A meta-analysis comparing the toxicity of sediments in the laboratory and in situ

Grant C. Hose*, Brad R. Murray, Margaux L. Park, Brendan P. Kelaher, Will F. Figueira

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

Sediment toxicity tests in the laboratory are an important part of ecological risk assessments, yet how they relate to sediment toxicity in situ has rarely been explored. Using meta-analysis, we examined differences in the toxicity of sediment tested in the laboratory and in situ. Data from four published studies were subjected to rigorous statistical analyses. Overall, the toxicity of sediments in laboratory tests was substantially less than their toxicity in situ. Differences between laboratory and in situ toxicity, expressed using the log odds ratio effect size, varied significantly among published studies. Effect size increased significantly with increasing sediment toxicity, showing that the more toxic the sediment, the greater the disparity between laboratory and field toxicities. Our findings may not apply to all laboratory/field comparisons; however, we consider that the overlying water in field situations is a significant contributor to this relationship through additional contamination and toxicity. Our findings also have important implications for the use of laboratory tests to assess improvements in sediment quality and remediation, because changes in laboratory toxicity may not reflect the true improvements to sediment quality in situ.

Original languageEnglish
Pages (from-to)1148-1152
Number of pages5
JournalEnvironmental Toxicology and Chemistry
Volume25
Issue number4
DOIs
Publication statusPublished - Apr 2006
Externally publishedYes

Keywords

  • In situ toxicity tests
  • Laboratory-field comparison
  • Meta-analysis
  • Sediment toxicity

Fingerprint

Dive into the research topics of 'A meta-analysis comparing the toxicity of sediments in the laboratory and in situ'. Together they form a unique fingerprint.

Cite this