What would it take to build a self-sustaining astronaut ecosystem on Mars?

In 1829, Nathaniel Bagshaw Ward, a doctor living near Wellclose Square in London, dropped a few seeds of fern and grass into a bottle partially filled with soil. Soon, he witnessed tiny blades of grass and one little fern sprouting from the soil, despite the bottle having been sealed. It turned out that plants, cycling through whatever water, minerals, nutrients, and atmosphere they had in their bottle, could live and grow almost completely isolated from the outside world, using sunlight as their only energy source. 

 

Today, after over six decades of researching bioregenerative life support systems, we’re edging closer to pulling the same trick off in habitats designed to support astronauts on alien worlds.

BIOS and CELSS

Bioregenerative life support systems are the product of two visionaries. In 1926, Vladimir Ivanovich Vernadsky, the founder and first president of the Ukrainian Academy of Sciences, elaborated the concept of the biosphere, a closed material cycle on a planetary scale that could be indefinitely sustained in part through life itself. In 1929, Konstantin Eduardovich Tsiolkovsky, a Russian rocket scientist, applied Vernadsky’s concept to space travel and proposed using small-scale closed ecosystems to support life on spaceships. The job of those ecosystems would be to produce air and food while recycling waste. 

 

The space race era saw the concepts developed by Tsiolkovsky and Vernadsky implemented in Moscow at the BIOS-1 facility. It consisted of two connected compartments: a microalgae cultivator and a living space for one crew member. The microalgae would take the CO2 exhaled by the human and release its oxygen via photosynthesis. It worked, more or less, but there were regular imbalances caused by the differences in metabolism between the human and algae. It was eventually fixed by changing the human’s diet, but everybody involved recognized that you can’t change a diet midflight on a spaceship. Any miscalculation would probably kill the astronauts on a real mission. 

 

In 1969, BIOS-1 evolved into BIOS-2, where a greenhouse with vegetables was added as a third compartment, soon followed others: a compartment for wheat and another with a microbial cultivator for oxidizing solid human waste. Experiments with humans were gradually lengthened from 12 hours to 24 hours, then to two weeks, and finally to 90 days spent in the facility. At the same time, another team built the BIOS-3 facility, which relied on the same tech but could be controlled from the inside by the crew and was arranged to resemble cabins of Soviet spaceships from that era. 

 

BIOS-3 experiments showed how much labor it took to operate this system. Results were bleak. Astronauts basically worked like full-time farmers just to keep it going. 

 

On the other side of the Iron Curtain, NASA was doing small-scale experiments in which algae colonies were used to produce atmosphere and food for mice. Those experiments developed into more advanced demonstrators built under the CELSS research program in the 1970s, which combined multiple organisms and higher plants to combat the oxygen balance issue already encountered by the Soviets. 

 

Both BIOS facilities and CELSS demonstrators relied on a similar architecture and underlying concepts. So it’s no surprise that they suffered from a similar limit: There was very little control over what exactly the biological component was doing. 

 

The plant chamber acted as a black box. American and Soviet engineers knew that a given input would yield an essentially predictable output and scaled the system based on the number of crew members. Processes already operating in nature were copied and pasted into confined, isolated spaces. It was like solving flight by imitating birds. 

 

Things didn’t change until 1987, when Claude Chipaux, a space engineer working for a company that later became Airbus, proposed building an entirely new bioregenerative life support system called MELiSSA.

MELiSSA is born

MELiSSA stands for Micro-Ecological Life Support System Alternative. Unlike its BIOS and CELSS predecessors, MELiSSA relies primarily on bacteria strains, not plants. The biological component was no longer a black box. Bacteria were to be cultivated in thoroughly engineered bioreactors under tightly controlled conditions. The functioning of those bioreactors was based on elaborate mathematical models and controlled by algorithms. 

 

The entire system was also dynamic. The crew could control its operation through a digital interface and assign priorities to different functions at will. Want to stabilize the atmosphere or change its composition? Just push a button. Want to make more food? Push another button. But this automation had its price. MELiSSA didn’t need years’ worth of funding. It needed decades. 

 

The European Space Agency greenlit it anyway. The project quickly grew into a gargantuan effort backed by 14 countries and over 50 institutes, universities, and companies.

 

“Several science groups started to gather around this concept providing the required know-how in biology and engineering. We all agreed that we needed to rely on a fundamental knowledge of how our organisms function, how their metabolism work, how they uptake nutrients, how they react to different lighting conditions or changes in the atmosphere composition, etc,” said Francesc Gòdia Casablancas, head of the MELiSSA Pilot Plant, a functional demonstrator of the MELiSSA located at the Autonomous University of Barcelona, Spain. “The goal was developing what we call the Melissa Loop,” says Gòdia. 

Dealing with waste

The first compartment in the Melissa Loop is a cylindrical bioreactor kept at constant 55º C that hosts bacterial cultures very similar to the ones found in a healthy human gut. It recycles waste generated by the crew, like human feces, urine, toilet paper, inedible parts of plants, biodegradable polymers—all kinds of ugly stuff a crew of astronauts would produce every day. It is constantly monitored by sensors and controlled by software that can adjust everything from temperature to pH. 

 

Finally, there is a filtration stage where any unprocessed solids are separated from the liquid, which contains minerals, ammonium, volatile fatty acids, and carbon dioxide. Another filtration system makes sure that no bacteria escape the bioreactor. 

 

The liquid is fed to the second compartment, an illuminated photobioreactor working with cultures of Rhodospirillum rubrum, a pink-colored bacteria that can either engage in photosynthesis or feed on fatty acids. The intensity of lighting is controlled automatically and used to regulate the growth of the bacteria. The main products of the second compartment are water-containing minerals and ammonium and biomass that could potentially be used as protein source. Rats fed with it for a couple of weeks during a food acceptability study turned out fine, so there’s that. 

Turning waste into air and food

The solution of minerals and ammonium flows to the third compartment, where two cultures of bacteria—Nitrosomonas europea and Nitrobacter winogradsky—await. First, the N. europea oxidize the ammonia into nitrite. Next, N. winogradsky oxidize nitrite into nitrate, which is a crucial nutrient for all living organisms. Together, the two stages form an important step (called nitrification) in the nitrogen cycle that operates on Earth. Since it’s an aerobic processes, the oxygen required is fed back to it from the next two compartments (4a and 4b in the diagram below), which are, in turn, fed nitrate-rich outflow.

“The purpose of 4a and 4b compartments is producing oxygen and a majority of biomass that is used as food,” explained Gòdia. 4a relies on cultures of Limnospira indica, cyanobacteria known as Spirulina, which produce oxygen and food. The choice of Spirulina was no accident; used as a diet supplement, it prevents bone loss, which is one of the most serious hazards in space travel. 

 

The 4b compartment is a hydroponic plant chamber. Carbon dioxide for the plants is fed from the first compartment and from the crew compartment. “At its current capacity, Melissa Pilot Plant can produce enough atmosphere to sustain one human. Currently, we use Wistar rats to emulate human respiration,” said Gòdia. 

MELiSSA-breathing rats

The rats live in an airtight isolator. Staff members working at MELiSSA Pilot Plant feed them and clean their cages using a transfer airlock. Carbon dioxide is transported out, and oxygen is fed in to keep the optimal atmosphere composition at all times. Artificial lighting is used to simulate the day-night cycle. “To test the system, we wanted living, breathing organisms that have their daily routines and circadian rhythms. We wanted to check if the system could react to constantly changing oxygen demand and variations in carbon dioxide production,” said Gòdia. 

 

It could. The rats spent hundreds of days breathing MELiSSA air with no issues.

MELiSSA Pilot Plant is probably the most technologically advanced bioregenerative life support system ever built. Given the lighting intensity and feed rate of carbon dioxide, minerals, and nitrate, it knows exactly how much oxygen will be produced. It can regulate all those parameters to achieve a target atmosphere composition and/or food output. The system is also surprisingly responsive, allowing it to react to emergencies. “If you need increase or decrease the oxygen level, just change the lighting conditions, and cyanobacteria will react in seconds,” said Gòdia. 

 

The MELiSSA project is all about optimizing metabolic processes by controlling the environment down to the tiniest details. The only thing the people involved in MELiSSA didn’t touch was life itself. “At the very beginning, we all agreed to refrain from using genetically modified organisms. I think this decision stemmed out of concerns about public perceptions. We wanted to stay on the safe side,” Gòdia explained. 

 

Not everyone was concerned about those perceptions, though.

Elon goes to Mars

The main reason that bioregenerative life support systems didn’t go mainstream in the 1990s was their excessive weight. In 2006, Harry Jones, a life support systems engineer at NASA Ames Research Center, did a study that showed a mission would need to last over 12 years for the bioregenerative system to break even and produce enough food to offset its weight. Nobody thought about missions that long at that time, so Jones concluded that despite significant funding and development time, bioregenerative life support systems appeared “surprisingly impractical.” 

 

Thinking shifted in 2016 when Elon Musk backed the idea of a permanent colony on Mars. Before that, plans for Mars involved another Apollo-like effort: go there, plant a flag, ferry some rocks back home, and never go back. Musk’s alternative was a long-term presence, long enough that bioregenerative systems would make way more sense.

 

In 2017, NASA founded the Center for Utilization of Biological Engineering in Space (CUBES), a conglomerate of federal agencies, industry, and academia with the goal of building a demonstration biosystem for a future Mars colony.

Engineering life

The bioprocessing system proposed by CUBES goes beyond traditional life support and is geared more toward manufacturing and in situ resource utilization (ISRU). Its key functions are the production of food, materials like bioplastics, and therapeutics. Those tasks are integrated with waste recycling. Optionally, it can produce breathable air, although this will most probably be taken care of by physico-chemical systems like the ones currently working on the ISS. 

 

What’s distinct about CUBES is its approach. While MELiSSA was focused on fine-tuning the hardware and software and left biology intact, CUBES involves engineering all three to make them work seamlessly together.

 

“It is not about a specific microbe or a specific crop. You can have a nice fruit or a nice vegetable that I want to eat a lot of, but we’re not going to use it if it needs eight thousand times more water than some alternative crops or needs huge amounts of light to grow. We look at all trade spaces in the system economy and choose most efficient solutions,” said Aaron J. Berliner, a bioengineering researcher at UC Berkeley and a member of CUBES. 

 

In terms of hardware, software, and automation, the system CUBES is aiming for will look much like MELiSSA with advanced, software-controlled bioreactors. The organisms in those bioreactors, though, are a different story.

Snacks, drugs, and plastic

“With genetic engineering we can address composing diverse diets,” said Shannon N. Nangle, a post-doctoral fellow in Pam Silver’s Lab at Harvard and founder of Circe Bioscience, a biotech startup working on decarbonizing food production. “You can engineer microbes to enhance certain amino acids; you can engineer them to make other things like fats, various nutrients, and sugars. Think of that as building blocks for making food to achieve a well-balanced protein content.”

 

Nangle built a fermentation technology that can make all those sugars, fats, and nutrients using proprietary, engineered microbes. Beyond supplying nutrients, the system’s output can be used to make meals with different textures, aromas, and flavors. 

 

This is a big step forward. The spirulina in MELiSSA’s bioreactors tastes like green, stagnant water. “What astronauts eat has profound effect on their well-being and motivation,” said Nangle. “During the first stages of Mars exploration we would probably go for the most calorie-dense foods because they take the least space. Later, we could move to less calorie-dense options that provide flavors, freshness, all those psychological enhancements people on the ground need,” she added. 

 

One of the first tests of this approach was a study Nangle led at Harvard back in 2020 in which she and her team engineered a single bacterium (Cupriavidus necato) to produce sugar (sucrose), polymers for plastics (polyesters), and lipo-chitooligosaccharides, which can act as fertilizer for plants like rice. All the bacteria needed was a mix of hydrogen and carbon dioxide. 

 

“The basic way this is done is by pulling some DNA fragments or genes from other organisms and simply transplanting them into a new organism to give it new functions,” said Nils Averesch, a senior researcher at CUBES and Stanford University. Those gene donors for Cupriavidus necato engineered by Nangle’s team were bacteria like E. coli, B. japonicum, and others. 

 

At the end of the day, CUBES is about engineering lifeforms to efficiently produce useful things like food, materials, or medication with whatever will be available on Mars. Microbes can even be modified to extract minerals and metals from regolith or alter Martian soil to make it usable for growing plants. “The thing to keep in mind with biology, though, is that the more shots at the goal you have, the better, because we don’t know what is really going on exactly,” said Nangle.

 

The problem is that life, when pushed, sometimes fights back.

Fighting evolution

“We call it the cheater problem,” said Berliner. “Genetic engineering moves organisms into an uncomfortable zone. They are not natural anymore—they become cheaters. In time, cheaters tend to evolve away from doing what you told them to do and toward doing what life does best: multiplying as fast as possible,” he added. Cupriavidus necato doesn’t benefit from producing plastics, table sugar, or fertilizers; it only thrives by producing more Cupriavidus necato. 

 

Given enough time, biology has a way of returning an organism to its evolutionary factory settings. But so do we. 

 

A standard way of combating evolution in synthetic biology is engineering the growth to happen only in a short time window so that evolution can’t kick in fast enough. There are other clever ways, too. “One technique used in the industry is called bolt-on armor,” said Berliner. It works by attaching multiple copies of any genes inserted into the target organism’s DNA. “Then, when it tries to evolve away from this gene—boom—there is another copy bolted on,” he explained. 

 

According to Berliner, however, bioreactors in space won’t rely on long-term growth of the same microbe population. “More likely, when you see drops in efficiency, you would just go to a freezer, take a new strain from storage, and use it to reboot the system,” Berliner said. One portable fridge can store fresh boot-up strains for hundreds of years, even if the crew were to reboot the bioreactors every day. 

Three stages of the Martian colony

The CUBES system is designed for a permanent base, one that will only be periodically occupied during the early stages of Martian exploration. So its development road map is additive—each new capability will be built on top of the ones already deployed. The CUBES team built their plan around Martian population numbers. 

 

The first stage is designed for a Martian headcount of less than a hundred people. The crew would rely mostly on prepackaged food brought from Earth. Bioreactors with engineered microbes would supplement that with micronutrients like iron, calcium, vitamins B and D, zinc, omega-3 fatty acids, and so on. Production of bioplastics would also be modest and limited to technology demonstrations (the majority of the colony’s materials being supplied from Earth). Local drug production would be limited to small molecules produced in bacteria. Water reclamation would be done through filtration, adsorption, and distillation of urine, very much as it is done today on the ISS. The brine that is leftover in this process, however, would be processed in spirulina bioreactors to provide a source of nitrogen.

The second stage would begin when the facility transitions to a permanent settlement that has to support over a hundred people. Photobioreactors producing various engineered algae and plants would be added and powered by artificial lighting and pressurized carbon dioxide retrieved from Martian atmosphere. The materials suitable for local production identified during the earlier trials would be mass-produced and used to expand the settlement, build greenhouses, and manufacture additional tools on demand. Engineered microbes would produce much-needed drugs. Other therapeutics made on the spot would include basic antibiotic precursors, insulin, and opioids, which are already made with biomanufacturing technologies on Earth. Reclamation systems will start processing solid human waste like feces, biodegradable plastics, and uneaten food. 

 

The third stage would begin when the Mars population exceeds 10,000 people. At this point, a Martian colony should have its own bio-foundry to engineer its own microbes on demand. “This won’t happen for a long time, though. It's a possibility only when the planet is colonized. We’re probably hundreds of years away from it,” said Averesh.

 

Microbes produced in this bio-foundry could be designed and tested on Earth, and the designs, in the form of DNA sequences, could be transmitted to Mars. Crops grown on Earth, like soybeans, potatoes, or peanuts modified to survive in Martian conditions, would be cultivated in large greenhouses directly in regolith pre-processed by microbes. Manufacturing of plastics would switch away from biology-based technologies, save for specialized materials; the rest will be made in chemical manufacturing plants. Biomanufacturing, however, will be kept as a primary technology for making drugs. Most of the waste will be recycled in large-scale, multi-stage bioreactors. 

 

NASA has to make some decisions about the basics before this grandiose vision can start materializing, however. “NASA needs to provide some metrics, like what do they want specifically? How long would the first mission last? What is the energy budget we have to work with? How much water we have? How many people we need to support? Knowing that, we could focus our work on designing organisms to make all that happen. So far, NASA funded CUBES with like 15 million USD in five years,” said Berliner. 

 

In the meantime, MELiSSA is sticking firmly to its European down-to-earth strategy of progressing slowly but surely. “We have been in talks with ESA recently and secured the funds for moving to the next stage with MELiSSA Pilot Plant, which is building a human-grade facility,” said Gòdia. “My estimate is within three to four years, we are going to be ready for first test campaigns with human subjects. I guess we are going to need volunteers.”