Premium
Quantifying Precision of in Situ Length and Weight Measurements of Fish
Author(s) -
Gutreuter Steve,
Krzoska Daniel J.
Publication year - 1994
Publication title -
north american journal of fisheries management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.587
H-Index - 72
eISSN - 1548-8675
pISSN - 0275-5947
DOI - 10.1577/1548-8675(1994)014<0318:qpoisl>2.3.co;2
Subject(s) - common carp , fish <actinopterygii> , zoology , cyprinus , lepomis macrochirus , carp , statistics , coefficient of variation , mathematics , biology , fishery
We estimated and compared errors in field‐made (in situ) measurements of lengths and weights of fish. We made three measurements of length and weight on each of 33 common carp Cyprinus carpio , and on each of a total of 34 bluegills Lepomis macrochirus and black crappies Pomoxis nigromaculatus . Maximum total lengths of all fish were measured to the nearest 1 mm on a conventional measuring board. The bluegills and black crappies (85–282 mm maximum total length) were weighed to the nearest 1 g on a 1,000‐g spring‐loaded scale. The common carp (415–600 mm maximum total length) were weighed to the nearest 0.05 kg on a 20‐kg spring‐loaded scale. We present a statistical model for comparison of coefficients of variation of length ( C l ) and weight ( C w ). Expected C l was near zero and constant across mean length, indicating that length can be measured with good precision in the field. Expected Cw decreased with increasing mean length, and was larger than expected C l by 5.8 to over 100 times for the bluegills and black crappies, and by 3 to over 20 times for the common carp. Unrecognized in situ weighing errors bias the apparent content of unique information in weight, which is the information not explained by either length or measurement error. We recommend procedures to circumvent effects of weighing errors, including elimination of unnecessary weighing from routine monitoring programs. In situ weighing must be conducted with greater care than is common if the content of unique and nontrivial information in weight is to be correctly identified.