Data was gathered experimentally at the Baruch Institute for Marine & Coastal Sciences. Proportion of the prey items consumed out of total provided (e.g. 1 of 3 mussels; 2 of 3 alternative prey) were used in analysis.

After finding no evidence of prey switching from bivalve prey to any of the previously mentioned diet items as a result of claw loss (see Results), another series of diet choice studies was conducted to assess the selection of various types and sizes of other prey in June 2013. They did this using 2 sequential experiments with a single set of 24 male stone crabs (mean CW ± SD, 81.3 ± 10.6 mm), and examined prey selection before and after removing the larger, crusher claw. Females were not included in these experiments because of low capture rates; however, no differences in consumption between male and female crabs were observed in either diet choice experiment 1 or the long-term consumption experiment. Because the same crabs were used for both experiments, the 2-clawed treatment of each experiment was conducted first (diet choice experiment 2, then prey size selection, as described later), and then the larger, crusher claw was removed fromthe crabs to conduct the 1-clawed treatment of each experiment (diet choice experiment 2, then prey size selection). Between each experimental treatment, crabs were starved for 48 h to standardize hunger. For both experiments, the crabs were housed in the same individual 5-gal buckets, each provided with a separate flow-through seawater source, allowing water temperature and salinity to fluctuate with ambient conditions. At the end of these 2 experiments, the remaining claw was removed from all crabs to assess the qualitative feeding habits of no-clawed crabs (details provided later).

For diet choice experiment 2, the crabs were provided with 3 small ribbed mussels (between 40 mm and 45 mm) and either 3 acorn barnacles (Balanus spp.) or 3 pieces of polychaete worms (Amphitrite ornata) for 48 h during both the 1- and 2-clawed treatments. These prey items were chosen because they are consumed by stone crabs (Powell & Gunter 1968) and are relatively abundant in the oyster reef in North Inlet estuary. For analysis, the results of 2-clawed and 1-clawed trials for each single crab were paired using a multivariate linear mixed-effects model (binomial distribution) with proportion consumed as the response variable; number of claws, prey type, and crab size as predictors; and crab identification number as the random factor to account for repeated measurements.