Dataset: Database summarizing independent studies of nekton taxa associated with oyster reefs

Literature review:
First, we determined the variety of reef characteristics reported by studies investigating nekton recruitment enhancement by oyster reefs using citations from the reference list of a recently completed meta-analysis (zu Ermgassen et al. 2016). We identified nekton families that were regularly reported at oyster reef and control habitats, including both resident (i.e., species that feed, breed, and shelter on reefs long after initial recruitment, Coen et al. 1999; Harding and Mann 2000) and transient (i.e., species that recruit to structured habitats but are more widely distributed across multiple habitats after recruitment, Harding and Mann 2001) reef-associated species. We also performed forward searches in Google Scholar on two published syntheses: Peterson et al. (2003) and zu Ermgassen et al. (2016). We retained studies that met the following criteria: 1) authors quantified density or relative abundance of target nekton families at both oyster reefs (or experimental units that contained oyster shell and served as a mimic for reef habitat; e.g., Humphries et al. 2011) and unstructured mud or sand habitats within the same study; 2) restored reefs used oyster shell, including shell piles, cultch, bagged shell, or shell piles from other species (e.g., surf clams) if topped by oyster shell; 3) restored reefs were within the tidal extent of natural reefs (< 10 m deep relative to mean low water [MLW] at the base of the reef; Kennedy and Sanford 1999), 4) fishing gear(s) quantitatively censused juveniles; and 5) authors reported densities or abundances of target nekton by species or family.

Data Extraction:
We extracted densities or abundances, measures of spread (standard deviation or standard error), and sample sizes of each nekton species from oyster reefs and their paired unstructured control habitat patches. We extracted data for nine nekton families, including reef residents: toadfish (Batrachoididae), blennies (Blenniidae), gobies (Gobiidae), and skilletfish (Gobiesocidae, which were later removed due to limited data availability); and reef transients: grunts (Haemulidae), snappers (Lutjanidae), swimming crabs (Portunidae), drums (Sciaenidae), and porgies (Sparidae; Table 1 of Davenport et al. 2021). We normalized densities to mean individuals m-2, abundances to mean individuals per sample (relative abundances), and measures of spread to one standard error of the mean (Appendix S2 of Davenport et al. 2021). We extracted tidal zone (subtidal or intertidal), reef type (restored or natural), restoration method (reefs restored with or without live oysters), and when available, reef size (standardized to m2), vertical relief (distance from bare sediment to the highest point on the reef, standardized to m), tidal elevation (at the base of the reef, in m relative to MLW), adult oyster density (individuals m-2 > 75 mm in shell height, or specified as adult by the authors), and the year of restored reef construction, from which we calculated reef age (Appendix S1: Table 1; Appendix S2 of Davenport et al. 2021).