Accounting for habitat-dependent observation error in bottom trawl survey indices for pelagic stocks using butterfish (Peprilus triacanthus) as a model

Principal Investigator: John Manderson
Co-Principal Investigators: David Richardson, Jonathan Hare, John Quinlan
External Collaborator: Nikitas Georgas (Stevens Institute), Josh Kohut (Rutgers)

 EXPERIMENTAL Prototype Thermal Habitat Model

EXPERIMENTAL Prototype Thermal Habitat Model


Habitats of small pelagic fish stocks central to marine food webs are defined by the spatially dynamic properties and processes of the oceans fluid.  The habitat dynamics that result can be particularly complex in temperate latitudes where rapid seasonal changes in the location and volume of thermal habitat cause ectothermic animals to undertake broad scale migrations from shallow inshore summer habitats to deep overwintering habitats.  These dynamics make population sizes of small pelagic stocks notoriously difficult to assess using traditional fishery independent bottom trawl surveys that are generally stratified in time and space by depth, location and calendar year.  Furthermore the survey ships are typically constrained to sample continental shelf waters at depths ranging from 20 to 240 meters and thus under sample shallow nearshore and deep offshore areas acknowledged to be ecologically important.  Because these important habitats are undersampled, changes in seasonal habitat dynamics can change the proportion of habitat and thus the associated population that is surveyed each year.  If changes in seasonal habitat dynamics are systematic, habitat dependent observation error can be misinterpreted as a systematic increase or decrease in population size.

We have assembled a diverse working group of government, academic, and fishing industry experts from disciplines of marine habitat ecology, oceanography, stock assessment and ecosystem modeling to develop methods for quantifying the overlap between trawl survey and pelagic habitat (e.g., habitat dependent observation error), with a specific focus on season-specific and ontogenic patterns in catchability through time.

The working group’s short term objectives are to:

1) Develop mechanistic and statistical habitat models for a small pelagic fish stock (butterfish, Peprilus triacanthus) that can be used as a tool to index habitat quality explicitly in space and time.

2) Use the habitat models to estimate biases related to spatial habitat dynamics in population indices calculated on the basis of traditional bottom trawl surveys.

3) Use the habitat models to develop a method to explicitly account for habitat dependent observation error.

4) To develop a cooperative scientific process that can provide information required for ecosystem assessment and management.