© 2004 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
RoxAnn bottom classification system, sidescan sonar and video-sledge: spatial resolution and their use in assessing trawling impacts
a Institute of Marine Research PO Box 1870 Nordnes, N-5817 Bergen, Norway
b Department of Fisheries and Marine Biology, Bergen High-Technology Centre N-5020 Bergen, Norway
*Correspondence to O.-B. Humborstad; tel: +47 55236939; fax: +47 55236830. e-mail: odd-boerre.humborstad{at}imr.no.
Three complementary seabed characterization tools with different spatial resolution were used to locate a research site and to assess physical effects of experimental otter trawling in the Barents Sea: an acoustic seabed classification system (RoxAnn), sidescan sonar and a video-sledge. The marine protected area (MPA) around Bear Island was chosen as it offered unfished reference sites. The area was topographically complex which resulted in certain challenges for choice of the experimental site due to the requirements of representativity and homogeneity and suitable sampling substrate. Systematic waylines with RoxAnn gave broad-scale patterns of bottom conditions, the more informative sidescan revealed topographic reliefs, whilst detailed information on sediment composition and small-scale seabed features was provided by the video-sledge. Accurate positioning of towed gears (trawl, sidescan and video-sledge) ensured unbiased data acquisition. Trawl doors and rockhopper gear created furrows that were visible by sidescan sonar and video. Intensive trawling also caused changes in the acoustic properties by increasing roughness and decreasing hardness. Results are consistent with a possible resuspension of the sediment and a homogenizing effect from the trawl doors and ground gear ploughing the area. The suitability and advantages of using spatially overlapping tools in trawl impact studies are discussed.
Keywords: impact studies, ITI positioning system, otter trawling, RoxAnn, sidescan sonar, video-sledge
Received 13 December 2002; accepted 3 October 2003.
 |
Introduction |
|---|
 Top Introduction Materials and methodsResultsDiscussionConclusionsReferences |
|---|
The environmental effects of fishing have aroused a growing degree of interest during the past few decades (Hall, 1999). The most obvious effect of fishing has been the decline in many major fish stocks (Myers et al., 1996), and fishing down the marine food web (Pauly et al., 1998), but secondary effects on non-target species and habitats have also caused growing concern (Kaiser and De Groot, 1999). Habitat alteration is believed to be having serious consequences for many species of invertebrates and fishes (see e.g. Jennings and Kaiser, 1998; Langton and Auster, 1999).
In order to meet the requirements of the fishing industry and its need to exploit new and deeper areas with rougher bottom conditions to meet the increasing competition for limited fish resources, heavier gears have gradually evolved (Van Beek et al., 1990; Jones, 1992), which in turn leaves fewer areas and habitat types undisturbed (Mortensen et al., 2000). The Barents Sea covers approximately 1.2 million km2, where bottom conditions vary widely from sand and mud to coarse ground consisting of cobble and rock (Figure 1), and which is the home of an important demersal trawl fishery. The majority of impact studies to date have been performed on relatively flat sandy and muddy sediments (Collie et al., 2000), although some have reported effects on rougher bottom types such as pebble, cobble and boulders (Collie et al., 1997; Freese et al., 1999). Established methods, tools and equipment for studies of environmental effects of fishing on relatively flat homogeneous bottoms may not therefore be applicable to these more topographically complex habitats, and an evaluation of such methods is therefore needed to ensure unbiased data acquisition.
|
An 18 month project with the aim of studying the impact and recovery of demersal bottom trawling started in May 2000 in the Barents Sea. In order to perform a quantitative BACI (before/after, control/impact) design for biological sampling, the experimental area had to fulfil the requirement of no influence by any recent previous fishing activity. Furthermore the area should be representative for commercial trawling grounds in terms of species composition, bottom habitat, topography, depth distribution and also be suitable for biological sampling. The marine protected area (MPA) around Bear Island in the Barents Sea (established 1977) was chosen as the only site, which could provide true undisturbed areas for control. Here we describe and discuss the resolution and spatial variability seen from the sidescan, video-sledge and RoxAnn data, and how to use the combined information for experimental site location, unbiased sampling and physical impact assessment. To the best of our knowledge this is the first investigation examining the environmental effects of fishing on the seabed in the Barents Sea.
|
Materials and methods |
|---|
|
TopIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Experimental area
The experiment was conducted from 10 to 24 May 2000 at 74°30'N 18°12'E, 9 nautical miles west of Bear Island in the Barents Sea (Figure 1). The exact location of the experimental area was based on acoustic mapping from the RoxAnn bottom classification system, sidescan sonar recordings and video-sledge observations (see descriptions below).
We selected one area (Ti) for intensive fishing with 10 continuous overlapping trawl tracks within a 200 m wide, 2.2 km long corridor, and one moderate fishing area (Tm) with 10 continuous overlapping trawl tracks within a 600 m wide, 2.2 km long corridor (Figure 2). All trawling was done in the east–west direction, evenly distributing effort with parallel tracks inside the corridors. Three reference areas of 100 m width (C1–3) with no fishing activity were also located between the trawled corridors. The distance between corridors was 200 m. Trawling was carried out by a commercial trawler equipped with a standard bottom trawl (Cotesi maxi 404/A) with Rockhopper gear (21'' diameter, 19 m long), bobbins (11'' diameter, 37.5 m on each side towards the sweeps), sweeps (140 m), otter boards (2300 kg each) and a door spread of approximately 140 m measured by Scanmar sensors.
|
Instrumentation
The ultrasonic processor, RoxAnn, offers real-time classification of seabed features via processing of signals from the ship's echo sounder. RoxAnn was connected to a Simrad EK 500 scientific echo sounder connected to a 38 kHz split-beam transducer positioned on the vessel's centreboard, and gathered data at 5-s intervals. RoxAnn discriminates between types of seabed material, and output data in a digital format ready for computer analysis, E1 (roughness) and E2 (hardness), are obtained by integrating different parts of the first and second echoes (Caddel, 1998). The system was calibrated during a test survey in a fjord outside Bergen in southwestern Norway where bottom conditions were known in detail, giving the track of the ship a colour on a mapping plotter that corresponded to the substrate below it (Figure 3). Additional ground-truthing was subsequently performed by acquiring sledge samples, video and sidescan sonar recordings at the study site. At approximately 95 m depth, the beam has a width of about 12 m and a footprint of approximately 106 m2 (at 7° 3 dB point).
|
A Simrad MS992 sidescan sonar was used for detailed monitoring of the seabed. A towfish operating at 120 kHz was used for acoustic observations and the data were recorded as hardcopy prints and DAT tapes. The towfish, which carried sensors for depth, bearing, temperature, pitch and roll, was towed about 10–30 m above the bottom and collected high-resolution data on bottom structure and topography from a sector covering about 200 m of the seabed on each side. A Simrad Integrated Trawl Instrumentation (ITI) sensor was connected to the wire for exact positioning.
The ITI system is a cordless trawl positioning and monitoring system. The system, which was used to position the trawl, video-sledge and sidescan sonar into the pre-selected corridors, is based on hydro-acoustic communication between a transducer mounted on the vessel and a sensor mounted on the device to be positioned. The sensor determined the position of the different devices relative to the vessel by measuring the depth, the distance to the vessel and the angular deviation of its position relative to the heading of the vessel. The system is operated using a menu of commands displayed on a colour video screen. The ITI accuracy is ±5 m in straight-line length calculation between ship and sensor and ±1° error in bearing (ITI Technical specification).
Sampling
A van Veen grab was tested for possible use as the main sampling tool, but it was soon recognized that it could not assure quantitative samples as it collected small unequal amounts of the substrate, and also suffered from the inability of visual supervising. In order to catch more mobile fauna an Agassiz trawl (a type of beam trawl) was tested. However, this gear was not appropriately constructed and was smashed and destroyed during a tow on coarse ground. A sledge (Sneli, 1998) was finally chosen to provide quantitative samples of benthic assemblages. The sampling sledge is designed to sample epifauna and parts of the infauna in the upper few centimetres of soft bottom. The sledge measures 200x80x20 cm and weighs about 80 kg. In order to prevent the sledge from flipping over, 10 buoys (12'') were connected on top of the sledge and four weights (10 kg each) to the lower part. A camera and light was placed in front of the sledge in order to monitor the seabed habitat and sledge performance. Batteries and a recording unit were placed on top of the sledge. Sledge samples data will be used to analyse the effect of trawling on epifauna in a future publication, whereas sledge data are used here for ground-truthing purposes.
|
Results |
|---|
|
TopIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Seabed description
An area of approximately 6.8x4.4 km on the northwest side of Bear Island was mapped using the RoxAnn system with tracks 300–400 m apart zigzagging the area (approximately 8% coverage). Track lines with different colours indicated variations in substrate types within this area (Figure 3). Hardpacked sand/mud and sand/gravel/stones were the only bottom types in the area as revealed by RoxAnn. The two bottom types were spatially separated and a smaller research area of 1.9x2.3 km with a more uniform bottom substrate was selected for further characterization. Additional tracks were conducted in the east–west direction giving Ti an estimated coverage of 54% and Tm 15%. No additional tracks were made in reference corridors due to time constraints. The mean depth of the research area was 94±7 m, and RoxAnn indicated a substrate that consisted of sand and mud according to the preset calibration. This area was divided into five corridors as described above (Figure 2).
Observations from 16 sidescan tows in the east–west direction, covering the whole research site, showed that the research area was not as uniform as the RoxAnn data indicated, but consisted of 5–6 parallel ridges (30–90 m wide) of rougher material oriented in a southwest/northeast direction in Ti (Figure 4) and dispersed smaller patches in Tm and reference corridors. By combining the sidescan prints the bottom could be seen like a snapshot throughout the research site.
|
Video and sledge samples revealed that the rougher areas consisted of small stones and rocks, occasionally up to 1 m diameter, in contrast to the areas in between which consisted of mixed silt/sand, gravel and shell fragments from Mya truncata and Balanus spp., often with a patchy distribution along the direction of the haul. Few signs of sessile three-dimensional building fauna were observed in the areas in between the ridges, yet some structurally forming fauna (erect hydroids, encrusting cirripeds and bryozoans) was seen at the ridges which provided hard substrate for their attachment. Visually recognized fauna from video was mostly echinoids, ophiuroids, and a range of mobile free-swimming crustaceans, cephalopods and small pelagic and demersal fish escaping in front of the sledge.
Sampling and positioning
The video recordings enabled us to perform a critical evaluation of every sledge haul in addition to visually confirming that the samples after trawling originated in an impacted area since trawl marks were readily seen at several locations. Patches of rough substrate (rocks and stones), sledge speed (both too low and too high) and wave action (during periods of fresh breeze with resulting wave height up to 2.5 m) had an adverse effect on sledge performance. In periods of harsher weather than fresh breeze no sampling could be adequately carried out. Of a total of 102 hauls, 65% were regarded as successful and suitable for quantitative analyses.
Positioning the trawl in the corridors Ti and Tm was done by inspection of real-time ITI positional data and gave a 100% success rate. This exercise gave Ti trawl coverage of 700% and Tm a coverage of 230%, calculated from the distance between otter boards (140 m). The coverage calculated from the width of the rockhopper gear and bobbins (approximately 40 m) gave 200% and 67% for Ti and Tm, respectively.
The average speed of the sledge hauls was approximately 0.5 m s–1 and the mean haul length was 65 m. No hauls were made outside the corridors, nor were two hauls made in the same position, so that all hauls were successful with respect to location. The differences in pre-selected and actual positions at bottom contact for the sledge hauls, as determined by ITI, are given in Figure 5. In terms of latitude the deviation was small, with 60% being less than 20 m away from the desired location. Along the towing direction and the corridors' longitude, 79% of hauls was more than 60 m away from the pre-selected longitude.
|
Post-processing and immediate physical effects
Post-processing of pre-trawling RoxAnn data showed only minor differences in mean values in roughness (E1) and hardness (E2) values between the corridors (Table 1). These parameters indicated a hard bottom devoid of any large structures, although later ground-truthing observations altered this interpretation. Some of the variations, especially the peaks in roughness in the intensive area (Figure 6), could be explained by comparing it with the position of ridges from the sidescan recordings (Figure 4). However, hardness failed to show increasing values at the ridges. The values of hardness were generally high before trawling, in that 49% and 42% for the intensive (Ti) and moderate (Tm) fishing areas, respectively, had values that reached the upper voltage capacity (4.095 V) of the RoxAnn system. Figures changed to 42% and 50% after trawling.
|
|
The spatial variation in E1 and E2 values corresponded well before and after trawling (Figure 6). In Ti, trawling caused an increase in surface relief (E1: sign-test, z=4.5643, p<0.001) and a decrease in sediment hardness (E2: z=3.8340, p<0.001). Mean difference overall in Ti before and after impact was 0.01 and 0.11 V for E1 and E2, respectively. In Tm, no differences between pre- and post-trawling values were found (E1: z=1.6431, p=0.10; E2: z=0.5477, p=0.58). The number of observations from control transects was too low to run any tests.
Sidescan sonar recordings showed no evidence of physical disturbance prior to trawling. After trawling, the tracks from otter boards were highly visible except at the ridges. Parallel tracks could often be seen about 140 m apart corresponding to the door spread. In several places, smaller depressions made by the rockhopper gear were also visible. In Figure 7b (lower part), at least 8 of a total of 10 hauls can be identified by the marks made by the otter boards.
|
From the video, trawl door tracks were seen as U formed depressions approximately 10 cm deep and 20 cm width and an adjacent rounded berm of sediment set off at one of the sides approximately 10 cm high. The rockhopper marks evident from sidescan sonar could, however, not be detected from the video with certainty.
|
Discussion |
|---|
|
TopIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Area description
We first employed the RoxAnn bottom classification system for systematic selection of a suitable homogeneous (in order to reduce variability) research area on a large spatial scale (kilometres). Although its performance has been found to be dependent on vessel speed (Hamilton et al., 1999), RoxAnn can be operated at high speed (10 knots in this study), and with wide spacing between track lines it covers a large area in a short period of time (Magorrian et al., 1995). A weakness was that with the preset calibration of RoxAnn, the recordings indicated a flat seabed consisting of hardpacked sand not showing the patchiness of rougher material present in the experimental site. However, post-processing of the RoxAnn data showed some variations along the research site that could be attributed to the ridges observed by sidescan and video (Figure 6). This deficiency may be due to the size of these patches being small relative to the size of the footprint (106 m2) of the echo sounder beam, or to RoxAnn bottom classes being difficult to define, as suggested by Hamilton et al. (1999). The reason why the ridges did not show up as harder bottom may be due to the system capacity, but also growth of biota (erect hydroids, encrusting cirripeds and bryozoans), which may decrease hardness. The use of rectangular boxes for bottom classification is a very crude way of grouping data and may result in misidentification (Greenstreet et al., 1997). The RoxAnn system, however, is cost-effective and has the advantage of offering a rapid classification of the seabed structure. Furthermore, RoxAnn data can be collected while the vessel is performing other activities.
Sidescan sonar was primarily intended to assure that the trawl disturbances had made the predetermined impact, and to reveal how the sediment surface structure was affected by the different components of the trawl. At least 8 tracks out of 10 were observed in the intensively trawled corridor, and it is likely that the latter two were disguised by consecutive trawl hauls (Friedlander et al., 1999). By combining the sidescan prints, the bottom could be seen like a snapshot, enabling us to get a good impression of the actual bottom contours and in turn aiding us when selecting areas for sampling with the sledge by avoiding the coarse stony areas. To give sonograms of sufficiently high resolution, the sonar had to be towed close to the bottom at a speed of 3–5 knots. This operation demanded the data provided by the RoxAnn on bottom topography and depth in order to plan waylines that would not be obstructed by rapidly changing depth or structures on the bottom. Sidescan sonars cover large areas in a short time (Brown et al., 2002) as RoxAnn does, but its operation demands the full attention since it is a vulnerable and expensive towed gear. Details on topography were better than revealed by RoxAnn, however, post-processing of sidescan data is a complicated task that involves image-processing software, and trawl tracks still have to be counted manually (Friedlander et al., 1999).
Neither the sidescan nor the RoxAnn data provided information about the actual composition of the seabed, which could only be provided by ground-truthing video observations and benthic samples. The video-sledge gave the final and detailed characterization of the area consisting of silt/sand and shell debris interspersed with longitudinal patches of gravels and stones. This heterogeneous pattern complicated sampling since only samples from the soft bottom were thought to be quantitative, and the assemblages were likely to differ between different types of bottom. Here, video was crucial in order to select only the hauls made outside the stony areas. The samples from the video-sledge also showed that the benthic assemblage was suitable for quantitative analyses. In a preliminary study, two sites further south were rejected, one because the fauna was too sparse and the other because the animals found could not be sampled quantitatively with the Sneli sledge.
Seabed description and impact assessment normally involve one or more of the sampling techniques we applied in this study (Kaiser and Spencer, 1994; Schwinghamer et al., 1998; Tuck et al., 1998). Conventional sampling may, however, provide an incorrect impression of the distribution of epibenthic communities (Magorrian et al., 1995). Our results clearly demonstrate that single tools cannot provide a decisive basis alone without comparison across observational platforms. The RoxAnn was capable of a broad-scale, rapid mapping of potential areas for conducting our experiment, and showed where the least rough areas were situated within the MPA. The sidescan revealed smaller features, although the height and size could not be provided from this tool. Sediment samples had to be collected in order to get detailed information. After ground-truthing we could zoom out and generalize over larger areas based on RoxAnn and sidescan data. With RoxAnn, we could cover a larger area within the MPA borders than with the other tools. Had we not used the sidescan, we would not have been able to evaluate the spatial distribution of the ridges that were not sufficiently mapped by RoxAnn. Finally there would have been a great risk of misinterpreting the spatial patterns of the area, if we only had used the video-sledge, which was essential for ground-truthing.
Positioning and sampling
Working in small research areas on tight time schedules (Figure 2) in order to avoid spatial and temporal differences between treatment and control sites (e.g. Morrisey et al., 1992a, b; Underwood, 1992; Hewitt et al., 2001) make accurate real-time observations of position of towed gears a necessity. The positioning of the trawl relative to the vessel has been shown to vary greatly between and within hauls, depending on the vessel's heading to wind and current and on warp length (Engås et al., 2000). Therefore, the ship's position could not be used to determine the position of the trawl and the sledge. Deviations of the position of the sledge in the longitudinal direction were not considered crucial, since no hauls overlapped and a good coverage of the transects was achieved. However, precise positioning in the latitudinal direction was crucial, as the smallest corridors were only 100 m wide and therefore substantial effort was put into this task. This was achieved by towing the sledge along the corridors in the east–west direction.
Generally, grab samples have been chosen in impact studies since they are known to be more quantitative than sledges and other towed gears. However, grabs are not suitable for sampling patchy distributed and low abundance fauna (Bergman and Van Santbrink, 1994). The main difficulty in obtaining quantitative sampling with towed gears seems to accurately determine the time the gear is in contact with the seabed and the length of a tow (Rice et al., 1982). Towing distances have previously been estimated by odometer wheels (Collie et al., 1997; Prena et al., 1999), although their accuracy is questionable due to their variable performance on soft sediments (Carney and Carey, 1980; Prena et al., 1996; M. J. N. Bergman pers. comm.). In our study, distance measurements of sledge hauls were made by post-processing of ITI data in conjunction with video observations of the time with bottom contact, which allowed us to get better and more reliable area estimates compared to estimates based on the duration of the hauls.
Post-processing and physical effects
Trawling was shown to cause physical disturbance in the intensively (700%) and moderately (230%) trawled area. Furrows and berms created by the trawl doors were clearly visible both on video recordings (single trawl door tracks) and sidescan sonograms (view of many tracks). Sidescan sonar was also used to demonstrate door tracks in a similar trawling experiment which was carried out on the Grand Banks off Newfoundland (Schwinghamer et al., 1998) and in other areas (Service and Magorrian, 1997; Tuck et al., 1998; Friedlander et al., 1999). Reports of scouring depths up to 0.3 m exist (Krost et al., 1990; Jones, 1992). However, the depth and longevity of furrows are dependent on the sediment type (DeAlteris et al., 1999), but also on weight, cable length, angle of attack, door type and depth. Reported scouring depths of trawl doors on sandy seabeds are in the 0–5 cm range (Brylinsky et al., 1994; Gilkinson et al., 1998). In our experiment the scouring depth was up to 10 cm, and is consistent with a sediment type between sand and mud, and the heavier trawl doors used in our experiment.
Physical effects of trawling were also reflected in the RoxAnn data in that overall hardness decreased after trawling, indicating that the hardpacked shellsand was resuspended and made less compact by the trawl (Churchill, 1989; Pilskaln et al., 1998). Most studies have concentrated on sediment biota and studies that examine the contribution of fishing to sediment resuspension are urgently required (Kaiser et al., 2002), hence the RoxAnn system may be an important tool for documentation of this phenomenon. RoxAnn also indicated an increase in surface roughness, and this observation was supported by the sidescan sonograms that showed small depressions made by the rockhopper gear in addition to the door marks. The ability to detect any changes with RoxAnn was dependent on the level of effort deployed, as consistent changes were not demonstrated for the moderately trawled area. Changes in sediment surface characteristics following intensive trawling were also demonstrated by RoxAnn data in similar experiments conducted on the Grand Banks (Schwinghamer et al., 1998) and in a Scottish sea loch (Tuck et al., 1998).
Effects of trawling are related to the level of natural variations (Auster and Langton, 1999; DeAlteris et al., 1999), and 5 months after the trawl disturbance, the door tracks could not be seen from either sidescan or video recordings (authors' observations). The shallow depth in the experimental area suggests a possible weather induced sediment transport (Pfirman, 1985; Amos and Judge, 1991 (cited in Solheim and Elverhøi, 1996)) in addition to winnowing by strong currents in the area (Huthnance, 1981; Solheim and Elverhøi, 1996), both factors contributing to a possible high level of natural variation in the sandy habitat. Sand is generally a mobile sediment type and long lasting topographical features are not expected (DeAlteris et al., 1999).
|
Conclusions |
|---|
|
TopIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
The need for spatial overlapping and complementary tools and instrumentation, to be able to evaluate any possible trawling impact on bottom fauna has been demonstrated. We used three different tools all of which provided valuable information within their respective resolutions, zooming and covering across in an area that fulfilled the preset requirements for the research site. Conclusions based on observations with a single tool in isolation would have lead to misinterpretation. In areas where detailed information of bottom conditions is not known, or where the topography is expected to vary, our approach should be appropriate. In areas of known homogenous substrate (e.g. Prena et al., 1999) this experimental setup may not be necessary for appropriate sampling, but may well be used for quantification of physical effects (this study; Schwinghamer et al., 1998). Accurate positioning of the towed equipments and visual observations during sampling is crucial for unbiased data acquisition.
|
Acknowledgements |
|---|
We thank skippers and crew of survey vessels. Jon-Arne Sneli for lending us the sledge. Tore Høisæter for planning and experimental design. Bjørn Totland, Svein Floen and Ingvald Svellingen for being helpful with instrumentation and post-processing. Anne Britt Skaar Tysseland for preparation of the map. Students and volunteers for help during cruises. The Norwegian Research Council for funding this study.
|
References |
|---|
|
TopIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
-
Amos C.L. and Judge J.T. (1991) Sediment transport on the eastern Canadian continental shelf. Continental Shelf Research 11:1037–1068.[CrossRef][Web of Science]
Auster P. J. and Langton R. W. (1999) In Benaka L. R. (Ed.). The effects of fishing on fish habitat. Fish Habitat: Essential Fish Habitat and Rehabilitation. American Fisheries SocietySymposium 22, Bethesda, Maryland pp. 150–187 400 pp.
Bergman M.J.N. and Van Santbrink J.W. (1994) A new benthos sledge. Netherlands Journal of Sea Research 33:129–133.
Brown C.J., Cooper K.M., Meadows W.J., Limpenny D.S., Rees H.L. (2002) Small-scale mapping of sea-bed assemblages in the eastern English Channel using sidescan sonar and remote sampling techniques. Estuarine, Coastal and Shelf Science 54:2263–278.[CrossRef]
Brylinsky M., Gibson J., Gordon D.C. Jr. (1994) Impacts of flounder trawls on the intertidal habitat and community of the Minas Basin, Bay of Fundy. Canadian Journal of Fisheries and Aquatic Sciences 51:650–661.
Caddel S.E. (1998) Application of an acoustic sea floor classification system for benthic habitat assessment. Journal of Shellfish Research 17:51459–1461.[Web of Science]
Carney R.S. and Carey A.C. (1980) Effectiveness of metering wheels for measurement of area sampled by beam trawls. Fishery Bulletin 78:791–796 National Oceanic and Atmospheric Administration of the United States.[Web of Science]
Churchill J.H. (1989) The effect of commercial trawling on sediment resuspension and transport over the Middle Atlantic Bight continental shelf. Continental Shelf Research 9:841–864.[CrossRef][Web of Science]
Collie J.S., Escanero G.A., Valentine P.C. (1997) Effects of bottom fishing on the benthic megafauna of Georges Bank. Marine Ecology Progress Series 155:159–172.[Web of Science]
Collie J.S., Hall S.J., Kaiser M.J., Poiner I.R. (2000) A quantitative analysis of fishing impacts on shelf-sea benthos. Journal of Animal Ecology 69:785–798.[CrossRef][Web of Science]
DeAlteris J., Skrobe L., Lipsky C. (1999) The significance of seabed disturbance by mobile fishing gear relative to natural processes: a case study in Narragansett Bay, Rhode Island. In Benaka L. R. (Ed.). Fish Habitat: Essential Fish Habitat and Rehabilitation pp. 224–237 American Fisheries Society, Symposium 22, Bethesda, Maryland. 400 pp.
Elverhøi A. and Solheim A. (1983) The physical environment. Western Barents Sea, sheet A. Surface Sediment DistributionNorwegian Polar Institute 23 pp.
Engås A., Godø O.R., Jørgensen T. (2000) A comparison between vessel and trawl tracks as observed by the ITI trawl instrumentation. Fisheries Research 45:297–301.[CrossRef][Web of Science]
Freese L., Auster P.J., Heifetz J., Wing B.L. (1999) Effects of trawling on seafloor habitat and associated invertebrate taxa in the Gulf of Alaska. Marine Ecology Progress Series 182:119–126.[Web of Science]
Friedlander A.M., Boehlert G.W., Field M.E., Mason J.E., Gardner J.V., Dartnell P. (1999) Sidescan-sonar mapping of benthic trawl marks on the shelf and slope off Eureka, California. Fishery Bulletin 97:786–801.[Web of Science]
Gilkinson K., Paulin M., Hurley S., Schwinghamer P. (1998) Impacts of trawl door scouring on infaunal bivalves: results of a physical trawl door model/dense sand interaction. Journal of Experimental Marine Biology and Ecology 224:2291–312.[CrossRef][Web of Science]
Greenstreet S.P.R., Tuck I.D., Grewar G.N., Armstrong E., Reid D.G., Wright P.J. (1997) An assessment of the acoustic survey technique, RoxAnn, as a means of mapping seabed habitat. ICES Journal of Marine Science 54:939–959.
Hall S.J. (1999) The Effects of Fishing on Marine Ecosystems and Communities(Blackwell Science, Oxford) 274 pp.
Hamilton L.J., Mulhearn P.J., Poeckert R. (1999) Comparison of RoxAnn and QTC-View acoustic bottom classification system performance for the Cairns area, Great Barrier Reef, Australia. Continental Shelf Research 19:1577–1597.[CrossRef][Web of Science]
Hewitt J.E., Thrush S.E., Cummings V.J. (2001) Assessing environmental impacts: effects of spatial and temporal variability at likely impact scales. Ecological Applications 11:51502–1516.[CrossRef][Web of Science]
Huthnance J.M. (1981) Large tidal currents near Bear Island and related tidal energy losses from the North Atlantic. Deep Sea Research 28:151–70.[CrossRef]
Jones J.B. (1992) Environmental impact of trawling on the seabed: a review. New Zealand Journal of Marine and Freshwater Research 26:59–67.[Web of Science]
Jennings S. and Kaiser M.J. (1998) The effects of fishing on marine ecosystems. Advances in Marine Biology 34:201–352.[Web of Science]
Kaiser M.J., Collie J.S., Hall S.J., Jennings S., Poiner I.R. (2002) Modification of marine habitats by trawling activities: prognosis and solutions. Fish and Fisheries 3:114–136.[CrossRef]
Kaiser M.J. and De Groot S.J. (1999) The Effects of Fishing on Non-target Species and Habitats. Biological, Conservation and Socio-economic Issues(Blackwell Science, Oxford) 416 pp.
Kaiser M.J. and Spencer B.E. (1994) Fish scavenging behaviour in recently trawled areas. Marine Ecology Progress Series 112:41–49.[Web of Science]
Krost P., Bernhard M., Werner F., Hukriede W. (1990) Otter trawl tracks in Kiel Bay (Western Baltic) mapped by sidescan sonar. Meeresforschung 32:344–353.
Langton R.W. and Auster P.J. (1999) Marine fishery and habitat interactions: to what extent are fisheries and habitat interdependent. Fisheries 24:614–21.
Magorrian B.H., Service M., Clarke W. (1995) An acoustic bottom classification survey of Strangford-Lough, Northern Ireland. Journal of the Marine Biological Association United Kingdom 75:4987–992.
Morrisey D.J., Howitt L., Underwood A.J., Stark J.S. (1992) Spatial variation in soft-sediment benthos. Marine Ecology Progress Series 81:2197–204.[Web of Science]
Morrisey D.J., Underwood A.J., Howitt L., Stark J.S. (1992) Temporal variation in soft sediment benthos. Journal of Experimental Marine Biology and Ecology 164:2233–245.[CrossRef][Web of Science]
Mortensen P.B., Roberts J.M., Sundt R.C. (2000) Video-assisted grabbing: a minimally destructive method of sampling azooxanthellate coral banks. Journal of the Marine Biological Association United Kingdom 80:2365–366.[CrossRef]
Myers R.A., Hutchings J.A., Barrowman N.J. (1996) Hypotheses for the decline of cod in the North Atlantic. Marine Ecology Progress Series 138:1–3293–308.[Web of Science]
Pauly D., Christensen V., Dalsgaard J., Froese R., Torres F. Jr. (1998) Fishing down marine food webs. Science 279:860–863.
Pfirman S. L. (1985) Modern sedimentation in the Barents Sea: input dispersal and deposition of suspended sediments from glacial meltwater. PhD thesis, Woods Hole Oceanographic Institution and The Massachusetts Institute of Technology. 375 pp.
Pilskaln C.H., Churchill J.H., Mayer L.M. (1998) Resuspension of sediment by bottom trawling in the Gulf of Maine and potential geochemical consequences. Conservation Biology 12:61223–1229.[CrossRef][Web of Science]
Prena J., Rowell T. W., Schwinghamer P., Gilkinson K., Gordon D. C. Jr. (1996) Grand Banks otter trawling experiment: 1. site selection process, with a description of macrofaunal communities. Canadian Technical Report of Fisheries and Aquatic Sciences 2094: 38 pp.
Prena J., Schwinghamer P., Rowell T.W., Gordon K.D. Jr., Gilkinson D.C., Vass W.P., McKeown D.L. (1999) Experimental otter trawling on a sandy bottom ecosystem of the Grand Banks Newfoundland: analysis of trawl bycatch and effects on epifauna. Marine Ecology Progress Series 181:107–124.[Web of Science]
Rice A.L., Aldred R.G., Darlington E., Wild R.A. (1982) The quantitative estimation of the deep sea megabenthos; a new approach to an old problem. Oceanologica Acta 5:63–72.[Web of Science]
Schwinghamer P., Gordon D.C., Rowell T.W., Prena J., McKeown D.L., Sonnichsen G., Guigné J.Y. (1998) Effects of experimental otter trawling on surficial sediment properties of a sandy-bottom ecosystem on the Grand Banks Newfoundland. Conservation Biology 12:61215–1222.[CrossRef][Web of Science]
Service M. and Magorrian B.H. (1997) The extent and temporal variation of disturbance to epibenthic communities in Strangford Lough, Northern Ireland. Journal of the Marine Biological Association United Kingdom 77:1151–1164.
Sneli J.A. (1998) A simple benthic sledge for shallow and deep-sea sampling. Sarsia 83:69–72.[Web of Science]
Solheim A. and Elverhøi A. (1996) Surface sediments of the north-western Barents Sea. In Stein R., Ivanov G.I., Levitan M.A., Fahl K. (Eds.). Berichte zur Polarforschung. Reports on Polar Research. Surface-Sediment Composition and Sedimentary Processes in the Central Arctic Ocean and Along the Eurasian Continental Margin(Alfred Wegner Institute for Polar and Marine Research, Bremehaven, Germany) pp. 144–153 315 pp.
Tuck I.D., Hall S.J., Robertson M.R., Armstrong E., Basford D.J. (1998) Effects of physical trawling disturbance in a previously unfished sheltered Scottish sea loch. Marine Ecology Progress Series 162:227–242.[Web of Science]
Underwood A.J. (1992) Beyond BACI: the detection of environmental impacts on populations in the real, but variable world. Journal of Experimental Marine Biology and Ecology 161:145–178.[CrossRef][Web of Science]
Van Beek F.A., Van Leeuwen P.I., Rijnsdorp A.D. (1990) On the survival of plaice and sole discards in the otter-trawl and beam-trawl fisheries in the North Sea. Netherlands Journal of Sea Research 26:1151–160.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. J. Smith, A. C. Banks, and K-N. Papadopoulou Improving the quantitative estimation of trawling impacts from sidescan-sonar and underwater-video imagery ICES J. Mar. Sci., December 1, 2007; 64(9): 1692 - 1701. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||