The Ocean Archive System searches our original datasets as they were submitted to us, not individual points or profiles. If you want to search and retrieve ocean profiles in a common format, or objectively analyzed fields, your better option may be to use one of our project applications. See: Access Data

OAS accession Detail for 0242339
<< previous |revision: 10
accessions_id: 0242339 | archive
Title: NOAA RESTORE Science Program: Ecosystem modeling to improve fisheries management in the Gulf of Mexico: model inputs and outputs for the West Florida Shelf, 1985-01-01 to 2018-12-31 (NCEI Accession 0242339)
Abstract: This dataset is a collection of files containing the necessary inputs to and relevant outputs from the West Florida Shelf ecosystem model used to estimate mortality rates on valuable reef fish species, and ecosystem indicators to estimate red tide mortality rates events. The model was developed using the Ecopath with Ecosim and Ecospace modeling software package. The spatial extent of the model is 25° to 30.5° N and 81° to 87.5° W and simulations were run in hindcast from 1985 to 2018 at a monthly timestep. Parameters include biomass, consumption, mortality, diet, landings, discards, dispersal rates, ecosystem indicators, environmental preferences, and habitat maps for 83 species or functional groups included in the model. Each input or output parameter type is included as its own csv or netcdf file with informative names.
Date received: 20211004
Start date: 19850101
End date: 20181231
Seanames:
West boundary: -87.5833
East boundary: -80.9166
North boundary: 30.5823
South boundary: 24.9156
Observation types:
Instrument types:
Datatypes:
Submitter:
Submitting institution: US DOC; NOAA; NOS; National Centers for Coastal Ocean Science
Collecting institutions:
Contributing projects:
Platforms:
Number of observations:
Supplementary information: To model the WFS ecosystem, we used the Ecopath with Ecosim software (EwE) (Christensen and Walters 2004). The EwE software suite is composed of three main components: Ecopath, Ecosim and Ecospace. Ecopath is the static version that provides a snapshot of the system. Ecosim is the time dynamic simulation module that allows for predicting ecosystem and species changes over time. And lastly, Ecospace is the spatial component that predicts the distribution of each functional group over a two-dimensional spatial grid at a monthly time step. The software is freely available at www.ecopath.org along with complete documentation of the model.

This model can be opened with Ecopath with Ecosim software version 6.7.0.17363 or later and requires the spatial-temporal plugin.

The first version of the WFS EwE model was developed at the Florida FWC Fish and Wildlife Research Institute to investigate forage fish interactions and was also used to evaluate the effects of shading by phytoplankton blooms on the marine community (Okey et al. 2004). This version was later modified to integrate with existing single-species stock assessments and management (Chagaris 2013, Chagaris et al. 2015) and modified again to include invasive lionfish (Chagaris et al. 2017). In the current version, we updated the WFS EwE and Ecospace model from Chagaris (2013) and Chagaris et al. (2017) by adding new species, splitting functional groups, adding fishing fleets, incorporating more age structure and updating time-series of biomass and catches. The current version of the WFS EwE model includes 17 fishing fleets and 83 functional groups and was calibrated to available time series on abundance, catch, fishing mortality, and fishing effort from 1985 to 2018.

The primary objective of the current WFS EwE model is to evaluate impacts of red tide and quantify mortality on valuable reef fish species. For this, we used the spatially explicit Ecospace model. The WFS Ecospace model encompasses an area ranging from 25° to 30.5°N from shore to about the 200m isobath. Ecosystem dynamics are simulated over a map with a spatial resolution of 10 minutes, (~20km) containing 38 rows x 40 columns from 1985 to 2018 at monthly steps. To spatially drive the foraging capacity of consumer groups through the habitat capacity model, we considered depth, rugosity, sea surface temperature (SST), sea bottom temperature (SBT), sea surface salinity (SSS) and red tide for its ability to influence the spatial distribution of marine species. Environmental monthly raster data were extracted from the Hybrid Coordinate Ocean Model (HYCOM) (Chassignet et al. 2007) except for depth which was obtained from NOAA (2001) and red tide (see details below). Environmental response functions for environmental layer were obtained by fitting generalized additive models (GAMs) to fisheries-independent data from multiple surveys. Baseline dispersal rates (km/year) were estimated from movement rates in published tagging studies or based on the general “300-30-3” rule which assumes 300 km·year-1 for large pelagic reef-associated species, 30 km·year-1 for small reef-associated and demersal functional groups, and 3 km·year-1 for benthic and planktonic functional groups. The WFS Ecospace model is publicly available at the University of Florida (UF) institutional repository (Chagaris 2021).

Red tide effects were incorporated into the WFS Ecospace model by first creating monthly red tide concentration (cells/L) maps using Karenia brevis cells concentration data collected by Florida Fish and Wildlife Conservation Commission. Monthly normalized fluorescent line height (nFLH) grids derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) aqua satellite (NASA Goddard Space Flight Center 2020) were used to define the spatial extent and duration of harmful algal blooms (Hu et al. 2005). The red tide cell concentration data were interpolated over the entire WFS using ordinary kriging. The kriged maps were clipped to the nFLH red tide polygons to obtain the monthly red tide maps and then resampled to match the 10-minute spatial resolution of the WFS Ecospace model.

Red tide mortality response functions were defined to generate mortality in each grid cell of the WFS Ecospace model. The red tide mortality response function determined the proportion of biomass killed in each grid cell per month as a function of the kriged K. brevis cell concentration in that grid cell. The mortality response assumed a logistic shape with varying slopes and intercepts to represent 20 different red tide mortality response functions with multiple sensitivity levels. The biomass loss due to red tides for each group, grid cell, and monthly time step was calculated as the product of proportion killed based on red tide concentration and biomass in each cell. The annual index of red tide mortality was computed by dividing the total red tide loss over all map cells by the total biomass and then averaging over all months.

Sub-lethal responses, expressed as the ability to forage and grow, were represented using Ecospace foraging response curves. The foraging responses serve two roles in the simulations. First, they reduce the foraging arena size in affected grid cells, which thereby reduces consumption and biomass growth. Additionally, the reduced foraging capacity in an affected cell will increase the movement rate out of that cell. This allows species to avoid red tide blooms and may mitigate direct mortality losses if cells with suitable habitat are nearby. The red tide foraging response was defined using a logistic curve that decreased with red tide cell concentrations. We assumed the foraging responses had lower inflection points than the mortality responses since sub-lethal effects and avoidance response are likely to be experienced at lower K. brevis concentrations (Landsberg et al. 2009). A total of 15 different foraging response curves were evaluated, representing multiple sensitivity levels.

Simulations were run from 1985 to 2018, using 160 different combinations of mortality and foraging sensitivities to red tide. Of those, only a subset was retained for summaries based on pre-defined model performance and fitting criteria. Here, we archive outputs from the ‘best fit’ Ecospace run.

Submission Package ID: 0P4L5E
Availability date:
Metadata version: 10
Keydate: 2021-10-25 17:31:08+00
Editdate: 2022-01-12 03:03:05+00