Skip to content

Hey, our world is a whole lot more complicated than we ecologists seem to believe.


I’ve been thinking a fair bit about the difficult challenges we face if we are going to become that responsible humanity we would like to see emerge in the Anthropocene.  I’m talking about the challenges in steering the planet successfully into a good future, meaning one not too different from the Holocene in which human civilization was born and came of age.  Don’t know what I mean about steering the planet?  Guess you have not been reading this blog.  Or much else about the global environmental crisis that surrounds us.

Like the navigators on board Hokule’a sailing the world’s ocean, we must learn how to steer our planet towards a safe, Holocene-like future.  Never been done before.  Can we do it?  Photo © Bryson Hoe & Polynesian Voyaging Society.

In a nutshell, we humans have become a potent planetary force causing big changes on this planet.  Changes so big that most environmental scientists now recognize present-day times as the Anthropocene, a new geological age in which the major planetary changes are being caused by one species of life – li’l ol’ us.  The problem, of course, is we have been changing the world in bad ways – ways that make it a less wonderful environment for creatures like us to live in.  Indeed, we have triggered what could become the sixth great mass extinction on the planet – average extinction rates for vertebrates are now estimated to be between 100 and 1000 times higher than typical over the grand sweep of geological history.  The last great mass extinction took place at the end of the Cretaceous age, 65 million years ago, and took out the last of the dinosaurs among many other creatures.  We have no idea what will disappear in the 6th extinction, although most non-domesticated vertebrates are high on the list.  Only hubris tells us this does not matter to us.

It’s sort of a no-brainer that we must do better than we have been doing if we want the planet in future to resemble the planet we have come to know and love, the planet of the Holocene, a time in which we invented agriculture and built our great civilizations, a time in which climate was remarkably stable so we grew to depend upon the weather, the seasons, and the natural environments that largely clothed our world.   Doing better means using our considerable power to steer the planet towards a future that is good for us to live in.  If this smacks of self-interest, so be it.  It will also be a good world for many other species that share it with us.  But there are reasons to be concerned that we may not succeed in this great challenge: we may never develop a will to change our ways, we may leave it so late that the planet’s ecology will deteriorate so far that even the mighty we will not be able to turn it around, or we may simply underestimate the challenge of steering effectively.

As somebody once said, “Ecology is not rocket science; it’s a whole lot more complicated than rocket science”.  Steering the planet towards a good future is a supremely ecological task, and I don’t know if we can do it.

I’m not talking here about our difficulty in agreeing on the task at hand, although that is a major challenge that seems to become larger instead of smaller as the first year of the Trump administration plods along.  I wonder if future historians will introduce a new calendar in which AD refers to ‘After Donald’ – but I digress…

I’m not even talking about the technical complexity of the challenges we will face, or the political challenges as we are forced to decide on a global path of action that will necessarily not benefit every corner of the globe equally.  I’m thinking about our ability to recognize the complexity of the world around us, so that we may plan and act in ways that are sufficiently subtle to be effective.  In what follows, I’ll set out why I think our own physiology limits our ability to see the complexity around us.  I’ll then take our thinking about coral reef systems as an example, looking at three recent papers that show how much more complex reefs are than in the models we create of them.  I then suggest some ways we might do better.

How we see our world

The human is an incredible creature.  It views its world using five senses of which vision is primary, with hearing and touch perhaps next in line.  Each of these senses is served by a suite of more or less complex sensory detectors – light receptors for example – organized into complex sensory organs that act to convert external stimuli (light in this case) into the language of the nervous system.  Activated sensory receptors stimulate nearby sensory neurons chemically.  Sometimes the stimulation is positive causing the neuron to become more active physiologically, and sometimes it is negative, inhibiting the activity of that neuron.  Sensory neurons, in turn, chemically stimulate or inhibit other nearby neurons, and so it goes until a modified pattern of neuronal activity reaches the brain.  It does not stop there, but instead becomes even more complex as neuronal pathways within the brain become activated or quietened down both because of chemical stimulation among the neurons coming in from the sense organ, and because of influences on that activity due to action of other neurons carrying their own messages from other sense organs, or the internal environment of the organism.  Miraculously, by the time this complex modulation of the activity of neural cells within the brain reaches the visual cortex, it is decoded as information about the outside world.  If the individual is ‘paying attention’ to its visual sensors, it recreates an image of reality and become aware of the world outside.

Vision – it is a sophisticated image analysis process, not a bit like building an image from pixels.  Image © London Neuro Physio Ltd

This process is radically different from what goes on in the camera in a cell phone – while the visual cortex of the human brain is laid out as a map that approximates the visual field of view, there is nothing resembling an array of pixels, or any other kind of image.  Instead, there are neurons that respond to edges, neurons that respond to small dark objects that are moving, but not to ones that are still, neurons that respond to intersection of two edges, and so on. By the time visual information reaches the visual cortex, it has already been simplified, codified and analyzed in ways that emphasize those things or events in the visual field that, over evolutionary time, have proved important to our survival – our visual system highlights those things that are important to us.  Yet we see – somehow creating an internal impression of a complete visual field, and our auditory, olfactory, gustatory, and tactile systems all operate in much the same way.  We don’t see perfectly, or accurately – that is, we don’t build a perfect replica of reality somewhere inside our heads.

In fact, our visual systems (or other sensory systems) do an enormous amount of filtering, compressing, and coding of the information that arrives at the visual cortex.  Our eyes have not evolved to present us with pretty pictures.  They, and the full network of neurons connecting them to the visual cortex, have evolved to detect movement, edges, color, and very simple shapes.  Out of these building blocks we somehow create sunsets.  But we do so by filling in gaps, by embroidering, and by glossing over the simplifications that have resulted as information travels from the eye to the brain.  And our conscious selves are totally unaware of all this processing.

It is possible, and plausible, to argue that our usual patterns of thought are a consequence of the way in which our neural system works.  We categorize, classifying stimuli into a small number of different types, such as the different colors of the rainbow, or crudely into just two types – lighter vs darker, larger vs smaller, moving to the left vs moving to the right and so on.  In classifying, we also simplify, emphasizing the features that distinguish one class of stimulus from another while minimizing or ignoring those that exist among members of the same class of stimulus.  For example, reality presents us with a continuous gradation of light frequencies across a rainbow, yet we persist in seeing bands of different color.  We are quick to detect movement, distinguishing moving from still objects, but quite poor at judging speed of movement.  This tendency to classify, categorize, contrast images extends also to conceptual thinking.

We find it easy to categorize similar, but not identical images as belonging to a single class of objects, or rabbits in this case.  We emphasize the differences among classes, but minimize (don’t see) those within classes.  Image © Planwallpaper

We all know what a rabbit is, but rabbits come in a great array of sizes, shapes and colors, and a rabbit facing you looks very different to one facing away.  Our concept of the species is a great example of our ability to cluster individual items into a single class that differs from other classes (other species).  We are remarkably good at throwing away information, so as to focus on what we choose to consider the important information – a rabbit is a mammal with a specific pattern of dentition, plus several other features that we use in classifying it in Order Lagomorpha.  Those features are quite similar to, but different from those of that other large group of species we call Rodentia.  There are a number of different species of rabbit, but they share long, floppy ears, and a hopping locomotion.  They also breed like, well, rabbits.

We also see what we expect to see, completing partial information to understand what we assume to be present.  See a fluffy white tail disappearing into the distance with a hopping movement, and we all report having seen a rabbit.  This ability to complete, or more generally to resolve, sensory information according to what we expect is why visual illusions are so captivating.  The illusion presents contradictory information according to what we expect to see, and we struggle when what we see does not seem to make sense.  Our ease in completing fragmentary information is considerable and mostly subconscious.  Marshall McLuhan said it well: “If I had not believed it, I never would have seen it”.

When people become scientists, they do not stop being humans.  Scientists use the same sensory systems as other people, and similar thought processes, when they collect data from the environment, or when they evaluate results of their observations or experiments they have run.  This remains true, even when the instruments they use for collecting those data are complex, sophisticated, and with great discriminatory power.  Part of the training as a scientist involves applying objective criteria for making the classificatory decisions we all make all the time.  There are many objective reasons why scientists agree on the specific features that make a rabbit a rabbit and not a squirrel, a kangaroo, or a praying mantis.  We could have used quite different criteria and classified rabbits differently.  I think our decisions re rabbits, framing them as a group with common evolutionary ancestry, have been logical; I admit we could have done things quite differently.  And this brings me finally back to environment and ecology.

We seduce ourselves to be content with too simple models

Ecology is a whole lot more complicated than rabbits.  To begin with, any ecological system is comprised of numerous individuals belonging to different species of life, each with specific requirements which it obtains from the environment, and each struggling to survive and reproduce.  These different individuals necessarily interact with each other as well as with the physical environment – most obviously there are the consumers and the consumed. But there are also those species which compete with each other for the things they need, and those species which cooperate with each other in order to obtain things they need (sometimes, but not necessarily the same things).

Reef-building corals (not all corals, because some live slowly in deep cold water and do not build reefs) compete with large fleshy algae and other organisms for space in the sunlight.  They sometimes shade each other out, and the shaded organism – coral or alga – dies.  But reef building corals also cooperate in an intricate intracellular relationship with single-celled algae which spend their lives within the coral’s tissues.  The corals cannot build reefs without the physiological boost they receive from their embedded cooperating algae, and the embedded algae, in turn, do better than they would if drifting freely in the water column, because some of the coral waste products are valuable nutrients for the algae.  Add in the fact that there are many different species of coral, and several different clones (effectively species) of symbiotic algae, as well as the fact that corals grow in a variety of environments from very shallow to deeper, from nearly nutrient-free to quite eutrophic waters, and from warm to very warm water, and it becomes clear that simply describing the relationship between corals and algae just got a lot more complicated.  Finally, add in the fact that there are lots of other creatures living on a coral reef that, in turn, interact in various ways with the corals and/or with the algae.  Suddenly we have a complex system with many interacting and changing parts.  Step back a bit, and the coral reef becomes just one of a number of ecosystems, each with its own complex assemblage of organisms and the nutrients and other resources they each need.

Under these circumstances, it would be naïve to believe that we are close to understanding how the biosphere operates, let alone how its operation will change as the climate changes.  We tend to complete our view of the world from fragmentary sensory information, and we tend to alternately magnify and ignore differences among species, among interactions, and among differing times and places as we build our conceptual models of how the world works.  Given these tendencies, the complexity of ecological systems absolutely demands a humble, precautionary approach as we build our understanding.  We are virtually guaranteed to be incorrect in our interpretations of some aspects, or simply to have completely ignored subtle but important processes.  Our enthusiasm for simple models encourages such errors, and can help lead us down some very unproductive paths.

I don’t deny that we can only advance our scientific understanding by developing simple models first, only adding to their complexity when they prove to be inadequate models of reality.  But I suggest that, too often, we become attached to our simple models, clinging to them long after it should be obvious to all that they are inadequate representations of reality.  And I suggest we do this partly because of our own sensory and analytical limitations.  We are too easily satisfied by overly simple models.  We need to examine them more rigorously than we usually do, and to work towards more complicated, but more realistic and reliable models than we usually do.

Coral reef decline and the interaction of corals and algae

Ecologists studying coral reefs know that these are particularly diverse systems.  One might expect we’d be particularly alert to the problem of too-simple models, and yet we do not appear to be so.  There are many coral reefs around the world that have lost a considerable proportion of the coverage of living coral they enjoyed as recently as the mid-20th century.  Reef ecologists refer to this broad pattern as reef decline or reef degradation.  Many such degraded reefs now support dense stands of foliose algae, and the presence of these algae makes it very difficult for new coral juveniles to survive and grow.  In many cases, the transition from a coral-dominated to an algal-dominated state was sudden, brought about by a catastrophic storm, bleaching event, or population explosion of crown-of-thorns starfish which prey upon corals.  A part of the complexity of reef systems is that not all storms, all starfish outbreaks, or even all bleaching events result in a profound, seemingly permanent, change in composition of the reef community.  Scientists refer to such abrupt transitions in ecosystem state as phase shifts, and recognize that changes to ecosystems are not always gradual, nor irreversible.  Sometimes reefs recover.

Reef ecologists prefer reefs with lots of living coral and few algae.  We spend time talking about the resilience of such systems.  Photo © Sea Monster

We explore what we call the resilience of ecological systems, and debate ways to restore algal-dominated reefs to their former coral-dominated state.  In pursuing greater understanding of these phase shifts, we have developed models.  Mostly, those models have contained just three interacting agents – the coral, the algae, and the algal predator, sometimes called ‘herbivore’ and sometimes ‘herbivorous fish’.  We have yet to adequately explain why some reefs that suffer massive die-off of corals and increases in foliose algae, get back to becoming coral-dominated systems after a decade or so, while others do not.  In other words, our understanding of reef resilience, how resilience is maintained or eroded, and how corals and algae share the space available on reefs has a long way to go.  Surely, given the global trends towards coral reef decline, building this understanding of resilience should be viewed as a critical need in reef science.  We might expect some clear evidence of progress in the sophistication of thinking on this topic, and yet simple three agent coral – algae – herbivore models continue to be used.  Why are we making so little progress?

I don’t have an answer to this problem, but I suspect it is now time to stop pretending we can characterize the complex interactions among corals, algae, and herbivores of various types in a simple, three-agent, model.  A casual survey of recent papers suggests reef scientists are well aware of this need for more sophistication, but continued references to the phase shift from coral- to algal-dominated state suggest that being aware of, does not yet mean we are adjusting our thinking towards a more nuanced, multifactorial view.

Multiple weak positive feedback loops

My evidence (if citation of three papers can be called that) comes from an off-the-cuff survey of papers in the last two issues of the journal Coral Reefs.  In April, Ingrid van de Leemput, Egbert van Nes and Martin Scheffer at Wageningen University, the Netherlands, and Terry Hughes at James Cook University, Australia, published a modeling study of the coral – algae – herbivore interaction to explore effects of multiple feedbacks on the coral – algae phase shift.  They begin by reviewing the nature of ecological phase shifts and point to the need for positive feedback loops if alternate stable states such as coral-dominated and algal-dominated reefs are to exist.  In a survey of earlier studies, they found arguments and evidence for about 20 different positive feedbacks that could be operating in the coral-algae relationship, and they point out that the roles of few of these have been investigated to any degree.  They suggest that multiple weak positive feedback loops could operate together and that feedback processes that, by themselves, would be too weak to set up the alternate states and sudden phase shift between them, could suffice to do this if operating in tandem.

Van de Leemput and colleagues then proceed to model such a situation:  Their model retained the simplicity of ‘coral’, ‘algae’ and ‘herbivore’ each of which grew in dutifully density-dependent ways because corals and algae were both limited by available space and the system had a fixed carrying capacity for herbivores.  But built into the model were three distinct types of positive feedback.  First, the algae experienced a reduced per capita rate of herbivory as they became more abundant (the herbivores became satiated).  Second, the algae had a greater crowding impact on survival and growth of coral recruits than that due simply to their preemption of space on the reef, so that coral reproduction was impeded more severely the more abundant algae became (direct physical or chemical interference with the recruits that reduced the recruits’ survival or growth could do this).  Third, the coral had positive effects on the abundance of the herbivore (perhaps because the corals provided shelter that enhanced herbivore survival).

The model was adjusted so that the magnitudes of each of these positive feedbacks were insufficient to lead to the condition of alternate stable states when acting alone.  Fishing pressure was varied as a way of modifying herbivore numbers, and, when only one of the three positive feedback mechanisms was operating, the change in fishing pressure always led to a gradual shift from coral to algae as fishing pressure grew.  When all three feedback mechanisms were operating, the result was alternative stable states with the chaos of a sudden phase shift between them.

In each of these graphs, fishing pressure varies along the x-axis, and abundance of herbivores (blue line) tracks that, being highest at lowest fishing pressure.  In cases a through d there is a switch from coral-dominated to algal-dominated at a particular low-herbivory level; case A is the case with no positive feedbacks operating, and cases B, C, and D are cases in which each of the three weak positive feedbacks operate alone.  In contrast, case E, which includes all three weak positive feedback mechanisms, contains a region of hysteresis at mid-levels of fishing pressure and herbivory in which either a coral-dominated or algal-dominated state can exist, with sudden and unpredictable phase shift between them.  Figure 3 from Van de Leemput’s paper, © Coral Reefs.

Does this study explain the interaction between coral and algae on real reefs?  Certainly not, but it does illustrate a) that real reefs are much more complex in the nature of the interactions among coral, algae and herbivores then has usually been claimed (20 different possible positive feedback mechanisms), and 2) that as you add complexity to simple models you jump quickly to a world of more complex outcomes.  Imagine if they had added some additional realism by recognizing several different functional groups of corals, of algae and of herbivores (a point they make explicitly in their paper).

Fragility of coral reef systems

In July, also in Coral Reefs, Camilo Mora of University of Hawaii, Nicholas Graham of Lancaster University and Magnus Nyström of Stockholm University, reviewed the nature of global coral reef degradation, and argued for considering several ways in which coral assemblages were particularly sensitive to the effects of common anthropogenic stressors.  They referred to this sensitivity as the particular fragility of coral reefs.

Mora and colleagues begin by enumerating the many stressors to which we are subjecting coral reefs at present, including warming due to climate change, associated acidification, and multiple more local, but still extensive, effects of human use of coastal regions including overfishing, siltation, arrival of invasive species, and various forms of pollution.  They note that these stressors operate differently in different locations or at different times, that some are slow and progressive while others are intermittent and rapid in occurrence, and that they act on different spatial scales.  They then turn to the ways in which coral reefs are fragile.

Mora and colleagues argue that coral species in general have narrow tolerances to environmental conditions including temperature, pH, and many aspects of water quality.  These animals are being pushed beyond their tolerances by the multiple impacts of human activities, and the stresses are so great that the corals are simply unable to evolve the capacities needed to adapt.

They suggest that the metapopulations that predominately characterize coral reef organisms are susceptible to being degraded.  As individual populations are reduced or eliminated, the distances between populations grow and the connectivity due to larval dispersal among these populations falls.  Metapopulation dynamics falter, and the capacity to maintain or restore local populations is diminished.  Populations that were viable in the context of a functioning metapopulation become less able to maintain themselves.

The bumphead, Bolbometopon muricatum, is the world’s largest parrotfish (up to 130 cm long).  It is a grazing herbivore, and the major bioeroder on Indo-Pacific reefs; each adult chomps up some 5 tonnes of living/dead coral rock each per year.  Photo © University of California, Berkeley.

Mora and colleagues also argue that while coral reefs are notoriously diverse at the species level, those numerous species occur in very much fewer functional groups of species.  One might expect that when a stressor adversely affects one species in a diverse system, other members of its functional group would be available to continue serving its role in the community, thus leading to considerable resilience at the community level.  However, on coral reefs, where the allocation of species among functional groups has been examined, it has frequently been found that many important functional groups contain few members, or are dominated by one or two abundant species, with lots of other group members being extremely rare.  Under these circumstances, stressors that affect important species can radically alter the nature of the community, despite its nominally high diversity.  As one example, they point to the two Acropora species in the Caribbean that together played the primary role in producing the high rugosity so characteristic of and important to coral reef systems.  Both these species are now on the IUCN Red List, and far less abundant throughout the Caribbean than they were in the mid-20th century.  Their second example concerns Bolbometopon muricatum, a large parrotfish which provides a major component of surficial bioerosion on Indo-Pacific reefs.  It is easily extirpated by targeted fishing, and there is nothing that could take its place in fulfilling this role.  (This parrotfish argument hinges on the assumption that surficial bioerosion is necessary to provide stores of new bare rock for colonization by corals and other sedentary organisms – I am far from convinced.)  I’m also not convinced that coral reefs are unusual in the pattern of distribution of their many species among functional groups, compared to other diverse systems, or indeed that there are many species (keystone species) whose roles in their communities are so important that their removal leads to accelerated decline of the system, but I’m a bit of an outlier among ecologists on this topic, and this idea deserves some reflection.

(In the middle of this section of their paper, I discovered citations to two of my early papers, produced to support the view that ecological specialization partitions resources among species thereby favoring coexistence!  I seem to remember writing the exact opposite – but then the arrow of scientific progress is not usually straight, and this observation is a complete digression.  Sorry.)

The fourth type of fragility presented is the likelihood that degraded reef systems will be impeded from recovering because of feedback effects and extinction vortices – another way of talking about phase shifts and alternative stable states.  While I find some things to quibble about in this paper, it is well worth a read because they have pulled together so many examples of patterns and processes that add to the complexity of what is currently happening to reefs.  After reading it, most reef scientists will find it difficult to look at a three-agent, coral-algae-herbivore model, with maybe a couple of positive feedback loops thrown in, as anything more than a stick figure sketch, pretty far removed from what is happening on real coral reefs.  After all, how likely is it that variations in intensity of herbivory are the sole, or even the major, driving force leading to the replacement of corals by algae on reefs around the world?

Complexity revealed in a tiny field experiment

My final discovery was published in August in Coral Reefs, by Michael A. Gil, formerly at University of Florida, Gainesville, and now at U.C. Davis, and four colleagues associated with CRIOBE, that wonderful French marine research station in Moorea, French Polynesia, where the work was done.  This paper, part, I presume, of Gil’s PhD dissertation, was a classic manipulative field experiment, of a type that used to be somewhat more common on coral reefs than today.  It was small-scale, and lasted just 6 months, but represents well the general point I am trying to make (that we need to stop clinging to our simple models of reality on coral reefs).

Gil and colleagues forced themselves to look at just one aspect of the reef degradation story.  Namely, what happens to those reef organisms that remain after a reef has been degraded by Crown-of-thorns starfish or a storm.  Well, that’s what they claim, but actually they looked at effects on two species of corals and on turf and foliose algae in a classic three-factor, randomized block design, with two levels of each of the three factors applied, and nine replicates.  For non-scientists, that means they set out 72 experimental units including corals and substrata for algae and coral recruits to grow on, in a shallow back-reef region which had experienced the damaging effects of a Crown-of-thorns outbreak, followed by a cyclone, three years prior to the experiment.  They experimentally manipulated three potential stressors acting on their experimental units: sedimentation, nutrient enrichment, and overfishing, applying two levels (high and low) of each of these in all possible combinations to individual units.  Nine units received each of the eight different combinations of the three stressors.

Lest you think this was an immense undertaking, the experiment lasted just 6 months, and the experimental units were individual concrete building blocks to each of which were affixed one small branch of Acropora pulchra, a similar sized branch of Porites rus, and four 10 x 10 cm terra cotta tiles on which coral recruits, algae and other organisms would settle and grow.  The stressor manipulations were managed as follows: effect of overfishing was represented by enclosing some units within coarse-meshed wire cages to exclude large fish herbivores, while uncaged units, with more herbivory, represented a less-intensely fished environment.  Nutrients were either at the ambient levels for their study site, or an elevated level provided by slow release plant food in open tubes secured within the openings of the concrete block.  Sediment load was at ambient levels on some units, and elevated on others by manual delivery of measured amounts of sediments in daily pulses over two to three days each week of the experiment.  At the end of the experiment they reweighed each of the corals, recorded the percent of the colony still living, removed and weighed each of turf algae and macroalgae on two of the four tiles, and counted coral recruit skeletons remaining after removing all organic matter from one of the remaining tiles.  These measurements provided estimates of coral survivorship and growth, coral recruitment, and turf and macroalgal growth on each of their experimental units.  So, what did they find?

Diagram summarizing the various interactions Gil and colleagues found acting in their experiment.  Figure 1 from their paper © Coral Reefs.

As usually happens in ecology when several factors are included in the experiment, they found significant interactions among stressors in their effect, and quite different patterns of effect on the different components of the system they had measured.  Several of the results were unexpected and some were complex.  For example, A. pulcra proved highly susceptible to predation, and were grazed down to nothing on all the uncaged units while surviving far better when caged (having them mounted, enticingly, on top of concrete blocks clearly made them more vulnerable than they would be out on the reef!).  This did not happen to P. rus growing just inches away from the A. pulcra, however many specimens of P. rus showed partial mortality.  They found an interesting interaction between sedimentation and caging on this P. rus mortality: enhanced sedimentation improved coral survivorship in caged units, but reduced survivorship in open units.  Their interpretation of this result is that sedimentation influenced coral survivorship because of its impacts on algal growth.  Specifically, sedimentation reduced algal growth in the caged units (acting to counter the effects of reduced herbivory), but might also have reduced herbivory on uncaged units because the sediment-covered algae were somehow less palatable.  Thus, enhanced sedimentation has opposite effects on algae depending on level of herbivory occurring, and opposite but indirect effects on coral mortality.  When they examined growth of the still-living portions of corals, they found both coral species responded positively to nutrient enrichment.  Corals are not all alike, and these two species behaved differently in the context of this experiment.

Photos of one of Gil’s experimental units – a small nubbin of each of two coral species, and four terra cotta tiles all mounted on top of one concrete block.  Figure 2 from the paper © Coral Reefs.

Gil and colleagues also found complicated responses by the algae.  Foliose macroalgae were less abundant on tiles from cages receiving enhanced sedimentation, but were more abundant on tiles from uncaged units that also received extra sediments (the opposite effects used to explain the results for P. rus mortality).  Gil and colleagues draw attention to the difference between this result and a previously published study to suggest that the relationship between macroalgae and sedimentation is non-linear, another way of saying ‘more complex than we might assume’.  Turf algae showed a three-way interaction among nutrient levels, sediment levels and grazing pressure, brought about by pronounced enhancement of growth when nutrients were enriched in caged units, suppression of growth by enhanced sedimentation in caged units, and negligible effects of either of these factors on turf algal growth in open units subject to herbivory.

If anything, Gil’s study illustrates there is lots we do not yet understand about the interactions among turf algae, macroalgae, and two species of coral, or about how these interactions are modified by variation in sedimentation, nutrient load, or level of herbivory.  Many will dismiss this study as a short-term, incredibly artificial experiment, but it has teased out some unexpected complexity in how these organisms interact.

How do we advance our understanding of coral reefs?

I am sure I could have found many other papers, published recently, which reveal complexities in the ecological interactions in coral reef systems.  But it was easy to find these three, quickly, within two issues of one journal.  Each takes a very different approach, and each reveals important complexities.  Coral reef scientists read such papers regularly, yet while we absorb the messages being delivered, we do not seem to be applying those messages to build a more robust understanding of the complexities that exist.  Instead, we continue to rely upon simple models of singular processes, among minimal numbers of actors, as our way of trying to explain the dynamics of coral reef systems.  If we are to have any hope of applying our science in useful efforts to sustain or restore coral reefs, we need to become much more sophisticated in how we think about them.

I am not suggesting that reef scientists have been particularly backward or inept.  One could make similar arguments concerning the ecology of most systems on the planet.  Nor am I suggesting our understanding has not improved markedly over the last few decades.  But if the need to actively manage this planet is becoming as urgent as I believe it is, it is imperative that we attempt to build the robust science that might let us do that steering.  The question is, are we capable, collectively, of the much more complex thinking that will be needed?

There are some steps that could help.  First is to strengthen our critical thinking capacities so that we can subject new reports to a detailed analysis to find out what they are really contributing to our developing understanding.  This means actually reading far more publications, instead of just their abstracts, and citing them because they say something important, rather than because our colleagues are talking about them.  I remember a time early in my career when we marine ecologists would joke about the ‘West coast opening paragraph’ which could be distinguished from an “East coast’ analog, or a “European’ analog by the particular string of citations to standard west coast ecological work.  Writing the opening sentences to the introduction of an article back then was a bit like reciting a specified piece of scripture that alerted everyone to the fact that the author did indeed belong to that particular sect.  We have not progressed very far beyond that approach, and with the growing tide of new articles, in new journals, available instantly on the web, we are probably doing an even poorer job of critical reading than we were back then.

Second is to guide students to design research topics to tackle specific questions, questions that are ripe for being answered, and to use the answers to build a stronger body of science.  Research should be more than winning a grant, collecting some data, finding some interesting patterns, preparing several eye-catching figures, and generating some text to wrap it all together as the next in a long stream of articles, each one carefully promoted on-line as the newest, greatest, most amazing breakthrough in the field this century.  There are strong forces at work, including those that can impede progress in scientific careers, acting to force us away from thoughtful, careful, precisely designed studies to answer clear, focused questions.  Scientists must resist those forces; with breathless self-promotion increasingly the order of the day, resisting will be difficult.

A third step, I suggest, is to ensure that, once in a while, we reflect on our limitations as data-evaluating beasts.  We need to recognize, for a start, that even with the most sophisticated statistical and graphing programs now available, we still rely on our eyes and ears, and especially on our brains to interpret our data.  Seeing the things that are there, which we do not expect, is particularly difficult.  Indeed, there may be ways to train scientists to think in more than three dimensions, to imagine numbers bigger than 7, and to remember that the systems they study usually have more than 7 different agents interacting in more than 7 ways.  Recognizing the complexity of real systems even while trying to focus on a subset of agents and processes (in order to make any progress at all), is something we need to do continually yet is something easily forgotten.

If not forewarned, we mostly fail to see the gorilla in this classic experiment.  Watch it on You Tube.  If we can miss a gorilla, what else can we miss when evaluating research results?

A fourth step is to learn from other areas of science.  Ecologists need to force themselves to step outside the boundaries of their chosen research system to learn about techniques and discoveries of ecologists working on different questions in different systems.  Learning from scientists who are not ecologist is also worthwhile – and this may be the major benefit of doing research within truly interdisciplinary, as opposed to multi-investigator research groups.  As one example of learning from other ecologists, take herbivory.  Algae of different types are eaten by a range of herbivores who process them quite differently and therefore have marked preferences (and effectiveness as grazers) among the different types.  Macroalgae differ in palatability.  Macroalgae start out small, but given the chance can grow too large to be grazed by many of the herbivores.  Intertidal ecologists have studied to death how algae on rocky shores can escape herbivory; theirs is a story of considerable complexity, and rocky shore systems are far simpler than coral reefs.  Herbivory in terrestrial systems is similarly complex and well-studied.  Yet, all three of the papers I discussed here treat herbivory as if it is a unitary process operating with the simplicity of action of a good ride-on lawnmower.  I guarantee that herbivory on coral reefs is way more complex than that, and I am also certain the authors of all three papers know that.  It is time to apply that knowledge.

My final suggestion (and, no, I did not apply all these steps consistently or particularly successfully in my own career) is that we try to preserve humility.  Each of us is struggling to apply our science to build knowledge of how the world works.  Each of us is operating with a sensory and cognitive system evolved to permit us to find food, shelter and mates, while avoiding predators, diseases, and environmental catastrophes.  Our evolution has done a remarkable job, but we do not have some god-like gift for evaluating complex multi-component, multi-process systems and quickly deducing how they work.  As I said near the beginning, ecology is not rocket science; ecology is far more difficult than rocket science.  Our urgent challenge is to recognize the need and apply the effort to do much better than we have been doing.

This impressive photo of a red sea reef captures the sheer complexity of the ecosystem reef ecologists strive to understand.  Can we visualize the myriad pathways for exchange of energy, nutrients, or information, especially when we understand that many of the creatures will live here, with their neighbors, for a decade or more?  It is definitely more complicated than rocket science!  Photo © Lynn Wu was the 2015 overall winner of the In Water Photographer of the Year competition.