My two favorite publications of all time

Marilyn and Mat Wooller in Belize-did we have a rigorous hypothesis?

            In 1980, I shared my office with visiting professor Dave Freeman, University of Maryland (Dave is Kate Freeman’s dad!). He was at a critical time in his career and wanted to assess where he’d been and what he’d accomplished.

            Reading over a list of his publications, he said, “Do you know what your best five publications are?”

            “Sure,” I said quickly. At that point in time, I only had 5 publications. It wasn’t difficult at all. Now with over 200 and some papers, a few stand out.

            The first one was published in 1999:

 Fogel, M. L., and N. Tuross, 1999. Transformation of plant biochemicals to geological macromolecules during early diagenesis. Oecologia 120:  336-346.

            I had been asked to give the keynote address to the first international meeting on isotope ecology—IsoEcol 1998. It was presented in the days before Powerpoint projectors were used. I had both photographic slides and overhead transparencies that I had made to give the talk.

            I was fascinated by new molecular techniques that my colleague Noreen Tuross was experimenting with at the Smithsonian Institution. When Noreen took a leave of absence to run the Watson Foundation, I spent a sabbatical in her lab learning Western blots, protein gels, and immunology. At that time, my husband was the Director of the Jug Bay Wetlands Sanctuary on the Patuxent River in Maryland. Every year, I’d watched the rapid growth of marsh plants in spring and summer, followed by their senescence in early fall. My family spent every weekend there for almost 8 years. I decided to design a litterbag experiment, similar to the ones I’d worked on with Ron Benner. My goal was to combine molecular biology, litterbags, and stable isotopes. I learned a lot doing this!

         Understanding the role of microorganisms in influencing sedimentary organic matter evolved rapidly in the 1980s. Today, there are more studies of the structure and isotopic compositions of microbial biomarkers and molecules than there are of compounds from higher plants (e.g. Schouten et al., 2013; Hopmans et al., 2004; Schouten et al., 2002).

         Plants were collected from the marsh in early September. I had a high school intern who was assigned the task of sewing the plants once they were dried and weighed, into mesh bags. Alex Feldman was not used to sewing, to put it mildly, but he learned rapidly and developed a unique style that looked nothing like how a matron like myself would sew. The bags were either buried in deep marsh muck or tethered to PVC pipes, where they sat on the sediment surface. Monthly, I would don hip waders and muck through the mud to pull out a couple of bags of rotting plants.

         It was a labor of love. As Ron Benner would say, “It’s like money in the bank.” The decomposed plants were gently washed free of sediment, dried, weighed, then  ground into a fine powder. I had hundreds of samples from this experiment. Next, we prepared them for “ash free dry weight”, a tedious measurement to quantify how much sediment was stuck to the plants that we couldn’t see. It was important for determining mass balance.

         Next we weighed out a milligram or two and determined the % carbon and nitrogen and the isotopic composition of each sample. The experiment was followed for about 500 days. During that time, I not only carried out the work, but also gave birth to my son, Evan. When he was a baby, Chris and I went together. I carried Evan in a front pack, while Chris did the marsh mucking for me. He would throw the samples from the marsh to me, standing on the dock. Once, there was a very near spill of young master Evan into the Patuxent River! We were more careful after that. 

Marilyn getting litter bags, circa 1991

 

         I also followed the progression of decomposition in my samples with compound specific isotope analysis (CSIA) using gas-chromatography-combustion-continuous flow methods, which had been used in my lab for about 5 years. Noreen Tuross taught me how to conduct ELISA (enzyme linked immunosorbant assay) using two monoclonal antibodies developed for the enzyme Rubisco and a lab-synthesized humic acid, a type of geo-polymer found in sediments. I tried out the modern techniques in protein purification and Western blotting, which visualized pieces of decomposed plant fragments by reaction with specific antibodies.

Black "bands" show proteins in fresh (upper panel) and rotted plants (lower panel)

         I was still fascinated with Rubisco and was curious how this major plant protein degraded along with the other major plant biochemicals.  In non-woody herbaceous plants, Rubisco declined by 40-80% in three months time, but we could still detect intact high molecular weight, 55,000 dalton Rubisco subunits (Fogel and Tuross, 1999). In upland plants with thicker, waxy leaves (e.g., mountain laurel), 80% of the original Rubisco remained intact after one year of incubation in anaerobic sediments. One of the plants I included in the experiment was leaves from locally grown corn plants. Corn, a C₄ plant, was incubated in the C₃ plant environment of the wetland. My hypothesis was that over time, I would be able to detect carbon isotope shifts from corn’s C₄ values to C₃ values originating from organic matter in the wetland. In fact, this indeed happened.

Methods of Western blots and isotopes: a first

         For the first time, and to my knowledge the only time, I purified the different subunits of Rubisco from fresh corn, then measured the carbon isotopes in the amino acids from the two subunits, as well as the amino acids in the bulk corn tissue. We found that the carbon isotopes in amino acids in purified Rubisco subunits had different compositions with respect to bulk protein. In addition, the amino acids within large and the small units in fresh Rubisco had different isotope signatures from each other. The small subunit is coded by a gene in the nucleus but is synthesized in the cytoplasm primarily in leaf tissue, as well as in stems and petals. The large subunit, however, is coded by a gene in the chloroplast and is synthesized within the chloroplast. The complete Rubisco molecule is assembled after the small subunit is imported into the chloroplast (Gutteridge and Gantenby, 1995). Thus, it is not surprising that the carbon isotopes of individual amino acids from each subunit should have different isotopic compositions.

            We also compared the isotope pattern of fresh plant material with decayed plant material. The simpler amino acids (e.g., serine, glycine, alanine) have more positive carbon isotope values, whereas the more complex ones (e.g. valine, leucine, phenylalanine) have more negative values (Fogel and Tuross, 2002). For geochemical studies, a sample’s amino acid fingerprint--the relative differences in carbon isotopes among amino acids--could be altered during microbial decomposition either by the preferential breakdown of amino acids or addition of microbial proteins leading to a heterogeneous mixture.

            When I presented the work to Ron Benner’s lab in Texas, he had one of his generally abrupt comments, “This work is kind of crude,” he remarked. Fortunately, I understood that he meant that I’d only measured the proteins and amino acids, and not the lipids and carbohydrates. Noreen and I learned what it takes to do something completely different like this. Although I didn’t repeat that type of study, I am proud of it today.

            My second favorite publication came from the investigations in Belize:

Fogel, M. L., M. J. Wooller, J. Cheeseman,  B. J. Smallwood, Q. Roberts, I. Romero,  and M. Jacobson Meyers,  2007. Unusually negative nitrogen isotopic compositions of mangroves and lichens in an oligotrophic, microbially-influenced ecosystem. Biogeosciences 5: 1639-1704. I described the work earlier, but want to tell the story of its publication.

         I submitted the first manuscript to the journal Biogeochemistry. It was long, had many figures, and was written in the historic way in which I figured out why we had measured such unusual isotope compositions in mangrove leaves. The paper was roundly rejected! 

“The manuscript suffers from unfocused objectives and poor organization. The dataset offers potential to address several objectives.  However, addressing them all within a single paper is not likely to complement the strengths of this dataset and rather detracts from a central message.”

More complicated studies often get reviews like this. A second review noted the following:

         “The data are interesting, overall — but I am not persuaded by the authors' conclusions based on their interpretation of the data. In particular, I am bit concerned that the authors' arguments for substantial uptake of NH₃ gas might not be plausible when put in the context of annual plant N demands and isotope mass balance. Moreover, the hypotheses put forth are not sufficiently teased apart by the authors; the manuscript comes across as very speculative at times… I would strongly urge the authors to identify a more rigorous question based approach, which provides mutually exclusive hypotheses and corollaries that are followed through to the end of the manuscript… The results section is way too long. It wanders considerably, filling the reader with information that is tangential to the manuscript. I would strongly recommend that the authors remove data and discussions of C isotopes from this manuscript, and focus on the N isotope patterns as they relate to foliage. ”

         What could we do? The work was finished, the data—thousands of data points—were collected. For young scientists, the message is—don’t give up!

            I revamped the manuscript, sent it out again, and once again, the work was criticized and viewed not publishable. It was a complete shock. In science, manuscripts are reviewed by our peers—it was clear to me that someone did not want the work published. Given the fact that I had ended the mangrove study on a sour note, it troubled me to think that those problems might have spilled over to publishing our findings.

            In response to this, the 3^rd version of the paper was submitted to the journal Biogeosciences, a journal that publishes not only the manuscript, but the reviews, and my response to those reviews. Interestingly, the reviews were now positive, non-personal, and the paper was accepted quickly. Isotope and mangrove expert Steve Bouillon wrote:

“The authors reasonably argue that ammonia from the atmosphere and in rainwater are likely important N sources, and that the d¹⁵N signatures observed in mangrove leaves reflect a balance between these N sources and N uptake from sediment porewaters. This balance is also demonstrated to be governed by the availability of P in this P-limited system. The paper is a nice contribution to our understanding of N cycling in mangrove systems, and relevant more generally for our understanding of N isotope patterns in vegetation and its utility to understand ecosystem nutrient cycling. I recommend publication in Biogeosciences, although the ms. could be improved if a number of issues are clarified.”

            Ecologist and isotope scientist Erik Hobbie wrote:

“Overall, the paper makes a solid case that several mechanisms other than soil N source isotopic signatures and mycorrhizal processes may control plant ¹⁵N patterns in mangrove systems. This paper points the way towards quantitative assessment of plant-atmosphere N fluxes as one fruitful avenue for providing us with a more complete picture of the causes of plant isotopic patterns.”

I think this is a great model for peer review—the reviewers here chose to reveal themselves and their reviews. Gone were the preachy reviews described above telling me to come up with a more rigorous hypothesis. If nothing else, I learned to be as polite as possible in a review and try to always hope to bring out the best in others’ work.