In 1799 Alexander von Humboldt went to see the world. The Sun fell straight down in front of his ship's bow, and moonlight rose all around him. He watched great pods of whales jump from the sea and surveyed the beauty of night skies bright with migrating stars. More striking to Humboldt than the beauty of the world, however, was the bounty of life it held. And more specifically, the patterns he saw in the distribution of life. The nearer he approached the tropics, he later wrote in Ansichten der Natur (Views of Nature), the greater "the variety of structure, grace of form, and mixture of colors, as also in perpetual youth and vigor of organic life." Humboldt had discovered the latitudinal gradient in biological diversity. All it took to see the pattern was traveling south for a few years. But as the next 200 would show, that was to be the easy part.

Time has added both detail and exclamation points to Humboldt's initial observations. We can now map the patterns of diversity in mammals, birds, amphibians, and reptiles for the entire world. Nearly all groups of organisms, from foraminifera to frogs, are most diverse in the tropics. And that gradient has great implications. There are not only more species in the tropics, but there are also more potential medicines (and conversely, more diseases), fruits, cultures, and languages. The gradient in diversity that Humboldt detected shapes human life, from our economies to our well-being.

In the article, "The Trouble with Biodiversity (link)" published in Seed Magazine I explore whether we will ever be able to say with certainty and consensus what governs large-scale patterns of biological diversity. As part of research for the article, each scientist below was asked a series of questions about their experience with and thoughts on patterns of biological diversity. Citation: Dunn, R. R. 2008. The Trouble with Biodiversity. Seed Magazine. October.

If you study large-scale patterns of species diversity and would like to add your response, please email me at Rob_Dunn "at" ncsu.edu.

"The main reason is probably that the opportunities to test all the ideas and hypotheses that have been put forward during the more than 100 years of study have been very difficult to test up to now. In that way, hypothesis was added to hypothesis with few being refuted. I’m confident that in 10 years we will consider the cause of global patterns in species diversity if not solved, close to a solution accepted by most. With much more large-scale biodiversity data, the advent of molecular phylogenetic methods to shed light on the relatedness among species, and much more environmental data at relevant spatial scales available, we are in a position to test the multitude of hypotheses. It is exciting times to be a student of biodiversity!"—Roland Jansson, Associate Professor, Department of Ecology and Environmental Science, Umeå Universitet, Sweden.

" One: It is easy to walk into a habitat and start counting species; it is challenging to do it in a careful, standardized way from tundra to rainforest. The datasets for any given critter thus vary widely in quality. When you then add the diversity of life to the mix, extracting pattern isn't comparing apples to oranges, it's comparing apples to grasshoppers to pond scum. Two: Of all the processes in ecology and evolutionary biology, speciation is the toughest nut to crack. Three: Species diversity is an oddly behaved statistic. Two islands can both have 10 species, with one having 91, 1,1,1,1,1,1,1,1,1 individuals, and the other in which each species has 10 individuals. An ecologist would see those two communities as being completely different, and yet, they have the same diversity. Four: Ecologists borrow shamelessly from the other sciences. With the diversity question, we are largely on our own. What other science asks the question "Why are there x forms of fillintheblank here, and y forms of fillintheblank there?"—Michael Kaspari, Professor, Department of Zoology, University of Oklahoma

"Well, that's an interesting question of course. I think that the prime reason why it's difficult to elucidate the mechanisms are the huge spatial and temporal scales. If we could take diversity gradients into the lab (in a realistic way) and evolve them in a few years it would probably be way easier to unravel things. I sometimes illustrate this witha silly example to my students: I hypothesize that the reason why there are more individuals and species around the equator lies in a Coriolis effect: the rotation of the earth gives pollen and seeds of plants a bias towards the equator when they move. Similarly, animals experience this force during every step they take. Now, over millions of years you will get most species around the equator. Quite silly, but try to falsify it! There are several approximately equally silly explanations in literature that received quite some attention, probably for the same reason: even the most stupid explanation is hard to falsify if you cannot do experiments. (Why did studies suggesting a principle role for area in promoting species richness never include Antarctica? There's several papers on the topic!) Another big problem, in my opinion, is the tendency of scientists to want to be the most clever guy in the classroom. It seems some people launch an idea, and even after it turned out not to be very useful, they continue to defend it as something of their own, which (through lost working time, publications, and refereeing of other work) may slow progress."— Folmer Bokma, Department of Ecology and Environmental Sciences, Umeå University

"It is not a single question, and it is about understanding webs of interactions bewteen different types of processes, and all those things involved events that occurred long ago. I think most people still do not appreciate how complex an complete answer would be for even a single group: it is easy to pretend that it is a simple question, but it isn't. It isn't obviously amenable to experimentation, and so relies on inferential evidence. Getting at historical processes relies on techniques that are new and still being developed. " Peter Mayhew, Department of Biology, University of York.

"A number of reasons: (i) As I mentioned before I agree with Bob Ricklefs, Dave Jablonski and others that history is a key part of this puzzle. Yet if you look at this literature history is often thought of as the stuff that is in the residuals of, say, a regression of temp and richness. To be fair, the evolutionary biologists have been slow to work on this, which is why we still don't have good estimates of diversification rates for most taxa but that's slowly changing. We really need better information on historical parameters of all sorts but that is a difficult challenge (ii) We have all focused on just species counts simply 'cause that's easy to get from the literature even though it really doesn't tell us much about the real biology. Ecologically and evolutionarily important parameters such as functional and life history data are largely lacking and data on things such as ecological interactions are virtually non-existent. If we had those information I bet you we would have had a much better sense of what is really going on (iii) There has been a strange lack of process-based models/thinking and we have readily accepted correlations as causations. As Ricklefs has pointed out, just because present day temp correlates well with present day diversity patterns doesn't mean temp drives the gradient. Many of us (myself included) have spent a fair bit of time on those correlations yet as far as I know nobody has been able to provide a satisfactory explanation of how more energy would translate into more species. Similarly it is not sufficient to show that diversification rates are higher in one area - that's just a proximate explanation. Ultimately we want to know why the rates are higher and for that we need more process-based models of speciation and extinction. So in summary, I don't think it is surprising that this has been a difficult problem - no matter how you look at it this is the most complex pattern in ecology. Because of that it also needs an interdisciplinary approach that includes ecologists, evolutionary biologists as well as earth scientists and climate modellers, and as we all know that's never easy." –Kaustuv Roy, Professor of Biology, University of California, San Diego.

"AGAIN TWO ANSWERS: 1) BECAUSE THERE MULTIPLE CAUSES, SO IT IS NOT A MATTER WOF WHICH MECHANISM OR WHOSE HYPOTHESIS IS 'RIGHT'. INSTEAD IT IS A MATTER OF SORTING OUT THE INDEPENDENT AND INTERACTING EFFECTS OF MULTIPLE PROCESSES. 2) BECAUSE WE ONLY HAVE ONE EARTH, AND IT WORKS IN SUCH A WAY THAT THE ENVIRONMENTAL AND HISTORICAL FACTORS THAT HAVE SHAPED DIVERSITY ARE CORRELATED. I.E., THE TROPICS ARE WARM, PRODUCTIVE, AND HAVE BEEN LESS PERTURBED BY PAST CLIMATIC FLUCTUATIONS. THIS MAKES FOR THE CHALLENGE OF ELUCIDATING CAUSATION IN THE FACE OF CORRELATION. "—James Brown, Distinguished Professor, University of New Mexico.

"The reasons are differences in the details of diversity gradients between groups, and the numerous secondary factors which modulate the effect of the primary ones. Thus, habitats differ considerably in the degree of heterogeneity, and this may conceal effects of evolutionary speed, etc.. Also, most authors, in my opinion, tackle the problem from the wrong angle. They look for correlations and arrive at fallacious "explanations". For example, great diversity is sometimes correlated with great productivity, and this supposedly explains diversity. However, experiments have shown that an increase in diversity causes greater productivity, and not vice versa. This is also shown by coral reefs: the surrounding waters, where reefs have evolved, have very low productivity, but the reefs themselves are highly productive as a consequence of photosynthesis by coral symbionts. In summary, then, we need a hypothesis which explains the causes of diversity gradients and does not just establish correlations, and the hypothesis of effective evolutionary time does just that. Various studies on plants and animals have provided support for it., e.g., Wright et al. (2006). The Road from Santa Rosalia: a faster tempo of evolution in tropical climates, PNAS 103, 7718-7722. It should also be pointed out that the metabolic theory of ecology (Brown etc.) is well in agreement with the hypothesis. Concerning your question: "Do you think we are getting closer to consensus as to the causes of diversity gradients?". I hope more and more people will see the light!" --Klaus Rohde, Professor Emeritus, University of New England, Armidale, Australia.

"I believe this difficulty is mainly based on three things: 1. The topic in itself requires a multi-scale and multi-factor approach. Thus, a factor might be very prominent in structuring biological diversity on one scale but not on others. Even when agreeing that e.g. speciation is a major driver of how diversity is distributed globally, there might be huge differences in why speciation is higher in some places than others. 2. Especially on large-scale diversity gradients, the research is too much focused on few organism groups. This has two connotations: First of all, it neglects a majority of the actual biodiversity of our planet – especially the small and the ugly (i.e. those with little morphological features) are highly underrepresented, whereas the big hairy and feathery stuff is overrepresented. Second, it is often suggested that if some organisms groups share a pattern (e.g. the latitudinal cline of diversity) they also share the same mechanism causing this gradient. This obviously does not have to be the case. 3. Macroecological patterns can rarely be investigated experimentally – that is, the majority of evidence for or against certain processes as drivers of diversity is simply based on correlations, which do not indicate any mechanistic causal link. Phylogenetic and molecular studies certainly open up the avenue for more mechanistic research here." –Helmut Hillebrand, Associate Professor, Institute for Botany, University of Cologne.

“ The largest stumbling block is the inability to do experiments. Also, it turns out we needed phylogenies to really understand what is going on, and they are only now becoming available. As you can tell from what I say in my essay, I think we now pretty much know the answer, although that may be a minority opinion. The bits that are still in doubt are details.”—Bradford Hawkins, Professor, Ecology & Evolutionary Biology School of Biological Sciences, University of California, Irvine.

"The problem is difficult because we cannot conduct the necessary experiments at the right spatial scales, and because ecologists have not been good about articulating additional patterns (beyond species richness differences) that would be predicted if alternative mechanisms are in operation. If anything, we seem to be further from consensus than we used to be, perhaps because the MDE debate has highlighted some of these issues (which has been a very good thing)."—Nick Gotelli, Professor, Department of Biology, University of Vermont.

"Both history and climate, which themselves are intertwined, are the obvious underlying factors driving geographic patterns of diversity on Earth. Historical patterns of vicariant geography provided opportunities for speciation and isolation, whereas climate determines productivity. Together, these offer ecological opportunities for resource partitioning.  A surprising amount of present day differences in diets among extant lizard species can be traced back to ancient phylogenetic events (Vitt, L. J. and E. R. Pianka. 2005. Deep history impacts present day ecology and biodiversity. Proc. Nat. Acad. Sci. 102: 7877-7881)."—Eric Pianka, University of Texas

"Difficult? It is the difficulty that makes it so intriguing! I guess in searching for generality we keep discovering exceptions. The complexity of such a large question dependent on so many processes at so many scales may ultimately prove overwhelming for consensus." Craig R. McClain, Postdoctoral fellow, Monterey Bay Aquarium Research Institute, California.

"As to why it has been so tough, maybe part of the answer is that too many of us have relied on gut feelings. And part of the answer is that one theory will always be wrong. The evidence is clear that we will need two, three or even four! And part of the answer is also the uncautious intermingling of evolutionary scales in our analyses. And part of the answer is how difficult it has been to get data and how long it would take to do a proper experiment. "—Michael Rosenzweig, Professor of Ecology and Evolutionary Biology, University of Arizona.

"I think we can go forward two main difficulties: 1) the processes are very intricate and scale-dependent (both in time and space), so it is difficult to tease them apart; 2) We do not have good tools and skill to measure macroevolutionary processes that allow us to better check some of these hypothesis. Despite these difficulties, I think this is a great moment and it is a very exciting field, especially after more than 200 years of research! I think that we are close to a consensus, or at least we have some agreements, as I pointed out before. The environment is important, but it strongly interacts with history. It is not clear how historical components in environment drive diversity at which scalesand we already saw some motions in this direction. Even if we do not have yet achieved a consensus, I think we are in a much better position now than we were 10 years ago, since now we at least know what we should try to improve to understand the gradients." –Alexandre Felizola Diniz Filho, Professor, Universidade Federal de Goiás, Brazil.

"I believe that we are making progress and that the impact of historical studies has been significant (but then you might say that a paleontologist would say that!). Thirty years ago studies of tropical high diversity were dominated by ecological approaches; resource partitioning and niche separation were thought to hold the key. But now we look much more into the past and the role of so-called dispersal-assembly mechanisms. We look for macroevolutionary sources and sinks, compare rates of evolution across latitude, and realize that events on huge timescales influence what we see today. Of course, ecology will always be important but so much of what we see today is built upon the framework of the past. Work in progress is beginning to show just when in the past the LDG expanded dramatically, and when it was much flatter. And this will in turn show that both the species-energy and area hypotheses cannot be ignored. But history shows us that, on the grandest of geographical scales, contraction is just as important as expansion, and extinction just as important as speciation. What we see today bears a historical legacy stretching back at least 100m.y. I should also emphasise the power of molecular phylogenetics to improve our understanding of the timing of major biogeographic events. This is the way forward with bipolar studies and there are certain species-rich clades, such as the flowering plants, phytophagous insects and birds, that clearly span the high- and low-latitudes. Using various molecular clock techniques it is possible to obtain the time of split between high- and low-latitude taxa that is completely independent of the fossil record. We need much more of this sort of work because it is vital in determining the timings of both tropical and temperate radiations. We are not there yet, but greater emphasis on both the deep time fossil record and the molecular phylogenetics of major clades is helping us move closer towards a consensus on the origin and maintenance of latitudinal diversity gradients."—Alistair Crame, Division Head Geological Sciences Division Cambridge University.

"Firstly, it is a very complex issue with many potential mechanism contributing to the observed diversity gradients. In addition, there are a number of other problems, noatbly: (1) No replicates (so far). We just have one Earth and regions on Earth have truly independent biota and therefre this is also true for their diversity patterns (as strongly illustrated by the molecular evidence for a lot of cross-oceanic dispersal between the continents). (2) Such complex large-scale phenoma are not easily to meaningfully address by the experimental approach that have dominated ecology for many years. This is especially true for biological patterns which cannot be understood reductionistically as a product of bottom-up processes, but generally also reflect large-scale evolutionary and historical phenomena (in my view). (3) Until the last 1-2 decades, we have had only scant biodiversity, environmental, and historical data to test our hypotheses on large-scale diversity gradients on. (4) Likewise, only recently have developments in computer science and statistics allowed us to begin handling and analyzing the relevant data with sufficient (or at least closer to sufficient) sophistication." –Jens-Christian Svenning, Associate Professor, Department of Biological Sciences, Aarhus Universitet, Denmark.

"Diversity gradients have proven to be a complex problem not easily solved and it is likely that any solution will be multifaceted. I think that if the “cause” of diversity gradients were based on a single factor, ecologists would have identified that variable by now. I do think that we are getting closer, reflecting the fact that we now have more tools at our disposal and that our approach is changing. First, as compared to two decades ago, ecologists have much more organismal and environmental data at their fingertips, made available via the internet and accessible databases, and have cultivated a general attitude that is more widely oriented towards sharing information than might once have existed. Further, the philosophical approach to diversity gradients has shifted from pattern description to the testing of specific mechanisms. Once, hypotheses developed to explain gradients only focused on the general, qualitative patterns already know to exist. Now, there is a much greater emphasis on the more quantitative aspects of gradients, such as their shape and metrics other than simple species richness, and on testing explicit predictions developed from competing mechanistic hypotheses." —Dawn Kauffman, Assistant Professor, Joint Science Department, the Claremont Colleges, Claremont, CA

"The question that has been the focus of much of the work on diversity gradients in past decades is: what factors are correlated with high diversity across geographic space? The main difficulties in answering this question are the multiplicity of potential (or actual) drivers of diversity patterns and the limitations to inference from correlative and comparative studies. Add to this the noisy and sometimes inconsistent nature of the patterns (between different taxonomic groups for example), and the fact that decent data on species distributions has only recently begun to accumulate, and it's not surprising that there has been so much uncertainty. Nevertheless, it is now pretty clear that area, temperature, available energy and topographic diversity have alot to do with species diversity patterns. Pinning down the mechanisms that link environment to diversity is even more of a challenge. Here there are two main questions: (1) why have there been such great evolutionary radiations in the tropics and other high-diversity regions? and (2) how are so many species (often close relatives) able to coexist at local scales? The second question is the stubbornly persistent problem of community ecology, and despite decades of theory and empirical work we don't seem a great deal closer to an answer (at least not a general answer) than we did 30 years ago. I am more optimistic about the first question. Although there are nutty problems like teasing apart speciation and extinction rates, I think our current uncertainty is largely the result of lack of data: as phylogenetic and geographic data accumulate, I am convinced that in 20 years' time we will know alot more about mechanisms of diversification than we do now."— Marcel Cardillo, ARC QEII Fellow, School of Botany & Zoology, ANU

" (i) it is a idifficult endeavour because we are simultaneoulsy attempting to solve multiple problems, without, in may cases recognising that this is the case. (ii) Causes of diversity gradients - yes I think we are making progress and in some key respects we have the basic building blocks of a comprehensive theory of diversity, but that diversity theory remains incompletely synthesised. (iii) I think 'never' is a long time but I would certainly expect that people will still be working on diversity patterns and processes when I am no longer around to read their work, so for me, they will never all be resolved."

— Robert Whittaker, Professor of Biogeography, School of Geography and the Environment, University of Oxford.