Jonathan Chase

The Shadow of Bias

effect size, from a single study, and then combines effect sizes between studies on the same topic (along with estimates of sample size and variance) to allow detection of the overall magnitude of effect. In this way, even if studies give somewhat conflicting answers to the same treatment, an overall effect among studies can be calculated to achieve a more conclusive answer. While meta-analysis is a powerful tool to overcome the variation among studies and arrive at an answer to a particular scientific question (e.g., does a particular intervention alleviate the symptoms of a disease?), it is less powerful in its ability to detect publication bias and the selective presentation of analyses. In the biomedical sciences, such biases not only slow the progression of science, but they could also result in bringing ineffective or harmful substances to clinical trial, creating considerable financial and health costs. Thus, it is important to understand just how rampant these biases are. In the current issue of PLOS Biology, Tsilidis and colleagues take the bold step of examining bias by employing a relatively new type of approach—a sort of meta-analysis of meta-analyses. This allowed them to assess whether the numbers of studies finding statistically significant effects of a biomedical intervention were higher than what would be expected if there were no bias. Specifically, they analyzed 160 separate meta-analyses comprising more than 1,000 studies that used animal models to evaluate the efficacy of interventions of six major neurological disorders—Alzheimer disease, multiple sclerosis, two types of stroke, Parkinson disease, and spinal cord injury—4,445 comparisons in all. A large proportion of these meta-analyses (nearly 70%) reported an overall positive effect of the tested interventions on the affliction. However, most of these meta-analyses also reported

Historical and contemporary factors govern global biodiversity patterns.

A novel hierarchical framework integrates the effects of time, area, productivity, and temperature at their respective relevant scales and successfully predicts the latitudinal gradient in global vertebrate diversity.

Which Came First: Burden of Infectious Disease or Poverty?

A few years ago, I was at an event where the CEO of a coal company was giving the keynote extolling the virtues of ‘‘green coal’’. The crux of his argument was that restrictions on carbon emissions from fossil fuels to generate electricity would in fact be quite detrimental, particularly to the world’s poorest people. He backed this up by showing positive correlations between the per capita electricity use in a given country and several indices of the quality of the human condition, which he interpreted to mean that electricity usage causes economic prosperity. Although it was clear that he was twisting cause and effect, my statistical rancor didn’t reach its limit until he showed that people from countries with lower per capita electricity use were also more likely to suffer from some of the world’s worst infectious diseases and had lowered life expectancies. His concluding slide stated ‘‘coal-generated electricity is not only good for the economy, but also good for your health’’. Clearly, economics plays a strong role in the burden of infectious diseases. Wealthier countries can invest more in immunizations, control of disease vectors, and treatment following infection, allowing people to live longer, healthier lives, while those in poorer countries are relegated to an often shortened lifetime filled with malady. However, our collective anthropocentric perspective often forgets that we are embedded within a complex ecological web. Nearly two-thirds of the pathogens and parasites that infect humans involve interactions with animals as vectors or alternative hosts. These include some of the worst chronic and emerging diseases we face, including malaria, cholera, and plague. While it’s well known that populations with high infectious disease burden are less prosperous, the cause and effect relationships between infectious diseases and prosperity are so intimately intertwined that we have not been able to disentangle whether pathogens drive economies or economies tame pathogens. Said another way, are poor people more likely to get sick, or are sick people more likely to be poor? A paper published in this issue of PLOS Biology by an interdisciplinary team of ecologists and economists led by Matthew Bonds, who has PhDs in both ecology and economics, makes a substantial advance in understanding the interplay between infectious diseases and economic prosperity. It does so by applying a form of structural equation modeling, which allows several different pathways of causality to be examined simultaneously, to disentangle the relative importance of each. The team jointly examined World Bank data on the per capita income of people from 139 countries and the burden of some of the most globally important parasitic and vector-borne diseases, measured as the per capita years of life lost due to mortality and the weighted equivalent of years lost due to morbidity. They also included several other variables known to influence income and infectious disease in a systematic way. For example, economies are known to be strongly influenced by several indicators of governance, including political stability and corruption, while the ecological burden of infectious disease is strongly influenced by latitude (tropical countries have higher burdens than temperate countries). With the caveat that disentangling cause and effect is always a bit tricky with comparative data, the beauty of the statistical approach employed is that it allowed the investigators to evaluate the relative importance of a potential causal link in the system of equations while controlling for other confounding variables. It also allowed them to evaluate the success of alternative scenarios by which ecology and economics interact. For example, on the economic side, there is a well-known positive relationship between per capita income and latitude; wealthier countries tend to be at higher latitudes, whereas tropical countries tend to be poorer. Economists have typically assumed this to be a historical artifact of colonial expansion from Europe and erroneously discounted the role of pathogens. The current study showed that the latitudinal income gradient most likely results because infectious diseases are more burdensome in the tropics, and that per capita income in many impoverished tropical countries could be doubled simply by reducing the burden of disease to the level found in temperate areas. On the ecological side, the authors examined an important, but controversial,

A Recipe for Achieving Aichi: Conservation Planning for 2020 Biodiversity Targets

While much remains to be known, decades of careful research have documented rapid global declines of biodiversity at the hands of humans, perhaps approaching rates not seen since the last mass extinction more than 65 million years ago. Furthermore, in addition to well-founded moral and ethical reasons to be concerned about and mitigate biodiversity loss, recent years have seen a skyrocketing recognition by scientists, governmental policy-makers, and the general public of more self-interested values of biodiversity, including economic returns, mitigation of global changes, and benefits to human health. Figure 1 A mismatch of priorities. The map shows the distribution of priorities for establishing new protected areas to meet the 17% targets under Aichi Target 11. Red indicates protection at minimal cost and ignoring ecological representation. Green indicates ... As a consequence of the recognition of the value of biodiversity and the need to quell its loss, several international consortia have devised various goals for the conservation of biodiversity. One of the more globally significant of these is the United Nations–initiated Convention on Biological Diversity, which in 2010 established a series of goals for the upcoming decade—collectively known as the Aichi Targets (after the location where the meeting took place: Aichi Prefecture, Japan). These targets include achieving greater awareness of biodiversity loss and the value of biodiversity, more sustainable use of resources for the protection of biodiversity, and safeguarding against future losses of biodiversity. Because the primary driver of biodiversity loss is habitat loss, one of the main strategic goals of the Aichi Targets includes increasing the amount of protected terrestrial habitat (excluding Antarctica) from the current 13% to 17% across the globe by 2020 (Aichi Target 11). With nearly 200 nations agreeing to the principles of the Aichi Targets, this could lead to the most rapid rate of land preservation in history, even if the targets are not fully achieved. Another key goal is to prevent the extinction of species already known to be threatened with future extinction and to achieve improvement towards sustainability in their populations by 2020 (Aichi Target 12). On the surface, achieving these two targets might seem quite complementary; preserving land should directly benefit the preservation of threatened species. Unfortunately, the ecological demons that have plagued scientists and policy-makers for decades by making what might seem to be a simple relationship on the surface into something much more complex are hard at work to thwart the Aichi targets. As pointed out by Venter and colleagues in this issue of PLOS Biology, the targets for protecting land and for protecting threatened species are not necessarily congruent, and in fact, a “business-as-usual” approach for land preservation to achieve the 17% target will do very little to increase the protection of threatened species. The crux of the argument boils down to the fact that not all land is equal when it comes to biodiversity preservation; there are biodiversity hotspots and biodiversity coldspots. And unfortunately, biodiversity hotspots, which tend to contain a higher proportion of threatened species, also have a tendency to have high economic value for uses other than preservation, such as agriculture. To quantify current and future protection of threatened species within preserved areas, Venter and colleagues overlaid publically available data on the distribution of protected areas across the terrestrial extent of the globe (excluding Antarctica) and the ranges of several International Union for the Conservation of Nature (IUCN) Red List “threatened” species. Of the 4,118 threatened species they considered, 17% of them are not contained in a single protected area and only 15% (603) are considered to be adequately protected by habitats currently set aside for biodiversity. Importantly, these numbers are not fundamentally different from those calculated a decade ago, indicating disappointingly little progress. Next, Venter and colleagues estimated the proportion of threatened species which could become protected through achieving 17% of land protected as mandated by Aichi Target 11. As a first pass, they assumed that governments would behave largely using a “business-as–usual” strategy for land preservation, meaning they would establish preserves in areas that have the least potential value for other uses (e.g., agriculture), allowing them to achieve the target area while minimizing the lost economic opportunity costs. These less economically valuable habitats, however, also tend to be less productive or otherwise constrained in the numbers of threatened species they house. As a result, Venter and colleagues predict that despite achieving the 17% land preservation of Aichi Target 11, only 249 more threatened vertebrate species would be adequately protected with this extended network, leaving 79% of these threatened species still at relatively high risk and doing little to achieve Aichi Target 12 (i.e., sustainability of threatened species). Notably, the Aichi Target 11 has wording designed to encourage land preservation strategies beyond a simple area target, including phrases for preservation such as “especially areas of particular importance for biodiversity and ecosystem services” and “ecologically representative.” To examine the potential of the latter criterion, Venter and colleagues examined a scenario where future land preservation was equally distributed among ecoregions of different vegetative communities. Again, they find only a marginal benefit in terms of numbers of threatened vertebrates which gained adequate protection, despite a 450% increase in lost-opportunity cost of setting land aside, relative to the “business-as–usual” scenario. They estimated the potential lost-opportunity costs of setting aside land that would ensure the adequate protection of all 4,118 threatened species to be nearly

Adapting to Change

43 billion (in US dollars), around 750% more opportunity cost lost than with “business-as–usual.” How then, can we reconcile the low economic cost but low conservation benefit scenario of “business-as–usual” designation of protected areas, given the exceedingly high economic costs that it would take to set aside land to achieve Aichi Target 11s 17% land protection goals—along with Target 12s mandate to minimize extinctions of already threat ened species? Importantly, Venter and colleagues identified a critical nonlinearity in the relationship between the costs of establishing new reserves and the benefits of achieving adequate protection of threatened species. That is, small increments of higher lost-opportunity cost lead to proportionately larger increments of adequate protection of threatened species. For instance, achieving a 400% increase in the adequate protection of threatened species only costs 50% more, in terms of lost-opportunity cost, than the “business-as-usual” strategy. There are, of course, many caveats inherent in the specific estimates put in place by Venter and colleagues. However, the nonlinearity is likely a robust result that points towards a “happy medium” where countries can gain considerable benefits in terms of biodiversity preservation with minimal lost-opportunity costs by incorporating considerations of threatened species into their networks of protected areas. Hopefully, the Aichi Targets will enjoy better success than many previous goals from the Convention on Biodiversity Conservation that were ultimately left unrealized, perhaps because the goals were unrealistic and/or because they were not well placed within socioeconomic constraints. By providing a better road map towards achieving biodiversity conservation goals within an explicit socioeconomic framework, the analysis by Venter and colleagues provides a step towards accomplishing these targets. And none too soon—the front page of the Convention on Biodiversitys website (http://www.cbd.int/) indicates that there are less than 2,500 days to achieve the Aichi Targets…and counting. Venter O, Fuller RA, Segan DB, Carwardine J, Brooks T, et al. (2014) Targeting Global Protected Area Expansion for Imperiled Biodiversity. doi:10.1371/journal.pbio.1001891

Does a Warmer World Mean a Greener World? Not Likely!

Biologists typically use the climatic conditions in which species currently live to project the likelihood that those climates will be available following a period of climate change, and these almost always predict dramatically high rates of species extinctions and biodiversity loss in future climates. In many instances, such predictions are warranted. Polar bears (Ursus maritimus), for example, are projected to be at high extinction risk as a result of the global warming-induced loss of arctic sea ice on which they depend for hunting their prey. For other species, however, the effect of climate change on extinction risk is less obvious. Because species can adapt to changing environments through both phenotypic plasticity (changes in behavior, physiology, and/or morphology) and micro-evolutionary adaptation, many can persist and even thrive in the face of changing climates. Newly hatched great tit Parus major nestlings. Data from a long-term study of this species are used to explore the adaptive significance of phenotypic plasticity in timing of breeding, and the importance of phenotypic plasticity in adjustment to climate ... A recent theoretical framework suggests that the likelihood of a species adapting to changing climates to avoid extinction results from the interaction between phenotypic plasticity and micro-evolutionary change, as well as environmental (rate of climate change) and demographic (life history) variables. While this theorys utility lies in its ability to predict which populations will escape extinction due to climate change through adaptation, the data necessary to make these predictions require explicit knowledge of both phenotypic plasticity and the strengths of selection in response to variable climates, which are rarely available at the appropriate scale and thus has not been attempted. That is, until the analyses by Vedder and colleagues published in this issue of PLOS Biology. Vedder and colleagues used data from a population of great tits (Parus major) at Wytham Woods, a moderately sized forested site near Oxford University that is one of the most intensively studied bird populations ever. This population has been monitored in the same way each spring since 1960, using more than 1,000 nestboxes on site. A key piece of data collected throughout this time period is the timing in which birds lay their eggs relative to spring temperatures, and in particular, how well they track the timing of peak caterpillar abundances, a critical food these birds need to successfully rear their young. These birds now lay their eggs on average two weeks earlier than they did 50 years ago, primarily as a result of phenotypic plasticity in response to a concurrent shift in the timing of peak caterpillar numbers. By using the relationship between egg-laying date and temperature among hundreds of individuals breeding each year, Veddar and colleagues calculated the parameters necessary to project population-level changes in the egg-laying date and the likelihood of population persistence or extinction. Specifically, they estimated the genetic variance of egg-laying date, the strength of selection on laying date (the numbers of offspring of a parent returning the next year), how that selection varied with temperature (and the timing of peak caterpillar numbers), and the phenotypic plasticity in egg-laying date with respect to temperature. From this parameterized model, they predicted that this great tit population could adapt and persist in an environment that was warming up to 0.5°C per year; much higher than even the most dramatic predicted warming from climate models (0.03°C per year). Even when they included uncertainty in the parameter estimates, the population had less than one-tenth of one percent chance of going extinct with the highest projected climate warming. Importantly, the incorporation of phenotypic plasticity was critical for these projections; without plasticity allowing the birds to adjust the timing of their egg-laying, the population had a 60% chance of going extinct. Although their study provided a compelling case study for the need to include both plasticity and microevolutionary dynamics to project extinctions due to climate change, Veddar and colleagues wanted to take their simulation models beyond great tits in Wytham Woods to see if they could learn something more general about adaptation to climate change. First, they varied the life history parameters in their model, including generation time and population growth rate, finding that species with slower life histories (long generations and low population growth) are less likely to be able to track changing environments and are more at risk of extinction than shorter-lived species with higher population growth. Second, they varied the degree of genetic variability and strengths of selection, again concluding that this great tit population and other species like it are likely to be able to adapt to even the most pessimistic climate change scenarios owing to the interaction between plasticity and microevolutionary adaptation. The great tits of Wytham Woods are able to adapt in the face of climate change through a combination of phenotypic plasticity and microevolutionary response in egg-laying date, allowing them to track changes in caterpillar densities. Other species might also be able to track changing climates through the interplay between plasticity and microevolution, although their ability to do so will depend critically on their life histories; shorter-lived species will adapt more readily than those that are longer-lived. Vedder O, Bouwhuis S, Sheldon BC (2013) Quantitative Assessment of the Importance of Phenotypic Plasticity in Adaptation to Climate Change in Wild Bird Populations. doi:10.1371/journal.pbio.1001605

A Fool to Do Your Dirty Work

Despite the “gloom and doom” scenarios depicted by most climate change scientists, the warmer world that we are creating can’t be all bad, can it? After all, hundreds of studies have shown that plant productivity is higher when temperatures are warmer and atmospheric carbon dioxide is high. Many models even predict an increase in global plant productivity, which has provided fodder for many to advocate the benefits of climate change to humans. Why wouldn’t we want a warmer and CO2-enriched world if it means higher productivity of plants, especially in impoverished regions where even slight increases in plant production could mean the difference between starvation and prosperity? Unfortunately, the simple idea that global warming could provide at least some benefits to humanity by increasing plant production is complicated by a number of factors. It is true that fertilizing plants with CO2 and giving them warmer temperatures increases growth under some conditions, but there are trade-offs. While global warming can increase plant growth in areas that are near the lower limits of temperature (e.g., large swaths of Canada and Russia), it can make it too hot for plant growth in areas that are near their upper limits (e.g., the tropics). In addition, plant productivity is determined by many things (e.g., sunlight, temperature, nutrients, and precipitation), several of which are influenced by climate change and interact with one another. And so while we have hundreds of models that are available to project future climate change based on a number of different scenarios, we really have little clue as to what this future climate-changed world might look like in terms of plant production—that is, until now. In this issue of PLOS Biology, Mora and colleagues provide a broader view of how global warming will likely change plant productivity in several ways. First, they include multiple factors that influence plant growth and are expected to be affected by climate change—temperature, soil water content, and sunlight levels –and their interactions. Second, they examine expected changes in different regions across the globe, examining the net increases and decreases in the number of days suitable for plant growth under climate change. Third, they examine the correlations between these expected changes in different regions and the vulnerability of people living in those regions. The answers they find, in a nutshell, are that if carbon emissions remain on their current trajectory (the status quo), the losses of plant production are likely to be far greater than any gains when examined at a global scale, and a substantial number of people, especially those who are most impoverished, will be at greater risk. Importantly, however, Mora and colleagues repeated their analysis for several global change scenarios and found that these changes will be much more moderate if society agrees to restrict carbon emissions in real, but manageable, ways. A look into the nitty-gritty of Mora and colleagues’ analyses shines some light on how they arrived at these conclusions. First, they used remote sensing satellite data to correlate how plant growth responded to varying levels of temperature, soil moisture, and solar radiation. From this, they estimated the range of climate conditions that enabled plant growth on our planet. These thresholds then allow them to calculate the numbers of days in a given year in which positive plant growth was expected in each region of the world. Next, they used daily projections of Earth Systems Models into the year 2100 for temperature, soil moisture, and radiation in regions across the world in order to simulate changes in the number of suitable plant growing days. They compared plant growth days from today to those under three climate change scenarios, including the status quo on carbon emissions, which would lead to more than a doubling of current amounts of atmospheric CO2 (~930 ppm) by 2100, and two scenarios in which emissions levels are actively reduced. If emissions remain unchecked, the analysis by Mora and colleagues shows that the number of days with climatic conditions suitable for plant growth would only increase in a few places (China, Russia, and Canada), whereas many other places, particularly in the tropics, will show dramatic declines, leading to an overall decline in plant growing days across the globe by more than 10% (Fig 1). Importantly, however, even if moderate controls on emissions are accomplished by society, this projected loss of plant growing days all but disappears. Fig 1 Many terrestrial ecosystems are vulnerable to losses in plant primary production, because of reductions in the number of days when climatic variables are suitable for plant growth. The second main thrust of Mora and colleagues’ analysis was to link their predicted changes in plant growth days in different places across the world to the human populations that live in those places. In tropical evergreen forests, for example, a status-quo scenario predicts that plants will lose up to 25% of their suitable growing days because of temperatures that are too warm. A huge number of people in these regions depend on plants (both from forests and agriculture) for food, fiber, and fuel, and many of these people are also impoverished; they have little means to adapt if these goods and services are impaired by climate change. Mora and colleagues examined how their projected estimates of changes in the numbers of plant growing days correlated with two aspects of the human populations exposed to these changes: dependency on plant resources and their social adaptability to change. They found that under the status-quo scenario, nearly 3.5 billion people could be exposed to reductions of plant growth days by 30% or more. Of those, nearly 3 billion are highly dependent on plant resources, and 2 billion are also in low-income countries that are likely to suffer the most from those changes. On the other side of the coin, only 270 million or so people live in countries that are projected to experience significant increases in plant growing days (e.g., Scandinavia). Mora and colleagues’ analysis contradicts the currently assumed “silver lining” of global warming—that, despite the many documented costs of global warming, at least plant growth will be enhanced. Instead, there are likely to be many more parts of the world that experience reduced, rather than increased, plant growth as a result of global warming, and this will likely have a negative impact on a large proportion of the world’s population. There is, however, a different silver lining in Mora and colleagues’ analysis. If our global society is able to come together and restrict emissions even a moderate amount, the magnitude of this predicted change in global plant productivity and human well-being will be substantially reduced.

A Sea of Change

Much of life on Earth owes its spectacular success to a rather important evolutionary transition—from single-celled organisms to multicellularity—which has occurred independently in many lineages, enabling the differentiation of cells to perform the highly specialized functions that we see in living fungi, plants, and animals. However, whereas all clones of single-celled organisms have a relatively equal chance of dividing and propagating their genes, most multicellular organisms entrust the propagation of their genes to a few select germline cells amidst a sea of non-reproductive somatic cells. At this point, the fitness of individual cells and the fitness of the entire organism become decoupled. Anytime complexity increases through evolution, one must ask how selection at the lower level of organization (i.e., the individual cell) doesn’t disrupt the integration at higher levels of organization (i.e., a multicellular organism) by favoring selfishness. There are some general evolutionary hypotheses that have been offered to explain why and how multicellularity and the division of labor between somatic and germline cells evolved, as well as the conditions under which these developments would be expected. Clearly, organisms with differentiated cells can experience many fitness advantages, such as the ability to grow larger and exploit novel resources. And along with these advantages come costs, such as the energy and materials that must be allocated towards growth and maintenance, rather than reproduction. However, there are more subtle, but no less important, constraints on an organism’s ability to acquire resources, grow, metabolize, and reproduce that might also influence the evolution of cellular differentiation. One idea that has been suggested, but not yet fully developed, is that the evolution of multicellular organisms with separate somatic and reproductive cells might be influenced by constraints on the preservation of genetic information. Most of the ‘‘work’’ performed by a cell—that is, the production and use of energy—takes place in the mitochondria and chloroplasts (in eukaryotes) or across membranes (in prokaryotes). As a byproduct of this work, reactive oxygen species such as hydrogen peroxide are generated. In turn, these byproducts can create oxidative stress in a cell, one result of which can be mutations to that cell’s DNA. Here, the idea of the so-called ‘‘dirty work’’ hypothesis is that the advent of cellular differentiation allows the organism to separate the energetically costly and potentially mutagenic processes into their somatic cells, while protecting their genomes within germline cells that need perform little work. While this and other theories about the evolution of multicellularity and cellular differentiation are intriguing, empirical evidence is less forthcoming. Some studies in yeasts and cellular slime molds, among others, have provided a few clues. But the time scales necessary to observe and manipulate the processes driving the evolution of cellular differentiation are typically prohibitive. Unless, that is, one could reproduce the evolutionary process in a realistic, but tractable way. That is just what Goldsby et al. did in this issue of PLOS Biology. To explore the role of the dirty work hypothesis in the differentiation of somatic and germline cells, Goldsby et al. performed a series of evolutionary experiments on populations of digital organisms. What’s a digital organism? In this case, digital organisms have a genome that comprises a fully functional computer program. These genetic programs can process numbers that flow into and out of their habitat to perform computational logic functions (e.g., AND, NOT, XOR) which order to gain resources. These genomes mutate at some defined probability, and the organisms differentially survive and reproduce as a function of their ability to acquire resources (i.e., when enough functions are executed). With this basic framework in hand, any number of evolutionary questions can be investigated simply by defining the parameters in which the digital organisms interact. In this case, the authors explored whether and how these simple multicellular individuals make the transition to having non-reproducing somatic cells and reproductive germline cells. They established a series of evolutionary experiments

Parasites in Food Webs: Untangling the Entangled Bank

​WhenWhen most people think about the influence of burning fossil fuels on the global ecosystem, they usually think of the “greenhouse effect” caused by a skyrocketing increase in the levels of CO2 in the atmosphere, which traps a higher percentage of radiant energy and raises global temperatures. Despite politically charged opposition to an overwhelming scientific consensus, public concern regarding global warming and associated climatic change has grown substantially in recent years. Solutions ranging from “cap and trade” economic policies, development and use of alternative energies, and minimizing destruction of forested ecosystems that absorb CO2 are now part of our everyday parlance. Mora and colleagues estimate that by the year 2100 rising carbon dioxide levels in the atmosphere will severely impact 470–870 million people who depend on the oceans for their livelihoods via adverse effects on oceanic biogeochemistry. Less frequently discussed, though no less important, are the many subsidiary effects of increased CO2 on the global ecosystem, including those that can dramatically influence humans. More than two-thirds of the planet is covered by ocean, and we humans extract a large proportion of our food and other resources from the sea. Awareness of the influence of increased atmospheric CO2 on oceanic resources is typically limited to concerns over loss of sea ice and rising sea levels from warming temperatures. However, the higher temperatures that result from increased atmospheric CO2 can also change oceanic circulation and stratification, as well as increase the rate of photosynthesis by algae, altering the productivity of the system upon which the entire food web is based. Furthermore, altered temperatures and productivity can influence the level of dissolved oxygen in the water, which is critical for the respiration of most organisms. Finally, perhaps the most important influence of increased CO2 in the ocean has more to do with chemistry than climate; increased levels of CO2 dissolved into seawater creates carbonic acid, ultimately reducing oceanic pH. In fact, the pH of the worlds oceans has already decreased by 0.1 since the industrial revolution (representing a ∼25% decrease, since pH is measured on a logarithmic scale). This reduces the availability of calcium carbonate in seawater, upon which a wide variety of organisms such as corals and shellfish depend for skeletal material. When the effects of manmade CO2 emissions on seawater pH, temperatures, productivity, and oxygen are examined in total, the magnitude of change in oceanic waters over the next 100 years is projected to exceed that observed at any time in the last 20 million years. What is less well-known, however, is just which parts of the worlds oceans will be more vulnerable to these changes, which ecosystems and organisms will be exposed to these changes, and how those changes might impact human populations that depend on oceanic resources. In this issue of PLOS Biology, Mora and colleagues accomplish just this through an interdisciplinary collaboration between climate modelers, biogeochemists, ecologists, and social scientists. After calibration, Mora and colleagues used the most recent and robust models of projected climate change (as part of the Coupled Model Intercomparison Project Phase 5 in the Fifth Assessment Report of the Intergovernmental Panel on Climate Change) to examine how oceanic biogeochemical parameters are likely to change by the year 2100. They projected change under two different CO2 emissions scenarios—“business as usual,” leading to an expected 900 parts per million (ppm) of CO2 in the atmosphere (more than double todays level and triple the level from pre-industrial times), and one of intense CO2 mitigation action, leading to only 550 ppm CO2 and more modest climate change. Though there was variation in the projections for each emissions scenario depending on how the climate system was modeled, the overall trends from all model scenarios were similar. For both emissions scenarios, sea surfaces across the globe were projected to increase in temperature with concomitant declines in pH, dissolved oxygen, and phytoplankton production; the magnitudes of expected change predictably depended on the intensity of emissions control. In contrast to the sea surface, most biogeochemical parameters were projected to change considerably less on the sea bottom. In addition, changes in biogeochemical parameters were projected to differ between regions. For example, pH was projected to be least changed in tropical areas, whereas temperature and productivity were projected to change least in temperate areas. In all, the majority of the ocean system, both surface and floor, was projected to change one or more of these biogeochemical parameters towards values that tend to reduce ecosystem functions (e.g., reductions in productivity or pH). Next, Mora and colleagues examined the co-occurrence of these biogeochemical changes among areas of the ocean representing different habitat types and hotspots of biodiversity within those habitat types. As might be expected, given the differences in exposure in different oceanic realms, deeper water habitats (e.g., sea mounts, deep sea benthic areas) were projected to experience less change than shallower water habitats like coral and rocky reefs, mangroves, and sea grass beds. Furthermore, hotspots of corals and mangroves were projected to be exposed to less biogeochemical change, while hotspots of whales, pinnipeds, squid, and krill were projected to experience much greater shifts in biogeochemistry. As a final step, Mora and colleagues examined which groups of people living near the worlds coasts would be exposed to the higher or lower degrees of projected biogeochemical change, the degree of dependency of these people on oceanic resources, and the potential for societal adaptation to environmental change (quantified by per capita gross domestic product, indicating the ability to access alternative resources). They projected that when assuming only moderate CO2 increases, ∼1.4 billion people will be exposed to medium-to-high oceanic change by 2100; of those, ∼690 million will be in countries with a medium-to-high ocean dependence and ∼470 million of these will also live in low-income countries. Under the “business as usual” scenario, however, ∼2 billion will live near oceans with medium-to-high biogeochemical change, of which 1.12 billion will live in countries with medium-to-high dependence on the ocean and ∼870 million will have little ability to adapt. Of course, throughout, Mora and colleagues are careful to recognize that just because the biogeochemistry of the ocean changes, that does not mean those changes are necessarily “bad”; this is particular important when bringing humans into the equation. For example, its difficult to predict just how changing temperature, productivity, oxygen, and/or pH will influence important subsistence fisheries upon which the less affluent and adaptable societies may depend. Indeed, in the game of global change, for every species that loses, others will gain. However, many of the changes projected by Mora and colleagues, such as reduced pH and productivity, are known to have negative consequences for the functioning of many important ecosystems that house high levels of biodiversity and supply a number of services to humans. While change can sometimes be good, humans, especially those from populations that rely heavily on oceanic resources for subsistence, are likely to be severely impacted by changing oceanic biogeochemistry within a relatively short time period. The good news is that this fate is still in our hands. However, it is clear that “business as usual” will wreak much greater havoc on the future of our oceans than will more moderate, but entirely doable, CO2 mitigation strategies. Mora C, Wei C-L, Rollo A, Amaro T, Baco AR, et al. (2013) Biotic and Human Vulnerability to Projected Changes in Ocean Biogeochemistry over the 21st Century. doi:10.1371/journal.pbio.1001682

How Much Lox Is a Grizzly Bear Worth

In the last paragraph of On the Origin of Species, Darwin described the interactions among birds, insects, and plants within an “entangled bank” as a metaphor to emphasize that although this might seem complex, these species and their interactions were formed by a set of basic laws (i.e., natural selection). Ecologists interested in the nature of webs of species interactions and their influence on biodiversity often use the entangled bank metaphor, but rather than distilling this complexity into a basic set of underlying laws as Darwin had attempted, the entangled bank is often used in ecology to describe the seemingly unbound complexity of nature wrought with idiosyncrasy. Nevertheless, a handful of brave ecologists are determined to find order amidst the morass of complexity, even though their critics might say they are tilting at windmills.