Pinning down the pesticides

California’s Central Valley — 20,000 square miles of lowland ringed by mountains — produces a quarter of the nation’s food. But this abundance comes with environmental consequences: there are so many chemicals sprayed across the valley’s croplands that the region’s waters are often tainted. 

That’s a broad statement, and for now, it has to be. The pesticide problem is so complex that it’s hard to even know the extent of the damage.

Bifenthrin is sprayed on almonds. Cypermethrin is sprayed on alfalfa. Fenpropathrin is sprayed on grapes. And none of these chemicals are sprayed alone. In aggregate, hundreds of chemicals are applied to hundreds of different crops — settling into soils, then tumbling into ditches, then running into creeks and rivers, until eventually these concoctions reach the Sacramento-San Joaquin Bay-Delta (Bay-Delta). Each chemical sticks in soils for a different length of time — and that length can vary by orders of magnitude for any single chemical, depending on the specific type of soil — and is carried by water at different rates. Each chemical impacts various species in various ways. 

Nonetheless, Nicol Parker, a doctoral student at the University of California, Santa Barbara, has decided it’s essential we make this swirl comprehensible — a necessary first step if we want to solve the problem.

When Parker received a 2020 Delta Science Fellowship, her plan was to adapt an existing data tool, provided by the U.S. Environmental Protection Agency, so it could consider the higher resolution data available in California. 

“That did not work out,” she concedes now. After discussing the tool — the Hydrologic and Water Quality System — with the EPA, Parker learned that tracking cumulative loads of pesticides would be more difficult than she had imagined. There weren’t any other tools that could do what she wanted. So she had to make her own.

That meant creating code that could wrangle data from the reports that the state of California requires farmers file about their pesticide use. It meant incorporating land use survey data, factoring in crop rotation, tracking down reliable toxicity data for thousands of unique pesticide products, and, wherever there were gaps in the data, finding reliable alternate sources.

The result is the Pesticide Management Prioritization Model (PMPM), which reveals which chemicals contribute the most to overall toxicity in the region — and, crucially, which crops correspond with these chemicals. Users can examine impacts locally, accounting for what’s been sprayed in a particular region. Parker is currently working to validate a second and more complex module that considers downstream effects, too, incorporating data on how each chemical moves throughout the watershed.

Parker grew up on a tree farm in rural Michigan — a place where almost everyone considered pesticides essential to their livelihood. She knows farmers aren’t likely to give up on pesticides any time soon, nor does she want them to. Her goal is to provide clear data so that farmers can determine which pesticides provide the most bang for their buck with the least impact on the surrounding environment. 

For example, Parker used the model to pinpoint bifenthrin, which is sprayed on almonds, as a particularly large source of toxicity. Growers might choose a different chemical that has less impact on non-target species. Because the model also considers economic impacts, Parker can identify crops that farmers in at-risk watersheds should avoid: safflower, for instance, contributes an outsize dose of toxicity given farmers’ relatively low gross profits. 

For anyone to actually use this data, it has to be friendly — “super simplified, clickable, colorable interfaces,” Parker says, and simple enough to use on a smartphone. She’s created such an interface for the Central Valley, and her methodology could be used by policymakers elsewhere to create similar tools.

Parker is now working to further improve the model. She wants to incorporate data on irrigation methods, which affect how chemicals are transported. Past efforts to “improve” pesticides have made them less toxic to humans and other mammals — but that resulted in concoctions that are more toxic to invertebrates and fish. So Parker wants to add data that does not treat toxicity as one general figure but breaks it apart for different classes of species. Finally, Parker thinks clearer information about the soil types across the valley — and how those soil types impact chemical transport — will help complete the picture.

Parker knows she could work on this model for the rest of her career, adding more and more bits of information for an ever-more-refined picture of the chemical world. But she plans to wrap up her work next year. Too many data projects are built and then never emerge from the academic world, Parker thinks. “My highest aim,” she says, “is not getting too caught up in just finding all the nuggets, but really getting this tool actually used on these issues.”