2.1 Main steps involved
2.2 Practical problems arising in obtaining the required information
2.3 Behaviour pattern of fish and its effect on biomass estimates
A summary of the main steps involved in the design and execution of acoustic survey programmes is listed below:
a) Objectives: The first task is to lay down in concrete terms the objectives of the acoustic survey programme. Failure to clarify the purpose of the programme will undermine its ultimate value; in the end it may be found that the results are not what was really wanted.b) Target population to be covered: The objectives of the acoustic programme should define the target population. The table below portrays the major structure groups of the target population and types of surveys employed for their coverage:
|
Fish population/Major structure groups |
Type of surveys. |
|
1. Surface schools only |
1. Aerial surveys |
|
2. Surface and sub-surface schools |
2. Sonar surveys |
|
3.1 Submerged schools and scattered pelagic fish |
3. Echo-integrator surveys (= Echo-sounder surveys) |
|
3.2 Demersal scattered fish |
|
c) The survey area; In order to cover the target population, it is important to define precisely the survey area (volume of water) in which the target population is associated. The sample area is identified by geographical rules associating them with the surface area in the sample. Specifically, the following two characteristics are taken into account for the demarcation of the survey area, space and depth.d) Sample selection; At this stage the stratification system, the type of sample (= regular line transect sample or random line transect sample), the sample size, the size of the elementary sampling distance unit (ESDU), and the estimation of population characteristics along with their margin of error are some of the technical problems that should receive the most careful attention.
e) The information to be collected: The question about the kind of acoustic and other information to be collected should be carefully considered at the early stages of an acoustic survey programme. Acoustic survey programmes are multi-purpose in character, i.e., data are collected from different sources of information and that estimates of the survey magnitudes are calculated after a proper processing, analysis and critical analysis of the sample observations. The following Table portrays sources of information used in acoustic programmes, data collected and results obtained.
|
Source of Information |
Data collected |
Results obtained |
|
1. Echo-sounder |
Kind of fish aggregation below the sample tracks |
Spatial and vertical distribution of survey fish population.
Total biomass estimates of the target population covered by the
echo-sounder’s sampling volume (after calibration). |
|
2. Integrator |
Value of echo-envelope integral (@ values of echo-integrator readings) |
|
|
3. Sonar |
Number and size of surface and sub-surface schools |
Spatial distribution of the survey schools. Total biomass of
the survey target population. |
|
4. Concurrent fishing operations |
Species identification, length, sex, stage of maturity of fish
and other related biological information |
Species composition of survey biomass and biological
characteristics of the surveyed stocks. |
|
5. Concurrent oceanographic survey |
Thermocline, surface temperature, oxygen, etc. |
Correlation of oceanographic parameters with size and spatial
distribution of survey stocks. |
Because of the multi-purpose character of acoustic survey programmes it is important from cost and efficiency points of view to integrate the survey system of the various components (= surveys) of the programme.f) Rotation system and reference period: A decision has to be made concerning how often the surveys should be conducted, and the reference period of the survey characteristics, i.e., the period for which information is collected from the sample units (day observations, night observations, etc.).
g) The source documents to be used: Having decided on the kind of information to be collected the schedules to be used for sonar and echo-integrator surveys and other concurrent sample surveys form an important part of the survey programme.
h) Editing: Procedures will have to be devised for the proper decomposition of the obtained total readings (echo-integrator surveys) into fish-readings and other (= noise, and signals from plankton, jelly fish and other scatterers). Also, provision must be made for a thorough and accurate identification of the echo traces in the obtainable echograms and sonargrams of the surveys.
i) Biomass species composition: Procedures will have to be devised for the separation of the total biomass estimates into different species (varieties of fish).
j) Processing, presentation, analysis and critical analysis of the findings of acoustic surveys are very important steps which will be carefully incorporated in the survey system of an acoustic survey programme.
k) Collection of concurrent oceanographic and biological data on an integrated basis and the processing, presentation, analysis and critical anlysis of the findings.
In this section we consider certain problems in the light of obtaining the required information.
1. Packing status of the target population
Temporal changes in the packing status characterize fish populations of small pelagic fish. From a statistical point of view the following two packing patterns can be distinguished
i) Dispersed fish: fish spread widely through a given volume of water. This is the case in the night-time packing pattern of small pelagic fish.Further, by taking as a criterion the level of packing density within the above established patterns, the following packing limits can be introduced:ii) Schooling fish: fish are grouped into schools of various sizes that are individualized. This is the case of the daytime packing pattern of small pelagic schools.
|
|
Packing limits |
|
|
Packing patterns |
Upper limit |
Lower limit |
|
1. Dispersed fish |
Dense layers |
Very scattered fish |
|
2. Schooling fish |
Dense schools |
Light dense schools |
a) Individualized schools2. Collection of required information
b) Clustering schools, forming layers of schools.
The methods of collecting the information are to a large extent conditioned by the material under survey and the type of information required. For survey operations the research vessel (R/V) is equipped with a complete acoustic system which includes an echo-sounder with T.V.G. and an echo-integrator. A sonar system is also included for the detection of surface and subsurface schools. The description and principles of operating the acoustic instruments are discussed in chapter 1 of this part of the report.
The method of line transect sampling1 is used for the collection of the required items of information. Estimates of the survey magnitudes are made by using the sample observations -
1 See chapter 5.In echo-integrator surveys (= echo-sounder surveys), quantitative items of information are collected, here called counts (= statistical meaning), along the sample tracks. Specifically, sound is transmitted vertically at regular intervals and a certain proportion of it is reflected from the various targets encountered within the insonified volume of water. The echoes received by the echo-sounder are accumulated by the echo-integrator and an aggregated total (mm) is produced by the echo-integrator at fixed successive periods of time or distance sailed, here called elementary sampling distance unit (ESDU).
Information on the vertical location of the detected fish is also recorded on the echo-sounder recording paper. Specifically, on an ESDU basis items of information are obtained on the location (upper, lower limit) and gravity centre of the detected fish concentrations.
As we have discussed in chapter 1, the sonar system works on the same acoustic principles as the echo-sounder, its main property is that its bearing and tilting can be changed accordingly. For fisheries surveys, the sonar is used horizontally and directed perpendicularly to the path of the vessel. Within the insonified volume of water (= sonar range) the sonar provides information on the number of detected schools, their horizontal size (cross section) and horizontal distribution of the schools (distance distribution).
At first sight, the raw material obtained by the sonar system can be considered as biased in its nature for the following reasons (see also chapter 4):
i) The insonified volume of water is not uniform within the sonar range, because of conical shape of the sonar beam, the volume of water sampled is bigger at long range than at short range.For estimation purposes, sample data which correspond to the “effective sonar range”2 are taken into account. This is the sonar range in which both the number and size (area) of schools traced are free from the effect of the above-mentioned biases.ii) The detection threshold of the sonar system varies with the density and size of schools: big schools can be detected at a longer range than small schools, the same goes as far as dense schools and light dense schools are concerned.
iii) At close range from the research vessel, the “avoiding effect” problem occurs. Also some of the schools split in their attempt to avoid the approaching vessel.
2 See sub-section 4.2Concurrent fishing operations are also conducted along the sample tracks aiming to provide information on the species composition (varieties of fish) of the survey biomass. In addition, useful biological information is collected on length, age, sex, sexual maturity, etc. of the survey stocks.
Concurrent items of information are also collected which are related to the environmental conditions, such as temperature, salinity, presence and depth of the thermocline, oxygen content; productivity data can also be collected and compared with the size of the estimated biomass and spatial distribution of fish.
It should be noted that, temperature and thermocline are two very important characteristics to be measured because of their effect on the behaviour of the sonar beam and the determination of its effective range.
3. Editing of raw sample data
It has been discussed that in echo-integrator surveys the total integrator readings (mm) per ESDU are proportional to the aggregated echoes returned from all targets sampled within the insonified volume of water. This is a compound value which consists of the following component values (mm)
- all species of fish coveredFor estimation purposes, the problem which arises is the decompositon of the received total integrator reading into the following three component groups
- plankton and larvae
- thermocline
- occasional bottom integration
- vessel noise
- sea noise (bubble echoes)
- all species of fish covered (mm)There are no pure quantitative methods for the proper separation of the total integrator readings into its main components. In practice, the contribution of unwanted targets to the total integrator readings are rejected by using subjective methods. For this purpose all the available information should be utilized, plus the proper selection of the settings of the acoustic equipment:
- plankton and larvae (mm)
- other (mm)
- a proper scrutiny of the echo-sounder paper (echograms) along with the integrator readings is required in order to determine the component value of the total integrator readings which corresponds to fish only.The separation of the received integrator readings into species (varieties of fish) can be made by observing the shape of the echo traces and their vertical location on the echo-sounder paper. Depth data help for identifying species with known bathymetric distribution. Also, catch data of concurrent fishing surveys are used for the same purpose.- vessel noise and sea noise can be eliminated up to a reasonable level by utilizing the threshold system in the integrator.
- echoes from thermocline, plankton and bubbles in bad weather conditions can be eliminated to some extent by comparing the observed integrator readings among the successive ESDU’s.
The vertical sampling gate of the echo-integrator is triggered by the echoes of the bottom which are stronger compared to fish echoes, and therefore bottom echoes are not integrated. However, in some cases, dense schools, e.g., daytime, may produce echoes as strong as those of the bottom and because of that might be excluded of integration. It is rather difficult to optimize the settings so that only school echoes should be taken into account and not bottom echoes. Some electronic systems have been developed for solving the problem. When such systems are not available judgement should be used by assessing echo values obtained from similar schools in the neighbourhood.
Experience in assessing echo traces and also experience in the behaviour of the survey stocks is very important for the proper editing of the sample raw material of large-scale echo-integrator surveys.
The proper editing of the sonar recordings requires a good deal of experience and attention in order to accurately assess the echo traces which correspond to schools of fish and those which correspond to the bottom. This task is extremely important when sonar surveys are conducted in shallow waters.
In this section we discuss in summary behaviour features of the survey stocks which affect biomass estimates.
1. Day and night behaviour
A critical assessment of the results of echo-integrator surveys covering the same target population (small pelagic fish) shows differences between daytime and night-time estimates. These differences can be attributed to a combination of factors related to behaviour of fish (see also chapter 3):
- tilting and changes in the physiology of fish affecting target strength.2. Effect of moon phases- the packing status of fish varies between daytime and night-time and this might affect the accuracy of the obtained measurements.
- during daytime fish schools are more active than the dispersal of fish during night-time and because of that the probability of encountering fish varies between day and night-time.
During the full moon period the fish tend to behave in the same way as they do under daytime conditions and this affects the accuracy of the obtained measurements (see above 1).
3. Avoidance problem
During the daytime the schooling small pelagic fish tend to escape from the path of the vessel. The “avoidance effect” will be more or less marked for different species of fish. In echo-integrator surveys the avoidance effect will be serious in the upper layers of water, close to the research vessel, producing biased counts. In sonar surveys, schools at a close distance from the vessel may also split when escaping and thus give a biased picture of their actual size. It should be noted that the avoidance problem is not so serious if it is known, from the interpretation of acoustic measurements always presuming a knowledge of behaviour of the survey stocks.
4. Vertical migration
Vertical migration of fish is a phenomenon which is related to seasons, spawning, feeding and light intensity (day and night). Vertical migration of fish affects the quality of the obtained measurements in a number of different ways (see also chapter 3).
- when fish move close to the upper layers of water they are subject to avoidance effect (see above 3).5. Geographical migration- when fish move very close to the bottom they are not properly detected by the acoustic equipment and thus give rise to biased results.
Part or all of the target fish population can move out of the survey area, for example, during spawning. Geographical migration also includes the movement of fish from the offshore to the inshore waters and the return migration. These movements are reflected in the temporal estimates of the survey biomass.
If the survey stock has not previously been studied using acoustic techniques, then the study must begin with the collation of all the existing sources of information concerning the fish stock. This process of collation should culminate in the compilation of seasonal charts which summarize the distribution and behaviour of the species.
