Mistakes made/lessons learned^†

I have made a number of mistakes during my scientific career and have also learned a number of lessons that I believe are worth passing on. One of the first lessons I learned was that scientists, by definition, need to make mistakes. My advice to students in the sciences, both graduate and undergraduate, is to make mistakes and learn from them. Supervisors should be encouraging students to make constructive mistakes.

To reinforce this advice, I share three of my mistakes. Early in my career, I was trained in aquatic oligochaete taxonomy by a world expert, my doctoral supervisor, Ralph Brinkhurst. One day I found what I believed to be a new species and, after writing a draft description for publication, took my findings to Ralph. He was very gentle when he told me that in fact all I had found was the polychaete Capitella capitata! Subsequently, I did find and publish on new species of oligochaetes, but I was a lot more knowledgeable and careful as a result of this mistake.

At the time I finished my PhD, there were no courses in environmental toxicology. I was doing work in this field because I was testing the tolerances of aquatic oligochaetes to both salinity and metals. I learned by reading the human toxicological literature. As a result, the first paper I published on environmental toxicology, which should have discussed lethal concentrations, instead incorrectly discussed lethal doses [1]. No one pointed out my mistake; rather, as my reading and understanding of environmental toxicology progressed, I realized what I had done. Again, I learned from that mistake and expanded my reading beyond human toxicology.

A more recent mistake made for several years as a practicing professional was to discount the importance of bioaccumulation testing compared to toxicity testing. I well remember any number of platform presentations in which I stated “bioaccumulation is a phenomenon, not an effect.” This is quite true, but I missed the possibility that this phenomenon might be useful as both an indicator and predictor of toxicity. In fact, today research into contaminant body residues for such purposes shows great promise [2].

In 1981, after finishing my PhD and working as a consultant, I was awarded a contract to work with the National Oceanic and Atmospheric Administration in Puget Sound, Washington State, USA. The National Oceanic and Atmospheric Administration contracting officer was Ed Long, who was then working out of an old seaplane control tower at the naval air base. I learned a lot from Ed, but the most valuable lesson I remember was articulated in a poster he displayed prominently on his wall: “There is no limit to what you can achieve if you do not care who gets the credit.” I truly believe that. Whenever I have held to this precept I have been amazed by what is achieved; whenever I have not, I have fallen short. I advise others to follow this principle and, more specifically, to avoid politics and personal attachments to theories and focus on the science. After all, a theory is only that. Scientific knowledge accumulates upon discarded theories, and widely held theories, especially our own, need to be challenged.

A more painful lesson I learned a few years later while working with Ed was that science can be censored by politicized managers. At that time, there was increasing focus in Puget Sound on the then-new problem of contaminated sediments and bottom fish with lesions. The news media seemed to be running alarming stories about new findings every month. At the same time, Metropolitan Seattle's sewage treatment plants were being pushed to upgrade from primary to secondary treatment, and the two issues were linked by certain individuals working for the U.S. Environmental Protection Agency. Those individuals were not happy with a report produced for Ed Long, of which I was a coauthor [3]. This report was commissioned by the National Oceanic and Atmospheric Administration to detail the status of contaminated sediments in Puget Sound and make predictions based on various possible scenarios. One of these scenarios involved sewage treatment plants upgrading to secondary treatment. Although the report acknowledged the environmental benefits of secondary treatment, it pointed out that this expensive undertaking would not solve the problem of contaminated sediments, which were largely due to historic inputs. Several hundred copies of the report were published and delivered to a senior manager, who promptly burned them all and shortly thereafter retired. I do not dispute the benefits of secondary treatment of sewage where required, but I do dispute the right of administrators to hide scientific findings to push their own agenda, as occurred in this case. I hope that future administrators will be more honest, and that scientists will loudly protest any efforts to bury the truth.

A lesson that frequently has been reinforced is that science does not always yield the results one expects or desires. The answers we get from our studies may not apply to the questions we are asking, and we need to be very careful not to be constrained by our expectations or desires. For example, the traditional dose-response curve that I learned and relied on as a student clearly is no longer the rule and may not even be the exception if the phenomenon of hormesis is as ubiquitous as it appears to be [4].

A lesson that certainly is not unique to the sciences is the fact that ease of publication is directly related to “name” recognition. This has been demonstrated repeatedly in the field of literature, when “name” authors who have written best sellers send manuscripts to publishers under assumed names and receive rejection notices, then resubmit with their own names and are immediately accepted for publication. It also applies in the sciences. I studied biological oceanography in graduate school, working on the uptake of dissolved organic matter by a crustacean. At that time it was believed that only soft-bodied aquatic organisms could take up dissolved organic matter for nutritional purposes through their skin. In fact, I showed that my crustacean, a marine copepod, could take up dissolved glucose from seawater through glands in its cuticle [5]. However, it took several years and many rejections before I could publish this work, because it did not fit with the generally accepted dogma of the time.

Many years later I discovered the reverse problem. I was working on a manuscript that would have benefited greatly from peer review. At that time I had presumed far too much on friends to read over and comment on my manuscripts, so I thought I would instead submit the manuscript to a journal. I knew it was neither ready nor good enough for publication, but figured that this would be an easy way to get peer reviews to help me finalize it. Imagine my surprise when the reviewers suggested only minor changes and the editor accepted my manuscript for rapid publication. Of course I withdrew it, in some embarrassment. I had learned the other part of the lesson regarding publication—sometimes it is too easy to publish, especially if you have a “name.”

I am not sure everyone has learned that lesson. For instance, too often in the literature I see papers that can be termed lowest publishable units or “LPUs.” Because of pressures to publish, too often high quality is sacrificed for volume. Thus, instead of publishing one extremely good, definitive paper, some authors will divide their work into LPUs, which are published separately, often in different journals. I find this tendency annoying to say the least. To understand the whole, one has to find all the pieces, which is not always easy.

I offer two pieces of general advice. First, peer review should be author-neutral. In other words, peer reviewers should receive manuscripts without knowing who wrote them. Second, “name” authors should not follow the LPU route. This advice is similar to that provided by Dr. René Schwarzenbach, in his presentation accepting the Society of Environmental Toxicology and Chemistry (SETAC) Environmental Education Award at the SETAC Europe meeting in Madrid, Spain, 2001. He advised against “average” papers. He also strongly recommended maximizing interdisciplinary courses and interactions.

I agree with both pieces of René's advice; for instance, we should be doing ecological toxicology, not environmental toxicology [6]. Ecological toxicology requires the involvement of ecologists as well as environmental chemists and toxicologists, and other disciplines as needed. The choice of taxa to test or study must be based in ecology. We must remember that species have different tolerances, even within species, in some cases because of acclimation or adaptation to chemical or other stresses. And we must also remember that there is no universal “laboratory white rat.”

We also must realize that a species may not be. Researchers need to know their organisms and know the limits of taxonomy. A species is typically defined as “related organisms potentially capable of interbreeding.” However, a more common definition, because we rarely investigate whether invertebrate species can interbreed would be any published designation by a taxonomist. Or to put it a little more elegantly, the species concept is an intellectual construct that “assumes that it is scientifically valuable to give a single name to a concept that includes organisms having arbitrarily selected characters in common” [7]. Sometimes species designations are very superficial. Some time ago, Ralph Brinkhurst and I [8] published a paper in which we changed species of oligochaetes based on simple manipulations of their culture water: pH, salinity, and water hardness. Basically, we simply induced them to grow or lose hairs or other calcareous growths that were being used by some taxonomists to publish descriptions of new taxa. We called our paper “Hair today, gone tomorrow”—no harm in having some fun doing science! The use of hairs or other external growths to distinguish taxa is not restricted to aquatic oligochaetes and, unfortunately, continues to this day. Similar problems in paleontology have led to calls for abolition of the species concept [7].

In some cases, a single species may be more. The best known example of this is the polychaete Capitella capitata, which was recognized by electrophoresis as consisting of several sibling species [9]. I remember asking Fred and Judith Grassle, shortly after their discovery, how benthic ecologists such as myself should deal with this finding because the sibling species distinctions were not amenable to standard taxonomic procedures. They had no good answer. Even today we have no good answer, other than to refer to the Capitalla species complex.

This brings me to one of my final points: Ecology (i.e., reality) is too complex for simple, all-encompassing rules. I agree with George Bernard Shaw that “the Golden Rule is that there is no Golden Rule” [10]. For example, an incredible amount of effort has been spent and continues to be spent, on deriving sediment quality values to replace other measures such as toxicity testing and benthic community structure analyses for management decision making. A few years ago my colleagues and I [11] reviewed all sediment quality values developed at that time by anyone, anywhere in the world, for metals and metalloids and compared them with sediment background values. In all cases, the range of sediment quality values spanned several orders of magnitude, and many sediment quality values were below sediment background values. The range would have been tightened if we had segregated sediment quality values designed to predict effects from those designed to predict no effects, but the range still would have been large and would still have varied between developers and jurisdictions. This is just one case where we should heed Aristotle's advice: “It is the mark of an instructed mind to rest satisfied with the degree of precision which the nature of the subject permits and not seek an exactness where only an approximation of the truth is possible.”

I conclude with two final lessons, which I consider the most important ones (of many) learned during my career. The first relates to discussions several years ago with Dr. Peter Landrum, of the National Oceanic and Atmospheric Administration. The discussions went along the vein of: “What we are doing is very enjoyable and useful in its limited way, but the world is in a lot of trouble and are we really doing anything meaningful in that bigger perspective?” I have thought long and hard on those discussions and that question, and I believe that what we do as ecologists, environmental chemists, and toxicologists is indeed worthwhile. But we need to make it worthwhile in the bigger context by two means. First, we must keep balance and perspective in choosing our research areas and in presenting our results, so that we are building upwards with others, not digging downwards in isolation. Second, and similarly, we need to ensure that we integrate our results appropriately with past work and with projected future work in a meaningful synthesis. In other words, our work needs to move the knowledge base forward as meaningfully as possible so that other researchers have a higher platform to start from. If we do this, there is no reason that science should not be fun. In fact, it should be.

My final lesson relates to priorities—I have four. My first priority, over and above everything else, particularly after September 11, 2001, is my loved ones. I made too many mistakes early in my career, putting work ahead of them. I will never make those mistakes again. I urge all scientists, be they students or established professionals, not to make this mistake. Nothing is more important in this world than our loved ones, and the greatest gift we can give them, particularly our children, is our time and our selves. My second priority is respect and good manners. There are too many cases of scientists making personal attacks on others, or of scientists taking professional criticism personally. Good scientists will have both respect and good manners [12]. My third priority is common sense, which is too uncommon in human affairs. Without common sense we can too often fail to see the bigger picture or make what we do as meaningful as possible. My final priority is of course good science, which is what we are all involved in.

My bottom line is simple. Be proud of your work and do it as well as you can without stinting your loved ones or demeaning others. We only live once (that we know of)—let's make it as good as we can for ourselves and others, both as scientists and as human beings.