How to read a marine biology paper as a beginner Image: U.S. Fish and Wildlife Service (public domain)

How to read a marine biology paper as a beginner

A scientific paper is not written to be read from the first word to the last, and the students who try to do that give up on page two. It is written in a fixed order that lets a reader take what they need and leave the rest, and once you know the order, a paper of twelve pages takes an hour and gives you more than a textbook chapter. This post goes through the method on one paper that the masterclass uses, Camilo Mora and colleagues' 2011 estimate of how many species there are on Earth and in the ocean, which is free to read in PLoS Biology. Open it in another tab.

Read the title and the abstract, and write down the claim

The abstract is a summary written by the authors, about 250 words, and it contains the question, the method in one sentence, the main result with a number, and what the authors think it means. Read it twice. Then write down, in your own words, the single claim the paper makes. For Mora's paper it is something like: there are about 8.7 million eukaryotic species on Earth, about 2.2 million of them in the ocean, estimated from the pattern in which higher taxonomic ranks have been described over time. If you cannot write that sentence, read the abstract a third time before going on. Everything else in the paper is support for it.

Go to the figures next

Skip the introduction for now and find the figures. Scientists read figures first, because a figure is where the data live, and a paper with good figures can be understood from them alone. For each one, read the caption, then the axes. What is plotted against what? What are the units? Are the axes logarithmic? In Mora's Figure 1 the number of described taxa is plotted against the year, for each rank from phylum to species, and the shapes of the curves are the whole argument: the high ranks have flattened, the low ones are still rising. Figure 2 fits a curve through the ranks and reads off where the species rank should end up. Look at the error bars, and if there are none, notice that too.

When you have looked at every figure, check your sentence. The figures should have shown you the claim. If they showed you something else, one of you has misunderstood, and it is usually worth finding out which.

Now the methods, for one question only

The methods section is written for someone who wants to repeat the work, and a beginner does not. Read it with one question in mind: what did the authors have to assume? Every method rests on assumptions, and the honest ones are stated. Mora's method assumes that the curve of ranks has the same shape in groups we know well, such as birds, and in groups we know badly, such as nematodes. It assumes that taxonomists have been consistent about what counts as a genus or a family. The authors say both. Write the assumptions down under your sentence, because they are what you will need when someone disagrees with the paper.

The results, with a pencil

The results section repeats the figures in words and adds the numbers. Read it with a pencil and underline every number that has an uncertainty attached: "8.7 million (±1.3 million SE)", "2.2 million (±0.18 million SE)". The uncertainty is the most informative part of a result, and it is the part that gets dropped when the paper is reported in the news. A result given as a range, or with a standard error, is telling you how far it could be wrong by its own method. It is not telling you how far it could be wrong if the assumptions fail, and that difference is where most scientific arguments live.

The discussion, and the paper that disagrees

The discussion is where the authors say what they think it means, compare it with earlier work, and admit the limits. Read it for the limits, which are usually in the last third. Then do the one thing that separates a reader from a student: find a paper that disagrees. For Mora's estimate it is the 2012 survey by Ward Appeltans and 120 taxonomists, which put the marine total at 700,000 to a million. Read its abstract and its first figure the same way. You now know what the two methods assume, where they agree, and why they differ, and that is a better understanding of the question than either paper gives alone. The post on counting species goes through the disagreement.

Where the uncertainty hides

Three places to look. The sample size, often given as \(n\), which tells you how much data the conclusion rests on; a result from six individuals is a different thing from a result from six thousand. The word "significant", which in a paper means that a statistical test passed a threshold and does not mean large or important. And the model. Many marine biology results are not measurements but the output of a model fed with measurements, and the model's assumptions are then part of the result. The global primary production of 50 billion tonnes of carbon a year, used throughout the masterclass, is a satellite estimate passed through a production model, and the papers that report it say so.

A second paper, for practice

Try the same method on Séverine Martini and Steven Haddock's 2017 paper on how common bioluminescence is, which is free in Scientific Reports and is the research spotlight of Unit 13. Its figures are easy: a bar chart of what fraction of the animals seen on 17 years of remotely operated vehicle dives off Monterey can make light, by depth and by group. The claim is one number, 76 per cent. The assumptions are about which animals count as luminous and how a camera survey samples the water column. It is a paper you can read in 40 minutes and then explain to someone else, which is the test of having read it.

Finding papers you can read

Most marine biology papers are behind a paywall, and most of those have a free copy somewhere. Unpaywall is a browser extension that finds the legal free version of a paper when one exists. Google Scholar shows "All versions" under each result, and one of them is often a PDF on an author's or a university's website. Semantic Scholar does the same with a summary. Preprints, which are papers posted before peer review, are free on bioRxiv, and a preprint is fine to read as long as you remember that nobody has checked it yet. Several marine journals are entirely open, among them Frontiers in Marine Science, and Marine Ecology Progress Series makes its articles free after five years. If all else fails, email the first author. Almost every scientist sends a PDF within a day to a student who asks politely and says why they want it.

Every unit of the Marine Biology Masterclass ends with a reading list of two to four papers, all with free full text, and a research spotlight on one of them. Unit 16 closes the program with a section on how to read and where to find the literature, and on where cruises and student programmes are advertised. The full outline is in the curriculum.

Study the ocean properly

The Marine Biology Masterclass: weekly units, quizzes, a final exam, and a certificate.
About the program