CONTENTS Abstract Introduction Materials & Methods Results Discussion Acknowledgements Literatured Cited Appendix A
Northeast Fisheries Science Center Reference Document 09-07
Henry O. Milliken and Michael J. Fogarty
Determination of Conversion Factors for Vessel Comparison Studies
NOAA National Marine Fisheries Service, Northeast Fisheries Science Center, 166 Water Street, Woods Hole MA 02543
Web version posted May 13, 2009Citation: Milliken HO, Fogarty MJ. 2009. Determination of Conversion Factors for Vessel Comparison Studies. US Dept Commer, Northeast Fish Sci Cent Ref Doc. 09-07; 18 p. Available from: National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543-1026.
Information Quality Act Compliance: In accordance with section 515 of Public Law 106-554, the Northeast Fisheries Science Center completed both technical and policy reviews for this report. These predissemination reviews are on file at the NEFSC Editorial Office.ABSTRACT
Northeast Fisheries Science Center has conducted groundfish surveys on the Northeast Continental Shelf of the United States in autumn since 1963. These surveys are used to compute relative abundance indices of fish and selected invertebrates. Since the inception of the survey program, the R/V Albatross IV has been the primary vessel conducting the work, except periodically when the R/V Delaware II has been utilized because the R/V Albatross IV was not available. When the R/V Delaware II was used, differences in vessel characteristics necessitate the development of conversion coefficients to adjust catch rates using the R/V Albatross as the standard.
A series of vessel calibration studies between the R/Vs Albatross IV and Delaware II employing a paired tow design were initiated in 1981. In 1997, the R/V Delaware II underwent a major refit. Accordingly, for these two research vessels, there are two variations in the experimental time series that require consideration in development of calibration coefficients. We explored the use of a ratio estimator for the development of a vessel conversion coefficient and compared these with a general linear model (GLM) approach used by Byrne and Forrester (1991).
For many decades the Federal Government has conducted groundfish surveys out of Woods Hole, MA. With the arrival of the newly built research vessel R/V Albatross IV in 1963, the Bureau of Commercial Fisheries (now National Marine Fisheries Service) began a time series that serves as the basis for some of the longest time series of standardized fishery-independent indices of relative abundance in the world. In addition to tracking abundance of mature animals, research surveys provide indices of juvenile abundance, which can indicate strong year classes before fish are vulnerable to commercial or recreational fisheries. Research surveys indicate the status of a stock over its entire range, not just in small areas of commercial or recreational concentration. Surveys also provide data on growth, maturity, predation, and mortality of a stock as well as trophic dynamics of fish communities (Azarovitz 1981).
The R/V Albatross IV’s autumn bottom trawl survey was conducted using a #36 Yankee net with a sixty-foot (18.3 m) headrope and eighty-foot (24.4 m) footrope. The sweep was comprised of sixteen-inch (40.6 cm) rollers that allow the gear to be fished on most bottom types. The net and fishing operation remained almost the same since the start of the autumn survey in 1963 (Azarovitz 1981). When changes were made, comparative tows were performed to quantify the changes in the catchability of the gear. Once determined, a conversion coefficient was applied to the data so that any differences in the catchability of the gear could be compared to future catches (Fogarty 1997).
Conversion coefficients between the R/V’s Albatross IV and Delaware II were developed by Byrne and Forrester (1991) using a general linear model (GLM) approach. When the R/V Delaware II was used for surveys, these estimates were used to transform catch data. In 1997 the R/V Delaware II underwent a major refit. This refit included changing the winches from slower direct drive to faster free spooling winches. After the refit, the previous conversion factors were still employed to transform the data. However, the GLM approach does not allow the consideration of tows with zero catch, thus those data were lost. An alternative approach using ratio estimators, which use a ratio of the sum of the paired data, allows the comparison of all catch data between the two vessels.
Because the GLM method does not permit the inclusion of paired tows where one of the tows has zero catch, we compared the use of the GLM and ratio estimator methods for deriving the conversion factors. Additionally we examined the conversion factors before and after the refit of the R/V Delaware II in 1997.
MATERIALS & METHODS
The R/V’s Albatross IV and Delaware II fished alongside each other five times since the R/V Delaware II was refurbished in 1997 (Table 1, Figure 1). During these cruises, the R/V Delaware II accompanied the R/V Albatross IV while the R/V Albatross IV was engaged in its standard bottom trawl survey. Three paired comparison cruises from the 1980s were used to compare the GLM and ratio estimators for developing conversion factors. One 1998 cruise and four cruises from the 2000s were utilized to construct the ratio estimator model to estimate conversion coefficients.
The conversion coefficient (a) was calculated as:
where w is a weighting coefficient that depends on the relationship between the variance of V2 (Vessel 2) and V1(Vessel 1); the variance is defined as:
If the variance of V2 and V1 is proportional, the weighting coefficient is V1-1 and the estimator for the conversion coefficient is:
We used the proportional variance estimator in our analyses where “V2” is the variance of the R/V Albatross IV and “V1” is the variance of the R/V Delaware II.
Conversion coefficients were developed for winter skate (Leucoraja ocellata) , silver hake (Merluccius bilinearis), Atlantic cod (Gadus morhua), haddock (Melanogrammus aeglefinus), white hake (Urophycis tenuis), red hake (Urophycis chuss), American plaice (Hippoglossoides platessoides), fourspot flounder (Hippoglossina oblonga), yellowtail flounder (Limanda ferruginea), witch flounder (Glyptocephalus cynoglossus), windowpane flounder (Scophthalmus aquosus), redfish (Sebastes fasciatus), and ocean pout (Zoarces americanus).
When we examined the conversion factors derived from both the GLM and ratio estimator models using the 1980s data, with the exception of haddock, there appears to be good correlation between the two methods for determining vessel conversion coefficients (Figure 2).
Ratio conversion coefficients calculated for catches before and after the R/V
Delaware ’s 1997 refit suggest that the vessel effect was reduced after the vessel was modified. Twelve of the 14 species’ coefficients calculated from the 1980s data showed a difference, while only 4 of the 14 species show a difference after the refit (Figure 3).The species that showed a difference included cod, haddock, pollock and redfish (Figure 4). For these four species, the post-fit conversion factors should be applied to the R/V Delaware II data when the R/V Delaware II was used in place of the R/V Albatross IV for the standardized groundfish surveys (Table 2).
Differences in catch efficiencies between the R/V Albatross IV and R/V Delaware II were reduced after the refit of the R/V Delaware II. It was unanticipated that the refit would reduce the difference in the catchability between the two vessels or, in other words, reduce the catchability of the R/V Delaware II. After the R/V Delaware II was modified, the winches were replaced with faster winches. We hypothesize that because the winches were faster, the amount of time the net spent on the bottom was reduced. The winch speed was designed to match the speed of the winches on the R/V Albatross IV which appears to have equalized the catchability between the two vessels.
The ratio conversion coefficients were calculated for fourteen species. The next step will be to apply it to the other principal groundfish species to determine if conversion coefficients are required.
This work would not be possible without the efforts of numerous members of the
Northeast Fisheries Science Center , particularly the Resource Survey Branch, who participated in these cruises and helped collect this data. A special thanks is due to Kathy Mays of the Northeast Fisheries Science Center who spent considerable time helping to develop the proper scripts to pull the data from the survey database.
Azarovitz TR. 1981. A brief historical review of the Woods Hole Laboratory trawl survey time series. Can Spec Pub Fish Aq Sci. 58:62-67.
Byrne CJ, Forrester JRS. 1991. Relative Fishing Power of NOAA R/V’s Albatross IV and Delaware II. In: Report of the Twelfth Northeast Regional Stock Assessment Workshop. US. Dept. Commer., NOAA,
Northeast Fisheries Science Center Ref. Doc. 91-03, 187 p.
Fogarty M.J. 1997. Standardizing fishing power in research vessel surveys. In: Proceedings of the Workshop on Maintaining Current & Future Fisheries Resource Survey Capabilities. Special Report No. 63 of the Atlantic States Marine Fisheries Commission, p 54-63.