Taking a deeper look: Quantifying the differences in fish assemblages between shallow and mesophotic temperate rocky reefs

Abstract

The spatial distribution of a species assemblage is often determined by habitat and climate. In the marine environment, depth can become an important factor as declining light and water temperature leads to changes in the biological habitat structure. To date, much of the focus of ecological fish research has been based on reefs in less than 40 m with little research on the ecological role of mesophotic reefs. We deployed baited remote underwater stereo video systems (stereo-BRUVS) on temperate reefs in two depth categories: shallow (20-40 m) and mesophotic (80-120 m), off Port Stephens, Australia. Sites were selected using data collected by swath acoustic sounder to ensure stereo-BRUVS were deployed on reef. The sounder also provided rugosity, slope and relief data for each stereo-BRUVS deployment. Multivariate analysis indicates that there are significant differences in the fish assemblages between shallow and mesophotic reefs, primarily driven by Ophthalmolepis lineolatus and Notolabrus gymnogenis only occurring on shallow reefs and schooling species of fish that were unique to each depth category: Atypichthys strigatus on shallow reefs and Centroberyx affinis on mesophotic reefs. While shallow reefs had a greater species richness and abundance of fish when compared to mesophotic reefs, mesophotic reefs hosted the same species richness of fishery-targeted species. Chrysophrys auratus and Nemodactylus douglassii are two highly targeted species in this region. While C. auratus was numerically more abundant on shallow reefs, mesophotic reefs provide habitat for larger fish. In comparison, N. douglassii were evenly distributed across all sites sampled. Generalized linear models revealed that depth and habitat type provided the most parsimonious model for predicting the distribution of C. auratus, while habitat type alone best predicted the distribution of N. douglassii. These results demonstrate the importance of mesophotic reefs to fishery-targeted species and therefore have implications for informing the management of these fishery resources on shelf rocky reefs.

Fig 1. Map of the study area from Port Stephens to Seal Rocks on the east coast of Australia.

The area that has been mapped using a swath acoustic sounder is indicated by the rainbow shaded area, while the 50m and 100m contours are represented by a differing shade of grey (source: GeoScience Australia). The brown shapes represent hand digitised reef with profile. Shallow stereo-BRUV deployments are delineated by the green circles and mesophotic stereo-BRUVs deployments are delineated by the blue circlers. Circles with a cross are stereo-BRUV deployments within a no-take area. Inset: Study location on the east coast of Australia.

Fig 2. An RDA triplot ordination of transformed relative abundance data constrained by depth, latitude, fished/no-take.

Filled circles represent mesophotic reef stereo-BRUV deployments and open circles represent shallow reef stereo-BRUV deployments.

Fig 3. The spatial distribution of species richness, total MaxN, speciose families and species of interest.

a) Distribution of species richness, b) total relative abundance, c) Labridae relative abundance, d) Monacanthidae relative abundance, e) Chrysophrys auratus, f) Nemadactylus douglasii, g) Pseudocaranx georgianus and h) Meuschenia scaber as observed by stereo-BRUVs across the study area. Bubble size and colour represents the species richness for each individual stereo-BRUV deployment. Inset plots: Mean (+/- SE) of species richness and MaxN for shallow and mesophotic stereo-BRUV deployments.

Fig 4. The spatial distribution of fishery targeted species.

Distribution of all commercial and recreationally targeted species as observed by stereo-BRUVs across the study area. Bubble size and colour represents the MaxN of all commercial and recreationally targeted species for each individual stereo-BRUV deployment. Inset plot: Mean (+/- SE) MaxN of all commercial and recreationally targeted species across shallow and mesophotic reef.

Fig 5. Lengths of two of the prominent fishery targeted species.

a) the length distribution histogram for the total lengths of C. auratus at shallow (green) and mesophotic (blue) depths. A boxplot below summarises the distribution of total lengths for C. auratus at shallow (green) and mesophotic (blue) depths. b) The length distribution histogram for the total lengths of N. douglassi at shallow (green) and mesophotic (blue) depths. Boxplots below summarise the distribution of total lengths for N. douglassi at shallow (green) and mesophotic (blue) depths. The MLL for each species has been indicated by a dashed line on each plot.