To explore whether cable presence, glass sponge cover or an
interaction between these variables affected megafaunal density
and select taxa abundance, we considered multiple regression
analysis. However, the assumption of collinearity could not be met
due to high correlation between cable presence and glass sponge
cover. Thus, the relationship between select taxa abundance (for
selection criteria see Table 1) and glass sponge cover was examined by calculating Pearson's correlation coefficients (a ¼ 0.05). Glass
sponge cover (live and total sponge area per m2 of substrate) was
calculated from a random subset of still images taken along two
cable (Cables 2 and 3) and two control (Controls 2 and 3) transects
over three spring surveys (n ¼ 120, 10 images per transect per
survey) and averaged per transect (n ¼ 12). For each of these
transects, basic megafaunal density and abundances of spot prawn
and “Other Arthropoda” (see Table 1) were calculated, in m2, from
corresponding video clips.