Piers are hotspots for benthic marine debris in an urbanised estuary

Abstract

Records of anthropogenic marine debris and the threats it poses are increasing worldwide, yet we know relatively little about the distribution of benthic debris. The seafloor is the final destination for a large proportion of debris due to the degradation and sinking of items. A more detailed understanding of debris distributions in hotspots such as urbanised estuaries can help decision makers target management and remediation activities. We selected sites frequented by fishers and boaters in Sydney Harbour, an urbanised estuary, to investigate the impacts of recreational activities on debris abundance. The aim of this study was to examine variation in macro debris (>5mm in diameter) type and abundance at two habitat types (piers and non-piers). We chose five locations at various distances from the estuary mouth. In each location SCUBA teams performed fixed transects at two sites, one under a pier and one over nearby soft-sediment habitat. Debris was recovered by the divers and brought to the surface for classification and disposal. Surveys were repeated multiple times at each location between November 2019 and February 2020, recording a total of 2803 debris items over 36 survey events. Overall, piers had more than ten times the debris abundance of soft-sediment sites, and much higher proportion of debris types related to recreational fishing. Over half of the debris items in this study were plastic (65%), and approximately 70% of the total debris was classified as related to recreational fishing. This trait was most prominent in debris at sites closest to the estuary mouth, likely reflecting increased fishing activity in this area. This study indicates that policy makers and community groups in urbanised estuaries should focus monitoring, reduction, and remediation efforts near artificial structures such as piers, and that public awareness campaigns should target the behaviour of recreational users of these structures.

Fig 1. Five survey locations in Sydney Harbour.

Coloured bands represent distance from the estuary mouth at 500m intervals, calculated using ‘cost distance analysis’ in ESRI ArcGIS. The location of Sydney Harbour Bridge is shown for reference.

Fig 2. Debris item types in both pier and soft-sediment habitats.

A) Multi-dimensional scaling (MDS) plot showing influence of environmental variables and similarity between survey events using ‘Bray-Curtis’ distance matrix. Red points represent pier surveys and blue points represent soft sediment surveys. Ellipses were calculated using standard deviation with ‘veganCovEllipse’, with red representing pier surveys and blue representing soft sediment surveys. Fit of distance to harbour mouth represents the strength of the association, with a longer segment more strongly correlated to the data. B) Total item count of abundant debris types per pier or soft-sediment habitat. Debris count was plotted with a log 10 scale.

Fig 3. Debris abundance.

Predicted means and standard errors from generalised linear mixed model (GLMM) examining the abundance of (A) total debris items, (B) total debris items classified as fishing-related, and (C) total debris items classified as non-fishing related. Shaded areas and error-bars represent standard error at 95% confidence interval.

Fig 4. Debris abundance by material type.

Predicted mean and SE from generalised linear mixed models (GLMM) examining the abundance of A) metal debris, B) glass debris, and C) plastic debris. Shaded areas and error-bars represent standard error at 95% confidence interval.

Fig 5. General linear latent variable model (gllvm) results.

A) Interactions between environmental variables and debris traits. Darker red indicates substantial positive interaction and darker blue indicates substantial negative interaction (E.g. fishing debris is strongly associated with increased abundance at pier habitats compared to non-fishing related debris). B) Plot of environment by debris trait interaction coefficients estimates marked by x and 95% confidence intervals marked by lines. Grey confidence interval lines include 0 while black lines were significant.