AR-024 species provide critical services to society (6), the role of biodiversity per se remains untested at the ecosystem level (1 4). We analyzed the Impacts of Biodiversity Loss on effects of changes in marine biodiversity on fundamental ecosystem services by combining available data from sources ranging from small Ocean Ecosystem Services scale experiments to global fisheries. 1 2 3 4 Experiments. We first used meta-analysis Boris Worm, * Edward B. Barbier, Nicola Beaumont, ). Emmett Duffy, 5 6 8 9 of published data to examine the effects of Carl Folke, • Benjamin S. Halpern/ jeremy B. C. ]ackson, • Heike K. Lotze/ Fiorenza Micheli/0 Stephen R. Palumbi/0 Enric Sala,8 Kimberley A. Selkoe/ variation in marine diversity (genetic or species John). Stachowicz,11 Reg Watson12 richness) on primary and secondary produc tivity, resource use, nutrient cycling, and eco system stability in 32 controlled experiments. Human-dominated marine ecosystems are experiencing accelerating loss of populations and species, with largely unknown consequences. We analyzed local experiments, long-term regional Such effects have been contentiously debated, time series, and global fisheries data to test how biodiversity loss affects marine ecosystem services particularly in the marine realm, where high across temporal and spatial scales. Overall, rates of resource collapse increased and recovery diversity and connectivity may blur any deter potential, stability, and water quality decreased exponentially with declining diversity. Restoration ministic effect of local biodiversity on eco of biodiversity, in contrast, increased productivity fourfold and decreased variability by 21%, on system functioning (1). Yet when the available average. We conclude that marine biodiversity loss is increasingly impairing the ocean's capacity to experimental data are combined (I 5), they reveal a strikingly general picture (Fig. I). In provide food, maintain water quality; and recover from perturbations. Yet available data suggest creased diversity of both primary producers that at this point, these trends are still reversible. (Fig. IA) and consumers (Fig. 1B) enhanced all examined ecosystem processes. Observed hat is the role ofbiodiversity in main directly caused by exploitation, pollution, and effect sizes corresponded to a 78 to 80% taining the ecosystem services on habitat destruction, or indirectly through cli enhancement of primary and secondary pro .W which a growing human population mate change and related perturbations of ocean duction in diverse mixtures relative to mono depends? Recent surveys of the terrestrial biogeochemistry (9- 13). Although marine cultures and a 20 to 36% enhancement of literature suggest that local species richness extinctions are only slowly uncovered at the resource use efficiency (Fig. I, A and B). may enhance ecosystem productivity and sta global scale (9), regional ecosystems such as Experiments that manipulated species di bility (1-3). However, the importance of bio estuaries (J 0), coral reefs ( 11), and coastal ( 12) versity (Fig. 1B) or genetic diversity (Fig. 1C) diversity changes at the landscape level is less and oceanic fish communities (1 3) are rapidly both found that diversity enhanced ecosystem clear, and the lessons from local experiments losing populations, species, or entire functional stability, here defined as the ability to withstand and theory do not seem to easily extend to long groups. Although it is clear that particular recurrent perturbations. This effect was linked term, large-scale management decisions (3). These issues are particularly enigmatic for the Fig. 1. Marine bio- 0.8 - world's oceans, which are geographically large diversity and ecosystem 0.7 A B and taxonomically complex, making the scal functioning in controlled ing up from local to global scales potentially experiments. Shown are 0.6 ! more difficult (4). Marine ecosystems provide a response ratios [tn(high/ ~ Q) 0.5 wide variety of goods and services, including low diversity} ±95% con- U) c 0.4 . vital food resources for millions ofpeople (5, 6). fidence interval (CI}] of I ! 8. U) A large and increasing proportion of our pop ecosystem processes to 0.3 ulation lives close to the coast; thus the loss of experimental manipula- ~ r:: 0.2 services such as flood control and waste de tions of species diversity _j 1 toxification can have disastrous consequences of (A) primary producers 0.1 I ! (7, 8). Changes in marine biodiversity are (plants and algae}, and 0.0 ...... (B) consumers (herbivores (5) (3) (14) (5) (6) (9) (8) c: 2-C and predators}. Incr eased G> 2-C: G> 1 0 {g.g _g.g 0epartment of Biology, Dalhousie University, Halifax, NS, !:! 2-'l3 !:! 'Co 2 diversity significantly en- c:g 8g -~ Canada B3H 4)1. Department of Economics and Finance, 0 0-o 0 ·~ :§ :.5 "' E "C "'0"' - u Univef5ity of Wyoming, Laramie, WY 82071, USA. 3Plymouth hanced aU examined eco- "' ., ·"'"'c:: e ~ ~ g~ a:., ::J"' ll.C. C/)Q. a:., ::l"' C/)Q. ~~ U5"' Marine Laboratory, Plymouth PL13DH, UK. "virginia lnstiMe system functions (0.05 > . of Marine Sdences, Gloucester Point, VA 23062-1346, USA P > 0.0001}. The number Producer diversity Consumer diversity 50epartment of Systems Ecology, Stockholm Univer5ity, of studies is given in 6 Stockholm, SE·106 91 Sweden. Beijer International Institute (0 §1400 parentheses. Genetic ?:- 70 of Ecological Economics, Royal Swedish Academy of Sdences, ·u; 1200 . D 7 diversity increased the 'g SE-104 OS, Stockholm, Sweden. National Center for r:: 65 recovery of seagrass eco- Ecological Analysis and Synthesis, Santa Barbara, CA ~ 60 ~ 1000 8 systems after overgrazing c. 93101. USA. Center for Marine Biodiversity and Conserva· 8 55 0>800 tion, Scrip,ps Institution of Oceanography, La Jolla. CA 92093- (solid drctes) and ctimatic .J::. 9 en 50 0202, USA Smithsonian Tropical Research lnstiMe, Box extremes (open drdes). :g ~600 1 45 'C 2072, Balboa, Republic of Panama. ~opkins Marine Station, 8_400 1 (D) Diet diversity en- !!! Stanford University, Padfic Grove, CA 93950, USA. 'section Cl 40 hanced reproductive ca- ~ 200 of Evolution and Ecology, Univer5ity of California, Davis, CA "'Q) 35 0 95616, USA. ll Fisheries Centre, Univer5ity of British pacity in zooplankton (/) (..) 0 Columbia. Vancouver, BC. Canada V6T 1Z4. www.sciencemag.org SCIENCE VOL 314 3 NOVEMBER 2006 787 :;XT"'IIOEO PDf rl'tMAT ~ l N30 r>y Travel GrantsAvailable Neurosc•t nee 2014 Ar'lfllleollttOI"' De t(11tllt kllf'y' 31"' A::>olv 1\Jt.~w !'-..!?....-=: www r.,r.hyst c-na. ta l~ Impacts of Biodiversity Loss on Ocean Ecosystem Services Boris Worm et a/. Science 314, 787 (2006); DOl: 10.1126/science.1132294 This copy is foryourpersonal, non-commercial use only. If you wish to distribute this article to others, you can order high-quality copies for your colleagues, clients, or customers by clicking here. Permission to republish or repurpose articles or portions of articles can be obtained by following the guidelines here. The following resources related to this article are available online at www.sciencemag.org (this information is current as ofJune 9, 2014 ): Updated information and services, including high-resolution figures, can be found in the online version of this article at: http://www.sciencemag.org/contenU314/5800/787.full .html Supporting Online Material can be found at: http://www.sciencemag.org/contenUsuppl/2006/1 0/31/314.5800.787. DC1.html A list of selected additional articles on the Science Web sites related to this article can be found at: http://www.sciencemag.org/contenU314/5800/787. full.html#related This article cites 23 articles, 13 ofwhich can be accessed free: http://www.sciencemag.org/contenU314/5800/787.full.html#ref-1 ist-1 This article has been cited by 362 article(s) on the lSI Web of Science This article has been cited by 100 articles hosted by HighWire Press; see: http://www.sciencemag.org/contenU314/5800/787. full.html#related-urls This article appears in the following subject co llections: Ecology . http://www.sciencemag .or g/cgi/collection/ecology Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright 2006 by the American Association for the Advancement of Science; all rights reserved . The title Science is a registered trademark ofAAAS. RESEARCH ARTICLE to either increased resistance to disturbance (/6) or ships are difficult to infer, these data suggest that the outer margins of the major current systems enhanced recovery aftetward (/7). A number of substantial loss ofbiodiversity (Fig. 2, A and C) (21). They are characterized by distinct bathym experiments on diet mixing further demonstrated is closely associated with regional loss of etry, hydrography, productivity, and food webs. the importance of diverse food sources for ecosystem services (Fig. 2D) and increasing risks Collectively, these areas produced 83% ofglobal sccondaty production and the channeling of that for coastal inhabitants (Fig. 2E). Experimentally fisheries yields over the past 50 years. Fish di energy to higher levels in the food web (Fig. !D). derived predictions that more species-rich sys versity data for each LME were derived inde Different diet items were requireCI to optimize tems should be more stable in delivering pendently from a comprehensive fish taxonomic different life-history processes (growth, survival, services (Fig. 1) are also supported at the database (22). and fecundity), leading to maximum total produc regional scale (Fig. 28). Globally, the rate offisheries collapses, defined tion in the mixed diet In swnmary, experimental Large marine ecosystems. At the largest here as catches dropping below I 0"/o of the results indicate robust positive linkages between scales, we analyzed relationships between bio recorded maximum (23), has been accelerating biodiversity, productivity, and stability across diversity and ecosystem services using the global over time, with 290/o of currendy fished species trophic levels in marine ecosystems. Identified catch database from the United Nations Food and considered collapsed in 2003 (Fig. 3A, diamonds). mechanisms from the original studies include com Agriculture Organization (FAO) and other sources This accelerating trend is best descnbcd by a power 8992 plementary resource use, positive in.teractions, and (I5, 20). We extracted all data on fish and in relation (y= 0.0168xl. , r = 0.96, P< 0.0001), increased selection ofhighly performing species vertebrate catches from 1950 to 2003 within all which predicts the percentage of currently col at high diversity. 64 large marine ecosystems (LMEs) worldwide. lapsed taxa as a fimction of years elapsed since 2 Coastal ecosystems. To test whether exper LMEs are large (>150,000 km ) ocean regions 1950. Cumulative collapses (including recovered imental results scale up in both space and time, reaching from estuaries and coastal areas to the species) amounted to 65% ofrecorded taxa ~t 2 3 we compiled long-term trends in regional bio seaward boundaries of continental shelves and 3A, triangles; regression fit: y =0.0227x · , diversity and services from a detailed database of 12 coastal and estuarine ecosystems (I(J) and A other sources (15). We examined trends in 30 to 80 (average, 48) economically and ecologically 0~------~~==~~k important species per ecosystem. Records over the past millennium revealed a rapid decline of native species diversity since the onset of industrialization (Fig. 2A). As predicted by experiments, systems with higher regional species richness appeared more stable, showing lower rates of collapse and extinction of i 20 . commercially important fish and invertebrate J taxa over time (Fig. 2B, linear regression. P < 0 0 u ------ 0.01). Overall, historical trends led to the present o 200 400 eoo soo 1ooo 1200 depletion (here defined as >50% decline over Species richno...ss baseline abundance), collapse (>90% decline), ro +------~------~------~------__j or extinction (100% decline) of91, 38, or 7% 1000 1200 1400 1600 1800 2000 of species, on average (Fig. 2C}. Only 14% Year 100 ..,.------,100 recovered from collapse (Fig. 2C); these species c D E were mostly protected birds and mammals. • 1000 These regional biodiversity losses impaired • I • • • ' 100 50 50 Cl) at least three critical ecosystem services (Fig. (I) C» Cl 2D): number of viable (noncollapsed) fisheries c • 10 1ij ctl .s:; (- 33%); provision of nursery habitats such as .s:; (,) (,) • oyster reefs, seagrass beds, and wetlands Hi9%); 0 ------0 '------ ~------'E 'E • (I) and filtering and detoxification services provided (I) ~ by suspension feeders, submerged vegetation. ~ I · (I) (I) a.. and wetlands (~3%). Loss of filtering services a.. -so • ·50 probably contnbuted to declining water quality I I (18) and the increasing occurrence of harmful (12) (l2) (10) (8) (3) 17) 13) ( 6) (9) (6) algal blooms, fish kills, shellfish and beach ·100 • 100 .. c; .. .. );.. );.. .. closures, and oxygen depletion (Fig. 2E}. i i ~ a! .\I! ~ l? .. .~ ! E \1) .~ 0 ;;; .c ::> $! .., o; "' Increasing coastal flooding events (Fig. 2E) are l!jl<~ s:: s:: ~ .. ""s:: ·; .!II "' .2 -§ ~ if s::cn 5i- ] > linked to sea level rise but were probably !aw~ ~ ~ "C .Si .,Q) Q; '! ~~ Oi ., jj .~ accelerated by historical losses of floodplains :s =u: ~ '§ ~8 "' <.> (I)"(; !2! 0 ., z ~ J: )( 0 c. and erosion control provided by coastal wetlands, "' 0 {/) reefs , and submerged vegetation (7). An increased number ofspecies invasions over time Biodiversity Services Risks (Fig. 2E) also coincided with the loss of native Fig. 2 . Regional loss of species diversity and ecosystem services in coastal oceans. (A) Trends of biodiversity; again, this is consistent with exper collapse (circles, >90% decline} and extinction (triangles, 100% decline) of species over the past 1000 imental results (/9). Invasions did not compen yeais. Means and standard errors are shown (n =12 regions in Europe, North America, and' Australia). sate for the loss of native biodiversity and (B) Percentage of collapsed (circles) and extinct (triangles) fisheries in relation to regional fish species services, because they comprised other species richness. Significant linear regression lines are depicted (p < 0.01). (C to E) Relative losses or gains in groups, mostly microbial, plankton, and small (C) biodiversity, (D) ecosystem services, and (E) risks that are associated with the loss of services. The invertebrate taxa (1 (J). Although causal relation number of studies is given In parentheses; error bars indicate standard errors. 788 3 NOVEMBER 2006 VOL 314 SCIENCE www.sciencemag.org RESEARCH ARTICLE fourfold average increase in catch perunit ofeffort Our findings further suggest that the elimination ty, coastal water quality, and ecosystem stability, in fishod areas around the reseiVes (Fig. 48). The of locally adapted populations and species not affecting current and future generations. difference in total catches was less pro.nounced only impairs the ability of marine ecosystems to (Fig. 48), probably because of restrictions on feed a growing human population but also References and Notes fishing effort around many reserves. Resistance sabotages their stability and recovery potential 1. M. Loreau et ol.• Science 294, 804 (2001). and recovery after natural disturbances from in a rapidly changing marine environment 2. M. Palmer et ol., Science 304, 1251 (2004). storms and thermal stress tended to increase in We recognize limitations in each of our data 3. D. U. Hoop~r t t o/., Ecol. Monogr. 75, 3 (2005). 4. I. E. Hendrik1, C. M. Duarte. C. H. R. Heip, Science 312, reserves, though not significantly in most cases sources, particularly the inherent problem of 1715 (2006). (Fig. 4C). Community variability, as measured by inferring causality from correlation in the larger 5. C. H. Peter1on. ). Lubchenco. in Noture·s Servfct$: the coefficient of variation in aggregate fish scale studies. The strength of these results rests Societal Dependence on Natural Ecosystems, G. C. Daily, biomass, was reduced by 21% on average (Fig. on the consistent agreement of theory, exper Ed. (Island Press. Washington. DC, 1997), pp. 177-194. 4C). Finally, tourism revenue measured as the iments, and obSCIVations across widely different 6. C. M. Holmlund, M. Hammer, Ecol. Econ. 29. 253 (1999). relative increase indive trips within 138 Canbbean scales and ecosystems. Our analysis may provide 7. F. Danielnn ~ ol.• Scitnce 310. 643 (2005). protected areas strongly increased after they were a wider context for the interpretation of local 8. W. N. Adger, T. P. Hughes. C. Folke, S. R. C.rptnter, established (Fig. 4D). For several variables, biodiversity experiments that produced diverging ). Rockstrom, Science 309, 1036 (2005). statistical significance depended on how studies and controversial outcomes(/, 3, 24). It suggests 9. N. K. Dulvy, Y. Sadovy, ). D. Reynolds, Fi5h Fish. 4, 25 (2003). were weighted (Fig. 4, solid versus open circles). that very general patterns emerge on progressive 10. H. K. Lotze et at. • Science 312. 1806 (2006). This is probably the result of large variation in ly larger scales. High- 790 3 NOVEMBER 2006 VOL 314 SCIENCE www.sciencemag.org