Anthropogenic sources of underwater sound can modify how sediment-dwelling invertebrates mediate ecosystem properties

Abstract

Coastal and shelf environments support high levels of biodiversity that are vital in mediating ecosystem processes, but they are also subject to noise associated with mounting levels of offshore human activity. This has the potential to alter the way in which species interact with their environment, compromising the mediation of important ecosystem properties. Here, we show that exposure to underwater broadband sound fields that resemble offshore shipping and construction activity can alter sediment-dwelling invertebrate contributions to fluid and particle transport--key processes in mediating benthic nutrient cycling. Despite high levels of intra-specific variability in physiological response, we find that changes in the behaviour of some functionally important species can be dependent on the class of broadband sound (continuous or impulsive). Our study provides evidence that exposing coastal environments to anthropogenic sound fields is likely to have much wider ecosystem consequences than are presently acknowledged.

Figure 1. The effect of sound field on the maximum mixed depth (^f-SPIL_max, cm) of sediment particles for Nephrops norvegicus.

Data points (open circles) have been horizontally jittered for clarity. Model predictions (open triangles) from the minimal adequate linear regression model (Supplementary Model S2) with GLS estimation (incorporating sound field as a variance covariate) are indicated. Positive values indicate increased particle reworking activity. Sound fields: Ambient, Gaussian sound spectrally shaped to mimic background sound; CBN, ambient + continuous broadband sound (Supplementary Sound 1); IBN, ambient + impulsive broadband sound (Supplementary Sound 2).

Figure 2. The effect of different sound fields on surface boundary roughness (SBR, cm) for Ruditapes philippinarum.

Data points (open circles) have been horizontally jittered for clarity. Model predictions (open triangles) from the minimal adequate linear regression model (Supplementary Model 3) with GLS estimation (incorporating sound field as a variance covariate) are indicated. Positive values indicate increased activity within the sediment-water interface. Sound fields: Ambient, Gaussian sound spectrally shaped to mimic background sound; CBN, ambient + continuous broadband sound (Supplementary Sound 1); IBN, ambient + impulsive broadband sound (Supplementary Sound 2).

Figure 3. The effect of different sound fields on bioirrigation activity (∆[Br⁻], mg L⁻¹) for (a) Ruditapes philippinarum and (b) Nephrops norvegicus.

Data points (open circles) have been horizontally jittered for clarity. Model predictions (open triangles) from the minimal adequate linear regression models (Supplementary Models 4 and 5) with GLS estimation (incorporating sound field as a variance covariate) are indicated. Negative values indicate increased bioirrigation activity. Sound fields: Ambient, Gaussian sound spectrally shaped to mimic background sound; CBN, ambient + continuous broadband sound (Supplementary Sound 1); IBN, ambient + impulsive broadband sound (Supplementary Sound 2).