Annex "Subtidal habitats"
In the Dutch Wadden Sea two long-term subtidal monitoring programs exist and in the past decade subtidal mussel beds were studied in two research programs:
Annual assessment of blue mussel stock
Long-term data on subtidal habitats in the Dutch Wadden Sea are provided since 1992 by the annual blue mussel stock assessment (van Stralen et al., 2015). Samples are taken with two types of bottom dredges along a stratified grid. Besides the target species Mytilus edulis, all shellfish species, starfish and crabs are registered in spring. The primary aim is to determine the quota of blue mussel seed that may be fished for mussel culture. Additionally, data of the spring survey are used to determine trends and distribution of subtidal habitats consisting of dense aggregations of shellfish. Another qualitative survey using a commercial mussel dredge is carried out in autumn to determine the quota of blue mussels that may be fished in areas designated as ‘instable areas’, where the mussel beds are exposed to storms and/or starfish predation. Both the spring and autumn surveys are carried out in the western part of the Dutch Wadden Sea because historically subtidal mussel beds are only rarely found in the subtidal of the eastern Dutch Wadden Sea.
Since the stratified sampling grid is based on the expected occurrence of mussel beds, distribution and abundance of species associated with mussel beds, e.g. the Pacific oyster (Crassostrea gigas) are also assessed efficiently.
Monitoring of Rottum
Another long-term monitoring of subtidal benthic communities is carried out in the eastern Dutch Wadden Sea covering an area close to the islands of Rottumerplaat and Rottumeroog (Fey-Hofstede et al., 2014). Sampling with a van Veen grab started in 2002 and is carried out annually in autumn. Two of four gullies are located in an area that is closed to human activities since 2005, while the other two gullies are located outside this area and are therefore still subject to human activities. The aim of this project is to assess differences between the gullies open and closed for human activities.
During the research project PRODUS (Project onderzoek duurzame schelpdiercultur; 2006-2012) the effects of seed fishery on subtidal mussel beds were studied in the western Dutch Wadden Sea (Smaal et al., 2014). Investigations were directed towards the development, survival, recruitment and biodiversity of wild mussel beds and the biodiversity was compared with mussel culture plots. Applied sampling equipment included side scan sonar, box corer and bottom dredges.
The aim of the Mosselwad research project (2009-2015) was to contribute to the recovery of mussel beds by studying natural processes that are of importance for the settlement of new beds and survival of existing beds (Dankers & Fey-Hofstede, 2015). Field surveys, laboratory experiments and mathematical models studied both intertidal and subtidal mussel beds. Emphasis was put on predation pressure by birds, starfish and crabs, as well as on the influence of storm events and environmental conditions reducing the stability of mussel beds. Sampling included side scan sonar, underwater video recordings, box corer and bottom dredges.
In 2009, the National Park Authority and the State Agency for Agriculture, Environment and Rural Areas of Schleswig-Holstein started an initiative for a coordinated subtidal habitat mapping and monitoring of the northern German Wadden Sea from the Elbe estuary in the south to the Danish border in the north. So far, approximately 70 % of the main tidal channels have been surveyed. Mapping and monitoring are implemented in a close cooperation of Kiel University with its Research and Technology Centre Westcoast in Büsum. Several research and development projects, such as 'AufMod', 'STopP' and 'WIMO' support the ongoing multidisciplinary habitat mapping and monitoring activities (Figure 1) by developing new and adapting proven methodologies to generate additional data and to achieve a better ecosystem understanding.
In view of the EU-Habitats Directive and Marine Strategy Framework Directive, the Lower Saxony Water Management, Coastal Defence and Nature Conservation Agency (NLWKN) and the Wadden Sea National Park Authority (NLPV) started in 2013 a long-term routine mapping program of the subtidal areas of the Lower Saxony Wadden Sea as well as adjacent areas of the North Sea. The program is carried out by enhancing an annual hydrographic survey plan implemented since 2011 by the Forschungsstelle Küste (FSK) and optimized to collect swath bathymetry and backscatter intensities in shallow and very shallow waters. At the same time, a second vessel has been equipped with several modern acoustic sensors to increase the survey efficiency (Mascioli et al., 2014). The subtidal habitat mapping project adds to a dredge and grab sampling monitoring conducted by NLWKN since 2006, aimed to assess the ecological status of the benthic communities.
During the last years, great effort has been made to define habitat mapping methods. The Scientific Monitoring Concepts for the German Bight (WIMO Project) gave a strong contribution to the development of a methodological concept to map intertidal and subtidal areas. Five test areas have been surveyed since 2010 by coupling airborne based remote sensing and hydroacoustic techniques. High-resolution habitat maps provide the base for modelling approaches on different temporal and spatial scales for elaborating tools to predict habitat dynamics (Capperucci et al., 2013).
From 1986 to 2009, the bivalve stocks of the Danish Wadden Sea have been monitored by the Danish Institute of Fisheries Research (now DTU Aqua). Since September 2013 all Danish mussel boats are obliged to have a “black box” installed. By monitoring the fishing activity and position, information on where the fisheries have taken place and on the extent of the affected areas was collected.
DTU Aqua is planning to resume monitoring of bivalves stocks, both in the intertidal and in the subtidal of the Danish Wadden Sea in 2016 by means of traditional techniques (dredging and grab sampling) as well as new remote sensing methods, which shall be tested and included in the stock surveys of blue mussels, Pacific oysters and common cockles (Cerastoderma edule).
The subtidal blue mussel stock is assessed annually in spring with a suction dredge along a stratified grid. The area sampled per station is 30 m2. Stations deeper than about 8 m are sampled using a towed bottom dredge (15 m2 per station). Density and biomass of all species (larger than 5 mm) of shellfish, starfish and crabs are registered. Since 2012 a hydraulic grab is used for dense Pacific oyster beds capturing a sampling area of 1 m2. In the qualitative autumn survey a mussel dredge is used. Mussel biomass is not measured but estimated and contours of mussel beds are mapped roughly. Data from the autumn survey can be used to reconstruct the development of seed beds, and was used alongside the spring monitoring data to construct a habitat suitability model (Troost et al., 2015), but is not used for trend analyses. In the short-term ZKO programme additional subtidal areas were sampled with the same survey techniques and an additional prototype of an ‘Ensis-dredge’ that has not been developed further since.
In the Rottum monitoring program sediment-dwelling animals are sampled annually in autumn with a van Veen grab (0.18 m2 sampling area) targeting small sized macrobenthos. In each of the four gullies included in the program 20 van Veen grab samples are taken, stratified over sediment composition (sand and shell fragments). In 2016 additional ecosystem components will be investigated: sediment grain size (van Veen grab sampling), dermersal fish inventory (2 m beam trawl) and larger (>5 mm) benthic species (dredge sampling). Also mapping of the seafloor of the gullies is planned by means of side scan sonar and multibeam echo sounding.
In the PRODUS project (Smaal et al., 2014) the effects of mussel seed fisheries were studied in 37 experimental plots where half of the plot was closed for fisheries and the other half open to fisheries. The development of mussel biomass and large macrobenthos was studied using the same techniques as in the annual stock assessment. Biodiversity was studied using box core samples (sampling area of 0.06 m2). In addition, a large biodiversity survey was carried out in 2008 throughout the entire western Dutch Wadden Sea using a box core. The same 397 stations were sampled that had been sampled before in 1981 and 1982 (Dekker 1989; Dekker & Drent 2013; Smaal et al., 2014). Of these, 53 stations were located inside a mussel culture plot. In addition to the PRODUS research plots, several wild mussel seed beds were closed for fisheries and their development over time was studied. The development in mussel bed structure was studied by side scan sonar recordings.
Many field surveys and laboratory experiments were conducted within the Mosselwad project. Effects of starfish predation were investigated in the laboratory, by field observations and mathematical models. Bird counts were combined with sublittoral survey data on molluscs to examine predation pressure and food preferences. The attachment strength of subtidal mussels was measured and compared with littoral mussels and effects of sulphate were investigated in the laboratory. Continuation of the monitoring of subtidal mussel beds that were closed for brown shrimp fishery and mussel seed fishery, was carried out within the Mosselwad program. Used techniques include underwater video recordings and dredge samples for monitoring mussel density and biomass, box core samples for benthic biodiversity and underwater video recordings for determination of epifaunal diversity (Glorius et al., 2014).
a) Technical approach
The investigation of the subtidal areas in the northern German Wadden Sea is mainly based on hydroacoustic methods like sides can sonar and/or single beam echo sounder based seabed discrimination systems (e.g. Echoplus). Higher frequency side scan sonars are operated with a range of 50 m to each side and a survey speed of 4-5 knots. In water depths less than 5 m the side scan fishes are pole mounted on the boat's side or bow rather than being town behind the research vessel. This mode of operation, on one hand, allows an area-wide mapping of the tidal channels and gullies within a reasonable time period. On the other hand, the unambiguous detection of pronounced habitat structures like mussel banks or dense Lanice conchilega meadows is ensured. However, the approach does not allow for the detection of sparsely distributed epibenthos. To verify the hydroacoustic findings van Veen grab and dredge samples were taken at predefined locations. Further ground truthing happens by means of underwater video observations.
In addition to the new area-wide hydroacoustic approach continuous dredge sampling takes place since 2007 (Büttger et. al., 2014). The selection of sampling locations is based on long lasting experiences regarding areas of existing mussel habitats or of the habitat potential of sites.
b) Theoretical approach
In Germany, a three steps approach is used to determine natural habitat types (Schwarzer et al., 2008):
- Identification of „probable“ habitat areas by using general data (satellite images, aerial photographs, bathymetric maps, maps of the seafloor with geomorphological/ sedimentological information, “grey” literature like internal reports, diploma thesis, BSc- and MSc thesis;
- Supplement information which is collected in „a“ by using/generating specific data sets from hydroacoustics (side scan sonar, multibeam echo sounder, subbottom profiler) and optical remote sensing (satellite, airborne). Validation by sediment sampling and analyzing, underwater video observation, diving observation, etc. with geological background;
- Validation of b) by sediment sampling and analyzing, underwater video observation, diving observation, etc. with biological background.
Hydro-acoustic remote-sensing in conjunction with ground-truthing provides a robust approach to characterize subtidal habitats. Recent technological progress of swath bathymetric systems, like multibeam echosounders (MBES) and phase measuring bathymetry systems (PMBS), drastically improve this approach, as they can collect simultaneously high-resolution bathymetry and high-quality backscatter imagery (Bartholomä et al., 2011; Les Bas & Huvenne, 2009; Lurton & Lamarche, 2015).
Within the WIMO project, MBES and side scan sonar (SSS) have been employed and coupled with airborne based technique to ensure the continuity of data between the intertidal and subtidal areas (Capperucci et al., 2013). The Forschungsstelle Küste employs MBES and PMBS. Both technologies optimize the effectiveness of the surveys in very shallow waters and maximize the quality of products. Since the bathymetry is provided by the sonar itself, such systems allow for computing absolute backscatter and comparing morphology with seabed composition (Lurton & Lamarche, 2015). This allows for applying quantitative and objective methods to characterize and segment the seabed, ensuring reproducible results.
Additional information on hard-substrates, reefs and sedimentary structures are collected by parametric subbottom profilers. They are based on low-frequency sound generation by non-linear interaction of two high intensity sound beams at higher frequencies (Wunderlich & Müller, 2003). The resulting signal has a high relative bandwidth, narrow beam profile and no side lobes. This results in a very high spatial resolution information on the lateral and vertical change of sediment type below the seabed in water depths of 2-500 m.
The different acoustic facies are linked to the seabed types from ground-truthing samples collected by means of van Veen grab, box corer, dredge and vibrocorer. Number, position and temporal replications of samples are planned on the basis of hydroacoustic surveys. Samples provide geological information on the seafloor sediments, which are then implemented with the analysis of biotic components carried out by NLWKN and NLPV biologists.
Habitat maps are created by the overlay of multiple segmented layers including abiotic and biotic information, extracted by bathymetry, backscatter intensities and samples. As the conventional interpretation carried out “by eye” involves a high degree of subjectivity, efforts are made to characterize and segment the hydro-acoustic data using objective and reproducible approaches. The analysis of bathymetry is carried out by geomorphometrical approaches, extracting geological and geomorphological information from the morphometric parameters (Jasiewicz et al., 2015). Calibrated and compensated backscatter provides substantial information to characterize the sediment by analysing the backscatter variation with respect to the incident angle (Lurton and Lamarche 2015). The availability of morphological data and backscatter allows the application of the most recent Object Based Image Analysis (OBIA) tools, as very powerful automated methods for the segmentation of acoustic datasets (Ismail et al., 2015).
Monitoring has been carried out using traditional biological methods and different sampling techniques have been applied (Kristensen et al., 2007). Sampling of blue mussels on subtidal beds in the Wadden Sea was carried out with help from local fishermen applying their traditional mussel dredge. For sampling of Pacific oysters an old type oyster dredge was used. All samples have been used to estimate the mussel biomass in the subtidal beds.
Information on the position and activity of the Danish mussel boats were derived by a black box, which logs the position every 10 seconds, and a sensor on the winch registering any activity, e.g., starting or ending of fishing time.