Date of Award
1986
Degree Type
Dissertation
Degree Name
Doctor of Philosophy
Abstract
Since sampling procedures for seed bank studies have long been ignored, my objective was to evolve a procedure which would give an unbiased estimate of the seed bank of a major weed species while minimizing the sampling variance and the sampling effort. 1024 soil cores (1.9 cm in diameter and 15 cm deep) were taken systematically over 1.35 ha in a corn field in Oxford County, Ontario. Intact seeds of Chenopodium spp. were extracted using a solution of sodium hexametaphosphate and sodium bicarbonate (2:1 w/w). Seed numbers were recorded to create a data bank from which repeated samplings with replacement were made to compare random, systematic, stratified random and cluster sampling in their capacity to minimize the sampling variance calculated by the Monte Carlo technique (MC S(,x)('2)). MC S(,x)('2) values decreased with increasing sample size regardless of the sampling method used. MC S(,x)('2) values for systematic and cluster sampling were greatly influenced by the sampling interval and the shape and size of the cluster respectively. This was attributed to the underlying aggregate seed distribution of Chenopodium spp. in the soil with its pattern of high and low seed density parallel to corn rows. There were some differences between MC S(,x)('2) values from random and stratified random sampling but either of these methods could be used to sample seed banks. Of the auger sizes tested (1.9, 2.7 and 3.3 cm in diameter and 15 cm deep), the smallest sampling unit gave the most precise estimate of the density of Chenopodium spp. seeds on a per volume basis. The minimum sample size needed to estimate the seed bank size for a common species ranged between 60 and 100 small sampling units. Fields under various crop rotations were found to have similar sized seed banks of Chenopodium spp. (802 to 2912 seeds/m('2)). Application of manure increased the number of Chenopodium spp. seeds in the soil (11 829 seeds/m('2)) the most. Two fallow fields with large populations of lamb's-quarters had similarly large seed banks (16 357 and 21 512 seeds/m('2)). The seed banks of Chenopodium spp. of all fields surveyed averaged 92% black seeds and 8% brown seeds. The contributions of the different categories of Chenopodium spp. seeds to the seed population in the soil were consistent throughout all fields with approximately 10% whole (viable) seeds, 56% damaged seeds and 34% underdeveloped seeds.
Recommended Citation
Benoit, Diane Lyse, "Method Of Sampling Seed Banks In Arable Soils With Special Reference To Chenopodium Spp" (1986). Digitized Theses. 1540.
https://ir.lib.uwo.ca/digitizedtheses/1540