Content area
Abstract
For many decades now, the use of pesticides is considered the most viable solution in conventiona 1 agriculture to in crea se food production. As a result, great amounts of pesticides are applied every year in agricultural fields. Amongst ail pesticides appl ied, glyphosate-based herbicides lead the market worldwide for more than two decades, with more than 825 000 tons sold globally in 2014, mostly as a result of the introduction of glyphosate-resistant (GR) crops in 1996 by Mon santo company. The huge and still increasing use of glyphosate-based herbicides is leading to its frequent occurrence in the different environmental compartments, as weil as in feed and food. Despite being sold by pesticide companies as an environmentally friendly product and of no concern to the environment and human health, its intensive use worldwide is resulting in high social concerns about its safety, especially a ft er being classified as possible carcinogenic to huma ns in 2015 by the International Agency for Research on Cancer (IARC).
Loess soils are amongst the most productive and fertile soils a nd, consequently, are intensively used for agriculture. They representa round 10% of the Earth's land surface, with major deposits occurring in China, USA, Argentina and Europe. Of particular interest are the loess deposits from Argentina and USA, which are intensively used for the production of GR crops and, consequently, are heavily subject to the application of glyphosate-based herbicides every year. Because loess soils are easily erodible by water and wind, the advent of GR crops promoted a change in agricultural management practices from tillage to nontillage systems. Consequently, weeds are controlled chemically rather than mechanically before and after sowing. Such an agricultural management change helped reducing water and wind erosion problems faced by farmers, but also resulted in the excessive use of glyphosate-based herbicides.
Despite being the most used pesticide worldwide, the environ mental fate of glyphosate and its main metabolite aminomethylphosphonic acid (AMPA) is still not weil understood. The decay behavior and persistence of these compounds is very variable across different soils and climatic regions, and AMPA is poorly studied sofa r. Although these compounds strongly adsorb to soi l particles, studies on their off-site transport with wind erosion were nonexistent until very recently (2018), and most studies on their transport wit h water erosion disregarded the particle-bound phase.
ln this context, this PhD thesis aims at better understanding the environmental fate of glyphosate and AMPA in the loess soil environment. Special focus is given to: 1) the decay kinetics of glyphosate and the formation and decay kinetics of AMPA; 2) t he off-site transport of glyphosate and AMPA associated to the particle-bound phase, as a result of wind and water erosion. These processes were studied under laboratory (cha pters 2 to 4) and field (chapter 5) conditions. The field study was performed in agribusiness fields of the loess Pampas of Argentina.
Chapter 2 investi ga tes the decay kinetics of glyphosate and the formation and decay kinetics of AMPA in loess soil under laboratory conditions. The combined effect of temperature (5 and 3o°c), soil moisture (20% WHC, 60% WHC, saturation) and light/darkness under biotic and abiotic conditions was addressed. Glyphosate and AMPA decay was mostly a microbiological process; abiotic (photodegradation) processes were negligible. Under biotic conditions, glyphosate followed first-order (SFO) kinetics at 5°c, but bi-phasic (FOMC) at 3o°c. Efforts are needed to apply the correct decay kinetics for glyphosate in environ mental fate models to avoid the underestimation of glyphosate's DT90. AMPA followed al ways SFO kinetics. A higher formation of AMPA was observed at 5°c (54-100%) than at 3o°c (37-48%). Persistence of glyphosate (DT 50: 1.5-54 days; DT 90 : 8-280 days) and AMPA (at 3o°c - DT sa: 26-45 days; DT go: 88-148 days; end points not achieved at 5°C) was very variable, depending on temperature and soil moisture conditions. From the DTso and DT90 values, it can be concluded that AMPA persisted longer than glyphosate for ali conditions. Glyphosate and AMPA decayed fastest under warm and moist soil conditions and slowest under cold and dry soil conditions. Therefore, repeated glyphosate applications under dry and cold conditions, as they happen in fallows or before sowing winter crops, may lead to glyphosate and AMPA accumulation in soil and to on-site soil pollution.
Chapter 3 investigates the role of wind erosion on the atmospheric transport of glyphosate and AMPA. This experiment was performed with loess soil in a wind-tunnel. The distribution of glyphosate and AMPA in different size fractions of the wind-eroded sediment and the relationship with clay, silt and organic matter (OM) were investigated. An estimation for the potential off-site transport of these compounds over large distances from the pollution source was also performed. The results indicated that glyphosate and AMPA contents are highest in the finest particle fractions of the wind-eroded sediment (<18 ~m ), decreasing significantly with increasing parti de size. The results a Iso showed that glyphosate and AMPA cor relate positively with the clay, OM and silt contents of the wind-eroded sediment (clay> OM > silt). The low soil moisture content of the sediment resulted in a very low decay of glyphosate and formation/decay of AMPA over time. We concluded that the risk of glyphosate and AMPA to be transported to the atmosphere over large distances is very high, due to the low decay of glyphosate in dry sediment and the highest glyphosate and AMPA contents in PM10 . Consequently, because PM10 or smaller particles are easily inha led, the risk of hum an exposure to contaminated dust is also high. Therefore, more attention needs to be pa id to this route of exposure in risk assessment studies.
Chapter 4 investigates the influence of small-scale sediment transport and of different soil surface micro-topographies on the off-site transport of glyphosate and AMPA with water erosion. The redistribution, transport and deposition onto the loess soil surface was also assessed. Smooth soil surfaces and soil surfaces resembling seeding lines on the contour were investigated, using soil flumes (1 x 0.5 m) with a 5% slope. Glyphosate (178 mg m-2 ) was sprayed on the upper 0.2 m of the flumes. Four 15-min rainfall events with 30-min interval in between and a total rainfall intensity of 30 mm h-1 were applied. When taking into account the off-site deposition to downslope soil surface areas, glyphosate and AMPA parti de-bou nd off-site transport was higher (9.4-17.8%) than the water-dissolved one (0.5- 2.8%). Wh en only considering the off-site transport to the out let of the fiume (i.e. exclu ding the downslope terrestrial deposition), glyphosate a nd AMPA water-dissolved off-site transport was higher (0.5-2.8%) than the particle-bound one (0.2-1.7%). The results have shown that the presence of micro-topographie disturbances ("seeding lines") reduces the runoff and erosion rates into the outlet of the flumes and, consequently, the off-site transport of glyphosate and AMPA to aquatic environments. However, their off-site deposition to downslope terrestrial environments is enhanced. The results also indicated that the off-site downslope transport of these compounds reduced with increasing distance from the source, for both types of soil surfaces. We concluded that the particle-bound transport of glyphosate and AMPA is as or more important than their wat er-dissolved transport. The pollution risk of off-target terrestrial and aquatic environments can be considerable. Buffer zones, promoted by soil (micro-)topographic barriers, within the treated areas as weil as at the downslope edges of treated fields are recommended to reduce the off-target transport of these compounds into terrestrial a nd aquatic environments.
Chapter 5 investigates the dynamics of glyphosate and AMPA in the soil surface of fields cultivated with GR crops (soybean and ma ize) in the loess Pampas of Côrdoba province, in Argentina. The cultivation a reas studied have a long-term history of repeated applications of glyphosate-based herbicides. Focus was given to the role of decay kinetic processes and to the off-site transport risk. Glyphosate decay kinetics and AMPA formation/ decay kinetics were investigated after a single application with the glyphosate commercial formulation Roundup Ultra Max© (soybean a rea: 1.0 kg a.e. ha-1 ; ma ize a rea: 0.81 kg a.e. ha-1 ). Moreover, glyphosate and AMPA concentrations were quantified in runoff water and eroded sediment after 2 erosive rainfall events. Under both cultivation areas, background resid ues in soil before application were 0.27-0.42 mg kg-1 for glyphosate and 1.3-1.7 mg kg-1 for AMPA. Glyphosate and AMPA contents were higher in the top 0-1 cm soil surface layer and were significantly lower in the subsequent soil surface layer (1-2 cm). Therefore, leaching ofthese compounds into deeper soil layers and groundwater seems to be limited. Glyphosate followed SFO decay kinetics in the soybean cultivation area, but bi-phasic (Hockey-StickHS) in the ma ize cultivation a rea. The latter was due to an abrupt change in the so il moisture conditions from dry to moist after a significant rainfall event, which resulted in an increase of the glyphosate decay rate from 0.06 to 0.37 days·1 . AMPA followed SFO decay ki netics, but its formation (5-24%) and decay rates (0.01 day5'1) were low. Glyphosate DT50 in soil varied between 6 (soybean) and 11 (maize) days, whereas its DT90 varied between 15 (maize) and 20 (soybean) days. AMPA persisted much longer in soil than glyphosate, with its DT50 varying between 55 (soybean) and 71 (maize) days and its DT90 between 182 (soybean) and 236 (maize) days. The low decay rates of AMPA in both cultivation areas indicate that it persists for long periods in soil and suggest that it may be accumulat ing in soil. During erosive rainfall events, higher glyphosate and AMPA contents were observed in eroded sediment than in runoff water. Glyphosate and AMPA contents were also 1-18 (glyphosate) and 0.8-8 (AMPA) times higher in water-eroded sediment than in soil. The results indicated that the risk of glyphosate and AMPA off-site transport with water erosion can be considera ble, even 2 months a ft er glyphosate application, particularly associated to sheet erosion. The tendency of glyphosate and AMPA to rem ain in the top soil surface layer a Iso suggests that wind erosion may play an important role on the off-site transport oft hese compounds, particularly when soil cover with stubbles is low, before sowing and after sowing until the crop covers sufficiently most of the soil surface.
Overall, this PhD thesis improved the knowledge and furthered the science of the environmental fate of glyphosate and AMPA in the loess soil environment. This is particularly true for AMPA, which until now has received little attention. This PhD t hesis is a major contribution to the understanding of glyphosate decay kinetics and particularly of AMPA formation and decay kinetics. Glyphosate decay kinetic studies presented in this PhD thesis indicate the need to incorporate bi-phasic kinetic models in environmental fate modelling studies in order to more reliably determine glyphosate persistence in soil. More and longer studies a re still required und er la boratory and field conditions to bet ter understand the formation and decay behavior of AMPA. Another major contribut ion of th is PhD research regards the off-site transport of glyphosate and AMPA with wi nd erosion, a transport pathway that has been disregarded so far. Field studies focusing on t he off-s ite and long-rangetra nsport of glyphosate and AMPA with wind erosion are needed in different types of soil and in fields under different tillage systems. Glyphosate and AMPA risk assessments need to consider the exposure to contaminated dust. This PhD t hesis also contributed for a better understa nding of glyphosate and AMPA off-site transport with water erosion, not only to aquatic environments but a Iso to terrestrial environments. Chapter 6 concludes with recommendations to pesticide regulatory and supervision authorities and farmers for a more sustainable agricultural management, suggest ing severa! alternatives to significa ntly redu ce the use of glyphosate-based herbicides and to promote a shift for a herbicide-free agricultural system.
