Distribution pattern of benthic invertebrates in Danish estuaries: The use of Taylor's power law as a species-specific indicator of dispersion and behavior

Erik Kristensen*, Matthieu Delefosse, Cintia Organo Quintana, Gary T. Banta, Hans Christian Petersen, Bent Jørgensen

*Corresponding author for this work

Research output: Contribution to journalJournal articleResearchpeer-review


The lack of a common statistical approach describing the distribution and dispersion pattern of marine benthic animals has often hampered the comparability among studies. The purpose of this study is therefore to apply an alternative approach, Taylor's power law, to data on spatial and temporal distribution of 9 dominating benthic invertebrate species from two study areas, the estuaries Odense Fjord and Roskilde Fjord, Denmark. The slope (b) obtained from the power relationship of sample variance (s 2) versus mean (μ) appears to be species-specific and independent of location and time. It ranges from a low of ~1 for large-bodied (>1mg AFDW) species (e.g. Marenzelleria viridis, Nereis diversicolor) to a high of 1.6-1.9 for small-bodied (<1mg AFDW) species (e.g. Pygospio elegans and Tubificoides benedii). Accordingly, b is apparently a valuable species-specific dispersion index based on biological factors such as behavior and intraspecific interactions. Thus, at the examined spatial scale, the more intense intraspecific interactions (e.g. territoriality) cause less aggregated distribution patterns among large- than small-bodied invertebrates. The species-specific interactions seem sufficiently strong to override environmental influences (e.g. water depth and sediment type). The strong linear relationship between the slope b and intercept log(a) from the power relationship is remarkably similar for all surveys providing a common slope of -1.63 with the present sampling approach. We suggest that this relationship is an inherent characteristic of Taylor's power law, and that b as a dispersion index may be biased by e.g. sampling errors when this relationship is weak. The correlation strength between b and log(a) could therefore be envisioned as a data quality check.

Original languageEnglish
JournalJournal of Sea Research
Pages (from-to)70-78
Publication statusPublished - Mar 2013



  • Aggregation
  • Biological Interactions
  • Estuaries
  • Marine Benthic Fauna
  • Random Distribution
  • Sample Size

Cite this