Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Model parameterization, parametric sequential sampling, and geostatistical analysis of weed seedling populations
Abstract
Environmental, social, and economic constraints on present weed control strategies have resulted in a significant change in weed management philosophies. Obtaining and interpreting weed population information is critical to implementing alternative weed management strategies. The objectives of this research were to determine the stability of several 2-parameter discrete numerical distribution models and associated parameters for use in sequential sampling methods. In addition, geostatistical analysis was used to describe the spatial aspects of weed seedling populations to better understand weed population biology and ecology. The negative binomial distribution model was found to consistently fit both broadleaf and grass weed species populations across time and space. The parameters of this distribution, specifically k, were not stable over time and space for all weed species populations tested. Interfield variability was greater than intrafield variability with respect to the negative binomial parameter k. Inherent variability of weed seedling populations between fields suggests that a field specific management philosophy is needed to manage weeds in the field. Since a common or stable k (over time and space) was not found, multistage estimation was used to estimate k for each weed species at the time of sampling. Mean weed seedling density estimates derived from sequential estimation procedures tended to underestimate the true population mean for most weed species. Low weed population density, coupled with very low k values, may have resulted in poor density estimates. Sequential hypothesis testing is discussed as well as sampling based on spatial arrangement of weeds in the field. Geostatistical analysis of weed species was performed in two eastern Nebraska corn and soybean fields. Weed seedling populations were shown to be highly patterned aggregation occurring in the direction of the crop row. In general, an application of preemergence herbicide treatment reduced weed seedling aggregation and density. Broadleaf weed populations appeared to be stable over years with minor fluctuations perpendicular to the crop row.
Subject Area
Agronomy|Biostatistics
Recommended Citation
Johnson, Gregg Alan, "Model parameterization, parametric sequential sampling, and geostatistical analysis of weed seedling populations" (1994). ETD collection for University of Nebraska-Lincoln. AAI9425286.
https://digitalcommons.unl.edu/dissertations/AAI9425286