Soil Biol. Biochem. Vol. 29, No. I, pp. 1125- l 133.1997 0 1997 ElsevierScienceLtd. All rights reserved zyxwvuts PII: soo3&0717(%)00306-9 Printed zyxwvutsrqponmlkjihg in &eat Br it ain 0038-0717/ 97 $17.00 + 0.00 SIZE-DENSITY FRACTIONATION FOR IN SITU MEASUREMENTS OF RAPE STRAW DECOMPOSITION- AN .ALTERNATIVE TO THE LITTERBAG APPROACH? JAKOB MAGID,* LARS STOUMANN JENSEN, TORSTEN MUELLER and NIELS ERIK NIELSEN Department of Agricultural Sciences, Royal Veterinary and Agricultural University, Thorvaldsensvej 40, DK 1871 FC, Frederiksberg, Denmark (Accepted I5 October1996) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ Summary-Decomposition rates have been elucidated by litterbag studies or in some instances by field- scale CO2 flux monitoring, but some deficiencies are apparent in these methods. In a randomized block experiment with additions of 0, 4 or 8 t of straw material the particulate organic matter (POM > 100,um) was monitored for 20 months. During the first 4 months the POM in the amended treatments decreased quite rapidly, and more slowly during the remaining 16 months. By separating the total POM into light (p < 1.4 g cme3) and heavy (p > 1.4 g cm-‘) fractions, further resolution of the residue decomposition was possible. The heavy fraction C was completely unaffected by the application of residues, and apparently consisted mainly of “native” organic matter with a low rate of decay. The differences in POM between treatments could be attributed completely to differences in the light POM fraction. Analysis of lignin and cellulose in light fractions from the 8 t treatment showed that cellulose was preferemially utilized in the early stages of decomposition. The analysis also indicated that the in- itial lignin concentration was high in the light fraction (20%) compared with that of the rape straw (15%). Thus the “native” light fraction must have been rich in lignin, which may explain the slower rate of decay of light fraction in the unammended treatment. The estimates of decomposition from POM were in qualitative agreement with estimates of decompo- sition based on field scale CO* fluxes, but indicated a considerably higher turnover in the initial phase of decomposition. This is of special interest, since it has been recognized that the static chamber method for estimating field scale CO2 fluxes underestimates high rates of CO2 evolution, and thus the study of POh4 may be complementary to this approach. Furthermore, it seems the study of POM may be an alternative to the litterbag method in some instances. In contrast to the litterbag method, POM fractionation approach allows the added residue to be completely exposed to the soil environment, and thus to the full range of fauna1 and other soil interactions. 0 1997 Elsevier Science Ltd INTRODUCTION An improved understanding of fieldscale decompo- sition of soil organi’: matter relies on measurements of biologically meaningful pools and fluxes. The development of simulation models for interpretation of measurements or for prediction of changes in management or environmental conditions on the soil-plant-atmosphere system has provided powerful tools (e.g. Jenkinson et al., 1987; Parton et al., 1987; Hansen et al.., 1991), but their reliability has been limited especially by a lack of accuracy with regard to the simulation of soil biological and bio- chemical processes (de Willigen, 1991). Only after the advent of the above cited models it was demonstrated that reliable in situ measurements of soil microbial biomass can be made under a wide range of conditions, e.g. after addition of organic matter (Ocio et al., 1991; Sparling and Zhu, 1993; Voroney et al., 1993; J’oergensen et al., 1994). While *Author for correspondence. field-scale N-mineralization and leaching has been used extensively in order to parameterize and test simulation models for short to medium-term dynamics, such models have not been tested against measurements of field-scale C mineralization, as far as we know. Jensen et al. (1996) compared a static chamber method using alkali-trapping of CO2 for 24 h and a dynamic chamber method using IR gas analysis for 2 min at each point for estimating field scale COz-fluxes from unplanted soils. They con- cluded that despite considerable limitations on sandy soils, the static method provided the best integrative measure, and that it would be necessary to monitor continuously in order to obtain a better estimate by use of the dynamic method. However, a major limitation of the static chamber method remains that it does not accurately reflect high rates of CO2 evolution from the soil due to a diffusional limitation of the trapping of CO1 in the alkali. Thus the initial COz evolution from decomposition of freshly added plant material will often be con- siderably underestimated, by this method. 1125