Development and application of techniques for prediction of the scallop Pecten maximus (L.) spatfall.

ABSTRACT Over five million spat of the king scallop, Pecten maximus (L.) were collected accidentally in the North Water of Mulroy Bay in County Donegal, Ireland in the summer of 1979. This large natural settlement, by far the greatest number obtained at the time in any European trial, led to research investigations in 1980 and 1981 to support the development of this site as a scallop spat collection center. A combination of gonad monitoring, plankton sampling, spat settlement and spat collection trials were used to develop a technique for predicting the date of the scallop spatfall to maximize the yield of scallop spat. A technique for forecasting the date of peak spatfall involving installation of spat collectors when the mean shell length of scallop larvae was 180 [micro]m and 50% of the larval cohort exceeded this size was developed. Unlike other sites studied, peak spat collection occurred at 4 m beneath the water surface. Since its development, this technique for prediction of the date of the peak collection has been used commercially, and for illustrative purposes data from 2002 have been provided. The development and application of this forecasting technique has assisted in the collection of millions of scallop spat for on growing trials by the scallop culture industry.

KEY WORDS: Pecten maximus, prediction, scallop settlement, spat collection, spatfall

INTRODUCTION

Many commercial scallop fisheries currently exploited in European waters are protected against overfishing by legislation applied to fishing gear; in terms of the number, type and specification of dredges; to landings in terms of the quantity and minimum shell length of the catch and to the duration of the fishing season. All of these precautionary measures are designed to prevent existing resources becoming further diminished or potentially exhausted beyond repair and to allow time for the natural recovery of the resource.

Faced with similar restrictive practices, fishermen in Japan adopted a more proactive role towards conservation and implemented measures to expand resources by the transplantation or reseeding of aquatic animals. The development of methods for the mass production of juveniles of selected species in waters suitable for breeding and the subsequent release of these juveniles to the oceans has resulted in dramatically increased yields from several fisheries. Noteworthy in this context is the successful expansion of Japanese scallop cultivation, annual landings increasing from approximately 9,000 t during the 1960s to consistently over 500,000 t since 1995. The success of this approach to scallop fishery management and resource expansion encouraged many European countries to implement scallop research programs during the 1980s with a view to emulating this Success.

Early spat collection trials in France, Spain, Scotland, Isle of Man and England showed considerable annual fluctuations in the number of spat per collector and generally disappointing yields compared with results in Japan (Buestel et al. 1976, Pickett 1977, Ventilla 1977a, 1977b, Mason & Drinkwaater 1978, Brand et al. 1980, Roman et al. 1985). Scottish trials in which spat collectors were installed at regular intervals over a 4-6 wk period demonstrated the existence of an optimum date for peak spat collection. Temporal variations in the date of the peak spat collection over five years ranged from late June to early August (Ventilla 1977a, 1977b, Slater, 1978, 1979, 1980). In some years this peak could rise and fall within a one-week period and in other years be extended over a two- to three-week period (Ventilla 1977a, 1977b). Installing spat collectors in the sea either before or after the date of peak spatfall could dramatically reduce the yield of spat by up to 75%. Clearly for natural spat collection to be successful in the provision of an abundant and reliable source of seed for aquaculture and fishery enhancement, there was a need for a spatfall prediction technique similar to that used in Japan.

Despite significant research effort on king scallops since the early 80's commercial scallop cultivation in Europe has not developed and in countries where fishery landings have been expanded, such increases have resulted only from further exploitation of already overfished natural resources. The major limitation to the scale of any scallop farming or stock enhancement program remains as the availability of a consistently large and reliable source of juvenile scallops or spat. Worldwide in those countries where commercial scallop cultivation has developed, for example Japan, China and Chile, problems associated with the provision of a regular supply of scallop spat have first been overcome.

In 1979, a local aquaculture co-operative, the North Water Co-operative Soc. Ltd., during attempts to capture mussel spat, Mytilus edulis (L.), accidentally collected on 12-mm diameter polypropylene rope over 5 million scallop spat in the North Water of Mulroy Bay in County Donegal, by far the greatest number obtained at the time in any European trial (Minchin 1980). As a result of this exceptionally large settlement, research trials commenced to support the exploitation of the site as a scallop spat collection center.

Research during 1980 and 1981 focused on gonad monitoring, plankton analysis, spat settlement and spat collection trials to develop a technique for prediction of the date and location of the peak scallop spatfall. This study provides results from these investigations that led to the development of techniques for prediction of the date of peak spatfall and provides for illustrative purposes results obtained during commercial application of this predictive technique in 2002.

MATERIALS AND METHODS

The Location

Mulroy Bay is located in County Donegal in the northwest of Ireland (Fig. 1). It is a sheltered, marine lough connected to the open sea by a series of long and narrow channels. The Bay is well sheltered from excessive exposure and no oceanic swells are able to penetrate into the two main parts, the Broad Water and the North Water. Seawater enters the Bay through the Narrows and flows into the Broad Water before passing through the Moross Channel into a further and even more enclosed sea lough, the North Water, which is approximately 3 km by 1 km in size. Water depths extend to 51 m in the North Water and the bathymetry of the lough is available on Admiralty Chart No. 2698. Resulting from this topography, the tidal range within the Bay and water exchange with the open sea is much diminished. The Bay provides an ideal location for shellfish spat production, because larvae produced by natural populations in the Bay are retained and settle within the system rather than being dispersed as occurs in more open waters.

[FIGURE 1 OMITTED]

The sea lough is surrounded by poor quality land with a low population density supporting an agricultural community dominated by small farms. At the head of the lough is the village of Milford (population 1,334) that forms the gateway to both the Fanad Peninsula down the eastern side of Mulroy Bay and the Rosguill Peninsula down the western side of Mulroy Bay. The only significant employers in the area are Marine Harvest (Fanad), a salmon farming company with approximately 150 staff.

Gonad Monitoring

Samples of adult king scallops for gonad assessment were dredged from around the Stookans, Massmount Bay, North Water of Mulroy Bay using a lightweight converted oyster dredge (Fig. 1). In 1980 samples comprising 22-27 individuals of mean shell height 113.4 [+ or -] 1.8 mm were collected at intervals ranging from 4-13 days (mean = 8.4 days) between June 5 and August 30, 1980. In 1981 samples comprising 23-25 individuals of mean shell height 113.8 [+ or -] 1.2 mm were collected at intervals ranging from 5-10 days (mean = 6.8 days) between June 10 and July 8, 1981. Soft tissues were dissected from each scallop, pooled and weighed. Gonad index was determined according to the following formula:

Gonad Index (%) = Total wt. of gonad (g)/Total wt. of soft tissue (g) x 100/1

To account for minor variations in the size of scallops from week to week, the relative gonad height (RGH) was calculated using the equation:

RGH (g.[mm.sup.-3]) = Mean wt. of gonad in g/[(Mean shell ht. in mm).sup.3] x [10.sup.6]

Following the introduction into Irish legislation of the Mulroy Bay (North Water and Moross Channel) (Prohibition on Dredging for Escallops) Bye Law No. 625, 1982 to protect broodstock scallops in this site from exploitation and hence conserve the area as a national center for scallop spat collection, farmed scallops from a single year-class and of a size greater than 90 mm shell height were held at 5-m depth on longlines at Deegagh Point in the Broad Water of Mulroy Bay (Fig. 1) and used as a source of adult stock for gonad monitoring.

Larval Sampling

In 1980 plankton samples for monitoring the occurrence and development of scallop larvae were obtained from a site at Lurgacloghan on the western side of the North Water (Fig. 1) within the site licensed to Deegagh Point Shellfish Ltd. for scallop aquaculture. Samples comprising a single vertical haul were collected using a 63-[micro]m mesh plankton net with a 30-cm diameter opening, hauled at a rate of 5 m.[min.sup.-1], from 20 m to the water surface. All planktonic larvae were added to a 1-L, wide-mouthed, plastic jar containing 100 mL of formaldehyde to immediately preserve the living material. In the laboratory, samples were placed in a 1-L beaker and gently stirred. Bivalve larvae, being slightly heavier than most other planktonic organisms, rapidly accumulated in the center of the beaker as a result of the gentle centrifugal force and were transferred to 25 mL Sterilin[R] storage containers using a 10 mL pipette. Larvae were examined on a Sedgewick Rafter counting cell using x40 and x100 magnification on an Olympus KH series microscope. Scallop larvae were identified from their morphological features and measured using an eyepiece graticule at x100 magnification (Slater 2005b). Identification was possible as a result of expertise gained during hatchery rearing studies with P. maximus (L.).

The distribution of scallop larvae in the North Water was investigated by measuring larval abundance at eight sites in the North Water of Mulroy Bay. Samples comprising a single vertical plankton haul from 15 m to the water surface were collected on July 16, 1980 between 10.00 PM to 12.00 PM and repeated on July 18, 1980 between 9.00 PM AND 11.00 PM. Weather and sea conditions on both these occasions were ideal for quantitative work of this type.

In 1981 plankton samples for monitoring the occurrence and development of scallop larvae were obtained from the same site at Lurgacloghan on the western side of the North Water using similar techniques to those outlined earlier. A single vertical haul from 20 m to the water surface provided sufficient numbers of scallop larvae for analysis. From 1985 onwards, following use of TBT (tributyl-tin) antifoulants by salmon farms in the Bay, three vertical hauls were collected at this site to provide sufficient scallop larvae for analysis (Minchin et al. 1987).

Indicator Collectors

Indicator collectors were used to provide data on time and intensity of settlement of scallop spat. These consisted of a piece of 6-mm Netlon[R] mesh of 0.8 m x 0.3 m dimensions, rolled twice along the long axis into a cylindrical shape and fixed using cable ties.

In 1980 indicator collectors were attached to ropes at 4 m, 8 m and 12 m above the seabed and installed in a 15-m deep site at the north of Croaghan Island (Fig. 1). After 10-11 days immersion, these indicator collectors were exchanged for a new set of collectors for spat settlement. Bivalve spat settling during the 10- to 11-day period were removed from the collectors by immersion in 20% sodium hypochlorite in fresh water for 3 min. The solution was sieved through a 90-[micro]m mesh sieve to collect the recently settled bivalve spat (Davies 1974). Examination of bivalve spat was performed on a Sedgewick Rafter counting cell at x40 and x100 magnification. Scallop spat were identified and counted from each indicator collector.

In 1981 trials with indicator collectors were repeated in a 23-m deep site in Bally Hork Bay, the site exhibiting higher spat collection results than the north of Croaghan island site in 1980 (Fig. 1). Collectors were attached at 5 m, 10 m and 15 m beneath the water surface, irrespective of the tidal state, based on the 1980 data on variation in settlement intensity with depth. Because the 1980 results showed a small settlement in early June, the indicator collector trial in 1981 was extended to cover the period May 27 to September 4.

Spat Collection

Spat collectors used in 1980 consisted of polyethylene woven-filament mesh bags of 0.45 m x 0.45 m dimensions and mesh size 6 mm x 3 mm. Each mesh bag contained 2 m of monofilament nylon salmon netting provided by the North Western Regional Fisheries Board, following confiscation from illegal fishing activities.

Collector bags were attached to individually anchored 12-mm diameter polypropylene ropes at 2-m intervals from 2-10 m above the seabed to provide information concerning the optimal depth for collection in the 15-m deep site north of Croaghan Island. A 20-cm diameter subsurface float was attached to each rope at 11 m above the seabed to reduce the action of wind and waves on collector bags. A 30-cm diameter marker buoy was attached to the top of the rope to mark the position of the collector line. A total of nine such ropes were positioned in the site from July 2 to September 2, 1980.

Following the larval distribution study, two ropes, each with five collector bags attached as previously mentioned, were positioned on July 17, 1980 in Bally Hork Bay, the site with the highest larval numbers. Because of limited supplies of the polyethylene woven-filament collectors, collector bags at this site comprised 1 m x 0.5 m Netlon[R] oyster bags of 6 mm x 6 mm mesh size containing 6 m of monofilament mesh salmon netting with the ends of the bags closed at right angles to each other. The spat collection efficiency of this type of collector, which had been successfully used in previous Scottish trials, was compared with the polyethylene woven filament collector in the north of Croaghan island site. Sorting of all collector bags commenced from late October onwards and continued over a two-week period. Collector bags and inner net filler were individually washed into tanks of seawater and live scallop spat from each bag counted individually.

Following the 1980 trials, which demonstrated increased numbers of Pecten spat with increased distance from the seabed at two sites in the North Water, collector bags used during 1981 were positioned at 2-m intervals from 2 m to 10 m beneath the water surface. All collector bags used were of the polyethylene woven-filament type and were positioned in Bally Hork Bay, the more productive of the two sites in the North Water and four sites in the Broad Water investigated during 1980.

Spat collection data from 2002 were provided by the pioneering scallop farming company, Deegagh Point Shellfish Ltd., established in 1980 to commercially exploit king scallop spat collection in Mulroy Bay. Collector bags used for commercial spat collection were 0.8 m x 0.3 m dimensions and 5 mm x 3 mm mesh size and filled with 2 m x 1 m of lightweight garden netting (Netlon[R]) of 6 mm x 6 mm mesh size. All collector bags were suspended from 300 m longlines, submerged 4 m beneath the water surface and attached to the shore at one end in a site at Lurgacloghan (Fig. 1). Spat collectors were installed over a 3-day period at the date of peak spatfall determined using gonad analysis and scallop larval monitoring.

RESULTS

1980 Gonad Monitoring

A large partial spawning was recorded from June 23 to July 1, GI decreasing from 16.0% to 11.9%. Following a rematuration phase over one week, a second smaller partial spawning was recorded from July 6 to 13, with GI decreasing from 13.9% to 13.0%. This smaller second partial spawning was more clearly evident in the RGH results, which accounted for shell size, there being a 3-mm increase in shell length of scallops used for the assessment on July 13, compared with July 6. Following gonad filling during the remainder of July, a larger and more complete spawning commenced in early August and continued throughout that month (Fig. 2A).

[FIGURE 2 OMITTED]

The trend in results for gonad index (GI), which relates the gonad weight to the total soft tissue weight, was very similar to that for relative gonad height (RGH), which relates the gonad weight to the size of the adult shell (Fig. 2A). There was a highly significant correlation between GI and RGH ([F.sub.1,11] = 300.158, P < 0.001).

1980 Larval Monitoring

Larvae of the scallop, Pecten maximus, were first recorded on July 7 at a mean shell length of 148.5 [micro]m and were present in all samples until July 31. The number of scallop larvae measured in each sample ranged from 3-100 during this period. Larval growth rate was approximately 4.54 [micro]m.[day.sup.-1] (Fig. 2B). Extrapolation of the larval growth trendline to a shell length of 100 [micro]m, a size usually achieved in the hatchery at three days old, suggested that these larvae originated from a spawning on June 22, corresponding with the first partial spawning recorded by gonad monitoring between June 23 and July 1. The final result on July 31 of 212.7 [micro]m, based on only three larvae, showed a decline from the previous result on July 23 of 221.7 [micro]m. This decline was attributed to the settlement of larger larvae from the plankton at the same time as the few remaining planktonic larvae continued to grow.

Quantitative assessment of the distribution of scallop larvae on July 16 and July 18 showed greater abundance of scallop larvae in Bally Hork Bay (Fig. 3).

[FIGURE 3 OMITTED]

Scallop larvae on the two sampling dates had an average shell length of 194.5 [micro]m and 208.5 [micro]m respectively (i.e., near settlement size).

1980 Indicator Collectors

Low numbers of scallop spat settled on the first set of indicator collectors immersed on June 10 and lifted on June 24 in the site north of Croaghan Island, showing that some spawning activity had occurred prior to June 10. Peak settlement intensity occurred between July 13-23 with an average of 204 spat per collector (Fig. 2C). Settlement intensity varied with distance above the seabed in this 15-m deep site with the indicator collectors situated 8 m above the seabed (i.e., in midwater exhibiting the highest settlement intensity).

1980 Spat Collectors

Results of the spat collection trial at the site north of Croaghan Island, a site exhibiting only 7 and 3 scallop larvae on July 16 and July 18 respectively are provided in Table 1. Peak scallop spat collection occurred from July 11-16 with an average of 450 Pecten spat per collector (Fig. 2D). Given a 20-day larval life for this species, a typical value from hatchery production of this species, peak spat collection corresponded to the large partial spawning recorded from June 23 to July 1. A much smaller peak in spat collection on September 4 yielding approximately 50 Pecten spat per collector corresponded to the more complete August spawning. Spat collectors installed in Bally Hork Bay, the site with the highest larval counts on July 16 and July 18, yielded an average of 2,100 Pecten spat per collector, although these results are not comparable because different collectors were used in the two sites. In both sites increased numbers of Pecten spat were obtained with increased distance of the collector bag above the seabed (Fig. 4).

[FIGURE 4 OMITTED]

1981 Gonad Monitoring

A small partial spawning occurred between June 10-20, with GI decreasing from 14.3% to 13.3%, followed by a short rematuration phase and a second larger partial spawning from 26 June to July 2 with GI decreasing from 14.0% to 11.4% (Fig. 5A). Similar to 1980 there was a highly significant correlation between GI and RGH in 1981 ([F.sub.1,5] = 1083.71, P < 0.001).

[FIGURE 5 OMITTED]

1981 Larval Monitoring

Larvae of the scallop, P. maximus, were first recorded on June 20 at a mean shell length of 146 [micro]m (Table 2). On the July 7 size distribution analysis of larvae showed the presence of two larval batches (Fig. 6). The larval batch comprising larger-sized larvae was not recorded after July 13. Between July 13 to July 30 size distribution analysis showed the presence of only one larval batch.

[FIGURE 6 OMITTED]

Scallop larval growth rates were approximately 4.36 [micro]m.[day.sup.-1] for the first larval batch and 5.68 [micro]m.[day.sup.-1] for the second larval batch (Fig. 5B). Extrapolation of the larval growth trendlines to a shell length of 100 [micro]m, a size usually achieved in the hatchery at three days old, suggested that the first larval batch originated from a spawning on June 6, a time prior to the commencement of scallop gonad monitoring and the second larval batch originated from a spawning on June 25 corresponding with the spawning recorded using GI between June 26 to July 2. Above 200 [micro]m both larval batches exhibited deviations from the straight line best-fit attributed to the settlement of larger larvae from the plankton at the same time as remaining planktonic larvae continued to grow.

Larval shell length data from the more abundant second batch of larvae (Table 2) were analyzed in terms of the percentage of larvae greater than or equal to 170 [micro]m, 180 [micro]m, 190 [micro]m, 200 [micro]m, 210 [micro]m and 220 [micro]m on each sampling occasion between July 7 and 27 (Fig. 7). The objective of this analysis was the development of a predictive model similar to that used in Japan in which a critical size and a percentage greater than or equal to that size are used to identify the date of the peak spat collection.

[FIGURE 7 OMITTED]

1981 Indicator Collectors

Data from indicator collectors in Bally Hork Bay demonstrated that scallop spat settlement during May and June was negligible. Heavy settlement intensity occurred on the two sets of indicator collectors used during the period June 30 to July 20, with a mean settlement intensity of 688 and 849 spat per indicator collector respectively (Fig. 5C). Settlement intensity varied with distance from the water surface in this 23-m deep site with the indicator collectors situated 10 m below the water surface exhibiting the highest settlement intensity.

1981 Spat Collectors

Spat collection in Bally Hork Bay peaked from July 8-14, slightly earlier than in 1980 (Table 3, Fig. 5D). Collector bags immersed during this period yielded over 3,000 spat per collector. Spatfall corresponded with the spawning period identified by GI from June 26 to July 2 and the spawning date on June 25 estimated by extrapolation of the larval trendline from the second larval batch monitored over the period July 7-30. Peak scallop spat collection at the Bally Hork Bay site in 1981 occurred 4 m beneath the water surface (Fig. 8).

[FIGURE 8 OMITTED]

Summary of the 1980 and 1981 Experimental Programs

The 1980 experimental program demonstrated that a spawning from June 23 to July 1 resulted in larvae that were first identified on July 7 at a mean shell length of approximately 150 [micro]m. Larval growth rate was 4.54 [micro]m.[day.sup.-1]. Spat collection intensity peaked from July 11-16, a time at which the mean shell length of larvae was between 174 [micro]m to 194.5 [micro]m. Increased spat collection intensity occurred in Bally Hork Bay, a site in which the highest larval concentration was recorded, compared with the north of Croaghan Island site. In both sites increased spat collection occurred with increased distance from the seabed.

The 1981 experimental program reported two spawnings from June 10 to 20 and June 26 to July 2 resulting in larvae from each that were first identified on June 20 and July 7 respectively at a mean shell length of approximately 150 [micro]m. Larval growth rate for the two larval batches were 4.36 [micro]m.[day.sup.-1] and 5.68 [micro]m.[day.sup.-1]. Spat collection intensity in Bally Hork Bay peaked from July 8 to 14, a time at which the mean shell length of larvae was between 156 [micro]m to 194.5 [micro]m. Scallop spat collection peaked 4 m beneath the water surface.

Both experimental programs demonstrated that peak spat collection could be predicted using a combination of gonad monitoring and larval development. Gonad monitoring demonstrated that a spawning had occurred, though this did not necessarily infer that significant numbers of larvae has been produced or that spat collection would occur. Larval development was more useful as a means of spatfall prediction, because it focused on the developmental stage immediately prior to settlement. Based on the data from the two years of experimental investigation a hypothesis was proposed that the date of peak spat collection occurred when the mean shell length of larvae was 180 [micro]m (Fig. 5B) and 50% of the larval cohort exceeded 180 [micro]m (Fig. 7). Using these two models the dates for peak spat collection in 1981 would have been predicted as the July 11-12 by the mean shell length model and July 11-12 by the percentage greater than 180-[micro]m model. The installation of collector bags on these dates would have resulted in excellent yields of scallop spat in 1981.

Commercial application of the technique in 2002

Since its development in the early 80s these techniques for prediction of the date of the scallop spatfall have been used commercially in the North Water of Mulroy Bay. Data from the 2002 commercial collection of spat have been provided for illustrative purposes. Monitoring of GI and RGH of scallops indicated that a partial spawning occurred from July 4 to July 11, 2002 with GI decreasing from 15.6% to 14.0% (Fig. 9A).

[FIGURE 9 OMITTED]

Low numbers of larvae of the scallop, Pecten maximus, were recorded on June 24, the first day of sampling at a mean shell length of 133.6 [micro]m (Table 4). The growth rate of these larvae was approximately 4.19 [micro]m.[day.sup.-1] (Fig. 9B). Extrapolation of the larval growth trendline suggested that this larval batch originated from a spawning on June 13, a time prior to the commencement of gonad monitoring. Size distribution analysis of larvae collected on July 9 showed the presence of an abundant larval batch of shell length 120 [micro]m to 140 [micro]m and the presence of at least one less abundant cohort corresponding to an earlier spawning(s). Microanalysis of the size distribution data of samples from July 15 to July 22 could be used to suggest that these scallop larvae represented one, two, three or even four larval cohorts with overlapping size distributions. However, because each larval size distribution in this period represented from 28-100 larval measurements, a number considered too low for microanalysis of data, this was not performed in this study (Fig. 10). Instead larvae from July 9 and July 12, samples have been considered as comprising two larval batches and larvae from July 15 onwards as one larval batch. The growth rate of larvae in the second, more abundant, larval batch was approximately 5.04 [micro]m.[day.sup.-1] (Fig. 9B). Extrapolation of the larval growth trendline suggested that the second larval batch originated from a spawning on June 30, somewhat earlier than the July 4-11 spawning period identified using GI monitoring. Growth rate of the second larval batch, as in 1980 and 1981, decreased above 200 [micro]m and was attributed to the settlement of larger larvae at the same time as the remaining planktonic larvae continued to grow. Using the prediction model developed in 1981, date of peak spat collection in 2002 was forecast as July 19, using both the mean shell length model (Fig. 9B) and the 50% greater than 180 [micro]m model (Fig. 11).

[FIGURES 10-11 OMITTED]

Commercial scallop spat collection yields in excess of 3,000 per collector bag were obtained from collector bags installed on July 19, 2002 and sorted in October 2002.

DISCUSSION

Determination of the spawning time of the scallop, P. maximus and thus the time at which larvae should be present in the water can be elucidated using either GI or RGH calculated from pooled tissue weights and mean shell size. The benefit of using more than one measure of reproductive condition was evident from the results of the second spawning in 1980; the small difference in GI results being more clearly evident in the RGH results, which accounted for the increased shell length between the samples. Although individual tissue weights and shell sizes would have provided a measure of the dispersion in reproductive condition within the sample, analytical equipment required for individual measurement was not available in 1980 and 1981. Frequent gonad monitoring at regular weekly intervals was used to ensure the provision of meaningful data. Partial spawnings during the summer, which can be all important in terms of scallop spat production, may be of short duration and be followed by rapid rematuration to pres-pawning levels (Slater 2005a). Sampling intervals greater than weekly could fail to record such partial spawnings. Even with a weekly sampling interval, gonad monitoring alone is not suitable as a tool for spatfall prediction. For example, in 1980 three spawnings were recorded over the duration of the sampling program, however significant numbers of scallop larvae were only recorded from the larger of the two partial spawnings occurring between late June to early July. These results demonstrate one inherent weakness of using gonad monitoring alone, namely that based on gonad results it would be reasonable to assume that three spawnings should have resulted in three larval batches and hence three spatfalls, with the larger more complete spawning in late August possibly providing the largest spat collection.

Scallop larval growth rates under natural conditions of 4.54 [micro]m.[day.sup.-1] in 1980, 4.36 [micro]m.[day.sup.-1] and 5.68[micro]m.[day.sup.-1] for the two batches in 1981, and 4.19 [micro]m.[day.sup.-1] and 5.04 [micro]m.[day.sup.-1] in 2002 were comparable with figures reported from hatcheries (Comely 1972, Gruffydd & Beaumont 1972, Le Pennec 1974, 1978). Back extrapolation of the larval growth rate trendline can be used to estimate the spawning date. Comparison of spawning dates provided by gonad monitoring with those provided by back extrapolation of the larval growth rate revealed a second weakness of the former, namely that GI/RGH identified a "spawning period" within which spawning was assumed to have occurred, whereas back extrapolation of the larval growth rate provided a "spawning date." The occurrence of a "spawning date" prior to the commencement of the "spawning period" (Table 5) may occur as an artifact of the sampling frequency used in gonad monitoring, spawning having already commenced before the high GI result at the commencement of the "spawning period" but not having been recorded because of the sampling frequency.

In summary, gonad monitoring can at best be used to provide an indication of the time period when larvae may be present, whereas larval monitoring can be used to provide confirmation that spawning has occurred and that scallop larvae have been produced and are developing normally.

To obtain an indication of the spatial distribution of scallop larvae, numbers of scallop larvae in a single vertical haul from 15 m to the water surface at eight sites in the North Water were determined. Other studies have reported that for quantitative work, plankton samples collected with a pump rather than plankton net provide a better estimate of abundance, however such equipment was not available for this study. Although significant variation in larval numbers between hauls have been reported in the literature, a single haul in the late evening was used in this investigation to allow rapid examination of all samples and the installation of collectors the following day in the site showing highest larval abundance. Quantitative analysis of the numbers of scallop larvae in a single vertical haul in eight sites demonstrated similar distribution patterns in the North Water on two dates in July 1980 (Fig. 3). The higher spat collection intensity in Bally Hork Bay compared with north of Croaghan Island, the former site having the highest larval abundance on both of the quantitative sampling dates, suggested that quantitative larval assessment using a single vertical haul with a plankton net could be used to provide an indication of larval distribution in this sheltered location and selection of more productive areas for scallop spat collection. A similar study into the distribution of larvae of Patinopecten yessoensis in Mutsu Bay, Japan reported that after initial widespread distribution in all regions, larvae accumulated in the eastern part of the Bay where spat settlement was highest (Ito et al. 1975). Changes in larval spatial distribution as larvae increased in size were attributed to regional water currents (Ventilla 1980).

The series of indicator collectors immersed for 10-day periods throughout the June to August period in 1980 exhibited three peaks in settlement, a small peak occurring in early June, the main settlement peak occurring in mid July and a small peak occurring in late August. In 1981, the main settlement peak again occurred in mid July, however the early June peak appeared to have moved to late June and the late August peak was of a smaller size. Experimental results for these two years demonstrated the dominant mid July peak settlement compared with other peaks in settlement activity. The importance of installing collectors at the time of maximum settlement is clearly exhibited in the 1980 results with a 9-fold increase at peak settlement time compared with the previous 10-day period and a 4-fold increase compared with the 10-day period after peak settlement. Similar 5-fold changes in spat settlement intensity within a 10-day period of the peak settlement have been reported with Japanese scallop, P. yessoensis in Mutsu Bay (Kanno 1970, Ito 1977). Such wide variation in spat settlement intensity over a short time period either side of the peak settlement raises doubts about the validity of results of much of the published work undertaken with natural spat collection where the poor settlements reported in some years could be attributed to the installation of collectors in the wrong time period (Mason 1969, Minchin 1981, 1983, Brand et al. 1980, Cashmore et al. 1998, McDonough 1998, Maguire & Burnell 1999).

Indicator collectors have no role in prediction of the time of the scallop spatfall because their results are only available postsettlement. Nevertheless, as a tool for optimizing site selection, for investigating peak settlement depths or providing an early indication of settlement intensity as a means of estimating equipment requirements for intermediate culture they can be of significant value.

Spat collection results from collector bags in 1980 and 1981 reflected the pattern of results obtained with indicator collectors. A sharper peak in spat collection intensity occurred in 1980 compared with 1981 and is attributed to movement of the early June settlement peak to late June during the 1981 season. The 1980 and 1981 results from Bally Hork Bay and the commercial result from 2002 demonstrated that the North Water of Mulroy Bay in County Donegal has significant development potential as a scallop spat collection center and is capable of exceeding the 1,000 spat per collector bag level identified in Japan as the standard for good and poor collection years (Taguchi & Walford 1976).

The experimental results obtained in 1980 and 1981 provided the foundation for a scallop spatfall prediction technique for use in Mulroy Bay. Since its development, the technique has been applied commercially resulting in the provision of millions of scallop spat for ongrowing trials by the fledgling scallop culture industry in Ireland. The prediction technique involved the installation of spat collectors when the mean shell length of scallop larvae was 180 [micro]m and 50% of the larval batch exceeded this size. By comparison, in Mutsu Bay with the Japanese scallop, P. yessoensis, the time recommended for the installation of collectors to correspond with peak collection is when 50% of umbonate larvae exceed 200 [micro]m in size (Ventilla 1980). In previous studies in Scotland with P. maximus, a mean larval shell length of 200 [micro]m was shown to correspond with the peak spat collection of this scallop (Slater 1980). These differences in the critical size, even with the same species, are attributed to the difference in settlement size of the larvae, settlement in Mulroy Bay occurring at approximately 220 [micro]m compared with 245-250 [micro]m in Scotland and 280-300 [micro]m with P. yessoensis in Japan.

Other techniques for prediction of the scallop spatfall in Japan have been developed. Yamamoto (1964) reported an association between the spatfall time and cumulative water temperatures over 4[degrees]C, which allowed prediction of the scallop spawning period and thus the spat settlement period. Ito et al. (1975) reported a relationship between the date of the scallop spatfall and the blooming of cherry blossom trees in Gappo Park, Aomori City. Because the Meteorological Department issued the blooming forecast one month in advance for the different regions of Japan, spat settlement dates could also be predicted. Although reported in the literature, these methods are capable of providing only a general guide to the settlement date (Ventilla 1980).

Peak collection of scallop spat in Bally Hork Bay, a site with a water depth of 23 m, occurred 4 m beneath the water surface in 1981. By contrast, most other studies have reported peak scallop spat collection with this species nearer the seabed (Buestel et al. 1976; Ventilla 1977a; 1977b; Slater 1978; 1979; 1980; Brand et al. 1980; Thouzeau 1991) or in midwater (Burnell et al. 1991, Maguire & Burnell 1999). Comparison of the peak depth of settlement on indicator collectors and the peak depth of collection on collector bags did not provide any clear explanation for the difference in depth profile between Mulroy Bay and the other sites reported in the literature. One factor that may contribute to the differences between sites may be the calmer less turbulent waters in Mulroy Bay, resulting in a reduced loss of spat nearer the water surface by comparison with the more open water sites reported elsewhere (Pearce et al. 1998). Recently the intensity of water column stratification, which might be more intense in calmer waters, was shown to influence the depth distribution of scallop larval settlement (Pearce et al. 2004).

On a note of caution regarding the commercial application of data of this nature, difficulties in the management of this unique scallop resource resulted when some of the results from the 1980 and 1981 investigations were made available. For example, local awareness of the fact that scallop larval concentrations were highest in Bally Hork Bay resulted in an over-focus on this area, with every scallop farming entrepreneur demanding access to this site. This resulted in the installation of longlines and collector bags across each other, threatening behavior between potential farmers, malicious damage to equipment and unfortunately compensation claims for damages through the courts. Good publicity directed at the potential of scallop farming combined with adverse media coverage focused on the court actions attracted the attention of commercial divers and resulted in the illegal removal of much of the scallop broodstock, despite the fact that diver removal of shellfish is prohibited in Ireland. The diminished broodstock population resulting from diver removal coupled with the effects of poor recruitment caused by the use of TBT antifoulants on salmon cage nets during the mid 1980s (Minchin et al. 1987) resulted in reduced numbers of scallop larvae and necessitated the collection of three vertical plankton hauls to ensure sufficient larval numbers in plankton samples for application of the prediction technique. Despite these difficulties the spatfall prediction techniques developed have been demonstrated over the last 20 years to be an effective technique for prediction of the date of peak scallop spat collection.

ACKNOWLEDGMENTS

Beirtreach Teoranta, a subsidiary of Udaras na Gaeltachta, the development agency for the Irish-speaking regions of Ireland, financed the experimental programs in 1980 and 1981. Deegagh Point Shellfish Ltd, a private company, supported at different times by financial grant assistance from Bord Iascaigh Mhara, the Irish Sea Fisheries Board and Udaras na Gaeltachta, provided results on the commercial application of the spatfall prediction technique. The Marine Institute in awarding post-doctoral fellowship number PDOC/01/004 to Dr. Gavin Burnell, University College Cork provided the impetus for the author to prepare this manuscript in part fulfilment of the requirements for a Ph.D. postgraduate award.

LITERATURE CITED 

Brand, A. R., J. D. Paul & J. N. Hoogesteger. 1980. Spat settlement of the scallops Chlamys opercularis (L.) and Pecten maximus (L.) on artificial collectors. J. Mar. Biol. Ass. UK 60:379-390.

Burnell, G., M. Barnett, T. O'Carroll & V. Roantree. 1991. Scallop spat collection and ongrowing trials in southwest Ireland. In: P. Lubet, J. Barret & J.C. Dao, editors. Acres de Colloques, No. 17. Presented at the 8th Int. Pectinid Workshop, Cherbourg, France. May 22 to 29, 1991. pp. 139-145.

Buestel, D., J.C. Dao & G. Lemarle. 1976. Collecte de naissain de pectinids en Bretagne. International Council for the Exploration of the Seas. C.M. 43. pp. 13.

Cashmore, D., M. McG. Learmouth & J. T. MacMillan. 1998. improving the efficiency of wild Pecten maximus spat collection: potential effects of spat bag design and of species temporarily settling in spat bags. Aquaculture 160:273-282.

Comely, C. A. 1972. Larval culture of the scallop, Pecten maximus (L.). J. Cons. Int. Explor. Mer. 34:365-378.

Davies, G. 1974. A method for monitoring the spatfall of mussels (Mytilus edulis [L.]). J. Cons. Int. Explor. Mer. 36(1):24-34.

Gruffydd, L. L. & A. R. Beaumont. 1972. A method for rearing Pecten maximus larvae in the laboratory. Mar. Biol. 15:350-355.

Ito, S., H. Kanno & K. Takahashi. 1975. Some problems on culture of the scallop in Mutsu Bay. Bull. Mar. Biol. Station of Asamushi. 15:89-100.

Ito, S. 1977. Present techniques and problems on the culture of scallops in Mutsu Bay. Presented at the 5th Japan-Soviet Joint Symposium on Aquaculture, September 1976, Tokyo and Sapporo.

Kanno, H. 1970. On relationship between the occurrence of swimming larvae of the scallop at the coast of Okunai and the degree of spat setting at the same place. The Aquiculture 17:121-134. (in Japanese)

Le Pennec, M. 1974. Morphogenese de la coquille de Pecten maximus (L.) eleve au laboratoire. Cahiers de Biologie Marine. 15:475-482.

Le Pennec, M. 1978. Genese de la coquille larvaire et postlarvaire chez divers bivalves marins. These de Doctorat d'etat, Universite de Bretagne Occidentale, Brest, France. 219 pp; 108p1.

Maguire, J. & G. M. Burnell. 1999. The potential for scallop spat collection in Bantry Bay, Ireland. Biology and Environment. Proc. R. Ir. Acad. [B] 99B: 183-190.

Mason, J. 1969. The growth of spat of Pecten maximus (L.). International Council for the Exploration of the Sea. C.M. 1969/K:32.

Mason, J. & J. Drinkwater. 1978. The settlement and early growth of the scallop, Pecten maximus (L.) and the queen, Chlamys opercularis (L.) in Scottish waters. 2nd Pectinid Workshop, Brest, France. May 8-13, 1978.

McDonough, N. A. 1998. An assessment of current scallop cultivation techniques with particular reference to Strangford Lough, Northern Ireland. Ph.D. Thesis. Queen's University Belfast.

Minchin, D. 1980. The escallop Pecten maximus L. in Mulroy Bay. Presented at 3rd Scallop Workshop, Port Erin, Isle of Man. May 13-16, 1980.

Minchin, D. 1981. The escallop Pecten maximus (L.) in Mulroy Bay. Irish Fisheries Research Report. 1:1-21.

Minchin, D. 1983. The escallop Pecten maximus (L.) in Mulroy Bay--Part 2. Presented at the 4th International Pectinid Workshop, Aberdeen. May 10-13, 1983.

Minchin, D., C. Duggan & W. King. 1987. Possible influence of organotins on scallop recruitment. Mar. Pollut. Bull. 18(11):604-608.

Pearce, C. M., S. M. Gallager, J. L. Manuel, D. A. Manning, R. K. O'Dor & E. Bourget. 1998. Effect of thermoclines and turbulence on depth of larval settlement and spat recruitment of the giant scallop Placopecten magellanicus in 9.5 m deep laboratory mesocosms. Mar. Ecol. Prog. Ser. 165:195-215.

Pearce, C. M., J. L. Manuel, S. M. Gallager, D. A. Manning, R. K. O'Dor & E. Bourget. 2004. Depth and timing of settlement of veligers from different populations of giant scallop Placopecten magellanicus in thermally stratified mesocosms. J. Exp. Mar. Biol. Ecol. 312:187-214.

Pickett, G.D. 1977. Artificial collection of pectinid spat: preliminary experiments in Start Bay, southwest England 1975 to 1977. I.C.E.S.C.M. 1977/E 49. pp. 1-6.

Roman, G., J. Cano & T. Garcia. 1985. A first trial with pectinid collectors in Malaga (S.E. Spain). Proc. 5th Pectinid Workshop, La Coruna, Spain. May 6-10, 1985.

Slater, J. 1978. Collection of naturally released scallop spat 1977. White Fish Authority Field Report No. 597.

Slater, J. 1979. Scallop spat collection off the Northern Ardnamurchan coast 1978. White Fish Authority Field Report No. 693.

Slater, J. 1980. Prediction of the time of peak settlement of the scallop, Pecten maximus by pectinid larval monitoring. Presented at the Scallop Workshop, Port Erin, Isle of Man. May 13-16, 1980.

Slater, J. 2005a. Spawning of king scallops, Pecten maximus (L.) in Mulroy Bay and the relationship with spatfall intensity. J. Shellfish Res. 24(4):951-958.

Slater, J. 2005b. Morphological identification of larval king scallops, Pecten maximus (L.) from natural plankton samples. J. Shellfish Res. 24(4):937-950.

Taguchi, K. & J. Walford. 1976. Techniques and economics of Japanese scallop culture in Mutsu Bay, Aomori Prefecture. Scallop Workshop, Baltimore, Ireland. May 11-16, 1976.

Thouzeau, G. 1991. Experimental collection of post larvae of Pecten maximus (L.) and other benthic macrofaunal species in the Bay of Saint Brieuc, France I. Settlement patterns and biotic interactions among species collected. J. Exp. Mar. Biol. Ecol. 148:159-179.

Ventilla, R. F. 1977a. A scallop spat collector trial off the Northern Ardnamurchan coast. White Fish Authority Field Report No. 485.

Ventilla, R. F. 1977b. Further investigations into the collection of natural scallop spat off the Northern Ardnamurchan coast. White Fish Authority Field Report No. 536.

Ventilla, R. 1980. The Scallop Industry in Japan. White Fish Authority Technical Report No. 175. August 1980.

Yamamoto, G. 1964. Studies on the propagation of the scallop, Patinopecten yessoensis (Jay) in Mutsu Bay. Fish. Res. Bd. Can. Translation Series No. 1054.

JOHN SLATER *

Letterkenny Institute of Technology, Department of Science, Port Road, Letterkenny, County Donegal, Ireland

TABLE 1.
The number of king scallop spat from collector bags installed between
2-10 m above the seabed at the site north of Croaghan Island between
July 2 to September 8, 1980.

                        Number of Scallop Spat per Collector Bag
Distance Above
Seabed (m)          Jul 2   Jul 11   Jul 16   Jul 18   Jul 21    Aug 2

10                    775      920      845      380      200       21
8                     440      685      455      270      145       22
6                     355      335      440      240      165        6
4                     135      320      145      160       90        5
2                     100      150      170      110       65        2
Total number
  from each date     1805     2410     2055     1160      665       56
Mean number per
  collector           361      482      411      232      133       11

                    Number of Scallop Spat
                      per Collector Bag          Total
Distance Above                                Number from
Seabed (m)         Aug 18   Sept 4   Sept 8   Each Depth

10                     43       41       31      3256
8                      38       63       30      2148
6                      36       51       41      1669
4                      16       66       36       973
2                      20       40       20       677
Total number
  from each date      153      261      158
Mean number per
  collector            31       52       32

TABLE 2.
Abundance and mean shell length data of two batches of scallop
larvae P. maximus in the North Water of Mulroy Bay during 1981.

                      Larval Batch 1    Larval Batch 2

           Ref. Day        Mean Shell         Mean Shell
             from            Length             Length
Date        Jun 01    n    ([micro]m)   n     ([micro]m)

01.06.81      1
20.06.81      20       8      146        --       --
22.06.81      22      19      155        --       --
24.06.81      24      22      166        --       --
26.06.81      26       8      175        --       --
28.06.81      28      59      178        --       --
28.06.81      30       9      190        --       --
02.07.81      32      16      200        --       --
03.07.81      33      16      202        --       --
07.07.81      37      12      206        88      154
08.07.81      38       8      210        92      156
10.07.81      40       6      217       144      168
13.07.81      43      --       --       100      186
15.07.81      45      --       --       144      198
18.07.81      48      --       --        25      200
20.07.81      50      --       --        70      201
24.07.81      54      --       --        28      196
27.07.81      57      --       --        79      202
30.07.81      60      --       --        15      197

TABLE 3.
The number of king scallop spat from collector bags installed between
2-10 m below the water surface at the site in Bally Hork Bay between
July 3-27, 1981. (N/A = Collector bags missing at the time of
retrieval.)

                     Number of Scallop Spat per Collector Bag
Distance Below
Water Surface (m)    Jul 3    Jul 8   Jul 10   Jul 13   Jul 15

2                     3060     4060     3260     3865      N/A
4                     3010     3640     3320     5070      N/A
6                     2905     3530     3320     3900     2670
8                     2540     3275     3230     3080     1970
10                    2750     2430     2135     2460     1555
Total number from
  each date          14265    16935    15265    18375     6195
Mean number per
  collector           2853     3387     3053     3675     2065

                    Number of Scallop Spat
                      per Collector Bag        Total Number
Distance Below                                  from Each
Water Surface (m)   Jul 18   Jul 20   Jul 27      Depth

2                      N/A     1355      620      16220
4                      N/A     2020      790      17850
6                      N/A     1730      930      18985
8                      N/A     1220      860      16175
10                    1080     1040      765      14215
Total number from
  each date           1080     7365     3965
Mean number per
  collector           1080     1473      793

TABLE 4.
Abundance and mean shell length data of two batches of scallop
larvae P. maximus in the North Water of Mulroy Bay during 2002.

                       Larval Batch 1     Larval Batch 2

           Ref. Day         Mean Shell         Mean Shell
             from             Length             Length
Date        Jun 01     n    ([micro]m)    n    ([micro]m)

24.06.02      24      29      133.6       --        --
01.07.02      31       7      158.3       --        --
09.07.02      39       5      196.2       95     131.9
12.07.02      42       5      196.4       95     143.5
15.07.02      45      --        --        28     153.8
17.07.02      47      --        --       100     167.4
18.07.02      48      --        --        84     175.4
19.07.02      49      --        --        63     183.1
22.07.02      52      --        --       100     191.9
24.07.02      54      --        --        66     206.5
26.07.02      56      --        --        40     202.6
01.08.02      62      --        --        13     223.8
06.08.02      67      --        --         9     214.9
09.08.02      70      --        --        13     215.5
13.08.02      74      --        --         7     219.6

TABLE 5.
Summary of scallop spawning "dates" using larval growth
extrapolation and spawning "periods" using gonad index in 1980,
1981 and 2002.

                Spawning "Date"
               from Larval Growth       Spawning "Period"
                 Extrapolation          from Gonad Index

1980 larvae         June 22         June 23-July 1
1981 larvae
  (batch 1)          June 6         Before sampling commenced
1981 larvae
  (batch 2)         June 25         June 26-July 2
2002 larvae
  (batch 1)         June 13         Before sampling commenced
2002 larvae
  (batch 2)         June 30         July 4-July 11