Gulf Research Reports

Transcription

1 Gulf Research Reports Volume 7 ssue 1 January 1981 Trens in Ex-Vessel Value an Size Composition of Reporte May - August Catches of an from the Texas, Louisiana, Mississippi, an Alabama Coasts, Charles Wax Caillouet National Marine Fisheries Service, Galveston Dennis Brian Koi National Marine Fisheries Service, Galveston DO: /grr Follow this an aitional works at: Part of the Marine Biology Commons Recommene Citation Caillouet, C. W. an D. B. Koi Trens in Ex-Vessel Value an Size Composition of Reporte May - August Catches of Brown Shrimp an from the Texas, Louisiana, Mississippi, an Alabama Coasts, Gulf Research Reports 7 (1): Retrieve from This Article is brought to you for free an open access by The Aquila Digital Community. t has been accepte for inclusion in Gulf an Caribbean Research by an authorize eitor of The Aquila Digital Community. For more information, please contact Joshua.Cromwell@usm.eu.

2 GulfResearch Reports, Vol., No. 1, 59-70, TRENDS N EX-VESSEL VALUE AND SZE COMPOSTON OF REPORTED MAY-AUGUST CATCHES OF BROWN SHRMP AND WHTE SHRMP FROM THE TEXAS, LOUSANA, MSSSSrPPl, AND ALABAMA COASTS, ' CHARLES WAX CALLOUET AND DENNS BRAN KO1 National Marine Fisheries Service, Galveston Laboratory Galveston, Texas ABSTRACT Exponential moels were use to characterize (1) ex-vessel value (in ollars) per shrimp by size category (count; Le., number of shrimp per poun, heas off?; (2) size composition (expresse as cumulative weight of the catch in pouns, heas off, by size category); an (3) ex-vessel value composition (expresse as cumulative ex-vessel value, in ollars, of the catch by size category) for reporte May-- August catches (inshore an offshore combine) of brown shrimp (Penaeus aztecus) an white shrimp (P. setiferus) from the Texas, Louisiana, Mississippi, an Alabama coasts (statistical areas 10-21) from 1960 to Exponents of the moels were use as inices to investigate trens in ex-vessel value per shrimp, sue composition, an ex-vessel value composition of the May-August catches uring this perio. This approach to analysis of catch statistics can be use to monitor these fisheries, an the results can be compare with changes that may be brought about by the closure of the fishery conservation zone off Texas, as propose by 1981 by the Gulf of Mexico Fishery Management Council, in the fishery management plan for the shrimp fishery of the Gulf of Mexico. NTRODUCTON The fishery management plan for the shrimp fishery of the Gulf of Mexico, prepare by the Gulf of Mexico Fishery Management Council (GMFMC 1980), propose a simultaneous closing of the territorial sea of the State of Texas an the ajacent fishery conservation zone (FCZ) to shrimping uring the time of the year when brown shrimp (Penaeus aztecus) in these waters are, for the most part, smaller than 65count (refers throughout this paper to number of shrimp per poun, heas remove). The territorial sea is the area uner state jurisiction extening from the coastal baseline to 9 nautical miles off Texas (Figure 1). The FCZ is the area uner feeral jurisiction beginning at the outer limit of Texas' territorial sea an extening 200 miles from shore. The closing of Texas' territorial sea to shrimping normally begins June 1 an extens to July 15. However, a S-ay flexibility in the closing an opening ates is allowe to accommoate effects of climatic variations on shrimp growth, within the restriction that the perio of closure oes not excee 60 ays. The inclusive ates for the closure in 1981 were May 22-July 15. The management plan encourage the State of Texas to continue its seasonal closure of the territorial sea, to eliminate minimum size restrictions on shrimp caught in open waters before an after the closure, an to evaluate the effect of allowing white shrimp (P. setifents) fishing to continue within the close areas uring the closure. Rationale for the propose closure was an expecte increase in yiel from aitional growth of the protecte brown shrimp, an from the elimination of waste ue to iscaring of unersize brown shrimp in the FCZ 'Contribution No G from the Southeast Fisheries Center, Galveston Laboratory, National Marine Fisheries Service, NOAA. Manuscript receive March 24, 198l;accepte June 15,1981. (GMFMC 1980). The management plan recognize that the closure might affect other fishing areas (e.g., the coasts of Louisiana, Mississippi, an Alabama) by shifting fishing effort to those areas. Therefore, it was the intent of the management plan that the biological, ecological, social an economic impacts of the closure be monitore in 1981 so that revisions coul be mae if warrante. 1JL e.7 7' Y L W 59 -=.,,.' N OFU U T Y u.0 Figure 1. Bounaries of statistical areas 10-21, the Texas territorial sea, an the fisheries conservation zone off Texas (base on information from GMFMC 1980). As might be expecte, the propose closure of the FCZ off Texas has become a highly controversial issue. There is consierable interest an concern on the part of the fishing inustry, the Gulf states, the GMFMC, the National Marine Fisheries Service (NMFS), an fishery scientists regaring the potential impacts of the propose closure. We expect that the reistribution of fishing effort, the changes in fishing strategy, an the aitional shrimp growth that may result from the closure will cause changes in size composition of the combine inshore an offshore catch.

3 60 CALLOUET AND KO1 nshore waters generally are consiere to be lanwar of the barrier islans, an are represente by bays or estuaries. Offshore waters are seawar of the barrier islans. Accoring to Henerson (1972) an Ricker (1975), an increase in average size of iniviuals in the catch coul inicate a ecrease in mortality (usually equate with a ecrease in fishing mortality) or an increase in growth (e.g., if recruitment were poor, an if population ensity were low as a consequence). A ecrease in average size might be brought about after the closure by retention an laning of large quantities of small shrimp, previously iscare at sea. Also a ecrease in average size might be cause by an intensification of fishing in offshore an inshore waters open to shrimping in other areas uring the closure. Socioeconomic factors leaing to changes in strategies of fishing, culling of the catch, an marketing of the lanings also coul influence size composition of the catch. Caillouet et al. (1980) evelope a simple exponential moel to characterize the size composition (expresse as cumulative percentage of weight of catch by size category) of annual catches of shrimp. They showe that the size of brown an white shrimp in the reporte annual catches from Texas an Louisiana ecrease from 1959 to Caillouet an Koi (1980) moifie the moel by applying it to cumulative weight by size category instea of cumulative percentage of weight by size category, an use it to investigate trens in size composition of the annual lanings of brown, pink (P. uoramm), an white shrimp from the Gulf an southeast coast fisheries of the Unite States from 1961 to Caillouet an Koi (1980) also use exponential moels to investigate trens in ex-vessel value per shrimp by size category, size composition, an ex-vessel value composition of these annual lanings. Using the methos of Caillouet an Koi (1980), Caillouet an Koi (1981) investigate trens in ex-vessel value per shrimp by size category, size composition, an ex-vessel value composition of reporte annual catches of pink shrimp from the Tortugs fishery off south Floria from 1960 to The effect of shrimp size on the ex-vessel value of the catch has also been recognize by Neal (1967), Griffin et al. (1974), Griffin an Nichols (1976), an Griffin et al. (1976). The NMFS has the responsibility for monitoring impacts of closing the FCZ off Texas. The purposes of this paper are to propose a proceure for monitoring the brown an white shrimp fisheries of Texas, Louisiana, Mississippi, an Alabama, base on the methos of Caillouet an Koi (1980), an to use these methos to investigate trens in ex-vessel value per shrimp by size category, size composition an ex-vessel value composition of the reporte May-August catches from 1960 to This approach can then be use as one means of assessing the impacts of closing the FCZ off Texas in The perio May-August was chosen for these analyses to assure that the perio of closure of Texas territorial sea an the FCZ woul be encompasse, consiering the allowe flexibility in the starting an ening ates for the closure. ncluing May an August in the time interval of coverage for the years will assure that some catch statistics will be available from the Texas coast for future comparison with those from Louisiana, Mississippi, an Alabama for the May-August perioin DESCRPTON OF DATA Summaries of the May-August catches of brown an white shrimp an their ex-vessel value were compile from ata files available from the NMFS, Southeast Fisheries Center (SEFC) Technical an nformation Management Services (TMS), Miami, Floria. The combine weight of the reporte May-August catches (inshore an offshore combine) was expresse in pouns (heas off) an the exvessel value in ollars, by year ( ); coastal area (statistical areas 10-12, 13-17, an 18-21, Figure 1); species (brown an white shrimp); an size category (< 15, 15-20, 21-25,26-30,31-40,41-50, 51-67, an 2 68 count, an pieces, representing parts of shrimp tails that coul not be assigne to a count category). Comparable ata for the years 1979 through 1981 were not available at the time of this writing. The three coastal areas are efine as (1) Texas coast (statistical areas combine); (2) Mississippi River to Texas (statistical areas combine), representing that part of the Louisiana coast west of the Mississippi River; an (3) Pensacola to the Mississippi River (statistical areas combine), representing that part of the Louisiana coast east of the Mississippi River, the Mississippi coast, the Alabama coast, an a small part of the upper coast of Floria (catches from Pensacola Bay are not inclue in this area; they are allocate to the ajacent Apalachicola area by TMS). Note that part of statistical area 17 is inclue in the area that was close in 1981 (Figure 1). Therefore, for the years 1960 to 1978, the May-August catch statistics for the Mississippi River to Texas coastal area represent a somewhat larger zone open to shrimping than was the case in 1981, as a result of the closure. This shoul be consiere in any future analyses applying our methos to ata for the Mississippi River to Texas coastal area. English rather than metric units are use throughout our paper because they have been use historically, an information woul have been lost in their conversion to metric units. Catches use herein represent those portions of the actual catches that were lane by omestic commercial fishermen at omestic ports an reporte by the National Marine Fisheries Service or its preecessor, the Bureau of Commercial Fisheries, U.S. Fish an Willife Service. May-August Catches by Year ANALYSES AND RESULTS The general trens in reporte May-August catches, an

4 VALUE AND SZE OF BROWN AND WHTE SHRMP 61 their ex-vessel value for both species an the three coastal areas are shown in Figures 2 through 7. n each coastal area, the catch of brown shrimp exceee that of white shrimp. The general trens in catch were upwar, except for white shrimp from Pensacola to the Mississippi River (Figure 7) for which the tren was ownwar. n all cases, the general tren in ex-vessel value of the catch was upwar, but this was not ajuste to account for inflation. May-August Ex-vessel Value per Shrimp by Size Cotegory We calculate the May-August average ex-vessel value per shrimp, V, by size category, C, for each year, accoring to the methos of Caillouet an Koi (1 980, 198 l), to obtain the following exponential moel which escribe the relationship between Vi an Ci for each species, coastal area, an year: qi = a (exp bci) (1) where Vi = May-August average ex-vessel value per shrimp for the ith size category; Ci = lower limit (count) of the ith size category(c1 = 15,C2 =21,C3=26,C4 =31& =41, C6 = 51, an C7 = 68); an i = 1, 2,..., 7. The logarithmic form of moel 1 was use to estimate parameters a an b by linear regression (Tables 1 through 3). The very high coefficients of etermination, r2, inicate that the straight lines fitte the ata very well. All slopes, b, were negative,showing that the value per TEXAS COAST BROWN SMRHP t DOLLARS / \ / '. 501 MSSSSPP RVER TO TEXAS BROWN SHRMP POUNDS = 20 ' S YEAR Figure 2. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May-August catches (inshore an offshore combine) of brown shrimp from the Texas coast (statistical areas combine), TEXAS COAST UHTE SHRMP 7 l l YEAR Figure 4. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May-August catches (inshore an offshore combine) of brown shrimp from the Mississippi River to Texas (statistical areas combine), MSSSSPP RVER TO TEXAS UHTE SHRMP i YEAR Figure 3. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May-August catches (inshore an offshore combine) of white shrimp from the Texas coast (statistical areas combine), YEAR Figure 5. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May-August catches (inshore an offshore combine) of white shrimp from the Mississippi River to Texas (statistical areas combine),

5 62 CALLOUET AND KO1 PENSACOLA TO THE MSSSSPP RVER BROWN SHRMP PENSACOLA TO THE MSSSSPP RVER WHTE SHRMP POUNDS 10 DOLLARS c-4 / t i 0, [,, YEAR Figure 6. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May-August catches (inshore an offshore combine) of brown shrimp from Pensacola to the Mississippi River (statistical areas combine), YEAR Figure 7. Weight (millions of pouns, heas off) an ex-vessel value (millions of ollars) of reporte May- August catches (inshore an offshore combine) of white shrimp from Pensacola to the Mississippi River (statistical areas combine), TABLE 1. Relationship between transforme ex-vessel value (ollars) per shrimp, nv, an count, C, for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast (statistical areas combine), * TABLE 2. Relationship between transforme ex-vessel value (ollars) per shrimp, nv, an count, C, for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from the Mississippi River to Texas (statistical areas combine), * Year a b r2 a b r2 Year a b r2 a b r OS *Base on the linear regression of 1nV on C, where V = May-August average ex-vessel value per shrimp in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(a) = intercept, b = slope, an r2 = coefficient of etermination; all slopes, b, were significantly ifferent from 0 at the 99% level of confience, an the high r2 values inicate a very goo fit of the straight lines to the ata points. *Base on the linear regression of 1nV on C, where V = May-August average ex-vessel value per shrimp in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(a) = intercept, b = slope, an rz = coefficient of etermination; all slopes, b, were significantly ifferent from 0 at the 99% level of confience, an the high r2 values inicate a very goo fit of the straight lines to the ata points.

6 TABLE 3. Relationship between transforme ex-vessel value (ollars) per shrimp, lnv, an count, C, for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from Pensacola to the Mississippi River (statistical areas combine), * VALUE AND SZE OF BROWN AND WHTE SHRMP 63 Year a b 2 a b r t t *Base on the linear regression of 1nV on C, where V = May-August average ex-vessel value per shrimp in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(a) = intercept, b = slope, an r2 = coefficient of etermination; all slopes, b, except one, were significantly ifferent from 0 at the 99% level of confience, an the high values inicate a very goo fit of the straight lines to the ata points. tthe slope, b, for white shrimp in 1972 i not iffer significantly from 0 at the 95% level of confience, an the r2 value was very low, because no catch was reporte for the > 68 count category. shrimp ecrease with increase in count (ecrease in size), as expecte. Lower limits rather than mipoints or upper limits of the seven size categories were use in constructing moel 1, as in Caillouet an Koi (1980, 1981). The < 15 category represente < 3% of the May-August catches of brown shrimp in each of the three coastal areas in any given year. However, for white shrimp, the < 15 category represente as high as 23% of the May-August catches from the Texas coast, 15% from the Mississippi her to Texas, an 28% from Pensacola to the Mississippi River in certain years. We i not inclue the < 15 size category in moel 1 to be consistent with previous work, an because the logarithmic form of moel 1 is not a straight line in the region of < 15 count (Caillouet ankoi 1980, 1981; Caillouet et al. 1980), The category pieces was exclue from the moel because it represente parts of shrimp tails which coul not be assigne to a count category. The constant, a, reflecte the elevation of the straight line which was influence in part by our use of lower limits of size categories an exclusion of the < 15 size category in fitting the moel. The slope, b, of the straight line is a simple inex of the ex-vessel price sprea among the size categories of shrimp, i.e., it is an inex of ex-vessel price structure. There were significant ownwar trens in b for brown shrimp in all three coastal areas, an for white shrimp in all coastal areas except the Texas coast from 1960 to 1978 (Table 4). For white shrimp from the Texas coast, the general tren was ownwar, but it was not statistically significant. The ownwar trens inicate that the May- August ex-vessel price sprea among the size categories Qf shrimp increase from 1960 to Whitaker (1973) also observe an increase in price sprea between large an small southern shrimp uring the perio from 1957 to The ata point for 1972 was exclue from calculation of the tren for white shrimp from Pensacola to the Mississippi River because no catch was reporte for the > 68 count category in 1972 an, therefore, the fit of the moel was poor (Table 3). May-August Cumulative Catch by Size Category We calculate the cumulative weight, P, of the May August catch in each of the same seven size categories, for each species, coastal area, an year (see Caillouet an Koi 1980, 1981). These catches were cumulate, starting with the size category of smallest shrimp (highest count, 2 68) an continuing towar the size category of largest shrimp (lowest count, 15-20). The following exponential moel escribe the relationship between Pi an Ci for each species, coastal area, an year: where Pi = cumulative weight of the May-August catch in the ith size category. The logarithmic form of moel 2 was use to estimate parameters c an by linear regression (Tables 5 through 7). The coefficients of etermination for the straight lines were very high. All slopes,, were negative, which reflecte the construction of moel 2 by cumulating catches from small- to large-shrimp size categories (see Caillouet an Koi 1980, 1981). There were significant upwar trens in for brown shrimp, but no significant trens in for white shrimp, in all three coastal areas from 1960 to 1978 (Table 4). The upwar trens inicate that the size of brown shrimp in the reporte May-August catches ecrease from 1960 to The values of for brown an white shrimp from the Texas coast (Table 5) were lower than those from the other two coastal areas (Tables 6 an 7), inicating that the shrimp in the May- August catch from the Texas coast generally were larger than those in the other two coastal areas. The ata point for 1972 was exclue from calculation of the tren for white shrimp from Pensacola to the Mississippi River (Table 7) as in the previous section.

7 64 CALLOUET AND KO1 TABLE 4. Trens in ex-vessel value (ollars) per shrimp by size category, in cumulative catch (pouns, heas off) by size category, an in cumulative ex-vessel value (ollars) of catch by size category, for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast, the Mississippi River to Texas, an Pensacola to the Mississippi River uring (base on ata from Tables 1-3,5-7 an 9-11). For ex-vessel value For cumulative For cumulative per shrimp by size catch by size ex-vessel value of Species Coastal Area category category catch by size category Texas Coast Trens Tren coefficients of etermination Mississippi River to Texas Trens Tren coefficients of etermination Pensacola to Mississippi River Trens Tren coefficients of etermination o.oo Texas Coast Trens Tren coefficients of etermination ~ Mississippi River to Texas Trens Tren coefficients of etermination Pensacola to Mississippi River Trens Tren coefficients of etermination Represents slopes of the linear regressions of b,,an h,respectively, on x, where x represents the last two igits of each year, The values b,, an h are efine in Tables 1-3, 5-7, an 9-1 1, respectively. Data for 1972 were exclue from regressions for white shrimp from Pensacola to the Mississippi River (see Tables 3,7, an 11). 2The tren (slope) was significantly ifferent from 0 at the 99% level of confience. 31nicates >O.OOO but <0.005, which woul not roun to The tren (slope) was significantly ifferent from 0 at the 95% level of confience. There were no significant correlations between the weight of the May-August catch (incluing pieces, Figures 2 through 7) each year an corresponing levels of (Table 8). A lack of correlation suggeste that size composition was not the major factor affecting the weight of the May- August catch. This woul be expecte if another factor (e.g., year-to-year variations in recruitment) playe a larger role than changes in size composition in etermining variations in weight of the May-August catch. May-August Cumulative Ex-vesel Value of Catch by Size Category For each species, coastal area, an year, we calculate the cumulative ex-vessel value, D, of the catch in each of the seven size categories, starting with the size category of smallest shrimp an cumulating towar the size category of largest shrimp (see Caillouet an Koi 1980, 1981). The following exponential moel escribe the relationship between Di an Ci for each species, coastal area, an year: Di = g (exp hci) (3) where Di = cumulative ex-vessel value of catch in the ith size category. The logarithmic form of moel 3 was use to estimate parameters g an h by linear regression (Tables 9 through 11). Very goo fits were inicate by the very high coefficients of etermination. All slopes, h, were negative, reflecting the construction of moel 3 by cumulating ex-vessel value of catch from small- to large-shrimp size categories. Only the upwar tren in h for brown shrimp from the Texas coast an the ownwar tren in h for white shrimp from Pensacola to the Mississippi River from 1960 to 1978 were statistically significant (Table 4). The upwar tren for brown shrimp from the Texas coast inicate that the proportions of the ex-vessel value of the May-August catch represente by the size categories of smaller shrimp increase from 1960 to The ownwar tren for white shrimp from Pensacola to the Mississippi River inicate that the proportions of the ex-vessel value of the May-August catch represente by the size categories of larger shrimp increase from 1960 to The ata point for 1972

8 VALUE AND SZE OF BROWN AND WHTE SHRMP 65 TABLE 5. Relationship between transforme cumulative weight (pouns, heas off) of catch, np, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast (statistical areas combine), * TABLE 6. Relationship between transforme cumulative weight (pouns, heas off) of catch, np, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from the Mississippi River to Texas (statistical areas combine), Year C r2 C rz ,811, ,182, ,575, ,873, ,996, ,666, ,600, ,724, ,618, ,706, ,567, ,765, ,003, ,536, ,211, ,559, , ,392, ,957, ,858, ,769, ,412, ,564, ,334, ,278, ,807, ,018, ,725, ,553, ,407, ,279, ,147, ,851, ,487, ,903, ,876, ,219, ,231, *Base on the linear regression of 1nP on C, where P = cumulative weight of May-August catch in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(c) = intercept, = slope, an = coefficient of etermination; all slopes,, were significant1 ifferent from 0 at the 99% level of confience, an the high r Y values inicate a very goo fit of the straight lines to the ata points. was exclue from calculation of the tren for white shrimp from Pensacola to the Mississippi River (Table 11) as in the two previous sections. Simulations Moels 1 an 2 provie information useful in simulating the impacts of preictable changes in moel parameters, barring any major changes in fishery management such as the closure of the FCZ off Texas. We conucte simulations to estimate what the overall average ex-vessel value per poun of the May-August catches of brown an white shrimp in the three coastal areas woul have been for selecte levels of b, to explore the possible consequences of changes in both the size composition of the catches an the ex-vessel price sprea among size categories. Because there were significant inverse relationships between ln(a) an b for both species in each coastal area (Table 8), we were able to estimate parameter a for selecte levels of parameter b for each species an coastal area, to simulate Vi in equation 1. We then calculate the corresponing ex-vessel value per poun by size category Year C r2 C ,792, ,683, ,121, ,298, ,538, ,842, ,312, ,665, ,600, ,210, ,922, ,789, ,351, ,561, ,059, ,688, ,812, ,701, ,423, ,735, , ,989, ,225, ,440, ,148, ,533, ,354, ,793, ,408, ,952, ,765, ,644, ,607, ,836, ,586, ,155, ,897,209 ~ ,211, *Base on the linear regression of np on C, where P = cumulative weight of May-August catch in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(c) = intercept, = slope, an r2 = coefficient of etermination, all slopes,, were significantly ifferent from 0 at the 99% level of confience, an the high 12 values inicate a very goo fit of the straight lines to the ata points. from the simulate Vi, n each case, we use the ex-vessel value per poun obtaine for the size category as an approximation of the minimum ex-vessel value per poun for the < 15 size category, because the moel i not encompass the < 15 size category. We then multiplie the simulate ex-vessel value per poun in each size category by the reporte pouns caught in each size category to simulate the ex-vessel value of the May-August catches by size category. The weight of catch in the category pieces was exclue from these calculations. The resulting values were summe over size categories to simulate the ex-vessel value of the May-August catches (pieces exclue). The simulate ex-vessel value was then ivie by the reporte May-August catch (pieces exclue) to obtain the simulate May-August average ex-vessel value per poun for each level of b for both species, for each coastal area, an for each year. Straight lines were fitte to the simulate ex-vessel value per poun versus by linear regression (Table 12, Figures 8 through 13). An increase in size of shrimp in the catches (as inicate by a ecrease in ), couple with an increase in price sprea

9 ~~ ~~ ~ 66 CALLOUET AND KO1 TABLE 7. Relationship between transforme cumulative weight (pouns, heas off) of catch, np, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from Pensacola to the Mississippi River (statistical areas combine), * Year C 12 C r ,688, ,116, ,525, , ,783, , ,786, ,028, ,320, ,610, ,107, , ,184, , ,420, , ,390, , ,867, , ,263, ,360, ,287, , ,473, ,844,069t t ,980, , ,348, , ,967, , ,660, , ,861, , ,224, , *Base on the linear regression of 1nP on C, where P = cumulative weight of May-August catch in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(c) = intercept, = slope, an r2 = coefficient of etermination; all slopes,, were significantly ifferent from 0 at the 99% level of confience, an the high 12 values inicate a very goo fit of the straight lines to the ata points.?both c an for white shrimp in 1972 are istorte because no catch was reporte for the 2 68 count category. among size categories (as inicate by a ecrease in b), clearly results in pronounce increases in the average ex-vessel value per poun for brown an white shrimp (Table 12, Figures 8 through 13). Decreases in b prouce greater increases in ex-vessel value per poun than equivalent ecreases in. Because catches also epen upon recruitment each year (Christmas an Etzol 1977), the simulate average ex-vessel value per poun can be use as a multiplier for estimating the ex-vessel value for a given weight of May-August catch of a given size composition, for selecte levels of b, for both species, an for each coastal area. DSCUSSON The extent to which the exclusion of unreporte catches from our analyses affecte our results an conclusions cannot be etermine. Because reporte catches of shrimp are not equivalent to actual catches, an because there are errors in assignment of catches to size categories, size composition of reporte catchesis not ientical to that of actual catches. Unknown portions of catches were not reporte, e.g., shrimp iscare because they i not meet minimum size limits or for economic reasons, catches by recreational fishermen, catches sol irectly to the consumer, an catches by foreign fishing craft (prior to 1976). Also unknown is the extent of errors of misclassification of catches by size category as a result of shrimp-graing practices. Such misclassification errors may average out in aggregate catches. However, a thorough investigation of the effects of shrimp graing practices ( machine graing an box graing ) on size istributions of shrimp assigne to various size categories woul be necessary to etermine the extent an magnitue of misclassification errors. TABLE 8. Linear regressions of catch (in millions of pouns, heas off; inclues pieces ) on, an h(a) on b for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast, the Mississippi River to Texas, an Pensacola to the Mississippi River, (base on ata from Tables 1-3 an 5-7). For catch on Texas Coast Mississippi River to Texas Pensacola to Mississippi River Slope ntercept Coefficient of Determination For n(a) on b Slope ntercept Coefficient of Determination Data for 1972 were exclue (see Tables 3,4, 7, an 11). 2The slope was significantly ifferent from 0 at the 99% level of confience. 3The slope was significantly ifferent from 0 at the 95% level of confience.

10 ~~ VALUE AND SZE OF BROWN AND WHTE SHRMP 67 TABLE 9. Relationship between transforme cumulative ex-vessel value (ollars) of catch, nd, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast (statistical areas Combine), * Year 8 h r2 g h r ,230, ,736, ,397, ,311, ,729, ,031, ,022, ,415, ,308, ,882, ,423, ,852, ,338, ,972, ,407, ,105, ,090, ,409, ,890, ,070, ,876, ,425, ,798, ,645, ,353, ,067, ,172, ,133, ,254, ,639, ,198, ,209, ,877, ,554, ,374, ,995, ,262, ,590, TABLE 10. Relationship between transforme cumulative ex-vessel value (ollars) of catch, nd, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from the Mississippi River to Texas (statistical areas combine), * Year g h r2 g h r ,294, ,791, ,367, , ,779, ,465, ,578, ,757, ,030, ,132, ,214, ,237, ,857, ,406, ,749, ,436, ,181, ,573, ,291, ,477, ,512, ,711, ,940, ,632, ,692, ,043, ,002, ,756, ,394, ,675, ,595, ,742, ,458, ,191, ,647, ,625, ,185, ,609, *Base on the linear regression of hd on C, where D = cumulative ex-vessel value of May-August catch in each of seven size categories, C = lower limit (count) of each of the seven size categories, h(g) = intercept, h = slope, an r2 = coefficient of etermination; all slopes, h, were significantly ifferent from 0 at the 99% level of confience, an the high r2 values inicate a very goo fit of the straight lines to the ata points. TABLE 1 1. *Base on the linear regression of lnd on C, where D = cumulative ex-vessel value of May-August catch in each of seven size categories, C = lower limit (count) of each of the seven size categories, ln(g) = intercept, h = slope, an r2 = coefficient of etermination; all slopes, h, were significant1 ifferent from 0 at the 99% level of confience, an the high r values inicate a very goo fit of the straight lines to the ata points. Relationship between transforme cumulative ex-vessel value (ollars) of catch, nd, an count, C, for reporte May- August catches (inshore an offshore combine) of brown an white shrimp from Pensacola to the Mississippi River (statistical areas combine), * Year g h r2 g h r2 Year g h r2 g h r ,761, , ,797, , ,500, , ,692, , ,749, , ,390, , ,541, , ,741, , ,320, , ,085, , ,598, ,749, ios ,545, ,432, ,185, ,592, ,694, ,094, ,804, ,304, , ,865, , , , , ,127, ,196, *Base on the linear regression of lnd on C, where D = cumulative ex-vessel value of May-August catch in each of seven size Categories, C = lower limit (count) of each of the seven size categories, ln(g) = intercept, h = slope, an r2 = coefficient of etermination; all slopes, h, were significantly ifferent from 0 at the 99% level of confience, an the high r2 values inicate a very goo fit of the straight lines to the ata points.?both g an h for white shrimp in 1972 are istorte because no catch was reporte for the 2 68 count category.

11 ~~ ~~~~ 68 CALLOUET AND KO1 TABLE 12. Linear regressions of simulate average ex-vessel value (ollars) per poun (heas off) on for reporte May-August catches (inshore an offshore combine) of brown an white shrimp from the Texas coast, the Mississippi River to Texas, an Pensacola to the Mississippi River, , an for selecte levels of b (base on ata from Tables 1-3 an 5-8). b' Species Coastal Area Brown shrimp Texas coast Slopes ntercepts Coefficients of etermination ' ' Brown shrimp Brown shrimp Mississippi River to Texas Pensacola to Mississippi River Slopes ntercepts Coefficients of etermination Slopes ntercepts Coefficients of etermination ' ' ' * ' ' White shrimp Texas coast ~ ~ - - Slopes ntercepts Coefficients of etermination ' ~~ ' ' White shrimp Mississippi River to Texas Slopes ntercepts Coefficients of etermination White shrimp Pensacola to Mississippi River Slopes ntercepts Coefficients of etermination ' ' ' ,2227' 'Levels of b selecte for the simulations encompasse the observe ranges in b, for the most part 'The slope was significantly ifferent from 0 at the 99% level of confience. TEXAS COAST BROUN SHRMP TEXAS COAST UHTE SHRMP 7.!, * b *. b b ,, - - -, - b--0.04, -,, Figure 8. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of brown shrimp from the Texas coast (statistical areas combine), at selecte levels of b over the range of (base on ata from Tables 1, 5, an 8). Lines fitte by linear regression (Table 12) J,.. mc* b=-0.07 b=-0.06 b= b= Figure 9. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of white shrimp from the Texas coast (statistical areas combine), at selecte levels of b over the range on ata from Tables 1, 5, an 8). Lines fitte by linear regression (Table 12).

12 VALUE AND SZE OF BROWN AND WHTE SHRMP 69 MSSSSPP RVER TO TEXAS BROWN SHRMP PENSACOLA TO THE MSSSSPP RVER BROWN SHRMP 001 n a LL : 1.00 Y) U J 2 a c - b=-0.07 b b=-0.05 b= , n 3 b E _? b=-0 05 b= Figure 10. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of brown shrimp from the Mississippi River to Texas (statistical areas combine), at selecte levels of b over the range of (base on ata from Tables 2, 6, an 8). Lines fitte by linear regression Gable 12) Figure 12. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of brown shrimp from Pensacola to the Mississippi River (statistical areas combine), at selecte levels of b over the range of (base on ata from Tables 3, 7, an 8). Lines fitte by linear regression (Table 12). MSSSSPP RVER TO TEXAS WHTE SHRMP WHTE SHRMP 75 l n 3 L bl-0.07 a 3 (L M m 0 00 i, bc U _1 J b= L ': b= b= Figure 11. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of white shrimp from the Mississippi River to Texas (statistical areas combine), at selecte levels of b over the range of (base on ata from Tables 2,6, an 8). Lines fitte by linear regression (Table 12). There were significant ecreases in size of brown shrimp in the reporte May-August catches from the three coastal areas from 1960 to Caillouet et al. (1980) etecte significant ecreases in size of brown shrimp in reporte annual catches from Texas an Louisiana from 1959 to 1976, an Caillouet an Koi (1980) etecte significant ecreases in sue of brown shrimp in reporte annual lanings from the northern Gulf from 1961 to Fishing effort has increase substantially in the northern Gulf coast since 1960 (Christmas an Etzol 1977, GMFMC 1980). Figure 13. Simulate average ex-vessel value (ollars) per poun (heas off) for reporte May-August catches (inshore an offshore combine) of white shrimp from Pensacola to the Mississippi River (statistical areas combine), at selecte levels of b over the range of (base on ata from Tables 3, 7, an 8). Lines fitte by linear regression (Table 12). For this reason, Caillouet et al. (1980) suggeste that the observe ecreases in size of brown shrimp may be the effects of increase fishing effort leaing to the harvesting of increasing quantities of small shrimp before they grow to larger sizes. However, in the absence of a ecline in total catch or conclusive evience that shrimp are being harveste at rates in excess of that which woul maximize yiel, this cannot be construe as growth overfishing. The ecrease in size of brown shrimp in catches from the Texas coast may be reverse as a result of closure of the FCZ off Texas ue

13 70 CALLOUET AND KO1 to postponement of fishing until the shrimp grow to larger sizes. Couple with continue increase in the price sprea among size categories, an increase in size of brown shrimp in the Texas coast catch coul greatly enhance the value of that catch. On the other han, the closure may increase fishing effort along the coasts of Louisiana, Mississippi, an Alabama (GMFMC 1980), with the possible consequence of exacerbating the trens towar ecrease of size of brown shrimp in the catches from these areas. n aition, the increase competition among offshore units coul force some of the smaller ones to fish inshore as an alternative, thereby increasing the fishing pressure inshore. To our surprise, there were no significant changes in size composition of reporte May-August catches of white shrimp in the three coastal areas from 1960 to However, if fishing pressure on the white shrimp stock were increase as a result of closure of the FCZ off Texas, the size of white shrimp in the May-August catch coul ecrease. Caillouet et al. (1980), an Caillouet an Koi (1980) etecte ecreasesin size of white shrimp in reporte annual catches an lanings, respectively, from the northern Gulf. Therefore, these ecreases in size must have been generate by an overwhelming influence of size composition of the catches uring months other than May-August. Our analyses o not account for the impact of overall inflation on the trens in ex-vessel value of shrimp catches. However, they o inicate that the rate of inflation in ex-vessel value per shrimp is higher for larger than for smaller shrimp, a phenomenon that shoul be consiere in stuies of inflationary effects on the ex-vessel value of shrimp catches. We have characterize the ex-vessel value per shrimp by size category, size composition, an ex-vessel value composition of the reporte May-August catches of brown an white shrimp from the Texas, Louisiana, Mississippi, an Alabama coasts from 1960 to Comparisons, by similar analyses, with catch statistics for 1979, 1980, an 1981, shoul be of particular use an interest as one means of assessing the impacts of the closure of the FCZ off Texas. ACKNOWLEDGMENTS We are especially grateful to those who ha the foresight to esign an implement the gathering of ata on weight an ex-vessel value of shrimp catches by statistical area, species, an size category, because they have mae our analyses possible. Notable among them are Charles H. Lyles, Director, Gulf States Marine Fisheries Commission (formerly of the USFWS), George W. Rounsefell (formerly Director, Galveston Laboratory, BCF, USFWS, ecease), Joseph H. Kutkuhn, Director, Great Lakes Fisheries Laboratory, USFWS, Ann Arbor, Michigan (formerly Assistant Director, Galveston Laboratory, BCF, USFWS), an George W. Snow (formerly Chief, Division of Statistics an Market News, NMFS, New Orleans. Louisiana, retire). The manuscript was reviewe by Dr. Ewar F. Klima, NMFS, Galveston Laboratory; John P. Wise, NMFS, Washington, D.C.; Dr. Clarence P. yll, National Council on Oceans an Atmosphere, NOAA, Washington, D.C.; an John War, NMFS, Miami, Floria, who provie many helpful suggestions. Beatrice Richarson, clerk-typist, NMFS, Galveston Laboratory, type the manuscript. REFERENCES CTED Caillouet, C. W. & D. B. Koi Trens in ex-vessel value an size composition of annual lanings of brown, pink an white shrimp from the Gulf an south Atlantic coasts of the Unite States. Mar. Fish. Rev. 42(12): Trens in ex-vessel value an size composition of reporte annual catches of pink shrimp from the Tortugas fishery, Gulf Res. Rept. 7(1): Caillouet, C. W., F. J. Patella & W. B. Jackson Trens towar ecreasing size of brown shrimp, Penaeus aztecus, an white shrimp, Penaeus setiferus, in reporte annual catches from Texas an Louisiana. Natl. Mar. Fish. Serv., Fish. Bull. 77(4): Christmas, J. Y & D. J. Etzol (es.) n e Shrimp Fishery of the Gulf of Mexico Unite States: A Regional Management Plan. Technical Report Series No. 2, Gulf Coast Research Laboratory, Ocean Springs, MS. 128 pp. Griffin, W. L., R. D. Lacewell & W. A. Hayenga Estimate costs, returns, an financial analysis: Gulf of Mexico shrimp vessels. Mar. Fish. Rev. 36(12):1-4. Griffin, W. L. & J. P. Nichols An analysis of increasing costs to Gulf of Mexico shrimp vessel owners: Mar. Fish. Rev. 38(3):8-12. Griffin, W. L., N. J. Warlaw & John P. Nichols Economic an financial analysis of increasing costs in the Gulf shrimp fleet. Natl. Mar. Fish. Serv., Fish. Bull. 74(2): Gulf of Mexico Fishery Management Council (GMFMC) Fishery management plan for the shrimp fishery of the Gulf of Mexico. Fe. Reg. 45(218): Henerson, F The ynamics of the mean-size statistic in a changing fishery. FA0 Fish. Tech. Paper No pp. Neal, R. A An application of the virtual population technique to penaei shrimp. Pages in Proceeings of the 21st Annual Conference, Southeastern Association of Game an Fish Commissioners. Rickcr, W. E Computation an interpretation of biological statistics of fish populations. Bull. Fish. Res. Boar Can. 191: 382 pp. Whitaker, D. R The U.S. shrimp inustry: past trens an prospects for the 1970 s. Mar. Fish. Rev. 35(5-6):23-30.