An overview of statistical methods applied to CPR data

• Published in: Progress in oceanography Vol. 58; no. 2-4; pp. 235 - 262
• Main Authors: BEAUGRAND, G, IBANEZ, F, LINDLEY, J. A
• Format: Conference Proceeding Journal Article
• Language: English
• Published: Oxford Elsevier 01.08.2003
• Subjects: Animal and plant ecology; Animal, plant and microbial ecology; Biological and medical sciences; Earth Sciences; Fundamental and applied biological sciences. Psychology; General aspects. Techniques; Marine; Methods and techniques (sampling, tagging, trapping, modelling...); Oceanography; Sciences of the Universe; Sea water ecosystems; Synecology; Atlantic Ocean; North Atlantic; AN, North Atlantic; Marine environment; Statistical method; Data analysis; Plankton; Ecology; Sampling; Biological monitoring; Ecological abundance
• ISSN: 0079-6611; 1873-4472
• DOI: 10.1016/j.pocean.2003.08.006

Since the beginning of the Continuous Plankton Recorder (CPR) survey in 1931, information on the abundance of a large number of plankton species or taxa has been obtained on a monthly basis in the northern North Atlantic. The many different ecological issues in which the survey has been involved have led to the application of a range of statistical methods. In this paper, we review some of the methods applied to the CPR data by presenting new and up-to-date analyses. Special attention is devoted to multivariate analysis, which has been used extensively to extract information from the CPR database. Results obtained from recently applied geostatistical methods on CPR data are then considered. An example of a time series decomposition by the use of Eigenvector filtering is presented to illustrate time-series analysis.