A comprehensive experimental assessment of glyphosate ecological impacts in riparian forest restoration.

Competition with invasive grasses is one of the most important drivers of tree planting failures, especially in tropical forests. A widely disseminated weeding approach has been glyphosate spraying, the most used herbicide globally in forestry and ecosystem restoration. However, glyphosate use in restoration is highly controversial and requires further studies to elucidate its effects on restoration processes and the environment. We evaluated the use of glyphosate in riparian forest restoration and its impacts on tree planting costs, weed control efficiency, planted seedling performance, herbaceous and woody species regeneration, soil bacteria, and environmental contamination, using mowing treatments as a reference and based on a controlled experiment established in the Brazilian Atlantic Forest. Glyphosate spraying reduced by one-half and one-third the accumulated aboveground biomass of, respectively, weeds in general and of the invasive grass Urochloa decumbens compared to mowing treatments, and it reduced the cost by half. The performance of planted tree seedlings was markedly favored by glyphosate spraying compared to mowing treatments, as expressed by improved seedling height (~twice higher), crown area (~5× higher), and basal area (~5× higher); the regeneration of both native woody and ruderal herbaceous plants were also enhanced. Neither glyphosate nor its metabolite Aminomethylphosphonic acid (AMPA) residues were detected in either water runoff or soil samples, but they were found at relatively high concentrations in the runoff sediments (from 1.32 to 24.75 mg/kg for glyphosate and from 1.75 to 76.13 mg/kg for AMPA). Soil bacteria communities differed before and after glyphosate spraying in comparison to mowing plots (without glyphosate). Glyphosate spraying was far more cost effective than mowing for controlling U. decumbens and greatly improved the performance of planted tree seedlings and natural regeneration, while not leaving residues in soil and water. However, the changes in the structure of bacterial communities and high concentration of glyphosate and AMPA residues in runoff sediments highlight the need for caution when using this herbicide in riparian buffers. We present alternatives for reducing glyphosate use and minimizing its risks in tree planting initiatives.

Effect of Soil Water Content on the Distribution of Diuron into Organomineral Aggregates of Highly Weathered Tropical Soils.

We evaluated the effects of soil water content on the retention of diuron and its residual distribution into organomineral aggregates in four Brazilian oxisols. (14)C-Diuron was incubated for days at 25, 50, and 75% of maximum water-holding capacity for each soil. After 42 days, the physical fractionation method was used to obtain >150, 53-150, 20-53, 2-20, and <2 µm aggregate sizes. Diuron retention increased with increasing soil water content for all soils. At lower soil water content, diuron's retention was higher in the sandier soil. It was mostly retained in the fine (<20 µm) aggregates of sandier soil, and for clayed soils, retention was higher in the coarse aggregates (>53 µm). The sorption coefficients (Kd and Koc) generated by batch studies should be carefully used because they do not provide information about aggregation and diffusion effects on pesticides soil sorption.

Organomineral Interactions and Herbicide Sorption in Brazilian Tropical and Subtropical Oxisols under No-Tillage.

We evaluated the effects of the soil organic matter (SOM) composition, distribution between soil aggregates size, and their interactions with the mineral phase on herbicide sorption (alachlor, bentazon, and imazethapyr) in tropical and subtropical Oxisols under no-till systems (NT). Using soil physical fractionation approach, sorption experiments were performed on whole soils and their aggregates. SOM chemistry was assessed by CP/MAS (13)C NMR. The lower sorption observed in tropical soils was attributed to the greater blockage of SOM sorption sites than in subtropical soils. When these sites were exposed upon physical fractionation, sorption of the three herbicides in tropical soils increased, especially for imazethapyr. High amounts of poorly crystallized sesquioxides in these soils may have contributed to masking of sorption sites, indicating that organomineral interactions may lead to blockage of sorption sites on SOM in tropical soils.

Effects of soil attributes and straw accumulation on the sorption of hexazinone and tebuthiuron in tropical soils cultivated with sugarcane.

Brazil is the largest sugarcane producer in the world in which hexazinone (3-cyclohexyl-6-dimethylamino-1-methyl-1,3,5-triazine-2,4-dione) and tebuthiuron (1-(5-tert-butyl-1,3,4-thiadiazol-2-yl)-1,3-dimethylurea) are heavily used. Sugarcane harvesting is changing from the manual system with previous straw burning to the mechanized system without straw burning. The lack of burning results in soil organic carbon accumulation mainly in clayey soils, which should affect herbicides availability and fate. Therefore, we evaluated sorption of these herbicides in soil samples with and without straw burning. Both herbicides presented low apparent sorption coefficients (mean K(d,app)= 0.6 and 2.4 L kg(-1) for hexazinone and tebuthiuron, respectively), suggesting that they may leach to groundwater. Moreover, their sorption correlated primarily with soil organic carbon (SOC), but iron oxide contents extracted with ammonium oxalate (Fe2O3(AOX)) also affected it (K(d,app) = -0.228 + 0.0397 SOC + 0.117 Fe2O3(AOX) for hexazinone and K(d,app) = -1.407 + 0.201 SOC + 0.348 Fe2O3(AOX) for tebuthiuron). Soil organic carbon accumulation due to straw maintenance in the field positively affected sorption of both herbicides, but its effects were not enough to classify them as "non-leachers."

Characterization of nicosulfuron availability in aged soils.

Sorption-desorption interactions of pesticides with soil determine their availability for transport, plant uptake, and microbial degradation. These interactions are affected by the physical-chemical properties of the pesticide and soil, and for some pesticides, their residence time in the soil. This research evaluated changes in sorption/availability of nicosulfuron (2-[[[[(4,6-dimethoxy-2-pyrimidinyl]amino]carbonyl]amino]sulfonyl]-N,N-dimethyl-3-pyridinecarboxamide) herbicide with aging in different soils, using a radiolabeled ((14)C) tracer. Aging significantly increased sorption. For instance, after the 41-day incubation, calculated K d,app increased by a factor of 2 to 3 in Mollisols from the Midwestern United States and by a factor of 5 to 9 in Oxisols from Brazil and Hawaii, as compared to freshly treated soils. In view of this outcome, potential transport of nicosulfuron would be overpredicted if freshly treated soil K d values were used to predict transport. The fact that the nicosulfuron solution concentration decreased faster than the soil concentration with time suggested that the increase in sorption was because the rate of degradation in solution and on labile sites was faster than the rate of desorption of the neutral species from the soil particles. It may have also been due to nicosulfuron anion diffusion to less accessible sites with time, leaving the more strongly bound neutral molecules for the sorption characterization. Regardless of the mechanism, these results are further evidence that increases in sorption during pesticide aging should be taken into account during the characterization of the sorption process for mathematical models of pesticide degradation and transport.

Sorption-desorption of atrazine and diuron in soils from southern Brazil.

The objective of this study was to investigate the behavior of sorption and desorption of the herbicides atrazine (6-chloro-N2-ethyl-N4-isopropyl-1,3,5-triazine-2,4-diamine) and diuron [3-(3,4-dichlorophenyl)-1,1-dimethyleurea] in soil samples from a typical lithosequence located in the municipality of Mamborê (PR), southern Brazil. Five concentrations of 14C-atrazine and 14C-diuron were used for both herbicides (0.48, 0.96, 1.92, 3.84, and 7.69 mg L(-1)). Sorption of both herbicides correlated positively with the organic carbon and clay content of the soil samples. Sorption isotherms were well described by the Freundlich model. The slope values of the isotherm (N) ranged from 0.84 to 0.90 (atrazine) and from 0.75 to 0.79 (diuron) for the lithosequence samples. Sorption of diuron was high regardless of the soil texture or the concentration added. The desorption isotherms for atrazine and diuron showed good fit to the Freundlich equation (R2 >or= 0,87). Atrazine slope values for the desorption isotherms were similar for the different concentrations and were much lower than those observed for the sorption isotherms. Significant hysteresis was observed in the herbicide desorption. When the two herbicides were compared, it was found that diuron (N = 0.06-0.22) presented more pronounced hysteresis than atrazine. The results showed that, quantitatively, a greater atrazine fraction applied to these soils remains available to be leached in the soil profile, as compared to diuron.

Influence of soil aging on sorption and bioavailability of simazine.

Characterization of pesticide bioavailability, particularly in aged soils, is of continued interest because this information is necessary for environmental risk assessment. However, pesticide bioavailability in aged soils has been characterized by a variety of methods with limited success, due in part to methodological limitations. The objective of this study was to use solvent extraction methods to correlate simazine residue bioavailability in aged soils to simazine mineralization using a simazine-mineralizing bacterium. Soils from Brazil, Hawaii, and the midwestern United States were treated with UL-ring-labeled [14C]simazine and incubated for up to 8 weeks. At the end of each incubation period, soils were either incubated further, extracted with 0.01 M CaCl2, or extracted with aqueous methanol (80:20 v/v methanol/water). In a parallel experiment, after each incubation period, soils were inoculated with the bacterium Pseudomonas sp. strain ADP, which is capable of rapidly mineralizing simazine, and 14CO2 was determined. The inoculated soil samples were then extracted with 0.01 N CaCl2 and with aqueous methanol. This allowed for the evaluation of the bioavailability of aged simazine residues, without the contribution of simazine desorption from soil. Results of these studies indicated that simazine sorption to soil increased with aging and that amounts of simazine in aged soils extracted by 0.01 M CaCl2 and aqueous methanol were highly correlated to amounts of simazine mineralized by Pseudomonas sp. strain ADP. Consequently, 0.01 M CaCl2/methanol-extractable simazine in aged soils can be used to estimate bioavailable residues. This technique may be useful in determining the bioavailability of other s-triazine compounds in soils.

Soil pH on mobility of imazaquin in oxisols with positive balance of charges.

The influence of soil pH on the leaching potential of the ionizable herbicide imazaquin was assessed on the profile of two highly weathered soils having a net positive charge in the B horizon, in contrast to a soil having a net negative charge in the whole profile, using packed soil column experiments. Imazaquin leached to a large extent and faster at Kd values lower than 1.0 L kg(-1), a much more lenient limit than usually proposed for pesticides in the literature (Kd < 5.0 L kg(-1)). The amount of imazaquin leached increased with soil pH. As the soil pH increased, the percentage of imazaquin in the anionic forms, the negative surface potential of the soils, as well as imazaquin water solubility also increased, thus reducing sorption because of repulsive electrostatic forces (hydrophilic interactions). For all surface samples (0-0.2 m), imazaquin did not leach at soil pH values lower than pKa (3.8) and more than 80% of the applied amount was leached at pH values higher than 5.5. For subsurface samples from the acric soils, imazaquin only began to leach at soil pH values > zero point of salt effects (ZPSE > 5.7). In conclusion, the use of surface K(oc) values to predict the amount of imazaquin leached within soil profiles having a positive balance of charges may greatly overestimate its actual leaching potential.

Sorption kinetics of atrazine and diuron in soils from southern Brazil.

Sorption kinetics of atrazine and diuron was evaluated in soil samples from a typical landscape in Paraná. Samples were collected (0-20 cm) in a no-tillage area from Mamborê, PR, which has been cultivated under a crop rotation for the last six years. Six sampling points of the slope were selected to represent a wide range of soil chemical and physical properties found in this area. Radiolabeled tracers (14C-atrazine and 14C-diuron) were used and the radioactivity was detected by liquid scintillation counting (LSC). Sorption was accomplished for increasing equilibration periods (0.5, 1.5, 3, 6, 12, 24, and 48 h). Kinetics data fitted adequately well to Elovich equation, providing evidences that soil reaction occurs in two distinct stages: a fast, initial one followed by a slower one. During the fast phase, 34-42 and 71-79% of total atrazine and diuron applied were sorbed to soil samples. No important differences were found among combinations of soil and herbicide sorption during the slow phase. The unrealistic conditions under batch experiments should be overestimating sorption in the fast phase and underestimating diffusion in the slow phase. Sorption of both herbicides was positively correlated to organic carbon and clay contents of soils, but atrazine was much less sorbed than diuron, showing its higher potential to contaminate groundwater, specially in sandy, low organic carbon soils.

Sorption of imazaquin in soils with positive balance of charges.

The herbicide imazaquin has both an acid and a basic ionizable groups, and its sorption depends upon the pH, the electric potential (psi0), and the oxide and the organic carbon (OC) contents of the soil. Sorption and extraction experiments using 14C-imazaquin were performed in surface and subsurface samples of two acric oxisols (an anionic "rhodic" acrudox and an anionic "xanthic" acrudox) and one non-acric alfisol (a rhodic kandiudalf), treated at four different pH values. Imazaquin showed low to moderate sorption to the soils. Sorption decreased and aqueous extraction increased as pH increased. Up to pH 5.8, sorption was higher in subsurface than in surface layers of the acric soils, due to the positive balance of charges resulted from the high Fe and Al oxide and the low OC contents. It favored electrostatic interactions with anionic molecules of imazaquin. For the subsurface samples of these highly weathered soils, where psi0 was positive and OC was low, it was not possible to predict sorption just by considering imazaquin speciation and its hydrophobic partition to the organic domains of the soil. Moreover, if Koc measured for thesurface samples were assumed to represent the whole profile in predictive models for leaching potential, then it would result in underestimation of sorption potential in subsurface, and consequently result in overestimation of the leaching potential.

Sorption interactions between imazaquin and a humic acid extracted from a typical Brazilian oxisol.

Soil sorption of most hydrophobic organic compounds (e.g., nonpolar pesticides) is directly related to soil organic matter (SOM) content. Humic substances are the major SOM components, containing carboxylic, phenolic, amine, quinone, and other functional groups, and specific structural configurations. In this paper, sorption interactions between imazaquin (2-[4,5-dydro-4-methyl-4-(1-methylethyl)-5-oxo-1H- imidazol-2-yl]-3-quinoline-carboxylic acid) herbicide (IM) and a humic acid (HA) extracted from a typical Brazilian Oxisol were studied with electron paramagnetic resonance (EPR) and Fourier-transform infrared (FTIR) spectroscopic techniques. A polarographic technique was used to quantify sorption. The IM amount sorbed on the HA was much higher than that on the whole soil within the pH range studied, emphasizing the prominent role played by SOM on IM sorption. Moreover, IM sorption increased as the soil-solution pH decreased. This enhancement in sorption was attributed to the hydrophobic affinity of the herbicide by the HA and to the electrostatic interaction between the protonated quinoline group of IM and the negative sites of the HA. Hydrophobic regions in the HA's interior at low pH (< 5.0) were recently demonstrated by an EPR detectable spin-label molecule. The FTIR and EPR spectroscopy and polarography data indicated weak interaction between IM and the soil and its HA, involving hydrogen bonding, proton transfer, and cation exchange (at low pH), and mainly hydrophobic interactions. However, no strong reaction mechanism, such as charge transfer, was involved. In addition, this research suggested that soil amendment with organic material might increase magnitude of IM sorption, consequently avoiding leaching and carryover problems usually found for mobile and persistent herbicides such as imazaquin.