Statistical inference about the relative efficiency of a new survey protocol, based on paired-tow survey calibration data

Paired-tow calibration studies provide information on changes in survey catchability that may occur because of some necessary change in protocols (e.g., change in vessel or vessel gear) in a fish stock survey. This information is important to ensure the continuity of annual time-series of survey indices of stock size that provide the basis for fish stock assessments. There are several statistical models used to analyze the paired-catch data from calibration studies. Our main contributions are results from simulation experiments designed to measure the accuracy of statistical inferences derived from some of these models. Our results show that a model commonly used to analyze calibration data can provide unreliable statistical results when there is between-tow spatial variation in the stock densities at each paired-tow site. However, a generalized linear mixed-effects model gave very reliable results over a wide range of spatial variations in densities and we recommend it for the analysis of paired-tow survey calibration data. This conclusion also applies if there is between-tow variation in catchability.

Modified sorting technique to mitigate the collateral mortality of trawled school prawns (Metapenaeus macleayi)

The potential for changes to onboard handling practices in order to improve the fate of juvenile school prawns (Metapenaeus macleayi) discarded during trawling were investigated in two Australian rivers (Clarence and Hunter) by comparing a purpose-built, water-filled sorting tray against a conventional dry tray across various conditions, including the range of typical delays before the start of sorting the catch (2 min vs. 15 min). Juvenile school prawns (n= 5760), caught during 32 and 16 deployments in each river, were caged and sacrificed at four times: immediately (T0), and at 24 (T24), 72 (T72), and 120 (T12 0) hours after having been discarded. In both rivers, most mortalities occurred between T0 and T24 and, after adjusting for control deaths (<12%), were greatest for the 15-min conventional treatment (up to 41% at T120). Mixed-effects logistic models revealed that in addition to the sampling time, method of sorting, and delay in sorting, the weight of the catch, salinity, and percentage cloud cover were significant predictors of mortality. Although trawling caused some mortalities and comparable stress (measured as L -lactate) in all school prawns, use of the water tray lessened the negative impacts of some of the above factors across both the 2-min and 15-min delays in sorting so that the overall discard mortality was reduced by more than a third. When used in conjunction with selective trawls, widespread application of the water tray should help to improve the sustainability of trawling for school prawns.

Effects of a large fishing closure on benthic communities in the western Gulf of Maine: recovery from the effects of gillnets and otter trawls

The recovery of benthic communities inside the western Gulf of Maine fishing closure area was evaluated by comparing invertebrate assemblages at sites inside and outside of the closure four to six years after the closure was established. The major restriction imposed by the closure was a year-round prohibition of bottom gillnets and otter trawls. A total of 163 seafloor sites (~half inside and half outside the closure) within a 515-km2 study area were sampled with some combination of Shipek grab, Wildco box corer, or underwater video. Bottom types ranged from mud (silt and clay) to boulders, and the effects of the closure on univariate measures (total density, biomass, taxonomic richness) of benthos varied widely among sediment types. For sites with predominantly mud sediments, there were mixed effects on inside and outside infauna and no effect on epifauna. For sites with mainly sand sediments, there were higher density, biomass, and taxonomic richness for infauna inside the closure, but no significant effects on epifauna. For sites dominated by gravel (which included boulders in some areas), there were no effects on infauna but strong effects on epifaunal density and taxonomic richness. For fishing gear, the data indicated that infauna recovered in sand from the impacts of otter trawls operated inside the closure but that they did not recover in mud, and that epifauna recovered on gravel bottoms from the impact of gillnets used inside the closure. The magnitudes of impact and recovery, however, cannot be inferred directly from our data because of a confounding factor of different fishing intensities outside the closure for a direct comparison of preclosure and postclosure data. The overall negative impact of trawls is likely underestimated by our data, whereas the negative impact of gillnets is likely overestimated.

The predicting power of models for eutrophication

Restoration of water-bodies from eutrophication has proved to be extremely difficult. Mathematical models have been used extensively to provide guidance for management decisions. The aim of this paper is to elucidate important problems of using models for predicting environmental changes. First, the necessity for a proper uncertainty assessment of the model, upon calibration, has not been widely recognized. Predictions must not be a single time trajectory; they should be a band, expressing system uncertainty and natural variability. Availability of this information may alter the decision to be taken. Second, even with well-calibrated models, there is no guarantee they will give correct projections in situations where the model is used to predict the effects of measures designed to bring the system into an entirely different ”operating point”, as is typically the case in eutrophication abatement. The concept of educated speculation is introduced to partially overcome this difficulty. Lake Veluwe is used as a case to illustrate the point. Third, as questions become more detailed, such as ”what about expected algal composition”, there is a greater probability of running into fundamental problems that are associated with predicting the behaviour of complex non-linear systems. Some of these systems show extreme initial condition sensitivity and even, perhaps, chaotic behaviour, and are therefore fundamentally unpredictable.

Spatial and temporal variability in the relative contribution of king mackerel (Scomberomorus cavalla) stocks to winter mixed fisheries off South Florida

King mackerel (Scomberomorus cavalla) are ecologically and economically important scombrids that inhabit U.S. waters of the Gulf of Mexico (GOM) and Atlantic Ocean (Atlantic). Separate migratory groups, or stocks, migrate from eastern GOM and southeastern U.S. Atlantic to south Florida waters where the stocks mix during winter. Currently, all winter landings from a management-defined south Florida mixing zone are attributed to the GOM stock. In this study, the stock composition of winter landings across three south Florida sampling zones was estimated by using stock-specific otolith morphological variables and Fourier harmonics. The mean accuracies of the jackknifed classifications from stepwise linear discriminant function analysis of otolith shape variables ranged from 66−76% for sex-specific models. Estimates of the contribution of the Atlantic stock to winter landings, derived from maximum likelihood stock mixing models, indicated the contribution was highest off southeastern Florida (as high as 82.8% for females in winter 2001−02) and lowest off southwestern Florida (as low as 14.5% for females in winter 2002−03). Overall, results provided evidence that the Atlantic stock contributes a certain, and perhaps a significant (i.e., ≥50%), percentage of landings taken in the management-defined winter mixing zone off south Florida, and the practice of assigning all winter mixing zone landings to the GOM stock should

Modification of the biological intercept model to account for ontogenetic effects in laboratory-reared delta smelt (Hypomesus transpacificus)*

We investigated age, growth, and ontogenetic effects on the proportionality of otolith size to fish size in laboratory-reared delta smelt (Hypomesus transpacificus) from the San Francisco Bay estuary. Delta smelt larvae were reared from hatching in laboratory mesocosms for 100 days. Otolith increments from known-age fish were enumerated to validate that growth increments were deposited daily and to validate the age of fish at first ring formation. Delta smelt were found to lay down daily ring increments; however, the first increment did not form until six days after hatching. The relationship between otolith size and fish size was not biased by age or growth-rate effects but did exhibit an interruption in linear growth owing to an ontogenetic shift at the postflexon stage. To back-calculate the size-at-age of individual fish, we modified the biological intercept (BI) model to account for ontogenetic changes in the otolith-size−fish-size relationship and compared the results to the time-varying growth model, as well as the modified Fry model. We found the modified BI model estimated more accurately the size-at-age from hatching to 100 days after hatching. Before back-calculating size-at-age with existing models, we recommend a critical evaluation of the effects that age, growth, and ontogeny can have on the otolith-size−fish-size relations

Length-weight Relationships of Dolphinfish, Coryphaena hippurus, and Wahoo, Acanthocybium solandri: Seasonal Effects of Spawning and Possible Migration in the Central North Pacific

Weight-on-length (W-L) relationships for 2,482 dolphinfish, Coryphaena hippurus, and 1,161 wahoo, Acanthocybium solandri, were examined. Data on fork length, whole (round) weight, and sex were collected for dolphinfish at the Honolulu fish auction from March 1988 through November 1989. Unsexed weight and length data for wahoo were collected at the auction from July 1988 through November 1989. We also used sex specific weight and length data of 171 wahoo collected during 1977–1985 research cruises for analysis. Coefficients of W-L regressions were significantly different between the sexes for dolphinfish. Coefficients did not significantly differ between the sexes for wahoo based on research cruise data. In a general linear model evaluating month as a categorical factor, month was significant for female dolphinfish, male dolphinfish, and wahoo with sexes pooled. W-L and length-on-weight (L-W) relationships were fitted by nonlinear regression for all dolphinfish, female dolphinfish, male dolphinfish, and all wahoo sexes pooled. W-L relationships for monthly samples of female dolphinfish, male dolphinfish, and all wahoo with sexes pooled were also fitted by nonlinear regression. Predicted mean weight at length for wahoo was highest at the beginning of the spawning season in June and lowest after the spawning season in September. Maximum and minimum predicted mean weight at length for both sexes of dolphinfish did not correspond with the peak spawning period (March–May). Plausible migration models in conjunction with reproductive behavior were examined to explain the variability in monthly predicted mean weight at length for dolphinfish.

In search of climate effects on Atlantic Croaker (Micropogonias undulatus) stock off the U.S. Atlantic coast with Bayesian state-space biomass dynamic models

Atlantic Croaker (Micropogonias undulatus) production dynamics along the U.S. Atlantic coast are regulated by fishing and winter water temperature. Stakeholders for this resource have recommended investigating the effects of climate covariates in assessment models. This study used state-space biomass dynamic models without (model 1) and with (model 2) the minimum winter estuarine temperature (MWET) to examine MWET effects on Atlantic Croaker population dynamics during 1972–2008. In model 2, MWET was introduced into the intrinsic rate of population increase (r). For both models, a prior probability distribution (prior) was constructed for r or a scaling parameter (r0); imputs were the fishery removals, and fall biomass indices developed by using data from the Multispecies Bottom Trawl Survey of the Northeast Fisheries Science Center, National Marine Fisheries Service, and the Coastal Trawl Survey of the Southeast Area Monitoring and Assessment Program. Model sensitivity runs incorporated a uniform (0.01,1.5) prior for r or r0 and bycatch data from the shrimp-trawl fishery. All model variants produced similar results and therefore supported the conclusion of low risk of overfishing for the Atlantic Croaker stock in the 2000s. However, the data statistically supported only model 1 and its configuration that included the shrimp-trawl fishery bycatch. The process errors of these models showed slightly positive and significant correlations with MWET, indicating that warmer winters would enhance Atlantic Croaker biomass production. Inconclusive, somewhat conflicting results indicate that biomass dynamic models should not integrate MWET, pending, perhaps, accumulation of longer time series of the variables controlling the production dynamics of Atlantic Croaker, preferably including winter-induced estimates of Atlantic Croaker kills.