Fluent Essays

1.The precis should be a minimum of 500 words.  2.Be sure to type your name, course number and section (400 or 401) at the top of the page. Type the précis number: Precis 1, 2, etc. and the date. 3.In a single sentence give the following:a.Name of author (if any), or title of the article, affiliation;b.Use an accurate verb (such as ‘asserts’, ‘describes’, ‘proves’, etc.)c.Followed by a that clause describing the core pointof the work (ex., Jane Goodall asserts that Global Warming is irrefutable) 4.The second sentence explains how the authors develop and support their major claim(s). 5.The third sentence gives a statement of the author’s purpose, followed by an in orderphrase. (ex., Jane Goodall, an ardent environmentalist, describes the major causes of global warming in order for decision makers to become aware of the problem and take action.) 6.The next paragraph should contain your personal conceptual Précis and contain the following:a.A clear statement of the significance and impact that the article had for you. “Liking” or not liking the text is irrelevant.b.Statement of some of your reasons that explain why (or why not) this article had some significance for you.c.What other people may think about the data/concepts put forward in this article, and how this differs or aligns with your interpretation. Do not leave this out or you will not receive a full grade. 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 www.sciencemag.org SCIENCE VOL 293 21 SEPTEMBER 2001 2207 I n the catalog of global environmental insults, extinctions stand out as irre- versible. Current rates are high and ac- celerating (1). After a recent conference (2), we concluded that preventing extinc- tions is practical, but requires innovative measures. Enforceable protection of re- maining natural ecosystems is an overar- ching recommendation. Our deliberations regarding the implementation of this led us to attempt to answer a number of ques- tions. The answers outlined here represent, if not consensus, then the opinion of the majority. Supplementary material summa- rizes unresolved debates (3). Is Saving Remaining Biodiversity Still Possible? Globally, the harm we inflict on biodiversi- ty often stems from impacts on vulnerable, biodiverse areas that contribute relatively little to overhall human well-being and of- ten diminish it. For example, tropical hu- mid forests house two-thirds of terrestrial species (4). Within half a century, tropical forests have shrunk by half, a loss of 9 mil- lion km2 (5), with several times more forest damaged than cleared each year (6). Yet clearing tropical forests has created only ~2 million km2 of the planet’s 15 million km2 of croplands (7). The poor soils under- lying many tropical forests soon degrade and are abandoned or contribute only marginally to the global livestock produc- tion (7, 8). About 10\% of these cleared forests are on steep mountain slopes where high rainfall has predictably tragic conse- quences to those who settle there (8). The Amazon, the Congo, and rivers in Southeast Asia hold almost half the world’s freshwater fish species. Their fates depend on the surrounding forest watersheds. Else- where, most accessible rivers are dammed and channeled (9), causing their faunas to be more threatened than terrestrial ones (10). Diversion of water for irrigation threatens ecosystems, such as the Mesa Central (Mex- ico) and the Aral Sea and its rivers (Central Asia). Irrigation projects are often economic disasters (11, 12), as salt accumulation quickly destroys soil fertility (13). Fishing contributes only 5\% of the glob- al protein supply, yet is the major threat to the oceans’ biodiversity. The multitude of fish species caught on coral reefs constitutes only a small, though poorly known, fraction of the total catch, but f isheries severely damage these most diverse marine ecosys- tems. Most major fish stocks are overfished (14); thus, mismanagement diminishes our welfare and biodiversity simultaneously. Conversely, protected areas enhance biodi- versity and fish stocks (15). It is at regional and local scales that human actions and bio- diversity collide. On land, 25 areas, called hotspots, contain concentrations of endemic species that are disproportionately vulnera- ble to extinction after regional habitat de- struction (16). These areas retain <10\% of their original habitat and have unusually high human population densities (17). Lo- cally, those who destroy biodiversity do so because they are displaced, marginalized, and perceive no alternative. Others do so for short-term profit (3). Is Protecting Biodiversity Economically Possible? Although a global reserve network covering ~15\% of each continent might cost ~$30 bil- lion annually (18), reserves in tropical wilderness areas and hotspots need only cost a fraction of this. Tropical wilderness forests, predominantly the relatively intact blocks of the Amazon, Congo, and New Guinea are remote and sparsely populated. Land values are low and sometimes equivalent to buying out logging leases. Recent conservation con- cessions suggest ~$10/ha for acquisition and management (3). Securing an additional ~2 million km2 and adequately managing the ~2 million km2 already protected for biodiversi- ty and indigenous peoples requires a one- time investment of ~$4 billion. Land prices for the densely populated hotspots are much higher. Of the 1.2 million km2 of unprotect- ed land, some will remain intact without im- mediate intervention, some is already too fragmented, and perhaps one-third consti- tutes the highest priority. A study of the South Africa fynbos hotspot (3) suggests a one-time cost of ~$1 billion, and so by ex- trapolation, ~$25 billion for the protection and adequate management of all hotspots. Additional marine reserves would likely re- quire ~$2.5 billion (3). These sums, although large, undercut arguments that saving biodiversity is unaf- fordable. They are of the same order of magnitude as the individual wealth of the world’s richest citizens—and 1/1000th the value of the ecosystem services that biodi- versity provides annually (19). This sug- gests a strategy of leveraging funding from governments and international agencies through private sector involvement. Will Protecting Areas Work? The pressures to destroy ecosystems are of- ten external (20). For example, the World Bank and the International Monetary Fund have indirectly encouraged governments to deplete their natural resources to pay off debt (21). Even when available, some coun- tries may view foreign purchase of conser- vation concessions as imperialism in a 21st- century guise. Almost all the hotspots were European colonies; one is still French terri- tory (3). Some countries have unstable gov- ernment, and others are at war. Some countries have welcomed pre-emp- tive purchasing of logging rights and other conservation actions, recognizing the advan- tages of protecting forests and receiving funds to do so. Unfortunately, cutting forests and otherwise depleting resources is too of- ten a way to personal aggrandizement among some government officials (22). How good is even a well-intentioned government’s guaran- tee of a forest’s security when its peoples need wood for cooking or land for farms? Will the government return the concession fees to those whose livelihoods are affected? Protected areas may be respected in one country, ignored in another, even attract ex- ploitation in a third. Although more money generally yields more protection, richly en- dowed parks may be severely threatened (Ev- erglades National Park; USA) and significant accomplishments are possible in the most economically unlikely places (Odzala Na- tional Park; Democratic Republic of Congo). Whereas overall assessments of what conservation actions work, what do not, and why they are long overdue, discus- sions of possible factors typically devolve into idiosyncratic case histories. Likely, there is no single answer to these multi- S. L. Pimm is at the Center for Environmental Re- search and Conservation, MC 5556, Columbia Uni- versity, New York, NY 10027, USA. Other author ad- dresses are available on Science Online (3). *To whom correspondence should be addressed. E- mail: [email protected] S C I E N C E ’ S C O M P A S S P O L I C Y F O R U M P O L I C Y F O R U M : E N V I R O N M E N T Can We Defy Nature’s End? Stuart L. Pimm,* Márcio Ayres, Andrew Balmford, George Branch, Katrina Brandon, Thomas Brooks, Rodrigo Bustamante, Robert Costanza, Richard Cowling, Lisa M. Curran, Andrew Dobson, Stephen Farber, Gustavo A. B. da Fonseca, Claude Gascon, Roger Kitching, Jeffrey McNeely, Thomas Lovejoy, Russell A. Mittermeier, Norman Myers, Jonathan A. Patz, Bradley Raffle, David Rapport, Peter Raven, Callum Roberts, Jon Paul Rodríguez, Anthony B. Rylands, Compton Tucker, Carl Safina, Cristián Samper, Melanie L. J. Stiassny, Jatna Supriatna, Diana H. Wall, David Wilcove 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 21 SEPTEMBER 2001 VOL 293 SCIENCE www.sciencemag.org2208 scaled problems. One process, however, emerges as a unanimous choice: to train and empower conservation professionals in each biodiversity-rich country. Should Conservation Research and Management Be Centralized or Distributed? At present, these capabilities are highly centralized in industrialized nations, while many key tropical areas have few conser- vation professionals. Our experiences point to the pressing need for more and better-trained people. Those at La Selva (Costa Rica), Comision Nacional Para el Conocimiento y Uso de la Biodiversidad (CONABIO), Mexico, the Humboldt Insti- tute (Colombia), the Centre for Ecological Sciences of the Indian Institute of Science (India), and the International Centre for Liv- ing Aquatic Resources (ICLARM), Philip- pines have been in place long enough to as- sist in training a new generation. Budgets for effective centers are a few million dol- lars per year. Roughly half a billion dollars would support 25 centers for a decade, enough for each hotspot and wilderness for- est without centers plus additional centers for marine and freshwater hotspots. Should Efforts Concentrate on Protection or on Slowing Harm? Most of us agree that immediate protec- tion of ecosystems and training of in-coun- try professionals is vital. Nonetheless, some effort should be allocated to actions to lighten the burden on future generations of conservation professionals (3). Others argue in favor of actions that stem the pro- cesses that harm biodiversity and encour- age those that protect it, with priority giv- en to actions yielding near-term results. Economic subsidies that degrade the environment are a common problem across terrestrial, freshwater, and marine ecosystems (23, 24). For instance, massive economic subsidies make unsustainable fishing practices possible (25, 26). Biodi- versity can be depleted if property rights give ownership to those whose “economic use” translates into short-term forest clear- cutting, transient crops or grazing, and longer-term land degradation. The public is often unaware of the costs of environmentally damaging policies. An- nually, subsidies for such policies cost $2 trillion globally (24). We recommend a fo- cused analysis on those governmental poli- cies that artificially alter market dynamics and that have the most detrimental impact on biodiversity. The overarching message is that sound economic and ecological strate- gies may often involve the same, and not conflicting, strategies. Alliances of the fis- cally frugal and the environmentally con- cerned are a still unexplored possibility. We recommend a major outreach to na- tional and international institutions that make loans for actions that degrade biodi- versity. Many of them could benefit from improved ecological standards that factor biodiversity protection into their decision- making. Obligations of parties to existing le- gal instruments (such as the United Nations Convention on Biological Diversity) should also be used to promote adequate incentives. Biodiversity-rich countries often lack le- gal mechanisms to encourage conservation. Tax savings, transferable development rights, and mitigation credits would at least allow private, public or indigenous landown- ers to secure economic benefits. Globally, the wilderness forests, if lost, would greatly exacerbate increasing atmospheric CO2. Their value as carbon sinks alone appears to be broadly similar to our estimates of what it would take to protect them (27). Capturing these values could save large areas through efforts designed to highlight their true value. Do We Know Enough to Protect Biodiversity? Most debate centers on identifying priority areas for conservation. Surely all remain- ing habitats across the species-rich tropics must be priorities, ones that do not depend on our knowing the scientific names for 1 of 10, or the geographical distributions of 1 of 100 species, or not having resolved complex issues of reserve selection (28). However, even modest scientific advances greatly improve the efficiency of our ac- tions. Knowing which areas within hotspots are especially important could re- duce costs considerably. Paradoxically, we are not limited by lack of knowledge, but by our failure to synthe- size and distribute what we know. Museums and herbaria are vast repositories of data on what species occurred where, while decades of remote-sensing imagery detail how fast the remains of species’ ranges are shrink- ing. Although a few of us question the utili- ty of these taxonomic repositories, the ma- jority emphasize the urgent need for more, globally distributed taxonomy (29). In contrast, there was broad consensus for a greatly expanded research effort into the links between biodiversity, ecosystems, their services, and people (30). Infectious diseases are entering human populations as our numbers increase and as we encroach upon tropical forests and other pathogen reservoirs. Global climate change will have major impacts on human health through changes in food production, access to fresh water, exposure to vector- and water-borne disease, sea-level rise and coastal flooding, and extreme weather events (31). In conclusion, we share mixed senses of concern, urgency, and optimism. Concern, because humanity’s numbers (and consump- tion) are increasing. Across several human generations, a transition to sustainable use of natural resources is essential, and we must protect biodiversity in the interim. The urgency is driven by the pending loss of a major portion of biological diversity in the first half of this century if we do not act im- mediately. Our optimism stems from the re- alization that greatly increasing the areas where biodiversity is protected is a clear and achievable goal, one potentially attainable by using funds raised in the private sector and leveraged through governments. References and Notes 1. S. L. Pimm et al., Science 269, 347 (1995). 2. See www.ldeo.columbia.edu/pimmlab. 3. Supplementary material is available on Science On- line at www.sciencemag .org/cgi/content/full/ 293/5538/2207/DC1 4. National Research Council, Research Priorities in Tropical Biology (National Academy of Sciences, Washington, DC, 1980). 5. N. Myers, The Primary Source: Tropical Forests and Our Future (Norton, New York, 1992). 6. D. C. Nepstad et al., Nature 398, 505 (1999). 7. S. L. Pimm, The World According to Pimm: A Scientist Audits the Earth (McGraw-Hill, New York, 2001). 8. A. Grainger, Int. Tree Crops J. 5, 31 (1988). 9. M. Dynesius, C. Nilsson, Science 266, 753 (1994). 10. B. A. Stein, L. S. Kutner, J. S. Adams, Eds., Precious Her- itage: the Status of Biodiversity in the United States (Oxford Univ. Press, Oxford, 2000). 11. M. Reisner, Cadillac Desert: The American West and Its Disappearing Water (Penguin USA, New York, 1993). 12. P. P. Micklin, Science 241, 1170 (1988). 13. United National Environment Programme (UNEP), Status of Desertification and Implementation of the United Nations Plan of Action to Combat Desertifi- cation (UNEP, Nairobi, Kenya, 2000). 14. Food and Agriculture Organization of the United Na- tions, “The state of world fisheries and aquaculture in 2 0 0 0 ” ; ava i l a b l e a t w w w. fa o. o rg / d o c re p / 0 0 3 / x8002e/x8002e00.htm (2000). 15. C. M. Roberts, J. P. Hawkins, Trends Ecol. Evol. 14, 241 (1999). 16. N. Myers et al., Nature 403, 853 (2000). 17. R. P. Cincotta et al., Nature 404 990 (2000). 18. A. James, K. J. Gaston, A. Balmford, Nature 404, 120 (2000). 19. R. Costanza et al., Nature 387, 253 (1997). 20. C. Kremen et al. Science 288, 1828 (2000). 21. N. Sizer, D. Plouvier, Increased Investment and Trade by Transnational Logging Companies in Africa, the Caribbean, and the Pacific (Joint Report for World Wide Fund for Nature-Belgium, World Resources In- stitute, and WWF-International, 2001). 22. See www.transparency.de/documents/newsletter/ 200.3/third.html 23. N. Myers, Nature 392, 327 (1998). 24. ———, J. Kent, Perverse Subsidies: How Tax Dollars Can Undercut Both the Environment and the Econo- my. (Island Press, Washington, DC, 2001). 25. M. Milazzo, Subsidies in World Fisheries (World Bank, Washington, DC, 1998). 26. D. Ludwig et al., Science 260, 17 (1993). 27. J. J. Hardner, P. C. Frumhoff, D. C. Goetze, Mitigat. Adapt. Strateg. Global Change 5, 61 (2000). 28. S. L. Pimm, J. H. Lawton, Science 279, 2068 (1998) 29. A. Sugden, E. Pennisi, Science 289, 2305 (2000) and E. O. Wilson, Science 289, 2279 (2000). That issue prompted responses in Letters [A. T. Smith et al., Sci- ence 290, 2073 (2000)] and a response to those re- sponses [W. J. Kress, Science 291, 828 (2001)]. 30. J. A. Patz, D. Engelberg J. Last, Annu. Rev. Publ. Health 21, 271 (2000). 31. D. J. Rapport, R. Costanza, A. J. McMichael, Trends Ecol. Evol. 13, 397 (1998). S C I E N C E ’ S C O M P A S S REVIEW doi:10.1038/nature09678 Has the Earth’s sixth mass extinction already arrived? Anthony D. Barnosky1,2,3, Nicholas Matzke1, Susumu Tomiya1,2,3, Guinevere O. U. Wogan1,3, Brian Swartz1,2, Tiago B. Quental1,2{, Charles Marshall1,2, Jenny L. McGuire1,2,3{, Emily L. Lindsey1,2, Kaitlin C. Maguire1,2, Ben Mersey1,4 & Elizabeth A. Ferrer1,2 Palaeontologists characterize mass extinctions as times when the Earth loses more than three-quarters of its species in a geologically short interval, as has happened only five times in the past 540 million years or so. Biologists now suggest that a sixth mass extinction may be under way, given the known species losses over the past few centuries and millennia. Here we review how differences between fossil and modern data and the addition of recently available palaeontological information influence our understanding of the current extinction crisis. Our results confirm that current extinction rates are higher than would be expected from the fossil record, highlighting the need for effective conservation measures. O f the four billion species estimated to have evolved on the Earth over the last 3.5 billion years, some 99\% are gone1. That shows how very common extinction is, but normally it is balanced by speciation. The balance wavers such that at several times in life’s history extinction rates appear somewhat elevated, but only five times qualify for ‘mass extinction’ status: near the end of the Ordovician, Devonian, Permian, Triassic and Cretaceous Periods2,3. These are the ‘Big Five’ mass extinctions (two are technically ‘mass depletions’)4. Different causes are thought to have precipitated the extinctions (Table 1), and the extent of each extinction above the background level varies depend- ing on analytical technique4,5, but they all stand out in having extinction rates spiking higher than in any other geological interval of the last ,540 million years3 and exhibiting a loss of over 75\% of estimated species2. Increasingly, scientists are recognizing modern extinctions of species6,7 and populations8,9. Documented numbers are likely to be serious under- estimates, because most species have not yet been formally described10,11. Such observations suggest that humans are now causing the sixth mass extinction10,12–17, through co-opting resources, fragmenting habitats, introducing non-native species, spreading pathogens, killing species directly, and changing global climate10,12–20. If so, recovery of biodiversity will not occur on any timeframe meaningful to people: evolution of new species typically takes at least hundreds of thousands of years21,22, and recovery from mass extinction episodes probably occurs on timescales encompassing millions of years5,23. Although there are many definitions of mass extinction and grada- tions of extinction intensity4,5, here we take a conservative approach to assessing the seriousness of the ongoing extinction crisis, by setting a high bar for recognizing mass extinction, that is, the extreme diversity loss that characterized the very unusual Big Five (Table 1). We find that the Earth could reach that extreme within just a few centuries if current threats to many species are not alleviated. Data disparities Only certain kinds of taxa (primarily those with fossilizable hard parts) and a restricted subset of the Earth’s biomes (generally in temperate latitudes) have data sufficient for direct fossil-to-modern comparisons 1Department of Integrative Biology, University of California, Berkeley, California 94720, USA. 2University of California Museum of Paleontology, California, USA. 3University of California Museum of Vertebrate Zoology, California, USA. 4Human Evolution Research Center, California, USA. {Present addresses: Departamento de Ecologia, Universidade de São Paulo (USP), São Paulo, Brazil (T.B.Q.); National Evolutionary Synthesis Center, 2024 W. Main Street, Suite A200, Durham, North Carolina 27705, USA (J.L.M.). Table 1 | The ‘Big Five’ mass extinction events Event Proposed causes The Ordovician event64–66 ended ,443 Myr ago; within 3.3 to 1.9 Myr 57\% of genera were lost, an estimated 86\% of species. Onset of alternating glacial and interglacial episodes; repeated marine transgressions and regressions. Uplift and weathering of the Appalachians affecting atmospheric and ocean chemistry. Sequestration of CO2. The Devonian event4,64,67–70 ended ,359 Myr ago; within 29 to 2 Myr 35\% of genera were lost, an estimated 75\% of species. Global cooling (followed by global warming), possibly tied to the diversification of land plants, with associated weathering, paedogenesis, and the drawdown of global CO2. Evidence for widespread deep-water anoxia and the spread of anoxic waters by transgressions. Timing and importance of bolide impacts still debated. The Permian event54,71–73 ended ,251 Myr ago; within 2.8 Myr to 160 Kyr 56\% of genera were lost, an estimated 96\% of species. Siberian volcanism. Global warming. Spread of deep marine anoxic waters. Elevated H2S and CO2 concentrations in both marine and terrestrial realms. Ocean acidification. Evidence for a bolide impact still debated. The Triassic event74,75 ended ,200 Myr ago; within 8.3 Myr to 600 Kyr 47\% of genera were lost, an estimated 80\% of species. Activity in the Central Atlantic Magmatic Province (CAMP) thought to have elevated atmospheric CO2 levels, which increased global temperatures and led to a calcification crisis in the world oceans. The Cretaceous event58–60,76–79 ended ,65 Myr ago; within 2.5 Myr to less than a year 40\% of genera were lost, an estimated 76\% of species. A bolide impact in the Yucatán is thought to have led to a global cataclysm and caused rapid cooling. Preceding the impact, biota may have been declining owing to a variety of causes: Deccan volcanism contemporaneous with global warming; tectonic uplift altering biogeography and accelerating erosion, potentially contributing to ocean eutrophication and anoxic episodes. CO2 spike just before extinction, drop during extinction. Myr, million years. Kyr, thousand years. 3 M A R C H 2 0 1 1 | V O L 4 7 1 | N A T U R E | 5 1 Macmillan Publishers Limited. All rights reserved©2011 www.nature.com/doifinder/10.1038/nature09678 (Box 1). Fossils are widely acknowledged to be a biased and incomplete sample of past species, but modern data also have important biases that, if not accounted for, can influence global extinction estimates. Only a tiny fraction (,2.7\%) of the approximately 1.9 million named, extant species have been formally evaluated for extinction status by the International Union for Conservation of Nature (IUCN). These IUCN compilations are the best available, but evaluated species represent just a few twigs plucked from the enormous number of branches that compose the tree of life. Even for clades recorded as 100\% evaluated, many species still fall into the Data Deficient (DD) category24. Also relevant is that not all of the partially evaluated clades have had their species sampled in the same way: some are randomly subsampled25, and others are evaluated as opportunity arises or because threats seem apparent. Despite the limita- tions of both the fossil and modern records, by working around the diverse data biases it is possible to avoid errors in extrapolating from what we do know to inferring global patterns. Our goal here is to high- light some promising approaches (Table 2). Defining mass extinctions relative to the Big Five Extinction involves both rate and magnitude, which are distinct but intimately linked metrics26. Rate is essentially the number of extinctions divided by the time over which the extinctions occurred. One can also derive from this a proportional rate—the fraction of species that have gone extinct per unit time. Magnitude is the percentage of species that have gone extinct. Mass extinctions were originally diagnosed by rate: the pace of extinction appeared to become significantly faster than background extinction3. Recent studies suggest that the Devonian and Triassic events resulted more from a decrease in origination rates than an increase in extinction rates4,5. Either way, the standing crop of the Earth’s species fell by an estimated 75\% or more2. Thus, mass extinction, in the conservative palaeontological sense, is when extinction rates accelerate relative to origination rates such that over 75\% of species disappear within a geologically short interval—typically less than 2 million years, in some cases much less (see Table 1). Therefore, to document where the current extinction episode lies on the mass extinction scale defined by the Big Five requires us to know both whether current extinc- tion rates are above background rates (and if so, how far above) and how closely historic and projected biodiversity losses approach 75\% of the Earth’s species. Background rate comparisons Landmark studies12,14–17 that highlighted a modern extinction crisis estimated current rates of extinction to be orders of magnitude higher than the background rate (Table 2). A useful and widely applied metric BOX 1 Severe data comparison problems Geography The fossil record is very patchy, sparsest in upland environments and tropics, but modern global distributions are known for many species. A possible comparative technique could be to examine regions or biomes where both fossil and modern data exist—such as the near-shore marine realm including coral reefs and terrestrial depositional lowlands (river valleys, coastlines, and lake basins). Currently available databases6 could be used to identify modern taxa with geographic ranges indicating low fossilization potential and then extract them from the current-extinction equation. Taxa available for study The fossil record usually includes only species that possess identifiable anatomical hard parts that fossilize well. Theoretically all living species could be studied, but in practice extinction analyses often rely on the small subset of species evaluated by the IUCN. Evaluation following IUCN procedures34 places species in one of the following categories: extinct (EX), extinct in the wild (EW), critically endangered (CR), endangered (EN), vulnerable (VU), near threatened (NT), least concern (LC), or data deficient (DD, information insufficient to reliably determine extinction risk). Species in the EX and EW categories are typically counted as functionally extinct. Those in the CR plus EN plus VU categories are counted as ‘threatened’. Assignment to CR, EN or VU is based on how high the risk of extinction is determined to be using five criteria34 (roughly, CR probability of extinction exceeds 0.50 in ten years or three generations; EN probability of extinction exceeds 0.20 in 20 years or five generations; VU probability of extinction exceeds 0.10 over a century24). A possible comparative technique could be to use taxa best known in both fossil and modern records: near-shore marine species with shells, lowland terrestrial vertebrates (especially mammals), and some plants. This would require improved assessments of modern bivalves and gastropods. Statistical techniques could be used to clarify how a subsample of well-assessed taxa extrapolates to undersampled and/or poorly assessed taxa25. Taxonomy Analyses of fossils are often done at the level of genus rather than species. When species are identified they are usually based on a morphological species concept. This can result in lumping species together that are distinct, or, if incomplete fossil material is used, over-splitting species. For modern taxa, analyses are usually done at the level of species, often using a phylogenetic species concept, which probably increases species counts relative to morphospecies. A possible comparative technique would be to aggregate modern phylogenetic species into morphospecies or genera before comparing with the fossil record. Assessing extinction Fossil extinction is recorded when a taxon permanently disappears from the fossil record and underestimates the actual number of extinctions (and number of species) because most taxa have no fossil record. The actual time of extinction almost always postdates the last fossil occurrence. Modern extinction is recorded when no further individuals of a species are sighted after appropriate efforts. In the past few decades designation as ‘extinct’ usually follows IUCN criteria, which are conservative and likely to underestimate functionally extinct species34. Modern extinction is also underestimated because many species are unevaluated or undescribed. A possible comparative technique could be to standardize extinction counts by number of species known per time interval of interest (proportional extinction). However, fossil data demonstrate that background rates can vary widely from one taxon to the next35,86,87, so fossil-to-modern extinction rate comparisons are most reliably done on a taxon-by-taxon basis, using well-known extant clades that also have a good fossil record. Time In the fossil record sparse samples of species are discontinuously distributed through vast time spans, from 103 to 108 years. In modern times we have relatively dense samples of species over very short time spans of years, decades and centuries. Holocene fossils are becoming increasingly available and valuable in linking the present with the past48,90. A possible comparative technique would be to scale proportional extinction relative to the time interval over which extinction is measured. RESEARCH REVIEW 5 2 | N A T U R E | V O L 4 7 1 | 3 M A R C H 2 0 1 1 Macmillan Publishers Limited. All rights reserved©2011 has been E/MSY (extinctions per million species-years, as defined in refs 15 and 27). In this approach, background rates are estimated from fossil extinctions that took place in million-year-or-more time bins. For cur- rent rates, the proportion of species extinct in a comparatively very short time (one to a few centuries) is extrapolated to predict what the rate would be over a million years. However, both theory and empirical data indicate that extinction rates vary markedly depending on the length of time over which they are measured28,29. Extrapolating a rate computed over a short time, therefore, will probably yield a rate that is either much faster or much slower than the average million-year rate, so current rates that seem to be elevated need to be interpreted in this light. Only recently has it become possible to do this by using palaeontology databases30,31 combined with lists of recently extinct species. The most complete data set of this kind is for mammals, which verifies the efficacy of E/MSY by setting short-interval and long-interval rates in a comparative context (Fig. 1). A data gap remains between about one million and about 50 thousand years because it is not yet possible to date extinctions in that time range with adequate precision. Nevertheless, the overall pattern is as expected: the maximum E/MSY and its variance increase as measurement intervals become shorter. The highest rates are rare but low rates are common; in fact, at time intervals of less than a thousand years, the most common E/MSY is 0. Three conclusions emerge. (1) The maximum observed rates since a thousand years ago (E/MSY < 24 in 1,000-year bins to E/MSY < 693 in 1-year bins) are clearly far above the average fossil rate (about E/MSY < 1.8), and even above those of the widely recognized late- Pleistocene megafaunal diversity crash32,33 (maximum E/MSY < 9, red data points in Fig. 1). (2) Recent average rates are also too high compared to pre-anthropogenic averages: E/MSY increases to over 5 (and rises to 23) in less-than-50-year time bins. (3) In the scenario where currently ‘threatened’ species34 would ultimately go extinct even in as much as a thousand years, the resulting rates would far exceed any reasonable estimation of the upper boundary for variation related to interval length. The same applies if the extinction scenario is restricted to only ‘critically endangered’ species34. This does not imply that we consider all species in these categories to be inevitably destined for extinction—simply that in a worst-case scenario where that occurred, the extinction rate for mammals would far exceed normal background rates. Because our computational method maximizes the fossil background rates and minimizes the current rates (see Fig. 1 caption), our observation that modern rates are elevated is likely to be particularly robust. Moreover, for reasons argued by others27, the modern rates we computed probably seriously underestimate current E/MSY values. Another approach is simply to ask whether it is likely that extinction rates could have been as high in many past 500-year intervals as they have been in the most recent 500 years. Where adequate data exist, as is the case for our mammal example, the answer is clearly no. The mean per-million-year fossil rate for mammals we determined (Fig. 1) is about 1.8 E/MSY. To maintain that million-year average, there could be no more than 6.3\% of 500-year bins per million years (126 out of a possible 2,000) with an extinction rate as high as that observed over the past 500 years (80 extinct of 5,570 species living in 500 years). Million-year extinction rates calculated by others, using different techniques, are slower: 0.4 extinctions per lineage per million years (a lineage in this context is roughly equivalent to a species)35. To maintain that slower million-year average, there could be no more than 1.4\% (28 intervals) of the 500-year intervals per million years having an extinction rate as high as the current 500-year rate. Rates computed for shorter time intervals would be even less likely to fall within background levels, for reasons noted by ref. 27. Magnitude Comparisons of percentage loss of species in historical times6,36 to the percentage loss that characterized each of the Big Five (Fig. 2) need to be refined by compensating for many differences between the modern and the fossil records2,37–39. Seldom taken into account is the effect of using different species concepts (Box 1), which potentially inflates the numbers of modern species relative to fossil species39,40. A second, related caveat is that most assessments of fossil diversity are at the level of genus, not species2,3,37,38,41. Fossil species estimates are frequently obtained by calculat- ing the species-to-genus ratio determined for well-known groups, then extrapolating that ratio to groups for which only genus-level counts exist. The over-75\% benchmark for mass extinction is obtained in this way2. Table 2 | Methods of comparing present and past extinctions General method Variations and representative studies References Compare currently measured extinction rates to background rates assessed from fossil record E/MSY*{ 7, 10, 15, 27, 62 Comparative species duration (estimates species durations to derive an estimate of extinction rate)*{ 14 Fuzzy Math*{ 44, 80 Interval-rate standardization (empirical derivation of relationship between rate and interval length over which extinction is measured provides context for interpreting short-term rates){ This paper Use various modelling techniques, including species-area relationships, to assess loss of species Compare rate of expected near-term future losses to estimated background extinction rates*{{ 7, 10, 14, 15 Assess magnitude of past species losses{{ 42, 45 Predict magnitude of future losses. Ref. 7 explores several models and provides a range of possible outcomes using different impact storylines{{ 7, 14, 15, 27, 36, 62, 81–84 Compare currently measured extinction rates to mass-extinction rates Use geological data and hypothetical scenarios to bracket the range of rates that could have produced past mass extinctions, and compare with current extinction rates (assumes Big Five mass extinctions were sudden, occurring within 500 years, producing a ‘worst-case scenario’ for high rates, but with the possible exception of the Cretaceous event, it is unlikely that any of the Big Five were this fast){ This paper Assess extinction in context of long-term clade dynamics Map projected extinction trajectories onto long-term diversification/ extinction trends in well-studied clades{ This paper Assess percentage loss of species Use IUCN lists to assess magnitude or rate of actual and potential species losses in well-studied taxa{ This paper and refs 6, 7, 10, 14, 15, 20, 36 and 62 Use molecular phylogenies to estimate extinction rate Calculate background extinction rates from time-corrected molecular phylogenies of extant species, and compare to modern rates 85 Fuzzy Math attempts to account for different biases in fossil and modern samples and uses empirically based fossil background extinction rates as a standard for comparison: 0.25 species per million years for marine invertebrates, determined from the ‘kill-curve’ method86, and 0.21 species35 to 0.46 species87 per million years for North American mammals, determined from applying maximum-likelihood techniques. The molecular phylogenies method assumes that diversification rates are constant through time and can be partitioned into originations and extinctions without evidence from the fossil record. Recent work has demonstrated that disentanglement of diversification from extinction rates by this method is difficult, particularly in the absence of a fossil record, and that extinction rates estimated from molecular phylogenies of extant organisms are highly unreliable when diversification rates vary among lineages through time46,88. * Comparison of modern short-term rates with fossil long-term rates indicate highly elevated modern rates, but does not take into account interval-rate effect. { Assumes that the relationship between number and kind of species lost in study area can be scaled up to make global projections. { Assumes that conclusions from well-studied taxa illustrate general principles. REVIEW RESEARCH 3 M A R C H 2 0 1 1 | V O L 4 7 1 | N A T U R E | 5 3 Macmillan Publishers Limited. All rights reserved©2011 Potentially valuable comparisons of extinction magnitude could come from assessing modern taxonomic groups that are also known from exceptionally good fossil records. The best fossil records are for near-shore marine invertebrates like gastropods, bivalves and corals, and temperate terrestrial mammals, with good information also available for Holocene Pacific Island birds2,33,35,42–44. However, better knowledge of understudied modern taxa is critically important for developing common metrics for modern and fossil groups. For example, some 49\% of bivalves went extinct during the end-Cretaceous event43, but only 1\% of today’s species have even been assessed6, making meaningful comparison difficult. A similar problem prevails for gastropods, exacerbated because most modern assessments are on terrestrial species, and most fossil data come from marine species. Given the daunting challenge of assessing extinction risk in every living species, statistical approaches aimed at understanding what well sampled taxa tell us about extinction risks in poorly sampled taxa are critically important25. For a very few groups, modern assessments are close to adequate. Scleractinian corals, amphibians, birds and mammals have all known species assessed6 (Fig. 2), although species counts remain a moving target27. In these groups, even though the percentage of species extinct in historic time is low (zero to 1\%), 20–43\% of their species and many more of their populations are threatened (Fig. 2). Those numbers suggest that we have not yet seen the sixth mass extinction, but that we would jump from one- quarter to halfway towards it if ‘threatened’ species disappear. Given that many clades are undersampled or unevenly sampled, magnitude estimates that rely on theoretical predictions rather than empirical data become important. Often species-area relationships or allied modelling techniques are used to relate species losses to habitat- area losses (Table 2). These techniques suggest that future species extinctions will be around 21–52\%, similar to the magnitudes expressed Cycadopsida Mammalia Aves Reptilia Amphibia Actinopterygii Scleractinia Gastropoda Bivalvia Coniferopsida Chondrichthyes Decapoda Big Five mass extinctions 0 25 50 75 100 Extinction magnitude (percentage of species) Figure 2 | Extinction magnitudes of IUCN-assessed taxa6 in comparison to the 75\% mass-extinction benchmark. Numbers next to each icon indicate percentage of species. White icons indicate species ‘extinct’ and ‘extinct in the wild’ over the past 500 years. Black icons add currently ‘threatened’ species to those already ‘extinct’ or ‘extinct in the wild’; the amphibian percentage may be as high as 43\% (ref. 19). Yellow icons indicate the Big Five species losses: Cretaceous 1 Devonian, Triassic, Ordovician and Permian (from left to right). Asterisks indicate taxa for which very few species (less than 3\% for gastropods and bivalves) have been assessed; white arrows show where extinction percentages are probably inflated (because species perceived to be in peril are often assessed first). The number of species known or assessed for each of the groups listed is: Mammalia 5,490/5,490; Aves (birds) 10,027/10,027; Reptilia 8,855/1,677; Amphibia 6,285/ 6,285, Actinopterygii 24,000/5,826, Scleractinia (corals) 837/837; Gastropoda 85,000/2,319; Bivalvia 30,000/310, Cycadopsida 307/307; Coniferopsida 618/618; Chondrichthyes 1,044/1,044; and Decapoda 1,867/1,867. 10,000 1,000 100 10 1 0.1 0.01 0 107 106 105 104 103 102 10 1 Time-interval length (years) E /M S Y Cenozoic fossils CR EN VU CR EN VU CR EN VU Pleistocene Extinctions since 2010 Minus bats and endemics Figure 1 | Relationship between extinction rates and the time interval over which the rates were calculated, for mammals. Each small grey datum point represents the E/MSY (extinction per million species-years) calculated from taxon durations recorded in the Paleobiology Database30 (million-year-or- more time bins) or from lists of extant, recently extinct, and Pleistocene species compiled from the literature (100,000-year-and-less time bins)6,32,33,89–97. More than 4,600 data points are plotted and cluster on top of each other. Yellow shading encompasses the ‘normal’ (non-anthropogenic) range of variance in extinction rate that would be expected given different measurement intervals; for more than 100,000 years, it is the same as the 95\% confidence interval, but the fading to the right indicates that the upper boundary of ‘normal’ variance becomes uncertain at short time intervals. The short horizontal lines indicate the empirically determined mean E/MSY for each time bin. Large coloured dots represent the calculated extinction rates since 2010. Red, the end-Pleistocene extinction event. Orange, documented historical extinctions averaged (from right to left) over the last 1, 30, 50, 70, 100, 500, 1,000 and 5,000 years. Blue, attempts to enhance comparability of modern with fossil data by adjusting for extinctions of species with very low fossilization potential (such as those with very small geographic ranges and bats). For these calculations, ‘extinct’ and ‘extinct in the wild’ species that had geographic ranges less than 500 km2 as recorded by the IUCN6, all species restricted to islands of less than 105 km2, and bats were excluded from the counts (under-representation of bats as fossils is indicated by their composing only about 2.5\% of the fossil species count, versus around 20\% of the modern species count30). Brown triangles represent the projections of rates that would result if ‘threatened’ mammals go extinct within 100, 500 or 1,000 years. The lowest triangle (of each vertical set) indicates the rate if only ‘critically endangered’ species were to go extinct (CR), the middle triangle indicates the rate if ‘critically endangered’ 1 ‘endangered’ species were to go extinct (EN), and the highest triangle indicates the rate if ‘critically endangered’ 1 ‘endangered’ 1 ‘vulnerable’ species were to go extinct (VU). To produce Fig. 1 we first determined the last-occurrence records of Cenozoic mammals from the Paleobiology Database30, and the last occurrences of Pleistocene and Holocene mammals from refs 6, 32, 33 and 89–97. We then used R-scripts (written by N.M.) to compute total diversity, number of extinctions, proportional extinction, and E/MSY (and its mean) for time-bins of varying duration. Cenozoic time bins ranged from 25 million to a million years. Pleistocene time bins ranged from 100,000 to 5,000 years, and Holocene time bins from 5,000 years to a year. For Cenozoic data, the mean E/MSY was computed using the average within-bin standing diversity, which was calculated by counting all taxa that cross each 100,000-year boundary within a million-year bin, then averaging those boundary-crossing counts to compute standing diversity for the entire million-year-and-over bin. For modern data, the mean was computed using the total standing diversity in each bin (extinct plus surviving taxa). This method may overestimate the fossil mean extinction rate and underestimate the modern means, so it is a conservative comparison in terms of …