Scientist Robert Leaf, of The University of Southern Mississippi, will lead a study on Gulf menhaden, a critical keystone fish species. The team will work with state resource agencies to collect fishery-independent data on age and length composition, which have been identified as data gaps. Scientists will compare the new data with fishery-dependent data. They also will determine minimum sample sizes for efficient age sampling efforts in the Gulf states. This study will provide more complete Gulf menhaden data that can be used in stock assessments.

Abstract

The Gulf Menhaden stock is the second largest fishery, by volume, in the United States. Effective assessment of the Gulf Menhaden (Brevoortia patronus) stock is critical due to the economic and ecological role that they play in the northern Gulf of Mexico (GOM). The most recent stock assessment of Gulf Menhaden indicates that the stock is not overfished and that overfishing is not occurring but that a “Very High” (highest) priority data need is to determine age composition information from the fishery-independent data collections programs administered by the states.

We propose a study that will address this need and is consistent with Mississippi-Alabama Sea Grant’s “Healthy Coastal Ecosystems” focus. We will evaluate length composition data and collect age composition data by partnering with the Gulf States’ resource agencies, determine the minimum necessary data or sample size to understand state-specific age compositions, compare the length and age composition data encountered in the fishery-independent (this study) and ongoing fishery-dependent sampling, determine the impact on Gulf menhaden management by integrating these data into the stock assessment and determine and assess the gaps in states’ data collection efforts and provide strategies to promote efficient and complementary sampling efforts among states.

We will achieve these objectives by working with our state partners to collect Gulf Menhaden using a stratified design. We will determine the age composition of Gulf Menhaden and derive selectivity patterns of the gear used in fishery-independent sampling. These data are directly and immediately applicable to the stock assessment efforts of this stock. We will present the results of this work to state agency partners and stakeholders concerned with Gulf Menhaden stock assessment for their respective agencies, four of those are supportive of this work.

Objectives

Methodology

To achieve the study objectives we will work collaboratively with state management agency personnel to collect Gulf Menhaden, maximizing the temporal and spatial sampling for each gear type used by the states. We will determine age of all fish using standard age-determination methods.

A Bayesian Hierarchical model will be constructed to describe the age composition of the stock at each of the multiple levels that will be sampled. Fish will be collected from each of the gear types deployed at various locations (sites) and times of the year (e.g. months) for each of the states.

We will make four trips (the first year) and two trips (the second year) to locations of collection from April to September. Collection locations are Panama City, Florida, Dauphin Island, Alabama, Grand Isle, Louisiana, Lake Charles, Louisiana, and Galveston, Texas. The role of our state partners will be limited to keeping the fish that they collect and putting them on ice in a cooler. The fish will be placed in individually-labeled bags that will denote the location, date, and gear type used in their collection. We will make four trips in year one and two trips in year two to locations where state resource agencies sample menhaden using their standard gear.

Age determination will be performed on each fish, by a minimum of two trained readers at the University of Southern Mississippi's Gulf Coast Research Laboratory. The distribution of ages and lengths will be analyzed using a Bayesian Hierarchical Model to evaluate the effects of gear type, location, and date of collection on the distribution of lengths and ages of fish.

We will evaluate the benefit (reduction in variance), that occurs by increasing sample size, in the estimation of the age- composition at each level of analysis: gear type, location, and date of collection. We will determine, using a Monte-Carlo resampling algorithm, how the parameters of the age distribution change with an increasing number of samples. Age composition information derived in this study will be integrated into the Beaufort Assessment Model (BAM) stock assessment model. The overall impact on reference points and thus stock status is difficult to evaluate without estimating the error matrix from good quality data.

Rationale