Forage fish exploitation

Are there cascading ecological effects associated with forage fish exploitation?

This overarching question can be explored using EwE to model different scenarios of forage fish exploitation and evaluate the ecosystem-wide consequences.

The question addresses a critical issue in marine ecosystem management, as forage fishes play an important role in transferring energy from lower to higher trophic levels – think “wasp-waist”^[1]. Their exploitation can have far-reaching consequences throughout the ecosystem, affecting not only the targeted species but also a wide range of predators, including many species of commercial and conservation importance.  Notably, we often see that exploitation of forage fish has direct and immediate consequences for their predators: increased fishing mortality may be compensated by reduced predation mortality due to decrease in predator abundance.  The result can be that forage fish are impacted less by fisheries than their predators.

Using EwE to explore this question may provide insights into the complex dynamics of marine ecosystems and help inform more inclusive, ecosystem-based management approaches for forage fish fisheries.

Potential policy questions

• Trophic impacts: How do fisheries for forage fish affect their predators, (e.g., seabirds, marine mammals, larger fish) and their prey?
• Energy flow alterations: Does forage fish exploitation change the energy transfer efficiency through the food web?
• Alternate stable states: Are there threshold levels of forage fish biomass that, if crossed, could lead to shifts in ecosystem structure?
• Competition effects: How does changes in forage fish population size impact the competitive relationships between other species in the ecosystem?
• Fishery interactions: What are the indirect effects on other fisheries that target predators of forage fish?
• Recovery potential: How resilient is the ecosystem to different levels of forage fish depletion, and what are the timescales for recovery?
• Spatial considerations: How do the impacts of forage fish exploitation vary spatially, especially in relation to important breeding or feeding grounds for dependent predators?
• Climate change interactions: How might the impacts of forage fish exploitation be exacerbated or mitigated by ongoing climate change?
• Management strategies: What harvest control rules could be used to ensure sustainable forage fish exploitation?
• Socio-economic trade-offs: How can the economic value of the forage fish fishery be balanced against the ecosystem services provided by abundant forage fish populations?

Trophic impacts: How do fisheries for forage fish affect their predators (e.g., seabirds, marine mammals, larger fish) and their prey?

The Anchovy Bay model can be used to explore this question, download if needed from this link. Open the model (Main menu, File > Open model), make a new Ecosim scenario (Main menu, Ecosim > New scenario). Reset fishing effort (Navigator, Ecosim > Input > Fishing effort > Reset all). Run Ecosim (Navigator, Ecosim > Output > Run Ecosim), and you’ll probably see the results of a 41 year simulation. On the plot, the lines indicate relative biomass, so biomass over time relative to the baseline Ecopath biomass. With the default setting, The whale biomass will increase asymptotically as seals are recovering from harvesting, and the seals will decrease asymptotically as they are being overexploited. An ecological effect is that mackerel will decrease due to higher predation by whales.

We now want to compare what happens to group biomasses with and without fishing for our forage fish, anchovy.  We could to this by setting fishing effort for the two fleets that catches anchovy (seiners and bait boats) to 0, but that would also change fishing for mackerel, (which also are caught by the seiners). Instead we can set the fishing mortality (F) for anchovy to 0. Go to Ecosim > Input > Fishing mortality, click 6. Anchovy and Set to 0.  Now run Ecosim again.

If you look at the plot, you’ll see that whiting and mackerel both increase. As does anchovy – but not as much as whiting^[2].  We need to store the results from this base run. Go to Ecosim > Output > Ecosim results and click Group, landed by and you’ll have a spreadsheet with catches and biomasses by group. Click the second column Biomass (end), and Ctrl+C to copy the end biomass to a suitable spreadsheet (paste with Ctrl+V). Also copy the Group names across.

Next, let’s increase fishing for our forage fish, anchovy.  We can do to do that by increasing the fishing mortality (F) for anchovy. The Ecopath baseline had a fishing mortality (F) of 0.2 year^-1 for anchovy. We can try to increase F to 0.4 year^-1. Go to Ecosim > Input > Fishing mortality, click 6. Anchovy and Set to value.  Now enter 0.4 on the pop-up screen. Run Ecosim, extract the Biomass (end), compare the ratio biomass ratios for the runs with high F and no fishing.

Some questions that may be interesting,

• Which groups are impacted most by the increase in fishing for anchovy?
• Is there a relationship between trophic level (TL) and how much a group is impacted?
  • You can get TL from Ecopath > Output > Basic estimates
  • Where the relationship is poor: is that because anchovy isn’t important for the group?
• Is the relative stability of anchovy in spite of increased F linked to reduced predation mortality?
  • Check the Ecosim > Output > Ecosim group plots, find anchovy where the predation mortality is represented from the baseline to red line. If you put your cursor on the red line, you can read off the predation mortality.  Run the model again with fishing mortality for anchovy set to 0.  Read off the predation mortality. Total mortality is fishing plus predation mortality. How much does the total mortality increase when fishing is increased from 0 to 0.4 year^-1?
  • Yes, it’s a bit cumbersome having to read off values from tables and plots, but it’s quick. It’s also not necessary. If you click the floppy disk icon in the second row at the top of the screen, and click Ecosim > Run results, then run Ecosim, and EwE will write Ecosim results out as CSV files, (which you can open with Excel). You find the files by clicking the floppy disk again, and then clicking the yellow folder symbol to the right of the Ecosim > Run results. There are many files, but the file names tell you what’s in them.
• So, do you find cascading ecological effects due to anchovy exploitation?
• Are there groups that increase when anchovy is exploited? Why might that be?
  • You may get an idea about indirect impacts if you check the Mixed Trophic Impacts graph (Ecosim > Output > Tools > Network Analysis > Mixed trophic impacts > Mixed trophic impacts plot > Options > Colors > Fit to available area. The row marked anchovy shows you the direct and indirect impacts of anchovy on other groups and on fleets with positive in blue and negative in red.
• What happens if you really overexploit anchovy?
  • Try setting anchovy F to 0.8 year^-1.
  • What group is impacted most now?
  • What’s the impact on predators?

The Anchovy Bay model cannot be used as-is to answer questions about impacts on seabirds (since seabirds aren’t included in the model), but see the One-third-for-the-birds study,^[3] which quantified impacts of forage fish exploitation on seabirds.  Or add seabirds to the model – it’s often required to modify a model in order to address specific policy questions.

Energy flow alterations: Does forage fish exploitation change the energy transfer efficiency through the food web?

Similar to the question above, the foundation for this question can be the Anchovy Bay model. There is, however, a problem. The obvious way of evaluating what happens to the energy transfer efficiency under different level of forage fish exploitation would be through time-dynamic simulations. Ecosim does have a number of network indicators being calculated over time, but energy transfer efficiency is not among those calculated.^[4]

Energy transfer efficiency is calculated in the base Ecopath model, see Ecopath > Tools > Network analysis > Flows and biomasses > Transfer efficiency.  To get estimates over time, the simple (but somewhat cumbersome) way to do this now is to extract an Ecopath model for the last year of the Ecosim run, and then extract transfer efficiencies from that model and compare to the original Ecopath model. For this, on Ecosim > Tools > Ecopath model from Ecosim check Enable model generation and check the last year. Then run Ecosim, and Ecosim will save an Ecopath model for that year, named here Anchovy Bay 41.ewemdb. The model will be saved in a folder “Ecopath model from Ecosim” where your output is stored (click the yellow folder icon in the second row at top). Load the new model, and compare the network analysis’ transfer efficiencies between the two models.

Alternate stable states: Are there threshold levels of forage fish biomass that, if crossed, could lead to shifts in ecosystem structure?

Alternate stable states happens in nature, think of the “gadoid outburst” in the North Sea^[5]  as an example. Over the period from the early 1960s up to around 1980, the cod, whiting, haddock and Norway pout all had exceptional good recruitment and their biomasses peaked.  There may never have been consensus on why this happened, but it may well be a consequence of the decimation of pelagic stocks (herring and mackerel) that happened when purse seining really took off around 1960 combined with favourable environmental conditions (a cold period). A possible explanation would be that mackerel and herring predated on the pelagic stages of the gadoids, and kept those down. Given that the gadoids prey on the pelagics as well, such a mechanism may indeed lead to alternate stable state.

We can explore this in Anchovy Bay if we modify the model a bit. For this it is necessary to introduce age structure in the model. We can start with a simple solution where the whiting and mackerel groups are changed to be multi-stanza groups and then let mackerel have a minuscule proportion of whiting larvae in their diet, and whiting feed a tiny bit on juvenile mackerel. We can see if this will lead to Anchovy Bay experiencing alternate stable states by increasing the fishing pressure on mackerel and see if this leads to an increase in whiting that may be sustainable even if the fishing pressure on mackerel is reduced.

To try this, download a version of the multi-stanza version of the Anchovy Bay model from this link.    … and do the following ….

Competition effects: How does changes in forage fish population size impact the competitive relationships between other species in the ecosystem?

Fishery interactions: What are the indirect effects on other fisheries that target predators of forage fish?

Recovery potential: How resilient is the ecosystem to different levels of forage fish depletion, and what are the timescales for recovery?

Spatial considerations: How do the impacts of forage fish exploitation vary spatially, especially in relation to important breeding or feeding grounds for dependent predators?

Climate change interactions: How might the impacts of forage fish exploitation be exacerbated or mitigated by ongoing climate change?

Management strategies: What harvest control rules could be used to ensure sustainable forage fish exploitation?

Socio-economic trade-offs: How can the economic value of the forage fish fishery be balanced against the ecosystem services provided by abundant forage fish populations?