In 2002, Vojtech Novotny, Yves Basset, Scott Miller, George Weiblen, Birgitta Bremer, Lukas Cizek and Pavel Drozd published a paper in Nature providing evidence for low host-specificity of tropical herbivorous arthropods. This finding was based on data from over 900 herbivores from 51 plant species in a rainforest in New Guinea. Based on this finding, Novotny and colleagues provided new global estimates of arthropod diversity in the range of 4 – 6 million species, which was substantially lower than the accepted figure at that point in time. Fourteen years after its publication, I spoke to Vojtech Novotny about the making of these findings and their validity today.
Questions sent by email on 19th June 2016; responses received by email on 17th July 2016.
Hari Sridhar: What was the specific motivation to write this paper? Was the fieldwork done with this paper in mind or did the idea for the paper come later, after the work was started?
Vojtech Novotny: The paper is called “Low host specificity…” and the study of host specificity of tropical herbivorous insects was the primary motivation for the underlying study – the grant proposal (for US NSF) and the field study. So there had always been a plan to write such a paper. It was only unclear what form it will take and for what journal. At the end we wrote two papers, one with more detail for J. Anim Ecol. and one for Nature. The latter benefitted from a plant phylogeny for our studied species – something we were not sure we could get at the beginning of the study. Further, the Nature paper also revised the global estimates of insect diversity that had always been based on host specificity estimates. In the first version of the paper (submitted to Nature), we only said that our host specificity results will lead to new, greatly revised, estimates of global diversity, but did not actually perform the calculation. Both reviewers and the editor suggested that we expand this section and revise the global diversity estimates ourselves, which we finally did. So that part of the paper was not foreseen when we were doing the field research.
HS: This paper has seven authors. How did this group come together?
VN: The authors include the team that wrote the original research proposal (Novotny, Basset, Miller, Weiblen), plus two postdocs who joined the field work (Cizek, Drozd) and a collaborator recruited during the project to help with the plant phylogeny (Bremer). This group (except Bremer) was involved – entirely or partially – in many previous and subsequent papers on the ecology of New Guinean insects, and to this day represents a collaborative team. Its senior members have complementary roles and specializations – Novotny and Basset: insect ecology, Miller: insect taxonomy, Weiblen: plant ecology and phylogeny and Bremer: plant phylogeny. The Novotny – Basset – Miller – Weiblen collaboration has been going on for 20 years now and remains focused on plant-insect food webs in Papua New Guinea (PNG).
My collaboration with Yves Basset goes back even further. Many years ago, I had applied for a University of Leeds position to study the insects on fig trees in Krakatoa. Though I didn’t get that position, in preparing for the job interview, I read a lot about figs and got very interested in the system. Sometime later I met Yves at an entomological congress. Yves was also interested in insects on fig trees and over a pint in a York pub our New Guinea collaboration started.
HS: The dataset used in this paper seems unique in its size and resolution. Can you tell me how this data was collected and over what period of time?
VN: The study has – in my view – three features that distinguished it from others at the time (and got it into Nature), viz. (i) the large scale of the data set, (ii) direct study of trophic interactions by feeding experiments, and (iii) plant phylogeny used in the measurement of host specificity. The data set was collected over seven years of continuous work in PNG rainforests, where we were sampling one set of (live) plants after another, using teams of trained paraecologists recruited from local villages. This approach was described in our papers in Bioscience and J. Appl Ecol.
© Vojtech Novotny
Another important aspect of the study was sustained presence of researchers – first Basset, then me, then also Cizek and Drozd. In this way this study created a basis of a research team that later expanded and presently constitutes the New Guinea Binatang Research Center.
HS: Was all the insect curating and identifying done in field? Where was the plant phylogeny work done?
VN: The insect sorting to morphospecies and databasing was done at our PNG base, and the final taxonomy overseas, mostly under Miller’s leadership.
© Vojtech Novotny
Plant phylogeny, in those days, was much more difficult than it is today. For our paper, it was partly done as a part of PhD studies by Weiblen (Moraceae) and partly by Bremer who specialized in another taxon we were interested in – Rubiaceae.
HS: Did this paper have an easy-ride through peer-review? Was Nature the first place you submitted this paper to? Did the paper change substantially from the first submitted draft?
VN: Nature was (as you might expect) our first journal of choice – and also our first attempt to send something to Nature from our PNG research. The manuscript was well-received by reviewers and the editor, Dr. Rory Howlett, who wrote in his response While all three reviewers find your work of considerable interest, as do we, they have raised points to which we would like to see your response… In addition to minor comments and queries, two reviewers, supported by the editor, wanted us to come up with revised global insect diversity estimates of insects based on our new estimates of tropical host specificity. In the original draft we submitted to Nature we had concluded by saying: Further, estimates of global biodiversity at ≥30 million species (Erwin 1982) need to be revised as they are based on the assumption of narrow host specificity of rainforest herbivores, which was disputed also elsewhere (Basset et al. 1996a, Ødegaard et al. 2000). Based on the reviews we received, we calculated and included our own estimates of global insect diversity in the revised version we submitted.
HS: Do the main findings of this paper – “low host specificity of tropical herbivorous insects” and “4-6 million of insect species globally” – hold true even today?
VN: Both findings still hold, by and large. There has been a subsequent discussion and even controversy as to how this low specificity compares to temperate zone patterns (i.e. what is the latitudinal trend in specificity), based on our and other people’s subsequent work, but the patterns on specificity uncovered here remain accepted. Likewise, further analyses of global insect diversity, including our own work, came close to our estimate.
HS: Did this paper create a buzz – within academia and outside – when it was published?
VN: It did have fairly significant popular science exposure, including BBC, Los Angeles Times etc. It has, currently, 548 citations on Google Scholar – pretty much evenly distributed since the time of publication to the present, suggesting that it is seen as a solid and still relevant piece of work, rather than generating a wave of excitement that would later wane. Interestingly – but perhaps not surprisingly – all media attention, and also a part of professional attention, focused on diversity estimates, i.e. the part of the paper that was developed only in response to the review process.
HS: How important has this paper been in your career? Did it have an influence on the course of your future research?
VN: It was very important as it validated my choice to focus on long-term ecological research in Papua New Guinea. I had left for PNG originally for a 6-month stay in 1995, which was extended to one year, followed by another year after only a 3-week break back in Europe, and followed by a series of half-year stays. By late 2001, when the manuscript was submitted, I had spent 4.5 of the past 7 years in PNG. This was not exactly appreciated by my employer, the Institute of Entomology of the Czech Academy of Science. This paper was the final proof to the institute that my PNG rainforest adventure was not a complete madness and could be productive scientifically.
Interestingly, the paper was less influential on the direction of our future research – it only confirmed that we were on the right course, studying plant-insect trophic interactions in rainforest food webs (not that we had any doubts about it). However, it did represent a closure of a particular research direction where, every year, we would pick another set of rainforest tree species for one-year sampling of their herbivorous communities, to add to our growing data set. After this paper we re-focused on beta-diversity of herbivore communities across large geographic scales, which in due course led to another Nature paper.
HS: Have you ever gone back and read this paper after it was published? When you read this paper now, what are the aspects about it that strike you first? What memories does it bring back?
VN: I have not read it in its entirety for a long time, but did go back occasionally to check some particular statement or detail. It is striking that after almost 15 years I am not aware of any comparably detailed replicated data set from any other tropical rainforest site, which, incidentally, is typical for many other studies of community ecology patterns. Plus, we still do not really know how many species of insects there are on the planet. There is also a degree of nostalgia generated by this paper for my early days of much more solitary research in PNG (with no mobile phones and email only starting to reach the island).
HS: If you compare this paper, to papers you write today, are there any differences, e.g. in writing style?
VN: One unfortunate, but rather predictable, development in my career since the paper’s publication is that I am writing fewer papers as the first author. This is compensated by being able to shape more papers, mostly written by my students, as the last author, but that is not as much fun as writing a paper from scratch. The 2002 paper was my first stab at a direct, concise, sparse style of writing, which I have tried to develop and improve in subsequent papers. So, although it has its imperfections, it was a start in the right stylistic direction.
HS: Have you had the opportunity to go back to these field sites after the paper was published? What is the status of these plots today?
VN: These are the rainforest plots closest to our research station where I am writing these lines now. I have been visiting them regularly for the past 20 years. These are village-based conservation areas and doing great. Some of my collaborators, recruited from these villagers and trained as paraecologists, have become co-authors on some of the subsequent papers, including the one on beta-diversity in Nature.
HS: Do you know what this paper gets cited for, mostly?
VN: I do take note of the citations (also because this is still my most-cited paper). I have not analysed closely the papers that cite the study, but I gather it is primarily the low host-specificity pattern that gets cited.
HS: What would you tell a student who is about to read this paper today? What should he or she take away from it?
VN: Students today will see that the paper is obviously rather primitive by current standards when it comes to plant phylogeny, but remains of high standard when it comes to insect data. Interestingly, we used these insect data to make some rather daring extrapolations in order to estimate global insect numbers. Despite the fact that these numbers are of interest to many biologists and non-biologists, we have not progressed much with making these estimates more precise over the past 14 years.
HS: Is this one of your favourite papers?
VN: It is one of my (three) favourites as it was the first milestone in the PNG research, brought results of general interest to both the wider public and researchers, and remains relevant to this day.