© 2004 International Council for the Exploration of the Sea
Using fish-processing time to carry out acoustic surveys from commercial vessels
National Institute of Water and Atmospheric Research Private Bag 14-901, Kilbirnie, Wellington, New Zealand
*Correspondence to R. L. O'Driscoll: tel: +64 4 386 0300; fax: +64 4 386 0574. e-mail: r.odriscoll{at}niwa.co.nz.
In some fisheries large factory freezer trawlers have periods of down time as the catch is processed. By utilizing this time, scientific acoustic surveys can be carried out between commercial-fishing operations without compromising fishing success. Examples are presented from three acoustic surveys for hoki (Macruronus novaezelandiae) in New Zealand waters during 2002 and 2003 conducted from a commercial vessel fitted with a scientifically calibrated SIMRAD ES-60 echosounder. These surveys confirmed the presence of a new spawning area for hoki and provided biomass estimates from known fishing grounds. The approach described works well for small-scale acoustic surveys adjacent to areas of high catch rates and is cost-effective because the vessel "pays for itself" by fishing commercially. The major limitation is that the boundaries of the survey area are determined by the time available during processing, which is related to the size of the catch and the time required to search for a suitable location for the next commercial trawl. In the New Zealand hoki surveys, processing time was typically 3–8 h, which was sufficient to carry out about 10–70 km of acoustic transects. Acoustic research was also limited to periods of relatively good conditions by the use of a hull-mounted transducer.
Keywords: acoustic survey, commercial fishing, hoki, New Zealand
Received 21 June 2004; accepted 9 November 2004.
 |
Introduction |
|---|
 Top Introduction MethodsResultsDiscussionReferences |
|---|
Acoustic surveys are used worldwide to provide estimates of fish and zooplankton abundance. Most acoustic surveys are carried out from research vessels using specialized scientific echosounders. However, echosounders capable of collecting scientific-quality data have recently become more widely available and are being installed on commercial fishing vessels. This provides the opportunity to collect acoustic data from these vessels in support of a range of research objectives. Acoustic data collected from commercial vessels have been used in studies of herring (Clupea harengus) in eastern Canada (Claytor and Clay, 2001; Melvin et al., 2002), widow rockfish (Sebastes entomelas) in British Columbia (Stanley et al., 2000), jack mackerel (Trachurus symmetricus murphyi) in Chile (Hector Peña, Institute of Marine Research, Norway, pers. comm.), walleye pollock (Theragra chalcogramma) in Alaska (Martin Dorn and Steven Barbeaux, Alaska Fisheries Science Center, pers. comm.), hoki (Macruronus novaezelandiae) and orange roughy (Hoplostethus atlanticus) in New Zealand (Kloser et al., 2000; Hampton and Soule, 2003; O'Driscoll, 2003; Soule and Hampton, 2003; O'Driscoll et al., 2004a) and Australia (Kloser et al., 2001; Ryan and Kloser, 2002), southern blue whiting (Micromesistius australis) in New Zealand (O'Driscoll and Hanchet, 2004), and toothfish (Dissostichus mawsoni) in Antarctica (O'Driscoll and Macaulay, 2003).
Commercial vessels have several advantages over research vessels as platforms to collect acoustic information. First, there are a large number of commercial vessels available, so their spatial and temporal capacity for data collection may exceed that of national research vessel fleets. Second, commercial vessels are more cost-effective than research vessels if research activities can be combined with commercial fishing which covers part or all of the running cost of the vessel. Finally, use of commercial vessels requires communication and collaboration between fishers and scientists, which is usually beneficial to both parties (Rose, 1997).
There have been three basic strategies for the collection of acoustic data from commercial vessels: i) undirected monitoring of the vessel's acoustic instruments (echosounder or sonar) during fishing operations (e.g., O'Driscoll and Macaulay, 2003); ii) using commercial vessels as "scouts" to locate aggregations, which are then surveyed by a research vessel (e.g., Stanley et al., 2000); and iii) directed acoustic surveys using the commercial vessel as the primary research platform (e.g., Hampton and Soule, 2003).
Undirected data collection is useful for studying the behaviour of fishers and fish, and can provide information on mark types from new areas, but it is difficult to generate abundance estimates from this type of data. A notable exception is in the Canadian herring fishery, where acoustic data collected from gillnet and purse-seine vessels have been used to develop relative abundance indices suitable for stock assessment (Claytor and Allard, 2001; Claytor and Clay, 2001). Directed research surveys which follow a pre-determined survey design are usually more suitable for biomass estimation, but the value of using a commercial vessel instead of a research vessel may be reduced if the vessel needs to stop its commercial-fishing operation to carry out the acoustic survey work.
In New Zealand, large (40–80 m) factory freezer trawlers participate in spawning fisheries for hoki, orange roughy, and southern blue whiting. These vessels target catches of 10–30 t, which are filleted and frozen at sea. Typically vessels can process 2–5 t green weight of fish per hour depending on the species and vessel. During the spawning period, catch rates are usually high, so the vessel does not have to deploy the trawl again immediately following a successful haul. This creates a window of "down time" during processing, when the vessel is not actively fishing. During this processing period, the vessel usually attempts to maintain contact with a known aggregation of fish or searches for new aggregations prior to the next trawl.
By utilizing the down time during processing, we have been able to carry out directed acoustic surveys, without compromising commercial-fishing success. This enables research to be carried out relatively cheaply as part of commercial fishing operations, allowing acoustic surveys of new areas, or at more frequent intervals than those carried out on research vessels. In this paper, we illustrate the utility of this approach using examples from three acoustic surveys for hoki during 2002 and 2003.
|
Methods |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
Background to hoki acoustic surveys
Hoki form New Zealand's largest fishery, with annual catches of 160 000–270 000 t since 1987 (O'Driscoll et al., 2004b). Two stocks of hoki are recognized based on morphometric and growth rate differences (Livingston and Schofield, 1996), and these have been assessed separately since 1990. The western stock resides primarily on the Campbell Plateau, south of New Zealand, and spawns on the west coast of the South Island (Figure 1). The eastern stock's "home ground" is the Chatham Rise, with spawning occurring mainly in Cook Strait (Figure 1). Juvenile hoki of both stocks mix together on the shallower areas of the Chatham Rise, and are believed to recruit to their respective stocks at maturity-at-age 3–8 (Livingston et al., 1997).
|
Acoustic surveys from research vessels have provided abundance indices for spawning hoki on the west coast South Island and in Cook Strait since 1988 (Coombs and Cordue, 1995; O'Driscoll, 2002), and these are an important input to the stock assessment model used to set the total allowable commercial catch (Francis, 2004). Hoki also spawn in other areas (e.g., Livingston, 1990), but there was little information about the magnitude of spawning aggregations away from west coast South Island and Cook Strait.
In 2002, a New Zealand fishing company, Independent Fisheries Limited, approached us with anecdotal evidence that marks of spawning hoki in Pegasus Canyon on the east coast of the South Island (Figure 1) were increasing. As part of a collaborative initiative, funded by the consortium of hoki-quota holders (The Hoki Fishery Management Company Ltd), we carried out the first acoustic survey of this area from 2 to 11 September 2002 using the commercial fishing vessel "Independent 1". This survey was repeated in July–September 2003.
The major hoki fishery is on the west coast South Island, accounting for almost half the catch in recent years (O'Driscoll et al., 2004b). The last acoustic survey of the west coast South Island spawning grounds was in 2000 (Cordue, 2002), and there are no plans for future research vessel surveys in this area. We carried out a pilot acoustic survey from 1 to 16 August 2003 from "Independent 1" to determine the feasibility of using a commercial vessel to survey spawning hoki on the west coast South Island. We also aimed to obtain estimates of the size of the main hoki spawning aggregation in Hokitika Canyon (Figure 1) in 2003.
Survey vessel and acoustic equipment
"Independent 1" is a 45.6 m factory freezer, stern trawler, fitted with a SIMRAD ES-60 echosounder with a hull-mounted 38-kHz split-beam transducer. The echosounder was calibrated on 6 August 2002 and on 26 July 2003, using standard scientific methods (Foote et al., 1987). Details of the acoustic system and its calibration are provided in Table 1. Calibration coefficients were similar (within 0.3 dB) between the two calibrations.
|
Survey design
The survey design in both areas followed the methods of Jolly and Hampton (1990) as adapted by Coombs and Cordue (1995) to obtain a biomass index for transient fish populations. It was similar to that used successfully in previous research-vessel acoustic surveys on the west coast South Island and in Cook Strait (Coombs and Cordue, 1995; O'Driscoll, 2004). Hoki have a long spawning season from July to early September and it is thought that during this period there is a turnover of fish on the spawning grounds (Coombs and Cordue, 1995). We therefore aimed to conduct a number of sub-surveys or "snapshots" of the areas during each survey to obtain estimates of the spawning biomass at different times.
The timing of spawning in Pegasus Canyon is uncertain. In 2002, the survey was carried out over a short time period relatively late in the spawning season (2–11 September). In 2003, the survey design was revised to enable acoustic snapshots to be spread over a wider period of time. "Independent 1" passed through Pegasus Canyon about every 25 days on the way to fishing on the west coast South Island and in Cook Strait. Snapshots were carried out at the beginning or end of trips in July to September. This timing overlapped with a research vessel acoustic survey of spawning hoki in Cook Strait from 17 July to 28 August 2003 (O'Driscoll and McMillan, 2004). Estimates of the timing of spawning on the west coast South Island were available from previous acoustic surveys and from commercial catches (Harley, 2002). These data suggested peak spawning in early August. Accordingly, the survey of Hokitika Canyon was carried out between 1 and 16 August.
Randomly allocated parallel transects were used to estimate the mean fish density in each snapshot. Survey boundaries were based on depth contours with transects oriented to run across canyon features in water depths greater than 200 m in Pegasus Canyon, and 300 m in Hokitika Canyon (Figure 2). Hokitika Canyon was divided into two strata because fishing in the inner canyon, within 25 nautical miles of the coast, is restricted to vessels under 46 m total length. A total of 9–11 transects were allocated to each snapshot of Pegasus Canyon, eight transects to each snapshot of inner Hokitika Canyon (Stratum 5A), and three transects in the outer canyon (Stratum 5B) (Figure 2).
|
Acoustic survey work was conducted in the processing time between commercial trawls as described earlier. Transects were run at 7–10 knots. It was not usually possible to complete an acoustic snapshot within a single processing window, so adjacent transects within an area were sometimes separated by several hours viz. the time required to complete a trawl and return to the next transect. To determine whether this strategy might bias abundance estimates, we conducted an experiment to investigate small-scale (hourly) temporal variation in the hoki distribution in Hokitika Canyon. During this experiment, a single 9.3 km transect (see Figure 2) was run 28 times over a period of 30 h.
Trawling
Commercial trawls were used for mark identification and the collection of biological data. The positions of tows were determined by the fishing officers (captain and first mate) and were usually targeted on relatively dense marks thought to contain hoki. Trawl catch weights and species composition were estimated from the vessel trawl-catch-effort-processing-returns, which give a tow-by-tow breakdown of the catch and are completed by the vessel crew as a legal requirement. A random sample of 100–200 hoki from every tow was measured, and the sex and macroscopic gonad stage determined.
A Seabird SM-37 Microcat conductivity-temperature-depth (CTD) datalogger was mounted on the headline of the net during five trawls in Pegasus Canyon and eight trawls in Hokitika Canyon to collect temperature and salinity data, which were then used to estimate the acoustic absorption coefficient during the survey using the equation of Doonan et al. (2003).
Acoustic data analysis and biomass estimation
Acoustic data were analysed using standard echo-integration methods, as implemented in the New Zealand National Institute for Water and Atmospheric Research Echo Sounder Package (ESP2) software (McNeill, 2001). A systematic, ping-induced variation (triangular wave of 1 dB amplitude with a period of 2721 pings) was identified in ES-60 data in 2003 (Ryan and Kloser, 2004), and a modified version of the ESP2 software which removed this error was used for the analysis.
Echograms were visually examined, and the bottom determined by a combination of an in-built bottom tracking algorithm and manual editing. Regions corresponding to various acoustic mark types were then identified. Marks were classified subjectively, based on their appearance on the echogram (shape, structure, depth, strength, etc.), and using information from commercial trawls. Backscatter from marks (regions) identified as hoki was then integrated to produce estimates of the mean area-backscattering coefficients (Sa) for each transect. No species decomposition of acoustic backscatter was carried out, i.e., backscatter from all hoki regions was assumed to be 100% hoki. This was a reasonable assumption because commercial and research fishing on spawning hoki aggregations typically results in very clean catches of hoki, with little or no bycatch. Transect Sa estimates were converted to hoki biomass using the ratio of mean weight to mean-backscattering cross-section, the linear equivalent of target strength, for hoki. This ratio was calculated from the scaled length frequency distribution of hoki from commercial trawls during the survey. Acoustic target strength (TS in dB) was derived using the TS–length relationship: TS = 18 log10 (L) – 74 (Macaulay, 2001); and mean hoki weight (w in kilogrammes) was determined from the length–weight relationship: w = 0.00000479 L2.89 (Francis, 2003), where L is fish total length in centimetres.
Biomass estimates and variances were obtained for each stratum and snapshot using the formulae of Jolly and Hampton (1990), as described by Coombs and Cordue (1995). The snapshots were averaged to obtain the biomass index (O'Driscoll, 2002). The sampling precision of the biomass index was calculated by assuming the snapshot biomass estimates are independent and identically distributed random variables. The sample variance of the snapshot means divided by the number of snapshots is therefore an unbiased estimator of the variance of the index, the mean of the snapshots. Note that the sampling precision will greatly underestimate the overall survey variability, which also includes uncertainty in survey timing, target strength, calibration, and mark identification (O'Driscoll, 2004).
Noise trials
One noise trial was carried out from "Independent 1" during the 2002 survey and two noise trials were carried out in 2003. The 2002 trial was carried out at vessel speeds of 8 and 10 knots in calm sea conditions in Pegasus Canyon over bottom depths greater than 500 m. The first trial in 2003 was also at 10 knots in calm conditions in Pegasus Canyon over a sloping bottom from 300–845 m. The second trial in 2003 was at 7 and 10 knots in Hokitika Canyon over bottom depths of 500–600 m, when the vessel was pitching into a 2 m swell. In all noise trials, the ES-60 echosounder was operated in passive mode with a ping interval and duration of 2.0 s and 1.024 ms, respectively (see Table 1). Data were recorded to about 1500 m range.
The noise levels were calculated based on the "SA method", taken from the SIMRAD EK500 manual (SIMRAD, 1993), using the following algorithm:
|
| (1) |
, the equivalent two-way beam factor (dB); and 
p, the pulse length (ms). The two-way transmission loss at range, R (m), is given by 2TL = 20log10R + 2
R, where is the absorption loss (dB m–1).
The transducer-specific parameters (Si and Z) for the transducer on the "Independent I" were unavailable and so data for other SIMRAD ES38B transducers were obtained and used to estimate these values. Absorption-loss values () were calculated from CTD data from the two areas (see Table 1). PTX, 10 log10 , and p values were recorded in the ES-60 data files. A summary of the parameters is given in Table 2.
|
|
Results |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
Acoustic data were recorded during 38 days at sea on "Independent 1" in 2002 and 2003. Seven acoustic snapshots of Pegasus Canyon were completed during the 2002 survey (Table 3). In 2003, four snapshots of Pegasus Canyon (Table 3), eight snapshots of inner Hokitika Canyon (Stratum 5A), and three snapshots of outer Hokitika Canyon (Stratum 5B) (Table 4) were carried out. Sea and weather conditions were good (swell height less than 1.5 m, windspeed less than 20 knots) for most of the survey days. Echogram quality deteriorated markedly when there was more than about 25 knots of wind and 2 m of swell, and there was about four days when conditions were deemed too rough for acoustic data collection.
|
|
Time available for acoustic-survey work
"Independent 1" targeted about 15 t of hoki per trawl to maximize product quality and processing efficiency. A SIMRAD ITI trawl-management system allowed accurate catch estimation, and catches were usually close to the target level (n = 110 trawls, mean = 14 t, range = 3–25 t). Most tows were of 15–30 min duration. The catch in both Pegasus and Hokitika Canyons was predominantly spawning hoki, with low levels of bycatch (Pegasus Canyon: n = 42 trawls, mean = 97% hoki by weight, range = 75–100%; Hokitika Canyon: n = 68 trawls, mean = 97% hoki, range = 83–100%). Processing time was related to the size of the catch (Figure 3a). The time between successive trawls was typically 3–8 h for catches of 10–20 t (Figure 3a). This usually allowed 10–70 km of acoustic transects to be completed (Figure 3b) and time for the vessel to return to the area of the next trawl.
|
Not all of the time between trawls was available for acoustic survey work. It typically took 1–2 h to locate and "line up" a suitable mark and then shoot the trawl. This is why no acoustic transects were carried out when the time between successive trawls was less than 2 h (Figure 3b). The variability in the relationship between processing time and the length of acoustic transects completed (Figure 3b) was due to the varying distance between the location of trawls and the start of the next acoustic transect. Commercial trawls were usually concentrated in the part of the survey area that gave best catch rates. For example, all 68 trawls carried out from "Independent 1" in Hokitika Canyon were in Stratum 5A, and most tows were on the southern side of the canyon (see Figure 2). The inner transects in Stratum 5A were adjacent to the area of commercial fishing, so almost all of the available time could be spent running these acoustic transects. However, transects in Stratum 5B were up to 40 km away, and the time required to steam to and from the transect was much greater than the time required to run the transect itself. In practice, the captain of the vessel modified his fishing practice to allow outer transects to be run (Allan Dillon, Independent Fisheries Limited, pers. comm.), by targeting a larger catch (20–25 t) as close as possible to the transect position.
Acoustic biomass estimates
Hoki biomass estimates by snapshot are given in Tables 3 and 4. The average biomass index from Pegasus Canyon was 100 000 t in 2002 and 79 000 t in 2003 (Table 3). It is difficult to compare biomass estimates between years because of the different timing of the two surveys. Snapshots on 11 September in both years gave very similar estimates of abundance (Table 3). Sampling precision (CV) of individual snapshots in Pegasus Canyon ranged between 12% and 55% (Table 3). The variance of the biomass estimates was 17% from the seven snapshots in 2002 and 21% from the four snapshots in 2003.
The estimates from Pegasus Canyon were 35% and 43% of the acoustic-biomass indices from research surveys in Cook Strait in 2002 and 2003, respectively (O'Driscoll and McMillan, 2004), confirming the importance of Pegasus Canyon as a spawning area for eastern stock hoki. The overlapping timing of the two surveys means it is unlikely that the fish in Pegasus Canyon were the same fish observed spawning in Cook Strait, and fish in spawning condition were observed simultaneously at both sites.
Hoki biomass in the inner Hokitika Canyon (Stratum 5A) was 96 000–120 000 t for the first five snapshots in 2003, declining to 52 000–68 000 t in snapshots six, seven, and eight (Table 4). All eight snapshots were averaged to obtain the abundance index for Stratum 5A of 89 000 t. Sampling precision of individual snapshots ranged between 13% and 65% (Table 4). The variance of the biomass estimates from the eight snapshots was 10%. Hoki biomass in outer Hokitika Canyon (Stratum 5B) was estimated as 10 000–27 000 t in three snapshots with snapshot CVs of 2–66% (Table 4). The mean estimate for Stratum 5B was 18 000 t with a variance of 27%.
To compare biomass estimates for Hokitika Canyon with those from the previous west coast South Island acoustic surveys, we re-integrated the 2003 data with an absorption coefficient of 8.0 dB km–1, which was the value, based on the formula of Fisher and Simmons (1977), used in 1988–2000 (Cordue, 2002). Estimates with the old absorption coefficient (Table 5) were about 87% of the estimates in Table 4. The biomass index from Stratum 5A in 2003 was similar to the equivalent estimate from this stratum in the 2000 survey (Table 5). However, the estimate for Stratum 5B was only 30% of the estimate in 2000 and the equal lowest estimate in the time-series (Table 5).
|
Temporal variation in hoki density
Visual examination of echograms from the 28 repetitions of the same experimental transect showed strong and consistent diurnal behaviour. Hoki occurred close to the bottom on the southern side of the canyon during the day, and were higher off the bottom on the north side at night. At dawn and dusk there was a transition period, where hoki were more dispersed and occurred along the whole transect.
These changes in hoki distribution are presented quantitatively in Figure 4, which shows mean hoki height (distance above bottom) and mean hoki latitude over time. During the day, hoki were an average of 30 m above the bottom and close to the southern end of the transect at 42°36.6'S, the transect running from 42°33.6' to 42°37.2'S. At night, hoki were about 70 m off the bottom and further to the north, at about 42°35.6'S.
|
There was a marked change in estimated hoki density associated with these diurnal changes in vertical and latitudinal distribution (Figure 4a). Mean hoki density along the experimental transect during the day (defined as 07:30 to 18:00 NZST; number of transects, n = 12; mean area-backscattering coefficient, Sa = 0.000036) was almost three times lower than mean density at night (n = 16; Sa = 0.000099). We were concerned that the difference in density estimates may be because hoki were somehow less detectable acoustically during the day (e.g., very close to the bottom in the acoustic deadzone, or lower acoustic target strength due to changes in fish orientation). Results from our survey transects did not indicate that this was the case. There was no systematic day–night difference in transect density estimates during the eight biomass snapshots of Stratum 5A. Of the eight highest transect densities, four were observed during the day and four at night. Rather, diurnal differences in mean hoki density along the experimental transect were probably a result of movement of fish into and out of the canyon, longitudinal east–west movement, occurring at the same time as the changes in vertical and latitudinal (north–south) distribution. It was not possible to detect this longitudinal movement using a single transect design.
Noise trials
Estimated noise levels are given in Table 6. The recording from Hokitika Canyon was an order of magnitude (10 dB) noisier than the data collected in Pegasus Canyon. This was because of the difference in sea conditions at the time of the experiments. Data from Pegasus Canyon in 2002 and 2003 were collected in calm conditions, but the recording on the west coast South Island was made when the vessel was pitching into a 2 m swell.
|
Trials at different speeds indicated that "Independent 1" was quieter at 10 knots than at slower speeds. The reduction in the noise level with increasing vessel speed was more apparent in the rougher conditions during the trial in Hokitika Canyon. This was probably due to the physical location of the transducer and the trim of the vessel. The transducer was located well forward and the vessel tended to be trimmed bow-up (up to 5°) to facilitate the draining of the factory deck. At higher speeds the bow "dug in", levelling the trim, reducing pitching, and increasing the depth of the transducer. The bow-up trim of the vessel also increased the incidence of side-lobe echoes, especially when depth was increasing.
|
Discussion |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
By utilizing the processing time between commercial trawls, we were able to carry out three acoustic surveys for New Zealand hoki from a commercial vessel during routine fishing operations. The acoustic research requirements did not appear to seriously compromise fishing success: "Independent 1" recorded a "record trip" during the survey in Hokitika Canyon, filling the holds to capacity in 17 days.
Surveys in 2002 and 2003 provided the first estimates of hoki spawning biomass in Pegasus Canyon. This spawning aggregation was much more extensive than previously thought, with biomass estimates about 35–43% of those from the main eastern spawning ground in Cook Strait. Following the presentation of results of our 2002 survey (O'Driscoll, 2003), interest in this area increased, and the commercial catch from Pegasus Canyon rose from 3000 t in 2002 to 7000 t in 2003 (O'Driscoll et al., 2004b).
The survey of Hokitika Canyon in 2003 provided another estimate of hoki spawning biomass for this area that could be compared with results from research-vessel surveys in 1988–2000. However, results from Hokitika Canyon probably do not provide a reliable estimate of hoki abundance for the entire west coast South Island spawning fishery. Biomass estimates for inner Hokitika Canyon (Stratum 5A) are much flatter than the overall west coast South Island acoustic index, which is derived from a much larger survey area of over 10 000 km2 (O'Driscoll et al., 2004a). This is consistent with the hypothesis that the inner Hokitika Canyon is a "preferred habitat". The density of fish in a preferred habitat may remain relatively constant, even if there is wide variation in abundance (MacCall, 1990). In years when abundance is high, fish "spill-over" into other areas. When abundance is low a much higher proportion of the population occur in the preferred area. The perception of many fishers was that overall abundance on the west coast South Island spawning grounds was low in 2003, yet our acoustic estimate for Stratum 5A was similar to 2000, and vessels fishing in this stratum maintained high catch rates.
The survey approach described in this paper works well for small-scale acoustic surveys adjacent to areas of high catch rates. The major limitation is that the boundaries of the survey area are determined by the time available during the processing window. The length of the processing window depends on the size of the preceding catch, which, in turn, is related to market-driven quality requirements and also the time required to locate a suitable mark for the next commercial trawl. In these hoki surveys, the processing window was typically 3–8 h. This was sufficient to cover an area the size of Pegasus Canyon (333 km2) or inner Hokitika Canyon (254 km2), where high densities of fish were present. It was more difficult to survey areas such as the outer Hokitika Canyon (529 km2) because hoki densities were usually too low to allow the vessel to remain in this stratum and fish commercially. Stratum 5B was surveyed on three occasions, which required careful planning to maximize the processing window, i.e. large catch as close as possible to the survey area.
Other characteristics of the Pegasus Canyon and inner Hokitika Canyon hoki fisheries make them well-suited for this type of survey approach. The distribution of hoki in both areas is related to bathymetry and so the location of aggregations is relatively predictable. This reduces the time required to locate suitable marks for commercial fishing. Aggregations are also dense so tow duration is short. The combination of these factors means that fish can be caught "on demand". In other fisheries with more dispersed aggregations or lower catch rates or both of these features, the down time between trawls is reduced because the vessel must search for aggregations, or because the trawl is shot earlier to allow for the longer tow duration. Hoki also form single-species schools which simplifies mark identification and reduces the need for targeted trawling to sample confusing mark types, which is typically part of acoustic surveys from research vessels.
This survey approach has also been applied to other New Zealand species with similar behaviour and fishing patterns. Hampton and Soule (2003) and Soule and Hampton (2003) successfully carried out two acoustic surveys of an orange roughy spawning plume on the Chatham Rise from the commercial factory trawler "San Waitaki" in 2002 and 2003. Like hoki, orange roughy in this area form a spatially predictable, dense, single-species aggregation occupying a relatively small area (about 100 km2). Catches were consistently large (25 t) and this created sufficient processing time to carry out repeated grid surveys (snapshots) of the spawning plume. O'Driscoll and Hanchet (2004) carried out a pilot acoustic survey of spawning southern blue whiting on the Campbell Plateau in 2003 from the factory freezer trawler "Aoraki". This survey was more difficult because, although southern blue whiting form dense aggregations, the location of these aggregations varies between years and the schools are also mobile prior to spawning. A two-phase strategy was devised, where the approximate boundaries of aggregations were first located by searching and mapping the location of other commercial vessels, and then biomass was estimated using a grid of random parallel transects. Two main aggregations were surveyed over an area of about 2250 km2. Southern blue whiting were slower to process than hoki and orange roughy, so there was usually longer between trawls (6–12 h for catches of 10–20 t).
The results from the repeated transect experiment have implications for survey design. It appears that the distribution of hoki in the inner Hokitika Canyon was relatively stable during the day and again at night, with rapid changes over a short period of about 2 h at dawn and dusk (see Figure 4). These differences were not related to tidal flow. Ideally all transects within a snapshot would be conducted during either day or night to avoid the transition periods at dawn and dusk. This strategy was not feasible with the current design, where transects were run in the processing time between commercial tows. There was never enough time to run all transects in a snapshot during a single processing period, so snapshots took 14–48 h, depending on how many commercial trawls were carried out. Because the timing, as well as the location, of transects was essentially random, the diurnal patterns we observed are unlikely to have biased the biomass estimates: there would have been some risk of "double counting", but a similar risk of "double missing". Temporal variability in spatial distribution, could have contributed to the relatively high sampling uncertainty (snapshot CVs) associated with snapshot biomass estimates (see Tables 3 and 4). This uncertainty is reduced by carrying out a large number of snapshots.
Acoustic data from the SIMRAD ES-60 echosounder were of generally high quality and could be analysed using the same software and methods as data from scientific echosounders. However, acoustic research from commercial vessels will be limited to periods of relatively good weather by their use of hull-mounted transducers. We were fortunate with the weather during our hoki surveys, and there were only four days when conditions were deemed too rough for acoustic-data collection. Noise trials indicated data quality was much worse when the vessel was pitching, so transect direction was sometimes manipulated to avoid running into the swell. Echogram quality was improved by running with the sea, but there would still have been some attenuation of the acoustic signal by wind-induced bubbles (Dalen and Løvik, 1981) which was not apparent from the echograms.
The noise levels for "Independent 1" in calm conditions were below the threshold of 95 dB re 1 µPa (1 Hz band) recommended as the acceptable vessel noise level at 38 kHz (Mitson, 1995), but about 10 dB higher at 1000 m than values reported by Hampton and Soule (2003) for the hull system on "San Waitaki". Ideally, noise levels should be subtracted from the measured backscatter following integration. There was no noise correction applied in this paper. However, a preliminary analysis indicated that the bias in our biomass estimates due to ambient noise was only about 1% in calm conditions. The echosounder transducer on "Independent 1" was also tilted up to 5° forward for much of the survey. The hoki target strength–length relationship (Macaulay, 2001) is based on average fish-tilt angles relative to a level transducer. If the transducer is not level, then the relative tilt-angle distribution of the fish changes and this may introduce a bias to the biomass estimates. In future surveys it would be useful to have pitch-and-roll sensors mounted on the commercial vessel. This would allow transducer orientation to be measured so that a new target strength–length relationship relative to mean transducer attitude could be calculated, and also allow correction for signal loss due to pitch and roll (e.g., Stanton, 1982).
|
Acknowledgements |
|---|
Thanks to Allan Dillon and Trevor Smith (captains), the crew of "Independent 1", and Independent Fisheries Limited for their cooperation. Neil Bagley and Matt Dunn assisted with data collection at sea. Adam Dunford helped with the analysis of results from noise trials. This paper was improved following critical reviews by Roger Coombs, Adam Dunford, and an anonymous referee. Funding was provided by the New Zealand Hoki Fishery Management Company Limited.
|
References |
|---|
|
TopIntroductionMethodsResultsDiscussionReferences |
|---|
-
Claytor R.R. and Allard J. (2001) Properties of abundance indices obtained from acoustic data collected by inshore herring gillnet boats. Canadian Journal of Fisheries and Aquatic Sciences 58:2502–2512.
Claytor R. and Clay A. (2001) Distributing fishing mortality in time and space to prevent overfishing. In Kruse G.H., Bez N., Booth A., Dorn M.W., Hills S., Lipcius R.N., Pelletier D., Roy C., Smith S.J., Witherell D. (Eds.). Spatial Processes and Management of Marine Populations(University of Alaska Sea Grant, Fairbanks, Alaska) pp. 543–558 AK-SG-01-02.
Coombs R.F. and Cordue P.L. (1995) Evolution of a stock-assessment tool: acoustic surveys of spawning hoki (Macruronus novaezelandiae) off the west coast of South Island, New Zealand, 1985–91. New Zealand Journal of Marine and Freshwater Research 29:175–194.[Web of Science]
Cordue P. L. (2002) An analysis of an acoustic survey of spawning hoki off the west coast of South Island during winter 2000. New Zealand Fisheries Assessment Report 2001/26. 51 pp.
Dalen J. and Løvik A. (1981) The influence of wind-induced bubbles on echo-integration surveys. Journal of the Acoustical Society of America 69:1653–1659.[CrossRef][Web of Science]
Doonan I.J., Coombs R.F., McClatchie S. (2003) The absorption of sound in seawater in relation to the estimation of deep-water fish biomass. ICES Journal of Marine Science 60:1047–1055.
Fisher F.H. and Simmons V.P. (1977) Sound absorption in seawater. Journal of the Acoustical Society of America 44:473–482.[CrossRef]
Foote K. G., Knudsen H. P., Vestnes G., MacLennan D. N., Simmonds E. J. (1987) Calibration of acoustic instruments for fish-density estimation: a practical guide. ICES Cooperative Research Report No. 144. 57 pp.
Francis R. I. C. C. (2003) Analyses supporting the 2002 stock assessment of hoki. New Zealand Fisheries Assessment Report 2003/5. 34 pp.
Francis R. I. C. C. (2004) Assessment of hoki (Macruronus novaezelandiae) in 2003. New Zealand Fisheries Assessment Report 2004/15. 95 pp.
Hampton I. and Soule M. (2003) Acoustic survey of orange roughy biomass on the northeast Chatham Rise, 9–22 July 2002. Report to the Orange Roughy Management Company Ltd, November 2002. Deepwater Working Group Document 2003/02. 63 pp. Unpublished report held by the Ministry of Fisheries, Wellington, New Zealand.
Harley S. J. (2002) Estimation of residence time for hoki on the Cook Strait and west coast South Island spawning grounds. New Zealand Fisheries Assessment Report 2002/1. 55 pp.
Jolly G.M. and Hampton I. (1990) A stratified, random-transect design for acoustic surveys of fish stocks. Canadian Journal of Fisheries and Aquatic Sciences 47:1282–1291.
Kloser R. J., Ryan T. E., Williams A., Lewis M. (2001) Development and application of a combined industry/scientific acoustic survey of orange roughy in the Eastern Zone: final report. CSIRO Marine Research, Hobart, Australia FRDC Project-99/111. 210 pp.
Kloser R. J., Ryan T. E., Williams A., Soule M. (2000) Development and implementation of an acoustic survey of orange roughy in the Chatham Rise spawning box from a commercial factory trawler, FV Amaltal Explorer. CSIRO Marine Research, Hobart, Australia. 90 pp.
Livingston M.E. (1990) Spawning hoki (Macruronus novaezelandiae Hector) concentrations in Cook Strait and off the east coast off the South Island, New Zealand, August–September 1987. New Zealand Journal of Marine and Freshwater Research 24:503–517.[Web of Science]
Livingston M.E. and Schofield K.A. (1996) Stock discrimination of hoki (Macruronus novaezelandiae, Merluccidae) in New Zealand waters using morphometrics. New Zealand Journal of Marine and Freshwater Research 30:197–208.[Web of Science]
Livingston M.E., Vignaux M., Schofield K.A. (1997) Estimating the annual proportion of non-spawning adults in New Zealand hoki, Macruronus novaezelandiae. Fishery Bulletin US 95:99–113.
Macaulay G. (2001) Estimates of the target strength of hoki. Final Research Report for the Ministry of Fisheries Research Project HOK1999/03 Objective 3. 12 pp. Unpublished report held by the Ministry of Fisheries, Wellington, New Zealand.
MacCall A.D. (1990) Dynamic Geography of Marine Populations(University of Washington Press, Seattle) 153 pp.
McNeill E. (2001) ESP2 phase 4 user documentation. NIWA Internal Report 105. 31 pp. Unpublished report held by the National Institute for Water and Atmospheric Research, Wellington, New Zealand.
Melvin G., Li Y., Mayer L., Clay A. (2002) Commercial fishing vessels, automatic acoustic-logging systems and 3D data visualization. ICES Journal of Marine Science 59:179–189.
Mitson R. B. (1995) Underwater noise of research vessels: review and recommendations. ICES Cooperative Research Report No. 209. 61 pp.
O'Driscoll R. L. (2002) Review of acoustic-data inputs for the 2002 hoki stock assessment. New Zealand Fisheries Assessment Report 2002/36. 67 pp.
O'Driscoll R. L. (2003) Acoustic survey of spawning hoki off the east coast South Island in September 2002. New Zealand Fisheries Assessment Report 2003/28. 26 pp.
O'Driscoll R.L. (2004) Estimating uncertainty associated with acoustic surveys of spawning hoki (Macruronus novaezelandiae) in Cook Strait, New Zealand. ICES Journal of Marine Science 61:84–97.[CrossRef][Web of Science]
O'Driscoll R. L., Bagley N. W., Macaulay G. J., Dunford A. J. (2004a) Acoustic surveys of spawning hoki off South Island on FV Independent 1 in winter 2003. New Zealand Fisheries Assessment Report 2004/29. 48 pp.
O'Driscoll R. L. and Hanchet S. M. (2004) Acoustic survey of spawning southern blue whiting on the Campbell Island Rise from FV Aoraki in September 2003. New Zealand Fisheries Assessment Report 2004/27. 31 pp.
O'Driscoll R. L. and Macaulay G. J. (2003) Descriptive analysis of acoustic data collected during the 2003 exploratory fishery for toothfish in the Ross Sea. CCAMLR WG-FSA-03/28. 25 pp. Unpublished report held by the Commission for the Conservation of Antarctic Marine Living Resources, Hobart, Australia.
O'Driscoll R. L. and McMillan P. J. (2004) Acoustic survey of spawning hoki in Cook Strait during winter 2003. New Zealand Fisheries Assessment Report 2004/20. 39 pp.
O'Driscoll R. L., Phillips N. L., Ballara S. L., Livingston M. E., Ayers D. (2004b) Catches, size, and age structure of the 2002–03 hoki fishery, and a summary of input data used for the 2004 stock assessment. New Zealand Fisheries Assessment Report 2004/43. 75 pp.
Rose G.A. (1997) The trouble with fisheries science!. Reviews in Fish Biology and Fisheries 7:365–370.[CrossRef][Web of Science]
Ryan T. E. and Kloser R. J. (2002) Analysis of industry acoustic observations of blue grenadier off the west coast of Tasmania. CSIRO Marine Research, Hobart, Australia. 40 pp.
Ryan T. E. and Kloser R. J. (2004) Improving the precision of ES60 and EK60 echosounder applications. Report of the Working Group on Fisheries Acoustic Science and Technology (WGFAST)20–23 April 2004Gydnia, Poland ICES CM 2004/B:06.
SIMRAD. (2003) EK500 scientific echosounder, P2170E(SIMRAD, Bergen, Norway) software version 4.01, September 1993.
Soule M. and Hampton I. (2003) Acoustic survey of orange roughy spawning plume on the northeast Chatham Rise New Zealand. Report to the Orange Roughy Management Company Ltd, July 2003. Deepwater Working Group Document 2003/66. 59 pp. Unpublished report held by the Ministry of Fisheries, Wellington, New Zealand.
Stanley R.D., Kieser R., Cooke K., Surry A.M., Mose B. (2000) Estimation of a widow rockfish (Sebastes entomelas) shoal off British Columbia, Canada as a joint exercise between stock-assessment staff and the fishing industry. ICES Journal of Marine Science 57:1035–1049.
Stanton T.K. (1982) Effects of transducer motion on echo-integration techniques. Journal of the Acoustical Society of America 72:947–949.[CrossRef][Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||