Controlling factors for glyphosate sorption: and the prediction of sorption coefficients for cultivated soils of Southern Greenland and elsewhere

Christensen, Laura Thode

4. term

Water and Environmental Engineering , Master

2017

84

Glyphosat er det mest anvendte ikke-selektive herbicid i verden og er den aktive ingrediens i bl.a. Roundup. Anvendelsen var i stigning i 2017, og der er bekymring for miljøsikkerheden omkring både glyphosat, dets nedbrydningsprodukt AMPA, de salte (kationer) det udsprøjtes som, samt tensider og andre hjælpestoffer. Man forventer normalt, at glyphosat nedbrydes før det kan nå grundvandet, men under det danske Pesticide Leaching Assessment Programme er stoffet fundet i grundvand over grænseværdien. Glyphosat er vandopløseligt og binder til jordoverflader på komplekse måder, som ikke er fuldt forstået. Denne afhandling undersøger, hvor stærkt glyphosat binder til jord (sorption), og hvilke jordegenskaber der styrer bindingens styrke. Vi diskuterer og forbedrer modeller, der forudsiger den lineære sorptionskoefficient ved en koncentration på 0,23 mg/L. Prøvesættet er omhyggeligt udvalgt: danske marker med tydelige gradienter i lerindhold, pH og organisk kulstof er med, og sammen med grønlandske lokaliteter opnås en gradient i jordens “alder” fra nyligt opdyrkede områder over jorde dyrket i få generationer til Danmarks århundredgamle landbrugsjorde. Der indgår jorde af forskellig geologisk oprindelse, fra danske glaciale aflejringer til grønlandske krystallinske bjergarter; forsøg på at inddrage vulkanske jorde mislykkedes for det meste på grund af manglende data. Jordparametre stammer fra Aarhus University, Foulums vidensbase eller laboratoriemålinger. Sorptionskoefficienter måles i batchforsøg med C14-mærket glyphosat og væskescintillationstælling, og der udføres lineære regressionsanalyser i Matlab. pH, indhold af ler og fin silt, oxalat-ekstraherbart fosfor og især oxalat-ekstraherbart jern er vigtige for at forudsige glyphosats sorption. Elektrisk ledningsevne, indhold af organisk kulstof og oxalat-ekstraherbart aluminium har kun svag eller ingen effekt. Jordens dyrkningshistorie og geologiske oprindelse påvirker relationerne mellem jordparametre og sorption. Enkle pedotransferfunktioner (forudsigelsesformler) baseret på oxalat-ekstraherbart jern og enten pH eller oxalat-ekstraherbart fosfor opnår forklaringsgrader (R2) over 0,6 på regionalt niveau eller højere. Resultaterne kan forbedre vurderingen af, hvor glyphosat sandsynligvis bindes stærkt i jorden, og hvor der kan være større risiko for udvaskning til grundvand.

Glyphosate is the world’s most used non-selective herbicide and the active ingredient in products such as Roundup. Its use was still increasing in 2017, and there are environmental concerns about glyphosate, its breakdown product AMPA, the salts (cations) it is applied with, and surfactants and other adjuvants. Glyphosate is generally expected to degrade before it reaches groundwater, yet the Danish Pesticide Leaching Assessment Programme has detected it in groundwater above the limit value. Glyphosate is water-soluble and attaches to soil surfaces in complex ways that are not fully understood. This thesis examines how strongly glyphosate binds to soils (sorption) and which soil properties control that binding. We discuss and improve models that predict the linear sorption coefficient at a concentration of 0.23 mg/L. The sample set was carefully chosen: Danish fields with strong gradients in clay content, pH, and organic carbon were included, and, together with Greenlandic sites, this provided a gradient in soil “age” from newly farmed areas and fields cultivated for only a few generations to Denmark’s centuries-old farmlands. Soils of different geological origin were included, from Denmark’s glacial deposits to Greenland’s crystalline rocks; attempts to include volcanic soils were largely unsuccessful due to lack of data. Soil parameters came from Aarhus University, Foulum’s knowledge base or laboratory measurements. Sorption coefficients were measured in batch experiments using C14-labeled glyphosate and liquid scintillation counting, and linear regression analyses were run in Matlab. pH, clay content, fine silt, oxalate-extractable phosphorus, and especially oxalate-extractable iron were important predictors of glyphosate sorption. Electrical conductivity, organic carbon content, and oxalate-extractable aluminum had weak or no influence. The soils’ management history and origin influenced the relationships between soil properties and sorption. Simple pedotransfer functions (predictive equations) based on oxalate-extractable iron and either pH or oxalate-extractable phosphorus achieved coefficients of determination (R2) above 0.6 at regional scale or higher. These findings help improve predictions of where glyphosate is likely to bind strongly in soils and where the risk of leaching to groundwater may be higher.

[This abstract was generated with the help of AI]

pH ; clay ; iron ; phosphate ; regression ; pedotransfer ; agriculture ; pesticide ; leaching ; retardation

Download
View record in AAU Student Projects