| Abstract: | The decomposition behaviour of glyphosate in four Victorian soils was investigated at two temperatures using non-steady-state compartmental analysis. At 25°C, glyphosate degradation was shown numerically to be derived from two different sources where the rate of release from each source behaved in accordance with first-order kinetics. Over the first 40 day period for each of the soils, glyphosate was derived simultaneously from the labile and non-labile phase, whilst after the first 40 days, glyphosate was derived solely from the non-labile phase. At this temperature, the amount of glyphosate partitioned into the labile phase ranged from 24·1 to 34·5%, whilst the amount partitioned into the sorbed, non-labile phase ranged from 67·2 to 74·9%. The half-lives for glyphosate within each phase was calculated and ranged from six to nine days for the labile phase to 222–835 days for the non-labile phase. Glyphosate appeared to be more strongly held in the acidic Rutherglen soil than in the alkaline soils studied, and this was thought to be related to the substantially lower pH and higher Fe content of the acidic soil. At 10°C, glyphosate was shown numerically to be derived from two different sources for two of the soils. However, for the two remaining soils, glyphosate appeared to be derived either from a single phase or from two phases at either the same rate or at differential rates where the rate of release from one phase was sufficiently fast to mask the rate of release from the other. At this temperature, more glyphosate was partitioned into the non-labile phase of the Walpeup and Rutherglen soils than at 25°C. However, the rate of release of glyphosate from this phase increased for the Walpeup soil relative to that at 25°C, but decreased substantially for the Rutherglen soil. This suggests that different mechanisms for the binding of glyphosate into the non-labile phase may exist between soils. © 1998 SCI. |
