In the summer of 2022, we wade with shovels and sampling tubes into rivers around Tübingen, crawl over wooden boards that have been laid over a drained lake and shovel sediment samples from its muddy bottom. We want to know what happens to glyphosate washed from fields by rain. Where do deposits form? We are surprised by the laboratory analysis: Some sediment samples contain as much glyphosate as farmland a few days after spraying the herbicide.
Could the sediment in the lakebed form a reservoir for glyphosate? We want to drill deeper and extract further sediment cores. This time we take palaeontologist Annett Junginger with us for sampling. She also arranged dating the sediment cores. Now we are puzzled: We find glyphosate at high concentrations all over the core dating back to the 1960s – to a time when the substance was not even approved. Has it really seeped down that deep over the years?
We would not have expected so much glyphosate in these layers. The concentrations also don't match the sales figures for the herbicide in agriculture. It is equally astonishing that we can detect the most important degradation product of glyphosate, aminomethylphosphonic acid, AMPA for short, in much higher concentrations throughout the core. How do both substances get there?
To understand the conundrum, I turn to the State Institute for the Environment in Baden-Württemberg and ask for data on glyphosate and AMPA, ideally from the neighboring Neckar. Monthly measurements have taken place at several points in the Neckar since 2004. This is impressive for glyphosate as it can only be detected by special analysis.
When I receive the data, I see a clear seasonal trend over almost 20 years. Every year, glyphosate concentrations rise in April and May, reaching a maximum in late summer, and fall again from October. Concentrations are low in winter. But how can this keep on happening? The increase in spring could be expected from preparing the seedbeds around this time. But why do the concentrations go down again in autumn, during the main period of glyphosate use, when harvesting is over and only stubble remains in the fields? And why are they still measurable throughout the winter when herbicides are not sprayed? AMPA and glyphosate occur in almost the same ratio over the entire time, which is even more surprising for us chemists. This is hardly to be expected with the different rates of degradation of the two substances, which we had already demonstrated in our work on agricultural soils. Looking for answers, I see that the data from a measuring station in Mannheim are almost identical.
I attend a workshop in Tübingen on the contamination of arable soils. It makes me think: What if the glyphosate and its degradation product AMPA are produced anew from another substance? A logical thought for a chemist. I now recall studies that mention another source for AMPA: Aminopolyphosphonates, which are used in detergents, among other things. Could these also be producing the glyphosate? They are chemically related.
I write to my colleague Professor Stefan Haderlein from the Center for Applied Geoscience. His answer: “Perhaps far-fetched but not implausible.” Our glyphosate task force
now has two investigative approaches: We continue looking at river water data and start our laboratory work. We receive a larger data set for Germany via contacts with authorities. It quickly becomes clear that the seasonal trend, which does not at all correspond to the use of glyphosate, occurs in almost all larger and smaller rivers. We see the same picture emerging from a huge dataset from France and later examine data from Luxembourg, Italy, the Netherlands, the United Kingdom and Sweden. Everywhere we look, it’s the same story, although the land use is so different.
