Elsevier

Ecological Modelling

Volume 359, 10 September 2017, Pages 1-24
Ecological Modelling

Modeling food web effects of low sardine and anchovy abundance in the California Current

https://doi.org/10.1016/j.ecolmodel.2017.05.007Get rights and content

Highlights

  • California Current sardine and anchovy are currently at low abundance.

  • Atlantis ecosystem model predicts food web impacts of this reduced forage base.

  • Declines predicted for predators such as dolphins, large flatfish, and some birds.

  • Weaker declines are predicted by Atlantis than an alternate ecosystem model type.

  • Atlantis is one approach in a multi-model toolbox for this region.

Abstract

Populations of sardine, anchovy, and other forage species can fluctuate to low levels due to climate variability and fishing, leading to indirect effects on marine food webs. In the context of recent declines of sardine (Sardinops sagax) and anchovy (Engraulis mordax) in the California Current, we apply an end-to-end Atlantis ecosystem model that is spatially explicit, includes trophic interactions, and allows high and low recruitment regimes (production of juveniles). Our simulations suggest that depleted sardine populations, whether caused by fishing or natural cycles, may lead to declines in predator groups such as dolphins and large piscivorous flatfish (e.g. California halibut Paralichthys californicus). Birds exhibited more moderate declines, and California sea lions (Zalophus californianus) exhibited relatively weak declines. The Atlantis ecosystem model also predicted indirect positive effects of sardine depletion, primarily for prey species such as zooplankton. Overall our model predicted moderate declines in most predators during simulated severe declines in sardine and anchovy, illustrating the important buffering role provided by forage species other than sardine and anchovy. This ‘buffered response’ is weaker than what would be suggested by another ecosystem model (Ecosim), as predicted by diet information and a global synthesis of Ecosim models (the PREP equation). One limitation of the Atlantis model is that it did not include processes that might give rise to localized depletion of sardine at scales relevant to central place foragers, such as birds and pinnipeds. This analysis will contribute to a collaborative multi-model approach that evaluates the role of sardine in the California Current.

Introduction

Sardine, anchovy, and other forage species support economically valuable fisheries and are a main prey source for many valuable predator species (Pikitch et al., 2014); therefore fluctuations in abundance can have large implications for both fisheries and food webs (Smith et al., 2011). For instance, in the California Current off the West Coast of the USA, revenue from the Pacific sardine (Sardinops sagax) fishery recently peaked at over $21 million in 2012 (Pacific Fishery Management Council, 2014). Pacific sardine and northern anchovy (Engraulis mordax) are also main prey sources for economically valuable predators such as salmon (Oncorhynchus spp.) and albacore tuna (Thunnus alalunga) and threatened and endangered species such as humpback whales (Megaptera novaeangliae) (Szoboszlai et al., 2015).

Populations of sardine, anchovy, and other forage species can fluctuate to low levels due to climate variability and fishing (Lindegren et al., 2013, Essington et al., 2015). For instance, sardine in the California Current have recently drastically declined to 9% of peak 2007 abundance (Hill et al., 2015), and anchovy in Southern California had declined to ∼1% of peak abundance by 2011 (MacCall et al., 2016). This has raised concerns about the implications for dependent predators and fisheries in this region. Of particular concern is that anchovy and sardine are simultaneously at low abundance. Concurrent periods of low abundance are atypical but problematic: Lindegren et al. (2016) analyzed the 60 year time series from CalCOFI surveys and associated data and suggested that it is the asynchrony between anchovy, sardines, and other small pelagic fish that leads to community-level stability. Functional complementarity of these small fish species in the diets of generalist predators may typically support resilience in this ecosystem, as long as some of these forage species remain abundant. Recently, synchronous declines in forage fish species have been linked to reduced pup weights of California sea lion (Zalophus californianus) (McClatchie et al., 2016) and breeding failure of brown pelican (Pelecanus occidentalis) in the Southern California Bight (Henry, 2015).

The consequences of depleted forage fish functional groups in the California Current have been tested previously using two ecosystem models (Kaplan et al., 2013). Both the Atlantis model (Horne et al., 2010) and an Ecosim model (Field, 2004) were consistent in predicting that forage fish depletion would lead to declines in large flatfish and increases in euphausiids, cephalopods, myctophids, and mackerel. The Ecosim model also suggested tradeoffs between forage fish harvest and seabird abundance, with compensatory increases in some planktonic prey of forage fish, not seen in Atlantis. This difference suggested higher responsiveness of the Ecosim model to changes in forage fish biomass. Each of the models was limited in different capacities; Ecosim lacks spatial resolution (i.e. no potential for local depletion via fishing) and this application of Ecosim assumed extremely low sardine abundance, and the Atlantis model contained poor taxonomic resolution for forage fish groups and some predator groups. While this earlier analysis was able to identify some consistent results between models, discrepancies and the overall higher responsiveness of Ecosim versus Atlantis were unresolved in the analysis.

Subsequent to the analyses of Kaplan et al. (2013), US West Coast fishery managers have called for improved ecosystem modeling tools to predict impacts of depletion of sardine and other forage fish (Pacific Fishery Management Council, 2013). Additionally, fishery managers have launched initiatives to protect forage fish species that are presently not targeted by fisheries (Pacific Fishery Management Council, 2016). Here, we address calls by fishery managers to consider the role of forage fish, improving upon earlier efforts (Kaplan et al., 2013) by developing a revised spatially-explicit Atlantis ecosystem model of the California Current (Marshall et al., 2017). In this data-rich system, the Atlantis approach allows us to mechanistically model food web impacts on predators, competitors, and prey of sardine and anchovy to ask: What are the consequences of depletion of sardine on sardine yield and spatial distribution, and on fished predators and on other species in the food web? How does this vary at high versus low regimes of sardine and anchovy productivity (Chavez et al., 2003, MacCall, 2009), including periods such as the apparent present dearth of both species? Do these results vary spatially, for instance near major sardine fishing ports?

To begin to make multi-model inference and to understand how model structure drives prediction uncertainty (Pinnegar et al., 2005, Gårdmark et al., 2013), we have conducted this work within a multi-model setting, the Ocean Modeling Forum (www.oceanmodelingforum.org). We compare our model results to those predicted using a different modeling approach, the PREP equation (Predator Response to the Exploitation of Prey). This equation generalizes results from a set of Ecosim models considered by the Lenfest Forage Fish Taskforce (Pikitch et al., 2012); the task force intended PREP as a tool in data-limited ecosystems where diet information was available but food web or ecosystem models were lacking. In a companion paper we more thoroughly compare results between model types within the Ocean Modeling Forum setting.

Section snippets

Methods

We briefly describe the Atlantis framework, and the revised implementation for the California Current. We detail the scenarios tested to understand the role of sardine and anchovy in the food web, the analyses of model output, and the comparison of Atlantis predictions to those from the PREP equation.

Results

As detailed in each subsection below, we first focus on the Base Recruitment scenarios, summarizing the effects of simulated levels of sardine depletion on the sardine stock, and the food web; predators such as Dolphins and Large piscivorous flatfish decline most strongly when sardine are depleted. We then translate these results into ecosystem indicators, finding that most of these indicators respond only weakly to these scenarios. We next focus on the broader range of sardine and anchovy

Discussion

Fishery managers in the California Current recently called for improved ecosystem modeling of forage fish (Pacific Fishery Management Council 2013), and to the extent practical our effort specifically addresses this request. In particular, our model includes the full geographic range of the sardine stock, identifies separate fisheries per region, represents fisheries that target species other than forage fish, and has a simple representation of climate variability (recruitment regimes). Our

Funding

The Ocean Modeling Forum was supported by the David and Lucile Packard Foundation. Model development benefited greatly from a grant from the NOAA Ocean Acidification Program and National Centers for Coastal Ocean Science. TEE was supported by Pew Marine Conservation Fellows program.

Acknowledgements

This work was undertaken as part of the Ocean Modeling Forum and California Current Integrated Ecosystem Assessment. As part of the Ocean Modeling forum, the following provided feedback and assistance of this project: André E. Punt, Alec D. MacCall, Tessa B. Francis, Felipe Hurtado- Ferro, Kelli F. Johnson, Phillip S. Levin, Richard Parrish, and William J. Sydeman. Blake Feist provided GIS expertise to assign anchovy and sardine distributions from CalCOFI to the Atlantis polygons.

References (73)

  • J.K. Pinnegar et al.

    Aggregation and removal of weak-links in food-web models: system stability and recovery from disturbance

    Ecol. Model.

    (2005)
  • A.E. Punt et al.

    Exploring the implications of the harvest control rule for Pacific sardine, accounting for predator dynamics

    A MICE model. Ecol. Model.

    (2016)
  • K.A. Rose et al.

    Demonstration of a fully-coupled end-to-end model for small pelagic fish using sardine and anchovy in the California, Current

    Prog. Oceanogr.

    (2015)
  • M.D. Smith et al.

    Using an Atlantis model of the southern Benguela to explore the response of ecosystem indicators for fisheries management

    Environ. Model. Softw.

    (2015)
  • A.I. Szoboszlai et al.

    Forage species in predator diets: synthesis of data from the California Current

    Ecol. Inform.

    (2015)
  • C. Walters et al.

    Predictions from simple predator-prey theory about impacts of harvesting forage fishes

    Ecol. Model.

    (2016)
  • T.R. Baumgartner et al.

    Reconstruction of the history of Pacific sardine and northern anchovy populations over the past two millennia from sediments of the Santa Barbara Basin, California

    CalCOFI Rep

    (1992)
  • P. Boveng

    Status of the California Sea Lion Population on the US West Coast. Admin. Rep. LJ-88-07

    (1988)
  • Brodeur, R.D., Buchanan, J.J., and Emmett, R.L., 2014. Pelagic and demersal fish predators on juvenile and adult forage...
  • J.V. Carretta et al.

    US Pacific Marine Mammal Stock Assessments: 2012 (NOAA Technical Memorandum NMFS-SWFSC 504)

    (2013)
  • F.P. Chavez et al.

    From anchovies to sardines and back: multidecadal change in the Pacific Ocean

    Science

    (2003)
  • D.M. Checkley et al.

    Climate, anchovy, and sardine

    Annu. Rev. Mar. Sci.

    (2017)
  • P.M. Cury et al.

    Global seabird response to forage fish depletion—one-third for the birds

    Science

    (2011)
  • P. Cury

    Small pelagics in upwelling systems: patterns of interaction and structural changes in wasp-waist ecosystems

    ICES J. Mar. Sci.

    (2000)
  • D.A. Demer et al.

    Sampling selectivity in acoustic-trawl surveys of Pacific sardine (Sardinops sagax) biomass and length distribution

    ICES J. Mar. Sci.

    (2013)
  • T.E. Essington et al.

    Fishing amplifies forage fish population collapses

    Proc. Natl. Acad. Sci.

    (2015)
  • J.C. Field

    Application of Ecosystem-Based Fishery Management Approaches in the Northern California Current

    (2004)
  • B.E. Fissel et al.

    Daily egg production, spawning biomass and recruitment for the central subpopulation of Northern anchovy 1981–2009

    Calif. Coop. Ocean Fish. Invest. Rep.

    (2011)
  • A.H. Fleming et al.

    Humpback whale diets respond to variance in ocean climate and ecosystem conditions in the California Current

    Glob. Change Biol.

    (2016)
  • E.A. Fulton et al.

    Lessons in modelling and management of marine ecosystems: the Atlantis experience

    Fish

    (2011)
  • E.A. Fulton et al.

    An integrated approach is needed for ecosystem based fisheries management: insights from ecosystem-level management strategy evaluation

    PLoS One

    (2014)
  • A. Gårdmark et al.

    Biological ensemble modeling to evaluate potential futures of living marine resources

    Ecol. Appl.

    (2013)
  • S.K. Gaichas et al.

    Beyond the defaults: functional response parameter space and ecosystem-level fishing thresholds in dynamic food web model simulations

    Can. J. Fish. Aquat. Sci.

    (2012)
  • G.P. Griffith et al.

    Effects of fishing and acidification-related benthic mortality on the southeast Australian marine ecosystem

    Glob. Change Biol.

    (2011)
  • G.P. Griffith et al.

    Predicting interactions among fishing, ocean warming, and ocean acidification in a marine system with whole-ecosystem models

    Conserv. Biol.

    (2012)
  • S. Henry

    Agenda Item G.3. a: USFWS Report

    (2015)
  • Cited by (24)

    • Pacific herring (Clupea pallasii) as a key forage fish in the southeastern Gulf of Alaska

      2022, Deep-Sea Research Part II: Topical Studies in Oceanography
      Citation Excerpt :

      Ecosystem models of the southeastern Gulf of Alaska and California Current also agree in emphasizing the roles of food web complexity and forage fish guild redundancy in modulating the impacts of herring depletion on predators (Kaplan et al. 2013, 2017; Koehn et al. 2016, 2017). In the California Current, simulated forage fish depletion in both EwE and Atlantis positively affected some zooplankton and planktivorous fish through release from predation and competition, respectively (Kaplan et al. 2013, 2017). These releases resulted from generalist feeding by most predators and functional redundancy among forage fish (Koehn et al., 2016; Kaplan et al., 2017).

    • Fisheries management tools to support coastal and marine spatial planning: A case study from the Northern Gulf of California, Mexico

      2020, MethodsX
      Citation Excerpt :

      The modeling platform summarizes biological components as functional groups aggregated by trophic, life history, or niche similarities. Further information on Atlantis can be found in the User's Guide [6], the Atlantis Wiki (https://research.csiro.au/atlantis/home/links/), and recently published Atlantis applications [29,49,50]. The Atlantis model for the Northern Gulf of California, which includes the Coastal Corridor, extends over 57,800 km2, represents ecosystem structure and function in 2008, current fishing effort, and provides a detailed representation of the Northern Gulf's oceanography, historical fishing patterns, migration and movement of key species, and variability in diet compositions [2,3].

    View all citing articles on Scopus
    View full text