To demonstrate conservation effects resulting from marine protected areas, many studies rely on spatial comparisons between areas afforded different levels of protection. These spatial comparisons can be confounded if the habitat and reef size are dissimilar and not accounted for in the statistical analysis. Taking into account reef size (obtained from multibeam sonar data) and benthic habitat structure, this research tested the effect of management zonation (No-take Sanctuary Zone = NTSZ; Controlled Pelagic Zone = CPZ) on the population structure (relative abundance and average biomass) of six fish species in the iSimangaliso Wetland Park, South Africa. Furthermore, this study tested the effect of ignoring reef size in spatial comparisons. Our results showed that reef size had a significant positive effect on the relative abundance and average biomass of most, but not all species. When reef size was included in the models, the results showed that two of the six species presented no effect of management zone; two appeared to be directly affected by the permitted (past and present) fishing activity in the CPZ; and the last two species appeared to be affected by the disturbance caused by the diving and/or boating activity in the CPZ. Excluding reef size from the analysis consistently resulted in the predicted relative abundance and average biomass decreasing in the CPZ and increasing in the NTSZ. This effect was most marked in the average biomass data, as the management zone effect changed from negligible to significant for five of the six species. Our results highlight the importance of accounting for the reef size, or area of suitable habitat, when conducting spatial comparisons among species and illustrate the potential impact of the trade-off required to accommodate human needs within protected spaces.