A tomato can double its fruit in a pot of fake Martian dust and still leave the farm behind. That is the awkward lesson sitting under a much louder week in space food. On September 24, 2026, NASA named the winners of its Deep Space Food Challenge: Mars to Table and handed the top prize, $300,000, to Chinyere “Chichi” Ukeje of Philadelphia for a design called Adaptive Nourishment Infrastructure. Two days earlier, a California nonprofit said it had evolved a bacterium that eats Martian dirt. Neither announcement grows a carrot in perchlorate. The greenhouse trial that still sets the floor for mixed crops on simulated Red Planet soil was published in 2024/00005/1171/1906 (英语)而番茄仍然是那泥土中唯一的赢家
Ukeje is 25. She grew up in Bowie, Maryland, studied computer science at the University of Maryland, Baltimore County, and now lives in Northeast Philadelphia. NASA judged 113 entries from 33 countries and 28 U.S. states. The mission on the page was blunt: feed 15 astronauts for 500 Martian sols, about 513 Earth days, on the surface, with a layout, a meal plan, a concept of operations, and a walkthrough video. The federal notice had advertised a $750,000 purse. The agency said it awarded a combined $650,000. Jennifer Edmunson, who runs Centennial Challenges at Marshall Space Flight Center, said the agency was “thrilled to keep advancing the future of space food systems with this challenge.”
Ukeje’s concept, named for Ani, a Nigerian earth goddess of harvest, is a modular loop: controlled-environment crops, fermentation, fungi, and bioreactors that recycle nutrients, with a limited stash of food shipped from Earth. NASA’s description says the system is meant to produce half the calories away from this planet and to cook a recognizable meal when power, water, hardware, or crew time runs short. Second place, $200,000, went to Cislune of Rosemead, California, for Fresh, Ferment, Reserve — grow selected crops, turn part of the harvest into familiar food in instrumented cassettes, and keep a protected Earth-loaded reserve for the day a batch fails. Smaller prizes went to a Hilo, Hawaii team using an ahupua’a, a land-to-sea management idea, as the model; to Orbital Health Systems of Evansville, Indiana, for a settlers’ cookbook; to a Bentonville, Arkansas collective for a closed-loop sustenance design; and, as international winner, to Astrofood of Ellezelles, Belgium.
Those are kitchens on paper. The dirt problem is older and meaner. Astrobiologist Rebeca Gonçalves, with Wieger Wamelink, Peter van der Putten, and Jochem Evers at Wageningen University, put cherry tomatoes, peas, and carrots into NASA’s MMS-1 Mars simulant, into sand, and into potting soil. Some pots were mixed, the old companion-planting pattern associated with Maya fields. Some were single crops. Rhizobia bacteria went in with the peas to pull nitrogen out of the air. The greenhouse was tuned to the air, heat, and humidity of a pressurized Martian grow-room. It was the first time that mix had been scored on a regolith stand-in.
豆子应该是英雄
Wamelink had bet on the legume. “We had expected the peas to grow the best but the opposite turned out to be true. The tomatoes grew the best,” he said. In the simulant, intercropped tomatoes roughly doubled the fruit of tomatoes grown alone — more berries, bigger berries, earlier flowers, thicker stems, higher above-ground biomass, and more potassium in the yield. Gonçalves told reporters she had not known what to expect, because nobody had run the method on space agriculture before, and later called the tomato result “a big find, one that we can now build further research on.”
The system did not double. In Mars simulant, carrots lost biomass and yield. Peas showed no significant gain and trended down. The relative yield total was 0.93, a net loss against growing each crop alone. The bacteria never formed the nodules they need. High pH, compacted grains, and missing nutrients appear to have beaten them. Tall tomato and pea canopies then shaded the carrots. Wamelink’s line was that the tomato took from the peas and “the carrot most certainly did not.” Put the same trio in sand, where nodules formed, and the math flipped: relative yield 1.32, a real advantage. The paper’s own conclusion is narrower than the headlines that followed it. Mix the crops after you fix the dirt enough for the bacteria to live. Before that, companion planting is a tomato subsidy.
毒药还在粉末里
MMS-1 copies texture and chemistry. It does not fully copy the perchlorate salts that ride in real Martian regolith, often cited around half a percent by weight and higher in places. Those salts are bad for plants and worse for people. A September 2026 open-access paper from the Plant Trek project took a Utah microbial community that reduces perchlorate and put it into a basaltic simulant spiked to about 2.25 percent calcium and magnesium perchlorate. Water-extractable perchlorate fell to roughly 7 to 9 percent of the starting amended level. It did not go to zero. Aerobic conditioning did something the airless jars did not: plant-available phosphorus jumped to about 222 milligrams per kilogram. A separate Soil Science Society of America Journal paper this year found that baking a spiked simulant at 470 degrees Celsius nearly wiped the salt out, and that three leaching cycles plus distillation of the rinse water also cleared it. A farm microbiome they pushed through directed evolution improved from about 35 percent reduction to about 52 percent. Half a toxin is not a garden.
On September 22, 2026, Pioneer Labs in Emeryville said it had a different microbe, designated sPL.001. Chief executive Erika Alden DeBenedictis described an organism evolved across more than a trillion cell divisions so it can pull nutrients, including trace elements, from Martian dirt, water, and processed air, then make a bioplastic building material. The group says a bioreactor small enough to ride one rocket could, on their numbers, turn local feedstock into shelter material, and that the strain does not need fertilizer shipped from Earth. A preprint they point to reports more than three times the polymer yield of the parent strain under simulated Martian conditions. That is a construction claim, not a salad. The organism is meant to live in a stirred, heated tank, shielded from radiation, cold, and thin air. It does not replace the nodulation failure in a tomato pot. It does put a live tool on the same poison the plants cannot yet ignore.
休斯敦在水中种植花椒 而不是灰尘
NASA’s second CHAPEA crew has been inside a 1,700-square-foot 3D-printed habitat at Johnson Space Center since October 19, 2025. Ross Elder commands. Ellen Ellis is the medical officer. Matthew Montgomery is the science officer. James Spicer is the flight engineer. The run is 378 days and is scheduled to end around October 31, 2026. On May 7 they crossed day 200, in the middle of a simulated two-week blackout meant to mimic Mars passing behind the Sun. Ellis said she was proud of a crew that kept finding “new ways to improve our performance and efficiency daily.” They have grown crops. Agency photos show peppers and other vegetables in a hydroponic rack, harvested by hand. Hydroponics skip the regolith. That is the point, and the limit. A box of nutrient solution in Houston does not answer what Gonçalves measured in the simulant.
Wamelink’s spinout, BASE, has kept building greenhouse prototypes for the Moon and Mars, including chambers and insect modules. None of that moves the 0.93. Ukeje’s prize is a serious design for a crew that will need tacos and pasta nights, not freeze-dried bricks, and fungi belong in a closed loop because astronauts will not eat the stems. The loop still assumes a substrate that is not trying to poison the roots. Until perchlorate is cooked, washed, or eaten down by microbes that also leave phosphorus behind, the first honest Martian acre is a chemistry bench. The tomato already told us which plant will steal the meal if we pretend otherwise.
必须将收获数据带回家的管子 和NASA的AEGIS联盟一样 灵敏的宇宙飞船在一亿四千万英里外,当 航天部队重新获得一颗已死亡的冷战气球卫星 过了四分之一世纪 水是死地的另一半 农场, 这就是为什么档案片 太阳能海水淡化 仍然属于这一个旁边。 更多文件在 常规.
