The impact of 85 years of coastal development on shallow seagrass beds (Posidonia oceanica L. (Delile)) in South Eastern France: A slow but steady loss without recovery

Abstract

Shallow Posidonia oceanica beds (0 to −15 m), the most common seagrass in the Mediterranean, were mapped from aerial photographs dating from the 1920's and from 2012 along 800 km of coastline in South-Eastern France (Provence-Alpes-Côte-d’Azur region). Changes in P. oceanica bed spatial distribution (limits and extent) during these 85 years were analyzed in terms of concordance (remaining areas), positive discordance (expanding areas) or negative discordance (lost areas). Lost areas were linked with direct or indirect impacts of coastal development (artificialized coastlines (namely harbours, ports of refuge, landfills, artificial beaches, groynes and pontoons, submarine pipelines and aquatic farms) visible on the photographs. The comparison showed that 73% of the shallow limits have declined. Considering spatial extent, remaining seagrass meadows areas accounted for the major part (85%), while lost areas accounted for 13% and expanding areas for 1.1%. Lost areas were mainly linked with artificial coastlines but 44% remained with undetermined causes (invisible pressures and/or mixed effects). The analysis of 96 coastal facilities creating the artificial (namely man-made) coastlines showed that the highest impact over the longest distance (5 km) was caused by harbours. Only artificial beaches had such a distant impact. Pontoons were the least surrounded by lost seagrass meadows areas. These quantitative data offer important information for marine conservation.

Introduction

With more than seven billion people on Earth (United States Census Bureau, 2014), human activities have global impacts on all oceans and seas (Jackson et al., 2001, Stachowitsch, 2003, Halpern et al., 2008). Coastal areas and coastal ecosystems are particularly affected as they concentrate rich marine biodiversity, an important human population and a wide range of human uses (Halpern et al., 2008). Population densities in coastal regions are now about three times higher than the average elsewhere, and the last 70 years with the industrial revolution and the population explosion were particularly demanding: rapid urban development, construction of new seaside resorts, marinas and extensions of existing ports (Small and Nicholls, 2003). However, marine ecosystems provide important and valuable goods and benefits (i.e. contributions that humans derive or create from ecosystem services (Millennium Ecosystem Assessment MEA, 2005, Haines-Young and Potschin, 2013)). For example, more than half of the total value of the world natural capital and services are considered to be related to a single marine ecosystem: seagrass beds (Costanza et al., 1997). In this context, marine conservation science needs to assess and understand the impacts of society on marine habitats in order to protect them. Approaches based on expert opinion (Halpern et al., 2007, Halpern et al., 2008, Claudet and Fraschetti, 2010, Parravicini et al., 2012) are often used as a proxy for real impacts on habitats, but they are not as significant as quantitative assessments, and the critical lack of empirical knowledge about marine systems impedes the implementation of effective conservation measures (Claudet and Fraschetti, 2010). The knowledge of historical reference points (the state of conservation of marine ecosystems prior to large-scale human impacts), and observation of the consequences of past pressures on their current state remains the best approach to reducing human impacts and moving along a sustainable development path, but we are lacking this knowledge (Underwood, 1992, Pauly, 1995, Micheli et al., 2013).

Seagrasses are often considered as biological sentinels because any change in their distribution (e.g. a reduction in the maximum depth limit or a loss of covered areas) implies an environmental change (Orth et al., 2006). Posidonia oceanica L. (Delile) is the most common seagrass species in the Mediterranean Sea (Boudouresque et al., 2012). It forms extensive meadows from the surface to 30–40 m depth (depending on water transparency and temperature). Over time, this long-lived plant builds up a set of rhizomes and roots which interstices are filled in by sediment; this structure is called ‘matte’ (Boudouresque et al., 2012). The plant can reproduce both sexually and asexually but its growth is very slow (a few cm per year). After the death of the plant, the deterioration of rhizomes is very slow, leading to a dead matte that may persist for millennia (Boudouresque et al., 2012). Because of the important ecological (nursery, spawning, feeding, oxygenation) and economic roles (coastal protection and sediment trapping) (Borum et al., 2004, Boudouresque et al., 2012), P. oceanica is protected by EU legislation (Habitats directive), the Bern and Barcelona Conventions, national legislation and is classified Least Concern on the IUCN Red List (Pergent et al., 2010).

As with numerous seagrass species (Short and Wyllie-Echeverria, 1996, Spalding et al., 2003, Waycott et al., 2009, Selig et al., 2014), Posidonia oceanica meadows have known a widespread decline over the last decades (Boudouresque et al., 2009); a decline characterized by a decrease of shallow seagrass beds and/or by a reduction of the deeper limits and thus a loss of spatial extent. Ten percent is the global decline (loss of area) generally accepted for P. oceanica over the last 100 years (Boudouresque et al., 2012) but a recent paper claims a reduction by 50% of the density or biomass within the Mediterranean over the last 20 years (Marbà et al., 2014a). Actually, the magnitude of the overall P. oceanica area loss over the last century ranges from 0 to 50 % depending on the author (González-Correa et al., 2007, Boudouresque et al., 2009, Bonacorsi et al., 2013) but could reach 8% per year with possible functional extinction in 2059 according to others (Marbà et al., 1996, Jordà et al., 2012). The reality is difficult to assess because of a lack of reliable baseline data: quasi-absence of historical data, studies often only focusing on small spatial and temporal scales and/or using uncertain old maps (Montefalcone et al., 2013, Bonacorsi et al., 2013). These observed declines are mainly located near urban areas (Thomas et al., 2005, Boudouresque et al., 2012) and mostly associated with human activities even if they can sometimes be related to natural processes (e.g. colonization and erosion dynamics, climate change, sea level change, weather events, exceptional tectonic events or diseases) (Duarte, 2002, Boudouresque et al., 2009, Pergent et al., 2014, Tuya et al., 2014). A recent review of the literature showed that the P. oceanica decline is attributed to human physical impacts by two thirds (67.6%) of the studies (Marbà et al., 2014b). Main declines of P. oceanica meadows are related to coastline engineering (Ruiz and Romero, 2003, Boudouresque et al., 2012, Roca et al., 2014), aquaculture (Pergent-Martini et al., 2006, Holmer et al., 2008, Rountos et al., 2012), solid and liquid waste (Morena et al., 2001, Pergent-Martini et al., 2002, Boudouresque et al., 2012), pleasure boats and cruise tourism (Montefalcone et al., 2006, Okudan et al., 2011, Boudouresque et al., 2012) and to the introduction of exotic species (Boudouresque et al., 2012, Marbà et al., 2014a). However, the relative quantitative influence of each of these causes on the overall decline remains unknown.

The present work estimates the changes that the shallowest part of Posidonia oceanica meadows have undergone in connection with coastal human activities over a large spatial (800 km) and temporal (85 years) scale. The objectives are thus: a) to assess old and present P. oceanica meadows (limits and spatial extent) using a unique methodology, b) to link the loss observed with human activities in order to estimate their direct and indirect impacts on the meadows, and c) to quantify the spatial scale of the impacts on adjacent seagrass meadows. Considering the available literature and the plant characteristics (slow growth, long-term persistence, high sensitivity) we expect to observe a decline of a large part of the shallow limits (an average loss of 10% of the initial area is expected) mostly located near urban areas, but also to highlight an overall stability of the meadows general spatial extent and small expanded areas.