Skip to main content
. 2007 Oct 26;104(45):17701–17706. doi: 10.1073/pnas.0707194104

Fig. 1.

Fig. 1.

Cross-plots of live-dead taxonomic similarity (Jaccard-Chao Index) and rank-order correlation of species abundances (Spearman's rho). In each plot, data sets occupying the upper-right quarter are characterized by the highest live-dead agreement, and those in the lower-left quarter the poorest. In both coastal embayments (A and B; 73 data sets analyzed here) and in open shelf settings (C and D; 34 large data sets used in ref. 18), molluscan death assemblages from pristine areas show consistently high agreement to the local living community by both metrics (A and C). Death assemblages from areas of suspected (AE0.5) and documented anthropogenic eutrophication (AE ≥1) overlap the pristine range but extend down to significantly poorer live-dead agreement by one or both metrics, with live-dead agreement generally decreasing as the intensity of AE increases (B and D). Lagoonal AE data sets having unusually high live-dead agreement are circled in B. Only the lowermost right corner of the cross-plot cannot be occupied.