© 2003 by ICES/CIEM International Council for the Exploration of the Sea/Conseil International pour l'Exploration de la Mer
Influence of improved performance monitoring on the consistency of a bottom trawl survey
Alaska Fisheries Science Center, Resource Assessment and Conservation Engineering Division, National Marine Fisheries Service NOAA, 7600 Sand Point Way NE Building 4, Seattle, WA 98115, USA
*Correspondence to M. Zimmermann; tel: +1 206 526 4119; fax: +1 206 526 6723. e-mail: mark.zimmermann{at}noaa.gov.
Numerous trawl hauls, made during the triennial bottom trawl surveys (1977–1998) conducted by the National Marine Fisheries Service off the U.S. West Coast, had unusually small catch rates of benthic fish and invertebrates (cpueB), probably because the trawl failed to contact the seabed (off-bottom). Technological advances in the equipment used to monitor trawl performance since 1986 have increased our ability to recognize off-bottom tows, and cpueB has risen. As direct trawl performance measurements were not available in earlier surveys, a minimum cpueB derived from the survey with the best monitoring of bottom contact of the time-series (1998) was used as a criterion to eliminate trawls with poor bottom contact from earlier surveys. The truncated data sets produce significantly larger biomass indices, especially in 1980, with increases of 43, 45, and 56% for Dover sole, petrale sole, and Pacific sanddab, respectively. The analysis suggests that changes in cpueB over the time-series may be related more to changing survey fishing methods than to changes in abundance. Other bottom trawl surveys, which have also added trawl monitoring equipment during their time-series, may have experienced similar changes in trawl performance.
Keywords: assessment surveys, benthic, bottom trawl, net monitoring, off-bottom, operational data, water hauls
Received 10 October 2002; accepted 27 February 2003.
 |
Introduction |
|---|
 Top Introduction Material and methodsResultsDiscussionReferences |
|---|
The primary utility of bottom trawl surveys is to provide a fishery-independent index of stock abundance for tuning assessment models. Therefore, an important assumption about them is that they have a catchability that is stationary with respect to time. In an attempt to maintain stationary catchability, most surveys use a standardized methodology which, when changes are necessary, is examined experimentally to allow calibration of the old catch per unit effort (cpue) against that of the new. During the past few decades, however, many technological innovations have been added to surveys that allow for continuous measurement of operational parameters such as net spread, headrope height, gear depth, and bottom contact. Such information has been used in three distinct ways. First, more specific operational data have allowed increasingly precise estimation of width (Engås and Godø, 1986; Koeller, 1991; Rose, 1993) and length (West and Wallace, 2000) of the area swept at each station. Second, operational data have allowed clearer assessment of trawl performance during a tow, which in some instances has led to changes in methodology that produce better, more consistent samples (Lauth et al., 1998). Third, operational data have allowed the recognition and rejection of individual hauls that have performed outside specified objective limits (e.g. bottom contact). The U.S. West Coast trawl survey has focused primarily on the use of operational data to improve swept-area measurements, but these modifications to bottom trawling practices have heightened our awareness of off-bottom hauls. The aim of this article is to examine the effects of these operational changes on survey abundance estimates.
Scientists of the National Marine Fisheries Service (NMFS) have conducted bottom trawl surveys of U.S. West Coast groundfish resources triennially since 1977 (Shaw et al., 2000). In response to different assessment priorities (Table 1) during the time-series, changes have been made to the survey strata, station positions, station allocations, and the depth and latitude ranges of the survey. The survey was started with a nylon Nor'eastern bottom trawl (1977–1986), but this was changed to a similar trawl constructed of polyethylene (1986–1998) for greater durability in hard-bottom areas (Wilderbuer, 1988). By adding trawl measurement devices as they became available (Table 2; see Zimmermann et al., 2001, for a detailed review), changes were also made to the methods used for determining events during trawl tows. The possibility that a net could be off-bottom during a tow was recognized during the early West Coast surveys, but there was no effective way to monitor the possibility or to determine the extent to which it affected catch rates (Weinberg et al., 2002).
|
|
Early research to develop trawl monitoring equipment demonstrated that trawls used during surveys could have problems with bottom contact (Wathne, 1977), but the equipment was impractical to use during trawl surveys. The first standardized collection of operational data in the time-series was in 1986, when Scanmar net mensuration units were first used (Coleman, 1988). In the 1992 survey, a depth and temperature recording unit (micro-bathythermograph or MBT) was added to the trawl, and 24 hauls were recognized as being off-bottom through analysis of operational data using revised plotting software following the survey (Zimmermann et al., 1994). The most important change in the 1995 survey was the consistent collection of high quality data on net height, the result of fixing the Scanmar headrope units to hard plastic boards fixed to the headrope as semi-permanent platforms. As a result, the net could be monitored from as high as 50 m off the seabed as it sank to the bottom and reached stable net configuration—a significant advancement over earlier surveys when the net was allowed to settle to the bottom by waiting 3–8 min. A bottom contact sensor was added to the suite of tools for the 1998 survey (Shaw et al., 2000), objectively documenting, for the first time, the moment the net contacted and left the bottom by events at the footrope (Somerton and Weinberg, 2001). Trawls that never sink to the seabed or rise off-bottom during a tow can be recognized at sea by monitoring the operational data. While the large-scale changes in survey design reported in Table 1 have been obvious sources of potential survey bias, the impact of the better trawl measurement equipment listed in Table 2 has not been assessed until now.
It was suspected that numerous hauls early in the triennial time-series—labeled ‘water haul’ on the original data sheets—were off-bottom because absolutely nothing was caught. As trawl measurement equipment was not used in the earlier years of the survey, it was not possible to assess directly whether trawl performance was the cause of these water hauls. Therefore, we used the cpue of selected benthic species (cpueB) from the 1998 survey as a proxy for assessing bottom contact of the earlier (1977–1995) surveys.
|
Material and methods |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
cpueB levels and deviations
Data from the 1998 survey were set as the standard for comparing possible changes in cpueB during the survey history, because gear monitoring in 1998 was at its most advanced state, compared with preceding years. There is no definitive method of determining the hauls that may have been off-bottom in the surveys prior to 1998 because of limited net performance measurements. Instead, we calculated a minimum combined catch rate of a specific group of the most bottom-orientated fish and invertebrates (cpueB) as a proxy for bottom contact. The selection of species for this group was based on personal experience with bottom trawling, scuba-diving, mini-submersibles, and underwater video (Table 3). The cpueB was calculated by summing the mass of all benthic species in a haul and dividing by the area swept of that haul.
|
The 1998 cpueB values were first used as benchmarks for comparison with earlier surveys to determine if cpueB, in general, had changed over time. As the surveys varied in their latitudinal and depth coverage (Table 1), and as station locations changed, direct comparisons of the cpueB at each geographic location were not possible. Instead, cpueB values from comparable latitudes and depths from each of the earlier surveys (1977–1995) were compared with the 1998 survey results. The cpueB values for each survey were sorted from lowest to highest and comparisons made between equivalent tow percentiles to account for the different numbers of hauls completed in each survey. To avoid a bias that might create an artificially large difference, the cpueB values from 1998 (e.g. the 50.1 percentile) were subtracted from the next highest cpueB from the earlier survey (e.g. the 50.2 percentile). The differences in cpueB between each earlier trawl survey and those of 1998 were then analysed in two-tailed t-tests to determine whether they were significantly different from zero.
Changes in biomass index
A second method of using the 1998 cpueB values as benchmarks was to recalculate area-weighted biomass indices after eliminating all hauls from previous surveys that had values of cpueB below an approximate minimum from 1998. The biomass index for a stratum is calculated by multiplying the average cpue of a species in a stratum by the area of the stratum, and that for a survey is the sum of the individual stratum biomass indices. The 1998 survey covered most of the depth and latitudinal ranges of previous surveys (Table 1) and was used for setting the minimal possible cpueB within the survey area. After truncating the data sets from each of the earlier surveys, the biomass indices for individual species were recalculated and percentage changes calculated. These changes in biomass indices were not calculated as corrections, but as indicators of the magnitude of change, to determine whether or not this different interpretation of trawl survey history may have an impact on abundance trends.
Vessel-year effects
A third method of examining potential changes in cpueB was to test the differences in the actual and expected number of on-bottom and off-bottom hauls between different vessel-years using a Chi-square test. The test examined whether off-bottom hauls were unevenly divided among different vessel-years, which would indicate that trawling methodology unique to particular vessel-years might have impacted the catch of benthic species.
Indices of flatfish biomass
A fourth method of examining the occurrence of water hauls was to plot the biomass indices of the seven most abundant flatfish species, and of total flatfish combined, for the triennial time-series, as estimated from the uncorrected survey data. This comparison was made to determine whether between-year fluctuations in abundance of total flatfish species could be causing water hauls. Each species was plotted in terms of a percentage of its own greatest biomass index over the time-series, because the between-species variation in abundance is great. For example, Dover sole (Microstomus pacificus) was approximately 10 times as abundant as petrale sole (Eopsetta jordani) in the 1998 triennial (Shaw et al., 2000). Therefore, if there was a survey where a major species like Dover sole was at a historical low while a minor species like petrale sole was at a historical high, then it would seem reasonable to conclude that low cpueB values in that year were caused by the absence of the major species. If the abundances of certain flatfish species and total flatfish combined fluctuate in synchrony, then it would be reasonable to conclude that years with low cpueB values were the result of the bottom trawling protocol followed.
|
Results |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
cpueB levels and deviations
An examination of the 100 lowest cpueB values of the 528 successful hauls from the 1998 triennial survey showed that there was always a catch of benthic fish and invertebrates (Figure 1). An arbitrary, but conservative, threshold was established for determining whether the trawl caught a sufficient quantity of benthic fish and invertebrates to indicate that it fished on the seabed. To set the threshold, the lowest cpueB of the 1998 survey (0.53 kg ha–1) was rounded up to 1.0 kg ha–1. This catch rate is roughly equivalent to a total catch of about 4 kg of benthic fish and invertebrates in a standard haul covering an area equal to 4.04 ha or 0.04 km2.
|
The differences between cpueB values from the same latitude and depth range of each survey year and 1998 were plotted on the same graph to illustrate comparable deficiencies in cpueB from 1977 to 1995 (Figure 2). By the 50th percentile of hauls completed in each survey, there were three surveys (1995, 1989, 1986) with cpueB values 5–7.5 kg ha–1 less than the 1998 survey. The deficiencies for these 3 years remained relatively constant, suggesting that there was a limited number of hauls with poor contact with the seabed. The 1992, 1977, 1983, and 1980 survey results were, respectively, 12.5, 17.3, 14.8, and 22.3 kg ha–1 less than the 1998 survey. These four surveys all had increasing deficiencies, suggesting that there were numerous poorly performing hauls. All surveys had consistently negative differences until reaching the upper end of their distributions, where there were some very large positive differences (above the 90th percentile). These large positive differences can be attributed to extremely large catches of benthic species that were not matched by similar catches in the 1998 survey. The mean differences between 1998 and the previous surveys were significantly less than zero (two-tailed t-tests=0.05, p<0.001), with the exception of 1989, where the difference was not significantly less than zero owing to several extremely large, positive differences at the upper end of its distribution (not shown). All differences in 1989 were negative until the 96th percentile (422 of 437).
|
Changes in biomass index
Hauls with values of cpueB that fell below the 1.0 kg ha–1 level of benthic species for each of the earlier triennial surveys (1977–1995) were deleted (Table 4), and recalculated biomass indices varied widely by species, year, and depth stratum (Table 5). Results are listed for five flatfish species and for shortspine thornyhead (Sebastolobus alascanus), which were included in the cpueB, for Pacific halibut (Hippoglossus stenolepis) and sablefish (Anoplopoma fimbria), which were not included in the cpueB, and for total flatfish. Changes in biomass index were large and positive (in value) across all species and both depth zones in 1980. Increases in biomass index were also large in all three depth zones in 1977, in the shallow stratum in 1983, and in the deep stratum in 1989. In 1992 and 1995, changes in total biomass index were low, and there were a few decreases, indicating that trawls that contained the species in question, but not enough cpueB to qualify as being on-bottom (less than
4 kg), had been deleted from the analysis.
|
|
Vessel-year effects
The Chi-square test showed that the ratio of on-bottom to off-bottom trawl hauls was significantly different between vessel-years (
2>481, p<0.001, d.f.=15). The NOAA ship "David Starr Jordan" cruise in 1977 (45% off-bottom) and the MV "Mary Lou" and MV "Pat San Marie" cruises in 1980 (35 and 27% off-bottom, respectively) had the greatest percentages of off-bottom hauls (Table 6). The MV "Tordenskjold" cruise in 1977, the FV "Alaska" cruises in 1986, 1992, and 1995, and the FV "Vesteraalen" cruise in 1995, had the lowest percentages of off-bottom hauls (3%). In general, off-bottom rates were comparable among vessels within each survey year. One significant exception was the 1977 survey, during which one vessel had the highest (45%) and one the lowest rates (1%) of off-bottom tows in the triennial survey history. If the off-bottom hauls detected in and removed from the post-cruise analysis of the "Green Hope" in the 1992 survey (Zimmermann et al., 1994) are reinstated, then the off-bottom rate climbs from 6 to 12%, substantially higher than the 1% for the FV "Alaska" from the same year. These two exceptions are further evidence that changes in cpueB can be linked to differences in trawling methodology among survey vessels in different years.
|
Indices of flatfish biomass
The (uncorrected) biomass index of total flatfish and the seven most abundant flatfish species was variable from survey to survey (Figure 3), but it was generally higher in recent surveys. The biomass index of all species remained roughly the same or declined from 1977 to 1980, the year with the greatest number of water hauls. In general, the abundance of all individual species, as well as total flatfish, increased from 1980 through 1989, with arrowtooth flounder (Atheresthes stomias) and petrale sole peaking in 1989. There was a sharp decline in 1992 for most species, which may be related to the problem with off-bottom tows that the post-cruise analysis found for one of the two vessels used (Zimmermann et al., 1994), and the greater number of low cpueB hauls remaining in the data set for the same vessel (Table 6). All species increased in abundance during 1995, which had the fewest number of water hauls of the earlier years (n=9; Tables 4 and 6). All species, except for Pacific sanddab (Citharichthys sordidus) and Pacific halibut, which peaked during 1995, continued to increase in 1998. Dover sole, rex sole (Glyptocephalus zachirus), English sole (Parophrys vetulus), and all flatfish species combined peaked during 1998, and petrale sole was at 99.7% of its peak in 1998.
|
|
Discussion |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
All previous U.S. West Coast surveys (1977–1995) caught smaller quantities of benthic fish and invertebrates than the survey of 1998. By removing hauls from earlier surveys that fell below a minimum cpue threshold of benthic species, the biomass index for all species examined increased substantially, especially for the first three triennial surveys, and the deep strata of the 1989 survey. Different rates of off-bottom hauls between two of the vessels used in 1977 and between both vessels in 1992 indicate that the vessels used different trawling methods. As the biomass indices of flatfish species generally varied in synchrony over time, low catch rates in some vessel-years are probably not attributable to differential availability of individual flatfish species, but are more likely related to the poor seabed contact that reduced the catch of all flatfish species similarly. While there is no definitive proof, we conclude that the increase in abundance indices of benthic fish and invertebrates over the time-series was most likely the result of improvements to the trawling methods. These improvements primarily reflect technological advances added to the survey to improve the estimates of area swept, but they also afford a better understanding of survey bottom trawling protocol and, in turn, changed the manner in which trawls were conducted.
It is speculated that few of the hauls with low cpueB were entirely off-bottom during the tow and that few of the remaining hauls were entirely on-bottom throughout the tow. It seems likely that most hauls landed on the seabed at some point and that there was a continuum, ranging from poor to excellent, of how well they tended the seabed. Our method of removing tows with low cpueB was intended to replicate the removal of off-bottom tows in a modern survey, such as in 1998, where analysis of operational data from the trawl measurement equipment detected a few off-bottom tows, which were removed.
Although the survey experimented with a polyethylene net as a possible replacement for the nylon net in 1986, the switch to polyethylene in 1989 was not based on a formal comparison of possible changes in fishing power of the data from the two nets. A second experiment was conducted with both nets off Kodiak Island, Alaska, in April 1986 (prior to the 1986 triennial survey), with larger (1.83x2.74 m), heavier (800 kg) doors (Wilderbuer, 1988) than those used traditionally in the surveys (1.5x2.1 m, 567 kg). Wilderbuer et al. (1998) found that the polyethylene net always had an equal or greater mean fishing power than the nylon net, although the 95% confidence intervals generally showed "no difference". It is possible that the increased catch rates of cpueB noted in this current study may be attributable, in part, to the use of the polyethylene net after 1986. To test for this possibility, we performed a two-sample t-test that demonstrated no significant difference (d.f.=194, 
=0.05, p>0.05) in mean cpueB values between the nylon and polyethylene nets fished by the same vessel during the 1986 survey of the West Coast shelf.
We assert that the increases in cpueB found during this study are most likely caused by changes in bottom trawling methods. However, the increase could also be due to regime shifts, natural fluctuations in abundance, or changes in commercial fishing effort and location. Beamish (1993) linked a change in climate over the years 1976–1978 in the North Pacific to increased production of several groundfish species off the west coast of the United States and Canada. This increased production of fish resulted from conditions that increased the survivorship of year classes in and around 1977, including such abundant and valuable groundfish as yellowtail rockfish (Sebastes flavidus), widow rockfish (Sebastes entomelas), Pacific ocean perch (Sebastes alutus), Pacific halibut, Pacific cod (Gadus macrocephalus), sablefish, and lingcod (Ophiodon elongatus; Beamish, 1993). Further, Anderson et al. (1997) described an epibenthic regime shift in the Gulf of Alaska from a community dominated by shrimp in the mid- to late 1970s to a community dominated by gadid and pleuronectid fish more recently. Another bottom trawl survey revealed a 13-fold increase in the estimated biomass of arrowtooth flounder in the Gulf of Alaska, ranging from approximately 146 000 t in the years 1973–1976, to a peak of 1 922 000 t in 1990 (Turnock et al., 2000). Clark et al. (1999) attributed an unusually high recruitment of Pacific halibut in the Gulf of Alaska and off British Columbia from 1985 to 1996 to the regime shift in the late 1970s. A large increase in the estimated biomass of northern rock sole (15-fold from 1975 to 1994; Walters and Wilderbuer, 2000) in the eastern Bering Sea has been attributed to changes in water temperature and climate. Further, the change in groundfish populations on Georges Bank from gadids and flounders to elasmobranchs has been attributed to intensive fishing (Fogarty and Murawski, 1998). From such a perspective, we conclude that it is unlikely that the changes in the cpueB through the West Coast triennial time-series, which includes such diverse taxa as Rajidae, Cottidae, Pleuronectiformes, Arthropoda, Actiniaria, Mollusca, Echinodermata, and Porifera, could solely be due to changes in abundance.
Consequences
If the changes in the catch of benthic species documented in this investigation are due to changes in fishing methods, as we conclude, then the results and conclusions of numerous studies using West Coast bottom trawl survey data may be compromised. For example, if increasing catches of flatfish during the survey are not the result of populations growing, then stock assessments that have tracked large increases since the 1977 survey (and especially since the 1980 survey) may need to be reconsidered with regard to the magnitude or direction of population trends. It is important to note that the percentage changes in biomass indices shown herein are most likely underestimates of real differences, particularly for the years 1977, 1980, 1983, and 1992. We simply recalculated biomass indices after removing the hauls with cpueB values that fell below an arbitrary, low threshold. We did not attempt to correct the cpue of individual hauls by, for example, artificially inflating low catch rates in areas with known high rates. These revised estimates are not intended to take the place of the original time-series (indeed, the water hauls still remain in the database as good performance hauls). The percentage changes in biomass indices are presented as examples of our interpretation of the data and can be used by stock assessment scientists to compare alternative model outputs. The most consistent stock assessment models, given the analysis presented here, would be those that closely match the biomass of benthic species in the 1998 survey (the most accurate) and also overestimate the biomass from the 1980 survey (the least accurate). We are uncertain if, or how, corrections could be applied for pelagic or semi-pelagic species. An examination of the West Coast bottom trawl survey catchabilities for several Sebastes species indicate low catchabilities for 1977 and 1980 (Millar and Methot, 2002).
This project resulted from comparing the comments on original, handwritten data sheets from earlier surveys with experiences in the field during modern surveys. The significant differences in catches between surveys are most likely attributable to differences in bottom trawling methods, which we link to increased collection of operational data, and therefore may not accurately represent true trends in abundance. These trends, however, have been accepted as real trends in abundance. By concentrating on ensuring best contact between the bottom trawl and the seabed and recording the area swept as accurately as possible, it is likely that we failed to recognize that the increasing use of operational data to accomplish these goals changed what was being caught.
Means of incorporating this new knowledge about trawl bottom contact in the triennial survey can be found in stock assessments for shortspine thornyhead (Piner and Methot, 2001), Dover sole (Sampson and Wood, 2001), and sablefish (Schirripa and Methot, 2001). Schirripa and Methot (2001) analysed data from the triennial and other assessment surveys, along with commercial catch data, and found that the models more closely fit the triennial survey abundance indices when the water hauls are excluded. The Piner and Methot (2001) assessment is an example of an assessment in which the commercial catch data could not be incorporated owing to problems with species identification, changes in fishing restrictions, and lack of knowledge of discards. Instead, the assessment was based on other assessment methods and on the revised triennial survey data presented in this study. Sampson and Wood (2001) incorporated commercial catch and effort data, other assessment survey indices, and the revised triennial survey indices in their assessment. Therefore, output from this work is already being taken into account in certain assessments. However, the main purpose of this work was to point out that conclusions from other bottom trawl surveys, that also added the collection of operational data to their sampling routine over time, may have developed a similar problem as found here.
|
Acknowledgements |
|---|
This project benefited greatly from discussions with stock assessment scientists involved with West Coast fisheries, including Gary Stauffer from the Alaska Fisheries Science Center, and Tom Helser, Rick Methot, and others from the Northwest Fisheries Science Center, as well as from discussions with field biologists too numerous to mention. Craig Rose, Tom Wilderbuer, Gary Stauffer, Jon Volstad, Ingvar Huse, and David Somerton provided critical reviews that improved the manuscript substantially.
|
References |
|---|
|
TopIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
-
Anderson P. J., Blackburn J. E., Johnson B. A. (1997) Declines of forage species in the Gulf of Alaska, 1972–1995, as an indicator of regime shift. Proceedings of the Forage Fishes in Marine Ecosystems Symposium(University of Alaska, Fairbanks) pp. 531–543 Alaska Sea Grant College Program, AK-SG-97-01.
Beamish R.J. (1993) Climate and exceptional fish production off the west coast of North America. Canadian Journal of Fisheries and Aquatic Sciences 50:2270–2291.
Clark W.G., Hare S.R., Parma A.M., Sullivan P.J., Trimble R.J. (1999) Decadal changes in growth and recruitment of Pacific halibut (Hippoglossus stenolepis). Canadian Journal of Fisheries and Aquatic Sciences 56:242–252.
Coleman B. A. (1988) The 1986 Pacific west coast bottom trawl survey of groundfish resources: estimates of distribution, abundance, length, and age composition. U.S. Department of Commerce, National Oceanic and Atmospheric Administration Technical Memorandum, NMFS F/NWC-152.
Engås A. and Godø O.R. (1986) Influence of trawl geometry and vertical distribution of fish on sampling with bottom trawl. Journal of Northwest Atlantic Fishery Science 7:35–42.
Fogarty M.J. and Murawski S.A. (1998) Large-scale disturbance and the structure of marine systems: fishery impacts on Georges Bank. Ecological Applications 8:Suppl., S6–S22.[CrossRef][Web of Science]
Koeller P.A. (1991) Approaches to improving groundfish survey abundance estimates by controlling the variability of survey gear geometry and performance. Journal of Northwest Atlantic Fishery Science 11:51–58.
Lauth R.R., Syrjala S.E., McEntire S.W. (1998) Effects of gear modifications on the trawl performance and catching efficiency of the West Coast upper continental slope groundfish survey trawl. Marine Fisheries Review 60:1–26.
Millar R.B. and Methot R.D. (2002) Age-structured meta-analysis of U.S. West Coast rockfish (Scorpaenidae) populations and hierarchical modeling of trawl survey catchabilities. Canadian Journal of Fisheries and Aquatic Sciences 59:383–392.
Piner K. and Methot R. (2001) Stock status of shortspine thornyhead off the Pacific west coast of the United States 2001(Pacific Fishery Management Council, Portland, OR 97220-1384, USA) 161 pp.
Rose C. S. (1993) Trawl width variation and its effects on groundfish trawl survey results. PhD dissertation, University of Washington, Seattle, WA.
Sampson D. B. and Wood C. (2001) Stock status of Dover sole off the U.S. west coast in 2000(Pacific Fishery Management Council, Portland, OR 97220-1384, USA) 110 pp.
Schirripa M. J. and Methot R. (2001) Status of the sablefish resource off the U.S. Pacific coast in 2001(Pacific Fishery Management Council, Portland, OR 97220-1384, USA) 58 pp.
Shaw F. R., Wilkins M. E., Weinberg K. L., Zimmermann M., Lauth R. R. (2000) The 1998 Pacific west coast bottom trawl survey of groundfish resources: estimates of distribution, abundance, and length and age composition. U.S. Department of Commerce, National Oceanic and Atmospheric Administration Technical Memorandum, NMFS-AFSC-114.
Somerton D.A. and Weinberg K.L. (2001) The effect of speed through the water on footrope contact of a survey trawl. Fisheries Research 53:17–24.[CrossRef][Web of Science]
Turnock B. J., Wilderbuer T. K., Brown E. S. (2000) Arrowtooth flounder. Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska(North Pacific Fishery Management Council, Anchorage, AL 99501-2252, USA) pp. 137–168 Chapter 4.
Walters G.E. and Wilderbuer T.K. (2000) Decreasing length at age in a rapidly expanding population of northern rock sole in the eastern Bering Sea and its effect on management advice. Journal of Sea Research 44:17–26.[CrossRef]
Wathne F. (1977) Performance of trawls used in resource assessment. Marine Fisheries Review 39:16–23.
Weinberg K.L., Somerton D.A., Munro P.T. (2002) The effect of trawl speed on the footrope capture efficiency of a survey trawl. Fisheries Research 58:303–313.[CrossRef][Web of Science]
West C. W. and Wallace J. R. (2000) Measurements of distance fished during the trawl retrieval period. ICES CM 2000/K: 37. 14 pp.
Wilderbuer T. K. (1988) Fishing power standardization in multi-vessel, multi-gear, resource assessment surveys. MS thesis, University of Washington, Seattle, WA.
Wilderbuer T.K., Kappenman R.F., Gunderson D.R. (1998) Analysis of fishing power correction factor estimates from a trawl comparison experiment. North American Journal of Fisheries Management 18:11–18.[CrossRef]
Zimmermann M., Wilkins M. E., Lauth R. R., Weinberg K. L. (1994) The 1992 Pacific west coast bottom trawl survey of groundfish resources: estimates of distribution, abundance, and length composition. U.S. Department of Commerce, National Oceanic and Atmospheric Administration Technical Memorandum, NMFS-AFSC-42.
Zimmermann M., Wilkins M. E., Weinberg K. L., Lauth R. R., Shaw F. R. (2001) Retrospective analysis of suspiciously small catches in the National Marine Fisheries Service West Coast triennial bottom trawl survey. U.S. Department of Commerce, Alaska Fisheries Science Center Processed Report, 2001-03.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||