Reproduced with the permission of the Scottish Naturalist
Copyright:
May be used for private research. All other rights
reserved
By JAN KUBECKA
Hydroacoustic Unit, Department
of Biology,
Royal Holloway University of London
ANNIE DUNCAN
Hydroacoustic Unit, Department
of Biology,
Royal Holloway University of London
and ALAN J. BUTTERWORTH
National Rivers Authority, Thames
Region
Introduction
A
great deal of what is already known about the biology
of Loch Ness derives from Sir John Murray and Laurence
Pullar's Bathymetrical Survey of the Scottish Fresh-Water
Lochs (Murray
and Pullar, 1910) and from Dr. Peter Maitland's
studies in 1977-80 on the ecology of Scotland's
five largest lochs (Maitland, 1981). Although he recorded seven species of fish in Loch Ness, Maitland
suggested that only three were predominant in the
open waters of the loch - Charr Salvelinus
alpinus, Brown Trout Salmo trutta
and Salmon Salmo
salar - all largely in the top 30 m. More recently, Shine and Martin (1988), of
the Loch Ness and Morar Project, provided some evidence
of the presence of fish in the profundal zone by
netting Charr from 200 m deep with the profundal
bivalve, Pisidium
conventus, in their guts.
With the advent of commercially available
scientific echo-sounders in the 1980s, the study
of fish in large deep lochs was suddenly made much
easier. These
instruments applied the physics of sound to be able
to discriminate between 'fish' and other targets
reflecting back sound echoes, to 'size' the echoes,
and to count them with confidence in a known volume
of water. By various means, the position of the 'fish'
target in the sound beam could be determined. The volume of water which could be sampled
for fish targets was very large indeed, - in the
May 1992 survey it was about 5% of the volume of
the pelagic region of Loch Ness down to 100 m depth.
Vol 105, The Scottish
Naturalist: Organisms Detected in Loch Ness
by Dual-Beam Acoustics p180
When Royal Holloway University of London
acquired a BioSonics Model 105 Dual-Beam Echo-Sounder,
we used it to carry out several preliminary acoustic
surveys in Loch Ness for the Loch Ness and Morar
Project, and were thrilled to find that it could
be used to provide an instantaneous acoustic picture
of the horizontal and vertical distributions of
the fish populations, as well as size composition,
in some two days of field work. The present account outlines some of the
difficulties encountered, but there can be no doubt
about the power of this new tool in the hands of
ecologists.
Acoustic
Methods
Of the three preliminary acoustic surveys
carried out in Loch Ness (May 1991, October 1991 and May 1992), the May
1991 survey was the best, and it is this survey
which will be most discussed. Figure
3a (5K) shows the position of the transects
which were surveyed during two consecutive days. Each transect was surveyed by three runs,
one pelagial in 200 m of water and two runs over
the basin walls (down to 70 m depths), in which
different ping rates were applied, 2.0 pings per
second in
the deeper water and 5.0 pings per second in the
shallower depths.
The BioSonics Model 105 Echo-Sounder was
operated from a Dory-17 boat (Figure
1, 17K photo) with a 420 kHz transducer
suspended from a towing body (Figure
2, 23K photo) at 2.0-3.0 m depth and beaming
vertically down to 70-100 m, at which depth the
signal-to-noise ratio became too low. Pulse durations used were 0.8 ms and single
targets were identified by selection criteria.
The instrument was calibrated in May 1991,
by using a long-life ping-pong ball (-42 dB) suspended
below the transducer, and re-calibrated later using a tungsten-carbide standard target (-43.7 dB). The noise thresholds applied were generally
very low in depths down to 50 m, thus allowing us
to record acceptable targets down to -81 dB.
The size and densities of single targets
were processed using the BioSonics Echo-Signal Dual-Beam
Processor (ESP_DB), whilst the total acoustic biomass
was determined by the BioSonics Echo-Integration
Processor (ESP_EI). The size analysis of single targets showed
a predominance of very small targets, and we made
an attempt to discriminate between these tiny targets
(or scatterers) and real fish targets by using two
kinds of integration strata simultaneously. The two strata were the primary, which integrated
all targets from -81 dB to the largest target (fish
plus the scatterers of the scattering layer), and
the secondary strata, which integrated those targets
larger than -55 dB (fish only).
Vol
105, The Scottish Naturalist: Organisms
Detected in Loch Ness by Dual-Beam Acoustics
p181
To convert echo-integrated acoustic biomass
from volts squared to the number of fish individuals
requires a scaling factor, which is influenced by
the size of the average back-scattering cross-section
of the targets. This was 2.974*10 -8
m2/ individual for the scatterers at all depths
(0-50 m) and 1.84*10-5 m2/individual for fish targets in different depths. These values result in widely differing scaling
factors (Figures
8, 7K graph and Figure
10, 7K graph).
The biomass of the fish targets was calculated using
the echo-integrated squared voltages divided by
a mean back-scattering cross section of 4.27*10-4 m2/kg. This scaling factor comes from stunned Trout
of various sizes exposed in front of a 420 kHz transducer
(unpublished data).
Results
and Discussion
Dual-Beam Echo-Counting and Fish Sizing
Despite the reputation
of Loch Ness, the densities of pelagic targets of
the size of fish were surprising low, if one takes
a fish target to be >-55 dB or larger than 4.0
cm length (Love, 1971). When the whole size composition of acoustic
targets in the open water of the entire loch is
considered, the undoubted fish targets contributed
a very small proportion (Figure
4, 6K). In the shallower, rather sheltered and more
productive Urquhart Bay, on the other hand, the
proportion contributed by fish was strikingly greater. The average acoustic size of fish in open
water of the loch was -47 dB, which, according to
Love (1971), is equivalent to a length of 8.5 cm.
In every transect, many
hundreds of scatterer targets were detected (Figure
3b, 8K) whose target strength ranged from
-79 dB to -73 dB in 1991 (Figure
4) and -74 dB to -70 dB in 1992 (Figure
5, 6K), concentrated in the depths between
12 m and 30 m. No-one yet knows what these are, although they
are most likely to form the scattering layer described
by Shine and Martin (1988).
For May 1992,
Figure 6 (11K) provides a three-dimensional
diagram of the frequency of target strengths of
all targets in every transect from north to south. As above, this figure also shows the small proportion of fish targets
and the consistency of the peak target throughout
the entire 23-mile loch, but with a tendency towards
an increased number of targets per transect from
north to south. This finding corresponds to earlier acoustic
records of increased acoustic counts in the South
Basin, as determined by the simultaneous deployment
of seventeen vessels during Operation Deepscan in
October 1987 (Shine and Martin, 1988).
Vol 105,
The Scottish Naturalist: Organisms Detected
in Loch Ness by Dual-Beam Acoustics p185
Figure
5 compares the target strengths of the loch's
scatterer targets with a 55 mm Three-spined Stickleback
Gasterosteus aculeatus suspended in the
scattering layer itself. Even such a small fish, were it in the full
pelagial, would be a much bigger target than the
scatterers. It
is known that young Charr inhabit the open water,
since in November 1992 Charr as small as between
3.5 mm and 7.0 mm were caught by mid-water trawl
(Shine, Kubecka, Martin and Duncan, 1993). However, sampling the scattering layer depths
by ichthyo-plankton tow net (1.0 m diameter; 1.0
mm mesh size) in May 1992 caught just one Lamprey
Lampetra fluviatilis larva, one small Charr and chironomid pupae at
densities of about 0.1 individuals/m3.
Larger targets are also present in Loch Ness;
between 67 m and 85 m depth of transect 17 a large
shoal of large targets was detected (Figure
7, 8K chart). The shoal was so dense that most of the targets
were not single. Tracked single targets at the edges of the
shoal are shown and these had target strengths of
about -35 dB to -24 dB, probably with a maximum
length of one metre! Such compact shoals of very large fish might
quite well have caused earlier mis-identification
of Monsters.
Echo-Integrations
for Biomasses
Figure
8 shows a marked trend of increasing fish
density from north to south. The same trend can be seen when the acoustic
density is expressed as fish biomass (Figure
9, 7K). The density distribution of total targets
(>-81 dB or greater) was calculated by applying
one scaling factor, that for small scatterers, since
these were much more numerous than fish targets;
the density distribution of these total targets
also showed a similar north-south trend, but much
less pronounced (Figure
10, 7K).
Simple subtraction of fish volts squared
from total volts squared was too crude a process
for estimating the density of the loch's scatterers. This is probably caused by the highest densities
of scatterers appearing as multiple targets of voltage
large enough to be taken by the Echo-Signal Processor
as fish targets, something which occurs more frequently
in the more southern transects. The bias caused to estimates of fish abundance
by identifying and incorporating small scatterers
as multiple targets is not too serious, because
of their very small 'acoustic weight', and
Figure 9 (7K) can be accepted as an upper
estimate of fish abundance down to 60 m depth in
transects along the loch.
Vol 105, The Scottish Naturalist:
Organisms Detected in Loch Ness by Dual-Beam
Acoustics p192
An average depth distribution of the so-called 'plankton
densities', arbitrarily
defined as targets between -75 dB and -55 dB, is
illustrated in Figure
11 (5K), and shows doubled densities between
12 m and 22 m. These densities were calculated as the difference in volts squared
between total V2 and fish V2 and converted to numbers/m3 by a 'scatterers'
scaling factor. Densities of fish targets (>-55 dB), given
as number/ha for each 2.0 m depth stratum (Figure
12, 4K), are also concentrated between 12
m and 28 m, with few in the top 10 m water layers. These were scaled by the same back-scattering
cross section at all depths. In the absence of better information on the
identity and target strengths of the small scatterers
and the smaller pelagic fish, we cannot fully exploit
the flexibility of BioSonics software to discriminate
further between these two classes of acoustic targets
in Loch Ness, or to better calculate their relative
biomasses. The actual densities shown in Figures
11 and Figure
12 should be considered to be relative rather
than absolute, because in each class of targets
the span in acoustic size is very great (the dB
scale is a logarithmic one), so that the use of
an average back-scattering cross section can only
be crude.
Summary
1. The entire 23-mile length of Loch Ness was
surveyed acoustically in May 1991, in October 1991
and in May 1992 using a BioSonics Echo-Sounder with
an operating frequency of 420 kHz. Detailed results are given of the May 1991 survey, since this was
the best and involved acoustic runs over the pelagic
region (200 m depth) in 22 transects.
2. Acoustic densities of fish targets (>-55
dB) in the pelagic region of the loch were low,
whereas there were hundreds of small scatterers
(-79 dB to -73 dB). The average acoustic size of fish was -47
dB. The mean density (SD) of fish targets in the loch pelagial was 97 91 fish/ha down to 60 m depth. The mean biomass of these fish was 4.23 4.00 kg/ha, giving a rough average individual size of
43.6 gm.
3. The density per transect of total targets
and of fish targets increased from north to south.
4. The average loch depth distribution of both
scatterers and fish targets peaked at between 12
m and 22 m for the former and between 12 m and 28
m for the latter.
Vol 105, The Scottish Naturalist: Organisms
Detected in Loch Ness by Dual-Beam Acoustics
p193
Acknowledgements
We should like to express our grateful thanks to Royal
Holloway University of London and the National Rivers
Authority for permission to use the sonar systems. We should also like to thank the Loch Ness
and Morar Project team for our work-time on their
boat Ecos, and for their hospitality during our visits; also Mr.
R. Bremner of Drumnadrochit for the loan of his
splendid Dory-17 boat in May 1991.
References
Love, R.H. (1971). Dorsal-aspect target
strength of an individual fish. Journal of the Acoustical Society of America, 49: 816-823.
Maitland, P.S. (Ed.) (1981). The
Ecology of Scotland's Largest Lochs: Lomond, Awe, Ness, Morar and Shiel. Monographiae Biologicae, Vol. 44. The
Hague: Junk.
Murray, J. and Pullar, L. (Eds.) (1910). Bathymetrical Survey of the
Scottish Fresh-Water Lochs. Vols. 1-6. Edinburgh:
Challenger Office.
Shine, A.J., Kubecka, J., Martin, D.S. and Duncan, A. (1993). Fish
habitats in Loch Ness. Scottish Naturalist, 105: 237-255.
Shine, A.J. and Martin, D.S. (1988). Loch Ness habitats
observed by sonar and underwater television. Scottish
Naturalist, 100: 111-199.
Received May 1993
Dr.
Jan Kubecka, Hydrobiological Institute,
Czech
Academy of Sciences, 370 05 Ceske Budejovice,
Na
Sadkach 7, CZECH Republic.
Present
address: Hydroacoustic Unit, Department of Biology,
Royal
Holloway University of London, EGHAM, Surrey TW20
0EX.
Dr.
Annie Duncan, Hydroacoustic Unit, Department of
Biology,
Royal
Holloway University of London, EGHAM, Surrey TW20
0EX.
Dr.
Alan J. Butterworth, National Rivers Authority,
Thames
Region, By-pass Road, GUILDFORD, Surrey GU1 1BZ.
