Relationships between suspended particulate material, light attenuation and Secchi depth in UK marine waters M.J. Devlin a, * , J. Barry b , D.K. Mills b , R.J. Gowen c,1 , J. Foden b , D. Sivyer b , P. Tett d a Catchment to Reef Research Group, ACTFR, James Cook University, Townsville, Queensland 4811, Australia b Centre for Environment, Fisheries and Aquaculture Science, Pakefield Road, Lowestoft, NR33 0HT, UK c Aquatic Systems Group, AFESD, Department of Agriculture and Rural Development, Newforge Lane, Belfast, BT9 5PX, Northern Ireland, UK d School of Life and Health Sciences, University of Napier,10 Colinton Road, Edinburgh, EH10 5DT, Scotland, UK article info Article history: Received 28 April 2008 Accepted 29 April 2008 Available online 16 May 2008 Keywords: light attenuation suspended particulate matter Water Framework Directive Secchi depth abstract Measurements of sub-surface light attenuation (K d ), Secchi depth and suspended particulate material (SPM) were made at 382 locations in transitional, coastal and offshore waters around the United King- dom (hereafter UK) between August 2004 and December 2005. Data were analysed statistically in relation to a marine water typology characterised by differences in tidal range, mixing and salinity. There was a strong statistically significant linear relationship between SPM and K d for the full data set. We show that slightly better results are obtained by fitting separate models to data from transitional waters and coastal and offshore waters combined. These linear models were used to predict K d from SPM. Using a statistic (D) to quantify the error of prediction of K d from SPM, we found an overall prediction error rate of 23.1%. Statistically significant linear relationships were also evident between the log of Secchi depth and the log of K d in waters around the UK. Again, statistically significant improvements were obtained by fitting separate models to estuarine and combined coastal/offshore data – however, the prediction error was improved only marginally, from 31.6% to 29.7%. Prediction was poor in transitional waters (D ¼ 39.5%) but relatively good in coastal/offshore waters (D ¼ 26.9%). SPM data were extracted from long term monitoring data sites held by the UK Environment Agency. The appropriate linear models (estuarine or combined coastal/offshore) were applied to the SPM data to obtain representative K d values from estuarine, coastal and offshore sites. Estuarine waters typically had higher concentrations of SPM (8.2–73.8 mg l 1 ) compared to coastal waters (3.0–24.1 mg l 1 ) and offshore waters (9.3 mg l 1 ). The higher SPM values in estuarine waters corresponded to higher values of K d (0.8–5.6 m 1 ). Water types that were identified by large tidal ranges and exposure typically had the highest K d ranges in both estuarine and coastal waters. In terms of susceptibility to eutrophication, large macrotidal, well mixed estuarine waters, such as the Thames embayment and the Humber estuary were identified at least risk from eutrophic conditions due to light-limiting conditions of the water type. Crown Copyright Ó 2008 Published by Elsevier Ltd. All rights reserved. 1. Introduction Many recent European and US directives aimed at the assess- ment of eutrophication in marine waters include some measure- ment of nutrients and phytoplankton and look to describe the ‘‘risk’’ of undesirable biological response to nutrient enrichment (Tett et al., 2007). Our understanding of the process of nutrient enrichment and its causative influence on eutrophication symp- toms is an important component of any eutrophication assessment of marine waters. Recent changes in our conceptual understanding of eutrophication (Cloern, 1999, 2001; Costanza and Mageau, 2001; Tett et al., 2007), suggest that there are complex direct and indirect responses to anthropogenic nutrient inputs (Nixon, 1995). In addition ‘filters’ play a role in determining the sensitivity to enrichment and, in marine waters, these include the light climate and advective loss (Cloern, 1987; Bricker and Stevenson, 1996). Given this complexity, the process of linking anthropogenic nutri- ent enrichment to biological response is not a trivial task. A detailed assessment that would be required to achieve this and to cover all UK near-shore marine waters does not seem feasible or indeed scientifically justified. A more pragmatic ap- proach is to first screen each water body to determine susceptibility to the impact of anthropogenic nutrient enrichment. Traditionally, identification of risk has relied on nutrient loading and/or observed winter nutrient concentration. While this approach is useful for initial screening, to identify those water bodies receiving high * Corresponding author. E-mail address: michelle.devlin@jcu.edu.au (M.J. Devlin). 1 Present address: Fisheries and Aquatic Ecosystems Branch, AFESD, Agri-Food and Biosciences Institute, Newforge Lane, Belfast, BT9 5PX, Northern Ireland, UK. Contents lists available at ScienceDirect Estuarine, Coastal and Shelf Science journal homepage: www.elsevier.com/locate/ecss 0272-7714/$ – see front matter Crown Copyright Ó 2008 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ecss.2008.04.024 Estuarine, Coastal and Shelf Science 79 (2008) 429–439