Journal article
Variability in Length–Weight Relationships Used to Estimate Biomass of Estuarine Fish from Survey Data
Transactions of the American Fisheries Society (1900), Vol.134(2), pp.481-495
03/2005
Handle:
https://hdl.handle.net/2376/116844
Abstract
The biomass of fish populations is often calculated from abundance‐by‐length data using length–weight (LW) relationships from separate studies (e.g., from the literature). Estimates of biomass determined this way have two principal sources of error: (1) error in total numbers and size distribution of fish due to sampling variability; and (2) prediction error, including that arising from the use of a LW relationship from another time, place, population, or species. We developed LW relationships from 6,390 measurements of fish of 24 species in the San Francisco Estuary. Our principal objective was to evaluate the errors that arise when calculating biomass from length data. Data were obtained from four sampling studies (none designed for this purpose) and analyzed with analysis of covariance on log‐transformed data. Differences in LW relationships among studies were apparent. Five tests were applied to assess the influence of these differences on predictions of biomass from length data. Three of these tests indicated some bias arising from several sources, including differences in the range of lengths used to develop the relationships. The remaining two tests compared the sampling variability of two common fish species with variability and bias introduced by means of different alternative LW relationships from our data and from the literature. Length–weight relationships from the literature introduced some bias and somewhat more variability into the biomass estimates compared with estimates based on LW relationships obtained from the San Francisco Estuary. However, sampling error was the largest source of error in all cases. Although it is preferable to calculate biomass from LW relationships of fish from the same area and time period, the error induced by using relationships from other time periods, other areas, or the literature is typically small compared with sampling error, particularly when only relative measures of biomass are needed.
Metrics
10 Record Views
Details
- Title
- Variability in Length–Weight Relationships Used to Estimate Biomass of Estuarine Fish from Survey Data
- Creators
- Wim Kimmerer - San Francisco State UniversitySean R Avent - San Francisco State UniversitySteven M Bollens - San Francisco State UniversityFred Feyrer - California Department of Water ResourcesLenny F Grimaldo - California Department of Water ResourcesPeter B Moyle - and Conservation Biology, University of California–DavisMatt Nobriga - San Francisco State UniversityTammie Visintainer - San Francisco State University
- Publication Details
- Transactions of the American Fisheries Society (1900), Vol.134(2), pp.481-495
- Academic Unit
- Environment, School of the (CAS)
- Publisher
- Taylor & Francis Group
- Number of pages
- 15
- Grant note
- Interagency Ecological Program and California Bay‐Delta Authority (CALFED) (ERP‐00‐E109) CALFED Ecosystem Restoration Program Suisun Marsh and Delta Predators studies BREACH II and Delta Wetlands studies Interagency Ecological Program
- Identifiers
- 99900547860601842
- Language
- English
- Resource Type
- Journal article