What I Study 2010

Files: IStudyFish.pdf

My name is Ben and I study the relationships and connections between fish and their environment in the ocean. I am studying how many of the fish that live off the coast of Massachusetts, and more broadly in the gulf of Maine, have reacted to man’s fishing efforts, and how global climate change will effect them in the years to come.

Gulf of Maine

These fish are important because they have been caught, here, off the east coast of the United States for hundreds of years, but historically were transported to both other areas of the US and also to Europe. Today, many of the fish are flown as far away as Japan before they are consumed.

The two fish I study are very different, they live different kinds of lives, and they even look very different!

Tuna Cod

The fish you see on the left is a very special kind of tuna fish! It is one of the fastest fish in the ocean, but it’s not the tuna fish that you had for lunch last Tuesday. Almost all of this type of tuna, the Atlantic bluefin tuna, is caught live, and then flash frozen, either on the boat itself or as soon as it is brought to shore and then sent all the way around planet to Japan, where it is served raw and used as sushi!

The fish you see on the right is one of the oldest fisheries in the New World. For hundreds of years European explorers and fishermen came to the Atlantic (East) coast of North America to catch this fish, even giving the name of the fish to “Cape Cod”. Can you guess the name of the fish on the right? If you guessed “Atlantic cod” you would be right! For a long time it was the basis for fish-‘n’-chips both in the US and Canada, and was a staple food in Great Britain, Spain, and the rest of Europe. It is now much rarer than it used to be, and even what we call “locally extinct” meaning some regions that it used to live in, no longer have any of this type of cod.

I am looking at the number of these fish in the Gulf of Maine and especially in the waters off Massachusetts, over a long time period, sometimes over a hundred years. Most data that is considered “current” didn’t start being collected until the 1930s, though a lot of data wasn’t kept scientifically and accurately until after World War II. Atlantic bluefin tuna are an example of this; fished for about 80 years off the coast of New England, reliable data did not begin until the 1950s, and then only on the number of pounds caught, but not the effort used to catch them: the number of boats out fishing, hooks used off the side of the boats, or days at sea that fishermen spent trying to catch the fish.

Fishing Hook

Atlantic cod have an even longer history, primarily from Canada, where data that includes both this effort (hooks, boats, days at sea) and the landings (pounds caught), go back to the middle of the 1800s. We try to “standardize” all of our data, meaning make it equivalent, to what we call “CPUE” or “Catch Per Unit Effort”. In other words, how much time, money, and effort does it take to catch a fish of a certain type in a certain year. By making all of the data equivalent across effort we can compare data from different types of fishing methods across hundreds of years.

While the Atlantic bluefin tuna and the Atlantic cod are the two species I am primarily interested in studying, I have to consider many factors that may have affected them. My research group, a collection of professors (college teachers) and students from four universities, are in the process of entering the data from the 1800s for over 20 species of fish, some that cod and tuna eat, some that eat cod and tuna when they are young, and some that just swim in the same area as cod and tuna. We are also entering data about many physical factors, like the temperature of the ocean along the seashore, at various points along the Northeastern United States and Canada, which has been recorded daily for hundreds of years.

I am trying to answer some pretty big questions, foremost, “What has happened to all the fish?” There are some well-accepted estimates that there used to be twenty times as much “biomass” in the Gulf of Maine as there is now. Biomass may be a new term to you; in general it means the amount of mass of all living organisms in an area or ecosystem at a specific point in time. I am trying to reconstruct the history of the two fish I study, using the historical data I discussed earlier. Think about this, 130 years ago, with only sailboats and hand-held, single-hook fishing gear, the fishermen in Massachusetts, and Maine brought in more Atlantic Cod in one season than is now caught by all of the motorized fishing boats in New England in one year, and not just more pounds of fish, but much bigger, older fish as well. They would salt them to preserve them and dry them on the docks as shown in the picture.

Cod on Pier

The analysis that I use is called “Causality Analysis,” the math is really weird. All scientists, at some point in their education, will encounter the idiom “Correlation is not Causation”; the idea behind causality theory is that you can predict one variable from a variety of other related variables without correlation! The easiest way to think of it is like a plot, you start with the standard X and Y axes for two dimensions, then you add the “Z” axis for depth, the third dimension. If you know a lot of pairings for (X, Y, and Z) points on a graph, with enough points, you can predict what one of the Y points will be if you know the X and the Z points for that same time point. With enough points, you get a narrower and narrower area where the missing “Y” data should be found. What I do is take those (X, Y, Z) points and add as many as 17 other dimensions (variables) to the math and construct a figure, which is called a “manifold,” which means “a figure in n-dimensional space”, though it is easiest to think of in three dimensions like the figure below. One of the things this math allows us to do is take time-series of data that have missing months, years, or even decades for one variable, and predict what the missing data might be based on the other variables in our mathematical model; it’s sort of like a time machine for information about fish.

Lorenz AttractorT

his figure is a simple manifold called the “Lorenz Attractor. If you can cross your eyes you will be able to see this image in 3D.

Another way to think about it is that with enough data you get clusters of related things, which are not necessarily correlated, put still plot on the manifold close to each other.

One Step Ahead

Think about this, in the figure to the right: we know the values of the three open squares in the left box, and we know that the dark circle is between them. In the right box, which in this example is predicting one year in the future, we still know where the squares are in this year. Where would you expect the circle to be? Using this math, because we know that the circle should be between the squares, based on many past observations, we can predict that the circle will still be between the squares, even though we do not have any actual data about the circle at this future time.

By combing the physical data from different points along the coast, like the temperature of the ocean, with the variability of many fish species, like the 20 species we are looking at; we are trying to pick out patterns that can be used to predict what the tuna and cod would have been doing if humans had not been fishing them. Our best estimates of the size of the fish and how many fish were in the area are not actually based on direct measurements of these fish through history, but instead are inferred based on what was caught and brought to market in a given week, month, or year by the fishermen. If we know how much of one type of fish should have been present naturally, and we know how much man took out in a given time period we can construct a better model of the history of the ecosystem.

Ultimately what I would like my research to do is help us set limits on where, when, and how many fish we can take out of a given area of the ocean. One of the things we can learn from this analysis is what variables: temperature, prey fish presence, predatory fish presence, etc., drive this ocean system in a certain direction. We know that man as the top predator plays a crucial role in the lives of both the species I study. But, maybe we are having indirect effects on the population as well; for example, catching all of the herring, a common food choice of both tuna and cod, which may have a time-lagged effect (one or more years later) on the number of adults that we find in the area. This is not something that would show up in a traditional analysis like a regression (what excel does when it fits a line to a set of points on a graph or what your mind is doing when you eyeball a “best fit line”). My research group and I are hoping to be able to advise people who make laws, rules, and policies on what is a safe level of fish to take out of different areas of the Gulf of Maine in a given year. If we know we want to prevent these fish from going extinct we need to not only protect the cod and tuna, but protect what they eat, and the whole ecosystem that they live in.