CRUISE REPORT

HALIBUT BYCATCH SURVIVAL/SORTING EXPERIMENT

F/T Northern Glacier

October 7-28, 1993

Conducted By

International Pacific Halibut Commission

National Marine Fisheries Service

Highliner's Association

December 15, 1993

HALIBUT BYCATCH SURVIVAL/SORTING EXPERIMENT

INTRODUCTION

The International Pacific Halibut Commission, the Highliners Association (with Natural Resource Consultants), and the NMFS Alaska Fishery Science Center (AFSC) conducted an experiment to evaluate methods of increasing halibut bycatch survival in bottom trawls. The experiment involved sorting and discarding halibut from the groundfish catch more rapidly than is now current practice, and estimating the savings in halibut discard mortality rates. The experiment took place aboard the F/T Northern Glacier from October 6 through 29.

Halibut are caught as bycatch by most gear types used in North Pacific groundfish fisheries, but the majority are taken by trawls, especially those targeting on Pacific cod. Bycatch mortality could be reduced by improving survival and several methods have been suggested to accomplish this goal. One way would be to sort the halibut from the catch on deck, before groundfish and halibut are dumped into the below-deck holding tanks. A screen or grid has been suggested as a means of filtering halibut, particularly large halibut, from the catch. Another possibility is to improve the sorting methods used in the factory, in a manner that returns halibut to the sea more quickly than is currently practiced. Termed enhanced sorting, this practice could improve survival for the smaller fish that previously passed through the grid. This experiment was designed to address these issues.

OBJECTIVES

The experiment involved sorting and discarding halibut from the groundfish catch more rapidly than is now current practice, and estimating the savings in halibut discard mortality rates.

The experiment addressed the following questions:

1) What percent of the total halibut bycatch can be screened by the grid?

2) What percent of the total halibut bycatch can be sorted during the period of enhanced sorting?

3) What is the survival rate of halibut discarded from the grid screening and the enhanced sorting, compared to normal discards?

4) How much additional operating time accrues from the sorting procedures?

5) Will grid screening or enhanced sorting increase overall survival of halibut bycatch from trawls?

Specific objectives were:

1) Determine the sorting capability of a grid or screen placed over the deck opening to the factory holding tanks.

2) Determine if overall halibut mortality is reduced by sorting large halibut out on deck and immediately returning them to the sea.

3) Determine if halibut mortality is reduced by "speed sorting" of bycatch from the groundfish in the factory.

EXPERIMENTAL DESIGN

The vessel targeted Pacific cod in a normal commercial manner over the full 24-hour period. The experiment focused on the bottom trawl Pacific cod fishery because it is allotted the greatest portion of bycatch in the Bering Sea. The vessel operated in the Bering Sea (NMFS areas 517 and 521) and on Sanak Bank in the Gulf of Alaska. Two NMFS observers, one supplied by the vessel and one by the AFSC, determined halibut viability from each haul and sampled the groundfish catch on most hauls.

Two specific experiments were conducted. The first experiment (the Grid Sorting Experiment) evaluated two improved methods of sorting halibut from groundfish against a control method. For many factory layouts, halibut and other prohibited species and discards transit a series of conveyor belts to reach the exit chute. Forty-five minutes or more may elapse for the discard to move from the hold to the exit chute. We considered this procedure for handling discards to be the control method. The second experiment (Live Tank Holding) examined the relative survival of halibut within the established condition categories of excellent, poor, and dead.

For the Grid Sorting Experiment, three treatments were performed: (1) deck sorting with a grid; (2) enhanced sorting of the catch in the factory; and (3) normal sorting in the factory (the control). On the Northern Glacier, a single, short conveyor led from the hold to the exit chute. Retained fish were selected from the conveyor, and all else was quickly discarded. The regular procedure on the Northern Glacier was designated the enhanced treatment, while the control treatment was simulated by delaying processing for 45 minutes. Thirty hauls for each treatment were conducted, for a total of 90 hauls. We randomized the order of treatments. Other factors monitored were tow duration, haul size, time on deck, and fish size. A factorial analysis will be conducted on the results to determine significance among these factors. In some cases, the data may be post-stratified for the analysis.

The Live Tank Holding Experiment was conducted to reaffirm relative differences in survival of the three condition categories. Halibut sorted from the catch on deck and in the factory were placed in holding tanks with running seawater for 3 days (72 hours) until the end of the trip, when holding time was reduced to about 12 hours. Differences in viability going in and coming out of the tanks will be compared among the 3 conditions (excellent, poor, and dead). Approximately 20 halibut at a time were selected for placement into a tank. Post-stratification will also be done on important factors, notably sorting method, tow duration, time on deck, and fish size. An ANOVA analysis is planned for the results.

The first four hauls on the first fishing day were used to set up specific sampling procedures, and the first haul tested appropriate grid dimensions. The two grid dimensions examined were 9 inches by 11 inches and 11 inches by 14 inches. These are based on an even division of the deck opening, the first yielding a grid 3 openings deep and 6 wide. The second provided 2 openings deep by 6 wide. The vessel had on-board welding equipment to modify the grid dimensions, which proved to be unnecessary.

Tow duration was not predetermined, but two duration strata of > 3 hr and < 3 hr were established. The distribution of tow times was adjusted so that equal numbers of short and long hauls occurred for each treatment.

While no limit was set on the catch of groundfish or halibut, we anticipated catching the following quantities of fish:

Groundfish (other than Pacific cod) 700 mt

Pacific cod 1,500 mt

Prohibited species

Pacific halibut less than 50 mt

The vessel was allowed to retain, process, and sell the groundfish caught. Only the traditional prohibited species (crabs, salmon, halibut, herring) were required to be discarded.

DATA COLLECTION

Grid Sorting Experiment

During this experiment, data on length (cm), condition factor (excellent, poor, or dead) observations, and time of observation from the net coming on board were collected from each halibut encountered. Such data will allow enumeration and frequency distributions for the treatments (total halibut, total halibut from grid screening or enhanced sorting, and total halibut missed by the experimental treatment). NMFS observers conducted basket sampling to define the groundfish catch and determined halibut condition, so that these data are consistent with data collected in commercial fishery situations.

A schedule of the treatment for each haul alerted the bridge and the factory so that hauls could be made with factory processing capacity available. As each codend came on board, a biologist started a stopwatch; time of each halibut was recorded to the nearest minute. The observer and the skipper each estimated the groundfish catch. For grid sort treatments, the grid was placed over the hold, the deck crew grabbed halibut prior to the hatch and on the grid, and passed them to biologists for measurement and viability determination by the observer. When deck sampling was completed, the biological team moved to the factory where length, viability and time data were collected for all remaining halibut. For enhanced and control treatments, the sampling process started in the factory. Enhanced treatments started processing groundfish and sorting halibut quickly after dumping to the hold, while control treatments started processing 45 minutes after dumping to simulate the time needed for halibut to transit the factory to the exit chute typical of most layouts.

Live Tank Holding

Three specially-constructed deck-mounted holding tanks, each about 80 square feet by 36 inches high, with seawater circulation, an inside lip, dump door, and water overflow sump were used for holding halibut. Originally, only halibut sorted on deck were scheduled for these tanks, but halibut sorted out from the factory were also placed in these tanks when the factory tanks proved impractical. Initially, halibut collected from the factory were held in one or two 4'x4'x15' holding bins fed with circulating water. Water flow rates exchanged bin volumes about once per hour. Unfortunately, water jets in the holding bins, designed to lubricate large volumes of dead fish flowing to an exit, churned the water significantly, greatly diminishing survival. Halibut from the factory were carried as quickly as possible to the holding tanks on deck.

When a fish was selected for holding, a round, uniquely-numbered ID tag was placed on the tail using a nylon electrical tie. Selected fish were measured, condition factor assessed, and ID number noted on a form. Halibut were released after three days, and date and time of release, ID number, and viability noted on a separate form.

PRELIMINARY RESULTS

Ninety five hauls made during the experiment included four test hauls, one invalid haul caused by a ripped net, and the ninety hauls specified in the experimental design (Table 1). Catch weight ranged from about 5 mt to 35 mt per haul, but most were in the 10 to 15 mt range. The experimental hauls were divided into 30 hauls for each treatment, and the hauls of each treatment partitioned equally among < 3 hr and > 3 hr tows. The number of halibut caught reached 13,861, at an estimated weight of 38,000 kg (2.75 kg/halibut). Groundfish harvest totalled 1,189 mt, of which the retained portion was 243 mt of Pacific cod and 496 mt of pollock. The remaining 450 mt, mostly arrowtooth flounder, other flatfish, and Atka mackerel, were discarded. The halibut bycatch rate was 32 kg/mt. Total Pacific cod and halibut were significantly below the anticipated catch of 1,500 mt of Pacific cod and the maximum 50 mt of halibut. Pollock and discarded groundfish somewhat exceeded the 700 mt anticipated for other groundfish. Bycatch rates were higher than expected, and had the anticipated 2,200 mt of groundfish been harvested, halibut catch would have reached approximately 70 mt.

Approximately equal numbers of halibut were caught in each of the three treatments, with 4,714 in the grid sorting, 4,244 in the control sorting, and 4,903 in the enhanced sorting. In the grid sorting, 1,927 halibut (41%) were collected on deck. While weights have not yet been calculated, larger sizes of halibut sorted on deck probably put the proportion of deck-sorted halibut at least at 50% by weight.

The grid selected for use, although the smaller of the two available, did not directly filter out many of the halibut. The high proportion of deck-sorted halibut was due to the slower rate of dumping catch from the cod end to the hold, and the opportunity for the deck crew to sort out halibut pouring from the cod end to the hatch. Time required to dump a cod end after the net came on board normally ranged from about 90 seconds to 2 1/2 minutes, while a grid sort took about 10 to 15 minutes to dump.

While condition factor data and survival estimates are not yet available, several obvious conclusions result from observing halibut in the treatments. Halibut collected on deck during the grid sort experienced a high proportion of excellent condition factors. Only a few poor condition halibut were encountered, and halibut in dead condition were rarely seen. For enhanced sorting or grid sorting in the factory, nearly all halibut were in poor condition for about the first 40-50 minutes after the net came on board. A few excellent and dead halibut were noted. For control sorting and for enhanced or grid sorting after about 40-50 minutes, nearly all halibut were in dead condition, with occasional poor and the rare excellent halibut.

Holding tank experiments did not provide as much useable data as anticipated, because of situations with high mortality of halibut in the tanks. Bleeding tanks in the factory did not work because the water flow system agitated the halibut. A sloped floor in the bleeding tanks that prevented halibut from resting without piling up may have also contributed to the mortality. Of three tanks on deck, only one provided consistent data. The best tank was nearly square, while the other two were long and narrow. Vessel movement caused traveling waves in the narrow tanks that disrupted the halibut. In cases of prolonged rough weather, nearly all halibut died, regardless of initial condition factor. A total of 320 halibut from 17 hauls were placed in the live tanks for the standard three day period. Eighty-one more from four hauls were held for 12 hours. Nine hauls of the long holding period were from grid sort hauls, three from control sort hauls, and five from enhanced sort hauls. Three hauls from the short holding period were grid sort, and the last was enhanced sort.

SUMMARY

Ninety hauls equally divided among three sorting treatments provided 13,861 halibut for which condition factor, length, and time on deck were collected. On-deck sorting provided the highest survival, and control sorting caused the most mortality. Pollock and Pacific cod made up the retained catch. About 62% of the total was retained, and the remaining 38% was discarded. At 32 kg/mt, the halibut bycatch rate was higher than expected.

Holding tank experiments were less successful than anticipated. Tanks in the factory could not be used because of excessive mortality, and periods of rough weather caused mortality not related to condition factor in two of the three deck tanks. Periods of good weather during several holding periods permitted useable data from several hauls.

PERSONNEL

Abbreviations:

IPHC International Pacific Halibut Commission, Seattle

NMFS/AFSC National Marine Fisheries Service, Alaska Fisheries Science Center, Seattle

NMFS/AKR National Marine Fisheries Service, Alaska Region Office, Juneau

NMFS/D Hbr National Marine Fisheries Service, Observer Program, Dutch Harbor

NRC Natural Resources Consultants, Seattle

NPFMC North Pacific Fishery Management Council, Anchorage

HANA Halibut Association of North America, Seattle

For further information, please call Bob Trumble or Gregg Williams at the International Pacific Halibut Commission, Seattle, Washington, (206)634-1838.

Table 1. Preliminary catch totals during 1993 Halibut Bycatch Survival/Sorting Study. Codes for treatment are CL=Control, ES=Enhanced Sort, and GS=Grid Sort. Haul 590 was considered invalid.