Date: August 12, 1998
To: copepoda@omnet3.omnet.com
QUESTION:
The measurement of copepod egg production rates has become a standard, and very convenient, method for estimating copepod growth and/or feeding in a time and site specific way. However, I do not believe we have come to consensus on the way in which we collect and treat our data. Some of us use incubations of several or many female copepods, and others separate incubations of single animals. I prefer the latter, but in this case a set of incubations always includes some individuals which do not produce eggs. Do we discard this information, on the basis that the females concerned are non-reproductive, or do we include this data, which results in a heavily skewed data set. In this case it depends to some extent what question we are asking; for example if recruitment is the issue then obviously the zero values must be left in; if feeding rate is the issue then the case could be made for excluding them.
Has anybody looked into the statistics of egg production data? What is
the feeling about the statistical power of the method, and how many
replicate incubations should we be running to get reliable data?
David McKinnon
Australian Institute of Marine Science
PMB No. 3, Townsville 4810, AUSTRALIA
Telephone : +61 7 4753 4444
FAX : +61 7 4772 5852
[SMTP:d.mckinnon@aims.gov.au]
http://www.aims.gov.au/
***************
COMMENTS:
I would like to bring to your attention the ICES Zooplankton Methodology Manual, which is (at last!) scheduled for publication next year and contains a section addressing issues brought up in that discussion. In brief, the manual recommends that, for most applications, a minimum of 30 (preferably 40) females are incubated separately (ie one female per container). This procedure yields data on clutch size and may reduce cannibalism. A simple mean and standard deviation is calculated, including all zeros (ie females that have not spawned during the incubation period). This gives an unbiased
estimate of the mean population egg production rate, for which the 95% confidence interval can be calculated from the theory of normal
distribution (even if the observations themselves are not normally distributed). The mean and variance of clutch size can be calculated from all individuals that spawned during the incubation period. However, the non-spawning females are not necessarily reproductively inactive and should in any case be included in the estimation of population egg production rate.
A draft (presently undergoing final revision) of the manual section dealing with egg production methodology is available upon request. Some of you may remember a survey of copepod egg production methods handed out at the ICES Zooplankton Meeting in Plymouth. The results of this survey were incorporated into the manual.
Jeffrey Runge
Fisheries and Oceans Canada
Laurentian Region
Maurice Lamontagne Institute
850, route de la Mer
P.O. Box 1000, Mont-Joli (Qc)
Canada G5H 3Z4
Tel.: (418) 775-0676
Fax: (418) 775-0546
E-Mail: RungeJ@dfo-mpo.gc.ca
http://www.qc.dfo-mpo.gc.ca/iml