REVIEW SUMMARY

SOIL SCIENCE

Soil and human security in the21st centuryRonald Amundson Asmeret Asefaw Berhe Jan W Hopmans Carolyn OlsonA Ester Sztein Donald L Sparks

BACKGROUND Earthrsquos soil has formed byprocesses that have maintained a persistentand expansive global soil mantle one that inturn provided the stage for the evolution of thevast diversity of life on land The underlyingstability of soil systems is controlled by theirinherent balance between inputs and losses ofnutrients and carbon Human exploitation ofthese soil resources beginning a few thousandyears ago allowed agriculture to become anenormous success The vastness of the planetand its soil resources allowed agriculture toexpand with growing populations or tomovewhen soil resources were depleted Howeverthe practice of farming greatly acceleratedrates of erosion relative to soil production andsoil has been and continues to be lost at ratesthat are orders ofmagnitude greater thanmech-anisms that replenish soil Additionally agri-cultural practices greatly altered natural soilcarbon balances and feedbacks Cultivation thusbegan anongoing slow ignition ofEarthrsquos largestsurficial reservoir of carbonmdashone that when com-

binedwith the anthropogenic warming ofmanybiomes is capable of driving large positivefeedbacks that will further increase the accu-mulation of atmospheric greenhouse gases andexacerbate associated climate change

ADVANCES The study of soil is now the do-main of diverse schools of physical and bio-logical science Rapid advances in empiricaland theoretical understanding of soil processesare occurring These advances have brought aninternational andglobal perspective to the studyof soil processes and focused the implicationsof soil stewardship for societal well-being Majoradvances in the past decade include our firstquantitative understanding of the natural ratesof soil production derived from isotopicmeth-ods developed by collaboration of geochemistsand geomorphologists Proliferation of researchby soil and ecological scientists in the northernlatitudes continues to illuminate and improveestimates of the magnitude of soil carbonstorage in these regions and its sensitivity and

response to warming The role of soil pro-cesses in global carbon and climate models isentering a period of growing attention and in-creasing maturity These activities in turn re-veal the severity of soil-related issues at stakefor the remainder of this centurymdashthe need torapidly regain a balance to the physical andbiological processes that drive and maintainsoil properties and the societal implicationsthat will result if we do not

OUTLOOK Both great challenges and oppor-tunities exist in regards to maintaining soilrsquosrole in food climate and human security Ero-sion continues to exceed natural rates of soilrenewal even in highly developed countriesThe recent focus by economists and natural sci-entists on potential future shortages of phos-phorus fertilizer offers opportunities for novelpartnerships to develop efficient methods ofnutrient recycling and redistribution systems

in urban settings Possi-bly the most challengingissues will be to better un-derstand the magnitudeof global soil carbon feed-backs to climate changeand to mitigating climate

change in a timely fashion The net results ofhuman impacts on soil resources this cen-tury will be global in scale and will have di-rect impacts on human security for centuriesto come

RESEARCH

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 647

The list of author affiliations is available in the full article onlineCorresponding author E-mail earthyberkeleyeduCite this article as R Amundson et al Science 348 1261071(2015) DOI101126science1261071

Large-scale erosion forming a gully system in the watershed of Lake Bogoria Kenya Accelerated soil erosion here is due to both overgrazing andimproper agricultural management which are partially due to political-social impacts of past colonization and inadequate resources and infrastructureThe erosion additionally affects the long-term future of Lake Bogoria because of rapid sedimentation This example illustrates the disruption of thenatural balance of soil production and erosion over geological time scales by human activity and the rapidity of the consequences of this imbalanceC

REDITBRENTSTIRTONG

ETTYIM

AGES

ON OUR WEB SITE

Read the full articleat httpdxdoiorg101126science1261071

on

June

18

201

5w

ww

sci

ence

mag

org

Dow

nloa

ded

from

o

n Ju

ne 1

8 2

015

ww

ws

cien

cem

ago

rgD

ownl

oade

d fr

om

on

June

18

201

5w

ww

sci

ence

mag

org

Dow

nloa

ded

from

o

n Ju

ne 1

8 2

015

ww

ws

cien

cem

ago

rgD

ownl

oade

d fr

om

on

June

18

201

5w

ww

sci

ence

mag

org

Dow

nloa

ded

from

o

n Ju

ne 1

8 2

015

ww

ws

cien

cem

ago

rgD

ownl

oade

d fr

om

on

June

18

201

5w

ww

sci

ence

mag

org

Dow

nloa

ded

from

REVIEW

SOIL SCIENCE

Soil and human security in the21st centuryRonald Amundson1 Asmeret Asefaw Berhe2 Jan W Hopmans3 Carolyn Olson4

A Ester Sztein5 Donald L Sparks6

Human security has and will continue to rely on Earthrsquos diverse soil resources Yet we havenow exploited the planetrsquos most productive soils Soil erosion greatly exceeds rates ofproduction in many agricultural regions Nitrogen produced by fossil fuel and geologicalreservoirs of other fertilizers are headed toward possible scarcity increased cost andorgeopolitical conflict Climate change is accelerating the microbial release of greenhousegases from soil organic matter and will likely play a large role in our near-term climatefuture In this Review we highlight challenges facing Earthrsquos soil resources in the comingcentury The direct and indirect response of soils to past and future human activities willplay a major role in human prosperity and survival

Soil is the living epidermis of the planet (1)Globally soil is themedium throughwhicha number of atmospheric gases are biolog-ically cycled and through which waters arefiltered and stored as they pass through

the global hydrological cycle (2) Soil is a largeand dynamic reservoir of carbon and the phys-ical substrate for most of our food productionProfound changes are on the horizon for theseinterconnected functionsmdashparticularly sparkedby changes to climate and food productionmdashthatwill likely reverberate through society this cen-tury Ultimately the way in which we directly andindirectly manage our planetrsquos soil will be inter-woven within our future success as a speciesSoil is commonly thought of as the ~1-m-thick

layer of biogeochemically altered rock or sedimentat Earthrsquos surface that has acquired numerousqualities during its exposure to the atmosphere thatgreatly distinguish it from its geological sources (3)Soil-forming chemical reactions createmicrometer-sized electrically negative clay minerals that im-part soil with plant nutrient retention capabilities(4) The electrical charge characteristics of soilcombined with its small particle size and highsurface area allow it to temporarily store rain andsnow melt for plant use and provide sufficientresidence time for a multitude of chemical reac-tions to occur that may remove or reduce the

toxicity of contaminants The water stored insoilmdashtermed greenwater (5)mdashserves as the sourcefor 90 of the worldrsquos agricultural productionand represents ~65 of global fresh water (5)Last the intimate intermingling of lifemdashplantanimal and microbialmdashwithin the soil matrixdrives redox reactions that control many elemen-tal cycles (6) and creates a reservoir of organicC that greatly exceeds the C in the global atmo-sphere and biosphere (7) The microbial com-munities that mediate these redox reactions arenow believed to represent much of Earthrsquos totalbiodiversity (8) but the nature function andeconomic potential of this soil biosphere is onlybeginning to be probed (6)Soil due to global variations in climate geol-

ogy and biota (3) has tremendous spatial diver-sity More than 20000 soil types (or soil series)have been identified and mapped in the UnitedStates alone (9) and the number identified in-creases as land area investigated increases If thesoil seriesndashtondashland area relationship (10) is extrap-olated to global ice-free land area the resultssuggest that there are more than 300000 serieson the planet The response of these soils to per-turbations can be extremely varied because of theirdiverse chemical physical and biological char-acteristics suggesting the importance as a sim-ple precautionary principle ofmaintaining segmentsof this diversity for the stability and resilience ofglobal biogeochemical systems in the face of an-thropogenic disturbances

Human Imprint on Soil

Humans altered the ecosystems they encoun-tered as they began their spread across the globeHowever the most momentous development inhuman landscape change occurred with the in-vention and adoption of agriculture (11) Mostagricultural practices involve the removal of thenatural flora the simplification of biodiversity tofavor monocultures and the physical disruption

of the soil Since the Industrial Revolution ex-panding populations have relied on the exploi-tation of more andmore soil for a correspondinggrowth in food production Today ~12 of ice-free land is in cropland and 38 is used for com-bined cropping and grazing (12) an area roughlyequivalent to the land area covered by ice andscoured or otherwise disturbed during the lastglacial maximum (Fig 1A) In addition to the sim-ilarity in area the agricultural impact on soilprocesses rivals or exceeds the effect of those icesheets in both rapidity and magnitudeUndisturbed soils have the characteristic as

result of a number of feedback mechanisms ofbeing able to retain many of their features indef-initely over timemdashtheir thickness C content andnutrients for examplemdasha condition that is equat-able to sustainability (Fig 2) Cultivated soils arehighly modified forms of their wild predecessorsand may thus be viewed as domesticated soils(9) One key characteristic is that domesticatedsoils seldom are able to maintain the qualities oftheir original conditions and these changesgreatly affect their productivity and their impacton surrounding geochemical cycles The effortsto improve themanagement and conservation ofthese domesticated soils and the preservation ofportions of their remaining wild ancestral stockwill be among themost important challenges thiscentury (9 13) Analyses of the combined agricul-tural and urban impact on soil series in theUnitedStates for example revealed large areas in theagricultural heartland where more than 50 ofthe soil series had been domesticated Soil diver-sity like biodiversity (14) provides an array ofhuman-valued goods and services Among themostapparent issues is the ability of soil to providesustained agricultural productionThe domesticated soil landscape is one ofEarthrsquos

most valuable commodities For example nearly$3816 billion (US dollars) in agricultural pro-ducts were produced globally in 2012 (15) How-ever agriculture is competing with increasingurban and suburban soil demands The conver-sion of soil to urban land is largely irreversible onhuman time scales There is uncertainty both inthe present and the future distribution of urbanland on Earth (Fig 1B) A recent meta-analysissuggests that between 1970 and 2000 an areagreater than the size of Denmark was urbanizedand that in the next 20 years 15 million km2 ofland (the size ofMongolia) will be urbanized (16)The conversion of farmland to urban areas mustbe weighed against the fact that our most pro-ductive soils have already been exploited andthat demand for food production will continueto increase

Soil and Climate Security

A relatively stable climate has been the stageon which the great human inventions of agri-culture and industrialization have evolved anddirect or indirect human impacts on soil C cy-cling processes will have much to do with atmo-spheric greenhouse gas concentrations and theassociated climate implications by the end of thiscentury

RESEARCH

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-1

1Department of Environmental Science Policy andManagement University of California Berkeley CA 94720USA 2Life and Environmental Sciences Unit University ofCalifornia Merced CA 95343 USA 3Land Air and WaterResources One Shields Avenue Davis CA 95616 USA4Climate Change Program Office Office of the ChiefEconomist US Department of Agriculture (USDA) 14th andIndependence SW Washington DC 20013 USA 5Board onInternational Scientific Organizations National Academy ofSciences 500 Fifth Street NW Washington DC 20001 USA6Plant and Soil Science Chemistry and Biochemistry Civiland Environmental Engineering and Marine Science andPolicy University of Delaware Newark DE 19716 USACorresponding author E-mail earthyberkeleyedu

Organic C stored in soil is the balance betweenplant inputs andmicrobiallymediatedmetaboliclosses as CO2 (Fig 2) In unperturbed conditionssoils achieve steady-state C pools on time scalesof centuries to a few millennia The total store ofsoil organic carbon is still uncertain but recentestimates suggest pools on the order of 2300 giga-tons (Gt) in the upper 3 m (7) Soil cultivationand clearing has caused a major fraction of totalanthropogenic greenhouse gas emissions since the19th century (17) Cultivation is amajor disruptionto the natural C balance in soil one that altersthe physical and biological structure of soil effec-tively igniting throughmicrobially mediated pro-cesses a vast store of labile C that has accumulatedover millennia (18) During the first few decadesthat soil is cultivated up to 50 of the carbonpool is oxidized to CO2 eventually a quasi-steady-state soil C pool is achieved (19) Based on theglobal agricultural land area cultivation has like-ly released between 50 and 70 Gt of C to the at-

mosphere over the course of human history (20)and the combined cultivation and biomass burn-ing contributions to atmospheric CO2 exceededthat of fossil fuel emissions well into the 20thcentury (17) However the agricultural imprinton atmospheric greenhouse gas concentrationsappearedmuch earlier in the Holocene (21) Earlyspikes in atmospheric CO2 and CH4 correspondedto agricultural expansion inMesopotamia and inChina (22) Much of the historical C loss was fromthe soils of forests and grasslands of the northernlatitudesHowever today the locus of land alterationhas shifted to the equatorial latitudes and up to10of global anthropogenic CO2 emissions are froma combination of biomass burning and soil culti-vation in the humid and subhumid tropics (23)Under changed management or through land

abandonment global agricultural soils have thecapacity to reapproach their original C storageand regain up to a half a decade of present fossilfuel emissions (over a multidecade period) Bet-

ter stewardship of domesticated soils that leadsto higher organic matter contents is a valuablepractice from an ecological perspective and froman agronomic point of view (24) There is now alarge body of research on the rates of C seques-tration under differing management practicesHowever there are limits to these practices as ameans of mitigating continued fossil fuel emis-sions First a serious concern withmanagement-based soil C sequestration strategies is that theyare dependent on restricted management op-tions in a highly decentralized and economicallydriven agricultural sector (25) A change in landownership or a change in factors driving agricul-tural practices can rapidly releasemuch of regainedC Second the effectiveness of soil C sequestrationis time dependent For example if all potentialsoil sequestration strategies were establishedthey would initially serve as a sink of about 13 Gtof C yearndash1 (Table 1) (26) but this sink termwould be expected to decline nonlinearly to low

1261071-2 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 1 The global patterns of soil properties and the human imprint onthem Expanding urban centers (55) remove soil from other uses whereasthe creation of cropland (56) is one of the major drivers of imbalances in thesoil carbon cycle (57) and accelerated rates of soil erosion (58) Phosphorus acritical plant nutrient is unevenly distributed and inherently low in warm andhumid climates (59)

RESEARCH | REVIEW

sequestration rates over a period of several dec-ades as a new soil C steady state is reached (27)Equally as important is the difficulty of actuallyachieving thismaximum potential which involvesmultiple governments and millions of individualland managers Last soil management effects onthe global C balance are inherently small relativeto the climate-driven changes to soil C storagethat will occur in this centurySoil C storage is well documented to decline

with increasing temperature and decreasing soilmoisture and soil C storage patterns mirror glob-al climate zones (28 29) with secondary impacts

by bedrock topography and soil age (20) Soil Cpools are the balance between plant inputs andmicrobial decomposition (Fig 3) and the responsesof these processes to anthropogenic climate changeare considered to be large (in the case of inputs)poorly constrained (30) and complicated by tem-perature and moisture interactions Anthropo-genic increases in atmospheric CO2 may driveincreased net primary production (NPP) as longas nutrient and water limitations do not occur(31) which ultimately may have a negative feed-back on atmospheric CO2 through increased in-puts to soil C (Fig 3) On the other hand increasing

air temperatures warm soil melt permafrost andstimulate biological metabolism of soil carbonpools driving what appears to be a large positivefeedback process (32) (Figs 1C and 3) Based oncurrent earth system modeling additions of soilC by increased NPP (relative to an 1850 BCE ref-erence date) are projected to be between 160 and1230 Gt by 2100 whereas C losses by increaseddecomposition are projected to be between 104and 629 Gt (31) Overall models suggest net soilC changes from a loss of 72 Gt to gains of 253 Gtby 2100 (31) However such exercises include greatuncertainties in both projected gains (by CO2-enhancedphotosynthesis) and losses (by soilwarm-ing) and inassumptions about long-termecosystemresponse to ever-increasing CO2 concentrationsOne important uncertainty is the response ofnorthern latitude soils to warming which couldresult in net soil C losses between 50 to 150 Gt(32 33) Last the current generation of earthsystemmodels has difficulty inmatching present-day soil carbon storage patterns (34) and tuningthe models is challenged by empirical uncertain-ties in the global soil C pool of more than 770 Gt(34) an uncertainty similar in size to the presentatmospheric C poolStill debated is the impact of soil erosion on

the global C cycle When agricultural soil is lostby water or wind erosion the surficial and mostC-rich material is preferentially removed whichaccelerates the decline in the soil C pool Rates ofsoil C replacement by crops and plants are rapidenough in certain situations to maintain soil Clevels at a steady state under the condition ofconstant erosionmdasheg creating an ongoing sink(35) This sink represents a net reduction inatmospheric CO2 only if the eroded C is not re-oxidized Because some depositional environmentsare conducive to partial preservation of buriedC (lakes reservoirs basins floodplains) the neteffect of accelerated agricultural erosionwas firstsuggested to be a global C sink of 06 to 15 Gtyearminus1 a rate similar to the total global land sink(35) If the eroded C is largely oxidized howeverit may result in no net sink (or possibly even anet source) (36) The most recent estimates sug-gest that agricultural erosion of soil C may be040 T020 Gt of C yearndash1 (37) If we benefit froman unintended C sink due to soil erosion anybenefits must be clearly balanced against therelated losses of nutrients and reduction of en-vironmental quality that require fossil fuel energyto remediate (38)The global soil C cycle has been greatly per-

turbed by human activity both directly throughfarming and indirectly through anthropogenicclimate change All projected soil C gains andlosses this century are highly uncertain becauseof economic population and political influences(which will largely affect carbon sequestrationefforts) and uncertainties in the magnitude of thesoil response to warming (because of the com-plexity of the soil C pool structure) (Fig 3) Hu-man changes to the global atmosphere andclimate are likely to simultaneously drive bothvery large gains and losses of soil Cmdashfluxes thatare equivalent to decades of emissions at present

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-3

Table 1 Published estimates of soil C sources and sinks for the 21st century

Management TypeMaximum Flux(Gt of C yearminus1)

ReferenceCumulative Flux

(Gt)

Carbon SinksIncreased net primary production (31) 160 to 1230Erosion 040 (37) 40

ManagementCropland management 036 (26 27)Grazing land management 037 (26 27)Restore degraded land 018 (26 27)Restore organic soils 036 (26 27)Total for management 126 164

Carbon SourcesLand clearingdagger 250Soil warmingBoreal regions (32) 50ndash270Globe (34) 104ndash629Total

Net balance ndash188 to +137Based on maximum new cropland by 2050 (10 billion ha) (68) and assumed loss of 25 of an average Ccontent of 10 kg mminus2 (29) daggerCalculated assuming exponential decline to an ultimate landscapesaturation after 50 years

Fig 2 Changes in the balance of important soil processes caused by human disturbance Manysoil characteristics are the balance or the result of a number of processes that respond to changes inenvironmental variables (3) However properties such as hillslope soil thickness organic carbon storageN content and other features attain steady state in intervals of a few centuries to millennia and appearcapable of regaining stability Human intervention in soil processes many times exceeds natural per-turbations and thus exceeds the resiliency of soil to recover to its original condition Viewed broadlysteady state is a quantitative measure of soil sustainability

RESEARCH | REVIEW

rates of fossil fuel consumption The presentlyunknown balance andmost importantly its signbetween the large fluxes represent considerableuncertainty for climate security (Table 1)

Soil and Food Security

The late 20th and early 21st centuries have beenfor industrialized countries an unprecedentedera of increasingly low food prices (39) There arenumerous factors that may reverse this trendmdashsuch as increased global demand climate change(40) and competition for soil by nonagriculturaluses such as biofuels or urbanization Abundantenergy has been the key driving force behind ourability to maintain food production apace withan expanding population that is estimated to reach11 billion by 2100 (41) Low-cost energy whichled to advanced agriculturalmachinery replacinghuman labor is causing migration to urban cen-ters Energy is used to replace the soil nutrientsremoved or lost by agricultural perturbations ofsoil Energy transforms atmospheric N2 to thebioavailable NH4 fertilizer through the energy-intensive Haber-Bosch processmdashconstituting thefirst and most important green revolution (42)(Fig 4A) one that allows us to feed the increas-ing global population (Fig 4A) Before the indus-trial fixation of N any increase in food productionfor a given country was largely due to increasedsoil used for production (43) and only after theadvent of N fertilizer (Fig 4A) did yields per areaof major crops begin their upward trajectory (43)

Last energy is essential to mine and transportessential plant nutrients such as P andK that canonly be accessed from limited geological reservoirsAgricultural soil erosion is one of the most de-

structive human perturbations to soil sustain-ability Given little opportunity or desirability forfurther agricultural expansion stewardship of ourexisting domesticated soil is essential for sus-

tained human prosperity Yet despite the impor-tance of soil conservation the implementation ofpractices to minimize soil erosion has not fol-lowed apace with the severity of the problem Themost pervasive mechanism of soil erosion is viawater Before European contact and the removalof native vegetation by plowing and cultivationthe geologicalmechanism of soil erosion onmost

1261071-4 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 3 A cause-effect diagram of the major soil-atmospheric CO2 feed-back processesThe values in parentheses in the circles are the approximatepool sizes (Gt) of C A solid arrow represents a direct response (eg as CO2

increases temperature increases) an arrow with a circle indicates an inverseresponse (eg as temperature increases soil carbon storagedecreases) Dashedlines are for processes that are less well understood (A) Environments withadequate moisture (B) Water-limited environments In (A) the a-b-c loop is apositive feedback process (even number of inverse relationship arrows) oneespecially important in regions of melting permafrostThe d-f-c loop is a negativefeedback one with less certain feedbacks between vegetation and soil (f-g loop)and temperature (h-i)The strength of the a-b-c- versus d-f-c loops on soil carbonpools will likely determine whether soil carbon losses in northern latitudes serve

as amajor source of CO2 and CH4 this century a balance that also hinges on theabilityof the soils to supply nutrients to plants (arrowg) in order to respond to theincreases in CO2 (arrow d) In (B) regions with limitedmoisture the strengths ofvegetation response toCO2 (d) and soil carbon response to temperature (b)maybe weakened (thinner arrow) In addition the vegetation response to increasingtemperature may become negative These figures reveal the importance of soilcarbon to the global CO2 balance this century as well as the uncertainties in thestrength and direction of important processes Arrow references are as follows a(60) b (32 33) c CO2 loss by respiration is the overwhelming pathway of Cremoval from soils d water efficiency response in (61) f soil C is the balancebetween plant inputs and decomposition losses g not well constrained but seediscussion in (62) h (63) i eg (64)

Fig 4 The postndashWorld War II rise in fertilizer production and cost coincides with the spike inglobal human populations The growth in world population in the late 20th century (65) mirrors theincreasing use of industrially derived N fertilizer Before N fertilization formajor US crops (66)wheat andother grain yields per acre increased only following World War II coincident with the rise in the use of Nfertilizer produced by the Haber-Bosch process Prices have sharply risen (67) because global demand isstraining the supplies

RESEARCH | REVIEW

uplands was by slow biologically driven creep(44) The removal of plant cover allows themecha-nism of erosion to change to raindrop dislodg-ment of soil particles and their subsequent removalbyoverland flowwhich is a farmore rapiderosionalprocess For example recent analyses of USerosion rates before European contact place theaverage rate at about 21mMyndash1 (Mymillion years)(45) Today erosion rates in the central UnitedStates can exceed 2000 mMyndash1 (45) whereas therates of soil erosion in portions of the loess pla-teau of China approach 10000 m Myndash1 (46) (Fig1D) This eroded sediment is ultimately replacedby the conversion of the underlying sediment orrock into new soils with the addition of organicmatter and nutrients through biological mecha-nisms Until the past decade the pace of this re-placement process was poorly known and theacceptable rates of soil erosion on agriculturallands were placed at 400mMyndash1 or more (47) Nu-merous studies of natural rates of soil productionnow suggest rates between 50 to 200 mMyndash1 formany environments indicating that the pace oferosion in numerous agricultural areas is or un-til recently has been unsustainable (47) Not onlydoes the loss of soil remove nutrients from thesite of agricultural production (38) but the sedi-ment generated adversely affects local drainageswater bodies and regional aquatic ecology Lastthe maintenance (or even the improvement) ofagricultural production in the face of acceleratedrates of soil erosion is energy intensive Althoughmicrobial symbiosiswithplants can fix atmosphericN to bioavailable forms and can substitute for Nfixed by the Haber-Bosch process there is no bi-ological or atmospheric source for rock-derivednutrients such as P K and CaAlthough natural processes of soil production

and formation replace or release nutrients thepaces of these processes are slow relative to ouranthropogenic use rate (Fig 2) The transport ofcrops from the site of production to other loca-tions remove plant-essential nutrients from thesoil potentially causing deficiencies that limitpotential production levels (48) This furtherdrives a dependence on the mining and distri-bution of macronutrients from geological sourceswhich can create economic inequalities or geo-political conflicts between nations (49) (Fig 1E)The growing demand for P has recently causedan increase in the cost of rock phosphate fromabout $80 per US ton in 1961 to up to $450 perton in 2008 Prices since then have fluctuatedbut are now at about $700 per ton (50) (Fig 4)In addition to cost is the related issue of accessMorocco is estimated to have the worldrsquos largestgeological P reserves much of it in disputed ter-ritory (49) The United States on the other handhas only about 2 of global P resources (51) Atcurrent rates of retrieval the most productivemine in the United States will be depleted in20 years (49) which will force it to become in-creasingly reliant on imports to sustain its agri-cultural and industrial sectorsBecause most other P-reliant countries lack the

geological resources to indefinitely sustain cur-rent use the only means to confront the decline

in reserves (other than conversion fromdomesticto imported P) is to develop amore coherent andintegrated program of P (and other nutrient) re-cycling The loss of nutrients in our human andanimal waste streams is environmentally damag-ing and economically problematic Regaining con-trol of these resources now largely consideredwaste would go far toward substantially lower-ing the demand for imported nutrients and otherresources (52) In addition to P other soil nu-trients appear to be entering periods of limitationor high demand (Fig 1F) For example K (potash)prices were ~$875 per metric ton in 2009 and areexpected to reach $1500 by 2020

The 21st Century Challenge

Humans have domesticated our soil resourcesand the planet (12 53) This domestication hasin turn perturbed a number of soil cycles suchthat they are no longer in balance and the im-balance is changing soil in ways that will affectfuture generations and their climate (Fig 2) Soilmanagement must be geared toward passing ahabitable albeit highly altered landscape to thegenerations that followmdashone where our exploita-tion of and impacts on soil resources is adjustedto the pace of our planetrsquos renewal These strat-egies should focus on regaining a balance in (i)organic C inputs and losses (ii) soil erosion andproduction and (iii) release and loss of nutri-ents Soil sustainabilitymdashbased on quantitativeprinciples andmeasurements of soil erosion andproduction soil nutrient loss and release and soilcarbon loss and returnmdashmust be the ultimate goalfor managing the global soil resource and shouldserve as the driving principle for soil researchthat will support this managementThese are challenging goals that will be diffi-

cult to achieve The solutions will require an ef-fort commensurate with the magnitude of theproblems First effective solutions to soil sustain-ability much like the approaches required to con-tendwith climate change (54) must involve highlymultidisciplinary research in novel intellectualsettings or institutions Second the ultimate suc-cess of any innovation requires a dialog and in-terfacewith policymakers and public institutionstheultimate ldquodecidersrdquo in broad-scale social changeThese linked efforts will depend on continuedand arguablymuch greater investments in knowl-edge and innovative knowledge transfer and simplydifferent ways of conceptualizing and approachingproblems From our vantage point the future ofEarthrsquos soil resources is tenuously in our controlor within our ability to sustain it into the futureOnly those on Earth in 2100 will know how wellwe succeeded

REFERENCES AND NOTES

1 ldquoWe might say that the earth has a spirit of growth that itsflesh is the soilrdquo From (69)

2 A Koch et al Soil security Solving the global soil crisisGlob Policy 4 434ndash441 (2013) doi 1011111758-589912096

3 H Jenny Factors of Soil Formation A System of QuantitativePedology (McGraw-Hill New York 1941)

4 G Sposito et al Surface geochemistry of the clay mineralsProc Natl Acad Sci USA 96 3358ndash3364 (1999)doi 101073pnas9673358 pmid 10097044

5 G Sposito Green water and global food securityVadose Zone J 12 0 (2013) doi 102136vzj2013020041

6 R D Bardgett W H van der Putten Belowground biodiversityand ecosystem functioning Nature 515 505ndash511 (2014)doi 101038nature13855 pmid 25428498

7 E G Jobbaacutegy R B Jackson The vertical distribution of soilorganic carbon and its relation to climate and vegetation EcolAppl 10 423ndash436 (2000) doi 1018901051-0761(2000)010[0423TVDOSO]20CO2

8 D H Wall R D Bardgett E Kelly Biodiversity in the darkNat Geosci 3 297ndash298 (2010) doi 101038ngeo860

9 R Amundson Y Guo P Gong Soil diversity and landuse in theUnited States Ecosystems 6 470ndash482 (2003) doi 101007s10021-002-0160-2

10 Y Guo P Gong R Amundson Pedodiversity in the UnitedStates Geoderma 117 99ndash115 (2003) doi 101016S0016-7061(03)00137-X

11 J Diamond Evolution consequences and future of plantand animal domestication Nature 418 700ndash707 (2002)doi 101038nature01019 pmid 12167878

12 J A Foley et al Solutions for a cultivated planet Nature 478337ndash342 (2011) doi 101038nature10452 pmid 21993620

13 M Tennesen Rare earth Science 346 692ndash695 (2014)doi 101126science3466210692 pmid 25378604

14 P R Ehrlich E Wilson Biodiversity studies Science andpolicy Science 253 758ndash762 (1991) doi 101126science2535021758 pmid 17835492

15 FAOSTAT production httpfaostatfaoorgsite613DesktopDefaultaspxPageID=613ancor

16 K C Seto M Fragkias B Guumlneralp M K Reilly A meta-analysis of global urban land expansion PLOS ONE 6 e23777(2011) doi 101371journalpone0023777 pmid 21876770

17 E T Sundquist The global carbon dioxide budget Science259 934ndash941 (1993) doi 101126science2595097934

18 L B Guo R M Gifford Soil carbon stocks and land usechange A meta analysis Glob Change Biol 8 345ndash360(2002) doi 101046j1354-1013200200486x

19 U Stockmann et al The knowns known unknowns andunknowns of sequestration of soil organic carbon Agric EcosystEnviron 164 80ndash99 (2013) doi 101016jagee201210001

20 R Amundson The carbon budget in soils Annu RevEarth Planet Sci 29 535ndash562 (2001) doi 101146annurevearth291535

21 W F Ruddiman The Anthropocene Annu Rev EarthPlanet Sci 41 45ndash68 (2013) doi 101146annurev-earth-050212-123944

22 W F Ruddiman The early anthropogenic hypothesisChallenges and responses Rev Geophys 45 RG4001 (2007)

23 P Cias et al in Climate Change 2013 The Physical BasisContributions of Working Group 1 to the Fifth AssessmentReport of the Intergovernmental Panel on Climate Change(Cambridge Univ Press Cambridge 2013) pp 465ndash570

24 R B Jackson W H Schlesinger Curbing the US carbondeficit Proc Natl Acad Sci USA 101 15827ndash15829 (2004)doi 101073pnas0403631101 pmid 15514026

25 P Smith An overview of the permanence of soil organiccarbon stocks Influence of direct human-induced indirectand natural effects Eur J Soil Sci 56 673ndash680 (2005)doi 101111j1365-2389200500708x

26 P Smith et al ldquoAgriculturerdquo in Climate Change 2007Mitigation Contribution of Working Group III to the FourthAssessment Report on the International Panel of ClimateChange (Cambridge Univ Press Cambridge 2007)

27 P Smith et al Greenhouse gas mitigation in agriculturePhilos Trans R Soc London Ser B 363 789ndash813 (2008)doi 101098rstb20072184 pmid 17827109

28 Center for Sustainability and the Global Environment(SAGE) Nelson Institute for Environmental Studies atthe University of Wisconsin - Madison The Atlas of theBiosphere wwwsagewisceduatlasmapsphpdatasetid=21ampincluderelatedlinks=1ampdataset=21

29 W M Post W R Emanuel P J Zinke A G StangenbergerSoil carbon pools and world life zones Nature 298 156ndash159(1982) doi 101038298156a0

30 P Friedlingstein et al Climate-carbon cycle feedback analysisResults from the C4MIP model intercomparison J Clim 193337ndash3353 (2006) doi 101175JCLI38001

31 K E O Todd-Brown et al Changes in soil organic carbonstorage predicted by Earth system models during the 21stcentury Biogeosciences 11 2341ndash2356 (2014) doi 105194bg-11-2341-2014

32 E J Burke I P Hartley C D Jones Uncertainties in globaltemperature change caused by carbon release from

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-5

RESEARCH | REVIEW

Organic C stored in soil is the balance betweenplant inputs andmicrobiallymediatedmetaboliclosses as CO2 (Fig 2) In unperturbed conditionssoils achieve steady-state C pools on time scalesof centuries to a few millennia The total store ofsoil organic carbon is still uncertain but recentestimates suggest pools on the order of 2300 giga-tons (Gt) in the upper 3 m (7) Soil cultivationand clearing has caused a major fraction of totalanthropogenic greenhouse gas emissions since the19th century (17) Cultivation is amajor disruptionto the natural C balance in soil one that altersthe physical and biological structure of soil effec-tively igniting throughmicrobially mediated pro-cesses a vast store of labile C that has accumulatedover millennia (18) During the first few decadesthat soil is cultivated up to 50 of the carbonpool is oxidized to CO2 eventually a quasi-steady-state soil C pool is achieved (19) Based on theglobal agricultural land area cultivation has like-ly released between 50 and 70 Gt of C to the at-

mosphere over the course of human history (20)and the combined cultivation and biomass burn-ing contributions to atmospheric CO2 exceededthat of fossil fuel emissions well into the 20thcentury (17) However the agricultural imprinton atmospheric greenhouse gas concentrationsappearedmuch earlier in the Holocene (21) Earlyspikes in atmospheric CO2 and CH4 correspondedto agricultural expansion inMesopotamia and inChina (22) Much of the historical C loss was fromthe soils of forests and grasslands of the northernlatitudesHowever today the locus of land alterationhas shifted to the equatorial latitudes and up to10of global anthropogenic CO2 emissions are froma combination of biomass burning and soil culti-vation in the humid and subhumid tropics (23)Under changed management or through land

abandonment global agricultural soils have thecapacity to reapproach their original C storageand regain up to a half a decade of present fossilfuel emissions (over a multidecade period) Bet-

ter stewardship of domesticated soils that leadsto higher organic matter contents is a valuablepractice from an ecological perspective and froman agronomic point of view (24) There is now alarge body of research on the rates of C seques-tration under differing management practicesHowever there are limits to these practices as ameans of mitigating continued fossil fuel emis-sions First a serious concern withmanagement-based soil C sequestration strategies is that theyare dependent on restricted management op-tions in a highly decentralized and economicallydriven agricultural sector (25) A change in landownership or a change in factors driving agricul-tural practices can rapidly releasemuch of regainedC Second the effectiveness of soil C sequestrationis time dependent For example if all potentialsoil sequestration strategies were establishedthey would initially serve as a sink of about 13 Gtof C yearndash1 (Table 1) (26) but this sink termwould be expected to decline nonlinearly to low

1261071-2 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 1 The global patterns of soil properties and the human imprint onthem Expanding urban centers (55) remove soil from other uses whereasthe creation of cropland (56) is one of the major drivers of imbalances in thesoil carbon cycle (57) and accelerated rates of soil erosion (58) Phosphorus acritical plant nutrient is unevenly distributed and inherently low in warm andhumid climates (59)

RESEARCH | REVIEW

sequestration rates over a period of several dec-ades as a new soil C steady state is reached (27)Equally as important is the difficulty of actuallyachieving thismaximum potential which involvesmultiple governments and millions of individualland managers Last soil management effects onthe global C balance are inherently small relativeto the climate-driven changes to soil C storagethat will occur in this centurySoil C storage is well documented to decline

with increasing temperature and decreasing soilmoisture and soil C storage patterns mirror glob-al climate zones (28 29) with secondary impacts

by bedrock topography and soil age (20) Soil Cpools are the balance between plant inputs andmicrobial decomposition (Fig 3) and the responsesof these processes to anthropogenic climate changeare considered to be large (in the case of inputs)poorly constrained (30) and complicated by tem-perature and moisture interactions Anthropo-genic increases in atmospheric CO2 may driveincreased net primary production (NPP) as longas nutrient and water limitations do not occur(31) which ultimately may have a negative feed-back on atmospheric CO2 through increased in-puts to soil C (Fig 3) On the other hand increasing

air temperatures warm soil melt permafrost andstimulate biological metabolism of soil carbonpools driving what appears to be a large positivefeedback process (32) (Figs 1C and 3) Based oncurrent earth system modeling additions of soilC by increased NPP (relative to an 1850 BCE ref-erence date) are projected to be between 160 and1230 Gt by 2100 whereas C losses by increaseddecomposition are projected to be between 104and 629 Gt (31) Overall models suggest net soilC changes from a loss of 72 Gt to gains of 253 Gtby 2100 (31) However such exercises include greatuncertainties in both projected gains (by CO2-enhancedphotosynthesis) and losses (by soilwarm-ing) and inassumptions about long-termecosystemresponse to ever-increasing CO2 concentrationsOne important uncertainty is the response ofnorthern latitude soils to warming which couldresult in net soil C losses between 50 to 150 Gt(32 33) Last the current generation of earthsystemmodels has difficulty inmatching present-day soil carbon storage patterns (34) and tuningthe models is challenged by empirical uncertain-ties in the global soil C pool of more than 770 Gt(34) an uncertainty similar in size to the presentatmospheric C poolStill debated is the impact of soil erosion on

the global C cycle When agricultural soil is lostby water or wind erosion the surficial and mostC-rich material is preferentially removed whichaccelerates the decline in the soil C pool Rates ofsoil C replacement by crops and plants are rapidenough in certain situations to maintain soil Clevels at a steady state under the condition ofconstant erosionmdasheg creating an ongoing sink(35) This sink represents a net reduction inatmospheric CO2 only if the eroded C is not re-oxidized Because some depositional environmentsare conducive to partial preservation of buriedC (lakes reservoirs basins floodplains) the neteffect of accelerated agricultural erosionwas firstsuggested to be a global C sink of 06 to 15 Gtyearminus1 a rate similar to the total global land sink(35) If the eroded C is largely oxidized howeverit may result in no net sink (or possibly even anet source) (36) The most recent estimates sug-gest that agricultural erosion of soil C may be040 T020 Gt of C yearndash1 (37) If we benefit froman unintended C sink due to soil erosion anybenefits must be clearly balanced against therelated losses of nutrients and reduction of en-vironmental quality that require fossil fuel energyto remediate (38)The global soil C cycle has been greatly per-

turbed by human activity both directly throughfarming and indirectly through anthropogenicclimate change All projected soil C gains andlosses this century are highly uncertain becauseof economic population and political influences(which will largely affect carbon sequestrationefforts) and uncertainties in the magnitude of thesoil response to warming (because of the com-plexity of the soil C pool structure) (Fig 3) Hu-man changes to the global atmosphere andclimate are likely to simultaneously drive bothvery large gains and losses of soil Cmdashfluxes thatare equivalent to decades of emissions at present

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-3

Table 1 Published estimates of soil C sources and sinks for the 21st century

Management TypeMaximum Flux(Gt of C yearminus1)

ReferenceCumulative Flux

(Gt)

Carbon SinksIncreased net primary production (31) 160 to 1230Erosion 040 (37) 40

ManagementCropland management 036 (26 27)Grazing land management 037 (26 27)Restore degraded land 018 (26 27)Restore organic soils 036 (26 27)Total for management 126 164

Carbon SourcesLand clearingdagger 250Soil warmingBoreal regions (32) 50ndash270Globe (34) 104ndash629Total

Net balance ndash188 to +137Based on maximum new cropland by 2050 (10 billion ha) (68) and assumed loss of 25 of an average Ccontent of 10 kg mminus2 (29) daggerCalculated assuming exponential decline to an ultimate landscapesaturation after 50 years

Fig 2 Changes in the balance of important soil processes caused by human disturbance Manysoil characteristics are the balance or the result of a number of processes that respond to changes inenvironmental variables (3) However properties such as hillslope soil thickness organic carbon storageN content and other features attain steady state in intervals of a few centuries to millennia and appearcapable of regaining stability Human intervention in soil processes many times exceeds natural per-turbations and thus exceeds the resiliency of soil to recover to its original condition Viewed broadlysteady state is a quantitative measure of soil sustainability

RESEARCH | REVIEW

rates of fossil fuel consumption The presentlyunknown balance andmost importantly its signbetween the large fluxes represent considerableuncertainty for climate security (Table 1)

Soil and Food Security

The late 20th and early 21st centuries have beenfor industrialized countries an unprecedentedera of increasingly low food prices (39) There arenumerous factors that may reverse this trendmdashsuch as increased global demand climate change(40) and competition for soil by nonagriculturaluses such as biofuels or urbanization Abundantenergy has been the key driving force behind ourability to maintain food production apace withan expanding population that is estimated to reach11 billion by 2100 (41) Low-cost energy whichled to advanced agriculturalmachinery replacinghuman labor is causing migration to urban cen-ters Energy is used to replace the soil nutrientsremoved or lost by agricultural perturbations ofsoil Energy transforms atmospheric N2 to thebioavailable NH4 fertilizer through the energy-intensive Haber-Bosch processmdashconstituting thefirst and most important green revolution (42)(Fig 4A) one that allows us to feed the increas-ing global population (Fig 4A) Before the indus-trial fixation of N any increase in food productionfor a given country was largely due to increasedsoil used for production (43) and only after theadvent of N fertilizer (Fig 4A) did yields per areaof major crops begin their upward trajectory (43)

Last energy is essential to mine and transportessential plant nutrients such as P andK that canonly be accessed from limited geological reservoirsAgricultural soil erosion is one of the most de-

structive human perturbations to soil sustain-ability Given little opportunity or desirability forfurther agricultural expansion stewardship of ourexisting domesticated soil is essential for sus-

tained human prosperity Yet despite the impor-tance of soil conservation the implementation ofpractices to minimize soil erosion has not fol-lowed apace with the severity of the problem Themost pervasive mechanism of soil erosion is viawater Before European contact and the removalof native vegetation by plowing and cultivationthe geologicalmechanism of soil erosion onmost

1261071-4 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 3 A cause-effect diagram of the major soil-atmospheric CO2 feed-back processesThe values in parentheses in the circles are the approximatepool sizes (Gt) of C A solid arrow represents a direct response (eg as CO2

increases temperature increases) an arrow with a circle indicates an inverseresponse (eg as temperature increases soil carbon storagedecreases) Dashedlines are for processes that are less well understood (A) Environments withadequate moisture (B) Water-limited environments In (A) the a-b-c loop is apositive feedback process (even number of inverse relationship arrows) oneespecially important in regions of melting permafrostThe d-f-c loop is a negativefeedback one with less certain feedbacks between vegetation and soil (f-g loop)and temperature (h-i)The strength of the a-b-c- versus d-f-c loops on soil carbonpools will likely determine whether soil carbon losses in northern latitudes serve

as amajor source of CO2 and CH4 this century a balance that also hinges on theabilityof the soils to supply nutrients to plants (arrowg) in order to respond to theincreases in CO2 (arrow d) In (B) regions with limitedmoisture the strengths ofvegetation response toCO2 (d) and soil carbon response to temperature (b)maybe weakened (thinner arrow) In addition the vegetation response to increasingtemperature may become negative These figures reveal the importance of soilcarbon to the global CO2 balance this century as well as the uncertainties in thestrength and direction of important processes Arrow references are as follows a(60) b (32 33) c CO2 loss by respiration is the overwhelming pathway of Cremoval from soils d water efficiency response in (61) f soil C is the balancebetween plant inputs and decomposition losses g not well constrained but seediscussion in (62) h (63) i eg (64)

Fig 4 The postndashWorld War II rise in fertilizer production and cost coincides with the spike inglobal human populations The growth in world population in the late 20th century (65) mirrors theincreasing use of industrially derived N fertilizer Before N fertilization formajor US crops (66)wheat andother grain yields per acre increased only following World War II coincident with the rise in the use of Nfertilizer produced by the Haber-Bosch process Prices have sharply risen (67) because global demand isstraining the supplies

RESEARCH | REVIEW

uplands was by slow biologically driven creep(44) The removal of plant cover allows themecha-nism of erosion to change to raindrop dislodg-ment of soil particles and their subsequent removalbyoverland flowwhich is a farmore rapiderosionalprocess For example recent analyses of USerosion rates before European contact place theaverage rate at about 21mMyndash1 (Mymillion years)(45) Today erosion rates in the central UnitedStates can exceed 2000 mMyndash1 (45) whereas therates of soil erosion in portions of the loess pla-teau of China approach 10000 m Myndash1 (46) (Fig1D) This eroded sediment is ultimately replacedby the conversion of the underlying sediment orrock into new soils with the addition of organicmatter and nutrients through biological mecha-nisms Until the past decade the pace of this re-placement process was poorly known and theacceptable rates of soil erosion on agriculturallands were placed at 400mMyndash1 or more (47) Nu-merous studies of natural rates of soil productionnow suggest rates between 50 to 200 mMyndash1 formany environments indicating that the pace oferosion in numerous agricultural areas is or un-til recently has been unsustainable (47) Not onlydoes the loss of soil remove nutrients from thesite of agricultural production (38) but the sedi-ment generated adversely affects local drainageswater bodies and regional aquatic ecology Lastthe maintenance (or even the improvement) ofagricultural production in the face of acceleratedrates of soil erosion is energy intensive Althoughmicrobial symbiosiswithplants can fix atmosphericN to bioavailable forms and can substitute for Nfixed by the Haber-Bosch process there is no bi-ological or atmospheric source for rock-derivednutrients such as P K and CaAlthough natural processes of soil production

and formation replace or release nutrients thepaces of these processes are slow relative to ouranthropogenic use rate (Fig 2) The transport ofcrops from the site of production to other loca-tions remove plant-essential nutrients from thesoil potentially causing deficiencies that limitpotential production levels (48) This furtherdrives a dependence on the mining and distri-bution of macronutrients from geological sourceswhich can create economic inequalities or geo-political conflicts between nations (49) (Fig 1E)The growing demand for P has recently causedan increase in the cost of rock phosphate fromabout $80 per US ton in 1961 to up to $450 perton in 2008 Prices since then have fluctuatedbut are now at about $700 per ton (50) (Fig 4)In addition to cost is the related issue of accessMorocco is estimated to have the worldrsquos largestgeological P reserves much of it in disputed ter-ritory (49) The United States on the other handhas only about 2 of global P resources (51) Atcurrent rates of retrieval the most productivemine in the United States will be depleted in20 years (49) which will force it to become in-creasingly reliant on imports to sustain its agri-cultural and industrial sectorsBecause most other P-reliant countries lack the

geological resources to indefinitely sustain cur-rent use the only means to confront the decline

in reserves (other than conversion fromdomesticto imported P) is to develop amore coherent andintegrated program of P (and other nutrient) re-cycling The loss of nutrients in our human andanimal waste streams is environmentally damag-ing and economically problematic Regaining con-trol of these resources now largely consideredwaste would go far toward substantially lower-ing the demand for imported nutrients and otherresources (52) In addition to P other soil nu-trients appear to be entering periods of limitationor high demand (Fig 1F) For example K (potash)prices were ~$875 per metric ton in 2009 and areexpected to reach $1500 by 2020

The 21st Century Challenge

Humans have domesticated our soil resourcesand the planet (12 53) This domestication hasin turn perturbed a number of soil cycles suchthat they are no longer in balance and the im-balance is changing soil in ways that will affectfuture generations and their climate (Fig 2) Soilmanagement must be geared toward passing ahabitable albeit highly altered landscape to thegenerations that followmdashone where our exploita-tion of and impacts on soil resources is adjustedto the pace of our planetrsquos renewal These strat-egies should focus on regaining a balance in (i)organic C inputs and losses (ii) soil erosion andproduction and (iii) release and loss of nutri-ents Soil sustainabilitymdashbased on quantitativeprinciples andmeasurements of soil erosion andproduction soil nutrient loss and release and soilcarbon loss and returnmdashmust be the ultimate goalfor managing the global soil resource and shouldserve as the driving principle for soil researchthat will support this managementThese are challenging goals that will be diffi-

cult to achieve The solutions will require an ef-fort commensurate with the magnitude of theproblems First effective solutions to soil sustain-ability much like the approaches required to con-tendwith climate change (54) must involve highlymultidisciplinary research in novel intellectualsettings or institutions Second the ultimate suc-cess of any innovation requires a dialog and in-terfacewith policymakers and public institutionstheultimate ldquodecidersrdquo in broad-scale social changeThese linked efforts will depend on continuedand arguablymuch greater investments in knowl-edge and innovative knowledge transfer and simplydifferent ways of conceptualizing and approachingproblems From our vantage point the future ofEarthrsquos soil resources is tenuously in our controlor within our ability to sustain it into the futureOnly those on Earth in 2100 will know how wellwe succeeded

REFERENCES AND NOTES

1 ldquoWe might say that the earth has a spirit of growth that itsflesh is the soilrdquo From (69)

2 A Koch et al Soil security Solving the global soil crisisGlob Policy 4 434ndash441 (2013) doi 1011111758-589912096

3 H Jenny Factors of Soil Formation A System of QuantitativePedology (McGraw-Hill New York 1941)

4 G Sposito et al Surface geochemistry of the clay mineralsProc Natl Acad Sci USA 96 3358ndash3364 (1999)doi 101073pnas9673358 pmid 10097044

5 G Sposito Green water and global food securityVadose Zone J 12 0 (2013) doi 102136vzj2013020041

6 R D Bardgett W H van der Putten Belowground biodiversityand ecosystem functioning Nature 515 505ndash511 (2014)doi 101038nature13855 pmid 25428498

7 E G Jobbaacutegy R B Jackson The vertical distribution of soilorganic carbon and its relation to climate and vegetation EcolAppl 10 423ndash436 (2000) doi 1018901051-0761(2000)010[0423TVDOSO]20CO2

8 D H Wall R D Bardgett E Kelly Biodiversity in the darkNat Geosci 3 297ndash298 (2010) doi 101038ngeo860

9 R Amundson Y Guo P Gong Soil diversity and landuse in theUnited States Ecosystems 6 470ndash482 (2003) doi 101007s10021-002-0160-2

10 Y Guo P Gong R Amundson Pedodiversity in the UnitedStates Geoderma 117 99ndash115 (2003) doi 101016S0016-7061(03)00137-X

11 J Diamond Evolution consequences and future of plantand animal domestication Nature 418 700ndash707 (2002)doi 101038nature01019 pmid 12167878

12 J A Foley et al Solutions for a cultivated planet Nature 478337ndash342 (2011) doi 101038nature10452 pmid 21993620

13 M Tennesen Rare earth Science 346 692ndash695 (2014)doi 101126science3466210692 pmid 25378604

14 P R Ehrlich E Wilson Biodiversity studies Science andpolicy Science 253 758ndash762 (1991) doi 101126science2535021758 pmid 17835492

15 FAOSTAT production httpfaostatfaoorgsite613DesktopDefaultaspxPageID=613ancor

16 K C Seto M Fragkias B Guumlneralp M K Reilly A meta-analysis of global urban land expansion PLOS ONE 6 e23777(2011) doi 101371journalpone0023777 pmid 21876770

17 E T Sundquist The global carbon dioxide budget Science259 934ndash941 (1993) doi 101126science2595097934

18 L B Guo R M Gifford Soil carbon stocks and land usechange A meta analysis Glob Change Biol 8 345ndash360(2002) doi 101046j1354-1013200200486x

19 U Stockmann et al The knowns known unknowns andunknowns of sequestration of soil organic carbon Agric EcosystEnviron 164 80ndash99 (2013) doi 101016jagee201210001

20 R Amundson The carbon budget in soils Annu RevEarth Planet Sci 29 535ndash562 (2001) doi 101146annurevearth291535

21 W F Ruddiman The Anthropocene Annu Rev EarthPlanet Sci 41 45ndash68 (2013) doi 101146annurev-earth-050212-123944

22 W F Ruddiman The early anthropogenic hypothesisChallenges and responses Rev Geophys 45 RG4001 (2007)

23 P Cias et al in Climate Change 2013 The Physical BasisContributions of Working Group 1 to the Fifth AssessmentReport of the Intergovernmental Panel on Climate Change(Cambridge Univ Press Cambridge 2013) pp 465ndash570

24 R B Jackson W H Schlesinger Curbing the US carbondeficit Proc Natl Acad Sci USA 101 15827ndash15829 (2004)doi 101073pnas0403631101 pmid 15514026

25 P Smith An overview of the permanence of soil organiccarbon stocks Influence of direct human-induced indirectand natural effects Eur J Soil Sci 56 673ndash680 (2005)doi 101111j1365-2389200500708x

26 P Smith et al ldquoAgriculturerdquo in Climate Change 2007Mitigation Contribution of Working Group III to the FourthAssessment Report on the International Panel of ClimateChange (Cambridge Univ Press Cambridge 2007)

27 P Smith et al Greenhouse gas mitigation in agriculturePhilos Trans R Soc London Ser B 363 789ndash813 (2008)doi 101098rstb20072184 pmid 17827109

28 Center for Sustainability and the Global Environment(SAGE) Nelson Institute for Environmental Studies atthe University of Wisconsin - Madison The Atlas of theBiosphere wwwsagewisceduatlasmapsphpdatasetid=21ampincluderelatedlinks=1ampdataset=21

29 W M Post W R Emanuel P J Zinke A G StangenbergerSoil carbon pools and world life zones Nature 298 156ndash159(1982) doi 101038298156a0

30 P Friedlingstein et al Climate-carbon cycle feedback analysisResults from the C4MIP model intercomparison J Clim 193337ndash3353 (2006) doi 101175JCLI38001

31 K E O Todd-Brown et al Changes in soil organic carbonstorage predicted by Earth system models during the 21stcentury Biogeosciences 11 2341ndash2356 (2014) doi 105194bg-11-2341-2014

32 E J Burke I P Hartley C D Jones Uncertainties in globaltemperature change caused by carbon release from

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-5

RESEARCH | REVIEW

sequestration rates over a period of several dec-ades as a new soil C steady state is reached (27)Equally as important is the difficulty of actuallyachieving thismaximum potential which involvesmultiple governments and millions of individualland managers Last soil management effects onthe global C balance are inherently small relativeto the climate-driven changes to soil C storagethat will occur in this centurySoil C storage is well documented to decline

with increasing temperature and decreasing soilmoisture and soil C storage patterns mirror glob-al climate zones (28 29) with secondary impacts

by bedrock topography and soil age (20) Soil Cpools are the balance between plant inputs andmicrobial decomposition (Fig 3) and the responsesof these processes to anthropogenic climate changeare considered to be large (in the case of inputs)poorly constrained (30) and complicated by tem-perature and moisture interactions Anthropo-genic increases in atmospheric CO2 may driveincreased net primary production (NPP) as longas nutrient and water limitations do not occur(31) which ultimately may have a negative feed-back on atmospheric CO2 through increased in-puts to soil C (Fig 3) On the other hand increasing

air temperatures warm soil melt permafrost andstimulate biological metabolism of soil carbonpools driving what appears to be a large positivefeedback process (32) (Figs 1C and 3) Based oncurrent earth system modeling additions of soilC by increased NPP (relative to an 1850 BCE ref-erence date) are projected to be between 160 and1230 Gt by 2100 whereas C losses by increaseddecomposition are projected to be between 104and 629 Gt (31) Overall models suggest net soilC changes from a loss of 72 Gt to gains of 253 Gtby 2100 (31) However such exercises include greatuncertainties in both projected gains (by CO2-enhancedphotosynthesis) and losses (by soilwarm-ing) and inassumptions about long-termecosystemresponse to ever-increasing CO2 concentrationsOne important uncertainty is the response ofnorthern latitude soils to warming which couldresult in net soil C losses between 50 to 150 Gt(32 33) Last the current generation of earthsystemmodels has difficulty inmatching present-day soil carbon storage patterns (34) and tuningthe models is challenged by empirical uncertain-ties in the global soil C pool of more than 770 Gt(34) an uncertainty similar in size to the presentatmospheric C poolStill debated is the impact of soil erosion on

the global C cycle When agricultural soil is lostby water or wind erosion the surficial and mostC-rich material is preferentially removed whichaccelerates the decline in the soil C pool Rates ofsoil C replacement by crops and plants are rapidenough in certain situations to maintain soil Clevels at a steady state under the condition ofconstant erosionmdasheg creating an ongoing sink(35) This sink represents a net reduction inatmospheric CO2 only if the eroded C is not re-oxidized Because some depositional environmentsare conducive to partial preservation of buriedC (lakes reservoirs basins floodplains) the neteffect of accelerated agricultural erosionwas firstsuggested to be a global C sink of 06 to 15 Gtyearminus1 a rate similar to the total global land sink(35) If the eroded C is largely oxidized howeverit may result in no net sink (or possibly even anet source) (36) The most recent estimates sug-gest that agricultural erosion of soil C may be040 T020 Gt of C yearndash1 (37) If we benefit froman unintended C sink due to soil erosion anybenefits must be clearly balanced against therelated losses of nutrients and reduction of en-vironmental quality that require fossil fuel energyto remediate (38)The global soil C cycle has been greatly per-

turbed by human activity both directly throughfarming and indirectly through anthropogenicclimate change All projected soil C gains andlosses this century are highly uncertain becauseof economic population and political influences(which will largely affect carbon sequestrationefforts) and uncertainties in the magnitude of thesoil response to warming (because of the com-plexity of the soil C pool structure) (Fig 3) Hu-man changes to the global atmosphere andclimate are likely to simultaneously drive bothvery large gains and losses of soil Cmdashfluxes thatare equivalent to decades of emissions at present

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-3

Table 1 Published estimates of soil C sources and sinks for the 21st century

Management TypeMaximum Flux(Gt of C yearminus1)

ReferenceCumulative Flux

(Gt)

Carbon SinksIncreased net primary production (31) 160 to 1230Erosion 040 (37) 40

ManagementCropland management 036 (26 27)Grazing land management 037 (26 27)Restore degraded land 018 (26 27)Restore organic soils 036 (26 27)Total for management 126 164

Carbon SourcesLand clearingdagger 250Soil warmingBoreal regions (32) 50ndash270Globe (34) 104ndash629Total

Net balance ndash188 to +137Based on maximum new cropland by 2050 (10 billion ha) (68) and assumed loss of 25 of an average Ccontent of 10 kg mminus2 (29) daggerCalculated assuming exponential decline to an ultimate landscapesaturation after 50 years

Fig 2 Changes in the balance of important soil processes caused by human disturbance Manysoil characteristics are the balance or the result of a number of processes that respond to changes inenvironmental variables (3) However properties such as hillslope soil thickness organic carbon storageN content and other features attain steady state in intervals of a few centuries to millennia and appearcapable of regaining stability Human intervention in soil processes many times exceeds natural per-turbations and thus exceeds the resiliency of soil to recover to its original condition Viewed broadlysteady state is a quantitative measure of soil sustainability

RESEARCH | REVIEW

rates of fossil fuel consumption The presentlyunknown balance andmost importantly its signbetween the large fluxes represent considerableuncertainty for climate security (Table 1)

Soil and Food Security

The late 20th and early 21st centuries have beenfor industrialized countries an unprecedentedera of increasingly low food prices (39) There arenumerous factors that may reverse this trendmdashsuch as increased global demand climate change(40) and competition for soil by nonagriculturaluses such as biofuels or urbanization Abundantenergy has been the key driving force behind ourability to maintain food production apace withan expanding population that is estimated to reach11 billion by 2100 (41) Low-cost energy whichled to advanced agriculturalmachinery replacinghuman labor is causing migration to urban cen-ters Energy is used to replace the soil nutrientsremoved or lost by agricultural perturbations ofsoil Energy transforms atmospheric N2 to thebioavailable NH4 fertilizer through the energy-intensive Haber-Bosch processmdashconstituting thefirst and most important green revolution (42)(Fig 4A) one that allows us to feed the increas-ing global population (Fig 4A) Before the indus-trial fixation of N any increase in food productionfor a given country was largely due to increasedsoil used for production (43) and only after theadvent of N fertilizer (Fig 4A) did yields per areaof major crops begin their upward trajectory (43)

Last energy is essential to mine and transportessential plant nutrients such as P andK that canonly be accessed from limited geological reservoirsAgricultural soil erosion is one of the most de-

structive human perturbations to soil sustain-ability Given little opportunity or desirability forfurther agricultural expansion stewardship of ourexisting domesticated soil is essential for sus-

tained human prosperity Yet despite the impor-tance of soil conservation the implementation ofpractices to minimize soil erosion has not fol-lowed apace with the severity of the problem Themost pervasive mechanism of soil erosion is viawater Before European contact and the removalof native vegetation by plowing and cultivationthe geologicalmechanism of soil erosion onmost

1261071-4 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 3 A cause-effect diagram of the major soil-atmospheric CO2 feed-back processesThe values in parentheses in the circles are the approximatepool sizes (Gt) of C A solid arrow represents a direct response (eg as CO2

increases temperature increases) an arrow with a circle indicates an inverseresponse (eg as temperature increases soil carbon storagedecreases) Dashedlines are for processes that are less well understood (A) Environments withadequate moisture (B) Water-limited environments In (A) the a-b-c loop is apositive feedback process (even number of inverse relationship arrows) oneespecially important in regions of melting permafrostThe d-f-c loop is a negativefeedback one with less certain feedbacks between vegetation and soil (f-g loop)and temperature (h-i)The strength of the a-b-c- versus d-f-c loops on soil carbonpools will likely determine whether soil carbon losses in northern latitudes serve

as amajor source of CO2 and CH4 this century a balance that also hinges on theabilityof the soils to supply nutrients to plants (arrowg) in order to respond to theincreases in CO2 (arrow d) In (B) regions with limitedmoisture the strengths ofvegetation response toCO2 (d) and soil carbon response to temperature (b)maybe weakened (thinner arrow) In addition the vegetation response to increasingtemperature may become negative These figures reveal the importance of soilcarbon to the global CO2 balance this century as well as the uncertainties in thestrength and direction of important processes Arrow references are as follows a(60) b (32 33) c CO2 loss by respiration is the overwhelming pathway of Cremoval from soils d water efficiency response in (61) f soil C is the balancebetween plant inputs and decomposition losses g not well constrained but seediscussion in (62) h (63) i eg (64)

Fig 4 The postndashWorld War II rise in fertilizer production and cost coincides with the spike inglobal human populations The growth in world population in the late 20th century (65) mirrors theincreasing use of industrially derived N fertilizer Before N fertilization formajor US crops (66)wheat andother grain yields per acre increased only following World War II coincident with the rise in the use of Nfertilizer produced by the Haber-Bosch process Prices have sharply risen (67) because global demand isstraining the supplies

RESEARCH | REVIEW

uplands was by slow biologically driven creep(44) The removal of plant cover allows themecha-nism of erosion to change to raindrop dislodg-ment of soil particles and their subsequent removalbyoverland flowwhich is a farmore rapiderosionalprocess For example recent analyses of USerosion rates before European contact place theaverage rate at about 21mMyndash1 (Mymillion years)(45) Today erosion rates in the central UnitedStates can exceed 2000 mMyndash1 (45) whereas therates of soil erosion in portions of the loess pla-teau of China approach 10000 m Myndash1 (46) (Fig1D) This eroded sediment is ultimately replacedby the conversion of the underlying sediment orrock into new soils with the addition of organicmatter and nutrients through biological mecha-nisms Until the past decade the pace of this re-placement process was poorly known and theacceptable rates of soil erosion on agriculturallands were placed at 400mMyndash1 or more (47) Nu-merous studies of natural rates of soil productionnow suggest rates between 50 to 200 mMyndash1 formany environments indicating that the pace oferosion in numerous agricultural areas is or un-til recently has been unsustainable (47) Not onlydoes the loss of soil remove nutrients from thesite of agricultural production (38) but the sedi-ment generated adversely affects local drainageswater bodies and regional aquatic ecology Lastthe maintenance (or even the improvement) ofagricultural production in the face of acceleratedrates of soil erosion is energy intensive Althoughmicrobial symbiosiswithplants can fix atmosphericN to bioavailable forms and can substitute for Nfixed by the Haber-Bosch process there is no bi-ological or atmospheric source for rock-derivednutrients such as P K and CaAlthough natural processes of soil production

and formation replace or release nutrients thepaces of these processes are slow relative to ouranthropogenic use rate (Fig 2) The transport ofcrops from the site of production to other loca-tions remove plant-essential nutrients from thesoil potentially causing deficiencies that limitpotential production levels (48) This furtherdrives a dependence on the mining and distri-bution of macronutrients from geological sourceswhich can create economic inequalities or geo-political conflicts between nations (49) (Fig 1E)The growing demand for P has recently causedan increase in the cost of rock phosphate fromabout $80 per US ton in 1961 to up to $450 perton in 2008 Prices since then have fluctuatedbut are now at about $700 per ton (50) (Fig 4)In addition to cost is the related issue of accessMorocco is estimated to have the worldrsquos largestgeological P reserves much of it in disputed ter-ritory (49) The United States on the other handhas only about 2 of global P resources (51) Atcurrent rates of retrieval the most productivemine in the United States will be depleted in20 years (49) which will force it to become in-creasingly reliant on imports to sustain its agri-cultural and industrial sectorsBecause most other P-reliant countries lack the

geological resources to indefinitely sustain cur-rent use the only means to confront the decline

in reserves (other than conversion fromdomesticto imported P) is to develop amore coherent andintegrated program of P (and other nutrient) re-cycling The loss of nutrients in our human andanimal waste streams is environmentally damag-ing and economically problematic Regaining con-trol of these resources now largely consideredwaste would go far toward substantially lower-ing the demand for imported nutrients and otherresources (52) In addition to P other soil nu-trients appear to be entering periods of limitationor high demand (Fig 1F) For example K (potash)prices were ~$875 per metric ton in 2009 and areexpected to reach $1500 by 2020

The 21st Century Challenge

Humans have domesticated our soil resourcesand the planet (12 53) This domestication hasin turn perturbed a number of soil cycles suchthat they are no longer in balance and the im-balance is changing soil in ways that will affectfuture generations and their climate (Fig 2) Soilmanagement must be geared toward passing ahabitable albeit highly altered landscape to thegenerations that followmdashone where our exploita-tion of and impacts on soil resources is adjustedto the pace of our planetrsquos renewal These strat-egies should focus on regaining a balance in (i)organic C inputs and losses (ii) soil erosion andproduction and (iii) release and loss of nutri-ents Soil sustainabilitymdashbased on quantitativeprinciples andmeasurements of soil erosion andproduction soil nutrient loss and release and soilcarbon loss and returnmdashmust be the ultimate goalfor managing the global soil resource and shouldserve as the driving principle for soil researchthat will support this managementThese are challenging goals that will be diffi-

cult to achieve The solutions will require an ef-fort commensurate with the magnitude of theproblems First effective solutions to soil sustain-ability much like the approaches required to con-tendwith climate change (54) must involve highlymultidisciplinary research in novel intellectualsettings or institutions Second the ultimate suc-cess of any innovation requires a dialog and in-terfacewith policymakers and public institutionstheultimate ldquodecidersrdquo in broad-scale social changeThese linked efforts will depend on continuedand arguablymuch greater investments in knowl-edge and innovative knowledge transfer and simplydifferent ways of conceptualizing and approachingproblems From our vantage point the future ofEarthrsquos soil resources is tenuously in our controlor within our ability to sustain it into the futureOnly those on Earth in 2100 will know how wellwe succeeded

REFERENCES AND NOTES

1 ldquoWe might say that the earth has a spirit of growth that itsflesh is the soilrdquo From (69)

2 A Koch et al Soil security Solving the global soil crisisGlob Policy 4 434ndash441 (2013) doi 1011111758-589912096

3 H Jenny Factors of Soil Formation A System of QuantitativePedology (McGraw-Hill New York 1941)

4 G Sposito et al Surface geochemistry of the clay mineralsProc Natl Acad Sci USA 96 3358ndash3364 (1999)doi 101073pnas9673358 pmid 10097044

5 G Sposito Green water and global food securityVadose Zone J 12 0 (2013) doi 102136vzj2013020041

6 R D Bardgett W H van der Putten Belowground biodiversityand ecosystem functioning Nature 515 505ndash511 (2014)doi 101038nature13855 pmid 25428498

7 E G Jobbaacutegy R B Jackson The vertical distribution of soilorganic carbon and its relation to climate and vegetation EcolAppl 10 423ndash436 (2000) doi 1018901051-0761(2000)010[0423TVDOSO]20CO2

8 D H Wall R D Bardgett E Kelly Biodiversity in the darkNat Geosci 3 297ndash298 (2010) doi 101038ngeo860

9 R Amundson Y Guo P Gong Soil diversity and landuse in theUnited States Ecosystems 6 470ndash482 (2003) doi 101007s10021-002-0160-2

10 Y Guo P Gong R Amundson Pedodiversity in the UnitedStates Geoderma 117 99ndash115 (2003) doi 101016S0016-7061(03)00137-X

11 J Diamond Evolution consequences and future of plantand animal domestication Nature 418 700ndash707 (2002)doi 101038nature01019 pmid 12167878

12 J A Foley et al Solutions for a cultivated planet Nature 478337ndash342 (2011) doi 101038nature10452 pmid 21993620

13 M Tennesen Rare earth Science 346 692ndash695 (2014)doi 101126science3466210692 pmid 25378604

14 P R Ehrlich E Wilson Biodiversity studies Science andpolicy Science 253 758ndash762 (1991) doi 101126science2535021758 pmid 17835492

15 FAOSTAT production httpfaostatfaoorgsite613DesktopDefaultaspxPageID=613ancor

16 K C Seto M Fragkias B Guumlneralp M K Reilly A meta-analysis of global urban land expansion PLOS ONE 6 e23777(2011) doi 101371journalpone0023777 pmid 21876770

17 E T Sundquist The global carbon dioxide budget Science259 934ndash941 (1993) doi 101126science2595097934

18 L B Guo R M Gifford Soil carbon stocks and land usechange A meta analysis Glob Change Biol 8 345ndash360(2002) doi 101046j1354-1013200200486x

19 U Stockmann et al The knowns known unknowns andunknowns of sequestration of soil organic carbon Agric EcosystEnviron 164 80ndash99 (2013) doi 101016jagee201210001

20 R Amundson The carbon budget in soils Annu RevEarth Planet Sci 29 535ndash562 (2001) doi 101146annurevearth291535

21 W F Ruddiman The Anthropocene Annu Rev EarthPlanet Sci 41 45ndash68 (2013) doi 101146annurev-earth-050212-123944

22 W F Ruddiman The early anthropogenic hypothesisChallenges and responses Rev Geophys 45 RG4001 (2007)

23 P Cias et al in Climate Change 2013 The Physical BasisContributions of Working Group 1 to the Fifth AssessmentReport of the Intergovernmental Panel on Climate Change(Cambridge Univ Press Cambridge 2013) pp 465ndash570

24 R B Jackson W H Schlesinger Curbing the US carbondeficit Proc Natl Acad Sci USA 101 15827ndash15829 (2004)doi 101073pnas0403631101 pmid 15514026

25 P Smith An overview of the permanence of soil organiccarbon stocks Influence of direct human-induced indirectand natural effects Eur J Soil Sci 56 673ndash680 (2005)doi 101111j1365-2389200500708x

26 P Smith et al ldquoAgriculturerdquo in Climate Change 2007Mitigation Contribution of Working Group III to the FourthAssessment Report on the International Panel of ClimateChange (Cambridge Univ Press Cambridge 2007)

27 P Smith et al Greenhouse gas mitigation in agriculturePhilos Trans R Soc London Ser B 363 789ndash813 (2008)doi 101098rstb20072184 pmid 17827109

28 Center for Sustainability and the Global Environment(SAGE) Nelson Institute for Environmental Studies atthe University of Wisconsin - Madison The Atlas of theBiosphere wwwsagewisceduatlasmapsphpdatasetid=21ampincluderelatedlinks=1ampdataset=21

29 W M Post W R Emanuel P J Zinke A G StangenbergerSoil carbon pools and world life zones Nature 298 156ndash159(1982) doi 101038298156a0

30 P Friedlingstein et al Climate-carbon cycle feedback analysisResults from the C4MIP model intercomparison J Clim 193337ndash3353 (2006) doi 101175JCLI38001

31 K E O Todd-Brown et al Changes in soil organic carbonstorage predicted by Earth system models during the 21stcentury Biogeosciences 11 2341ndash2356 (2014) doi 105194bg-11-2341-2014

32 E J Burke I P Hartley C D Jones Uncertainties in globaltemperature change caused by carbon release from

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-5

RESEARCH | REVIEW

rates of fossil fuel consumption The presentlyunknown balance andmost importantly its signbetween the large fluxes represent considerableuncertainty for climate security (Table 1)

Soil and Food Security

The late 20th and early 21st centuries have beenfor industrialized countries an unprecedentedera of increasingly low food prices (39) There arenumerous factors that may reverse this trendmdashsuch as increased global demand climate change(40) and competition for soil by nonagriculturaluses such as biofuels or urbanization Abundantenergy has been the key driving force behind ourability to maintain food production apace withan expanding population that is estimated to reach11 billion by 2100 (41) Low-cost energy whichled to advanced agriculturalmachinery replacinghuman labor is causing migration to urban cen-ters Energy is used to replace the soil nutrientsremoved or lost by agricultural perturbations ofsoil Energy transforms atmospheric N2 to thebioavailable NH4 fertilizer through the energy-intensive Haber-Bosch processmdashconstituting thefirst and most important green revolution (42)(Fig 4A) one that allows us to feed the increas-ing global population (Fig 4A) Before the indus-trial fixation of N any increase in food productionfor a given country was largely due to increasedsoil used for production (43) and only after theadvent of N fertilizer (Fig 4A) did yields per areaof major crops begin their upward trajectory (43)

Last energy is essential to mine and transportessential plant nutrients such as P andK that canonly be accessed from limited geological reservoirsAgricultural soil erosion is one of the most de-

structive human perturbations to soil sustain-ability Given little opportunity or desirability forfurther agricultural expansion stewardship of ourexisting domesticated soil is essential for sus-

tained human prosperity Yet despite the impor-tance of soil conservation the implementation ofpractices to minimize soil erosion has not fol-lowed apace with the severity of the problem Themost pervasive mechanism of soil erosion is viawater Before European contact and the removalof native vegetation by plowing and cultivationthe geologicalmechanism of soil erosion onmost

1261071-4 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

Fig 3 A cause-effect diagram of the major soil-atmospheric CO2 feed-back processesThe values in parentheses in the circles are the approximatepool sizes (Gt) of C A solid arrow represents a direct response (eg as CO2

increases temperature increases) an arrow with a circle indicates an inverseresponse (eg as temperature increases soil carbon storagedecreases) Dashedlines are for processes that are less well understood (A) Environments withadequate moisture (B) Water-limited environments In (A) the a-b-c loop is apositive feedback process (even number of inverse relationship arrows) oneespecially important in regions of melting permafrostThe d-f-c loop is a negativefeedback one with less certain feedbacks between vegetation and soil (f-g loop)and temperature (h-i)The strength of the a-b-c- versus d-f-c loops on soil carbonpools will likely determine whether soil carbon losses in northern latitudes serve

as amajor source of CO2 and CH4 this century a balance that also hinges on theabilityof the soils to supply nutrients to plants (arrowg) in order to respond to theincreases in CO2 (arrow d) In (B) regions with limitedmoisture the strengths ofvegetation response toCO2 (d) and soil carbon response to temperature (b)maybe weakened (thinner arrow) In addition the vegetation response to increasingtemperature may become negative These figures reveal the importance of soilcarbon to the global CO2 balance this century as well as the uncertainties in thestrength and direction of important processes Arrow references are as follows a(60) b (32 33) c CO2 loss by respiration is the overwhelming pathway of Cremoval from soils d water efficiency response in (61) f soil C is the balancebetween plant inputs and decomposition losses g not well constrained but seediscussion in (62) h (63) i eg (64)

Fig 4 The postndashWorld War II rise in fertilizer production and cost coincides with the spike inglobal human populations The growth in world population in the late 20th century (65) mirrors theincreasing use of industrially derived N fertilizer Before N fertilization formajor US crops (66)wheat andother grain yields per acre increased only following World War II coincident with the rise in the use of Nfertilizer produced by the Haber-Bosch process Prices have sharply risen (67) because global demand isstraining the supplies

RESEARCH | REVIEW

uplands was by slow biologically driven creep(44) The removal of plant cover allows themecha-nism of erosion to change to raindrop dislodg-ment of soil particles and their subsequent removalbyoverland flowwhich is a farmore rapiderosionalprocess For example recent analyses of USerosion rates before European contact place theaverage rate at about 21mMyndash1 (Mymillion years)(45) Today erosion rates in the central UnitedStates can exceed 2000 mMyndash1 (45) whereas therates of soil erosion in portions of the loess pla-teau of China approach 10000 m Myndash1 (46) (Fig1D) This eroded sediment is ultimately replacedby the conversion of the underlying sediment orrock into new soils with the addition of organicmatter and nutrients through biological mecha-nisms Until the past decade the pace of this re-placement process was poorly known and theacceptable rates of soil erosion on agriculturallands were placed at 400mMyndash1 or more (47) Nu-merous studies of natural rates of soil productionnow suggest rates between 50 to 200 mMyndash1 formany environments indicating that the pace oferosion in numerous agricultural areas is or un-til recently has been unsustainable (47) Not onlydoes the loss of soil remove nutrients from thesite of agricultural production (38) but the sedi-ment generated adversely affects local drainageswater bodies and regional aquatic ecology Lastthe maintenance (or even the improvement) ofagricultural production in the face of acceleratedrates of soil erosion is energy intensive Althoughmicrobial symbiosiswithplants can fix atmosphericN to bioavailable forms and can substitute for Nfixed by the Haber-Bosch process there is no bi-ological or atmospheric source for rock-derivednutrients such as P K and CaAlthough natural processes of soil production

and formation replace or release nutrients thepaces of these processes are slow relative to ouranthropogenic use rate (Fig 2) The transport ofcrops from the site of production to other loca-tions remove plant-essential nutrients from thesoil potentially causing deficiencies that limitpotential production levels (48) This furtherdrives a dependence on the mining and distri-bution of macronutrients from geological sourceswhich can create economic inequalities or geo-political conflicts between nations (49) (Fig 1E)The growing demand for P has recently causedan increase in the cost of rock phosphate fromabout $80 per US ton in 1961 to up to $450 perton in 2008 Prices since then have fluctuatedbut are now at about $700 per ton (50) (Fig 4)In addition to cost is the related issue of accessMorocco is estimated to have the worldrsquos largestgeological P reserves much of it in disputed ter-ritory (49) The United States on the other handhas only about 2 of global P resources (51) Atcurrent rates of retrieval the most productivemine in the United States will be depleted in20 years (49) which will force it to become in-creasingly reliant on imports to sustain its agri-cultural and industrial sectorsBecause most other P-reliant countries lack the

geological resources to indefinitely sustain cur-rent use the only means to confront the decline

in reserves (other than conversion fromdomesticto imported P) is to develop amore coherent andintegrated program of P (and other nutrient) re-cycling The loss of nutrients in our human andanimal waste streams is environmentally damag-ing and economically problematic Regaining con-trol of these resources now largely consideredwaste would go far toward substantially lower-ing the demand for imported nutrients and otherresources (52) In addition to P other soil nu-trients appear to be entering periods of limitationor high demand (Fig 1F) For example K (potash)prices were ~$875 per metric ton in 2009 and areexpected to reach $1500 by 2020

The 21st Century Challenge

Humans have domesticated our soil resourcesand the planet (12 53) This domestication hasin turn perturbed a number of soil cycles suchthat they are no longer in balance and the im-balance is changing soil in ways that will affectfuture generations and their climate (Fig 2) Soilmanagement must be geared toward passing ahabitable albeit highly altered landscape to thegenerations that followmdashone where our exploita-tion of and impacts on soil resources is adjustedto the pace of our planetrsquos renewal These strat-egies should focus on regaining a balance in (i)organic C inputs and losses (ii) soil erosion andproduction and (iii) release and loss of nutri-ents Soil sustainabilitymdashbased on quantitativeprinciples andmeasurements of soil erosion andproduction soil nutrient loss and release and soilcarbon loss and returnmdashmust be the ultimate goalfor managing the global soil resource and shouldserve as the driving principle for soil researchthat will support this managementThese are challenging goals that will be diffi-

cult to achieve The solutions will require an ef-fort commensurate with the magnitude of theproblems First effective solutions to soil sustain-ability much like the approaches required to con-tendwith climate change (54) must involve highlymultidisciplinary research in novel intellectualsettings or institutions Second the ultimate suc-cess of any innovation requires a dialog and in-terfacewith policymakers and public institutionstheultimate ldquodecidersrdquo in broad-scale social changeThese linked efforts will depend on continuedand arguablymuch greater investments in knowl-edge and innovative knowledge transfer and simplydifferent ways of conceptualizing and approachingproblems From our vantage point the future ofEarthrsquos soil resources is tenuously in our controlor within our ability to sustain it into the futureOnly those on Earth in 2100 will know how wellwe succeeded

REFERENCES AND NOTES

1 ldquoWe might say that the earth has a spirit of growth that itsflesh is the soilrdquo From (69)

2 A Koch et al Soil security Solving the global soil crisisGlob Policy 4 434ndash441 (2013) doi 1011111758-589912096

3 H Jenny Factors of Soil Formation A System of QuantitativePedology (McGraw-Hill New York 1941)

4 G Sposito et al Surface geochemistry of the clay mineralsProc Natl Acad Sci USA 96 3358ndash3364 (1999)doi 101073pnas9673358 pmid 10097044

5 G Sposito Green water and global food securityVadose Zone J 12 0 (2013) doi 102136vzj2013020041

6 R D Bardgett W H van der Putten Belowground biodiversityand ecosystem functioning Nature 515 505ndash511 (2014)doi 101038nature13855 pmid 25428498

7 E G Jobbaacutegy R B Jackson The vertical distribution of soilorganic carbon and its relation to climate and vegetation EcolAppl 10 423ndash436 (2000) doi 1018901051-0761(2000)010[0423TVDOSO]20CO2

8 D H Wall R D Bardgett E Kelly Biodiversity in the darkNat Geosci 3 297ndash298 (2010) doi 101038ngeo860

9 R Amundson Y Guo P Gong Soil diversity and landuse in theUnited States Ecosystems 6 470ndash482 (2003) doi 101007s10021-002-0160-2

10 Y Guo P Gong R Amundson Pedodiversity in the UnitedStates Geoderma 117 99ndash115 (2003) doi 101016S0016-7061(03)00137-X

11 J Diamond Evolution consequences and future of plantand animal domestication Nature 418 700ndash707 (2002)doi 101038nature01019 pmid 12167878

12 J A Foley et al Solutions for a cultivated planet Nature 478337ndash342 (2011) doi 101038nature10452 pmid 21993620

13 M Tennesen Rare earth Science 346 692ndash695 (2014)doi 101126science3466210692 pmid 25378604

14 P R Ehrlich E Wilson Biodiversity studies Science andpolicy Science 253 758ndash762 (1991) doi 101126science2535021758 pmid 17835492

15 FAOSTAT production httpfaostatfaoorgsite613DesktopDefaultaspxPageID=613ancor

16 K C Seto M Fragkias B Guumlneralp M K Reilly A meta-analysis of global urban land expansion PLOS ONE 6 e23777(2011) doi 101371journalpone0023777 pmid 21876770

17 E T Sundquist The global carbon dioxide budget Science259 934ndash941 (1993) doi 101126science2595097934

18 L B Guo R M Gifford Soil carbon stocks and land usechange A meta analysis Glob Change Biol 8 345ndash360(2002) doi 101046j1354-1013200200486x

19 U Stockmann et al The knowns known unknowns andunknowns of sequestration of soil organic carbon Agric EcosystEnviron 164 80ndash99 (2013) doi 101016jagee201210001

20 R Amundson The carbon budget in soils Annu RevEarth Planet Sci 29 535ndash562 (2001) doi 101146annurevearth291535

21 W F Ruddiman The Anthropocene Annu Rev EarthPlanet Sci 41 45ndash68 (2013) doi 101146annurev-earth-050212-123944

22 W F Ruddiman The early anthropogenic hypothesisChallenges and responses Rev Geophys 45 RG4001 (2007)

23 P Cias et al in Climate Change 2013 The Physical BasisContributions of Working Group 1 to the Fifth AssessmentReport of the Intergovernmental Panel on Climate Change(Cambridge Univ Press Cambridge 2013) pp 465ndash570

24 R B Jackson W H Schlesinger Curbing the US carbondeficit Proc Natl Acad Sci USA 101 15827ndash15829 (2004)doi 101073pnas0403631101 pmid 15514026

25 P Smith An overview of the permanence of soil organiccarbon stocks Influence of direct human-induced indirectand natural effects Eur J Soil Sci 56 673ndash680 (2005)doi 101111j1365-2389200500708x

26 P Smith et al ldquoAgriculturerdquo in Climate Change 2007Mitigation Contribution of Working Group III to the FourthAssessment Report on the International Panel of ClimateChange (Cambridge Univ Press Cambridge 2007)

27 P Smith et al Greenhouse gas mitigation in agriculturePhilos Trans R Soc London Ser B 363 789ndash813 (2008)doi 101098rstb20072184 pmid 17827109

28 Center for Sustainability and the Global Environment(SAGE) Nelson Institute for Environmental Studies atthe University of Wisconsin - Madison The Atlas of theBiosphere wwwsagewisceduatlasmapsphpdatasetid=21ampincluderelatedlinks=1ampdataset=21

29 W M Post W R Emanuel P J Zinke A G StangenbergerSoil carbon pools and world life zones Nature 298 156ndash159(1982) doi 101038298156a0

30 P Friedlingstein et al Climate-carbon cycle feedback analysisResults from the C4MIP model intercomparison J Clim 193337ndash3353 (2006) doi 101175JCLI38001

31 K E O Todd-Brown et al Changes in soil organic carbonstorage predicted by Earth system models during the 21stcentury Biogeosciences 11 2341ndash2356 (2014) doi 105194bg-11-2341-2014

32 E J Burke I P Hartley C D Jones Uncertainties in globaltemperature change caused by carbon release from

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-5

RESEARCH | REVIEW

uplands was by slow biologically driven creep(44) The removal of plant cover allows themecha-nism of erosion to change to raindrop dislodg-ment of soil particles and their subsequent removalbyoverland flowwhich is a farmore rapiderosionalprocess For example recent analyses of USerosion rates before European contact place theaverage rate at about 21mMyndash1 (Mymillion years)(45) Today erosion rates in the central UnitedStates can exceed 2000 mMyndash1 (45) whereas therates of soil erosion in portions of the loess pla-teau of China approach 10000 m Myndash1 (46) (Fig1D) This eroded sediment is ultimately replacedby the conversion of the underlying sediment orrock into new soils with the addition of organicmatter and nutrients through biological mecha-nisms Until the past decade the pace of this re-placement process was poorly known and theacceptable rates of soil erosion on agriculturallands were placed at 400mMyndash1 or more (47) Nu-merous studies of natural rates of soil productionnow suggest rates between 50 to 200 mMyndash1 formany environments indicating that the pace oferosion in numerous agricultural areas is or un-til recently has been unsustainable (47) Not onlydoes the loss of soil remove nutrients from thesite of agricultural production (38) but the sedi-ment generated adversely affects local drainageswater bodies and regional aquatic ecology Lastthe maintenance (or even the improvement) ofagricultural production in the face of acceleratedrates of soil erosion is energy intensive Althoughmicrobial symbiosiswithplants can fix atmosphericN to bioavailable forms and can substitute for Nfixed by the Haber-Bosch process there is no bi-ological or atmospheric source for rock-derivednutrients such as P K and CaAlthough natural processes of soil production

and formation replace or release nutrients thepaces of these processes are slow relative to ouranthropogenic use rate (Fig 2) The transport ofcrops from the site of production to other loca-tions remove plant-essential nutrients from thesoil potentially causing deficiencies that limitpotential production levels (48) This furtherdrives a dependence on the mining and distri-bution of macronutrients from geological sourceswhich can create economic inequalities or geo-political conflicts between nations (49) (Fig 1E)The growing demand for P has recently causedan increase in the cost of rock phosphate fromabout $80 per US ton in 1961 to up to $450 perton in 2008 Prices since then have fluctuatedbut are now at about $700 per ton (50) (Fig 4)In addition to cost is the related issue of accessMorocco is estimated to have the worldrsquos largestgeological P reserves much of it in disputed ter-ritory (49) The United States on the other handhas only about 2 of global P resources (51) Atcurrent rates of retrieval the most productivemine in the United States will be depleted in20 years (49) which will force it to become in-creasingly reliant on imports to sustain its agri-cultural and industrial sectorsBecause most other P-reliant countries lack the

geological resources to indefinitely sustain cur-rent use the only means to confront the decline

in reserves (other than conversion fromdomesticto imported P) is to develop amore coherent andintegrated program of P (and other nutrient) re-cycling The loss of nutrients in our human andanimal waste streams is environmentally damag-ing and economically problematic Regaining con-trol of these resources now largely consideredwaste would go far toward substantially lower-ing the demand for imported nutrients and otherresources (52) In addition to P other soil nu-trients appear to be entering periods of limitationor high demand (Fig 1F) For example K (potash)prices were ~$875 per metric ton in 2009 and areexpected to reach $1500 by 2020

The 21st Century Challenge

Humans have domesticated our soil resourcesand the planet (12 53) This domestication hasin turn perturbed a number of soil cycles suchthat they are no longer in balance and the im-balance is changing soil in ways that will affectfuture generations and their climate (Fig 2) Soilmanagement must be geared toward passing ahabitable albeit highly altered landscape to thegenerations that followmdashone where our exploita-tion of and impacts on soil resources is adjustedto the pace of our planetrsquos renewal These strat-egies should focus on regaining a balance in (i)organic C inputs and losses (ii) soil erosion andproduction and (iii) release and loss of nutri-ents Soil sustainabilitymdashbased on quantitativeprinciples andmeasurements of soil erosion andproduction soil nutrient loss and release and soilcarbon loss and returnmdashmust be the ultimate goalfor managing the global soil resource and shouldserve as the driving principle for soil researchthat will support this managementThese are challenging goals that will be diffi-

cult to achieve The solutions will require an ef-fort commensurate with the magnitude of theproblems First effective solutions to soil sustain-ability much like the approaches required to con-tendwith climate change (54) must involve highlymultidisciplinary research in novel intellectualsettings or institutions Second the ultimate suc-cess of any innovation requires a dialog and in-terfacewith policymakers and public institutionstheultimate ldquodecidersrdquo in broad-scale social changeThese linked efforts will depend on continuedand arguablymuch greater investments in knowl-edge and innovative knowledge transfer and simplydifferent ways of conceptualizing and approachingproblems From our vantage point the future ofEarthrsquos soil resources is tenuously in our controlor within our ability to sustain it into the futureOnly those on Earth in 2100 will know how wellwe succeeded

REFERENCES AND NOTES

1 ldquoWe might say that the earth has a spirit of growth that itsflesh is the soilrdquo From (69)

2 A Koch et al Soil security Solving the global soil crisisGlob Policy 4 434ndash441 (2013) doi 1011111758-589912096

3 H Jenny Factors of Soil Formation A System of QuantitativePedology (McGraw-Hill New York 1941)

4 G Sposito et al Surface geochemistry of the clay mineralsProc Natl Acad Sci USA 96 3358ndash3364 (1999)doi 101073pnas9673358 pmid 10097044

5 G Sposito Green water and global food securityVadose Zone J 12 0 (2013) doi 102136vzj2013020041

6 R D Bardgett W H van der Putten Belowground biodiversityand ecosystem functioning Nature 515 505ndash511 (2014)doi 101038nature13855 pmid 25428498

7 E G Jobbaacutegy R B Jackson The vertical distribution of soilorganic carbon and its relation to climate and vegetation EcolAppl 10 423ndash436 (2000) doi 1018901051-0761(2000)010[0423TVDOSO]20CO2

8 D H Wall R D Bardgett E Kelly Biodiversity in the darkNat Geosci 3 297ndash298 (2010) doi 101038ngeo860

9 R Amundson Y Guo P Gong Soil diversity and landuse in theUnited States Ecosystems 6 470ndash482 (2003) doi 101007s10021-002-0160-2

10 Y Guo P Gong R Amundson Pedodiversity in the UnitedStates Geoderma 117 99ndash115 (2003) doi 101016S0016-7061(03)00137-X

11 J Diamond Evolution consequences and future of plantand animal domestication Nature 418 700ndash707 (2002)doi 101038nature01019 pmid 12167878

12 J A Foley et al Solutions for a cultivated planet Nature 478337ndash342 (2011) doi 101038nature10452 pmid 21993620

13 M Tennesen Rare earth Science 346 692ndash695 (2014)doi 101126science3466210692 pmid 25378604

14 P R Ehrlich E Wilson Biodiversity studies Science andpolicy Science 253 758ndash762 (1991) doi 101126science2535021758 pmid 17835492

15 FAOSTAT production httpfaostatfaoorgsite613DesktopDefaultaspxPageID=613ancor

16 K C Seto M Fragkias B Guumlneralp M K Reilly A meta-analysis of global urban land expansion PLOS ONE 6 e23777(2011) doi 101371journalpone0023777 pmid 21876770

17 E T Sundquist The global carbon dioxide budget Science259 934ndash941 (1993) doi 101126science2595097934

18 L B Guo R M Gifford Soil carbon stocks and land usechange A meta analysis Glob Change Biol 8 345ndash360(2002) doi 101046j1354-1013200200486x

19 U Stockmann et al The knowns known unknowns andunknowns of sequestration of soil organic carbon Agric EcosystEnviron 164 80ndash99 (2013) doi 101016jagee201210001

20 R Amundson The carbon budget in soils Annu RevEarth Planet Sci 29 535ndash562 (2001) doi 101146annurevearth291535

21 W F Ruddiman The Anthropocene Annu Rev EarthPlanet Sci 41 45ndash68 (2013) doi 101146annurev-earth-050212-123944

22 W F Ruddiman The early anthropogenic hypothesisChallenges and responses Rev Geophys 45 RG4001 (2007)

23 P Cias et al in Climate Change 2013 The Physical BasisContributions of Working Group 1 to the Fifth AssessmentReport of the Intergovernmental Panel on Climate Change(Cambridge Univ Press Cambridge 2013) pp 465ndash570

24 R B Jackson W H Schlesinger Curbing the US carbondeficit Proc Natl Acad Sci USA 101 15827ndash15829 (2004)doi 101073pnas0403631101 pmid 15514026

25 P Smith An overview of the permanence of soil organiccarbon stocks Influence of direct human-induced indirectand natural effects Eur J Soil Sci 56 673ndash680 (2005)doi 101111j1365-2389200500708x

26 P Smith et al ldquoAgriculturerdquo in Climate Change 2007Mitigation Contribution of Working Group III to the FourthAssessment Report on the International Panel of ClimateChange (Cambridge Univ Press Cambridge 2007)

27 P Smith et al Greenhouse gas mitigation in agriculturePhilos Trans R Soc London Ser B 363 789ndash813 (2008)doi 101098rstb20072184 pmid 17827109

28 Center for Sustainability and the Global Environment(SAGE) Nelson Institute for Environmental Studies atthe University of Wisconsin - Madison The Atlas of theBiosphere wwwsagewisceduatlasmapsphpdatasetid=21ampincluderelatedlinks=1ampdataset=21

29 W M Post W R Emanuel P J Zinke A G StangenbergerSoil carbon pools and world life zones Nature 298 156ndash159(1982) doi 101038298156a0

30 P Friedlingstein et al Climate-carbon cycle feedback analysisResults from the C4MIP model intercomparison J Clim 193337ndash3353 (2006) doi 101175JCLI38001

31 K E O Todd-Brown et al Changes in soil organic carbonstorage predicted by Earth system models during the 21stcentury Biogeosciences 11 2341ndash2356 (2014) doi 105194bg-11-2341-2014

32 E J Burke I P Hartley C D Jones Uncertainties in globaltemperature change caused by carbon release from

SCIENCE sciencemagorg 8 MAY 2015 bull VOL 348 ISSUE 6235 1261071-5

RESEARCH | REVIEW

permafrost thawing The Cryosphere 6 1063ndash1076(2012)

33 E A G Schuur et al Expert assessment of vulnerability ofpermafrost carbon to climate change Clim Change 119359ndash374 (2013) doi 101007s10584-013-0730-7

34 K E O Todd-Brown et al Causes of variation in soil carbonsimulations from CMIP5 Earth system models and comparisonwith observations Biogeosciences 10 1717ndash1736 (2013)doi 105194bg-10-1717-2013

35 R F Stallard Terrestrial sedimentation and the carbon cycleCoupling weathering and erosion to carbon burial GlobalBiogeochem Cycles 12 231ndash257 (1998) doi 10102998GB00741

36 R Lal et al Soil erosion A carbon sink or source Science 3191040ndash1042 (2008) doi 101126science31958661040pmid 18292324

37 S Doetterl K Van Oost J Six Towards constraining themagnitude of global agricultural sediment and soil organiccarbon fluxes Earth Surf Process Landf 37 642ndash655 (2012)doi 101002esp3198

38 J N Quinton G Govers K Van Oost R Bardgett The impactof agricultural erosion on biogeochemical cycling Nat Geosci3 311ndash314 (2010) doi 101038ngeo838

39 USDA Economic Research Service data for 2014 wwwersusdagovdata-productsfood-expendituresaspxU2vmj17lNZF

40 FAO FAO in the 21st Century Ensuring Food Security in aChanging World (Food and Agriculture Organization of theUnited Nations Rome 2011)

41 P Gerland et al World population stabilization unlikely thiscentury Science 346 234ndash237 (2014) doi 101126science1257469 pmid 25301627

42 V Smil Nitrogen cycle and world food production WorldAgriculture 2 9ndash13 (2011)

43 USDA Economic Research Service wwwersusdagovdata-productswheat-dataaspx

44 M J Kirkby Measurement and theory of soil creep J Geol 75359ndash378 (1967) doi 101086627267

45 B H Wilkinson B J McElroy The impact of humans oncontinental erosion and sedimentation Geol Soc Am Bull119 140ndash156 (2007) doi 101130B258991

46 W Sun Q Shao J Liu J Zhai Assessing the effects of landuse and topography on soil erosion in the Loess Plateau in

China Catena 121 151ndash163 (2014) doi 101016jcatena201405009

47 D R Montgomery Soil erosion and agricultural sustainabilityProc Natl Acad Sci USA 104 13268ndash13272 (2007)doi 101073pnas0611508104 pmid 17686990

48 D L Jones et al Nutrient stripping The global disparitybetween food security and soil nutrient stocks J Appl Ecol50 851ndash862 (2013) doi 1011111365-266412089

49 D Cordell J-O Drangert W White The story of phosphorusGlobal food security and food for thought Glob Environ Change19 292ndash305 (2009)

50 USDA Economic Research Service wwwersusdagovdata-productsfertilizeruse-and-priceaspx

51 httpmineralsusgsgovmineralspubscommodity52 J Elser E Bennett Phosphorus cycle A broken

biogeochemical cycle Nature 478 29ndash31 (2011) doi 101038478029a pmid 21979027

53 E C Ellis et al Used planet A global history Proc Natl AcadSci USA 110 7978ndash7985 (2013) doi 101073pnas1217241110 pmid 23630271

54 National Research Council ldquoAdvancing the science of climatechangerdquo in Americarsquos Climate Choices Panel on Advancing theScience of Climate Change (National Academies PressWashington DC 2010)

55 Center for International Earth Science Information Network(CIESIN) Columbia University International Food Policy ResearchInstitute (IFPRI) the World Bank and Centro Internacional deAgricultura Tropical (CIAT) Global Rural-Urban Mapping ProjectVersion 1 (GRUMPv1) Urban Extents Grid [Socioeconomic Dataand Applications Center (SEDAC) Columbia University PalisadesNY 2011] available at httpsedacciesincolumbiaedudatadatasetgrump-v1-urban-extents

56 N Ramankutty A T Evan C Monfreda J A Foley GlobalAgricultural Lands Pastures 2000 (2010) httpsedacciesincolumbiaeduesaglandshtml

57 P Gottschalk et al How will organic carbon stocks in mineralsoils evolve under future climate Global projections usingRothC for a range of climate change scenarios Biogeosciences9 3151ndash3171 (2012) doi 105194bg-9-3151-2012

58 L R Odelman R T A Hakkeling W G Sombroek GlobalAssessment of Soil Degradation GLASOD International SoilReference and Information Centre (1991)

59 Yang X WM Post PE Thornton A Jain 2014 GlobalGridded Soil Phosphorus Distribution Maps at 05-DegreeResolution (data set) doi httpdxdoiorg

60 IPCC Climate Change 2013 The Physical Science BasisContribution of Working Group I to the Fifth AssessmentReport of the Intergovernmental Panel on Climate ChangeT F Stocker et al Eds (Cambridge Univ Press Cambridge2013)

61 N Gedney et al Detection of a direct carbon dioxide effectin continental river runoff records Nature 439 835ndash838(2006) doi 101038nature04504 pmid 16482155

62 C D Koven et al The effect of vertically resolved soilbiogeochemistry and alternate soil C and N models on Cdynamics of CLM4 Biogeosciences 10 7109ndash7131 (2013)doi 105194bg-10-7109-2013

63 L Cao G Bala K Caldeira R Nemani G Ban-WeissImportance of carbon dioxide physiological forcing tofuture climate change Proc Natl Acad Sci USA 1079513ndash9518 (2010) doi 101073pnas0913000107pmid 20445083

64 H Lieth Primary production Terrestrial ecosystemsHum Ecol 1 303ndash332 (1973) doi 101007BF01536729

65 ldquoWorld Population Prospects The 2012 Revisionrdquo (XLS)Population Division of the Department of Economic andSocial Affairs of the United Nations Secretariat June 2013

66 wwwersusdagovdata-productswheat-dataaspxU2ktw17lNZE67 wwwersusdagovdata-productsfertilizer-use-and-price

aspxU2kz-l7lNZE68 H K Gibbs et al Tropical forests were the primary

sources of new agricultural land in the 1980s and 1990sProc Natl Acad Sci USA 107 16732ndash16737 (2010)doi 101073pnas0910275107 pmid 20807750

69 Jean Paul Richter Ed The Notebooks of Leonardo Da Vinci(Dover Publications New York 1970)

ACKNOWLEDGMENTS

The paper resulted from discussions within the US NationalCommittee for Soil Sciences P Bertsch Chair We thank I FungC Koven and G Sposito for discussions and P Gottschalk forfigure data

101126science1261071

1261071-6 8 MAY 2015 bull VOL 348 ISSUE 6235 sciencemagorg SCIENCE

RESEARCH | REVIEW

DOI 101126science1261071 (2015)348 Science

et alRonald AmundsonSoil and human security in the 21st century

This copy is for your personal non-commercial use only

clicking herecolleagues clients or customers by you can order high-quality copies for yourIf you wish to distribute this article to others

herefollowing the guidelines

can be obtained byPermission to republish or repurpose articles or portions of articles

) June 18 2015 wwwsciencemagorg (this information is current as of

The following resources related to this article are available online at

httpwwwsciencemagorgcontent34862351261071fullhtmlversion of this article at

including high-resolution figures can be found in the onlineUpdated information and services

httpwwwsciencemagorgcontent34862351261071fullhtmlref-list-1 14 of which can be accessed freecites 48 articlesThis article

httpwwwsciencemagorgcgicollectiongeochem_physGeochemistry Geophysics

subject collectionsThis article appears in the following

registered trademark of AAAS is aScience2015 by the American Association for the Advancement of Science all rights reserved The title

CopyrightAmerican Association for the Advancement of Science 1200 New York Avenue NW Washington DC 20005 (print ISSN 0036-8075 online ISSN 1095-9203) is published weekly except the last week in December by theScience

on

June

18

201

5w

ww

sci

ence

mag

org

Dow

nloa

ded

from