Pesticide use and off-site risk assessment: a case study of glyphosate fate in Chinese Loess soil
Abstract: Repeated applications of pesticide may contaminate the soil and water, threatening their quality within the environmental system and beyond it through water erosion related processes and leaching. Taking into account the situation of intense farming, pesticide use and erosion in China, and in the loess area in particular, this study aims to further the science of pesticide risk assessment by increasing the understanding of the fate of glyphosate and its main metabolite AMPA while also assessing their risks related to soil erosion in Chinese loess soil. Four main objectives are conducted:
1) Assessment of farmers’ knowledge and awareness of pesticide use in the Wei River catchment
In this study, we presented the information for the knowledge and awareness of pesticide risks via face-to-face interview among farmers (209) and retailers (20) in two rural regions (Qianyang County (S1) and Chencang County (S2)) of the Wei River catchment in China where the modes of farming and the state of erosion are very different. The results showed that farmers learned about the use and risk related to of pesticide use mainly by oral communication (p<0.01), rather than by introduction labelled on pesticide package bag/bottle and media tools (TV/newspaper/Internet/books). Protective measures were inadequate and washing hands (>70%) was the most common mode of personal hygiene after spraying, comparing to wearing masks, showering, and changing clothes. Regarding to dealing with the packages or containers of pesticide, farmers dumped them directly onto the land or into water, and over 85% of farmers claimed to use illegal pesticides. Compared to farmers, pesticide retailers were well-informed and highly conscious of their responsibility for the safe use of pesticides. Educational programmes targeted to age groups, proper disposal of pesticide waste, and sufficient supervision from authorities should consequently be considered for improving the levels of knowledge and awareness of the dangers of pesticides to human health and environmental pollution in the Wei River catchment, China.
2) Investigation of off-site transport of glyphosate and AMPA by runoff and erosion in Chinese loess soil
In this study, we assessed the short-term transport (1h duration, 1 mm min-1 rainfall) of glyphosate and its metabolite aminomethylphosphonic acid (AMPA) related to erosion and runoff on bare loess soil with different slopes (10° and 20°)and application rates (3.6 kg ha-1 and 7.2 kg ha-1) in hydraulic flumes. Significant power and exponent function described a significantly negative relation between rainfall duration and the content of glyphosate and AMPA (p<0.01) in runoff and suspended load, respectively. The transport rate of glyphosate (including AMPA) by runoff and suspended load was approximately 4% and 10% of the applied amount, respectively indicating that it mainly underlies particulate transport. The glyphosate and AMPA content in the flume soil at the end of the experiment (after 1h) decreased significantly with depth (p<0.05), and 72, 2, and 3% of the applied glyphosate (including AMPA) remained in the 0-2, 2-5, and 5-10 cm soil layers, respectively. Thus, the risk of leaching was low. In view of loss of 14 % of the initially applied glyphosate after 1 h, “buffer zone”, referred as the protection area for ecosystem discharge between farming land and public rivers, is strongly recommended, especially in regions where rain-induced soil erosion is common.
3) Investigation of the on-site fate of glyphosate and AMPA formation in Chinese loess soil
In this study, we observed glyphosate decay, erosion and runoff related transport of glyphosate and AMPA at different slopes (10° and 20°) and application rates (3.6 kg ha-1 and 7.2 kg ha-1) under field condition during 35 d, September-October in 2012 (16.8±2.1°C with 35 mm rainfall in total) and July-August in 2013 (27 ± 2.3°C with 74 mm rainfall in total). The initial glyphosate decayed rapidly (half-life of 3.5 d) in the upper 2 cm of soil following a first-order rate of decay. AMPA content in the 0-2 cm soil layer correspondingly peaked 3 d after glyphosate application and then gradually decreased. The residues of glyphosate and AMPA decreased significantly with soil depth (p<0.05) independently of the slope inclination and application rate. About 0.36% of the glyphosate initially applied was transported off-site plots after one erosive rain 2 days after the application. Glyphosate and AMPA concentration in runoff were low while the contents in the sediment was much higher than in the upper 2 cm of the soil. This suggested that even though the transported glyphosate is limited, the off-site risk of glyphosate and AMPA is high. Consequently, a realistic erosion-pesticide model should be developed that can simulate the particulate-facilitated transport of glyphosate and its off-site risks involving decay processes.
4) Off-site risk assessment of particle-facilitated pesticide transport related to erosion
In this study, we have developed a parsimonious integrative model of pesticide displacement by runoff and erosion that explicitly accounts for water infiltration, erosion, runoff, and pesticide transport and degradation in soil. The conceptual framework was based on broadly accepted assumptions such as the convection-dispersion equation and lognormal distributions of soil properties associated with transport, sorption, degradation, and erosion. To illustrate the concept, a few assumptions are made with regard to runoff in relatively flat agricultural fields: dispersion is ignored and erosion is modelled by a functional relationship. A sensitivity analysis indicated that the total mass of pesticide associated with soil eroded by water scouring increased with slope, rain intensity, and water field capacity of the soil. The mass of transported pesticide decreased as the micro-topography of the soil surface became more distinct. The timing of pesticide spraying and rate of degradation before erosion negatively affected the total amount of transported pesticide. The mechanisms involved in pesticide displacement, such as runoff, infiltration, soil erosion, and pesticide transport and decay in the topsoil, were all explicitly accounted for, so the mathematical complexity of their description can be high, depending on the situation.
Overall, the outcome of this study indicates that it is urgent to improve farmers knowledge and awareness on pesticide use in order to reduce pesticide risks before they are applied in environmental system. The rapid decay of glyphosate to AMPA in loess soil suggests that on-site risks of glyphosate is mainly related to AMPA, especially under warmer climatic conditions. The off-site transport of glyphosate in our study was about 14% of the applied dose after 1 h of rainfall indicating that the off-site risks of glyphosate should be taken into account in regions highly prone to soil erosion. The concept of particulate transport elaborated in objective 4 should be incorporated in existing erosion models, such as LISEM, to predict off-site effects of glyphosate applications.