Enhancing multiple scales of seafloor biodiversity with mussel restoration

Abstract

Restoration projects are underway internationally in response to global declines in shellfish beds. As diverse biological assemblages underpin a variety of ecosystem services, understanding broader changes in biodiversity associated with mussel restoration becomes increasingly valuable to scientists and restoration practitioners. Studies generally show bivalve beds increase species richness and abundance, but results are scale-dependent and conditional on the mobility of specific communities observed. We examined biodiversity at multiple scales to determine how communities with varying levels of mobility are influenced by subtidal mussel restoration. Significant changes in assemblage structure were observed in both mobile fish and epifaunal communities, with enhanced species richness and total abundance of associated individuals. In contrast, we observed site-dependent effects of bivalve restoration on macrofaunal community structure and composition, with sheltered, harbour mussel bed communities numerically dominated by detritivores accustomed to organically enriched, muddy sediments. Sediment organic matter significantly increased within mussel beds, and distance-based linear models showed that sediment organic matter was an important predictor of macrofaunal assemblage structure on mussel beds, highlighting the significance of benthic-pelagic coupling and biodeposition to soft-sediment communities. This study contributes novel methods and ecological insights on the role of species mobility and site selection in structuring restoration outcomes, better informing future mussel restoration efforts aimed at emphasising functionally-driven ecosystem services.

Figure 1

Each rectangle represents a paired study site, including a mussel bed and soft-sediment control (~ 5 m away), located either in Mahurangi Harbour or Kawau Bay, New Zealand. Pukapuka = PP, Lagoon Bay = LB, New Lagoon Bay = NLB, Motuora = MR, and Motoketekete = MK.

Figure 2

Conceptual diagram illustrating effects of subtidal mussel restoration on biological communities, differentiated by scales of mobility. Mobile communities of triplefins and commercially important fish species such as snapper (Chrysophrys auratus) are found in higher abundances on mussel beds than soft-sediment control sites, while highly transient elasmobranchs were not found to significantly differ with habitat type. Species richness and abundance of epifaunal invertebrates and encrusting species significantly increase with additional organic matter and hard substrate provided by mussels. Location-dependent changes in macrofaunal assemblage structure result from restoration efforts, with bivalve and polychaete detritovores more abundant beneath muddy, organically enriched mussel beds. Some symbols adapted from Integration and Application Network (ian.umces.edu/media-library).

Figure 3

Two-dimensional non-metric multidimensional scaling (nMDS) plots for visualisation of differences in assemblage structure observed between sites (colour) and mussel bed status (fill). (a) Mobile community (points represent all available reps from un-baited video cameras); (b) epifauna/benthic invertebrate community (ordination showing two transects per site); and (c) macrofaunal community (calculated from the distance between centroids where n = 8 sediment cores for the combined factor SiteStatus). Ordinations created from Bray–Curtis similarity matrices on square-root transformed data using PRIMER v. 7 (Clarke & Gorley 2015; available at http://www.primer-e.com/).

Figure 4

Ranked species abundance plot for pooled macrofauna cores (n = 8), split by site (colour) and mussel bed status (fill). Inset: dominance curves for all macrofauna cores (n = 64), split by mussel beds (light blue, filled) and soft-sediment controls (dark blue, open), highlighting increased dominance percentage in mussel bed communities. Labels for y-axis (relative abundance as a percentage of total abundance observed) and x-axis (species rank on log scale) are the same for both plots.