As a relatively new practice of seasonal productivity management in the rangelands of the Flooding Pampas, spraying glyphosate at the end of the summer is usually tied up to fertilizer application, adding N to promote the growth in autumn-winter of cool season grasses and P to correct a common deficiency in the soils of the area. Although there were previous efforts on the subject evaluating the changes in productivity, composition and quality of these rangelands, there was a void in studies that also investigate what processes in soil could be affected by the abovementioned practices, which can at its turn condition the performance of the plant component in these systems. To this regard, we concentrated on the study of the residue decomposition dynamics provided that the amount and quality of the residues generated by glyphosate spraying could be different from the natural senescence processes acting on the dominant warm season grasses stand and this fact would origin different decomposition rates. There were also some questions about the effect of the herbicide on the soil biotic components although they are not the specific target of glyphosate.

A field experiment was set up on a rangeland in a cattle raising ranch in Azul, Buenos Aires province, Argentina (36°40'S, 59°52'W) in February 2003, in an area previously fenced to avoid cattle accession. A first sampling was done in February in order to characterize the dominant soils and the initial composition of the plant community. Climate information was recorded (rainfall and mean, maximum and minimum daily air temperature at 5 cm from the soil surface) from February to December 2003. Naturally standing dead material (SDM) from the rangeland and plant dead matter generated by glyphosate (Glypho) was collected in February, oven dried and cut to yield 1.5 to 2-cm length debris. The experiment followed a split plot design with three replications: in the principal (main) plots the vegetation was managed according to conventional practices (defoliation by cutting trying to simulate grazing) and with glyphosate in order to have a canopy around the decaying material similar to managed conditions. Four subplots in each main plot – replication combination received the addition of N and P at a rate of 0-0, 0-70; 70-0 and 70-70 kg ha^-1 of N and P respectively as broadcasted urea or STP in March. In April, residues of each source were put in nylon bags in the subplots, in small bare patches where carefully plant material was previously removed; five samplings were performed in order to evaluate residue mass loss and global biological activity (CO₂ trapped in alkali beakers) the 24 hours following the bag removal. Results were analyzed using the SAS package.

There was no effect neither of the residue source (SDM vs. Glypho) nor the nutrients added on the decomposition rate (p> 0.05) at any sampling date. Plant analysis carried on both residues revealed that differences were not significant between structural carbohydrates for SDM and Glypho, except by the total N content that was higher in Glypho debris. The absence of enhanced decomposition due to nutrient addition is leading to suspect losses in case of N by volatilization due to the drought that followed urea application and some P provision in the 0-0 NP glyphosate treated plots by the herbicide that is used in a high rate (4 l ha^-1). Rangeland soil microorganisms could be also resilient to change, with a flora not trained to react to big amounts of external nutrients in autumn, whereas the normal natural nutrient trade off occur commonly in spring-summer. The cumulative mass loss from the nylon bags during one year of study was equivalent to graminoid residues evaluated by the same procedure (64.0 to 70.8% remained undecomposed), with an average 45% of this decay expressed at the first sampling (cumulative 200 degrees-day). Comparing these results to the residue mass loss without encryption recorded in a parallel study, there is an important underestimation of the decomposition rate using the nylon bag procedure (about 40 to 50%), mainly due to the disruption of the soil-residue interface, the limited opportunities to fauna to access to the debris and the missed soil microbial – alive plants interactions. The global biological activity adjusted mainly to climatic fluctuations, with no long term impact of herbicide or nutrient provision.