Appendix - Monitoring of biological attributes

Macrofauna

Birds

Fish

Macrofauna

Intertidal mud and sandflats

i. Sediment cores

On intertidal areas core samples using a 0.01m corer can be used to take quantitative samples of the sediment. Generally five replicates are recommended for statistical analysis. Samples should be taken up to 5m either side of the site location but not up or down the shore. Each sample should be placed in a sealable plastic bag which is clearly labelled with a waterproof marker. The samples should be washed through a 0.5mm mesh within 24 hours and fixed in a 10% buffered saline formalin solution for storage after which the samples can be sorted aided by an illuminated magnifier or low power binocular microscope and the animals identified with high power compound microscopes and up to date taxonomic literature. In coarser sediments an additional five 0.01m cores or three 0.1m box cores should be taken depending on the level of silt present. If data is to be collected on the abundance of specific species then prior knowledge on the density and aggregation of the species is necessary to determine the number of samples required. Additional information on the condition of the sediment, anoxic layer and surficial features should be noted. A 1mquadrat should also be used at the site to record numbers of worm casts and algal cover and the area dug over to within 20-30 cm depth to record larger species. 1m quadrats can also be used as part of an ACE survey to determine abundance scales of conspicuous species and epifauna noted over a 10m area followed by site photographs. Species which cannot be identified in the field can be taken for subsequent identification. Information should be noted on an MNCR recording form. Site positions can be recorded using shore transit mark or DGPS as appropriate.

ii. Remote sampling

This technique generally involves all equipment lowered from a boat to take a quantitative sample of the sediment. Samples are taken by grabs e.g. the Van Veen or Day grab. Other equipment includes the Reineck box corer, Knudsen corer (particularly in softer sediments), Hamon or Shipek grabs (in coarser sediments) and Forster anchor dredge (for semi quantitative sampling of mobile megafauna). Site conditions, size of vessel and previous experience will dictate which equipment is most suitable. Details on these and other models are given in Holme (1971). Samples are generally 0.1m and sieved through 0.5 or 1mm mesh depending on substratum and between five and ten replicates should be taken. Notes on the sediment sample should be taken backed up by photographs if possible. For Day and Van veen grabs samples should be at least 7 cm deep in finer sediments and 5cm in sands. After sieving samples should be fixed in a 10% saline formalin solution. Site positions should be recorded by DGPS.

Subtidal mobile sandbanks

i. Towed and remote operated video

The cameras are generally mounted at 45 degrees with video lights and flash strobes pointed vertically downwards. With ROV systems the cameras are mounted on a submersible vehicle which is controlled by a surface operator via an umbilical cable. Towed systems require a vessel which can maintain steerage at low speeds of around 1 to 1.5 knots and the water depth needs to be monitored and the length of cable checked accordingly.

Both systems usually operate along predefined transects along which the seabed characteristics are recorded and the benthic animals counted, still photographs may be taken of interesting features. With towed systems the ships DGPS can give the precise location of the equipment but this is more problematical with remote systems though in some areas it is possible to lay a marked transect line onto the seabed with marker buoys at the surface. ROV systems are hampered by having to remain within a certain radius of the ship but have the advantage that they may hover over a particular area or retrace its path and avoid obstructions, some models can also take benthic samples.

The recording of animals with towed systems usually involve time series counts by which the length of video recording is broken into segments of equal times and either the presence of animals in each segment noted (rapid visual count) or the total number of animals counted (visual fast count). ROV systems may also operate in this fashion or if reliable knowledge on the size of the video frame and a straight transect is possible (allowing the total distance travelled to be calculated) then the number of target species in a known area can be calculated. Problems in the quantification of animals with ROVs may arise because the ROV may not always be a fixed distance from the substratum so the field of view may change.

ii. Acoustic survey

The underlying principle behind these methods is based on the fact that acoustic energy is reflected off the sea floor and the amount of reflection depends on the density of the bed material and the angle of incidence at which the acoustic waves meet the sea floor. It should be remembered that these systems measure the acoustic properties of the sea bed not other properties such as grain size although these are related. The most widely used system available is the RoxAnn system. This utilises commercial echosounders set at frequencies of 200 KHz in waters shallower than 60m or in deeper waters frequencies of 50-120 KHz. The first and second echoes (E1 and E2) are used with E1 used as a measure of roughness i.e. topographic irregularities and E2 as a measure of hardness which is related to substratum type.

Transects across the study area are made at moderate speeds with the data logged as an average over a set time e.g. 5-10 seconds. The two parameters are plotted against each other to derive sediment type and this information together with depth and positional information from DGPS allow a 3 dimensional representation of the seabed to built up and the main substratum types determined. The differences in acoustic properties of the sea floor are related to factors such as grain size which is itself an important factor in determining benthic communities and as such these methods can give an idea of biotope. The benthos itself affects the substratum e.g. through bioturbation, tube building etc so the system can also detect biological features such as Modiolus beds, seagrass, Nephrops burrows, Sabellaria reefs etc. The information gathered from these systems needs to be ground-truthed either by grab sample or towed or remote operated video and post processing of the data may be required to recalibrate the system to get the most accurate representation. Successful studies of candidate SACs have been made using RoxAnnTM e.g. Strangford Lough (Magorrian et al, 1995).

The advantages of systems such as RoxAnnTM is that they can map extremely large areas of substratum fairly quickly with no depth constraints in coastal areas. Other factors such as turbidity are unimportant. With accurate ground-truthing a much more representative idea of the main substratum types in a SAC can be gained than from spot grab samples and the information can be used stratify other direct sampling regimes for an area and reduce the cost and time spent undertaking these surveys. At large areas acoustic methods are probably the only way to effectively map the main biotope complexes. However the system is an indirect representation of the substratum and detailed ground-truthing is required as no direct information is produced on the sediment type or of the detailed composition of the community. The equipment is also expensive an needs a skilled operator to derive accurate results. Care should be taken not to interpret the data at too higher a level of precision as the results may be misleading and the differences between the echosounders and frequencies used may mean that results from different surveys may not be entirely comparable.

Birds

WeBS counts are probably the most effective method. Current WeBS counts are carried out by volunteers and cover the majority of UK wetland sites including estuaries and some open coastal areas. A very large dataset over the last 30 years has been built up for many areas so there is a large amount of existing data available. These counts can be made on a monthly or yearly basis (in the winter) depending on the size of the area, although in extremely large areas e.g. the Wash it may not be possible for financial reasons to sample this regularly.

Bird counts on the intertidal areas are made around high water (usually 2 hours depending on location) with the area divided into sectors and counted separately. These counts give a good indication of usage in general and counting is easy as the birds are pushed to the top of the shore. High water counts are also inexpensive. However depending on the topography of the site and tidal conditions birds can be pushed off sites into different areas at high water and give misleading usage. Low water counts may be used in smaller areas and these give a good indication of preferred feeding areas on the flats not just general usage and identifies which species are using the site as predators rather than roosting. However there is a shorter historical dataset and the method is difficult over large areas. Goss Custard (1985) showed that there was often a good correlation between high and low water counts although this is not true in all areas. In large areas remote sensing techniques although costly can provide useful information on habitat usage and may give initial information on the changes in feeding distribution reflecting changes in sediment type.

Fish

A quantitative survey of the juvenile fish populations using the intertidal areas can be carried out with push nets. Small beam trawls should also be made of the subtidal sand banks whilst minimising any damage to the epibenthic population. The beam trawl will usually be around 2m diameter with tickler chains across the front of the net to disturb fish buried in the substratum. Transects of around 20 minute duration can be made with positions fixed by DGPS. Demersal fish are rarely evenly distributed throughout an area so a random sampling method is usually applicable. With beam trawls it is fairly easy to calculate the absolute abundance per area using the width of the trawl and the distance trawled to determine the area trawled, a catchability component may also be included.

Seasonal components should be taken into account in the sampling strategy and analysis of production and age classes (cohort analysis) carried out. Analysis of the stomach content of the fish population may also prove useful and may dispense with the need to sample the infauna.

References