Simulated Economic Impact of TED Regulations on Selected Vessels in the Texas Shrimp Fishery

JOY CLARK, WADE GRIFFIN, JERRY CLARK, and JAMES RICHARDSON

Introduction of a Turtle Excluder Device (TED) it was called a Turtle Excluder Device which, when placed in the shrimp trawl, (TED) (Watson et aI., 1985). Shrimp fishermen trawling in the Gulf prevents turtle mortality. However, the All sea turtles are listed as endangered ofMexico and the south Atlantic inadver­ shrimp industry has been concerned or threatened by the Endangered Species tently capture and kill sea turtles which about the possible impact of the TED Act. Underthis Actit is illegal to import, are classified as endangered species. Re­ upon vessel shrimp catch. export, take, possess, sell, or transport cent Federal legislation requires the use This analysis was designed to address endangered species without a permit this issue by evaluating the impact ofthe unless these activities are specifically Joy Clark is with the Economics Department, TED regulations upon the economic allowed by regulation (USDC, 1978; Auburn University at Montgomery, Montgomery, AL36193-0401. WadeGriffin and James Richard­ viability ofrepresentative shrimp vessels Yaffee, 1982). Fivespecies ofsea turtles son are with the Agricultural Economics Depart­ in the Texas shrimp fishery. are caught in shrimp trawls in the waters ment, Texas A&M University, CoUege Station, TX This analysis, however, does not ex­ ofthe southeast United States. They are 77843. Jerry Clark was with Texas Parks and Wildlife and Fisheries, P.O. Box 98000, Baton plicitlyconsiderthe interactiveaspects of the loggerhead, Caretta caretta; Kemp's Rouge, LA 70898-9000. Views or opinions ex­ the shrimp fishery, both among the ves­ ridley, Lepidochelys kempi; green, Che­ pressed or implied are those of the authors and do sels and between the vessel catch and loniamydas; leatherback, Dermochelys not necessarily reflect the position ofthe National Marine Fisheries Service, NOAA. remaining shrimp stock. An implicit coriacea; and hawksbill, Eretmochelys assumption ofthis analysis is that the in­ imbricata (Dean and Steinbach, 1981; l dividual vessel's fishing behavior would Anonymous ). In 1978, when the green not change as a result ofthe TED regula­ and loggerhead sea turtles were listed ABSTRACT-Shrimpfishermen trawling in tions, in response to either increases or under the Endangered Species Act (the the Gulf of Mexico and south Atlantic in­ decreases in catch per tow. Rather, this other three species were listed in earlier advertentlycaptureandkillsea turtles which analysis is based on the resultant impact rulemakings), the problem ofincidental are classifiedas endangeredspecies. Recent legislation requires the use ofa Turtle Ex­ on the catch ofthe representative vessel take ofthese species in the shrimp fishery cluderDevice (TED) which, when in place in in the Texas shrimp fishery given all the was addressed in a Final Environmental the shrimp trawl, reduces sea turtle mortal­ interactive effects. That is, after all other Impact Statement (USDC, 1978). At that ity. The impactofthe TED on shrimpproduc­ considerations, if the vessel catch has time, methods to reduce the incidental tion is notknown. This intermediate analysis changed, what is the impact on the eco­ take were not available. ofthe TED regulations using an annualfirm level simulation model indicated that the nomic viability ofthe vessel? Finally, this In 1983, NMFS began a formal pro­ average Texas shrimp vessel hada lowprob­ is an intermediate analysis ofthe impact gram to encourage voluntary adoption of ability ofbeing an economic success before ofthe TED regulations. Future analysis TED's by the shrimp industry. Through regulations were enacted. An assumption that should be directed at examining the inter­ the voluntary program, TED's werecon­ the TED regulations resulted in decreased production aggravatedthis condition andthe active effects within the fishery. structed under contract and distributed to change in Ending Net Worth andNet Present shrimpers who agreed to use them. Modi­ Value ofEnding Net Worth before and after History fication and evaluation ofthe TED con­ a TED was placed in the net was significant In 1981 the National Marine Fisheries tinued, resulting in a smaller, lighter, col­ at the 5 percent level. Service (NMFS), as the result ofan ongo­ lapsible NMFS TED, as well as other However, thedifference in the IntemalRate ofRetumfor the TED and non-TED simula­ ing research and development program, non-NMFS TED's. Despite numerous tions was not significant unless the TED introduced a shrimping gear design extension programs, publicity and train- caused a substantial change in catch. This aimed at reducing the capture of sea analysis didnotallowfor interactionsbetween turtles. This device would be sewn into I Anonymous. 1983. Environmental assessment of the fishermen in the shrimp industry, an a shrimp traw1(Fig. 1) and was designed a program to reduce the incidental take ofsea turtles assumption which couldsignificantlyalterthe by the commercial shrimp fleet in the southeast impact ofTED use on the catch andearnings to provide a way for sea turtles to exit the United State. U.S. Dep.Cornmer., NOAA,NMFS ofthe individual shrimp vessel. trawl. Because ofits proposed function, Southeast Reg. Off., 51. Petersburg, Fla., 20 p.

53(2),1991 1 Fla., prior to 1988 (Table 1) identified four TED's which satisfactorily excluded turtles from commercial shrimp nets: 1) The standard 30-inch opening and 25­ inch opening NMFS TED's, 2) the "Georgia" TED, 3) the "Cameron" TED, and4) the" Matagorda Bay" TED. In the time period subsequentto 1988, ad­ ditional devices have been approved but they are not considered as part of this analysis. The NMFS and Cameron TED's are three-dimensional, implying that a sec­ tion ofthe net must be removed to install the device. The NMFS devices are rec­ tangular in shape whereas the Cameron TED is circular (Fig. 2). TheGeorgiaand Figure I.-Position of the TED in the shrimp trawl. Matagorda TED's are two-dimensional and are sewn directly into the net. The Georgia TED has a long oval shape with parallel bars creating a barrier across the ing activities, the voluntary program was the TED March though November, and surface of the device. The Matagorda not effective (Anonymous2). As oflate Atlantic vessels from May through Aug­ TED is rectangular and also has parallel 1986, less than 3 percent of the shrimp ust. Inshore regulations have similar bars which function as a barrier to entry fleet had used or continued to use a TED seasonal requirements but all shrimp intothecod end ofthe trawl (Fig. 3). The (Oravetz3) . trawls musteither use aTED orlimittow positioning of the opening in the trawl, Specific regulations concerning use of time to 90 minutes. which allows turtles and other large TED's were developed in 1986 during organismsto escape, varies by TED. The mediation between members ofthe south­ Turtle Excluder Devices NMFS and the Matagorda devices have eastern U.S. shrimp industry and inter­ Testing conductedoffCapeCanaveral, top openings. The Georgia TED has a ested members of the environmental community. Proposed regulations were published in the March 1987 Federal Register and final regulations were pub­ Table 1.-Summary01 testing conducted todeterminecapabilities ollourTurtleExcluder lished in the July 1987 Federal Register. Devices. Sources: Text footnotes 4,5,6,8, and 10.

As summarized by the National Marine Turtles Shrimp Fisheries Service (Anonymous4), as of 1 No. caught (no.) catch (lb.) By-catch (lb.) Test and of May 1989 in offshore waters, use ofthe type of TED tows Control TED Control TED Control TED TED was to be required of all vessels Cape Canaveral tests 1 measuring 25 feet or longer. For vessels NMFSTED 10 14 0 26.00 24.00 7,488 4,164 of less than 25 feet in length, the option Cameron TED 10 21 0 26.75 26.50 4,551 3,026 Georgia TED 10 16 0 13.75 17.25 5,275 4.014 of towing 90 minutes was available. Matagorda TED 10 17 0 31.75 29.50 7,771 4,312

There are seasonal requirements by re­ North Carolina Sea Grant gion, with Canaveral and southwest NMFS TED with 45° grid angle 8 696' 291 158' 55.50 Florida vessels required to pull the TED NMFS TED with year-round, Gulfvessels required to pull 37° grid angle 10 1,393 1,513 73.10 42.65

SI. Simon's Island NMFSTED 18 158.00 175.38 Georgia TED 18 15988 156.88 2Anonymous. 1986. Report from the turtle excluder Matagorda TED 18 143.25 194.75 device workshop. U.S. Dep. Commer., NOAA, Cameron TED 18 17588 200.75 NMFS, Southeast Fish. Cent., Pascagoula, Miss., 15 p. Texas testing' 'Chuck Oravetz, National Marine Fisheries Ser­ Std. NMFS TED 49 3,073 3,127 186.8 230.3 vice, NOAA, Southeast Regional Office, St. Peters­ Mini NMFS TED 5 170 249 0 2.2 burg, Fla. Presentation at TED meetings in Georgia Jumper 14 753 793 29.6 40.1 Pascagoula, Miss., Oct. 1986. 'Data recorded for 7 tows wfTED, while 10 tows were recorded for control. 4Anonymous. 1989. Summary of TED/tow time 2Measured as number of shrimp rather than pounds of shrimp caught. regulation. U.S. Dep. Commer., NOAA, NMFS 'Measured in kg. Southeast Reg. Off., St. Petersburg, Fla. I p. 'By-catch measured in bushels.

2 Marine Fisheries Review bottom opening and the Cameron TED can be used with either a bottom or top opening. Meanwhile, TED development and use has also been proceeding in other na­ tions. In July 1986, following a publicity and coordination trip in March, a work­ shop and vessel demonstration was held in Mazatlan, Mex. In Indonesia, over 1,000TED's are in use in the western area onjoint-ventureJapanese vessels. Indo­ nesia has sent fishing gear experts for training at the NMFS Harvesting Sys­ Cameron TED tems Division, Mississippi Laboratories (USDC, 1985). Georgia TED Tests for Shrimp Exclusion NMFSTests As early as September 1983, NMFS was testing its new TED to determine the impact on shrimp catch and by-catch reduction. These tests were directed at evaluating the TED design modifications to improve finfish by-catch reduction rates. The TED used in theseexperiments was the original solid NMFS TED (Anonymous5) . NMFS continued testing and improv­ ing the TED, and additional tests were made in 1984. From July through Sep­ NMFS TED Matagorda TED tember, the FRV Jeanie conducted tests off Florida, Alabama, and Mississippi. Initially, tests were directed at perfor­ Figure 2.-General schematic of Figure 3.-General schematic of the mance ofseveral different designs ofthe Cameron TED (circular) and NMFS Georgia TED (circular) and Mata­ TED (rectangular). gorda TED (rectangular). NMFS TED. These included: Fiberglass frame and PVC joints, steel-frame col­ lapsible, aluminum frame, and minia­ ture-frameTED. PVCjoints ofthe alum­ North Carolina Test shrimp catch was measure in numbers of inum-frame TED's were found to be too individual shrimp. Large quantities of weak to withstand normal shrimping The University ofNorth Carolina Sea shrimp were not encountered, possibly operations. Grant program was also involved in test­ due to the half-hour tow limitation. This Other cruises were made to compare ing shrimp and finfish retention rates of made it difficult to measure the impact on shrimp retention rates ofvarious TED's the various TED's. In April and May shrimp catch. vs. a control net without a TED. These 1986, two cruises were made aboard the In July 1986, undertheauspicesofthe cruises were done both day and night and Carolina Coast (Table 1) (Anonymous7). University of North Carolina Sea Grant incorporated different types of finfish Ofthe two NMFS TED's tested, the one Program, tests were conducted on the deflectors (Hummer wire, types A and H with the 37° grid angle appeared to be 8). 6 "Georgia Jumper" (Anonymous A and solid bar, type D) (Anonymous ). more effective at eliminating by-catch total of24 tows were made by the Caro­ with no overall loss ofshrimp. Although lina Coast. Nineteen tows were com­ SAnonymous. 1983. Cruise Report for FRS OREGON II Cruise 137 - 9/6/83 - 9/21/83. Dep. by-catch was measured in kilograms, pared with a standard net and 5 tows were Commer., NOAA, NMFS Southeast Fish. Cent., Pascagoula, Miss., 4 p. 7 Anonymous. 1986. Cruise Report for CARO­ "Anonymous. 1986. Cruise Report for CARO­ 6Anonymous. 1984. Cruise reports for FRY LINA COAST Cruises: TED Testing- April 28-30, LINA COAST Cruise: TED Testing -July 14-18, JEANIE. U.S. Dep. Commer., NOAA, NMFS 1986. Univ. N.C. Sea Grant, N.C. Mar. Resour. 1986. Univ. N.C. Sea Grant, N.C. Mar. Resour. Southeast Fish. Cent., Pascagoula, Miss., 4 p. Cent., Atlantic Beach, N.C., 2 p. Cent., Atlantic Beach, N.C., 2 p.

53(2),1991 3 compared with the NMFS TED. Several Texas shrimp production (GrahamI2). TestsoffSt. Simon'sIsland, Ga., were adjustments were made to the TED be­ Many of the tows done in Texas were the only ones to use all four certified fore it was operating properly. When all faulty, for one reason or another, and TED's. The Georgia Jumper had a 1.9 19 tows were considered, the TED ex­ thus, rigorous statistical analyses have percent production gain, whereas the perienced an 18 percent shrimp loss, not been carried out (Table 1). The results other three TED's showed losses (NMFS however, for the 15 tows when the TED ofthis testing were useful in indicating the a 9.9 percent loss, Cameron a 12.4 per­ was operating properly there was only a learning process which must occur when cent loss, and Matagorda a 26.4 percent 9 percent shrimp loss. The tested TED a TED is introduced into the trawl. loss). However, only 18 tows were made was very effective in reducing horseshoe with each TED. crab, cannonballjellyfish, stingray, and TED Impacts The final series oftests, offthe Texas tunicate catches, with anoverall by-catch coast, are reported here. Again, the vari­ When reviewing testing that has been reduction of24percent (21 percentforthe ety ofvessels used and variation in equip­ completed on various TED's, a wide 15 tows above). Forthe 5tows againstthe ment, as well as change in location, made range of results is apparent. Tests have NMFS TED there was an overall 9 per­ accurate interpretation of this data dif­ been conducted in different areas offthe cent shrimp loss with 40 percent less by­ ficult. In addition, a report by Byrne, et. eastern and southern U. S. coasts using catch (Anonymous9). al. l4 argues that the experimental design different types of vessels and different ofthe TED testing does not allow specific TED's. Theapproach to testing is consis­ Georgia Test shrimp retention rates to be identified. tent, i.e., simultaneous drags with a con­ The Marine Extension Service of the trol net and a TED net, but the gear and Georgia Sea Grant College Programcon­ nets vary across all experiments. As a Economic Impact of TED's ducted tests off Georgia on the four result, it is not surprising to find that TED TED's tested at the Cape Canaveralchan­ capabilities varied with the experiments. Method of Analysis nel: Georgia Jumper, Cameron TED, Cape Canaveral testing certified four The intent of this analysis was to ex­ Matagorda TED, and NMFS Collapsible TED's as possessing the capability of amine the impact ofTED regulations on TED without funnel. The purpose of "kicking out" sea turtles from the trawl. the earning capacity of a single repre­ these tests was to determine shrimp ex­ Further testing conducted by the NMFS sentative vessel. The method ofanalysis clusion rates ofTED's (Anonymous 10). determined finfish reduction capability of is to simulate conditions that are repre­ A total of 72 double-rig trawls were the TED and helped in refinement ofthe sentative of a vessel in the Texas' Gulf conducted with 9 port and 9 starboard TED. Finfish reduction rates as high as shrimp fishery, using the firm-level tows conducted with each ofthe four dif­ 80 percent were recorded during the day simulation model FLEETSIM, underthe ferent TED's (Table 1). Sampling was although night reduction rates were proposed TED regulations (Clark et conducted 3-4 miles east ofthe south por­ lower. There was no significant reduc­ aI 15). FLEETS1M is a firm level, recur­ tion ofSt. Simon's Island, Ga., in 30 feet tion in shrimp catch during these tests sive, simulation model which simulates ofwater. The Georgia Jumper had a 1.9 (Anonymous13). the annual production, costs, and income percent gain compared with a standard Testing by the University of North aspects ofa fleet, by vessel, overa multi­ trawl without a TED. The NMFS TED Carolina Sea Grant program resulted in year planning horizon. had a 9.9 percentloss, the Cameron TED extreme values for average shrimp loss FLEETSIM is capable ofsimulating a a 12.4 percent loss, and the Matagorda (58 percent) over all tests. The TED was hypothetical fleet for 1-10 years. The ll TED a 26.4 percent loss (Anonymous ). mounted in the net at a 45 ° angle; subse­ model recursively simulates a typical quent North Carolina testing showed this fleet by using the ending financial posi­ Texas Test to be an inappropriate angle. In addition, tion for year one as the beginning position Tests with the four TED's presently the results were for experiments where for the second year, and so on. FLEET­ certified for use wereconducted in Texas the total number ofshrimp encountered SIM does not includean overall objective waters using both bay and Gulfvesselsto were small, and this makes any analysis function to be optimize, but rather ana­ determine the impact ofTED devices on suspect. Furthertestingwitha37° angle lyzes the outcome ofa given set ofinput for TED installation gave a shrimp in­ data and assumptions for a typical fleet. crease of8.6 percent, and more shrimp Accounting equations and identities con­ "Anonymous. 1986. Cruise Report for CARO­ LINA COAST Cruises: TED Testing - April 28-30, were consistently encountered during 1986. Univ. N.C. Sea Grant, N.C. Mar. Resour. these experiments. 14R. Byrne, W. Griffin, and J. Clark. 1987. Four Cent., Atlantic Beach, N.C., 2 p. TED's analysis of variance. Nat. Resour. Work. IOAnonymous. 1987. Preliminary results from Pap. Ser. Nat. Resour. Workgroup, Dep. Agric. shrimp retention studies on four different turtle ex­ 12G. Graham. 1986. Summary of TED cruise re­ Econ., Tex. A&M Univ., Coll. Sta., IS p. c1uderdevices offthe Georgia Coast. Univ. Ga. Sea ports #1 - #10 and individual cruise accountings. 15 J. L. Clark, J. W. Richardson, and C. J. Nixon. Grant, Mar. Ext. Serv., Brunswick, IS p. Tex. Agric. Ext. Serv., Tex. A&M Univ., ColI. 1987. Description ofFLEETS1M :A general firm "Anonymous. 1987. Preliminary results from Sta., Tex., II p. level policy simulation model fora shrimp fleet. Nat. shrimp retention studies on four different turtle ex­ I3Anonymous. 1984. Cruise reports for FRV Resour. Work. Pap. Ser. Nat. Resour. Workgroup, c1uder devices offthe Georgia Coast. Univ. Ga. Sea JEANIE. U.S. Dep. Commer., NOAA, NMFS Dep. Agric. Econ., Tex. A&M Univ., Coll. Sta., Grant, Mar. Ext. Serv., Brunswick, 15 p. Southeast Fish. Cent., Pascagoula, Miss., 4 p. SOp.

4 Marine Fisheries Review stitute most of the computational com­ changes in numberand size ofvessels for historical data in much the same manner ponents of the model. the period 1969-86. Vessel size and con­ as those for landings and prices. struction varied considerably from 20- to Procedure 30-foot wood up to 90-foot steel ones. Assumptions Three separate simulations were con­ Some firms had vessels that were 20-30 The impactofthe TED on the produc­ ducted in an effort to analyze the impact years old, while others culled vessels tion capability ofthe vessel is potential­ of the TED regulations. These three after 10 years. ly manyfold. It is expected that initially models were: 1) Historical, 2) Baseline, Baseline and TED simulation analyses there will be a negative impact while the and 3) TED Simulations. were stochastic, as opposed to determin­ fisherman learns how to use the device. The first FLEETS1M analysis exam­ istic. In Baseline and TED policy anal­ This"learning period" will vary by ves­ ined the historical situation in the Gulfin­ yses, pounds landed by the vessel were sel and captain and, therefore, a measure­ dustryfrom 1978to 1986. No analysis is stochastically set about a mean value, ment of its impact is difficult. included for bay vessels on the assump­ based on a 5 year (1982-86) average of Contributing to this measurement tion that bay shrimpers will take advan­ fishery landings. Prices were similarly problem is the range ofdifficulty ofuse tage ofthe 90-minute tow time exception set but included an adjustment based on across various TED devices. The Geor­ and will not be required to use TED's. changes in the Consumer PriceIndex. No gia Jumper and Matagorda TED's are This analysis was deterministic and was attempt wasmade to accountfor any price relatively simple devices, whereas the conducted to obtain a starting financial changes which might occur ifthe use of NMFS and Cameron TED's are bulky position for the baseline simulation. The a TED leads to a decline in total produc­ and more difficult to work with. Many program was run using actual values for tion of shrimp, Gulfwide. Price would vessel captains already employ sometype changes in inflation, interest rates, land­ likely increase if production declined, ofexcluderdevice (' 'jellyballshooters' ') ings and prices, and production costs for especially for large shrimp which are the during certain portions ofthe year. This a representative vessel. In addition, the mainstay of the Gulf shrimp fleet ana­ experience should make efficient use of model began with a new vessel in 1978 lyzed here. a TED more likely. In this analysis, and was run withoutconsidering income Inaneffortto make stochastic variables because of the inability to establish a tax. Therefore, all values generated rep­ a function ofactual relationships in the in­ reasonable estimate of the "learning resent a before-tax situation. dustry, ananalysis ofhistorical priceand period" impact of adopting a TED, no The second FLEETSIM analysis in­ landing trends for Gulf shrimp vessels impact is included. volved the development of a baseline was conducted. To generate random Another point is the widespread per­ simulation modelforthefishery. Thefirst landings and prices, it was necessary to ception that some ofthe TED'sare bulky yearofanalysis for this model was 1987. provide cumulative deviations around and, therefore, unsafe to handle and may Thepurpose ofthe baseline model was to mean values for these variables. This was have an effect on insurance rates. Con­ predict what would occur in the industry accomplished by regressing average an­ versely, there may be positive benefits without TED regulations. The result of nuallandings and prices against time and associated with the use of a TED. Im­ the baselinesimulations was laterused to taking resulting deviations and dividing proved fishing efficiency and fewer safe­ determine the impactofthe TED regula­ them by the value for mean landings ty problems which arise from heavy by­ tions. Thethird step, following the devel­ and prices. An additional explanation of catch, may result in favorable changes in opment ofthe baseline simulation model this procedure can be found in the insurance rates. No analysis is incor­ andestimates offuture costs and returns FLEETSIM user manual. porated in this study because ofan inabil­ for the representative vessel, was to Values for future interest rates and in­ ity to establish a specific impact of the build effects ofTED regulations into the flations rates, associated with prices and TED on insurance rates. model. production costs, were obtained from For similar reasons, issues dealing COMGEM17, a macroeconomic simula­ with possible improved shrimping effi­ Data tion model developed by Penson, ciency associated with use oftheTEDare Production and budget information for Hughes, and Romain. The "best" pre­ not dealt with in this analysis. This im­ Gulfshrimperswasobtained from inter­ dictions available were annual predic­ proved efficiency centers around possi­ views conducted with members of the tions of inflation rates approximating ble higher percentage of shrimp in the Texas Gulf shrimp industry (Anony­ 4 percent for 1987 through 1990 and, catch when a TED is used, resulting in mous16). This information included past thereafter a constant of 4 percent. The longer tow times, improved shrimp qual­ production, costs and returns, and cumulative deviations of these macro­ ity, and reduced sorting time. exogenous variables, i.e., interest rates A final issue is that research to date has 16Anonymous. 1987. Regulatory impact review and and fuel prices, were generated based on been limited to tow-by-tow comparisons regulatory flexibility analysis for regulations which for one vessel and do not represent the require the use of turtle excluder devices by situation when all vessels will pull a TED. shrimpers to conserve sea turtles. U.S. Dep. Com­ '7Mention oftrade names or commercial firms does oc~an mer., NOAA, NMFS Southeast Reg. Off., SI. not imply endorsement by the National Marine For example, ifan area of bottom Petersburg, Fla., 25 p. Fisheries Service, NOAA. contains, at anyone time, a fixed amount

53(2),1991 5 ofshrimp, severalpasses through the area periments where more than a few shrimp aside for the purchase of a TED 2 years might catch a fixed proportion of avail­ are encountered and technical difficulties in the future. able shrimp. Ifeach net is catching less are absent. Therefore, four scenarios are shrimp on the first pass (because of the reported in this research: No change, a 5 Method ofEvaluation TED), subsequent passes will be associ­ percentdecrease, a 10 percentdecrease, Variables used in evaluating the impact ated with larger remaining population and a 5 percent increase in shrimp pro­ ofTED regulations on the representative level because less shrimp were taken on duction associated with the use ofa TED. vessel were ending net worth in year 10, the previous tow. Thus, although catch This range encompasses the mean reten­ internal rate of return of the analysis, per unit ofeffort might be reduced for a tion of the most efficient TED's. present value ofending net worth, equity­ given tow because ofthe TED, the catch to-asset ratio and the probability net pres­ per unit of effort would not decrease as TED Costs ent value will be greater than zero. Net much as suggested by the sample data worth was determined by subtracting At the time ofthis study there were five because more population remains for the total liabilities from total assets. Ending certified TED's: the NMFS, Georgia next tow. The argument is perhaps best net worth reflected owner's equity in the Jumper, Cameron, Matagorda, and a seen in reverse. Some ofthe research in­ vessel and in other personal property at "soft" TED. This paper was essentially dicates that the use of a TED increases the end of the planning horizon. The complete before the soft TED was cer­ shrimp catch, yet it is not sensible to internal rate ofreturn is often referred to tified, and therefore, no economic anal­ believe that the total shrimp catch Gulf­ as "marginal efficiency ofcapital. " By ysis ofthis particulardevice is included. wide could increase annually by 5percent definition, internal rate of return is the Highest estimates of acquisition costs if TED's are employed. The impact of discount rate that equates present value were for the NMFS TED, with quotes TED's on the industry cannot be extrap­ ofbenefits with the present value ofcosts. between $375 and $475, whereas price olated from an isolated vessel pulling a An investment is selected as long as estimates for the Georgia and the TED, which has been the research to internal rate of return exceeds cost of Cameron TED's ranged between $150 date. capital. and $250 (Clark and GriffinI8). For pur­ The studies that have been conducted The present value technique puts the poses of this analysis, each TED is as­ to date do not focus on the above issues net worth at the end of the planning sumed to cost $300 (a number between and, therefore, data are limited. Although horizon in real dollars. The equity-to­ the lesser cost TED's and the NMFS these issues are recognized as being im­ asset ratio is onemeasure ofsolvency. A TED). portant, they are not considered in this one-to-one ratio means a vessel/boat Total cost of using a TED varied de­ analysis. ownerdid not have any debt. A ratio less pending on type of vessel and type of Itwas assumed that the vessel was pur­ than one-to-one would indicate abusiness TED. In the Gulfitis possible to see both chased new at the beginning ofthe histor­ had not paid off all debt owed on assets double-trawl and twin-trawl rigs, indi­ ical simulation (1978). The vessel was of (Osburn and Schneeberger, 1978). cating between three and six TED's could steel construction and 73 feet in length. Another issue when examining eco­ be required for a Gulfvessel. These num­ In purchasing the vessel, it was assumed nomic viability ofthe vessel is the prob­ bers include a spare TED for each two that 50percent ofthe purchase price was ability that net present value ofa stream TED's used. The expected life of the paid down and the rest of the purchase of income during the period of analysis TED is assumed to be 2 years. price was financed over a 10 year period was greater than zero. The discount rate The model was run for a double-trawl at 9 percent per annum. used for calculating netpresentvalue was rig which required an investment in three To examine the impactofthe TED reg­ set at 7 percent. It was not unreasonable TED's at a cost of $300 per TED. This ulations, assumptions had to be made toexpectthis rate ofreturn onalternative resulted in a total purchase cost for the about the impact ofthe TED on produc­ investments outside the fleet. These five Gulf vessel of $900 or an annual cost of tion capabilities ofthe vessels and cost of variables are obtained from stochastic $450. various TED devices. These assumptions simulations for thebaselineand four TED An additional annual maintenancecost were based on production impactanalysis scenarios. of $50 was assumed, resulting in an an­ presented earlieras well as cost informa­ nual TED cost of $500. An accounting tion obtained for the four TED devices. Discussion of Simulation Results technique, which entered the TED cost on Considerable differences exist in the the cost side ofthe model and reduced ini­ studies that have been conducted about Deterministic Simulation 1978-1986 tial cash by one-halfthe purchase price of the impact oftheTED on shrimp produc­ A representative Gulfvessel was simu­ the TED, forced the model to put cash tion. Itis assumed that shrimpers will use lated for 9 years. An outstanding balance those TED's that are mosteffectiveat re­ on the vessel of $120,000 was financed taining shrimp. The mean shrimp reten­ 18J. L. Clark and W. L. Griffin. 1987. Update of during 1°years at a rate of9 percent per tion, however, has generally ranged from costs and returns for seven Texas shrimp vessels. annum. Beginning cash reserve was set Nat. Resour. Work. Pap. Ser. Nat. Resour. Work­ a small increase to about a 1°percent group, Dep. Agric. Econ., Tex. A&M Univ., ColI. at $26,000 and resulting beginning net decrease when using a TED in those ex­ Sta., 2 p. worth (market value) was $117,692.

6 Marine Fisheries Review Table 2.-Annual produclion cosls Table 3.-Comparison of output variables althe end of a 1O-yearperiod across baseline and TED policy simula· lor a represenlalive Gull shrimp lions lor Ihe Gull 01 Mexico, vessel (73 leel, sleel hull), Percent impact PVo! Cost on shrimp Ending Internal rate ending Equity! Probability (season Simulation production net worth of return net worth assets NPV>O Item total) Baseline $896,801 0.0372 $455,888 0.945 0,78 Ice $ 3,106 Scenario 1 0 886,062 0.0361 450,429 0.945 0,78 Fuel 23,271 Scenario 2 -5 781,577 0.0231 397,314 0.936 0.74 Repair and replacement 12,746 Scenario 3 -10 680,316 0.0091 345,838 0929 0.58 Other 14,063 Scenario 4 +5 998,247 0.0500 507,458 0.960 0,82 Packing $0.09/1b. Dock rental 663 Insurance 7,738

line simulation was $898,801 (Table 3), plished by a statistical comparison ofthe Although the value for average internal means. The hypothesis tested was These starting val ues are required by the rate of return is low (3.72 percent), the model, but will not affect the comparative use ofa representative vessel in the sim­ analysis of using a TED. ulation model contributed to this low val­ Annual values for those costs which ue. The practice by many Gulfshrimpers with an alternative hypothesis of varied with level ofproduction are pre­ ofspreading the cost ofrunning a vessel sented in Table 2. In addition to these across many different operations, allows costs, dock space was set at an annual the vessel to be run less efficiently and still rentalof$663. Fixed costs, those which remain solvent; hence, alow internal rate The test statistic used are setatthe beginning ofthe year and do ofreturn for the vessel. Many vessel op­ not vary during the period under con­ erators, however, are very efficient and sideration, included only vessel insur­ generated much higher rates of return. ance at $7,738. Other costs, such as Average present value of ending net depreciation and interest, werecalculated worth was $455,888 and average equity by FLEETSIM. to asset ratio was 0.945, indicating the By simulating the model for the nine operator had a low debt level. The base­ which is an approximation ofthe Student year period, 1978-86, the ending finan­ line simulation model had a 78 percent t statistic, and can be used when sample cial position for the vessel was obtained. chance ofgenerating a net present value sizes are sufficiently large (DeGroot, The vessel ends the 9 years with $122,469 greater than zero (with a discount rate of 1975). The sample sizes, nl and n2, are in cash on hand and vessel assets worth 7 percent). equal to 50. This test statistic allows $211,692, giving total asset value of determination ofsignificant differences $334,161. The only liability associated Simulation with TED from the baseline results but does not with the vessel was an intermediate-term Four different simulations were run to allow any comparisons to be made across debt of$17 ,155. Net present value for 9 examine impacts ofTED regulations on all simulations. Since the primary pur­ years was $54,739. Internal rate ofreturn atypical Gulfvessel. Again, 50 iterations pose ofthis analysis was to determine if was 8 percent. of each simulation allowed average use ofthe TED resulted in a change in the values for the statistics of interest to economic well-being ofthe vessel, it was Baseline Simulation be generated. These different scenarios felt this test statistic would be sufficient. The baseline simulation was set up to were run assuming no impact on shrimp The computedz-values for the various simulate what would occur to the Gulf catch (Scenario 1), a 5 percent decrease economic indicators are presented in vessel during 1987-97. This was accom­ in catch (Scenario 2), a 10 percent de­ Table 4. Each row examines the com­ plished withoutconsideringthe impactof crease in catch (Scenario 3), and finally puted z-value for a TED scenario against TED regulations and results were later a 5 percentincrease incatch (Scenario 4). the baseline scenario. Those test statistics used to compare against those analyses These results were as expected in that all which were significant at the 95 percent where the TED policy was simulated. economic indicators declined for nega­ level are marked as footnote 2. It is ap­ The baseline simulation was run for 50 tive impacts on shrimp productionand in­ parent that all of the TED Scenarios different iterations which allowed aver­ creased for the positive impact on shrimp where there is a change in the level of age values to be generated for the statis­ production (Table 3). catch will cause a significant change in tics of interest. The next step in the analysis of the both Ending NetWorth in Year 10 as well The vessel, purchased in 1978 had a simulation results was to determine ifthe as Present Value of Ending Net Worth. market value in 1987 of $211 ,692 and a decrease or increase in the economic in­ This is important because it indicates that replacement value was $400,000. Aver­ dicators was a significantchange from the even the discounted Net Worth is signif­ age ending net worth in year 10 for base­ baseline simulation. This was accom- icantly different. However, Internal Rate

53(2),1991 7 Table 4.-Calculated z-valuesforoutput variablesused indetermining ifTED scenario of Return (IRR) is not significantly dif­ values are significantly different from baseline values. ferent from the baseline until Scenario 3. Percent impact Ending Internal PVol In Scenario 2, where the impact on the on shrimp net rate 01 ending Equityl shrimp catch is a loss of 5 percent, the Simulation production worth return net worth assets calculated z-value is significant at the 80 Scenario 1 0 1.517' 0.100 0.48 0.01 Scenario 2 -5 19.433' 1.365' 6.146' 0.281 percent level. There is no significant dif­ Scenario 3 -10 18.289' 2.797' 12.168' 0.774 ference in the values forthe Equity/Asset Scenario 4 +5 -15.155' -1.18 -4.792' 0.645 ratio between the baseline and any ofthe 'Indicates difference is significant at the 20 percent level. TED scenarios. 'Indicates difference is significant at the 5 percent level. What this suggests is that the impact of the TED is significant ifthe vessel owner is primarily interested in vessel earnings. However, ifalternativeenterprises are to Scenario 3 (10 percent decline in catch) bined effect leads to shrimp loss to all beconsidered then there is no appreciable were Ending Net Worth in Year 10, IRR, vessels, then further statistically valid difference in the rate ofreturn associated and Present Value ofEnding Net Worth side-by-side tests ofa TED orTED's on with the decision to operate a shrimp boat all significantly different from the Base­ shrimp retention need to be done. The or an alternative enterprise until the im­ line Scenario at a 5 percent level. Re­ otherissues ofshrimp quality, safety and pact upon the shrimp catch is very large. search to date, however, indicates that deck handling procedures could possibly It is also notable that the impact upon there are TED's, specifically the Georgia be included in these further tests. How­ shrimp catch and therefore earnings, TED, which have experimental results ever, ifthere is no significantshrimploss does not result in significant increases in consistently better than a 10 percent to all vessels when considering the above debt levels in the TED scenarios. Because shrimp loss. In fact, the Georgia TED (or combinedeffects, then further expensive ofthe way the FLEETSIM modelgener­ Georgia Jumper) has increased shrimp side-by-side retention tests of TED's is ates Probability of Net Present Value retention in all experiments reported not warranted. greater than Zero, it was not possible to here. If there is no effect upon shrimp use the above test to determine significant landings as a result of pulling the TED, Acknowledgment differences from the Baseline scenario. the only economic effect will be the cost ofpurchasing and maintaining the TED. The authors gratefully acknowledge Summary, Discussion, These costs will have only a very minor the assistance of the National Marine and Conclusion impact on the shrimp industry. Fisheries Service, financial assistance This analysis is an intermediate anal­ It is important to note that no research award#NA86WC-8-06129 (MARFIN). ysis ofthe impactoftheTED regulations. has been undertaken on other impacts a Itdoes not explicitly consider the interac­ TED may have on shrimping operations. Literature Cited tive aspects of the shrimp fishery both For example, in addition to the impact Dean, R., andD. W. Steinbach. 1981. Endangered among vessels and between vessel catch which by-catch reduction may have on marine turtles ofthe Gulfcoast. Tex. Agric. Ext. Serv., Tech. Rep. L-1867, 5 p. and the remaining shrimp stock. This shrimp catch, the TED may also impact DeGroot, M. 1975. Probability and statistics. analysis is based on the resultant impact quality, onboard safety, and onboard Addison-Wesley Press, Reading, Mass., ch. 8. on the catch ofa representative vessel in handling ofgear and shrimp fleet catches Osburn, D. D., and K. C. Schneeberger. 1978. Modern agriculture management. Reston Publ. the Texas shrimp fishery given all inter­ when all vessels are pulling a TED. None Co., Inc., Reston, Va., p. 81-113. active effects. of these potentially significant impacts Penson, J.P. Jr., D. W. Hughes, and R. F. J. Four scenarios were analyzed for the have been studied here or elsewhere. Romain. 1984. An overview of COMGEM: A macroeconomic model emphasizing agriculture. effect ofthe TED on shrimp production. Asdiscussed, no conclusive statements Tex. Agric. Exp. Sta., Dep. ofAgric. Econ. In­ These were, no change, a 5 percent de­ could be made about the impact of the form. Rep., DIR 84-1, DP-12, 25 p. USDC. 1978. Final Environmental Impact State­ crease, a 10 percent decrease and a 5 TED upon shrimp retention using the ex­ menton Listing and Protecting the Green Sea Tur­ percent increase. The results were as ex­ isting data. A first step in the analysis of tle, Loggerhead Sea Turtle, and Pacific Ridley pected in that all economic indicators the impact of TED regulations on the Sea Turtle Underthe Endangered SpeciesAct of 1973. U.S. Dep. Commer., NOAA, Natl. Mar. declined for negative impacts on shrimp shrimp industry should be to analyze the Fish. Serv., St. Petersburg, FLa., 25 p. production and increased for the positive combined effects of shrimp gain or loss _-,----,---_ . 1985. Annual report FY85 - trawling impact on shrimp production. A statis­ by an individual vessel with biomass efficiency device project. U.S. Dep. Commer., NOAA, Natl. Mar. Fish. Serv., Southest Reg. tical analysis ofthe economic indicators changes when all vessels in the industry Off., St. Petersburg, Fla., 10 p. pointed outthat in Scenario2 and 4, both use a TED and the individual vessel im­ Watson, J. W., J. F. Mitchell, and A. K. Shah. ofwhich represented a 5 percent impact pacts from reduced by-catch. These com­ 1985. Trawling Efficiency Device, a new con­ ceptfor selective shrimp trawling gear. U.S. Dep. upon shrimp catch, Ending Net Worth binedeffectsentail lookingatthe impacts, Commer., NOAA, Natl. Mar. Fish. Serv., and Present Value ofEnding Net Worth in the limit, ofshrimp gain or loss when Southeast Fish. Cent., Pascagoula, Miss. Yaffee, S. L. 1982. Prohibitive policy: Implement­ were significantly different from the tows are successively applied to a fixed ing the Federal Endangered Species Act. MIT baseline at a 5 percent level. Only in biomass ofshrimp overtime. Ifthe com­ Press, Cambridge, Mass.

8 Marine Fisheries Review