07-5061. Fisheries of the Caribbean, Gulf of Mexico, and South Atlantic; Shrimp Fisheries of the Gulf of Mexico and South Atlantic; Revision of Bycatch Reduction Device Testing Protocols

AGENCY:

National Marine Fisheries Service (NMFS), National Oceanic and Atmospheric Administration (NOAA), Commerce.

ACTION:

Proposed rule; request for comments.

SUMMARY:

In accordance with the framework procedures for adjusting management measures specified in regulations implementing the Fishery Management Plan for the Shrimp Fishery of the Gulf of Mexico (Gulf FMP) and the Fishery Management Plan for the Shrimp Fishery of the South Atlantic Region (South Atlantic FMP), NMFS proposes to consolidate and make modifications to the Bycatch Reduction Device Testing Manuals (Manual) for the Gulf of Mexico and the South Atlantic regions. This proposed rule would also revise the bycatch reduction device (BRD) certification criterion for the western Gulf of Mexico and would certify additional BRDs. The intended effect of this proposed rule is to improve bycatch reduction in the shrimp fisheries and better meet the requirements of national standard 9.

DATES:

Comments must be received no later than 4:30 p.m., eastern time, on November 13, 2007.

ADDRESSES:

You may submit comments, identified by 0648-AU59, by any one of the following methods:

• Electronic Submissions: Submit all electronic public comments via the Federal eRulemaking Portal http://www.regulations.gov.
• Fax: [phone], Attn: Steve Branstetter.
• Mail: Steve Branstetter, Southeast Regional Office, NMFS, 263 13th Avenue South, St. Petersburg, FL 33701.

Instructions: All comments received are a part of the public record and will generally be posted to http://www.regulations.gov without change. All Personal Identifying Information (for example, name, address, etc.) voluntarily submitted by the commenter may be publicly accessible. Do not submit Confidential Business Information or otherwise sensitive or protected information.

NMFS will accept anonymous comments. Attachments to electronic comments will be accepted in Microsoft Word, Excel, WordPerfect, or Adobe PDF file formats only.

Copies of the proposed regulatory amendment, which includes an Environmental Assessment, an Initial Regulatory Flexibility Analysis (IRFA), a Regulatory Impact Review (RIR), and a Social Impact Assessment/Fishery Impact Statement, may be obtained from the Gulf of Mexico Fishery Management Council, 2203 North Lois Avenue, Suite 1100, Tampa, FL, 33607; phone: [phone]; fax: [phone]; email: [e-mail].

Copies of the proposed consolidated and revised Bycatch Reduction Device Testing Manual and the associated IRFA, RIR, and Social Impact Assessment/Fishery Impact Statement are available from the Southeast Regional Office, NMFS, 263 13th Avenue South, St. Petersburg, FL 33701; phone: [phone]; fax: [phone].

Comments regarding the approved collection-of-information requirements contained in this rule should be submitted in writing to Jason Rueter at the Southeast Regional Office address (above) and to David Rostker, Office of Management and Budget (OMB), by e-mail at [e-mail], or by fax to 202-395-7285.

FOR FURTHER INFORMATION CONTACT:

Steve Branstetter, telephone: [phone], fax: [phone], e-mail: [e-mail].

SUPPLEMENTARY INFORMATION:

The fisheries for shrimp in the exclusive economic zone (EEZ) of the Gulf of Mexico and the South Atlantic are managed under the authority of the Magnuson-Stevens Fishery Conservation and Management Act (Magnuson-Stevens Act) and regulations at 50 CFR part 622. The regulations implement the Gulf FMP prepared by the Gulf of Mexico Fishery Management Council (GMFMC) and the South Atlantic FMP prepared by the South Atlantic Fishery Management Council (SAFMC).

Background

Regulations implementing Amendment 9 to the Gulf FMP were published April 14, 1998 (63 FR 18139). The final rule established a requirement, with limited exceptions, for the use of certified BRDs in shrimp trawls towed in the Gulf of Mexico EEZ shoreward of the 100-fm (183-m) depth contour west of 85° 30′ W. longitude (western Gulf), the approximate longitude of Cape San Blas, FL. Regulations implementing Amendment 9 also required NMFS to develop a Manual for the Gulf of Mexico outlining testing procedure for examining the bycatch reduction performance of additional BRD designs. BRDs tested under such a procedure and determined to reduce bycatch mortality of juvenile red snapper by a minimum of 44 percent from the average level of mortality on these age-0 and age-1 groups during the years 1984-1989 would be certified for use in the western Gulf shrimp trawl fishery. A final rule implementing the requirements for this testing procedure was published and became effective July 13, 1999 (64 FR 37690), except for the collection-of-information requirements which became effective September 29, 1999 (64 FR 52427).

NMFS had already published similar regulations (62 FR 18536, April 16, 1997), to implement Amendment 2 to the South Atlantic shrimp FMP, requiring the use of BRDs in the South Atlantic penaeid shrimp fishery. Amendment 2 established a bycatch reduction certification criterion based on 40-percent reductions in the number of Spanish mackerel and weakfish. The final rule implementing Amendment 2 also established a Manual for the South Atlantic shrimp fishery.

To better address the requirements of national standard 9 of the Magnuson-Stevens Act, regulations implementing Amendment 10 to the Gulf FMP (69 FR 1538, January 9, 2004) required BRDs in shrimp trawls fished in the EEZ east of 85°30′ W. longitude (eastern Gulf). To be certified for use in the EEZ of the eastern Gulf, a BRD has to reduce finfish bycatch by at least 30 percent, by weight. NMFS established this new criterion because juvenile red snapper are not common in the eastern Gulf. Therefore, evaluating the effectiveness of a BRD in the eastern Gulf, under a red snapper criterion, would not be feasible. A general finfish reduction, addressing national standard 9, was the more appropriate measure to establish for this region.

The final rule implementing Amendment 6 the South Atlantic FMP (70 FR 73383, December 12, 2005) transferred authority to the NMFS Southeast Regional Administrator (RA) to modify the SAFMC's Manual, as needed, after consultation with the SAFMC. The final rule implementing Amendment 6 also modified the South Atlantic BRD certification criterion to match the eastern Gulf criterion of a 30-percent finfish reduction, and expanded the BRD requirement to include the rock shrimp fishery.

BRD Certification Criterion

In accordance with the BRD framework procedures of regulations implementing the Gulf FMP, the proposed rule would modify the existing BRD certification criterion for the western Gulf to be consistent with the existing criterion for the eastern gulf and the South Atlantic—a 30-percent reduction in total finfish catch by weight. The existing criterion, established in Amendment 9, is based on a 1995 stock assessment model no longer applicable to the revised red snapper rebuilding target. The 1995 assessment recommended a 50-percent reduction in fishing mortality on age 0 and age 1 red snapper from the average mortalities during the 1984 to 1989 period. The model estimated a fishing mortality rate for the 1984 to 1989 period at 2.06. Recognizing a 10-percent reduction in effort had occurred in the shrimp fishery since 1989, NMFS established a target for a 44-percent reduction from BRDs, which achieved the goal of reducing fishing mortality to approximately 1.03. The rationale for this action assumed that such reductions, beginning in 1997, would meet the existing goal of a 20-percent spawning potential ratio for red snapper by 2019.

This approach was valid based on the modeling techniques used for red snapper at the time; however, recent stock assessments used different models, and the rebuilding target for red snapper has changed. For example, with changes to the red snapper stock and to the red snapper and shrimp fisheries, in combination with refined assessment techniques, the 2005 red snapper stock assessment estimated fishing mortality on age 0 and age 1 red snapper at 0.74 for the 1984-1989 time period. This does not mean the 1995 assessment overestimated fishing mortality, rather that the 2005 assessment utilized updated information which revised estimates of natural mortality (M). In running the models with that revised estimate of M, other parameters, including F, also changed. The 2005 assessment went through a rigorous SEDAR/peer review process. The fishing mortality rate for juvenile red snapper attributable to the shrimp fishery still needs substantial reduction to rebuild the red snapper stock by the new 2032 target; however, the existing BRD certification criterion of a 44-percent reduction in fishing mortality rate to a level of 1.03 is no longer appropriate.

Although the 1995 assessment model could still be used, with a change in scaling, to develop a revised BRD reduction criterion based on a reduction in fishing mortality, there are still problems with using a mortality rate target as the criterion. The annual fishing mortality rates for juvenile age 0 and age 1 red snapper are dependent on seasonal recruitment and the quantity of shrimp fishing effort taking juvenile red snapper. These variables, in turn, affect the ability of a given BRD to reduce annual fishing mortality to a specific level. Thus, the overall goal of reducing the annual juvenile red snapper mortality rate in the shrimp fishery could be achieved from a high reduction of red snapper by BRDs, or by a lesser reduction of red snapper by BRDs in combination with an overall reduction in fishing effort. Under the current certification criterion, based on the mortality rate for one year compared to previous years, it is not possible to independently distinguish the contribution of the BRD from the contribution of overall shrimp effort reductions between the two time periods.

A more appropriate measure of the efficacy of a BRD to reduce bycatch is to evaluate the reductions in catch or catch-per-unit-effort (CPUE) of a species or species group on a real-time basis. Doing so isolates the contribution by the BRD and removes the interactions of total shrimping effort and annual fluctuations in recruitment. The catch rate of a net with a BRD can be directly compared to the catch rate of a net without a BRD, to give reduction levels at any given time. Fishing mortality reductions can then be calculated based on the documented total effort by the fleet and the estimates of recruitment for any given time frame. This is a more appropriate approach than attempting to apply mortality rate values for a specific year against a previous benchmark value, given the fluctuations in recruitment, effort, and CPUE values which affect estimates of annual mortality rates attributable to the shrimp fishery.

In addition, because of the existing statistical procedures prescribed in association with the bycatch reduction criterion, it is difficult to certify new BRDs. Only two BRD types have been certified since 1998 for use in the western Gulf. New BRD designs need to be available to shrimp fishermen to better reduce bycatch of red snapper and achieve recovery goals of this overfished stock, to better reduce overall finfish bycatch to meet the requirements of national standard 9, and to improve shrimp retention for a more efficient fishery under current economic conditions.

Several potentially effective BRD designs could not meet the very specific and rigorous mortality-based criterion established for the western Gulf. However, these experimental BRD designs have been demonstrated to achieve substantial levels of overall finfish reduction, and a moderate and consistent level of red snapper reduction, exceeding the red snapper reduction being achieved by the most commonly used configuration of the fisheye BRD. In addition, these BRDs are similar to the fisheye BRD in terms of overall shrimp retention.

A change in the bycatch reduction criterion west of Cape San Blas, FL, from a reduction in fishing mortality of red snapper to a reduction in finfish catch would increase the opportunity to certify a greater variety of BRDs for use in the fishery, provide a uniform bycatch reduction criterion and list of certified BRDs for the Gulf of Mexico and the South Atlantic regions, and improve the overall reduction in juvenile red snapper bycatch mortality. BRDs may have different capabilities under different fishing conditions, and having a wider variety of BRDs for use in the fishery would allow fishermen to choose the most effective BRD for the specific local fishing conditions.

Revisions to the BRD Testing Protocol Manual

Background

BRD testing is conducted by comparing the differences in the catch and bycatch of two nets that are towed simultaneously by a single vessel. One net (control net) is a standard rigged shrimp trawl without a BRD, and one net (experimental net) is identically configured, except it contains the experimental BRD. Assuming the two nets have equal or similar fishing efficiencies, the differences in catch and bycatch between the two nets can be attributed to the inclusion of the experimental BRD in one net. Since the Gulf of Mexico and South Atlantic Manuals have been in effect, several experimental BRDs have been tested for certification, but none have been certified. Two specific issues appear to be impeding the successful testing and eventual certification of experimental BRDs.

To be certified by the RA, the BRD candidate must demonstrate an observed reduction rate meeting the bycatch reduction criterion with some degree of statistical certainty. Currently, a modified Student t-test, a standard statistical approach, is used to evaluate the data collected during an experimental BRD evaluation. The criterion for the western Gulf requires there be no more than a 5-percent probability the true reduction rate is less than one standard deviation from the observed mean reduction rate. The magnitude of any standard deviation of a sample is dependent on the data set in question, and the analysis is based on the assumption the individual data points reflect a consistent result among sampling trials during a test. In the case of evaluating a BRD candidate in the marine environment, where organisms in the environment are not randomly distributed, catch rates can be highly variable among successive trawl tows or even between nets during a single tow. This variability increases the standard deviation, and this increase is exacerbated by the small minimum sample size required by the Manual, 30 comparative tows. A sample size of 30 is a recognized minimum standard for conducting a Student t-test, but this standard assumes the data being analyzed have relatively similar values, which as noted, is not often true in the marine environment. However, this sample size was considered necessary to minimize the cost and effort involved in conducting an experimental BRD test.

The variability among data points and the resulting uncertainty regarding the observed sample mean can be reduced in two ways: (1) the researcher must ensure the sampling effort will generate consistent results between samples, thus reducing variability (increase precision); or (2) the sample size must be increased to better ensure the resulting sample mean or average value is more likely representative of the true mean value (increase accuracy). The first of these options is not feasible for most biological sampling efforts; as noted, organisms are not randomly distributed and collections of these organisms would never be expected to produce consistent results. The second alternative can be achieved, but only with a greatly increased cost to the researcher; initial estimates suggest that between a four-fold and eight-fold increase in sample size would be needed.

A second and equally critical issue for the initial development of experimental BRDs involves the field sampling procedures prescribed in the Manuals. These rigorous procedures were established with the intent of reducing the inherent variability and uncertainty in the data stemming from a small, 30-tow sample size. Several field tests were not completed successfully (e.g., 30 successful comparative tows could not be completed) because of a failure to meet one or more of the procedural requirements set forth in the Manuals. However, not being able to complete a field test on potentially effective BRDs because of logistical constraints has substantial negative consequences for conservation. Further development of particularly productive concepts may cease, and BRD efficiency might never rise above the current level. This discourages innovative developments to improve BRDs.

These issues were identified at a 1999 shrimp fishery stakeholder's workshop sponsored by the Gulf and South Atlantic Fisheries Foundation, Inc. Recommendations stemming from the workshop were made available to the GMFMC and SAFMC for their consideration. Based on this information as well as additional public input regarding the existing bycatch Manuals, the GMFMC and SAFMC requested that NMFS develop alternative procedures to address and alleviate these impediments to testing and certifying new BRD candidates, while maintaining the statistical confidence BRDs will meet the established bycatch reduction criterion and achieve bycatch reduction goals.

NMFS is proposing to consolidate and make revisions to the Manuals for the Gulf of Mexico and the South Atlantic region. The new, combined BRD Testing Manual implemented under this proposed rule would establish alternative statistical procedures and field sampling procedures. The new statistical procedures would address the issue of statistical uncertainty due to limited sample size when evaluating the effectiveness of experimental BRD designs. Additionally, the proposed rule would modify the Manual to incorporate additional flexibility in the field sampling procedures. Coupled with the proposed modification to the statistical approach, alternative sampling procedures provide flexibility to better meet the logistical constraints of field sampling while maintaining an acceptable level of statistical precision and accuracy.

Gear Changes During a BRD Test

According to the current Manuals, if the fishing gear used at the start of the test incurs damage and requires replacing, then the certification test of a BRD candidate must begin anew. Under actual field conditions, damage to fishing gear often occurs before the completion of 30 tows. Even if 30 consecutive tows are completed during a test without incident, the data represent results collected aboard a single vessel using only one trawl configuration in a limited area and during a specific time frame. Results from such a test might not be applicable to other vessels fishing at other times of the year, in other areas, or using other shrimp trawl configurations.

The proposed rule would modify the procedure to allow the compilation of results from a series of tests to meet the 30-tow minimum sample size for a complete BRD test. This alternative would eliminate the need to reinitiate tests after a gear failure. Under the proposed modifications, should gear failure occur, the applicant would replace the damaged gear, conduct “tuning” tows (see Gear Tuning below) to determine the new gear did not affect the fishing efficiency between the two nets, and then continue the test. Minor repairs to the gear (e.g., sewing holes in the webbing; replacing a broken tickler chain with a new one of the same configuration) would not be considered a gear change. Additionally, under this proposed procedure, it would be possible to conduct the test over a longer period, aboard different vessels, using different fishing gear configurations (with the same BRD design), or while fishing in different areas. Should the data collected in this manner demonstrate the BRD meets the bycatch reduction criterion, there is a greater likelihood the BRD would be effective under a broader array of actual commercial fishing conditions.

Tow Times

Currently, the sampling procedures require the selection of a fixed tow time before beginning a test. Each tow may not deviate more than 10 percent from the selected tow time. The fishing efficiency of a net changes (decreases) during a tow as the catch in the net increases. The fixed-tow time requirement was intended to reduce that source of variability in the data set, thus reducing the resulting uncertainty associated with the sample mean reduction rate.

However, because of the non-random and patchy distribution and abundance of organisms in the marine environment, a decrease or increase in the tow time may be necessary during a specific BRD test. For example, the total catch taken during a tow may be greater than anticipated. If so, it may not be possible to keep the catch from each net separated for sampling, thus precluding a successful sample of the catch from a specific net. Under such conditions, shorter tow times would produce manageable quantities of catch for sampling. Conversely, catches of shrimp may be lower than anticipated, and the vessel captain may want to increase the tow time. In either case, under the current requirements, the test would have to be aborted and reinitiated if the tow time were changed in increments greater than 10 percent of the original tow time.

This proposed rule would allow the tow time to be changed after the initiation of a test. The applicant would still be required to propose a preferred tow time in the operations plan submitted to the RA as part of the application for a Letter of Authorization (LOA). However, the applicant would be allowed to make reasonable adjustments to the tow times during a given test to adapt to local fishing conditions and successfully complete the test.

Because the fishing efficiency of a trawl will change depending on the amount of catch in the net, and the efficiency of the experimental BRD similarly may be affected by the amount of catch in the net, excessive differences in tow times for segments of a complete 30-tow test sample could introduce a bias in the overall results. Therefore, any tow time changes would need to be described and justified in a report submitted to the RA at the conclusion of the test. The RA would have to approve the changes before the data would be evaluated for certification. The RA would consult with scientific and technical staff, including the SEFSC, regarding the acceptability of any alterations prior to making a final determination.

Gear Tuning and Fishing Efficiency Bias

As noted in the “Background” section, the basic assumption in assessing the bycatch reduction efficiency of the BRD candidate during paired-net tests is the BRD candidate in the experimental net represents the only factor causing a difference in catch from the control net. Therefore, prior to beginning a test series, the nets to be used in the tests must be calibrated (tuned) to minimize, to the extent practicable, any differences in catch efficiency, or “bias”. Nets would need to be tuned again after any gear modification or change.

Even so, some efficiency bias may remain between nets, or biases may develop during the test. To address the issue of potential biases in fishing efficiency between nets, the current procedures require rotation of the functioning experimental BRD between the port and starboard nets every four to six tows (Gulf of Mexico) or daily (South Atlantic). The intent of this requirement was to negate any remaining bias by introducing that bias into both the control and experimental data on a regular basis, thus reducing the uncertainty associated with the resulting sample mean reduction rate.

To move a complex BRD candidate integrated into the structure of the trawl (e.g., a soft turtle excluding device) may require moving large sections of the net, or even the entire net, on each side of the vessel. This would require loading the trawl doors onboard, disconnecting, moving and re-connecting the nets, and re-deploying the doors and nets overboard. This activity can take several hours to complete. Not only does this increase non-fishing time for the commercial vessel, it increases the amount of time required to complete a BRD test. The need to load and handle the heavy trawl doors and other fishing equipment on a frequent basis increases concerns about vessel crew safety. All of these issues can be alleviated by allowing greater flexibility in the establishment of a rotational schedule best meeting the needs of the specific proposed test.

This proposed rule would remove the static requirement to rotate the BRD every few tows, and allow the applicant to propose, as part of the application for a LOA from the RA, a reasonable gear rotation schedule to accommodate the complexity of the gear being tested. The proposed rotational schedule would still need to ensure equal numbers of tows are conducted with the BRD candidate in both the port and starboard nets. Because the applicant would be monitoring the catch rates in each net after each tow, if a substantial bias develops, the applicant could take action to re-tune the gear or increase the rotational schedule as needed. The applicant's proposed rotational schedule would have to be approved by the RA before the LOA would be issued. If the rotational schedule is changed during the test, the applicant would need to provide a rationale for the action in the final report submitting the data for certification. The RA would consult with scientific and technical staff regarding the acceptability of any changes to the rotational schedule prior to making a final determination regarding the acceptability of the data.

Use of a Try Net During a BRD Test

A try net is a separate, small net pulled for brief periods by a shrimp trawler during an extended trawling effort to test for shrimp concentrations or determine fishing conditions. In the case of vessels fishing four nets (quad-rigged), the nets being used to evaluate the experimental BRD are positioned beyond the influence of the try net, thus the use of a try net on a quad-rigged vessel is allowed under the current procedures. However, on a vessel pulling only two nets (twin-rigged) the try net is fishing in front of the main net on the same side of the vessel. In that case, the try net is removing or diverting some catch before the catch could enter the main net, and introducing bias.

To avoid that bias, the current requirements in the Gulf of Mexico prohibit the use of a try net during BRD tests conducted aboard twin-rigged vessels. Nevertheless, the use of a try net is an integral part of normal shrimping activities, ensuring the vessel is fishing on commercial quantities of shrimp during each extended tow. Because BRD candidate tests are intended to be conducted aboard actively fishing commercial vessels, even if a state government, academic institution, or other entity is the applicant of record, the quantity of shrimp and incidental catch should reflect real fishing conditions. Use of a try net is necessary to ensure the catch levels reflect those expected during normal commercial shrimping operations.

The proposed rule would modify the procedures in the Manual to allow the use of a try net during BRD tests aboard twin-rigged vessels with the try net fishing directly in front of one of the main test nets. To minimize and negate the potential bias, NMFS is proposing a condition requiring the fishing time for the try net to remain a consistent percentage of the total tow time for each tow throughout the course of the test. This condition would expose both the control and experimental nets (as they are rotated) to equivalent effects introduced by the try net. This requirement should adequately address the shrimp fishermen's need to use a try net as part of the commercial operation while negating any potential bias introduced from the use of the try net.

Data Collection

The current procedures require the collection of information on a variety of species taken as catch and bycatch in shrimp trawls. The current SAFMC Manual requires the collection of information on 25 species or species groups of finfishes. However, the certification criterion is a 30-percent reduction, by weight, in total finfish, in aggregate, not individual species. Therefore, the species specific data requirement is outdated, and while informative, is not needed to determine whether a BRD meets the existing certification criterion. For the western Gulf, currently a BRD is certified only on its ability to reduce the bycatch mortality of juvenile red snapper. However, this proposed rule would revise the western Gulf criterion to also be a 30-percent reduction in total finfish, and the specific requirements to sample red snapper would no longer be appropriate.

The proposed rule would reduce mandatory data collection requirements for tests conducted to certify a BRD. Mandatory data collection during a certification test would be limited to recording the total catch of each net, the total catch of commercial shrimp in each net, and the total catch (or total catch in a pre-determined sample) of all finfish species in aggregate. For tests conducted in the western Gulf, applicants would be encouraged to record the total catch of red snapper in each net, but these data would not be used in making a decision to certify a BRD. Similarly, for all areas, data collection for any other specific portions of the catch (i.e., specific finfish species) is encouraged but voluntary, as this information is not required for the certification of the BRD candidate.

Statistical Evaluation

The current certification approach was developed from the procedures used in the Congressionally-mandated BRD research program of the early 1990s. From a statistical standpoint, the goal is to develop a procedure that has zero chance of passing a device with a true reduction less than the target value, and zero chance of failing a device with true reduction greater than the target value. Realistically, there will always be some probability a BRD with true reduction less than the target criterion will pass (Type I error), and some probability a BRD with true reduction greater than the target criterion will fail (Type II error). In a certification context, a Type II error (rejecting an acceptable BRD) has important negative conservation consequences, i.e., not being able to use a more effective BRD, or not having a wider variety of BRD types available for use. A Type I error (accepting an unsatisfactory BRD) may also have negative conservation consequences.

The concept of Type II errors is of general concern to the statistical community, and has prompted substantial statistical research and scientific publications on the properties of Type II error. The probability of a Type II error of a hypothesis test is known as the power of the test. Power analyses of the existing BRD data indicated, because of the inherent variability, certification of devices was unlikely unless the BRD demonstrated a 60- to 70-percent sample mean reduction rate. This was not the intent when NMFS established certification criteria of substantially lesser values.

It is preferable to be able to evaluate an experimental BRD via probability statements of the form “There is at least 'X' probability the true reduction meets the target.” SEFSC scientists have recommended the use of a statistical standard, based on a Bayesian approach, as a more applicable method than the current use of the “classical” Student-t test, or frequentist approach. The Bayesian approach is more instructive about how competing risks (Type I and Type II errors) can be controlled, given the new information now available regarding the statistical power of the data and approaches. Additionally, the Bayesian approach allows for the development and evaluation of the capabilities of an experimental BRD in terms of probability statements.

The proposed rule would replace the current “classical” statistical approach with a Bayesian approach. Under a Bayesian approach, two probability statements would address the existing null hypothesis regarding the certification of a BRD. These probability statements would be: (1) The probability the true reduction meets the target is at least 'A'; and (2) The probability the true reduction is less than some minimum threshold is not more than 'B'. The probability statements are based on observed data sets.

To be certified, the data set for a BRD candidate would need to demonstrate a best point estimate (sample mean) meeting the certification criterion. Additionally, the BRD candidate would have to satisfy both probability statements above. The statistical properties of the data being collected dictate a 50-percent probability value for 'A'. For any BRD, even if it were tested indefinitely under identical conditions, there would be an ever-narrowing probability distribution on either side of the mean observed reduction rate. Nevertheless, half the probability distribution would include values less than the mean, and half of the distribution would include values greater than the mean. Therefore, to certify BRDs capable of meeting the target, NMFS has determined the first probability statement can be adequately expressed as: “There is at least a 50-percent probability the true reduction meets the bycatch reduction criterion.” This would be similar to other NMFS actions that have at least a 50-percent probability of achieving a stock rebuilding target.

There will always be some risk the data set generated for a specific device will result in a sample mean reduction rate meeting the certification criterion, when the device's true reduction rate is less than the certification criterion. Therefore, selecting a value for a minimum threshold and a value for 'B' is a greater focus to managing the risk of accepting a BRD not meeting the criterion. To address this issue, it is necessary to establish a minimum threshold level, below the target criterion, which is completely unacceptable, and set 'B' accordingly, such that there is only a low risk of accepting a BRD because of chance variation in the available data.

Based on the statistical results generated from data sets certifying the BRDs currently in use in the South Atlantic and Gulf of Mexico shrimp fisheries, SEFSC scientists have determined the second probability statement can be adequately expressed as: “There is no more than a 10-percent probability the reduction rate of the BRD candidate is more than 5 percent less than the bycatch reduction criterion.” In other words, for the current 30-percent finfish reduction target, there is no more than a 10-percent probability the true reduction rate of the BRD candidate is less than 25 percent.

The proposed change would increase the opportunity to certify a greater variety of BRDs for use in the fishery, while maintaining a statistical confidence in regard to the efficiency of the BRD. BRDs may have different capabilities under different fishing conditions, and having a wider variety of BRDs for use in the fishery would allow fishermen to choose the most effective BRD for the specific local fishing conditions. This would enhance compliance with national standard 9 of the Magnuson-Stevens Act, and in the western Gulf of Mexico, potentially accelerate the rebuilding efforts for the overfished red snapper resource in the Gulf of Mexico.

Provisional Certification

In addition to revising the statistical evaluation for BRD certification, NMFS proposes to create a “provisional certification” category for experimental BRDs. A provisional certification would apply to an experimental BRD not quite meeting the criteria for certification, but deemed likely to meet the criteria with further testing. To be provisionally certified, statistical analyses of the test results for an experimental BRD must demonstrate there is at least a 50-percent probability the true reduction rate of the BRD candidate is no more than 5 percent less than the bycatch reduction criterion.

In other words, the BRD candidate must demonstrate a best point estimate (sample mean) within 5 percent of the certification criterion.

A provisional certification of a BRD would be effective for 2 years from the date of publication in the Federal Register of any final rule determining provisional certification. This time period would allow additional wide-scale industry evaluation of the BRD candidate. The intent would be to further refine the design or application of the experimental BRD so it could eventually meet the certification criterion.

Certification of New BRDs

The new BRD certification criterion to be established with this proposed rule, along with the revisions to the Manual, especially the addition of a “provisional certification,” would allow new and more effective BRDs to be certified for use in the fishery. There would be no change to the status of the existing certification of the Jones Davis BRD in the southeast shrimp fishery. The original data used to certify that BRD indicate it achieves a 58-percent reduction in total finfish bycatch; there is a 100-percent probability the true reduction rate meets the certification criterion.

The proposed rule would certify the Modified Jones Davis BRD for use by the shrimp fishery throughout EEZ of the Gulf and South Atlantic. This device has been demonstrated to provide a 33-percent reduction in total finfish bycatch. The power test indicates this device has a 98-percent probability the true reduction rate of the BRD is greater than the certification criterion, and there is less than a 1-percent probability the true reduction rate of the BRD is 25 percent or less.

The proposed rule would also provisionally certify the extended funnel BRD for use in the western Gulf. The extended funnel BRD is currently certified for use in the eastern Gulf and South Atlantic. The data set from the 1990's certifying the extended funnel BRD indicated it reduced total finfish by 30 to 35 percent. Newer information collected during 2001 through 2003 in the Gulf indicates the extended funnel BRD is reducing finfish by only about 27 percent. Therefore, the extended funnel BRD would not meet the proposed new certification criterion. However, consistent with the proposed criterion for provisional certification, there is a 74-percent probability the true reduction rate of the BRD is at least 25 percent. Therefore, this proposed rule would change the status of the extended funnel BRD in the Gulf to a provisional certification which would remain effective for two years from the date of publication of any final rule to implement this regulatory amendment. NMFS anticipates additional work on the extended funnel BRD would improve its performance, and allow it to meet the certification criterion. No new information is available regarding the efficacy of the extended funnel BRD in the South Atlantic. The shrimp fishery in the South Atlantic tends to operate in shallower water and has a different species composition to its bycatch. The new information on the extended funnel BRD was all collected in the Gulf of Mexico; there are no new data collected from the South Atlantic fishery to indicate the BRDs are not meeting the bycatch reduction targets. Therefore, the BRD will remain certified in the South Atlantic based on prior determinations the BRD meets the criterion in that part of the fishery.

This proposed rule would also provisionally certify one new design, the composite panel BRD, for use in the Gulf and South Atlantic shrimp fisheries. This BRD design has only been tested in the Gulf, but with a provisional certification, this BRD can be more extensively evaluated for its use in the South Atlantic. The mean sample reduction rate is 25.1 percent. There is a 52- percent probability the true reduction rate of this BRD design is at least 25 percent. Therefore, NMFS proposes to provisionally certify this BRD design. This provisional certification would remain effective for two years from the date of publication of any final rule to implement this regulatory amendment; NMFS anticipates this would allow sufficient time to further test this design in both the Gulf and South Atlantic fisheries.

The fisheye BRD was one of two BRD designs originally certified under the existing criterion for use in the western Gulf. Because of its simplistic design and low cost, it became the industry standard. The most common configuration and placement in the trawl is greater than 10.5 ft (3.2 m) from the trawl's cod end tie-off. According to NMFS' SEFSC estimates, the fisheye BRD in this configuration is achieving between 11- and 25-percent reductions in fishing mortality on juvenile red snapper and a 14- to 23-percent reduction in finfish bycatch by weight. Thus, it does not meet the current red snapper morality target or the proposed 30-percent finfish reduction criterion. Whether the criterion is changed or not, NMFS would not be able to maintain the certification of the industry-standard fisheye BRD placed 10.5 ft (3.2 m) forward because it does not meet the existing red snapper criterion or the proposed 30-percent finfish reduction criterion. However, placed in other areas of the cod end, this type of BRD is more effective, and NMFS is developing subsequent rulemaking to modify the allowable placement of the fisheye BRD in trawl nets. The analysis in this proposed rule discusses indirect impacts arising from the change in the certification criterion, and its potential impact on the future certification and possible decertification or revision to allowable BRDs. For example, it appears at this time that the fisheye BRD would be restricted in its allowable placement in the shrimp trawl net. NMFS is developing separate rulemaking to address this additional change, and the potential direct economic impacts associated with Gulf shrimp vessels having to change or modify the current placement of BRDs in their shrimp trawl nets will be fully analyzed in the subsequent rule.

Similarly, it appears the efficiency of the expanded mesh BRD, currently certified for use in the eastern Gulf and South Atlantic, has decreased. During the original tests of the expanded mesh BRD in the mid-1990s, the BRD achieved between 30- and 35-percent reduction in total finfish. Recent tests of the expanded mesh BRD in the Gulf indicate it is only achieving about a 17-percent reduction in total finfish, thus, it does not meet the criteria to be certified or provisionally certified. NMFS may revise the certification status of the expanded mesh BRD in a separate rulemaking.

For all of these BRD designs, the potential of the BRDs has not changed, but it appears fishing behavior, or some other factor in the fleet, has changed. Actions to maximize shrimp retention, without concurrently maintaining fish reductions, have diminished the BRDs' effectiveness to reduce bycatch. There have been numerous technological changes to the overall construction of shrimp trawl gear, such as new turtle excluder devices and longer nets. In addition, there have been changes in fishing practices to help increase shrimp retention, such as faster towing speeds and modified retrieval procedures. The exact reasons for the BRDs' change in efficiency are not known.

The new BRDs would actually improve red snapper bycatch reduction and general finfish reduction relative to what the industry is currently achieving with its use of the forward-placed fisheye BRD because these new BRDs have a better exclusion rate than the industry standard. The forward-placed fisheye BRD reduces fishing mortality on juvenile red snapper by about 11 percent and reduces the biomass of finfish by about 14 percent. The Modified Jones Davis BRD reduces red snapper mortality by approximately 31 percent and reduces finfish by 33 percent. The extended funnel BRD reduces juvenile red snapper mortality by approximately 25 percent and reduces finfish biomass by about 27 percent.

NMFS is now addressing red snapper management through measures proposed in the Joint Amendment 27 to the FMP for the Reef Fish Resources of the Gulf of Mexico and Amendment 14 to the FMP for the Shrimp Fishery of the Gulf of Mexico. NMFS has initiated review of this joint amendment and announced the availability of this joint amendment for public comment on July 26, 2007 (72 FR 41046). Given the current declines in the number of participants and effort expended by the shrimp fishery, it is more practicable to control red snapper mortality in the shrimp fishery through effort controls of that fishery versus the use of BRDs. However, BRDs still play an important role in addressing national standard 9 for total bycatch reduction potential.

Classification

Pursuant to section 304(b)(1)(A) of the Magnuson-Stevens Act, I have determined that this proposed rule is consistent with the regulatory amendment proposing these BRD-related revisions, other provisions of the Magnuson-Stevens Act, and other applicable law, subject to further consideration after public comment.

This proposed rule has been determined to be significant for purposes of Executive Order 12866.

NMFS prepared an IRFA, as required by section 603 of the Regulatory Flexibility Act, for this proposed rule. The IRFA describes the economic impact this proposed rule, if adopted, would have on small entities. A description of the action, why it is being considered, and the legal basis for this action are contained at the beginning of this section in the preamble and in the SUMMARY section of the preamble. A copy of the full analysis is available from NMFS (see ADDRESSES). A summary of the IRFA follows.

The Magnuson-Stevens Act provides the statutory basis for the proposed rule. The proposed rule would modify the procedures for field testing BRD candidates for use in the Gulf of Mexico and South Atlantic EEZ commercial shrimp fisheries and would modify the bycatch reduction criterion for certifying BRDs for use in the penaeid shrimp fishery in the Gulf EEZ west of Cape San Blas, FL.

The purpose of this proposed rule is to implement more practical field testing procedures for BRD certification candidates and to establish a realistic bycatch reduction threshold for the Gulf EEZ commercial shrimp fishery.

No duplicative, overlapping or conflicting Federal rules have been identified.

The primary entities that are expected to apply for the BRD certification process are state government, academic, and not-for-profit entities. Independent commercial shrimping operations in either the Gulf or South Atlantic may also be included among applicants. NMFS estimates up to 24 applicants will apply for the BRD certification process during the first year and a smaller number in following years. While the identity of entities that might pursue future BRD testing cannot be determined with any certainty, based on past applicants, BRD testing is expected to be undertaken by NOAA Fisheries Service, the Texas Parks and Wildlife Department, the Florida Department of Environmental Protection, Texas A&M University, the University of Georgia, other institutions, and owners of shrimp vessels in the Gulf.

There are approximately 700 vessels permitted to operate in the South Atlantic EEZ commercial shrimp fishery. The most current assessment of the South Atlantic commercial shrimp fishery covers the period 2000-2002 and encompasses vessels that operated in both state and EEZ waters. While this assessment covered a larger universe of vessels, an average of approximately 1,900 vessels per year, and different economic conditions, it represents the best profile available at this time. Over this period, average gross revenue per vessel ranged from approximately $71,000 to approximately $81,000. The highest gross revenue per vessel from all commercial harvesting activities did not exceed $1.0 million.

For the Gulf EEZ, as of March 26, 2007, a moratorium permit is required to fish for shrimp. Although it is unknown how many eligible applicants will apply for a moratorium permit, 2,666 vessels would qualify for the permit and are assumed to constitute the universe of indirectly affected shrimping vessels.

An evaluation of revenue distribution by vessel size indicates substantial differences in yearly average revenues between large (at least 60 ft (18.3 m) in length) and small vessels in the Gulf EEZ commercial shrimp fishery. For the large vessel group, average annual revenues per vessel in 2004 was approximately $140,000, while the comparable value for small vessels was approximately $27,000. Across all vessels, the average annual gross revenue per vessel was approximately $110,000. Maximum yearly gross revenue reported by a qualifying vessel was approximately $1,046,000.

On average, “small” vessels are also “smaller” in regards to almost all of their physical attributes (e.g. they use smaller crews, fewer and smaller nets, have less engine horsepower and fuel capacity, etc.). Small vessels are also older on average. Larger vessels also tend to be steel-hulled. Fiberglass hulls are most prominent among small vessels, though steel and wood hulls are also common. Nearly two-thirds of large vessels have freezing capabilities while few small vessels have such equipment. Small vessels still rely on ice for refrigeration and storage, though more than one-third of large vessels also rely on ice. Some vessels are so small that they rely on live wells for storage.

An important difference between large and small Gulf EEZ commercial shrimp vessels is with respect to their dependency on the food shrimp fishery. The percentage of revenues arising from food shrimp landings is approximately 81 percent for large vessels, but only approximately 58 percent for small vessels. Thus, on average, large vessels are more dependent than their smaller counterparts on the food shrimp fishery. However, dependency on food shrimp is much more variable within the small vessel sector than the large vessel sector. Many small vessels are quite dependent on food shrimp landings, while others illustrate little if any dependency.

Finally, according to recent projections, on average, both small and large Gulf EEZ commercial shrimp vessels are experiencing significant economic losses, ranging from a -27 percent rate of return in the small vessel sector to a -36 percent rate of return in the large vessel sector (-33 percent on average for the fishery as a whole). Therefore, almost any but the most minor additional financial burden would be expected to generate a significant adverse impact on affected vessels and potentially hasten additional exit from the fishery.

The Small Business Administration (SBA) defines a small organization as any not-for-profit enterprise that is independently owned and operated and not dominant in its field of operation. This definition includes private educational institutions. The SBA also defines a small governmental jurisdiction as the government of cities, counties, towns, townships, villages, school districts, or special districts with a population less than 50,000. Finally, the SBA defines a small business in the commercial fishing activity as an entity that is independently owned and operated, is not dominant in its field of operation (including its affiliates), and has average annual total receipts not in excess of $4.0 million annually (NAICS codes 114111 and 114112, finfish and shellfish fishing).

While the identity of entities that might pursue future BRD testing cannot be determined with any certainty, based on past applicants, BRD testing is expected to be undertaken by NOAA Fisheries Service, the Texas Parks and Wildlife Department, the Florida Department of Environmental Protection, Texas A&M University, the University of Georgia, other institutions, and owners of shrimp vessels in the Gulf. The respective state agencies are extensions of the respective state governments and, as such, clearly exceed the SBA population thresholds for small government entities. Similarly, both Texas A&M University and the University of Georgia are, as public universities, extensions of the respective state government educational systems, with staff being state employees, and, therefore, would similarly be appropriately classified as large entities. Although no private colleges or universities that might apply for the BRD testing process have been identified, as private rather than public educational institutions, while some exceptions may exist, private educational institutions generally are understood to be smaller in terms of student population, staff, and operational budgets than public institutions and, as such, are determined for the purpose of this analysis to be small entities. Given the aforementioned maximum annual revenue figures for Gulf and South Atlantic commercial shrimping operations, vessels that would be expected to participate in the certification program are determined to be small business entities for the purpose of this analysis. Thus, most entities that may apply for the BRD certification process are likely to be small entities, and only a maximum of 24 entities would be expected to apply the first year, with fewer entities applying in subsequent years.

All entities that would qualify for the Gulf EEZ commercial shrimp fishery moratorium permit, 2,666 vessels, would be expected to be indirectly affected by the proposed Gulf bycatch reduction criterion. Given the maximum revenue provided above for Gulf EEZ commercial shrimping operations, all shrimp vessels that have the potential to be indirectly impacted by the proposed change in the Gulf bycatch reduction criterion are determined to be small entities for the purpose of this analysis.

The outcome of “significant economic impact” can be ascertained by examining two issues: disproportionality and profitability.

The disproportionality question is: do the proposed regulations place a substantial number of small entities at a significant competitive disadvantage to large entities? Revision to the Manual would not be expected to result in any direct or indirect adverse economic impacts to any affected entities since the reporting burden per applicant will not increase and the revisions, in and of themselves, will not cause any BRDs to be certified, provisionally certified, or decertified in future actions. Therefore, the issue of disproportionate impacts would not apply to this action.

Similarly, the proposed change to the Gulf EEZ commercial shrimp fishery bycatch reduction criterion would not result in any direct adverse economic impacts on participants in the Gulf EEZ commercial shrimp fishery. However, the change in the bycatch reduction criterion would be expected to generate indirect impacts on vessels in the Gulf EEZ commercial shrimp fishery as a result of future certification, provisional certification, and/or decertification actions. All of these vessels have been determined to be small business entities. Hence, the issue of disproportionality would also not apply to this action.

The proposed certifications and provisional certifications would also impact all vessels in the Gulf EEZ commercial shrimp fishery, as well as vessels in the South Atlantic EEZ commercial shrimp fishery in some cases. As all of these entities were determined to be small entities, the issue of disproportionality would not apply to these proposed actions.

The profitability question is: do the regulations significantly reduce profit for a substantial number of small entities?

The proposed revision of the Manual would not directly affect fishery participation or harvest because it merely establishes procedures under which research and gear development may proceed. The proposed bycatch reduction criterion for the Gulf EEZ commercial shrimp fishery is not expected to result in any direct adverse economic impacts the participants in this fishery because it is an administrative action.

The proposed criterion would, however, be expected to result in decertification of some currently used BRDs/configurations through subsequent regulatory action. This decertification would require the use of alternative certified or provisionally certified BRDs and would result in increased operating costs. Among the BRDs currently in use, the maximum increase in operating costs that would be incurred as a result of future decertification would be the first-year BRD replacement costs, ranging from $2,550 to $4,250 per vessel per year, associated with the Jones-Davis BRD--the most expensive of the remaining certified BRDs. This increase would represent between 2.3 percent and 3.8 percent of an average vessel's annual revenues. Industry-wide, the re-gearing costs for the Gulf EEZ commercial shrimp fishery would be expected to range from approximately $2.8-$10.1 million for all moratorium permit qualifiers, or approximately $2.2-$7.7 million if only active qualifiers elect to obtain moratorium permits. However, these costs would directly accrue only to a subsequent rule and not to the current proposed action.

The proposed criterion would also allow for the Modified Jones-Davis BRD to be certified for use in the Gulf of Mexico and South Atlantic EEZ shrimp fisheries, the extended funnel BRD to be provisionally certified for use in the western Gulf EEZ shrimp fishery, and the composite panel BRD to be provisionally certified for use in the Gulf of Mexico and South Atlantic EEZ shrimp fisheries, as is proposed in this rule. However, these three BRDs are used by few shrimp vessel owners at present, are more costly to purchase, and attain higher levels of shrimp loss on average relative to the predominantly used fisheye BRD. As such, no shrimp vessel owners would be expected to voluntarily switch from their currently used BRDs to these BRDs. As such, no direct impacts would result from their certification or provisional certification. Therefore, this proposed rule would not be expected to result in any direct impact on the profitability of any small business entities in the shrimp fishery or associated industries. However, substantial reductions in annual gross revenues could occur as a result of subsequent BRD decertification associated with future rulemaking. Depending upon the BRD type currently used and the availability of replacements, small vessels could lose from approximately $300 to $4,000, or from less than 1 percent to more than 8 percent of annual gross revenues, while large vessels could experience a small gain of approximately $600 to a loss of $26,000, or a less than 1 percent gain to a greater than 14 percent loss. Even assuming net shop supply is able to meet demand, if all vessels are able to switch to certified BRDs, the range of impacts is only reduced to a maximum projected annual loss of $1,400 (3 percent) for small vessels and $14,000 (8 percent) for large vessels, though this last figure would apply to relatively few vessels, with the majority of large vessels projected to experience a loss of $3,500 to $4,000 (2 percent) reductions in annual gross revenues.

The management measures considered in this proposed rule do not affect the reporting or record-keeping requirements for shrimp vessels. This proposed action, which only modifies the performance standards used in BRD certification, does not require additional records or report preparation.

Two alternatives, the proposed alternative and the status quo, were considered for the action to modify the Manual. The status quo would continue overly restrictive and inflexible testing procedures and would not achieve NMFS' objectives.

Three alternatives, including the status quo, were considered for the action to change the BRD bycatch reduction criterion. Two alternatives contained multiple options, resulting in seven effective alternatives. As previously discussed, changing the criterion is an administrative action and would not simultaneously decertify BRDs currently in use or require immediate replacement. Decertification, with attendant costs, however, could be expected to occur through subsequent action.

The status quo would be expected to result in the decertification of the fisheye BRD for use in the Gulf commercial shrimp fishery, inducing industry-wide replacement costs of approximately $6.0-$10.1 million for all moratorium permit qualifiers, or approximately $4.6-$7.7 million if only active qualifiers elect to obtain moratorium permits. The minimum range of these costs is greater than that of the proposed rule because while the proposed rule could also lead to the decertification of the fisheye BRD via subsequent action, it would allow the use of the cheaper modified Jones-Davis BRD.

The second alternative would continue to base the bycatch reduction target on juvenile red snapper, similar to the status quo, but considered three different minimum thresholds. The two lower thresholds (12 percent and 20 percent) would be expected to allow continued use of the fisheye BRD, which is the most commonly used BRD, resulting in no direct adverse economic impacts and no increased indirect costs. Neither threshold, however, would meet the objective of national standard 9, which requires that bycatch be reduced to the extent practicable. Hence, these lower thresholds would not meet the Magnuson-Stevens Act's requirements. The highest threshold (30 percent) would be expected to result in the same effects as the status quo, resulting in greater indirect adverse economic impacts than the proposed rule.

The third alternative would base the bycatch reduction criterion on all finfish and considered four minimum thresholds, ranging from 10-40 percent. The two lower thresholds (10 percent and 20 percent) would be expected to allow continued use of fisheye BRDs, resulting in no direct adverse economic impacts and increased indirect gear costs. However, neither threshold would meet the Magnuson-Stevens Act requirement of achieving bycatch reduction to the extent practicable. The highest threshold (40 percent) would not be expected to result in any direct adverse economic impacts but would be expected to result in indirect increased gear costs equal to those of the status quo, which are higher than those of the proposed rule. This alternative would also set an excessive standard that few BRD designs could achieve.

This rule contains approved collection-of-information requirements--namely, the BRD certification process, consisting of applications for pre-certification or certification of a new BRD, pre-certification adjusting, the testing itself, the submission of the test results, application for observer position, and references for observers, subject to the Paperwork Reduction Act (PRA). These collection-of-information requirements have been approved by OMB under Control Number 0648-0345. The public reporting burden for this collection of information which includes the application, pre-certification phase, testing, and submission of results, is estimated to average 194 hours per test. The public reporting burden for applying for an observer position will average 1 hour per response, and the burden for obtaining references will average 1 hour per response. The collection consists of an Application Form, Vessel Information Form, Gear Specification Form, TED/BRD Specification Form, Station Sheet Form, Species Characterization Form, Length Frequency Form, and Condition and Fate Form. The average response time for each of these forms is 20 minutes, except for the Species Characterization Form which has a 2.8-hour response time and the Application Form which has a 2.3-hour response time. In addition, 4 hours will be needed to prepare the final report. These burden estimates include the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding these burden estimates or any other aspect of the collection-of-information requirement, including suggestions for reducing the burden, to NMFS and to OMB (see ADDRESSES).

Notwithstanding any other provision of law, no person is required to respond to, nor shall a person be subject to a penalty for failure to comply with, a collection of information subject to the requirements of the PRA unless that collection of information displays a currently valid OMB control number.

List of Subjects in 50 CFR Part 622

Fisheries, Fishing, Puerto Rico, Reporting and recordkeeping requirements, Virgin Islands.

Dated: October 9, 2007.

John Oliver,

Deputy Assistant Administrator for Operations, National Marine Fisheries Service.

For the reasons set out in the preamble, 50 CFR part 622 is proposed to be amended as follows:

PART 622—FISHERIES OF THE CARIBBEAN, GULF, AND SOUTH ATLANTIC

1. The authority citation for part 622 continues to read as follows:

Authority: 16 U.S.C. 1801 et seq.

2. In § 622.41, paragraph (h) is removed and reserved and paragraph (g) is revised to read as follows:

§ 622.41

Species specific limitations.

* * * * *

(g) BRD requirement for Gulf and South Atlantic shrimp. On a shrimp trawler in the Gulf EEZ or South Atlantic EEZ, each net that is rigged for fishing must have a BRD installed that is listed in paragraph (g)(2) of this section and is certified or provisionally certified for the area in which the shrimp trawler is located, unless exempted as specified in paragraphs (g)(1)(i) through (iv) of this section. A trawl net is rigged for fishing if it is in the water, or if it is shackled, tied, or otherwise connected to a sled, door, or other device that spreads the net, or to a tow rope, cable, pole, or extension, either on board or attached to a shrimp trawler.

(1) Exemptions from BRD requirement—(i) Royal red shrimp exemption. A shrimp trawler is exempt from the requirement to have a certified or provisionally certified BRD installed in each net provided that at least 90 percent (by weight) of all shrimp on board or offloaded from such trawler are royal red shrimp.

(ii) Try net exemption. A shrimp trawler is exempt from the requirement to have a certified or provisionally certified BRD installed in a single try net with a headrope length of 16 ft (4.9 m) or less provided the single try net is either pulled immediately in front of another net or is not connected to another net.

(iii) Roller trawl exemption. A shrimp trawler is exempt from the requirement to have a certified or provisionally certified BRD installed in up to two rigid-frame roller trawls that are 16 ft (4.9 m) or less in length used or possessed on board. A rigid-frame roller trawl is a trawl that has a mouth formed by a rigid frame and a grid of rigid vertical bars; has rollers on the lower horizontal part of the frame to allow the trawl to roll over the bottom and any obstruction while being towed; and has no doors, boards, or similar devices attached to keep the mouth of the trawl open.

(iv) BRD certification testing exemption. A shrimp trawler that is authorized by the RA to participate in the pre-certification testing phase or to test a BRD in the EEZ for possible certification, has such written authorization on board, and is conducting such test in accordance with the “Bycatch Reduction Device Testing Manual” is granted a limited exemption from the BRD requirement specified in this paragraph (g). The exemption from the BRD requirement is limited to those trawls that are being used in the certification trials. All other trawls rigged for fishing must be equipped with certified or provisionally certified BRDs.

(2) Procedures for certification and decertification of BRDs. The process for the certification of BRDs consists of two phases--an optional pre-certification phase and a required certification phase. The RA may also provisionally certify a BRD.

(i) Pre-certification. The pre-certification phase allows a person to test and evaluate a new BRD design for up to 60 days without being subject to the observer requirements and rigorous testing requirements specified for certification testing in the “Bycatch Reduction Device Testing Manual.”

(A) A person who wants to conduct pre-certification phase testing must submit an application to the RA, as specified in the “Bycatch Reduction Device Testing Manual.” The “Bycatch Reduction Device Testing Manual”, which is available from the RA, upon request, contains the application forms.

(B) After reviewing the application, the RA will determine whether to issue a letter of authorization (LOA) to conduct pre-certification trials upon the vessel specified in the application. If the RA authorizes pre-certification, the RA's LOA must be on board the vessel during any trip involving the BRD testing.

(ii) Certification. A person who proposes a BRD for certification for use in the Gulf EEZ or South Atlantic EEZ must submit an application to test such BRD, conduct the testing, and submit the results of the test in accordance with the “Bycatch Reduction Device Testing Manual.” The RA will issue a LOA to conduct certification trials upon the vessel specified in the application if the RA finds that: The operation plan submitted with the application meets the requirements of the “Bycatch Reduction Device Testing Manual”; the observer identified in the application is qualified; and the results of any pre-certification trials conducted have been reviewed and deemed to indicate a reasonable scientific basis for conducting certification testing. If authorization to conduct certification trials is denied, the RA will provide a letter of explanation to the applicant, together with relevant recommendations to address the deficiencies resulting in the denial. If a BRD meets the certification criterion, as determined consistent with the “Bycatch Reduction Device Testing Manual”, NMFS, through appropriate rulemaking procedures, will add the BRD to the list of certified BRDs in paragraph (g)(3) of this section; and provide the specifications for the newly certified BRD, including any special conditions deemed appropriate based on the certification testing results.

(iii) Provisional certification. Based on data provided consistent with the “Bycatch Reduction Device Testing Manual”, the RA may provisionally certify a BRD if there is at least a 50-percent probability the true reduction rate of the BRD is no more than 5 percent less than the bycatch reduction criterion. Through appropriate rulemaking procedures, NMFS will add the BRD to the list of provisionally certified BRDs in paragraph (g)(3) of this section; and provide the specifications for the BRD, including any special conditions deemed appropriate based on the certification testing results. A provisional certification is effective for 2 years from the date of publication of the notification in the Federal Register announcing the provisional certification.

(iv) Decertification. The RA will decertify a BRD if NMFS determines the BRD does not meet the requirements for certification or provisional certification. Before determining whether to decertify a BRD, the RA will notify the appropriate Fishery Management Council in writing, and the public will be provided an opportunity to comment on the advisability of any proposed decertification. The RA will consider any comments from the Council and public, and if the RA elects to decertify the BRD, the RA will proceed with decertification via appropriate rulemaking.

(3) Certified and provisionally certified BRDs—(i) Certified BRDS. The following BRDs are certified for use in the Gulf EEZ and South Atlantic EEZ unless indicated otherwise. Specifications of these certified BRDs are contained in Appendix D to this part.

(A) Fisheye.

(B) Gulf fisheye.

(C) Jones-Davis.

(D) Modified Jones-Davis.

(E) Expanded mesh.

(F) Extended funnel -South Atlantic EEZ only.

(ii) Provisionally certified BRDs. The following BRDs are provisionally certified for use in the areas and for the time periods indicated. Specifications of these provisionally certified BRDs are contained in Appendix D to this part.

(A) Extended funnel- Gulf EEZ only; through the date that is 2 years after the date of publication of the final rule implementing this regulatory amendment.

(B) Composite panel -Gulf EEZ and South Atlantic EEZ; through the date that is 2 years after the date of publication of the final rule implementing this regulatory amendment.

* * * * *

3. In Appendix D to part 622, sections F and G are added to read as follows:

Appendix D to Part 622—Specifications for Certified BRDs

* * * * *

F. Modified Jones-Davis.

1. Description. The Modified Jones-Davis BRD is a variation to the alternative funnel construction method of the Jones-Davis BRD except the funnel is assembled by using depth-stretched and heat-set polyethylene webbing instead of the flaps formed from the extension webbing. In addition, no hoops are used to hold the BRD open.

2. Minimum Construction and Installation Requirements. The Modified Jones-Davis BRD must contain all of the following.

(a) Webbing extension. The webbing extension must be constructed from a single piece of 1 5/8-inch (4.1-cm) stretch mesh number 30 nylon 39 1/2 meshes by 150 meshes. A tube is formed from the extension webbing by sewing the 39 1/2-mesh side together.

(b) Funnel. The funnel must be constructed from two sections of 1 5/8-inch (4.1-cm) heat-set and depth-stretched polypropylene or polyethylene webbing. The two side sections must be rectangular in shape, 25 meshes on the leading edge by 21 meshes deep. The 25-mesh leading edge of each polyethylene webbing section must be sewn evenly two meshes in from the front of the extension webbing starting 25 meshes from the top center on each side. The 21-mesh edge must be sewn to the extension webbing on a 9-bar and 1-mesh angle in the top and bottom, forming a V-shape funnel.

(c) Cutting the escape opening. The leading edge of the escape openings must be located within 18 inches (45.7 cm) of the posterior edge of the turtle excluder device (TED) grid. The area of the escape opening must total at least 635 inches² (4,097 cm²). Two escape openings, 6 meshes wide by 12 meshes deep, must be cut 4 meshes apart in the extension webbing, starting at the top center extension seam, 7 meshes back from the leading edge, and 30 meshes to the left and to the right (total of four openings). The four escape openings must be double selvaged for strength.

(d) Cone fish deflector. The cone fish deflector is constructed of 2 pieces of 1 5/8-inch (4.1-cm) polypropylene or polyethylene webbing, 40 meshes wide by 20 meshes in length and cut on the bar on each side forming a triangle. Starting at the apex of the two triangles, the two pieces must be sewn together to form a cone of webbing. The apex of the cone fish deflector must be positioned within 12 inches (30.5 cm) of the posterior edge of the funnel.

(e) 11-inch (27.9-cm) cable hoop for cone deflector. A single hoop must be constructed of 5/16-inch (0.79-cm) or 3/8-inch (0.95-cm) cable 34 1/2inches (87.6 cm) in length. The ends must be joined by a 3-inch (7.6-cm) piece of 3/8-inch (0.95-cm) aluminum pipe pressed together with a 1/4-inch (0.64-cm) die. The hoop must be inserted in the webbing cone, attached 10 meshes from the apex and laced all the way around with heavy twine.

(f) Installation of the cone in the extension. The apex of the cone must be installed in the extension within 12 inches (30.5 cm) behind the back edge of the funnel and attached in four places. The midpoint of a piece of number 60 twine (or at least 4-mesh wide strip of number 21 or heavier webbing) 4 ft (1.22 m) in length must be attached to the apex of the cone. This piece of twine or webbing must be attached within 5 meshes of the aft edge of the funnel at the center of each of its sides. Two 12-inch (30.5-cm) pieces of number 60 (or heavier) twine must be attached to the top and bottom of the 11-inch (27.9-cm) cone hoop. The opposite ends of these two pieces of twine must be attached to the top and bottom center of the extension webbing to keep the cone from inverting into the funnel.

G. Composite Panel.

1. Description. The Composite Panel BRD is a variation to the alternative funnel construction method of the Jones-Davis BRD except the funnel is assembled by using depth stretched and heat set polyethylene webbing with square mesh panels on the inside instead of the flaps formed from the extension webbing. In addition, no hoops are used to hold the BRD open.

2. Minimum Construction and Installation Requirements. The Composite Panel BRD must contain all of the following:

(a) Webbing extension. The webbing extension must be constructed from a single piece of 1 5/8-inch (4.1-cm) stretch mesh number 30 nylon 24 1/2 meshes by 150 meshes. A tube is formed from the extension webbing by sewing the 24 1/2-mesh side together. The leading edge of the webbing extension must be attached no more than 4 meshes from the posterior edge of the TED grid.

(b) Funnel. The V-shaped funnel consists of two webbing panels attached to the extension along the leading edge of the panels. The top and bottom edges of the panels are sewn diagonally across the extension toward the center to form the funnel. The panels are 2-ply in design, each with an inner layer of 1 5/8-inch (4.1- cm) heat-set and depth-stretched polyethylene webbing and an outer layer constructed of 2-inch (5.1-cm) square mesh webbing (1-inch bar). The inner webbing layer must be rectangular in shape, 36 meshes on the leading edge by 20 meshes deep. The 36- mesh leading edges of the polyethylene webbing should be sewn evenly to 24 meshes of the extension webbing 1 1/2 meshes from and parallel to the leading edge of the extension starting 12 meshes up from the bottom center on each side. Alternately sew 2 meshes of the polyethylene webbing to 1 mesh of the extension webbing then 1 mesh of the polyethylene webbing to 1 mesh of the extension webbing toward the top. The bottom 20-mesh edges of the polyethylene layers are sewn evenly to the extension webbing on a 2 bar 1 mesh angle toward the bottom back center forming a v-shape in the bottom of the extension webbing. The top 20-mesh edges of the polyethylene layers are sewn evenly along the bars of the extension webbing toward the top back center. The square mesh layers must be rectangular in shape and constructed of 2-inch (5.1-cm) webbing that is 18 bars or squares on the leading edge and 32 bars or squares down each side. The 18 bar leading edge of each square mesh layer must be sewn evenly 1 bar to 2 meshes of the 36-mesh leading edge of the polyethylene section and the 32-bar sides are sewn evenly (in length) to the 20-mesh edges of the polyethylene webbing. This will form a v-shape funnel using the top of the extension webbing as the top of the funnel and the bottom of the extension webbing as the bottom of the funnel.

(c) Cutting the escape opening. There are two escape openings on each side of the funnel. The leading edge of the escape openings must be located on the same row of meshes in the extension webbing as leading edge of the composite panels. The lower openings are formed by starting at the first attachment point of the composite panels and cutting 9 meshes in the extension webbing on an even row of meshes toward the top of the extension. Next, turn 90 degrees and cut 15 points on an even row toward the back of the extension webbing. At this point turn and cut 18 bars toward the bottom front of the extension webbing. Finish the escape opening by cutting 6 points toward the original starting point. The top escape openings start 5 meshes above and mirror the lower openings. Starting at the leading edge of the composite panel and 5 meshes above the lower escape opening, cut 9 meshes in the extension on an even row of meshes toward the top of the extension. Next, turn 90 degrees, and cut 6 points on an even row toward the back of the extension webbing. Then cut 18 bars toward the bottom back of the extension. To complete the escape opening, cut 15 points forward toward the original starting point. The area of each escape opening must total at least 212 in² (1,368 cm²). The four escape openings must be double salvaged for strength.

Note:

The “Bycatch Reduction Device Testing Manual” is published, excluding the Manual's appendices, as an appendix to this document. See the contact under ADDRESSES to obtain a complete Manual. This appendix will not appear in the Code of Federal Regulations.

Appendix—Bycatch Reduction Device Testing Manual Definitions

Bycatch reduction criterion is the standard by which a BRD candidate will be evaluated. To be certified for use by the shrimp fishery in the Exclusive Economic Zone off the southeastern United States (North Carolina through Texas), the BRD candidate must demonstrate a successful reduction of total finfish bycatch by at least 30 percent by weight.

Bycatch reduction device (BRD) is any gear or trawl modification designed to allow finfish to escape from a shrimp trawl.

BRD candidate is a BRD to be tested for certification for use in the commercial shrimp fishery of southeastern United States.

Certified BRD is a BRD that has been tested according to the procedure outlined herein and has been determined by the RA as having met the bycatch reduction criterion.

Control trawl means a trawl that is not equipped with a BRD during the evaluation.

Evaluation and oversight personnel means scientists, observers, and other technical personnel who, by reason of their occupation or scientific expertise or training, are approved by the RA as qualified to evaluate and review the application and testing process.

Experimental trawl means the trawl that is equipped with the BRD candidate during an evaluation.

Net/side bias means when the net(s) being fished on one side of the vessel demonstrate a different catch rate (fishing efficiency) than the net(s) being fished on the other side of the vessel during paired-net tests.

Observer means a person on the list maintained by the RA of individuals qualified (see Appendix H) to supervise and monitor a BRD certification test.

Paired-net test means a tow during certification trials where a control net and an experimental net are fished simultaneously, and the catches and catch rates between the nets are compared.

Provisional Certification Criterion means a secondary benchmark which would allow a BRD candidate to be used for a time-limited period in the southeastern shrimp fishery. To meet the criterion, the BRD candidate must demonstrate a successful reduction of total finfish bycatch by at least 25 percent by weight.

Provisionally certified BRD means a BRD that has been tested according to the procedure outlined herein and has been determined by the RA as having met the provisional certification criterion. A BRD meeting the provisional certification criterion would be certified by the RA for a period of 2 years.

Regional Administrator (RA) means the Southeast Regional Administrator, National Marine Fisheries Service.

Required measurements refers to the quantification of gear characteristics such as the dimensions and configuration of the trawl, the BRD candidate, the doors, or the location of the BRD in relation to other parts of the trawl gear that are used to assess the performance of the BRD candidate.

Sample size means the number of successful tows (a minimum of 30 tows per test are required).

Shrimp trawler means any vessel that is equipped with one or more trawl nets whose on-board or landed catch of shrimp is more than 1 percent, by weight, of all fish comprising its on-board or landed catch.

Successful tow means that the control and experimental trawl were fished in accordance with the requirements set forth herein and the terms and conditions of the letter of authorization, and there is no indication problematic events, such as those listed in Appendix D-5, occurred during the tow to impact or influence the fishing efficiency (catch) of one or both nets.

Tow time means the total time (hours and minutes) an individual trawl was fished (i.e., the time interval beginning when the winch is locked after deploying the net overboard, and ending when retrieval of the net is initiated).

Trawl means a net and associated gear and rigging used to catch shrimp. The terms trawl and net are used interchangeably throughout this Manual.

Try net means a separate net pulled for brief periods by a shrimp trawler to test for shrimp concentrations or determine fishing conditions (e.g., presence of absence of bottom debris, jellyfish, bycatch, and seagrasses).

Tuning a net means adjusting the trawl and its components to minimize or eliminate any net/side bias that exists between the two nets that will be used as the control and experimental trawls during the certification test.

I. Introduction

This Bycatch Reduction Device Testing Manual (Manual) establishes a standardized process for evaluating the ability of bycatch reduction device (BRD) candidates to meet the established bycatch reduction criterion, and be certified for use in the EEZ by the southeastern shrimp fishery. BRDs are required for use in shrimp trawls fished shoreward of the 100-fathom (183-meter) depth contour in the Gulf of Mexico, and within the EEZ of the South Atlantic region.

Various BRD requirements also exist in state waters in the South Atlantic and off Florida and Texas in the Gulf of Mexico. Persons wishing to conduct BRD candidate evaluations exclusively in state waters do not need to apply to NMFS for authorization to conduct these tests, but should contact the appropriate state officials for authorizations. However, for data collected in such evaluations to be considered by NMFS for certification, the operations plan and data collection procedures must meet the criteria established in this Manual.

II. BRD Candidate Evaluations

A. Application

Persons interested in evaluating the efficiency of a BRD candidate must apply for, receive, and have on board the vessel during the evaluation, a Letter of Authorization (LOA) from the Regional Administrator (RA). To receive an LOA, the applicant must submit the following documentation to the RA: (1) a completed application form (Appendix A); (2) a brief statement of the purpose and goal of the activity for which the LOA is requested; (3) an operations plan (see Section C below) describing the scope, duration, dates, and location of the test, and methods that will be used to conduct the test; (4) an 8.5- inch x 11-inch (21.6-cm x 27.9-cm) diagram drawn to scale of the BRD design; (5) an 8.5-inch x 11-inch (21.6-cm x 27.9-cm) diagram drawn to scale of the BRD in the shrimp trawl; (6) a description of how the BRD is supposed to work; (7) a copy of the testing vessel's U.S. Coast Guard documentation or its state registration; and (8) a copy of the testing vessel's Federal commercial shrimp vessel permit.

An applicant requesting an LOA to test an unapproved turtle excluder device (TED) as a BRD (including modifications to a TED that would enhance finfish exclusion) must first apply for and obtain from the RA an experimental TED authorization pursuant to 50 CFR 223.207(e)(2). Applicants should contact the Protected Resources Division of NMFS' Southeast Regional Office for further information. The LOA applicant must include a copy of that authorization with the application.

Incomplete applications will be returned to the applicant along with a letter from the RA indicating what actions the applicant may take to make the application complete.

There is no cost to the applicant for the RA's administrative expenses such as reviewing applications, issuing LOAs, evaluating test results, or certifying BRDs. However, all other costs associated with the actual testing activities are the responsibility of the applicant, or any associated sponsor. If an application for an LOA is denied, the RA will provide a letter of explanation to the applicant, together with relevant recommendations to address the deficiencies that resulted in the denial.

B. Allowable Activities

Issuance of an LOA to test a BRD candidate in the South Atlantic or Gulf of Mexico allows the applicant to remove or disable the existing certified BRD in one outboard net (to create a control net), and to place the BRD candidate in another outboard net in lieu of a certified BRD (to create an experimental net). All other trawls under tow during the test must have a certified BRD, unless these nets are specifically exempted in the LOA. All trawls under tow during the test must have an approved TED unless operating under an authorization issued pursuant to 50 CFR 223.207(e)(2), whereby the test is being conducted on an experimental TED. The LOA, and experimental TED authorization if applicable, must be on board the vessel while the test is being conducted. The term of the LOA will be 60 days; should circumstances require a longer test period, the applicant may apply to the RA for a 60-day extension.

C. Operations Plan

An operations plan should be submitted with the application describing a method to compare the catches of shrimp and fish in a control net (net without a BRD candidate installed) to the catches of the same species in an experimental net (a net configured identically to the control net but also equipped with the BRD candidate).

The applicant may choose to conduct a pre-certification test of a prototype BRD candidate. A pre-certification test would be conducted when the intent is to assess the preliminary effectiveness of a prototype BRD candidate under field conditions, and to make modifications to the prototype BRD candidate during the field test. For pre-certification testing, the operations plan must include only a description of the scope, duration, dates, and location of the test, along with a description of methods that will be used to conduct the test. No observer is required for a pre-certification test, but the applicant may choose to use an observer to maintain a written record of the test. The applicant will maintain a written record for both the control and experimental net during each tow. Mandatory data collection is limited to the weight of the shrimp catch and the weight of the total finfish catch in each test net during each tow. These data must be submitted to NMFS at the conclusion of the test. Although not required, the applicant may wish to incorporate some or all the certification test requirements listed below.

For a BRD candidate to be considered for certification, the operations plan must be more detailed and address the following topics:

(1) The primary assumption in assessing the bycatch reduction efficiency of the BRD candidate during paired net tests is that the inclusion of the BRD candidate in the experimental net is the only factor causing a difference in catch from the control net. Therefore, the nets to be used in the tests must be calibrated (tuned) to minimize, to the extent practicable, any net/side bias in catch efficiency prior to beginning a test series, and tuned again after any gear modification or change. Additional information on tuning shrimp trawls to minimize bias is available from the Harvesting Technology Branch, Mississippi Laboratories, Pascagoula Facility, 3209 Frederic Street, Pascagoula, MS 39568 1207; phone [phone].

(2) A standard tow time for a proposed evaluation should be defined. Tow times must be representative of the tow times used by commercial shrimp trawlers. The applicant should indicate what alternatives will be considered should the proposed tow time need adjustment once the test begins.

(3) A minimum sample size of 30 successful tows using a specific BRD candidate design is required for the statistical analysis described in Section F. No alterations of the BRD candidate design are allowed during a specific test series. If the BRD candidate design is altered, a new test series must be started. If a gear change (i.e., changing nets, doors, or rigging) is required, the nets should be tuned again before proceeding with further tests to complete the 30-tow series. Minor repairs to the gear (e.g., sewing holes in the webbing; replacing a broken tickler chain with a new one of the same configuration) are not considered a gear change.

(4) Biases that might result from the use of a try net should be reduced to the extent practicable. Total fishing times for a try net must be a consistent percentage of the total tow time during each tow made in the test.

(5) To incorporate any net/side bias that remains after the tuning tows (e.g., the effect of a try net), or to accommodate for bias that develops between the control and experimental nets during the test, the operations plan should outline a timetable ensuring that an equal number of successful tows are made with the BRD candidate employed in both the port and starboard nets.

(6) Mandatory data to be collected during a test includes: (1) detailed gear specifications as set forth in Appendices B and C, and (2) pertinent information concerning the location, duration and catch from individual tows as set forth in Appendices D and F.

(7) Following each paired tow, the catches from the control and experimental nets must be examined separately. This requires that the catch from each net be kept separate from each other, as well as from the catch taken in other nets fished during that tow. Mandatory data collections include recording the weight of the total catch of each test net (control and experimental nets), the catch of shrimp (i.e., brown, white, pink, rock, or other shrimp by species) in each test net, and the catch of total finfish in aggregate in each test net.

(8) When recording the detailed information on the species found in the catch, if the catch in a net does not fill one standard 1-bushel [ca. 10 gallon] (30 liters) polyethylene shrimp basket (ca. 70 lb) (31.8 kg), but the tow is otherwise considered successful, data must be collected on the entire catch of the net, and recorded as a “select” sample (see Appendices D and F), indicating that the values represent the total catch of the particular net. If the catch in a net exceeds 70 lb (31.8 kg), a well-mixed sample consisting of one standard 1-bushel [ca. 10 gallon] (30 liters) polyethylene shrimp basket must be taken from the total catch of the net. The total weight of the sample must be recorded, as well as the weights (and numbers as applicable) of the various species or species groups found within that sample. These sample values can then be extrapolated to estimate the quantity of those species or species groups found in the total catch of the particular net.

(9) Although not a criterion for certification, applicants testing BRD candidates are encouraged to collect additional information that may be pertinent to addressing bycatch issues in their respective regions. For example, In the western Gulf of Mexico applicants are especially encouraged to collect information on red snapper. If the applicant chooses to collect these data, the total (“select”) catch of the target species from each test net (not just from the sample) should be recorded along with lengths for as many as individuals per net per tow as set forth in Appendices E and F. Additional information in regard to the catch can be recorded on forms such as Appendix G.

The operations plan should address what the applicant will do should it become necessary to deviate from the primary procedures outlined in the operations plan. The plan should describe in detail what will be done to continue the test in a reasonable manner that is consistent with the primary procedures. For example, it may become necessary to alter the pre-selected tow time to adapt to local fishing conditions to successfully complete the test. Prior to issuing a LOA, the RA may consult with evaluation personnel to review the acceptability of these proposed alterations.

D. Observer Requirements

It is the responsibility of the applicant to ensure that a qualified observer (see Appendix H) is on board the vessel during the certification tests. A list of qualified observers is available from the RA. Observers may include employees or individuals acting on behalf of NMFS, state fishery management agencies, universities, or private industry who meet the minimum requirements outlined in Appendix H. Any change in information or testing circumstances, such as replacement of the observer, must be reported to the RA within 30 days. Under 50 CFR 600.746, when any fishing vessel is required to carry an observer as part of a mandatory observer program under the Magnuson-Stevens Fishery Conservation and Management Act (16 U.S.C. 1801, et seq.), the owner or operator of the vessel must comply with guidelines, regulations, and conditions to ensure their vessel is adequate and safe to carry an observer, and to allow normal observer functions to collect information as described in this Manual. A vessel owner is deemed to meet this requirement if the vessel displays one of the following: (i) a current Commercial Fishing Vessel Safety Examination decal, issued within the last 2 years, that certifies compliance with regulations found in 33 CFR chapter I, and 46 CFR chapter I; (ii) a certificate of compliance issued pursuant to 46 CFR 28.710; or (iii) a valid certificate of inspection pursuant to 46 U.S.C. 3311. The observer has the right to check for major safety items, and if those items are absent or unserviceable, the observer may choose not to sail with the vessel until those deficiencies are corrected.

E. Reports

A report on the BRD candidate test results must be submitted by the applicant or associated sponsor before the RA will consider the BRD for certification. The report must contain a comprehensive description of the tests, copies of all completed data forms used during the tests, and photographs, drawings, and similar material describing the BRD. The captain, vessel owner, or the applicant must sign and submit the cover form (Appendix I). The report must include a description and explanation of any unanticipated deviations from the operations plan which occurred during the test. These deviations must be described in sufficient detail to indicate the tests were continued in a reasonable manner consistent with the approved operations plan procedures. Applicants must provide information on the cost of materials, labor, and installation of the BRD candidate. In addition, any unique or special circumstances of the tests, such as special operational characteristics or fishing techniques which enhance the BRD's performance, should be described and documented as appropriate.

F. Certification

The RA will determine whether the required reports and supporting materials are sufficient to evaluate the BRD candidate's efficiency. The determination of sufficiency would be based on whether the applicant adhered to the prescribed testing procedure or provided adequate justification for any deviations from the procedure during the test. If the RA determines that the data are sufficient for evaluation, the BRD candidate will be evaluated to determine if it meets the bycatch reduction criterion. In making a decision, the RA may consult with evaluation and oversight personnel. Based on the data submitted for review, the RA will determine the effectiveness of the BRD candidate, using appropriate statistical procedures such as Bayesian analyses, to determine if the BRD candidate meets the following criteria:

(1) There is at least a 50-percent probability that the true reduction rate of the BRD candidate meets the bycatch reduction criterion (i.e., the BRD candidate demonstrates a best point estimate [sample mean] that meets the certification criterion); and

(2) There is no more than a 10-percent probability that the true reduction rate of the BRD candidate is more than 5 percent less than the bycatch reduction criterion.

To be certified for use in the fishery, the BRD candidate will have to satisfy both criteria. Criterion 1 will ensure that the observed reduction rate of the BRD candidate has an acceptable level of certainty that it meets the bycatch reduction criterion. Criterion 2 will ensure BRD candidates meeting the bycatch reduction criterion also demonstrate a reasonable degree of certainty that the observed reduction rate represents the true reduction rate of the BRD candidate. This determination ensures the operational use of the BRD candidate in the shrimp fishery will, on average, provide a level of bycatch reduction that meets the established bycatch reduction criterion. Interested parties may obtain details regarding the hypothesis testing procedure to be used by contacting the Harvesting Technology Branch, Mississippi Laboratories, Pascagoula Facility, 3209 Frederic Street, Pascagoula, MS 39568 1207; phone [phone]. Following a favorable determination of the certification analysis, the RA will certify the BRD (with any appropriate conditions as indicated by test results) and add the BRD to the list of certified BRDs in the Federal Register through appropriate rulemaking procedures.

In addition, based on the data provided, the RA may provisionally certify a BRD candidate through appropriate rulemaking procedures based on the following criterion:

There is at least a 50-percent probability that the true reduction rate of the BRD candidate is no more than 5 percent less than the bycatch reduction criterion (i.e., the BRD candidate demonstrates a best point estimate [sample mean] within 5 percent of the certification criterion).

A provisional certification will be effective for 2 years from the date of publication in the Federal Register of a determination of provisional certification. This time period will allow additional wide scale industry evaluation of the BRD candidate, during which additional effort would be made to improve the efficiency of the BRD to meet the certification criterion.

III. BRDs Not Certified and Resubmission Procedures

The RA will advise the applicant, in writing, if a BRD is not certified. This notification will explain why the BRD was not certified and what the applicant may do to either modify the BRD or the testing procedures to improve the chances of having the BRD certified in the future. If certification was denied because of insufficient information, the RA will explain what information is lacking. The applicant must provide the additional information within 60 days from receipt of such notification. If the additional information is not provided within 60 days, the application will be deemed abandoned. If the RA subsequently certifies the BRD, the RA will announce the certification in the Federal Register.

IV. Decertification of BRDs

The RA will decertify a BRD whenever NMFS determines a BRD no longer satisfies the bycatch reduction criterion. Before determining whether to decertify a BRD, the RA will notify the appropriate Fishery Management Council in writing, and the public will be provided an opportunity to comment on the advisability of any proposed decertification. The RA will consider any comments from the Council and public, and if the RA elects to proceed with decertification of the BRD, the RA will publish proposed and final rules in the Federal Register with a comment period of not less than 15 days on the proposed rule.

A provisionally certified BRD is valid for use in the fishery for 2 years from the date of publication of a notice in the Federal Register. If no new data are submitted to indicate the efficiency of the BRD has been improved, the RA will remove the BRD from the list of provisionally certified BRDs.

V. Interactions with Sea Turtles

The following section is provided for informational purposes. Sea turtles are listed under the Endangered Species Act as either endangered or threatened. The following procedures apply to incidental take of sea turtles under 50 CFR 223.206(d)(1):

“Any sea turtles taken incidentally during the course of fishing or scientific research activities must be handled with due care to prevent injury to live specimens, observed for activity, and returned to the water according to the following procedures:

(A) Sea turtles that are actively moving or determined to be dead (as described in paragraph (B)(4) below) must be released over the stern of the boat. In addition, they must be released only when fishing or scientific collection gear is not in use, when the engine gears are in neutral position, and in areas where they are unlikely to be recaptured or injured by vessels.

(B) Resuscitation must be attempted on sea turtles that are comatose or inactive by:

(1) Placing the turtle on its bottom shell (plastron) so that the turtle is right side up and elevating its hindquarters at least 6 inches (15.2 cm) for a period of 4 to 24 hours. The amount of elevation depends on the size of the turtle; greater elevations are needed for larger turtles. Periodically, rock the turtle gently left to right and right to left by holding the outer edge of the shell (carapace) and lifting one side about 3 inches (7.6 cm) then alternate to the other side. Gently touch the eye and pinch the tail (reflex test) periodically to see if there is a response.

(2) Sea turtles being resuscitated must be shaded and kept damp or moist but under no circumstance be placed into a container holding water. A water-soaked towel placed over the head, carapace, and flippers is the most effective method in keeping a turtle moist.

(3) Sea turtles that revive and become active must be released over the stern of the boat only when fishing or scientific collection gear is not in use, when the engine gears are in neutral position, and in areas where they are unlikely to be recaptured or injured by vessels. Sea turtles that fail to respond to the reflex test or fail to move within 4 hours (up to 24, if possible) must be returned to the water in the same manner as that for actively moving turtles.

(4) A turtle is determined to be dead if the muscles are stiff (rigor mortis) and/or the flesh has begun to rot; otherwise, the turtle is determined to be comatose or inactive and resuscitation attempts are necessary.

Any sea turtle so taken must not be consumed, sold, landed, offloaded, transshipped, or kept below deck.”