GFDL - Geophysical Fluid Dynamics Laboratory

Reconciling Ocean Productivity and Fisheries Yields

January 23rd, 2017

Charles A. Stock, Jasmin G. John, Ryan R. Rykaczewski, Rebecca G. Asch, William W. L. Cheung, John P. Dunne, Kevin D. Friedland, Vicky W. Y. Lam, Jorge L. Sarmiento, and Reg A. Watson. PNAS. DOI: 10.1073/pnas.1610238114


The authors explore the complex relationship between phytoplankton production and fish, using recent critical advances in our knowledge of global patterns in fish catch and fishing effort, as well as the plankton food webs that connect phytoplankton and fish. A high-resolution global earth system model, developed at GFDL, was used to assess the potential magnitude of future changes in fish yield under climate change. This model has ten times the resolution of a typical climate model and includes comprehensive plankton dynamics.

The same phytoplankton production and food web dynamics that act together to create very large regional differences in fish catch in today’s ocean also act together to amplify trends in phytoplankton production expected under climate change. The projected changes in fish catch generally go in the same direction as the changes in phytoplankton productivity, but the changes in fish catch are larger. This is because warming and a decline in phytoplankton production also lead to longer, less efficient food webs connecting phytoplankton and fish. Many low to mid-latitude regions projected to experience modest to moderate declines in phytoplankton production (e.g., decreases ~5-15%) may experience catch decreases which, in some cases, may exceed 50% (Figure 1).

Figure 1: Projected percentage change in phytoplankton production (left) and maximum fisheries catch (right) for a high greenhouse gas emissions scenario (RCP8.5) between the latter half of the 20th century (1950-2000) and the latter half of the 21st century (2050-2100).
Figure 1: Projected percentage change in phytoplankton production (left) and maximum fisheries catch (right) for a high greenhouse gas emissions scenario (RCP8.5) between the latter half of the 20th century (1950-2000) and the latter half of the 21st century (2050-2100).

Microscopic phytoplankton provide the energy that supports nearly all marine life. It is thus tempting to assume that the number of fish we catch from a patch of ocean may vary in proportion to phytoplankton production. The conceptual simplicity of this idea has caused it to linger in the popular and scientific literature. However, even sparse catch and phytoplankton production data available half a century ago suggested that phytoplankton production was acutely insufficient to explain fish catch differences across globally distributed ecosystems.

Dr. John Ryther, publishing in the journal Science in 1969, noted that differences in fish catch were far more dramatic than differences in phytoplankton production might suggest. Specifically, coastal systems where large amounts of nutrients critical for phytoplankton growth were “upwelled” from depth via wind driven currents made a contribution to global fish catch that far exceeded what one would expect from phytoplankton production alone. Ryther hypothesized that differences in primary production and food web properties must act together to create such large catch disparities. In short, phytoplankton production in upwelling regions is higher, and it is delivered more efficiently to fish by the marine food web.

The authors’ analysis of the relationship between phytoplankton production and fish catch, using several new tools, generally supported Ryther’s hypothesis. Phytoplankton production alone could not explain regional patterns in fish catch per ocean area that varied by a factor of 100 even after controlling for fishing effort. Using this factor alone would always overestimate catch in low-catch systems and underestimate catch in high catch systems.

However, these large catch differences could be reconciled if one accounted for the tendency for short food webs that efficiently transfer energy from phytoplankton and fish in colder, higher productivity systems, and longer food webs that transfer energy from phytoplankton to fish less efficiently in warmer, low productivity systems.

The climate sensitivity of ocean productivity arises from phytoplankton’s need for nutrients (e.g., nitrogen, phosphorus, iron) and light to grow and create the energy that eventually supports fisheries. In the ocean, however, nutrients tend to accumulate in deeper waters while light is prevalent only at the ocean surface. Furthermore, the oceans upper layers tend to be warmer and, in many cases fresher than deeper waters. These properties make surface water less dense than deeper waters. This stable “stratification” works to keep nutrient rich deep waters and the well-lit euphotic zone segregated. The ocean relies on a variety of processes to work against stratification and bring nutrients and light together. For example, cooling of waters in the winter and high winds can break down stratification. Winds also move surface waters and, if oriented correctly, can create “upwelling” currents that carry deep nutrients to the surface.

Climate change warms surface waters and increases stratification. In the low- to mid-Latitude oceans, this tends to reduce the supply of nutrients and phytoplankton productivity (note the prevalence of blue regions in Fig. 1 above), though changes in other factors at regional scales can overcome this mean trend. At very high latitudes (e.g., the Arctic), light is a far more important limiting factor then nutrients, so a bit more stratification and reductions in sea ice tend to have a positive impact on phytoplankton productivity. Food web dynamics act to accentuate changes in both decreases and increases.

The data advances that this study benefits from are notable in several respects:

  1. One of the greatest challenges Ryther faced in linking fish catch and ocean productivity is separating out the effects of changes in fishing effort from differences in ocean productivity. Co-author Reg Watson has developed and recently published a global fishing effort database measuring the number and power of fishing vessels and days spent at sea, that allows us to control for this factor.
  2. While catch data from industrial fisheries is reported to the Food and Agricultural Organization, estimates of catch from discards and small-scale artisanal fisheries can require extensive engagement at local levels. We benefitted from a recently published catch reconstruction led by Daniel Pauly at the University of British Columbia that does just that.
  3. While satellites have provided a global perspective on phytoplankton production from space, they cannot observe the plankton food web dynamics that connect phytoplankton and fish. To address this, the authors used a prototype high-resolution global earth system simulation developed at GFDL. With comprehensive plankton dynamics and ten times the resolution of a typical climate model, its computational costs quite high. The authors used a short run to explore patterns in today’s ocean, while using coarser simulations to explore climate change implications.

Assuming a simple relationship between phytoplankton production and fish catch risks overestimating the resilience of fish catch to changing ocean productivity. We need to develop management strategies that accommodate the possibility for larger changes in baseline catch levels to ensure the continued social and economic benefits from these resources.