Development of longline mussel farming and the influence of sleeve spacing in Prince Edward Island, Canada

Aquaculture 281 (2008) 56–62

Contents lists available at ScienceDirect

Aquaculture j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / a q u a - o n l i n e

Development of longline mussel farming and the influence of sleeve spacing in Prince Edward Island, Canada L.A. Comeau a,⁎, A. Drapeau b, T. Landry a, J. Davidson b a b

Department of Fisheries and Oceans, Gulf Fisheries Center, P. O. Box 5030, Science Branch, Moncton, New Brunswick, Canada E1C 9B6 Department of Health Management, Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3

A R T I C L E

I N F O

Article history: Received 23 October 2007 Received in revised form 23 May 2008 Accepted 26 May 2008 Keywords: Mytilus edulis Mussel Farming Husbandry Longline Carrying capacity

A B S T R A C T This paper describes the historical development of blue mussel (Mytilus edulis) farming in Tracadie Bay (Prince Edward Island, Canada) and relates the spacing of suspended sleeves (SS) on longlines to seston uptake. From 1990 to 2001, mussel biomass in Tracadie Bay increased by a factor of four (from 1137 to 4743 t). By 2001, seston uptake rates were approximately three-fold water renewal rates, and harvest yields (kg sleeve− 1) were significantly lower than in the early 1990s. A one-year field experiment was carried out to determine whether a change in SS could restore harvest yields. We found that SS (10, 20, 40, 60, 80 cm) had no significant effect on the condition index of mussels. However, high SS positively affected shell growth and abundance for small seeds that were densely packed within sleeves. A complete husbandry shift toward high SS and high seeding densities (within sleeves) may enhance farm productivity (production per unit of effort) and curtail seston uptake at the bay scale. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved.

1. Introduction Bivalves are filter-feeding organisms that extract suspended food particles from the water column with an extraordinary filtration capacity. Dense bivalve beds can deplete available food resources faster than the ecosystem can replace them through primary production and tidal currents (Wildish and Kristmanson, 1979; Fréchette and Bourget, 1985a,b; Newell, 1990; Dolmer, 2000; Petersen, 2004). The determination of optimal rearing densities for bivalve aquaculture has become an important—and increasingly studied— issue. A predictable impact of overstocking bivalves is a reduction in farm productivity due to curtailed growth and delayed harvests (e.g., Aoyama, 1989; Héral et al., 1986; Boromthanarat and Deslous-Paoli, 1988). Therefore, there is a need to identify the stocking density at which the demand for food particles is well matched to the supply. This optimal density level is often referred to as the production carrying capacity (PCC) of a system (McKindsey et al., 2006). Using mathematical models optimal density was estimated to be approximately 0.26 oysters m− 3 in the Carlingford Lough in Ireland (Ferreira et al., 1998) and 50 scallops m− 3 in Sungo Bay in China (Bacher et al., 2003). The most sensitive model input parameters are normally bivalve biomass, phytoplankton turnover rates and hydrographic data. However, one key element that is currently not well

⁎ Corresponding author. Tel.: +1 506 851 2723; fax: +1 506 851 2079. E-mail address: [email protected] (L.A. Comeau).

parameterized is the influence of gear configuration on the PCC. Current models assume that individual bivalves are homogeneously distributed within an embayment (or sub-system), a simplification that discounts any effect of gear configuration on bivalve feeding behavior. In a recent review of carrying capacity models, McKindsey et al. (2006) highlighted this shortcoming and recommended the inclusion of rearing techniques (e.g. gear configuration) to improve veracity in model simulations and predictions. For now, modeling the influence of gear configuration is hindered by a scarcity of documented information on the subject. The effect of gear configuration on seston depletion rates is still poorly defined both in quantitative and predictive terms. To date, the limited research on this topic has been directed almost exclusively toward a better understanding of the mussel raft culture (e.g. Cabanas et al., 1979; Navarro et al., 1991; Boyd and Heasman, 1998; Heasman et al., 1998; Fuentes et al., 2000). There is an emerging consensus that the close spacing of mussel ropes (1 rope m− 2) results in localized food reduction, such that mussels at the downstream end of the raft are left with fewer food particles to consume than upstream mussels. However, such generalizations are not straightforward for submerged longline systems. Reports of food reductions associated with longline culture show variable results (Rosenberg and Loo, 1983; Fréchette et al., 1991; Ogilvie et al., 2000; Pilditch et al., 2001; Strohmeier et al., 2005), perhaps because of major differences in stocking density and longline configuration between study sites. Sleeve spacing (Ss) along individual longlines may be an important regulator of food uptake. SS is reportedly set at 120 cm in France's Pertuis Breton region (Garen et al.,

0044-8486/$ – see front matter. Crown Copyright © 2008 Published by Elsevier B.V. All rights reserved. doi:10.1016/j.aquaculture.2008.05.031

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

2004), 60 cm in Scottish sea lochs (Okumu and Stirling, 1998), and 44 cm in Canada's Prince Edward Island (PEI) region (Drapeau et al., 2006). To date, there have been few attempts to study relationships between SS and food uptake. Drapeau et al. (2006) reported a positive field correlation between SS and mussel yield; however, the correlation was tenuous, appearing only for a single year (in a three-year study). The principal objective of this investigation was to evaluate the relationship between SS, mussel yield, and seston uptake in Tracadie Bay, PEI, Canada (Fig. 1). An extensive grower interview process was used to gather detailed historical information on longline configuration in Tracadie Bay. The influence of SS was then investigated as part of a field experiment. Mussel yield at different SS was monitored at specific intervals over one year at four experimental sites. Yield was converted into seston uptake rates, and the potential impact of SS on bay-wide seston utilization was modeled. 2. Materials and methods 2.1. Study embayment Blue mussel (Mytilus edulis) farming in PEI (Fig. 1a) is carried out using a longline system of suspended polyethylene sleeves (Scarratt, 2000). Mussels are cultivated on the northern and the eastern sides of the island (Fig. 1b), where a total of 4351 ha of estuarine waters have been leased out to individuals and companies. Presently, there is a societal consensus that the PEI mussel industry is completing its developmental phase and entering into a management phase. This view is evidenced by a moratorium on the granting of new leases since 1999. Tracadie Bay is situated on the north shore of PEI (Fig. 1c) and represents an important mussel producing bay, contributing 20% of the island's total production.

57

2.2. Grower interviews Tracadie Bay has 34 mussel leases exploited by 24 leaseholders. In the winter of 2001, the majority (23/24) of these leaseholders agreed to share historic production and husbandry information, with the understanding that the data would not be divulged in a way that could be traced back to individual growers. Data from 1990 to 2000 was captured on a standardized questionnaire during the interview process. In general, the data originated from personal logbooks and sales records that leaseholders brought with them to the interviews. Similar interviews were repeated in the winter of 2002 to capture data for 2001 and extend the time series by an additional year. The interview dataset was used to calculate standardized indices of stocking density. We first examined seed abundance within individual sleeves. Sleeving is normally carried out in autumn when seed on collector ropes ranges in size between 4 and 30 mm. Leaseholders provided the initial (at deployment) seed abundance within polyethylene sleeves (SDIsleeve) and we reported SDIsleeve as the number of seed contained in a one meter section of sleeve. Leaseholders also provided detailed information that was used to calculate seeding density on an area basis within individual leases. Average seeding density within leases (SDIlease) was calculated as the number of seed m− 2. SDIlease was computed for each year of the dataset using the following equation: 0 SDIlease

n

∑

Blease¼1 ¼B @

h


1 Ss Ls



 Sl  SDIsleeve n

i1 C C A

ð1Þ

where Ss and Ls are the reported spacing between sleeves and longlines, respectively, and Sl is the sleeve length. Note that SDIlease does not take

Fig. 1. Map of study area (a) showing mussel farming sites in Prince Edward Island (b) and four experimental sites in Tracadie Bay (c). Grey boxes in Tracadie Bay represent leases. Experimental longline holding five “blocks” of sleeves is shown in panel d. Within blocks, 15 sleeves were hung at a fixed distance (10, 20, 40, 60 or 80 cm) from one another. Blocks were set in triplicate along longlines.

58

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

into account empty areas within the lease, which according to leaseholders represents between 42 and 58% of the total lease area. SDIlease is an indicator of seed density in areas where new crop was deployed. Also, SDIlease may be viewed as a maximum density index because seed mortality will inevitably occur during the grow-out phase. The interview dataset was also used to estimate the cultivated mussel biomass at the bay scale. For each year of data, seed biomass deployed for grow-out purposes was calculated as follow: SEEDB ¼

n

∑ lease¼1

Sn  Sl  0:76 1000

ð2Þ

where SEEDB represents the seed biomass (t), Sn is the number of seed sleeves reportedly deployed in the lease, and 0.76 is a weight conversion factor (0.76 kg per meter of sleeve) obtained from the interview dataset. In addition to seed, a lease may contain 1-year old and 2-year old crops. TOTALB represents the sum of seed biomass and older crop biomass. Because these parameters vary seasonally in accordance to seeding and harvest schedules, TOTALB was calculated for each month and reported as a yearly average. 2.3. Field experiment A field experiment was conducted to gauge the influence of SS on yield. Wild mussel larvae first settled onto collectors in the spring of 2003. They were placed into polyethylene sleeves by leaseholders in the fall of 2003 and thereafter remained under the ice cover during the winter of 2004. The field experiment was initiated on May 11th 2004 (Day 0). Four sites within Tracadie Bay were selected for the study. Sites were identified as LS1 (46°24.373′N, 62°59.385′W), LS2 (46°23.664′N, 62°59.431′W), HS1 (46°24.652′N, 62°59.703′W) and HS2 (46°23.658′N, 62°59.247′W). Their locations within Tracadie Bay are shown in Fig. 1c. A single longline of approximately 150 m was selected at each of the four sampling sites. On Day 0 baseline mussel samples were collected by randomly retrieving 10 sleeves from each experimental longline. After the baseline samples were collected, SS was rearranged along each of the selected longlines according to a randomized block design (see Fig. 1d). Within each block, 15 sleeves were hung at a fixed distance from one another, either at 10, 20, 40, 60 or 80 cm apart. Blocks were set out in triplicate along each longline: for example, there were three blocks of 15 sleeves, with each of the sleeves separated by 10 cm. Thereafter, samples were collected on three separate dates over a one-year period: August 19th 2004 (Day 100), November 2nd 2004 (Day 175) and on May 4th 2005 (Day 358). On each sampling date, one of the three blocks described above was targeted for sampling. The two peripheral sleeves in every block were ignored to avoid potential interaction effects with adjacent blocks and, among the 13 remaining sleeves, 10 were randomly selected for measurements. The number of sleeves collected as part of the experiment totaled 640 [(10 baseline sleeves × 4 sites) + (10 experimental sleeves × 5 SS treatments × 4 sites × 3 dates)]. All measurements were restricted to a particular segment of the sleeves. The bottom 0.3 m portion of each sleeve was discarded to eliminate possible interactions with the benthos. The next 0.3 m bottom section was removed, labeled and stored frozen at −20 °C until it was processed. At the laboratory, mussel abundance was determined by counting the number of live mussels in the 0.3 m sample. Results were extrapolated to a 1.0 m section of sleeve and reported as SDsleeve. Shell length was determined by measuring the maximum posterior–anterior axis of the shell using a Mitutoyo Digimatictm electronic caliper (±0.02 mm). Shell length was measured on 100 randomly selected individuals contained in the 0.3 m sample; all individuals were measured in samples containing less than 100 individuals. A total of

62,902 individuals were measured during the study. Similarly, the condition index (CI) was determined on a total of 5251 mussels, with samples ranging in size from 30 to 100 individuals treatment− 1. Sample size was a function of the sampling date, SDsleeve and outcome of power analyses. Dry tissue weight (DTW) and dry shell weight (DSW) were obtained following a 12 h exposure to 60 °C, and the CI was calculated according to the formula given in Abbe and Albright (2003): DTW  100 DSW

CI ¼

ð3Þ

Dry yield at the sleeve scale (DYsleeve) was calculated as the sum of dry tissue weight and dry shell weight. Results were extrapolated to a 2.0 m sleeve using the following equation: DYsleeve ¼ ðDTW þ DSWÞ  SDsleeve  2

ð4Þ

2.4. Seston uptake Mussel filtration rates were estimated using the allometric equation reported in Møhlenberg and Riisgård (1979). Filtration rates were scaled upwards to represent a sleeve of 2.0 m in length. This extrapolation was achieved by adding the SDsleeve term to the allometric equation below: Fsleeve ¼ 7:45  DTW0:66  SDsleeve  2

ð5Þ

where Fsleeve represents the estimated filtration rate of a single sleeve and DTW is the mean dry tissue weight of individuals contained in the sleeve. DTW and SDsleeve were measured in the laboratory as described above. An index of seston depletion (ID) was calculated following Dame (1996) and Grant et al. (2005). ID provides an indication of how important seston uptake may be in relation to estuarine volume and tidal flushing. First, the number of days required for mussels to filter the total estuarine volume was determined by applying a clearance time (CT) equation: CT ¼

VT Fsleeve  Nsleeve

ð6Þ

where VT is the Tracadie Bay volume (4.58 × 107 m3, Grant et al., 2005), Fsleeve is the filtration rate of a single sleeve (determined above), and Nsleeve is the total number of sleeves in the estuary (obtained from the interview dataset). ID was calculated as: ID ¼

CT RT

ð7Þ

where RT is the residence time (flushing time), estimated at 3.4 days for Tracadie Bay (Grant et al., 2005). The combination of Eqs. (5), (6) and (7) leads to equation: ID ¼

VT 7:45  DTW0:66  SDsleeve  2  Nsleeve  RT

ð8Þ

In this paper, a mean ID was obtained by applying the mean values of DTW and SDsleeve to Eq. (8). As indicated earlier, both DTW and SDsleeve were measured parameters. A range for ID (min ID and max ID) was obtained by applying DTW and SDsleeve min/max values to Eq. (8). 2.5. Statistical analysis The interview dataset was analyzed with the objective of determining whether longline configuration or stocking practices changed over the course of the time series (1990–2001). Annual means were computed and then compared using the Kruskal–Wallis non-parametric test.

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

59

Table 1 Summary of grower interview dataset FarmA

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 P=

SEEDB

TOTALB

Sl

Ss

Ls

SDIsleeve

SDIlease

FarmU

ID

WYsleeve

UH

ha

%

t

t

m

cm

m

Mussels m− 1

Mussels m− 2

%

Max

Mean

Min

kg sleeve− 1

%

412 486 486 486 538 542 586 620 620 620 620 620

26 31 31 31 35 35 38 40 40 40 40 40

60 112 113 95 135 236 406 388 451 392 502 577

1137 1506 1516 1385 1677 2399 3614 2972 3518 4178 4684 4743

1.90 (0.11) 1.89 (0.10) 1.89 (0.10) 1.91 (0.09) 1.98 (0.07) 2.00 (0.07) 2.01 (0.05) 2.01 (0.05) 2.03 (0.05) 2.00 (0.04) 2.03 (0.05) 1.96 (0.07) 0.86

39.7 (1.6) 39.8 (1.4) 39.8 (1.4) 39.4 (1.3) 38.4 (1.2) 37.0 (1.2) 36.5 (1.0) 36.8 (0.9) 37.2 (0.9) 37.0 (0.9) 37.7 (1.1) 37.8 (1.3) 0.43

10.9 (1.1) 10.2 (1.2) 10.2 (1.2) 10.1 (1.1) 10.2 (0.8) 9.4 (0.7) 9.4 (0.6) 9.7 (0.6) 9.3 (0.6) 9.7 (0.6) 10.6 (0.6) 11.5 (0.7) 0.64

559 (36) 561 (31) 561 (31) 567 (28) 572 (23) 589 (18) 613 (23) 606 (22) 601 (22) 608 (21) 603 (17) 584 (30) 0.91

282 (47) 307 (48) 307 (48) 317 (43) 319 (35) 383 (36) 400 (33) 378 (29) 389 (28) 388 (30) 360 (27) 283 (45) 0.34

45 (10) 43 (09) 43 (09) 43 (09) 45 (10) 42 (08) 44 (07) 52 (10) 44 (06) 44 (07) 51 (06) 58 (04) 0.93

1.00 0.77 0.77 0.75 0.73 0.46 0.49 0.44 0.47 0.47 0.37 0.38

0.90 0.69 0.69 0.67 0.66 0.42 0.44 0.40 0.43 0.43 0.33 0.34

0.77 0.59 0.59 0.58 0.56 0.36 0.37 0.34 0.37 0.37 0.28 0.29

10.0 (0.8) 10.0 (0.8) 10.0 (1.1) 10.0 (1.1) 13.0 (0.9) 13.0 (0.9) 9.4 (0.6) 9.1 (0.5) 8.1 (0.4) 8.0 (0.4) 7.9 (0.4) 6.7 (0.3) 0.003

7.0 7.0 6.9 21.8 23.1 31.5

FarmA: farming area (ha) and coverage (%) in relation to total estuarine area. SEEDB: mussel seed biomass in the bay. TOTALB: total (combined year-classes) mussel biomass in the bay. Sl: sleeve length. Ss: sleeve spacing. Ls: longlines spacing. SDIsleeve: seed abundance within sleeves (i.e. number of mussels contained in 1 m of suspended sleeve at deployment). SDIlease: seeding density within leases (i.e. seed m− 2) where longlines were present. FarmU: farm usage (i.e., % area within lease where longline gear was present). ID: seston depletion index. WYsleeve: standardized wet yield (kg sleeve− 1 after a 20-month growth period). UH: percentage biomass that was un-harvested after a 24-month growth period (calculated as biomass at 24 months/biomass at 12 months × 100).

With respect to the field experiment, comparisons between sites were considered of little interest because mussel size and abundance were initially different amongst sites. Consequently statistical analyses were applied to each site separately. Individual sleeves served as the measurement unit for SDsleeve and DYsleeve. A linear model was used to gauge fixed effects (SS, sampling date, interaction). For shell length and CI, however, individual mussels were considered as measurement units and individual sleeves were viewed as experimental units. Also, linear mixed models (LMM) were used to account for both fixed effects (SS, sampling date, interaction) and random effects within treatments (potential clustering in sleeves) (Dohoo et al., 2003). Inter-class correlation coefficients were low (b0.08) for all sampling dates, indicating there was very little clustering within individual sleeves. All models relied upon maximum-likelihood estimating procedures and hypotheses were tested using the Wald test. Multiple comparisons between treatments for a given sampling date were adjusted using the Bonferroni procedure. Statistical analyses were performed using Stata software (version 9; Stata Corporation, College Station, Texas). The significance level was set at P b 0.05.

Table 1 also shows that mussel filtration was balanced with water renewal in 1990 (ID ∼ 0.90). Filtration, however, progressively became more rapid than water renewal during the 1990s. By 2001, filtration rates were approximately three times greater than water renewal rates (ID ∼ 0.34). Falling ID values were accompanied by a declining trend in mussel productivity. The wet yield index (see WYsleeve) reported by growers fell from a mean value of 10.0 kg in 1990 to 6.7 kg in 2001

Table 2 Mussel abundance within sleeves at onset of experiment (SDIsleeve); mussel abundance (SDsleeve), shell length and sleeve yield (DYsleeve) in relation to sleeve spacing (Ss) on day 358

Site

Ss (cm)

LS1

10 20 40 60 80

LS2

10 20 40 60 80

HS1

10 20 40 60 80

HS2

10 20 40 60 80

3. Results 3.1. Grower interviews A summary of the interview dataset is presented in Table 1. Fourteen new leases were added in Tracadie Bay during the 1990s, raising the cultivation area from 412 to 620 ha. Farming area explained most of the variability associated with cultivated biomass in the bay: it was retained as the sole predictor for TOTALB (stepwise linear regression, r2 = 0.81, P b 0.001, β = 16.71, se β = 2.6), which increased by a factor of four (1137 to 4743 t) during the 1990s. In terms of gear configuration, the spacing of both sleeves and longlines was similar (P N 0.05, Kruskal–Wallis) between years. Likewise, there were no inter-annual differences in seed abundance within individual sleeves (SDIsleeve) or within leases (SDIlease). Therefore, growth in mussel aquaculture in Tracadie Bay during the 1990s is attributed to the addition of new leases as opposed to changes in stocking practices within individual leases.

SDIsleeve (mussels m− 1)

SDsleeve (mussels m− 1)

Shell length (mm)

DYsleeve (kg sleeve− 1)

Day 0

Day 358

Day 358

Day 358

Mean

433

333

923

723

SE

Mean

SE

Mean

SE

Mean

SE

43

297a 290a 223a 240a 223a

37 20 33 13 10

51.5a 52.5a 52.7a 52.0a 52.6a

0.19 0.17 0.18 0.20 0.20

3.10a 3.08a 2.28a 2.42a 2.28a

0.32 0.18 0.32 0.16 0.06

10

223a 230a 250a 230a 240a

17 13 10 13 17

52.5a 52.8a 52.4a 52.9a 53.4a

0.18 0.18 0.18 0.19 0.17

2.26a 2.34a 2.58a 2.36a 2.50a

0.16 0.14 0.10 0.16 0.18

33

327a 380ab 300a 330a 560b

50 53 50 57 43

40.4a 41.2a 39.5a 42.0ab 43.8b

0.17 0.20 0.23 0.20 0.16

1.97a 2.36a 1.84a 2.14a 3.80b

0.28 0.32 0.26 0.40 0.32

23

400ab 343a 383ab 383ab 487b

23 27 37 23 43

45.1a 46.0ab 46.5b 46.6b 48.5c

0.13 0.14 0.14 0.14 0.15

2.60ab 2.44ab 2.76ab 2.92ab 3.80bb

0.18 0.16 0.28 0.14 0.38

For each site, means without a common superscript were significantly different (P b 0.05).

60

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

Fig. 2. Estimated volumes of water filtered by mussels at the four experimental sites (LS1, LS2, HS1 and HS2). Data were scaled to a 2.0 m mussel sleeve and were grouped according to sleeve spacing and sampling period. Error bars represent the SEM associated with the last phase of the experiment (day 175 to 358, open bars); means without a common superscript were significantly different (P b 0.05) within sites.

(P b 0.05, Kruskal–Wallis). Moreover, growers reported delayed harvests for more recent years (see UH). In 2001, approximately 32% of the cultivated mussels were un-harvested following a 24-month growth period, compared to 7% from 1996–1998. 3.2. Field experiment The interaction term “Ss × sampling date” was often significant in the different model outputs, indicating that the effect of Ss on mussel performance depended on which sampling date was being considered. We decided to focus our report on the final sampling date (Day 358), because it was the closest date to harvest and consequently the most relevant to the industry. On Day 358, Ss had no significant effect on the condition index (CI) of mussels at the four study sites. Table 2 contains data on other response variables. At LS1 and LS2, where SDIsleeve was initially low (433 and 333 mussels m− 1), Ss had no significant influence on SDsleeve, shell length and DYsleeve. In contrast, Ss had a positive effect on the latter variables at HS1 and HS2, where SDIsleeve was initially high (923 and 723 mussels m− 1). For example, Ss 80 cm mussels tended to be more abundant than Ss 40cm mussels. Moreover, shell length was greater at Ss 80cm by 11% (HS1) and 4% (HS2) in comparison to shell length at the standard Ss 40cm applied by the industry. The combined growth and survival enhancements at Ss 80cm increased DYsleeve by 107% (from 1.84 to 3.80 kg sleeve− 1 at HS1) and 38% (from 2.76 to 3.80 kg sleeve− 1 at HS2). The experiment ended in May 2005, prior to mussels reaching a commercial size (55 mm) in all sites and Ss. Fig. 2 shows the estimated mean volumes of water filtered by mussels according to site and Ss. At LS1 and LS2, filtration rates were similar between different Ss, ranging between 18,437 and 23,912 m3 of water sleeve− 1 year− 1. However, at HS1 and HS2, filtration rates were distinctly elevated at Ss 80cm (27,799 and 25,649 m3 of water sleeve− 1 year− 1) compared to Ss 40cm (15,506 to 21,729 m3 of water sleeve− 1 year− 1). 4. Discussion 4.1. Farming development in Tracadie Bay Our study shows a declining trend in mussel productivity in Tracadie Bay from 1995 to 2001. It appears unlikely that the declining trend was driven by changes in the environment. Although long-term environmental data are non-existent for Tracadie Bay, useful informa-

tion is available for the adjacent southern Gulf of St. Lawrence. Satellite imagery (http://www.mar.dfo-mpo.gc.ca/science/ocean/ias/ seawifs/seawifs_1.html) reveals no apparent pattern in chlorophyll or temperature that could explain the reported fall in mussel productivity in the late 1990 s. Moreover, no major shift in bivalve productivity was recorded outside Tracadie in the late 1990s (The Shellfish Monitoring Network, http://www.glf.dfo-mpo.gc.ca/os/smn-rmm/ intro-e.php), suggesting that environmental conditions were favorable for mussel growth in neighboring estuarine systems. Thus it seems that factors other than large-scale environmental forcing were responsible for the reported decline in mussel productivity in Tracadie Bay. Our study also shows a rapid development of mussel culture in Tracadie Bay during the 1990s. Cultivated mussel biomass increased by a factor of four, reaching nearly 5000 t in 2001. In comparison, the biomass of wild mussels in the same embayment was recently estimated at approximately 120 t (M. Ouellette, DFO, unpublished data). Our data suggest that the production carrying capacity (PCC) in Tracadie Bay was exceeded in the late 1990s when cultivated mussel biomass surpassed approximately 3000 t. This conclusion is supported by falling sleeve yields in the late 1990s, and also by growers having decided to extend the grow-out period for some of their crop. Together these observations suggest that intensive farming led to severe intra-specific competition for food resources. Water sample analyses indicated low seston (b2 mg l− 1) and chlorophyll (b1 μg l− 1) concentrations in 2000 (Grant et al., 2005). Our conclusion on intensive farming negatively affecting mussel productivity is in broad agreement with recent modeling work showing that the grazing potential in the embayment has exceeded water renewal rates (Grant et al., 2005). There is, however, a discrepancy between previous studies and our results regarding the magnitude of the exceedence. Grant et al. (2005) reported that mussel filtration rates were 10 to 16 times greater than food renewal rates in Tracadie Bay (ID ∼ 0.05 to 0.09), whereas we suggest a more modest three-fold factor (ID ∼ 0.34). The discrepancy may be due to the use of different data sources for quantifying standing stocks and thus grazing potential. We calculated grazing using stocking information derived from grower logbooks, which were highly aligned with measurements made by divers (Table 1 in Drapeau et al., 2006). Nonetheless, despite scaling down the value of seston utilization in Tracadie Bay, an ID of 0.34 falls in the lower range of values reported for 11 other coastal bays or estuaries (Dame and Prins, 1998), including areas characterized by intensive bivalve culture (e.g. ID ∼ 0.54 in Ria de Arosa, Spain). 4.2. Sleeve spacing and yield Little is currently known about the benefits of re-arranging suspended sleeves along a longline. Drapeau et al. (2006), reported a positive correlation (r2 = 0.41) between SS and yield based on field observations made in Tracadie Bay in 2002; however, no such correlation was detected in subsequent years (2003 and 2004). The lack of inter-annual replication raised doubts about the potential benefits of increasing SS. Our results supported a link between SS and mussel productivity at two of the four study sites. The doubling of the industry's standard (from Ss 40cm to Ss 80cm ) led to increased shell growth at sites HS1 and HS2 (but not at LS1 and LS2). The exact reasons for the different outcomes cannot be determined with certainty. One interpretation relates to the environment and more specifically to food availability. Food was perhaps exceedingly limiting at both HS1 and HS2, such that the doubling of SS significantly alleviated intra-specific competition and enhanced growth at those two sites. On the other hand, such an interpretation is inconsistent with the modeling of phytoplankton distribution within the embayment (Grant et al., in press). Apart from the environment, cultures practices, such as mussel abundance within individual sleeves, may

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

explain the observed differences between sites. It is possible that high SS (SS 80 cm) enhanced water exchanges and food particle renewal at the local (cm) scale for all sites, but that only highly aggregated mussels benefited from these conditions. This interpretation is consistent with our results: Ss 80 cm produced superior shell growth only where sleeves were densely packed (HS1 and HS2). Presumably, densely packed mussels experienced less competition for food resources at Ss 80 cm than at SS b 80 cm . For the raft culture technique, Heasman et al. (1998) reported increased shell growth for mussels spaced at 90 cm compared to 60 cm, and attributed the increased growth to a greater availability of food particles at local scales. Our study also examined the effect of SS on mussel abundance. It is well documented that severe intra-specific competition for food and space leads to mortality. This process is often referred to as selfthinning (Westoby, 1984; Hughes and Griffiths, 1988; Fréchette et al., 1992; Fréchette and Lefaivre, 1995). Lauzon-Guay et al. (2005) recently documented self-thinning in PEI cultivated mussels and concluded that self-thinning was attributable to competition for space on individual sleeves. Our results are unique because they show a link between SS and self-thinning, at least at sites where mussels were densely packed in the sleeves (HS1 and HS2). It is likely that these highly aggregated mussels competed severely for food resources. Under such circumstances, it appears that seed survival during the self-thinning process can be regulated, to some extent, by one simple husbandry variable, SS. When both shell growth and abundance were taken into account, the net benefit of increasing SS became evident at HS1 and HS2, where average sleeve yield was significantly greater at SS 80cm than at SS 40cm (by 107% at HS1 and 38% at HS2). We found no indication that the added weight at SS 80 cm was associated with lower mussel quality. The condition index indicated similar meat yields between Ss groups, while shell length coefficient of variations (not shown) indicated similar population size structures within sleeves. Nevertheless, it is important to re-iterate that our experiment ended a few months prior to mussels reaching a commercial size (55 mm). It is possible that the added weight at SS 80 cm would have ultimately led to increased falloffs in the final months of the production cycle. Fall-offs can be controlled by a procedure called “double-sleeving”. This procedure is commonly employed in PEI, where clumps of mature mussels detach easily from the sleeves. 4.3. Potential implications for industry In 2002, Tracadie Bay growers collectively developed a bay management plan that limits the stocking density to 1235 sleeves per leased hectare. This number was obtained by averaging stocking density at the time and was viewed by stakeholders as a reasonable middle ground. The management plan was mainly intended to even out crop distribution within the bay as opposed to reducing the total crop biomass in the bay. Following the plan's implementation, Tracadie Bay growers began positioning individual sleeves at a greater distance from one another: divers' measurements indicated that SS increased by 11 cm (from 38 to 49 cm) from 2002 to 2004 (Drapeau et al., 2006). Our results imply that the modest SS increase (11 cm) carried out by the industry had little or no impact on mussel productivity. We base this conclusion on the lack of significant differences in mussel performance between SS 40cm and SS 60 cm. Tracadie Bay growers may ultimately decide to reduce grazing biomass (standing stocks). We calculated that a 50% reduction in standing stocks (from 5000 to 2500 t), without any deviation from current sleeving practices, may potentially increase the ID from 0.34 (Table 1, 2001) to 0.51 (0.44–0.57). However, if accompanied by a high SDIsleeve / / high SS strategy, the same biomass reduction could lead to an ID of 0.70 (0.68–0.73). In the latter scenario, fewer food particles would be allocated to mussels that die prior to harvest, given that elevated SS would curtail the self-thinning process. We caution that

61

the projection model is based upon data from pre-market size (44– 49 mm) mussels at two of four study sites; it does not take into account possible fall-offs from the densely packed sleeves in the final months of the production cycle. Nonetheless, the projection outcome does show that crop arrangement in the water column is nearly as important as absolute biomass in terms of bay-wide grazing potential. This conclusion supports the view that husbandry techniques and their effects on filtration rates should be considered when developing carrying capacity models for bivalve culture. Acknowledgements The authors are grateful to Tracadie Bay leaseholders for providing detailed husbandry data; H. Stryhn, L. Spangler and G. Johnson for codirecting the project; and J. Hill, G. Arsenault, B. Thompson, M. Hardy and D. Bourque for the technical support. We also thank R. Fortune, D. Roberts and M. Habbi for their cooperation and assistance during the experimental phase of the project. Comments made by anonymous reviewers were very helpful. This study was funded by DFO's Aquaculture Collaboration Research and Development Program (ACRDP project MG-01-06-026). References Aoyama, S., 1989. The Mutsu Bay scallop fisheries: Scallop culture, stock enhancement, and resource management. In: Caddy, J.F. (Ed.), Marine Invertebrate Fisheries: Their Assessment and Management. Wiley, New York, pp. 525–539. Abbe, G.R., Albright, B.W., 2003. An improvement to the determination of meat condition index for the eastern oyster Crassostrea virginica (Gmelin 1971). J. Shellfish Res. 22, 747–752. Bacher, C., Grant, J., Hawkins, J.S., Fang, J., Zhu, M., Besnard, M., 2003. Modelling the effect of food depletion on scallop growth in Sungo Bay (China). Aquat. Living Resour. 16, 10–24. Boyd, A.J., Heasman, K.G., 1998. Shellfish mariculture in the Benguela system: water flow patterns within a mussel farm in Saldanha Bay, South Africa. J. Shellfish Res. 17, 25–32. Boromthanarat, S., Deslous-Paoli, J.M., 1988. Production of Mytilus edulis L. reared on bouchots in the Bay of Marennes-Oléron: comparison between two methods of culture. Aquaculture 72, 255–263. Cabanas, J.M., González, J.J., Mariño, J., Perez, A., Román, G., 1979. Estudio del mejillón y de su epifauna en los cultivos flotantes de la Ría de Arosa: III. Observaciones previas sobre la retención de partículas y la biodeposición de una batea. Bol. Inst. Esp. Oceanogr. 5, 45–50. Dame, R.F., 1996. Ecology of marine bivalves: an ecosystem approach. CRC Press, Boca Raton, FL. Dame, R.F., Prins, T.C., 1998. Bivalve carrying capacity in coastal ecosystems. Aquat. Ecol. 31, 409–421. Dohoo, I., Martin, W., Stryhn, H., 2003. Mixed models for continuous data. Veterinary Epidemiologic Research, AVC Inc, pp. 473–498. Charlottetown, PE, Canada. Dolmer, P., 2000. Algal concentration profiles above mussel beds. J. Sea Res. 43, 19–36. Drapeau, A., Comeau, L.A., Landry, T., Stryhn, H., Davidson, J., 2006. Association between longline setup and mussel productivity in Prince Edward Island, Canada. Aquaculture 261, 879–889. Ferreira, J.G., Duarte, P., Ball, B., 1998. Trophic capacity of Carlingford Lough for oyster culture — analysis by ecological modelling. Aquat. Ecol. 31, 361–378. Fréchette, M., Bourget, E., 1985a. Energy flow between the pelagic and benthic zones: factors controlling particulate organic matter available to an intertidal mussel bed. Can. J. Fish. Aquat. Sci. 42, 1158–1165. Fréchette, M., Bourget, E., 1985b. Food-limited growth of Mytilus edulis L. in relation to the benthic boundary-layer. Can. J. Fish. Aquat. Sci. 42, 1166–1170. Fréchette, M., Lefaivre, D., 1995. On self-thinning in animals. Oikos 73, 425–428. Fréchette, M., Booth, D.A., Myrand, B., Bernard, H., 1991. Variability and transport of organic seston near a mussel aquaculture site. ICES Mar. Sci. Symp. 192, 24–32. Fréchette, M., Aitken, A.E., Page, L., 1992. Interdependence of food and space limitation of a benthic suspension feeder: consequences for self-thinning relationships. Mar. Ecol. Prog. Ser. 83, 55–62. Fuentes, J., Gregorio, V., Giráldez, R., Molares, J., 2000. Within-raft variability of the growth rate of mussels, Mytilus galloprovincialis, cultivated in the Ría de Arousa (NW Spain). Aquaculture 189, 39–52. Garen, P., Robert, S., Bougrier, S., 2004. Comparison of growth of mussel, Mytilus edulis, on longline, pole and bottom culture sites in the Pertuis Breton, France. Aquaculture 232, 511–524. Grant, J., Cranford, P., Hargrave, B., Carreau, M., Schofield, B., Armsworthy, S., Burdett-Coutts, V., Ibarra, D., 2005. A model of aquaculture biodeposition for multiple estuaries and field validation at blue mussel (Mytilus edulis) culture sites in eastern Canada. Can. J. Fish. Aquat. Sci. 62, 1271–1285. Grant, J., Bacher, C., Cranford, P.J., Guyondet, T., Carreau, M., in press. A spatially explicit ecosystem model of seston depletion in dense mussel culture, J. Mar. Syst. doi:10.1016/j.jmarsys.2007.10.007.

62

L.A. Comeau et al. / Aquaculture 281 (2008) 56–62

Heasman, K.G., Pitcher, G.C., McQuaid, C.D., Hecht, T., 1998. Shellfish mariculture in the Benguela system: raft culture of Mytilus galloprovincialis and the effect of rope spacing on food extraction, growth rate, production, and condition of mussels. J. Shellfish Res. 17, 33–39. Héral, M., Deslous-Paoli, J.M., Prou, J., 1986. Dynamiques des productions et des biomasses des huîtres creuses cultivées (Crassostrea angulata et Crassostrea gigas) dans le bassin de Marennes–Oléron depuis un siècle. ICES CM86/F:14. Hughes, R.N., Griffiths, C.L., 1988. Self-thinning in barnacles and mussels: the geometry of packing. Am. Nat. 132, 484–491. Lauzon-Guay, J.S., Dionne, M., Barbeau, M.A., Hamilton, D.J., 2005. Effects of seed size and density on growth, tissue-to-shell ratio and survival of cultivated mussels (Mytilus edulis) in Prince Edward Island, Canada. Aquaculture 250, 652–665. McKindsey, C.W., Thetmeyer, H., Landry, T., Silvert, W., 2006. Review of recent carrying capacity models for bivalve culture and recommendations for research and management. Aquaculture 261, 451–462. Møhlenberg, F., Riisgård, H.U., 1979. Filtration rate, using a new indirect technique, in thirteen species of suspension-feeding bivalves. Mar. Biol. 54, 143–148. Navarro, E., Iglesias, J.I.P., Perez Camacho, A., Labarta, U., Beiras, R., 1991. The physiological energetics of mussels (Mytilus galloprovincialis Lmk) from different cultivation rafts in the Ria de Arosa (Galicia, N.W. Spain). Aquaculture 94, 197–212. Newell, C.R., 1990. The effect of mussel (Mytilus edulis, Linnaeus, 1758) position in seeded bottom patches on growth at subtidal lease sites in Maine. J. Shellfish Res. 9, 113–118.

Okumu, B., Stirling, H.P., 1998. Seasonal variations in the meat weight, condition index and biochemical composition of mussels (Mytilus edulis L.) in suspended culture in two Scottish sea lochs. Aquaculture 159, 249–261. Ogilvie, S.C., Ross, A.H., Schiel, D.R., 2000. Phytoplankton biomass associated with mussel farms in Beatrix Bay, New Zealand. Aquaculture 181, 71–80. Petersen, J.K., 2004. Grazing on pelagic primary producers—the role of benthic suspension feeders. In: Nielsen, S.L., Banta, G.T., Pedersen, M.F. (Eds.), The influence of primary producers on estuarine nutrient cycling. Kluwer, pp. 129–155. Pilditch, C.A., Grant, J., Bryan, K.R., 2001. Seston supply to sea scallops (Placopecten magellanicus) in suspended culture. Can. J. Fish. Aquat. Sci. 58, 241–253. Rosenberg, R., Loo, L.O., 1983. Energy-flow in a Mytilus edulis culture in western Sweden. Aquaculture 35, 151–161. Scarratt, D.J., 2000. Development of the mussel industry in eastern Canada. Bull. Aquacult. Assoc. Can. 100, 37–40. Strohmeier, T., Aure, J., Duinker, A., Castberg, T., Svardal, A., Strand, Ø., 2005. Flow reduction, seston depletion, meat content and distribution of diarrhetic shellfish toxins in a long-line blue mussels (Mytilus edulis) farm. J. Shellfish. Res. 24, 15–23. Westoby, M., 1984. The self-thinning rule. Adv. Ecol. Res. 14, 167–225. Wildish, D.J., Kristmanson, D.D., 1979. Tidal energy and sublittoral macrobenthic animals in estuaries. J. Fish. Res. Board Can. 36, 1197–1206.