Mobility risk and desorption irreversibility of glyphosate and aminomethylphosphonic acid in highly weathered tropical soils of Sri Lanka

Technical Summary: Mobility Risk and Desorption Irreversibility of Glyphosate and AMPA in Highly Weathered Tropical Soils of Sri Lanka

Problem Statement
Glyphosate (GLY) and its primary metabolite, aminomethylphosphonic acid (AMPA), are widely used globally, yet their environmental dissipation behaviors in mineral-rich, weathered tropical soils remain insufficiently studied. While extensive data exists for temperate soils dominated by minerals like illite and smectite, there is a significant knowledge gap regarding tropical soils characterized by kaolinite, gibbsite, and high iron/aluminum (Fe/Al) oxide content. Furthermore, conventional mobility risk indices (e.g., $K_{OC}$, GUS) may be inadequate for these environments because they often fail to account for sorption-desorption hysteresis—the non-coincidence of adsorption and desorption isotherms. This hysteresis influences pesticide persistence, bioavailability, and the risk of groundwater contamination. In Sri Lanka, where glyphosate is extensively used in tea and rubber plantations, understanding the specific mobility risks in wet-zone tea estate soils is critical for assessing environmental safety and potential residue transport to commercial products.

Methodology
The study characterized the sorption-desorption hysteresis of GLY and AMPA across 15 wet-zone tea estate soils in Sri Lanka.

• Sampling: Composite soil samples were collected from the A horizon (10 cm depth) of tea estates using transect sampling.
• Characterization: Soils were analyzed for physicochemical properties including pH, organic carbon (OC), clay content, and total Fe, Al, Ca, and Mg concentrations via ICP-MS.
• Experimental Design: Batch equilibrium experiments were conducted following OECD 106 guidelines at 27±2°C with a soil-to-solution ratio of 1:5.
  • Kinetics: Adsorption and desorption rates were monitored over 24 hours.
  • Isotherms: Single-point Freundlich isotherms were generated using concentrations ranging from 5 to 150 mg L⁻¹.
• Analysis: Residual glyphosate and AMPA were quantified using LC-MS/MS.
• Modeling: Data were fitted to pseudo-first-order (PFO) and pseudo-second-order (PSO) kinetic models (linear and non-linear) and Freundlich/Langmuir isotherm models.
• Risk Assessment: The study evaluated mobility using the conventional Gustafson Groundwater Ubiquity Score (GUS) and introduced two novel indices: the Hysteresis Control Index (HCI) and the normalized Mobility Risk Index ($MRIn,100$).
• Statistics: Cluster analysis, Pearson correlation, heatmaps, and Partial Least Squares Regression (PLSR) were employed to identify pedological drivers of sorption behavior.

Key Results

• Adsorption Efficiency: Adsorption was rapid, reaching equilibrium within 24 hours. In 12 of the 15 soil samples, adsorption exceeded 98.4% for GLY and 96.5% for AMPA. The remaining three samples showed lower adsorption (75.6% and 82.3%, respectively).
• Desorption Irreversibility: Desorption was notably slow and partial, failing to reach equilibrium within 24 hours in most soils. A strong reciprocal relationship was observed: soils with high adsorption exhibited the lowest desorption.
• Hysteresis Patterns: Significant hysteresis was observed.
  • Positive Hysteresis: 73% of soils for GLY and 27% for AMPA showed positive hysteresis (adsorption > desorption), indicated by Hysteresis Index (HI) values greater than unity.
  • Negative/Crossover Hysteresis: A crossover pattern was observed where the desorption isotherm exceeded the adsorption isotherm at low concentrations but converged at higher loadings, indicating concentration-dependent irreversibility.
• Kinetic and Isotherm Modeling: Adsorption followed the pseudo-second-order (PSO) kinetic model and the Freundlich isotherm, suggesting a hybrid chemisorption process on heterogeneous surfaces. Desorption kinetics were best described by non-linear PSO models.
• Pedological Drivers:
  • Positive Correlations: Distribution coefficients ($K_d$) showed strong positive correlations with Fe content, Al content, clay content, organic carbon, and moisture.
  • Negative Correlations: A sharp negative correlation was found between soil pH and $K_d$. Alkaline conditions promote surface deprotonation, increasing competition for metal oxide binding sites and reducing adsorption.
  • Surfactant Impact: Statistical analysis (ANOVA, Tukey, Dunnett) confirmed that commercial surfactants in glyphosate formulations did not significantly alter the adsorption behavior compared to the reference standard.
• Risk Classification:
  • High Risk: Alkaline, sandy-textured soils with low Fe/Al oxide content demonstrated very high mobility risk ($MRIn,100 = 100$).
  • Low Risk: Acidic, mineral-rich, high-clay tropical soils effectively immobilized both compounds ($MRIn,100 < 0.02$), resulting in lower bioavailability.

Key Contributions

1. Novel Risk Indices: The study introduced the Hysteresis Control Index (HCI) and the normalized Mobility Risk Index ($MRIn,100$). These indices integrate hysteresis behavior, desorption potential, and mineralogical buffering capacity into a unified framework, addressing limitations in conventional mobility indices for tropical soils.
2. Tropical Soil Specificity: The research provides critical data on GLY and AMPA behavior in highly weathered tropical soils (kaolinite/oxide-dominated), contrasting with the existing literature focused on temperate soils.
3. Mechanistic Insight: The study confirms that inner-sphere complexation at Fe and Al surfaces is the primary mechanism driving the observed hysteresis and irreversibility.

Significance and Claims
The paper claims that the environmental fate of glyphosate in tropical agricultural systems cannot be accurately predicted using models derived from temperate soils. The authors assert that the specific mineralogy (high Fe/Al oxides) and acidity of Sri Lankan tea estate soils generally lead to strong immobilization of glyphosate and AMPA, thereby reducing mobility risk and bioavailability in these specific contexts. However, the study highlights that alkaline, sandy soils within the same region pose a significant mobility risk. By characterizing the desorption irreversibility and introducing the $MRIn,100$ index, the study offers a more nuanced tool for evaluating site-specific environmental risks, suggesting that in highly weathered tropical soils, the risk of groundwater contamination is generally low due to strong chemisorption, provided the soil remains acidic and mineral-rich.