Soil sampling studies for estimation of weed seedbanks

Abstract

Studies were conducted during 1996-97 to develop guidelines for soil sampling to best estimate the weed seedbank. Results based on samplings from 15 arable sites showed large within site variations. For predicting a given seed population with some accuracy, either the species has to be present in large numbers (>1000/m2) and/or large numbers of soil cores need to be taken. Comparison of results from 25 mm and 75 mm diameter cores (from 5 arable sites), showed that collecting the same quantity of soil with a greater number of smaller cores would give better estimates. For practical purposes, approximately 25 cores, each 25 mm diameter, may provide a sufficient size sample for fields of annual crops such as maize