From Estuary To Open Ocean: Oceanographic Findings From Falmouth
Welcome to Group 6
Gareth Bennett

Madeleine Brasier

Thomas Clarke

Freya Garry

Alex Kesser

Figure 1.1

 

 

Lucy Martin

Louisa Payne

Michael Pownall

Julia Robinson

Suzie Tooke

 

 

Intro To Falmouth: Aims & Schedule

This webpage provides an overview of the scientific procedures and findings undertaken by Group 6 (Figure 1.1) during a field course in the Falmouth Estuary. The field course aim was to gain an overview of the physical chemical and biological systems in the Falmouth area (Figure 2.1). Three boat practicals were undertaken throughout the fortnight, one in the Falmouth Estuary, one offshore and one side scan habitat mapping exercise (Table 2.1). The physical, chemical and biological data collected during the surveys are processed and integrated to give an understanding of the estuarine and offshore systems. Particular aims of each exercise are outlined in the relevant sections below.

The Fal Estuary is an area of high conservation value with important biological habitats, for instance maerl beds. The estuary was formed as the river valley/ria became flooded. As a consequence there is a deep channel in the centre of the estuary. The geology of the area which principally comprises of Carnmellis granite and other metamorphic rocks to the west of the Fal influence the characteristics of the estuary. Due to the high levels of mining in the past, Restronguet Creek -which leads onto the Fal Estuary- has become the most metal polluted estuary in the UK and this has caused contamination of the sediments in other parts of the Fal. The Wheal Jane mine discharge in 1992 is a notable source of contamination, but residual drainage and leakage from spoil heaps lead to contamination of Restronguet and therefore the Fal (Langston et al. 2003).

The Fal Estuary, a busy harbour economically and recreationally, is subject to various pressures from pollution, for instance from organotin contamination originating mainly from Falmouth Dockyard. Industry and dredging, as well as sewage and contamination from tri-butyl tin, are additional anthropogenic factors that affect the physical, chemical and biological characteristics of the estuary. The Fal Estuary is generally high in nutrients, leading to hyper nutrification and risk of eutrophication in parts of the estuary (Langston et al. 2003).

Falmouth is additionally an ideal base from which to conduct offshore surveys in the Western English Channel. The vertical mixing processes in the Channel can be studied whilst aiming to discover how this affects the plankton communities in the region, with particular reference to the change in stratification around tidal fronts.

Figure 2.1

Date Description
28/6/2011 Web Prep am
29/6/2011 Estuarine Boat
30/6/2011 Chem Lab am
Bio Lab pm
1/7/2011 Data Lab
2/7/2011 Offshore Boat
3/7/2011 Catch Up Day
4/7/2011 Bio Lab am
Chem Lab pm
5/7/2011 Data Lab
6/7/2011 Geophys boat
7/7/2011 Geo Data
8/7/2011 Data Lab
9/7/2011 Submit Web page

Table 2.1

Boats & Equipment
R.V Callista R.V Conway L.C. Grey Bear

Figure 3.1

Figure 3.2

Figure 3.3

R.V. Callista (Figure 3.1) is a twin hulled purpose built scientific research vessel owned by the University of Southampton.  It has a large rear deck and 3 separate deployment points with on-board wet and dry labs enabling chemical and physical data to be collected and processed on the water. 

Specifications:

Length: 19.75m
Breadth: 7.40m
Draft: 1.80m
Depth Midship: 2.85m
Max Speed: 15 knots
Range: 400 Nautical Miles
Max Passengers: 30 + 4 crew

Equipment:

1 ‘A’ frame with 4 tonne winch at stern.
2 Side mounted Davits with 100kg capacity hand winch.

1 Capstan 1.5 tonne pull

R.V. Bill Conway (Figure 3.2) is a small purpose built scientific vessel owned by the University of Southampton. In the cabin there is a small lab bench and a covered area of the rear deck where data processing can take place. 

Specifications:

Length: 11.74m
Breadth: 3.96m
Draft: 1.30m
Depth Midship: 2.85m
Max Speed: 10 knots
Range: 150 Nautical Miles

Max Passengers: 12 + 2 crew

Equipment:

1 ‘A’ frame with 750kg winch.
2 Davits with 50kg capacity

Grey Bear (Figure 3.3) is a multipurpose, shallow draft vessel owned by FD Marine Ltd.  With its very large deck area it is primarily used as a landing craft vessel.  It is able to go into shallower waters than most vessels and so is highly suited for water front construction projects, cable laying, salvage work and surveying.

Specifications:

Length: 15m
Breadth: 6.1m
Draft: 1.14m
Max Speed: 7.5 knots
Max Passengers: 12 + 2 crew

Equipment:

1 port side Hiab 1250 crane fitted with a 2.5 tonne winch.
1 starboard side HS Marine AK10 crane fitted with a 1.25 tonne winch.
1 stern side 3 tonne winch with roller.

2 Deck winches with 5 tonne capacity

CTD FLUOROMETER TRANSMISSOMETER ADCP

Figure 3.4

Figure 3.5

The CTD (Figure 3.4) is deployed from the deck and measures three vital physical parameters; conductivity, temperature and depth. Salinity is then derived from the known relationships of pressure, temperature and conductivity. Other parameters may also be measured by instruments attached to the CTD, and sampling bottles may be attached using a rosette system.
The CTD is attached to the vessel by a conducting cable and data are electronically uploaded to the vessel in situ allowing scientists to sample based on data.
A fluorometer was deployed on the CTD rosette system. It emits light of a certain wavelength and records the amount of light returned as a result of fluorescence. It is therefore used to measure chlorophyll as light excites the fluorophores. The data can then be used as an indication of the phytoplankton biomass. A transmissometer probe measures the amount of suspended or particulate matter in the water column by the measurement of attenuation of a laser beam (usually 660nm) through a known volume. This provides a measurement of turbidity in NTU. The Acoustic Doppler Current Profilers aboard Bill Conway and Callista measure the current speeds throughout the water column below the vessel. An ADCP uses 3 or 4 sonar beams to measure any non-perpendicular current. The acoustic Doppler shift reflected by the sediment is removed from the Doppler water shift to give an accurate reading of current speed and direction. The ADCP also provides an insitu measurement of the backscatter which can be used to identify structure in the water column and provides an indication of zooplankton populations. The output data from the ADCP is displayed on a screen as shown in Figure 3.5.
SECCHI DISK

NISKIN BOTTLES

HORIZONTAL NISKIN BOTTLE

YSI PROBE

Figure 3.6

Figure 3.7

Figure 3.8

Figure 3.9

A Secchi disk (Figure 3.6) is used to calculate the approximate depth of the euphotic zone.  The disk is circular with black and white segments. It is lowered through the water column at 1 metre intervals, until the disk can no longer be seen - this is known as the Secchi depth.  The depth of the euphotic zone is calculated as 3 times the Secchi depth. Water sampling bottles such as the Niskin bottle (Figure 3.7) are deployed on a rosette (often in conjunction with a CTD) and are used to collect samples at a range of depths. The bottles are deployed open so water can flow through them reducing contamination such as biofilms and ensuring an accurate representation of the water column. A horizontal Niskin bottle (Figure 3.8) was used for collecting water samples on the pontoon in the Fal Estuary. It has a unique end stopper release mechanism which allows the sampling bottle to be used in a horizontal position. A multiprobe (Figure 3.9) measures a range of parameters including salinity, temperature (oC) and depth (m). It may also possess probes that measure pH, turbidity (NTU) and chlorophyll (mg g-1). Readings are electronically transported from depth through a conducting cable to the vessel's data systems or to a hand held logger.
CLOSING NET VAN VEEN GRAB SIDE SCAN TOW FISH VIDEO CAMERA

Figure 3.10

Figure 3.11

Figure 3.12

Figure 3.13

A closing net (Figure 3.10) allows specific sampling between two defined depths. It is made up of a 200µm mesh with a screw on a vented collecting bottle. A weight keeps the net vertical in the water column. After hauling the net from the lower to the upper depth the messenger weight is used to trip catch the primary haul lines which transfer the load to a draw string causing the closure of the net. This prevents contamination of the sample during the recovery to the surface. The Van Veen grab (Figure 3.11) is designed to take large sediment samples and associated biota from areas of soft sediment.  The grab itself consists of two stainless steel, weighted, sharp edged scoops positioned to act like jaws. On the top of the scoops there are 4 lids which when opened allow subsampling of the undisturbed sample in order to view stratification of the sediment of the sediment in its original setup before the grab is emptied and the sediment mixes. The Towfish (Figure 3.12) is the housing for the side scan sonar transducer, which is towed through the water behind the vessel. By towing the transducer the sonic wave emissions can be emitted closer to the seafloor. When the emitted acoustic signal reaches the seafloor it is reflected back to the Towfish where it is received, the time elapsed between emission and reception of the signal allows for the determination of the depth, the images created using the sonar are based upon the reflectivity of the sediments, as different forms of material provide different ranges of reflection. An underwater camera (Figure 3.13) allows a glimpse of the seafloor through optical imaging. Deployed from the vessel over the side, the camera provides real time images of the seabed; this allows us to view the ecosystem change beneath the boat. Submersible video cameras are usually attached to an object that can be controlled from the boat, and also a cable that is attached to a small monitor.

SIEVES

SPECTROPHOTOMETER

WINKLER APPARATUS

Figure 3.14

Figure 3.15

Figure 3.16

Sieves (Figure 3.14) are used to separate samples by their size. From course to fine grained material, the size of the mesh is used to class what is caught at that sieves size.

 

 

A spectrophotometer (Figure 3.15) consists of two instruments, a spectrometer to produce light of any selected wavelength and a photometer to measure the intensity of light.  The instrument is arranged to allow a cuvette of a liquid sample to be placed between the spectrometer beam and the photometer. The light from the spectrometer as it passes through the cuvette is measured by the photometer, which is shown on the screen. The signal on the screen changes as the amount of light absorbed by the liquid changes. The winkler apparatus (Figure 3.16) is used to measure the concentration of dissolved oxygen. This is done by titration; sodium thiosulphate is added to the sample in the bottle until the solution becomes clear.
Lab Methods PhosphateSilicateDissolved O2ChlorophyllZooplanktonPhytoplankton
Phosphate

1.   1. Preparation of phosphate working standard   The standard must be freshly prepared. The stock solution was diluted 400 fold by taking 1ml and made up to 100ml with MQ water in a volumetric flask. Then 25ml was separated and made up to 100ml with MQ water, giving a solution containing 15µmol per litre which is used to prepare the calibration standards.

2.   2. Preparation of Calibration and Blank solutions    Using a 5ml hand pipette, 10ml of MQ water was carefully added to three tubes labelled as blanks. Using appropriate pipettes, 50µl, 100µl, 200µl, 500µl, 1000µl and 2000µl was added each to 3 separate sample tubes, giving 21 test tubes. All tubes were made up to 10ml using MQ water. These tubes were treated the same as the sample tubes and then the analytical methods below were followed. These calibration tubes will have the following phosphate concentrations as in Table 4.1.

Addition volume (µl)

50

100

200

500

1000

2000

Phosphate conc. (µmol l-1)

0.07

0.15

0.3

0.75

1.5

3.0

Table 4.1

Fresh mixed reagent was prepared by combining  20ml Ammonium Molybolate, 50ml Sulphuric Acid (2.5M), 20ml Ascorbic Acid, 10ml Potassium antimonyl tartrate producing 100ml total. 1ml mixed reagent was added to every sample tube (blanks, calibration tubes and samples alike) then mixed well and left for 1 hour.
Then the spectrophotometer set at 840nm was used to determine the absorbance of each sample, calibration and blank tube. The calibration and blank tubes are used to calculated the phosphate concentration of the sample tubes based on the measured absorbencies.

Silicate

The laboratory analysis to determine the amount of silicon present in each sample was carried out as in Mullin & Riley (1955); a slight modification was that the river end member samples were diluted by a factor of 5. A sub-set of standard samples were used to perform a calibration using silicon solutions of known concentrations, which were also placed into the spectrophotometer and their absorbencies measured.

1. Standard preparation    A stock silicon solution of 35.6mmol l-1 was diluted using a 25 times dilution to create the working standard solution.
Further dilution was then carried out to create the following standards:
1.4μmol l-1 , 2.8μmol l-1 , 7.1μmol l-1  , 14.2μmol l-1, 21.4μmol l-1.

2. Sample preparation      2ml of molybdate reagent was added to 5ml of each silicon sample and left to stand for 10 minutes. 2ml of molybdate was also added to 5ml of MQ water to be used as blanks. A mixed reducing agent (MRR) was prepared by mixing metol sulphite, oxalic acid, sulphuric acid and MQ water. 3ml of the MRR was added to all samples, standards and blanks, and left to stand for 2hours. The absorbance of the samples were measured on a spectrophotometer (U1800) at a wavelength of 810nm, a 4cm cell was used.

Dissolved Oxygen - Winkler Titration

Glass bottles are used for dissolved oxygen measurements in order to prevent contamination.  Samples were taken from Niskin bottles before any other samples were taken to avoid contamination, and were siphoned through rubber tubing with a tight seal on the valve. Air bubbles were removed from the tubing and the glass bottles were filled to a third and rinsed then emptied to clean which avoids contamination. The bottles were then filled to overflowing to avoid trapping any air within the bottle. Winkler reagents were added to the sample; 1ml manganous chloride and 1ml alkaline iodide solution using a pipette. This fixes the oxygen in the bottles, which are then stored in water to prevent leaching from air through the stopper until lab analysis.

In the lab, 1-2ml of sulphuric acid 10M was added to and mixed in each bottle to release the oxygen, turning the solution to a clear yellow. The bottles were placed into the end point detector and sodium thiosulphate (normality 0.22) was added using a metrohm device until the solution began to clear. Further additions were monitored on the Servoscribe 1s until the needle ceased movement and plateaued. The metrohm reading was recorded. As the amount of sodium thiosulphate added to the sample was known, the dissolved oxygen concentration could therefore be calculated.

Chlorophyll

Chlorophyll a concentration is currently the best index for estimating phytoplankton biomass (Huot et al., 2007) although the fluorometry readings obtained in the fieldcourse did not correlate very well to the CTD fluorometer readings; the CTD readings should be assumed to be the more accurate reflection of chlorophyll. In order to chemically analyse water samples for chlorophyll however, seawater was filtered into glass and plastic bottles for oxygen and nutrient analysis respectively on the boat. The porous glass fibre filters that the water was filtered through was then stored in 90% acetone; the acetone acts as a solvent and extracts the chlorophyll. The test tubes were frozen overnight. Using a fluorometer in the laboratory the following day, the fluorescent properties of the chlorophyll pigment in the acetone were measured and determine the amount of chlorophyll present using the following equation:

Chlorophyll (µg l-1) = ( Volume acetone / Volume seawater filtrate ) * Fluorometer Reading

As there were two filters for each water sample taken, the results gained in the lab could be compared to give a reasonable indication of the chlorophyll levels in the sample generally.

Zooplankton

Formalin was initially added to each 500ml zooplankton sample bottle in order to preserve the zooplankton. In the laboratory after mixing the sample by simply inverting the bottles, 10ml of the sample was pipetted into a measuring cylinder.  Then 5ml of this sub-sample was pipetted into a Borgorov chamber and viewed under a light microscope, and each zooplankton was identified with the aid of photo guide books.  The individual organisms were tallied into a table under the major taxa groups found in the area: Copepoda, Copepoda Nauplii, Cladocera, Mysidacea, Chaetognatha, Hydromedusae, Siphonophorae, Ctenophora, Appendicularia, Decopoda, Cirripedia, Polychaeta, Gastropod, Echinoderm and fish larvae. 

Phytoplankton

Water samples were added to bottles containing Lugols solution which preserved the phytoplankton.  100ml of solution was then left in a settling column overnight to allow the phytoplankton to settle.  In the laboratory the top 90ml of each sample was removed with a vacuum pump to leave a concentrated sample.  From this 10ml sample a smaller subsample of 1ml was taken and added to a Sedgewick-Rafter Counting Chamber which was placed under an optical microscope to be viewed. 5 vertical transects of 20 1μl squares were viewed with the number of cells in each square being recorded and the different species were identified and logged.

 
 
EstuarinePhysicalLight AnalysisPhosphate AnalysisSilicate AnalysisO2 AnalysisPlankton AnalysisChlorophyll AnalysisConclusion

Introduction


Physical, chemical and biological data were collected in the Fal Estuary (Figure 5.1) in order to establish an understanding of how the Fal Estuary behaves. As an estuary, the area is a transition zone between the freshwater input of the rivers into the estuary and the coastal sea, and therefore can be expected to have differing characteristics to the coastal sea observed in the offshore practical. The Fal Estuary is a high nutrient region subject to various pollution pressures.

Physical data provide an indication of the physical structure of the Fal Estuary and provide information on whether the estuary is mixed, partially mixed or stratified. The data collected throughout the estuary will provide an overview of how the physical structure changes throughout the estuary and enable examination of how the tides affect the conditions in the estuary. The chemical data collected will give an indication of the behaviour of nutrients in the estuary. Studying the biological characteristics in conjunction with the rest of the data will be used to indicate what is controlling or limiting the phytoplankton in the estuary. The data can be compared to data collected in the offshore practical.

Data on the estuary were collected in two ways in order to determine how the physical, chemical and biological characteristics of the estuary vary temporally and spatially. Lagrangian measurements were collected using the R.V. Bill Conway along the Fal Estuary commencing up the estuary at low water in the morning and travelling down the estuary throughout the day during the flood and ebb tidal cycle. Physical data were collected using a CTD to measure the temperature, salinity, turbidity and chlorophyll, and an ADCP system to collect data on the flow direction and velocity. Water samples were collected via Niskin bottles mounted on a rosette. Water samples were filtered and then used in the laboratory to determine the concentrations of phosphate, silicon, dissolved oxygen and chlorophyll as outlined in the lab techniques section. Phytoplankton samples were also collected from the Niskin bottles. Zooplankton was collected using a 210 micron net and towed for 5 minutes at selected stations. A Secchi Disk  was used to estimate the depth of the euphotic zone and is used in conjunction with the transmissometer on the CTD which indicates the light attenuation of the water column.

Eulerian measurements were taken at the King Harry pontoon, using an YSI multiprobe to measure the salinity temperature, chlorophyll, turbidity and pH through vertical profile at a fixed point on the estuary.  A Niskin bottle was also used at the surface to measure the same biological properties as aboard the R.V. Bill Conway. The data from the pontoon were used to observe the changes in the properties over time and the tidal cycle.

The data were collected in conjunction with Group 11: Group 6 collected boat data in the morning, whilst group 11 were on the pontoon. Group 6 then took over measurements at the pontoon at 12:00 GMT whilst group 11 continued collecting data on the boat further down the estuary. Ancillary data for the survey can be found in Tables 5.1 and 5.2.

Date 29/6/2011
General Weather Sunny with some cloud cover
Visibility Good, clear day
Sea State Very calm
Cloud Cover 3/8 in the morning to 2/8 in the afternoon

Table 5.1

29/6/2011 Tide Times GMT Tidal Height (m)
Low Water

0952

1.5

High Water 1548 4.8
Low Water 2221 1.5

Table 5.2

Figure 5.1

Physical Characteristics


The sampled area of the Fal estuary is well mixed. The lowest salinity was sampled at 25 at the furthest point up the estuary that could be sampled and the highest salinity measured was 34 towards the mouth of the estuary. The top part of the estuary sampled by group 6 had the biggest change of salinity of 8 found mostly at the station 1 and 2; salinity at station 3 and 4 is around 33. This compared with the bottom of estuary sampled by group 11 having a change of 2 where surface salinity stays around 35 from stations 2-5.

Surface temperature down the estuary changes from 15.7°C to 13.4°C at the last station (Figure 5.8). The top of the estuary sees little change in surface temperature of 0.4°C whereas the bottom of the estuary shows a change from 13.7 to 13.4°C. The largest change in temperature change is at group 6 station 3 (Figure 5.7) where in 8 meters the temperature goes from 15.2 to 13.5°C – this stratification could be due to intense heating together with this region not being as well mixed. Little change in temperature at bottom of the estuary could be due to incoming tide going up the estuary and mixing the surface water at the lower end of the estuary (particularly as this station was sampled at high water).

Fluorescence, used to indicate chlorophyll levels in the estuary, varies from surface values of 0.3 at the top of the estuary to 0.13 volts at the bottom of the estuary. Fluorescence stays relatively low at the bottom of the estuary with the maximum value being 0.1 at the surface. At the top of the estuary there is a peak of 0.6 volts at station 1 (Figure 5.2) and then fluorescence stays fairly constant around 0.4 volts. This indicates that there might be higher phytoplankton populations near the top of the estuary.

Figure 5.2 - Station 1

Figure 5.3 - Station 2

Figure 5.4 - Station 3

Figure 5.6 - Group 11 Station 1

Figure 5.7 - Group 11 Station 3

Figure 5.8 - Group 11 Station 5

Richardsons number calculations (see offshore) of Group 6 station 2 and Group 11 station 5 located up-estuary and at the estuary mouth respectively are shown in (Figure 5.9).  At station 2 low Ri numbers describe turbulence throughout the water column, except at 4m where Ri=0.252. This corresponds with the depth of the surface warmed water boundary as seen in Figure 5.3. On a large temporal scale, for example offshore, a high Ri number depicting laminar flow is predictable at such a boundary. However, stability might not be maintained at station 2 in a tide dominated estuary such as the Fal. Here, some gravitational stability has been reached, however shear instability prevails over laminar flow (with Ri>1).

At Group 11 station 5 (Figure 5.8) near the mouth of the estuary there is no synonymous warmed surface water layer and values of Ri<1/4 describe turbulent flow in the entire sampled water column. At the time of sampling the water depth was twice as deep at Group11 station 5 when compared to station 2. This is due to a deeper channel and because this station survey, at 15:34 occurred 14 minutes before high water making the water column ~3.1m higher than the low water height when station 2 was sampled. Both vertical, horizontal, wind and tidal mixing occurs at the estuary mouth which is in close proximity to open ocean.

Figure 5.9

Pontoon

EAST

Temperature

Solar surface warming was evident from 09:30GMT throughout the day heating up the surface layer. The warming effect was countered by the rising tide which drew cooler seawater up into the estuary. From low water (09:52GMT) to the end of sampling at 16:00GMT the influx of seawater into the estuary cooled the deep waters and created a significant level of stratification in the water column, limiting solar heating to the uppermost 0.5m. As the tide rose the cool deep seawater layer moved shallower due to the volume of seawater being forced up the estuary displacing the freshwater; this can be seen on the contour plot (Figure 5.10) as a region of cooler water below 1.5m. This seawater intrusion can be identified on the salinity contour plot (see below). At high water (15:48GMT) the cool seawater layer dominates the water column starting at 1.5m and extending to the bed.

Salinity The temporal proximity of the morning samples to low water resulted in a region of high salinity in water below 4m; this is due to the tidal forcing drawing the denser seawater back seaward from the estuary below the freshwater as the tide fell. Salinity was lower in the surface layer throughout the sampling time due to freshwater laying above the denser seawater though the level of stratification varied across the tidal cycle. At 14:30GMT a protrusion of saline water pushed through the layer of freshwater at depth 1m. This may have been linked with weakening tidal forcing as the tidal cycle approached high water. A salinity hotspot is seen in Figure 5.11 - this is due to sigmaplot extrapolation and so ignored in analysis.
Chlorophyll Figure 5.12 shows a high concentration of chlorophyll measured at approximately 2m for the majority of the tidal cycle. At 13:00GMT the chlorophyll maximum moved upwards to 1m. At the end of sampling (16:00GMT) a chlorophyll maximum was measured deeper than before below 4m. This is caused by the drawing in of phytoplankton with the seawater as the tide rises to high water.
Turbidity

Figure 5.13 shows that there was a high level of turbidity in the bottom water layer due to turbulence suspending silt and detritus from the bed. The extreme variations between measurements taken before and after 12:00GMT may indicate the different depths of water that group 6 and group 11 were sampling in. Group 11 were sampling in shallower water due to the tidal cycle which meant that more of the water column was being influenced by turbulence with the bed and resulting resuspension of sediment. However group 6 were sampling closer to high water when the water column was deeper and therefore turbulence with the seabed was less significant.

Figure 5.10 Temperature

Figure 5.11 Salinity

Figure 5.12 Chlorophyll

Figure 5.13 Turbidity

WEST

Temperature Figure 5.14 shows solar surface warming was evident from 09:30GMT throughout the day heating up the surface layer. As the tide rose the cold deep seawater layer moved progressively shallower until at 14:00GMT it displaced the freshwater at the surface leading to a cold surface layer existing for a short period of approximately one hour. Simultaneously the warm freshwater surface layers are pushed up the estuary by tidal forcing as the tide rises.
Salinity Figure 5.15 shows that at low water (09:52GMT) river flow was dominant and so the salinity was low. As the tide rises tidal forcing pushes seawater up the estuary and this can be seen on the contour plot as a region of high salinity. There was a lag between the turning of the tide at 09:52GMT and the rise in salinity; this may have been due to the counteracting force of river water coming down the estuary.
Chlorophyll Figure 5.16 shows that chlorophyll concentrations increase at approximately 1m throughout the period of sampling. A maximum was measured at 2.5m and below from 09:00GMT till 13:00GMT. This was caused by the falling tide in the first 52 minutes of sampling pulling chlorophyll rich water down from upstream. A second maximum was observed at 14:00GMT where tidal forcing pushed chlorophyll rich seawater up the estuary beneath the freshwater.
Turbidity Figure 5.17 shows similar trends were observed as on the east side of the pontoon for turbidity.

Figure 5.14 Temperature

Figure 5.15 Salinity

Figure 5.16 Chlorophyll

Figure 5.17 Turbidity

 

Variation in East and West measurements from the pontoon

The measurements taken using the YSI probe on the East and West side of the pontoon show variations in data collected, despite their close spatial proximity.  It is important to analyse the differences between these two sites before any assumptions are made about the relationship between measurements made on the pontoon and on RV Bill Conway. Variations on this small spatial scale can be used to assess the significance of variations seen between stations sampled on RV Bill Conway. The East and West sides of the pontoon experience differing levels of disturbance from passing vessels: the East side of the pontoon being disturbed and mixed by ferries every 15 minutes or so whereas the West lies relatively undisturbed which may have lead to the stabilisation of water layers that were unable to develop on the East side. The two sides may have been affected in different ways by tidal forcing due to the interruption of the tidal flow by the pontoon structure. There was also the issue of depth; the West side of the pontoon was approximately 2m shallower than the East side so the development of layers in the column would be affected.

 

Light Analysis


The Secchi disk measurements (Figure 5.18) show a clear trend between the upper sampling stations and the lower stations. The shallowest Secchi depth was measured at station 1 with a depth of 1.26m and the deepest at Group 11’s station 5 with a depth of 7.5m which partly reflects the deepening of the natural river channel and hence the varying proximity of the euphotic zone to the sediment of the seabed. Group 11's station 5 was sampled 14 minutes before high water therefore the Secchi depth is deeper due to reduced bottom turbidity, as highlighted by the pontoon time series. There is a very slight decline in depth between station 3 and 4 of 0.1m but all other measurements show an increasing trend which reflects a larger euphotic zone (recognised as approximately three times the Secchi disk depth unless this exceeds the depth of the bed).

The attenuation coefficient, k, is an indication of the rate at which light is absorbed within the water column. The highest k value and therefore fastest attenuation rate is shown at station 1 with a value of 1.14 (Table 5.3). There is then a general decline down the estuary towards the mouth, with the lowest value at group 11 station 5 with a k value of 0.19 indicating a deeper euphotic zone.

Station Number Secchi Depth (m) k - Attenuation Coefficient
Station 1 1.26 1.14
Station 2 1.57 0.92
Station 3 2.54 0.57
Station 4 2.40 0.60
Group 11, Station 1 4.00 0.36
Group 11, Station 2 No Data
Group 11, Station 3 5.00 0.29
Group 11, Station 4 6.30 0.23
Group 11, Station 5 7.50 0.19

Table 5.3

Figure 5.18

 

Phosphate Analysis


Figure 5.19

The mixing diagram is a diagram which shows the concentration of a solute against an index of conservative mixing in the estuary - salinity is commonly used.  The diagram can be used to understand the behaviour of solutes, in this case phosphate, as they flow through the estuary.

The four assumptions for estuarine mixing diagrams were met: the constituents were assumed to be at steady state, the end member concentrations are constant on a timescale somewhat greater than the residence time of the estuary, there is only one riverine and one end member and there were no additions of pore waters.

Samples taken at stations up the estuary were taken over a relatively narrow salinity range as this was the extent of the estuary that could be surveyed in the R.V. Bill Conway; the low number of riverine inputs mean that the estuary is relatively saline for quite some distance up the estuary. Freshwater riverine end members were obtained by staff members separately.

The theoretical dilution line in the graph (Figure 5.19) between the riverine and saline end members represents the conservative behaviour of solutes - changes in concentration are solely due to mixing. If there is deviation of the data points away from the TDL addition or removal of the solute is indicated. The graph shows that there is a dramatic increase above the TDL. This demonstrates that phosphate is behaving non-conservatively  and is being added to the water column. This could be a result of possible anthropogenic factors such as sewerage outfalls and agricultural inputs further up the estuary. A mussel farm near King Harry pontoon may also cause a change in levels of dissolved phosphate in the water column. Mussels have been farmed near the King Harry Pontoon and close to station 3 by West Country Mussels of Falmouth since 1993 (accessed: www.westcountrymussels.co.uk, 7th July 2011).

An example of an input of phosphate in the Fal Estuary is a survey of the mussel farm by Envirogene which used DNA to establish that there was persistent contamination of the waters by human and bovine faeces (there are a large number of cattle/dairy farms in the area). In addition significant inputs of human faecal matter were detected entering the Fal at the King Harry Ferry.                                                         (accessed: www.envirogene.co.uk/downloads/casestudies/envirogene_case_fal.pdf , 7th July 2011).

Figure 5.20

Phosphate Time Series at King Harry Pontoon

Low tide was at 09:52GMT and salinity was at its lowest following this because the influence of salt water was lowest at low tide, although a small tidal lag was observed. With the flood tide, the salinity began to increase again with a peak around high tide (Figure 5.20).

Phosphate concentration increased from low tide reaching a peak at 13:00GMT of 5.9µm. From low tide to high tide the increase in phosphate could be due to nutrient inputs further downstream being washed up the river. Nutrient inputs could be due run off from agricultural practices in the region and sewage treatment farms. Phosphate in the Fal Estuary is also known to be discharged from mine drainage sites, notably in the Carnon Valley.

  
Dissolved Silicon Analysis

Estuary

The estuarine mixing diagram (Figure 5.21) shows that dissolved silicon was behaving conservatively between the salinity 0 and 25; this suggests that the extent of mixing between these two points was the determining factor for the concentration of dissolved silicon at this time. This observed conservative behaviour does not guarantee that no biological uptake was occurring; there could have been a low rate of uptake which was not observable as non- conservative behaviour. This slow removal could be due to turbidity caused by the heavy rainfall during June 2011, this could affect the growth of diatoms; such effects have been shown in the River Zaire (Cadee 1978). Another suggestion could be made that the behaviour is conservative but has been represented as slightly non-conservative due to long term temporal changes of dissolved silicon concentrations at one of the end member. If this change occurs on a different scale to the residence time of the estuary then a conservative solute can be plotted as non-conservative on estuarine mixing diagrams (Lorder & Reichard 1981).

There may, however, be some removal of dissolved silicon between salinities 30.2 and 34.8 indicated by the slight bowing of the data points beneath the Theoretical Dilution Line (TDL). Biotic uptake by siliceous diatoms may be the reason for this due to late spring phytoplankton growth.

 

Figure 5.21

Pontoon

As observed in Figure 5.15 the salinity decreased an hour after low water, when the silicon concentration was at a maximum of 1.2µ mol l-1. When salinity was highest at 15:00GMT the silicon had a minimum concentration of 0.4 µ mol l-1 (Figure 5.23). Figure 5.22 demonstrates that silicon was behaving conservatively over the tidal cycle, illustrated by the scattering around the theoretical dilution line. Therefore the silicon concentrations observed at the pontoon were a product of tidal mixing.

Figure 5.22

Figure 5.23

Dissolved Oxygen Analysis


Vertical profiles of oxygen saturation from stations 1-3 are shown in Figure 5.26. Changes in oxygen saturation with depth may have been due to zooplankton respiration. For example at station 1 there was a decline in saturation at 1.8m, after which there was a gradual increase with depth. At station 2 there was a decline in the surface saturation from station 1 from 92.4% to 77.8%; this could have been due to an influx of deoxygenated water or substance with a high biological demand. Saturation increased with depth but then declined again at 4m. This could be due to respiration or decomposition. Station 3, down river of the mussel farm, had a surface saturation of 79.6%. Souchu et al (2001) demonstrate how mussel farming can affect the chemical properties of nearby water bodies.

The oxygen saturation of the surface water is shown in Figure 5.24; this graph combines data from our 3 stations and group 11’s 5 stations. The overall trend of surface oxygen saturation showed an increase from the upper estuary at station 1 to the lower stations sampled by group 11. There was a decline at stations 2 and 3, the highest surface saturations were Group 11’s station 2 with 106.1% and station 5 with 104.2%.

The oxygen saturation over the tidal cycle was shown in Figure 5.25. The data were collected at the King Harry Pontoon from 09:00 to 14:00GMT. There is a distinct change in oxygen saturation over the tidal cycle. The highest saturation (102.4%) at 10:00 GMT corresponded with low water at 09:52. There was a decline in saturation until 13:00GMT after which the saturation increased again.

Figure 5.24

Figure 5.25

Figure 5.26

Phytoplankton & Zooplankton Analysis


Estuary

Phytoplankton

Phytoplankton abundance in the Fal estuary was estimated by taking four samples throughout the estuary. Group 6 collected samples at station 1 and station 3 at the top of the estuary. Group 11 collected their samples at group 11 station 1 and station 5 near Black Rock.

It was found that the highest abundance of phytoplankton was at station 1 (Figure 5.27) which corresponds with the zooplankton data, which show a high number of copepods. At station 1 the most abundant phytoplankton was the Chaetoceros genus as they were the largest percentage (47%) of the sample at station one. Chaetoceros. spp are planktonic diatoms, and would use the dissolved silicon within the water to build their frustules; this could cause dissolved silicon to be lower in areas of high Chaetoceros abundance. This can be seen in Figure 5.21, which shows that dissolved silicon is being removed within the estuary at salinities above 30. This was the area where large numbers of diatoms were recorded.  The smallest sample was at station 3; this could be due to the nearby mussel farm. Studies have shown a relationship between mussel aquaculture growth and a decline in nearby phytoplankton numbers (H.F. Kasper, 1995). This is due to feeding processes of the mussel; the water and the organisms contained within it are passed through siphon and particles are trapped within the gills.

Station 5 was the last station for phytoplankton sampling, and is near the mouth of the estuary. The percentage of the overall sample was quite varied at station 5 where many genera were abundant, for example Thalassiosira comprised 17% of the phytoplankton population.

Figure 5.27

Zooplankton

Zooplankton abundance was estimated by sampling at various stations throughout the estuary.  Group 6 collected samples at stations 1 and 4 located at the head of the estuary.  From these stations the highest abundance of zooplankton was at station 1 with the most abundant type being holoplanktonic copepods which made up 45% of the sample (Figure 5.28).  This corresponds with the phytoplankton data collected at the same station which show the highest abundance sampled, the most abundant of which was Chaetoceros curvisetus, a centric diatom.  The lowest zooplankton abundance was found at station 4 which was located near to the mouth of the Truro River, close to where the Channals and Tolcarne Creeks enter the river.  Again the most dominant zooplankton type were copepods however the low abundance could be the result of the station being downriver from a mussel farm meaning there were less nutrients available and fewer phytoplankton.

Figure 5.28

Pontoon

Phytoplankton

From 09:00 to 11:00GMT the phytoplankton samples were dominated by Pyrophacus, a small thecate dinoflagellate (Figure 5.29).  After 12:00GMT the samples' compositions appeared to vary with the dominant genera becoming diatoms, the prevalent genera being chaetoceros. This change in overall composition from dinoflagellates to diatoms could be attributed to the change in the tidal cycle. It could be suggested that the rising tide pushed the dinoflagellates up the estuary. However, it could also be suggested that the dinoflagellates moved deeper in the water column so would not have been sampled as all water samples were taken from the surface. It is also important to note that the Pyrophacus may have been misidentified in the laboratory as small thecate dinoflagellates are often difficult to differentiate from each other.

 

Figure 5.29

Chlorophyll Analysis


The chemical chlorophyll analysis conducted from water samples in the lab shows a trend from low chlorophyll concentrations (< 2 µg l-1 ) in the uppermost estuary increasing to a maximum at station 2 (> 6 µg l-1) (Figure 5.30). The concentrations begin to decrease to ~ 4 µg l-1 at station 3 and ~ 3 µg l-1 at station 4. In the lower estuary concentrations are lower at Group 11 station 1 < 1 µg l-1, increasing to ~ 2 µg l-1 at Group 11 station 2 and ~ 3µg l-1 at Group 11 stations 3-5. The sampling was taken from low tide at the top of the estuary to high tide at group 11 sampling stations.

The chlorophyll temporal analysis at the pontoon (Figure 5.31) shows a relationship between the chlorophyll concentrations and the tidal cycle as the chlorophyll decreases at the tide ebbs and increases as the tide floods. There is a slight tidal lag but a clear relationship as more phytoplankton were being washed up the estuary with the tide.

Calibration of the chlorophyll data plotted from laboratory analysis against the voltage gained by the fluorometer on the CTD shows an unclear and weak relationship (Figure 5.32), but not enough to develop a clear relationship between the voltages and the chemical chlorophyll analysis. The chlorophyll laboratory equipment has provided some inconsistent results, but the trends observed and described above should still show relative patterns, though the exact Figures might not be reliable. 

The chlorophyll analysis does not correspond very well to the phytoplankton count analysis; for example the phytoplankton count demonstrated a higher number of phytoplankton at station 1 compared with station 3 when the chlorophyll analysis suggests that in fact at station 3 the chlorophyll concentrations and therefore the phytoplankton populations may have been twice as great. This might be partially explained by the differences in chlorophyll concentration in different phytoplankton cells, and therefore the most numerous species not necessarily being the most chlorophyll rich. In addition there are inconsistencies with chlorophyll data and replicates which is becoming apparent as a general trend across all groups and may be due to methodology or equipment error. The oxygen data however corresponds well as there are lower dissolved oxygen concentrations at Group 6 stations 1 and 2 – the highest concentration chlorophyll stations. This is an expected finding as chlorophyll concentrations indicate a higher population of phytoplankton which deplete the oxygen concentrations due to photosynthesis. Nutrient analysis shows addition of phosphate, but no depletion of either phosphate or silicon. Therefore the chlorophyll concentrations in the estuary do not indicate a phytoplankton population of enough significance to cause nutrient removal, though any removal may be countered by agricultural/sewage inputs in the case of phosphate (discussed above).

Figure 5.30

Figure 5.31

Figure 5.32

Conclusion

The data obtained in the estuary demonstrate the characteristics of the Fal Estuary at the time of sampling and reflects expected characteristics for a temperate estuary such as the Fal. The sampling up the estuary reflected a greater freshwater influence at the top compared to towards the mouth, but the salinity variation was not large due to freshwater inputs not being sizable in the region. Warmer temperatures were sampled at the top of the estuary than at the bottom, but the tide was low when sampling the shallow top of the estuary so the water column would have been much easier to warm than near the deeper mouth at high tide. Greater phytoplankton populations are demonstrated at the top of the estuary, which appear to support greater zooplankton populations near the top. The high nutrients are not utilised to a degree that indicates significant removal and indeed phosphate addition was observed due to high agricultural, sewerage and mine drainage inputs into the Fal. The high productivity near the top of the estuary has depleted the dissolved oxygen relative to the bottom of the estuary where values measured are much higher. Particularly at station 3 dissolved oxygen was depleted but the phytoplankton also remained low; station 3 was taken by a mussel farm which may cause the oxygen depletion and also restrict the phytoplankton population.

 

OffshorePhysicalLight AnalysisSilicate AnalysisPhosphate AnalysisO2 AnalysisPlankton AnalysisChlorophyll AnalysisConclusion

Introduction


The offshore boat practical aimed to investigate the vertical mixing processes in the waters of the Western English Channel off Falmouth. They usually become vertically stratified during the summer months due to lower wind mixing and higher irradiance levels, but the shallow waters next to the coast remain mixed. Fronts therefore form in the channel, with warmer water above cold water on the stratified side and mixed cooler water on the coastal side. Fronts often tend to have large plankton communities due to the nutrients provided by the mixed side but the stratified waters allowing the plankton to remain in the high irradiance surface waters. The systems offshore along the coast around Falmouth were therefore investigated to observe how the vertical processes affect the plankton communities. The physical measurements made by instruments aboard a CTD would be compared with chemical data from water sampling and the biological data from phytoplankton samples and zooplankton trawls to observe how the physical processes control the biological productivity.

The primary plan for the offshore day was formulated based on data from the previous day's group (Group 3) who had travelled west from Falmouth along the coast towards Lizard Point and discovered frontal systems along the 30m contour. Based on this, Group 6 planned to zig-zag across the frontal system out to 50m and back in to 20-30m along the coast from the Manacles to Lizard Point. Samples were collected from stations along the frontal systems appropriate to the findings in the field indicated by the thermo-salinometer and the acoustic doppler current profile. An overview of each station is provided below in Table 6.1. Although a number of stations were fully sampled, due to time constraints only a CTD drop occurred at others in order to gain a larger set of physical data relating to the front.

In case of the primary plan not yielding a change in physical processes (being constantly monitored onboard by the ADCP and thermo-salinometer) the secondary plan would be to head out directly offshore until a change from mixed waters to stratified waters was observed and then to sample either side of and on the frontal system. In the event, the secondary plan was not required and the path taken is displayed in Figure 6.1.

Figure 6.1

 

Station Number Latitude (WGS84) Longitude (WGS84) Activity
1 - Black Rock 50°08.662 N 005°01.486 W A full sample station was taken at Black Rock. This station has been sampled by every group on Callista and therefore together the data will provide a temporal series of data at the site over 10 days.
2 - The Manacles 50°02.693 N 005°01.151 W CTD drop only. A continuously stratified station.
3 - Offshore of the Manacles 50°00.927 N 004°58.975 W Full sample station. Stratified
4 50°00.385 N 005°01.641 W CTD drop only. Stratified station. A high backscatter anomaly seen on the ADCP prompted the group to deploy the CTD at this station to investigate.
5 50°00.991 N 005°04.455 W Full station. Stratified station with high chlorophyll levels throughout suggested that the station had recently been mixed.
6 50°00.524 N 005°05.536 W Full station. Water column mixed indicating the station was inshore of the frontal systems.
7 49°58.017 N 005°09.850 W CTD drop only. Partially mixed station investigated whilst heading towards Lizard Point due to multiple crossings of a frontal system indicated by a marginal temperature rise and ACDP data, as well as visually indicated by a change in water colour causing stripes to form across the ocean.
8 - Offshore Lizard Point 49°54.715 N 005°10.230 W Full station. Stratified station directly offshore Lizard Point.

Table 6.1

Physical Characteristics


Figure 6.2 - Station 1

Figure 6.3 - Station 2

Figure 6.4 - Station 3

Figure 6.5 - Station 4

Figure 6.6 - Station 5

Figure 6.7 - Station 6

Figure 6.8 - Station 7

Figure 6.9 - Station 8

Richardson's Number

The Richardson's (“Ri”) number can be calculated and used to describe the nature of the water column; Ri<1 describes turbulent flow, Ri<1/4 describes laminar flow and an Ri number in-between these values expresses gravitational shear. It can be calculated using the following equation:

Ri is dimensionless

h= depth (m) over which density/ velocity are compared

g=9.81 m s-1 (gravitational force)

ρ= average density (kg m-3) in the water column
dρ/dz= change in density (kg m-3) with depth du/dz= change in velocity (m s-1) with depth

Ri depth profiles were plotted for each fully sampled station and were used with CTD profiles to help describe the balance of the stabilizing forces, such as buoyancy, on flow over the destabilizing forces, such as vertical shear.

Figure 6.2 shows that the water column at Black Rock was quite stratified with a warmer layer on top of a cooler layer with salinity and chlorophyll remaining relatively constant through both of the layers. The gradual thermocline seen between 15-20m in Figure 6.2 has high levels of laminar flow as shown by an increase in Richardson number in Figure 6.10. The stratified water above has low Ri values and thus is shown to be turbulent; with mixing present.

Figure 6.4 shows a stepwise thermocline in stratified water, with sharp decreases in temperature at 10m, 18m, and 22m. The main thermocline boundary is at 40m after which there is a mixed water layer. Cross analysis with the Ri number in Figure 6.11 shows laminar flow at each of these depths and turbulent water bodies in between. Turbulent water persists once the mixed layer boundary has been crossed, as is shown by a Ri < 0.25. Some fluctuations in the Ri number may be explained by this ‘snap shot’ view of the water column which captures where small variations in shear or buoyancy dominate, but does not describe the general water body trend.

Figure 6.6 shows a frontal water column that has recently become stratified. This is evident due to uniform levels of chlorophyll throughout the water column. A steep thermocline at 5m corresponds to high Ri numbers on Figure 6.12. The thermocline weakens and becomes more gradual to 15m, where a low Ri number depicts more turbulent flow and mixing. The likely recent mixing of this water column may explain other fluctuations in the data, as does the snap shot nature of this survey as described above.

Figure 6.7 shows a steep temperature gradient at 9m, corresponding with a high Ri number as would be expected in a thermocline. As the temperature gradient is crossed and the water cools turbulent flow becomes predominant once more. At most other depths in the water column low Ri values describe turbulence and a well-mixed system.

Figure 6.9 was an additional offshore stratified station to complement station 3. However, station 8 lacks the steep thermocline seen in station 3, instead a gradual thermocline occurs from the surface to 20m. Within this surface stratification, high Ri numbers as shown in Figure 6.14, and may fluctuate due to internal currents. Once the mixed layer boundary has been crossed low Ri numbers describe turbulence and deep water mixing.

Figure 6.10 - Station 1

Figure 6.11 - Station 3

Figure 6.12 - Station 5

Figure 6.13 - Station 6

Figure 6.14 - Station 8

ADCP Analysis

Figures 6.15, 6.16, 6.17 and 6.18 show the ADCP data for station 3 where the water was stratified. The backscatter shows a band of high backscatter around 20-30m reflecting zooplankton populations feeding on phytoplankton which reside around the thermocline where they can access nutrients in the lower layer. The magnitude of velocity was low at this station with minimum values around 0.4m s-1 and maximum around 0.9m s-1.

Figure 6.15 Backscatter

Figure 6.16 Velocity Direction

Figure 6.17 Velocity Magnitude

Figure 6.18 Stick

The ADCP data for station 6 (illustrated in Figures 6.19, 6.20 and 6.21) demonstrate that the water column was mixed. Backscatter here was high across the water column compared with stratified stations as at mixed stations the higher nutrient concentrations (see phosphate and silicate) throughout the water column allow phytoplankton and therefore zooplankton populations to remain high. Shear flow can be observed in the velocity direction plot with direction varying from 0 to 360o. The velocity observed at station 6 is lower then at station 3 with maximum value around 0.5ms-1 and a minimum of 0.05ms-1. The velocity of tidal flow may vary depending on the location and position along the coast with features such as headlands affecting the velocity measured.

Figure 6.19 Backscatter

Figure 6.20 Velocity Direction

Figure 6.21 Velocity Magnitude

Figures 6.22, 6.23, 6.24 and 6.25 show a tidal front that was observed on the ADCP. The backscatter plot shows lower backscatter than at either side of the front.  Velocity direction goes from around 200° on the eastern side of the front to around 100° on the western side with shear flow in the middle at the front. In addition, the velocity at the front is low compared to either side with the eastern side having a flow of around 0.7ms-1 and the front having a flow of 0.1ms-1.

Figure 6.22 Backscatter

Figure 6.23 Velocity Direction

Figure 6.24 Velocity Magnitude

Figure 6.25 Stick

The position of the tidal front was obtained by using ADCP data obtained whilst zig-zagging between Black Head and Lizard Point. The latitude and longitude (using WGS 84) of three points along the front are plotted on the following Google Earth plot (Figure 6.26). The points plotted on the admiralty chart in Figure 6.27 show that the front is positioned between the 27 - 33m contour.

Figure 6.26

Figure 6.27

Stratification Parameter

The stratification parameter log(h/u^3) (where h is the depth and u the tidal current) gives an indication of the degree of mixing in the water column due to tides and wind or the degree of stratification enhanced by heat input, freshwater inputs, and increased water depth. As a general rule, the stratification parameter tends to be < 2 at mixed stations and >3 at stratified stations. At a tidal front the value may be approximately log (h/u3) = 2.7± 0.3 for a tidal front (as suggested by Simpson and James (1986)). At station 1, the station was well stratified: the stratification parameter value 3.65 (Table 6.2) is corroborated by the CTD data. Stations 3,5 and 6 are all relatively close to the tidal front and so show stratification parameter between 2 and 3. For example, station 5 was observed in the CTD data to have a stratified water column with respect to temperature, but high chlorophyll levels throughout the water column suggested that the water column had been recently mixed. This demonstrates how the tidal front moves as the balance of mixing due to tide and winds and heat input varies. Station 8, offshore of Lizard Point, was well stratified and the stratification parameter corroborates the data gained from the CTD at this station. As R. V. Callista neared station 8, a temperature of >17°C was recorded on the thermo-salinometer which was the highest surface water temperature that had been recorded offshore for the preceding week. This illustrates the increasing solar irradiance/heat input and how its balance with the tidal and wind mixing would affect the position of the tidal front during the fieldcourse (especially as the weather was generally fine with low winds).

Station Depth, h (m) Tidal Strength, u (ms-1)

(h/u3)

Stratification parameter
log(h/u3)
1 31.5 0.191 4495 3.653
3 61.5 0.628 248 2.394
5 30.8 0.536 200 2.301
6 14.3 0.367 289 2.460
8 62.0 0.295 2418 3.383

Table 6.2

Light Analysis


Secchi disk measurements were used to estimate the depth of the euphotic zone for the offshore stations. This is defined as the area with sufficient light for photosynthesis. The LUP measurements from the CTD can also be used as a euphotic depth estimate by finding the 1% light level. Both estimates suggested that the deepest euphotic zone was found at station 8 (Table 6.3), this station also had the lowest k value and therefore slowest attenuation of light. Station 8 was strongly stratified suggesting lower mixing in the surface layer so reduced turbidity. As shown in Figure 6.28 there is a difference of 11m between the Secchi and LUP estimates, the LUP 1% is considered to be more accurate, and the fluorescence CTD readings indicated the presence of chlorophyll at the LUP depth as shown in Figure 6.9.

These estimates should be used as a rough guide to the euphotic depth and this cannot be understood fully until chlorophyll and fluorescence are analysed. In all cases the LUP depth should be considered as the most reliable when identifying relationships.

Note that the Secchi disk euphotic depth estimates for station 6 and 7 are deeper than the water column. Therefore the entire water column was contained within the euphotic zone and could expect to receive sufficient light for photosynthesis from the surface to the bed. However the LUP 1% was much shallower at station 6 suggesting an inaccurate Secchi estimate. The 1% light level was not reached for stations with marked * in Table 6.3.

Station Secchi Depth Euphotic Depth Estimation (m) k - Attenuation Coefficient LUP 1% Light Depth (m)
1 9.1 27.3 0.15 20
2 7.9 23.7 0.18 24
3 9.9 29.7 0.15 25
4 11.0 33.0 0.13 *
5 8.8 26.4 0.16 *
6 8.6 25.8 0.17 15
7 9.5 28.5 0.15 *
8 11.7 35.1 0.12 46

Table 6.3

Figure 6.28

Dissolved Silicon Analysis


Station 1 (Figure 6.29) showed a decrease in measured dissolved silicon concentration at 17m which increased down to 20m to 4.8 µmol l-1. On the CTD data a strong thermocline was illustrated down to 21m, this prevented the dissolved silicon mixing below the thermocline. This correlates with Figure 6.32 which shows an increase in phytoplankton abundance at this depth and suggests that they are utilising the silicon source, resulting in a decreased silicon concentration at this depth.

At station 3 measured dissolved silicon concentration increased to 6.4 µmol l-1 at 18m, then decreased with depth. The CTD data showed a very shallow surface warmed layer and a deep thermocline, below which chlorophyll increased (Figure 6.4). An increase in chlorophyll is a proxy for phytoplankton abundance and suggests the decrease in silicon concentration at 32m may have been due to the utilisation of dissolved silicon.

At station 5 silicon concentration increased to 5.2 µmol l-1 at 17.3m. The shallowest warmed layer at station 5 (Figure 6.6) may have forced the immotile phytoplankton to the very surface which is where lower silicon concentrations were observed indicating utilisation. The dissolved silicon concentration showed an increase at depth at 17m indicating replenishment just below the thermocline.

From surface to 14.3m, measured silicon concentrations at station 6 decreased by 0.5 µmol l-1. Figure 6.7 showed that the water column at this station had weak stratification. This was reflected by the low concentrations measured as there is no strong stratification keeping the silicon in the shallower depths sampled.

At station 8 silicon concentration showed no measured changes with depth after 3.5m, with slight depletion in the surface waters. Figure 6.9 showed a strong thermocline down to 20m below which temperature and salinity data indicated well mixed waters. Also below 20m fluorescence remained homogenous with depth, suggesting phytoplankton were also mixed within the water column, which may have explained the lack of measured changes in dissolved silicon concentrations.

Figure 6.29

Phosphate Analysis


At Station 1 no measured changes in phosphate concentration were observed with depth (Figure 6.30). This is reflected in the CTD data (Figure 6.2) which shows no significant changes in fluorescence with depth, indicating very low biological activity at this point.

At station 3 measured phosphate concentration showed a general increase with depth. However at 32.5m phosphate decreased to 0.09 µmol l-1. At this depth (Figure 6.4) shows an increase in fluorescence implying increase of phytoplankton abundance. Phosphate being a macronutrient may be being utilised at this depth and has the potential to become limiting over time.

Station 5 shows a measured increase of phosphate concentration of 0.6 µmol l-1 at depth 30.8m. (Figure 6.6) shows that this depth is in the mixed layer below the thermocline suggesting that the increase in phosphate was due to remineralisation. However also at this depth the phytoplankton data show abundance in Thalassiosira. This genus of phytoplankton are surviving at this depth as they are in the photic zone, however it is likely they have not got sufficient light to utilise the phosphate to the point of depletion.

Phosphate concentration at station 6 increased from 0.02 µmol l-1 at depth 1.7m to 0.12 at depth 14.3m. (Figure 6.7) shows that the thermocline exists down to about 15m, above which phosphate was more depleted than below where it is being replenished.

Measured phosphate concentration at station 8 showed a very similar profile to measured silicon concentrations (Figure 6.29); showing no measured change in concentration with depth from 19.8m. There was depletion in the surface waters, with 0 µmol l-1 at 6.1m. (Figure 6.9) shows a strong thermocline down to 20m below which temperature and salinity data indicate well mixed waters. Also below 20m fluorescence remains homogenous with depth, suggesting phytoplankton are also mixed within the water column therefore unable to utilise phosphate which can explain the lack of observed changes in phosphate concentration.

Figure 6.30

Dissolved Oxygen


Station 1 (Black Rock) was well-mixed with low background concentrations of chlorophyll and no significant chlorophyll peaks; this is reflected in the homogeneity of the oxygen measurements taken over the range of depths (7.3m to 21.1m) (Figure 6.31). Due to the attenuation of light through the water column, chlorophyll concentration decreases with depth as does rate of photosynthesis and therefore the drop in oxygen concentration may be explained.

Station 3 was offshore at a stratified water column with the strongest recorded thermocline of the offshore stations. The water was supersaturated with oxygen (260.1% at surface, 466.6% at 32.5m) and the peak in oxygen saturation coincided with the peak of chlorophyll at 32.5m (24.8µg l-1).

Station 5 was inshore on a body of water that had recently been mixed and was weakly stratified at the time of sampling. Variation between the oxygen measurements taken across the warmed surface layer was small. A small rise in oxygen concentration was measured at 17.3m which coincided with a small peak of chlorophyll at the same depth.

Station 6 was inshore on a well-mixed body of water. The oxygen measurements showed little variation and CTD data show the homogeneity of the water column.

Station 8 was on a stratified water body and yet the oxygen data show there was little variation over a large depth range (6m to 69m). Despite the presence of a chlorophyll maximum, turbidity minimum and strong thermocline at 12.5m, the oxygen saturation remained stable.

Figure 6.31

Phytoplankton & Zooplankton Analysis


Phytoplankton

The highest abundance of phytoplankton was recorded at station 5, with 211000 cells l-1, the most abundant phytoplankton was the Thalassiosira genus with 135000 cells l-1 (Figure 6.32). The Thalassiosira are a wide spread diatom that can be found throughout the world’s oceans. The high abundance at station 5 could have been due to the depth - the shallower water meant the seabed forced the thermocline up to the surface, allowing the nutrients from the deeper cold water to be evenly distributed throughout the water column and utilised to a greater extent by photosynthetic organisms in the eutrophic zone. Station 5 was also a stratified station with corresponding high levels of chlorophyll that suggested that the station was recently mixed and nutrients freshly available; other areas that have been stratified for longer had become nutrient depleted in the surface as the nutrients are rapidly used by surrounding phytoplankton.

The lowest abundance was recorded nearby at station 6 (48000 cells l-1), an inshore station, and the dominant genus was Chaetoceros (39000 cells l-1). Chaetoceros spp. were consistently present at all sampled stations except for station 3 which had only 1700 cells l-1.  Station 6 had the lowest phytoplankton population where the water column was mixed indicating it was inshore of the frontal systems - this could have been due to the nutrients having been utilised throughout the water column. Figure 6.30 illustrates that phosphate concentrations were relatively low.

The samples from the two offshore stations (3 and 8) appeared to have the lowest phytoplankton populations of the sampled stations; this could have been due to nutrients not being mixed to the upper water column as the waters are stratified. This explains why there was a reduced number of phytoplankton within the sample. Using CTD data the chlorophyll maximum can be shown e.g. 20-30m at station 3 around the point where the water temperature fell as the layers of warm nutrient poor water interacted with the nutrient rich cold water. At this point there was sufficient light and heat for photosynthesis from above and the nutrients were accessible either by minor diffusion or by plankton migration through the water column. The chlorophyll maximum at station 8 was around 25m but the chlorophyll levels were kept low due to predation by zooplankton - using the CTD data in Figure 6.30 high turbidity can be observed, caused by the larger plankton feeding on the photosynthetic organisms.

Figure 6.32

Zooplankton

Knowledge of variability in zooplankton biomass is important for understanding the ocean food webs and energy flow in the oceans; as  zooplankton contribute to the  transport of carbon and nitrogen to the deep sea via production of faecal pellets and active transport by vertical migration, which are important components of the biological pump (Longhurst et al., 1990). Using information displayed on the computer in real time from acoustic backscatter intensity derived from Acoustic Doppler Current Profilers (ADCPs) and zooplankton biomass from net collected zooplankton samples, variability in the distributions of zooplankton biomass are described. This is possible as zooplankton are large enough to provide an image on the ADCP. The zooplankton samples collected were taken from areas that showed high acoustic backscatter in relation to the surrounding areas.

Sample sites were also selected based on fluorometer readings, which indicated areas of high chlorophyll concentrations at various depths. This allowed comparisons of chlorophyll concentration with zooplankton biomass. All zooplankton samples were collected from stations where physical data from the CTD, chemical and biological (phytoplankton) data from Niskin bottles attached to the rosette system were all collected.

Samples were collected using a vertical closing net allowing specific depths within the water column to be targeted for sampling. All zooplankton samples were stored in the dark with formaldehyde added after collection to preserve the sample.

As seen in Figure 6.35 the highest abundance of zooplankton was measured at stations 3 and 8: 1613 m-3 and 1103 m-3, respectively. The dominant species at both of these stations were hydromedusae, 728 m-3  at station 3 and  467 m-3  at station 8. The phytoplankton data obtained (Figure 6.32) may explain the higher numbers of zooplankton at station 3 and 8.

Stations 3 and 8 have different secondary dominant species; station 3 has decapod larvae and at station 8 copepods are the second largest group recorded.

As zooplankton feed on phytoplankton in areas of well mixed coastal waters, with larger communities of phytoplankton more zooplankton can be supported. In this data the larger zooplankton samples collected were situated further offshore, this could be due to the chlorophyll maximum at depth and the development of  thermocline (Figure 6.9). Nutrients were constantly replenished from deeper waters, the phytoplankton were able to grow thus providing a food source for the zooplankton. This can be seen in the CTD data as the fluorometer reading at station 8 shows a minor increase at a depth of 20m, however there is a large amount of turbidity which could possibly be zooplankton which feed on the phytoplankton keeping the chlorophyll levels low.

At station 5, the zooplankton were sampled at two depths: 0-15m and 15-30m. High chlorophyll concentration throughout the water column suggested that the station may have been recently mixed allowing the chlorophyll concentration to rise in response. However the recent increase in levels of irradiance (http://www.metoffice.gov.uk/climate/uk/anomalygraphs) may have caused stratification of the water column with a shallow steep thermocline. High levels of acoustic backscatter were also observed. For these reasons two samples were taken from either side of the thermocline; to study any differences in zooplankton compositions. From 0-15m station 5 had a larger number of hydromedusae present (239 m-3) with a smaller number of copepods (144 m-3), whereas at 15-30m station 5 was dominated by copepods (167 m-3). It could be suggested that this is due to the different communities of phytoplankton at the thermocline allowing for different niches for individuals to develop within.

Station 1 is situated at Black Rock near the mouth of the Fal estuary, there is a low amount of phytoplankton recorded at this station (Figure 6.32) although there is a high chlorophyll concentration (Figure 6.2).  This could possibly be due to phytoplankton cells having high chlorophyll content but are few in number therefore unable to support high zooplankton abundance.

 

Figure 6.33 - Teleost Egg photo from zooplankton sample

Figure 6.34 - Copepod photo from zooplankton sample

Figure 6.35

Comparison of phytoplankton and zooplankton populations in the estuary and offshore

In the estuary the vast majority of phytoplankton were found to be diatoms, with a few other forms present. The main species of diatom sampled were Chaetoceros. This species was also observed in the offshore sampling, but in smaller numbers, in addition to many other diatom species that were sampled in similar sized numbers (Thalassiosira, Cylindrotheca, Leptocylindrus). Offshore a large number of dinoflagellates were recorded in addition to diatoms; this could have been due to the environment being more osmotically stable with stratified environments, unlike in the estuary, creating a greater number of niches for organisms to fill and allowing dinoflagellates to grow where diatoms cannot as they can use their flagella for motility to obtain nutrients.

The zooplankton sampled in the estuary and offshore show a large variation in population composition; the estuarine zooplankton community had a very high proportion of copepods and copepod nauplii possible due to a greater tolerance of changing salinities compared to the zooplankton found offshore. Offshore zooplankton are less tolerant of changing environments, as they are used to a more stable environment and so hydromedusae numbers were much greater than those in the estuary.

Chlorophyll Analysis


The chlorophyll measurements from the samples collected reflected the general characteristics at the different stations and the fluorescence readings on the CTD profiles. Stratified stations 3 and 8 (Figure 6.36) show a distinctive increase in chlorophyll l-1 just beneath the thermocline (shown by comparison to the CTD profiles). The greatest chlorophyll concentration was at station 3 with a concentration of 25mg l-1 at 32m. Both stations show a decline in chlorophyll after the peak as depth increased. Results at station 5, which was also stratified, suggested recent mixing within the surface water as there was no distinct increase in concentration surrounding the thermocline. The thermocline was also much weaker compared to stations 3 and 8. Mixed stations 6 and 1 showed an gradual increase in chlorophyll with depth.

Figure 6.36

Conclusion


The offshore data reflect the concept of a frontal system around half a mile to a mile offshore. ADCP data show positions where the front was crossed to be between the 27-33m depth contours. The stratification parameter reflects the position of each station regarding the front well. Stations offshore of the front reflect a stratified water column with warm water above colder water and a distinct thermocline, just beneath which is where the chlorophyll tends to reach a maximum value as nutrients become available to phytoplankton residing in the euphotic zone. Mixed stations show higher chlorophyll values throughout as the chlorophyll is well mixed throughout the water column as nutrients are freely available. Nutrient data follows an expected trend with depth; biomass in the upper water column causes depletion near the surface of silicon and phosphate. Offshore the nutrients remain higher beneath the thermocline where chlorophyll is low as phytoplankton are unable to utilise nutrients because the light levels are too low at depth. Offshore the levels of dissolved oxygen tend to be much higher than inshore with an increase below the thermocline where biological activity is limited. At the inshore stations oxygen has a more homogeneous distribution in the water column due to physical processes mixing the water column. Surface fluorescence data is high at most stations but are particularly high at stations near the front e.g. 5 and 6 reflecting the frontal conditions providing ideal conditions for phytoplankton blooms, but the sampling was not fine enough to draw firm conclusions about how phytoplankton distributions vary across the front.

 

GeoPhysics Sidescan SonarGrabs AnalysisVideo AnalysisConclusion
Introduction

The aim of this survey was to compile a habitat map from sidescan sonar interpretation and the use of direct sampling to gain biological information.

Habitat maps allow the wide scale geology and present day sedimentary processes to be determined and understood (Laban 1998). This understanding of the seabed structure and distribution is vital to enable effective management of the impact of human activity and climate change on ecosystem form and function.

Habitat mapping is a powerful tool, allowing comparison of sites; therefore a global view on the variety of habitats can be gained and areas important for conservation of species can be identified. Jordan et al. (2004) mapped the Kent Group of islands, south-eastern Australia. This habitat mapping generated a capability to define the boundary and size of potential MPA.

The Fal estuary is a Special Area of Conservation (SAC) accredited by the Joint Nature Conservation Committee (JNCC). One of the reasons for this status is due to the Annex 1 habitat ‘sandbanks covered by water all the time’. Within this classification the JNCC has made a specific comment on the importance of the extensive areas of maerl gravel (dead and alive maerl) which extend throughout the Carrick Rhoads and Falmouth Bay. The JNCC recognises the importance of maerl beds due to the diversity of species living among these beds and the general scarcity of this habitat in UK waters. (www.defra.gov.uk)

The geophysical sidescan survey was undertaken on L. C. Grey Bear. Using a sidescan towfish a selected area was first surveyed, followed by grabs being undertaken at interesting sites using a Van Veen Grab. Video survey was also used in order to ground truth the data from the sidescan sonar and to confirm that areas of uncertainty were suitable for taking a grab. The habitats present between Pendennis Point, Black Rock and Falmouth Harbour were surveyed and the findings were presented on a poster and are described below on this website.

Trimble Hydropro software was used to create the desired track for survey following consultation of WGS84 Admiralty charts. This software was used in conjunction with the vessel’s GPS in order to provide the skipper with information about how far port or starboard the vessel was from the planned track and the course could then be adjusted accordingly. After the initial survey, the grab sites were plotted and the ship could navigate to them accordingly. During adverse weather conditions, the captain would orientate bow or stern to the wind as appropriate in order that the grab could be completed as safely and comfortably as possible. Weather conditions for this survey are shown in Table 7.1

Date 6/7/2011
General Weather

Bright and sunny morning, deteriorating to cloudy and rainy afternoon.

Visibility Clear morning, deteriorating slightly in the afternoon
Sea State 2-5
Cloud Cover 55%
Wind Speed 15.11
Wind Direction SW

Table 7.1

6/7/2011 Tide Times GMT Tidal Height (m)

Low Water

0244

0.6

High Water

0832 4.8

Low Water

1459 0.8

Table 7.2

Figure 7.1. Tidal curve for 06.07.2011

The three transects undertaken on the survey are shown in Figure 7.2 and Table 7.3. A fourth transect was also completed, however due to technical difficulties the record was not printed; therefore no further reference will be made to this transect. All three grab sites sampled are also specified in Figure 7.2. Two video transects were completed which are shown in Figure 7.2, an attempt to video at the first grab site (32011.71N, 182909.50E) was made, however weather conditions were poor rendering the footage unsatisfactory for viewing.

Figure 7.2

Transect Number

Start time (GMT)

End Time  (GMT)

Start Position

End Position

1

08:54

09:08

31644 N
182936 E

32222 N
182252 E

2

09:16

09:27

33002 N
182419 E

31657 N
182999 E

3

09:32

09:47

31716 N
1830334 E

33250 N
182370 E

Table 7.3

Sidescan Sonar


The sidescan system consists of three components: a towfish, a transmission cable and the topside processing unit. The towfish was lowered into the water off the stern of the boat as the area to be surveyed was reached. The towfish was then transmitting when we entered the survey area. The towfish was then towed behind and below L.C Grey Bear with the vessel heading as steady as possible to prevent the image blurring.  The towfish maintained a constant depth due to the length of the transmission cable.

The towfish has an acoustic transducer which sends out a wide angle pulse, this wave then propagates across the seabed. Every sonar pulse is one line of information as soon as the pulse is sent the tow fish then ‘listens’ for an echo, this is called backscatter .This information then gets passed along the transmission cable. The sidescan sonar system then displays the intensity of the sound scattered back to the towfish from the seafloor; the data received was displayed in real time on a laptop computer screen on the deck of L.C Grey Bear. This allowed us to detect any objects or bedforms on the seafloor. The sidescan sonar works based on the first law of reflection which states that the incident ray, the reflected ray, and the normal all lie on the same plane. Because this only applies to perfectly smooth surfaces, backscatter is created.

Whilst watching the screen we were able to view the backscatter data and discuss how the properties of the seabed are what determine the strength of the backscatter echo. Rock, gravel, coarse sand and wood are better at reflecting so they create more backscatter and are therefore recorded as darker areas than soft muddy sediments; which produce less backscatter, so appear lighter in colour on the sonar record.

We also discussed how the angle of the seafloor slope influences the backscatter; slopes facing away from the transducer appear dark, and slopes facing towards the transducer appear paler. Another feature of the sidescan system we noted whilst carrying out the survey was how a 3-dimensional quality can be achieved by acoustic shadows (Figure 7.3). These acoustic shadows occur alongside areas that stand out or dip down in the sea-floor; these are very useful for indicating the shape and nature of the mound/groove/bed-form.

Figure 7.3 Sidescan theory

The operating frequency on the sidescan sonar can be altered depending on the resolution and swath required. A low frequency, such as 100 kHz, will give a wider swath width but the resolution will be lower. This is useful for surveying large areas.  A higher frequency, such as 500 kHz, will give better resolution but a narrower swath width. An even higher frequency of 2.5 MHz, for example, can be used for a really well defined image; however this will significantly lower the swath range, so would be used in small sites. This survey was operated at 110 kHz.

Swath systems, such as sidescan sonar, are most likely to provide the best high-resolution maps, particularly over wide areas. They provide information on sediment texture and bedform structure and allow dynamic processes (e.g. sediment transport) to be deduced. Disadvantages associated with swath systems are their high costs and the need to have skilled interpretation. In addition, the output often requires considerable post-processing time and expense to obtain appropriate classifications (Kenny et al., 2003).

Limitations

There are a few limitations with sidescan sonar. One is the interference produced by the ship’s echo sounder which produces sharp marks on the sidescan sonar when running at high frequency. Other sources of interference include other sonar sources as well as dolphins. Dolphins use a frequency of 400nm for hunting and for communication they have use a frequency of 100nm which is at the lower end of sidescan and produces weird traces.

Another problem is due to the fact that the sidescan beams widen with distance from the tow fish and so objects may merge if for example there were two closely placed objects from the perspective of the tow fish it would look like one, for example seaweed appears like single white lines rather than many plants (see Figure 7.4).

Figure 7.4

Lastly is the uncertainty with the sidescan, when analysing bedforms can only be suggested and the scan only gives an indication of what the bed could be formed from (substrate/bedrock).  This emphasises the need for ground truthing by taking grabs and video footage.

Grabs


Grab 1

Time 1033GMT
Northings 23011.71
Eastings 182909.5
Sediment Description Fauna and Flora
Sediment was made up of mainly coarse and very coarse sand, 70% of which was retained in the 2mm sieve and 25% in the 1mm sieve. Rocks and pebbles made up the remaining 5% and were retained within the 10mm sieve. Sediments also contained broken Mollusca shells and dead maerl. Live maerl
Hermit crabs
Long clawed porcelain crabs              Kelp
Rock with attached keel worms, pink coralline algae, green algae, pink ray limpets

Figure 7.6 - Long Clawed Porcelain Crab

Figure 7.7 - Live Maerl

Table 7.4

Figure 7.5

Figure 7.8

Figure 7.9

Maerl is a free living, calcified, coralline red algae. Maerl beds are composed of either living or dead unattached maerl which form branching and interlocking structures (Birkett et al. 1998).  Maerl beds form in areas without high levels of siltation, as this leads to smothering and burial of the maerl which will reduce photosynthesis, laboratory experiments show that smothering by fine sediment and lowered oxygen concentrations are particularly damaging to maerl algae (Wilson et al. 2004).  Maerl is an important habitat for numerous species and thus support high biodiversity and high trophic group diversity (Barbera et al. 2003).The main ecological role for maerl is to provide shelter for animals and attachment  for plants (Castriota et al. 2005).


Grab 2

Time 1121GMT
Northings

32339.25

Eastings

182695.08

Sediment Description Fauna and Flora
No sediment was recovered in the grab however the fur bellows hold fast was attached to a 250mm rock. Furbelows
Sugar kelp
Irish moss
Red and Brown algae
Painted and Grey top shells
Whelks
Snakelocks anemone
Porcelain and Edible crabs
Squat lobsters
Barnacles
Cowrie

Figure 7.11 - Edible Crab

Figure 7.12 - Bryozoa

Table 7.5

Figure 7.10

Furbelows kelp

Phylum: Orchrophyta
Class:
Phaeophyceae
Species:
Saccorhiza polyschides
Habitat description: 
Saccorhiza polyschides grows from extreme low water springs to a depth of 35 m. It normally attaches to rocks but is occasionally found loose-lying on small stones or shells. It can form dense stands in sheltered areas and can tolerate strong currents.

Description: Saccorhiza polyschides is kelp species with a distinctive large warty holdfast and a flattened stipe with a frilly margin. The stipe is twisted at the base and widens to form a large flat lamina, which is divided into ribbon-like sections. The species is an annual, and very fast growing. It is opportunistic and colonises available hard substrata in the sublittoral (www.marlin.ac.uk).

Sugar Kelp

Phylum: Chromophycota
Class:
Phaeophyceae
Order:
Laminariales
Species:
Saccharina latissima
Habitat description:
Sometimes found in littoral rock pools but more commonly in the sublittoral below 20m anchored to rocky substrate using a large holdfast. Widespread global distribution in temporal waters
Description:
Can grow up to 4 m long, attached to rock by strong cylindrical stipe. Long, leathery, unbranched blade about 15cm wide without a midrib. Blades are flat but wrinkly with wavy margins.

Grab 3

Time 1209GMT
Northings 33013.9
Eastings 182538.2
Sediment Description Fauna and Flora
Sediment was made up of mainly small pebbles and coarse sands, 50% of which was retained in the 10mm sieve, 30% in the 2mm sieve and 20% in the 1mm sieve. Sediments also contained broken Mollusca shells and dead maerl.

Brown algae
Sand mason worms
Grey and Flat top shells
Cowries
Brittle star
Crabs                                                Live maerl
Brittlestar
Towershell
Rock with attached keel worms, pink coralline algae, platyhelminthes

 

Figure 7.14 - Grey and Flat Top Shell

Figure 7.15 - Sandmason Worm

Table 7.6

Figure 7.13

Figure 7.16

Figure 7.17

Video Analysis

The video trawls show that the habitat surveyed with the video was a sandy/shingle bottom, with a third to a quarter covering of empty larger shells. Red (for example Dudresnaya verticillata) and green seaweeds and some kelp covered approximately a tenth of the seafloor. There were occasional large features covered in kelp and rhodophyta. As the long trawl following the second grab progressed more of the seabed became covered in maerl and occasional shoals of juvenile fish were observed (maerl is known as an important nursery ground for juveniles).

Figure 7.18

Conclusion


Interpretation

Figure 7.19

 

The most Northerly point of the transect (33295N, 182410E) consists of sediment including broken maerl, shells and gravel, as indicated by grab 3 which was taken nearby (Figure 7.19). This bed type persists in a South East direction, until position 32490N, 182480E. Here, the bedform is divided and sediment is present only in the East of the tracks. In the West a new bedform exists; a mix of bedrock, boulders, gravel and seaweeds. A decision was made to sample at the division between two substrate types, grab 2. This made it possible to see the transitional species that inhabit this area. This division between sediment and bedrock runs from 32490N, 182640E to 31895N, 182895E, with some gullies and peninsulas extending from the main body of the bedrock.

Grab sample 1 was taken at 32011N, 182909E in an area that had been identified on track 3 as having sidescan signatures for fine sediment. To achieve a contrast between grabs and understand the variations of fauna and flora found between sediment and rock substrates.

Key notable features in the sediment include lobster pots, which are absent in the bedrock. At point 32800N, 182440E and extending to 326710N, 182500E lies what is likely to be a pipeline from the disused sewage treatment facilities on the nearby shore.

At the extreme South East end an area exists from; 31860N, 182840E to 31770N, 182865E across all three transects. Here, bedrock dominates with no presence of sediment and three prominent dips have been identified on the sidescan trace. Pythagorean equations have shown that the depth and length of these differ as indicated above. However when analysing the sidescan trace one must bear in mind that from different angles the same bedform may appear different. This is one of the limitations of sidescan: the nature of bedforms and substrate may only be suggested and grabs are required to provide verification. This verification process is known as ground truthing. Grab 3 was taken from the same sediment type as grab 1 but at some distance away to allow analysis of any spatial gradients in flora and fauna that may exist. This can only be done because the type of organisms that were sampled move at a negligible rate with comparison to the progress of the surveying vessel.

 

Discussion of Further Possible Investigation

Further investigations in the estuarine system that spring from our findings may include investigating how the mussel farm on the Fal Estuary affects the chemical and biological parameters in the estuary. Our data may have been influenced by the mussel farm on several occasions including phytoplankton and oxygen concentrations. The phosphate additions in the estuary, presumed to be from agricultural practices, sewerage outputs and possible mine drainage, could be monitored to observe any change over time and how this relates to current practices at potential input sites. Phytoplankton and zooplankton populations in the estuary could be further investigated and compared to other estuarine systems in the area such as the Helford Estuary and up Restronguet Creek.

In the offshore system, the frontal systems could be further investigating by finer sampling closer to the front – this would provide more detailed data on phytoplankton distributions across the system and how this varies to offshore data generally. Time series sampling of the front over the course of the year would investigate how the front moves.

Geophysical surveying and habitat mapping in the same area that we sampled could be repeated over time to give a time series that could be used to show how the habitat varied, particularly with reference to the maerl as an important long living organism of conservation value. The area that we sampled in was close to the outer harbour wall, parts of which are currently disintegrating, and contain various substances likely to cause pollution.

References

 

Barbera C. (and 17 others) 2003. Conservation and management of northeast Atlantic and Mediterranean maerl beds. Aquat. Conserv.Mar. Freshwat. Ecosyst: 13, 65–76

 

Birkett D.A. Maggs C.A. & Darling, M.J.  1998. Maerl, (Volume V) An overview of dynamic and sensitivity characteristics for conservation management  of Marine SACs. Scottish Association for Marine Science (UK Marine SACs project). 116

 

Boyer I.C. Gillespie P.A. Kaspar H.F. MacKenzie A.L. 1995. Effects of mussel aquaculture on the nitrogen cycle and benthic communities in Kenepuru Sound, Marlborough Sounds, New Zealand. Marine Biology: 85:2, 127-136

 

Cadee G.C. 1978. Primary production and chlorophyll in the Zaire River, estuary and plume, Netherlands. Journal of Sea Research : 12, 368-381

 

Castriota L. Agamennone F. & Sunseri G. 2005. The mollusc community associated with maerl beds of Ustca Island. Cah.Bio.Mar: 46, 289-297

 

Engel C.R. Guillemin M.L. Jacob A.M. Valero M. Viard F.  2008. Permanent Genetic Resources: Isolation of microsatellite loci from the kelp, Saccorhiza polyschides (Heterokontophyta, incertae sedis) Molecular Ecology Resources: 8, 406–408

 

Huot Y. Babin M. Bruyant F. Grob C. Twardowski M.S. Claustre H. 2007. Does chlorophyll a provide the best index of phytoplankton biomass for primary productivity studies? Biogeosciences Discuss: 4, 707-745

 

Jordan A. Lawler M. Halley V. and Barrett N. 2005. Seabed habitat mapping in the Kent Group of islands and its role in Marine protected area planning. Aquatic Conservation: Marine and Freshwater Ecosystems: 15, 51–70

 

Kenny A.J. Cato I. Desprez M. Fader G. Schuttenhelm R.T.E. Side J. 2003. An overview of seabed mapping technologies in the context of marine habitat classification. ICES Journal of Marine Science: 60:2

 

Laban C. 1998. Seabed mapping. Hydro International: 2:1, 4

 

Langston W.J. Chesman B.S. Burt G.R. Hawkins S.J. Readman J. Worsfold P. 2003. Site Characterisation of the South West European Marine Sites: Fal and Helford cSAC. Marine Biological Association

 

Lorder C.T & Reichard R.P. 1981 The dynamics of conservative mixing in estuaries. Estuaries and Coasts: 4:1, 64-69

 

Mullin J.B. & Riley J.P. 1954. The colorimetric determination of silicate with special reference to sea and natural waters. Analytica Chimica Acta:12, 162-176.

 

Simpson J.H. & James J.D. 1986. Coastal and Estuarine Fronts. Baroclinic Process on Continental Shelves. Geophys. Union: 63-94.

 

Souchu P. Vaquer A. Collos Y. Landrian S. Deslous-Paoli J.M. Bibent B. 2001. Influence of shellfish farming activities on the biogeochemical composition of the water column in Thau lagoon. Marine ecology progress series: 218, 141-152

 

Vesci A. 2001. Fore-reef carbonate production: Development of a regional cencus based on method & first estimates. Palaeogeology, Palaeoclimatology, & Palaeoecology: 175, 185-210

 

Welschmeyer N.A. & Lorenzen C.J. 1985. Chlorophyll budgets: Zooplankton grazing and phytoplankton growth in temperate fjord and central Pacific Gyres

 

Longhurst A.R. Bedo A.W. Harrison W.G. Head E.J.H. Sameoto D.D. 1990. Vertical flux of respiratory carbon by oceanic diel migrant biota. Deep-Sea Research: 37:4, 685–694.

 

Wilson S. Blake C. Berges J.A. Maggs C.A.  2004. Environmental tolerances of free-living coralline algae (maerl): implications for European maerl conservation. Biological Conservation: 120, 283–293

 

Websites Accessed

Accessed: www.envirogene.co.uk/downloads/casestudies/envirogene_case_fal.pdf , 7th July 2011

Accessed: www.westcountrymussels.co.uk, 7th July 2011

MarLIN - Marine Life Information Network [Online] Available:  http://www.marlin.ac.uk/speciesfullreview.php?speciesID=4284 [Accessed 2011, July 7th]

JNCC – Joint Nature Conservation Committee [Online] Available: http://jncc.defra.gov.uk/ProtectedSites/SACselection/habitat.asp?FeatureIntCode=H1110 [Accessed 2011, July 7th]

Tri tech innovative underwater technology. [online] Available: http://www.tritech.co.uk/products/info/products-info-sidescan_sonars.htm [Accessed 2010, November 23rd]

Disclaimer: The ideas expressed in this website are those of the group and do not reflect those of the University of Southampton or the National Oceanography Centre (NOC).