Co-culture of the sea urchin Paracentrotus lividus and the edible mussel Mytilus edulis L. on the west coast of Scotland, United Kingdom.
ABSTRACT The sea urchin Paracentrotus lividus was grown on commercial mussel long-lines with the blue mussel, Mytilus edulis, at a farm in Loch Beag, on the west coast of Scotland, to investigate if enhanced sea urchin survivorship and performance (somatic and gonadal) resulted from sea urchin-mussel co-culture system. The sea urchins were red four diets including: two class sizes of M. edulis, (1) large mussels (31.16 [+ or -] 5.25 mm shell length), (2) small mussels (18.47 [+ or -] 4.86 mm shell length), (3) the kelp Laminaria spp. and (4) no additional feed. The experimental period lasted for 12 mo. No significant difference was observed in survivorship among treatments, ranging from 95.8% to 100%. Final test diameter, linear growth rates (LGR) and specific growth rates were significantly greater for the sea urchins red the kelp diet compared with sea urchins red either of the mussel diets and given no additional feed. No significant difference in growth rates was seen between the sea urchins red the two size classes of mussel. A seasonal variation in growth rates was observed for all the treatment groups with a greater LGR in September to November 2005 compared with January to March 2006. Sea urchins red on kelp showed significantly greater gonadal growth than the other treatment groups and no significant difference was observed between the sea urchins red the two size classes of mussel. Gonad coloration in the sea urchins grown on the kelp diet was acceptable or excellent. Minimal roe material in the other treatments prevented color assessment. The results show that P. lividus exhibits high survivorship and linear growth rates when grown on long-lines used for commercial mussel cultivation, even at this northerly latitude. The results suggest that Laminaria spp. is a superior food source for P. lividus compared with the mussel M. edulis, however, a preharvest diet would have to be used to increase roe content prior to harvest even when red a diet of Laminaria spp. The co-culture of sea urchins and mussels could potentially be implemented globally, wherever rope-grown mussel culture is practiced and sea urchin hatcheries are present. This would enable mussel farms to diversify in the production of a second commercially valuable product, with minimal requirement for new equipment or infrastructure, and would reduce the pressure on the already depleted wild stocks of sea urchins.
KEY WORDS: Paracentrotus lividus, sea urchin, Mytilus edulis, blue mussel, co-culture
Sea urchin roe is a commercially valuable food item and, as many sea urchin fisheries are overexploited, there is an increasing effort into commercializing sea urchin cultivation methods (Kelly 2004). Whereas the biological basis for sea urchin culture has been long established, research continues to refine the hatchery production of the juveniles (Liu et al. 2007a, 2007b). If the industry is to develop further, however, cost effective grow-out systems ate also required, as are suitable diets for the grow-out phase that enhance gonad biomass, color, taste, and texture. In contrast to Japan, where hatchery-reared juveniles ate mainly released to managed areas of sea floor (Hagen 1996), researchers in other countries continue to experiment with a wide range of grow-out systems for juvenile and adult urchins, ranging from relocation from poor to good feeding grounds (Moylan 1997) and ranching urchins caged on the sea floor (Cuthbert et al. 1995, Bridger et al. 1998, Vadas et al. 2000) to sea cages suspended from a supporting structure (Robinson & Colborne 1997, Mendes & Becarra 2004, James 2006). Hatchery reared juveniles have been grown in suspended culture with Atlantic salmon, Salmo salar (Kelly et al. 1998, Cook & Kelly 2007a, Cook & Kelly 2007b), in closed recirculation systems (Grosjean et al. 1998), in land-based integrated systems (Shpigel et al. 2004) and in rock pools in southern Ireland (J. Chamberlain, Dunmanas Seafoods Ltd., pers. comm.). Systems that accelerate growth to market size, whereas producing a uniform size class, would give an economic advantage. One possible route to obtaining sustainable and environmentally beneficial systems for urchin culture is to further examine their potential in co-culture systems.
Although at the extreme northernmost limits of its range in north-west Scotland, Paracentrotus lividus is being evaluated as a potential new species for aquaculture. This region, including the Shetland Isles, produces around 5,000 tons per annum of the blue mussel, Mytilus edulis. The mussels are typically grown in long-line systems where long topes of "drops," seeded with mussels, hang vertically in the water column at intervals of 25 cm to 1 m from a horizontal top line. The long-lines are anchored to the seabed at either end and buoyed up in the water column, to around 30 cm subsurface, by large barrel floats. According to market demand, mussels are harvested and graded to size throughout the year. Approximately 25% of the harvest weight can be comprised of small (shell length <30 mm, depending on time of year) or cracked mussels and other fouling organistas (I. MacKinnon, Loch Beag Shellfish, pers. comm.).
P. lividus is often described as herbivorous, although there are documented instances of its feeding on artificial diets containing fish meal (Fernandez & Boudouresque 2000), sponges, hydrozoa, copepods, dead fish and mussels (Boudouresque & Verlaque 2007). Diet quality can significantly influence the somatic and gonadal growth in sea urchins (Frantzis & Gremare 1992, Cook et al. 1998, Meidel & Scheibling 1999, Schlosser et al. 2005), particularly elevated protein levels, such as those found in animal-derived diets (Cook et al. 1998, Schlosser et al. 2005). Increased growth rates have been observed in Strongylocentrotus droebachiensis fed on a combination of kelp and mussel tissue compared with a high ration of kelp alone (Meidel & Scheibling 1999), although the size of whole mussels has been shown to influence feeding rates in this species (Briscoe & Sebens 1988) and Psammechinus miliaris (Otero-Villanueva et al. 2006), which has been attributed to prey-handling capability. In a short term study, Haya and Regis (1995) found that P. lividus would readily consume, and had a significantly higher absorption efficiency, when fed whole mussels (74.2%) compared with a primarily soya based diet (42.4%). To date, however, no long-term study has been undertaken on the effect of a whole mussel diet on the performance of P. lividus.
The aim of this study was to examine the long-term suitability of the long-line mussel culture system and whole mussels, graded unsuitable for harvest, as a diet for P. lividus. The effect of the culture system and diet on survivorship, somatic and gonadal growth and gonadal color was assessed.
The study was conducted for 12 mo (August 31, 2005-2 August 2006) at a commercial mussel farm operating a 0.06 [km.sup.2] site in Loch Beag on the west coast of Scotland (N 56[degrees]53.06', W 05[degrees]44.21'). The company operates a "long-line" system comprising of a single 100-m and two 300-m horizontal lines, which are 25 m apart and arranged in rows parallel to the shore in a seabed depth of approximately 19-m. The 100-m long-line nearest the shore was supported by smaller 25-L air-filled containers placed at 25-cm intervals and the two 300-m lines were supported by large 400-L barrel-like flotation buoys (Explora) at 5-m intervals. Vertical lines supporting the mussels are attached to the long-lines at 50-cm intervals. The annual production of mussels at this site is 40 ton and the grow-out time for the mussels is between 24 and 30 mo. The mussels are graded at harvest and the grading process produces approximately 10 tons of waste material (or 25% of total production) over the production cycle (I. MacKinnon, Loch Beag Shellfish, pers. comm.).
Experimental Design and Sampling Procedures
Hatchery-reared Paracentrotus lividus (17.60 [+ or -] 2.03 mm (test diameter); n = 720) were selected from stocks reared at the Scottish Association for Marine Science and red on kelp (Laminaria spp.). Previous research has shown that hatchery-reared urchins can be successfully transferred to sea at a smaller size (Kelly et al. 1998). The sea urchins were stocked into 36 pyramidal "pearl" nets (mesh size 5.0 mm; dimensions 40 x 40 x 30 mm), commonly used in the shellfish industry, at a density of 20 sea urchins per net and were deployed at the study site on the August 31, 2005. The density was equivalent to 65 ind. [m.sup.-2] of net surface area, which corresponded to half the density of juvenile P. miliaris (Kelly et al. 1998) and P. lividus (Cook & Kelly 2007a) that has previously been shown to significantly enhance sea urchin growth in a co-culture system with Atlantic salmon. This reduced density was chosen as the initial size and biomass of the sea urchins was greater than those used in the previous studies and because space in the pearl nets would also be occupied by the experimental diets.
Experimental diets included: (1) large mussels (LM; shell length 31.16 [+ or -] 5.25 mm; mean [+ or -] SE, n = 360) (2) small mussels (SM; shell length 18.47 [+ or -] 4.86 mm; mean [+ or -] SE, n - 360), (3) kelp, Laminaria spp. (K), and (4) no additional food (NF), with the exception of fouling organisms on the pearl nets. The mussel size classes were based on previous laboratory studies of sea urchin grazing on Mytilus edulis (Otero-Villanueva et al. 2006) and the two sizes were chosen to see if there was an optimal size of mussel for urchin consumption. The whole (i.e., shell intact) and damaged (i.e., shell cracked or partially missing) mussels were collected from the commercial long-lines, graded on site and a new batch of mussels from the graded "waste" material was fed to the sea urchins bimonthly. Quantities of mussels were based on an average feeding rate of 2 large mussels and 4 small mussels [urchin.sup.-1] [day.sup.-1] (Otero-Villanueva et al. 2006) and the pearl nets were checked on a weekly basis to ensure ad libitum food availability throughout the study. Sea urchins were observed feeding on the mussels throughout the experimental period (E. Cook, pers. obs.) and the mesh size of the pearl nets was small enough to prevent mussel consumption by other predators. Bimonthly triplicate samples, each consisting of 30 mussels, were removed from the graded waste material to confirm differences in shell length between the LM and SM treatment groups and to calculate the proportion of damaged mussels. Photographs of the mussels were taken using an Olympus C2020 digital camera at time of sampling and later analyzed by measuring the maximum shell length (to 0.5 mm) using digital software (ImageJ; http://rsb.info.nih.gov/ij/), previously calibrated with caliper measurements and by recording the number of damaged shells, as a proportion of the total number of mussels in the sample. Laminaria spp., including Laminaria digitata and Sacchoriza latissima (=Laminaria saccharina), was collected in Loch Beag and the remaining macroalgae in the pearl nets, from the previous sampling event, was replaced by freshly collected fronds without any visible signs of epiphytic growth. Laminaria spp. was fed to the sea urchins ad libitum on a bimonthly basis, although occasionally additional macroalgae was added if pearl nets were found empty on the weekly checks. Food rations were increased throughout the experimentas size and consumption rates of the sea urchins increased.
The pearl nets containing the sea urchins were arranged along the three parallel long-lines, with 12 pearl nets on each long-line at a depth of 4 m from the surface. Three replicates of the four feeding treatments were randomly assigned to the pearl nets on each long-line. Replicates on each line were at least 2 m apart. All the pearl nets were between 100-150 m from the shore. Water temperature was measured daily throughout the experimental period, using a TinyTag temperature data logger, and varied from 7.5[degrees]C to 15.5[degrees]C. All pearl nets were thoroughly cleaned on a bimonthly basis, once the urchins and remaining diets had been removed, using a stiff brush.
The suitability of the long-line system for sea urchin cultivation was assessed by monitoring the survivorship and performance of the sea urchins on the mussel diets. Several variables that are influenced by food availability and quality, including somatic (test diameter and linear growth rate) and gonadal growth (gonad index) were monitored. Survivorship (S) at each location was calculated from the formula:
S (%) = [[U.sub.L]/[U.sub.D]] X 100
Where [U.sub.L] is the total number of live sea urchins within each pearl net and [U.sub.D] is the total number of dead or missing urchins.
Individual test diameters of all the sea urchins within each pearl net (TD; maximum equatorial axis of the test) were measured on a bimonthly basis using the method used by Cook and Kelly (2007a). Test diameter was measured at the study site and the urchins were returned to the same pearl net within 5 mins of emergence. Wet weights ([+ or -] 0.01 g) were measured, following 5 mins drying on absorbent paper, at Month 0 and Month 12. For each replicate pearl net, the average TD and wet weight of the sea urchins was calculated and used in the analysis. Sea urchin linear growth rates (LGR) were recorded as [micro]m [day.sup.-1] and specific growth rates (SGR) as % (ww) [day.sup.-1] and were determined by the following formulas:
LGR = ([L.sub.f] - [L.sub.i])/t
SGR=(ln [W.sub.f] - ln [W.sub.i] x 100)/t
where [L.sub.f] and [L.sub.i] were the final and initial test diameters ([micro]m), [W.sub.f] and [W.sub.i] were the final and initial wet weights (g) and t was time in days.
The gonad content was assessed using the gonadal index (GI). Gonad indices were calculated at Month 0 by sacrificing 20 sea urchins and at Month 12 using a subsample of five individuals from each of the pearl nets remaining at the end of the experiment. GI was calculated from the formula:
GI(%) = [W.sub.g]/Wu x 100
where [W.sub.g] is the wet weight of the gonad and [W.sub.u] is the total wet weight (g) of the urchin. For each replicate pearl net, the average GI of the sea urchins was calculated and used in the analysis.
Sea urchin gonad color was assessed at month 12 on a subsample of five sea urchins from each pearl net using a PANTONE Color Card (Cook 1999) and the same observer in natural daylight. In addition, the gonad was analyzed for pigment color using a Konica Minolta CR-300 Chroma Meter to quantitatively determine the color using the quantitative three dimensional color method developed by the Commission Internationale de l'Eclairage 1976 (CIE 1976); Lightness ([L.sup.*]), redness ([a.sup.*]), and yellowness ([b.sup.*]) as used by Agatsuma (1998), Robinson et al. (2002), and Cook and Kelly (2007a). Four replicate measurements were taken for each gonad sample and averaged to determine the three color parameters.
Sea urchin survivorship, test diameter, total biomass, and gonad index (GI) were assessed at Month 12 using a two-way analysis of variance (ANOVA) with dietary treatment and long-line as the fixed factors. Linear and specific growth rates were analyzed for the 12-mo period using a two-way ANOVA as above. In addition, the LGR were analyzed using a three-way ANOVA with dietary treatment (4 levels), Line (2 levels), and Time (5 levels) as fixed factors using the general linear model. Unfortunately, only the data from the pearl nets deployed on the two 300-m long-lines could be used in the statistical analysis because of damage to the 100-m long-line and associated pearl nets early on in the experiment. A paired t-test was used to assess shell length and damage between the two mussel treatments for each sampling event. Prior to analysis, all percentage data were arcsine square root transformed and the data were tested for normality (Kolmogorov-Smirnov test) and homogeneity of variances (Bartlett test, (Underwood 1997). If necessary, the data were transformed to meet statistical assumptions. Significant treatment effects (P < 0.05) were further investigated to identify the sonrce of the differences (Tukey multiple comparison test). All ANOVA and posthoc tests were performed using MINITAB, Release 14 for Windows.
Mussel Shell Length and Proportion of Damaged Shells
The size of the mussels in the LM and SM treatment groups remained consistent throughout the 4 sampling events, with the exception of January 2006, when the mean shell length of the mussels in the SM treatment group increased slightly. The mean ([+ or -] SE) shell length for the LM and SM treatment groups was 31.16 [+ or -] 5.25 mm and 18.47 [+ or -] 4.86 mm, respectively, and there was a significant difference in length between the two treatment groups at each sampling event (Table 1).
The proportion of damaged mussels was not significantly different between the LM and SM for the sampling events, with the exception of January 2006 (Table 1). The proportion of damaged mussels was generally less than 10% and ranged from 6.44 [+ or -] 2.57% to 9.96 [+ or -] 3.58%. In the SM treatment group, however, this proportion increased to 20.57 [+ or -] 1.60% in January 2006.
On Line 1, a total of 5 pearl nets were lost over the experimental period and the remaining pearl nets were badly damaged during poor weather, which may have been exacerbated by the buoyancy system on this line, between November 2005 and January 2006. As a consequence, Line 1 was not included in the statistical analysis. No nets were lost from Line 2 and 1 net was lost from Line 3 over the experimental period. Survivorship of Paracentrotus lividus in the remaining pearl nets on Lines 2 and 3 at Month 12 was not significantly affected by treatment group or long-line and was greater than 95.8% (Kruskall Wallis; H = 2.34; P > 0.05). A 100% survival rate was recorded for sea urchins in the K treatment group (Table 2).
Final Test Diameter and Linear Growth Rate
The average final test diameter (TD) was 29.0 [+ or -] 3.9 mm (LM), 28.6 [+ or -] 3.0 mm (SM), 40.5 [+ or -] 3.9 mm (K) and 23.9 [+ or -] 2.4 mm (NF), with a significantly greater final test diameter exhibited by sea urchins on treatment K compared with treatments LM, SM, and NF (ANOVA; F = 483; P < 0.001; Tukey P < 0.05) (Table 2). No significant difference in test diameter was observed between the LM and SM treatment groups and between the long-lines (P > 0.05). In addition, no significant interaction was seen between treatment and long-line (P > 0.05).
A highly significant difference was seen in the linear growth rate (LGR) between the 4 treatment groups over the 12 mo experimental period (ANOVA; F = 39.8; P < 0.001), with the sea urchins red the kelp diet exhibiting the highest LGR of 61.8 [+ or -] 2.4 [micro]m TD [day.sup.-1] (mean [+ or -] SE) compared with the urchins that were not given any additional feed (17.8 [+ or -] 4.4 [micro]m TD [day.sup.l]). No significant difference was observed between the LM and SM treatment groups (Table 2). In addition, a highly significant difference was shown with time (ANOVA; F = 6.75; P < 0.001), with the highest LGR for all treatment groups observed between September and November 2005 (45.5 [+ or -] 26.9 84.1 [+ or -] 22.4 [micro]m TD [day.sup.-1]; equivalent to 1.4-2.5 mm [month.sup.-l]) compared with the lowest between January and March 2006 (2.43 [+ or -] 5.15 - 31.55 [+ or -] 13.8 gm TD [day.sup.-1]; Fig. 1). The maximum average linear growth rate (LGR) was 97.26 [+ or -] 19.11 [micro]m TD [day.sup.-1] (mean [+ or -] SE) for P. lividus on Treatment K between March and May 2006. No significant difference was observed in LGR between the 2 experimental long-lines (ANOVA; F = 0.14; P = 0.872) or interaction between diet and long-line (P > 0.05).
Total Biomass and Specific Growth Rates (SGR)
The total biomass of P. lividus was significantly different between dietary treatments at Month 12 (ANOVA; F = 21.11; P < 0.001). The total biomass of sea urchins fed on the kelp diet (1168 [+ or -] 312 g) was over double the biomass of sea urchins in either the LM or SM treatment groups (427 [+ or -] 183 g and 490 [+ or -] 130 g, respectively) and three times greater than the biomass of sea urchins given no additional feed (314 [+ or -] 48 g). No significant difference was seen between the sea urchins on the LM and SM dietary treatments (Fig. 2).
A highly significant difference in specific growth rates (SGR) of P. lividus was seen between the 4 treatment groups at the end of the experimental period (Table 3). P. lividus had a significantly higher SGR on the K treatment compared with the other treatment groups. No significant difference was exhibited between the sea urchins fed on LM and SM, however, these urchins had a SGR of approximately double the SGR of the sea urchins given no additional feed (NF). No significant difference in SGR was observed between the long-lines or interaction between diet and long-line.
Gonad Index, Condition, and Coloration
A highly significant difference was observed in gonad index between the dietary treatment groups at Month 12 (ANOVA; F = 71.51; P < 0.001), whereas no significant difference was seen between the long-lines (ANOVA; F = 1.15; P = 0.29). In addition, no significant interaction was seen between the long-lines and diets (P > 0.05). P. lividus on the K treatment had a GI approximately six times greater (3.0 [+ or -] 2.0%) than the sea urchins on the LM (0.5 [+ or -] 0.4%), SM (0.6 [+ or -] 0.6%) diets and NF treatment (0.3 [+ or -] 0.5%) (Fig. 3). At Month 12, 95% of the urchins on the K treatment spawned upon dissection, whereas < 20% of the urchins on the remaining diets spawned.
The roe coloration for P. lividus fed on the kelp diet at the end of the experimental period ranged from yellow orange to bright orange/red (i.e., acceptable to excellent colors, respectively). No unacceptable roe coloration (i.e., light to dark brown) was observed in this treatment group. The lightness coefficient ([L.sup.*]) was measured as 43.19 [+ or -] 15.05 (mean [+ or -] SE; n = 30), the redness coefficient ([a.sup.*]) was 19.87 [+ or -] 4.70 and the yellowness coefficient ([b.sup.*]) was 29.68 [+ or -] 6.00. Roe coloration for the remaining three treatments could not be calculated because of the minimal gonad material available (GI < 1.0%).
The survivorship and linear growth rates of hatchery reared Paracentrotus lividus show that this species can thrive in the culture environment offered by a commercial mussel farm. The results suggest that the co-culture of P. lividus on mussel long-lines is a viable grow-out method, providing the sea urchins are given a diet of macroalgae, such as Laminaria spp. and a preharvest diet to increase roe content.
[FIGURE 1 OMITTED]
The growth rate of the sea urchins fed on the two mussel diets was similar and they grew to a final test diameter of 29 mm, equivalent to 1.0 mm [month.sup.-1] over the 12 mo experimental period. Evidence that P. lividus was feeding on the mussels was shown by grazing marks around the edges of empty shells. This growth rate was comparable to juvenile P. lividus (initial test diameter 10.1 mm) held with no additional feed adjacent to Atlantic salmon in north-west Scotland (1.2 mm [month.sup.-1]) at similar seawater temperatures (Cook & Kelly 2007a). However, in the latter study, P. lividus was held at a density of 40 individuals in each pearl net (130 ind. [m.sup.-2]), twice the density as in the present study. The growth rate in this study for urchins fed on the mussel diets was also lower than the those observed for P. lividus by Fernandez and Caltagirone (1994), when the sea urchins, held at a density of 200 ind. [m.sup.-2], were fed an animal-based artificial diet in a lagoon in France over a 20-mo period (1.4 mm [month.sup.-1]), thus suggesting that the two experimental whole mussel diets were less than optimal for enhanced somatic growth in this species. Further investigation, however, is required into the provision of a greater proportion of crushed mussel shells in the diet, which may reduce handling time and energy expenditure by the sea urchins and thus indirectly increase somatic growth.
[FIGURE 2 OMITTED]
In contrast, the growth rate of P. lividus fed on the kelp diet was significantly greater than observed in the urchins on the two mussel diets, equivalent to 1.9 mm [month.sup.-1] over the experimental period. Analysis of the gut contents of P. lividus in the field has shown that this species is predominantly herbivorous, when resources are nonlimiting (Verlaque & Nedelec 1983, Boudouresque & Verlaque 2007). A recent study by Cook and Kelly (2007b) has also found that P. lividus is less capable of benefiting from high dietary protein levels (>23%), such as those found in animal derived diets, compared with the more omnivorous sea urchin Psammechinus millaris.
[FIGURE 3 OMITTED]
A seasonal variation in growth rates was observed, with maximum growth rates in the autumn (September to November 2005), coinciding with elevated seawater temperatures (12.0[degrees]C to 15.3[degrees]C). Minimum growth rates for all the treatment groups were observed between January and March 2005, when the seawater temperature ranged between 7.5[degrees]C and 8.5[degrees]C. This supports the seasonal variation observed in somatic growth rates of P. lividus grown adjacent to Atlantic salmon (Cook & Kelly 2007a). Sea urchins fed on the kelp diet, however, exhibited growth rates over 5 times higher than the urchins fed the two mussel diets during this period of "minimum" growth and this may be attributed to reduced prey-handling capability in P. lividus when feeding upon mussels at the lower temperatures, as previously observed in Strongylocentrotus droebachiensis (Scheibling & Hatcher 2007). In contrast, similar elevated growth rates for all treatments were observed between September and November 2005 and May and August 2006, which could be attributed to the increased algal biofouling that grew on the pearl nets during these periods.
The specific growth rates (SGR) of P. lividus red on the two size classes of mussels were not significantly different (0.25% [day.sup.-1]) and were approximately half the SGR exhibited by the sea urchins fed the kelp diet (0.48% [day.sup.-l]). This can be primarily attributed to differences in the accumulation of gonadal tissue, with the sea urchins fed the kelp diet exhibiting a gonad index of 3.0% compared with a gonad index of <0.6% for the sea urchins on the two mussel diets at the end of the experimental period. In contrast, the SGR of juvenile P. lividus held adjacent to caged Atlantic salmon, at twice the stocking density, was double the SGR exhibited by the sea urchins fed on the kelp diet in the present study (Cook & Kelly 2007b). This difference, however, may be attributed to loss of gonadal material through spawning, because the former study ended during the period of maximum GI for P. lividus and the present study ended towards the end of the spawning period when GI is typically at a minimum (Byrne 1990). Further investigation into changes in GI throughout the reproductive cycle using this co-culture system are required though to determine the most appropriate harvest time for the sea urchins and whether a preharvest diet was necessary.
The higher GI observed in the P. lividus fed on the kelp diet compared with the mussel diet, however, was unexpected, particularly because previous studies have suggested that diets with a higher protein value will support increased somatic and gonadal growth rates (Cook et al. 1998, Schlosser et al. 2005, Cook & Kelly 2007b). It is possible though, that with only 6.4% to 10.0% of the mussels broken or damaged in the mussel "waste" fed to the sea urchins, there was a greater energetic cost in handling mussels as a food item compared with grazing on the macroalgae, which may have contributed to the difference in gonad index.
The coloration of the roe was categorized as acceptable or excellent for the Laminaria spp. treatment group, however, because of minimal roe content in the other three treatment groups, roe coloration could not be determined. Dietary beta-carotene is the likely precursor to 9-cis echinenone, the primary pigment in the gonads of P. lividus (Symonds et al. 2007) and is found in algae (McDermid & Stuercke 2003). The similarity in CIE color values between P. lividus fed kelp in the present study and co-cultured with Atlantic salmon in a previous study also supports the suggestion that gonad coloration in the latter study was attributed to the ingestion of macroalgal biofouling (Cook & Kelly 2007a).
The significant loss of pearl nets from the long-line closest to the shore demonstrates that the buoyancy system used for supporting the long-lines on the mussel farm is critical is minimizing the loss of nets, with the 400-L barrel-like flotation buoys providing a far more stable long-line than the 25-L buoys. The high survivorship of P. lividus in the remaining pearl nets, however, suggests that the environmental conditions for growing mussels (i.e., relatively sheltered, but well flushed sites) and the long-line mussel system, which is established throughout Europe, Chile, and Australasia provides an ideal physical platform for growing sea urchins.
To conclude, the co-culture of sea urchins and the blue mussel, Mytilus edulis, provides a viable method for the on-growing of hatchery-reared urchins to market size. This study has shown that P. lividus with ah initial test diameter of 18-20 mm and stocking density of 65 individuals m 2 could reach market size (>55 mm test diameter) within two years, if fed on Laminaria spp. ad libitum. The existing mussel farming infrastructure provides an ideal platform to deploy the nets containing the sea urchins, although scaling up the system to a commercial scale will provide its own technical challenges. One of the main challenges will be the harvest and delivery of macroalgae to each of the urchin cages and basket design, so the sea urchins themselves can prevent fouling organisms from smothering the nets and reducing water flow. Further research is also required into the use of mussel waste with a higher proportion of damaged shells and to evaluate the maximum gonad index that can be obtained for P. lividus fed M. edulis in this type of system. The co-culture of sea urchins and mussels, however, provides the grower with a second commercially valuable product and if adopted in areas where long-line culture is practiced, could significantly reduce the pressure on wild sea urchin stocks.
This research was supported by European Union SPIINES2: Sea urchin production in integrated systems, their nutrition and roe enhancement, Project number 512627 and AAAG: Atlantic Arc Aquaculture Group (Interreg IIB), Project number 091--AAAG. The authors thank the Director of SAMS for the provision of facilities, Ian MacKinnon, the owner of Loch Beag Shellfish for his invaluable support and patience, and Hui Liu of the Yellow Sea Fisheries Institute, China for her invaluable help with the field work.
Agatsuma, Y. 1998. Aquaculture of the sea urchin (Strongylocentrotus nudus) transplanted from coralline flats in Hokkaido, Japan. J. Shellfish Res. 17:1541-1547.
Boudouresque, C. F. & M. Verlaque. 2007. Ecology of Paracentrotus lividus. In: Lawrence, J.M. (Ed.), Edible sea urchins: biology and ecology. Amsterdam: Elsevier. pp. 243-285.
Bridger, C. J., R. G. Hooper & T. J. McKeever. 1998. Pilot-scale commercial sea urchin roe enhancement: Ocean corral trials. Bull Aquacult. Assoc. Can. 98:99-101.
Briscoe, C. S. & K. P. Sebens. 1988. Omnivory in Strongylocentrotus droebachiensis (Muller) (Echinodermata, Echinoidea): Predation on subtidal mussels. J. Exp. Mar. Biol. Ecol. 115:1-24.
Byrne, M. 1990. Annual reproductive cycles of the commercial sea urchin Paracentrotus lividus from an exposed intertidal and a sheltered subtidal habitat on the west coast of Scotland. Mar. Biol. 104:275-289.
Cook, E. J. 1999. Psammechinus miliaris (Gmelin). Factors affecting its somatic growth and development, and its suitability as a species for sea urchin cultivation. PhD Thesis. Edinburgh: Napier University. pp. 168.
Cook, E. J. & M. S. Kelly. 2007a. Enhanced production of the sea urchin Paracentrotus lividus in integrated open-water cultivation with Atlantic salmon Salmo salar. Aquaculture 273:573-585.
Cook, E. J. & M. S. Kelly. 2007b. Effect of variation in the protein value of the red macroalga Palmaria palmata on the feeding and growth of the sea urchins Psammechinus miliaris and Paracentrotus lividus (Echinodermata). Aquaculture 270:207-217.
Cook, E. J., M. S. Kelly & J. D. McKenzie. 1998. Somatic and gonadal growth of the sea urchin Psammechinus miliaris (Gmelin) fed artificial salmon feed compared with a macroalgal diet. J. Shellfish Res. 17:1549-1555.
Cuthbert, F. M., R. G. Hooper & T. McKeever. 1995. Sea urchin Aquaculture Phase I: sea urchin feeding and ranching experiments. In: Government of Newfoundland, editors. Canadian Centre for Fisheries Innovation. Report No. AUI-503.
Fernandez, C. & C. F. Boudouresque. 2000. Nutrition of the sea urchin Paracentrotus lividus (Echinodermata: Echinoidea) fed different artificial food. Mar. Ecol. Prog. Ser. 204:131-141.
Fernandez, C. M. & A. Caltagirone. 1994. Growth rate of adult Paracentrotus lividus in a lagoon environment: the effect of different diet types. In: B. David, A. Guille, J. P. Feral, M. Roux, editors, Echinoderms through time. Rotterdam: Balkema. pp. 655-660.
Frantzis, A. & A. Gremare. 1992. Ingestion, absorption, and growth rates of Paracentrotus lividus (Echinodermata: Echinoidea) fed different macrophytes. Mar. Ecol. Prog. Ser. 95:169-183.
Grosjean, P., C. Spirlet, P. Gosselin, D. Vaitilingon & M. Jangoux. 1998. Land-based, closed-cycle echiniculture of Paracentrotus lividus (Lamarck) (Echinoldea: Echinodermata): A long-term experiment at a pilot scale. J. Shellfish Res. 17:1523-1531.
Hagen, N. T. 1996. Echinoculture: From fishery enhancement to closed cycle cultivation. World Aquacult. 27:6-19.
Haya, D. & M. B. Regis. 1995. Comportement trophique de Paracentrotus lividus (Lam.) (Echinodermata: Echinoidea) soumis a six regimes alimentaires dans des conditions experimentales. Mesogee 54:35-42.
James, P. J. 2006. A comparison of roe enhancement of the sea urchin Evechinus chloroticus in sea-based and land-based cages. Aquaculture 253:290-300.
Kelly, M. S. 2004. Sea urchin aquaculture: a review and outlook. In: Munchen-Heinzeller, Nebelsick, editors. Echinoderms. London: Taylor & Francis Group, pp. 283-289.
Kelly, M. S., C. C. Brodie & J. D. McKenzie. 1998. Somatic and gonadal growth of the sea urchin Psammechinus miliaris (Gmelin) maintained in polyculture with the Atlantic salmon. J. Shellfish Res. 17:1557-1562.
Liu, H., M. S. Kelly, E. J. Cook, K. D. Black, H. Orr, J. X. Zhu & S. L. Dong. 2007a. The effect of diet type on growth and fatty acid composition of the sea urchin larvae, II. Psammechinus miliaris (Gmelin). Aquaculture 264:263-278.
Liu, H., M. S. Kelly, E. J. Cook, K. D. Black, H. Orr, J. X. Zhu & S. L. Dong. 2007b. The effect of diet type on growth and fatty acid composition of sea urchin larvae, I. Paracentrotus lividus (Lamarck, 1816) (Echinodermata). Aquaculture 264:247-262.
McDermid, K. J. & B. Stuercke. 2003. Nutritional composition of edible Hawaiian seaweeds. J. Appl. Phycol. 15:513-524.
Meidel, S. K. & R. E. Scheibling. 1999. Effects of food type and ration on reproductive maturation and growth of the sea urchin Strongylocentrotus droebachiensis. Mar. Biol. 134:155-166.
Mendes, C. & R. Becarra. 2004. Cultivation of Loxechinus albus in lantern nets. In: J. M. Lawrence, O. Guzman, editors. Sea urchin fisheries and ecology. Lancaster, USA: DEStech Publications. pp. 374.
Moylan, E. 1997. Gonad conditioning and wild stock enhancement of the purple sea urchin Paracentrotus lividus on the west coast of Ireland. Bull. Aquacult. Assoc. Can. 97:38-45.
Otero-Villanueva, M. D. M., M. S. Kelly & G. Burnell. 2006. How prey size, type and abundance affects foraging decisions by the regular echinoid Psammechinus miliaris. J. Mar. Biol. Ass. UK. 86:773-781.
Robinson, S. M. & L. Colborne. 1997. Enhancing roe of the green sea urchin using an artificial food source. Bull. Aquacult. Assoc. Can. 97:14-20.
Robinson, S. M. C., J. D. Castell & E. J. Kennedy. 2002. Developing suitable colour in the gonads of cultured green sea urchins (Strongylocentrotus droebachiensis). Aquaculture 206:289-303.
Scheibling, R. E. & B. G. Hatcher. 2007. Ecology of Stronglyocentrotus droebachiensis. In: J. M. Lawrence, editors. Edible sea urchins: biology and ecology. Amsterdam: Elsevier. pp. 353-392.
Schlosser, S. C., I. Lupatsch, J. M. Lawrence, A. L. Lawrence & M. Shpigel. 2005. Protein and energy digestibility and gonad development of the European sea urchin Paracentrotus lividus (Lamarck) fed algal and prepared diets during spring and fall. Aquacult. Res. 36:972-982.
Shpigel, M., S. C. McBride, S. Marciano & I. Lupatsch. 2004. Propagation of the European sea urchin Paracentrotus lividus in Israel. In: J. M. Lawrence, O. Guzman, editors. Sea urchins: fisheries and aquaculture. Pennsylvania: DEStech publications, pp. 386.
Symonds, R. C., M. S. Kelly, C. Caris-Veyrat & A. J. Young. 2007. Carotenoids in the sea urchin Paracentrotus lividus: occurrence of 9'-cis-echinenone as the dominant carotenoid in gonad colour determination. Comp. Biochem. Physiol. B. Biochem. Mol. Biol. 148:432-444.
Underwood, A. J. 1997. Experiments in ecology: their logical design and interpretation using analysis of variance. Cambridge: Cambridge University Press. 522 pp.
Vadas, R. L., B. Beal, T. Dowling & J. C. Fegley. 2000. Experimental field tests of natural algal diets on gonad index and quality in the green sea urchin, Strongylocentrotus droebachiensis: a case for rapid summer production in post-spawned animals. Aquaculture 182:115-135.
Verlaque, M. & H. Nedelec. 1983. Biologie de Paracentrotus lividus (Lamarck) sur substrat rocheux en Corse (Mediterranee, France): alimentation des adultes. Vie Milieu 33:191-201.
E. J. COOK * AND M. S. KELLY
Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll, PA37 1QA, United Kingdom
* Corresponding author. E-mail: email@example.com
TABLE 1. (A) Shell length (mm; mean (SE)) and (B) Percentage (%) of damaged mussel shells for "large" (LM) and "small" (SM) Mytlius edulis sampled four times throughout the experimental period. P value highlighted in bold indicates statistically significant differences between LM and SM within columns (Paired t-test; n = 3). Sept 05 Nov 05 (A) LM 32.03 (5.25) 30.47 (5.27) SM 17.75 (4.69) 16.08 (3.80) t 10.61 112.0 p 0.009 0.000 (B) LM 7.11 (1.68) 8.94 (3.58) SM 9.85 (1.76) 9.96 (3.58) t -1.49 -0.39 p 0.275 0.731 Jan 06 Mar 06 (A) LM 31.23 (5.45) 30.90 (4.92) SM 22.26 (3.80) 17.78 (4.73) t 4.12 15.09 p 0.05 0.004 (B) LM 6.44 (2.57) 7.28 (0.81) SM 20.57 (1.60) 7.43 (3.61) t -4.65 0.07 p 0.043 0.952 TABLE 2. Effect of dietary treatment: large mussels (LM), small mussels (SM), kelp, Laminaria spp. (K) and no additional feed (NF) on survival and growth, including Linear Growth Rate (LGR) of the sea urchin Paracentrotus lividus (mean (SE), n = 6). Treatment Survival (%) Initial TD (mm) (excl. Line 1) LM 99.0 (2.0) (a) 18.4 (2.1) (a) SM 96.7 (7.5) (a) 16.9 (1.7) (a) K 100 (a) 17.4 (2.0) (a) NF 95.8 (4.5) (a) 17.6 (2.0) (a) Treatment Final TD (mm) LGR ([micro]m TD [day.sup.-1]) LM 29.0 (3.9) (b) 29.8 (4.5) (bc) SM 28.6 (3.0) (b) 36.2 (4.9) (b) K 40.5 (3.9) (a) 61.8 (2.4) (a) NF 23.9 (2.4) (c) 17.8 (4.4) (c) Means in the same column sharing the same superscript letter were not significantly different as determined by Tukey test (P > 0.05). TABLE 3. Effect of dietary treatment: large mussels (LM), small mussels (SM), kelp, Laminaria spp. (K) and no additional feed (NF) on weight and specific growth rates (SGR) of the sea urchin Paracentrotus lividus (mean (SE), n = 6). Treatment Initial WW (g) Final WW (g) SGR (% day (1)) Baseline 3.72 (0.56) LM 9.22 (2.09) (b) 0.24 (0.07) (b) SM 9.42 (1.85) (b) 0.25 (0.06) (b) K 21.90 (3.76) (a) 0.48 (0.05) (a) NF 5.88 (1.43) (c) 0.12 (0.05) (c) ANOVA F 340.4 255.3 p 0.000 0.000 Means in the same column sharing the same superscript letter were not significantly different as determined by Tukey test (P > 0.05).