REVIEW SUMMARY

CONSERVATION

Merging paleobiology withconservation biology to guide thefuture of terrestrial ecosystemsAnthony D Barnosky Elizabeth A Hadly Patrick Gonzalez Jason HeadP David Polly A Michelle Lawing Jussi T Eronen David D Ackerlydagger Ken AlexEric Biber Jessica Blois Justin Brashares Gerardo Ceballos Edward DavisGregory P Dietl Rodolfo Dirzo Holly Doremus Mikael Fortelius Harry W GreeneJessica Hellmann Thomas Hickler Stephen T Jackson Melissa Kemp Paul L KochClaire Kremen Emily L Lindsey Cindy Looy Charles R Marshall Chase MendenhallAndreas Mulch Alexis M Mychajliw Carsten Nowak Uma RamakrishnanJan Schnitzler Kashish Das Shrestha Katherine Solari Lynn StegnerM Allison Stegner Nils Chr Stenseth Marvalee H Wake Zhibin Zhang

BACKGROUND The pace and magnitude ofhuman-caused global change has accelerateddramatically over the past 50 years overwhelm-ing the capacity of many ecosystems and spe-cies to maintain themselves as they have underthe more stable conditions that prevailed for atleast 11000 years The next few decades threat-en even more rapid transformations becauseby 2050 the human population is projected togrow by 3 billion while simultaneously increas-ing per capita consumption Thus to avoidlosing many species and the crucial aspects ofecosystems that we needmdashfor both our physicaland emotional well-beingmdashnew conservationparadigms and integration of information fromconservation biology paleobiology and theEarth sciences are required

ADVANCES Rather than attempting to holdecosystems to an idealized conception of thepast as has been the prevailing conservationparadigm until recently maintaining vibrantecosystems for the future now requires newapproaches that use both historical and novelconservation landscapes enhance adaptive ca-pacity for ecosystems and organisms facilitateconnectedness and manage ecosystems forfunctional integrity rather than focusing en-tirely on particular species Scientific break-throughs needed to underpin such a paradigmshift are emerging at the intersection of ecol-ogy and paleobiology revealing (i) which speciesand ecosystems will need human interventionto persist (ii) how to foster population connec-tivity that anticipates rapidly changing climateand land use (iii) functional attributes thatcharacterize ecosystems through thousands tomillions of years irrespective of the species thatare involved and (iv) the range of compositionaland functional variation that ecosystems have

exhibited over their long histories Such infor-mation is necessary for recognizing which cur-rent changes foretell transitions to less robustecological states and which changes may signal

benign ecosystem shifts that will cause no sub-stantial loss of ecosystem function or servicesConservation success will also increasingly

hinge on choosing among different sometimesmutually exclusive approaches to best achieve

three conceptually distinctgoals maximizing biodi-versity maximizing ecosys-tem services and preservingwilderness These goals varyin applicability dependingon whether historical or

novel ecosystems are the conservation targetTradeoffs already occurmdashfor example man-aging to maximize certain ecosystem servicesupon which people depend (such as foodproduction on farm or rangelands) versus main-taining healthy populations of vulnerable species(such as wolves lions or elephants) In the fu-ture the choices will be starker likely involvingdecisions such as which species are candidatesfor managed relocation and to which areas andwhether certain areas should be off limits forintensive management even if it means losingsome species that now live there Developing thecapacity to make those choices will require con-servation in both historical and novel ecosystemsand effective collaboration of scientists govern-mental officials nongovernmental organizationsthe legal community and other stakeholders

OUTLOOK Conservation efforts are currentlyin a state of transition with active debate aboutthe relative importance of preserving historicallandscapes with minimal human impact on oneend of the ideological spectrum versus ma-nipulating novel ecosystems that result fromhuman activities on the other Although thetwo approaches are often presented as dichot-omous in fact they are connected by a con-tinuum of practices and both are needed Inmost landscapes maximizing conservation suc-cess will require more integration of paleo-biology and conservation biology because ina rapidly changing world a long-term perspec-tive (encompassing at least millennia) is necessaryto specify and select appropriate conservationtargets and plans Although adding this long-term perspective will be essential to sustain bio-diversity and all of the facets of nature thathumans need as we continue to rapidly changethe world over the next few decades maximizingthe chances of success will also require dealingwith the root causes of the conservation crisisrapid growth of the human population increasingper capita consumption especially in developedcountries and anthropogenic climate changethat is rapidly pushing habitats outside thebounds experienced by todayrsquos species

RESEARCH

594 10 FEBRUARY 2017 bull VOL 355 ISSUE 6325 sciencemagorg SCIENCE

The list of author affiliations is available in the full article onlineCorresponding author Email tonybarnoskystanfordedudaggerAuthors after the first seven are listed alphabeticallyCite this article as A D Barnosky et al Science 355eaah4787 (2017) DOI 101126scienceaah4787

Fewer than 900 mountain gorillas are leftin the world and their continued existencedepends upon the choices humans makeexemplifying the state of many species andecosystemsCan conservation biology save bio-diversity and all the aspects of nature thatpeople need and value as 3 billion more of usare added to the planet by 2050 while climatecontinues to change to states outside the boundsthat most of todayrsquos ecosystems have everexperienced P

HOTOEAHADLY

ATVOLC

ANOESNATIO

NALPARKRWANDA

ON OUR WEBSITE

Read the full articleat httpdxdoiorg101126scienceaah4787

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

REVIEW

CONSERVATION

Merging paleobiology withconservation biology to guide thefuture of terrestrial ecosystemsAnthony D Barnosky12 Elizabeth A Hadly3 Patrick Gonzalez45 Jason Head6

P David Polly7 A Michelle Lawing8 Jussi T Eronen910 David D Ackerly11daggerKen Alex12 Eric Biber13 Jessica Blois14 Justin Brashares15 Gerardo Ceballos16

Edward Davis17 Gregory P Dietl1819 Rodolfo Dirzo20 Holly Doremus21

Mikael Fortelius2223 Harry W Greene24 Jessica Hellmann25 Thomas Hickler26

Stephen T Jackson27 Melissa Kemp28 Paul L Koch29 Claire Kremen30

Emily L Lindsey31Dagger Cindy Looy32 Charles R Marshall33 Chase Mendenhall3435

Andreas Mulch3637 Alexis M Mychajliw38 Carsten Nowak39 Uma Ramakrishnan40

Jan Schnitzler4142 Kashish Das Shrestha43 Katherine Solari44 Lynn Stegner45

M Allison Stegner46sect Nils Chr Stenseth47 Marvalee H Wake48 Zhibin Zhang49

Conservation of species and ecosystems is increasingly difficult because anthropogenicimpacts are pervasive and accelerating Under this rapid global change maximizingconservation success requires a paradigm shift from maintaining ecosystems in idealizedpast states toward facilitating their adaptive and functional capacities even as speciesebb and flow individually Developing effective strategies under this new paradigm willrequire deeper understanding of the long-term dynamics that govern ecosystempersistence and reconciliation of conflicts among approaches to conserving historicalversus novel ecosystems Integrating emerging information from conservation biologypaleobiology and the Earth sciences is an important step forward on the path to successMaintaining nature in all its aspects will also entail immediately addressing the overarchingthreats of growing human population overconsumption pollution and climate change

Local and global stress on ecosystems spe-cies and populationsmdashalready severemdashwillintensify in the near future as the numberof people land use and consumption ofnatural resources increase and as anthro-

pogenic climate change continues (1ndash4) Even

now it is no longer possible to effectively man-age most ecosystems to maintain them in a his-torical state as has been prevailing practice andtheory in conservation biology (4ndash6) Rather aparadigm shift is under way with new conser-vation goals aiming to maximize the capacity of

ecosystems to adapt to current and impendingchanges (4ndash11)Recent work highlights that achieving such

goals will require understanding how ecologicaldynamics play out over time scales much longerthan a human lifetime (12) Such information haslong been known to be available from historicalpaleobiological and geological records (10 13)However scientists and land managers are stillgrappling with how to more fully integrate paleo-biological information into conservation theoryand decisions (Fig 1)

The current conservation landscape

More than half of Earthrsquos ice-free land has beenconverted for human use (14) tens of thousandsof species have been transported around the globe(15) and species and populations are going ex-tinct at highly elevated rates (16) As a resultmuch if not most of the planet is now coveredby novel ecosystems (Fig 1D) (6)mdashthat is thosewith assemblages of species or other character-istics that did not exist before preindustrial timeswhich human activities have created either in-tentionally or inadvertently Here we include asnovel ecosystems cropland pastureland timberplantations and land modified by logging andhuman-caused erosion and sedimentation in totalcovering ~47 of ice-free land (14) Urban andrural communities roads reservoirs railways andmining areas account for an additional ~7 (14)Historical ecosystems are defined loosely as

those presumed to be operating as they have forat least centuries (Fig 1B) (6) They include somenational parks and other large tracts of land thatare often protected to some degree and are atmost lightly inhabited by people Although allhistorical ecosystems exhibit signs of humanmodification ranging from millennia of manip-ulation by indigenous people to present-day an-thropogenic climate change their dynamics andspecies compositions seem to closely resemblethose present for at least centuries and to bemore heavily influenced by nonanthropogenic

RESEARCH

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 1 of 10

1Jasper Ridge Biological Preserve Stanford University Stanford CA 94305 USA 2Department of Integrative Biology and Museums of Paleontology and Vertebrate Zoology University of California Berkeley BerkeleyCA 94720 USA 3Department of Biology Department of Geological Sciences Woods Institute for the Environment Center for Innovations in Global Health Jasper Ridge Biological Preserve Stanford University Stanford CA94303 USA 4Natural Resource Stewardship and Science US National Park Service Berkeley CA 94720 USA 5Department of Environmental Science Policy and Management University of California Berkeley CA94720 USA 6Department of Zoology University of Cambridge Downing Street Cambridge CB2 3EJ UK 7Department of Geological Sciences Indiana University Bloomington IN 47405 USA 8Department of EcosystemScience and Management Texas AampM University College Station TX 77845 USA 9Senckenberg Biodiversity and Climate Research Centre (BiK-F) Senckenberganlage 25 60325 Frankfurt Germany 10Department ofGeosciences and Geography University of Helsinki Post Office Box 64 FIN-00014 Finland and BIOS Research Unit Kalliolanrinne 4 Helsinki Finland 11Department of Integrative Biology University and Jepson HerbariaUniversity of California Berkeley CA 94720 USA 12California Governorrsquos Office of Planning and Research Post Office Box 3044 Sacramento CA 95812-3044 USA 13School of Law University of California Berkeley CA94720 USA 14School of Natural Sciences University of California Merced CA 95343 USA 15Department of Environmental Science Policy and Management University of California Berkeley CA 94720 USA 16Instituto deEcologia Universidad Nacional Autonoma de Mexico CU Ap Post 70-275 Mexico DF 04510 Mexico 17Department of Geological Sciences University of Oregon Eugene OR 97403 USA 18Paleontological ResearchInstitution 1259 Trumansburg Road Ithaca NY 14850 USA 19Department of Earth and Atmospheric Sciences Cornell University Ithaca NY 14853 USA 20Department of Biology Stanford University Stanford CA 94303USA 21School of Law University of California Berkeley Berkeley CA 94720 USA 22Department of Geosciences and Geography Post Office Box 64 FI-00014 University of Helsinki Finland 23Centre for Ecological andEvolutionary Synthesis University of Oslo Post Office Box 1066 Blindern NO-0316 Oslo Norway 24Department of Ecology and Evolutionary Biology Cornell University Ithaca NY 14853 USA 25Institute of the EnvironmentUniversity of Minnesota Minneapolis MN 55455 USA 26Senckenberg Biodiversity and Climate Research Centre (BiK-F) and Institute of Physical Geography Goethe-University Frankfurt Senckenberganlage 25 D-60325Frankfurt am Main Germany 27US Geological Survey Department of Interior Southwest Climate Science Center and Department of Geosciences University of Arizona 1064 E Lowell Street Tucson AZ 85721 USA28Center for the Environment Harvard University Cambridge MA 02138 USA 29Department of Earth and Planetary Sciences University of California Santa Cruz CA 95064 USA 30Department of Environmental SciencePolicy and Management University of California Berkeley Berkeley CA 9472 USA 31Department of Integrative Biology and Museum of Paleontology University of California Berkeley Berkeley CA 94720 USA32Department of Integrative Biology Museum of Paleontology University and Jepson Herbaria University of California Berkeley Berkeley CA 94720 USA 33Department of Integrative Biology and Museum of PaleontologyUniversity of California Berkeley Berkeley CA 94720 USA 34Center for Conservation Biology Stanford University Stanford CA 94305 USA 35The Nature Conservancy Arlington VA 22203 USA 36SenckenbergBiodiversity and Climate Research Centre (BiK-F) 60325 Frankfurt am Main Germany 37Institute of Geosciences Goethe-University Frankfurt 60438 Frankfurt am Main Germany 38Department of Biology StanfordUniversity Stanford CA 94303 USA 39Senckenberg Research Institute and Natural History Museum Frankfurt Clamecystrasse 12 D-63571 Gelnhausen Germany 40National Centre for Biological Sciences Tata Institute ofFundamental Research GKVK Bellary Road Bangalore 560065 India 41Institute of Biology Leipzig University Johannisallee 21-23 04103 Leipzig Germany 42Senckenberg Biodiversity and Climate Research Centre (BiK-F)Senckenberganlage 25 60325 Frankfurt am Main Germany 43City Museum Kathmandu 44600 Nepal 44Department of Biology Stanford University Stanford CA 94303 USA 45Continuing Studies Program StanfordUniversity Stanford CA 94305 USA 46Department of Integrative Biology and Museum of Paleontology University of California Berkeley CA 94720 USA 47Centre for Ecological and Evolutionary Synthesis (CEES)Department of Biosciences University of Oslo Post Office Box 1066 Blindern NO-0316 Oslo Norway 48Department of Integrative Biology and Museum of Vertebrate Zoology University of California Berkeley CA 94720USA 49State Key Laboratory of Integrated Management of Insects and Rodents Institute of Zoology Chinese Academy of Sciences 1 Beichen West Road Chaoyang District Beijing 100101 PR ChinaCorresponding author Email tonybarnoskystanfordedu daggerAuthors after the first seven are listed alphabetically DaggerPresent address La Brea Tar Pits and Museum 5801 Wilshire Boulevard Los Angeles CA 90036 USAsectPresent address Department of Zoology University of Wisconsin 430 Lincoln Drive Madison WI 53706 USA

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

processes than by humans Here we regard his-torical ecosystems as those that still have at least70 of the habitats that were present 500 yearsago and that contain fewer than 5 peoplekm2

(17) A subset of historical ecosystems retains atleast ~90 of habitats that have characterizedthem over at least the past five centuries andcontain lt1 personkm2 (17) We refer to theseplaces as wilderness such wilderness comprises~26 of ice-free land (17)Conservation efforts typically target at least one

of three key goals maximizing biodiversity main-taining ecological structure and function andormaximizing ecosystem services In attempts toconserve historical ecosystems achieving all threegoals simultaneouslymdashwith the approach typicallybeing to minimize human impactsmdashis usuallyimplicitly or explicitly intended and can be verysuccessful A case in point is preserving intacttropical forests which support the majority ofEarthrsquos terrestrial species retain a general eco-logical structure that has persisted for at leasttens of thousands of years (even considering thelong integration of indigenous peoples into suchsystems) and as a result offer myriad ecosystemservices ranging from carbon sequestration topurifying water and air and provisioning foodto providing diverse aesthetic emotional andwilderness experiences for peopleIn contrast in novel ecosystems particularly

under conditions of rapid global change theconservation goals of maximizing biodiversitymaintaining particular ecological structures andfunctions and provision of particular ecosystemservices can diverge substantially (6 18) althoughin most cases the three goals exhibit at leastsome overlap (Fig 1E) For instance rescuing aspecies threatened by climate change or habitatloss may require manipulating its genetic diver-sity (Fig 1 example E1) (19) andor managed re-location into a geographic region and ecosystemin which the species has never lived (Fig 1 ex-ample E2) (20) Although effective in conservingbiodiversity and perhaps promoting certain eco-system services such as tourism such actionscontradict other goalsmdashfor example the philos-ophy of maximizing wilderness attributes (21)Less nuanced examples include zoos botanic gar-dens and well-designed urban landscapes whichwill be essential in maximizing biodiversity butwhich completely replace wilderness At anotherend of the spectrum are agricultural landscapeswhich maximize a necessary ecosystem service(food production) and can either severely de-press biodiversity relative to pre-anthropogenicconditions as in the case of monoculture farmingor help maintain biodiversity if designed to doso as in some coffee farms in Costa Rica (Fig 1example E3) (22 23) Thus in novel landscapesdeciding which goal to optimize can engendermuch debate among stakeholders as can the verydecision about whether a given landscape shouldbe regarded as historical or novel (Fig 1A)Such concerns lead to conflicting opinions

about the relative importance of preserving largelandscapes with minimal human impact (24ndash26)versus designing human-dominated ecosystems

in ways that maximize particular conservationgoals (27) In practice the two schools of thoughtboth recognize a gradient of human impactsthat conservation only works when human valuesare articulated by multiple stakeholders to guideany given effort and that threatened species andspecial landscapes and ecosystems should bepreserved (18 28 29) Nevertheless the differentviewpoints have led to developing and implement-ing a diverse menu of conservation approaches(18) some of which appear contradictory or evenmutually exclusive (Table 1) Yet when put in the

context of making choices (Fig 1) that recognizeboth historical and novel ecosystems and rapidglobal change most have a place

Critical information from paleobiology

Under current global change successful conser-vation outcomes for most approaches (Table 1)depend onmeaningful comparisons betweenmod-ern conditions and long-term histories The emerg-ing discipline of conservation paleobiology (10 30)is supplying necessary data insights and tech-niques through (i) specifying long-term fluctuating

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 2 of 10

Fig 1 Critical conservation decisions (A to E) Decision points at which data from paleobiology areessential Here ldquobiodiversityrdquo refers to all levels of the biological hierarchy Novel ecosystems can serveto maximize biodiversity particular ecosystem services or ecosystem structure which can be mutuallyexclusive or overlap to varying degrees The overlap of these conservation goals is typically broad inhistorical systems For conserving historical ecosystems taxon-based methods from paleobiology have al-ready proven essential and taxon-free methods also are useful For novel ecosystems taxon-free me-thods hold much potential for linking past present and future to help formulate effective conservationprograms

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

baselines required to understand how ecosystemscommunities species populations and geneticstructure vary naturally through time and spaceand how they respond to major perturbations(10 31ndash36) (ii) identifying scalable taxon-freemetrics that allow attributes of ecosystems tobe tracked over seasons decades centuries mil-lennia and millions of years (10 37) (iii) dem-

onstrating biotic outcomes of many ldquonaturalexperimentsrdquo in global change (30) ranging frommajor extinctions (10 16 29 38) to rapid climatechange (10 39 40) to species invasions (10 41)(iv) testing and refining models of biotic responseto future environmental change (10 35 42) and(v) tracking the long-term dynamics of ecosystemsin ways relevant to assessing continued potential

for ecosystem services (12 43 44) and early warn-ing signs of ecological state-shifts (10 45)The kinds of fossils that have provided such

insights most commonly include phytoplanktonand zooplankton many kinds of plants (repre-sented by fossil pollen seeds leaves and wood)invertebrate animals with hard parts (such asmollusks) and vertebrate animals (representedby bones and teeth) (Fig 2) Depositional envi-ronments in which these taxa commonly are fos-silized include lakes river valleys rock sheltersand caves The fossil samples recovered fromsuch deposits are particularly useful because theyoften show high fidelity to the living commun-ities from which they were drawn in terms oftaxonomic composition and relative abundanceas demonstrated by taphonomic studies that as-sess how the same sampling vectors that buildfossil assemblages sample modern communities(10 46ndash49) Some of these favorable depositionalenvironments occur in most biomes openingpossibilities to use paleobiological data in manyconservation settings therefore it should not beassumed that paleontological information is notavailable if there has not been targeted explora-tion for appropriate fossil sitesLess available for conservation uses are species

restricted to areas where fossilization potential islow (such as upland areas lacking lakes riverscaves or rock shelters) those whose taphonomyis not well understood (10 50) and those whosebody parts are too fragile to fossilize on a regularbasis such as most insects and birds Even sosome taxa with low fossilization potential areinformative for conservation efforts when theyare present for instance beetles found in lake sed-iments and peat illustrate a response to climatechange that differs from the mammalian response(51) and bones of California condors have beencritical in identifying their pre-anthropogenicdiet (10) In general Quaternary fossils (Pleistoceneand Holocene) have proven especially informa-tive for addressing conservation questions butuseful information has also come from mucholder fossil deposits reaching back millions ofyears (34 37)

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 3 of 10

Table 1 Multiple conservation approaches

Approach Examples

Address root causes of conservation crisis through reducing the human footprint (96)

Increase the number of national parks and other protected areas and connections between them (26)

Quantify ecosystem services regionally and globally and value them in economic terms (97)

Abandon the idea of ldquopristine wildernessrdquo and maintain high biodiversity

(at genetic population and species levels) in cultural landscapes(27)

Triage species interventions according to importance specificity or likelihood of success (98)

Relocate species populations or genotypes whose habitats are disappearing in one place

but emerging in another especially where corridors are lacking(8 20)

Reconstruct or restore damaged or extinct ecosystems (99)

Manipulate populations and genetics of endangered species to enhance their survival (19 100)

Create ecosystems that simulate long-past conditions such as Pleistocene rewilding (65)

Produce facsimiles of extinct species through emerging techniques in molecular biology

such as ldquode-extinctionrdquo (which is not considered a viable conservation strategy in this Review)(29 75 76)

Fig 2 Tracking community fluctuations through millennia (A) Deposits accumulated by wood rats(Neotoma spp) as they drag bone-laden carnivore scat and raptor pellets into their middens are par-ticularly useful in sampling the vertebrate and plant community with high fidelity to taxonomiccomposition and relative abundance (46ndash48) Such records can provide successive snapshots of taxonpresence absence and abundance through thousands to millions (33) of years and genetic variation (73)through thousands of yearsWood rat middens occur through much of North America species in generaother than Neotoma construct paleoecologically useful middens elsewhere (B) Bones excavated froma wood rat midden after being concentrated by screening and sorting (C D and E) Sediment coresfrom lakes contain fossil pollen and spores that allow reconstruction of vegetation changes throughhundreds to hundreds of thousands of years (10)

PHOTO(A

)AND

(B)ANTHONYDB

ARNOSKY

(C)AND

(D)CIN

DYLO

OY

(E)STEPHEN

TJA

CKSON

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

The analytical methods that allow comparingpresent with past fall into two main categoriestaxon-based and taxon-free Taxon-based methodsare those that rely on the presence absence orabundances of certain taxa and their underlyingdiversity including genetic or phylogenetic dif-ferentiation Taxon-free methods use metrics thatreflect ecosystem function rather than structureexamples used in modern systems include assess-ing ecological network structure (52) and measur-ing biomass (53) functional traits (54) nutrientflow (43) net primary productivity (55) or eco-system services (7 43) Quantifying taxon-basedand taxon-free attributes in the present alonedoes not capture the full range of conditionsunder which a given ecosystem can thrive be-cause most ecosystems (at least those that havenot been created by humans) have persisted forthousands to millions of years The taxon-basedapproach to link conservation paleobiology andbiology has been used most so far but currentefforts to develop and use taxon-free methodshold considerable promise Depending on theavailability of fossils and the type of conservationquestion being asked one or the other approachmay be more appropriate

Historical or novel

Taxon-based paleontological data are critical indeciding whether a given so-called ldquonaturalrdquo land-scape represents a historical or a novel ecosystem(Fig 1A) These methods rely on direct com-parisons of the taxa that occupied a region inthe past to those living there presently By usingsuperposed taphonomically understood andwell-dated fossil samples (Fig 2) to reconstructsuccessive snapshots of the past it is possibleto outline the range of taxonomic and relative-

abundance variation that characterizes ecosys-tems as they fluctuate over thousands to tensof thousands of years sometimes much moreFor a modern ecosystem to be considered his-

torical (Fig 1B) its taxon assemblage and theirabundances should fall within the range of pastmillennial-scale variation For example in theworldrsquos first national park Yellowstone NationalPark USA paleontological data influenced crit-ical management decisions by demonstrating thatYellowstonepreservesahistoricalecosystemFossildepositsverifiedthattheareaproposedin1995 forthe reintroduction of the gray wolf (Canis lupus)had indeed harbored wolves (up to their extirpa-tioninthe20thcentury) formorethan3000yearsthat a principle prey speciesmdashelkmdashused the areafor calving thousands of years ago as they still dotoday and that almost all of the mammal spe-cies that had occupied the region for millenniaare still present (Fig 1 example C1) (13) In thiscase the conservation question arose first Werewolves and elk native to the region before thepark was established Exploration for the requi-sitefossilsitespreviouslyunknownensuedaspartof the data-gathering exercise beforemanagementactions Further verification that Yellowstone stillrepresents ahistoric ecosystemcame fromassess-ing impacts of climate change on small mam-mals Ancient DNA obtained from fossil rodentsconfirmed that although genetic diversity popula-tion sizes and gene flow had fluctuated throughtime in response to climatic conditions genotypesin the park now have been there for millennia(56) From the botanical perspective palynologi-cal (fossil pollen) records show that the currentvegetation has persistedwith only minor fluctu-ations in abundance of dominant taxa for atleast 8000 years (57)

Assessing whether historical ecosystemscan be maintainedA critical question for many historical ecosystemsis whether they will be able to persist in the samestates in which they have existed for thousands ofyears given rapid intensifying environmentalchanges (Fig 1C) The taxon-based approachessummarized above provide useful answers throughestablishing the range of variation that taxon-based attributes exhibit through the perturbationsan ecosystem experiences over thousands of yearsThe nature and magnitude of the perturbationscan be assessed from the contemporaneous geo-logic recordmdashfor example isotopic proxies fortemperature lake-level or tree-ring analyses forprecipitation or charcoal records in alluvial de-posits and lake cores to track fire frequencyModern-day changes in the taxon-based metricor in the suspected perturbing agent (for instanceclimate change) that exceed the variation evidentthrough millennia or longer may warn that ashift to a new ecological state is imminent Insuch cases the management choice is to eitherattempt to hold the system to historical condi-tions which would require ever-more intensiveinterventions and may be impossible or managethe system for ldquoadaptive capacityrdquo (4) Adaptivecapacity in this context is the ability of an eco-system to ldquore-configure without significant changesin crucial functions or declines in ecosystemservicesrdquo (58) Put another way adaptive capac-ity is the ability of an ecosystem to avoid col-lapse as it makes the critical transition from itshistorical ecological state to a new state and to beas resilient in its new state as it was in its pre-vious stateRobustly assessing adaptive capacity requires

combining informationaboutprehistoric conditions

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 4 of 10

Fig 3Transformation of historical ecosystems (A) Near Parque Nacionalde Anavilhanas Brazil Palynological data revealed that increased precipita-tion coincided with southern expansion of Amazon rainforest ~3000 yearsago to its early 20th-century position (85) Reduced rainfall across two thirdsof the rainforest related to changes in the El NintildeondashSouthern Oscillation (86)combined with deforestation (87) have recently contributed to shifting therainforest-savanna border northward reducing vegetative productivity (88)increasing fire frequency and lengthening the fire season (89) Climate modelssuggest that future transformation of vast areas of Amazon rainforest may be

imminent (90 91) (B) Las Conchas Fire near Bandelier National MonumentNew Mexico USA Paleontological data show that western North Americanforest ecosystems depend on high-frequency low-severity fires and that wildfirefrequencies increased with drought (92 93) Fuel buildup from fire suppressionand anthropogenic climate change increased fire frequencies in the late 20thcentury (94) and climate projections suggest that by 2100 fire frequency mayincrease to levels far above those to which historical and current vegetation haveadapted (91)These considerations have been used to justify prescribed burningto preempt catastrophic crown fires (95)P

HOTO(A

)PATRIC

KGONZALEZ(B

)KARIGREER

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

modern and historic observations and futureprojections An illustrative example is the con-servation of Joshua trees (Yucca brevifolia) inand around Joshua Tree National Park California(Fig 1 example C2) The distribution of Joshuatree fossils preserved in wood rat middens (Fig 2)and in the dung of extinct Shasta ground sloths(Nothrotheriops shastensis) demonstrated the sen-sitivity of the trees to increased temperature andaridity ~11700 years ago and also revealed thatdispersal of the species occurred only slowly ~1to 2 myear (59) Today Joshua trees are onlyrarely reproducing within the park because ofincreased temperatures and drought and climateprojections under a medium greenhouse gas emis-sions scenario indicate that 90 of their suitablehabitat in the park could be lost by 2100 (59)Climatically suitable areas may shift northwardand upslope but the slowmigration rate of Joshuatrees limits their ability to track their suitableclimate space especially because one of their dis-persal agentsmdashShasta ground slothsmdashis extinct Inthis case the fossil information implies that con-serving Joshua Tree National Park in its present

ecological state may not be possible because ofclimate-triggered species turnover including lossof its namesake species and colonization by cur-rently exotic species The implication is that con-serving Joshua tree ecosystems may require moreactivemanagement in protected areas outside thenationalpark acquisitionofnew lands andperhapstargeted planting This landscape-scale manage-ment of Joshua trees could nurture the adaptivecapacity of the species across its range even ifthe national park loses its suitable habitat whilestillmaintaining thewilderness character of JoshuaTreeNational Park ifwilderness character dependsmoreon the low level of local human impacts thanon the presence of Joshua treesPaleontological data show that other histori-

cal ecosystems have already begun to change andare vulnerable to future change including fire-dependent ecosystems across western North Amer-ica and Amazon tropical rain forests the latterof which can rapidly shift to savanna (Fig 3) Com-parisons of species dispersal rates to past andpresent velocity of climate changemdashthe distanceper unit of time that a species would need to

move to remain in its current conditions of tem-perature or other climate variables (60)mdashhavealso proven useful in determining areas and spe-cies most at risk of ecological transformationsIn general such studies reveal that the speed atwhich species will need to move over the comingdecades far outstrips the pace at which they ac-tually did move in response to the most rapidclimate change documented in the fossil recordthe transition from the last glacial period into theHolocene (60) Such information indicates thatthe need to manage historical ecosystems for theiradaptive capacity will increase mandating increaseduse of taxon-free metrics to assess how well thatmanagement is proceeding

Conservation in novel ecosystems

Landscapes that already fall outside historicalnorms (Fig 1D)mdashwhich in many cases is evidenteven without consideration of fossil data (6 14)mdashmay be candidates for restoration to a desiredhistorical state if the kinds of data described aboveverify that the system has not been pushed ir-reparably beyond its millennial-scale variation in

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 5 of 10

Fig 4 Ecometrics in paleobiology (A) Proportions of certain bones arelinked to land cover land use and topography through locomotor performance(B) In mammalian carnivore communities locomotor diversity can be measuredby using the in- and out-levers of the limbs and is linked to vegetation cover(68 70) in snakes the same relationship can be measured with the ratio of tailto body length (67) (C) Changes in the variance and mean of these traits canbe assessed for congruence with changes in community composition and landcover For example when land acquired by the University of Kansas wasallowed to revert from agricultural grassland to forest between 1947 and

2006 turnover in the herpetofauna changed the mean (black line) andstandard deviation (gray bar) of tail-to-body length ratios (D) The changeillustrated in (C) was congruent with ecometric values associated with grass-land and forest ecosystems elsewhere (67) (E) Conversely 19th-century de-forestation of Indiana extirpated many large mammalian carnivores resultingin a loss of locomotor diversity measured as the standard deviation (graybars) of the out-to-in-lever ratio (F) The loss of locomotor diversity can bemapped to identify other regions (dark gray shading) that may have beensimilarly affected (68)

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

taxonomic composition and abundance (10) Thismay be the case for example if the current sys-tem can be returned to its long-term state byreintroducing key taxa or genotypes Where res-toration is not possible or desirable (Fig 1E)mdashthe situation for perhaps most of the ecosystemson the half of the planet that humans have trans-formed and that are changing even more undercurrent anthropogenic pressuresmdashnovel ecosys-tems also provide many conservation opportunities(6) By definition novel ecosystems are distinctwith respect to past ecosystems in terms of tax-onomic composition thus in these cases taxon-free instead of taxon-based approaches providethe most effective applications of paleontologicaldata to inform conservation strategiesFor example paleontological analyses have

shown that certain body-mass distributions (34)biomass patterns (29) numbers of species withintrophic and size categories (33 34) abundancepatterns (61) and ecological networks (62 63)are characteristic of mammal communities thatpersist for thousands to millions of years irre-spective of the constituent species This knowl-edge can answer critical questions which aboundabout the design and long-term viability of novelecosystems by considering the constituent speciesprimarily in terms of ecological function ratherthan taxonomic identity In urban settings forexample do domestic cats carry out the functionof extirpated or extinct meso-predators keepingrodent and bird populations in check whichwould indicate healthy ecological function oris their impact greater than previously presentmeso-predators which might degrade ecosystemhealth In ranchlands is the biomass of live-stock within the bounds of long-term megafaunavariation which once included mammoths andother extinct large mammals or is the biomassof livestock presently greater In managed relo-cation experiments how will the transferredspecies affect trophic structure and ecologicalnetworks of the target ecosystems And in re-wilding initiativesmdashwhich can range from re-placing ldquomissingrdquo taxa with the same species[for example wolves in Yellowstone and the Re-wilding Europe effort (64)] to building ecosys-tems from scratch by using functional analogs ofextinct species (65)mdashwhat trophic structures andecological networks will maximize biodiversityand ecosystem services and yield a system thatis functionally robust to perturbations thus keep-ing maintenance costs at a minimumBecause taxon-free metrics can often be related

to environmental parameters with statistical sig-nificance they offer opportunities for under-standing which kinds of species are likely to thrivein which regions as biota adjust to rapidly chang-ing environmental conditions (Fig 4) Such mea-sures have been applied in modern communityecology (54 66) In conservation paleobiologythey have been called ldquoecometricsrdquo (37 67ndash69)and include studies of both plants and animalswith a focus on functional traits that are frequent-ly preserved in the fossil record For plants thisincludes leaf size and shape (reflects precipitationpatterns) stomatal index (measures equilibrium

with atmospheric carbon dioxide) and phytolithshape (a proxy for resistance to herbivore useand whether or not the leaf wax hardened in asunny or shady environment) Animal-based traitsinclude dental morphology (which is a proxy fordiet) locomotor attributes (which show distinctdifferences in different environments) (Fig 4)and body size (which can reflect climate varia-bles and nutrition) By focusing on such traitsit becomes possible to assess the ability of taxato persist in particular places under particularscenarios of rapid environmental change Thisin turn helps in identifying suitable candidatesand locations for managed relocation restorationand rewilding programs (Fig 1 example E2)For example in mammalian carnivore com-

munities locomotor diversity is known to belinked to vegetation cover (68 70) which pro-vides a valuable predictor of which carnivore spe-cies will be best suited to areas where climatechange or other human impacts substantiallyalter plant communities and also a metric bywhich to identify ecologically impoverished sys-tems (Fig 4) The application of such techniquesrequires that the linkage between a given trait

and environmental parameter be firmly estab-lished which so far has only been done for rel-atively few traits especially in vertebrate animalsFuture research that expanded the suite of usefultraits would be valuableTaxon-free paleontological measures can also

reveal whether the potential for delivery of eco-system services is being sustained in novel eco-systems by tracking metrics that reflect ecologicalprocesses over centennial to millennial time scalessuch as nutrient cycling biomass crop productionwater supply climate regulation timber and coast-al protection (43) Geologically based proxies cantrack nutrient cycling soil formation and stabi-lization and erosion (43) As an example a suiteof 50 paleoenvironmental proxies demonstratedthat since the year 1800 rapid economic growthand population increases since the mid-20th cen-tury coincided with environmental degradationin the lower Yangtze Basin China (44)Last taxon-free paleontological data are crit-

ical for understanding whether certain ecosystemsare approaching ecological thresholds (10 45)mdashso-called ldquotipping pointsrdquo as demonstrated by anal-ysis of diatoms pollen and sediments from lake

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 6 of 10

Fig 5 Pressures affecting wild tigers (A) Only ~3800 wild tigers remain confined to only 7 (darkgreen) of their historic (light green) geographic range (the white arrow shows the region of RanthamboreNational Park) (B) Present geographic range of wild tigers (white outline) overlain on a map of crop andpasture lands (darker shades of purple indicate more intensive agricultural use) and on (C) (red outline)a map of human population density where darker blues indicate higher density ranging from le1 personkm2

in the lightest colored regions to gt10000 peoplekm2 in the darkest (D) Tigers remain mainly in the leastdensely populated areas or in reserves located in novel ecosystems such as this individual in RanthamboreNational Park India By 2050 at least one half billion more people are projected to populate regions thatinclude tiger reservesP

HOTO(D

)ABHIN

AVTYA

GI

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

cores which identified a match between math-ematical models and an ecological state-shift inYunnan China Another example comes fromPennsylvania USA where feedbacks that causeddeforestation in one area triggered an ecologicalstate-shift in an adjacent area (10)The utility of such taxon-free approaches for

conservation paleobiology and predictive ecol-ogy has been demonstrated over the past dec-ade by many case studies (71) A challenge goingforward will be to develop a coherent theoreticalframework that takes into account such impor-tant relationships as the underlying trait distri-bution performance filters that define trait fitnessin varying environments how traits will performas environments change (71) spatial and temporalscaling and demography and inter- and intra-specific variability in trait distribution and per-formance (72)

Emerging conservation applicationsfor paleobiologyConservation genetics

Conservation genetics is now being enhancedthrough studies of ancient DNA (56 73) Besidesestablishing the long-term range of genetic di-versity population fluctuation and gene flow asnoted above for Yellowstone rodents (Fig 1 ex-ample E1) (73) paleontological studies also haveresulted in new methods applicable to contem-porary conservation problems notably coalescentsimulation analysis This technique was devel-oped to understand the relative contributionsof gene flow and population size in explainingobserved fluctuations in genetic diversity chron-icled in ancient DNA (73 74) but is now informingconservation strategies for presently threatenedspecies A case in point is one of the worldrsquos iconicmammals tigers (Panthera tigris) (Fig 5) Mosttigers live in zoos and other captive situationsonly ~3800 remain in the wild and many ofthose are confined to novel ecosystems such asRanthambore National Park which has beenheavily used by humans for more than a thousandyears Such small reserves can support just a fewindividuals which has led to dwindling geneticdiversity within populations It has been unclearwhether such bottlenecks presage extinction oftigers even in the few remaining habitats setaside for them Coalescent simulation analysesused to forecast into the future instead of inter-preting the past indicate that without substantialgene flow between reserves reduced diversitywill likely imperil tigers by the next century butthat diversity can be maintained and perhapseven enhanced by aggressively maintaining func-tional connectivity physically moving individualsand prioritizing breeding among reserves world-wide (19) Global conservation efforts thus farhowever tend to prioritize tiger numbers overconnectivity or focus on maintaining the ldquopurityrdquoof the genetic composition of tiger subspeciesFossils have also figured prominently in ex-

perimentation with so-called ldquode-extinctionrdquo (75)mdashefforts to reconstruct facsimiles of species that hu-mans have driven to extinction either recently (pas-senger pigeons) or in the deeper past (mammoths)

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 7 of 10

Canada

national parks

provincial parks

USA

national parks

national forests

wildlife refuges

Bureau of Land Management

Department of Defense

state parks

Mexico

national parks

other federal protected

state protected

Private Initiative

Yellowstone to Yukon

bird sanctuarieswildlife areas

0 500

Km

CANADA

USA

MEXICO

USA

USA

Fig 6 The importance of conservation corridors Climate change and jurisdictional differenceschallenge corridor design The Yellowstone to Yukon Conservation Initiative (purple outline) spans eco-systems rapidly transforming from increasing wildfire frequency and forest mortality both triggered byglobal climate change and two nations where private land confers varying property rights and federalprotected areas are managed by different government agencies Although multiple jurisdictions com-plicate enhancing connectivity such diversity can also contribute to success when the goals of maximizingbiodiversity ecosystem services and preserving wilderness come into conflict because each stakeholdermay choose to optimize a different goal while still contributing to the overall effect of providing a piece ofthe corridor

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

Although such efforts may eventually create sci-entific curiosities their conservation applicationsare at best limited (29) given that (i) the createdgenomes would be mostly composed of the basepairs of the nearest living relatives of the extinctspecies (ii) epigenetic effects are not yet wellunderstood (iii) only a few individuals of a givenspecies could be engineered because the processis both time-consuming (because of gestationtimes) and very expensive (iv) imparting thelearned behavior that offspring gain from parentalteaching would be impossible because that knowl-edge went extinct with the lost species (v) theecosystems that supported many extinct speciesno longer exist so survival outside of captivitywould be difficult or impossible and (vi) prevent-ing the extinction of extant species and habitatsnumbering in the thousands already is challeng-ing so the prospects of sustaining ldquode-extinctedrdquospecies are poor at best Genetic engineering tosimulate extinct life also raises ethical and legalconcerns for many (76)

Invasive species

Whether invasive species substantially alter eco-logical structure and function is a critical conser-vation question that can only be answered witha paleontological perspective For instance inCalifornia grassland ecosystems historic cattleintroduction transformed historic ecosystems intonovel ones as cattle populations grew grazingmegafauna biomass rose far above prehistoriclevels precipitating a functional shift in grazingpressure that favored replacement of native an-nual grasses by invasive species (Fig 1 exampleE4) (77) The fossil record also can help informcontroversial management decisions (41) suchas whether wild horses on western North Amer-ican ranch lands are invasive because they havebeen absent for most of the Holocene or nativebecause they evolved in those regions and werefor millions of years an integral component ofthe ecosystems in which they are now thriving

Enhancing connectivity

Corridors designed to connect protected areassuch as the Yellowstone to Yukon ConservationInitiative (Fig 6) are critical today (4 78) andwill be become even more so in the near futurebecause one tenth to one half of global terres-trial area is highly vulnerable to biome shifts in the21st century (79) whereas refugia in existing pro-tected areas cover only 1 to 2 of global land (78)Therefore a new perspective is that effective cor-ridor design (Fig 1 example E5) besides takinginto account present land-use will need to iden-tify key areas that have served as refugia in pre-history (80) and anticipate ecological changes thatwill inevitably take place as climate changes (81)Anticipating the future efficacy of corridors

generally uses species distribution modeling(82) Most species distribution models rely onmatching present or near-historic occurrencesof a given species with nearby climatic param-eters to estimate the ecological niche Recentwork that uses the same models combined withpaleontologic geologic and paleoclimatic data

to hindcast where species could have occurredover the past several thousand years (35 42) re-veals that in many cases existing models do notadequately project where species may move inthe future In addition incorporating prehistoricdistributional information helps quantify the prob-ability of errors (10) Using the fossil record torefine species distribution models requires pa-rameterizing the climate models with appropriateboundary conditions as well as adequate datingcontrol which is now routinely achievable withaccelerator mass spectrometry radiocarbon datesthat place the age of critical fossils within decadesThe paleobiological approach can further im-

prove species distribution models by incorporat-ing information on trait-environment connectionsandor persistent associations of taxamdashthat isgroups of two or more taxa that co-occur in fossillocalities distributed widely through time andspace Current models rely primarily on climaticparameters alone to estimate niche space Suchpaleontologically enhanced species distributionmodels can also be helpful in informing effortsto relocate species into suitable environmentsranging from managed relocation experimentsthat aim to save threatened species to choosingwhich trees to plant in urban and suburban land-scaping in order to jump-start dispersal in antic-ipation of future climatic conditionsEven with ideal corridors however species

will not all respond in concert as climate changesa lesson made clear by the fossil record (10)Some species will move quickly some slowlyand some not at all and species will key on dif-ferent aspects of global change such as temper-ature humidity or biotic interactions Effectivecorridors will maximize the opportunities for suchnatural adjustments to proceed even though theend result will be species assemblages almost cer-tainly different than current or historical ones

Conservation policy implications

Laws and governmental policies have played acritical role in conservation Examples are numer-ous ranging from the court-mediated EndangeredSpecies Act in the United States to extremes suchas the ldquoshoot-to-killrdquo policy for poachers in SouthAfrica and Kenya An open question under veryrapid global change however is whether exist-ing policies and laws are adequate to facilitatemanaging for the adaptive capacity of ecosys-tems as opposed to simply mandating the pres-ence of certain species (9 83 84) Answeringthat question will require concerted interactionsamong conservation biologists paleobiologistsand the policy and law communities nationallyand internationally A key challenge for paleo-biologists and conservation biologists will be iden-tifying ecological metrics that are meaningfulfor legislation

Conclusions and outlook

Effective conservation of biological resources nowinvolves understanding and anticipating changein ecological systems in terms of adaptive ca-pacity and ecosystem structure and functionknowledge that will become even more impor-

tant in the future The path forward requiresenhanced use of information from the fossil androck records in conservation planning and prac-tice combined with the coordination of conser-vation efforts situated in historical and novelecosystems Future efforts need to clearly dif-ferentiate between historical and novel ecosys-tems identify key resiliencies and features ofpast ecosystems that may be generally applica-ble to the future and characterize the functionalinteractions that persist in ecosystems for at leastthousands of years All of these tasks requireintegrating information from paleobiology Earthsciences and conservation biology through useof both taxon-based and taxon-free analyses thatallow parallel characterization and comparisonof contemporary and past ecosystems Taxon-freemethodsmdashwhich allow comparisons of functionalattributes of past present and future ecosystemsregardless of the species involvedmdashmay proveespecially useful for conservation efforts in novelecosystems and for calibrating the extent offunctional change that historical ecosystemswill experience under ongoing and future globalpressures In addition to implementing thesenew approaches to conservation it will be es-sential to deal with the root causes of the con-servation crisismdashrapid human population growthoverconsumption of goods and resources andclimate changemdashin order to keep nature diverseadaptive and able to fulfill the needs of the bil-lions of people for whom Earth is the only home

REFERENCES AND NOTES

1 UNDESA World Population Prospects The 2012 RevisionVolume II Demographic Profiles (United Nations Departmentof Economic and Social Affairs Population Division 2013)pp 439 httpsesaunorgunpdwpppublicationsFilesWPP2012_Volume-II-Demographic-Profilespdf

2 J R Oakleaf et al A world at risk Aggregating developmenttrends to forecast global habitat conversion PLOS ONE 10e0138334 (2015) doi 101371journalpone0138334pmid 26445282

3 Y Lu et al Addressing Chinarsquos grand challenge of achievingfood security while ensuring environmental sustainabilitySci Adv 1 e1400039 (2015) doi 101126sciadv1400039pmid 26601127

4 B A Stein P Glick N Edelson A Staudt Climate-SmartConservation Putting Adaptation Principles into Practice(National Wildlife Federation 2014)

5 R Colwell et al Revisiting Leopold Resource Stewardship inthe National Parks Report of the National Park SystemAdvisory Board (National Park Service Advisory Board 2012)pp 1ndash24

6 R J Hobbs et al Managing the whole landscape Historicalhybrid and novel ecosystems Front Ecol Environ 12557ndash564 (2014) doi 101890130300

7 E A Chornesky et al Adapting Californiarsquos ecosystemsto a changing climate Bioscience 65 247ndash262 (2015)doi 101093bioscibiu233

8 J J Hellmann M E Pfrender Future human intervention inecosystems and the critical role for evolutionary biologyConserv Biol 25 1143ndash1147 (2011) doi 101111j1523-1739201101786x pmid 22070271

9 A Camacho H Doremus J S McLachlan B A MinteerReassessing conservation goals in a changing climateIss Sci Technol 26 4 (2010) httpissuesorg26-24p_camacho

10 G P Dietl et al Conservation paleobiology Leveragingknowledge of the past to inform conservation andrestoration Annu Rev Earth Planet Sci 43 79ndash103 (2015)doi 101146annurev-earth-040610-133349

11 J J Lawler et al The theory behind and the challenges ofconserving naturersquos stage in a time of rapid change ConservBiol 29 618ndash629 (2015) doi 101111cobi12505pmid 25922899

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 8 of 10

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

12 A L Davies M J Bunting Applications of palaeoecology inconservation Open Ecol J 3 54ndash67 (2010) doi 1021741874213001003020054

13 E H Barnosky Ecosystem dynamics through the past 2000years as revealed by fossil mammals from Lamar Cave inYellowstone National Park USA Hist Biol 8 71ndash90 (1994)doi 10108010292389409380472

14 R L B Hooke J F Martiacuten-Duque J Pedraza Landtransformation by humans A review GSA Today 22 1ndash10(2012)

15 E C Ellis E C Antill H Kreft All is not loss Plantbiodiversity in the anthropocene PLOS ONE 7 e30535(2012) doi 101371journalpone0030535 pmid 22272360

16 G Ceballos et al Accelerated modern human-inducedspecies losses Entering the sixth mass extinction Sci Adv1 e1400253 (2015) doi 101126sciadv1400253pmid 26601195

17 R A Mittermeier et al Wilderness and biodiversityconservation Proc Natl Acad Sci USA 100 10309ndash10313(2003) doi 101073pnas1732458100 pmid 12930898

18 G M Mace Ecology Whose conservation Science 3451558ndash1560 (2014) doi 101126science1254704pmid 25258063

19 R A Bay U Ramakrishnan E A Hadly A call for tigermanagement using ldquoreservesrdquo of genetic diversity J Hered105 295ndash302 (2014) doi 101093jheredest086pmid 24336928

20 M W Schwartz et al Managed relocation Integrating thescientific regulatory and ethical challenges Bioscience 62732ndash743 (2012) doi 101525bio20126286

21 88th Congress Second Session in Public Law 88-577(16 US C 1131-1136) US Congress Ed (United StatesCongress 1964) vol 3 September 1964 wwwwildernessnetNWPSdocumentspubliclawsPDF16_USC_1131-1136pdf

22 C D Mendenhall D S Karp C F J Meyer E A HadlyG C Daily Predicting biodiversity change and avertingcollapse in agricultural landscapes Nature 509 213ndash217(2014) doi 101038nature13139 pmid 24739971

23 C Kremen Reframing the land-sparingland-sharing debatefor biodiversity conservation Ann NY Acad Sci 135552ndash76 (2015) doi 101111nyas12845 pmid 26213864

24 A S Leopold S A Cain C M Cottam I N GabrielsonT Kimball Wildlife Management in the National ParksThe Leopold Report The National Park Service (1963) httpswwwnpsgovparkhistoryonline_booksleopoldleopoldhtm

25 C Meine M Souleacute R E Noss ldquoA mission-driven disciplinerdquoThe growth of conservation biology Conserv Biol 20 631ndash651(2006) doi 101111j1523-1739200600449x pmid 16909546

26 E O Wilson Half Earth Our Planetrsquos Fight For Life (LiverightW W Norton and Company 2016)

27 P Kareiva M Marvier R Lalasz Conservation in theAnthropocene beyond solitude and fragility The BreakthroughWinter (2012) httpthebreakthroughorgindexphpjournalpast-issuesissue-2conservation-in-the-anthropocene

28 H Tallis J Lubchenco Working together A call for inclusiveconservation Nature 515 27ndash28 (2014) doi 101038515027a pmid 25373659

29 A D Barnosky Dodging Extinction Power Food Money andthe Future of Life on Earth (Univ of California Press 2014)

30 J L Gill et al A 25-million-year perspective on coarse-filter strategies for conserving naturersquos stage ConservBiol 29 640ndash648 (2015) doi 101111cobi12504pmid 25924205

31 J L Blois J L McGuire E A Hadly Small mammal diversityloss in response to late-Pleistocene climatic change Nature465 771ndash774 (2010) doi 101038nature09077pmid 20495547

32 E A Hadly Influence of late-Holocene climate on northernRocky Mountain mammals Quat Res 46 298ndash310 (1996)doi 101006qres19960068

33 A D Barnosky et al Exceptional record of mid-Pleistocenevertebrates helps differentiate climatic from anthropogenicecosystem perturbations Proc Natl Acad Sci USA 1019297ndash9302 (2004) doi 101073pnas0402592101pmid 15197254

34 M A Stegner M Holmes Using palaeontological data toassess mammalian community structure Potential aid inconservation planning Palaeogeogr PalaeoclimatolPalaeoecol 372 138ndash146 (2013) doi 101016jpalaeo201204019

35 J L McGuire E B Davis Conservation paleobiogeographyThe past present and future of species distributionsEcogeography 37 1092ndash1094 (2014)

36 NRC The Geological Record of Ecological DynamicsUnderstanding the Biotic Effects of Future Environmental Change(National Research Council of the National Academies 2005)

37 J T Eronen et al Ecometrics The traits that bind the pastand present together Integr Zool 5 88ndash101 (2010)doi 101111j1749-4877201000192x pmid 21392327

38 P L Koch A D Barnosky Late Quaternary extinctions Stateof the debate Annu Rev Ecol Evol Syst 37 215ndash250(2006) doi 101146annurevecolsys34011802132415

39 J L Blois E A Hadly Mammalian response to Cenozoicclimatic change Annu Rev Earth Planet Sci 37 181ndash208(2009) doi 101146annurevearth031208100055

40 S T Jackson J L Betancourt R K Booth S T GrayEcology and the ratchet of events Climate variability nichedimensions and species distributions Proc Natl AcadSci USA 106 (suppl 2) 19685ndash19692 (2009) doi101073pnas0901644106 pmid 19805104

41 J J Crees S T Turvey What constitutes a lsquonativersquo speciesInsights from the Quaternary fossil record Biol Conserv 186143ndash148 (2015) doi 101016jbiocon201503007

42 K C Maguire D Nieto-Lugilde M Fitzpatrick J WilliamsJ L Blois Modeling species and community responses topast present and future episodes of climatic and ecologicalchange Annu Rev Ecol Evol Syst 46 343ndash368 (2015)doi 101146annurev-ecolsys-112414-054441

43 E S Jeffers S Nogueacute K J Willis The role ofpalaeoecological records in assessing ecosystem servicesQuat Sci Rev 112 17ndash32 (2015) doi 101016jquascirev201412018

44 J A Dearing et al Extending the timescale and range ofecosystem services through paleoenvironmental analysesexemplified in the lower Yangtze basin Proc Natl AcadSci USA 109 E1111ndashE1120 (2012) doi 101073pnas1118263109 pmid 22499786

45 K J Willis R M Bailey S A Bhagwat H J B BirksBiodiversity baselines thresholds and resilience Testingpredictions and assumptions using palaeoecological dataTrends Ecol Evol 25 583ndash591 (2010) doi 101016jtree201007006 pmid 20800315

46 E A Hadly Fidelity of terrestrial vertebrate fossilsto a modern ecosystem Palaeogeogr PalaeoclimatolPalaeoecol 149 389ndash409 (1999) doi 101016S0031-0182(98)00214-4

47 S Porder A Paytan E A Hadly Mapping the origin of faunalassemblages using strontium isotopes Paleobiology 29197ndash204 (2003) doi 101017S0094837300018066

48 R C Terry On raptors and rodents Testing the ecologicalfidelity of cave death-assemblages through live-deathanalysis Paleobiology 36 137ndash160 (2010) doi 1016660094-8373-361137

49 S M Kidwell Biology in the Anthropocene Challenges andinsights from young fossil records Proc Natl AcadSci USA 112 4922ndash4929 (2015) doi 101073pnas1403660112 pmid 25901315

50 S T Jackson Representation of flora and vegetation inQuaternary fossil assemblages Known and unknown knownsand unknowns Quat Sci Rev 49 1ndash15 (2012) doi 101016jquascirev201205020

51 S A Elias Differential insect and mammalian response toLate Quaternary climate change in the Rocky Mountainregion of North America Quat Sci Rev 120 57ndash70 (2015)doi 101016jquascirev201504026

52 J M Tylianakis E Laliberteacute A Nielsen J BascompteConservation of species interaction networks Biol Conserv143 2270ndash2279 (2010) doi 101016jbiocon200912004

53 I A Hatton et al The predator-prey power law Biomassscaling across terrestrial and aquatic biomes Science 349aac6284 (2015) doi 101126scienceaac6284pmid 26339034

54 D D Ackerly et al The evolution of plant ecophysiologicaltraits Recent advances and future directions Bioscience 50979ndash995 (2000) doi 1016410006-3568(2000)050[0979TEOPET]20CO2

55 F Krausmann et al Global human appropriation of netprimary production doubled in the 20th century Proc NatlAcad Sci USA 110 10324ndash10329 (2013) doi 101073pnas1211349110 pmid 23733940

56 E A Hadly et al Genetic response to climatic changeInsights from ancient DNA and phylochronology PLOS Biol2 e290 (2004) doi 101371journalpbio0020290pmid 15361933

57 M A Huerta C Whitlock J Yale Holocene vegetationndashfirendashclimate linkages in northern Yellowstone National Park USA

Palaeogeogr Palaeoclimatol Palaeoecol 271 170ndash181 (2009)doi 101016jpalaeo200810015

58 Resilience_Alliance Key Concepts Resilience Alliance (2016)wwwresallianceorgkey-concepts

59 K L Cole et al Past and ongoing shifts in Joshua treedistribution support future modeled range contraction EcolAppl 21 137ndash149 (2011) doi 10189009-18001pmid 21516893

60 N S Diffenbaugh C B Field Changes in ecologically criticalterrestrial climate conditions Science 341 486ndash492 (2013)doi 101126science1237123 pmid 23908225

61 B J McGill E A Hadly B A Maurer Community inertia ofQuaternary small mammal assemblages in North AmericaProc Natl Acad Sci USA 102 16701ndash16706 (2005)doi 101073pnas0504225102 pmid 16260748

62 J D Yeakel et al Collapse of an ecological network inAncient Egypt Proc Natl Acad Sci USA 111 14472ndash14477(2014) doi 101073pnas1408471111 pmid 25201967

63 M M Pires et al Pleistocene megafaunal interactionnetworks became more vulnerable after human arrivalProc R Soc B 101098rspb20151367 (2015)

64 F Schepers W de Bruijn Rewilding Europe 2013 AnnualReview Rewilding Europe 2014 1ndash68 (2014)

65 H W Greene in After preservation Saving American Naturein the Age of Humans B A M a S J Pyne Ed (Univ ofChicago Press 2015) pp 105ndash113

66 B J McGill B J Enquist E Weiher M Westoby Rebuildingcommunity ecology from functional traits Trends Ecol Evol21 178ndash185 (2006) doi 101016jtree200602002pmid 16701083

67 A M Lawing J J Head P D Polly in Paleontology in Ecology andConservation J Louys Ed (Springer-Verlag 2012) p 276

68 P D Polly J J Head in Earth-Life Transitions Paleobiologyin the Context of Earth System Evolution The PaleontologicalSociety Papers P D Polly J J Head D L Fox Eds(Yale Press 2015) vol 21 pp 21ndash46

69 I Žliobaitė et al Herbivore teeth predict climatic limits inKenyan ecosystems Proc Natl Acad Sci USA 11312751ndash12756 (2016) doi 101073pnas1609409113pmid 27791116

70 P D Polly in Carnivoran Evolution New Views on PhylogenyForm and Function A Goswami A Friscia Eds (CambridgeUniv Press 2010)

71 C T Webb J A Hoeting G M Ames M I PyneN LeRoy Poff A structured and dynamic framework toadvance traits-based theory and prediction in ecologyEcol Lett 13 267ndash283 (2010) doi 101111j1461-0248201001444x pmid 20455917

72 A Escudero F Valladares Trait-based plant ecology Movingtowards a unifying species coexistence theory Features ofthe Special Section Oecologia 180 919ndash922 (2016)doi 101007s00442-016-3578-5 pmid 26897604

73 U Ramakrishnan E A Hadly Using phylochronology toreveal cryptic population histories Review and synthesis of29 ancient DNA studies Mol Ecol 18 1310ndash1330 (2009)doi 101111j1365-294X200904092x pmid 19281471

74 C N K Anderson U Ramakrishnan Y L Chan E A HadlySerial SimCoal A population genetics model for data frommultiple populations and points in time Bioinformatics 211733ndash1734 (2005) doi 101093bioinformaticsbti154pmid 15564305

75 C J Donlan De-extinction in a crisis discipline FrontBiogeogr 6 25ndash28 (2014)

76 A E Camacho Going the way of the dodo De-extinctiondualisms and reframing conservation Wash Law Rev 92849ndash906 (2015)

77 J Hillerislambers S G Yelenik B P Colman J M LevineCalifornia annual grass invaders The drivers or passengersof change J Ecol 98 1147ndash1156 (2010) doi 101111j1365-2745201001706x pmid 20852668

78 F Eigenbrod P Gonzalez J Dash I Steyl Vulnerability ofecosystems to climate change moderated by habitatintactness Glob Chang Biol 21 275ndash286 (2015)doi 101111gcb12669 pmid 25059822

79 P Gonzalez R P Neilson J M Lenihan R J Drapek Globalpatterns in the vulnerability of ecosystems to vegetationshifts due to climate change Glob Ecol Biogeogr 19755ndash768 (2010) doi 101111j1466-8238201000558x

80 J R Stewart A M Lister I Barnes Refugia revisitedIndividualistic responses of species in space and timeProc R Soc London B 277 661ndash671 (2010)

81 J L McGuire J J Lawler B H McRae T A NuntildeezD M Theobald Achieving climate connectivity in a fragmented

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 9 of 10

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

landscape Proc Natl Acad Sci USA 113 7195ndash7200 (2016)doi 101073pnas1602817113 pmid 27298349

82 D G Hole et al Projected impacts of climate change on acontinent-wide protected area network Ecol Lett 12420ndash431 (2009) doi 101111j1461-0248200901297xpmid 19379136

83 E Long E Biber The Wilderness Act and climate changeadaptation Environ Law 44 623ndash694 (2014)

84 E Biber Adaptive management and the future ofenvironmental law Akron Law Rev 46 933ndash962 (2013)

85 F E Mayle R Burbridge T J Killeen Millennial-scale dynamicsof southern Amazonian rain forests Science 290 2291ndash2294(2000) doi 101126science29055002291 pmid 11125139

86 D C Morton Y Le Page R DeFries G J Collatz G C HurttUnderstorey fire frequency and the fate of burned forests insouthern Amazonia Philos Trans R Soc London B Biol Sci368 20120163 (2013) doi 101098rstb20120163pmid 23610169

87 T Fanin G R Werf Relationships between burned areaforest cover loss and land cover change in the BrazilianAmazon based on satellite data Biogeosciences 126033ndash6043 (2015) doi 105194bg-12-6033-2015

88 T Hilker et al Vegetation dynamics and rainfall sensitivityof the Amazon Proc Natl Acad Sci USA 11116041ndash16046 (2014) doi 101073pnas1404870111pmid 25349419

89 A A Alencar P M Brando G P Asner F E Putz Landscapefragmentation severe drought and the new Amazon forestfire regime Ecol Appl 25 1493ndash1505 (2015) doi 10189014-15281 pmid 26552259

90 Y Malhi et al Exploring the likelihood and mechanism of aclimate-change-induced dieback of the Amazon rainforest

Proc Natl Acad Sci USA 106 20610ndash20615 (2009)doi 101073pnas0804619106 pmid 19218454

91 J Settele et al in Intergovernmental Panel on ClimateChange Climate Change 2014 Impacts Adaptation andVulnerability (Cambridge Univ Press 2014)

92 T W Swetnam Fire history and climate change in giantsequoia groves Science 262 885ndash889 (1993) doi 101126science2625135885 pmid 17757357

93 J R Marlon et al Long-term perspective on wildfires in thewestern USA Proc Natl Acad Sci USA 109 E535ndashE543(2012) doi 101073pnas1112839109 pmid 22334650

94 A L Westerling M G Turner E A H SmithwickW H Romme M G Ryan Continued warming couldtransform Greater Yellowstone fire regimes by mid-21stcentury Proc Natl Acad Sci USA 108 13165ndash13170(2011) doi 101073pnas1110199108 pmid 21788495

95 M P North et al Reform forest fire management Science349 1280ndash1281 (2015) doi 101126scienceaab2356pmid 26383934

96 P R Ehrlich P M Kareiva G C Daily Securing naturalcapital and expanding equity to rescale civilization Nature486 68ndash73 (2012) doi 101038nature11157 pmid 22678281

97 R Costanza et al Changes in the global value of ecosystemservices Glob Environ Change 26 152ndash158 (2014)doi 101016jgloenvcha201404002

98 M C Bottrill et al Is conservation triage just smart decisionmaking Trends Ecol Evol 23 649ndash654 (2008)doi 101016jtree200807007 pmid 18848367

99 Society for Ecological Restoration International in EcologicalRestoration A Means of Conserving Biodiversity and SustainingLivelihoods Society for Ecological Restoration InternationalInternational Union for Conservation of NaturendashCommission on

Ecosystem Management Eds (Clung Wicha Press Company2004) httpsportalsiucnorglibrarynode8715

100 S Mondol M W Bruford U Ramakrishnan Demographicloss genetic structure and the conservation implications forIndian tigers Proc Biol Sci 280 20130496 httpdxdoiorg2013041020131098rspb2013201320130496 (2013)doi 101098rspb20130496 pmid 23677341

ACKNOWLEDGMENTS

We thank the financial sponsors of the workshop at which theseideas were formulated the Integrative Climate Change BiologyGroup (a scientific program of the International Union of BiologicalSciences) the Museum of Paleontology Berkeley Initiative forGlobal Change Biology and Office of the Vice Chancellor forResearch at the University of California Berkeley the ConservationPaleobiology Group at the Department of Biology StanfordUniversity and the Senckenberg Biodiversity and Climate ResearchCentre Frankfurt Germany J Head and PDP were additionallyfunded by National Science Foundation ELT (Earth-Life Transitions)awards NSF EAR-1338028 (J Head) and NSF EAR-13388298(PDP) ADB and CRM were partly funded by NSF grantEAR-1148181 We are especially grateful to V Bowie for logisticalsupport and graduate student helpers A Poust S Elshafie andP Kloess We thank D Thomas for his production of tiger mapsthat were modified for Fig 5 and the Howard Hughes MedicalInstitute for granting permission to use the maps (original versionsare available at wwwBioInteractiveorg copyright 2014 all rightsreserved) E Holt N Spano B Stein Zixiang Zhang and anonymousreviewers provided constructive comments

101126scienceaah4787

Barnosky et al Science 355 eaah4787 (2017) 10 February 2017 10 of 10

RESEARCH | REVIEW

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from

(6325) [doi 101126scienceaah4787]355Science (February 9 2017) Stegner Nils Chr Stenseth Marvalee H Wake and Zhibin ZhangKashish Das Shrestha Katherine Solari Lynn Stegner M Allison Mychajliw Carsten Nowak Uma Ramakrishnan Jan SchnitzlerCharles R Marshall Chase Mendenhall Andreas Mulch Alexis M Paul L Koch Claire Kremen Emily L Lindsey Cindy LooyHellmann Thomas Hickler Stephen T Jackson Melissa Kemp Holly Doremus Mikael Fortelius Harry W Greene JessicaGerardo Ceballos Edward Davis Gregory P Dietl Rodolfo Dirzo D Ackerly Ken Alex Eric Biber Jessica Blois Justin BrasharesHead P David Polly A Michelle Lawing Jussi T Eronen David Anthony D Barnosky Elizabeth A Hadly Patrick Gonzalez Jasonfuture of terrestrial ecosystemsMerging paleobiology with conservation biology to guide the

Editors Summary

this issue p eaah4787Scienceecological resilience is maintained even in the face of changeargue that the best way to do this is to look back at paleontological history as a way to understand how

et alsystems so that they can continue to adapt and function into the future In a Review Barnosky approaches to take in conservation ultimately we need to permit or enhance the resilience of naturalchange have occurred throughout the history of life Although there is much debate about the best

The current impacts of humanity on nature are rapid and destructive but species turnover andLooking back to move forward

This copy is for your personal non-commercial use only

Article Tools

httpsciencesciencemagorgcontent3556325eaah4787article tools Visit the online version of this article to access the personalization and

PermissionshttpwwwsciencemagorgaboutpermissionsdtlObtain information about reproducing this article

is a registered trademark of AAAS ScienceAdvancement of Science all rights reserved The title Avenue NW Washington DC 20005 Copyright 2016 by the American Association for thein December by the American Association for the Advancement of Science 1200 New York

(print ISSN 0036-8075 online ISSN 1095-9203) is published weekly except the last weekScience

on

Febr

uary

22

201

7ht

tp

scie

nce

scie

ncem

ago

rg

Dow

nloa

ded

from