Experimental removals of the dominant canopy-forming kelp Ecklonia radiata were conducted at two sites on rocky walls in New Zealand and monitored for approximately 1.5 years. We hypothesized that the removal of the E. radiata canopy would affect the structure of subcanopy assemblages, such that there would be a reduction in sponge species richness and abundance. Furthermore, we investigated the biological and physical (predictor) variables that best explained variability in sponge assemblages after canopy removal. Canopy removal led to a community dominated by turf algae, which corresponded with a decrease in sponge abundance and richness. Our results suggest that the Ecklonia canopy may positively influence the presence of sponge species such as Crella incrustans; we propose that the canopy may allow its coexistence with turf algae underneath the canopy by altering the light regime and other environmental factors, which may be detrimental for some species. Our results highlight how any loss of canopy-forming species might have negative effects on sponge assemblages, which could affect the energy flow and the overall biodiversity found in these habitats.