Concerted Action "Improvements of Tagging Methods for Stock Assessment and Research in Fisheries" (CATAG)
FAIR. CT.96.1394, FINAL REPORT May 1999 (DRAFT)
5. ELECTRONIC TAGS
In recent years, the most significant advances in tagging and tagging methodologies have come about with the development of electronic tags. The range and versatility of the these tags has meant that new applications are constantly being discovered and the full benefits to be derived are barely being exploited at present. It is because of this that a significant amount of attention should be drawn to this area in future fisheries related studies. As electronic tagging techniques are constantly being developed and improved, consideration is given in the following section to current methodologies and progress with capture, handling and recovery of tagged fish. Particular attention has been paid to attachment methods, especially with regard to modifications to the normal behaviour of the fish which may make interpretation of data difficult to apply to whole populations. Apart from providing information on fish location and position in the water column at fixed times, it is now possible using electronic tags to provide details of fishes’ immediate environment in real time and over long time periods, thus allowing study of the factors which most influence their subsequent behaviour. Some of the recently developed electronic tags depend on being recovered subsequently, and for this reason information is presented on the approach to be taken when organising studies which require intensive and systematic tag recovery programmes.
To illustrate the enormous potential for fishery based applications that could be developed, specific examples of electronic tag application are given which have already provided significant input into fisheries related surveys and fish passage investigations. Finally, a comprehensive examination is presented outlining the future developments needed to sustain necessary technological developments and to consolidate recent advances.
5.2 TAG AND SENSOR CLASSIFICATION
In recent years there has been a proliferation of electronic tag types and systems for tracking fish at sea and in freshwater. Many of these tags have been designed with specific applications in mind. The more generally applicable tags are described here together with their operational details. The list is not meant to be exhaustive but to give a broad overview of the most commonly used tags. Specific details on individual makes and manufacturers can be obtained from the CATAG website which has links to most of the main manufacturers of electronic tags world-wide. A summary classification of the tags types discussed below is given in Figure 5.2.1. Details on dimensions and characteristics are given in Table 5.2.1.
5.2.1 Transponder Tags
18.104.22.168 Passive Integrated Transponder (PIT) Tags
The PIT tag consists of a small glass-encapsulated electromagnetic coil and microchip that is inserted into the body cavity of a fish using a veterinary syringe. The tag is inert until it is activated inductively by a tag reader, which provides the power for the tag to transmit a unique alpha-numeric code. The system offers 34 billion codes and operates at 125 or 400 kHz but the detection range is very small. Tags can be decoded with a portable hand-held reader which has a range of 10-15 cm. Automatic readers are also available with either a tunnel detector (up to 30 cm diameter) or a strip detector, which can be placed on the bed of a stream (up to 20 cm depth). PIT tags are made in three sizes ranging from 11 to 28 mm in length and 2.1 to 3.5 mm in diameter.
PIT tags may last throughout the life cycle of their "hosts" and the tagging system allows rapid retrieval of transmitted information from large numbers of tagged fish. They can be detected and decoded in living fish in fresh and salt water, and they eliminate the need to anaesthetise, handle, restrain or kill the fish during data retrieval. Used with computer stations, they allow repeated identification and measurement of individuals within a population. As each PIT tag can carry a short unique code, it provides a good basis for many types of survey where the fish are able to come in very close contact to detecting equipment.
22.214.171.124 Sonar Transponding Tags
A transponding tag allows the position of a free-ranging fish to be fixed accurately relative to a research vessel. Transponding tags differ from other electronic tags in that they only transmit an acoustic signal when they receive an interrogation pulse from a sonar. Ultrasonic frequencies are produced by stimulating an annular ceramic transducer at its resonant frequency. Tag size is governed by the size of the transducer, whose diameter is inversely proportional to frequency. Range is also inversely proportional to frequency, so that, while a large diameter 30 kHz tag may have a range in excess of 1 km, a small 300 kHz tag usually has a range of less than 400 m. Frequencies of 34 to 50 kHz are commonly used for tracking large pelagic fish, while 60-80 kHz is commonest in coastal and estuarine waters. The higher frequencies (150 to 300 kHz) are used in freshwater, for studies where a small tag is required, or where specialised high-frequency imaging sonars are available.
5.2.2 Transmitter Tags
This is a large family of tags which is increasing in size due to new developments for specific applications. These tags are larger than PIT tags and require an internal battery to power the transmitter and microchip (if present). The lifetime of the tag is a critical consideration in telemetry studies and depends on the trade-off between transmitter size, power supply, range and rate of the signals. Telemetry studies on free swimming fish are generally short term studies ranging over periods of hours to months. Apart from pulsed and coded signals which identify the individual fish, some tags can carry sensors which can transmit data to the receiving station. Sensors can record depth, swimming direction and speed, or even heart rate. The behavioural and physiological data sampled via transmitting tags can be used to study the activities of fish in relation to their immediate environment and also in relation to anthropogenic factors (fishing gears, dams, oil rigs, effluents etc). Microchip technology allows for specific instructions to be placed on some types of tag which allow the tag to be switched on or off under specified conditions (e.g. entry into freshwater). These features can be used to increase the longevity of the tags or to transmit under certain environmental conditions. Accurate geolocation is possible by a variety of methods. The detection range for some of these tags is up to a kilometre in some instances but is generally less than 100m. Attachment of the tag can be internal or external (see Section 5.4).
126.96.36.199 Pulsed tags
Radio and acoustic transmitting tags can transmit a simple pulsed signal at selected pulse rate. Theoretically, large numbers of fish can be monitored simultaneously, using multiple frequencies or pulse rates. In practice, however, it is very difficult to distinguish more than four or five pulse rates on an individual frequency.
Radio tags, which can only be used in water of very low salinity, are useful in freshwater because radio waves are less affected by physical obstacles, turbidity, turbulence and thermal stratification than acoustic (non-electromagnetic) waves. Radio signals also radiate through the water surface and can be detected at great distances because there is little loss of signal strength in air. Receivers can be fitted in boats, aircraft or land-based listening stations. Radio tags operate at high frequencies (20-250 MHz), so there is little signal drift.
Acoustic tags are used in the sea because sound is transmitted over long distances in salt water, whereas radio waves are attenuated very rapidly. Frequencies of 30-300 kHz are used. Pulsed acoustic tags have been used to follow a number of species in the open sea, often using a simple receiving system comprising a hand-held directional hydrophone, a portable receiver and headphones. This method provides only a rough indication of the position of the fish relative to the tracking boat and accurate position fixing with a pulsed tag requires triangulation using an array of fixed hydrophones.
a) Non-Programmable Pulsed Radio Tags
Non-programmable transmitter tags are used to transmit a simple radio pulse at set intervals. They require radio receivers operating at 30-50 MHz frequency range to detect the signals.
b) Programmable Pulsed Radio Tags
Programmable micro-processor tags are used to transmit simple radio pulsed signals at user defined intervals. Specific on/off sequences can be set which can be useful for preserving the battery life of the transmitter. New developments include the ability to include sensors which can telemeter information relating to the behaviour or physiology of the fish including electromyograph (EMG) and tail beat to the receiving station.
c)Non-Programmable Pulsed Acoustic Tags
These tags telemeter a simple acoustic pulsed signal to an acoustic receiver. These tags are generally used in saline or semi-saline conditions where radio signals cannot be transmitted. There is a limitation on the number of individual fish which can be identified as each receiver can only differentiate a single frequency.
d) Combined Acoustic/Radio Tags
CART tags are hybrid tags combining components of both radio and acoustic tags. This allows tracking of individual animals to be carried out between salt and freshwater. A conductivity sensor is incorporated to detect the salinity of the water body around the fish and a microprocessor can automatically switch between acoustic transmission and radio transmission as appropriate.
188.8.131.52 Coded Tags
Coded tags operate by emitting a digitally encoded signal on specific radio and acoustic frequencies. Each signal can, theoretically, be unique. This offers the advantage that many individual fish can be tracked separately on a single frequency and that the information can be automatically recorded and downloaded to a PC. Coding has great potential for increasing data acquisition rates and increasing sample sizes in telemetry experiments. Digitally coded tags are also available, which allow as many as 170 tags to operate at one frequency.
a) Coded Radio Tags
These tags contain a programmable micro-processor and transmit a digitally encoded radio pulse at user defined intervals
b) Coded Acoustic Tags
These tags are similar to coded radio tags but transmit a digitally encoded acoustic pulse at user defined intervals. They are used for monitoring fish passage in marine and freshwater.
5.2.3 Data Storage Tags
Also known as archival tags, these tags have ranged from simple data loggers, capable merely of recording depth or temperature of an individual fish, to sophisticated programmable devices capable of providing a direct estimate of geographical position at regular intervals over periods of many months. Developmental work over the last five years has led to the production of a number of tags that are beginning to be used very successfully with free-ranging fish in the open sea. The species have included tuna, salmon, cod and plaice. The most exciting and rapid advances in both technology and biology in recent years have been associated with data storage tags which can be programmed to record details of temperature, depth, salinity, pressure, light, chemical and physiological indicators at set intervals. Other sensors tags in development include tilt, heading and position fixing. Some of these tags can record data for up to five years and store this information for up to twenty years. However, in order to retrieve the information the tags must be recovered from the fish. Normally this involves establishing an intensive recapture operation or relying on commercial or recreational catch returns. Therefore, an external mark or tag is usually applied to the test animal to facilitate identification of fish carrying DST tags. Incentives (money, prizes) are often offered to improve the frequency of tag returns (see Section 5.6). Due to the high cost of production only relatively small numbers of animals are usually tagged. However, the cost is offset by the enormous amount of data which can potentially be generated from single tags on recovery.
Geolocation of fish may also be achieved by underwater light intensity measurements which can be used to estimate times of sunrise and sunset. The data are used, in turn, to calculate latitude and longitude using interactive software. An independent check on latitude can be obtained from the temperature measurements made by the tag.
DST tags have been produced in many shapes and have been developed for both round and flat fish applications.
5.2.4 Satellite Tags/Popup Tags/Chat Tags
A relatively new application has developed for tags that telemeter stored data to remote receiving stations, rather than relying on tag recovery to retrieve data from data storage tags.
Recent developments in oceanic tracking have been possible with the development of the ARGOS data collection and location system (CLS), a joint venture between France (Centre National d’Etudes Spatiales - CNES) and the USA (National Aeronautics & Space Administration - NASA, and National Oceanic & Atmospheric Administration - NOAA). This provides complete world coverage with receivers on board NOAA satellites orbiting the earth in near-polar orbit at a height of 850 km (Taillade, 1992). The system uses UHF radio frequencies and its Doppler location system depends on a very stable transmitter frequency (401.650 MHz). The location of the platform transmitter terminal (PTT) carried by an animal is calculated from the shift in frequency of the transmitted radio signal as the satellite approaches and then moves away from the PTT (Harris et al., 1990, Taillade, 1992). Accuracy of location improves with the number of successful ‘uplinks’ during each satellite overpass and Service Argos classifies the quality of location (class 1, 2 or 3) achieved with each fix.
The size of PTTs has precluded the use of the Argos system with all but the largest fish, and this difficulty has been compounded by the severe attenuation of UHF radio signals in salt water. Applications have been confined to large sharks, which surface sufficiently often to be detected by a passing satellite.
Very recently results have been obtained in the North Atlantic from bluefin tuna fitted with the first generation of ‘pop-up’ tags that detach from the fish at a predetermined time and float to the surface from where they transmit to the Argos satellite (see Section 5.4 and 5.7).
Combinations of transponding and data storage tags are being developed to increase the versatility of the applications. Communicating history acoustic transponding (CHAT) tags allow researchers to locate and track animals using transponding tags and retrieve data without recapturing the animal. The information is telemetered from the tag to a tracking receiver or a fixed monitoring station which can search for and locate any tag within range of the receiver, then store real time information to disk with GPS position. The receiving station can also send commands to the tag remotely to reset data recording intervals.
5.3. CAPTURE, HANDLING AND FISH RECOVERY
In order to get experimental fish in good physical condition for laboratory or in situ experiments, the methods of capture and the handling of the fish prior to and during tagging is of particular importance. Different capture and handling methods inflict different damages and stressors on the fish and different species have different tolerance for capture and handling. In addition, the vulnerability to handling may vary during different life stages. For example, salmonids, especially Atlantic salmon, vary greatly in their resistance to handling during their life span. While the fresh water related stages (parr, maturing and kelt stages) are less vulnerable, the skin of smolts, post-smolts and immature fish is very sensitive to handling. Atlantic halibut (Hippoglossus hippoglossus) are known to be difficult to handle without causing lethal damage (Midling, pers. com.), and several pelagic species such as herring (Clupeids) and mackerel (Scombrids) are similarly easily damaged (e.g. Wardle, 1968 and Blaxter & Holliday, 1963).
This review of fish capture and handling methods in relation to electronic tagging is based on a selection of experiments mainly restricted to marine fish with emphasis on Atlantic species. Fresh water and non-Atlantic marine species are included where either the capture method or the physiological observations made are of particular interest.
5.3.2 Damage during capture and handling
Most capture methods result in abrasion of the skin. The mucous layer protecting the epidermis and the scales is particularly delicate in most fish species. This layer protects the fish against fungal, bacterial and viral invasions and, together with skin and scales provides a barrier against leakage or dilution of body fluids. An undamaged mucous layer is essential for the well being of the fish after capture. Damage such as scale loss and skin wounds will cause problems of increasing seriousness depending on the degree of body cover lost. Prolonged struggles or swimming activity during capture leads to exhaustion with subsequent conversion of muscle glycogen to lactate acid. In the case of severe exhaustion lactates are released into the blood stream from the muscles and cause lethal metabolic acidosis. The post-capture metabolism of accumulated lactates in the muscles will also lead to an elevated oxygen demand, which must be considered during subsequent transport and handling of the fish. Stress may also result in a reduction of immune responses.
A particular problem to overcome when working with physoclist (closed swimbladder) fish, like gadoids, is the expansion or reduction of the gas contained in the swimbladder if the external pressure changes. Physoclists can overcome this by absorbing or secreting the gas in order to keep neutrally buoyant. The compensatory mechanisms are rather slow, and depend on temperature and pressure (Harden Jones & Scholes 1985). As a result, even relatively small involuntary upward movements transports cause substantial expansion, leading to rupture of the bladder and compression of internal organs (see section 5.5 and 7.4.6).
Solomon & Hawkins (1981) and Wardle (1981) give an overview of the damage that may be inflicted on the fish during the capture and handling processes. The physiological and bacteriological processes set off by capture and handling will act in a similar manner after release back to nature. Experiences with methods used to capture fish for aquaria or aquaculture are therefore very relevant also in this context, even though they may not yet have been applied to electronic tagging.
5.3.3. Capture methods
The choice of a particular fishing method will depend on the species sought, fish density, location and possible legal restrictions. Solomon & Hawkins (1981) discuss some general advantages and drawbacks of various capture methods for obtaining good quality fish for aquarium use. Bottom dragnets (trawls, seines etc.) and midwater trawls, in which the fish is forced to swim with the gear during capture, may lead to exhaustion if towed too fast, or for too long. The risk of skin damage and scale loss is always present, although this effect can be alleviated to a certain extent by using a lined codend. Despite these disadvantages, towed nets often are the only possible practical method. Gillnets, either stationary or drifting, enmesh the fish or entangle them and cause damage where the fish have been held, often by the head, the gills or the trunk. If enmeshed at the gill region, the fish may either suffocate or bleed to death. Encircling nets like purse seines have several advantages over gillnets, but if the catch is large, crowding during the final pursing may cause oxygen depletion and abrasions when the fish hit each other. Baited or unbaited trapping gear may be very effective. The fish enter voluntarily and are seldom damaged or severely stressed. However, the fish may be damaged if other fish enter later and when they are taken out of the trap. Solomon & Hawkins (1981) recommend that particularly delicate species are removed under water into a holding tank, or that the lower part of the trap is lined. In fresh water electrofishing may prove effective and inoffensive because the fish recover quickly. However, Solomon & Hawkins warn against the possibilities of spinal fracture and haemorrhage that can be caused if the voltage is not properly adjusted. Such effects have also been observed by other authors (see section 5.3.3.b.). Angling and handline fishing are singled out as methods with many advantages over other methods. Damage is often slight and confined to the jaws and can be further minimised by using barbless hooks. Struggling time can be reduced by using heavy fishing tackle. The disadvantage of angling is the low number of fish that can be caught. Longlines, set lines and drifting lines catch many more fish but have the disadvantage that the hook may be swallowed with the bait by many species.
In many experiments the descriptions of how the fish were captured and handled prior to electronic tagging are rather non-specific, often only stating which gear was used for capture. The time elapsed from start of capture to landing on deck, the method of handling the fish on board, or in the hatchery, and recovery times before tagging are often omitted. The most widely applied method for obtaining experimental fish from natural environment, however, seems to be to catch large numbers of fish, place them in a tank and, after an observation period of relatively short duration, choose perfect looking specimens for tagging.
(a) Demersal fish and shellfish
Capture methods reported in electronic tagging experiments with demersal (bottom dwelling) fish have included hook-and-line fishing for lingcod, Ophiodon elongatus (Matthews, 1992), handline fishing for cod, Gadus morhua L. (Arnold et al., 1990; 1994), trawl fishing for plaice, Pleuronectes platessa (Greer Walker et al., 1978; Harden Jones et al., 1981,1982; Metcalfe et al. 1993; Metcalfe & Arnold 1997) and cod (Engås et al., 1991, Godø & Michalsen, 1997), Danish seining for cod (Thorsteinson, 1995), and seine netting for cod, and plaice (Isaksen & Midling, 1999). Due to their robustness and economic importance cod have been the targets for many tagging studies.
Decompression of swimbladder gases of physoclists has commonly been dealt with by catching the fish at depths less than 10 m, or catching the fish with gear (e.g. pots, other cage-type gear, or hook and line) that enables the catch to be lifted slowly up to the surface. Tytler & Blaxter (1973) suggest a 5 hour decompression halt for gadoids for every 50 % reduction in external pressure. Engås, et al. (1991) captured cod by jigging in shallow water and let them recover in net pens for 3-8 days before tagging with hydroacoustic tags. Arnold et al., (1992, 1994) caught cod by rod-and-line or long-line in shallow water (< 8 m) being careful to bring the fish slowly towards the surface, the maximum pressure reduction always lying well below the 50% recommended by Tytler and Blaxter 1973. Fish were kept and fed in a large laboratory tank for several months until taken on board the tracking vessel, transported to the release site, tagged and released from cages.
Another method commonly used to reduce mortality caused by over-inflation of the swimbladder is to release the internal gas by puncturing the body wall and bladder with a hypodermic needle once the fish is on deck (Midling, pers.com.; Olsen pers. com). Positive effects of decompression and swimbladder puncture are reported by Keniry et al. (1996), who conducted experiments on yellow perch, Perca flavescens, collected at 10 and 15 m depths in Lake Michigan. Decompressed fish had higher survival than non-decompressed fish and, as would be expected, this effect was greater for fish caught at 10 than 15 m. Puncturing the swimbladder had a significant, positive effect on 3 day survival; long-term survival was not affected.
There are no restraints on the speed at which demersal fishes without a swimbladder can be brought to the surface and flatfish like plaice and sole are relatively robust with respect to handling in general. Plaice have been electronically tagged by the Lowestoft Laboratory since the early 1970s and until recently the technique has been to select undamaged fish from trawl catches and return them to laboratory tanks until viability was confirmed (Greer Walker et al., 1978; Harden Jones et al., 1981,1982; Metcalfe et al., 1993; Metcalfe & Arnold, 1997). Recently fish have been tagged at sea with data storage tags immediately after capture to avoid disrupting natural patterns of movement and avoid problems with disease in the laboratory.
Other non- physoclist fish such as sea wolves (Anarchicas lupus variants), anglerfish (Lophius piscatorum) and halibut (Hippoglossus hippoglossus) are known to be difficult to handle without causing skin abrasions (Midling, pers. com.). At the Dept. of Fisheries and Aquaculture, Fisheries Research Centre (FRC), Tromsø, northern Norway, where all these species have been captured and kept in conjunction with various fish holding experiments, the impression is that the grey wolf fish is best captured by Danish seine, while the spotted wolf fish is most easily taken in a trawl. If caught on hooks Anarchichas risk excessive bleeding from the large arteries in the head and mouth region and need several weeks for adaptation (Midling, pers com.). Anglerfish are also difficult to handle without damaging the skin, although some individuals caught in a Danish seine survived in captivity for several weeks (Midling, pers com.). Lumpsuckers (Cyclopterus lumpus) are easily damaged in both the coastal and the pelagic phase, and require special observation when captured. Nets and trawl both cause lethal skin damage. Plaice and lemon sole on the other hand cause few problems and have been captured with seine net and transported with little mortality in special holding tanks (Midling pers. com.).
In order to avoid adverse effects of capture and to secure observation of entirely natural food search and reactions to olfactory stimulants, Løkkeborg & Fernö (1998) and Løkkeborg (1998) set up experiments where cod were allowed to voluntarily swallow tags wrapped in various types of bait. This technique has also been applied with success to several deep-sea species, which are often stenothermal and stenohaline (Solomon & Hawkins, 1981), and could not otherwise be tagged because of the slow decompression rate and the time needed to get them to the surface. Grenadiers (Coryphaenoides yaquinae, C. armatus), deep sea eels (Synaphobranchus bathybius) and the deep sea gadoid Antimora rostrata have all been successfully tagged with acoustic tags after ingestion of bait hung beneath the cameras of a deep-sea lander (Armstrong et al., 1991, 1992; Bagley et al., 1994; Priede et al., 1990, 1991, 1994).
Shellfish have mostly been obtained by trapping the animals in pots or cages, or in some cases with tangle nets (González-Gurriarán & Freire 1994) or at shallower depths by divers. The risk of damage is small if the gear is carefully hauled; decompression is not a problem for shellfish. The attachment of tags is fast and the animals can rapidly be returned to their normal environment, if not tagged in situ by divers. Details of capturing and electronic tagging of Norway lobster, Nephrops norvegicus L., are given by Chapman et al. (1975). González-Gurriarán & Freire 1994 give similar details for the spider crab (Maja squinado). Collins & Jensen (1992) and van der Meeren (1997) have tagged and tracked European lobster (Homarus gammarus).
(c) Pelagic fish
Most pelagic species are susceptible to handling. As far as is known, the smaller schooling species (e.g. Atlanto-Scandinavian herring (Clupea harengus L.) and Atlantic mackerel (Scomber scombrus L.)) have not been used in electronic tagging studies so far. But as tags continually get smaller and sensors more varied the possibility of tagging these species increases and methods of handling will be of interest. Since 1968, the Norwegian Institute of Marine Research has been quite successful in tagging large numbers of mackerel and herring using conventional (internal) steel tags. Mackerel are caught by jigging, carefully unhooked and placed in tanks for observation prior to tagging. Bleeding or wounded fish are discarded (Myklevoll, 1994). Public aquaria, such as the North Sea Centre (NSC) in Hirtshals in Denmark regularly obtain these species for display in tanks. The NSC relies on professional fishermen, who use very fine meshed purse seines to catch schools of herrings and mackerel close to the coastline. Fish are transferred to holding tanks, and viable looking specimens chosen for transport to the aquarium. As mackerel are extremely sensitive to touch, great care has to be taken to avoid skin damage and this is achieved by only handling the fish when they are immersed in water. Herring are caught by similar methods as the mackerel (Flintegård, pers. com).
Special methods have been developed for capturing and tagging large pelagic species such as sharks, tunas, marlins and sailfish, which are difficult to handle and sedate on board a boat because of their size and strength. Pole and line fishing from vessels using lures with special barbeless hooks is the main method of capture. The fish are handled rapidly without anaesthesia and care is taken not to cause skin damage by using soft plastic covered tagging/ measuring cradles (Williams, 1992). Carey & Robinson (1981) and Carey & Scharold (1990) carried out pioneering work to develop methods for handling and tagging swordfish (Xiphias gladius) and blue sharks, (Prionace glauca). Holland et al. (1990a, 1990b), tracked yellow and bigeye tunas and blue marlins (Makaira nigrans) caught by trolling and pole-and line fishing. The chosen tag attachment method enabled release of fish after approximately one minute out of water. Block et al. (1992) caught blue marlins for tracking by trolling artificial lures with rod and reel from boats. Block et al. (1998) have developed a successful method of capturing and handling Atlantic bluefin tuna (Thunnus thynnus) for use in archival tagging and acoustic tracking studies. The fish are caught by heavy tackle using circle hooks and bait presented in a chum stick ("chunk fishing"), a technique which allows chasing down the fish in order to keep fight times less than 15 minutes. The fish are taken on board a boat with specially designed leaders through a "tuna door" in the stern and tagged and released immediately. The method is suitable also for handling large individuals (> 50 kg) with low risk of damaging the fish. A similar approach has been used with southern bluefin tuna (Gunn et al. (1994).
Sharks lack a swimbladder and must swim to maintain position in the water column. Muscular movement assists in venous return of the blood and oxygenation at the tissue level is maintained in many by swimming at some optimum speed. Care is therefore needed when capturing and handling sharks to minimise the time for which the shark is restrained (Gruber & Keyes, 1981) and the amount of struggling. Prolonged struggles affect blood serum protein deleteriously and accumulate lactates. Capture by trapping or trawling should therefore be avoided and Gruber & Keyes recommend the use of a handline. This method reduces the risk of injuring the mucous layer, skin and eyes and keeps the time for capture short. Reference to capture of shark species for electronic tagging by handline or rod and line is made by Nelson (1978), Carey et al. (1981), and Stevens (1996).
Four main lifestages of Atlantic salmon are recognised with different spatial distributions and different vulnerability to capture and handling. These stages must be considered separately in relation to the use of electronic tags (Anon, 1997).
Smolts. These fish are in transition from the fresh water phase to salt water tolerance and have started their down-river migration towards the sea. Wild fish (10 - 17 cm fork length depending on river environment and genetic origin) are generally too small to be tagged with electronic tags at present. Many experimenters have used hatchery fish instead, although smolts from wild stocks have been tagged where they are large enough to tolerate the application of the smallest available electronic tags. Tytler et al. (1978) used wild smolts caught in a trap in the river N. Esk. Holm et al. (1982) obtained a few wild fish from a fish trap in the river Imsa in south-western Norway. After capture, the smolts were stored for 2 - 14 days in a hatchery trough before tagging; they were released within 24 h of tagging. Moore & Potter (1992) and Moore et al. (1990b, 1990c, 1994, 1995) used wild fish, which they caught in streams in Wales and southern England using fyke-nets and a keep box for the trapped fish. The fish were anaesthetised, tagged and put in oxygenated water for a recovery period of 30 - 60 min. before release into the river. Other techniques used for capturing wild fish for electronic tagging in rivers include electrofishing and beach seining (Knutsson, unpublished). However fish traps and trapnets have advantages over electrofishing as trapping will capture only actively migrating smolts, while electrofishing takes all fish including those not yet in the active migratory phase (Anon, 1997).
Post-smolts are salmon in their first year after leaving fresh water. Depending on genetic origin and the time of capture after entering the marine environment, Atlantic salmon in this stage range from approx. 15 - 35 cm in length. Until recent years few captures of post-smolts had been made but they are now regularly caught in surface trawls (Holst et al., 1993; Hvidsten et al., 1995). Trawl caught post-smolts lose 50 - 100% of their scales even in short tows (Hvidsten, pers. com.; Holm et al., 1998; Holst et al., in prep.). Other reported capture methods include floating long-lines and drifting gillnets (Reddin & Short, 1988, Sturlaugsson & Thorisson, 1995), although none of these methods of capturing post-smolts has yet produced fish in a fit condition for tagging. Instead, most tracking studies have been performed with hatchery fish, or with wild fish trapped as smolts in freshwater (Moore et al., 1995, 1999; Lacroix, 1996, pers. com.) and then released in rivers, estuaries, or fjords. Recently a device for obtaining post-smolts in viable condition from trawl catches has been developed and tested with promising results (0 -6 % scale loss) (Holst & MacDonald, in prep.). It is hoped this device will allow post-smolts to be caught in the open sea in a fit state to be tagged with electronic tags.
Adult stage - immature fish. Immature salmon (both one- and multi-sea-winter fish) are found in feeding areas in the open ocean. Handling must be done with great care as the risk of scale loss is substantial (Hansen & Jacobsen 1997). Adult immature salmon are occasionally caught by surface-trawling in the Norwegian Sea (Holm et al., 1998), but this method is unsuitable for obtaining fish for electronic tagging because of the large loss of scales that occurs. Drifting gill nets have been used to catch salmon for tagging in the Pacific. Fish caught by long-line have been used for electronic tagging studies in the north Atlantic (Jakupstovu, 1988) and experiences from a Carlin tagging programme in the Faeroes give valuable indications of how to handle the fish. The lines were patrolled constantly to remove hooked salmon. The fish were carefully lifted over the ship side with a scoop-net and placed in a recovery tank where undamaged, viable looking fish were chosen for immediate tagging and release. It is sometimes more deleterious to remove a long-line than it is to leave it in the fish. Hansen & Jacobsen (1997) stress the importance of hook shape for ease of removal and recommend using non-galvanised material in case the hooks have to be left in place.
Adult fish- maturing salmonids and kelts. Maturing fish homing to their natal streams have been captured in coastal and estuarine waters using gear such as bag-nets, trap-nets, other fixed engines or beach seines. These methods are relatively harmless and, in addition, the salmon are much more resistant to handling at this stage in their life history as a result of physiological changes to skin and mucus, which occur in conjunction with maturation. Several authors (Westerberg, 1982, 1984; Potter & Solomon, 1988; Potter et al., 1992; Heggberget et al., 1993; Smith & Smith, 1995; and Karlsson et al., 1996) have used salmon obtained from trapping gear in the vicinity of the rivers to study various aspects of the homing behaviour of Atlantic salmon. Fish were tagged and released when they had regained their equilibrium after anaesthesia. Brawn (1982) caught Atlantic salmon with a mackerel net and lure in an estuary and kept them in cages for around 1 day prior to anaesthesia and tagging with acoustic tags. After tagging the fish were left in a cage for up to 1 day to recover. Kelts are post-spawning fish that will return to the sea. Like maturing fish, they are relatively resistant to handling because of the condition of their mucus. Where they are installed, fish ladders provide excellent facilities for capturing fish in rivers and good survival of ladder-caught adult Atlantic salmon and rainbow trout is reported by Peake et al. (1997).
When tagging anadromous trout (Salmo trutta L.) and arctic char (Salvelinus alpinus L.) with data storage tags in north-western Iceland, Sturlaugsson & Johansson (1996) and Sturlaugsson et al. (1998) captured the fish by angling from a boat in a lagoon (Sturlaugsson pers. com.). The tags were immediately implanted in the fish under anaesthesia; the char were released after a short recovery in a deck tank.
5.3.4 Handling and recovery
It is well known that anaesthetics cause physiological effects that can be measured as changes in levels of corticosteroid and other parameters (see Chapter 7), which in turn may lead changes in the behaviour of the fish for a varying time after sedation. On the other hand, the handling stress will be reduced under anaesthesia and tagging can be carried out more rapidly with less risk of the fish damaging themselves when trying to get loose. Anaesthesia and anaesthetics are discussed in chapter 7. Legal requirements are dealt with in Chapter 6.
Anaesthetics are easy to apply in the hatchery. Kreiberg & Powell (1991) identified the netting and capture phase of various hatchery operations as the major contributor to overall stress and developed a standard procedure for lightly sedating fish with metomidate before any major handling disturbance. They recommend the procedure for handling of all sensitive fish such as chinook and other salmonids.
In field experiments, the ideal conditions for handling the fish cannot always be met. Setting up facilities for anaesthesia and recovery may be difficult because of spatial restrictions or poor weather at sea. The experimenter must then evaluate the relative difficulties of applying anaesthesia against possible trauma and damage caused by handling unanaesthetised fish, although legal considerations may be paramount.
When electronic tags can be attached rapidly and non-intrusively, anaesthesia has often been replaced by simpler methods of keeping the fish quiet during tagging. Arnold et al. (1992, 1994) blindfolded cod with wet paper over the eyes and Thorsteinsson (1995) used a wet soft cloth with the same species. Blindfolding is also commonly used when tagging adult salmon. These fish are relatively easily calmed if kept in their natural swimming position, for example in a moist handling cradle with the head covered with a wet soft cloth. Handling of unanaesthetised salmonids smaller than 60 cm is, however, not recommended (Sturlaugsson et al., 1995; Hansen & Jacobsen, 1997).
Anaesthesia has in general not been applied when tagging large pelagic species such as tunas and sharks. The capture process is likely to be much more stressful and time consuming than attaching the tag, which generally only requires a minor incision, and the fish are instead quietened by covering the eyes. Special devices to ease the process and minimise handling time have been developed by Block et al. (1992, 1998), Carey & Robinson (1981), Carey & Scharold (1990), Stevens (1996), Holland et al. (1990a; 1990b) and Williams 1992 (see section 5. 3.2.c).
(b) Recovery from capture and handling
McCleave & Stred (1975), Moore et al. (1990) and Lacroix & MacCurdy (1996), among others, have investigated experimentally the effects of tagging and handling on salmonids using dummy tags. In most cases it was shown that the fish recovered quickly from the handling process.
Once the fish has been released it is difficult to assess the impact of the capture, handling and tagging process, although information from data storage tags may provide some useful indications. The various studies performed to estimate survival of fish escaping from fishing gear may, however, aid the assessment of short and long term effects of capture on the survival of electronically tagged fish.
A number of studies have been made on demersal fish escaping from codends of trawls, although estimates of mortality vary according to circumstance. Soldal et al. (1991) found no mortality of cod (Gadus morhua) and less than 10 % mortality of haddock (Melanogrammus aeglefinus) that were kept in cages anchored on the sea bed bottom and observed for 12 to 16 days after escaping from the codend. Jacobsen et al. (1996) observed saithe (Pollachius virens) for 6-7 days by underwater television in cages drifting freely at 40 m depth. Only low mortalities were recorded from these fish, which had escaped from a trawl at 150 m depth. On the other hand, Sangster and Lehmann (1994) recorded 11- 52 % mortalities of haddock and whiting (Merlangius merlangus) escaping from codends when collected and stored in cages on the seabed for 60 days. No mortality was observed in the controls and there were no significant differences between the two species. In trawl simulation studies Soldal et al. (1993) and DeAlteris & Reifsteck (1993) recorded 100% survival of cod after escapement, while haddock suffered 10% mortality. Additional mortality occurred in all groups due to infection of wounds. Jonsson (1994) studied survival and scale damage of long-line caught haddock in aquarium after simulating escape through the meshes of cod-end; the survival rate in these experiments was only 30-50%.
The swimbladder of gadoids is observed to heal relatively rapidly. Experiments made at the University of Tromsø in the early 1980s (Olsen, pers. com) show that healing started 2- 3 days after capture in cod caught in a trawl at 100 m depth. Nevertheless, the use of fish with recently ruptured swimbladders should be avoided (Solomon & Hawkins, 1981), particularly if the aim is to use hydroacoustics to observe natural behaviour in the short term (Mohus & Holand, 1983).
The time the fish have been subjected to a fishing operation will also have consequences for tagging and must therefore be considered. After seven days of post-capture observation in cages Oddsson et al. (1994) recorded significant differences in survival of Pacific halibut (Hippoglossus stenolepis) subject to towing durations of 30 and 120 minutes
The capture process affects small and large fish differently. Hansen and Jacobsen (1997) and Anon (1998) found evidence of size dependent vulnerability to long line capture and subsequent handling in Atlantic salmon. Larger salmon had significantly better Carlin-tag recovery rates than smaller fish, which during tagging were observed to lose scales more easily than the larger ones. The deleterious effects of capture on small fish have also been demonstrated for other species. Soldal et al. (1991) examined scale loss of escaped cod and haddock compared to a control group. On average, less than 1% of the total body surface of cod was injured, while haddock, particularly those smaller than 40 cm, showed substantial scale loss and therefore greater mortality.
Harrell & Moline (1992) have assessed the effects of electrofishing. Striped bass (Morone saxatilis) captured by electrofishing showed significantly lower effects of stress and shorter recovery times than striped bass caught in gillnets. Dalbey et al. (1996) observed that rainbow trout (Oncorhynchus mykiss) suffered significantly more incidents of spinal injury if pulsed rather than smooth DC was used. The severity of injuries was increasing with increasing fish length and, although long term survival was not affected, 28% of the fish had markedly lower growth and condition.
Angling appears to be a good way of catching some species of fish. Pankhurst & Dedual (1994) found no mortality in rainbow trout as a result of capture or any of the handling protocols. In most fish initially elevated blood plasma levels returned to normal within 24 h of capture indicating that metabolic recovery had occurred.
Tytler et al. (1978) gave wild smolts a recovery time of 3 - 48 h after anaesthesia and tagging in a portable holding tank before transporting the tank to the release site, where they were given minimum one hour to adapt to local river conditions. Moore et al. (1990) conclude that consideration must be given to a satisfactory recovery time before the fish are released from the controlled experimental conditions. Their results indicate that smolts can be safely released as soon as they are fully recovered from anaesthesia. Recovery from anaesthesia was judged to have occurred when full equilibrium was regained, and the fish reacted to external stimuli. Far better results have also been obtained for several Pacific salmonids (Oncorhynchus spp.) released immediately (e.g. Mellas and Haynes, 1985) instead of after prolonged recovery. Keeping wild salmonids for extended periods in tanks to recover after handling may give adverse results and may not improve fish survival (Nettles 1983).
In contrast, with hatchery fish, survival appears to be improved by the provision of a recovery period after handling. Sharpe et al. (1998) studied the effects of various hatchery practices, including tagging and fin- clipping, on juvenile chinook salmon. No lethal effects were observed, and though indeed stressful the physiological effects measured as elevated cortisol levels were of relatively short duration. Sharpe et al., nevertheless recommend that fish to be released into a more challenging environment than a hatchery should be given a recovery time of at least 24 h. The work of Hansen & Jonsson (1988) supports this observation as the survival of 1 and 2-year old hatchery smolts was reduced if they were handled immediately prior to release for sea ranching. Of the various treatments given, dipnetting significantly reduced the survival of the younger smolts, although it did not affect the older smolts significantly.
5.4 TAG ATTACHMENT METHODS
Electronic tags have been attached to fish both externally, in a variety of locations, and internally by insertion in the stomach or by surgical implantation in muscle or in the peritoneum. There are advantages and disadvantages for each of these methods and choice depends on the type of tag, the type of fish and its lifestyle and the purpose of the research.
5.4.2 External attachment
Internal tagging is not feasible with flatfish, such as plaice (Pleuronectes platessa), which have a tightly coiled gut and a small peritoneum; for these species external tagging is essential. External tagging may also be desirable in other species for reasons of tag or data recovery, even though internal tagging may be possible biologically. External tagging is simpler and quicker than most internal tagging, avoids surgery and anaesthesia and may also entail a shorter refractory period. External tags may be attached directly to the surface of the fish, or by a trailing lead that allows the tag to stream free when the fish is swimming. Tags have been attached dorsally, both anterior and posterior, dorso-laterally and ventrally. A few authors (e.g. Carr & Chaney, 1977) have used tags attached to the caudal peduncle, although this method is undesirable because it interferes with swimming.
184.108.40.206 Directly-attached tags
(a) Methods of attachment
Tags may be sutured directly to the body of the fish and this technique has been used with cod (Gadus morhua) to fasten ultrasonic transmitters to the dorsal surface ahead of the first dorsal fin (Sintef, 1983). Plaice (Pleuronectes platessa) have been tagged in a similar way with a transmitter fastened to the upper surface of the body (Sintef, 1983). Most external tags are, however, attached with fine wires or nylon cords, which pass through the body muscles and are attached to plastic discs or plates on the other side of the fish. The plate may be cushioned with foam to minimise scale damage.
One of the commonest positions for directly attached external tags is ventro-lateral to the dorsal fin. Usually this involves a single tag on one side (e.g. Gray & Haynes, 1979; Mellas & Haynes, 1985), although some studies have used a pannier arrangement (Thorpe et al., 1981; Greenstreet & Morgan, 1989) to equalise the load on the two sides of the body. Typically, the tag is attached immediately below the dorsal fin. This arrangement has been used successfully to fit salmonids (Salmo and Oncorhynchus spp.) with radio tags (Gray & Haynes, 1979), and also with data storage tags (Sturlaugsson, 1995). A transponding acoustic compass tag has been fitted to salmon in a similar fashion, using a plastic plate on both sides of the fish (Potter 19??). Bradbury et al. (1995) describe an interesting variant of the one-sided tag layout, which involves two tubes mounted one above the other. The lower tube contains the transmitter, while the other is partially filled with water to make the unit neutrally buoyant. The transmitter can be replaced when its batteries are exhausted or exchanged for a dummy transmitter of identical size and weight, while the fish recovers from the tagging process.
Cod have been fitted with acoustic tags in the same position (Arnold et al., 1992, 1994), although with the tag attached more loosely to the fish. Plastic spaghetti tags were passed through the dorsal muscle at either end of the first dorsal fin, using a surgical needle, and the ends tied in a reef knot. A 300 kHz transponding acoustic tag was tied to the spaghetti tags using a nylon cord at each end of the tag. Tesch (1974) used a similar arrangement to fasten an acoustic pinger alongside the anterior end of the dorsal fin of eels (Anguilla anguilla), although in this case a single perlon thread was used and the tag was coated in balsa wood to make it neutrally buoyant.
Nylon cable ties have been used to attach ultrasonic transmitters to yellowfin (Thunnus albacares) and bigeye (T. obesus) tuna immediately behind the last dorsal fin, where the body slopes down to the caudal peduncle (Holland et al., 1990). This method is probably only useful for large robust species.
In recent years, the Lowestoft Laboratory has attached 300 kHz transponding acoustic tags to plaice using a light ‘saddle’ made from a single stainless steel wire, which is inserted through the ‘dorsal’ muscles. A numbered Petersen disc is fitted to the underside of the fish, the wire is cut to length to allow for growth and the end twisted to form two or three rings as with a conventional Petersen tag. The acoustic tag is attached to the saddle by a nylon cable-tie and, as a safety precaution, a fine nylon cord is used to join the end of the tag to the top of the Petersen wire. This arrangement, which allows the tag to rotate a little, separates the tag from the upper surface of the fish and keeps the transducer clear of the sand when the fish buries into the bottom. A neoprene disc can be used to cushion the tag and protect the surface of the fish.
A similar arrangement was used to attach the Mk 1 Lowestoft Data Storage Tag (DST) to plaice (Metcalfe & Arnold, 1998). Two stainless steel wires were passed through small lugs on opposite sides of the circular tag and two Petersen discs were used on the under side of the fish. This system has been modified for the cylindrical Mk 3 DST. The wire passes through a rib moulded around the case of the tag and is held in position by a single large Petersen disc on the under side of the fish. The rib is flattened to allow the tag to rest on the surface of the fish and the Petersen disc has two holes. A similar arrangement was devised for a large acoustic tag that telemetered the compass heading of the fish back to the tracking ship (Mitson et al., 1982; Pearson & Storeton West, 1987; Metcalfe et al. 1993). The Petersen wires passed through a small hole at each end of a flat plastic plate, which was glued to a tapered wedge on the lower surface of the tag; two standard Petersen discs were fitted to the under side of the fish.
An unusual arrangement is possible with blue sharks, which often swim at the surface with the dorsal fin in air. The late Frank Carey of the Woods Hole Oceanographic Institution (WHOI) used a combined data logger and satellite transmitter to track the movements of three fish in the Gulf Stream from Cape Hatteras northwards. His design, which was based on a transmitter developed by the Sea Mammals Research Unit (Cambridge, UK), consisted of two aluminium pressure tubes cast into a polyurethane saddle, which rested on the back of the fish and a flange, which bolted through the dorsal fin. A streamlined mast raked back at the same angle as the leading edge of the fin carried a radio antenna at the top and a small propeller half way up the rear edge (Kingman, 1996).
There are a number of well-recognised problems with tags that are attached directly to the body of the fish with two or more attachment points, as described above. These problems, which include chafing, abrasion and ulcerated wounds, also arise routinely with conventional tags and are discussed further in Chapter 7. Chafing may be avoided initially by cushioning the tag on a thin layer of high-density foam, but often, as the fish grows, the space between the tag and the body wall disappears and the tag grows into the flesh of the fish. To date, this has not been too much of a problem with electronic tags, because most radio and acoustic tags have only a limited life. It is likely to become much more of a problem in the future with the use of archival tags with potential lives of 10 to 20 years. External tags can adversely affect various aspects of the behaviour and physiology of swimming animals, particularly if they have not been designed for minimal drag. There is scope for substantial improvement in this area, particularly when developing smaller tags. Shape needs consideration, as well as the method of attaching and mounting the tag. The work that has been done in recent years to improve the streamlining and positioning of tags on the backs of turtles (Watson & Granger, 1998) and penguins (Wilson et al., 1986; Gales et al., 1990; Culik & Wilson, 1991; Wilson & Culik, 1992; Bannasch et al., 1994) demonstrating the gains to be obtained from minimising tag drag.
220.127.116.11 Trailing tags
(a) Methods of attachment
For many years the Lowestoft Laboratory attached 300 kHz transponding acoustic tags to plaice and other flatfish using a nylon cord, which passed through the body of the tag just below the end cap, and was tied to the upper ring of a Petersen disc wire. This arrangement was very effective when the fish was in midwater. Flume studies (Arnold & Holford, 1978) showed that the tag streamed free of the body when the fish was swimming. On the bottom, the tag lay on the upper surface of the fish with the transducer close to the marginal (dorsal) fin. This was a poor arrangement when the fish was buried in sand, as the acoustic signal was often attenuated and difficult to detect. This problem was mitigated by the use of the saddle attachment described in section 18.104.22.168.
Similar single-point trailing attachments have recently been used to fasten positively buoyant data storage tags to the upper surface of cod just ahead of the first dorsal fin (Godø & Michalsen, 1997). In this case, the tags were attached in the same way as conventional Lea tags, using monofilament line inserted through the dorsal muscles.
Tethered tags require a strong permanent anchor point. With large fish this can be achieved by using a dart with an arrowhead that resists extraction from the muscles. Darts are commonly used with tuna, swordfish (Carey, 1973, 1981) and marlin (Holland, et al., 1990). An even better solution is provided when the dart is placed in the muscles at the base of a dorsal fin so that the barb penetrates the bony extensions at the base of the fin rays (CSIRO, 1992). Titanium (Block, 1997) and nylon darts (Prince, 199?) of this type have recently been developed in the USA for use with tuna and large billfishes (see section 22.214.171.124).
Trailing tags avoid many of the problems associated with close-coupled tags and, if properly designed, should produce limited drag. The original 300 kHz transponding acoustic tag developed at Lowestoft, for example, which had quite a high frontal drag coefficient (Cd0 = 0.6), was shown to have little effect on the swimming performance of medium size plaice (Pleuronectes platessa, 36-52 cm) and cod (Gadus morhua 50-70 cm). The majority of these fish would have been slowed down by rather less than 5% and the extra power output required for a tagged fish to maintain the same steady speed as an untagged fish of the same size was shown to be about 3-5% (Arnold & Holford, 1978). Little attention has, however, been paid to minimising drag either by optimising the shape of the tag, by or determining the optimal attachment point and tether length and this is particularly so for small and medium size fish. Pop-up satellite-detected tags are an exception and TECHNION in Israel has recently produced designs for a low-drag bomb-shaped towed body for use with large tunas and billfishes (reference ?).
126.96.36.199 Detachable tags
A pop-up tag satellite tag has recently been developed that will detach itself from the fish after a pre-set interval and float to the surface, from where it transmits a radio signal to an Argos satellite. The satellite determines the pop-up position and the tag transmits a limited amount of stored data after it has reached the surface. This is currently the average daily temperature for the first 60 days after release plus the average temperature on the day before the tag releases from the fish. The tag, which is designed for use with large pelagic species, is too large (34 x 4 cm, 65-68 g) for use with most exploited species in European waters. However, it has been used successfully on bluefin tuna (Thunnus thynnus) in the North Atlantic (Block et al., 1998; Boyan, 1998; Lutcavage et al., 1999) and Mediterranean (de Metrio et al., 1998). The tag is contained in a composite, positively buoyant, low-drag housing towed by a short (25-30 cm) leader attached to a tagging dart. The buoyancy is moulded to the rear of the tag, which floats vertically at the surface with a 16-cm aerial projecting vertically upwards above the surface. Prior to release, the tags, which are placed near the rear of the second dorsal fin, trail freely behind the fish with both tag and aerial horizontal.
The attachment dart, which is made of titanium (Block et al., 1998) or medical grade nylon (Floy, Inc.), can be inserted in the dorsal muscle (Lutcavage et al., 1998) or at the base of the second dorsal fin, where it can be anchored through the bony projections and connective tissue radiating ventrally from the fin (Block et al., 1998).
Block et al. (1998) caught large bluefin tuna on rod and reel with heavy tackle and tagged the fish on board a small angling boat. Lutcavage et al. (1998) caught large bluefin by rod and line, or purse seine, and tagged the fish in the water, using a custom-built applicator, or a harpoon (Chaprales et al., 1998). Lutcavage et al. (1999) and Block et al. (1998) respectively report success rates of 85 and 95% for data retrieval from batches of 20 and 37 pop-up tags released on large tuna in the western North Atlantic.
External tags can also be deliberately detached to avoid adverse long-term effects of tagging, or to recover the tag before it stops transmitting. One simple way of doing this is to use absorbable filaments. Baras (pers. comm.), for example, has used catgut 3.0 filaments to attach tags to grayling (Thymallus thymallus). Different release times can be achieved through use of filaments with different rates of absorption.
5.4.3 Internal attachment
Internal tagging is only suitable for a fish with a large stomach, or space in the body cavity into which a tag can be inserted without impeding or damaging the internal organs. Internal tagging avoids the causes of tag loss associated with external tags and has a number of positive advantages, not least of which is the proximity of the tag to the centre of gravity of the fish. But the method is not suitable for all applications and may produce signal attenuation if acoustic tags are used with large fish. It is usually also necessary to mark the fish externally, so that fishermen are aware of the presence of the internal tag, if recovery of the tag is required.
188.8.131.52 Stomach insertion
The stomach is the natural location in which to impose extra mass on the fish and this may explain why the use of stomach tags is often very successful, particularly if tags have been ingested voluntarily by the fish.
(a) Voluntary ingestion
Some fish will ingest acoustic transmitters embedded in baits deployed close to the sea floor. This technique has been used to excellent effect to study the short term movements of grenadiers (Coryphaenoides spp.) and other abyssal demersal fish in the Atlantic and Pacific Oceans (Priede & Smith, 1986; Armstrong & Badwin, 1990, Armstrong et al., 1992). The work was done with Aberdeen University’s free-fall vehicle, AUDOS, using approximately 0.2 kg of bait (mackerel flesh) wrapped around a number of ultrasonic transmitters or transponders. The baited packages were tied to a scaled cross in the field of view of a 35 mm underwater camera using fine thread. The remainder of the mackerel carcass was tied to the centre of the cross to help attract fish to the rig. Fish taking the baited tags triggered the camera and were subsequently identified from the processed photograph after the vehicle was recovered. The same technique has also been used with cod (Løkkeborg, 1998) and saithe (Engas et al. 1996?).
(b) Forced ingestion
Forced insertion of a telemetry tag into the stomach is readily achieved with a glass or plastic rod or tube (e.g. Monan et al., 19??), using glycerine as lubricant (Mellas & Haynes, 1985). This method of attachment is more commonly used with radio than acoustic tags and often involves an aerial wire fastened to the top of the mouth with a dart, or fed back through the gill slits and allowed to trail free in the water. Forced insertion is possible even with quite small fish and drinking straws have apparently been used to implant tags in the stomachs of 5-6 cm American shad. Insertion is easiest with a hollow tube fitted with a plunger (Nielsen, 1992). The tag is placed in the open end of the tube, flush with the end, and expelled when the plunger is depressed as the tube is withdrawn from the stomach.
184.108.40.206 Oviduct insertion
In salmonids and some other species, in which it is not connected to the ovary, it is possible to insert tags into the body cavity through the oviduct. Peake et al. (1997) have recently shown that it is possible to insert dummy radio transmitters into Atlantic salmon (Salmo salar) in this way without affecting survival, behaviour or egg development, provided insertion is done prior to egg formation, or after the eggs have been shed. The leading end of the tag was tapered to assist insertion. The radio aerial was allowed to trail freely from the oviduct. Some fish expelled the transmitter via the oviduct within 7-13 days of insertion but Peake et al. (1997) reported retention times of 60 days for salmon (~70%) that retained the tags for more than 14 days. Dissection showed that the tags were positioned well forward of the internal opening of the oviduct at, or near, the pelvic girdle. Tags were encapsulated in, and anchored by, a thin, transparent sheet of tissue. Similar trials with rainbow trout showed that reproductive success was compromised when the tags were inserted into fish with already developing egg masses. The technique may also be possible in female sturgeons (Acipenseridae), lungfish (dipnoans) and bowfins (Amia), which also shed eggs into the body cavity, and male hagfish and lampreys (agnathans), which similarly deposit sperm in the body cavity and have urinogenital ducts leading into the body cavity (Peake et al., 1997).
220.127.116.11 Intra-peritoneal surgery
Because of problems of regurgitation, abrasion and possibly predation, neither stomach tags nor external tags can offer long-term security of attachment. For long-term experiments, the solution is to insert the tag internally in the peritoneum. This has been done successfully over the last twenty years with a number of marine and freshwater species, both with and without the use of anaesthetics. Insertion need entail no more than making a small surgical incision in the body wall and the whole process can often be completed within a few minutes. Having had extensive experience of tagging several thousand southern bluefin tuna (Thunnus maccoyii), CSIRO has perfected the technique to the stage where a trained operator can insert an archival tag into an unanaesthetised fish in 50 s, a time that includes injecting antibiotics and suturing the wound (Williams, 1992). Surgery is, however, a delicate operation and field technicians need to be carefully chosen for their manual dexterity and seaworthiness in order to ensure the quality control vital for a successful tagging experiment (Gunn, et al., 1994; Gunn, pers. comm.). Similar protocols have recently been described for cod (Thorsteinsson, 1995) and Atlantic bluefin tuna (Block et al., 1998). Longer surgical operations have been carried out equally successfully in the laboratory with both cod (Pedersen & Andersen, 1985) and rainbow trout (e.g. Kaseloo et al., 1992) using controlled anaesthesia. The advantaged and disadvantages of anaesthetics are considered in Chapter 7, which also provides criteria for selecting the appropriate compound.
(a) Incision site and length
Once the fish is anaesthetised, an incision is made in the body wall with a scalpel blade. For some species, such as serrasalmids, which have a midventral cartilaginous structure and long ribs, only one site is possible for the incision (Baras & Westerloppe, 1995). For other species there is a choice that can be made on the basis of a number of criteria, such as innocuity, healing dynamics and minimum expulsion risk. Because the viscera lie in the dorsal part of the body cavity when the fish is turned upside down, midventral incisions are unlikely to cause direct internal damage. They are thus more frequently chosen (e.g. Hart & Summerfelt, 1975; Bidgood, 1980) than lateral incisions, which may puncture the gonads and prove more difficult to close because they involve a thicker body wall, longer healing times and lower survival rates. By contrast with midventral incisions, lateral incisions also cause systematic damage to bundles of striated muscle, for which degenerative processes often outstrip tissue reconstitution (e.g. Roberts et al., 1973; Birtles et al., 1995; Knights & Lasee, 1996). However, lateral incisions can be advantageous because the transmitter exerts less pressure over tissues weakened by the incision (Tyus et al., 1984) and there is less risk of expulsion through the wound than with a midventral incision.
In order to minimise the trauma to the fish, the duration of healing and the risk of expulsion of the tag through the wound, the surgical incision should be as short as possible. Key factors governing incision length are the diameter of the transmitter, its length and the flexibility of the fish body wall. The ratio of incision length to tag diameter is a convenient index. Feasibility studies (Baras, 1992; Baras & Westerloppe, 1995; Birtles et al., 1995; Thoreau & Baras, 1997) indicate that a ratio of 1.4-1.5 is appropriate for catfishes, which have a flexible body wall. A ratio of 2.5 is more suitable for serrasalmids, which have a thick body wall and in which a lateral incision is unavoidable (see above). For most cyprinids, salmonids and cichlids a ratio of 1.6 to 1.8 is suitable (Baras, pers. comm.).
(b) Implant size and weight
There are finite limits on the size and weight of an implant, which are determined by the size and species of fish to be tagged. The relevant factors are considered in Chapter 7.
(c) Internal position of implant
Internal transmitters may move within the body cavity and cause a variety of damage, such as gonad alteration (Chamberlain, 1979), internal haemorrhage (Bidgood, 1980), bruised liver, eroded rectum (Schramm & Black, 1984), and puncture of the intestine (Baras et al., 1996). Tags should therefore be put in a position that offers the least probability of movement, such as over the pelvic girdle, and as far as possible from hazardous locations, such as the pericardium or the incision wound. Internal movements can be restricted by suturing the tag to the body wall and this technique works well with the Atlantic cod (Pedersen & Andersen, 1985) but not with channel catfish, in which it induces systematic expulsion (Marty & Summerfelt, 1986). An aerial or umbilical passing through the body wall can limit the movements of the tag inside the body, although the benefits of reduced movement are offset by a higher risk of bacterial infection in the incision wound. In species in which the implant becomes encapsulated by host tissues, tag movements are in practice often restricted to the first days or weeks after surgery.
(d) Closing the incision
Surgical incisions can be closed with absorbable or non-absorbable sutures or stainless steel staples. With larger fish (< 5 kg) it may be appropriate to close the incision with a double row of sutures, one each for the peritoneum and skin (Summerfelt & Smith, 1990). There are several common suture patterns, of which the ‘simple interrupted’ and ‘interrupted horizontal mattress’ sutures are the strongest and most suitable for closing the skin of fish. Each comprises a series of independent knots. The continuous suture, which involves less trauma, is suitable for soft internal tissue but is less secure than the interrupted sutures because it has only two knots. The knot is the weakest point of a suture and if one knot becomes untied the entire suture will pull apart. Summerfelt & Smith (1990) give details of common sutures and knotting (their figures 8.2 to 8.4).
The commonest technique for closing abdominal incisions in fish is to suture at about 8 mm intervals with separate stitches right through the body wall (Hart & Summerfelt, 1975). For species with a rigid body wall or very thick skin, such as cichlids and catfish, round ‘atraumatic’ needles should not be used because their use may result in more damage and increase the time needed for tissue reconstitution (Baras & Westerloppe, 1995; Thoreau & Baras, 1997). There is a choice between absorbable (plain or chromed catgut) and non-absorbable (nylon or silk) filament. This often entails a trade-off between the risks of expulsion of the tag through an unhealed incision at the time the filament becomes dissolved (Schardt & Nallm 1981) and the risks of infection associated with a transcutaneous foreign body (Baras, 1992; Knights & Lasee, 1996). The removal of permanent suture filaments after the incision has healed may be advantageous but fish have to be held in captivity for longer and this is often detrimental to the fish or the experiment. Braided silk may be an undesirable suture material because fish have been observed to interfere with the healing process by grazing on stitches on which algae start to develop (Thoreau & Baras, 1997).
Suturing is time consuming and alternative methods of closing the wound may be desirable. Surgical staples can be used to close long incisions quickly (Mulford, 1984; Filipek, 1989; Mortensen, 1990) but necessitate the removal of more rows of scales than suturing and may render the fish more liable to fungal infection (Mellas & Haynes, 1985). They are also permanent transcutaneous foreign bodies.
Commercial grade cyanoacrilate adhesives, applied to the opposed edges of a blotted dry incision, enable the wound to close quickly and almost always suppress the inflammatory response at the incision site (Nemetz & MacMillan, 1988). However, they remain in place for a few days only and may result in more frequent loss of tags through the incision (Petering & Johnson, 1991). Baras & Jeandrain (in press) used cyanoacrilate to close incisions in European eels, for which suturing induces frequent necrosis of the body wall. They found, however, that the eels removed the adhesive within a few hours unless a biological bandage (a freshly cut fin fragment) was applied over the incision before the cyanoacrilate had dried.
A final option is to leave the abdominal incision open. This may cause ethical problems and favour bacterial infection, but it compares favourably with suturing in respect to survival, healing (Carmichael, 1991) and implant retention rate (Baras, 1992), especially for short incisions in small fish (Baras et al., 1996).
(e) Healing rates
Wound healing is a process of tissue reconstitution (Roberts et al., 1973; Marty & Summerfelt, 1990), whose dynamics are governed by factors, such as species, age, temperature and food availability, that control fish growth. Fast-growing tropical fish heal incisions within less than two weeks (Baras & Westerloppe, 1995; Thoreau & Baras, 1997). Temperate species require four to six weeks (e.g. Pedersen & Andersen, 1985; Baras, 1992; Birtles et el., 1995) and much longer at low temperatures (Ross & Kleiner, 1982; Knights & Lasee, 1996). Juvenile fish heal much faster than adults do. Wound healing in juvenile African cichlids and catfish (Baras et al., 1996; Baras & Westerloppe, 1996) and Atlantic salmon parr and smolts (Moore et al., 1990) occurs in 7-8 days and 14 days, respectively, equivalent to the resorption times of plain and chromed catgut.
18.104.22.168 Muscle implantation
To date most archival tag experiments with tuna have used tags implanted in the body cavity, although the NMT archival tag was originally designed to be inserted into the dorsal muscles of these fish. A series of trials with small yellowfin tuna (T. albacares) and small dummy tags (Brill et al. 1995) suggests that implants will also be retained successfully in dorsal muscle.
5.4.4 Effects of electronic tags on fish behaviour and physiology
There have been a number of studies on the impact of electronic tags on fish behaviour and physiology since acoustic and radio tags first began to be used in the early 1970s. These are reviewed in Chapter 7 (section 7.4).
5. 5 Recovery of data storage tags (DSTs)
5.5.1 Publicity and Rewards
In general, intensive tag recovery is not essential for either transmitting or transponding tags as these tags require specialised receivers. Research programmes are generally designed to ensure that the tagged fish come in contact with these receivers either by placing receivers in strategic locations or by actively tracking using mobile tracking equipment. However, tag recovery for data storage tags is essential as they do not transmit their position or information and each individual tag may contain an enormous amount of data. Despite this, few studies using electronic tags have directly examined whether the rate of recovery is sufficient to produce data which is representative of the problem being studied.
As the cost of each individual data storage tag is high, only a relatively small number may be used in any given research programme. This is offset by the amount of information that can be retrieved from even a few tags. The number of tags recovered will improve considerably with good publicity and reward systems in association with a good catch/stock scanning programme. Therefore, recovery programmes for data storage tags should include:
22.214.171.124 Investigation of the likely geographic area where tags will be recovered
Generally, programmes involving DSTs take into account the probability of re-encountering tagged fish subsequently. In marine fisheries, the area of encounter is potentially vast but can be reduced significantly with backup information from catch data or conventional tagging studies. Pre-tagging surveys with conventional tags should be carried out to provide a rough estimate of where the electronic tags will be recovered and what the target fisheries are likely to be. Subsequently, standard fishing techniques can be applied to recover tags or catches can be scanned in a similar manner to conventional tags.
For migratory fish species, the area of encounter can be more accurately predicted if the migration routes are known and the fish can be intercepted at specific points geographically or at certain points in the life cycle. However, DSTs may be particularly useful for studying anadromous fishes such as the Atlantic salmon which home with a high degree of accuracy to their natal rivers. Tagging of kelts (spent adult salmon) in rivers prior to their return to the sea has been suggested as a way of providing information on oceanic migrations of salmon. Provided these kelts survive to return as repeat spawners sufficient tags should be generated to provide important information on migrations of salmon at sea.
126.96.36.199 Advertising the tagging programme
Initially, the objectives, tag type, secondary tag type if used and the rewards (if any) should be clearly advertised. Prospective individuals likely to recover tags or be aware of recovered tags (fishermen, fish processors, anglers etc) should be informed that tags of different types may be present in the fish they handle. It is important to emphasise the scientific value of the information contained in the tags rather than the tag itself, and the overall benefits of the data for protecting and possibly enhancing stock assessment and management.
Methods of advertisement include:
Advertisements in national or local newspapers – if the tagging programme is locally based it is probably best to advertise only in local papers to emphasise the probable recovery location of the tagged fish.
Posters – these should show the features which will identify a tagged fish (presence of an external tag, fin-clip, mark etc) and a clear contact for return of the fish or the tag. They have been used extensively in conventional and electronic tagging studies and placed prominently in fish processors and fishing ports. A selection of typical posters which advertise tagging programmes and rewards for recovery of various types of tags are included in the web-site developed for this concerted action.
Public presentations – Experience has shown that direct interactions between scientists and commercial fishermen or the public improve the rate of recovery of tags and provides a more lasting impression of the objectives of the programme. Public presentations should be directed at fishermen and fishing organisations, processors, local representative groups and all users of the resource being studied.
Local interviews/contacts - again, direct contact with fishermen or other local contacts allows any queries to be dealt with expediently and creates a valuable dialogue between the scientists and public.
Subsequent reinforcement - reinforcing both the original message and the initial contacts has been shown to be effective in obtaining tags which might otherwise not be recovered, especially if tags may be recovered in more than one fishing season.
188.8.131.52 Adequate tag scanning programmes and sufficient sample sizes
Even if the general area of encounter has been identified, there is still the problem of tag retrieval. For marine fisheries, where shoal sizes may be large relative to the number of tagged fish, large numbers of fish may need to be captured to ensure a tag recovery. In general then, marine tagging programmes are normally associated with commercial fisheries where large numbers of fish are available for examination. For anadromous fishes, recoveries can be made in drift nets, traps, fish ladders or by angling and a systematic scanning programme at these recovery sites will greatly improve tag recovery.
Ideally, the entire catch should be examined for tags. If this is not feasible then a sufficient proportion of the catch should be examined which will be dependant on the estimated size of shoals, the temporal and geographic distribution of shoals and the number of tagged fish released initially. Significant improvements could be made if entire catches were routinely scanned for tags on board fishing vessels or in processing plants.
184.108.40.206 Simple identification of tagged fish in catches or samples
Clearly, catch scanning will only be effective if the tagged fish can be easily identified from non-tagged fish. This implies that a tag is clearly identifiable, or that the tagged fish is marked clearly with a secondary tag or mark. A message should be contained within the DST to inform the captor of the country of origin, tagging agency, the contact for tag return, information on rewards which may be available and any other instruction to the captor.
220.127.116.11 Clear instructions to fishermen and processors
Instructions on removing the tags and the procedures to follow for recording relevant information or retaining the actual fish should be issued well in advance of the tagging period and then reinforced while the fishery is taking place.
During intensive commercial fishing operations and in busy fish processing plants, retrieval of tags should not interfere substantially with routine processing or interfere with commercial operations. If tag removal is simple, then more co-operation can be enlisted from fishermen or fish processors who are more likely to come into contact with tagged fish. This can be done on a contract basis or by organising a fee for tags recovered. In some instances, the time available to fishermen or processors to retrieve tags may be short and it may be better to rely on trained technical personnel to scan landings and remove tagged fish.
18.104.22.168 Incentive to declare tags
Clearly, the data value of even a single DST is significant. Therefore the incentive to return tags should be high particularly if tag recovery is dependent on commercial fishermen or processors. The following incentives have been used extensively in conventional tag recovery programme with varying degrees of success.
A time honoured standard, but often it is difficult to decide on an adequate monetary reward. If the intention is to retrieve transmitting tags for reuse then certainly the reward should not be as high as the cost of a replacement tag. For DSTs the value must be decided in relation to the cost of the tagging programme, the value of the data and the effort needed to obtain tag recoveries. This may not be estimable in terms of direct cost benefit.
Gifts are often preferred as they are easier to administer and are often more acceptable, particularly if they have a high ‘popularity’ value. Institutes are moving towards offering T-shirts, sweatshirts, badges and peaked caps, all of which have a collectable appeal.
Often, the incentive to return tags can be increased if there is a corresponding return of information back to the individual recovering the tag, particularly if he/she is working within the fishing industry. Generally, the information would be in the form of an information leaflet outlining the objectives of the tagging study, any information on the tagged fish that was recovered and information on the overall results of the programme.
Publication of a list of individuals who have recovered tags in an institute or fishing newsletter is often useful to advertise the tag programme and encourage tag recovery.
As a general incentive, a lottery scheme can be a useful method to improve return rates for tagged fish. The names of people who have returned tags are entered into a draw and an overall winner or winners is randomly picked. This has the advantage that a substantially more attractive prize can be offered for the return of tags or tagged fish. A tag recovery lottery was carried out for a number of years by the North Atlantic Salmon Conservation Organisation (NASCO) to provide an incentive to fishermen to return conventional tags and improve the rate of tag return (ICES CM/1993 Assess:10, Ref. M), while the Iceland’s Institute of Freshwater Fisheries operates a lottery for the return of DST tags as well as conventional tags on salmon.
5.5.2 International Collaboration
Tagging programmes involving highly migratory fish species or stocks that are exploited by several different national fleets need special approaches for tag recovery. Again, a high degree of advertisement and publicity should be established between national co-ordinating agencies as outlined above. A separate reward scheme could be considered for tags returned from non-national fisheries.
The tagged fish should be readily identifiable to the captor particularly if the fish has an internal DST. At the very least, a message should be contained within the DST to inform the captor of the country of origin, tagging agency, the contact for tag return and information on rewards that may be available.
Considering the widespread use of conventional tags and the similarity of these tags being used internationally, it is recommended that a special conventional tag is used with fish containing internal electronic tags. These could be differentiated by colour, code or shape and should be advertised widely, both nationally and internationally as being specifically for this purpose.
Electronic mail and the World-Wide Web should be encouraged as a method of advertising tagging programmes which have the potential to generate tag returns in non-national waters. The Web-site developed within this Concerted Action will provide an international forum for informing other agencies of ongoing or new tagging programmes and should go some way to stimulating co-operation in returning tags.
5.6 APPLICATION TO FISHERIES
Electronic tags are now being widely applied in many areas of fish biology and fisheries management. Generally, electronic tags are used to provide information which cannot be obtained using conventional tags. The main areas of application are given below with some specific examples quoted to illustrate the type of information which can be obtained from electronic tags.
5.6.1 Investigating fish behaviour in relation to fishing activities
a) Behaviour of fish in relation to vessels and gears
The examination of fish behaviour in relation to fishing vessels and fishing gear is one of the most important areas of application for electronic tags and one which is likely to develop significantly in future. Fish senses are highly developed and apart from sight and smell, fish can be extremely sensitive to even minute vibrations in the water or on the sea bed. Electronic tags allow for real time tracking of fish or groups of fish and provide information on the reactions of these fish as the fishing vessels and gears are operating.
Telemetry studies in the late 1960s clearly showed that fish could detect and avoid fishing gear by sight and by other senses when light intensities were inadequate for the fish to see the gear. Shad (Alosa sapidissima) migrating up the Connecticut river reacted to drifting commercial gillnets at ranges of 1-2 m and few were caught. Similar Norwegian studies in the early 1990s showed that cod (Gadus morhua) could detect and avoid a 30 m trawler approaching at a speed of 1 m s-1. The fish reacted to the noise of the vessel at a range of 200 m and accelerated and swam out of its path when the range decreased to 100 m. In more recent experiments, Norwegian fisheries scientists have made further observations of the reactions of cod to trawls using a fixed hydrophone array with radio-telemetry buoys to transmit data to a research vessel. The system has also been used to investigate the reactions of Norway lobsters (Nephrops norvegicus) to baited pots. An electromagnetic tracking system has been used in Australia to investigate the effects of baited traps on the foraging movements of juvenile western rock lobster (Palinurus cygnus).
b) Improving fishing gear efficiency
Leading on from studies which investigate the avoidance behaviour of fish to vessels and gears, specific studies to improve the efficiency of fishing gears have developed from electronic tag applications. In the 1970s, the Directorate of Fisheries Research at Lowestoft, UK, carried out a major investigation to measure the efficiency of the Granton otter trawl on a flat sandy ground in the southern North Sea. The work was carried out over seven years and involved releasing several hundred plaice (Pleuronectes platessa) tagged with small transponding acoustic tags. One research vessel with a sector scanning sonar was used to observe the fish; a second vessel was used to tow the trawl. The results indicated that modifications of the gear could increase the efficiency of the trawl from 44% to 80%. These fishery-independent estimates of gear efficiency appear to be unique and more applications should be developed using the available technology.
c) Improving estimates derived from acoustic survey
Stock assessments of many fish species are now routinely carried out using acoustic technology. The results of these assessments are used to provide management advice for many of the most important marine stocks. However, validating the results of the acoustic trials is extremely time consuming and results in large expenditure of capital, shiptime and manpower. Specifically, biomass assessments from interpretation of the acoustic signals from shoals of fish may alter significantly during active migrating and feeding periods.
Data storage tags have been applied to investigate the accuracy of acoustic assessments for gadoid fishes. It has become apparent over the last fifteen years that acoustic estimates of the biomass of gadoids may be underestimated if no account is taken of the reductions of 2-5 dB in average target strength (TS) caused by changes of attitude (pitch or tilt angle) of the fish, or because of changes in swimbladder volume caused by feeding or gonad maturation. Ultrasonic tracking studies in the southern North Sea (which have shown that cod are neutrally buoyant at the top of their vertical range but negatively buoyant on the sea bed) indicate that vertical migration may be accompanied by systematic and possible even larger changes in TS than those associated with feeding or gonad maturation. Negative buoyancy is accompanied by compensatory changes in attitude that may further reduce TS but no attempts have as yet been made to measure them, although at least one telemetry tag has been developed for this purpose.
Considering that acoustic data are now used extensively to provide information on biomass, applications which lead to improving the efficiency and the applicability of these estimates are essential.
5.6.2 Investigating fish migration, migration routes and distribution
a) Vertical and horizontal movements of oceanic fish
While conventional tagging studies have provided much information on the extent of migrations of many oceanic fish species, they provide little information on the behaviour of the fish from the time the fish is tagged to the time it is recaptured. Ultrasonic telemetry is now being used extensively to study both vertical and horizontal movements of a wide range of oceanic fish, including salmon (Onchorhynchus spp.) and large pelagic species such as tuna, billfish and sharks. It has been possible to link behavioural changes with specific feeding events or reactions to thermoclines.
Japanese studies of the behaviour of the six species of Pacific salmon (sockeye, chum, pink, coho, chinook and steelhead) in the central Bering Sea and North Pacific indicate that these fish occur mostly in the upper 50 m of the water column, with occasional forays to greater depths (150-200 m max.) Sockeye, pink, coho and steelhead are restricted to the top 10 m for over 70% of the time and chum salmon also swim near the surface; chinook salmon occur at depths of 20-40 m. Pacific salmon show few regular patterns of vertical migration.
In contrast, the large pelagic species - tuna, billfishes, sharks - exhibit a number of patterns of vertical migration, which appear to be associated with feeding, thermoregulation, or the avoidance of limiting oxygen levels. Some species, such as skipjack (Katsuwonus pelamis), yellowfin (Thunnus albacares) and giant bluefin (T. thynnus) tuna (Lutcavage et al., 1997) and blue marlin (Makaira nigricans), appear to be confined to the thermocline and the mixed surface layer of the ocean, swimming nearer to the surface at night than by day. Others, such as the bigeye tuna (T. obesus), move rapidly up and down the water column apparently without regard for the thermocline, though the species forages in deeper colder water by day and makes regular, rapid ascents back into warmer surface waters in order to recover lost heat. Swordfish (Xiphias gladius) also make extensive diel vertical migrations, swimming deep by day and coming near the surface at night. They appear to follow isolumes and have been recorded at midday depths of over 600 m in well oxygenated water in the Atlantic. X. gladius has a large ‘brain-heater’ behind the eye and can maintain its brain temperature at 10-12° C above ambient temperature in water of 8-17° C, allowing it to forage in cold water below the thermocline (Block, 1986).
Significant further progress in this area will accelerate with the increasing use of satellite tracking and data storage (archival) tags. Blue sharks often break the surface with the dorsal fin and the Woods Hole Oceanographic Institution (USA) tracked sharks using a satellite transmitter, based on a design developed by the Sea Mammal Research Unit (Cambridge, UK), which bolted through the fin and carried a radio antenna on top of a long raked and streamlined mast (Kingman, 1996).
US scientists began a large tagging programme with Atlantic bluefin tuna (Thunnus thynnus) in the western North Atlantic in 1996. The fish are tagged with archival tags. The aim is to investigate migration and spawning site fidelity and test the current ICCAT management hypothesis that there are discrete eastern and western stocks (Block, et al., 1998).
b) Behaviour of shelf seas fishes
There has been a substantial amount of fish tracking work on the European continental shelf over the last 25 years and this has significantly advanced our understanding of the behaviour of free-ranging fish in the open sea. A wide range of species has been studied, including salmon (Salmo salar), eel (Anguilla anguilla), dogfish (Scyliorhinus canicula), plaice (Pleuronectes platessa), sole (Solea solea) and cod (Gadus morhua). There has also been some limited work on the Norway lobster (Nephrops norvegicus), the European lobster (Homarus gammarus) and the spider crab (Maja squinado).
Elsewhere around the world there has been research on the of local movements of a variety of shellfish and finfish. Shellfish tracking studies have included work on queen conchs (Strombus gigas), American lobsters (H. americanus), spiny lobsters (Panilurus argus), king crabs (Paralithodes kamtschatica), portunid crabs (Scylla serrata) and prawns (Macrobrachium rosenbergii). Research on finfish has included the overwintering behaviour of cod (G. morhua) in near-shore waters in Newfoundland, studies of lingcod (Ophiodon elongatus) and yellowtail rockfish (Sebastes flavidus) in the Pacific Northwest, and investigations of the movements of white goatfish (Mulloides flavolineatus) in a fisheries conservation zone in Hawaii.
Sonic tags have also been used to study the homing behaviour of sharks in tropical lagoons and on a variety of reefs, banks and seamounts in temperate and tropical waters. These studies have encompassed space utilisation and social behaviour, as well as diel patterns of movement and metabolic rates.
In Europe, open sea tracking of Atlantic salmon has focused on coastal movements during the return spawning migration and has been carried out both in UK and Swedish waters. The work has shown that adult salmon can maintain a compass course over quite large distances irrespective of current or tidal stream direction. It has also shown that salmon may exhibit diel vertical migrations, swimming close to the surface during daylight but descending to depths of as much as 40 m at night. Individual fish may also show large vertical movements near river mouths and it has been suggested that this behaviour may be related to olfactory discrimination of fine scale hydrographic features during the search for the home stream. Some data also suggest that salmon swimming at the sea surface may be orientating to surface swell patterns. The European salmon tracking work has been paralleled by similar work with Pacific salmon on the west coast of North America and also in Japan.
Ultrasonic telemetry has been used to demonstrate selective tidal stream transport in plaice, sole and dogfish, as well as in cod and silver eels. Its importance as a migratory mechanism for adult plaice in areas of fast and directional tidal streams has been confirmed by comparative fishing experiments with midwater trawls. Similar experiments with smaller nets have demonstrated selective tidal stream transport in newly metamorphosed plaice larvae, as well as juvenile plaice and sole, and shown that the mechanism plays an important role in the movements of young fish to and from their coastal nursery grounds.
Tidal streams are essentially deterministic and research on selective tidal stream transport has provided the basis for a computer simulation model which can be used to predict rates and scales of movement of demersal fish on the European continental shelf. The model has been validated by reconstructing the ground tracks of fish tracked in the open sea using sector scanning sonar and transponding acoustic tags. Potential applications include estimating transfer coefficients for use in spatial assessment models, assessing the exchange of fish between fishable areas and areas closed to fishing for conservation reasons and providing a theoretical framework for analysing returns from conventional fish tagging experiments. The model is being used predictively to forecast the movements of both individual fish and populations and retrospectively to reconstruct the ground tracks of fish fitted with data storage tags. Data storage tags have recently begun to accelerate the rate of progress of research on fish migration and movement in the North Sea and have begun to provide data on the behaviour of individual fish over periods of many months.
c) Estuaries and coastal waters
Data storage tags have recently been used on salmonids in Iceland and the Baltic. The main aim of the Icelandic work (Sturlaugsson, 1995, 1997) has been to study the final stages of the homing migration of salmon and the Baltic study (Karlsson et al., 1996) had similar aims. The Icelandic investigations also included studies of growth and movement of sea trout in the sea (Sturlaugsson, 1996). In the Icelandic studies, Atlantic salmon (Salmo salar) were captured in the estuaries of their home rivers, tagged and released at a number of sites at distances up to 200 km away. In the Baltic, fish were caught much further away from their spawning rivers and released in the same location. Both studies confirmed that migrating salmon swim within a few metres of the surface for much of the time, allowing satellite measurements of sea surface temperature (SST) to be used to deduce location and movement. Recapture rates of 50 to 70% were achieved in these data storage tag experiments (Karlsson et al., 1996; Sturlauggsson, 1997).
Pioneering work with ultrasonic tags in northern California in the 1960s established minimal acceptable levels of dissolved oxygen, water temperature and flow for the upstream passage of migrating king salmon (O. tshawytscha) in the San Joaquin delta. However, the application of telemetry to this type of environmental management has not been extensive and subsequent studies in Europe along similar lines have tended to be small and incomplete.
Many important insights into the migratory behaviour of anadromous and catadromous fish have come about with telemetry using radio tags. Specific applications have been to investigate the movements of adult and juvenile salmon in relation to fish passes, hydroelectric generation stations, barrages and a wide variety of man-made obstacles including thermal and chemical effluents. Information from tracking studies has provided information on exploitation rates by commercial and recreational fishermen and estimates of illegal catches of salmon in some rivers. Recent telemetry studies have shown that much of the mortality associated with smolt migration occurs within the first weeks of migration as the fish pass through freshwater and into the estuary. This has important implications for life-history models which are used for providing management advice to ICES and NASCO.
Telemetry studies have been carried out to monitor the bahaviour of important coarse fish populations in Ireland. In particular, the homing and territorial behaviour of pike has been well described. Transplantation studies have shown that adult pike can travel long distances to return to their own territory. Recent investigations have been carried out in Ireland to investigate multiple capture of coarse fish in competition stretches of important coarse angling venues and to assess the impact of such angling on the populations.
5.6.4 Assessments of predation and other multi-species interactions
Despite the significance of multi-species interactions, this area of fisheries biology and management is very poorly described or understood. Most assessments are carried out on a stock by stock basis. Attempts to add extra parameters to account for interactions between stocks generally lead to extremely complex analyses and increased uncertainty in results. This is mainly due to the lack of reliable data on real rather than simulated interactions. Despite this, fishery scientists are becomingly increasingly dependent on the results of these analyses to provide advice to managers. Studies using electronic tags can be applied to describe real interactions and the scale on which these interactions occur.
a) Predation by other fish species
With the exception of deep-sea scavengers, electronic tags have been little used to date to study feeding and predation of fish in the open sea.
Aberdeen University in the UK has used its AUDOS autonomous free-fall vehicle and baited acoustic tags to study the foraging behaviour of grenadiers (Coryphaenoides spp.) and other deep-sea fish at various oligotrophic and eutrophic sites in the Pacific and Atlantic oceans. Using this technique to measure times of arrival and departure of fish from the vehicle, it has been possible to show that grenadiers are active scavengers which move independently of the abyssal currents. Population densities are higher at eutrophic sites than oligotrophic sites but, in accordance with optimal foraging theory, staying times are significantly longer at oligotrophic sites. Staying times also vary seasonally and appear to reflect seasonal variations in the supply of food reaching the ocean floor. Application of this technique might be expected to produce useful results in the study of multi-species interactions in shelf seas.
b) Interactions between fish and sea birds
Archival tags have been used to study the feeding ecology of oceanic sea birds alone or in combination with satellite telemetry tags. Foraging location has been provided by the satellite tag or by a light sensor in the archival tag. Foraging behaviour has been recorded by an external temperature-depth sensor or an ingestible stomach-temperature sensor, designed to be recovered by stomach flushing when the bird returns to the nest. These techniques have revealed quite a lot about seasonal and diurnal patterns of feeding activity in sea birds which are known to prey on important commercial fish species. Important information has been obtained on rates of capture of individual prey items and the quantity of food ingested during foraging excursions. Temperature recorders have proved particularly useful by revealing a characteristic pattern in the change of stomach temperature associated with feeding.
Oceanic seabirds may capture pelagic prey that avoid ships and fishing gear often occurring in the upper 5-10 m of the water column that cannot routinely be sampled by acoustical means. This attribute has been employed recently using birds tagged with electronic tags to acquire physical and biological data that cannot be obtained in conventional ways. Albatrosses (Weimerskirch et al., 1994) and penguins (Ancel et al., 1992) have both been used to assess the distribution and abundance of inaccessible biological resources in Antarctic waters. Pygoscelid penguins fitted with data logging tags have been used to determine the distribution of krill in relation to depth and water temperature and obtain catch per unit effort indices of krill abundance (Wilson et al., 1994). Two species of albatross fitted with satellite telemetry tags have similarly been used to locate concentrations of ommastrephid squid and study trophic interactions in relation to the mesoscale oceanography of the Southern Ocean (Cherel & Weimerskirch, 1995; Rodhouse et al., 1996).
Radio telemetry has been used to investigate the foraging activities of cormorants and shags (Phalacrocorax spp.), foot-propelled pursuit divers that feed on sandeels (Ammodytes spp.) and other marine fish, and do not range too far from their breeding colonies. Changes in signal characteristics indicate when the bird is at the colony or away feeding; breaks in signal transmission indicate when the bird is diving in pursuit of prey (Wanless & Harriss, 1992; Wanless et al., 1993). Combined with automatic electronic balances, which measure adult body mass before and after a foraging trip (Grémillet et al., 1997a) it is now possible to measure daily food intake and foraging effort. This in turn allows calculation of catch per unit of effort, gross foraging efficient and parental investment at different breeding stages (Grémillet et al., 1997b).
c) Interactions between marine mammals and fish
Telemetry investigations using electronic tags on seals in UK and Antarctic waters and whales in the Arctic, have enormous potential for investigating predation of marine mammals on fish and determining how feeding distribution compares with that of fish and fishing fleets (Harwood, 1992). The diving and foraging behaviour of grey seals (Halichoerus grypus) has been studied in the northern North Sea and off the Hebrides, using a combination of depth-telemetering acoustic tags, VHF radio tags, which transmit when the animal is on the surface, and satellite telemetry (Thompson et al., 1991; McConnell et al., 1992). Analysis of diet and dive data shows that grey seals feed almost exclusively on benthic or demersal fish, foraging exclusively on or near the sea bed. Sandeels and large gadoids (cod, whiting, haddock, saithe and ling) dominate the diet. Off the east coast of England grey seals concentrate their foraging activities over areas of gravelly sand. Grey seals are observed to dive directly beneath dense assemblies of feeding seabirds - mostly gannets (Sula bassana), kittiwakes (Rissa tridactyla), puffins (Fratercula arctica), guillemots (Uria aalge) and shags (Phalacrocorax aristoteles) - and may then be feeding on the deeper parts of shoals, or on predatory fish (Thompson et al., 1991).
These mammalian studies emphasis the importance of understanding behaviour and spatial dynamics in any quantitative analysis of predator-prey interactions (Croxall et al., 1985). They give some idea of what might be possible with the multi-species problem, if electronic telemetry was applied to investigating fisheries problems .
5.6.5 Aquaculture and sea ranching
Electronic tags have only been used in aquaculture relatively recently, with the application of fixed omnidirectional hydrophone arrays to the problem of determining the position of fish within or around fish farm cages. Recent studies have also begun to investigate how activity varies with social factors, such as fish density, and environmental factors, such as wind, rain and temperature. PIT tags have also been used to monitor activity of fish at demand-feeders and investigate the effects of dominance hierarchies on feeding and growth rates. Sea ranching applications are also recent and have been used in comparative studies of the local migratory behaviour of wild and farmed Atlantic salmon and the extent of upstream migrations between ranched, escapee and wild salmon.
5.7 FUTURE APPLICATIONS AND DEVELOPMENTS
Electronic tags have a major role to play in fisheries science in the next century and will provide solutions to many currently intractable problems. In recent years there has been a marked resurgence of interest in bio-telemetry as newer equipment has become available and novel research possibilities have been created. The development of faster and cheaper microprocessors, coupled with the development of sophisticated software, means that complex algorithms can easily be incorporated into new tracking systems. New batteries, and smaller, more powerful and more efficient transmitters, have overcome many of the earlier problems of longevity and reliability. Data can now be recorded from a number of animals over long periods and many different kinds of environmental and physiological information can be obtained simultaneously. Furthermore, the continued refinement of surgical procedures and the development of new and safe anaesthetics have permitted an increase in the size and diversity of fish species that can now be successfully tagged.
This section sets out to identify the areas of research that are likely to benefit from the application of electronic tags, specify some relevant research objectives and identify the technical developments that are needed to realise them.
22.214.171.124 Migration and distribution
Because most conventional sources of data (surveys and simple tagging experiments) are biased by the distribution of fishing effort, good descriptions of migration and distribution are lacking for many commercially exploited species of fish in the open sea. This lack applies as much to demersal species, such as cod, in shelf seas, as it does to far-ranging diadromous species, such as salmon and eels, or other ocean migrants, such as tuna. Much less is generally known, though, about distribution and migration in the open ocean, and for many large pelagic species it is often not possible to provide a description of geographical distribution for all stages of the life history. This type of information is, however, essential for effective fisheries management and will be increasingly required as a result of the UN Agreement on Highly Migratory Species and Straddling Stocks. It is also required to understand ecological processes and how different species and size-classes of fish interact with one another. Similar arguments apply to vertical migration, which is a major feature of fish behaviour in most marine environments, and which appears to serve a number of different ecological functions.
Important objectives in this area of research include: describing migratory pathways and seasonal changes in vertical and horizontal distribution; identifying guidance mechanisms; and describing and understanding the functions of vertical migration. Major technical challenges are posed by determining geographical location, identifying the water mass (e.g. by temperature and salinity) in which fish are swimming, and recording orientation and swimming speed. Physiological measures of condition and reproductive state are also highly desirable.
126.96.36.199 Methods of estimating fish abundance
Fish behaviour strongly influences estimates of population abundance derived from static fishing gear, survey trawls, and acoustic instruments (echo sounders & sonars). It may bias results in a variety of ways, which may or may not be systematic. Fish that encounter the gear may avoid capture by reacting to individual parts of the gear or to the noise it produces. These reactions may vary with the size of the fish and ambient environmental conditions. ‘Natural’ behaviour, which governs horizontal and vertical distribution, determines the extent to which sampling gear encounters fish at all. For acoustic surveys changes in swimbladder volume or tilt angle, which occur naturally during vertical migration, or as the fish reacts to the noise of an approaching survey vessel, can cause major variations in target strength (TS) measurements and any estimates of abundance derived from them.
Research on ‘natural’ behaviour needs to focus primarily on the vertical and horizontal movements that determine spatial distribution and thus availability and accessibility to sampling gear. The rates and extents of such movements vary with both biological and environmental factors (e.g. light intensity, temperature and tidal currents) and the effects of these factors need to be determined. It is similarly important to determine how fish regulate buoyancy in relation to depth and how they compensate for negative buoyancy by tilting the body in the vertical plane. Static and mobile fishing gears work in different ways and studies of vulnerability accordingly have different objectives for each type of gear. Static gears work by chance encounter (e.g. gillnets) or by attracting fish or shellfish from a distance with bait (e.g. pots and longlines) and an odour trail. For gillnets the research objective is to determine how fish move as they approach the net and how visibility of the net affects their avoidance reactions. For baited gear the aims are to define the shape and size of the odour trail in relation to the prevailing currents and the concentration of the olfactory stimulant, and to determine whether catch per unit effort is a reliable measure of population density. With towed fishing gear the principal objective is to study avoidance reactions and determine how capture efficiency differs between sizes and species of fish and varies with physical factors, such as temperature, light intensity and underwater visibility.
To meet these various objectives we need to measure one or more of the following physical quantities: noise, temperature, light intensity, turbidity, depth, rate of ascent or descent, tilt angle, swimming speed and reaction distance. In most cases the measurements must be made at the fish rather than at the research vessel and for some projects it may also be necessary to measure a physiological parameter, such as heart rate. Measuring and recording, or telemetering, these variables is difficult at present, particularly when real-time observations are needed for more than just a few fish at any one time.
188.8.131.52 Species interactions
Electronic tags have the potential to tell us a great deal about how and when fish eat and how much food they consume. Knowledge of the natural behaviour of free-ranging fish would provide a major impetus to the study of multispecies interactions and reduce the current over-reliance on theoretical models. It could also be used directly to correct or tune the multi-species VPA models used to provide advice for fisheries management. Improved knowledge of natural behaviour would also advance our understanding of a number of important ecological processes, such as habitat selection (particularly important for small fish) and partitioning of resources between apparently sympatric species. For most marine species, habitat changes markedly as individuals grow from one size class to the next, alter their physiological optima, change their prey and become susceptible to larger predators. Changes usually occur in three dimensions, not just two. In addition to providing a better understanding of ecological processes, quantitative estimates of feeding rates in free-ranging fish would provide an important practical tool for interpreting gut content data collected during multispecies fisheries surveys. These data are currently difficult to interpret within the confines of existing knowledge, which is based almost entirely on laboratory studies.
Initially, any investigation of habitat selection and resource partitioning needs good descriptions of the horizontal and vertical distributions of the fish in question, by size and species. The second objective is to describe natural feeding behaviour and measure rates of encounter between predators and prey, rates of predator avoidance and feeding success. These quantities are needed to estimate costs of predator avoidance in relation to lost feeding opportunities. Locomotory costs are also important in establishing energy budgets and, in this area, measurements are needed of burst swimming speeds during predator escape reactions and cruising speeds during feeding. Direct measurements are also needed of basal metabolic rates and quantities of food consumed. There are major technical challenges in developing devices to measure and record these parameters.
184.108.40.206 Growth and reproduction
On-line estimates of growth and reproductive condition could be extremely useful both in understanding ecological processes and for practical applications such as sea ranching and stock enhancement (see 220.127.116.11). In this context a thermal history of the fish throughout its time at liberty would be of great interest, particularly if temperature measurements recorded by data storage tags could be correlated with a direct estimate of growth rate from scales or otoliths. The identification of feeding locations in relation to the productivity of different water masses would also be highly informative. A measure of gonad fullness would aid studies of reproduction and a means of identifying specific spawning events could also be most useful. Tail beat frequency is a good correlate of spawning in salmonids, as is the noise made by the fish when they cut redds in gravel spawning beds. Other species, such as cod and haddock, have a repertoire of sounds that they produce during spawning. Biosensors in the blood would allow us to measure and record hormone levels and correlate them with different patterns of behaviour. This capacity would significantly advance our understanding of the links between physiology and behaviour. These requirements, which are to a large degree shared by the other research areas identified in this section, provide a major challenge for sensor development and miniaturisation.
18.104.22.168 Aquaculture, sea ranching and enhancement
Electronic tags have so far made a relatively limited contribution to aquaculture, although there is considerable scope to apply the technology to the investigation of a number of physical and biological factors that control production. The aquaculture industry should be encouraged to investigate these opportunities. Applications include feeding and energetic studies, which could serve as a useful precursor to similar studies with fish in the open sea (see 22.214.171.124). Studies of interactions of fish in rearing cages with predators or wild fish outside the cage would also be useful, as could a cheap identification tag that allowed escapees to be quickly and readily identified. Fish health is probably the highest priority in the fish farming industry and techniques for long-term monitoring of fish condition have considerable potential, particularly if data could be recovered regularly without removing the tag from the fish. Stress resulting from handling is obviously an important factor for fish kept at high densities and biosensors linked to data logging tags have considerable potential in this field, as they do, for example, in relation to studies of growth, migration and reproduction of free-ranging fish (see sections 126.96.36.199 and 188.8.131.52). Stress also arises in relation to slaughter and saltwater/freshwater transfer.
Ranching and enhancement studies are concerned with where hatchery fish go in the wild and how they interact with wild stocks. Objectives are thus very similar to studies of migration and distribution in wild stocks and involve descriptions of local and migratory movements, geographical location, swimming behaviour and the measurement of appropriate environmental factors, as discussed in previous sub-sections.
184.108.40.206 Anthropogenic effects
Existing research has already identified the benefits of using electronic tags to study the impact of man-made structures on the distribution and abundance of fish. The construction of dams and other barrages in rivers has had a major impact on fish populations through disruption of migration and reproduction and electronic tags have been widely used to test the effects of mitigating measures, such as fish passes. They have been used less frequently to assess the impact of proposed structures before construction. This is an important area for the future, however, particularly in relation to the impending development of hydropower in big tropical rivers (e.g. in Southeast Asia), where large proportions of the human population depend on fisheries, and where most of the fish are highly migratory. There are similar opportunities in the sea in relation to policy decisions on the future of decommissioned oilrigs and studies to assess the uptake of pollutants by fish attracted to feed in the vicinity of drilling platforms. Electronic tags offer an ideal way of studying local movements of fish in the neighbourhood of these and other structures (e.g. effluent discharge pipes) and also the migrations and seasonal movements that are capable of dispersing disease and pollutants over wide areas.
5.7.2 Biological improvements
Engineering will provide many of the technical advances needed to improve tag performance and reliability and this topic is discussed in section 5.7.3. Technology will not realise its full potential, however, unless biologists also make significant improvements to the way in which they capture and handle fish and attach tags. Some of these issues are addressed here; others are dealt with in Chapter 7.
220.127.116.11 Capture and handling fish
Reviews of the effects of capture, handling and tagging inevitably focus on the negative aspects of these procedures and tend to obscure the fact that, in many cases, it is already possible to obtain fish in excellent condition (see Appendix 1 of Chapter 7). Often, however, the relevant expertise is passed on by word of mouth and much useful knowledge never finds its way into the ‘grey literature’, let alone refereed scientific publications. There is a general need, therefore, for improved documentation of the various capture procedures and a codification of general principles. This clearly needs to be done in respect of each type of fishing gear (lines, trawls, traps etc.) and capture method. The incompleteness of existing information identified in section 5.3, however, means that there is also a need for systematic investigations to determine the effects of capture and handling on the condition and survival of different species of fish at various stages of their life history. While there is a general need for more research on the effects of these processes on commonly tagged fish in temperate waters, there is an even greater need research on tropical species. Special attention also needs to be paid to methods of handling endangered species and delicate species of delicate life history stages.
During research, careful records must be kept of the size and condition of the fish, as well as environmental conditions, and any factors relevant to the specific method of capture. For trawls these factors include speed of towing, haul duration, and depth. The size and composition of fish catch is also important, as is the quantity and type of by-catch. By-catch can significantly increase mortality, especially in bottom trawls, where sharp objects such as shells and spiny fish and invertebrates can do a great deal of damage. Qualitative observations suggest that it is probably possible to define levels of debris in trawl catches above which it is not possible to use fish for tagging at all. Other, comparable constraints may apply in the case of line- and trap-caught fish.
Laboratory studies offer one way of recording mortalities and observing the condition of fish after capture, which is generally regarded as a more serious cause of damage than tagging. The fish must, however, be returned to the laboratory from sea and this process may exacerbate any problems caused during capture. It may therefore be better to make the observations at sea using cages to monitor condition and survival, as described in section 5.3.4(b). Further work of this type should be encouraged, even though it is not easy to do for logistic reasons. A third option is to obtain information from data storage tags, which may reveal how long fish exhibit atypical behaviour after tagging and release before resuming natural activities such as migration and spawning. Confirmation of spawning can also be obtained by recovery of the carcass of the tagged fish, which can also reveal the state of any wounds associated with tag attachment. Pilot projects with dummy tags are recommended before starting DST tagging programmes as they can clarify the effects of capture, handling and tagging on the fish and indicate the expected recovery rates of the electronic tags.
While systematic studies of the effects of conventional methods of capture are essential, encouragement should also be given to the adoption of new, less traumatic approaches where the fish are tagged underwater. Baited tags, which have been used on a range of species (see section 5.3), allow the fish to ingest a tag voluntarily without being caught. This could clearly be advantageous in many situations, although there is limited control over the size, or even species, of fish that is caught and an individual fish may ingest more than one tag. Tagging underwater may eventually become a routine way of avoiding the problems of catching fish with closed swimbladders, although the cost of deploying a team of scuba divers makes it impractical at present. One commercial company is, however, developing an automated device capable of withstanding depths of 1000 m and able to automatically tag large numbers of fish of different types and sizes. If successful, such a device would revolutionise the whole process of tagging fish at sea.
18.104.22.168 Design and attachment of external tags
Weight, which is an important consideration in the design of electronic tags, is discussed more fully in Chapter 7. The main design aim is to minimise the ratio of tag weight to fish weight by reducing the weight of the tag in water. In many cases it is possible to increase the volume so that the tag becomes neutrally buoyant and imposes no extra weight on the fish. Slight positive buoyancy may actually be advantageous, provided the increase in tag volume does not result in excessive drag.
For external tags, drag, which is a function of shape, is generally more important than weight and should be minimised wherever possible. Although some progress has recently been made with hydrodynamic designs of pop-up tags for use with bluefin tuna (Block et al. 1998, Lutcavage 1999), there have been very few similar studies with other externally attached tags. Systematic studies are therefore urgently needed to devise the most appropriate hydrodynamic shape for the tag and, perhaps more importantly, the best position of attachment on the fish. Sensor design and attachment must be included as an integral part of this programme, which requires assistance from hydrodynamicists and access to flumes for work with swimming fish. Some idea of the improvements that can be expected from a programme of this type can be gained from the work done with penguins e.g. (Wilson & Culik, 1962; Bannasch et al., 1994) and turtles (Watson & Granger, 1998). Significant advances have been made by matching the shape of the tag to the morphology of the animal and this approach should be adopted with fish. One possibility worth considering would be to design a blister shaped tag that would be equally streamlined in all directions and could be used with flatfish.
Fish with external tags may be more prone to predation than untagged fish and the shape and colour of the tag may influence the risk of predation. Some species of fish use sexually selected traits (for example coloured or swollen body parts) as signals during mating rituals, which could - theoretically be confused by the presence of an external tag. Both subjects should be studied and tag designs modified in the light of findings.
Tag loss is a common problem with electronic tags and there is a clear need to develop more permanent methods of tag attachment, particularly for DSTs, which potentially have a life of several years. The development needs to be done in conjunction with the hydrodynamic investigations identified above and with full consideration of welfare implications. The problem of tag loss is unlikely to be solved completely, however, and an alternative solution may be useful in tracking studies where the fish may not move for long periods. Stationary tag signals are difficult to interpret in this situation and it would be useful to be able to distinguish between a tag that is still attached to a live and a tag that has fallen off, or is attached to a dead fish. A tag that could differentiate between these situations – with an internal accelerometer or other sensor – would be a useful development.
22.214.171.124 Swimming performance and behaviour
In addition to developing new attachment procedures there is a need to develop challenge tests to evaluate how tags modify the normal behaviour and responsiveness of the fish. To date the most commonly used challenge test is the comparative swimming trial. This has been used successfully to evaluate various attachment techniques, using both critical swimming speed and fatigue trials (Beddow & McKinley 1998, 1999; Peake et al. 1997). There is clearly scope for significant development in this area and one approach will be to use physiological sensors to record the recovery of fish from tagging as a step in developing standard procedures for commonly tagged species. Tags that measure muscle activity (EMG) and heart rate are commercially available and are already suitable for this purpose. Existing techniques are probably not suitable for long-term measurements in the open sea, however, without significant further development and one of the main challenges for the future will be to devise ways of recording ‘natural’ behaviour and physiology without resorting to invasive surgery.
As discussed in more detail in Chapter 7, concern is growing about the welfare of experimental animals. Experimental procedures are strictly regulated and authorities in most countries are increasingly scrutinising the number of experimental animals used in individual studies. In this context, electronic tags – particularly data storage tags - have the great advantage over simple identity tags that much more information can be gained from fewer fish over much longer periods. Concomitantly, it becomes more important to demonstrate that results from a relatively small number of tags are representative of the whole population. One way of doing this is to use large numbers of identity tags in parallel with the electronic tags. Another very effective approach is to predict the behaviour of the population from the electronic tag observations and test the prediction by independent means. This method has been used to demonstrate the importance of selective tidal stream transport in the life cycle of plaice (Pleuronectes platessa) in the Southern North Sea and eastern English Channel (Harden Jones, 1979; Arnold & Metcalfe, 1998). The phenomenon was first demonstrated with a small number of acoustically tagged fish; its importance to the population at large was confirmed by a series of comparative fishing experiments with large midwater trawls.
5.7.3. Engineering developments
Major advances in the research areas identified in section 5.7.1 will depend to a large extent on improvements in the design and performance of electronic tags. The most important of these are telemetry tags that transmit more data over longer ranges and data storage tags that can record more information and store it for longer periods. Individual coding and remote, fishery-independent data retrieval are also becoming increasingly important. Improvements in telemetry may come from smaller, more powerful tags. More sophisticated retrieval of data from noisy backgrounds may, however, offer a more effective solution by avoiding the need to increase the transmitting power of the tag. A significant reduction in the cost of tags, particularly data storage tags, would provide a major impetus to the use electronic tags and would see them used to solve a wider range of fisheries and ecological problems. Cheapness, however, conflicts with the need for more sophisticated and smaller devices. Technological factors that will affect the development of better and cheaper tags are considered in this section.
126.96.36.199 Tag performance
The development of significantly smaller tags is an almost universal requirement, although small tags currently have some disadvantages, such as reduced life and increased weight in water. They may also be harder to find when the fish is recaptured. Smaller tags are, however, needed for use with smaller fish, especially juveniles, and smaller circuits are needed to allow greater sophistication. Successful development of smaller tags will depend primarily on the availability of smaller electronic components, batteries and sensors. At present most microsystems are built from a large number of components, none of which are tailor-made for the needs of wildlife telemetry. Integrated circuits may perform more functions than are actually required in the specific tag application, and consume more current than necessary thus reducing battery life. One solution may be to develop custom-built integrated circuits, although costs are high because of the need to manufacture these devices in sufficiently large quantities to justify development. Another option would be to use custom-built silicon chips for whole tags, although with continuing advances in microcontroller technology this approach is not likely to be cost effective. A prime requirement is to use small batteries and to do this it is necessary to minimise power consumption. Greatest reductions in power consumption are likely to be achieved by the use of quick response sensors that can be switched on for short periods only and can be sampled within a few milliseconds of being switched on.
(b) Life and memory size
Telemetry tags often only require a relatively short life measured in days rather than months. Most biological cycles are seasonal, however, and the majority of data storage tags therefore need to be able to record several items of information for at least a year and store the data for several years to maximise the chances of capture and tag recovery. The use of flash memory, with an expected life of 10-20 years, overcomes the need to provide power for long-term data storage but, as more sensors are included, tags will need more memory. Certain applications, such as the investigation of spawning site fidelity will require tags that record data for several years in succession and tags of this capacity will be needed for many applications with large pelagic species that range extensively through the oceans. Data management will also be important to ensure that best use is made of the available memory, either through data compression or intelligent data recording (e.g. by not recording new data until a sensor reading changes significantly).
Tag size and life is currently determined largely by the size of batteries, which will continue to be a limiting factor for the foreseeable future. Silver oxide cells offer a number of important advantages, such as the ability to deliver a high peak current (8 mA) for a short period from a small cell. Lithium cells are unable to do this and have other technical disadvantages, such as a tendency to passivation, which often leads to premature tag failure. They do, however, operate over a much wider temperature range (-30° to >100° C) than silver oxide cells (-2° to 65° C) and most battery development is now devoted to lithium cells.
188.8.131.52 Tracking systems
(a) Short-range systems
Fixed arrays of hydrophones are currently used to investigate the movements of fish or shellfish static in the vicinity fishing gear (Løkkeborg, 199?; Løkkeborg et al., 199?) and similar techniques have been used to study the effects of dams and barrages on the passage of fish in rivers. The simple systems used to date have depended on three or four hydrophones anchored several hundred metres apart in a triangular pattern with a maximum effective range of about 1000m. Early systems depended on electric cables to bring data ashore. More recently data has been sent by radio telemetry to the shore or to a research vessel and this development has allowed these systems to be used in the open sea. Fish are tagged with acoustic transmitters and their position estimated from the time of arrival of the sound pulses at each of the hydrophones. The depth of the fish is measured with a pressure sensitive telemetry tag. Traditionally each tag has worked on a different frequency and up to 10 tagged fish have been kept under surveillance at any one time. Recently new systems have become available that use binary codes – pseudo random (PN) numbers – to code the tag signals (Cote et al., 1998). This development allows the arrival time of the signal to be measured much more accurately and also enhances the signal to noise ratio. The resulting coding gain provides increased range, allows the tags to operate in noisier environments and can track up to 212 fish simultaneously on a single acoustic frequency. There is no restriction on the number of hydrophones and the operating area can be readily increased to match the type of investigation to be undertaken. These developments are to be encouraged because they will open up the possibility of investigating predator-prey behaviour in the open sea in a cost-effective way.
(b) Long-range systems
Fixed arrays of acoustic listening stations on the seabed have been used with great success in the North Atlantic to monitor ocean currents by recording the tracks of neutrally buoyant SOFAR floats. A similar system would be highly desirable for use with highly migratory oceanic fish. Unfortunately, the SOFAR system operates at low frequencies and the transmitters are far too large to be used on any fish. The receivers of the RAFOS system, which operates on the same principle, but in reverse with fixed transmitters and mobile receivers, are also too large for use with fish.
Fisheries research vessels undertaking regular trawl surveys or acoustic surveys might be used as mobile listening stations. Sonars could be used to search for acoustically tagged fish and it might be possible to recover stored data using a sonar or hydrophone. The concept could be tested with single frequency tags designed merely to record the presence or absence of fish. Even with this simple configuration, however, it would be necessary to released very large numbers of tagged fish, given the depth of the sea, the power of existing acoustic tags, the range of existing sonars and the limited width of search swath that could be achieved.
184.108.40.206 Physical sensors
Temperature: Thermistor technology is well developed and small beads are readily available. Calibration and stability are not a problem. The main difficulty is to avoid spuriously high readings when the tag is heated by sunlight and the thermistor is mounted inside the tag. The problem is probably soluble using a miniature thermistor bead coupled as closely as possible to the water.
Pressure: small electronic sensors are available that are capable of measuring pressure at depths in excess of 3000 m. They are suitable for use with most adult fish but smaller devices are needed for use with juveniles. Lower prices would also be desirable. Technically there are a number of problems, including zero drift, undesirable variation in sensitivity between individual devices and variation in output (offset voltage) with temperature, all of which create problems for accurate calibration.
Light: a number of data storage tags incorporate sensors that measure ambient light intensity and are used to estimate geographical position. Some tags use large area (5-10 mm²) silicon diodes, either singly or joined together; others use a point source diode and a separate collector with a light pipe to focus the light on the diode. These devices are fairly sensitive and can detect light down to several hundred metres in clear oceanic water. They can satisfactorily measure light levels around dawn and dusk when the fish is swimming near the surface. They are not sensitive enough, however, for fish that dive to greater depths at dawn and dusk, or swim consistently well below the surface. Greater sensitivity is required for these applications and to avoid excessive power consumption switched mode operation is desirable. This requires the development of high input-impedance amplifiers with fast settling times to cope with the behaviour of the sensor at low light levels. The frequency response of existing semiconductor sensors is also not ideal, peaking as it does at approximately 900 nm. Devices with maximum sensitivity in the range 450-550 nm would provide greater sensitivity for most marine applications, except turbid coastal waters, whose peak transmission may be as high as 600 nm.
Salinity: One available DST on the market can measure salinity. These tags, which have been used to study the movements of adult salmonids returning to spawn in Icelandic rivers (Sturlaugsson, 199?), identify whether the fish is still at sea or has entered estuarine waters. Much greater sensitivity is needed before it will be possible to use DSTs to identify the type of water in which a fish is swimming in the open ocean.
Tilt angle: One commercially available DST can measure tilt angles in the range ±40° from horizontal with a resolution of <5°, using a mechanical sensor. These tags have been used in the Barents Sea to investigate the attitude of free-swimming cod in relation to vertical migration and target strength (TS). Sampling rates are rather low, however, in relation to the problem and a sampling interval of the order of 1 s would be desirable for this type of research. It would also be desirable to use an electronic sensor with an increased angular range and a resolution of <1°. To achieve this resolution in practice and make reliable measurements, however, it would be necessary to develop a much better system of tag attachment and this is a significant challenge for the biologists.
Speed and acceleration: direct measurement of swimming speed is currently difficult and most devices (e.g. turbines, hot films, electromagnetic sensors, Doppler shift instruments) that have been used are either too large, too vulnerable, too unstable, or consume too much power, to be of practical use. An alternative approach might be to measure speed indirectly through tail-beat frequency, although further research is needed to fully elucidate the relationship between the two quantities in many species of fish. Another approach could be to measure acceleration, which, if sufficiently accurate, might open up the prospect of reconstructing through-water movements of the fish by inertial navigation techniques. Small low-power accelerometers would, however, be required and there would probably be difficulties in operating them for sufficiently long periods to sample the movements of the fish adequately.
Activity: for a number of applications, it may be sufficient to record activity rather than speed or acceleration. This approach has already been adopted with spawning Atlantic salmon, in which spawning behaviour was identified with an electrolytic tilt sensor inside the fish (reference needed). A similar approach may be useful in other situations.
Compass heading: miniature magneto-resistive devices are now available that are capable of resolving the compass heading of a fish to <1°, provided the sensor can be kept within a few degree of horizontal. Fish rarely swim in this fashion, however, and an accurate electronic compass requires the output of the compass sensor to be corrected for pitch and roll. Accelerometers that measure pitch and roll are now available in miniature form and an electronic compass would seem to be a realistic development goal in the near future.
Magnetic field sensors: geomagnetic sensors could be useful in helping to determine the geographical position of oceanic fish that migrate large distances to feed and spawn, particularly if combined with light-based or other methods of geolocation. Such devices might be less useful in shelf seas, where fish generally cover shorter distances during migration. Small low-power devices capable of measuring the total intensity of the local magnetic field of the earth would be appropriate, if they existed, and magneto-optical devices might be able to provide a solution. An alternative approach would be to use a sensor capable of measuring magnetic dip and suitable magneto-resistive devices already exist. As with an electronic compass sensor, however, measurement will be complicated by the need either to provide a stable platform or to compensate for movements of the fish in all three dimensions.
Sound: sound recordings require a large amount of memory. A typical CD of classical music, for example, can use 650 MB. Simple calculations show that even with a limited bandwidth (e.g. 10-40 Hz) and 4-bit analogue to digital (A-D) conversion, a miniature recorder is impractical for the present, except possibly for use with very large pelagic species of fish, or sea mammals.
220.127.116.11 Physiological sensors
Surprisingly little physiological research has yet progressed to the stage where it is possible to envisage worthwhile programmes with free-ranging fish in the open sea. In most cases, even with relatively simple subjects (e.g. the relationship between swimming speed and tail-beat frequency) a substantial amount of laboratory research is still an essential pre-requisite. There are a few areas, however, where some progress could be made fairly soon. The most obvious of these is feeding, where the ability to record feeding patterns and rates of food intake in the sea would significantly advance our knowledge of predator-prey relationships and where some work has already been attempted. Cardiac output and EMG are others.
Mechanical feeding sensors: A simple design for a mechanical jaw angle sensor for use with sharks was proposed in the late 1970s but apparently not used. The device consisted of two rods sutured to the upper and lower jaw and connected, respectively, to the case and rotor of a miniature one-turn trimpot (reference). More recently, progress has been made in Denmark with the development of a mechanical probe to detect food intake in free-swimming cod (Lundgren, pers. comm.). The probe consisted of a piezoelectric film encapsulated in a sheath of silicon rubber, which was surgically implanted and attached to the wall of the oesophagus. A device of this type will provide information about when and where fish eat and the frequency of ingestion. Coupled with a strain gauge the device might also provide an estimate of the size of individual food organisms. Further development work is required in this area. An alternative approach is to measure changes in pressure in the buccal cavity with a differential pressure sensor. This technique should require only minimal surgery and might therefore be more appropriate for use with free-ranging fish. Patterns of buccal pressure will clearly vary, however, with the mode of food ingestion and not much is yet known about this subject except for fish that feed by suction.
Physiological sensors: Mechanical sensors may not offer the best long-term solution to recording feeding events and a radically different approach is to measure one or more of a variety of physiological parameters that should change predictably following feeding. Stomach, or visceral, temperature is one possibility, especially in warm-blooded fish. Some encouraging advances have recently been reported with Southern Bluefin Tuna (Gunn et al., 1999) and there may be scope to pursue the idea with other species, including possibly some poikilotherms. Heart rate, blood flow, gut pH, bile colour and blood chemistry are other options; they would also be relevant to studies of bioenergetics, stress and the reactions of fish to changes in the surrounding environment.
Physiological sensors will increasingly be needed to measure the responses of to environmental changes induced by anthropogenic activities as well as performance under natural conditions. Existing techniques, such as EMG and heart rate monitors can already be used in these areas, although there are limitations to the usefulness of both. EMG can be used to measure metabolic expenditure when swimming behaviour dominates but not when the fish is resting during recovery from a stressful event. Heart rate measurements, which can be used during both types of event, do not give a true measure of metabolic activity because of the changes in stroke volume, which occur in many species. A sensor that could record both heart rate and stroke volume and provide a measure of cardiac output would be a big step forward.
In addition to sensors to measure the physiological state of the fish, there is a continuing need to for more sensors to monitor the surrounding environment. Depth, temperature and conductivity sensors have all been incorporated in transmitting or recording tags but information is now needed on the chemical environment as well. Sensors are required to measure quantities such as pH, chlorine and ammonia and these factors must be monitored at the same time as the physical measurements. There is therefore a continuing drive to develop a range of smaller, more accurate devices that can be assembled in small multi-function tags. A variety of devices exist to measure these quantities, including electromagnetic sensors, thin-film electrodes and biosensors. But most have so far only been used under controlled conditions in the laboratory and substantial research programmes are required to transfer the technology for use with that will be allowed to range freely in the wild. Animal welfare considerations will be extremely important in this context.
18.104.22.168 Remote data retrieval
Remote data retrieval is already possible in certain situations. In freshwater, for example, signals from radio tags can be detected by mobile receivers in aircraft or by fixed detectors on the riverbank. Fixed listening stations can in return relay information to the laboratory by radio or telephone. A similar approach can be adopted with acoustic tags if the fish are within a short distance of a hydrophone on a moored or drifting buoy, which converts the acoustic signals into radio signals before transmission.
a) Remote radio telemetry
Satellite data recovery is currently possible with pop-up tags that detach from the fish at pre-programmed times and float to the sea surface (see Section 5.6). Second generation pop-up tags, which will be capable of measuring several physical quantities and will have substantial data storage capacity, are currently undergoing field trials.
The size of satellite tags is unlikely to decrease much in the short-term, however, and pop-up tags are unlikely to be suitable for use with the small to medium sized fish found in European waters until use can be made of cellular phone systems. Rapid advances are currently underway in this field and several worldwide systems are under development, each of which will employ 60-80 satellites. A number of problems need to be solved before the prospect becomes a reality, however. These include miniaturising the phone and the pop-up system and providing enough power to transmit to the satellite against a background of increasing radio noise, especially in heavily populated areas. The service provider will have to agree to this use of the network. There may also be problems of data corruption when information is transferred between satellites and possible a lack of cover over the open ocean at some times. Developing a miniature pop-up system is likely to be difficult, as is provision of sufficient battery power. Phone miniaturisation may not be a problem if manufacturers develop a single chip for the purpose in response to the very high demand for cellular phones.
b) Remote acoustic telemetry
Recovery of data by an acoustic link demands a lot of power, is slow and is limited to short range. It is also susceptible to multi-path reflections, which can corrupt the data. Acoustic data recovery is thus best suited to situations where fish regularly and reliably return to a known position or remain within the vicinity of a sonabuoy for a sufficiently long period. Communication history acoustic transponding tags (CHAT tags, see Section 5.2) tags are beginning to show interesting results. Recovery of large amounts of data from data storage tags may be possible via sonabuoys or static listening stations, if fish can be induced to stay close to the hydrophone for as long as it takes to download the data. Alternatively, it may be possible to recover the data in limited (say 100k) blocks by sending a command signal to the tag by a low power radio link when the fish is in the immediate vicinity of the hydrophone.
5.7.4 Costs of production and sustaining development
The wildlife telemetry market is small compared to the mobile telephone market, for example, and has not benefited from mass production. High prices of existing electronic tags are preventing realisation of the full potential of the technology and ways need to be found of reducing costs. This may be possible in some areas such as aquaculture, where mass markets may become feasible. There are proposals, for example, in Scandinavia that consideration should be given to individually marking all cultivated fish to provide a quality assurance system for the aquaculture industry, and to facilitate the identification of fish that escape to the wild. Electronic tags would be ideal for this purpose. The aquaculture industry could also benefit from the use of electronic tags to monitor the health and condition of stocks, without the need for handling or otherwise interfering with the fish. Similar possibilities may also exist in relation to cheap unsophisticated tags capable of worldwide use for a range of very simple applications. This approach will, however, not satisfy most of the research objectives defined in section 5.7.1, for which advanced devices are needed. Thus, whilst mass markets may become possible for some applications, development costs are likely to remain very high and continued public funding and pre-market investment will be essential to ensure rapid and sustained development of new technology.
Due to the limitations of conventional tagging methods, electronic tags will play an increasing role in fishery science and management in the next century. In the following section the requirements and recommendations for maintaining progress in this field are identified.