Matsimäe bog in Estonia. Photo by L. Michelson
The working rooms of Martin Zobel and Lauri Laanisto, both scientists from Tartu, are less than two kilometres apart as the crow flies. One can walk from Tähtvere district to the Botanical Garden in 15 minutes. Discussing scientific matters by phone or having an argument in an office seems like the easiest thing.
Still, so that the views of ecologists all over the world could be influenced, the discussion between the two men developed on the pages of Science — one of the most well-known scientific journals in the world. This marks the first time that two scientific teams from Estonia are having a discussion in a science magazine that belongs to the top worldwide.
If it were still summer of last year, I would’ve written how “the scientists have proven a long-sought-after ecological correlation”, but by now every sane journalist would use the word “dispute” in the headline of the story. However, neither journalistic angle would actually present a true picture. In fact, it is quite a typical tale about discussions playing a central part in science.
Although ecologists study nature, in this field there is still a dearth of so-called natural laws, proven universal linkages that could be used in describing and predicting the properties of the ecosystem. For decades it has been suspected that one of these links would be the correlation between the productivity of an ecosystem and the variety of species in it.
A study published in 1973 proposed that the connection would look like a classic bell curve when laid out on a graph, although the authors preferred to called it “the hump-backed model”. According to this model, the areas of medium fertility have the most species. The areas where the productivity (the amount of plants growing) is low have few species, because many could not survive in such conditions. But when the productivity is high, a limited number of species proliferate, taking all the resources.
Since then, quite a few studies both confirming and rebutting the hump-backed model have been published. In 2011, one of the skeptical studies was published in the journal Science. A team of plant ecologists that had gathered for a major conference in Tartu concluded with one accord that the study was inaccurate and methodically inadequate.
Right there and then it was agreed to conduct a new global study. At 30 sites spanning 19 countries on six continents, spots of 8 x 8 meters were measured out. In these spots, all plants were determined, cut down, and then even the dried blades of grass gathered together, so it all could be weighed. “It’s the kind of dull manual work that has to be done with a big team”, says Martin Zobel, a professor at the University of Tartu and one of the leaders of the study.
After the data was added up and the spots were included in the figures, Science, one of the two most prestigious journals in the field of natural sciences, accepted the study. The article assured that the global study of grasslands, designed to involve as many climatic zones as possible, proves the hump-backed model right.
When the article saw the light of day, Lauri Laanisto, a senior researcher at the Estonian University of Life Sciences, was doing field work at Spitsbergen. On the last night before returning to Estonia, he sat on a comfy couch in the Czech polar station, started reading the data that the study was using, and performed his own analysis. But in addition to the two parameters — diversity and productivity — he took into account how many different species of plants would grow in the area altogether. And the hump was gone!
Laanisto invited Michael Hutchings, a phytologist at the University of Sussex, to be his co-author, and sent a comment letter to Science, where he stressed that any result would be lopsided if one did not consider the total amount of species in the area. Both Laanisto’s commentary and an answer from the aforementioned work group were published in Science last December.
Publishing comments as such is quite common of the top journals, constituting a form of scientific discussion. Similar disputes where skeptical colleagues question the relevance of theories, methods, and interpretations take place at conferences and seminars. Before publication, every article is objected to critique by expert reviewers.
But it is still usual that after the article is published there are more questions hanging in the air than is possible to answer using the collected data. The questions will await the next scientists and studies that, whether it be through new data or new methods, will search for exhaustive answers.
The study by Zobel and his co-researchers was ignited by discontent with previous studies. Together, data was collected and analysed in a more global way. Yet, as Laanisto points out, even using this data, we still cannot claim that the hump-backed relationship between diversity and productivity of the ecosystem is universal to all ecosystems.
On one hand, it is a technical dispute. Laanisto saw the hump disappearing in his analysis; in the answer where the authors of the original study had performed a statistical analysis of their own, the hump became flatter but, according to scientists, remained. From this point the discussion becomes loaded with technical details about the statistical methods used.
“In our opinion, the solution offered by our colleague [Lauri Laanisto] wasn’t the most elegant”, Zobel says. “The same thing was calculated over again, at the initiative of [a professor at the University of Tartu] Meelis Pärtel, and we thought that we found a more elegant answer”. After being published in Science, the debate has continued on the blogs of both research teams (see the posts by Laanisto and Robert Szava-Kovats) as well.
A more principal question would be: what is the thing influencing the species richness of an area? Is productivity of the ecosystem the most important factor, or are there others playing a main part?
“I just wanted to show with my simple analysis that the importance of the number of species is almost two magnitudes greater than productivity”, Laanisto says. “When there are species, the productivity starts to influence it, not the other way around”.
This means that in a given area there can maximally grow only as many species as there are present in this region. All the species that can grow in the region might not be present. Estonia, as well as the middle latitude as such, would be a good example. “The ice age washed all the mould away from here and types of really fertile combinations haven’t yet had time to reappear,at all”, Laanisto explains. “There just aren’t species in the species pool that would prefer the very productive combinations”.
Zobel and his colleagues agree with this idea. They are its originators, in fact, and Meelis Pärtel, a member of the research group, is a former supervisor of Laanisto.
“We actually do agree fully. We even tried to say it in our consortium, but it was thought that it is some novel, complicated thing, something to be tackled next, not now”, Zobel says. “Fortunately, public opinion didn’t let it go this way, so it had to get solved immediately”.
“Variations in diversity can be greatly explained with evolutionary history”, he adds. “We are showing that where there are less productive ecosystems, the print of history is stronger”, he adds. “But mathematically, it is a tough question. The attempt to evaluate the role of different factors is not trivial”.
In the meantime, another step forward has been taken. Another international research team, with Meelis Pärtel representing the University of Tartu, has published an article in the other top journal — Nature — attempting to untangle the links between diversity and productivity more elaborately.
“The goal was to find a suitable statistical model that wouldn’t simply bind together two parameters being studied (productivity and diversity) but more”, Pärtel says. “We found that a complex model described the situation much better”.
Among other things, the article concludes that the explanation that uses the amount of species doesn’t automatically exclude the explanation based on the interactions of the local level. “We had the same idea in the Science debate”, Pärtel points out. “Nature is often more complicated than we assume. At the same time, philosophy and mathematical methodology are catching up, so we could be able to make predictions about the behaviour of ecosystems, so needed at the moment”.
Looking back at the Science debate, Zobel says that he was happy for the feedback: “The discussion moved science forward. Based on unique global material, an additional study was conducted — otherwise it wouldn’t have happened.”
Arko Olesk is a science journalist.
The Estonian version of this article was first published in the Estonian daily Postimees.
