RT Dissertation/Thesis
T1 Glyphosate use in agro-ecosystems : identification of key factors for a better risk assessment
A1 Tesfamariam,Tsehaye
WP 2010/10/21
AB Glyphosate ([N-phosphonomethyl] glycine) is a non-selective, post-emergence, organo-phosphorous, broad-spectrum herbicide used worldwide for controlling weeds in horticulture, agriculture, silviculture, and urban landscapes. It effectively controls most annual and perennial weed species and is the world´s biggest-selling herbicide. One reason for the popularity of glyphosate is its effect on roots and rhizome systems of weed following foliar application. After coming in contact with soil, glyphosate will be strongly adsorbed and this sorption behavior makes glyphosate unique as compared to most other herbicides and has elicited a general belief that it is rapidly adsorbed to the soil without any residual effect. However, glyphosate adsorption to the soil matrix seems a reversible process and glyphosate conserved in roots of treated target plants has been overlooked in most previous risk assessments. 
  Therefore, in face of the increasing number of yet unexplained observations of negative side effects after glyphosate application, this thesis was initiated to identify possible risk factors associated with the frequent use of glyphosate in agro-ecosystems. For this purpose: (1) relevance of waiting time between weed desiccation by glyphosate and subsequent crop planting, (2) remobilization risk of glyphosate fixed in the soil matrix mediated by pH change in the rhizosphere, (3) glyphosate preservation in target plant roots and (4) contribution of glyphosate released from decaying weed residues for intoxication of following non-target plants were investigated in controlled greenhouse conditions using two contrasting soils: a weakly buffered acidic Arenosol (top soil) and a highly buffered calcareous Luvisol (subsoil). Furthermore, field experiment was conducted to partially confirm the found results of controlled model experiments under greenhouse conditions.
  These model experiments as well as the experiment in farmer´s field revealed that the residual toxicity of glyphosate increased with a declining waiting time between glyphosate weed desiccation and subsequent crop planting. In the greenhouse experiments, seedling growth and biomass production of sunflower plants were strongly impaired by pre-sowing application of glyphosate in the variants with less than 21 days waiting time. The inhibitory effects on seedling growth were associated with a corresponding increase of shikimate accumulation in the root tissue as physiological indicator for glyphosate toxicity and impairment of the manganes-nutritional status of the sunflower seedlings. 
  Results of the field experiment at Hirrlingen/Tübingen confirmed the relevance of waiting time. Stunted development and heterogeneous emergence of winter wheat plants occurred at field plots where the wheat sowing was done 2 days, compared to plants sown 14 days after foliar application of glyphosate to weed plants. At a short waiting time (2 d), data on visual scoring showed up to 50% of the culture damage that was visually persistent still after 6 months at harvests. This was also associated with a reduced nutritional status of wheat plants Ca, Mg, Zn and Cu, particularly expressed when glyphosate application rate was elevated from 2L to 6 L ha-1. 
  Since glyphosate shows a similar pattern of reaction like that of phosphate in soil, it has been hypothesized that rhizosphere processes responsible for P mobilization are likely to co-mobilize also glyphosate. To test this hypothesis, an experiment was conducted using the two soils with contrasting properties pre-incubated with different rates of glyphosate and supplied with stabilized NH4+-N or NO3--N to induce the different changes in rhizosphere pH. From the results of this experiment, however, it was not possible to confirm this hypothesis. No glyphosate phytotoxicity of sunflower seedlings on the Luvisol with NH4+ could be detected due to observed minor rhizosphere acidification. In agreement, also no shikimate accumulation in root was measured. However, there was a distinct decrease in biomass of the sunflower seedlings at NH4- supply, possibly due to a missing NO3- signal. In contrast in the Arenosol no difference in growth could be shown between both supplied N-forms despite a clearly expressed difference in rhizosphere pH. 
  Root exudation of organic carboxylates has also been considered to assists the release of adsorbed phosphate in the rhizosphere from the soil matrix via exchange chelation. A similar phenomenon was expected for glyphosate. In the present study, however, supplementation of Na-citrate or citric acid to both contrasting soils, pre-incubated with different levels of glyphosate, did not show a clear evidence for an adequate glyphosate remobilization and the subsequent plant damage. On the acidic Arenosol, there was no difference in growth of sunflower seedlings between the treatments. In contrast, on the Luvisol, supplementation of Na citrate (10µmol g-1 soil) but not citric acid indicated some promotion of root growth on glyphosate free treatment. This could not be easily explained because no intracellular shikimate accumulation as bio-indicator for glyphosate could be detected in the treatments with glyphosate pre-incubated soil. 
In many plant species, glyphosate is not readily metabolized, but preferentially translocated to young growing tissues of roots and shoots, where it can be accumulated in millimolar concentrations. In soil-grown target plants, this inhomogeneous distribution of glyphosate within the root tissues may lead to the formation of hot spots of glyphosate containing root residues in soils. Subsequently this stored glyphosate as hot spots can be released during microbial degradation of root material. To evaluate the potential of roots of target plant in stabilization and subsequent release of glyphosate with intoxication of subsequent crop plants, model experiments were conducted with application of glyphosate either via rye grass as target weed plants or directly to the soil. Sunflower seeds were sown at different waiting times (0-21 days) for both glyphosate application modes. Toxicity of glyphosate applied shortly before sowing of sunflower as non-target was strongly dependent on the mode of glyphosate application. When glyphosate was sprayed on pre-cultured rye grass seedlings as model weed, detrimental effects on plant growth and the Mn nutritional status, as well as increased intracellular shikimate accumulation in root tissue were more strongly expressed than at a direct soil application of the same amount of glyphosate. The increased extent of toxicity after a glyphosate pre-sowing application to pre-cultured rye grass compared with a direct soil application might indicate that the root tissue of glyphosate-treated weeds represents a storage pool for glyphosate in the pots. 
  The globally increasing adoption of no-till or reduced tillage systems are becoming a driving force for an increase of glyphosate use. In such systems, glyphosate is applied pre-sowing for weed control and glyphosate may remain in root and shoot residues. Usually in these reduced tillage systems, only a minimal soil disturbance occurs at sowing, which might lead to limited incorporation of the glyphosate contaminated straw to the upper soil layer where germination of following non-target crop will take place. To evaluate such risk, a pot experiment was conducted under controlled greenhouse conditions with the two contrasting soils. Glyphosate was supplied via glyphosate pre-treated shoot or root material of rye grass applied either as chopped plant material ?straw? or as homogenate. Analysis of physiological parameters such as intracellular shikimate accumulation as metabolic indicator for glyphosate toxicity, biomass production and micronutrient status revealed, that a detrimental effect could be only with treated rye grass shoot material as straw or homogenates incorporated into the Arenosol but not into the Luvisol. This is most probably related to the difference in soil property between the two soils. At this level of glyphosate supply, the detoxification capacity of the highly buffered calcareous subsoil might have played a primary role in preventing glyphosate toxicity, while this glyphosate supply level seems beyond the detoxification capacity of the weakly buffered acidic Arenosol. 
  All together, the achieved results of the model pot experiments are in correspondence with that of the reported field experiments. Further, the results revealed the important role of glyphosate stored in root and shoots of weed plants as a glyphosate pool in soils for intoxication of following crops. More information on transformation of these glyphosate enriched crop residues and its glyphosate release during microbial decomposition in different soils are urgently needed for a better precaution and risk assessment of glyphosate use for weed control for farmer´s practice.    
K1 Glyphosat 
K1 Toxizität
K1 Mineralstoffhaushalt
K1 Nutzpflanzen
PP Hohenheim
PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
UL http://opus.uni-hohenheim.de/volltexte/2010/501