Casey Cox, who hails from a family with five generations of farming history along the Flint River in the southeastern U.S. state of Georgia, never expected to come back home. She’d graduated from the University of Florida in Gainesville with a natural resources degree. A big thinker, Cox could have taken her desire to make a difference in any number of challenging directions.

But she felt a pull back to her roots.

“I grew up playing in these woods and boating on the river,” Cox said, as we prepared to launch a double kayak to explore the Flint. “And it hit me that I had such a unique experience growing up on the farm.”

Cox plans at some point to take over the reigns from her father, Glenn, who not only turned sweet corn into a valuable crop for southwest Georgia, but whose homestead harbors one of the few remaining parcels of native long-leaf pine, among the most biologically diverse forests outside of the tropics.

But for now Cox is deep into another challenge. As executive director of the Flint River Partnership, a collaboration of the Flint River Soil and Water Conservation District, the U.S. Department of Agriculture’s Natural Resources Conservation Service, and The Nature Conservancy, she aims to figure out how to make agriculture in the lower Flint River Basin more ecologically sustainable. With populations of rare freshwater mussels dwindling in the lower Flint and Florida’s leveling of a lawsuit against Georgia claiming overuse of shared waters, there’s some urgency to the matter.

Cox and her partners believe smarter irrigation can be a big part of the solution.

A land of stately live oaks, swaying pines, and old cypress trees draped with Spanish moss, southwestern Georgia is a slice of rural America that displays a rare beauty wrapped in southern charm. Fields of peanuts, cotton and corn stretch as far as the eye can see. But the lower Flint’s $2 billion farm economy now faces a threat much more associated with the arid West than the humid Southeast: shrinking river flows and water disputes with neighbors.

The source of both the region’s bounty and trouble is the proliferation since the 1970s of an irrigation system called the center-pivot sprinkler. It consists of a giant horizontal sprinkler arm, often stretching the length of a football field, which pivots on wheels around a central point. While some systems rely on surface water, most in the lower Flint draw from the aquifer below. The irrigated field takes the shape of a giant circle, easily visible from the air. A typical center pivot might irrigate 200 acres (80 hectares).

As the center pivots rolled into southwestern Georgia, the farm economy took off. But so did water use. Peanuts and other field crops in the region typically require 22-25 inches of water. In a dry year, rain provides only about half of that amount. To make up the difference, farmers activate their sprinklers and apply about a foot of water to their fields, or nearly 326,000 gallons per acre.

Today some 8,900 center pivots dot the landscape of the lower Flint. Farmers mainly draw water from an aquifer called the Upper Floridan, a highly productive karst formation that underlies most of Georgia’s coastal plain. The aquifer is shallow, so drilling a well into it costs about a quarter as much as drilling down to a deeper formation. But the Upper Floridan is hydrologically connected to the Flint River and its tributaries. Especially during the low-flow period of June to October, groundwater discharge can account for much of a stream’s total flow. As irrigation pumping increased, stream levels during this critical period declined. When scientists compared recent minimum streamflows with pre-irrigation levels, they found declines in some tributaries of 30-40 percent. Long stretches of Spring Creek, historically a perennial stream, dried completely during eight summers between 2000 and 2011.

The Flint River Basin was once home to twenty-nine species of freshwater mussels, one of the most diverse assemblages in southeastern North America. But periods of drought combined with irrigation withdrawals have caused a major decline in mussel diversity and numbers. As flows diminish, critical habitat along the stream banks disappears. Further drops in flow expose shoals and eventually the streambed. Due to their limited mobility mussels can get stranded, exposing them to dehydration, high temperatures, low oxygen levels and predation.

While known by playful names like Oval Pigtoe (Pleurobema pyriforme), Little Spectaclecase (Villosa lienosa), and Shiny-rayed Pocketbook (Hamiota subangulata), mussels do serious work to help keep rivers clean and healthy. They filter pollutants from the water, move and cycle nutrients, and provide food for otter, muskrats, birds and fish.

“Mussels are a critical component of stream food webs and one piece of the equipment that makes a stream work well,” said Stephen Golladay, an aquatic biologist at the J. W. Jones Ecological Research Center in Newton, Georgia. While Golladay emphasizes that there’s a lot we don’t know about mussels, his surveys performed with colleagues at universities and state and federal agencies find that populations are declining in the water-stressed middle reaches of Flint tributaries. “Spring Creek has seen pretty extensive die-offs,” Golladay said.

Today, the U.S. Fish and Wildlife Service lists seven species of mussels in the Flint River Basin as threatened or endangered, along with one species of fish, the Gulf sturgeon.

In 2000, in the midst of a severe three-year drought, the University of Georgia (UGA) opened a new research facility near the rural town of Camilla, Georgia. Called the C.M. Stripling Irrigation Research Park, its mission is to improve the efficiency of irrigation in the southwestern part of the state.

“The water wars were going on and the university realized it needed to do more in the lower Flint,” said Calvin Perry, an agricultural engineer who runs the research center.

Perry was referring to the ongoing dispute between Alabama, Florida and Georgia over the use and allocation of the waters that make up the Apalachicola-Chattahoochee-Flint (ACF) watershed, which extends from north of Atlanta south through Florida’s panhandle to Apalachicola Bay. The Flint joins the Chattahoochee, which flows along the southern half of the Alabama-Georgia border, to form the Apalachicola, home to one of the highest concentrations of imperiled species in the United States. Apalachicola Bay ranks among North America’s most productive estuaries, yielding highly prized harvests of blue crabs, shrimp, and more than 90 percent of Florida’s commercial oysters. Successful reproduction of these fish depends upon the salinity levels maintained by freshwater flowing into the bay.

The three states signed a compact in 1997 setting out a framework to develop a water-sharing formula, but the compact expired in 2003 with no resolution. In October 2013, Florida sued Georgia in the U.S. Supreme Court, effectively asking the court to decide how much water each state can use. A decision is expected in 2017.

Meanwhile, Perry got busy with his UGA colleagues and the Flint River Partnership to develop and spread more-efficient ways of watering farm fields in the lower Flint. Many farmers have now added drop hoses with low-pressure sprays to their center pivot sprinkler arms. Instead of spraying water high into the air, where wind and evaporation can rob up to 40 percent of the irrigation water, the upgraded sprinklers deliver bigger drops at lower pressures and closer to the crops, boosting efficiency from 60 percent to over 80 percent.

The team also saw great potential in better scheduling how much water got delivered to different parts of the farmers’ fields. The result was an innovative technology called “variable rate irrigation,” or VRI, which essentially tailors water application to field conditions. On average, about a tenth of each field in the lower Flint is taken up by roads, wildlife corridors or wetlands, and is not growing a crop. VRI involves programming a GPS-equipped center pivot to shut off as it passes over those non-crop areas. That can reduce farm water use by 10 percent.

Using a similar on-off mechanism, VRI can also enable a farmer to apply less irrigation water where soils naturally retain more moisture. That typically brings water savings to 15 percent. If adopted widely in the lower Flint Basin, these irrigation upgrades could bring about a sizeable reduction in water withdrawals from the Upper Floridan Aquifer and area streams.

Despite its water-saving potential, VRI, which was commercialized about a decade ago, has been slow to permeate the market. “Manufacturers will tell you it hasn’t taken off yet,” Perry said during my visit to the research park in August 2016. “Despite our many years of effort, it’s still seen as an add-on.”

VRI now qualifies as a standard conservation practice under U.S. farm programs, which enables irrigators to get financial assistance to purchase it. But more incentives and technical assistance seem to be needed to expand its adoption. The cost of installing VRI varies with factors such as the brand and length of the center pivot, but runs about $13,500. Farmers can typically offset that investment through reduced pumping and fertilizer costs over the life of the system.

“The technology is ahead of the support capacity,” said Cox, who works closely with the UGA team. She says many farmers need guidance in setting up and operating a VRI system. “It’s all about helping farmers make better decisions.”

With the lawsuit pending, the state of Georgia appears hesitant to take much action. It did, however, re-institute a moratorium on permits for new wells drilled into the Upper Floridan – essentially a cap on groundwater use, albeit at levels already harmful to streams.

George Vellidis, a UGA professor of crop and soil sciences, figures it would take $120 million to upgrade all the center pivots in the lower Flint with standard VRI technology. “If the state said we’re going to try to solve the water wars by investing in VRI, you’d have an explosion of use,” he said.

Out in the field at the Stripling Research Park, Vellidis and Perry showed me the next generation of their technology, which they call “dynamic VRI.” It combines real-time soil moisture monitoring with VRI to schedule the timing and volume of irrigation much more precisely. The sensors, which are strategically placed in different parts of the field, give a numeric reading of soil moisture. That information is used to create a “prescription map,” a representation of how much water to apply to each zone in the field that day. The map can be sent remotely to the pivot’s VRI control panel.

In 2015, Vellidis and his colleagues worked with a local farmer to test the dynamic VRI process on a 230-acre (93-hectare) field of peanuts. Over the entire growing season, the dynamic VRI system recommended applying 30 percent less water to the field than did a commonly used irrigation program. The experimental and control fields got nearly the same yield of peanuts. Overall, dynamic VRI boosted water productivity – yield per unit water, or crop per drop – by an impressive 43 percent.

But can dynamic VRI restore stream flows and help mussels and fish hang on in the lower Flint Basin?

Cox, Perry and Vellidis took me to the farm of Marty Tabb outside the town of Colquitt, Georgia, where another dynamic VRI project is attempting to shed light on that question. Tabb’s farm is near Spring Creek, with its dangerously low flows and threatened mussels. Tabb’s wife works for the NRCS, one of Cox’s collaborators in the Flint River Partnership. At Tabb’s farm, Cox and the UGA team have equipped a 180-acre peanut field with their latest precision irrigation scheduling equipment and will compare the results with Tabb’s control fields.

Whether Spring Creek gets a bump in flow is the million-dollar question, and it will likely take more than one season and one field to answer it. Along with the Georgia Environmental Protection Division, the Change the Course initiative I helped launch provided funding to the Flint River Partnership for this project.

It’s clear that restoring healthy flows to the Flint River and its tributaries will not be easy. The people of Georgia and the lower Flint must decide if they want to take action to save the diversity of life in their streams. Perhaps a Supreme Court decision will say that additional flow from the Flint must make it to the Florida border. But one way or another, as Vellidis put it, “it’s got to be both a policy and technology solution.”

In the late afternoon light, the cypress trees lining the Flint cast dancing shadows along the river banks. Dark clouds gathered overhead as Cox and I kayaked down a river that seemed far too placid to be the cause of so much fighting. Just as my thoughts turned to the challenge that lay before Cox and her team, an alligator swam stealthily across the river not far in front of us. If we stayed our course, the kayak would meet it head on. Just as we planned a cut to the right, the alligator slinked under the water. With no way to tell which way the gator had gone, Cox suggested we just pick a path, paddle stronger, and continue on toward our destination.

Stay calm and paddle on. It seemed a fitting metaphor for the challenge Cox and her partners have taken on in the lower Flint River Basin.

I am grateful to the C. S. Mott Foundation for support to undertake research and travel for this piece.