MARINE DIVERSITY AND SEA-LEVEL CHANGES - NUMERICAL TESTS FOR ASSOCIATION USING EARLY JURASSIC BIVALVES

Past attempts to substantiate the species-area effect by correlating c hanges in sea-level and marine diversity have met with limited success . Partial rank correlation and concordance analyses are used as two co mplementary numerical methods to examine the association between sea-l evel changes and diversity as predicted by the species-area effect. Wh en applied to Early Jurassic bivalve species from northwestern Europe, the numerical analyses failed to discriminate an association (rho = - 0.298 for the partial rank correlation and (p) over cap = 0.45 for the concordance probability). Additional analysis using subsets of the da ta or recoding periods of anoxic water as periods of reduced habitable area (in addition to marine regression) also failed to show a signifi cant association. The absence of significant correlation is likely to be due to numerous biotic and abiotic factors that cannot directly be measured from sediments or fossil assemblages. A multitude of interrel ated cause-and-effect relationships renders the species-area effect a poor predictor of the influence of sea-level changes on marine diversi ty.