Decades of Data Reveal How Herbicides Really Fight Invasive Plants in a Tidal Estuary

In the vast tidal labyrinth of California’s Sacramento–San Joaquin Delta, an expensive and decades-long war is being waged against floating invasive plants that choke waterways, smother native habitat, and clog the infrastructure that moves water to millions of people. Now, one of the most comprehensive assessments ever attempted has delivered a verdict that is both encouraging and sobering: the two workhorse herbicides used against these invaders do work, but only under specific conditions, and their effectiveness depends far more on where and how they are applied than managers might have assumed.

A research team led by Shruti Khanna of the California Department of Fish and Wildlife and the University of California, Davis, together with Jereme Gaeta, Krista Hoffmann, and remote sensing pioneer Susan Ustin, analyzed the Delta’s aquatic invasive plant control program across two five-year windows, 2004 to 2008 and 2014 to 2018. Their study, published in the journal Environmental Management, combined more than six hundred treated and reference sites with multi-year airborne hyperspectral imagery capable of mapping floating aquatic vegetation at the genus level. By fusing these vegetation maps with detailed herbicide application logs and boat path records, the team could ask questions that no small-scale field trial could answer: does the type of herbicide matter, does the dose matter, does the number of visits matter, and does any of it last?

The stakes are enormous. California spends more than 13 million US dollars each year treating invasive aquatic plants in the Delta, a 2,220-square-kilometer estuary formed by the confluence of the Sacramento and San Joaquin rivers. The main villains are water hyacinth (Pontederia crassipes), a Brazilian native that has plagued the Delta since 1904, and water primrose (Ludwigia hexapetala), present since 1949 and increasingly dominant since 2014. Two newer arrivals, spongeplant detected in 2008 and alligator weed in 2017, add to the threat. These plants form dense floating mats that darken water, deplete oxygen, block fish movement, and displace native communities, while their drifting fragments seed new infestations throughout the tidal system.

The control program itself has a long institutional history. Funded after 1982 legislation, it began as the Water Hyacinth Control Program and was renamed the Aquatic Invasive Plant Control Program in 2019, administered by California State Parks’ Division of Boating and Waterways. For most of its history it has relied on two systemic herbicides with remarkably long pedigrees: 2,4-dichlorophenoxyacetic acid, patented in 1945, which disrupts plant growth hormones and kills broadleaf dicots while sparing monocots, and glyphosate, patented by Monsanto in 1974, which blocks the shikimate pathway enzyme EPSPS essential to plant survival. The program straddled a major transition, relying mainly on 2,4-D in the early period and shifting to glyphosate in the later years, a natural experiment the researchers exploited to evaluate each chemical independently.

The analytical approach was rigorous. The team fitted mixed-effects logistic regression models, treating the probability of floating vegetation presence at each site as a function of herbicide application rate, spray frequency within a season, and consecutive years of treatment, with random intercepts for site and year to account for repeated observations. Separate models were built for each of the Delta’s three hydrologically distinct habitat types: fast-flowing tidal channels, slow shallow waters between remnant wetland patches, and flooded islands, lake-like expanses created by levee failures where water lingers and tidal action is weak.

The results were strikingly habitat-dependent. Both herbicides performed best in slow, shallow, open-water habitats and flooded islands, and worst in fast-flowing channels. Glyphosate was essentially ineffective in channels at any application rate, while 2,4-D showed significant impact there only at the highest rates used in the field. The likely explanation lies in the Delta’s relentless tidal dynamics. Previous work using GPS trackers showed that water hyacinth mats drift constantly in all directions with the tides, so treated plants in channels are quickly replaced by fresh propagules floating in from untreated areas. Flooded islands, isolated from these tidal conveyor belts, allow treatment effects to persist and become detectable months and even years later.

Perhaps the most actionable finding concerns dose and frequency. Across habitats and time periods, applying a higher concentration of herbicide in fewer visits consistently outperformed delivering the same total amount in smaller, more frequent doses. Low-dose applications were not merely ineffective; in some cases they were associated with a higher probability of floating vegetation than untreated reference sites. The authors point to two possible mechanisms. The first is hormesis, a well-documented phenomenon in which sublethal chemical stress actually stimulates plant growth and defense mechanisms, potentially triggering a post-treatment rebound. The second is herbicide resistance: sites treated with glyphosate for more than six consecutive years required noticeably higher application rates to achieve significant reductions than sites treated for only one or two years, a pattern consistent with resistance evolving under repeated sublethal selection pressure.

The legacy effects, meaning how long benefits persisted after spraying stopped, differed sharply between the two chemicals. Sites treated with 2,4-D showed no detectable reduction in floating vegetation one, two, or three years after treatment ended, in any habitat. Glyphosate, by contrast, left a measurable legacy in flooded islands, where the probability of floating vegetation remained significantly below reference levels even three years after the last application. In slow shallows, longer histories of glyphosate treatment, five to six consecutive years, produced stronger lasting suppression than shorter histories. In channels, neither herbicide left any lasting benefit, and longer treatment histories there appeared if anything counterproductive.

The study also revealed an unsettling background trend: floating vegetation cover and occurrence probability were higher across all habitats in 2014–2018 than in 2004–2008, suggesting the invasion problem has intensified even as treatment continued. Slow shallows consistently showed the highest occurrence, likely because their nutrient-rich, slow-moving waters combine continuous propagule input with ideal conditions for fragments and seedlings to take root.

For managers, the recommendations are concrete. Avoid sublethal doses that risk stimulating growth or breeding resistance; concentrate herbicide into fewer, higher-dose applications; match herbicide choice to habitat, favoring glyphosate’s longer-lasting action in stable flooded islands while recognizing that neither chemical offers much hope in fast channels; and rotate active ingredients, a strategy the Division of Boating and Waterways has begun testing with newer herbicides such as imazamox and penoxsulam in demonstration zones. Above all, the authors argue, chemical control must be embedded in a holistic program of monitoring, integrated methods, and native plant restoration. What this study demonstrates most powerfully is the value of scale: only by combining ten years of treatment records with a decade-spanning remote sensing archive could the true contours of herbicide efficacy in a tidal ecosystem come into focus, and only at that scale can the difference between money spent and money wasted finally be seen.

Subject of Research: Herbicide efficacy against invasive floating aquatic vegetation in the Sacramento–San Joaquin Delta

Article Title: Multi-Year Regional-Scale Efficacy Assessment of Two Herbicides for Treatment of Invasive Floating Aquatic Vegetation in a California Estuary

Article References: Khanna, S., Gaeta, J. W., Hoffmann, K., & Ustin, S. L. (2026). Multi-Year Regional-Scale Efficacy Assessment of Two Herbicides for Treatment of Invasive Floating Aquatic Vegetation in a California Estuary. Environmental Management, 76(10), Article 326. https://doi.org/10.1007/s00267-026-02625-8

Image Credits: AI Generated

DOI: 10.1007/s00267-026-02625-8

Keywords: glyphosate, 2,4-D, invasive species, water hyacinth, water primrose, Sacramento-San Joaquin Delta, remote sensing, estuary management, herbicide resistance, aquatic plants, tidal wetlands, environmental management