A statistical analysis was performed to investigate if, and to what extent, the response of sugarbeet and potatoes to fertilizer N depended on the amount of mineral N already present in the soil, soil type, and prior application of organic manures. For this purpose the results of 150 field trials with sugarbeet and 98 with potatoes were used. The analysis was focussed on the within-block stratum of variation in yield, where regression models were fitted to describe the response to N. For both sugarbeet and potatoes the best fit was obtained when not only fertilizer N was taken into account, but also soil mineral N, soil type and prior application of organic manures. The response to fertilizer N was weaker as the amount of soil mineral N was larger. The optimum amount of fertilizer N plus soil mineral N required was larger on sandy soils than on loam and clay soils. The difference was about 20 kg N/ha for sugarbeet and 100 kg N/ha for potatoes. When organic manures were applied prior to the application of fertilizer N, the optimum for both sugarbeet and potatoes was 15-50 N/ha lower than without application of organic manures. (Abstract retrieved from CAB Abstracts by CABI’s permission)

Soil food webs from conventional and integrated management practices at an experimental site (Netherlands) were analysed using multivariate statistical procedures and simulation modelling, so as to identify patterns in species interactions and material transfers. Cluster analysis, canonical discriminant analysis and canonical correspondence analysis of the dynamics of biomass-N of functional groups within the food webs indicated that the webs could be compartmented into functional groups based on food choice and trophic level. The degree of compartmentalization depended on management practice. Consumers of fungi were separated in time from consumers of bacteria under the integrated management practice whereas little separation was observed under conventional practice. Simulation modelling was used to estimate the flux rates of nitrogen among functional groups within the food webs. The modelling demonstrated that more flow occurred in the integrated plot than in the conventional plot. More material flow occurred in the upper 10 cm of the integrated plot than in the 10-25 cm layer, whereas there was no such difference in the conventional plot. The effect may be due to the differences in the tillage practice on each plot. (Abstract retrieved from CAB Abstracts by CABI’s permission)