Limited Evidence for Depth Specialism in Isolated Seamount Reef Predators

Abstract

Gradients in light, temperature and hydrodynamics associated with water depth are important determinants of ecological communities in marine environments. While depth specialism in coral reef fishes has been extensively studied in shallow (< 30 m) coastal reef systems, less is known about how depth-associated drivers operate over the larger depth ranges on isolated pinnacle and seamount reef systems, which are known to support abundant assemblages of predatory fishes. Using remotely operated vehicles, we surveyed predatory fish assemblages across a 100 m depth gradient on three seamount reefs in the Coral Sea. We tested for declines in abundance and diversity, as well as differences in assemblage structure of predatory fishes among depth strata. Species richness and abundance decreased significantly with depth, with predator abundance declining fourfold between the shallowest (5 m) and deepest (95 m) depths surveyed, while species richness was halved. Despite this, compositional differences among depth zones were minimal, with most taxa spanning the full depth range, suggesting adaptations to the limited horizontal habitat available on seamounts. Depth-associated shifts in taxonomic composition were primarily attributed to a single predator family, reef sharks (Carcharhinidae), which increased in abundance at mesophotic depths. The capacity of a large number of predatory fish taxa to utilize a wide range of depths allows these organisms to access favoured thermal environments and may be a potential resilience mechanism under future environmental change. Further studies are needed to assess the implications of depth use for predator behaviour, trophodynamics and conservation strategies.

Keywords: MCE; depth; pinnacle; predatory fishes; seamount coral reef.

FIGURE 1

(a) Location of Osprey, Bougainville and Lihou reefs in the central and northern Coral Sea. (b) ROV survey sites at Osprey Reef (3 sites). (c) Bougainville Reef (2 sites), (d). Lihou Reef (12 sites). Total numbers of transects per reef are indicated in parentheses and relative reef transect numbers indicated by pie size. Proportions of transect conducted at depth zones per site are indicated by pie colour, with depth zone coded in blue shades. Shallow = 0–30 m, Upper (mesophotic) = 31–60 m, Lower (mesophotic) = 61–100 m depth.

FIGURE 2

Infographic showing the remotely operated vehicle (ROV) field survey techniques for quantifying fish assemblages within deep reef habitats. Transects were conducted from max 100 m–surface, with 5–10 separation (horizontal and vertical) between transects. Adapted from Galbraith et al. (2022).

FIGURE 3

Effect of depth on predatory fish abundance. (a) Modelled trend. Each line represents a posterior draw from the model. (b) Distribution of proportional decline in abundance per 1SD increase in depth. Median decline (34% per 1SD depth) represented black circle. Narrow/wide horizontal bars indicate 95% and 80% area under the curve, respectively. Dashed line at zero represents no effect. (c) Pairwise contrast between the shallowest and deepest depths surveyed in the study (4 and 99 m). Dashed line indicates a 1:1 ratio.

FIGURE 4

Effect of depth on predatory fish asymptotic species richness. (a) Modelled trend. Each line represents a posterior draw from the model. (b) Distribution of proportional decline in asymptotic species richness per 1SD increase in depth. Median decline (18% per 1SD depth) represented black circle. Narrow/wide horizontal bars indicate 95% and 80% area under the curve, respectively. Dashed line at zero represents no effect. (c) Pairwise contrast between the shallowest and deepest depths surveyed in the study (4 and 99 m). Dashed line indicates a 1:1 ratio.

FIGURE 5

Nonmetric multidimensional scaling of family assemblage composition at each depth zone.

FIGURE 6

Mean abundance per transect (±S.E.) by depth zone for each of the 17 predatory fish families used in the multivariate model. Panels ordered by most–least abundant families in thestudy.