Opinion by Merrill Rudd

As scientist John Shepherd aptly pointed out, “managing fisheries is hard; it’s like managing a forest in which the trees are invisible and keep moving around.” Unquestionably, as bioGraphic reported last October, counting fish is a nearly impossible task. Perhaps an even greater obstacle in managing fisheries, however, is managing people.

The world’s fish stocks are our last wild food resource, but are under increasing pressure as global populations expand. Billions of people now rely on fish as a source of protein and income. In many of the world’s poorest countries, fish is the cheapest and most accessible protein source, providing up to 60 percent of total dietary protein. With an increasing global population and rising demand for seafood products, maintaining resource sustainability desperately requires effective management systems.

One important part of fisheries management is the monitoring of fish catches. Collecting information about the harvests people take from the sea each year might seem like a simple endeavor, since plenty of tools and processes exist to help people report how much they catch. Indeed, in many fisheries, boats are monitored in ports as they offload their catch. However, logistical challenges, misaligned incentives, distrust, and illegal operations often make it exceedingly difficult to count the total number of fish of each species caught per year.

Reported catch in some cases may be a vast underestimate of the true catch, and this is generally perceived to lead to overfishing. To combat this issue, many agencies have tried over the years, with varying success, to improve monitoring programs and to construct more accurate estimates of the total number of fish caught. While I think these efforts are worthwhile, I would argue that it’s actually more important to focus on the trends in these numbers—tracking their change or consistency over time—than it is to improve total counts.

To inform fishery management decisions, my colleagues and I use computer models of fish populations to ask “what if” questions around the big uncertainties—from people’s behaviors to environmental fluctuations. Such models have become the basis for both setting sustainable catch thresholds and for ensuring that the rules for managing fisheries hold up in the face of uncertainty. A key assumption in most of these models is that the total catch is known, without error. If the catch data used in the model are incorrect, model estimates of population size, exploitation rate, sustainable yield, and stock status may also be incorrect—but not necessarily.

Misreported catches come in many forms that can be difficult to track. The most common types of unreported catch are undesired or threatened species thrown overboard that don’t survive; small-scale operations that sell, barter, or subsist on their catch; and illegal operations in which large-scale fleets may be operating covertly. It’s also possible for fish catches to be over-reported—an outcome that sometimes occurs when fishermen are incentivized to establish a history of a larger harvest.

While there has been increased effort globally, via on-board observer programs, to estimate how many fish are thrown overboard, the decentralized or illegal nature of the other unreported sectors make them inherently difficult to quantify. In some cases, fisheries scientists attempt to reconstruct what a previous year’s catch might have been using clues from market prices, government documents, interviews with fishermen, photographs, and many other sources. But these methods rely on an entirely different set of assumptions that could also lead to biases in estimated catch.

Our recent study in Fish and Fisheries used population models to determine how sustainable harvest estimates might be biased under different patterns of catch misreporting. The study showed that when catch is misreported at a constant rate, the sustainable harvest numbers recommended by the scientific community are biased proportionally to the level of misreporting—so the exploitation rate targets are actually set appropriately.

For example, if 50 percent of fish catches are consistently unreported, estimates of population size and sustainable yield will be 50 percent lower than they should be. Despite the misreporting, estimates of the current exploitation rate and status of the fishery will still be accurate because they are ratios between the reported catch, estimated population size, and estimated sustainable yield, all estimated to be 50 percent lower than they should be. To put it another way, in terms of sustainability, it’s really no different to harvest 100 tons of fish from a population size of 1,000 tons than it is to harvest 200 tons from a population size of 2,000 tons. In both scenarios, the exploitation rate and fishery status estimates will be the same. This is how consistent under- or over-reporting of fish catches can still result in management recommendations that are sustainable. While this does not solve social or economic equity problems in the fishery, management based on this misreported level of catch would prevent overfishing.

In contrast, when catch misreporting changes over time, the model becomes unreliable. This can lead to lower-than-necessary catch limits if reporting rates improve and overfishing if illegal harvests increase. Neither scenario is ideal: If catch limits are unnecessarily low, legal fishermen lose out on revenue; and if they’re not high enough, fish may not be able to reproduce quickly enough to withstand the fishing pressure, potentially causing the population to collapse.

Just like weather models, fisheries models attempt to illustrate both current conditions and future projections. Ultimately, the goal is to estimate how many fish can be sustainably caught. A common misconception is that if illegal operations are harvesting fish in addition to the legal catch being hauled in by upstanding operations, we must be overfishing. This is not necessarily the case. As an example, the Commission for the Conservation of Southern Bluefin Tuna recently discovered that actual catches in recent decades were much higher than had been reported. Fear of overfishing shook the community. However, when scientists re-ran the assessment models with each year’s new catch numbers, the estimated level of sustainable harvest was actually greater than they had previously concluded.

During the years with higher illegal harvest, the population models indicated that the fishery was declining, and the legal harvest was accordingly ratcheted down. When the illegal catches were halted, the population models correctly predicted a population recovery under the (lower) legal catch limits. While illegal fishing resulted in some operators wrongfully taking a greater piece of the pie than others, managing the population with the lower estimate of sustainable yield buffered the population from overfishing.

Tens of millions of dollars have been spent by various agencies worldwide, both government and private, to reconstruct the total number of fish we harvest from our oceans each year. These reconstructions help us learn what has been removed from the oceans over time, how local economies use their fish resources, and how to improve food security in coastal areas. However, perhaps the most important use of these reconstructions for fisheries management is in assessing the consistency of misreporting rates over time. If consistency has been high, management based on population models has likely been sustainable. In fisheries where catch reporting has been inconsistent, we probably need to take a closer look at those management plans. Simply put, we can manage our world’s fisheries more effectively by using existing data differently to carefully track reporting trends, rather than trying to measure the magnitude of misreported catches.

This is not to say that illegal harvest is insignificant. It’s quite the opposite, in fact. Illegal fishing operations impact large-scale economics, taxation, local revenue, employment; they also take economic advantage of legal fishermen, and benefit from providing low wages and poor work conditions to crew members from developing nations. Additionally, these operations often use illegal fishing gear, which can damage habitats and impact ecosystems by catching non-target species and juvenile fish. We should absolutely continue to put resources into eliminating illegal fishing operations. In the meantime, however, we need to pursue management strategies that include monitoring fish populations, a model-based assessment of each fish population’s status, and a rule for controlling how many fish can be caught in a given time period.

Managing people is no small feat in our effort to sustainably manage fisheries, and navigating social-ecological systems can be fraught with complexities. Efforts to better understand how misreported catch has changed over time may not only be cheaper and more feasible than reconstructing the total harvest but nearly as valuable. We should be asking whether catch data today is more reliable now than in the past, or if catch data has been consistently under or over-estimated. We can find answers to these questions by talking to local experts, or digging into the catch reconstructions and examining whether the assumptions made to estimate total catch adequately answer these questions and reflect the potential variability over time.

The catch data can then be used alongside other pieces of information in population models to tell a story about population status. Are decreases in catch attributed to management interventions, a decline in the population size due to high fishing pressure, or years of reproductive failure? Are increases in catch attributed to unregulated increases in fishing, years of reproductive success, or changes in the ecosystem? These models should be our starting point for making tough harvest decisions to keep wild fisheries sustainable into the future.