SCRS/2004/130

1446

Col. Vol. Sci. Pap. ICCAT, 58(4): 1446-1465 (2005)

STANDARDIZED CATCH RATES BY SEX AND AGE FOR SWORDFISH (XIPHIAS GLADIUS) FROM THE U.S. LONGLINE FLEET 1981-2003

Mauricio Ortiz1

SUMMARY

Swordfish (Xiphias gladius) catch and effort data from the U.S. Pelagic longline fleet operating in the Western North Atlantic were used to update indices of abundance for the North Atlantic swordfish stock. Standardized catch rates were estimated using a Generalized Linear Mixed modeling approach assuming a delta-lognormal error distribution. Indices of abundance in units of biomass (dressed weight) were estimated for fish greater than 33 lbs due to U.S. size restrictions implemented in 1991. For comparison, indices of abundance in numbers of fish for ages 3-10+ combined sexes were estimated. The age-sex slicing algorithms used were the size at age by sex relationship employed during the 1999 swordfish stock assessment. The explanatory variables considered for standardization included geographical area, seasonal trimesters, fishing target species, and a fixed factor operational procedure (OP) that classifies the U.S. longline fishing fleet according to boat and fishing gear characteristics, and fishing styles.

RÉSUMÉ

Les données de prise et d’effort de l’espadon (Xiphias gladius) de la flottille palangrière pélagique des Etats-Unis opérant dans l’Atlantique Nord-Ouest ont été utilisées pour actualiser les indices d’abondance du stock d’espadon nord-atlantique. Les taux de capture standardisés ont été estimés à l’aide d’une approche de modèle linéaire généralisé mixte postulant une distribution d’erreur delta-lognormale. Les indices d’abondance dans les unités de biomasse (poids manipulé) ont été estimés pour les poissons supérieurs à 33 lbs (13,5 kg) en raison des restrictions sur les tailles en vigueur aux Etats-Unis depuis 1991. A titre de comparaison, on a estimé les indices d’abondance en nombres de poissons pour les âges 3-10+, sexes combinés. Les algorithmes de découpage des âges-sexes utilisés se basaient sur le rapport de taille à l’âge par sexe employé pendant l’évaluation du stock d’espadon de 1999. Les variables explicatives considérées pour la standardisation incluaient la zone géographique, les trimestres saisonniers, l’espèce-cible de la pêche, et une procédure opérationnelle (OP) à facteur fixe qui classifie la flottille de pêche palangrière nord-américaine en fonction des caractéristiques du bateau et de l’engin de pêche, et des modes de pêche.

RESUMEN

Se utilizaron los datos de captura y esfuerzo para el pez espada (Xiphias gladius) de la flota palangrera pelágica estadounidense que opera en el Atlántico noroccidental para actualizar los índices de abundancia del stock de pez espada del Atlántico norte. Se estimaron las tasas de captura estandarizadas utilizando un enfoque de modelado lineal mixto generalizado, asumiendo una distribución de error delta-lognormal. Se estimaron los índices de abundancia en unidades de biomasa (peso canal) para los especimenes de más de 33 lbs, debido a las restricciones de talla estadounidenses que se implementaron en 1991. A efectos comparativos, se estimaron índices de abundancia en número para peces de edades 3-10+, para ambos sexos. Los algoritmos de separación de edad-sexo utilizados se basaron en la relación de talla por edad y por sexos empleada durante la evaluación de stock de pez espada de 1999. Las variables explicativas consideradas para la estandarización incluían zona geográfica, trimestres estacionales, especies objetivo de la pesca y un procedimiento operativo de factor fijo (OP) que clasifica la flota estadounidense de pesca al palangre en función de las características de los buques y los artes de pesca y de las modalidades de pesca.

KEY WORDS

Catch/effort, abundance, longline, pelagic fisheries, swordfish

1 U.S. Department of Commerce National Marine Fisheries Service, Southeast Fisheries Science Center. Sustainable Fisheries Division 75 Virginia Beach Drive. Miami, Florida 33149 USA Contribution SFD-2003/043. Email: Mauricio.ortiz@noaa.gov

1447

1. Introduction Indices of abundance of swordfish from commercial fisheries have been used to tune stock assessment models (Anonymous 2000). Data collected from the U.S. pelagic longline fishery was used to develop age and sex specific standardized catch per unit effort (CPUE) indices for the North Atlantic swordfish stock. This report updates the methods applied to the available U.S. longline fleet data through 2003 and presents age-sex specific standardized CPUE indices for swordfish. Standardized catch rates were estimated using the Generalized Linear Mixed Model (GLMM) approach. 2. Materials and methods All fishers that fish for and land swordfish in U.S. are required to report their catch through logbooks. Each report includes the catch in numbers of all caught species and general fishery settings for each longline set (Pelagic Longline Logbook data). They are also required to submit weight-out sheets for each trip, which include individual carcass weights for swordfish and other large pelagic species landed and marketed in the U.S. (Weight-out data). The Pelagic longline fleet has also an observer program, established in 1992 that monitored the fishing activities of the fleet, recording detailed information on fishing operations, gear characteristics and deployment, environmental related conditions and biological information from all longline catch (Lee and Brown 1998). Implementation of U.S. regulations, in conformity with the ICCAT recommendations, limit the allowable landings of swordfish by U.S. fishers, resulting in changes in both the type of data obtained and in the protocols in which the data are used for analysis. Regulatory norms that affect the present analysis include: first the implementation(s) of the minimum size of 125 cm LJFL with a 15% tolerance in mid 1991, subsequently modified to 119 cm LJFL with 0% tolerance in mid 1996; second, the implementation of a total annual allowable catch (TAC) since 1995; and third, due to management regulations related to swordfish and or other species, time-area closures that were in effect since late 1999. These time-area restrictions include two permanent closures to pelagic longline; the Desoto Canyon in the Gulf of Mexico (effective since November 1, 2000) and the Florida east coast (effective since March 1, 2001) (Figure 1). There are also three time-area closures for longline in the U.S. Atlantic coast: the Charleston Bump that is closed from February 1 to April 30, effective in 2001, the Bluefin tuna protection area that is closed from June 1st to June 30, effective in 1999, and the Grand Banks that was closed from July 17, 2001 to January 9, 2002, as a result of an emergency rule implementation (Cramer 2002). Age-sex specific indices were developed after ageing the swordfish catch at size data. The age slicing method used the Ehrhardt’s size at age growth models for males and females (Ehrhardt et al. 1995). Since swordfish sex ratios differ in a spatio-temporal scale (Mejuto et al. 1998) estimated sex ratios at size by area and time variation were incorporated into the age-slicing procedures (Ortiz et al. 2000). Similar procedures were used in the latest stock assessment of North Atlantic swordfish (Anonymous 2000). The swordfish weight-out data set extends from 1981 through 2003. Each record represents information of catch by vessel-trip, including date, geographical area of the catch (Figure 1), catch in numbers and weight for swordfish, tunas and other market species, and fishing effort estimated as number of sets per trip times the average number of hooks per set. Prior to 1991, reporting of fish sizes and fishing effort was voluntary and incomplete for many longline vessels. The U.S. longline pelagic fleet includes at least 1,714 different registered vessels from 1981 to 2003. This fleet has changed in terms of gear technology and fishery operations, Hoey et al. (1988) characterized the swordfish fleet into nine different vessel-groups based on boat size-power and fishing operations. This classificatory factor has shown to be an important explanatory variable of several species catch rates including swordfish (Ortiz and Cramer 2000). The longline fishing grounds of the U.S. fleet extend from the Grand Banks in the North Atlantic to 5°-10° latitude south, off the South America coast, including the Caribbean and the Gulf of Mexico. Eight geographical areas have been defined for spatial classification of this fishery (Figure 1). These include: the Caribbean (CAR, area 1), Gulf of Mexico (GOM, area 2), Florida East coast (FEC, area 3), South-Atlantic Bight (SAB, area 4), Mid-Atlantic Bight (MAB, area 5), New England coastal (NEC, area 6), Northeast Distant waters (NED, area 7) and the Southern offshore (OFS, area 8). Trimesters were used to account for seasonal fishery distribution through the year (Jan-Mar, Apr-Jun, Jul-Sep, and Oct-Dec).

1448

Fishing effort is reported as total number of hooks per set times the number of sets per trip, therefore nominal catch rates were calculated as numbers of swordfish caught per 1000 hooks. In addition, a classificatory variable Size-set was defined as the mean number of hooks per set, grouping observations into 3 levels: a) trips with an average of 100 to 300 hooks/set, b) trips with 300 to 500 hooks/set, and c) trips with 500 or more hooks/set. Set size was assumed to control for changes in gear deployment hypothesized to affect catch rates. Swordfish is a main target species of the U.S. pelagic longline fleet. However this fleet also targets tunas (yellowfin, bigeye and bluefin tuna) and to a lesser extent other pelagic species including sharks. A proxy for targeted species was defined based on the proportion of swordfish catch to total catch per trip and grouped into categories, corresponding to the quartiles 0-25%, 25-50%, 50-75%, and 75-100%. This target variable was assumed to control for effects on swordfish catch rates associated with the diverse species targeted by the fleet. Standardized indices of abundance were estimated for Age-Sex classes and as combined sex age groups. Catch at age by sex (CAAS) were generated using sex-ratios at size and size-at-age slicing methods used in the 1999 swordfish stock assessment (Anonymous 2000, Ortiz et al. 2000). For females, CAAS was estimated for ages 0 to 10+, while for males CAAS was estimated for ages 0 to 5+. Due to size restrictions implemented in 1991, standardized CPUE were restricted for ages 0, 1 and 2 from 1981 to 1990. For ages 3 and older, standardized CPUE were estimated from 1981 to 2003. Combined sex standardized CPUE was estimated for ages 0, 1, and 2 also from 1981 to 1990, and for ages 3 and older from 1981 to 2003. As per recommendation of the SCRS, a swordfish biomass index was estimated using the Weight-Out data. This biomass index was restricted to fish ≥ 13 kg (due to size restrictions implemented in 1991) and estimated as total pounds landed per thousand hooks. Due to the recent implementation of time-area closures on pelagic longline fishing, it was considered important to evaluate its possible effects of swordfish catch rates. A limitation with the Weight-out data, that generates the biomass index, is that it includes only general geographic area(s) of the trip-catch, and given that time-area closure restrictions are smaller than the geographic area definitions (Figure 1), it was not possible to directly allocate the swordfish catch to a non-closure or closure location within the general geographic areas. However the Pelagic Logbook data have specific latitude longitude information for most of the sets, and it is possible to link each set in the Pelagic Logbook data to the corresponding trip record in the Weight-Out database, although only from 1996 on. Thus, we estimated an average lat-lon position for each trip, based on sets from the Pelagic Logbooks. Preliminary results of standard swordfish catch rates show non-significant differences between analysis including or not observations from the time-area closures (Ortiz and Scott 2001). However, due to the limited number of years for comparison it was recommended to exclude observations from the time-area closure areas for 1996 on. Relative indices of abundance of swordfish were estimated by Generalized Linear Modeling approach assuming a delta lognormal model distribution. The standardization protocols assumed a delta model with a binomial error distribution for modeling the proportion of positive sets, and a lognormal error distribution for modeling the mean catch rate of successful (i.e. positive swordfish catch) sets. The lognormal frequency distributions age aggregated by sex and data source are shown in Figure 2. Parameterization of the models used the GLM structure; for the proportion of successful sets per stratum is assume to follow a binomial distribution where the estimated probability is a linear function of fixed factors and interactions. The logit function was used as a link between the linear factor component and the binomial error. For successful sets, estimated CPUE rates assumed a lognormal distribution of a linear function of fixed and random effect interactions when the year term was within the interaction. A step-wise regression procedure was used to determine the set of systematic factors and interactions that significantly explained the observed variability. As the deviance difference between two consecutive nested models follows a chi-square (χ2) distribution, this statistic was used to test for the significance of an additional factor(s) in the model. Deviance analysis tables are presented for each data set analysis. Each table includes the deviance for the proportion of positive observations, and the deviance for the positive catch rates. Final selection of the explanatory factors was conditional to: a) the relative percent of deviance explained by adding the factor in consideration, normally factors that explained more than 5% were included in the final model, b) the χ2 test significance, and c) type III test significance within the final specified model. Once a set of fixed factors was specified, possible first level interactions were evaluated in particular random interactions between the year effect and other factors. The significance of random interactions was evaluated between nested models by using the likelihood ratio test (Pinheiro and Bates 2000), the Akaike information criteria (AIC), and the Bayesian

1449

information criteria (BIC) (Littell et al 1996). Analyses were done using GLIMMIX and MIXED procedures from the SAS® statistical computer software (SAS Institute Inc. 1997) Relative indices of abundance were estimated by age-sex or age groups, and for biomass of fish ≥ 13 kg. Within age-sex analyses, the age component was included as fixed factor in the model. Relative indices were calculated as the product of the year (year*age in age-sex indices) effect least square means (LSmeans) from the binomial and the lognormal components. LSmeans estimates were weighted proportional to observed margins in the input data, and for the lognormal estimates, a log-back transformed bias corrections was applied (Lo et al. 1992). 3. Results and discussion The deviance analyses tables for the swordfish CPUE standardization by age-sex results are shown in Table 1 for males (Age 0 to 5+), Table 3 for females (Age 0 to 10+), and Table 5 for combined sex, age groups 3-10+, respectively. Table 7 shows the deviance table for the swordfish biomass index derived from the weight-out data. In this case analysis excluded those trips that were within an area designated as management area from 1996 on. The age-sex index standardization analyses indicated that area, OP and target where the main explanatory factors for the proportion of positive sets models, both for males and females. While for the positive catch sets models, the main explanatory factors were OP, size set (Szst), target and area. Of the interactions evaluated, the year*Area, and year*OP were also important explanatory factors primarily for the positive catch sets models. Tables 2 and 4 present the evaluation of these interactions as random components in the mixed models. For combined sex and age aggregated indices, also area, OP and target as well quarter were the main explanatory factors for the proportion of positive sets, while for the positive catch sets, the factors area, OP, target and size set did explain most of the observed variability (Table 6). The biomass index analyses also reiterated area, OP, target and quarter and the random interactions year*Op and area*quarter as main explanatory factors for the proportion of positive trips (Table 7). While area, OP and target and random interactions year*area and year*OP were the main explanatory factors of catch rates for trips with catches of swordfish (Table 8). Table 9 and Figure 4 show the nominal and standardized CPUE by age for female swordfish, while Table 10 and Figure 3 present correspondent results for males. Figure 5 and Tables 11 and 12 show the nominal and standardized CPUE for combined sex-age groups 0-2 and 3-10+, respectively. Reviewing catch trends by sex and age, for males ages 0-1 and females 0-1 the catch trends were different for the early years (1981-1987) (Figure 3 and 4). For older fish, ages 3 and above, both males and females show a general decline trends from higher values in the early period (1981/82) to lower values in the early 1990’s, follow by a rather constant low catch rates through the 1990’s with some slight indication of recovery by the latest years (2000/02) particularly for fish age 3. Figure 7 shows the trends for ages 3, 4 and 5 (the main component of the catch) for males and females swordfish adjusted to the cohort-group year. In general cohorts follow similar trends comparing by contiguous ages. For ages 3 both males and females showed an increase in catch rates in the latest year, while ages 4 and 5 showed lower catch rates compared to 2002, but overall there is a continuous increasing trend from the lower values in the mid-1990s. Figures 8 and 9 present diagnostic plots; mainly for the positive observations component of the delta lognormal model fits. The cumulative normalized residuals (qq-plots) indicate possible deviations from the assumed error distribution of the model. In this case the expected residual pattern should follow the straight line. Figure 8 shows the qq-plots for females and males by age, respectively, only ages 0 and 1 show some deviations, particularly at the lower end of the distribution. This pattern also repeated with the combined sex-age 0-2 group and the biomass index. Distribution of residuals by year show no trend or bias.

References ANONYMOUS. 2000. Report of the ICCAT Swordfish Stock Assessment Session (Madrid, Spain, September

27-October 4, 1999). - Col. Vol. Sci. Pap. ICCAT 51(1):1001-1209.

CRAMER, J. 2002. Large Pelagic Logbook Newsletter 2000. NOAA Tech. Mem. NMFS SEFSC 471, 26 p.

1450

CRAMER, J. and A. Bertolino. 1998. Standardized catch rates for swordfish (Xiphias gladius) from the U.S. longline fleet through 1997. Col. Vol. Sci. Pap. ICCAT 49(1):449-456.

EHRHARDT, N.M, R.J. Robbins, and F. Arocha. 1995. Age validation and growth of swordfish, Xiphias gladius, in the Northwest Atlantic. Col. Vol. Sci. Pap. ICCAT 45(2):358-367.

LEE, D.W. and C.J. Brown. 1998. SEFSC Pelagic Observer Program Data Summary for 1992-1996. NOAA Technical memorandum NMFS-SEFSC-408:21 p.

LITTELL, R.C., G.A. Milliken, W.W. Stroup, and R.D Wolfinger. 1996. SAS® System for Mixed Models, Cary NC:SAS Institute Inc., 1996. 663 pp.

LO, N.C., L.D. Jacobson, and J.L. Squire. 1992. Indices of relative abundance from fish spotter data based on delta-lognormal models. Can. J. Fish. Aquat. Sci. 49: 2515-2526.

MEJUTO, J., J.M. de la Serna, and B. Garcia. 1998. Some considerations on the spatial and temporal variability in the sex ratio at size of the swordfish (Xiphias gladius). Col. Vol. Sci. Pap. ICCAT 48(1):205-215.

ORTIZ, M., V.R. Restrepo, and S.C. Turner. 2000. North Atlantic swordfish sex-ratios at size keys: Analysis and development. Col. Vol. Sci. Pap. ICCAT 51(1):1480-1509.

ORTIZ, M, and J. Cramer. 2000. Standardized catch rates by sex and age for swordfish (Xiphias gladius) from the U.S. longline fleet 1981-1998. Col. Vol. Sci. Pap. ICCAT 51(1):1559-1620.

PINHEIRO, J.C. and D.M. Bates. 2000. Mixed-effect models in S and S-plus. Statistics and Computing. Springer-Verlag New York, Inc.

SAS Institute Inc. 1997, SAS/STAT® Software: Changes and Enhancements through Release 6.12. Cary, NC:Sas Institute Inc., 1997. 1167 pp.

1451

Table 1. Deviance analysis table of explanatory variables in the delta lognormal model for swordfish catch rates by Age and sex (number of fish per thousand hooks) from the US Pelagic Longline fishery. Percent of total deviance refers to the deviance explained by the full model; p value refers to the Chi-square probability between consecutive models (alpha = 0.05). Swordfish Females by Age ( 0-3 1981-1990 / 3-10+ 1981-2003)

Model factors positive catch rates values d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 94419.5409Year 22 80094.8532 14324.69 24.3% < 0.001Year Age 10 57935.3178 22159.54 37.5% < 0.001Year Age Op 6 47223.9474 10711.37 18.1% < 0.001Year Age Op Area 6 45831.8102 1392.14 2.4% < 0.001Year Age Op Area Qtr 3 45430.5611 401.25 0.7% < 0.001Year Age Op Area Qtr Szst 2 41663.4592 3767.10 6.4% < 0.001Year Age Op Area Qtr Szst Targ 3 38826.0438 2837.42 4.8% < 0.001Year Age Op Area Qtr Szst Targ Year*Age 181 38435.4826 390.56 0.7% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Area 124 37415.0747 1020.41 1.7% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Op 118 36970.9085 444.17 0.8% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Area 27 36849.2729 121.64 0.2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Qtr 65 36578.0848 271.19 0.5% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Qtr 18 36382.7235 195.36 0.3% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Qtr 18 36324.9408 57.78 0.1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Szst 40 35954.9321 370.01 0.6% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Szst 12 35864.3346 90.60 0.2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Szst 12 35808.9386 55.40 0.1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Qtr*Szst 6 35759.9908 48.95 0.1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Targ 64 35594.4095 165.58 0.3% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Targ 18 35498.5427 95.87 0.2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Targ 18 35432.0415 66.50 0.1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Qtr*Targ 9 35397.4056 34.64 0.1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Szst*Targ 6 35378.1563 19.25 0.0% 0.004

Model factors proportion positives d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 121473.634Year 22 114293.586 7180.05 11% < 0.001Year Age 10 93217.749 21075.84 33% < 0.001Year Age Op 6 74661.323 18556.43 29% < 0.001Year Age Op Area 6 69342.392 5318.93 8% < 0.001Year Age Op Area Qtr 3 67899.239 1443.15 2% < 0.001Year Age Op Area Qtr Szst 2 67353.643 545.60 1% < 0.001Year Age Op Area Qtr Szst Targ 3 60206.777 7146.87 11% < 0.001Year Age Op Area Qtr Szst Targ Year*Age 181 59838.260 368.52 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Szst 41 59446.507 391.75 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Targ 66 59321.950 516.31 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Op*Area 30 59148.800 689.46 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Qtr 65 59128.461 709.80 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Area*Qtr 18 58937.013 901.25 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Op 119 58572.151 1266.11 2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Area 124 57899.034 1939.23 3% < 0.001 Table 2. Analysis of mixed model formulations for swordfish catch rates by age-sex from the U.S. pelagic longline fishery. Likelihood ratio tests the difference of –2 REM log likelihood between two nested models.

Swordfish Females by Age [0-3 1981-1990 / 3-10+ 1981-2003] GLMixed Model

-2 REM Log

likelihood

Akaike's Information

Criterion

Schwartz's Bayesian Criterion

Proportion Positives * Year Age Year*Age Area Target OP 145193.8 145195.8 145204.6

Year Age Year*Age Area Target OP Year*Area 145430.3 145434.3 145434.3 -236.5 N/AYear Age Year*Age Area Target OP Year*Area Year*OP 145229.6 145235.6 145244.7 200.7 0.0000

Positives catch ratesYear Age Year*Age OP Area Szst Target 146539.6 146541.6 146550.6Year Age Year*Age OP Area Szst Target Year*Area 145423.3 145427.3 145433.4 1116.3 0.0000

* Year Age Year*Age OP Area Szst Target Year*Area Year*Szst 144548.1 144552.1 144559.9 875.2 0.0000Year Age Year*Age OP Area Szst Target Year*Area Year*Szst Year*OP 144537.3 144545.3 144557.5 10.8 0.0010

Likelihood Ratio Test

1452

Table 3. Deviance analysis table of explanatory variables in the delta lognormal model for swordfish catch rates by age and sex (number of fish per thousand hooks) from the U.S. pelagic longline fishery. Percent of total deviance refers to the deviance explained by the full model; p value refers to the Chi-square probability between consecutive models (alpha = 0.05). Swordfish Males by Age ( 0-2 1981-1990 / 3-5+ 1981-2003)

Model factors positive catch rates values d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 40169.4244Year 22 35534.971 4634.45 22.9% < 0.001Year Age 5 33885.8058 1649.17 8.1% < 0.001Year Age Op 6 27489.1978 6396.61 31.6% < 0.001Year Age Op Area 6 26682.8581 806.34 4.0% < 0.001Year Age Op Area Qtr 3 26447.2978 235.56 1.2% < 0.001Year Age Op Area Qtr Szst 2 24571.1315 1876.17 9.3% < 0.001Year Age Op Area Qtr Szst Targ 3 22355.8582 2215.27 10.9% < 0.001Year Age Op Area Qtr Szst Targ Year*Age 71 22190.8518 165.01 0.8% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Area 124 21522.3662 668.49 3.3% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Op 116 21153.2461 369.12 1.8% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Area 29 21065.1485 88.10 0.4% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Year*Qtr 65 20721.1315 344.02 1.7% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Qtr 18 20602.8436 118.29 0.6% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Qtr 18 20572.1548 30.69 0.2% 0.031Year Age Op Area Qtr Szst Targ Year*Age … Year*Szst 39 20413.5063 158.65 0.8% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Szst 12 20339.1424 74.36 0.4% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Szst 12 20312.4434 26.70 0.1% 0.009Year Age Op Area Qtr Szst Targ Year*Age … Qtr*Szst 6 20291.6109 20.83 0.1% 0.002Year Age Op Area Qtr Szst Targ Year*Age … Year*Targ 63 20174.9683 116.64 0.6% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Area*Targ 18 20053.1879 121.78 0.6% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Op*Targ 18 19970.1632 83.02 0.4% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Qtr*Targ 9 19938.8517 31.31 0.2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age … Szst*Targ 6 19928.7716 10.08 0.0% 0.121

Model factors proportion positives d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 43064.380Year 22 41724.915 1339.47 8% < 0.001Year Age 5 40891.801 833.11 5% < 0.001Year Age Op 6 34846.375 6045.43 37% < 0.001Year Age Op Area 6 30931.245 3915.13 24% < 0.001Year Age Op Area Qtr 3 30891.552 39.69 0% < 0.001Year Age Op Area Qtr Szst 2 30848.056 43.50 0% < 0.001Year Age Op Area Qtr Szst Targ 3 27777.470 3070.59 19% < 0.001Year Age Op Area Qtr Szst Targ Year*Age 71 27668.678 108.79 1% 0.003Year Age Op Area Qtr Szst Targ Year*Age Op*Area 30 27501.457 167.22 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Szst 41 27473.091 195.59 1% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Targ 66 27334.280 334.40 2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Area*Qtr 18 27283.287 385.39 2% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Qtr 65 27106.937 561.74 3% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Op 119 27089.721 578.96 4% < 0.001Year Age Op Area Qtr Szst Targ Year*Age Year*Area 124 26839.766 828.91 5% < 0.001 Table 4. Analyses of mixed model formulations for swordfish catch rates by age-sex from the U.S. pelagic longline fishery. Likelihood ratio tests the difference of –2 REM loglikelihood between two nested models. * indicates the final delta mixed model.

Swordfish Males by Age [0-5+ 1981-1990 / 3-5+ 1981-2002] GLMixed Model

-2 REM Log

likelihood

Akaike's Information

Criterion

Schwartz's Bayesian Criterion

Positives catch ratesYear Age Year*Age Area OP Szst Targ 73607.8 73609.8 73618Year Age Year*Age Area OP Szst Targ Year*Area 73114.6 73118.6 73124.6 493.2 0.0000Year Age Year*Age Area OP Szst Targ Year*Area Year*OP 72813.4 72817.4 72826.2 301.2 0.0000

* Year Age Year*Age Area OP Szst Targ Year*Area Year*OP Year*Qtr 72439.3 72445.3 72458.5 374.1 0.0000

Proportion Positives * Year Age Year*Age OP Area Targ 80347.5 80349.5 80357.4

Year Age Year*Age OP Area Targ Year*Area 80355.3 80359.3 80365.3 -7.8 N/AYear Age Year*Age OP Area Targ Year*Area Year*OP 80461.1 80465.1 80474.1 -105.8 N/A

Likelihood Ratio Test

1453

Table 5. Deviance analysis table of explanatory variables in the delta lognormal model for swordfish catch rates combined sex and age groups (number of fish per thousand hooks) from the U.S. pelagic longline fishery. Percent of total deviance refers to the deviance explained by the full model; p value refers to the Chi-square probability between consecutive models (alpha = 0.05). Swordfish Combined Age 3-10+ 1981-2003

Model factors positive catch rates values d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 29321.4395Year 22 26707.5983 2613.84 13.4% < 0.001Year Op 6 19081.2958 7626.30 39.0% < 0.001Year Op Area 6 16709.0543 2372.24 12.1% < 0.001Year Op Area Qtr 3 16029.9434 679.11 3.5% < 0.001Year Op Area Qtr Szst 2 14710.1212 1319.82 6.8% < 0.001Year Op Area Qtr Szst Targ 3 11534.3442 3175.78 16.3% < 0.001Year Op Area Qtr Szst Targ … Year*Area 124 11104.2781 430.07 2.2% < 0.001Year Op Area Qtr Szst Targ … Year*Op 119 10854.8583 249.42 1.3% < 0.001Year Op Area Qtr Szst Targ … Op*Area 30 10802.4636 52.39 0.3% 0.007Year Op Area Qtr Szst Targ … Year*Qtr 65 10684.347 118.12 0.6% < 0.001Year Op Area Qtr Szst Targ … Area*Qtr 18 10532.3656 151.98 0.8% < 0.001Year Op Area Qtr Szst Targ … Op*Qtr 18 10481.5335 50.83 0.3% < 0.001Year Op Area Qtr Szst Targ … Year*Szst 40 10376.3951 105.14 0.5% < 0.001Year Op Area Qtr Szst Targ … Area*Szst 12 10308.6029 67.79 0.3% < 0.001Year Op Area Qtr Szst Targ … Op*Szst 12 10281.3959 27.21 0.1% 0.007Year Op Area Qtr Szst Targ … Qtr*Szst 6 10246.703 34.69 0.2% < 0.001Year Op Area Qtr Szst Targ … Year*Targ 63 10154.9912 91.71 0.5% 0.011Year Op Area Qtr Szst Targ … Area*Targ 18 9880.34319 274.65 1.4% < 0.001Year Op Area Qtr Szst Targ … Op*Targ 18 9806.2918 74.05 0.4% < 0.001Year Op Area Qtr Szst Targ … Qtr*Targ 9 9792.46276 13.83 0.1% 0.129Year Op Area Qtr Szst Targ … Szst*Targ 6 9788.27555 4.19 0.0% 0.651

Model factors proportion positives d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 11056.624Year 22 10651.099 405.53 7% < 0.001Year Op 6 8611.360 2039.74 34% < 0.001Year Op Area 6 7777.176 834.18 14% < 0.001Year Op Area Qtr 3 7448.568 328.61 5% < 0.001Year Op Area Qtr Szst 2 7367.853 80.72 1% < 0.001Year Op Area Qtr Szst Targ 3 5323.522 2044.33 34% < 0.001Year Op Area Qtr Szst Targ Year*Szst 41 5247.925 75.60 1% < 0.001Year Op Area Qtr Szst Targ Op*Area 30 5212.377 111.14 2% < 0.001Year Op Area Qtr Szst Targ Year*Targ 66 5199.032 124.49 2% < 0.001Year Op Area Qtr Szst Targ Year*Qtr 65 5153.386 170.14 3% < 0.001Year Op Area Qtr Szst Targ Area*Qtr 18 5073.954 249.57 4% < 0.001Year Op Area Qtr Szst Targ Year*Op 119 5038.573 284.95 5% < 0.001Year Op Area Qtr Szst Targ Year*Area 124 5019.284 304.24 5% < 0.001 Table 6. Analysis of mixed model formulations for swordfish catch rates combined-sex and age groups from the U.S. pelagic longline fishery. Likelihood ratio tests the difference of –2 REM log likelihood between two nested models. * indicates the final delta mixed model.

Swordfish Combined Age3-10+ GLMixed Model-2 REM

Log likelihood

Akaike's Information

Criterion

Schwartz's Bayesian Criterion

Deviance Dispersion scale

Proportion Positives Year Area OP Qtr Target 27528.1 27530.1 27536.7 5560.37 1.255Year Area OP Qtr Target Year*OP 27425.7 27429.7 27435.7 102.4 0.0000 5369.77 1.211

* Year Area OP Qtr Target Year*OP Area*Qtr 27242.1 27248.1 27257.1 183.6 0.0000 5113.83 1.145

Positives catch ratesYear Area OP Szst Targ 45634.1 45636.1 45643.9Year Area OP Szst Targ Year*Area 45268.6 45272.6 45278.7 365.5 0.0000

* Year Area OP Szst Targ Year*Area Year*OP 45094.5 45100.5 45109.6 174.1 0.0000

Likelihood Ratio Test

1454

Table 7. Deviance analysis table of explanatory variables in the delta lognormal model for swordfish biomass (pounds dressed weight/ thousand hooks) from the U.S. pelagic longline fishery. Swordfish ( = 13 kg dressed weigth) biomass CPUE Index

Model factors positive catch rates values d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 53745.4038Year 21 50305.7216 3439.68 8.8% < 0.001Year Area 7 33310.4251 16995.30 43.7% < 0.001Year Area Op 6 29076.93 4233.50 10.9% < 0.001Year Area Op Targ 3 17089.4182 11987.51 30.8% < 0.001Year Area Op Targ Qtr 3 16976.6222 112.80 0.3% < 0.001Year Area Op Targ Qtr Mngarea2 1 16924.3984 52.22 0.1% < 0.001Year Area Op Targ Qtr Mngarea2 ... Year*Area 130 16376.0884 548.31 1.4% < 0.001Year Area Op Targ Qtr Mngarea2 ... Year*Op 118 16105.2789 270.81 0.7% < 0.001Year Area Op Targ Qtr Mngarea2 ... Area*Op 34 15986.9323 118.35 0.3% < 0.001Year Area Op Targ Qtr Mngarea2 ... Year*Qtr 63 15783.0876 203.84 0.5% < 0.001Year Area Op Targ Qtr Mngarea2 ... Area*Qtr 21 15712.6861 70.40 0.2% < 0.001Year Area Op Targ Qtr Mngarea2 ... Op*Qtr 18 15655.2056 57.48 0.1% < 0.001Year Area Op Targ Qtr Mngarea2 ... Year*Targ 63 15453.0241 202.18 0.5% < 0.001Year Area Op Targ Qtr Mngarea2 ... Area*Targ 21 15171.3539 281.67 0.7% < 0.001Year Area Op Targ Qtr Mngarea2 ... Op*Targ 18 15029.2758 142.08 0.4% < 0.001Year Area Op Targ Qtr Mngarea2 ... Targ*Qtr 9 14967.8605 61.42 0.2% < 0.001Year Area Op Targ Qtr Mngarea2 ... Year*Mngarea2 7 14948.2539 19.61 0.1% 0.006Year Area Op Targ Qtr Mngarea2 ... Area*Mngarea2 7 14855.8478 92.41 0.2% < 0.001Year Area Op Targ Qtr Mngarea2 ... Op*Mngarea2 6 14848.3665 7.48 0.0% 0.279Year Area Op Targ Qtr Mngarea2 ... Qtr*Mngarea2 3 14844.2479 4.12 0.0% 0.249Year Area Op Targ Qtr Mngarea2 ... Targ*Mngarea2 3 14833.728 10.52 0.0% 0.015

Model factors proportion positives d.f.Residual deviance

Change in deviance

% of total deviance p

1 1 10011.483Year 21 9505.526 505.96 6% < 0.001Year Area 7 6541.351 2964.18 36% < 0.001Year Area Op 6 6260.795 280.56 3% < 0.001Year Area Op Targ 3 2325.139 3935.66 48% < 0.001Year Area Op Targ Qtr 3 2200.592 124.55 2% < 0.001Year Area Op Targ Qtr Mngarea2 1 2189.144 11.45 0% < 0.001Year Area Op Targ Qtr Mngarea2 Year*Mngarea2 7 2164.163 24.98 0% < 0.001Year Area Op Targ Qtr Mngarea2 Year*Targ 63 2069.861 119.28 1% < 0.001Year Area Op Targ Qtr Mngarea2 Year*Qtr 63 2039.266 149.88 2% < 0.001Year Area Op Targ Qtr Mngarea2 Area*Qtr 21 1886.841 302.30 4% < 0.001Year Area Op Targ Qtr Mngarea2 Year*Area 130 1882.101 307.04 4% < 0.001Year Area Op Targ Qtr Mngarea2 Year*Op 118 1864.677 324.47 4% < 0.001 Table 8. Analysis of mixed model formulations for biomass swordfish catch rates (> 33 lbs dressed wgt/ thousand hooks) from the U.S. pelagic longline fishery. Likelihood ratio tests the difference of –2 REM log likelihood between two nested models. * indicates the final delta mixed model.

Swordfish Combined Age3-10+ GLMixed Model-2 REM

Log likelihood

Akaike's Information

Criterion

Schwartz's Bayesian Criterion

Deviance Dispersion scale

Proportion Positives Year Area OP Qtr Target 27528.1 27530.1 27536.7 5560.37 1.255Year Area OP Qtr Target Year*OP 27425.7 27429.7 27435.7 102.4 0.0000 5369.77 1.211

* Year Area OP Qtr Target Year*OP Area*Qtr 27242.1 27248.1 27257.1 183.6 0.0000 5113.83 1.145

Positives catch ratesYear Area OP Szst Targ 45634.1 45636.1 45643.9Year Area OP Szst Targ Year*Area 45268.6 45272.6 45278.7 365.5 0.0000

* Year Area OP Szst Targ Year*Area Year*OP 45094.5 45100.5 45109.6 174.1 0.0000

Likelihood Ratio Test

1455

Table 9. Nominal and standard swordfish CPUE by age-sex (females) from the weight-out data (fish/1000 hooks).

Age Year N obs NominaL CPUE

Stdard Low Upp coeff var

std error

0 1981 36 1.550 1.583 0.780 3.212 36.5% 0.916

0 1982 89 1.667 1.293 0.760 2.200 27.0% 0.554

0 1983 128 1.327 1.189 0.777 1.819 21.5% 0.405

0 1984 162 0.625 0.604 0.376 0.970 24.0% 0.230

0 1985 168 0.891 0.824 0.548 1.237 20.5% 0.268

0 1986 320 1.236 1.465 1.055 2.035 16.5% 0.384

0 1987 729 0.844 0.954 0.695 1.310 15.9% 0.241

0 1988 930 0.742 0.736 0.534 1.014 16.1% 0.188

0 1989 728 0.541 0.655 0.474 0.906 16.3% 0.169

0 1990 793 0.577 0.696 0.502 0.966 16.5% 0.182

1 1981 36 1.318 1.004 0.563 1.791 29.6% 0.937

1 1982 89 1.171 0.851 0.517 1.400 25.3% 0.680

1 1983 128 1.309 1.011 0.700 1.461 18.5% 0.593

1 1984 162 0.806 0.807 0.568 1.148 17.7% 0.452

1 1985 168 0.703 0.735 0.517 1.043 17.7% 0.410

1 1986 320 1.012 1.296 0.952 1.763 15.5% 0.635

1 1987 729 1.007 1.117 0.838 1.489 14.4% 0.510

1 1988 930 0.987 1.131 0.851 1.505 14.3% 0.512

1 1989 728 0.936 1.066 0.800 1.421 14.4% 0.487

1 1990 793 0.751 0.981 0.736 1.309 14.5% 0.449

2 1981 36 1.898 1.383 0.752 2.540 31.1% 1.168

2 1982 89 0.902 0.628 0.385 1.024 24.8% 0.423

2 1983 128 1.019 0.853 0.589 1.236 18.7% 0.433

2 1984 162 0.799 0.847 0.595 1.207 17.8% 0.410

2 1985 168 0.987 0.958 0.684 1.342 17.0% 0.441

2 1986 320 0.868 1.081 0.794 1.472 15.5% 0.456

2 1987 729 0.916 1.122 0.842 1.495 14.4% 0.439

2 1988 930 1.029 1.118 0.841 1.488 14.3% 0.435

2 1989 728 0.785 0.944 0.707 1.261 14.5% 0.373

2 1990 793 0.797 1.064 0.798 1.420 14.5% 0.418

3 1981 36 3.120 1.586 0.864 2.909 31.1% 0.814

3 1982 89 1.759 0.898 0.541 1.491 25.7% 0.382

3 1983 128 1.398 0.812 0.545 1.212 20.2% 0.271

3 1984 162 1.272 0.958 0.661 1.390 18.8% 0.297

3 1985 168 1.784 1.171 0.821 1.670 17.9% 0.347

3 1986 320 1.364 1.045 0.753 1.449 16.5% 0.284

3 1987 729 1.171 1.064 0.792 1.429 14.8% 0.261

3 1988 930 1.296 1.068 0.796 1.432 14.8% 0.260

3 1989 728 1.048 0.994 0.738 1.339 15.0% 0.246

3 1990 793 0.904 0.914 0.675 1.237 15.2% 0.230

3 1991 1219 0.845 0.959 0.714 1.288 14.8% 0.235

3 1992 1760 0.638 0.788 0.589 1.053 14.6% 0.190

3 1993 2006 0.556 0.721 0.539 0.963 14.6% 0.174

3 1994 2117 0.605 0.728 0.545 0.974 14.6% 0.176

3 1995 2195 0.559 0.826 0.619 1.102 14.5% 0.198

3 1996 1789 0.555 0.829 0.618 1.112 14.7% 0.202

3 1997 1844 0.620 0.955 0.715 1.275 14.5% 0.230

3 1998 1555 0.583 0.876 0.653 1.175 14.8% 0.214

3 1999 1446 0.612 1.048 0.784 1.402 14.6% 0.253

3 2000 1512 0.584 0.997 0.746 1.332 14.6% 0.240

3 2001 1227 0.530 1.179 0.880 1.581 14.7% 0.287

3 2002 1113 0.584 1.316 0.978 1.771 14.9% 0.325

3 2003 1004 0.612 1.268 0.943 1.706 14.9% 0.313

4 1981 36 2.771 1.489 0.724 3.064 37.3% 0.519

4 1982 89 2.263 1.149 0.668 1.975 27.6% 0.297

4 1983 128 1.485 0.919 0.585 1.443 22.9% 0.197

4 1984 162 1.418 0.991 0.652 1.506 21.2% 0.196

4 1985 168 2.106 1.355 0.933 1.967 18.8% 0.238

4 1986 320 1.644 1.191 0.840 1.687 17.5% 0.195

4 1987 729 1.129 1.008 0.735 1.382 15.9% 0.150

4 1988 930 1.153 0.958 0.703 1.307 15.6% 0.140

4 1989 728 0.998 0.911 0.661 1.256 16.1% 0.138

4 1990 793 0.841 0.861 0.623 1.190 16.3% 0.131

4 1991 1219 0.719 0.882 0.645 1.206 15.7% 0.130

4 1992 1760 0.635 0.816 0.601 1.108 15.4% 0.117

4 1993 2006 0.556 0.741 0.546 1.006 15.3% 0.106

4 1994 2117 0.543 0.719 0.528 0.979 15.5% 0.105

4 1995 2195 0.473 0.744 0.548 1.011 15.4% 0.107

4 1996 1789 0.460 0.735 0.537 1.005 15.7% 0.108

4 1997 1844 0.570 0.921 0.680 1.246 15.2% 0.131

4 1998 1555 0.504 0.859 0.629 1.173 15.7% 0.126

4 1999 1446 0.603 1.097 0.812 1.483 15.2% 0.156

4 2000 1512 0.502 0.939 0.692 1.274 15.3% 0.135

4 2001 1227 0.494 1.176 0.867 1.597 15.4% 0.169

4 2002 1113 0.593 1.439 1.058 1.956 15.5% 0.208

4 2003 1004 0.539 1.099 0.804 1.502 15.7% 0.162

5 1981 36 3.161 1.825 0.887 3.755 37.3% 0.412

5 1982 89 3.213 1.675 0.971 2.889 27.8% 0.282

5 1983 128 1.795 1.050 0.650 1.696 24.4% 0.155

5 1984 162 1.399 1.038 0.647 1.665 23.9% 0.151

5 1985 168 2.024 1.370 0.892 2.102 21.7% 0.180

5 1986 320 1.344 1.083 0.734 1.599 19.7% 0.129

5 1987 729 1.051 0.936 0.663 1.322 17.4% 0.098

5 1988 930 0.945 0.787 0.556 1.113 17.5% 0.083

5 1989 728 0.934 0.869 0.612 1.235 17.7% 0.093

5 1990 793 0.732 0.782 0.545 1.122 18.2% 0.086

5 1991 1219 0.786 1.002 0.721 1.393 16.6% 0.101

5 1992 1760 0.527 0.746 0.532 1.044 17.0% 0.077

5 1993 2006 0.513 0.714 0.512 0.994 16.7% 0.072

5 1994 2117 0.442 0.633 0.450 0.892 17.2% 0.066

5 1995 2195 0.367 0.641 0.454 0.905 17.4% 0.067

5 1996 1789 0.417 0.605 0.423 0.863 18.0% 0.066

5 1997 1844 0.439 0.789 0.563 1.105 17.0% 0.081

5 1998 1555 0.458 0.800 0.567 1.129 17.3% 0.084

5 1999 1446 0.558 1.068 0.773 1.474 16.2% 0.105

5 2000 1512 0.483 0.983 0.712 1.357 16.2% 0.097

5 2001 1227 0.463 1.165 0.841 1.615 16.4% 0.116

5 2002 1113 0.529 1.460 1.057 2.017 16.2% 0.144

5 2003 1004 0.419 0.980 0.694 1.384 17.4% 0.103

6 1981 36 4.701 2.688 1.265 5.713 39.1% 0.441

6 1982 89 2.359 1.280 0.663 2.472 33.8% 0.182

6 1983 128 1.996 1.259 0.735 2.154 27.4% 0.145

6 1984 162 1.680 1.265 0.775 2.065 24.9% 0.132

6 1985 168 1.814 1.262 0.793 2.009 23.6% 0.125

6 1986 320 1.053 0.935 0.588 1.486 23.5% 0.092

6 1987 729 1.025 0.934 0.637 1.368 19.3% 0.076

6 1988 930 0.996 0.787 0.536 1.155 19.4% 0.064

6 1989 728 0.849 0.800 0.539 1.187 20.0% 0.067

6 1990 793 0.766 0.778 0.520 1.163 20.3% 0.066

6 1991 1219 0.615 0.813 0.556 1.189 19.2% 0.066

6 1992 1760 0.495 0.783 0.544 1.127 18.4% 0.060

6 1993 2006 0.462 0.675 0.464 0.980 18.8% 0.053

6 1994 2117 0.369 0.596 0.405 0.878 19.5% 0.049

6 1995 2195 0.390 0.654 0.448 0.955 19.1% 0.052

6 1996 1789 0.341 0.540 0.353 0.825 21.4% 0.049

6 1997 1844 0.342 0.669 0.453 0.988 19.7% 0.055

6 1998 1555 0.356 0.673 0.448 1.009 20.5% 0.058

6 1999 1446 0.539 0.982 0.688 1.403 18.0% 0.074

6 2000 1512 0.494 1.003 0.706 1.425 17.7% 0.075

6 2001 1227 0.342 1.027 0.712 1.481 18.5% 0.080

6 2002 1113 0.572 1.565 1.114 2.199 17.1% 0.113

6 2003 1004 0.442 1.033 0.710 1.504 18.9% 0.082

7 1981 36 3.010 1.964 0.677 5.700 57.3% 0.276

7 1982 89 3.221 1.667 0.804 3.457 37.7% 0.154

7 1983 128 2.847 1.623 0.895 2.944 30.4% 0.121

7 1984 162 1.964 1.452 0.827 2.550 28.7% 0.102

7 1985 168 1.843 1.327 0.749 2.351 29.2% 0.095

7 1986 320 0.894 0.772 0.419 1.424 31.3% 0.059

7 1987 729 0.926 0.855 0.527 1.387 24.6% 0.051

7 1988 930 0.871 0.705 0.430 1.156 25.1% 0.043

7 1989 728 0.905 0.786 0.479 1.289 25.1% 0.048

7 1990 793 0.701 0.728 0.436 1.216 26.1% 0.047

7 1991 1219 0.592 0.776 0.486 1.241 23.8% 0.045

7 1992 1760 0.528 0.758 0.480 1.196 23.1% 0.043

7 1993 2006 0.487 0.655 0.413 1.038 23.4% 0.037

7 1994 2117 0.384 0.620 0.386 0.995 24.0% 0.036

7 1995 2195 0.392 0.715 0.455 1.123 22.9% 0.040

7 1996 1789 0.317 0.478 0.269 0.849 29.3% 0.034

7 1997 1844 0.357 0.620 0.373 1.032 25.8% 0.039

7 1998 1555 0.376 0.617 0.364 1.045 26.8% 0.041

7 1999 1446 0.552 1.092 0.723 1.650 20.8% 0.056

7 2000 1512 0.498 1.090 0.727 1.635 20.5% 0.055

7 2001 1227 0.375 1.139 0.741 1.750 21.7% 0.061

7 2002 1113 0.518 1.513 1.016 2.253 20.1% 0.075

7 2003 1004 0.442 1.048 0.664 1.652 23.1% 0.059

8 1981 36 4.415 3.244 1.072 9.818 59.9% 0.294

8 1982 89 3.441 1.974 0.773 5.045 49.6% 0.148

8 1983 128 2.768 1.638 0.757 3.540 40.0% 0.099

8 1984 162 1.789 1.285 0.597 2.768 39.8% 0.077

8 1985 168 1.696 1.048 0.474 2.319 41.3% 0.066

8 1986 320 0.939 0.901 0.431 1.882 38.1% 0.052

8 1987 729 0.819 0.797 0.419 1.517 33.0% 0.040

8 1988 930 0.826 0.657 0.338 1.276 34.1% 0.034

8 1989 728 0.650 0.585 0.279 1.223 38.2% 0.034

8 1990 793 0.574 0.599 0.288 1.246 37.8% 0.034

8 1991 1219 0.533 0.753 0.414 1.369 30.6% 0.035

8 1992 1760 0.420 0.715 0.396 1.292 30.2% 0.033

8 1993 2006 0.380 0.582 0.311 1.092 32.2% 0.028

8 1994 2117 0.372 0.650 0.358 1.180 30.5% 0.030

8 1995 2195 0.403 0.696 0.386 1.256 30.2% 0.032

8 1996 1789 0.351 0.393 0.169 0.915 44.2% 0.026

8 1997 1844 0.300 0.506 0.241 1.060 38.3% 0.029

8 1998 1555 0.267 0.527 0.249 1.116 38.8% 0.031

8 1999 1446 0.509 1.021 0.603 1.731 26.8% 0.041

8 2000 1512 0.425 0.979 0.576 1.664 27.0% 0.040

8 2001 1227 0.334 1.065 0.608 1.864 28.6% 0.046

8 2002 1113 0.408 1.471 0.901 2.402 24.9% 0.055

8 2003 1004 0.381 0.913 0.500 1.668 30.8% 0.043

9 1981 36 5.545 4.090 1.149 14.556

70.4% 0.283

9 1982 89 3.301 2.302 0.841 6.303 53.7% 0.122

9 1983 128 3.105 2.032 0.849 4.863 45.8% 0.092

9 1984 162 1.438 1.097 0.369 3.258 58.7% 0.063

9 1985 168 1.070 0.687 0.186 2.533 72.9% 0.049

9 1986 320 0.801 0.713 0.248 2.051 56.7% 0.040

9 1987 729 0.869 0.870 0.396 1.913 41.0% 0.035

9 1988 930 0.831 0.676 0.300 1.525 42.4% 0.028

9 1989 728 0.793 0.707 0.299 1.672 45.1% 0.031

9 1990 793 0.601 0.601 0.233 1.552 50.2% 0.030

9 1991 1219 0.498 0.693 0.312 1.542 41.6% 0.028

9 1992 1760 0.551 0.700 0.324 1.511 39.9% 0.027

9 1993 2006 0.350 0.501 0.204 1.229 47.3% 0.023

1456

9 1994 2117 0.241 0.543 0.233 1.267 44.3% 0.024

9 1995 2195 0.259 0.515 0.206 1.289 48.4% 0.025

9 1996 1789 0.253 0.381 0.127 1.136 59.0% 0.022

9 1997 1844 0.409 0.534 0.208 1.368 49.8% 0.026

9 1998 1555 0.215 0.377 0.120 1.189 62.4% 0.023

9 1999 1446 0.331 0.793 0.361 1.743 40.9% 0.032

9 2000 1512 0.440 1.006 0.512 1.980 34.8% 0.034

9 2001 1227 0.292 0.900 0.415 1.952 40.2% 0.036

9 2002 1113 0.399 1.295 0.678 2.475 33.3% 0.042

9 2003 1004 0.408 0.987 0.476 2.048 37.7% 0.037

10 1981 36 5.107 3.737 1.349 10.356

54.5% 0.327

10 1982 89 3.283 2.130 0.900 5.041 45.2% 0.155

10 1983 128 3.730 2.230 1.169 4.254 33.2% 0.119

10 1984 162 1.486 1.206 0.548 2.650 41.0% 0.079

10 1985 168 1.562 1.191 0.571 2.488 38.1% 0.073

10 1986 320 0.884 0.914 0.441 1.898 37.8% 0.056

10 1987 729 0.947 0.945 0.529 1.687 29.6% 0.045

10 1988 930 0.833 0.771 0.426 1.398 30.4% 0.038

10 1989 728 0.803 0.752 0.397 1.422 32.7% 0.040

10 1990 793 0.632 0.716 0.369 1.389 34.0% 0.039

10 1991 1219 0.554 0.789 0.441 1.412 29.7% 0.038

10 1992 1760 0.384 0.680 0.374 1.236 30.5% 0.033

10 1993 2006 0.306 0.520 0.273 0.989 33.0% 0.028

10 1994 2117 0.277 0.558 0.294 1.062 33.0% 0.030

10 1995 2195 0.269 0.645 0.350 1.189 31.3% 0.032

10 1996 1789 0.272 0.400 0.180 0.887 41.4% 0.027

10 1997 1844 0.212 0.336 0.137 0.824 47.3% 0.026

10 1998 1555 0.197 0.403 0.174 0.933 43.8% 0.028

10 1999 1446 0.235 0.647 0.332 1.263 34.3% 0.036

10 2000 1512 0.270 0.820 0.465 1.447 29.0% 0.038

10 2001 1227 0.198 0.743 0.387 1.428 33.6% 0.040

10 2002 1113 0.300 1.100 0.631 1.919 28.3% 0.050

10 2003 1004 0.259 0.765 0.402 1.456 33.1% 0.041

Table 10. Nominal and standard swordfish CPUE by age-sex (males) from the weight-out data (fish/1000 hooks). Age Year N obs Nominal

CPUE Standard CPUE

Low CI 95%

Upp CI 95%

Coeff Var

Std Error

0 1981 36 0.708 0.914 0.432 1.930 38.7% 0.496

0 1982 89 1.627 1.359 0.801 2.305 26.9% 0.512

0 1983 128 0.942 0.903 0.512 1.594 29.0% 0.367

0 1984 162 0.770 0.826 0.500 1.365 25.5% 0.295

0 1985 168 0.795 0.606 0.376 0.977 24.2% 0.206

0 1986 320 1.342 1.370 1.000 1.876 15.8% 0.304

0 1987 729 0.948 1.029 0.770 1.376 14.6% 0.210

0 1988 930 1.050 1.011 0.765 1.337 14.0% 0.199

0 1989 728 0.842 1.083 0.820 1.431 14.0% 0.212

0 1990 793 0.975 0.899 0.676 1.197 14.4% 0.181

1 1981 36 1.127 0.986 0.460 2.113 39.5% 0.737

1 1982 89 1.064 1.048 0.615 1.786 27.2% 0.538

1 1983 128 0.731 0.628 0.371 1.063 26.8% 0.318

1 1984 162 0.987 0.921 0.622 1.364 19.8% 0.346

1 1985 168 0.652 0.526 0.356 0.779 19.8% 0.197

1 1986 320 1.082 1.137 0.829 1.559 15.9% 0.342

1 1987 729 1.020 0.987 0.749 1.301 13.9% 0.259

1 1988 930 1.257 1.341 1.032 1.742 13.1% 0.333

1 1989 728 1.165 1.382 1.060 1.802 13.3% 0.348

1 1990 793 0.913 1.043 0.797 1.366 13.5% 0.267

2 1981 36 1.667 1.083 0.505 2.322 39.6% 0.599

2 1982 89 1.370 1.319 0.782 2.226 26.6% 0.490

2 1983 128 0.767 0.672 0.397 1.138 26.8% 0.251

2 1984 162 0.879 0.984 0.659 1.469 20.2% 0.278

2 1985 168 0.810 0.745 0.504 1.101 19.7% 0.205

2 1986 320 0.900 0.980 0.709 1.354 16.3% 0.223

2 1987 729 0.858 0.934 0.703 1.243 14.3% 0.187

2 1988 930 1.124 1.171 0.896 1.530 13.4% 0.220

2 1989 728 0.826 1.087 0.826 1.431 13.8% 0.209

2 1990 793 0.799 1.024 0.778 1.348 13.8% 0.197

3 1981 36 1.920 1.194 0.421 3.389 55.9% 0.518

3 1982 89 1.850 1.350 0.764 2.384 29.0% 0.304

3 1983 128 0.667 0.478 0.217 1.049 40.9% 0.152

3 1984 162 1.152 0.968 0.620 1.513 22.6% 0.170

3 1985 168 1.490 0.905 0.590 1.388 21.6% 0.152

3 1986 320 1.163 0.967 0.652 1.434 19.9% 0.149

3 1987 729 1.003 0.825 0.592 1.150 16.7% 0.107

3 1988 930 1.295 0.966 0.712 1.310 15.3% 0.115

3 1989 728 1.093 0.968 0.710 1.319 15.6% 0.117

3 1990 793 0.930 0.857 0.622 1.180 16.1% 0.107

3 1991 1219 0.882 0.945 0.700 1.275 15.1% 0.111

3 1992 1760 0.747 0.895 0.674 1.189 14.2% 0.099

3 1993 2006 0.690 0.833 0.630 1.102 14.1% 0.091

3 1994 2117 0.657 0.773 0.581 1.030 14.4% 0.086

3 1995 2195 0.765 0.849 0.643 1.123 14.0% 0.092

3 1996 1789 0.883 1.002 0.761 1.320 13.8% 0.108

3 1997 1844 0.947 1.140 0.867 1.500 13.8% 0.122

3 1998 1555 0.984 1.186 0.903 1.558 13.7% 0.126

3 1999 1446 0.935 1.342 1.018 1.769 13.9% 0.145

3 2000 1512 0.717 1.151 0.867 1.529 14.3% 0.127

3 2001 1227 0.552 1.036 0.779 1.380 14.4% 0.116

3 2002 1113 0.683 1.099 0.822 1.470 14.6% 0.125

3 2003 1004 0.995 1.267 0.954 1.682 14.2% 0.140

4 1981 36 3.231 2.202 0.845 5.738 50.8% 0.566

4 1982 89 2.530 1.679 0.921 3.060 30.7% 0.261

4 1983 128 0.610 0.522 0.226 1.208 43.9% 0.116

4 1984 162 1.451 1.267 0.785 2.046 24.3% 0.156

4 1985 168 1.476 1.043 0.652 1.669 23.8% 0.126

4 1986 320 1.180 1.023 0.664 1.577 21.9% 0.113

4 1987 729 0.937 0.807 0.559 1.164 18.5% 0.075

4 1988 930 1.134 0.881 0.631 1.230 16.8% 0.075

4 1989 728 0.920 0.911 0.646 1.285 17.3% 0.080

4 1990 793 0.694 0.718 0.496 1.039 18.6% 0.068

4 1991 1219 0.795 0.865 0.620 1.206 16.7% 0.073

4 1992 1760 0.639 0.835 0.611 1.140 15.7% 0.066

4 1993 2006 0.629 0.769 0.566 1.045 15.4% 0.060

4 1994 2117 0.654 0.761 0.559 1.036 15.5% 0.060

4 1995 2195 0.614 0.734 0.539 1.000 15.6% 0.058

4 1996 1789 0.726 0.823 0.607 1.116 15.3% 0.064

4 1997 1844 0.790 1.008 0.748 1.359 15.0% 0.077

4 1998 1555 0.774 1.031 0.764 1.390 15.0% 0.078

4 1999 1446 0.717 1.057 0.776 1.438 15.5% 0.083

4 2000 1512 0.681 1.052 0.774 1.430 15.4% 0.082

4 2001 1227 0.490 0.980 0.716 1.340 15.8% 0.078

4 2002 1113 0.560 0.968 0.702 1.336 16.2% 0.080

4 2003 1004 0.769 1.065 0.779 1.455 15.7% 0.085

5 1981 36 3.574 2.678 1.322 5.424 36.4% 1.375

5 1982 89 4.664 3.197 1.947 5.250 25.2% 1.136

5 1983 128 1.010 0.842 0.488 1.453 27.8% 0.330

5 1984 162 1.533 1.240 0.826 1.864 20.6% 0.360

5 1985 168 1.685 1.117 0.739 1.686 20.8% 0.328

5 1986 320 1.040 0.867 0.598 1.257 18.7% 0.229

5 1987 729 0.677 0.632 0.458 0.873 16.2% 0.145

5 1988 930 0.824 0.628 0.466 0.845 15.0% 0.133

5 1989 728 0.886 0.909 0.675 1.223 14.9% 0.191

5 1990 793 0.637 0.675 0.498 0.915 15.3% 0.145

5 1991 1219 0.650 0.783 0.587 1.045 14.5% 0.160

5 1992 1760 0.472 0.729 0.555 0.956 13.7% 0.140

5 1993 2006 0.495 0.657 0.501 0.862 13.6% 0.126

5 1994 2117 0.413 0.555 0.421 0.732 13.9% 0.109

5 1995 2195 0.451 0.630 0.480 0.826 13.6% 0.121

5 1996 1789 0.473 0.631 0.480 0.829 13.7% 0.122

5 1997 1844 0.555 0.813 0.622 1.063 13.5% 0.154

5 1998 1555 0.501 0.812 0.620 1.064 13.6% 0.155

5 1999 1446 0.572 0.958 0.732 1.255 13.5% 0.183

5 2000 1512 0.444 0.886 0.674 1.166 13.8% 0.172

5 2001 1227 0.421 0.882 0.669 1.164 13.9% 0.173

5 2002 1113 0.453 0.937 0.708 1.240 14.1% 0.186

5 2003 1004 0.571 0.940 0.712 1.242 14.0% 0.186

Table 11. Nominal and standard swordfish CPUE combined sex and age groups (Age 0-2) from the weight-out data (fish/1000 hooks). Year

Numb obs

Nominal CPUE

Standard CPUE

Low CI 95%

Upp CI 95%

Coeff Var

Std Error

1981 36 30.14 17.55 11.85 26.01 19.8% 3.483

1982 89 27.17 11.57 8.16 16.39 17.6% 2.032

1983 128 23.03 12.39 9.35 16.42 14.1% 1.752

1984 162 17.18 13.09 10.11 16.95 13.0% 1.700

1985 168 17.37 14.17 11.17 17.96 11.9% 1.688

1986 320 22.81 18.30 14.63 22.87 11.2% 2.049

1987 729 20.18 16.39 13.24 20.29 10.7% 1.754

1988 930 21.72 18.05 14.56 22.38 10.8% 1.947

1989 728 18.16 16.72 13.51 20.70 10.7% 1.787

1990 793 16.89 13.60 11.00 16.80 10.6% 1.442

1457

Table 12. Nominal and standard swordfish CPUE combined sex and age groups (Age 3-10+) from the weight-out data (fish/1000 hooks). Year

Numb obs

Nominal CPUE

Standard CPUE

Low CI 95%

Upp CI 95%

Coeff Var

Std Error

1981

36 19.30 9.78 6.55 14.60 0.202 1.979 198

2 89 18.18 8.34 6.09 11.43 0.158 1.323

1983

128 8.68 4.20 3.20 5.52 0.137 0.575 198

4 162 8.32 3.87 3.00 5.00 0.128 0.498

1985

168 9.77 4.04 3.19 5.12 0.119 0.480 198

6 320 6.50 2.91 2.34 3.63 0.110 0.321

1987

729 4.97 2.74 2.24 3.36 0.102 0.279 198

8 930 5.22 2.77 2.27 3.38 0.099 0.276

1989

728 4.91 2.62 2.15 3.20 0.100 0.263 199

0 793 3.86 2.59 2.12 3.16 0.100 0.259

1991

1219

3.69 3.13 2.58 3.80 0.096 0.302 199

2 176

0 2.89 2.76 2.29 3.33 0.093 0.257

1993

2006

2.75 2.49 2.07 3.00 0.093 0.231 199

4 211

7 2.45 2.27 1.88 2.74 0.094 0.213

1995

2195

2.39 2.30 1.91 2.76 0.093 0.214 199

6 178

9 2.45 2.05 1.70 2.47 0.094 0.193

1997

1844

2.77 2.54 2.11 3.07 0.094 0.238 199

8 155

5 2.59 2.56 2.12 3.09 0.094 0.241

1999

1446

3.08 3.21 2.66 3.88 0.094 0.303 200

0 151

2 2.64 3.22 2.66 3.89 0.095 0.306

2001

1227

2.30 3.06 2.53 3.70 0.095 0.291 200

2 111

3 2.77 3.45 2.85 4.17 0.096 0.330

2003

1004

2.84 3.16 2.61 3.83 0.096 0.304

Table 13. Nominal and standard swordfish biomass CPUE (pounds dressed weight/ thousand hooks) from the weight-out data.

Year N Obs Nominal CPUE

Standard

Low Upp coeff var

std error

1982 90 3762.0

1769.8

1256.1

2493.6 17.3% 305.6

1983 128 2197.2

1170.2

868.2 1577.2 15.0% 175.6

1984 162 1816.2

1003.3

755.1 1333.0 14.3% 143.3

1985 168 2091.6

1047.8

805.9 1362.3 13.2% 138.1

1986 320 1569.2

919.8 721.1 1173.3 12.2% 112.4

1987 729 1305.0

772.7 612.6 974.6 11.6% 90.0

1988 930 1374.9

752.8 598.2 947.3 11.5% 86.8

1989 731 1191.5

686.4 545.1 864.2 11.6% 79.3

1990 796 1066.4

683.8 543.9 859.6 11.5% 78.5

1991 1221 897.6 664.1 530.3 831.8 11.3% 75.0

1992 1768 719.9 611.4 490.3 762.5 11.1% 67.7

1993 2014 669.8 536.9 430.8 669.0 11.0% 59.2

1994 2126 637.6 501.8 402.5 625.6 11.1% 55.5

1995 2252 612.1 523.4 420.0 652.2 11.0% 57.8

1996 1382 473.7 385.6 308.3 482.2 11.2% 43.3

1997 1543 527.8 461.6 369.3 577.0 11.2% 51.7

1998 1198 500.7 454.2 362.7 568.7 11.3% 51.2

1999 1077 464.9 532.7 423.3 670.3 11.5% 61.4

2000 1113 408.2 531.0 421.1 669.7 11.6% 61.8

2001 1113 445.1 461.5 367.5 579.4 11.4% 52.7

2002 1059 560.0 523.4 417.1 656.7 11.4% 59.6

2003 926 619.9 516.5 410.2 650.4 11.6% 59.7

1458

Figure 1. Geographic area classification for the US Pelagic longline fishery: CAR Caribbean, GOM Gulf of Mexico, FEC Florida east coast, SAB south Atlantic bight, MAB mid Atlantic bight, NEC north east coastal, NED north east distant waters, SNA Sargasso area, and OFS offshore waters. Shaded areas represent the current time-area closures affecting the pelagic longline fisheries. Permanent closures: the DeSoto area in the Gulf of Mexico, and the Florida east coast area. Time-area closures: the Charleston Bump in the SAB area closed Feb-Apr, the Bluefin tuna protected area in the MAB and NEC areas closed Jun, and the Grand Banks in the NED area closed from Oct 10/00 to Apr 9/01.

100°

-90°

-80°

-70°

-60°

-50°

-40°

-30°

-20°

-10° 0°

-10°

0°

10°

20°

30°

40°

50°

60°

CAR

GOMFEC

SAB

MAB

NEC

NED

SNASNA

OFS

OFS

1459

-3.25 -1.41 0.44 2.28 4.12

-3.25 -1.41 0.44 2.28 4.12

Nominal logCPUE female swordfish (fish/1000 hooks)

0.00.10.20.30.40.5

0.00.10.20.30.40.5

0.00.10.20.30.40.5

0.00.10.20.30.40.5

age: 0 age: 1 age: 2

age: 3 age: 4 age: 5

age: 6 age: 7 age: 8

age: 9 age: 10

-3.09 -2.20 -1.31 -0.42 0.47 1.36 2.25 3.14 4.03 4.92

-3.09 -2.20 -1.31 -0.42 0.47 1.36 2.25 3.14 4.03 4.92

Nominal logCPUE males swordfish [fish/1000 hooks]

0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.4

0.0

0.1

0.2

0.3

0.4

age: 0 age: 1

age: 2 age: 3

age: 4 age: 5

1.16 2.14 3.13 4.11 5.09 6.07 7.06 8.04 9.02

Nominal logCPUE swordfish [fish>= 13 kg] biomass (lbs/1000 hooks)

0.0

0.1

0.2

0.3

-2.16 -1.26 -0.35 0.56 1.46 2.37 3.28 4.18 5.09

Nominal logCPUE swordfish combined sex Age 0-2 (fish/1000 hooks)

0.0

0.1

0.2

0.3

0.4

-2.75 -1.97 -1.19 -0.41 0.37 1.15 1.93 2.71 3.49 4.27

Nominal logCPUE swordfish combinedsex Ages3-10+ (fish/1000 hooks)

0.0

0.1

0.2

0.3

Figure 2. Swordfish density frequency distributions of positive catch trips (log CPUE) by sex and age (top), combined sex and ages biomass for fish >= 33 lbs (middle left), and combined sex age 2-0 group (middle right) and age 3-10+ (bottom) from the weight out data.

1460

Figure 3. Nominal (solid circles) and standard CPUE for swordfish by age-sex (males) from the U.S. elagic longline fishery 1981-2002. Bars represent upper and lower estimated 95% confidence intervals for the scaled CPUE value. Series are scaled to their corresponding mean for each age class.

Swordfish Males Age 0

0.00

0.50

1.00

1.50

2.00

2.50

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Males Age 1

0.00

0.50

1.00

1.50

2.00

2.50

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Males Age 2

0.00

0.50

1.00

1.50

2.00

2.50

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Males Age 3

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Males Age 4

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Males Age 5

0.00

1.00

2.00

3.00

4.00

5.00

6.00

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

1461

Swordfish Females Age 0

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 1

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 2

0.00

0.50

1.00

1.50

2.00

2.50

3.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 6

0.00

1.00

2.00

3.00

4.00

5.00

6.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 7

0.00

1.00

2.00

3.00

4.00

5.00

6.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 8

0.00

2.00

4.00

6.00

8.00

10.00

12.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Figure 3. (continued).

Swordfish Females Age 3

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 9

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

16.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

1462

Figure 4. Nominal (solid-circles) and standard CPUE for swordfish by age-sex (Females) from the US Pelagic longline fishery 1981-2002. Bars represent upper and lower estimated 95% confidence intervals for the scaled CPUE value. Series are scaled to their corresponding mean for each age class.

Swordfish Females Age 4

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 5

0.00

0.50

1.00

1.50

2.00

2.50

3.00

3.50

4.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

Swordfish Females Age 10

0.00

2.00

4.00

6.00

8.00

10.00

12.00

1980 1985 1990 1995 2000

Scal

ed C

PUE

[fis

h/10

00 h

ooks

]

1463

Figure 5. Nominal and standard swordfish CPUE for combined sex and age groups (0-2 top panel, 3-10+ bottom panel) from the US Pelagic Longline fishery. Series are scaled to their corresponding mean for each age group.

Swordfish Standardized CPUE Combined Age3-10+ Pelagic Longline US Fishery 95% CI

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

1980 1985 1990 1995 2000 2005

Sca

led

CP

UE (f

ish/

1000

hoo

ks)

Swordfish Standardized CPUE Combined Age0-2 Pelagic Longline US Fishery 95% CI

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

1980 1985 1990 1995 2000 2005

Sca

led

CP

UE

(fis

h/10

00 h

ooks

)

1464

0.5

1

1.5

2

2.5

1975 1980 1985 1990 1995 2000

Fem Age 4Fem Age 3Mal Age 4

Mal Age 3

0.5

1

1.5

2

2.5

3

3.5

1970 1975 1980 1985 1990 1995 2000

Fem Age 4Fem Age 5M al Age 4M al Age 5

Swordfish Standardized biomass CPUE Pelagic Longline US Fishery (Non-closure areas only)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

1980 1985 1990 1995 2000 2005

Scal

ed C

PUE

(dre

ssed

wgt

lbs/

1000

hoo

ks)

Figure 6. Nominal and standard biomass CPUE for swordfish (≥ 33 lbs) from the US Pelagic Longline fishery. Bars represent upper and lower 95% confidence intervals.

Figure 7. Comparison of Ages 3, 4 and 5 for females (solid lines) and males (broken lines) index trends adjusted to their correspondent cohort-year group.

1465

-1.000 4.000

-1.000 4.000

-1

4

-1

4

-1

4

-1

4Res

idua

ls p

ositi

ve o

bser

vatio

ns m

odel

fit

age: 0 age: 1 age: 2

age: 3 age: 4 age: 5

age: 6 age: 7 age: 8

age: 9 age: 10

-1.000 4.000

-1.000 4.000 -1.000 4.000

-1

4

-1

4

Res

idua

ls p

ositi

ve o

bser

vatio

ns m

odel

fit

age: 0 age: 1 age: 2

age: 3 age: 4 age: 5

1981 1982 1983 1984 1985 1986 1987 1988 1989 1990

-1

3

Res

idua

ls p

ositi

ve o

bser

vatio

ns

Swordfish Age 0-2 combined sex

19811982

19831984

19851986

19871988

19891990

19911992

19931994

19951996

19971998

19992000

20012002

2003

-2

2

Res

idua

l pos

itive

obs

erva

tions

Swordfish Age 3-10+ Combined sex

Figure 8. Diagnostic plots: Cumulative normalized residual plots (qq-plot) for the positive observations delta lognormal model fit to swordfish catch rates by sex and age: Males (right) ages 0 to 5+, and females (left) ages 0 to 10+.

Figure 9. Diagnostic plots: Cumulative normalized residual plots (qq-plot) for the positive observations delta lognormal model fit to swordfish catch rates Age groups and combined sex (right panel), and box plot distribution of residuals from the positive trips by year for swordfish catch rates by age groups and combined sex (left panel, box size is proportional to the number of observations by year).

-3 0 3

-2

2

Res

idua

ls p

ositi

ve o

bser

vatio

ns m

odel

fit

Swordfish Biomass [fish >= 33 lbs]

-3 0 3

-1

3

Res

idua

ls p

ositi

ve o

bser

vatio

n m

odel

fit

Swordfish Age 0-2 sex combined

-3 0 3

-2

2

Res

idua

ls p

ositi

ve o

bser

vatio

ns m

odel

fit

Swordfish Age 3-10+ sex combined