Growing Food in a Desert City Using Rainwater and Urban Runoff

The following is one of my favorite water-harvesting stories. It comes from one of my mentors, Russ Buhrow, and has inspired me in much of my work. It is amazing what Russ produced with stormwater, something too many people consider to be a waste or a liability, but as Russ shows, is actually a great resource.

In summer 1980, plant sciences graduate student Russ Buhrow decided to take a break from the books to gain "hands-on" knowledge by dryfarming in the middle of the simmering desert city of Tucson, Arizona, where annual rainfall averages 12 inches (305 mm). Taking his cue from the ancient traditions of indigenous Tohono O'odham, Russ raised his crops solely on direct rainfall and runoff harvested from short bursts of sporadic summer monsoon rains. Yet Russ's situation was somewhat different from his Native American neighbors. He didn't farm alluvial flats of a healthy desert ecosystem where runoff from surrounding mountains flows down a braiding arroyo to a field. Instead, he farmed a semi-urban vacant lot between a dry riverbed and cinder block apartment buildings. Rather than intercepting runoff from low desert mountains and foothills, Russ learned to harvest runoff from rooftops, yards, parking lots, and a city street.

Russ began by observing the gradually sloping arable vacant lot. Wheel ruts crisscrossed it. Random piles of compacted debris were strewn about, and dense weeds grew in depressions where rainwater and organic matter collected. "Ah ha!" thought Russ. "The weeds grow where the water is, so that's where my garden will go!"

With permission from the landowner he dug several 8- x 8-foot (2.4- x 2.4-m) sunken garden beds where the weeds grew tallest. Berms stretched to either side on the downslope side of his basins like open arms of a big welcoming hug for rainwater runoff (see Fig. 11, after article, for the multi-year progression of Russ' fields). He planted seeds as the summer storms rolled in. Thunder cracked, lightening flashed, and the rain came down in sheets. Russ stood in the middle of it all and watched. He saw water pool in the garden basins. He also noticed water pouring off a parking lot just upslope of the basins and quickly dug a ditch as a diversion swale, directing runoff from parking lot to garden.

After 40 minutes the rain stopped. With the soil now saturated, his seeds germinated in just 3 to 5 days. Russ was full of adrenaline. As he says, "When you see the rain flow like that, it's a EUREKA moment. You see the water and realize, this really works!" The parking lot ditch had just increased his water resources five-fold. He went right to work, expanding his planting area to 700 square feet (65 m²).

From then on, no matter how far away he was, Russ always ran to the garden when rain started to fall. "It's amazing how much water you can catch in neighborhoods," he says. "You just need to watch the sheet flow when it rains." Rain and runoff revealed the land's subtle slopes and depressions. He saw how much water flowed and where. Then he figured out how to catch and use it.

For Russ this was a rush - like playing "flood" in a huge sandbox, with lightening! The unlocked car was always nearby so he could leap in when the lightening struck too close.

By fall, drought-hardy tepary beans, black-eyed peas, corn, squash, and devil's claw (a fiber plant with edible seeds and okra-like immature fruit) were harvested–all irrigated only by rain (Figs. 1, 2, 3).

In winter the fields went fallow, but Russ stayed active. Once, standing in the rain, he saw excessive runoff rushing down Columbus Boulevard, an asphalted arterial street that dead ended 400 yards (360 m) from his garden. "FREE WATER!" Russ yelled. He dug a 1/4-mile (0.8 km) long diversion swale/ditch from the street to his garden (Fig. 4A and Fig. 4B). He reworked the old garden beds and added new ones. When finished, the garden basins ranged from 100 to 700 square feet (9-63 m²) each. Together they resembled a series of stepped terraces that directed overflow water from the upper gardens to the lower gardens (Figs. 5, 6, 7, 8). The combined planting area was now 1/2 acre (0.1 ha).

That summer two big storms flooded the long swale with water 8 inches (20 cm) deep and 2-3 feet (60 - 90 cm) wide. Both storms lasted less than an hour and made Russ run through the garden, euphoric from the water, terrorized by lightening, and exhausted by effort. With the diversion swale running above (upslope of) the top of the terraced garden basins, Russ could temporarily divert the flow into a specific field by cutting an opening in the swale's berm, then blocking the water's flow within the swale at a location just downslope of the cut, to force more flow into this field. He created this "blockage" by throwing wads of weeds and brush into the swale and packing them down with his shovel, his feet, and rocks, then piling dirt behind the brush. Next, he'd run down the swale to the next field, and make a new cut in the berm and a new barrier just downslope of it. Running back up to the first diversion Russ would wait until the first field was well watered, then remove the original barrier of weeds, rock, and soil, and use this material to plug the cut in the berm so the water flowed down the main swale again. That done, he'd run down the swale to make sure the second cut and barrier were diverting water into his second field. If all was well, he'd dash off to the next field for another diversion cut and barrier. And so on, until the water stopped flowing.