AB 14:185-200 (2012)  -  DOI: https://doi.org/10.3354/ab00391

ABSTRACT: We quantified the biological response of 2 exploited species to marine reserve (MR) protection by comparing meta-analysis results based on response ratio (RR) and Hedges’ g statistics. To determine the effect of MR area and age on biological responses, a RR analysis was performed on density and size data for both species from sites inside versus outside 13 MRs. Most MRs supported a greater density of larger individuals than unprotected areas. Linear and non-linear plots of MR age and area with respect to species-specific density and size were used to quantify the trajectories of the responses. In the RR meta-analysis but not in the Hedges’ g meta-analysis, MR age explained significant variation in the density and size of both species, while no effect of MR area was detected in either analysis. Comparison of the performance of RR with that of Hedges’ g revealed no overall evidence of a relationship between them, probably because the RR does not include an estimate of variance, whereas the Hedges’ g analysis does. While RR analysis is an appropriate alternative to Hedges’ g statistic meta-analysis assessments of MR effectiveness because of its ease of use and interpretation, we recommend that future RR analyses include both an estimate of variance and a test for ‘effect size’. Finally, we recommend that research be aimed at determining why different MRs produce different biological responses—that is, address why significant ‘effect size’ exists.

KEY WORDS: Response ratio · Meta-analysis · Marine conservation · Biological response · Marine management