*2.4. Reef Fish Observations on Individual Structures*

Reef fishes' use of each individual structure was estimated using time-lapsed videophotography. A single, anchored GoPro was placed facing the structure and set to take a picture every minute for 60 min, the maximum time possible, given the power of the batteries. The camera was mounted 50 cm above the substrate and 150 cm away from the center of the structure. Every photograph after the first 5 min except the last 5 min was analyzed for fish species and for the relation of the fish to the structure (near or far). Only the photos captured after the first 5 min and before the final 5 min were analyzed to avoid any diver influence on reef fishes' use of the structure. Juvenile and adult fishes were treated the same for this study. Only those fish near the camera and directly above, below, beside, or in front of the structure were considered to be associated with it. These observed fish were then assigned to the 4 functional feeding guilds referred to above (Table S1). Since observations could be influenced by multiple images of the same individual, we considered the individual structure surveys as a measure of coral use rather than an estimate of fish abundance.

Over the course of 2 summers, we visited the 8 reefs once per year, recording 64 artificial structures and 112 natural structures so as to obtain 10,496 photos, resulting in 29,279 fish observations. After eliminating those fish considered not to be directly in the vicinity of the structure, our dataset consisted of 18,881 counts of 109 species of fish from 31 families, divided among the 4 functional feeding guilds as follows: 6047 individual herbivores from 23 species, 6225 individual omnivores from 48 species, 5725 individual invertivores from 12 species, and 884 individual piscivores from 26 species (Table S1).
