Group 6

INTRODUCTION THE TEAM BOATS & EQUIPMENT METHODS
ESTUARY OFFSHORE GEOPHYSICS CONCLUSION
INTRODUCTION TO THE FAL ESTUARY  

Figure 1.1 Map of Cornwall (inset the Fal Estuary)

The Fal estuary is a Ria situated on the South coast of Cornwall. Rising on the Goss Moor, various tidal creeks and rivers join to form Carrick Roads, where the estuary spans at least a mile wide from Turnware Point to Black Rock as it meets the English Channel.  The Fal is the third deepest natural harbour in the world, with a significant 127km coastline. Falmouth is situated towards the mouth of the estuary, and has historically (dating back to ~14th century) acted as an important port and docks for the West Country, especially with the development of the Cornish Railways and the tourist industry in the early 19th century.

The Fal is macrotidal, with a max. spring tide of 5.3m at Falmouth (with mesotidal spring tide of 3.5m at Truro) (Pirrie et al., 2009). There are various important habitats, with 93ha of saltmarsh and 653ha of intertidal mudflat; this has lead to the Fal and Helford estuaries being classified a special area of conservation (SAC) under Annexes I and II of the Habitats directive (Fal and Helford Management Forum, 2006). The SAC status refers to the presence of various rare and interesting habitats like eelgrass meadows (Zostera marina) and maerl beds (Phymatolithon calcareum).  The total estuary area is 2482ha with 6387.8 ha (JNCC) including the estuary and surrounding area classified under SAC.

Figure 1.2 Map of Fal Estuary

Most river borne material from rivers passes through estuaries; they are important transition zones for continentally derived material and can change the fluxes of materials entering the sea. The chemical distribution may influence the biological processes and biological production. In the Fal however, anphropogenic inputs have disrupted the natural balance of estuarine constituents. Large levels of arable farming and the Truro Newham Sewage Treatment Works  towards the estuary head and along the tributary rivers has lead to greater inputs of nutrients leading to eutrophication incidents that affects the delicate habitats present, such as the 1995-1996 bloom Alexandrium tamarense toxic dinoflagellate bloom (Langston et al., 2003).

There is also a history of metal pollution in the Fal estuary arising from the large amount of Sn, Fe, Pb and Cu mining occurring from the bronze age to present day. Tributyl tin (TBT) from antifoaling paints used on the hulls of boats has also had a significant impact, leading to ‘progressive change in the benthic invertebrate communities’ (Langston et al., 2003), although it is now banned from vessels. A notable example of metal pollution is the Wheal Jane incident in 1991; highly acidic water was released into the Carnon River and Fal Estuary (Bryan & Lagston, 1992), from which the estuary still hasn’t fully recovered.

In this investigation the aim is to collect data on the physical, chemical and biological characteristics of the Fal Estuary and surrounding offshore environment of the near shore English Channel and to gain an understanding of the interactions and an investigation on the benthic environment of Falmouth Bay.


BOATS & EQUIPMENT

RV Bill Conway This 11.74m single hull vessel  has a full load draught of 1.3m and was used for the estuarine data collection allowing us to reach stations further up the estuary towards King Harry Ferry that the larger vessels would be unable to reach. An onboard ADCP 6000 unit allowed for cross sectional current profiles of the estuary at various positions, with help from navigational software onboard.

 

Length

11.74m

Speed

10 knots

Figure 1.3 RV Bill Conway

Beam

3.96m

Passengers

2 crew + 12 scientists

Draft

1.3m

Equipment

‘A’ frame – 3m height & 750kg max. lift

Range

150nm

 

Trawl winch – 70m

SV Xplorer A fast survey support vessel, this mid size vessel is used for the geophysics data collection; side scan sonar and grab samples can be taken from the stern deck.

 

Range

100nm

Top Speed 25knots

Figure 1.4 SV Xplorer

Length

11.88m

Passengers 12 + crew

Beam

5.20m

Equipment Capstan & hydraulic crane – 0.5m

Draft

1.2m

  Bathymetric surveying side scan & boomer

 

 

 

Van veen grab

RV Callista is the largest of the three, and allows for offshore data collection. The twin-hulled Callista is just under 20 metres in length and a purpose built research vessel with a large rear deck and A frame allowing for deployment of heavier scientific equipment.

 

Length

19.75m

Max. passengers 30

 

 

 

 

 

Figure 1.5 RV Callista

Max. Breadth

7.40m

Range 400nm

Draft

1.8m

Equipment ‘A’ frame and winch – 4 tonne

Max. speed

15 knots

  Capstan – 1.5 tonne

 

 

 

Hull mounted ADCP – 600kHz

 

 

 

CTD and Rosette

Figure 1.6 How an ADCP works

ADCP - acoustic doppler current profiler, 400 kHz ADCP fitted to RV Bill Conway, and a 600 kHz ADCP to RV Callista. The ADCP measures the Doppler shift of a returning sound pulse emitted by the ADCP's transducers into the water column which is scattered, reflected and returned to the ADCP by particles within the water. The ADCP measures the Doppler shift of the sound pulse by combining 3 or 4 directional transducers; a profile of water currents over a range of depths can be created.

CTD Probe - stands for Conductivity-Temperature-Depth probe, and is the oceanographers ‘workhorse.' It measures salinity from the electrical conductivity of seawater, temperature with a thermocouple and depth by using a pressure transducer. CTD probes are usually fitted to a Rosette system with Niskin bottles, allowing ‘intelligent’ sampling of a water column. The probe used on RV Bill Conway was produced by Ocean Science International.

Figure 1.7

Rosette Frame with Niskin bottles and CTD probe

Niskin Bottle - used to collect samples at different water depths; they are hollow tubes with caps held closed with strong elastic. During deployment the caps are held open and then at the required depth they are triggered shut; either by messenger weights (if deployed on a hydro-line) or if on a rosette system they are shut electronically. The Niskin bottles used on RV Bill Conway are made by General Oceanic Inc. (model number 1010). On RV Bill Conway, the bottles were used in conjunction with a rosette system.

Rosette Frame - The Niskin bottles and CTD probe are mounted on a rosette frame along with other instruments like a Fluorometer and Transmissometer. The frame creates a platform in which multiple instruments can be deployed simultaneously from a vessel. The Frame on Bill Conway carries 5 niskin bottles, a CTD probe, a Fluorometer and a transmissometer. It is made by General Oceanic Inc. and is a Submersible Array Electronic Module.

Fluorometer -  this primarily measures the wavelength of light and fluorescence. It works by emitting a wavelength of light, with a receiver measuring the resultant change in wavelength. This change then provides a measurement phytoplankton biomass. The fluorometer used on RV Bill Conway was produced by Wet Labs, with the serial number WS3S400P; the fluorometer was attached to the rosette frame to provide insitu measurements along with the CTD system and to complement data from the water samples collected by the Niskin bottles.

Transmissometer - this is used to measure suspended particulate matter; this is achieved by emitting a wavelength of light through a path length of sea water about 20cm in size, towards a receiving sensor. Suspended particles within the water column block the light source and so the receiving sensor measures different wavelengths of light, giving a quantifiable value of suspended particulate matter.

Figure 1.8 Secchi Disk

Secchi Disk - this is a disk split into alternate black and white quarters or all white and is used to measure water turbidity and euphotic zone depth. The depth of the euphotic zone is usually 3 times greater than that of the secchi depth. The secchi disk is normally complemented with a light probe, as the values obtained for the secchi depth a usually ‘user’ biased.

Figure 1.9 'Bongo' zooplankton nets

 

Zooplankton net - this is a 200 micron net with a screw bottle attached to the end. It is towed behind a vessel at about 1.5 knots. It is used to sample surface waters to determine zooplankton abundance per metre cubed. The zooplankton net used on RV Bill Conway had a diameter of 53cm and mesh size of 200microns. Notes: the diameter of the opening of the zooplankton net used was 53cm. The area of the net opening was calculated by [(53/2)2]*π=0.2202m2. To calculate the volume of water through the net in the tow time: Final flowmeter reading  * 0.3 = distance travelled horizontally in tow time (m). Note: 0.3 is a fixed value converting revolutions into distance travelled.

T/S Probe - T/S stands for Temperature/Salinity, with the probe working in the same way that temperature and salinity are measured on the CTD system. Salinity is measured by the conductivity of the water column and temperature being measured from a platinum thermistor.

METHODS

The following methods were used to collect and analyse samples to ascertain levels of chlorophyll, silicon, nitrate, phosphate and oxygen - methods described below were relevant to both estuarine and offshore data.

Chlorophyll - Sub samples (60ml) extracted from the Niskin bottle samples collected by the CTD rosette system and surface pump were filtered with the filters collected and placed in 6ml of 90% acetone solutions over night in a fridge. The acetone solutions having been left over night to dissolve chlorophyll from phytoplankton collected on the filters were analysed for fluorescence levels to determine amounts of chlorophyll and therefore relative phytoplankton abundance. A calibrated flourometer  with pure acetone(serial….) was used to determine fluorescence of the samples (decanted into flourometer cuvettes), normalized chlorophyll concentrations taking into account volume of acetone (v) and media filtered (V) can be found by applying the following equation with the flourometer reading (P, in µg/L): Chl. In sample water = P x (v/V)

Nitrate - Samples were collected at a number of depths at known CTD stations using Niskin bottles mounted on a rosette frame.  Sub samples were filtered into glass bottles, left overnight and refiltered to ensure no obstruction to the analytical machinery. All vessels used were rinsed with a small amount of each sample. 2-3 mL of sample was added to a system of flow-rated PVC tubes with salt water and 2 reagents and through a catalyst of copperised cadmium, converting nitrate to nitrite. Values for transmission for the subsequent solution were measured and converted to millimetres using a chart recorder. The instrument used had a lower detection limit of 0.1µM

Phosphate - Samples were collected as for Nitrate above and prepared by the addition of 1ml of a mixed reagent, bottles were then incubated for one hour to allow colour to develop and a spectrophotometer was then used to determine absorbance of light by the solution. A calibration graph created from samples of known phosphate concentration was then used to determine the phosphate concentration in the water samples.
Silicon - Again samples were collected as above and subsequently prepared by pipetting 5ml of each sample into a tube and adding 2ml of Molybdate reagent. After ten minutes 3ml of the mixed reducing agent (MRR-metol sulphite, oxaic acid, sulphuric acid [50%v/v] and milli Q water in the ratio 10:6:6:8 respectively) was added and the solution left to stand for two hours for full colour development. A spectrometer was used to determine absorbance of light by the solution. A calibration graph created from samples of known Si concentration was then used to determine the silicon content in the water samples.

Oxygen - The Winkler method was utilized to determine Oxygen Concentration as outlined by Grasshoff et al (1999). The reagents added to the samples were 1ml manganese chloride and 1ml alkaline iodide, in respective order. This was done immediately after extraction from the Niskin bottles so to limit biological processes within the sample bottles. Care was taken to ensure no air bubbles were present in the sample bottles when inserting bottle stopper. Samples were then stored and kept cool in a bucket of seawater.

Zooplankton Trawl - three trawls were carried out, 2 towed from the Conway and a third from a pontoon as follows:

 

Table 1. Locations of zooplankton trawls

 

1

2

3

Start

Lat: 50° 08.818N  Long: 05° 01.518W

Lat: 50° 10.316N  Long: 05° 01.706W

Lat: 50° 14.404N  Long: 05° 01.878W

Finish

Lat: 50° 08.818N  Long: 05° 01.630W

Lat: 50° 10.402N  Long: 05° 01.976W

 

Meters Travelled Horizontally (M):

47.1

176.7

2.4

Volume Of Water Filtered (M3):

10.41562

39.07518

0.530732

Irradiance: A light meter was used to determine light attenuation in the water and deployed at each of the CTD stations in the estuary, for the offshore data a light meter was attached to the Rosette Frame. To determine the light attenuation (k) at each depth the following equation was used to determine the irradiance(ez) at each depth: Ez=eo-ekz 

Secchi Disk

The light meter data were corroborated  using a secchi disk to monitor light attenuation at depth in both the estuary and offshore at each of the stations at which the CTD was deployed. The depth at which the secchi disk was no longer visible from the surface was recorded and used to calculate the k value using the following equation: 

K (light attenuation)=

The CTD, measuring temperature, conductivity (salinity) and depth, is attached to the Rosette Frame with the mid point of the Niskin Bottles being 65cm above the CTD for RV Bill Conway and 1m for RV Callista hence all bottle sample measurements had to be adjusted to reflect the actual sample depths. Also attached to both Rosette Frames, at the same level as the CTD, were fluoremeters, measuring fluorescence and transmissometers, measuring SPM (suspended particulate matter) or turbidity. The RV Callista Rosette Frame also had a light meter. Measurements from these instruments was recorded directly to a computer programme and subsequently manipulated to provide the results below.

ADCP

The ADCP was used to conduct a number of transects during the estuarine study and on a continuous basis during the offshore study.

ESTUARY  ESTUARY RESULTS   ESTUARY DISCUSSION

INTRODUCTION

The aim of this study was to investigate the interactions within the estuary between the chemistry, biology and physics and to understand the horizontal and vertical changes within the estuary characterised by the mixing properties of the freshwater and seawater fronts.     

 

Data were taken from various stations along the estuary spanning from the mouth at Black Rock (50o08. 444N, 5o01.098W) to the pontoon at King Harry Reach (50o14. 444N, 5o00.834W), allowing for a comparison of the behaviour horizontally up the estuary. Data collection began in the morning of 30 June 2010.

At 5 stations up the estuary (as marked on fig. 2.0) CTD profiles (mounted on a rosette system with 5 Niskin bottles) were taken, with Niskin bottle sample collection for nutrient analysis and data on fluorescence and transmission taken from rosette mounted sensors; at 4 of these positions ADCP transects were taken across the estuary to determine the physical properties. At each station light meter readings were taken with depth to determine light attenuation (k) values. The light meter data were corroborated using a secchi disk whereby the depth at which the secchi disk was no longer visible from the surface was recorded and used to calculate the k value.

Due to its Ria status, the estuary receives relatively small freshwater input, so collection of freshwater end member samples was done an hour before high tide on the 30/06/10 from Truro. This extra data will allow for estuarine mixing diagrams to be made to study the behaviour of nutrients in the estuary, along with the vertical profiles, light attenuation data and data on plankton.


CTD Data were collected at 5 stations, details of these stations, conditions and tides were as follows:

 

Table 2 CTD station locations, weather and tide information

CTD Stations

Lat

Long

 

 

Tide times

 

 

Figure 2.0 Map of CTD stations

1

50°08.444' N

5°01.574' W

Wind speed

4.9knots

0149

Low

1.2m

2

50°09.605' N

5°02.147' W

Direction

SW

0723

High

4.8m

3

50°10.373' N

5°01.656' W

Sea state

Slight

1401

Low

1.3m

4

50°11.839' N

5°02.611' W

visibility

good

1930

High

5.0m

5

50°14.444' N

5°00.834' W

Cloud cover

8/8

 

 

 

 

Four transects were designed cross-sections of the lower, middle and upper estuary at the following locations.

Table 3 ADCP transect locations

 ADCP Stations

Lat/Long (START)

Lat/Long (FINISH)

Notes.

Lower

Transect 1 (East-West)

50o08.440N  05o01.098W

50o08.561N  05o02.495W

8:32-8:48 GMT

Transect 2 (East-West)

50o10.417N  05o01.430W

50o10.224N  05o02.646W

10:30-10:43 GMT

At 338m, 020o alteration towards port.

Middle

Transect 3 (West-East)

50o11.377N  05o01.805W

50o11.339N  05o02.897W

12:06-12:17 GMT

Upper

Transect 4 (East-West)

50o12.267N  05o02.192W

50o12.342N  05o02.299W

14:12-14:16 GMT


ESTUARY RESULTS

Figure 2.0 Chlorophyll results

Chlorophyll - at Station 1 chlorophyll concentration decreased from the surface to 4m, then increased to a peak at 6m. After this peak, the chlorophyll concentration remained constant with depth. The same general pattern was repeated at Station 2, although both the minimal and maximal peaks were deeper. This could indicate that the thermocline deepened between Stations 1 and 2. Station 1at Black Rock is the most seaward of the stations and marks the mouth of the Fal estuary. Chlorophyll concentration in Station 3 shows an increase with depth, to a maximum of 1.2uM at 4m followed by a sharp decrease to 0.9µM at 8m and a further slight increase with depth. Station 4 shows a steady chlorophyll concentration from surface to 2m, then a sharp increase until a peak at 6m. Below 6m, the concentration remains relatively constant. Station 5 shows a constant rate of decrease from surface to 4m.

Figure 2.1 Nitrate results

Nitrate: the highest NO3 concentration (424.61µM) can be found at the riverine end member, likely to be due to the introduction of anthropogenic inputs such as sewage and agricultural run-off. Towards the lower end of the estuary samples plot below the TDL indicating nitrate is being lost from the system and behaving non conservatively. The marine end member (35.07) has a concentration of 1.04uM suggesting there is  rapid removal of nitrate from the estuary due to planktonic blooms, where nitrate is being used as an essential element towards growth.

Figure 2.2 Phosphate results

Phosphate: the data show that phosphate is acting conservatively in the estuary. One value was discarded in order to produce the phosphate mixing diagram, which was an anomalous result. This came from bottle 10 which was one of the random samples from the onboard pump. This result was determined to be an outlier, is still visible in the raw data but not included in the final estuarine mixing diagram.

Figure 2.3 Silicon results

Silicon: Most plots lie along the Theoretical Dilution Line showing that Si is conservative with respect to the Fal estuary. This means that Si concentration is solely dependant on mixing processes, and that there is no addition of removal by any chemical or biological processes.

Dissolved oxygen saturation has been plotted against temperature, salinity and chlorophyll concentration for each station. Data have been calibrated and depths have been corrected. Falmouth estuary is predominantly saturated in dissolved oxygen with only the seabed at the riverine end of the estuary showing under saturation. *Note samples were not taken at station 4 due to sampling limitations, an ADCP profile only was taken.

Figure 2.4 Station 1 Temp °C

Station 1 shows a well mixed water column at the estuary mouth, with salinity stable at 33.9 and temperature range of 15°C at surface decreasing to 14°C at 15m depth where it remains stable with depth.  Dissolved oxygen saturation decreases with depth to a minima of 100.3% at 15m, then increasing to a maxima of 104.7% at 20m. Chlorophyll increases with depth from 2.51μg/L at 4m to 7.27μg/L at 20m.

Figure 2.5 Station 2 Temp °C

Station 2 shows a strong thermocline at a depth of 2.3m, with a drop of temperature from 15.7°C at surface to 14.7°C. Salinity increases from 31.9 at the surface to 35.3 at depth. There is a dissolved oxygen saturation maxima of 107.1% at 5m, this datum point is erroneous and could be due to machine error. Ignoring this datum point, the general trend shows water column is oxygen enriched with a lower saturation of 102.5% at surface increasing to 105.2% at 10m and 104.8% at 20m. Chlorophyll concentration increases with depth except for a low at 5m, just below the diurnal thermocline.

Figure 2.6 Station 3 Temp °C

Station 3 shows increasing salinity with depth, increasing from 33.4 at surface to 33.9, with a halocline at approximately 5m. Temperature decreases with depth in the mixed layer down to 10m, where temperature is stable with depth at 14.4°C. Oxygen saturation increases with depth down to 15m with saturation of 105.7%, and decreases to 104.8% at 20m. Chlorophyll concentration follows a similar trend with a surface value of 6.51μg\L increasing to 9.67μg/L at 8m and decreasing to 7.71μg/L at 20m.

Figure 2.7 Station 5 Temp °C

Station 5* the most riverine station, was shallow with a sampled depth of 4.3m. Over this depth salinity increases with a value of 31.1 at surface to 31.6 at 4m. Temperature drops from a surface temperature of 20°C to 19.6°C at 4m. Only 2 water samples were collected, one at surface and one at 4m. Surface oxygen saturation is the highest of all the stations with a saturation of 105.7%, this drops to the lowest saturation level of all stations, with a saturation of 98.7% at 4m. The chlorophyll concentrations are the highest of all the stations with 107μg/L at surface dropping to 79.6μg/L at 4m.

Zooplankton Trawl

Table 4 Zooplankton Trawl

Tow Number

Flowmeter Reading start (rotations)

Flowmeter Reading end (rotations)

Final Flowmeter Reading (rotations)

Bottle Number

Water Through Net in total tow time (m3)

Figure 2.8 Zooplankton abundance

The first two zooplankton tows were 5mins in duration, the last 3mins as it was taken alongside a fixed floating pontoon.

1

12951

13108

157

B

10.39

2

13109

13698

589

C

38.91

3

13702

13710

8

A

0.53

Overall zooplankton abundance is much higher in tow 2 than any other tow; this tow was carried out mid estuary, and for a length of 176meters, with 39m3 of water being filtered. Tow 1 then has the second highest overall abundance. This tow was carried out at the mouth of the estuary; the tow was 47m long, with 10m3 of water being filtered. Tow 3 was carried out at the head of the estuary and displays the lowest overall abundance.      

Across the species identified and quantified, a pattern can be seen in abundance across the sites. Copepods have the highest abundance within all 3 tows, being most abundant in tow 2, with 8675 copepods per m3. Certain species were also only found at the seaward end of the estuary; these species in particular are Ctenophores, Echinoderm larvae, Appendicularia, Mollusca larvae and Siphoniphorae, with neither species being found at the head of the estuary.

Horizontal distance (m) * area of net opening (m2) =  volume of water through net (m3).

Phytoplankton trawl

The phytoplankton counts have been conducted twice for each sample, except for the one collected at station 2, for which a single count has been made. The phytoplankton counts for all the samples range from 2 cells in sample 1 to 24 in sample 3. 8 species have been recognised under microscope analysis, with Karenia mikimotoi being identified at all stations, while Alexandrium, Dinophysis and Polytrictos have only been found in one over three stations.  Up to 16 Alexandrium species of phytoplankton were identified on the first count in sample 3; the occasional bloom of this type of phytoplankton is known as “red tide”, and occurred for the first time in 1995 in the upper estuary, as an indirect consequence of heavy metal pollution (Langston et al, 2003). The recurrence of Alexandrium in the sample collected at the pontoon could be related to the same cause.

 

Table 5 Zooplankton Abundance

Plankton Tow 1

Plankton Tow 2

Plankton Tow 3

 

N. in sample

N. in sample

N. in sample

Groups/Order

Count 1

Count 2

Count 1

Count 2

Count 1

Count 2

Karenia mikimotoi

1

 

 

2

6

1

Guinardina flaccida

 

 

 

2

1

1

Guinardina delicatala

 

 

 

 

 

1

Mesodinium rubrum

 

 

 

1

 

4

Alexandrium

 

 

 

 

16

4

Dinophysis

 

 

 

 

1

 

Polytrictos

 

 

 

 

 

1

Dinoflagellates

7

2

 

 

 

 

Total

8

2

 

5

24

12

 

Irradiance

Figure 2.9 Ez/Eo

A graph showing the natural log of Ez/Eo against depth was constructed (figure 2.10), the inverse gradient of this line was used to determine k at each station.

The light attenuation changes significantly with distance up the estuary with the attenuation coefficient at station 5 from the light meter reading almost an order of magnitude greater than that at station 1.

Irradiance: A graph showing the ratio of Ez/Eo was constructed and displayed in figure 2.9.

Secchi disk results show differences between the k values of the secchi disk and light metre between stations have a range of 0.27m-1.

Table 6 Comparison of Secchi disk and light meter readings

Station

k m-1 (2d.p.) secchi disk

k m-1 (2d.p.) light meter

1

0.21

0.20

2

0.32

0.25

3

0.37

0.26

4

0.46

0.44

5

1.13

1.51

 

Figure 2.10 Ln (Ez/Eo) against depth

Figure 2.11 Station temp profiles

The highest surface temperature reading is at station 1 (20oc) due to the high volume of warm freshwater entering near Truro.  Water column temperature decreases downstream as it is mixed with the colder water from Falmouth Bay. Below 2m, temperature decreases by approximately 1oC for each station.

Stations 1, 3 and 4 show little variability in salinity, with a range between 34.5 and 35.2. Station 2 has a lower salinity in the surface waters (31.5) which decreases with depth below 1.5m. Station 5, has the lowest salinity reading (31.1) as it is closest to the riverine input. Salinity increases below 3m as the underlying saline water mixes with the overlying fresh water.

Figure 2.11a Station

salinity profiles

Figure 2.12 Station fluorescence profiles

Fluorescence is generally consistent with depth for stations 1-4, with a small increase in values collected from stations further upstream. Station 5 has a much greater fluorescence reading, with a maximum value of 0.59v at 0.3m.

Turbidity decreases with stations further upstream, although remains consistent with depth. Station 1 has the highest turbidity reading (4.47v) due to its proximity to the tidal influence at the marine end of the estuary. Station 5 has the lowest turbidity reading (3.75v) although increases to 3.9v at 5m

 

 

 

 

 

 

Figure 2.12a Station

turbidity profiles

ADCP

At the lower estuary (Transect 1), the flow of the ebb tide shows a strong bias  towards the west bank of the estuary which is clearly shown by the Stick Ship Track for Transect  1(Fig. 2.14). This circulation pattern is present further up the estuary, shown by the Ship Stick Track for Transect 3 (Fig. 2.16). The Velocity Magnitude Profile for Transect 1 (Fig.2.15) illustrates that the area of high current energy is restricted to the western bank and that there is a large area of low energy present in the centre and east of the estuary. The Velocity Magnitude Profile for Transect 3 (Fig.2.17) indicates that further up the estuary the ebb flow, while still showing a bias for the west bank, makes up a greater proportion of the water mass within the estuary. The Ship Stick Track for Transect 3 (Fig. 2.16) and the Velocity Magnitude Profile for Transect 3 (Fig.2.17) also illustrate that the net water flow in depths of less than approx. 5m is retarded compared with the unidirectional flow  of the ebb-tide.

Figure 2.13 Velocity direction profile for transect 1

Figure 2.14 Ship stick track for transect 1

Figure 2.15 Velocity magnitude profile for transect 1

Figure 2.16 Ship stick track for transect 3

Figure 2.17 Velocity magnitude profile for transect 3

Richardson Number: The circulation pattern has important consequences for the Fal estuary. The Velocity Magnitude Profile for Transect 1 (Fig. 2.15) and Velocity Direction Profile for Transect 1 (Fig. 2.13) suggest that the ebb flow moves on top of a stationary water mass generating two-layer flow in the estuary. Two layer flow appeared to be present on the east bank of Transect 2 and Transect 3 also. Using the UNESCO Equation of State (1980) to covert the Temperature and Salinity readings from the nearest CTD profile to density and the magnitude and direction of the flows from the ADCP data, the Richardson number (Ri) was calculated to determine whether the circulation pattern of the estuary enforces mechanical mixing of the water mass specifically during an ebb tide.

The Ri numbers are all below 1.0, suggesting that the water flow in the estuary is turbulent. The Fal estuary is a combination of a fully mixed and partially mixed estuary (Table 7).

The circulation pattern also has important pollution and other consequences for contaminants in the estuary. Any river source pollutants are going to be restricted to the west bank of the estuary while the effects of any marine source pollution will be concentrated on the eastern side. 

The Ri number was then plotted as a function of depth from the CTD profiles conducted at the corresponding ADCP transect (fig 2.17). The plot of Ri profile of CTD profile 1 for T1 (Fig. 2.17.T1.4) illustrates a laminar flow at the surface <3m that changes to a turbulent/partially mixed water structure >3m.  It should be noted that the CTD profile data used to plot of Ri profile of CTD profile 1 for T1 (Fig. 2.17.T1.4) was collected from another group because of the anomalous readings collected by CTD profile 1. Further up the estuary the plot of Ri profile of CTD profile 2 for T2 (Fig. 2.17.T2.4) shows a reversed physical structure. The majority of the water column, from the surface to 15m is partially mixed/turbulent . Below this depth, layers of laminar flow are present and the plot of Ri profile of CTD profile 3 for T3 (Fig. 2.17.T3.4) shows a similar physical structure. At the furthest point up the estuary, the plot of Ri profile of CTD Profile 5 for T4 (Fig. 2.17.T4.4) shows that the water column is turbulent with a partially mixed layer along the estuary bed. A recurring feature of the Fal estuary that the Ri plots have identified is that physical structure of the water column alternates between turbulent and partially mixed. Importantly, there no bottom boundary turbulence layer in the estuary as might be expected. Fig. 2.17.T2.4-Fig. 2.17.T4.4 all clearly show bottom boundary induced turbulence is suppressed by a laminar flow which is a characteristic of partially mixed estuaries (Geyer, 1993).

Table 7 Richardson numbers for Fal Estuary ADCP transects

Transect

CTD Profile

Richardson number (Ri)

1

1

0.07 Turbulent

2

3

0.74 Partially mixed

3

4

0.59 Partially mixed

Figure 2.17 Profiles for Ri as a function of depth for transects 1 to 4. (T1.4, T2.4, T3.4, T4.4)

Fal Estuary Residence Time  

Smean = Mean Estuarine Salinity – Calculated from Stations 2,3,4 and 5 as  equipment failure at station 1 prevented correct results.

 Ssea = Sea Salinity – Arbitrary figure of 35 used due to equipment failure at station 1 prevented collection of true marine end member

 Vtotal = Total volume of estuary (m3), R = Discharge rate of estuary (m3s-1) (Rough estimation from Admiralty charts 5602.5 & 5602.3).

 The rate of discharge of the estuary was calculated as a 10 year average of July from 1999 – 2008

Year

Rate of discharge for July (m3s-1)

1999

0.766

2000

0.988

2001

0.926

2002

1.006

2003

1.131

2004

0.678

2005

0.648

2006

0.586

2007

2.051

2008

1.948

AVERAGE

1.074

NOTE: Data was collected from the Centre for Ecology and Hydrology archives for the Fal station at Tregony. 

 = 687949.4s

Tres = 7.96 days

 Due to naturally low rate of river discharge, the Fal estuary has a long residence time. This has significant consequences as pollutants from upstream tailings dams and sewage outfalls are confined to the estuary for prolonged time periods. The recent low precipitation levels mean that for 2010 the residence time can be expected to be even longer, exacerbating the detrimental effects of the upstream pollution inputs.


ESTUARY DISCUSSION

The 5 depth profiles show a slight decrease in salinity with distance up the estuary, this will be because of a decrease in the saline member further up the estuary and an increase in the freshwater member. Fig. 2.9-2.12 show the vertical profiles of salinity, temperature, fluorescence and transmission at the 5 CTD stations within the estuary. The position of these stations is noted in table 2. The fluorescence profiles show that stations 1-4 have relatively similar levels of phytoplankton (remaining ~ 0.15-0.2v), with some increase with distance up the estuary and decrease with depth at station 4. Station 5 however shows much higher levels of fluorescence (ranges between 0.4 – 0.6v), this may be down to the decreased size and depth of the estuary at this point leading to increased nutrient concentrations allowing for increased phytoplankton growth, this apparent increase in fluorescence may also be helped by the decreased amount of grazing zooplankton (see fig. 2.8).

 

Transmission: decreases with distance up the estuary, ranging from 4.2 - 4.4v at stations 1-4 to 3.8v at station 5, showing increased levels of turbidity, this may be down to the shallow nature further up the estuary leading to increased sediment suspension or simply may arise from the increase organic input and phytoplankton at this point in the estuary; it’s worth noting that these data agree with the secchi disk and light meter data of greater light attenuation with depth further up the estuary (see figs. 2.9 and 2.10).

 

Salinity: behaves in a typical manor decreasing with distance up the estuary from the seaward member, however with the small freshwater input the variation is small, from ~35 at the stations in the main Carrick Roads channel (other than the anomalous values from station 2) to ~31 at station 5. The station 1 data were taken from the 30/06/10 Callista data set as the Bill Conway CTD initially gave clearly incorrect values (~19). The estuary seems relatively well mixed as there is no sign of a strong halocline, other than at station 2 which seems anomalous, however could be data from a surface freshwater pocket.

 

Temperature: behaves similarly but in inverse, with distinct differences with distance up the estuary. The thermocline changes in depth, shallowing with distance up the estuary as temperature increases. The thermocline seems a lot less prominent at stations 4(4*)-5, with less than a 0.5oC change with depth, as opposed to the 1-1.5oC change with depth at stations 1-3.

 

The dominant force within the estuary is tidal mixing, so where station 1 was taken at slack water of the high tide, there was no mechanical mixing occurring allowing a slight thermocline to form, as tidal currents increased throughout the day and as data were collected further up the estuary the temperature profiles show a more homogenous water column, with relatively little fluctuation.

 

The CTD data show typical partially/well mixed estuarine behaviour, with the values of components as a function of tidal mixing and the meeting of seawater and freshwater fronts.

 

Coriolis: From the data recorded by the ADCP, the ebb flow the Fal estuary can be seen to be restricted to the western bank of Carrick Roads. This corresponds with the circulation patterns of large bays and enclosed seas in the Northern Hemisphere where the coriolis force generated by the Earth’s rotation directs tides to circulate anti-clockwise. It can be inferred that the flood tide of the Fal estuary enters along the east bank and, while travelling up the estuary, is restricted to the eastern side of Carrick Roads. The tide then turns at the top of the large water mass above Transect 3 an the ebb tide leaves along the western bank. At Transect 1 the ebb tide is forced into the western English Channel. The ADCP Transects should be repeated on a flood tide to confirm that this phenomenon is indeed present. However, it should be noted that coriolis force is NOT the single determining factor in terms of the circulation pattern of this estuary.

To see if the coriolis deflection could be influencing the circulation of the estuary a simple equation can constructed. Where a is the coriolis acceleration, t is the length of time of the deflection.  If result  is greater than the width of Carrick Roads the coriolis deflection is able to act for a long enough time over a large enough distance to could be a contributing factor to the circulation pattern of the estuary (width of Carrick Roads approx. 1.7km).

a= fv = 10-4 x 0.5m

t2 = (2.16 x 104)2 = 4.67 x 108s

 = 11.675 x103m   = 11km. Coriolis can effect the tidal flow.

The effect of coriolis is most likely balanced by a pressure gradient across the estuary. The water on the eastern side of Carrick Roads is a greater density than the water on the western side because the highly saline seawater that enters the estuary on a flood tide is restricted to the east bank whereas the low salinity freshwater discharge is restricted to the western bank. This density difference between the water on both sides of Carrick Roads forms a pressure gradient across the estuary that balances the effect of the coriolis deflection. Other factors, such friction, could balance the effect of the coriolis deflection.

The effect of coriolis is most likely balanced by a pressure gradient across the estuary. The water on the eastern side of Carrick Roads is a greater density than the water on the western side because the highly saline seawater that enters the estuary on a flood tide is restricted to the east bank whereas the low salinity freshwater discharge is restricted to the western bank. This density difference between the water on both sides of Carrick Roads forms a pressure gradient across the estuary that balances the effect of the coriolis deflection. 2

OFFSHORE OFFSHORE RESULTS  OFFSHORE DISCUSSION

Introduction

The aim of this study was to determine how the vertical profile of the water column changes at a set offshore point over a 6 hour period.  The time series data gained will allow a study on the change in thermocline depth, and biological and chemical activity with respect to this change throughout a tidal cycle.

Table 8: Weather and Tidal data for 3 July 2010

Wind speed

4.9knots

Direction

SW

Sea state

Slight

visibility

good

Cloud cover

3/4

0318

Low

1.5m

  Figure 2.18 Location of offshore station

0912

High

4.5m

 

1527

Low

1.7m

 

2117

High

4.6m

 

 

A time series of data were taken over a 6 hour period on 3 July 2010 at (50. 05. 11 N, 4. 54. 95 W) to look at the structure of the vertical water column, phytoplankton and zooplankton present at specific positions within the column. Vertical profiles were taken using a rosette system with 5 Niskin bottles (see methods) a fluorometer, transmissometer and CTD at 30 minute intervals to a depth of ~60m. Surface to depth data were analysed to determine sample collection points where changes were seen in the water column throughout a tidal cycle. A continuous ADCP profile was taken allowing for judgment on the changing zooplankton population and tidal flows beneath the vessel.

A continuous ADCP profile was taken allowing for judgement on changing zooplankton population and tidal flows beneath the vessel. Data on tides and weather was taken into account when choosing the station position, along with an abundance of biological activity present; large number of diving seabirds feeding indicated areas of food, in turn indicating fish abundance and therefore high levels of phytoplankton. There was a moderate westerly wind, with easterly tidal flow throughout the morning to one hour before slack water at 1527 GMT. Niskin bottle samples were taken at profiles on the hour, with zooplankton vertical tows taken at hourly intervals on the half past profiles.

The transmissometer and ADCP are able to identify areas of higher SPM (suspended particulate matter) indicating zooplankton or suspended sediment. Using the data from the downward CTD profile the depth of zooplankton sampling was determined from maxima in turbidity below the chlorophyll maxima (that were inferred as being layers of zooplankton).

Ideally a closing zooplankton net would have been used, however this was unavailable so was replaced by two vertical trawls, one beginning below the maxima to surface, and one above to surface, so these abundance values could be subtracted to find the approximate zooplankton abundance in the layer of interest. A ‘bongo’ system was used, with a 100µm and 200µm mesh net.  Secchi disk depths to determine the euphotic depth at the station throughout the day were also taken on the hour.

  For a description of equipment used and analytical methods please follow the METHODS link.

OFFSHORE RESULTS
Chlorophyll: The chlorophyll results from all CTDs show a deep chlorophyll maximum in the morning, shallowing towards the middle of the day and then deepening again in the afternoon. In all CTDs sampled, the chlorophyll maxima equates to approx. 1.4μg/L. All three CTD profiles also show low levels of chlorophyll at the surface. CTD1 shows a small chlorophyll peak at ~12m depth, which correlates with nutrient graphs. After a small decrease, the main peak occurs at ~20m, below which the chlorophyll decreases with depth. The second CTD shows that the chlorophyll maxima has shallowed to ~15m and shows a very sharp decrease with depth, but within a few metres the rate of decrease with depth slows and the chlorophyll concentration continues to decrease until the end of the profile, where an increase occurs. The third CTD shows a deeper chlorophyll maximum than CTD2, and then follows the general trend of decreasing chlorophyll with depth common to all the stations.

Figure 3.1 CTD Chlorophyll profile

Nitrate: The data show that at the beginning of the day, the nitrate concentration showed a clear minima at 12.23m, which correlates well with the chlorophyll data. The general trend followed by the data is decreasing concentrations from surface to the thermocline, with the minima being just above the thermocline, then either increasing or remaining the same with depth below the thermocline.

Figure 3.2 Nitrate profile

Phosphate: The graph indicates that phosphate concentration increases with depth. The maximum phosphate concentration (0.31uM) recorded is found at 19.8m

Figure 3.3 Phosphate profile

Silicon: The silicon concentration is plotted against depth for stations 1,2 and 3. All transects show an increase in silicon concentration with depth, ranging from values of 0.007 µM to about 0.011 µM, down to a maximum depth of 50 meters at station 1. The rate of change of silicon concentration with depth varies within each transect: at transect 1, the silicon concentration remains approximately constant throughout the top 13 meters; it increases significantly between 13 and 24 meters, roughly reflecting the thermocline, while below 24 meters the concentration decreases linearly at a lower rate. The Si concentration trend at station 2 shows similar behavior to station 1, with a constant Si value throughout  the surface mixed layer down to 10 m depth, followed by a significant increase at the thermocline (approximately between 10 m and 24 m), going back to constant concentration below 24 m, reflecting a well mixed bottom layer. The trend in transect 3 shows an increase in Si concentration with depth from surface to 20 m depth.

Figure 3.4 Silicon profile

Oxygen: The percentage saturation of dissolved oxygen is highly variable at the near surface waters ranging from 91-109% within the top 3 metres. There is then an area of high percentage saturation at around 10-20m depth, though the range is still relatively large with values from 99 to 111% with no specific correlation between them. Percentage saturation decreases sharply at 20-30m, just below the thermocline, with values of 82-85% saturation. After this the concentration increases slightly to around 91%at 31m and decreases steadily to 89% at 47m.

Figure 3.5 Oxygen profile

Temperature: Throughout the day the defined thermocline shallows dramatically (fig. 2.1) from ~25oC at 0840GMT to 18oC at 1330GMT, the temperature change in the water column remains relatively constant, so the data have been displayed as relative thermocline shape, with the values of each respective profile +1oC of the previous. Surface waters remained at ~16.4oC, changing to ~11.7oC at the thermocline.

Figure 3.6 Time Series temp profile

Zooplankton tow: The offshore zooplankton data were inconclusive, with clear errors that may have arisen from the different methods to evaluate the depths of the tow nets. The first tow was with a weight, with the second tow without and the rest was with a depth meter so the actual depths of the tows are unknown and in some cases negative values for zooplankton numbers were found at depths with low transmission (high turbidity) readings. This data set does show however the taxas present even if the depth of the tows and zooplankton maximas and abundance per unit volume cannot be determined. As with the estuary, the offshore water column was dominated by copepods, with much higher numbers of Hydromedusae jellyfish. There was seemingly less Cirripedia larvae, but similar amounts of copepod nauplii, Appendicularians and Cladocera.
Turbidity: Turbidity levels in the surface 10m are constistent with depth for all 4 CTD stations, with a small decrease around 10m. The turbidity for all 4 stations then increases to above the surface levels, at between 16 and 24m depth. The turbidity falls between 25 and 27m, increasing again to another peak between 29 and 44m. In CTD3 this peak is not so sharply evident as in the other profiles, and the eventual peak is delayed to below 40m. Below 44m the turbidity decreases roughly linearly with depth to a minimum at the bed in all profiles apart from CTD3, which shows high turbidity at the bed.

Figure 3.7 Turbidity profile

Fluorescence: The general trend followed by the fluorescence data are of low surface fluorescence, increasing to a peak roughly corresponding to the depth of the 1% surface PAR depth, then a decrease with depth below this. 1% surface PAR marks the bottom of the euphotic zone, where the most common wavelength of light is visible blue light. Red light is attenuated in the surface μm layer whilst blue light persists to the bottom of the euphotic zone.  CTD3 shows a main fluorescence peak around 20m and a smaller peak between 40m and 50m depth, which then sharply decreases again to follow the trend. CTD4 shows a main peak in fluorescence at between 15 and 20m, and another smaller peak just below 60m.  CTD4 also shows the highest surface fluorescence, possibly as this CTD was taken at 1330 GMT when the surface PAR was greatest.

Figure 3.8

irradiance profile

ADCP

A continuous ADCP was run for the duration of the time series (5hrs 16mins) to establish the physical properties of the water column. This enabled the investigation of any correlation between the changes in the hydrodynamic conditions on the variation of a number of parameters (Temperature, Salinity, Chlorophyll, etc.) that were measured over the course of the time series by comparing the ADCP data with CTD data sets that were taken at the same times during the time series. The continuous ADCP data were then separated into 5 data sets: A1, A2, A3, A4 and A5:

 

Table 9: ADCP transect information

ADCP Data Set

Time Start (GMT)

Time End (GMT)

Latitude

Longitude

CTD Profiles

A1

08:55

10:04

50o05.023 N

04o55.085 W

1/2/3

A2

10:04

11:12

50o05.111 N

04o54.945 W

4/5

A3

11:12

12:11

50o05.109 N

04o54.945 W

6/7

A4

12:11

13:11

50o05.109 N

04o54.941 W

8/9

A5

13:11

14:11

50o05.123 N

04o54.941 W

10

 

At the start of the time series the A1 Data Set shows that the water column is clearly stratified.The Velocity Magnitude Profile of A1 (Fig. 3.9.A1.1) and Velocity Direction Profile of A1 (Fig 3.9.A1.2) illustrates that the water column has a clearly defined physical structure. Above 20m the water column oscillates vertically as internal waves; gravity waves that oscillate within the water column rather than at the surface. Internal gravity waves are common phenomena in offshore coastal regions (Helfrich and Melville, 1986) and have a significant role in the hydrodynamic properties of the water column, increasing the mixing and transport components (Klymak and Moum, 2003). For example, upwards transport of nutrients from depths into the euphotic zone without disruption of the thermocline. The source of the internal waves is likely to be the river plume of the Fal estuary as river plumes have been identified as a source of large-amplitude internal waves in coastal seas. (Nash and Moum, 2005).

 Internal waves propagate along layers of different density in the water column and in A1 the oscillation is confined to a layer of reduced density above 20m. This indicates the presence of a thermocline at 20m.

 

The profile of the Backscatter Contour Profile of A1 (Fig. 3.9.A1.3) supports this conclusion. Fig. 3.9.A1.3 shows  a high level of backscatter in the top layer down to a depth 20m, below which there is little reflectance . The presence of a thermocline at 20m would explain this phenomenon. Dense populations of zooplankton in euphotic zone cause the high levels of reflectance recorded by the ADCP and as such, a chlorophyll max measured by the CTD would be expected at this depth. The presence of phytoplankton indicates the euphotic zone, the bottom limit of which is limited by the depth of the thermocline.

 

To quantify the physical structure of the water column for the A1 Data Set the Ri number was calculated a function of depth for the CTD profiles that were conducted between 08:55 and 10:04 average velocity recorded at each depth from the ADCP (Fig.3.9.A1.1). The plot of the Ri profiles for each of the corresponding CTD profiles for A1 (Fig. 4.0.A1.4) clearly shows the structure of the water column. The thermocline is laminar flow that is present at around 20m that separates layers of turbulent to partially mixed water dynamics. This is analogous with data collected from each of the CTD profiles 1, 2 and 3 that the water column is clearly stratified. Importantly, Fig. 4.0.A1.4 shows the continuous vertical oscillation of the thermocline over the course of the time series due to the oscillation of the internal wave.

 

To determine the strength of the stratification the Brunt-Vaisala Frequency was calculated and plotted as a function of depth (Fig. 4.1.A1.5). The Brunt-Vaisala Frequency is the maximum frequency that the internal wave can oscillate and the results conclude that the stratification is stable.

 

The next section of the time series was recorded by data set A1. In this data set the physical structure of the water columned remained very similar. The Velocity Magnitude Profile of A2 (Fig. 3.9.A2.1) and the Velocity Direction Profile of A2 (Fig 3.9.A2.2) clearly show the presence of the internal wave is restricted to the layer above the thermocline, suggesting that an upper layer and a deep layer of greater density is present. The profile of the Backscatter Contour Profile of A2 (Fig. 3.9.A2.3) illustrates that the upper layer corresponds with the limits of the euphotic zone, suggesting that the boundary between the upper and lower layers is the thermocline at approx. 20m. The plot of the Ri profiles for both of the corresponding CTD profiles for A2 (Fig. 4.0.A2.4) showed a stratified water column with the laminar thermocline sits in between the turbulent to partially mixed upper and lower levels. The Brunt-Vaisala Frequency was again plotted (Fig. 4.1.A2.5) and classified the stratification as stable.

 

As the time series progressed in data set A3 the internal waver structure was lost but the Velocity Magnitude Profile of A3 (Fig. 3.9.A3.1) and Velocity Direction Profile of A3 (Fig 3.9.A3.2) indicate two layer flow and thus, a thermally stratified water column and the Backscatter Contour Profile of A3 (Fig. 3.9.A3.3) supports this by still showing a definite euphotic zone.  However the plot of the Ri profiles for both of the corresponding CTD profiles for A3 (Fig. 4.0.A3.4)and Brunt-Vaisala Frequency (Fig. 4.1.A3.5) are less conclusive when trying to determine the level of stratification of the water column.

As the time series came to a conclusion the Velocity Magnitude Profile of A4 and A5 (Fig. 3.9.A4.1; Fig. 3.9.A5.1) and Velocity Direction Profile of A4 and A5 (Fig 3.9.A4.2; Fig 3.9.A5.2) became less conclusive when trying to determine the stratification of the water column. However the plots of the Ri profiles for both of the corresponding CTD profiles for A4 and A5 (Fig. 4.0.A4.4; Fig. 4.0.A5.4) and Brunt-Vaisala Frequency for both A4 and A5 (Fig. 4.1.A4.5, Fig. 4.1.A5.5) suggest that not only is the water column is still thermally stratified and stable, but that the thermocline is migrating upwards towards the end of the time series.

 

The time period, T, the frequency, f, and the phase velocity, c, of the internal wave were calculated for data sets A1 and A2 to see how the properties of the inertia wave change over the time series.

                 Table 10: ADCP calculated values

Data Set

T(s)

f(s-1)

c(ms-1)

Λ (m)

A1

2070

0.00048

0.15

310.5

A2

2622.9

0.00038

0.38

996.7

The wave length, phase velocity and Time period increase over time. Due to the fact that the source of the internal wave is likely to be the river plume of the Fal estuary, the tidal cycle of the estuary could be the determining factor of the physical properties of the resulting internal waves in the immediate coastal zone.

Figure 3.9.A1.1  The Velocity Magnitude Profile of A1

Fig 3.9.A1.2 Figure Velocity Direction Profile of A1

 

Fig 3.9.A1.3 Figure The Backscatter Contour Profile of A1

Fig 3.9 A2.1 The Velocity Magnitude Profile of A2

Fig 3.9.A2.2 Velocity Direction Profile of A2

Fig 3.9.A2.3 Backscatter Contour Profile of A2

 

 

     

Fig 3.9.A3.2 Velocity Direction Profile of A3         Fig 3.9.A3.3 Backscatter Contour Profile of A3   Fig 3.9.A4.1 Velocity Magnitude Profile of A4   Fig 3.9.A4.2 Velocity Direction Profile of A4

               

Fig 3.9.A5.1 Velocity Magnitude Profile of A5         Fig 3.9.A5.2 Velocity Direction Profile of A5   Fig 3.9.A3.1 Velocity Magnitude Profile of A3

 

Figure 4.0.A1.4, A2.4, A3.4, A4.4, A5.4 Richardson Number (Ri) profiles: 

  

Figure 4.1.A1.5, A2.5, A3.5, A4.5, A5.5 Brunt-Vaisala frequency profile

  

OFFSHORE DISCUSSION

Temperature, fluorescence and turbidity: In the offshore environment the water column is characterized by temperature differences as opposed to the haline controlled conditions within estuaries. This means that the data on temperature is crucial, as the thermocline determines the mixing ability of the water column. At this depth and position offshore, tidal mixing plays a small role, with surface wind mixing the defining factor for water column mixing. It is worth noting that the data were collected 2 days after a small storm event which would potentially have lead to localized upwelling from the swell and therefore a phytoplankton bloom.

The fluorescence maxima at just above the thermocline is explained by the thermocline trapping nutrients in the surface layers, the maxima in turbidity (minimum transmission values) can be explained by the presence of a migrating zooplankton population grazing on the phytoplankton bloom from below, this may be down to the shallowing of the thermocline from the movement of internal waves throughout the day compressing the stratified top layer. In the top 3 metres there is some slight diurnal thermocline formation from solar heating, however it is hardly noticeable.

Silicon: The rate of change of silicon concentration with depth varies within each transect: at transect 1, the silicon concentration remains approximately constant throughout the top 13m, for mixed by winds in the surface layer of the water column; it increases significantly between 13m and 24m, roughly reflecting the temperature changes in the thermocline, while below 24m the concentration decreases linearly at a lower rate, reflecting slight mixing in the bottom layer.

Nitrate: The highest concentration of phytoplankton is around the thermocline, which could explain the low nitrate values as the phytoplankton utilise the nitrate throughout the day. An exception is Niskin 4 (1300GMT) which shows an increase in nitrate concentration from surface to thermocline and a further increase from thermocline to depth. This accumulation of nitrate indicates that the phytoplankton are no longer using up all the nitrate available in the water column, possibly because of decreasing phytoplankton numbers. For Niskin 1 the nitrate minima equates to chlorophyll maxima indicating that phytoplankton were using up the nitrate. As the time series continued, the vertical stratification of the nitrate became less pronounced, perhaps indicating that the phytoplankton uptake rate had decreased. Throughout the morning, up until 1200 GMT, the nitrate concentration increased overall, and then decreased again from 1200GMT to 1300 GMT. This could indicate that the phytoplankton uptake rate had decreased. Overall throughout the time series, the nitrate values do not show any strong trend, however they do show small variations in concentration and maximal position within the water column, all of which can be correlated with the numbers of phytoplankton present.

Phosphate: is being used in the euphotic zone as an essential element to phytoplankton growth. As light attenuation diminishes with depth,  the abundance of phytoplankton also decreases, leading towards a build up of phosphate. The thermocline (20m) acts as a barrier, by trapping phosphate in the upper 20m, whilst also stopping the surface waters being replenished by nutrient upwellings from the deep.

Irradiance: As there was no data on surface irradiance levels and levels at depth the attenuation coefficient can only be found from secchi disk data. This data didn’t change throughout the day, which is unsurprising in an offshore environment where sunlight is not a limiting factor and remains relatively constant for extended periods. Data on changes in irradiance with depth are more relevant within estuaries, where sunlight is often a limiting factor and nutrients rarely. The offshore environment is the inverse, with a consistent euphotic depth of ~27m and k of 0.053m-1.

The current and behaviour of the water column at the time series seems to mainly be controlled by the tidal regime of the English Channel, surface wind mixing and the movement of internal waves, with upwelling and then trapping of nutrients in surface layers by thermocline formation.

 
GEOPHYSICS

Introduction

The aim of this study was to complete a bathymetric survey of the benthic environment and create a biotope map on the benthos in an area within the small bay to the west of Pendennis Point and the mouth of the Fal estuary, and determine the relationship between bathymetry and biota present at the benthos.

 In a region of Falmouth Bay a bathymetric survey was completed, using side scan sonar tow fish to determine sea bed bathymetry and a towed camera and Van Veen Grab to determine biota and sediment characteristics. The three methods can be compared to determine the reliability of data achieved from each and how an overview of the sediment biota interactions can be gained from employing all three of them.

 

The sampling took place on SV Xplorer on 7/07/10, with a side scan sonar tow fish towed behind the vessel over an area, along 4 horizontal tracts, giving an area shown in figure 4.1. Analysis of the side scan sonar read out allowed for recognition of bedforms and areas of interest in order to decide sample sites for grab samples. At applicable sites, a towed camera system was deployed to determine where in the area a grab should (if at all) be taken. Four grabs were taken at 50o06.839N, 05o05.800W (grab 1), 50o06.934N, 05o04.745W (grab 2) and 50o08.2604N, 05o03.3417W (grab 3 and 4). Grabs 1 and 2 were taken at either side of a transition zone between coarse and fine sediment, in order to compare the affect of substrate on biota. Grab 3 and 4 were taken further east along the transect in an area of homogenous ripples, grab 3 however misfired and failed to collect a representative sample.

Methods

Side scan sonar

This tow fish system was deployed from the stern of the boat using the hydraulic crane and towed along mapped transects. It allows for determination of seabed characteristics; sediment type and bedforms (see equipment for details). The side scan ran at a swath width of 150m, meaning data on the seabed to 75m perpendicular to the tow fish could be taken.

 

Video transect

At areas of interest as determined by the side scan sonar, the towed video camera was deployed using the hydraulic crane. This allowed for live streaming of footage of the seabed to the vessel and made it possible to check for applicability of site; i.e. is it a sensible site, will it damage any protected species is it worth taking a sample.

 

Van Veen Grab

Using the hydraulic crane, the Grab was deployed at the 3 sample sites, with the collected 0.5m3 of sediment placed into a collection box. This sample was then hand checked for any macrofauna, or organisms that may be damaged or easily desiccate, before being sieved through a sieve stack of 1cm, 0.5cm and 0.1cm mesh. The taxonomy of collected biota was determined before the sample was returned to the water at as similar a position as possible that the grab was taken.

 

Video Transect 1.

This transect showed a large plain of uniform ripples with large red algae present at the ripple peaks, at the peaks there was also a aggregation of bivalve shell and larger sediment.  

Video Transect 2.

This transect initially showed similar results to video 1, however the vessel moved across the transition between ripples and flat, finer sediment, with seemingly little biota.

Video Transect 3.

This was altogether different, with large red algae assemblages, and large numbers of echinoderms

Figure 4.1: Still from camera tow at site 1. Notice parallel ripples with reg macroalgae.

Figure 4.2: Still from camera tow at site 2.

Figure 4.3: Still from camera tow at site 3.

 

Grab 1

VIDEO: 50o06.840N, 05o04.800 W

 

An area of linear ripples with red macroalgae dominating at the ripple peaks and causing build up of bivalve shells. A few Asteroidea were visible with some juvenile fish sheltering in the algae.

 

GRAB: 50o06.839N, 05o04.784 W

Time: 1103GMT

Depth: 16m

 

Fauna: Maerl, Tube worms, Bivalves, Amphioxious, Amphipod

 

The grab showed the sediment to be coarse fragmented shell sand, with large amounts of dead and alive Maerl. Unidentified Polychaete tubes were found on a few larger shell and Maerl fragments.

Grab 2

VIDEO START: 50o06.932N, 05o04.811 W

VIDEO FINISH: 50o06.947 N, 05o04.722 W

 

A flat bedded area of fine sand, indicating a fast moving underlying current near the benthos.

 

GRAB: 50o06.934N, 05o04.745 W

Time: 1143GMT

Depth: 16.5m

 

Fauna: Bivalve – veneroidia, Polycheate – neridie, Dead maerl

 

Few organisms present due to low nutrient availability and unstable substrate.

Grab 3

VIDEO: 50o 08.258 N, 05o 03.335 W

GRAB 3: 50o 08.2604 N, 05o03.3417W                                            

Depth: 17.5m

Time: 1235GMT GMT                                                                             

Fauna: Ascidacia, Bryozoan mat, Red & Brown varieties of Fucus, Polychaete tubes          Maerl (living and dead)                                                                

 

Grab 4

500 08.2604 N 050 03.3417 W

 

Depth: 17.5m

 Time: 12:42 GMT                                                                                            

Fauna: Ophiuroidea sp, Nemertean worm, Nereid worms, Psmmanchus, Gastropod, Bivalve Chlamys, Ulva sp, Sipunculid

 

Note: At site 3, the first benthic grab was a partial mis-fire, so the sample volume was small. In order to gain a larger sample size, a second grab was taken.

 

Figure 4.4 Lithothamnion sp Figure 4.5

 veneroidia, Polycheate – neridie, Dead maerl

 

Figure 4.6 Maerl, living and dead Figure 4.7 Ophiuroidea sp

Site 3 marks a benthic boundary with ripple formations on one side and none on the other. One grab was taken either side of the benthic boundary. Both grab samples were dominated by Maerl, with the second grab being over 80% Maerl. The benthic substratum at this site was hard with many rocks for sessile fauna such as the Sipunculid to attach to, with fine sandy mud in between the rocks. Most of the fauna recovered from the samples were either mobile or sessile epifauna, with a few burrowing fauna also recovered. Grab 3 was dominated by Ophiuroids and Grab 4 by tube-building polychaetes. The wide variety of fauna recovered at this site can be interpreted as evidence that even the smallest change in benthic conditions can completely change the biotope.

Discussion

The data show a variation in biota with sediment type. The change in biota between sites may be down to food availability and stability of substrate, both a function of sediment type and bottom currents. Food availability changes in a variety of ways, the feeding mechanism of the organism is an important factors, whether it is a deposit or suspension feeder or detrtivore etc, meaning for example, if the organism concerned is a suspension feeder, there needs to be certain levels of suspended biogenic particulate material to feed on, or if the organism is a detritivore there needs to be certain levels of benthic detritus available. Food availability is in turn affected by sediment type, as from differing bacterial breakdown rates from different sediments.  So, sediment type is affected by bottom currents, which in turn affects the potential organisms present from the feeding methods made available, which are in turn reliant on the correct sediment type, which is controlled by bottom currents. There is therefore a very succinct feedback loop between biota and sediment which can be seen from the change in biotopes between sites. Grab 1 shows an area limited on potential substrate for algae, which therefore limits the other organisms able to survive, the lack of large numbers of echinoderms as with grab 3 shows that either the flow rate is too great, not allowing settlement of biogenic matter for suspension or deposit feeding invertebrates, and therefore not providing the diet for echinoderms. Grab 2 shows a site of little to no biota, with the fine sediment present being unable to sustain many macrofauna. This fine sediment may be present due to low flow rates, which may also be affected the present biota. The lack of substrate available explains the sudden decline in algae. Grab 3 as mentioned above seems to be the most diverse biotope, with the greatest abundance of organisms.

CONCLUSION

The estuarine and offshore environments are characterized by differing physical conditions. The estuarine environment is dependent on the mixing of fresh and seawater members, with the biological, physical and chemical interactions often a function of this frontal mixing system. If the estuary is large enough then tidal pumping will be affected by coriolis and also affect conditions by skewing the flow. Estuaries are therefore halocline controlled, with little affect from temperature. Nutrients are consistently high and rarely a limiting factor, whereas in the offshore environment nutrients are often limited in surface waters towards the end of summer, where waters masses are trapped in surface layers by seasonal thermoclines. Phytoplankton growth is therefore mainly a function of nutrient availability, whereas within the estuarine environment, where attenuation coefficients are an order of magnitude greater, light is often a limiting factor, with the well mixed tidally pumped waters relatively consistent with nutrient levels. Both systems are significantly influenced by the tides, with diurnal changes in the offshore environment from internal waves and changes in tidal currents, and with the estuarine currents dominated by tidal flow. The tide is the one factor that ties the two together and has consequences in both.

In the estuarine environment, the tide dictates influences the physical structure of the water column. The tidal flow in the Fal estuary is part of a two layer water column with the high energy tidal flow moving on top of a low energy body of water. This two layer flow was a quantified as turbulent at the mouth of the estuary and partially mixed further towards the source. Furthermore, it was established that the entire water column was partially mixed/turbulent apart from a bottom boundary laminar layer that prevented bottom-boundary turbulence, a characteristic of partially mixed estuaries.

The offshore environment was found to be a continuously stratified throughout the tidal cycle. Internal waves caused by the river plume of the estuary were shown to effect key mixing and transport parameters of the offshore water column. The properties of the internal wave (period, wavelength, phase velocity and frequency) were shown to change in accordance with the tidal cycle.

Where these systems meet, the transition zone has a variety of characteristics, common in both. This investigation has allowed for data to be taken on the differences between the systems, and the dynamic character of the estuarine environment in relation to the more stable offshore conditions.

References

Callista - http://www.soes.soton.ac.uk/resources/boats/vessels.html Accessed on 4 July 2010

Fal and Helford Management Forum, Fal and Helford: Marine Special Area of Conservation Management Scheme, Fal and Helford SAC Management Scheme, 2006, pp. 8-15

Geyer, W. 2000. the dynamics of a partially mixed estuary. Journal of Physical Oceanography. 30. 2035-2048

Grasshoff, k., K. Krenling, M. Ehrhardt (1999). Method of Seawater analysis. 3rd ed. Wiley – VCH.

Helfrich, K., Melville. 1986. Transcritical two-layered flow over topography. Journal of fluid mechanics. 178. 31-52.

Klymak, J. & Moum, J. 2003. Internal solitary waves of elevation advancing on a shoaling shelf. Geophysical research letters. 30

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

Nash Moum, J. 2005.  

Pirrie, D., Power, M.R., Rollinson, G., Hughes, S.H., Camm, G.S. and Watkins, D.C. no date, Mapping and visulisation of the historical mining contamination in the Fal Estuary,
 Cornwall
, [online], Available: http://projects.exeter.ac.uk/geomincentre/estuary/Main/intro.htm [accessed 2010, July 6th].

JNCC Fal and Helford SAC Site details, [online], Available: http://www.jncc.gov.uk/protectedsites/sacselection/sac.asp?EUCode=UK0013112 [Accessed 6 July 2010]

Map images from Google Earth www.earth.google.co.uk [Accessed 3 July 2010]

Tides and Weather: www.bbc.co.uk/weather/coast/tides/southwest [Accessed 5 July 2010]

Disclaimer: all views expressed on this website are those of the students involved and not necessarily those of the University of Southampton or the National Oceanography Centre, Southampton.

All figures and photographs have been produced by the named students unless otherwise stated