CONTENTS Introduction Methods Results Summary and Conclusions References Acknowledgments
Northeast Fisheries Science Center Reference Document 05-01
James R. Weinberg1, Eric N. Powell2, Chris Pickett1, Victor A. Nordahl, Jr.1, and Larry D. Jacobson1
Results from the 2004 Cooperative Survey of Atlantic Surfclams
166 Water St., Woods Hole, MA 02543, USA
2Haskin Shellfish Research Laboratory,
Rutgers University, Port Norris, NJ 08349, USA
Web version posted February 10, 2005Citation: Weinberg, J.R.; Powell, E.N.; Pickett, C.; Nordahl, V.A., Jr.; Jacobson, L.D. 2005. Results from the 2004 cooperative survey of Atlantic surfclams. U.S. Dep. Commer., Northeast Fish. Sci. Cent. Ref. Doc. 05-01; 41 p.
Information Quality Act Compliance: In accordance with section 515 of Public Law 106-554, the Northeast Fisheries Science Center completed both technical and policy reviews for this report. These predissemination reviews are on file at the NEFSC Editorial Office.
The Atlantic surfclam (Spisula solidissima) supports a multimillion dollar annual fishery along the Mid-Atlantic coast of the USA. The history of the Atlantic EEZ clam fishery is described in Murawski and Serchuk (1989), Weinberg (1999), NEFSC (2000, 2003) and Weinberg et al (2002a). Federal surveys of the Atlantic surfclam resource in the EEZ are conducted by NOAA/NMFS Northeast Fisheries Science Center (NEFSC) every 3 years. The next federal clam survey on the RV Delaware-II (i.e., DE-II) is scheduled for spring of 2005. Due to concerns about the status of the surfclam stock in the Mid-Atlantic region, a cooperative survey of the EEZ was planned and then successfully carried out in late June - early July 2004 with contributions from the surfclam industry, Rutgers University and the NEFSC. The survey area included the Mid-Atlantic coast from the Hudson Canyon, off northern New Jersey, to Virginia. All fieldwork was conducted from the FV Lisa Kim, a commercial clamming vessel operated by a professional captain and crew and equipped with a commercial ocean quahog dredge. A stratified random sampling design, utilizing NEFSC clam strata, was used during the survey, and dredge efficiency was measured with depletion experiments.
The 2004 cooperative survey was preceded by a brief shakedown leg, April 6-9, 2004 off the coast of New Jersey to determine the appropriate tow duration to use during the upcoming survey. Results indicated that 5-min was the optimal tow duration for capturing 5-10 bushels of surfclams at locations which had low-moderate clam densities in 2002. In addition, two depletion experiments (sc04-1, sc04-2) were completed during the shakedown to estimate dredge efficiency.
This report presents results from the 2004 cooperative surfclam survey and three dredge efficiency experiments (2 during the shakedown; 1 during the survey). The report describes catches at each station as well as regional and stratum-based analyses of population size-structure and biomass. Results from 2004 are compared with historical survey data, when appropriate. Particular attention is given to changes in the depth distribution of surfclams and to the surfclam resource in the Delmarva region, where a temperature-related die-off was detected in 2002 (NEFSC 2003; Kim and Powell 2004).
In this report, some abbreviations and terms are used repeatedly. Region names are: Northern New Jersey (NNJ), Southern New Jersey (SNJ), Delmarva (DMV), Southern Virginia (SVA) and Georges Bank (GBK). Reference is made to 3 depth zones: shallow (9 - 28 m), mid (29 - 46 m), and deep (47 - 55 m). Surfclams are grouped into three length classes: large (120 mm+), medium (88 - 119 mm), and small (1 - 87 mm).
The 2004 surfclam survey was conducted from the FV Lisa Kim, a commercial clamming boat equipped with an ocean quahog dredge (Table 1). The dredge is constructed of metal bars with 1.25" spaces, although in some places, spaces as large as 2" were noted. Commercial ocean quahog dredges have smaller openings than surfclam dredges. The openings on this dredge were similar in size to those on the NEFSC clam dredge, which was used in previous NMFS surfclam surveys with the RV Delaware-II.
The 2004 survey used a stratified random sampling design (Figure 1) based on NEFSC clam strata, and the number of samples per stratum was the same as in the 2002 clam survey with the RV Delaware-II (NEFSC, 2003). Station locations (Figure 2) were selected by the NEFSC Ecosystems Survey Branch using a standard computer program. In addition to sampling at randomly chosen sites within strata, nonrandom tows were made to monitor dredge efficiency.
The sampling procedure at random stations was similar to that on RV Delaware-II clam surveys. Total number of surfclams per tow was determined and the maximum shell lengths of a subset of surfclams were measured to the nearest mm with manual measuring boards. In general, tows were made in the direction of the next station. The boat arrived at each station doing 4-6 knots, but quickly slowed down to make the 5-minute timed tow at 3 knots, once the gear hit the bottom and started fishing. The fast winches and heavy dredge on the FV Lisa Kim made the time to set and haul back the dredge very short compared to the RV Delaware-II (Weinberg, et al. 2002b).
When the dredge was hauled aboard the vessel, the catch was passed over a mechanical shaker to retain whole and broken clams. Spacing of the bars on the shaker was set at 1.25 inches, although some spaces up to 1.5 inches were present. Surfclams were then moved on a conveyor belt and collected in bushel baskets, each of which could hold about 50-60 adult surfclams. Total number of full, level baskets was recorded. For small catches (i.e., 0 - 3 bushels), all clams were counted and measured. For medium catches (i.e., 4 - 5 bushels), the number of clams in each bushel was recorded separately, and 3 of those bushels were selected at random for shell measurements. For large catches (>5 bushels), the total number of level bushels was recorded, the number of clams in each of 5 randomly selected bushels was recorded separately, and 3 of those bushels were randomly selected for shell measurements. For large catches, the total number caught was estimated by multiplying the average number per bushel times the total number of bushels. At every station, partial bushels were counted and added in separately. Broken clams were counted, but not measured.
Due to the small number of scientists on the FV Lisa Kim, compared to the RV Delaware II, some standard sampling was not done. Surfclam shells were not collected for age analysis and counts, but no measurements, were made of ocean quahogs (Arctica islandica), southern quahogs (Mercenaria campechiensis), surfclam clappers, and ocean quahog clappers. Data were not collected on other by-catch species, such as crabs, starfish or gastropods. Data were not collected on the amount of shell hash or sediment in the dredge. Sediment samples were not collected from the bottom.
At many stations, scientists from Rutgers University collected surfclams to examine the relationship between shell length and meat weight. These results are not available yet.
Dredge performance on the FV Lisa Kim was monitored with the NMFS Survey Sensor Package (SSP). In addition, Rutgers provided WindPlot GPS software and Vemco pressure-temperature sensors. The SSP measured dredge angle, depth, temperature, manifold pressure and ship's position. SSP sampling frequency was 1-sec for all variables except for position, which was every 2-sec. There was an SSP equipment failure after Station 20. During subsequent survey stations, WindPlot was set to sample ship's position every 10 sec, and the Vemco sensor on the dredge sampled every 5-sec. Due to the timing of message packets, the actual GPS times from WindPlot were 9-12 seconds apart. The dredge sensor data from each station were used to determine the starting and ending position of each tow. Distance sampled was determined from the GPS latitudes and longitudes along each tow track. The GPS measured the boat's position, which was assumed to be representative of the position of the dredge. Additional details about the survey are given in Table 1.
Audited SSP sensor data from the shakedown leg and 2004 surfclam survey were loaded into the NEFSC database with Oracle table names LK_200415 and LK_200416, respectively. Audited biological and station data from the 2004 Surfclam Survey were assigned Cruise code 200416 and loaded into the NEFSC survey database. Original survey log sheets are stored with the Ecosystem Survey Branch of the NEFSC.
Three depletion experiments were conducted to estimate the FV Lisa Kim's dredge efficiency in capturing surfclams (Figure 3; Table 2, Table 3, and Table 4). The tables give information on the location of each experiment, bottom depth, dates of operation, number of tows per experiment, and sensors that were used to monitor dredge behavior and vessel location. The first two experiments were done during the shakedown leg, in NMFS Strata 21 and 88. The third was carried out during the survey leg, in NMFS Stratum 13.
Each experiment consisted of making repeated tows in a rectangular area, counting the catch per tow, monitoring size composition, and logging the location of the vessel during each tow (Figure 4).
The Rago Patch model, described in NEFSC (2003) and NEFSC (2004), was used to analyze the depletion experiments. The model estimates dredge efficiency and clam density. The model was run using a cell size of 20 feet, which is twice the width of the FV Lisa Kim dredge (10 feet). Gamma was fixed at 0.5, which assumes no "indirect" effects. In analyzing sc04-3, it was necessary to interpolate positions to every 10 feet to match the scale of the location data to dredge and cell size (Table 4). No interpolation was needed for the sc04-1 and sc04-2 experiments, because the position of the ship had been measured every 10 feet.
Based on the likelihood profiles, the Patch model was able to estimate dredge efficiency and surfclam density in all three experiments (Figure 5; Table 5). The mean of the three estimates of dredge efficiency was 0.792 (SD=0.036, CV=4.6%). There was little variation among the three efficiency estimates even though the surfclam density varied among experiments by an order of magnitude and the experiments were done at different depths and in two regions (Table 5 and Table 6).
For comparison, estimates of commercial dredge efficiency from 2004 are listed along with estimates from earlier experiments involving surfclams and ocean quahogs (Table 6). Those values range from 0.46 to 0.95. Estimates from 2004 are well within the range of these earlier estimates.
"Repeat" Lisa Kim Stations to Examine Dredge Efficiency
At six stations, tows were repeated to check for gross changes in efficiency between the start and end of the 2004 cooperative survey. Pairs of catches were similar at the two times, with no consistent bias, suggesting that dredge efficiency did not change during the survey (Table 7, Figure 6). To test for a gross change in the catches between the two times, a one-sample T-test was run on the six differences, testing the null hypothesis that the mean difference = 0. This null hypothesis could not be rejected (T = 0.98, Pr >0.1). Although the correlation between catch at time 1 and time 2 was positive, it was not statistically significant (r = 0.577, Pr >0.1, n=6). This test probably had low statistical power owing to the small sample size. Overall, the statistical tests do not indicate gross changes in gear efficiency between the start and end of the survey.
Selectivity of the dredge was examined by pushing surfclams of known sizes through the bars on the dredge (Table 8). Although clams of all sizes tested could fit through the grate directly behind the blade, this is not likely to be a place where many clams are lost because the clams are being scooped up and quickly pass over this grate. Both the floor of the dredge and the mechanical shaker probably have a large effect on size selectivity. The data, shown in the table, suggest that at least some clams as large as 89 mm in shell length could pass through the shaker; this could vary depending on the thickness of the shell. Overall, the data suggest that clams 90 mm and larger in shell length were retained by the gear used during the 2004 survey. Depending on their size and shell shape, a fraction of surfclams, smaller than 90 mm in length, was not retained.
Summary statistics on tow distances are given in Table 9. The average distance sampled at random survey stations was 0.244 nmi. Tow distance did not vary appreciably with depth or from the start to the end of the survey (Figure 7). The few cases where tows had exceptionally short distances were the result of retrieving the dredge early when it was full or had encountered bad bottom.
Catch per tow was standardized to a tow distance of 0.15 nmi, the distance used in recent surfclam assessment reports (NEFSC 2000, 2003). Standardizing catch to a common distance allows for a better comparison of the catches from stations at various locations within a survey. We note, however, that these standardized catches per tow for 2004 are not directly comparable to those reported for earlier surveys (NEFSC 2000, 2003) because the dredge on the FV Lisa Kim sampled a larger area per tow and had a higher efficiency than the RV DE-II.
Surfclam catch per tow, standardized for tow distance, is shown for three length classes: large (120 mm+) (Figure 8), medium (88 - 119 mm) (Figure 9), and small (1 - 87 mm) (Figure 10). As mentioned earlier, the smallest length class was not retained consistently by the survey dredge.
In NNJ, large clams were most abundant in the shallow (Strata 88 and 89) and mid-depth strata (Strata 21 and 25). There were large catches of surfclams in the deeper portions of Strata 21 and 25, a pattern that occurred very rarely in the 1980's and early 1990's (see Fig. C42 in NEFSC 2003).
In DMV, large clams were most abundant in mid-depth strata (Strata 9 and 13). Within Stratum 9, larger catches occurred in deeper water. This pattern was also seen in 2002 (see Fig. C36 in NEFSC 2003).
In 2004, mid-sized clams were most abundant in mid-depth strata of both NNJ (Strata 21 and 25) and DMV (Strata 9 and 13) (Figure 9). Larger catches occurred in deeper water. Mid-sized clams were not captured in high numbers in shallow strata of NNJ in 2004 (Figure 9) or in 2002 (see Fig. C37 in NEFSC 2003).
Efficiency Corrected Swept Area Biomass (ESB)
Methods used here for calculating ESB and 80% confidence intervals for ESB are described in two recent laboratory reference documents in this series: NEFSC (2003, pages 299-301) and NEFSC (2004, pages 27-30). Terms in the ESB calculation include survey-specific dredge efficiency (e), sensor tow distances (ds), area swept per standard tow (a), total area of region (A), percent suitable habitat (u), and catch. The CVs for area swept per tow and habitat area in each region, given in Table A17 of NEFSC (2003), were also assumed in this report when computing region specific ESBs for 2004.
Fully-recruited ESB estimates and confidence intervals were computed for four Mid-Atlantic EEZ regions, by year (Table 10, Figure 11 and Figure 12). The ESB time series begins with 1997, the first year with a dredge efficiency estimate. Only those regions that were surveyed in 2004 are listed in the table. Other regions that may have considerable surfclam biomass, such as GBK, are not shown.
Of the four regions surveyed in 2004, the greatest biomass was in NNJ. Considering the confidence intervals, fully recruited biomass in NNJ has remained fairly stable from 1997 to 2004 at about 500,000 mt (Figure 11). There may have been a dip in biomass between 2000 and 2004. It is difficult to pinpoint when changes occur because surveys have not been conducted annually.
The region with second highest biomass in 2004 was DMV, with approximately 143,000 mt (Table 10, Figure 12). This biomass has held steady since 2002, after the large decline that took place between 1999 and 2002.
Neither SNJ nor SVA had significant biomass (<20,000 mt combined; Table 10) in 2004, and biomass appears to have declined in both regions since 2002.
For the regions surveyed in 2004, NNJ and DMV had 77% and 21% of the biomass, respectively (Figure 13). The biomass is concentrated in the NNJ region, and the degree of concentration has increased since 1997 (Figure 14).
Changes have occurred in population size-structure within each region over time, based on the time series going back to 1982 (Figures 15 and Figure 16). Surveys were conducted before 1982, but they used different collecting gear (see Table C6 in NEFSC 2003), and are not directly comparable to more recent data. We also note that a different dredge was used in 2004 than in previous surveys. Though the two dredges probably have similar selectivity of surfclams >100 mm in length, there may be some differences in retention of smaller clams.
At the large regional scale, surfclams in NNJ increased in shell size over time, through growth. The mode in NNJ shifted from 140-145 mm in 1997 to 150-155 mm in 2004 (Figure 15). Furthermore, the NNJ size structure was bimodal in 2002 and 2004, suggesting a pulse of recruitment. Relative to NNJ, surfclams in the DMV region were smaller (Figure 16). DMV clams were commonly in the 110 - 140 mm length interval. Based on the size-frequency distribution, there was some evidence of recruitment in 2002.
Although there was both recruitment of small clams and growth of large clams in NNJ, these processes did not occur uniformly throughout that region. This is evident from data from certain strata in NNJ. Strata 88 and 89, two of the main shallow NNJ strata, show the increase in body size of large surfclams over time, but little or no evidence of recruitment (Figure 17 and Figure 18). The modal size in 2004 in these shallow strata was 155-160 mm. In contrast, there was evidence of growth of large clams over time and recent strong recruitment in the two NNJ mid-depth strata, 21 and especially 25 (Figure 19 and Figure 20). Stratum 25 is the most northern large stratum in NNJ, located on the edge of the Hudson Canyon. The large clams in the mid-depth strata were approximately 140 - 155 mm in length in 2004, making them smaller than the large clams in shallow Strata 88 and 89.
A detailed examination of the catch per tow data from strata in the NNJ region in 2004 was made to determine where biomass per tow and density per tow were highest (Table 11 and Table 12). For the analysis, strata within the region were classified into three depth zones: shallow, mid, deep (Table 11).
For large (>= 120 mm) clams, shallow Stratum 89 had the highest catch per tow (Table 12A). The two mid-depth strata (21 and 25) ranked second and third, ahead of shallow Stratum 88. This latter result is significant because historically, the commercial surfclam fishery was active in and around Stratum 88.
For recruiting (88 - 119 mm) clams, the two mid-depth strata (25 and 21) ranked first and second in catch per tow (Table 12B). Stratum 25 had much greater catches than Stratum 21. Thus, catches of recruiting surfclams in 2004 were higher in both of the mid-depth NNJ strata than in the shallow NNJ strata.
Few surfclams of any size were captured in the deepest NNJ strata sampled in 2004, Strata 22 and 26 (Table 12).
A robust index for monitoring gross changes in the DMV population, where thermal stress may have killed off surfclams recently, is the percentage of random tows that captured no surfclams (Table 13). This index was high in 1999 and 2002, 0.3 and 0.39 respectively. The index in 2004 (0.24) must be interpreted cautiously because in 2004 the probability of capturing a clam in a tow was greater than in earlier surveys, irrespective of sufclam density. The 2004 survey was conducted with a wider dredge and a dredge which has much greater catching efficiency. Thus, the value for 2004 is a lower bound estimate for that year.
Several signals in the data suggested a shift in the surfclam population over time to deeper water. For example, in 2004 in NNJ more recruit-size surfclams were captured per tow in mid-depth strata than in shallow strata. Also, there was evidence of a die-off in the shallow portion of DMV.
To examine this more directly, we computed the proportion of surfclam biomass in three depth zones (shallow, mid, deep) for each combination of region and year (Table 14, Table 15, and Table 16, Figure 21). This calculation could not be made for the NJ region in 1984, because random samples were not taken from the deep zone in that year (Table 15). For NNJ and SNJ, the values shown in Figure 21 for 1984 are the averages of the values from 1983 and 1986.
In NNJ, there was a clear change over time in the distribution of surfclam biomass with respect to water depth. In the 1980s, over 70% of the biomass was in the shallow zone. This percentage declined continuously throughout the late 1980s and the 1990s to its historical low value of 29% in 2004. In 2004, 71% of the NNJ biomass was in the mid-depth zone. The other regions did not show clear trends regarding depth.
Future analyses, with finer partitioning of depth zones and taking into account the locations of commercial landings, might yield additional results.
SUMMARY AND CONCLUSIONS
- The 2004 Cooperative surfclam survey sampled from the Hudson Canyon to approximately 36.5 degrees latitude. The survey collected data from 4 regions: NNJ, SNJ, DMV, and SVA. The survey covered the full depth range of the species. Most stations were at depths of 10 to 50 m.
- Of the four regions surveyed in 2004, the greatest biomass was in NNJ. Considering the confidence intervals, fully recruited biomass in NNJ has remained fairly stable from 1997 - 2004 at about 500,000 mt.
- The region with second highest biomass in 2004 was DMV, with approximately 143,000 mt. There was a large decline in the resource in this region between 1999 and 2002. No change in biomass was detected from 2002 to 2004.
- Neither SNJ nor SVA had significant biomass (<20,000 mt combined) in 2004, and biomass appears to have declined in both regions since 2002.
- Fully-recruited surfclam biomass is concentrated in the NNJ region. The degree of concentration in 2004 was greater than in the recent past.
- The biomass of surfclams in NNJ has not changed greatly since 1997; however,most of the NNJ resource is located in deeper water now.
- At the large regional scale, the fully recruited clams in NNJ have increased in shell size over time. Surfclams in NNJ are larger than those in the DMV region.
- Strong recruitment occurred recently in the two NNJ mid-depth strata (21 and 25). Strong recruitment did not occur in the shallower strata of NNJ or DMV.
Murawski, S. A. and F. M. Serchuk. 1989. Mechanized shellfish harvesting and its management: the offshore clam fishery of the eastern United States. In: Caddy JF (ed.) Marine Invertebrate Fisheries: Their Assessment and Management. Wiley, New York, pp 479-506.
NEFSC (Northeast Fisheries Science Center). 2000. 30th Northeast Regional Stock Assessment Workshop (30th SAW). E. Surfclams. pp 311-477. NEFSC Ref. Doc. 00-03.
NEFSC (Northeast Fisheries Science Center). 2003. 37th Northeast Regional Stock Assessment Workshop (37th SAW). C. Atlantic Surfclam. pp.284-433. NEFSC Ref. Doc. 03-16.
NEFSC (Northeast Fisheries Science Center). 2004. 38th Northeast Regional Stock Assessment Workshop (38th SAW). A. Ocean quahogs. pp.1-173. NEFSC Ref. Doc. 04-03.
Kim, Y. and E. Powell. 2004. Surfclam histopathology survey along the Demarva mortality line. J. Shellfish Res. 23: 429-441.
Weinberg, J.R. 1999. Age-structure, recruitment, and adult mortality in populations of the Atlantic surfclam, Spisula solidissima, from 1978 to 1997. Marine Biology 134:113-125.
Weinberg, J. R., T. G. Dahlgren, and K. M. Halanych. 2002a. Influence of rising sea temperature on commercial bivalve species of the U.S. Atlantic coast. American Fisheries Society Symposium 32: 131-140. In: Fisheries in a changing climate. N. A. McGinn, ed.
Weinberg, J. R., P. Rago, W. Wakefield, and C. Keith. 2002b. Estimation of tow distance and spatial heterogeneity using data from inclinometer sensors: an example using a clam survey dredge. Fisheries Res. 55:49-61.
We are grateful to the clam industry for their strong support of this survey. In particular we thank the captain and
crew of the F/V Lisa Kim for their professional work. This successful project was carried out cooperatively by
the clam industry, Rutgers University, and NOAA's National Marine Fisheries Service (NMFS).