"Eastern Pennsylvania is home to beautiful farmlands and countryside, but it's also a gold mine of information for geneticists, who have studied the region's Amish culture for decades. Because of their closed population stemming from a small number of German immigrants -- about 200 individuals -- the Amish carry unusual concentrations of gene mutations that cause a number of otherwise rare inherited disorders, including forms of dwarfism."

"One form of dwarfism, Ellis-van Creveld syndrome, involves not only short stature but polydactyly (extra fingers or toes), abnormalities of the nails and teeth, and, in about half of individuals, a hole between the two upper chambers of the heart. The syndrome is common in the Amish because of the 'founder effect.'"


Oh shit, I forgot mycorrhizae.... That's a boatload right there.
All this business about a germ-free human seems like nonsense to me: we have intestinal symbiotes for a reason, namely to aid with the enzymatic breakdown of our food. The reason the germ-free rats look silly (re: HUGE stomach) is that their gut is incapable of breaking down food as well on their own. (ok ok, citation needed.) Sure, we could c-section us some sterile human, but why on earth would we want to do that? Maybe some kind of standardized human-gut population assay could be done and a big vat of yogurt-like substance with the right mix could be kept on standby, but in general... If it ain't broke don't fix it?

Also regarding animals as bioreactors: correct me if I'm wrong but I believe it's still necessary to compost the shit in order to make it easily usable? I'm sure worms would love to helm verimcompost, but perhaps, if turnover is not to be as high, unneccesary?

Phil@87 Culture GSV's are normally run by three ship Minds, voting. I can't actually think of a book in which this has been depicted, but it is mentioned in passing in Excession that the Sleeper Service had somehow got rid of its other two minds...


Heinlein hinted at the complexity in Farmer in the Sky.

Acres and Hectares: a near-enough approximation is 2.5 acres per hectare. A hectare is 10,000 square metres, and an acre is 10 square chains, or one-tenth of a square furlong.

The advantage of an artificial environment is that you can run different sections at different stages of the season-cycle.

Don't forget the space needed to access and harvest crops.


Are the crew going to want to get off this vehicle eventually?

If so, I'd advise against the germfree approach, because you don't develop much of an immune system in a sterile environment. Look at a nice, rich microbial under-layer (compost heaps, soil bacteria, mmm!). Not only does this give us a (reasonably) constant gas exchange (balance that with H's tree-derived CHO binding) but also gives us somewhere to dispose of human waste.

For the meat eaters, we could get away with just chickens (in a simple system). Chickens and goats, maybe, if you want to get more complex.


I have to say the germfree human thing seems like a huge red herring. Loads of unknown unknowns and known problems. Benefits not obvious. The Soviets didn't encounter any problems as a result of not doing it.


@19: If you put "fire and forget ecosystem" on left end of the scale and "constant monitoring and tweaking by humans" on the right end, then there is still a lot more room farther on the right. Theoretically, ALL substances humans need could be artificially synthesized in bulk, and then the whole problem of ecosystem with its nightmarishly complex nonlinear relationships goes away. So the answer to Charlie's original question is ZERO. (Or one, if you count humans.)

I suspect this approach requires at least one genetic modification in humans -- remove the desire for food to have taste.



Erm... ecosystems tend to be self organising and self stabilising. Trying to design and build one and have it not be a spectacular clusterfail is the wrong approach. How about trying evolve one, using natural selection.

Get a great big rotating habitat filled with dirt, stuff a huge variety of organisms in there seal it up and give it a year or ten with sunlight input. Once the ecosystem has sorted itself out, introduce a human burden to the system *slowly* and sustainably. The CO2 a human cycles could be buffered while the ecosystem catches up. Likewise other compounds elements. But the secret is to allow for adapation.

Yeppers. This is one of those descriptive science things as opposed to predictive. Also, hideously expensive, but then, the low-hanging fruit in other sciences has also been extensively picked over as well. I'm surprised no one has written any good short stories on the subject(though there were a raftload of Settlements . . . In Space! items back in the 70's.) Otoh, I grew up on about 200 acres with no running water and no electricity; we - two adults, five children, three dogs, two goats, chickens, and assorted geese and cats - seemed to get along pretty well without too much in the way of external inputs. Of course, along with stuff like canning jars and paraffin, that would also include coffee, sugar, etc. Perhaps the best way to do this sort of thing would be to completely isolate, say, 1/4 of an acre(modulo water and a few other inorganics) and see how long it takes to die off, and why. Even that would be fairly expensive if one was determined to impose an absolutely impassible barrier.


@94 "Theoretically, ALL substances humans need could be artificially synthesized in bulk, and then the whole problem of ecosystem with its nightmarishly complex nonlinear relationships goes away. So the answer to Charlie's original question is ZERO. (Or one, if you count humans.)"

Well, it needs to be "capable of supporting humans," not actually contain them, so zero is the right answer. A bit past currently implementable technology, however.

"remove the desire for food to have taste"

If you can synthesize complete human nutrition, you can synthesize various flavoring elements. Not to do so sounds like something a government space agency would do to 'cut costs.'


@89: Absolutely Chris. I've only heard of inverted trophic pyramids in aquatic systems. Whether it would work for people (i.e. a small vessel of fast-growing algae feeding the crew) is one of the crazy solutions I mentioned above. A ludicrously high turnover rate for food production might work, but the failure modes would be interesting. That is true of the slow turnover systems too, I guess.

@90: I didn't forget the mycorrhizae. However, they might not be needed. BTW, where did you think that 20% of photosynthate into the soil came from? Mycorrhizal studies.

Here's the experimental results for AM mycorrhizae (the most common type): if you pair your average variety of plant and ecotype of AM fungus, you'll form a mycorrhiza (which is the symbiosis, not the partners). Do this many times, with random partner choice. Then chart plant growth against growth of uninfected partners. What you'll see is that some plants up to double in size over the uninfected states. Some plants are half the size of their uninfected partners. Overall, the growth benefit of AM mycorrhizae is on average zero, plus or minus 16% (in the system studied). The kicker is that it depends, not on species, but on ecotypes of plant and fungus, and the effect varies unpredictably by environment. Some researchers believe there are super fungi in particular systems that benefit the growth of all plants, but that might be simple luck of the draw.

Mycorrhizae might work, but they're another layer of complexity. If you're running a mycorrhizal system, you need to test all possible combinations of plant and fungal partners in your biosphere before you set out. Since most modern crop plants are actually bred to grow without mycorrhizae, I'd use non-mycorrhizal varieties as the default for a space mission. Keep it simple.

On this planet, I'm mostly in favor of using mycorrhizal inoculum. The stuff they sell...not so much. Get a soil sample from under a healthy plant instead. The random pairing approach implicit in using commercial inoculum is (on average) worthless, although it might work occasionally.

@94: Bulk synthesis might work, but the question is whether you can make devices that run on the equivalent of sunlight (~5000 watts per person, equivalent to a half hectare per person of sunlight), require simple inputs that can be recycled from their outputs, are easily fixable, with the parts buildable using materials available in the ship, do not require very sterile conditions (as with most current cell culturing) or toxic reagents (as with conventional organic chemistry), since both of these are hard to maintain long-term in a spaceship. Oh yeah, and this system has to be superior to a farm system in at least one of the following: price, weight, or durability.

It's theoretically possible. Whether it's a reasonable solution is unclear, because our technology isn't there yet.


It occurs to me -- and I can't believe I didn't see it before -- that the answer to this problem is: "I don't know, let's simulate it."

Which, as it happens, would make for a dandy sequel to Glasshouse.


5000 watts per person, equivalent to a half hectare per person of sunlight

This is out by a factor of between 200 and 1400, depending on the location of your hectares.


So far as the human side of the equation goes - I seem to remember that historic evidence suggests that any human population in an environment not capable of supporting about 2,000 for most of the time is likely to go extinct within a few centuries. A smaller population may well survive a single disaster, but is likely to be wiped out by a second one.


Recycling waste products back into the cycle is key, so bacteria for that. Then you can probably survive farming oysters? They're edible raw so you don't lose all the vitamins.


Oops, no - recycling their shells might be a problem. What about fish eggs though? Not a lot of waste products and they would be nutrient-rich.


Oops, no - recycling their shells might be a problem. What about fish eggs though? Not a lot of waste products and they would be nutrient-rich.


@100: Yup. Math error. Thanks Nick. I'm assuming 5000 m2 per person, so obviously it's not 1 w/m2. The World Meteorological Association defines ambient sunlight as 120 w/m2, and I learned it as 180 w/m2 (at the bottom of the atmosphere), so I think 200-1400 times too low isn't accurate either. My assumption was supposed to be 600,000-900,000 watts/person, though technically that could be divided in half, because plants experience nights as well. You could probably get by with a small fraction of that if you made sure most of the energy went into lighting plant leaves, but my original 5000 watts/person is giggle material. Thanks for catching that, Nick.


All-algal or bacterial diets are not an option; even a microbial source for a glucose drip is more complex than you may assume - because they may not use glucose in their metabolism.

Where large multicellulars are variable in our structure - microorganisms are variable in their biochemistry.

A lot of algaes and bacteria don't even store energy as starch ( the way land plants do.) They use other polysaccharides, which we can't digest. Green algae do produce starch, but have other metabolic differences. We are evolved to eat land plants, we don't have the enzymes to digest other photosynthesizers.

Even if they give you enough stuff you can digest, the stuff you can't is a problem. 'Cause there's something in your gut that can, and will reproduce out of all normal proportion. Giving ya painful gas, for example - that's what lactose intolerance is, your bacteria digesting a bunch of sugar 'cause you can't. Note the digestive problems in the failed NASA experimental attempt at algal feeding mentioned earlier.

Maybe you could put a little algae in the diet...if you were really careful.

And reworking other photosynthesizers to be digestible is not a minor genetic engineering project like golden rice. It involves whole multi-enzyme metabolic pathways.

And there's a larger point here: the blithe assumption that we can live on algae and bacteria is one of a category of blithe assumptions that we can throw a complex past out the window and live totally differently.

Maybe in some cases there is a much simpler solution. But maybe not. There will be many unforeseen pitfalls.

Speaking of which:
I would suggest the need for agricultural and other work to maintain the ecosystem is a plus: it's likely to help keep the humans and their on-board society sane.


OK, when I said "they may not use glucose in their metabolism" that was nonsense. That is the lowest common denominator of metabolic variation. But that glucose may not be stored in a human-digestible form like starch. Heck cellulose is made of glucose; I invite ya to live on that.


1366 W/m^2 at the top of the atmosphere (that was my 1400). 1000 W/m^2 at sea level, for a square metre perpendicular to the sun on a clear day. Divide by 4 for the Earth being a sphere to get 250 W/m^2. Then clouds bring it down to your 120-180. First job for our hypothetical space farmers is to choose or create plants which are efficient 24/7 at higher light levels.



I think you're allowed to assume some level of processing to extract nutrition from the food-source species. I suspect that the equipment required to convert cellulose into human-edible food would need less space and energy than a spacebound reproduction of the arrangements that render olives edible*. Alternatively, of course, you go the Quorn way, and choose/design your food organism so that the human-food is the organism's waste product. Then you simply clean up what it leaves behind :)

* Olives are easy for a low-technology culture because you only need (fresh-and-or-salt) water, water-tight containers, and lots of time.


Maybe we could hack our gut bacteria to process cellulose? Eh,what could go wrong :/


@108: Nick, increasing photosynthetic rates is a great idea.....BUT.

Running plants 24/7 isn't a problem so long as you don't want them to flower. Many cue their flowering to the about of darkness they get. To some degree this could be bred out, but you do need plants to time their lifecycles in a way that maximizes their productivity for us. For example, a big plant produces more fruit than a small one, so you would probably benefit more from the plant growing for months and then fruiting extravagantly, rather than growing an inch and producing a single, rice-grained sized fruit.

Bigger problem is photosynthesis itself. The system is quite efficient, but it has two flaws. One is that it normally maxes out aroun 20% ambient sunlight. I learned this in light, not energy terms, so I'm not sure if that really means 40 watts or not. Second problem is it gets wonky if there's too much oxygen around. Plants have all sorts of mechanisms for shunting energy aside and keeping oxygen from poisoning the reaction too much, and C4 plants like corn can push up above that 20% line, but not indefinitely. There are upper limits to the chemical reactions in photosynthesis, and above those limits, things start overheating and breaking down.

The bottom line is that you'd have to design a new photosynthetic system from scratch, and making it edible would be an additional challenge.

You can grow a plant in ridiculously high levels of light (think deserts), but said plants are going to be highly reflective, full of antioxidants (which repair the damage caused by excess energy oxidizing things it shouldn't), and so forth. Probably this is less productive than just spreading the light over a greater surface area of plants at lower intensity.


I can certainly imagine that a few extra antioxidants in your diet might come in handy if you're exposed to cosmic radiation. (Or has that one been debunked as Daily Mail quackery?)


Regarding power: if you are on a spaceship, odds are power to make food is not significant compared tot he problem to make it go.

It occurs to be that the original question just specifies "stability." Later on, Our Host mentions a Mars mission, which makes me think "stable for about a decade." I think this is pretty doable simply because you can cheat shamelessly bringing redundant supplies of everything but the macro nutrients. There's plenty of ways for it to go wrong, sure, but a 20 year well funded and managed RnD program starting right should be able to make something pretty darn good, IMO. (The main remaining problem being testing for 10 year endurance takes at least ten years.)

If we start talking about multiple decades, say a 50 year hermetically sealed Cthulhu shelter, things get iffier. Testing the system takes longer, more things need to be locally produced, and there's more opportunities for things to go horribly, horribly wrong. Additionally, all the machinery required to run things needs to be insanely rugged or repairable on site. Also, the crew needs to be capable and stable enough to keep it all up with no outside help.

If we make the "stability" requirement multi-century, as with a generation ship that will take millenia to arrive, then we have to make stuff up. We don't have good biodata on that scale. Modern machinery that is centuries old is an oxymoron, yet the system will need high tech stuff that lasts centuries. Even if you have that covered, the sociology of stable cultures over millenia is unprecedented, even among hunter gatherers, AFAICT.

There's interesting noodling room in the question.


@113: You nailed it, Martin. In fact, our flying ecosphere really doesn't look remotely like anything we have a good analog for (other than Biosphere II and similar experiments).

One thing I'd quibble about is stable. I really prefer recoverability over stability, by which I mean that you can keep at least some people alive under all but the most catastrophic conditions, and always have something to rebuild with. This argues for the equivalent of a damn good seed bank, and the willingness to use and restock it.

However we look at it, making and living in one of these things is probably more complicated than creating an aircraft that can be rebuilt from the inside in midflight. Maybe we should try that first as a warm-up exercise?


It seems to me that a number of experts in growing plants in high artificial light levels concentrate on a species which is not particularly reflective or anti-oxidant, and succeed in producing a great number of flowers. Or at least buds. I guess they know one or two things about 24-hour growth too. I might ask my neighbours.

Deserts are unusual not in their light levels but in their dryness. The subtropics can be greened by the application of water and clever farming techniques (see 20th-century Israel) or desertified by the absence of same (see 1st-19th century Israel).

Possibly the anti-oxidant etc etc applies to high levels of high-frequency photons. I'm assuming we can tune our grow-lamps however we want.


@115: Too bad you're using a name, Nick, or we could talk more specifically about the odd physiology of a certain bud-forming plant that grows well under lights. If I recall correctly, the buds form when there's a certain level of darkness. You also need to look at the energy flux hitting the leaves, not going out of the plants. The experimental plant physiologists continually gripe about how hard it is to make good fake sunlight--until LEDs came along, it involved really expensive, energy hungry, very hot lamps.

As for other plants, antioxidants are normal. The basic problem is that the photon hitting a leaf either goes through the leaf or into it, and if it goes into the leaf and doesn't do useful work (photosynthesis or heating the leaf), then the energy has to be caught and/or dissipated, and/or the damage caused has to be fixed. Plants have a bunch of ways of doing this, and these unsurprisingly include things we plant-eaters use as vitamins (notably vitamin A precursors).

As for the tropics, the plants there keep cool by blowing off huge amounts of water (yes, there's 100% humidity--the trick is the inside of the leaf is slightly hotter than the surroundings, so there's a continuous outflow). In the desert, plants deal with having surplus light by reflecting it off, which is why they tend to be gray or white. The basic point stands, which is you can't arbitrarily crank up the light and expect plants to crank up production.


Well, if we have cheap energy on our space mission we can keep our plants cool, wet, and brightly-lit as well. And we won't even have to steal electricity from the streetlight circuit. I can see that there are limits to this, and they might well be heat-related, but I'm not convinced that we are bumping up against them, at least for food crops, if only because of the 4x factor we can get by running the grow-lamps overhead 24 hours per day.


Some plants schedule their lives by hours of daylight, and others are a little more baroque. (The California buckeye can go completely dormant in midsummer if the previous winter was dry.)


How do plants that grow briefly in the boreal midnight sun deal with that?


You're forgetting the higher plants need periods of dark otherwise they won't flower, set seed/fruit. Different species require different balances of light and dark. You can fiddle with this to some degree - it's been too long and I'm too lazy to dig out the relevant paper but you can reset the plants 'clock' for day/night by using infra red illumination in an otherwise dark setting if you use far infrared you set the clock one way near infrared the other, but you still require a couple of hours of dark.

Agreed you can't 'crank up' the illumination to increase yield. The C4 plants (Maize - corn in the US) is a more efficient at higher light intensities than Calvin cycle (C3) but in climax situation C4 plants are shaded out. At low light levels photosynthetic efficiencies can reach 30% for C3.

Simplest system is to stagger two or more biomes into different time zones

Then there's the entropy factor Larger more complex systems have a longer natural lifespan. Biosphere 2 just survived a two year closure. With additional experimental setups, and the simple expedient of increasing the biome sizes the lifespan could be reasonable doubled or tripled. Beyond that uncertainties stack up.

Practical realworld example; a tank of tropical fish or a marine aquarium if you prefer For a mature well maintained freshwater tropical tank the smallest tank that requires only weekly maintenance is around 60 litre capacity. The interval can be extended to two to three weeks if the volume is around 100 litres (as a practical measure you maintain on a weekly cycle but while a 60 litre aquarium gets a 10% water change/top up a 100 litre gets a 5% change and 200 litre can survive months with basic top ups plus filter maintenance. Marine tanks requirements are more rigorous and it's very hard to maintain a healthy aquarium below approximately 100 litres and a more reasonable minimum volume would be 150 litres.


@ 113:

I'll have to beg a dispensation from Charlie for mentioning religion on his blog, but given the very limited number of organizations which have managed to last a millenium or more, if you want to send out a multi-generation spaceship it would seem that you might do best to have it crewed by either Roman Catholics or Buddhist monastics. (Alternatively, you might try having it crewed by a Japanese construction company or inn, a French winery, or a German restaurant or brewery.)

@116: Doesn't the peculiar chemical synthesized by the plant under discussion function precisely as a way of blocking and absorbing high levels of UV without causing leaf damage? ISTR that was why it was traditionally grown at high altitudes in or near the tropics.


Chris @109. If it was economical to produce food by organic-chemical manufacturing processes, wouldn't it be done here on Earth? All costs will be higher in space. And yes, economical matters - at some point the product of one human's labor falls below the needs of one human's survival....

I think the "laboratory food" idea is an example of the same kind of blithe assumption as the algal/bacterial diet idea....


The only currency that matters on a space ship is mass - If you have the choice between feeding a crew of a hundred with one farmer + 100 tonnes of soil and assorted insects, or feeding the same crew of a hundred by bringing 33 chemistry/biology double phds + 50 tonnes worth of hydropondics, chemisty kit, and living space for those 33 extra workers, the second choice is superior, because it saves mass.


Thomas@123: That kind of depends on the size of the ship. The discussion here has kind of cemented my idea that if we ever built interplanetary ships worthy of the name, they're going to be big suckers, long-lived, largely independent of whatever mission they happen to be carrying cargo for this year. Lots of skin to soak up micrometeorites, lots of redundancy in the life-support systems, lot of radiation shielding.


@123: Of course, you have to feed and house 1/3 more people, the numbers are going to get pretty close to break even with Mr. Muddy Farmer. I agree that hydroponics or even aeroponics might be a reasonable way to save weight.

@124: Yup. I'm not sure that you want to carry people and cargo in the same ship, though. Human and organismal interplanetary transport would be reasonable with this kind of system though. Cargo ships might be better unmanned.


Well if you can't simply increase and expect higher conversion rates by upping the light levels, you can at least up the CO2 concentration. From memory, I think it was at +-1000 ppm of CO2, additional growth thanks to specifically the abundance of CO2 ensure up to 7% in average for a variety of edible plants. A measure simpler to implement than to manipulate the number or nature of the chloroplasts within the plants.


So perhaps CO2 is another good reason to include food animals, besides variety & B12. I'm thinking if you need to raise the oxygen level having a Thanksgiving Feast maybe better on morale than putting people to sleep or dipping(?) into any emergency tanks.


P.S. a couple cites on my tongue-in-cheek suggestion of Buddhist monastics for generation ships: White Horse temple in Luoyang, China was founded 68 AD (though I don't think it houses any monks since the Cultural Revolution.) Shaolin Monastery was founded in 477 and still has monks in residence. Shitenno-ji temple in Osaka, Japan was founded in 593 AD and is still functioning, as is Horyuji built in 607 AD. Kenchoji temple in Kamakura is still a Zen training monastery just as it was when founded in 1253 AD, and so on. I figure people have heard of the Catholic church.


@128: Clifton, that's a suggestion that's analogous to suggesting we send skeletons because they last longer than bodies. In both cases (and in others), you have a tradition that eventually became monastic (i.e. NON-BREEDING) that served various purpose as part of a larger society. We need the whole body, not the bones.

Also, both institutions are a bit like grandpa's axe. You know the one where we've replaced the head twice times and the handle four times, but it's still the axe that grandpa has used. The Catholic Church has a lineal descent from ancient times, but so do almost all other lineages of Christianity. It just has the treasury and the name. Ditto with the people using the Buddhist monasteries. Same name, same building site, but what else is the same? Not that change isn't necessary for continuity, but you really need to look at how continuous these institutions really were before using them as examples.

I agree that China's an interesting example. There's an old book (Farmers of Forty Centuries that's worth reading). It's still in print, and readily available from Amazon. The question of how they managed to cultivate the same fields for centuries is relevant here.


Sheesh, is the explicit phrase *tongue-in-cheek* not clear enough for you?


@130: Nope. Tongue-in-cheek is not understood. I don't read carefully, especially when you point out a real problem.

Continuity actually looks like a really nasty problem to solve. Losing all your engineers or agronomists to a palace coup or random killing spree is kind of bad, if you're in a STL starship or on a remote space station, where relief is years away if it's available at all. Even if there's no massive crisis, just making sure each new generation has the raw talent and the training to do the necessary work is a major hurdle.


A bit off-topic, but related to the question of counting the number of organisms in a biosphere: Expert's map the body's bacteria

"They identified more than 4,200 species of bacteria, but only about five were shared by all 51 participants.... our human genomes vary so little but our repertoire of microbial genes vary so much"

Even counting, much less understanding the number of organisms in and on a normal healthy human body is still challenging.


@Will (87) - whereas the Banks reference to the three ship minds is clear now, I would recommend Swanwicks /Vacuum Flowers/ for some space trees.


@Phil (87) - whereas the Banks reference to the three ship minds is clear now, I would recommend Swanwicks /Vacuum Flowers/ for some space trees.

@Clifton (121) - the oldest German brewery, Weihenstephan, actually is/was a monastery. /Anathem/, anyone?


Is institutional continuity really such an issue ?

compare those institutions mentioned that have persisted to ones that haven't (eg. corporations, governments, etc) and you notice a trend: the ones that survived largely have an unfulfilled or uncompleted purpose (or are outliers: some companies that have continued for centuries after their peers).

The ones that "failed" often did so because they outlived their role: eg financial backing moves from one company to another in search of profit; the backers have no particular loyalty to keeping the company alive in modern capitalism.
Individual governments (or even countries) over the millenial timescale are the power vehicles of rulers or dynasties.

If you compare that to the task of ensuring continuity of skills and infrastructure on a century-scale starship colony, people have a definite vested interest in keeping that going. Separating that from the day-to-day government of the starship (attractions of power) would be a good idea, but doable, I'd say.


So. Anything large enough and complex enough to sustain humans for decades with no external inputs are going to be about the size of an O'Neill habitat, it seems. No generation star ships, alas. Anything much smaller and you're better off carrying your own groceries. Back to the frozen sleep alternative.

Having 20,000 people in cold storage with only 100 people awake at a time seems much more sustainable than having all 20K awake. I'll go out on a limb here and say that putting people into suspended animation for years at a time will be a solved problem long before a sustainable ecology keeping the same amount of people alive and active is gotten right.


@135: Hi Alastair, it's equally possible to argue that long-lived institutions are a black swan in Taleb's original sense. Their success is totally unpredictable at the time, but in retrospect, people make up stories to explain why they succeeded, whether or not the explanations make sense when compared to what happened. Taleb's Black Swan is instructive reading for those of us who attempt to make predictive science out of descriptive science because of this problem.

This is the (deliberately) scary interpretation, because it says that we may not be able to predict which ones will succeed, or why the ones that succeed do so.

So far as life support goes, the hypothetical generation ship needs to be able to be run and maintained by average people. If the system needs people of very high intelligence or a highly specific innate talent set, there will be problems, because the children of such parents tend to be closer to average than their parents were--and there have to be enough extra people to handle reproduction. Obviously we can compensate through technology, but then we switch the continuity problem from life support to tech support.

Not an insoluble problem, of course, but it is a chronic problem.



(Almost) Everything is more expensive in space, yes, but the relative costs are what matters. My point is that making edible olives the traditional way is cheap on Earth because you only need lots of water and time, both of which are, to a first approximation, free to the farmer - he has them anyway (the water can be reused afterwards, and the farmer can get on with other things while the olives soak). But making edible olives like that in space is totally impractical; transporting tonnes of extra mass just to turn sour-tasting berries into yummy food in small quantities? Tying up large chunks of your spaceship's mass-restricted internal volume for weeks at a time in the process? Fantastically expensive; it might well be cheaper in space to have a machine that takes raw hydrocarbons and a bit of nitrogen and turns them into olive-flavoured paste. Much more expensive to do that on Earth, but a tiny fraction of the mass to boost into orbit and transport from planet to planet. Especially when you count the soil that you don't need for the olive trees you aren't growing :)


The discussion here has kind of cemented my idea that if we ever built interplanetary ships worthy of the name, they're going to be big suckers

Like ship ships?

BTW, the long-lasting institution question, including breweries and churches, was thrashed out on the High Frontier thread. Basically, they are possible, but you might want to think about planning for the institution to change its purposes over time.


Alex: zactly. Only more so.


Till @133, for old space trees, try Niven's The Integral Trees. I don't like the sequel The Smoke Ring as well.


Chris L @ 124:

One suggestion that's been made for large-scale interplanetary transport is to put great big ships, perhaps made out of asteroids, into permanent resonant orbits ("ballistic cycle trajectories"). By putting several of them in various trajectories, you can guarantee having 1 inbound and 1 outbound vehicle coming up to a short period arc between the planets (4 vehicles are the minimum for Earth/Mars cycles). Of course they never go into orbit around the destination planets, so you'd need shuttles to carry passengers and cargo between vehicle and planet.

As far as the long-term stability of (relatively) small ecosystems is concerned, the experiment has been done at least once already, at Easter Island. There's strong evidence that removal of all the palm trees took a major source of protein out of the food supply, overfishing in the immediate area of the island removed the only other source, and the human population crashed as a result. Will algae produce all the amino acids humans usually get from animal tissue? How about taking along an animal tissue culture instead of carrying livestock?


heteromeles: Raw IQ of the children is only half the problem. Most children of smart and educated parents on Earth today have a choice of career. The generation ship leaders will likely have to assign jobs based on a particular aptitude. Therefore, offspring traits such as loyalty and dedication should be added to the desired mix.


I wonder what the ideal age would be for generation ship career assignment. Maybe it should be 14, which is late enough to measure ability but before the typical "rebellion" phase.


Actually, setting up a stable ecosystem fits in really well with protecting its residents from cosmic radiation. Put a thick layer of dead or frozen water and soil around the warm and live stuff, and keep the people in the middle. I also like the idea of setting up two independent ecosystems for redundancy. Tether them together and spin them up for gravity. It will be a whole lot easier if we don't have to deal with microgravity at the same time we're trying to debug the ecosystem-in-a-can.

There need to be very substantial safety margins. It isn't worth going for faster, smaller, cheaper in ecosystems unless you're willing to keep throwing away entire missions and their crews and trying again until you get it right. The best possible safety margin would be to include other climax predators in the ecosystem. Maybe wolves and jaguars on land, salmon and a pod of orcas in the aquaculture system. If you can keep them alive, keeping the people alive along with them is a piece of cake. If things get tight, you can start throwing predators out the airlock and eat what they were eating, with the caveat that the more we shorten the food chain the greater the risk of catastrophic failure.


@143: Aside from keeping a population of jaguars happy (Orcas are a wee bit, I don't know, big), why not use, say army ants. You can prune a lot of the biomass out of a swarm and it will still regenerate.

That said, I think the Trobriand islanders (among others) perform the same service in a simpler, more useful way. In traditional times, and to some extent today, they normally generated an agricultural surplus. They competed to generate an agricultural surplus, and display their surpluses with great pride, letting the excess yams rot in storage.

The neat ecological trick they're performing is that said surplus *could* be used to do things like feed warriors if someone attacked them.

Similarly, I'd suggest that you want to have an spaceship ecosystem (or ecosystems) that normally generates various useful surpluses that have to be recycled periodically. The surplus is the system's safety margin.

Personally, I wish our society would stop fetishizing efficiency and realize that surplus capacity in some functions can be a really good thing. I can dream.

In any case, potlatching might become a hallowed custom in space. There's something to think about.



I wonder what the ideal age would be for generation ship career assignment. Maybe it should be 14, which is late enough to measure ability but before the typical "rebellion" phase.

This touches upon a reason why I don't care for the generation ship idea. There seems to be something very dicey, ethically speaking, about imposing these sorts of conditions on one's children and grandchildren (and for that matter, one's great-great grandchildren.) Somebody or several thousand somebodies want to go and live The Dream? Fine. But that doesn't given them the right to dictate these same options to their children. And practically speaking, how long would it take for a system like this to devolve into a hereditary aristocracy? ObSF: "Rogue Ship", wherein the ships officers are found to have a 'natural aptitude' for officer-type duties while the poor shmoes down in hydroponics just can't seem to have any offspring capable of grasping the subtleties encountered in command positions.


Do the testing before sending out the ships. It is relatively simple to do RnD before making the production model, and almost always worth it. Testing via interplanetary canned pumas is kind of fun in a mad science sort of way, but a rampantly silly.

Re: Institutional longevity - the question includes safety (how much revolution can the system take) and mission continuity. A mission that is going to mutate en route is hard to get funding for. I think this will kick in before generation ships - a 10 year mission without options to swap out crew needs serious psych prep to make sure everyone stays committed for the whole decade.

@145 Generation ship morality. In practice, all parents make long term choices for their descendants. I see no reason a reasonably safe and clement generation ship is more immoral than, say, immigrating to a dangerous country.

The SF problem is safe and clement genships are boring, so we see unsafe ones, and/or ones with collapsing governments and/or ecosystems.


ScentOfViolets: If all children are clones of each original passenger, would that be more ethical? The clones would still be subject to early career manipulation, but they would be less like new individuals and more like an extension of their originals. Pseudo-immortality can have advantages! And this would mostly solve the problem of gifted parents producing mediocre offspring.


Josh @147: would it be ethical for you to make decisions like that about your (hypothetical) twin brother? The answer is "no" if we're consistent about the ethical framework, and identical twins are probably more similar than clones.

FWIW I agree with martin@146: parents make all sorts of decisions that affect the lives of their children, sometimes even grandchildren.


One general point, whether it's about a generation ship or a space base, is that there's a cultural issue here. Literally, it will work if there's the culture of the ship says that running the ship and keeping everyone alive is the one right and proper way and reason to live a life.

We're sitting in front of our computers looking at all these problems, and rightly so. We're no more a spacefaring culture than Joe Average Papuan is. Most of us would go nuts on even a working generation ship.

It's not just a matter of recruiting and training, it's also a matter of people wanting to live a ship life. It ain't us.

So, there's yet another interesting question: is there some culture or subculture on this planet has the precursor elements of a spacefaring culture? Astronauts? Freegans? Technomads? Off-the-grid millenial nutcases? Who would we shove in a tin can and send off to infinity?


Chris, there's a big difference between a sibling relationship and a parent/clone relationship. The parent of each clone would still be a learned adult making decisions for an inexperienced child. Maybe you thought I meant a fully formed adult clone that doesn't have a standard childhood.


When people say "generation ships," I hear "Columbine in space."

Seriously. Generation ships have all the problems of a small town, plus the constant stress of ecology maintenance and energy rationing, and you can never leave. I love the concept of a space-based ecology that can sustain the existence of multiple generations of humans, I do. But "we'll have to tweak the kids, of course," sounds a lot like "we'll be greeted as liberators," in terms of actual planning.

Some of your kids will want to get off the boat. And they'll probably have good reason. If you're lucky, you'll find an opportunity for them to leave. If you're unlucky, the least of your problems will be their deep and ceaseless resentment of you. This is an environment where even non-violent resistance can lead to instant famine and disease, and a community small enough that a single psychopath can make a world of difference.

Whether this is an ethical decision to make for one's children is, I think, irrelevant. In this situation, you'll be lucky if the decision to have children has anything whatsoever to do with you personally. What is a relevant question of ethics is how to plan for the system's failure.


Madeline: I'm coming to the conclusion that generation ships are unlikely to work because what they need is a bimodal society -- one that operates the ship during cruise, and then flips over into generating colonists when it gets to the other end of the journey. And suppressing the currently-undesired mode is going to set up some ... interesting ... social tensions. (Cruise mode: population growth is a really bad idea, going off and doing your own thing is a really bad idea -- this society would need to be sufficiently disciplined to make Japan or Sweden today look like crazy hippy anarchies. Whereas in colony mode, the whole frontier ethos starts to look good, as does having lots of children. Spot the difference between these societies?)

I think I have some thoughts on this subject to jot down, but right now I'm getting ready to check out of a hotel in Oslo and I won't have time to do it before tomorrow.


>We're no more a spacefaring culture than Joe Average Papuan is. Most of us would go nuts on even a working generation ship.

Well... we're all sitting on quite a big spaceship, even though it moves rather slowly and we have no control over it's destination. Much like the descendants of a generation ship that has gone wrong.

We have am awful lot of redundancy here on spaceship Earth, things can go badly wrong but people still survive.

Again I'll say that any self sustaining generation ship, which is what we're really interested in, has to be a minimum of an order of magnitude larger than most people have suggested.

We don't even have the beginnings of the infrastructure required to build such a ship. A space elevator or anti gravity drive is going to be needed at the least, just to get the earth and water needed to fill the sucker into orbit.


Humans are certainly capable of surviving for centuries in isolated populations of a few hundred. We've done it many times before. For that matter, for much of history in Europe, almost everyone would live and die without ever leaving a small community. Just because you can't imagine living in a village without going crazy doesn't mean that it's beyond the limits of human endurance.

The kids will be fine. Kids are incredibly adapatable.

For shorter periods, of up to a decade or so, we have ample maritime historical basis for managing those even with small groups, even exclusively-male groups, even in appalling conditions of brutal mis-treatment, ill-health, dietary deficiencies, and physical extremes.

I say, stop worrying about what's going to go wrong and just get on with it (except, of course, it's essentially impossible, see High Frontier thread).


Surely the best way to solve the passenger / colonist problem is for the ecosystem and the ship to grow with the population, rather than to see it as a functioning system which must maintain a constant size and complexity from day one?

So yr thousand-year voyage starts out with a ship which is a solid mass of CHON with a minimally-complex functioning ecosystem - rather more like a zoo than a farm, and lots of trace elements, and a seed, egg and sperm (animal) bank aboard, to deal with the problem of genetic narrowness in small herds. It's got the minimum number of humans as well, to start with (but see below). So you don't have to worry about population control, because you've got a lot of extra stuff with which to make extra space.

Gradually, you grow the ship, and increase the complexity of the ecosystem. This is going to help deal with some discontent. Don't like it in Bubble Two? Well, take your birthright of elements, some chickens and rabbits, your friends, and go off to blow another bubble at the back.

This MO means that you can establish intermediate goals. Sure, you won't be around for Arrival Day, but the default plan has an option for someone to unfreeze the orcas on or around year 450 - YOU could be that person! Orca-unfreezing is cool. Orca-unfreezers are written about in gossip columsn, eat steak, get to choose boyfriends, etc.

Also, if you want to score big community points and gain kudos, why not borrow some sperm or an egg from the _human_ seed banks, and have kid who brings in some more genetic diversity?

Perhaps the designers were thoughtful enough to include some redunancy in the form of a Bussard Ramjet starter kit, which you could unpack and give to the kids to play with. Extra hydrogen! Fast decceleration means a shorter trip time, which is a good secular goal to keep people happy.

Year 800 rolls around, and it's time to start experimentally building the dropships and the comet miners. With a bit of effort and dropships which aerobrake in the target system's gas giants (handwave...), you can overshoot and deccelerate the habitat _after_ you've passed throuh the system, meanwhile entertaining yourself by teleoperating the probes, and corresponding with the cazy fools who deccelerated early.

The design philosophy, therefore, is to give the crew something to do other than be farmers - though if 90% want to do just that, it's cool. They can expand the ecosystem actively, increase genetic relatively diversity passively, and play with machines and designs so as to get their name on a dropship.

So by the time the ship arrives, we've had a playful thousand year interstellar history, all for the piffling price of a measly few extra gigatons accelerated to 0.1 c! Oh, hang on.

ObSF - _The Beat Cluster_, by, I think, Fritz Leiber.


there is an obituary of the guy who built the big_bad_boat from 'the Jennifer Morgue' in today's Guardian (nov 9)


@154: Charlie, I have to disagree with you on this one. If you're colonizing a new world, you're almost certainly going to have to live in a little bubble on said world, and expand by building new bubbles. This goes for whether the new world has a biosphere or not.

Living in a bubble is what the generation ship people have been doing for a long time, so while some fundamental things will change (such as gravity, daylength, and access to some raw materials), the ship culture should be able to handle it.

This is akin to the Polynesian expansion. Their culture evolved on coral atolls, and they: a) lost a lot of technology (such as ceramics, and potentially metal working) that wouldn't work on an atoll, and b) evolved some technologies (ships) and farming and fishing techniques that allowed them to survive in the limited environment of the atolls. When they got to New Zealand, they had the full range of their former resources, but they didn't spontaneously revert to their former Lapita culture. Rather, they created new forms based on their Polynesian playbook.

I'd expect something similar from spacers. The kind imperial expansionist culture we've seen in the western US, Brazil, and elsewhere is too cluelessly wasteful to work anywhere else. It doesn't really work here either, except in the short term. That doesn't mean that spacers won't be expansionist. However, I'd expect something more like a franchise network, where they set up lots of daughter colonies that are basically copies of the ship systems, adapted to local conditions.


I remember an old generation ship SF novel where the "in transit" culture as modeled after the Aztecs, because the author postulated it was the most brutally repressive culture known that didn't explode into revolution.

My understanding is that the author was wrong about the revolution part, at least.

I also recall a novel where there were two castes that didn't interbreed on pain of death. Both castes were monomaniacal about their carefully circumscribed jobs, and incurious and unimaginative. However, crossbreeding led to curiosity, high intelligence, and ambition. May have been the same book.

ISTR that Our Hero (a secret crossbreed) also discovered that some previous generation had actually arrived at the colony system, but decided to go on, since by then leaving the genship was unthinkable...

...ah Harrison's _Captive_Universe_.

(Tangent) It seems one possible approach here is to colonize the outer comets, and just let peeps (in whatever form) diffuse through the galaxy. _Hyperion_ postulated something like this.


@160: Check out Brin and Benford's Heart of the Comet. That dealt with the idea of colonizing a comet.

Also Robinson's Red/Green/Blue Mars trilogy dealt with creating domed cities throughout the solar system. While some of it was goofy (like not having them insulated against the ground temperature), he still thought about how colonization of the outer planets and, eventually, stars might occur.


Ken Macleod's Learning the world is the best treatment of this I've read, but he uses a fair amount of hand-wavium. In particular a crew who don't age, at least not in any way we'd recognise.


Charlie: The difference is easy to spot. Those two cultures think differently on just about every important issue -- energy usage, reproduction, rights and freedoms. So either you suppress Civ 1 so harshly that you end up creating a Civ 2 with a Manifest Destiny attitude, or one group is actually the steward for another. That might work: a group of sedate cruisers whose sole mission is to seed colonists without ever really interacting with them. They could travel along blowing their colonial load on friendly planets, spamming and sporing humans this way and that, meanwhile maintaining a critical distance from the end result.

At that point, though, you don't even need humans. The ship itself could do the job of ferrying your Manifestation (man-infestation?) along. And if the probes are just in spore-style habs attached to the main body of the ship, they don't have to wake up or work until they land. That might eliminate the culture issue.


Robin @155 points to a possible answer – we are a spaceship which (mostly) meets the requirements Charlie sets forth. So reformulate the question to read thus:

What is the largest subset or quantity one could subtract from Spaceship Zero (Terra) without measurable detriment to its continued ability to function? Note that I am not suggesting that the reduced ship be capable of supporting all life currently on the planet.
Can we ditch ten or twenty nines of its mass? Are mosquitoes essential, or auks? (How about houses of Congress/Parliament?)

Re mission handover: allow me to recommend Ken McLeod's Learning the World for some interesting ideas on the subject. Suffice to say that the process need be no more traumatic than live mammalian birth... at best.

Perfection is achieved, not when there is nothing more to add, but when there is nothing left to take away.
    - Antoine de Saint Exupéry


re @137, et al - not just getting people with the right intelligence but actually maintaining enough people to maintain a living, breathing expertise in something. Goes far beyond enough people to just have a functioning social unit.

Even with modern data storage you really do need to be taught - not in the lecturing classroom sense, although that also benefits from having a teacher not a video recording - but practical experience under guidance, with occasional access throughout your career (hopefully less over time), to experts in particular niches.

Otherwise you gradually lost the why of what you are up to. Just becomes "because that's how we do it", which is fine until you have a problem. Then you realise you've dropped so far you can't understand the solutions because the solution/explanation is too complicated.

@156 - humans may naturally live in a group of up to a few hundred (and maybe even tend to recreate it with work and friends and family, carving out an urban tribe in the big city), but I'd suggest you need really quite a lot of people if you aren't going to see your tech level gradually decline and your society ossify. Hard to prove I'm right- Jared Diamond argues that the

Then add in your various organisation structures gradually becoming a seniority rules, dead-mans shoes promotion, type situation, unless you can come up with some quite novel social orgs to stop it. And that's aside from the problem of getting enough people at a high IQ level discussed above.

Not really a society with living, breathing, knowledge. End up a gerontocracy following rituals that keep the ship working.


argh... not sure what happenned there, think it disliked the "less than" symbol.

I meant to say:
Jared Diamond argues that the less than 10,000 tasmanian aborigines gradually lost the tech that they brought with them from the mainland, but wikipedia disagrees.


@166: I think when you add a less than, it turns into a tag.

Not sure about the Tasmanian aborigines either. A better example comes from McGhee, Ancient People of the Arctic. The polar Inuit, an isolated group in northwestern Greenland, los the use of bow and arrow, fish spear, the kayak, and other parts of their traditional technology. "They told early Inuit and European explorers of the area that the use of these artifacts had been lost when a epidemic swept through their small population, killing all of the adult hunters and craftsmen who knew how to construct things such as kayaks and bows."

Several people have brought up "Spaceship Earth" and the example is apt, since our culture isn't doing a great job as long-term astronauts either. We're also vulnerable, because a relatively small group of people (all highly trained, highly specialized engineers) maintain the important bits of our civilization including power plants, water systems, power networks, and the internet. There are also logistics experts making sure we get fed and supplied. An epidemic could put us in about as bad shape as the polar Inuit. Good reason to nurture future engineers, I say, and have a backup plan or ten.


Speaking of "Spaceship Earth" ...

As noted earlier by Our Host, the odds of inhabitable planets being nearby are not looking good.

Note from me: a culture capable of making any reasonable generation ship is capable of building quite nice real estate in the Solar System out of asteroids, moons, etc. Some sort of Nivenesque "Belters" if you will.

It seems to me that a more reasonable target for a generation ship would be extra-solar asteroid belts. In this case, the colonization is less likely to be traumatic for our intrepid pioneers.

At least, until the folks they left behind (with millenia of economic head start) get the Nicol-Dyson laser online and start radioing in the equivalent of "dance monkey dance."



Generation ship morality. In practice, all parents make long term choices for their descendants. I see no reason a reasonably safe and clement generation ship is more immoral than, say, immigrating to a dangerous country.

So then you're entirely comfortable with the idea of having to clean toilets for the rest of your life - by law - because your great-great-grandmother wanted to live like a queen and it was legal for her to forward her debt to her descendants? Something tells me(despite what you may say) that you would feel much more comfortable being the spendthrift matriarch than you would being toilet cleaner, third class, just like your da and mum before you, because you are legally obliged to discharge her debts. Like it or not, that's what you're doing with this generation ship scheme.


@157 You may not need to bring live organisms, or frozen sperm/eggs at all at least if you store the individual genomes as digital copies, you could probably take the entire planetary ecosystem with you. With Craig Venter et al. feat of replacing genomes, and creating an artificial organism, you may use these prototype organisms, and advanced DNA synthesis machines to print sperm/egg or even embryos on demand. It also follows that the entire human race could be represented at the genome level as well.
Genome Transfer paper (pay walled): http://www.sciencemag.org/cgi/content/full/325/5948/1693
Minimal Organism: http://en.wikipedia.org/wiki/Mycoplasma_laboratorium

I guess I was also wondering about the prospect of competition. If a large amount of your ship mass is dedicated to a biosphere than you will spend an inordinate amount of energy/mass on marginal dV. If a hundred years later, a ship implements some iteration of the above scheme, they could spend their mass budget on more reaction mass or whatever and beat you to the planet by years with even disgustingly small gains in fractions of C.



Humans are certainly capable of surviving for centuries in isolated populations of a few hundred. We've done it many times before. For that matter, for much of history in Europe, almost everyone would live and die without ever leaving a small community. Just because you can't imagine living in a village without going crazy doesn't mean that it's beyond the limits of human endurance.

The kids will be fine. Kids are incredibly adapatable.

Whether or not the kids will be 'fine' is beside the point, particularly since 'fine' doesn't seem to be any more than a euphemism for "They can always get out and walk back to Earth if they don't like it." The point is whether or not this is ethical, or if you like, moral. We've got lot's of experience today with kids who don't want to continue to live their parent's choice of life style, whether it's a kibbutzim or a commune. Here on Earth at least, they do have the option to leave. A half light year out from Sol? Not a chance.


141 Trees?
Like the two-off specials by Donald Moffitt?

BIG ships?
Like hollowed-out asteroids?
IIRC there's a problem with that called "Wind from a burning woman" - unfortunately I've lost my copy of that one ....

153 onwards - Culture to CONTROL a limited environment with NO ESCAPE PERMITTED?
Try Tokugawa Japan?
Complete with the VERY harsh penalties for stepping out of line.....

167/8 But we DON'T value engineers. I could tell you a bitter story about that, but won't here.
As for othe continuity problems there is a discussion about the difference between the "Flyers" and the "Landers" in another one-off - where the McGuffin is that the Ship has a human brain, and doesn't want to land and be dismantled ("Mayflies" by K. O'Donnell)


Most kids here on Earth don't, in fact, have any meaningful choice. I agree that there's a moral question here, but I don't think it's at all clear-cut.

Besides, if unborn generations counted in any truly effective moral calculus, we would be living in a very different world. Some people, sometimes, plan ahead for their children or grandchildren, or very occasionally more distant descendants, and consider the effects of their decisions upon them. Mostly not so much.

Maybe generation ship pioneers will think that their children's world will be considerably better than what they themselves are leaving behind. For instance, untainted by some heresy. And quite likely they will imagine that they are giving their more distant descendants a very great gift.

I think we have strayed a long way from CS's original question, which I think probably wasn't referring to generation ships at all. Consider this variation: when we settle distant star systems with replicating nanoprobes, what genetic information do we need to send with them, for them to form a functioning biosphere into which they can decant freshly-grown humans? The nanoprobes will get there many centuries ahead of the generation ships anyway....


And my answer to that question is: concentrate on soil organisms: prokaryotes, nematodes, lichens, grasses, mosses, mycorrhizal fungi, mites, earthworms. If you can make soil, you're sorted.


Lance @170: it's a bit more difficult than that -- an organism's genome isn't all there is to being able to produce one; you've got a bunch of epigenetic stuff, methylation patterns, chromosome structure, siRNAs, and other tools you need to bring along. And even if you can manufacture a zygote, you need to be able to bring it to term -- good luck with that camel if your only other large ruminant is a goat or a kangaroo!

Taking along a complete genome library -- with epigenetic extras, and hopefully a frozen sample bank -- is a really good idea, but it's not the end of the story (except for single-celled species, which appear to be a damn sight easier to rebuild from source code than vertebrates). And it does suggest to me that any generation ship is going to need a whole bunch of cladistics experts working with this stuff. Hmm.


I think that what I've learnt from this thought experiment is that a lot of science fiction that deals with this subject is written with very little knowledge of the real science requirements to deal with the problem.

Not that that's a bad thing in and of itself, as most scifi is more about stimulating the imagination anyways.


Nick @ 156: The 2,000 figure I mentioned was intended to be for a totally isolated population - some of the more scattered Pacific islands, for example.

European village populations might often be only a few dozen people but have never been that isolated. A few people in each generation would always be moving from one village to the next (or even further) and breeding there. A village that was too small to support a technical specialist (potter, blacksmith, miller) by itself might simply try to make do without but might either share with a neighbouring village or depend on an itinerant one visiting every now and again. If one village, or even an entire region, hit a disaster, its neighbours would start moving in to the now under-occupied territory.

So perhaps a generation fleet, with several dozen ships of about 100 people each, might work as well or better than a single generation ship with the same total number of people - provided that there is a slow circulation of people between the ships, to keep up genetic diversity, maintain necessary long-term survival skills and give any kids who decide they can't face a whole lifetime in their native ship a limited opportunity to see if they can fit in better elsewhere.



Most kids here on Earth don't, in fact, have any meaningful choice. I agree that there's a moral question here, but I don't think it's at all clear-cut.

Argument by assertion and false equivalence? Care to back up the statement that most kids don't have any meaningful choice? And do you really think that growing up knowing that you are destined from birth to die grubbing for potatoes while the hereditary leadership class grows ever more detached from reality and ever more demanding of the lower orders is the same as your lack of 'meaningful' choice? Unless you can say with some truth that you'd be entirely happy if such were your lot(I'm a believer in the golden rule), I don't think you have much of a case.

Maybe generation ship pioneers will think that their children's world will be considerably better than what they themselves are leaving behind. For instance, untainted by some heresy. And quite likely they will imagine that they are giving their more distant descendants a very great gift.

And as I've already pointed out, if the children here on Earth disagree, they at least have the option to leave. On a generation ship, not so much. And as I have also pointed out, this is precisely equivalent to being legally obligated to paying off your spendthrift great-grandparents debts, by potato farming if you are told to do so by the duly constituted authorities. Somehow, I don't think you're down with that :-)

I think we have strayed a long way from CS's original question, which I think probably wasn't referring to generation ships at all. Consider this variation: when we settle distant star systems with replicating nanoprobes, what genetic information do we need to send with them, for them to form a functioning biosphere into which they can decant freshly-grown humans? The nanoprobes will get there many centuries ahead of the generation ships anyway....

I think we've threshed out the conclusion that the original question did in fact have the obvious answer: no one has the slightest idea. But that it's good to err on the safe side :-) My own personal caveat was that suspended animation was a more likely alternative to life support for interstellar travel than an O'Neil style generation ship. There's a reason why physics became the first discipline to be thoroughly grounded in mathematics.



I think that what I've learnt from this thought experiment is that a lot of science fiction that deals with this subject is written with very little knowledge of the real science requirements to deal with the problem.

Yeah, pretty much. There hasn't been much, if any progress on that front since the fifties as far as generation ship stories are concerned. Part of the problem I think is that there really isn't much call for that sort of research into minimal sizes other than for something like space colonies or long space voyages. I suspect that this information is going to be very hard won.


Robin @176: your conclusion is essentially correct.

SoV @178: yup, nobody knows yet. And I tend to agree with you that suspended animation with crew rotation en route would work out a lot better than a generation ship (which adds cultural engineering on top of environmental engineering and spaceflight).

On the other hand? Suspended animation: it ain't easy. We have just got the inkling of a clue about reducing human metabolic rates and inducing hibernation using hydrogen sulfide (but: it's not been used in humans yet, and when it is, the first candidate application won't be spaceflight -- it'll be stabilizing critically ill patients while they get to an operating theatre). It probably won't be good (from a medical angle) to stay under for more than weeks to months at a time, and the stars are a long way away. Cryonics ... useless, without some breakthroughs in how we prep the patients beforehand (and, ahem, figure out how to bring them back). Right now it's a Hail Mary pass for folks who really don't want to rot when they die, not a viable strategy for interstellar exploration.

I think Ken MacLeod nailed it in "Learning the World": what we really need is something that looks like an autonomous space habitat, crewed by folks who are functionally immortal. Alas, barring some unprecedented breakthroughs (see also: singularities) we're about as far away from building self-propelled city-states crewed by tens of thousands of immortals as we are from the Holy Roman Empire.


Care to back up the statement that most kids don't have any meaningful choice?
Well, let's start with the fact that most kids grow up in absolute poverty, under a dollar a day, in countries where there is no safety net, either in small village communities or urban slums, under the thumb of custom, caring for their siblings and then forced by circumstances into the family trade. If you think most kids have the choice to run away to a kibbutz or commune, it's down to you to explain exactly how most kids are going to manage that.


200 years? I'd guess longer.


Charlie@180: Well, yes, saying that suspended animation is more likely than a generation ship with it's rather cumbersome life support doesn't mean that it's all that likely[1]. Me, I'm a believer in long-range observation and robotic probes.

Of course, those might not be the sorts of stories you want to tell :-)

[1]Otoh, I can see lots of money being thrown at cryonics and suspended animation by people who have absolutely no interest at all in space flight. I think there's a story or three about superannuated baby boomers throwing their considerable support behind this sort of research for all the usual selfish reasons. In fact, without my tongue in cheek, I'd predict that by 2040 in real life there will be considerable research into the subject for just this sort of reason. I can't say the same thing for sealed, self-sustaining biomes capable of supporting hundreds or thousands of people.



Well, let's start with the fact that most kids grow up in absolute poverty, under a dollar a day, in countries where there is no safety net, either in small village communities or urban slums, under the thumb of custom, caring for their siblings and then forced by circumstances into the family trade. If you think most kids have the choice to run away to a kibbutz or commune, it's down to you to explain exactly how most kids are going to manage that.

Game, set, match. I win. Nick, it's considered poor form to talk about the people who wouldn't be going anyway, particularly since I think it's pretty implicit that the people who would be are those with skillz, i.e., middle-to-upper class types. Now, care to actually argue the point this time? Instead of talking about people thousands of miles or thousands of years away?

And it's not kids running away to a kibbutz, it's kids running away from kibbutzim. Or communes. Finally, not to put to fine a point on it, but since you're the one making the assertion, it's up to you to defend it(something that's come up here recently.) If you don't know this, perhaps you should stop posting until you do the research and find out that yes, this is the case. As it stands now, I can only conclude that you know I'm dead right, but you just can't bring yourself to say so. An all too common affliction, alas.


Hmm, SoV, Nick appears to be closer to the money than you are on the rather crucial definition of 'most'.


Uh-huh. Right. So of course, you think that stuff like minimum wage, health care, public education, and so on and so forth are luxurious frills, because after all, most people throughout history have never had them? And that an average life expectancy of 50 is just dandy, because most people have had a life expectancy of 40.

So, no Chris, he's not. Not even close. Not unless you want to argue that these are the people that will actually be captaining and crewing this generation ship. In fact, arguing in this fashion is generally considered dishonest, and indicative of the fact that one doesn't have a leg to stand on.


So which will come first and what are our likely timescales?

1. Digitised humans in virtual realities inside tiny spaceships as in Accelerando

2. Life extension into the hundreds of years needed for a long space flight

3. Suspended animation

4. Arkships


Robin@187: 5. Not enough information?
6. None of the above?

I'd ask about not only the plausibility, but what sort of economic reasons not related to space flight would you have for wanting to do any of the above. Offhand, the teeny tiny digital people seems like a sort of generic spinoff from medical/computational advances, and suspended animation seems likely for people wealthy enough to not want to have their ticket punched just yet(double ditto for extended life.) Your 4. Arkships seems the least plausible on that scale, because I can't think of any other reason to want them.


SoV, 'is' does not imply 'should be'. On that point, I'm invoking Brooks' Law and leaving this particular discussion.


Chris@189: And you are wise to do so, since your case simply doesn't have any merit. Why is it so hard for a certain type of person to simply admit that they were wrong? Personally, I'd rather admit that I was wrong, as opposed to still being wrong and having other people think I was small and/or dishonest to boot.


@175: Charlie, I was thinking about the problems of storing eggs/genetic information, etc. recently. Let's not forget that organelles such as mitochondria and plastids have their own genomes.

That said, I think there's a way to store genetic information in a fairly compact way.

What you need:
1. "the genome" for a species (anyone who realizes how little diversity is actually captured in the 1-5 organisms sequenced should read until that phrase brings up a smile). Anyway, we need a standard template.
2. Methylation coding library. I suspect that this will need to be annotated as temporary vs. permanent, and some sort of importance scale. For example, in polyploid genomes, often there's only a single active copy of a gene, and multiple silent copies. Some silencing is more important than others.
3. A library of important variations. There's a lot of variation in your average genome, because in many places, the folding is more important than the coding, and a number of different sequences will fold in the same way. Such variation needs to be coded as unimportant. Conversely, variation in important regions (such major histocompatibility complexes) needs to be highlighted.
4. For crops and domesticates, we also want a construction template that describes which specific sequences provide the characteristic and useful traits of the varieties we want. We don't need to know what they have in common with other potatoes, since that is covered in the standard potato genome.

With information like this, we could have, say, a "potato library" with one standard genome and all other libraries, and make not only every potato variety that has been sequenced, but also to experiment to make varieties that can, say, grow in low gravity.

Individuals can be coded the same way. For instance, to clone you, I don't necessarily need to know your genome. However, I do need to know how you differ from my template human genome. I think it's probably simpler to describe how you differ from average than to describe you in total.

As for bacteria...there was an experiment ~15-20 years ago (reference escapes me, but I think it was in Nature) where a scientist did an important ecological experiment. He took his used toilet paper and sequenced the bacteria that he was excreting. What he found was that the community of organisms in his intestines changed on a daily to weekly time period, and there was a strong signal for bacterial evolution going on in his intestines.

Note that rumens in animals like cows and goats have much more complex microbial communities than human intestines, and we don't really know all the organisms that live there, or how they inter-relate.

In any case, my vote would be for having a big library of bacteria and other microbes, on the assumption that they can assemble communities and adapt to local conditions about 100x faster than we can describe said communities, at least with current technology. Provide a lot of diversity and let them solve the problem themselves.


SoV, you asserted (171) that kids on earth have a choice; I replied (173) that you were wrong. You accused me (178) of arguing by assertion and of false equivalence. I provided (181) the trivial argument to back up my statement. You claimed (184) that this evidence was irrelevant, that these weren't the kids we were looking for, and that you therefore somehow "win". Since you're not capable of sticking to the point, I'm not going to bother.


Chris, SoV, Nick: you've wandered off-topic in an uncivil direction that's likely to attract cranks and trolls if it continues, and I'd rather not see any more discussion along these lines (at least not on this thread), KTHX?

Robin @187 asks an interesting question, though.

#1: "Digitised humans in virtual realities inside tiny spaceships as in Accelerando" ... this is very hard to distinguish from "robot AI probes". (Okay, the method you use to get your AI is non-standard, but functionality-wise, it's a tie.)

#2: Life extension into the hundreds of years. Actually, I'm kind of partial to this (because, ahem, I want it for myself :) so I'm an unreliable commentator. Life extension per se doesn't seem impossible but it's not a problem that's obviously amenable to the magic bullet treatment. We'd also need other supplementary medical breakthroughs: repair of genetic damage from high-energy cosmic radiation, fixes for cancer, and so on. If we get the bundle, however, interstellar missions at much-slower-than-light speed looks a lot less daunting. (Hell, even getting bootprints on Pluto right now is a bit of a long slog ...)

#3: Suspended animation: I'm going to come down on the "no" side for this one. You still need a crew (rotating in shifts, probably) to keep things running, you still need to manage your biosphere, and it's not something that looks easy to manage. Also, without life prolongation, it isn't going to crack the problem. (Do bears or squirrels or other hibernating mammals live longer if allowed to hibernate than if they're kept in a warm environment year-round? Anyone know? My guess would be a qualified "no" but I'm opining in the absence of evidence ...)

#4: Arkships: you need this if you're going to send canned primates outside near-earth orbit, because there's no friendly biosphere anywhere else. (Exception: if #1 comes along first, we can make do without.)



Me: "reasonably safe and clement generation ship"

SoV: "So then you're entirely comfortable with the idea of having to clean toilets for the rest of your life - by law"

I think we have different ideas about "reasonably safe and clement."

A generation ship that will subject the kids to horrific conditions such as endless toilet cleaning is in fact a cruel thing to do to said children. (Assuming there are better options elsewhere, which is a good bet if you can afford to make a generation ship).

A generation ship that will subject the children to similar conditions to what they would have back in the Solar System (and I find the idea of generation ships built by folks who don't have significant space habitats already quite unlikely) suffers morally mainly from the lack of free emigration. I think the proper approach is to cast the question as "Is it immoral to bear children into this 'world'?" Let the *parents decide* if the environment is ok to bear children into. *This is exactly what we do today on our current 'spaceship'.*

And yes, there are moral issues with letting the parents decide, no question. I haven't seen an alternate approach I think is better.

Addendum: I am presuming the ship in question will be sufficiently clement that the question of "kids or no kids" is not a matter of survival into old age for said parents. The "first world - in space!"

Aside: I am also getting a serious vibe that the mental model most people have for genships is "third world - in space!" Or possibly "soviet gulag - in space!" I agree that these are bad plans. If you can't build a nice place to live in space, you shouldn't build a crappy one and launch it at distant stars.

Aside to Aside: Of course, plenty of crappy plans have been executed in the past. I can certainly see a quarter-assed generation ship made for the express purpose of politically exiling someone who couldn't be killed or imprisoned safely - ex. Napoleon. Not enough delta V to return, not enough stability to last to destination, but enough to last for the politically significant lifetime of the passengers. Worse, said passengers are all partisan volunteers, with little in the way of applicable skills. Sounds like a decent background for a SF story to me.

As for paying for my ancestor's debts, we currently collectivize this into a "National Debt" that I am in fact in some sense responsible for.

Returning to the general vicinity of the original subject: The minimum spaceship size for "first world - in space!" is probably pretty damned big for canned monkeys. I suspect that sociology is the limiter rather than biology, however. You're going to need a *lot* of people, unless axiom changing technology is involved. Of course, a plausible genship more or less *requires* axiom changing technology, so I might as well posit black is white and get offed at the next zebra crossing.

As is common in far future SF, we are flailing around in the dark, making wild predictions about abused elephants. Hopefully the elephant is patient and the flailers amused.


CS@193: our comments crossed in the post. I've stopped now.

I wonder how hard it is to construct a biosphere given a boatload of bits and some good nanotech replicators. Is it easier or harder than building a generation ship? It seems harder, because it relies on *two* things we don't know how to do (nanotechnology and synthetic life) rather than just one (propelling a generation ship). But if the bootstrap-biosphere-from-nanoprobe magic technologies are possible, we're likely to develop them on earth anyway, as side-benefits of other projects.


Not sure that leaving LEO requires an arkship. Say the ecosystem on a Mars-earth cycler - or in the Mars system itself, or in the outer Jovian moons - is good for 10 years. You can model it before you try it, and if it gets out of kilter ou can always reboot it with new genes - perhaps even new flora and fauna sent from your test articles on earth. No such reboot is possible for a generation ship.


Martin@194: The "partisans in dire exil" (not a spaceship, but a bareley habitable moon) trope is the base for LeGuins The Dispossessed, isn't. Thus it could become a realistisc utopia, with the right kind of partisans ...


@193: As for hibernation extending lifespans, a quick google reveals that hibernating black bears live 25-30 years in the wild (we're talking about animals that weigh 90-125 pounds without hibernation fat. Non-hbernating wolves get up to about 80 pounds and live about 8 to 10 years. I'm including the weight scaling because bigger animals tend to live longer. Wolves and bears are somewhat close evolutionarily, so the first cut is that hibernation correlates with longer lifespan. The real problem is that the *cool* hibernation tricks are performed by things that are much smaller (turtles, frogs, squirrels), which tend to not live very long anyway due to predation. Sorting out whether longer lifespan is due to decreased exposure to predation is tricky. Another comparison is that non-hibernating pikas live about three years, and hibernating golden-mantled ground squirrels live about four years. They're both high mountain species. Unfortunately, they're both prey animals, but again, it looks like the hibernators live a little longer.

Humans are supposedly non-hibernating. The supposedly part comes from the Independence 1 culture I mentioned above (first known settlers of the high Arctic), which, if you believe the archeologists, got through high Arctic winters on a few cubic meters of willow twigs for fuel while living in tents of some sort. It looks as if they were close to hibernating, only eating maybe once per week, and staying warm by huddling under musk-ox hides. Scary existence, but I do wonder what would be possible, particularly if you could use sedatives to make staying in a sleeping bag for a week at a time bearable. That might cut down drastically on the resources needed for a voyage.


A lot is known about partially isolated ecosystems from studies of biodiversity on small islands. Basically, remote small islands have very low biodiversity; if you take out mobile species (like birds and plants that spread via the sea) they are often very simple ecosystems. Lack of biodiversity extends into the surrounding seas; for example there are few species of seaweed compared to the shores of bigger land masses.

Obviously, they are not closed systems - bulk and trace elements interchange with their surrounds to greater or lesser extents.

The most important conclusion: extinction is more likely for small isolated populations.

Active management of species may avoid this for macro-fauna and flora, but it would be difficult or even impossible to manage the micro-fauna effectively; useful bacteria and fungi will die out.

For a medium range ship (say hundreds of years), keeping mixed reference cultures in the freezer might work, but long range travel sounds very hard.


@ chris, 157
"This MO means that you can establish intermediate goals. Sure, you won't be around for Arrival Day, but the default plan has an option for someone to unfreeze the orcas on or around year 450 - YOU could be that person! Orca-unfreezing is cool. Orca-unfreezers are written about in gossip columsn, eat steak, get to choose boyfriends, etc."

I would love to see the propaganda posters valorizing the orca unfreezers!!!
made me laugh.


Charlie@193: My apologies. I tend to get frustrated at the style of argument that amounts to the accusation that you didn't say your words right("You're wrong to say that dogs are four-legged animals, because, hey!, here's a three-legged one.")

That being said, I don't think that any of the four scenarios are especially likely (for various flavors of AI in the first case.) Although, what does 'long life' mean in this case? One way to Alpha-C would take approximately 90 years at 0.05 c. This kind of life-extension might not be in the not too far future.

500 years at .01 c, maybe not so much. Even for the die-hard enthusiasts. Of course, the ObSF would be "Cities in Flight" where space travelers still have to live hundreds of years if they want to get to anyplace interesting. I just reread this one last summer, and it still holds up well.



I think we have different ideas about "reasonably safe and clement."

A generation ship that will subject the kids to horrific conditions such as endless toilet cleaning is in fact a cruel thing to do to said children. (Assuming there are better options elsewhere, which is a good bet if you can afford to make a generation ship).

Unfortunately, there will always be hewers of wood and drawers of water . . . even on a generation ship. The problem here is that it's all too easy to imagine a situation where this becomes hereditary. 'Born in hydro, die in hydro' they say in the aft decks, F and back. Or do you think it's more plausible that it's just as likely that it will be the Captain's son who will end up grubbing for spuds?

A generation ship that will subject the children to similar conditions to what they would have back in the Solar System (and I find the idea of generation ships built by folks who don't have significant space habitats already quite unlikely) suffers morally mainly from the lack of free emigration. I think the proper approach is to cast the question as "Is it immoral to bear children into this 'world'?" Let the *parents decide* if the environment is ok to bear children into. *This is exactly what we do today on our current 'spaceship'.*

Since a generation ship requires, you know, generations, you're not making any sort of a real argument at all. And unless you can give me some sort of assurance that you would have no problem at all personally doing the garbage detail because that's what your mom and dad did, and their parents before them, well, you haven't really addressed the argument.

And yes, there are moral issues with letting the parents decide, no question. I haven't seen an alternate approach I think is better.

This is just throwing up your hands and saying that even if it is immoral and unethical, well, what can you do? I don't buy that here on Earth as any sort of valid argument, so I don't think I will buy it in space.

Addendum: I am presuming the ship in question will be sufficiently clement that the question of "kids or no kids" is not a matter of survival into old age for said parents. The "first world - in space!"

Aside: I am also getting a serious vibe that the mental model most people have for genships is "third world - in space!" Or possibly "soviet gulag - in space!" I agree that these are bad plans. If you can't build a nice place to live in space, you shouldn't build a crappy one and launch it at distant stars.

Not at all. At least, not me. I'm sure that for the more well-off types, section heads, commanders, and whatnot, their kids will think everything is just fine. The problem is in assigning the unpleasant work that nobody wants to do, and if you think that's going to be completely eliminated as any sort of future concern, well, I've got a bridge I'd like to sell you. Here in the Western world at the beginning of the 21st century at least, we like to think that even a garbage man's daughter can rise in the world if that's what she wants, even become fabulously wealthy or hold high office. On a generation ship? I just don't see it.


heteromeles@198: Waaaay OT here, but what else is new... ectotherms (refuse to use the term cold-blooded) in captivity clock up some fairly impressive life-spans, even little ones. There are also geckos on New Zealand's offshore island that are definitely known to be over 30 years old. That's pretty well off the chart on the body size/longevity front. People talk about tuatara reaching 100, but I've yet to see conclusive evidence. Oldest known vertebrate is, I think, a giant tortoise, although there are some big numbers waved around for deep-sea fish (that's cheating though, they practically live in a fridge).

Humans, incidentally, are already an outlier on mammalian survival: we live longer than other 80kg-mammals.



I am also done discussing things with you. HAND.


@203: Human survival "in the wild" actually seems to be equivalent to chimps. In fairly primitive conditions, without medical care, we tend to croak around 40 or 50. That's why elders were so valued. They made it past that point.

Says a lot for good medical care that humans in captivity (did I say that, I meant towns) live a lot longer than do those in the wild, just as many animals do. In other words, read those things about humans living an abnormally long time with a small boulder of NaCl, and thank providence for doctors and public health workers. I suspect that, with lack of predation and decent living conditions, most species can live far longer than they normally do.

That said, there are lots of critters that screw up the age/weight scale: tuataras, bats, etc. Even small dogs in certain breeds routinely live longer than wolves, at least in captivity.

That's the problem with trying to figure out whether hibernation is a useful strategy for extending life. Getting a good matched set of hibernators vs. non-hibernators is an ugly exercise. What I posted previously was five minutes with google for a rough first cut. I'm glad Charlie asked the question though, because I didn't even think about it, and I do know better.


martin@204:Shrug. Stamp your feet all you like. It doesn't change the fact that you really don't have an argument addressing the ethical diceyness such a generation ship would entail. Or that after acknowledging 'concerns', you just don't care.

A depressing thought. As I get older, I find myself more worried about doing the right thing, doing right by people, being a good person. Unfortunately, this tends to conflict with some early sf wish-fulfillment. Take robots, for example. Wouldn't it be nice to have a robotic Jeeves to smooth one's way through life? Or at least a Daneel Olivaw? A fourteen-year-old me would have jumped at the thought, to say nothing of owning a stable of fembots. But I've come around to the idea that owning a device capable of doing all that Daneel could would be no different than owning a slave, and would be objectionable for precisely the same reasons.

So, whatever the technical feasibility of robot man- and lady-servants, ethically, they're a no go. Unless of course, one pays for their services at the going market rate . . .


@ 202
"A Gift From Earth"


SoV@202 John Scalzi in his story in "Metatropolis" proposes an interesting social solution to the problem of "someone needs to clean toilets". Here a more or less closed* city-state uses very strict aptitude tests with enforces job-placement following the score of this tests - even the son of a very influential politician who fluked his tests has to learn this in the story.

* Of course, one big difference between this semi-autonomous high-tech city-state and a generation ship is the fact that the city state can set people who do not take part in the job assessment "out of doors" into the wilderness - that would be rather extreme in a spaceship.



"The real problem is that the *cool* hibernation tricks are performed by things that are much smaller (turtles, frogs, squirrels), which tend to not live very long anyway due to predation"

Tortoises (that's land-turtles for you colonial types) actually tend to live a very long time. Those species that hibernate (mostly the Mediterranean ones like T. Graecas) have been recorded with lifespans over the 100 year mark - Which is why ours are in the will! - In nature, once they get past the size where they are at risk of being picked up and dropped by Eagles onto Aeschylus head (or a rock, whichever is convenient) they are pretty well defended from natural predators.

The generally held view in Herpetological circles (according to what my wife has read) is that regular over-wintering tortoises does indeed reduce their lifespan.



John Scalzi in his story in "Metatropolis" proposes an interesting social solution to the problem of "someone needs to clean toilets". Here a more or less closed* city-state uses very strict aptitude tests with enforces job-placement following the score of this tests - even the son of a very influential politician who fluked his tests has to learn this in the story.

That's not an ideal solution, but it's at least acknowledging the problem. Unfortunately, despite all the high-flying rhetoric at launch, I suspect that any such scheme will devolve very quickly into 'showing' that those sons of influential political leaders just have a natural 'aptitude' for leadership, while the gardeners kids will have a natural aptitude for gardening.

In short, a project like this would among other things deliberately create a caste of untouchables. Some people are apparently all right with this, or enough all right that they can say it's a problem, but hey, what are you going to do, can't make an omelette without breaking eggs and all that. And that these perfunctory acknowledgments are all you need to do to deal with the issue. I disagree. Emphatically.


Generation ship shopping list:

POWER: 1x fusion reactor (actually, better make that two)
PROPULSION: 1x lightsail (again...)
ISRU, on-site repair capability: 1x universal RepRap (better take a spare)
IT, CONTROL, NAVIGATION, SCIENCE MISSION SUPPORT: 1x giant data centre full of computers
ECLOSS: 1x murky water aquaculture ecosystem in a can
CREW, ORGANISATION: 1x Rawlsian utopia

Shit, we were doing so well until halfway down!


SoV, I don't really see much of a difference there. Whether you make people clean toilets because of some arbitrary inborn skill or lack of skill or because of their ancestry doesn't seem to make things more fair in any direction.

And keep in mind that cleaning toilets, and much dirtier work than that, is not in itself deeply problematic. It's the social status that hurts. I had a garbage man job in college, and the smell is something you get used to very quickly. But the difference between being a garbage man with the social status of a student and garbage man with the social status of a garbage man is almost tangible.

I can imagine social status issues on a starship, but not necessarily any stronger than in the normal world.


@Low caste cleaners: this is a western concept, where the poor do the dirty work.

Other societies use different standards for wealth and status, such as how productive their gardens are. This isn't alien to our culture (see state fairs, or people bragging about their backyard gardens). You can also use the Andean trick of segmenting any repetitive task (such as cleaning the toilets), allocating the bits to work teams (each gets X toilets to clean), and making it a competition to see who finishes first. People who don't do the work get socially penalized.

Mind you, these systems still end up generating social hierarchies, but they do have the virtue that jobs are not assigned on the basis of social status.


I agree with h@213 and would like to add that shouldn't we consider it a failure of imagination to have, OTOH, some sort of society with the capacity to develop and maintain a fully functioning biosphere, but but on the other hand can't eliminate, at least in concept, 'dirty work'?

I'd go so far as to put it on the list of design requirements.


The toilets need cleaning wherever you live. And children who don't like it on earth do not presently have the option of leaving.

Ken Macleod laid stress on the complexity of the evolved constitution of his generation ship. Writing that is a non-trivial task. It might have spinoff benefits for earth though.


What pressure are we keeping this biosphere at?



I agree with h@213 and would like to add that shouldn't we consider it a failure of imagination to have, OTOH, some sort of society with the capacity to develop and maintain a fully functioning biosphere, but but on the other hand can't eliminate, at least in concept, 'dirty work'?

I'd go so far as to put it on the list of design requirements.

Unfortunately, the idea that this is just some sort of Western idea(and thus, presumably a local cultural tic that can be easily modified or eliminated) is just plain wrong. From the wiki:

In the context of traditional Hindu society, Dalit status has often been historically associated with occupations regarded as ritually impure, such as any involving butchering, removal of rubbish, removal of waste and leatherwork. Dalits work as manual labourers, cleaning latrines and sewers, and clearing away rubbish.[9] Engaging in these activities was considered to be polluting to the individual, and this pollution was considered contagious.

I suspect that if you have robotics good enough to completely automate all the dirty work, then you probably have automation good enough that you don't need to bring the meatbags along. As Heinlein noted in one of his novels (manual) dishwashing, for example, is a complex task that requires a fair degree of intelligence, which is precisely what the dirty jobs are going to need.

And to forestall anyone trying to pick a fight by trying out a "you didn't say your words right" type of argument, yes, I'm aware that there are 'automatic' dishwashers that do the same job in a different way for various values of 'same'. But that doesn't mean that all such jobs can be automated.


@216: Good question, Kelson. On the one hand, keeping it low makes things light. On the other, mammals cannot reproduce above ballpark 50% sea-level pressure (*very roughly* I'm back calculating from where the highest babies are born in the Andes and Himalayas). The problem for mammals is getting oxygen across the placenta, even in organisms evolved to low O2. And cooking at low gas pressure gets tedious too. When you need a pressure cooker to make a decent cup of tea or cook a potato, it's annoying.

So, as low as your crops can stand. Fortunately, the Andeans and Himalayans have left us a good playbook of how crops and critters deal with different altitudes, so that won't derail our biosphere.

@217: I'm not going to quibble about India. In my book, Aryan is western, but it's certainly not worth fighting over, because there were/are untouchable castes the world over. The basic point is that there are other ways to structure society so that social status/rank is based on the amount and quality of work you do, rather than how degrading the work is. Alternatively, we can go with Mike Rowe's observation on Dirty Jobs which is that the people doing the crap jobs often seem to be happier and better adjusted than the suits. It may turn out that being the Only Guy on the Spaceship who's willing to fix the sewage pipes will save you from the Cull when things get bad....


Why should the "day job" that your putative spaceship crew do be affecting that much of their lives? We've already established that we'll want lots of people for a sane society and a decent gene pool. If any of them are working 60 hour weeks, something is wrong!


Chris L: good point. American working hours are a bizarre cultural aberration -- a very local one. More to the point, designing a space habitat/generation ship with the implicit parameter that the crew are expected to work 40-60 hours a week is a really bad idea; efficiency is the enemy of redundancy, and multiple redundancy is absolutely vital to any such project (because it provides resiliency that is essential to have any hope of recovering from a disaster).

A sensibly designed long-duration hab would require the crew to do just enough work to maintain the necessary skill set (you don't want them to go rusty), but leave lots of time available for education, recreation, and socialisation. You can't build a stable hab culture on material acquisition because it has to function in a resource-bounded environment (although soft goods/intellectual property is another matter, if you want to provide an escape valve for acquisitive urges, or a "training wheels" environment for the market-mediated culture that you might need to revive after arriving in another solar system).

Consumer capitalism along our current model simply won't work as a way of running a long-duration generation ship (the failure modes are lethal and non-recoverable). Communism (or rather, Leninism) has a slightly better prospect, but is still a long way from optimal. Monarchism is just a pretty word for "hereditary dictatorship supported by military caste". What are the alternatives?


Consumer capitalism along our current model simply won't work as a way of running a long-duration generation ship (the failure modes are lethal and non-recoverable). Communism (or rather, Leninism) has a slightly better prospect, but is still a long way from optimal. Monarchism is just a pretty word for "hereditary dictatorship supported by military caste". What are the alternatives?

Those forms are all very much nation-state sized concepts, for groups of people numbering 100,000s at least, and millions really. I can imagine large colony ships, but something that large would not really be a purpose-built ship anymore, more a society accidentally on the move.

At smaller sizes, thousands or tens of thousands, politics works different somehow. I guess we don't have much experience anymore with independent, self-sufficient units that size, but odds are it would resemble village townhall politics a lot. All the 'important' people know each other rather well, and that alone makes decision making less formal.

Economics works different too, even for intellectual goods. In a large society, it makes sense to spend much effort on very accomplished, polished versions that are not aimed at particular persons, that are then cheaply multiplied. Like IKEA furniture design, or I guess Mr. Stross's books.

In a small community, the optimum is more towards less perfected designs that are aimed to a particular need and situation. Like hand-made furniture and oral stories.


Charlie@220: What are the alternatives?

Some form of "proper" anarchy, obviously. There are powerful incentives to co-operate already built into the system, and you could enhance those with the right initial design of infrastructure. (I'm still thinking about specifics, but say a serious attempt to eliminate single points of control combined with the least silly bits of David Brin's "transparent society".) And you're already positing limited limited resources and less emphasis on efficiency.

Failure modes would be roughly the same as democracy - probably more chance of factionalism, but less of transforming into a dictatorship. You could also end up with a caste system if most people only do one kind of work - better plan the work-allocation process carefully...


Sam: this echoes some ideas I've been (inconclusively) batting around with Karl Schroeder -- how do you design a society for the really long term? Democracy is prone to mutation into some other form (kakistocracy, oligarchy, populist dictatorship). Monarchy has a single point of failure and historically only worked when there was a draconian enforcement regime backed up by Malthusian pressure (if the lid came off -- e.g. with the opening of a new frontier for emigration -- the oppressed vote with their feet). We are somewhat intrigued by the idea of a society with multiple designed-in local attractors, so that over time it can oscillate between different modes of governance (but returning eventually to previous patterns); also by the idea of designed-in escape valves.

The social pressure on a generation ship is going to be fierce; but if there's a designed-in expectation that, say, 20-50% of the inhabitants at any given time will be preoccupied by non-functional distractions such as the arts or sports, that might go some way to defusing social stresses. Arts and sports can act as vectors for social competition and status-seeking, while being channeled easily in directions that don't consume excessive physical resources.

The protestant work ethic underlying American-style capitalism, with its added dog-eat-dog ethos, would be a recipe for disaster aboard a generation ship -- regardless of whether it's run as a democracy or a dictatorship. (And don't get me started on libertarianism. Economic libertarianism -- in the contemporary American sense -- aboard a generation ship would be just plain suicidal.)

Hmm. I think this particular sub-thread deserves it's own posting. But I'm about to drive 350 miles to an SF convention in the next few hours, so it'll have to wait until tomorrow.


Charlie: As far as I'm concerned, libertarianism is just another kind of oligarchy - but that's classic flame bait, so I won't elaborate.

Monarchies are odd... if you take a wide view then most of them look like warts on the side of various kinds of oligarchy. Which raises the point that social order is not the same thing as government. That's potentially two attractors for your social oscillator right there.

Another slightly obvious thought: if you're looking at meta-systems, then the failure modes for any society depend heavily on what your status indicators are and what kinds of conflict-resolution are available. Your initial physical conditions will have a big effect on those even if it's hard to track the long-term consequences, which is why I'm thinking choosing your infrastructure is likely to be more useful than trying to engineer-in an obsession with sports.

And a less-obvious one: the kind of built-in redundancy you're talking about sounds like a classic utopia of abundance: lots of resources, just enough work to keep you engaged, plenty of leisure-time activities. So at the very least we have an answer to why anyone would volunteer for a generation-ship.


Stiglitz has an interesting take on such issues: What we are actually deciding is how to decide projects that get funded (encouraged), and which are not, also by whom. Projects can be anything that is relevant to the survival of the human race.

Now: if the failure of a project means extinction of human race, those decisions are better-off centralized. If however, we need to "try" bunch of projects, don't care of failures, but success will push everyone forward in some way, there we need decentralized projects, that is, people start them on their own, take the risks, and the benefits as well. The latter would be capitalism (or any other free enterprise based system), fhe former central, government. The paper is here http://bit.ly/3uUhkf


Hey guys, first time poster, just wanted to float an idea:

I know we are discovering new gut flora every day, but has anyone tried to establish a bare minimum? Could we kill off our internal friends with high doses of antibiotics, and introduce say 15 or 20 of our favorite ones (maybe picked for endurance and usefulness) and then give ourselves a vitamin shot once a week so make up for the fact that we no longer absorb niacin?

Side note: do you think intestinal flora differs greatly by geography? Would there be like a American Standard Gut Flora pill?

Glad I found this place.