California groundwater priorities shift with nitrogen data choice, UC Davis finds

The data used to estimate nitrogen surplus can change which crops and regions receive priority for California groundwater protection, a University of California, Davis-led study found. The research, published Nov. 10 in Nature Sustainability, tested 24 scenarios using different combinations of agricultural datasets.
The result matters because nitrogen-surplus estimates help regulators identify where farm nitrogen management or conservation measures may be most needed. A surplus is the nitrogen remaining after crop harvest removes nutrients from a field; some can move below the root zone and contribute to groundwater nitrate loading. If the underlying data changes the estimated surplus, a map of priority areas can change too.
Different nitrogen data produce different estimates
Researchers used California’s Central Valley as a data-rich test case. The study compared five datasets covering fertilizer application and nitrogen removed in harvested crops: U.S. Department of Agriculture surveys and census data, grower-reported records, UC Davis Cost and Return Studies, and county crop acreage and harvest records collected by the California Department of Food and Agriculture.
Each source captures different information and has limits. USDA surveys provide less detail for specialty crops and are not collected as frequently as annual farm records. Grower reports can differ because of recall or interpretation, while the UC Davis studies provide representative crop-level estimates rather than measurements from each field. County records document acreage and harvest, but do not directly measure fertilizer applied on individual farms.
The team held its nitrogen-balance framework and land-use maps constant while varying the source data and assumptions about manure and nitrogen stored in perennial plants. The spatial model ran at a 50-meter resolution across 17 crop categories, representing about 20 of the Central Valley’s dominant crops. Those crops covered about 75% of the valley’s irrigated cropland and accounted for more than 90% of estimated potential nitrogen loading below the root zone.
The scenarios define a bounded range of plausible outcomes, not 24 equally likely forecasts. Rice was excluded because flooded fields in the region typically lose little nitrogen through groundwater leaching, with denitrification accounting for more of the nitrogen loss.
Field crops showed the widest variation
Estimates diverged most for corn and other field crops that can receive dairy manure, including cotton, wheat, and oats. The study calculated average nitrogen surplus for corn at about 300 kilograms of nitrogen per hectare per year, with a standard deviation of 190 kilograms across the scenarios. That spread reflects model estimates under varying data and assumptions, not direct measurements at every corn field.
Orchard and vineyard crops generally had narrower uncertainty ranges and lower estimated surpluses. The researchers found that differences between datasets could also change how townships ranked for mitigation. In some cases, the variation in estimated surplus was large relative to the average, making a single-point estimate an unreliable guide to the precision of a local priority map. Other field studies have measured how management choices can cut nitrogen losses, and how large the payoff from reducing nitrogen losses from farmland could be.
Uncertainty affects California groundwater planning
California’s nitrogen rules require growers in affected areas to report nitrogen use, giving regulators a locally generated data stream alongside federal surveys and other records. The study supports using multiple sources, while cautioning against treating one dataset as definitive. Thomas Harter, a UC Davis professor and Cooperative Extension specialist and the study’s corresponding author, said regulators should “be cautious” because differences among datasets can rearrange the map used to target conservation efforts.
For fertilizer and farm advisers, the findings underline a distinction between estimating nutrient balances and measuring nitrate in a particular well. The model assesses potential nitrogen surplus and loading, not the concentration of nitrate in every groundwater source. Local decisions still depend on site conditions, farm records, and monitoring data, as well as the chosen method for estimating applied and harvested nitrogen.
The authors said the Central Valley’s data problem may also matter in regions that rely on only one or two agricultural datasets. There, uncertainty could remain hidden rather than appear as a visible range. Their conclusion is that nitrogen assessments should carry data-source uncertainty through to the final estimates, so regulators can see how robust a hotspot ranking is before directing protection resources.
More consistent farm reporting and improved information on fertilizer use, harvest, and manure could narrow the gaps. Until then, the study argues, groundwater protection plans should account for the possibility that a different but credible dataset may produce a different set of priorities.
Source: Global Agriculture

Enjoyed this story?
Every Monday, our subscribers get their hands on a digest of the most trending agriculture news. You can join them too!









Discussion0 comments