Fish associated with fish aggregation devices off the Canary Islands ( Central-East Atlantic ) *

The tendency of fish to aggregate around floating objects has been known for some time (Mortensen, 1917; Uda, 1933; Kojima, 1956; Galea, 1961; Hunter, 1968). The first commercial fish aggregation devices (FADs) were deployed in the waters of the Philippines in the early 1970s (Kihara, 1981) to attract Thunnus albacares (Bonnaterre, 1788). Due to the success of FADs in aggregating fish, these have come to play an important role in the commercial, subsistence, and recreational fisheries of all the tropical and sub-tropical oceans of the world (Pollard and Matthews, 1985; Frusher, 1986; Wilkins and Goodwin, 1989; Biais and Taquet, 1990; Holland et al., 1990; Buckley and Miller, 1994; Friedlander et al., 1994; Higashi, 1994; Kitamado and Kataoka, 1996). This fishing method has most specifically been used in large scale commercial tuna fisheries over the last three decades and has given a higher yield of fish (Greenblatt, 1979; Fonteneau, 1992). The method yields an average catch of 40 tonnes of tuna per fishing operation around flotsam in comparison with 20 tonnes for free schools (Fonteneau and Hallier, 1993). Sometimes up to 200 tonnes of tuna have been harvested in a single fishing operation (Sacchi, 1986). The first studies designed to ascertain the mechanisms leading to aggregation under floating objects were carried out by Gooding and Magnuson (1967) and Hunter and Mitchell (1967). Subsequent to SCI. MAR., 63 (3-4): 191-198 SCIENTIA MARINA 1999


INTRODUCTION
The tendency of fish to aggregate around floating objects has been known for some time (Mortensen, 1917;Uda, 1933;Kojima, 1956;Galea, 1961;Hunter, 1968).The first commercial fish aggregation devices (FADs) were deployed in the waters of the Philippines in the early 1970s (Kihara, 1981) to attract Thunnus albacares (Bonnaterre, 1788).Due to the success of FADs in aggregating fish, these have come to play an important role in the commercial, subsistence, and recreational fisheries of all the tropical and sub-tropical oceans of the world (Pollard and Matthews, 1985;Frusher, 1986;Wilkins and Goodwin, 1989;Biais and Taquet, 1990;Hol-land et al., 1990;Buckley and Miller, 1994;Friedlander et al., 1994;Higashi, 1994;Kitamado and Kataoka, 1996).
This fishing method has most specifically been used in large scale commercial tuna fisheries over the last three decades and has given a higher yield of fish (Greenblatt, 1979;Fonteneau, 1992).The method yields an average catch of 40 tonnes of tuna per fishing operation around flotsam in comparison with 20 tonnes for free schools (Fonteneau and Hallier, 1993).Sometimes up to 200 tonnes of tuna have been harvested in a single fishing operation (Sacchi, 1986).
Several hypotheses have been suggested to explain aggregation under floating objects (Kingsford, 1993).Amongst many other factors, the most appealing are: Fish congregate around flotsam looking for refuge from predators (Hunter and Mitchell, 1968;Feigebaum et al., 1989).Fish may aggregate because more food is available under flotsam (Gooding and Magnuson, 1967).The disturbance produced by the flotsam in the uniformity of the ocean may be a reason for the attraction (Holland et al., 1990;Hunter and Mitchell, 1967).However, it seems to be the consequence of many factors which influence fish behavior (Fonteneau, 1992;Hall, 1992;Kingsford, 1993).
Several fish species approach FADs looking for food, and feed on the invertebrate that grow on the FADs (i.e.hydrozoans, cirripedes, amphipods and crabs) (Kojima, 1967;Ida et al., 1967;Rountree, 1990;Massutí and Reñones, 1994).It is thought that large predators like tunas, sharks and dolphin-fish, are attracted by the fish aggregated under the FADs (Wickham et al., 1973, Arenas et al., 1992).However, there are doubts that these species prey on the fish fauna associated with the FADs (Brock, 1985;Cort, 1990;Fonteneau, 1992;Massutí and Reñones, 1994).It is also possible that some species like dolphin-fish use FADs as spawning places, due to the fact that drifting objects indicate the presence of oceanic currents where eggs and larvae may drift more efficiently towards better nursery areas.
Studies carried out on the fish communities associated with FADs off the West coast of Africa are few, and deal mostly with the large scale tuna fisheries under flotsam (Wood, 1989;Fonteneau, 1992;Kwei and Bannerman, 1993).In the Canary Island waters, there is no traditional commercial fishing around flotsam, and there is only one previous study on the fish communities associated with artificial habitats (i.e.FADs and cinder blocks on the bottom) (Bortone et al., 1994).In this paper, we describe the faunal composition and abundance of fishes associated with FADs anchored at the south of the island of Gran Canaria, using census data and information from commercial catches over two years.The aim of this paper is to study the fish community associated with FADs, and the variations in the number of species and biomass aggregated in relation to season, depth of anchoring and immersion time.Observation of the behaviour of the species associated with FADs was also recorded to provide further details on the fish community structure.

MATERIALS AND METHODS
In 1995, 1996 and 1997, 16 Fish Aggregation Devices (FADs) were constructed (four in 1995, ten in 1996, and two in 1997) using blocks of foam of expanded polystyrene (200x100x12.5 cm).These blocks were covered with a black plastic film to protect the "foam" from the sun and damage caused by seagulls.Each unit was netted to hold the flotsam to the anchor.The submerged float volume was increased by adding fragments of old nets (about 1.5 m in length) hanging from the float (Fig. 1).No float unit surpassed 10 Kg in weight.
The anchorage of each FAD was composed of four concrete blocks, each weighing 50 Kg, held together with chains.The floats were tied to the anchorages using synthetic ropes about 20% longer than the depth of the area, to avoid sinking the FAD due to current or tidal drag.The FADs were deployed at depths between 50 and 500 m (Fig. 2).The cost of the FADs ranged between 200 and 370 US $, depending on the depth of the anchorage.
From April until October 1995 and August 1996 to May 1997, observations of fish associated with the FADs were carried out almost fortnightly.The number of visits to the FADs in 1995 and 1996-97 were 26 and 29 respectively.During each of these visits, two divers undertook a visual census of all the fish within the range of visibility from the FADs (visibility around the FADs ranged from 15 to 30 m).Divers recorded the number of fish of each species and their approximate size.The size (total length) recorded was transformed to weight, using the length-weight relationships of each species obtained from the literature (Isidro, 1990;García-Gómez, 1993;Gordo, 1996) and from our own non-published data.During observations, video recordings were also taken to verify the census and to provide records for use in subsequent studies of the behavior and spatial distribution of fish around the FADs.
We carried out a statistical analysis in order to test the following hypotheses: 1) Fish community change with the season of the year, as a consequence of the recruitment processes of the local fish fauna (sensus Rountree, 1990).2) The depth of anchorage (50-100, 120-160 and >300 m) influences the fish community structure, so that we can expect to find fish biomass and the number of species associated to FADs increasing from shallow to deep waters.3) The fish community is expected to increase the diversity of the species as immersion time and maturity increase (sensus Vinogradov, 1983).
We also logged the fish captures carried out 100 m around FADs by artisanal fleet based in nearby fishing ports.This fleet was composed of wooden vessels of 11-16 m length, using live-bait and trawling lines.Fishermen were provided with form sheets where they recorded the daily catches by FAD (all FADs were numbered to facilitate their location on a chart).

Composition by species
Nine families and 15 species of fish were observed under or near the FADs (Tables 1 and 2), eight of which had commercial value in the Canary Islands.Katsuwonus pelamis (Linnaeus, 1758) and Thunnus alalunga (Bonnaterre, 1788) were the most important species for the local market, followed by Coryphaena spp.(only after genetic analysis were we able to isolate C. hippurus Linnaeus, 1758 from C. equiselis Linnaeus, 1758, since visual census did not allow for segregation of both species), Seriola spp., and Pseudocaranx dentex (Bloch and Schneider, 1810).The most frequently observed family was Carangidae.

Aggregated fish biomass
There were significant differences in the fish biomass under FADs by month (ANOVA, F=2.17, P=0.039), with the maximum aggregations in September (Fig. 3).Although the number of species underneath FADs increased significantly with the immersion time (ANOVA, F=2.90, P=0.023; Pearson's correlation r=0.4503,P<0.001, N=53; Fig. 4), the fish biomass aggregated did not (ANOVA, F=0.94, P=0.46).The average biomass found was 8.93 Kg (SD=23.6);however, when only baitfish (juvenile fish) were taken into account, the mean biomass aggregated was 3.38 Kg (SD=4.84),with a maximum of 23.37 Kg.The census where Coryphaena spp. was present offered a mean aggre-194 J.J. CASTRO et al. gated biomass of 53.9 Kg (SD=53.6), with a maximum of 150 Kg.The FADs deployed in shallower waters gave an aggregated biomass which was significantly lower for baitfish than for those deployed in deeper areas (ANOVA, F=3.457, P=0.039; Pearson's correlation r=0.4417,P<0.001, N=53; Fig. 5).The largest number of species registered in any one census of a single FAD was five (mean=2.72;SD=1.31).However, we observed a significant difference in the number of species present as a function of the anchoring depth of the FADs (ANOVA, F=4.484, P=0.016).Generally, the FADs deployed on shallower waters gave a lower number of species aggregated.

Fish behaviour around FADs
Depending on the size of the individuals, the schools of juvenile P. dentex and Trachurus spp.aggregated under the protection of fragments of net hanging from the flotsam (Fig. 1) or deeper, always close to the rope (the juveniles smaller than 5 cms. in length, of both species, frequently established mixed schools).The juvenile P. dentex even aggregated around other fish such as B. carolinensis, when they were close to the FAD.These juveniles, when they reached a body length of between 15 and 20 cms, moved no further than 20 m from the FAD, aggregating around whatever object passed closeby, even the divers, and later returning to the proximity of the FAD when the object moved further than 20 m from the same.Frequently, these individuals swam to deep waters following the anchorage rope, out of the visible range of divers (over 30 m).
The juvenile of Seriola spp.and Schedophilus ovalis were solitary or constituted small loose groups of 2-5 individuals (Fig. 6).Those smaller than 10 cms.did not move away from the refuge of the submerged structure of the float.Juvenile of K. sectator less than 15 cm long took refuge between the hanging nets of the FAD, while larger specimens were observed up to 20 m away.Aggregated juveniles (5-10 cm) and adults (30-50 cm) of S. ovalis remained a few centimetres under the float.
A male and two females of Coryphaena spp were observed and video recorded during courting/ spawning around a FAD.This species was normally observed, aggregated in small groups of under 15 individuals (Fig. 6).

The fouling community
The fouling community of the FADs was composed of algae and invertebrates, especially Cirripedia and Hydrozoans.The most abundant invertebrate was Lepas anatifera (Linné) which began the colonization of the FADs two week after deployment and completed it after four months (colonizing the flotsam, nets and the rope).Between the hanging nets of the FAD, colonized by hydrozoans and algae, the amphipod Caprella acutifrons Latreille and the crab Planes minutus (Linnaeus) were frequently detected.

Fish captures around FADs
Two hundred tons of tuna (K.pelamis and T. alalunga) were caught in the vicinity of one FAD between August and September 1996.These catches had a market value of US$ 323 thousand, thus the cost of this FAD represented only 0.13% of the obtained catch value.During the 1995 and 1996 fishing seasons (from June to October), commercial catches of Katsuwonus pelamis around FADs fluctuated between 200 and 7000 kg per fishing day and per FAD, whilst the average catch was approximately 1000 kg.The catches of Coryphaena spp.fluctuated between 10 and 400 kg per fishing day and per FAD, giving an average catch of 100 kg.

DISCUSSION
During our observations, fish under or near the FADs were generally juveniles belonging to eight species and five families (although we registered nine families and 15 species of fish).Seriola spp., Naucrates ductor and Pseudocaranx dentex were the most common species.
In agreement with Rountree (1990), our observations indicated that seasonally regulated juvenile fish availability, as a consequence of the natural recruitment processes in the region, is apparently the most important factor determining FAD species composition and abundance.Most of the species registered around FADs in the Canary Islands spawn during Spring or at the beginning of Summer (i.e.carangids) (Fischer et al., 1981), so recruitment to the area peaks at the end of the Summer.
However, there are other factors which also seem to have an effect.The abundance of the fauna under FADs was affected by the depth of the anchorage.But, contrary to what we initially expected, the FADs deployed in shallower waters showed a lower biomass and a reduced number of aggregated species.This result may be biased by the character-istics of the shelf of the island (very narrow and abrupt) where deeper anchorage areas are not far enough from the shore when compared with shallower ones.On the other hand, floats undergo a maturation process (Arenas et al., 1992) and it is possible that since older FADs are more completely colonized by a fouling community (algae, hydrozoans and crustaceans, hence available food resources), they can support a higher aggregated biomass, and with increasing maturity, the diversity of species in the communities increases.Nevertheless, the age of FADs (immersion time) had no effect on the aggregated biomass, unlike the number of species aggregated,which increased proportionately in the way we expected.
As pointed out by Hunter andMitchell (1968) andFeigebaum et al. (1989) juvenile fish probably congregate around flotsam looking for the shelter that FADs provide.In general, the fish observed around our FADs were distributed spatially in accordance with their body size, with the smaller individuals to the FAD, taking refuge between the hanging net and the same.It is also possible that fish found local high concentration of available food under flotsam (Gooding and Magnuson, 1967) or around the FADs.Most of the natural drifting devices (i.e.drift algae, limbs, etc) are carried away by currents which come from places where floats are frequently generated (i.e. a river mouth or coastal areas) and introduced into the pelagic environment (Hunter and Mitchell, 1967).It is usually assumed that currents carry away the phytoplankton production from the region of its formation into oligotrophic regions (Vinogradov, 1983).Concentrations of biotic structures are often intensified by oceanographic features, such as fronts (Kingsford, 1993).So, some pelagic fish species could associate drifting floats and the area around them with currents where larvae and juveniles will find suitable food availability during their early life stages.This could explain the observed courting/spawning behaviour of Coryphaena spp.around FADs.We have also found anomalous and significant higher biomass of zooplankton under several anchored FADs as compared to the surrounding area (ANOVA P=0.03; authors unpublished data) which could support the hyphothesis proposed by Gooding and Magnuson.This anomalous concentration of zooplankton may be due to eddies generated by the interaction of floats and ropes of anchorage with the current (in the same way as the current field around a boat in agreement with Lindquist and Pietrafesa (1989)).However, this is an issue which needs confirmation in further studies.
Most of the species observed by divers under the FADs in this study did not occur in sizes large enough to become attractive for fishing, except Coryphaena spp.and possibly N. ductor.However, FADs aggregate baitfish which are thought to attract other species of higher economic value into the general area (Rountree, 1990).
FIG.1.-Fish Aggregation Devices and anchorages used during the study period.FAD was a mattress of foam (200x100x12.5 cm), covered with a black plastic film and a piece of net.The submerged volume was increased with fragments of old nets (about 1.5 m of longitude).The anchorage was composed of four 50 kg concrete blocks.

FADs
FIG. 4. -Variation in the number of fish species with the length of time that the devices were in place (Pearson's correlation r= 0.4503, P<0.001, N=53).
FADs OFF THE CANARY ISLANDS 193 FIG. 2. -Location of the FADs deployed in waters off Gran Canaria (Canary Islands, Spain).

TABLE 1 .
-Fish species associated with Fish Aggregation Devices in waters off Gran Canaria, recorded from diver visual counts from April 1995 to May 1997 (*fish species only recorded through fishing).
FIG. 3. -Changes in abundance in fish biomass over the study period (the figures on top of bars indicated the maximum number of species counted each month).