sign in sign up
<<
Home , Lessepsian migration

Research Article

Mediterranean Marine Science Volume 10/1, 2009, 35-54

Influence of Levantine Artificial Reefs on the assemblage of the surrounding seabed

D. EDELIST and E. SPANIER

Department of Maritime Civilizations and the Leon Recanati, Institute for Maritime Studies, University of Haifa, Mount Carmel, Haifa 31905, Israel e-mail: [email protected]

Abstract

Four (AR) units were deployed at a 20m depth on a flat hard substrate 3 km west of Haifa, Israel and then surveyed for fish for 12 months. AR units supported 20 times the biomass of con- trol quadrates and their enrichment impact was still significant at a radius of 13m away from units. The 13m values were also significantly higher than those of quadrates adjacent to units, suggesting the exis- tence of a halo of relative depletion within the outer enrichment halo. The main species contributing to this pattern was the migrant herbivore . A decrease in grazing resources is thus suggest- ed as an explanation for creation of this halo. The most consistent AR residents were also Lessepsian migrants - Sargocentron rubrum, nocturnal predators which displayed high microhabitat fidelity and a steady increase in density. The 6 species of migrants recorded accounted for 65.3% of the commercially exploitable biomass and 25.2% of the specimens in the AR site. Other constant AR residents were the groupers Epinephelus costae and Epinephelus marginatus, which are rare and commercially important species. Site protection from fishing and storms were found to be of utmost importance, and design and deployment considerations are discussed.

Keywords: Artificial reefs; Mediterranean Levantine basin; Fish assemblage; Lessepsian migration; Influence radius; Production vs. attraction.

Introduction enhancement of the fishing yields. This en- hancement, however, is not to be taken for The magnitude of the world fishery granted, as ARs are assumed to function yield makes the practice of deployment, in a combination of two mechanisms: ag- monitoring and harvesting Artificial Reefs gregation of scattered specimens and sec- (ARs) a subject of active interest globally ondary biomass production through in- (BOHNSACK & SUTHERLAND, 1985). creased survival and growth of juveniles Today ARs are used for diverse applica- (e.g., BOHNSACK & SUTHERLAND, tions even though the principal one remains 1985; PRATT, 1994; BOHNSACK et al.,

Medit. Mar. Sci., 10/1, 2009, 35-54 35 1997; SEAMAN, 2000; JENSEN et al. 2000; ing the AR biomass. Their radii indisputably JENSEN, 2002; OSENBERG et al., 2002). vary with AR size, design, material, loca- An AR may even entirely deplete stocks by tion, depth and distance from natural re- merely concentrating the fishing effort lief – both the supply source of adult set- (POLOVINA, 1989). This conflict has been tlers and potential gene pool (CARR & dubbed the ‘attraction production debate’. HIXON, 1997). TURNER et al. (1969) sug- Yet, there are great variations in ARs and gested leaving 15-18m diameter open spaces the behaviour of different species of fish between AR units. STONE et al. (1979) may vary depending on the locations, oc- noted that an AR placed within 25 m of a casions and conditions of the ARs and thus natural habitat recruited juveniles, and did the predominating mechanism (attrac- not reduce the population of the existing tion or production) varies accordingly (e.g., natural reef. OGAWA (1982) concluded SPANIER, 1996). One of the main goals that for benthic species influence radii of ongoing AR research is to spatially and change from 1-100m and recommended 'a temporally chart these differences in order few meters' as a good choice for distance to gain a deeper understanding of the mech- between AR and natural habitat. For pelag- anisms through which ARs attract and fa- ic he determined this radius stretch- cilitate the production of fish. Recruitment es up to 800m. FRAZER and LINDBERG is a key factor which has to be quantified (1994) proposed a 60m gap between units. in order to study an AR's ability to produce For hard substrate habitats in the south- new individuals. Recruitment is the addi- eastern Mediterranean, SPANIER (2000a) tion of new individuals to populations or has suggested 3m3 of AR for every 1000m2 to successive life-cycle stages within popu- of seabed as optimal AR density. Assum- lations (CALEY et al., 1996). As ARs de- ing three separate 1m3 cubic units per 1000m2, pend greatly on import (both juvenile re- a 10.3m influence radius is to be expected. cruitment and addition of adults) from near- The present study focuses on these close- by existing stocks, a study of their inter-re- range inter-relations between fish and ARs lations with the ecotone is indispensable. in the eastern Mediterranean. As this basin Fish and invertebrates use both natu- is comparatively poor in both nutrients and ral and artificial surfaces for shelter, feed- fishing yields (e.g., SAURNIA, 1973; BER- ing, spawning, energy economy and orien- MAN et al., 1984, AZOV, 1986; HERUT tation (BOHNSACK et al., 1994; CARR et al., 2000), ARs are a subject of great & HIXON, 1997). Their accumulation interest in its waters. Relative scarcity of around ARs is a stupendous outcome of fish is also presumed to result in vacant eco- behavioural . Nevertheless, a great logical niches, which allow species of Indo- portion of the enhanced biomass comes Pacific origin that migrate into the Levan- from materials consumed in forage areas tine basin from the through the outside the AR complex. Depending on Canal (GOLANI, 1998) to establish each species' association with the AR and and develop permanent and considerable its foraging range and behavioural patterns, populations. This phenomenon, called feeding halos are formed around the AR Lessepsian migration, has intensified in re- (BOHNSACK, 1989; CARR & HIXON, cent years and fish of Red Sea origin were 1997; BORTONE et al., 2000; SHENG, observed as far as Sardinia (PAIS et al., 2000). These halos are critical to sustain- 2007). Species from tropical origin are con-

36 Medit. Mar. Sci., 10/1, 2009, 35-54 sidered more competitive than autochtho- nous Atlanto-Mediterranean species (GOLANI, 1998). Thus, the appearance of a Lessepsian migrant in the basin may result in a competitive exclusion of in- digenous species or their displacement to another habitat or depth range (e.g., SPANIER & GALIL, 1991). The present study also demonstrates how small AR units can provide an in situ look at Lessepsian migrants in their vicinity. Previous studies in nearby sites showed AR deployment to concentrate fish in larg- er amounts and diversity than surrounding natural reefs (SPANIER, 2000a; 2000b; SPANIER et al., 1983; 1985a; 1985b; 1989; 1990). The present study set out to exam- ine the nature in which deployment of small Fig. 1: Haifa AR Location map and illustration of the deployment scheme (map adapted from AR units affects the ichthyofauna of the SPANIER et al. 1989, with permission). surrounding seabed.

Materials and Methods suitable properties was located 3km NW Artificial Reef – Location and Structure of the Carmel promontory at 32Æ51'02''N Four 1.2m sided cubical steel reinforced- - 034Æ56'33''E. Twelve 5X5m rope frames concrete structures, weighing 1500kg in wa- were laid in the AR field as survey quadrates: ter were used as ground units. 16 sections Four were set one around each ground unit of 25 cm diameter polyethylene pipes were (unit quadrates). The next four were set fitted into each such structure. Floating alongside them (adjacent quadrates) and units, or FADs (Fish Attracting Devices), the last four 13m (C-T-C) away from units were tethered to the ground units. These (detached quadrates). Two more frames were 1m steel profile cubes, into which 16 were set as controls on similar hard sub- polyethylene pipes were similarly fitted. strate, 500m to the south of the AR field Without the concrete ballast, they supplied on a rocky substrate at the same depth as 160kg of buoyancy. The AR field was com- the AR units (Fig. 1) at 32Æ50'46''N – 034Æ posed of four such double units of a ground 56'27''E. These control quadrates varied unit and a FAD suspended 10m above the morphologically – one represented com- seafloor (Fig.1). Typical seafloor property plex control features with dense porous out- was of the flattest ground with as little com- crops of 0.4-0.8m height, much like those plexity as possible. Distances between units of a complex natural reef. The second had were 25-40m, which enabled the whole 4 plain control features, with few small 0.1m unit field to be monitored within one dive outcrops, representing similar bathymetry and were presumed far enough to avoid to that on which AR units and survey overlapping influence radii. A site with the quadrates were deployed.

Medit. Mar. Sci., 10/1, 2009, 35-54 37 Data collection Length-Weight Tables in FROESE and Fourteen surveys were conducted by PAULY (2006) from the nearest sighting two divers as point count surveys, adapted of the species to Haifa. Shannon's species from SHENG (2000) during 7 sub-seasons: diversity index (H') was calculated for the summer 2004 – summer 2005. Each count quadrates according to SHANNON (1948): lasted 2 minutes per quadrate. One diver H' = ™ (Pi*lnPi), where Pi represents the concentrated on quantifying the bigger proportion of the i'th species. schools of fish and the more abundant For inter-quadrate comparisons, data species, the other on identification of rare was pooled from all surveys for every and cryptic species. After the completion quadrate type. A non-parametric Wilcox- of the visual census, a third diver recorded on signed rank-test was employed the fish and macro-invertebrates on video (WILCOXON, 1945) in order to determine and still photography for later aid in tax- whether differences in abundance, species onomy, comparisons and study. Survey re- richness, biomass and diversity were sig- sults were written on pre-designed PVC nificant between quadrate pairs and a lev- slates in order to save time underwater and el of P<0.05 was determined as significant then transferred to excel sheets for pro- for comparison. cessing. Underwater visibility was meas- ured using both horizontal (at the seafloor) Results and vertical secchi disk depth. Water tem- peratures were measured by a Nitrox Su- Thirty species belonging to 18 families unto Solution gauge. Currents were meas- were observed during surveys (Table 1). ured with an Interocean S4 current meter. Twenty seven of the species were record- Video and still photo records were exam- ed in AR unit quadrates, as opposed to on- ined by ichthyologists and compared with ly 11 and 18 species in the plain and com- data from LYTHGOE and LYTHGOE plex control quadrates respectively (rep- (1971) and GOLANI and DAROM (1999), resenting similar seabed to AR deployment to identify cryptic species and back up in site with no AR unit, and a high-relief nat- situ counts. Fourteen census sorties were ural reef). Unit quadrates also supported executed in a back to back day format. Da- a mean of 85.7 specimens and a mean bio- ta was pooled from each such pair of con- mass of 237g/m2 per survey, whereas plain secutive surveys and then into 7 seasonal and complex control quadrates held a mean data sets. All surveys were executed at of only 15 and 36.4 specimens and a mean the same time of day (between 0900 and biomass of 18 and 68.1g/m2 respectively. 1200). Biomass estimates were shown to The abundance, species richness, Shannon's facilitate approximation of the magnitude diversity index and estimated biomass of of AR fauna (BORTONE et al., 2000) and fish in the site during the 7 seasons are pre- were therefore employed in the present sented in Figures 2a-d. Values peaked dur- study. ing the first summer and autumn in the post deployment phase, then declined during Statistical analysis winter and increased again the following Biomass estimates of fish in the AR site summer. Unit quadrates generally displayed were based on diver records of LT in cm, higher abundance, richness and biomass taken in situ, and a later calculation via values than both control quadrates and ad-

38 Medit. Mar. Sci., 10/1, 2009, 35-54 (continued) d nd feeding habits. Car – Fish, Habits c g/specimen b Level a Total No. of fish Trophic Biomass in Feeding Estimated a Table 1 (Lowe, 1834) 9 A 3.7 587 Car- Cru, some fish a Ruppell 1 R 4.3 288 Car – fish (Delaroche) 7 A 3.4 645.2 Car – Fish, Cru (Valenciennes) 72 A 3.8 380.5 Car/Her (Forssk l) 302 R 3.5 150.7 Car- Inv, Cru, Mol. Steindachner 56 A 3.9 515.3 Car – mostly fish Linneaus 19 A 3.4 103.7 Cru, Mol, Worms Lowe 2 A 3 284 Car – Fish, Mol, Cru (G. Saint Hilaire) 249 A 3.2 63.3 Car - worms, cru (Linneaus) 2966 A 3 21.1 Z (Linneaus) 1 A 3.9 13.3 Z + Inv, Cru (Linneaus) 7 A 3.3 44.3 Car- Fish, Inv Valenciennes 710 C 3 26.2 Z- Cru, some fish Linneaus 12 A 4.2 531.2 Car- Fish, Cru Risso 15 A 3.5 52.6 Car- Inv, fish (Linneaus) 22 A 3.8 117.7 Car – Fish, Inv (Forssk l) 655 R 3.3 42.2 Z (Linneaus) 34 A 4.2 34 Car- Inv (Linneaus) 27 A 3 36.4 Car- Inv BEN-TUVIA & Golani 2 R 3.3 77.6 Car - Inv, Cru Spicara maena Spicara smaris Mullus surmuletus djedaba Apogon imberbis Serranus cabrilla Serranus scriba Epinephelus marginatus Sargocentron rubrum Epinephelus costae Muraena helena Gymnothorax unicolor Diplodus cervinus Diplodus vulgaris Pagellus acarne Pagrus coeruleostictus Sardinella aurita Fistularia commersonii Chromis chromis f f List of species recorded in The AR and FAD site their total abundance, including origin, trophic level, estimated biomass a Pomacentridae Mediterranean Damsel fish Centracanthidae Blotched picarel Centracanthidae Picarel Mullidae Brownband Mullidae Striped red mullet Carangidae Shrimp's Scade Apogonidae Cardinal fish SerranidaeSerranidae Comber Painted comber Serranidae Dusky grouper HolocentridaeSerranidae Red Squirrelfish Golden grouper FamilyCommon Name Scientific Name observed Origin Muraenidae Clupeidae Round Sardinella Muraenidae Brown moray SparidaeSparidaeSparidae Zebra sea bream Sparidae Two-banded sea bream Axillary sea bream Blue-spotted sea bream Fistulariidae Cornetfish

Medit. Mar. Sci., 10/1, 2009, 35-54 39 d Her-Car - Alg, Habits c = Carnivore; Her = Herbivore; g/specimen b Level a A 5-10 Car – Inv. Total No. of fish Trophic Biomass in Feeding Estimated (Linneaus) 44 A 3.1 5.1 Car – Mol, Cru, Inv a Fraser-Bruner 25 R 2.8 46.6 Car- Inv (Linneaus) 57 A 2.9 64.2 Her/Car Table 1 (continued) (Risso) 44 A 3.5 9 – Mol, Inv, Cru Car (Linneaus) 562 A 3.5 26.3 Car – Mol, Cru Steindachner 2 A 3.1 6.7 Inv, Cru (Linneaus) 1 A 3.5 31 Car – Mol, Inv, Cru Forssk l 394 R 2 124.9 Her – Alg (Rüppell) 129 R 2 171 Her – Alg (Linneaus) 1547 A 3.2 14.1 Z Parablennius sp. Sparisoma cretenese Thalassoma pavo Gobius bucchichi Symphodus tinca Symphodus roissali Symphodus mediterraneus Coris julis Siganus luridus Siganus rivulatus f f Filefish/ Reticulated Feeding Habits, as per GOLANI and DAROM (1999), LYTHGOE and LYTHGOE (1971) and FROESE and PAULY (2006): Z = Zooplanktivore; Car Trophic level Data – as per FROESE and PAULY (2005) Observed exclusively around the FADs Origin: A = Atlantic-Mediterranean; R Red Sea (Indo-Pacific); C Cosmopolitan Biomass was estimated using L-W Tables from FROESE and PAULY (2006) Observed around the FADs and AR units Siganidae Blennidae Unidentified Blenny Scaridae Parrot fish Labridae Ornate wrasse Labridae Peacock Wrasse Labridae Five-spotted wrasse FamilyCommon Name Scientific Name observed Origin Labridae Rainbow wrasse Monacanthidae leatherjacket Labridae Axillary Wrasse Siganidae Marbled spinefootf Gobiidae Bucchich's goby Inv.=Invertebrates; Mol.=Mollusks; Cru=Crustaceans; Alg = Z = Zooplanktivore; Car Carnivore; Her Herbivore; Inv.=Invertebrates; Mol.=Mollusks; Cru=Crustaceans; Alg Algae a b c d e f

40 Medit. Mar. Sci., 10/1, 2009, 35-54 Fig. 2a-d: Mean Fish Abundance, Species Richness, Shannon's Diversity Index and Biomass ± S.E. during 12 months of census (divided into 7 sub-seasons) in the Haifa AR Site quadrates: Unit quadrate, Adjacent quadrate, Detached quadrate, Complex control, Plain control.

Medit. Mar. Sci., 10/1, 2009, 35-54 41 jacent and detached quadrates, with Shan- aspros was, much like S. rubrum, closely as- non's diversity index values showing a greater sociated with the AR units, however only variability (Fig. 2c). AR ecotone (adjacent single specimens or couples were record- and detached quadrates) normally displayed ed and their numbers did not increase with slightly higher values than the plain con- time. F. commersonii was recorded hover- trol, although complex control values ex- ing in close proximity and parallel to quadrate ceeded those of the ecotone (Fig. 2). lines, presumably mimicking them for cam- Thirteen species were of commercial ouflage. The detached quadrates assem- importance (according to SNOVSKY and blage, also showing a large proportion of SHAPIRO 2003), thirteen were piscivores migrants (Fig. 3), was comprised mostly of and only two were obligatory herbivores S. rivulatus specimens, usually observed in –Siganus rivulatus Forsska l, and S. luridus motion, displaying foraging behavior. S. (Rüppell), both Lessepsian migrants. Seven rivulatus was also the only migrant species of the 30 species observed were of Red Sea observed in control quadrates. origin. Lessepsian migrant percentage, cal- AR unit quadrates were compared with culated for the 6 benthic species recorded the complex control, in order to determine in surveys, is presented in Figure 3. A sep- whether the AR provides a superior habi- arate "inner unit" data series is presented, tat to that of a natural fully- developed reef to underline the massive presence of the Red (Table 2). They were indeed found to have squirrelfish Sargocentron rubrum (Forsska l) significantly higher diversity and biomass in the inner AR assemblage. Along with (Wilcoxon, P=0.047) than complex control another migrant – the Filefish, Stephanolepis (Wilcoxon P=0.006). Unit quadrates also diaspros Fraser-Bruner, Red-Sea species carried almost 3 times the abundance and accounted for the majority of individuals 50% more species than complex control observed inside AR units (Fig. 3). S. di- quadrates, however differences for abun-

Fig. 3: Mean Lessepsian migrants percentage during 12 months of census in Haifa AR site in the study quadrates: Inner unit, Unit quadrate, Adjacent quadrate, Plain control, Complex control. Detached quadrate.

42 Medit. Mar. Sci., 10/1, 2009, 35-54 ndance), mean no. of species/quadrate (Richness), and PAULI, 2006). * Significant effects (P < 0.05) are indicated. Table 2 5.201.17 3.161.32 0.34 10 1.762.70 100.70 39.90 8 1.570.33 42.90 0.45 10.88 10 0.793.55 100.78 13.00 8 2.650.58 12.50 0.57 6.38 22 0.76 22 21.20 13 20.30 9.77 Richness 7.64 2.87 56 32.36 399 1.1506 0.125 DiversityBiomass 1.36Richness 5.17Diversity 0.39Biomass 2.41 4.08 56 1.03Richness 31 2.12 1.15 31.82Diversity 0.41 10.88Biomass 25 1.28 4.08 429 25 1.03 87 20.00 1.6736 16 2.12 1.15 0.0471* 20.20 0.41 2.5216 130 15.56 0.0058* 25 1.28 125 1.8412 25 51 0.0328* 26.46 1.9941 16 0.0231* 27.26 2.9713 467 19.25 0.0015* 447 1.3215 127 0.0932 1.732 0.0416* 2.9615 0.0015* Quadrate pair comparison: Wilcoxon two-sample signed-rank test scores and summary statistics for mean no. of fish/quadrate (Abu Quadrate PairUnit Quadrate -Complex Control Variable Abundance Mean 85.70 St. Dev. 61.78 29.20 N 31 42.98 Mean Score 21.10 8 Statistic z 126 15.75 p 1.166 0.1218 Detached Quadrate - AbundancePlain Control 17.70Detached Quadrate - 20.78 AbundanceAdjacent Quadrate 10.60 16 17.70 19.28 13.69 20.78 16.11 8 81 16 21.30 10.13 1.1411 18.66 0.1269 13 136 10.50 2.5555 0.0053* Shannon's Diversity Index (H') (Diversity) and estimated mean biomass of fish, drawn from Species Length-Weight tables (FROESE

Medit. Mar. Sci., 10/1, 2009, 35-54 43 dance and richness were not statistically blotch grouper, Epinephelus costae Stein- significant (Wilcoxon, P=0.121 and 0.125 dachner and the Dusky Grouper, Epinephelus respectively). In order to determine at- marginatus Bloch and Schneider. Figure 4 traction radius, detached quadrates were displays the patterns exhibited by the most compared to plain control, and then to ad- common dominant large predators, i.e. the jacent quadrates (Table 2). Richness, di- groupers and squirrelfish, in the AR unit versity and biomass values were significantly quadrates throughout the study period, con- (Wilcoxon, P=0.03, 0.023 and 0.001 re- trasted with water temperature. Groupers spectively) higher in detached quadrates lurked mostly in the lower rows of pipes in- than in plain control; but surprisingly, abun- side units or in the crevices formed under dance diversity and biomass (Wilcoxon, AR units before scouring closed them. Like P=0.005, 0.042 and 0.001 respectively) were squirrelfish, none were viewed over flat sub- also significantly higher in detached quadrates strate and were altogether absent from non- than adjacent ones (Table 2). During sum- AR quadrates. Groupers reached a maxi- mer, the great number of fish shoaling mum of 12 fairly large individuals (30-50cm around units occasionally ‘spilled’ into ad- LT) inhabiting AR units in the early winter jacent quadrates and yet in 19 out of a to- of 2005. By this stage, specimens could be tal of 25 observations in which both quadrates individually identified by their size, color were surveyed, more specimens were record- patterns and choice of microhabitat with- ed in detached quadrates than adjacent in units. This changed when in the spring ones. of 2005, the numbers of groupers in the site The most common fish in surveys were dropped within 48 hours from nine speci- the Mediterranean damselfish, Chromis mens in 28.2.05 to none in 2.3.05 (Fig. 4). chromis (Linneaus) (Table 1), most of which E. marginatus was never sighted again af- were observed shoaling in great proximity ter this date and all E. costae observed here- to AR units. The Rainbow wrasse Coris after were relatively small (max. 25cm LT). julis (Linneaus) and the Ornate wrasse Tha- In contrast to groupers, the number of S. lassoma pavo (Linneaus) were also very rubrum, (302 observations – Table 1) grew common in surveys, however they were not steadily with time throughout the study pe- as tightly grouped around AR units as C. riod (Fig. 4). chromis were. Other common fish (Table Damselfish, wrasse and ju- 1) included the Two-banded sea bream, veniles were recorded almost exclusively in Diplodus vulgaris (G. Saint Hilaire), and the warm seasons (Fig. 5), following their the Blue-spotted sea bream, Pagrus spawning period in spring and summer. Ju- coeruleostictus (Valenciennes), as well as veniles of C. chromis were observed exclu- the Painted comber, Serranus scriba (Lin- sively within 1m of the structures. neaus). All Sparids and Serranids were The four FADs showed little resilience common around AR units in summer and to winter storms. The FADs caused ground autumn but disappeared completely in win- units to scour, flip or altogether break in- ter and spring. to pieces. Therefore, the FAD units were The larger predators found in unit removed after only five months at sea. crevices were the Brown moray, Gymnothorax The species composition of the FADs unicolour (Delaroche), the Mediterranean in these five months is displayed in Fig- moray, Muraena Helena Linneaus, the Gold ure 6. In summer they were occupied by

44 Medit. Mar. Sci., 10/1, 2009, 35-54 Fig. 4: Total abundance of AR residents - Epinephelus marginatus and E. costae as well as Sargocentron rubrum vs. water temperatures.

Fig. 5: Juveniles/females recorded in censuses. Siganus rivulatus and Chromis chromis figures refer to juveniles (easily distinguishable by size and colour) whereas females of the wrasses Tha- lassoma pavo and Coris julis may highly resemble juveniles in both size and morphology.

only one species – The Lessepsian migrant aurita Valenciennes, as well as C. chromis, , Alepes djedaba (Forsska l). which were observed ascending from bot- Within a couple of months it was joined by tom units. When winter storms first hit the large shoals of Round Sardinella, Sardinella site, FADs were lowered, as a protective

Medit. Mar. Sci., 10/1, 2009, 35-54 45 measure, to only 5m above the seafloor and <3m, except during summer, when the in- this apparently enabled species more close- crease in abundance and richness caused ly associated with the benthos, such as S. spillover into adjacent quadrates. Detached rivulatus and C. julis to ascend to FADs quadrates, located 13m away from the units (Fig. 6). The enlarged surface area of FADs showed significantly greater richness, di- facilitated settlement of a thick epybiota, versity and biomass than the flat, plain con- mostly of the Pearl oyster, Pinctada radia- trol site (Table 2). Thus, AR induced en- ta (Leach). No fish were observed feed- richment was still discernible at this dis- ing on it and none were viewed under or tance. Nevertheless, these detached quadrates inside the FADs. displayed significantly higher values than those of quadrates adjacent to AR units. Discussion This finding suggests the existence of a ha- lo of relative depletion within the outer en- Artificial Reef unit deployment was richment halo. The prominent species ex- found to affect the fish assemblage in dif- hibiting behaviour which fitted this pattern ferent intensities and radii for different was the Lessepsian herbivore S. rivulatus. species. AR units themselves provided habi- Unlike its congener S. luridus, more fre- tats for several species that were rare or ab- quently observed in AR unit quadrates, sent from other quadrates. These were most- S. rivulatus was observed in schools of 5-20 ly cryptic species and/or nocturnal carni- specimens, grazing farther away from units. vores, but also juveniles of reef-associated It is therefore suggested that upon sensing species. Units were shown to significantly unit presence directly (by sight, lateral line, raise the ecotone carrying capacity for fish- and/or smell) these fish elect to approach es. Although capacity exceeded that of a to within 1-2m and benefit from shoaling flat control site, it did not match that of the advantages (e.g. PITCHER et al. 1982). complex control (Fig. 1 and Table 2). This It is further speculated that they used AR means absolute enhancement occurred at units as navigational benchmarks (e.g.

Fig. 6: Species composition and density recorded for the Fish Aggregating Devices in Haifa AR site before FADs were lost in winter storms - Coris julis, Chromis chromis, Sardinella aurita, Alepes djedaba, Siganus rivulatus.

46 Medit. Mar. Sci., 10/1, 2009, 35-54 BRAITHWHAITE, 1998) in their grazing groupers' prey items (Fig. 2a). Why then excursions into the ecotone, when near AR had vertical migration not occurred earli- resources were exhausted. er? Enquiries among local fishermen sug- The attractive properties of AR units gested an alternative explanation. The site towards the 20m depth Eastern Mediter- was familiar to SCUBA divers and was vis- ranean fauna were demonstrated at first by ited quite often by spear fishermen. It is a post deployment overshoot, in the sum- highly probable then that the disappear- mer of 2004. This type of overshoot has ance of its larger inhabitants was indeed been described in previous AR studies fishing related. Grouper absence from con- (BOHNSACK and SUTHERLAND, 1985; trol quadrates stressed these overexploit- MORENO et al., 1994). It is thought to rep- ed species' demand for additional relief and resent an early curiosity of fish towards the complexity. Nevertheless, unless protected newly established habitat, prior to reach- from fishing, such habitat erection is futile. ing a seasonal, dynamic equilibrium. As recommended by PITCHER and In the ensuing winter some species de- SEAMAN (2000), AR deployment in no- clined in numbers while others disappeared take zones can and should play a positive altogether. The following summer saw most role in future restoration and fishery man- of them return, although in slightly lower agement programs. This is also the case in numbers. BOMBACE (1989) and the Levant region (SPANIER, 2000a) where BOMBACE et al. (1994) suggested that in site protection must be given high priority, the , the decrease in winter so that ARs can produce, rather than mere- species richness is due to migration into ly attract, fish. deeper water. This phenomenon is docu- In contrast to the groupers, the popu- mented for groupers along the Israeli coast lation of S. rubrum continued to grow as well (DIAMANT et al., 1986; GOLANI throughout the study period (Fig. 4). Squir- and DAROM, 1999; ARONOV and relfish are not as highly prized as groupers, GOREN, 2003). It may account for the win- due to their smaller size, hard scales and ter decline in grouper numbers (Fig. 4) as sharp dorsal spines (GOLANI and DAROM, well lower abundances of other species. 1999). They are thus not targeted by spear- Most notable among these was the sparid fishermen. Consequently, they have been population, which vanished in the cold sea- able to establish themselves as dominant son and recuperated the following summer. Levantine cave and AR dwellers in re- Nonetheless, had vertical migration cent decades (GOLANI and BEN-TUVIA, been the reason for grouper decline, why 1985; DIAMANT et al., 1986; SPANIER did their population fail to recover the fol- et al., 1989; SPANIER and GALIL, 1991; lowing summer as the sparids' did? No cli- SPANIER, 2000a,b; BARICHE et al., 2004; matic changes were noted during this pe- GOREN and GALIL, 2005). So far as di- riod, the current was weak (0.1Kn.), water urnal species are concerned, however, temperature albeit low at 16ÆC, remained the most frequent protagonists of the as- steady (Fig. 4) and visibility was excellent semblage were wrasses and damselfish. (horizontal Secchi depth >30m). Water C. julis, T. pavo and C. chromis (Table 1), temperatures had dropped 2 months prior are similar to the most abundant species to this survey (Fig. 4), and with it AR bio- found by AZZURRO et al. (2007) in an mass and the abundance of many of the AR in the Straits of Sicily.

Medit. Mar. Sci., 10/1, 2009, 35-54 47 CHARBONNEL et al. (2002) deduced in the Mediterranean (GOLANI et al., 2002) that the increased density and biomass of and by now common throughout the Le- predators in a north Mediterranean AR vantine basin (KARACHLE et al., 2004; did not result solely from the increased food PAIS et al. 2007). A single winter observa- availability offered by surfaces of the AR, tion was made of the Brownband goatfish, but from sheltering in the interstices as well. Upeneus pori Ben-Tuvia and Golani. The present AR data concurs with this mod- Although it habitually prefers soft bottoms el, as inner and under unit crevices of AR (LYTHGOE and LYTHGOE, 1971; units were indeed densely and rapidly in- GOLANI and DAROM, 1999), it was habited by squirrelfish (Figs. 2 and 4) and recorded over the sandstone ridge, in vicin- other larger predators. SEAMAN (2000) ity to ARs and further contributed to the noted that in the south-east Mediterranean, increase in Lessepsian species richness. Mi- with no sea-grass meadows, a filter feeder grants accounted for 25.2% of total speci- dominated AR is likely to develop. The im- mens and 65.3% of the commercially ex- portance of herbivores will thus diminish, ploitable biomass in AR quadrates. whereas predation intensity will increase. The higher figures of the present study Despite an active presence in the vicinity may be explained by differences in the sam- of AR units, herbivores were not observed pling methods and/or habitat depth. Addi- feeding directly off units. Grazing activity, tional explanations can originate in the high- most notably for S. rivulatus, was indeed er efficacy of ichthyocide in exposing cryp- observed mostly in the ecotone. Neverthe- tic species (mainly members of the families less, these Lessepsian herbivores' wide range bleniidae and gobiidae). Since most of these diets and their ability to feed off various in- species are indigenous, they decrease the digenous biogenic resources (STERGIOU, relative proportion of Lessepsian species. 1988), do contribute significantly to her- However, the large migrant proportions bivory in the ARs of the Levantine basin. may also point to two trends: a spatial one, When assemblages of two natural rocky which reflects a competitive edge migrants habitats and a small Mediterranean AR have over indigenous species in AR sites, were compared in the mid 1980s similar to the one demonstrated by (DIAMANT et al., 1986), Red Sea migrant TYRRELL and BYERS (2007) for fowl- species constituted only 7.4% of the fish, ing species, and a temporal trend - the in- but contributed >20% of the standing crop. crease in time of the rate of Lessepsian col- SPANIER (2000a,) has found migrants con- onization. stituted 16.7% of the species composition Other than S. rubrum, the other fore- in tire ARs deployed in 1985 on a similar most migrants to benefit from AR presence substrate at the same depth and 18.9% in were rabbitfish. In their original, Indo-Pa- the same ARs in 1995 (SPANIER 2000b). cific habitat, rabbitfish are found in small GOLANI et al. (2007), using rotenone ichthy- schools in shallow water close to the bot- ocide in rocky coastal littoral found a 12.16% tom (FROESE and PAULY, 2006). They migrant species ratio. In the present AR feed on a wide range of benthic algae (troph- study, the percentage of migrant species ic level 2 – Table 1) and their success as mi- was 22.6%. A newcomer to the assemblage grants is attributed to the scarcity of in- was the Blue spotted cornet fish, Fistular- digenous herbivores in the Levant ia commersonii Rüppell, recently detected (LUNDBERG and GOLANI, 1995;

48 Medit. Mar. Sci., 10/1, 2009, 35-54 BARICHE et al., 2004). Their high feed- rocky outcrop in the complex control in the ing intensity and high competitive poten- summer of 2004. tial (STERGIOU, 1988) have enabled them The short duration of the FAD study to become dominant in the Levantine her- did not allow a complete colonization pat- bivore niche. The absence of the indige- tern to be described and only partial con- nous herbivore Saupe, Sarpa salpa (Lin- clusions can be drawn. During summer, naeus), which was not sighted in this study, FADs attracted schools as big as 110 spec- as well as its absence from the Israeli fish- imens of the Lessepsian Shrimp scad, A. ing yield reports in recent years (SNOVSKY djedaba (Fig. 6). Its dominance as well as and SHAPIRO, 2003) provides some sup- the utter lack of indigenous pelagic fishes port to the hypothesis raised by BARICHE near FADs may be evident of the suscep- et al. (2004) regarding its exclusion by rab- tibility of the to the bitfish. Lessepsian invasion. It was joined in au- Large numbers of juveniles, mostly dam- tumn by the transient Round sardinella, S. selfish, were very closely associated with aurita, as well the demersal damselfish C. the structures in summer (Fig. 5). This close chromis. FADs were then lowered to 15m range interaction stressed the advantage of depth in an attempt to prevent unit de- AR units over natural reef control sites in struction after the first winter storm. The their role as nurseries. Mediterranean reefs assemblage then became more heteroge- are non-living rocky outcrops and thus neous when at this height, rabbitfish and resemble ARs in temperate or less stable wrasses joined the damselfish to ascend environments (SEAMAN, 2000). Where- from ground AR units to the FADs. Since as tropical reef recruitment is chiefly gov- natural depth distribution of all three nec- erned by juvenile fish, reefs in temperate to-benthic species does not limit them to seas gravitate towards adult colonization 20m (GOLANI and DAROM, 1999; (SEAMAN and SPRAGUE, 1991). Our FROESE and PAULY, 2006), an isobath findings, however, detected large numbers must exist therefore, between 4-10m above of damselfish and wrasse juveniles during the seafloor, below which fish relate to the warm season. Therefore, since the sea- FADs as close enough to function as a sin- sonal temperature gradient in the Levan- gle structure. This ascent also supports the tine basin is so acute (from 15ÆC in winter hypothesis that vertical relief plays a key to 30ÆC in summer) and since the appear- role in AR's success as habitat (RILOV ance of juveniles was witnessed only in the and BENAYAHU, 2000). In contrast to warm season, it is suggested that a season- bottom units, the lack of fish inside FADs ally alternating recruitment mechanism suggests it was not shelter from predation took place in the AR field: a limited tem- fishes sought, but current lee, as suggested perate adult recruitment to the assemblage for fish larvae by LINDQUIST et al. (2005). every winter-spring and a more sub-tropi- This dominance in biomass of species not cal like juvenile recruitment every summer- trophically dependent on the bio-fowling autumn. This is possibly a magnification of accumulated on FADs is also in accordance the similar mechanism occurring in natu- with findings by DEUDERO et al. (1999). ral reefs, as the only record of C. chromis The present study was carried out over juveniles (n=11) other than in unit quadrates only 12 months of sampling. MONTEIRO was in great proximity (<1m) to a large and SANTOS (2000) found the cumulative

Medit. Mar. Sci., 10/1, 2009, 35-54 49 species richness in a Portuguese AR took of the two remaining bottom units, once 5 years to reach the succession plateau. severed from their FADs, advises against CHARBONNEL et al. (2000) found that such a tether of bottom to midwater struc- an AR assemblage still evolved and densi- tures in the future. ty and biomass continued to grow after 7 years. RELINI et al. (2002) reported species Acknowledgements diversity and richness were still steadily in- creasing even after 10 years. However, This study was partially supported by GOLANI et al. (2007) as well as DIAMANT a grant from the Italian Ministry for En- et al. (1986) demonstrated that after a com- vironment and Territories to E. SPANIER plete defaunation, it only took ARs along and by the Sir Maurice Hatter award fund the Israeli coast 1 year to return to pre-de- for Maritime Studies and the Joshua Salti faunation values. The data gathered for the Scholarship Foundation to D. Edelist. Mr. Haifa AR project, although it provided S. Breitstein, Mr. A. Yurman and other col- some good basic and comparative figures, leagues and students are acknowledged for was insufficient for long-term predictions their assistance in the diving operations. of ecological processes. For example – the We are grateful to Dr. D. Golani for his decline in both adult and juvenile numbers useful comments on draft of this manu- over the study period can only be fully script. Thanks are due also to Dr. N. Barkan understood in a multi-annual study. A longer from the Department of Statistics, Univer- duration of data collection is therefore high- sity of Haifa, for the statistic consulta- ly advised. tion. Twenty years ago, BOHNSACK and SUTHERLAND (1985) termed AR con- References struction 'more of an art than a science' and stressed the need for inexpensive, effective, ARONOV, A. & GOREN, M., 2003. Vari- long-lasting, easily handled, easily trans- ation in aggregation behaviour and ported structures. Marine structures are by anatomy resulting from differences in and large planned to withstand storms of a degree of sperm competition in groupers certain repetition probability, based on the (Epinephelinae, Serranidae). Israel Jour- probability of a worse storm occurring dur- nal of Zoology, 49: 72. ing the suggested period. Structure resilience, AZOV, Y., 1986. Seasonal patterns of phy- cost-benefit and potential damage consid- toplankton productivity and abundance erations must be taken into account in these in near shore oligotrophic water of the planning stages. The concrete structures Levant Basin (Mediterranean). Jour- used as ground units in the Haifa AR proj- nal of Plankton Research, 8: 41-53. ect were originally prefabricated sewer AZZURRO, E., PAIS, A., CONSOLI, P. ponds. They were thus cheap and relative- & ANDALORO, F., 2007. Evaluating ly small and were easily fitted with the pipes day–night changes in shallow Mediter- for enhanced complexity. ranean rocky reef fish assemblages by Nevertheless, when coupled with the visual census. , 151: 2245- FAD units, they proved inadequate for use 2253. as ARs, as their life span was only 3-5 months. BARICHE, M., LETOURNEUR, Y. & The relative longevity (over 3 years so far) HARMELIN-VIVIEN, M., 2004. Tem-

50 Medit. Mar. Sci., 10/1, 2009, 35-54 poral fluctuations and settlement pat- In: Seaman, W.J. (ed.) Artificial Reef terns of native and Lessepsian her- Evaluation with Application to Natural bivorous fish on the Lebanese coast Marine Habitats. CRC Press, Boca Ra- (Eastern Mediterranean). Environ- ton, Florida 127-164. mental Biology of Fishes 70: 81-90. BRAITHWHAITE, V. A., 1998 Spatial BERMAN, T., TOWNSEND, D. W., EL- memory, landmark use and orientation SAYED, S. Z., TREES, C. C. & AZOV, in fish. In: Healy, S. (ed.) Spatial Rep- Y., 1984. Optical transparency, chloro- resentation in Animals. 86-102. phyll and primary production in the CALEY, M. J., CARR, M. H., HIXON, eastern Mediterranean near the Israeli M. A., HUGHES, T. P., JONES, G. P. coast. Oceanologica Acta 9: 439-447. & MENGE, B. A., 1996. Recruitment BOHNSACK, J. A., 1989. Are high densi- and the local dynamics of open marine ties of fish at artificial reefs the result populations. Annual Review of Ecolog- of habitat limitation or behavioural ical Systems 27: 477-500. preference? Bulletin of Marine Science CARR, M. H. & HIXON, M. A., 1997. Ar- 44: 631-645. tificial reefs: the importance of com- BOHNSACK, J. A. & SUTHERLAND, parisons with natural reefs. Fisheries D. L., 1985. Artificial reef research: a 22: 28-33. review with recommendations for fu- CHARBONNELL, E., FRANCOUR, P., ture priorities Bulletin of Marine Science HARMELIN, D. O. & BACHET, F., 37: 11-39. 2000. Effects of artificial reef design on BOHNSACK, J. A., HARPER, D. E., associated fish assemblages in the Côte MCCLELLAN, D. B. & HULSBECK, Bleue marine park (Mediterranean Sea, M., 1994. Effects of reef size on colo- ). In: Jensen, A.C., Collins, K.J. nization and assemblage structure of and Lockwood, A.P.M. (eds) Artificial fish at artificial reefs off southeastern Reefs in European Seas. Kluwer Aca- Florida, U.S.A. Bulletin of Marine Science. demic Publishers, Dordrecht, Nether- 55: 796-823. lands. 365-377. BOHNSACK, J. A., ECKLUND, A. M. & CHARBONNEL, E., SERRE, C., SZMANT, A. M., 1997. Artificial reef RUITTON, S., HARMELIN, J. G. & research: Is there more than the at- JENSEN, A., 2002. Effects of increased traction-production issue? Fisheries 22: habitat complexity on fish assemblages 14-16. associated with large artificial reefs BOMBACE, G. 1989. Artificial Reefs in units (French Mediterranean coast). the Mediterranean Sea. Bulletin of ma- ICES Journal of Marine Science 50: 208- rine Science 44: 1023-1042. 213. BOMBACE, G., FABI, L., FIORENTINI, DEUDERO, S. C., MERELLA, P., L. & SPERANZA, S., 1994. Analysis MORALES-NIN, B., MASSUTI, E. & of the efficacy of ARs located in five ALEMANY, F., 1999. Fish communi- different areas of the Adriatic Sea. Bul- ties associated with FADs. Scientia Ma- letin of Marine Science 55: 559-580. rina 63: 199-207. BORTONE, S. A., SAMOYLIS, M. A. & DIAMANT, A., BEN-TUVIA, A., FRANCOUR, P., 2000. Fish and BARANES, A. & GOLANI, D., 1986. macroinvertebrate evaluation methods. An analysis of rocky coastal eastern

Medit. Mar. Sci., 10/1, 2009, 35-54 51 Mediterranean fish assemblages and a ology 21: 364 comparison with small adjacent arti- HERUT, B., ALMOGI-LABIN, A., ficial reef. Journal of Experimental Marine JANNINK, N. & GERTMAN, I., 2000 Biology and Ecology 97: 269-285. The seasonal dynamics of nutrient and FRAZER, T. K. & LINDBERG, W. J., chlorophyll a concentrations on the SE 1994. Refuge spacing similarly affects Mediterranean shelf-slope. Oceanolog- reef associated species from three phy- ica Acta 23: 771 – 782. la. Bulletin of Marine Science 55: 388- JENSEN, A. C., 2002. Artificial reefs of 400. Europe: perspective and future. ICES FROESE, R. & PAULY, D. (Eds) 2006. Journal of Marine Science 59: 3-13. FishBase. World Wide Web electron- JENSEN, A. C., COLLINS, K. J. & ic publication. www.fishbase.org, ver- LOCKWOOD, A. P. M. (eds.) 2000. sion (06/2006).\ Artificial Reefs in European Seas. Kluw- GOLANI, D., 1998. Impact of Red Sea fish er Academic Publishers, Dordrecht, migrants through the on Netherlands. 508. the aquatic environment of the eastern KARACHLE, P. K., TRIANTAPHYLLI- Mediterranean. Yale F&ES Bulletin 103: DIS, C. & STERGIOU, K. I., 2004. 375-387. Fistularia com- GOLANI, D. & BEN-TUVIA, A., 1985. mersonii Rüppell, 1838: a Lessepsian The biology of the Indo-Pacific squir- sprinter. Acta Ichtiologica et Piscatoria relfish, S. rubrum, a Suez Canal mi- 34: 103-108. grant to the eastern Mediterranean. LINDQUIST, D. C., SHAW, R. F. & Journal of Fish Biology 27: 249-258. HERNANDEZ, F. J., 2005. Distribu- GOLANI, D. & DAROM, D., 1999. Hand- tion patterns of larval and juvenile fish- book of the Fishes of Israel, Keter Pub- es at offshore petroleum platforms in lishing Jerusalem, Israel. 269. the north-central Gulf of Mexico. Es- GOLANI, D., ORSI RELINI, L., tuarine, Coastal and Shelf Science 62: MASSUTI, E. & QUINGARD, J. P., 655-665. 2002. CIESM Atlas of Exotic Species in LUNDBERG, B. & GOLANI, D., 1995. the Mediterranean. CIESM Publishers, Diet adaptations of Lessepsian migrant Monaco. 256. Rabbitfish, Siganus luridus and S. rivu- GOLANI, D., REEF-MOTRO, R., latus, to the algal resources of the EKSHTEIN, S., BARANES, A. & Mediterranean coast of Israel. Marine DIAMANT, A., 2007. Ichthyofauna of Ecology 16: 73. the rocky coastal littoral of the Israeli LYTHGOE, J. & LYTHGOE, G., 1971. Mediterranean, with reference to the Fishes of the sea. Blandford Press, Lon- paucity of Red Sea (Lessepsian) mi- don UK. 320. grants in this habitat. Marine Biology MONTEIRO, C. C. & SANTOS, M. N., Research 3: 333-341. 2000. Portuguese artificial reefs. In: GOREN, M. & GALIL, B., 2005. A review Jensen, A.C., Collins, K.J. and Lock- of changes in the fish assemblages of wood, A.P.M (eds.) Artificial reefs in Levantine inland and marine ecosys- European Seas. Kluwer Academic Pub- tems following the introduction of non- lishers, Dordrecht, Netherlands 249- native fish. Journal of Applied Ichthy- 261.

52 Medit. Mar. Sci., 10/1, 2009, 35-54 MORENO, I., ROCA, I., RE ONES, O., Science 59: 132-137. COLL, C. & SALAMANCA, M., 1994. RILOV, G. & BENAYAHU, Y., 2000. Fish Artificial reef program in Balearic wa- assemblage on natural versus vertical ters (Western Mediterranean). Bulletin artificial reefs: the rehabilitation per- of Marine Science 55: 667-671. spective. Marine Biology 136: 931-942 OGAWA, Y., 1982. Jinko gyosho, tsukiiso SAURNIA, A., 1973. La production pri- and marine organisms. Japanese artifi- maire planctonique en Méditerranée. cial reef technology technical report 604: Newsletter of Cooperation Investment. 42-48. Aquabio, Inc., 2957 Sunset Blvd., Mediterranean, Species Issue, 5: 1-128. Bellair Bluffs, Fl. SEAMAN, W. J., 2000. Artificial Reef Eval- OSENBERG, C.W., ST. MARY, C. M., uation with Application to Natural Marine WILSON, J.A. & LINDBERG, W. J., habitats. CRC Press, Boca Raton, Flori- 2002. A quantitative framework to eval- da, US. 246. uate the attraction–production con- SEAMAN, W. J. & SPRAGUE, L. M., 1991. troversy. ICES Journal of Marine Science Artificial Habitats for Marine and Fresh- 59: 214-221. water Fisheries. Academic Press, Lon- PAIS, A., MERELLA, P., FOLLESA, M. don, UK. 285. C. & GARIPPA, G., 2007. Westward SHANNON, C. E., 1948. A mathematical range expansion of the Lessepsian mi- theory of communication. Bell System grant Fistularia commersonii (Fistu- Technical Journal 27: 379-423 lariidae) in the Mediterranean Sea, with SHENG, Y. P., 2000. Physical characteris- notes on its parasites. Journal of Fish tics and engineering at reef sites In: Biology 70: 269-277. Seaman, W.J. (ed.) Artificial Reef Eval- PITCHER, T. J., MAGURRAN A. E. & uation with Application to Natural Marine WINFIELD, I. J., 1982. Fish in larger Habitats. CRC Press, Boca Raton, Flori- shoals find food faster. Behavioral Ecol- da, US. 51-94. ogy and Sociobiology 10:149-151. SNOVSKY, G. & SHAPIRO, J. (eds.) 2003. PITCHER, T. J. & SEAMAN, W.J., 2000. Israel fisheries in figures, State of Israel. Petrarch's principle: how protected hu- Ministry of Agriculture – Department man-made-reefs can help the recon- of Fisheries struction of fisheries and marine ecosys- SPANIER, E., 1996. Assessment of habi- tems. Fish and fisheries,1:73-81. tat selection behavior in macro-organ- POLOVINA, J. J., 1989. Artificial reefs: isms on artificial reefs. In: Jensen, A.C. nothing more than benthic fish aggre- (ed.) European Artificial Reef Research gators. California Cooperative Oceanic - Proceedings of the 1st EARRN (Euro- Fisheries Investigations 30: 37-39. pean Artificial Reef Research Network). PRATT, J. A., 1994. Artificial habitats and Ancona, Italy. Southampton Oceano- restoration: managing for graphic Centre, UK, 323-336., the future. Bulletin of Marine Science SPANIER, E., 2000a. Artificial reefs off 55: 268-275. the Mediterranean coast of Israel. In: RELINI, G., RELINI, M., TORCHIA, G. Jensen, A.C., Collins, K.J. and Lock- & PALANDRI, G., 2002. Ten years of wood, A.P.M (eds.) Artificial reefs in censuses of fish fauna on the Loano ar- European Seas. Kluwer Academic Pub- tificial reef. ICES Journal of Marine lishers, Dordrecht, Netherlands. 1-19.

Medit. Mar. Sci., 10/1, 2009, 35-54 53 SPANIER, E., 2000b. Changes in the ichthy- fish assemblage on a coralligenous shal- ofauna of an artificial reef in the south- low shelf off the Mediterranean coast eastern Mediterranean in one decade. of Northern Israel. Journal of Fish Bi- Scientia Marina 64: 279-284. ology 35: 641-649. SPANIER, E. & GALIL, B. S. 1991. Lessep- SPANIER, E., TOM, M., PISANTY, S. & sian migration: a continuous biogeo- ALMOG SHTAYER, G., 1990. Arti- graphical process. Endeavour 15: 102- ficial reefs in the low productive ma- 106. rine environment of the southeastern SPANIER, E., PISANTY, S. & TOM, M. Mediterranean. P.S.Z.N.I: Marine Ecol- 1983. Artificial reef in the eastern ogy, 11: 61-75. Mediterranean, preliminary results. In: STERGIOU, K. I., 1988. Feeding habits of Shuval, H. (ed.) Developments in Ecol- the Lessepsian migrant Siganus luridus ogy and Environmental Quality Vol. II. in the Eastern Mediterranean, its new Balaban ISS. Philadelphia, 317-325. environment. Journal of Fish Biology SPANIER, E., TOM, M. & PISANTY, S. 33: 531-543. 1985a. Enhancement of fish recruit- STONE, R. B., PRATT, H. L., PARKER, ment by artificial enrichment of man- R. O. & DAVIS, G. E., 1979. A com- made reefs in the southeastern Mediter- parison of fish populations on an arti- ranean. Bulletin of Marine Science, ficial and a natural reef in the Florida 37: 356-363. Keys. Marine Fisheries Review 41: 1-11. SPANIER, E., TOM, M., PISANTY, S., TURNER, C. H., EBERT, E. E. & GIVAN, BREITSTEIN, S., TUR-CASPA, Y. R. R., 1969. Man-made reef ecology. & ALMOG, G., 1985b. Development California Department of Fish and Game of artificial habitats for commercial Fisheries Bulletin 146: 221. species and open sea fish culturing in TYRRELL, M. & BYERS, J. E., 2007. Do the southeastern Mediterranean. Pro- artificial substrates favor non-indige- ceedings of the joint international scien- nous fouling species over ? tific diving symposium October 31 to Journal of Experimental Marine Biolo- November 3, 1985. Academy of Un- gy and Ecology 342: 54-60. derwater Sciences, La Jolla, California, WILCOXON, F., 1945. Individual Com- 123-135. parisons by Ranking Methods. Bio- SPANIER, E., PISANTY, S., TOM, M. & metrics Bulletin 1: 80-83. ALMOG SHTAYER, G., 1989. The

Submitted: May 2008 Accepted: April 2009 Published on line: June 2009

54 Medit. Mar. Sci., 9/2, 2008, 35-54