Environmental and Ecological Statistics

, Volume 4, Issue 2, pp 95–115

Natural variability of benthic species composition in the Delaware Bay

Authors

  • Dean Billheimer
    • Boeing 155
  • Tamre Cardoso
    • Department of StatisticsUniversity of Washington
  • Elizabeth Freeman
    • Department of StatisticsUniversity of Washington
  • Peter Guttorp
    • Department of StatisticsUniversity of Washington
  • Hiu-Wan Ko
    • Department of StatisticsUniversity of Washington
  • Mariabeth Silkey
    • Boeing 155
    • Department of StatisticsUniversity of Washington
Article

DOI: 10.1023/A:1018514226420

Cite this article as:
Billheimer, D., Cardoso, T., Freeman, E. et al. Environmental and Ecological Statistics (1997) 4: 95. doi:10.1023/A:1018514226420

Abstract

Biological monitoring of aquatic biota is used to assess the impact of changes in the environment. Critical to the development of a sound biological monitoring protocol is the judicious selection of organisms and organism characteristics to be monitored. Accurate interpretations of change necessitate description of the natural variability of the system. We introduce a state-space model for compositional monitoring data, and illustrate how one can incorporate spatial structure and covariates to assess natural variability. The methods are illustrated on benthic survey data from Delaware Bay, and applied to proportional composition at the genus level. The distribution of benthic macroinvertebrates in Delaware Bay depends significantly on salinity. There is residual spatial dependence in the data after accounting for the salinity effect.

benthic invertebratesbiological monitoringspatial modelstate-space model

Copyright information

© Chapman and Hall 1997