Response of soil microbial biomass dynamics, activity and plant litter decomposition to agricultural intensification over a seven-year period

Soil microorganisms and the processes that they govern are essential for lo ng-term sustainability of agricultural systems, but most studies on agricul tural effects on the soil microflora are inherently short-term. We investig ated the effects of three aspects of agricultural intensification, i.e. cul tivation (disturbance), herbicide addition (modification of plant compositi on) and mulching (resource addition) on soil biological properties such as microbial biomass and activity over 7 yr in annual (maize) and perennial (a sparagus) cropping systems. The mulching treatment had strong, usually posi tive effects on both substrate-induced respiration (SIR) and CO2-C release from chloroform-fumigated soil throughout the study. In the perennial crop, treatments allowing high weed biomass caused large increases in microbial biomass and respiration after yr 3, and in both sites microbial biomass was positively correlated with weed biomass and negatively with crbp plant bio mass. This latter effect appears due to the high decomposability of weed re sidues relative to those from crop plants. Microbial biomass was also enhan ced in atrazine-treated plots in the annual crop but only during the final year, presumably due to beneficial effects of plot invasion by herbicide-to lerant weeds. Mulching often also enhanced the microbial metabolic quotient (qCO(2)), the bacteria-to-fungal biomass ratio and within-year temporal va riability of the microbial biomass, all of which are indicative of greater turnover and instability of the microbial biomass. Other treatments general ly had smaller effects on these properties, although in the perennial crop an intense summer drought in yr 4 caused a large elevation in the metabolic quotient in the herbicide-treated (low weed) plots relative to the other p lots, suggesting that high quality weed residues have stabilising effects. Temporal variability across years of both SIR and C02-C release from fumiga ted soil was greatest in the herbicide-treated plots in the perennial crop, suggesting that high weed biomass (producing easily degradable organic mat ter) has stabilising effects. Decomposition rates of added litter were part ially consistent with the microbial biomass data, with the highest litter b reakdown rates usually occurring in the mulched plots. Our study shows that soil biological properties such as microbial biomass and activity are not necessarily adversely affected by agricultural intensification and that con sequences of intensification mainly depend upon practices which alter the q uality and quantity of residue inputs. Further, our results underline the n eed for long-term field experiments, and several of the effects we identifi ed could only have been detected through an experiment of several years dur ation. (C) 1999 Elsevier Science Ltd. All rights reserved.