Document Type


Publication details

Schultz, AL 2016, 'Fish assemblages of unconsolidated substrata in the Solitary Islands Marine Park using baited remote underwater', PhD thesis, Southern Cross University, Lismore, NSW.

Copyright AL Schultz


Unconsolidated sedimentary habitat is the world’s dominant benthic sub‐tidal habitat, however fish assemblages associated with unconsolidated habitats are generally poorly understood when compared with assemblages found in other habitats. In the Solitary Islands Marine Park (SIMP), approximately 80% of the seafloor is composed of unconsolidated habitats, however fish assemblages associated with these habitats are poorly described, restricting objective management of these extensive habitats and the species therein. Recent improvements in video technologies provide new non‐extractive methods to rapidly gather data on fish assemblages in this habitat.

Fish assemblages associated with unconsolidated habitats were broadly surveyed across the SIMP using Baited Remote Underwater Videos (BRUVs) which have been proven to be a robust, non‐destructive, fishery‐independent and cost effective method of rapidly collecting fish assemblage data. The effects of reef proximity were assessed at six distances from reef, to understand the transition between reef‐associated and sediment‐associated fish assemblages, and to determine the distance from reef required to ensure only a sedimentassociated assemblage was sampled. Assemblages were then broadly sampled across depths 10‐70m to assess the effectiveness of the existing Habitat Classification System (HCS), which was previously only based on data for reef fish, at representing fish assemblages associated with unconsolidated substrata. High resolution swath imagery and granulometrics from grab samples, in conjunction with BRUVs, were employed to test the relationship between sediment type, backscatter indices, and fish assemblages. Shallow‐water (5‐8 m) 5 fish assemblages, which are not currently represented in the HCS, were surveyed across multiple beaches with different lengths, aspects and exposures.

BRUVs were highly effective at discerning differences in fish assemblage structure in unconsolidated habitats in the SIMP, and when used in conjunction with high resolution swath imagery were able to resolve differences at relatively small spatial scales (10s to 100s of metres). Reef proximity was found to be strongly influential in determining fish assemblage structure, and assemblages close to reef did not overlap with those at 400 m from reef. At intermediate distances there was a ‘halo’ assemblage comprised of both reef‐associated and sediment‐associated species. It was determined that at least 200 m was required to negate any noticeable effect on assemblage structure, and this was incorporated into all further sampling. Subsequent surveys of fish assemblages at depths 10‐70m found differences in assemblage structure relating to depth and longshore position in the park. However, multivariate regression modelling suggested that backscatter intensity, which could itself be a surrogate for sediment type or hardness, was also influential. Targeted sampling at locations with different backscatter intensity revealed markedly different assemblages, and multivariate analyses found that ‘gravel’ assemblages were discrete from ‘fine sand’ assemblages. Granulometrics and swath imagery metrics were found to be closely correlated, confirming that backscatter metrics could be reliably used as a proxy for sediment type in further studies. The current HCS only divides unconsolidated sediments by depth (<25 m, 25‐50 m, >50 m), and a further division into ‘gravel’ and ‘fine sand’ habitats would better represent fish assemblages in conservation planning. Assemblages in shallow water were 6 discrete from those in deeper waters, and mostly differed between beaches with aspect and exposure. Species richness in this depth range was higher than that found in 10‐70m.

BRUVs have been an effective method at sampling many species in unconsolidated habitats which would otherwise be difficult to non‐destructively sample. Many species would not be amenable to diver‐based sampling, and BRUVs have great application in areas too deep for divers to access. This thesis demonstrates that BRUVs are highly applicable to fish assemblage research in unconsolidated habitats, and have an important role in future research in this habitat, both in this region and globally.