SCIENCE’S COMPASS                            REVIEW
                                                              ●

   REVIEW: ECOLOGY


                 Economic Reasons for Conserving
                     Wild Nature
     Andrew Balmford,1* Aaron Bruner,2 Philip Cooper,3 Robert Costanza,4† Stephen Farber,5 Rhys E. Green,1,6 Martin Jenkins,7
       Paul Jefferiss,6 Valma Jessamy,3 Joah Madden,1 Kat Munro,1 Norman Myers,8 Shahid Naeem,9 Jouni Paavola,3
            Matthew Rayment,6 Sergio Rosendo,3 Joan Roughgarden,10 Kate Trumper,1 R. Kerry Turner3

                                                                matched estimates of the marginal values of
     On the eve of the World Summit on Sustainable Development, it is timely to assess                   goods and services delivered by a biome
     progress over the 10 years since its predecessor in Rio de Janeiro. Loss and degradation                when relatively intact, and when converted
     of remaining natural habitats has continued largely unabated. However, evidence has                  to typical forms of human use. To ensure
     been accumulating that such systems generate marked economic benefits, which the                    we did not neglect private benefits of con-
     available data suggest exceed those obtained from continued habitat conversion. We                   version, studies were only included if they
     estimate that the overall benefit:cost ratio of an effective global program for the                   covered the most important marketed
     conservation of remaining wild nature is at least 100 :1.                               goods, as well as one or more nonmarketed
                                                                services delivering local social or global
                                                                benefits. We cross-validated figures for in-


 H
       umans benefit from wild nature (1) in         per hectare for each of 17 services across 16        dividual goods and services with other es-
       very many ways: aesthetically and cul-        biomes and then extrapolating to the globe by        timates from similar places. Finally, we
       turally; via the provision of ecological       multiplying by each biome’s area, the Costanza       checked that the comparisons across differ-
   services such as climate regulation, soil forma-       team estimated the aggregated annual value of        ent states of a biome used the same valua-
   tion, and nutrient cycling; and from the direct        nature’s services (updated to 2000 US$) to lie       tion techniques for particular goods and
                                  in the range of $18 trillion to $61 trillion (1012 ),
   harvest of wild species for food, fuel, fibers,                                      services. Our survey uncovered only five
   and pharmaceuticals (2). In the face of increas-       around a rough average of $38 trillion. These        examples that met all these criteria. Here,
   ing human pressures on the environment, these         figures are of similar size to global gross na-       we summarize their findings, with all fig-
   benefits should act as powerful incentives to         tional product (GNP), but have been criticized       ures expressed as net present values (NPVs,
                                                                in 2000 US$ ha 1), and using the discount
   conserve nature, yet evaluating them has           by some in the economic community (5–9).
   proved difficult because they are mostly not           One problem is that such macroeconomic         rates considered by the authors [see Fig. 1
   captured by conventional, market-based eco-          extrapolations are inconsistent with microeco-       and supplemental online material (10) for
   nomic activity and analysis.                 nomic theory: extrapolation from the margin to       further details].
     In 1997, Costanza et al. published a synthe-       a global total should incorporate knowledge           Two studies quantified net marginal ben-
   sis (3) of more than 100 attempts to value          about the shape of the demand curve (3, 5–8). In      efits of different human uses of tropical forest
   ecosystem goods and services using a range of         practice, it is very likely that per-unit demand      areas. Kumari compared the values obtained
   techniques including hedonic pricing, contin-         for nonsubstitutable services escalates rapidly as     from timber plus a suite of nontimber forest
   gent valuation, and replacement cost methods         supply diminishes, so that simple grossing-up of      products (NTFPs), as well as the values of
   (4). Using case studies to derive average values       marginal values (as is also done in calculating       water supply and regulation, recreation, and
                                  GNP from prices) will probably underestimate        the maintenance of carbon stocks and endan-
                                  true total values. On the other hand, high local      gered species, for forests under a range of
    Conservation Biology Group, Department of Zoology,
   1
                                  values of services such as tourism may not be        management regimes in Selangor, Malaysia
   University of Cambridge, Cambridge CB2 3EJ, UK. 2Cen-
                                  maintained if extrapolated worldwide. In addi-       (11). Compared with two methods of re-
   ter for Applied Biodiversity Science at Conservation In-
                                  tion, while some policy decisions are made us-       duced-impact logging, high-intensity, unsus-
   ternational, 1919 M Street, NW, Suite 600, Washington,
   DC 20036, USA. 3Centre for Social and Economic Re-      ing macroeconomic indicators, many others are        tainable logging was associated with greater
   search on the Global Environment (CSERGE), School of     made at the margin, and so are more appropri-        private benefits through timber harvesting (at
   Environmental Sciences, University of East Anglia, Nor-
                                  ately informed by marginal rather than total        least at high discount rates and over one
   wich NR4 7TJ, UK. 4Center for Environmental Science,
                                  valuations (9).                       harvesting cycle), but reduced social and
   Biology Department and Institute for Ecological Eco-
   nomics, University of Maryland, Box 38, Solomons, MD       Another problem with the original estimate       global benefits (through loss of NTFPs, flood
   20688, USA. 5Graduate School of Public and Internation-    is that landscapes can yield substantial (albeit      protection, carbon stocks, and endangered
   al Affairs, University of Pittsburgh, Pittsburgh, PA 15260,  rather different) flows of goods and services        species). Summed together, the total econom-
   USA. 6The Royal Society for the Protection of Birds, The
                                  after, as well as before, conversion by humans       ic value (TEV) of forest was some 14%
   Lodge, Sandy, Bedfordshire SG19 2DL, UK. 7UN Environ-
                                  (which is of course why people convert them).        greater when placed under more sustainable
   ment Programme–World Conservation Monitoring Cen-
   tre (UNEP-WCMC), 219 Huntingdon Road, Cambridge        A clearer picture of the value of retaining habitat     management (at $13,000 compared with
   CB3 ODL, UK. 8Green College, Woodstock Road, Oxford                                    $11,200 ha 1).
                                  in relatively undisturbed condition might there-
   OX2 6HG, UK; and Upper Meadow, Old Road, Heading-
                                  fore be obtained by estimating not the gross          A study from Mount Cameroon, Cam-
   ton, Oxford OX3 8SZ, UK. 9Department of Zoology,
                                  values of the benefits provided by natural bi-       eroon, comparing low-impact logging with
   University of Washington, 24 Kincaid Hall, Box 351800,
   Seattle, WA 98195–1800, USA.10Department of Biolog-      omes, but rather the difference in benefit flows      more extreme land-use change again found
   ical Sciences, Stanford University, Stanford, CA 94305,    between relatively intact and converted versions      that private benefits favor conversion, this
   USA.
                                  of those biomes.                      time to small-scale agriculture (12). Howev-
   *To whom correspondence should be addressed. E-                                      er, a second alternative to retaining the forest,
   mail: a.balmford@zoo.cam.ac.uk                Net Marginal Benefits                    conversion to oil palm and rubber planta-
   †Address after Sept. 2002: Gund Institute of Ecolog-
                                  To address these concerns, we reviewed           tions, in fact yielded negative private benefits
   ical Economics, The University of Vermont, Burling-
   ton, VT 05405, USA.                      more than 300 case studies, searching for          once the effect of market distortions was

950                            9 AUGUST 2002 VOL 297 SCIENCE www.sciencemag.org
                             SCIENCE’S COMPASS
removed. Social benefits from NTFPs, sedi-    did sustainable fishing (15). The social ben-  vices outweighs the marketed marginal ben-
mentation control, and flood prevention      efits of sustainable exploitation, arising from efits of conversion, often by a considerable
were highest under sustainable forestry, as    coastal protection and tourism, were also lost  amount. Across the four biomes studied,
were global benefits from carbon storage     upon dynamiting reefs. As a consequence,     mean losses in TEV due to conversion run
and a range of option, bequest, and exis-     the TEV of retaining an essentially intact    at roughly one-half of the TEV of relatively
tence values. Overall, the TEV of sustain-    reef was almost 75% higher than that of     intact systems (mean      54.9%; SE
able forestry was 18% greater than that of    destructive fishing (at $3300 compared      13.4%; n   4). This is certainly not to say
                         with $870 ha 1).
small-scale farming ( $2570 compared                                that conversion has never been economical-
with $2110 ha 1), whereas plantations had       One clear message from our survey is     ly beneficial; in most instances, past clear-
a negative TEV.                  the paucity of empirical data on the central   ance of forests and wetlands for prime
  Three other biomes yielded single studies   question of the changes in delivery of      agricultural land and other forms of devel-
meeting our criteria. Analysis of a mangrove   goods and services arising from the conver-   opment probably benefited society as a
system in Thailand revealed that conversion    sion of natural habitats for human use. For   whole. But unless the present case studies
for aquaculture made                                                    or the range of ser-
sense in terms of                                                      vices and biomes
short-term    private                                                  examined in the lit-
benefits, but not once                                                   erature are extreme-
external costs were                                                     ly unrepresentative
factored in (13). The                                                    (and we know of no
global benefits of                                                     reason why this
carbon sequestration                                                    should be the case),
were considered to be                                                    our synthesis indi-
similar in intact and                                                    cates that at present,
degraded systems.                                                      conversion of re-
However, the sub-                                                      maining habitat for
stantial social bene-                                                    agriculture, aquac-
fits associated with                                                    ulture, or forestry
the original man-                                                      often does not make
grove cover—from                                                      sense from the per-
timber, charcoal,                                                      spective of global
NTFPs, offshore                                                       sustainability.
fisheries, and storm
                                                              Continuing Losses
protection—fell to
almost zero follow-                                                     These results there-
ing conversion. Sum-                                                    fore provide a clear
ming all measured                                                      and compelling eco-
goods and services,                                                     nomic case, alongside
the TEV of intact                                                      sociocultural and
mangroves exceeded                                                     moral arguments (16–
that of shrimp farm-                                                    18), for us to strength-
ing by around 70%                                                      en attempts to con-
( $60,400 com-                                                       serve what remains of
pared with $16,700                                                     natural ecosystems.
ha 1).                                                           Yet, when we sum-
  van Vuuren and                                                     marized available es-
Roy (14) reported that                                                   timates of recent
draining freshwater Fig. 1. The marginal benefits of retaining and converting natural habitats, expressed as NPV (in trends in the global
              2000 US$ ha 1) calculated using the discount rates ( ) and time horizons presented. Values of
marshes in one of measured goods and services delivered when habitats are relatively intact and when converted status of natural habi-
Canada’s most pro- are plotted as green and black columns, respectively. [From (11–15); see (10) for further tats and free-ranging
ductive agricultural details.]                                               vertebrate popula-
areas yielded net pri-                                                   tions, we found that
vate benefits (in large part because of substan- 10 of the largely natural biomes (including   although key data are again disturbingly scarce,
tial drainage subsidies). However, social bene-  rangelands, temperate forests, rivers and    they show that rates of conversion are high
fits of retaining wetlands, arising from sustain- lakes, and most marine systems) in        across most biomes (10).
able hunting, angling, and trapping, greatly ex- Costanza et al. (3), we found no studies that    We included in our survey any estimate of
ceeded agricultural gains. Consequently, for all met all of our criteria. For the four biomes   global trend in habitat cover based on a series
three marsh types considered, TEVs were high-   which were analyzed, only a handful of      which began in 1970 or later and included a
er when the wetlands remained intact, exceed-   well-established ecosystem services were     period of at least 5 years after the 1992
ing figures for conversion by a mean of around  considered, and some particularly valuable    United Nations Conference on Environment
60% ( $8800 compared with $3700 ha 1).      services, such as nutrient cycling, waste    and Development in Rio de Janeiro. We sup-
  Finally, a synthesis of economic studies   treatment, and the provision of cultural val-  plemented this with biome-specific indices
examining Philippine reef exploitation dem-    ues, were not examined at all.          based on time-series data on populations of
onstrated that despite high initial benefits,     Despite the limited data, our review also  wild vertebrates, derived from the World
destructive techniques such as blast fishing   suggests a second broad finding: in every    Wildlife Fund (WWF) 2000 Living Planet
had a far lower NPV of private benefits than   case examined, the loss of nonmarketed ser-   Index (LPI) and UN Food and Agricultural

                                                                            951
                     www.sciencemag.org SCIENCE VOL 297 9 AUGUST 2002
                                 SCIENCE’S COMPASS
   Organization (FAO) fisheries data (19, 20).     standing to gain immediate private benefits    should in due course result in public and
   For three biomes, we found two estimates       from land-use change. Hence, conserving rel-    private decision-makers acting to reduce con-
   derived by different methods and from either     atively intact habitats will often require com-  version of remaining habitats worldwide.
   largely or wholly independent data. In each     pensatory mechanisms to mitigate the impact    More immediately, given concerns about the
   case, the two estimates were remarkably sim-     of private, local benefits foregone, especially  practicalities of exploiting natural resources
   ilar (10), and so were averaged to yield single   in developing countries. We see the develop-    sustainably, one of the most important strat-
   estimates of rates of change. Data such as      ment of market instruments that capture at a    egies to safeguard relatively intact ecosys-
   these, quantifying trends in areal coverage     private level the social and global values of   tems is the maintenance of remaining habitats
   and in populations, in some ways provide a      relatively undisturbed ecosystems—for in-     in protected areas. This costs money, and
   more tractable measure of the scale of the      stance, through carbon or biodiversity credits   predictably, our current undervaluation of na-
   ongoing crisis facing nature than do estimates    or through premium pricing for sustainably     ture is reflected in marked underinvestment
   of extinction rates, which are harder to doc-    harvested wild-caught fish or timber (22,     in reserves. To the best of our knowledge, the
   ument and more difficult to link to monetary     23)—as a crucial step toward sustainability.    world spends (in 2000 US$) $6.5 billion
   values.                         Third, the private benefits of conversion   each year on the existing reserve network
     Overall, we found that five of the six biomes  are often exaggerated by intervention fail-    (28). Yet, half of this is spent in the United
   measured have experienced net losses since the    ures. In the Cameroon study, for example,     States alone. Globally, despite increased ex-
   Rio summit, with the mean rate of change       forests were cleared for plantations because    penditure since the Rio Summit by both
   across all measured biomes running at –1.2%     of private benefits arising from government    international institutions and private founda-
   per year, or –11.4% over the decade (Fig. 2)     tax incentives and subsidies (12). The same is   tions, available resources for existing re-
   (10). Hence, the capacity of natural systems to   true for the Canadian wetland example (14),    serves fall far short of those needed to meet
   deliver goods and services upon which we de-     as well as for many other wetlands across the   basic management objectives (29). Moreover,
   pend is decreasing markedly. Costing the over-    United States and Europe (24). While over     terrestrial and marine reserves currently cov-
   all value of these losses is fraught with the    the short term these programs may be rational   er only around 7.9% and 0.5% of Earth’s land
   problems of extrapolation and data availability   with respect to public or private policy ob-    and sea area, respectively (30, 31), well be-
   already discussed. Nevertheless, it is sobering   jectives, over the longer term many result in   low the minimum safe standard considered
   to calculate that if the aggregate figures of                              necessary for the task of maintaining wild
   Costanza et al. (3) and our estimate of the                               nature into the future (32–34).
   proportion of TEV lost through habitat change                                To estimate the resources needed to meet
   are roughly representative, a single year’s hab-                            this shortfall on land, we reworked recent
   itat conversion costs the human enterprise, in                             calculations (28, 35) of the costs of properly
   net terms, of the order of $250 billion that year,                           managing existing terrestrial protected areas
   and every year into the future (10). Why then is                            and expanding the network to cover around
   widespread habitat loss still happening, and                              15% of land area in each region. We found
   what can we do about it?                                        that a globally effective network would re-
                                                       quire an approximate annual outlay of be-
   Reasons for Continued Conversion                                    tween $20 billion and $28 billion [includ-
   In economic terms, our case studies illustrate                             ing payments to meet private opportunity
   three broad, interrelated reasons why the                                costs imposed by existing and new reserves,
   planet is continuing to lose natural ecosys-                              spread out over 10 and 30 years, respectively
   tems despite their overall benefits to society                             (10)]. New work derived from the costs of
   (21). First, there are often failures of infor-                             existing marine reserves suggests that an
                              Fig. 2. Recent global estimates of the annual
   mation. For many services, there is a lack of                              equivalent initiative for the world’s seas, this
                              rate of change in area or the abundance of
   valuations of their provision by natural sys-                              time covering 30% of total area (34, 36),
                              associated vertebrate populations for six bi-
   tems, and particularly of changes in this pro-                             would cost at most $23 billion/year in re-
                              omes. Note that the biomes that have declined
   vision as human impacts increase. Although                               current costs, plus $6 billion/year (over 30
                              deliver valuable ecosystem services (3). *Values
   this is an understandable reflection of sub-                              years) in start-up costs (10). The estimated
                              plotted are the mean of habitat and popula-
   stantial technical difficulties, we believe that                            mean total cost of an effective, global reserve
                              tion-based estimates; †Little confidence can be
                              attached to this value (10).
   future work needs to compare delivery of                                program on land and at sea is some $45
   multiple services across a range of competing                              billion/year. This sum dwarfs the current $6.5
   land uses if it is to better inform policy      both economic inefficiency and the erosion of   billion annual reserve budget, yet could be
   decisions. Our examples show that even        natural services. Globally, the subset of sub-   readily met by redirecting less than 5% of
   when only a few ecosystem services are con-     sidies which are both economically and eco-    existing perverse subsidies (25, 26). The cru-
   sidered, their loss upon conversion typically    logically perverse totals between $950 billion   cial question is whether this is a price worth
   outweighs any gains in marketed benefits.      and $1950 billion each year [depending on     paying.
     Second, these findings highlight the fun-    whether the hidden subsidies of external costs     Although limited data make the answer
   damental role of market failures in driving     are also factored in (25, 26)]. Identifying and  imprecise, they indicate that conservation in
   habitat loss. In most of the cases we studied,    then working to remove these distortions      reserves represents a strikingly good bargain.
   the major benefits associated with retaining     would simultaneously reduce rates of habitat    We assumed that the mean proportional loss
   systems more or less intact are nonmarketed     loss, free up public funds for investing in    of value upon conversion recorded in our
   externalities, accruing to society at local and   sustainable resource use, and save money      case studies is representative of all biomes
   global scales. Conversion generally makes      (25–27).                      and services, and that previous gross per-
   narrow economic sense, because such exter-                               hectare values of those services are roughly
                              Costing Conservation
   nal benefits [or related external costs, as in                             correct (3). If these assumptions are valid,
   the case of the damage caused by shrimp       Tackling these underlying economic prob-      then our hypothetical global reserve network
   farming (13)] have very little impact on those    lems requires action on many levels, but      would ensure the delivery of goods and ser-

952                       9 AUGUST 2002 VOL 297 SCIENCE www.sciencemag.org
                             SCIENCE’S COMPASS
                                                         20. UN Food and Agricultural Organization, The State of
vices with an annual value (net of benefits   case: our relentless conversion and degra-
                                                           the World Fisheries and Aquaculture (UN FAO, Rome,
from conversion) of between $4400 billion    dation of remaining natural habitats is erod-
                                                           2000).
and $5200 billion, depending on the level of   ing overall human welfare for short-term             21. R. K. Turner et al., Ecol. Econ. 35, 7 (2000).
resource use permitted within protected ar-   private gain. In these circumstances, retain-          22. J. Hardner, R. Rice, Sci. Am. 286 (no. 5), 71 (2002).
eas, and with the lower number coming from    ing as much as possible of what remains of            23. S. Scherr, A. White, D. Kaimowitz, Policy Brief: Making
                                                           Markets For Forest Communities (Forest Trends,
a network entirely composed of strictly pro-   wild nature through a judicious combina-
                                                           Washington, DC, and Center for International Forest-
tected reserves [for working, see (10)]. The   tion of sustainable use, conservation, and,
                                                           ry Research, Bogor, Indonesia, 2002).
benefit:cost ratio of a reserve system meeting  where necessary, compensation for result-            24. R. Turner, T. Jones, Eds., Wetlands: Market and Inter-
minimum safe standards is therefore around    ing opportunity costs [as called for at the             vention Failures (Four Case Studies) (Earthscan, Lon-
100 :1.                     Rio Summit (40)] makes overwhelming                 don, 1991).
                                                         25. C. P. van Beers, A. P. G. de Moor, Addicted to Subsi-
  Put another way, the case studies, the    economic as well as moral sense.
                                                           dies: How Governments Use Your Money to Destroy
service values of Costanza et al. (3), or our
                                                           the Earth and Pamper the Rich (Institute for Research
                           References and Notes
reserve costs would have to be off by a                                        on Public Expenditure, The Hague, Netherlands,
                         1. By “wild nature” we mean habitat in which biodiver-
factor of 100 for the reserve program en-                                       1999).
                           sity, nonbiotic components, and ecosystem function-
                                                         26. N. Myers, J. Kent, Perverse Subsidies (Island Press,
visaged to not make economic sense. We        ing are sufficiently intact that the majority of eco-
                                                           Washington, DC, 2001).
consider errors of this size to be highly      system services typically derived from such a habitat
                                                         27. S. L. Pimm et al., Science 293, 2207 (2001).
                           are still being sustainably and reliably supplied. Our
unlikely, because most of our assumptions
                                                         28. A. James, K. J. Gaston, A. Balmford, BioScience 51, 43
                           usage differs from other usages, such as those adopt-
are conservative [for other sensitivity anal-                                     (2001).
                           ed in cultural or anthropological studies. Because our
yses, see (10)]. For example, in terms of the    focus is on wild nature, we excluded the cropland and    29. A. N. James, M. J. B. Green, J. R. Paine, Global Review
                           urban biomes when using data from table 2 of (3).        of Protected Area Budgets and Staff ( WCMC, Cam-
values of services, we assume that unit
                         2. G. C. Daily, Ed., Nature’s Services (Island Press, Wash-     bridge, 1999)
values will not increase as supply dimin-      ington, DC, 1997).                      30. International Union for Conservation of Nature and
ishes, that nature reserves do not increase    3. R. Costanza et al., Nature 387, 253 (1997).           Natural Resources (IUCN), 1997 United Nations List
                         4. The hedonic price method values environmental ser-
the flow of services beyond their bound-                                       of Protected Areas ( WCMC and IUCN, Cambridge,
                           vices by comparing market prices (e.g., for residential
aries [whereas some clearly can (34, 37 )],                                      UK, and Gland, Switzerland, 1998).
                           housing) across situations which differ in the provi-
                                                         31. G. Kelleher, C. Bleakley, S. Wells, A Global Represen-
and that all of a biome’s services not in-      sion of those services. Contingent valuation involves
                                                           tative System of Marine Protected Areas ( The World
                           asking respondents how much they would be pre-
cluded in the Costanza et al. survey (3) are
                                                           Bank, Washington, DC, 1995).
                           pared to pay for a particular environmental benefit
worthless. On the reserve costs side, we                                     32. IUCN, Parks for Life: Report of the IVth World Con-
                           (such as ensuring the survival of a species or habitat)
assume that management costs do not de-                                        gress on National Parks and Protected Areas (IUCN,
                           or how much compensation they would demand for
                                                           Gland, Switzerland, 1993).
crease once local communities’ private op-      its loss. The replacement cost technique quantifies
                                                         33. M. E. Soule, M. A. Sanjayan, Science 279, 2060
                                                                 ´
                           the cost of restoring or synthetically replacing an
portunity costs are met, and that expanding                                      (1998).
                           ecosystem service.
reserve systems yield no cost savings                                       34. J. Roughgarden, P. Armsworth, in Ecology: Achieve-
                         5. M. Toman, Ecol. Econ. 25, 57 (1998).
                                                           ment and Challenge, M. Press, N. Huntly, S. Levin, Eds.
through economies of scale or dissemina-     6. R. K. Turner, W. N. Adger, R. Brouwer, Ecol. Econ. 25,
                                                           (Blackwell Science, Oxford, 2001), pp. 337–356.
                           61 (1998).
tion of best practice. Because all of these
                                                         35. A. N. James, K. J. Gaston, A. Balmford, Nature 401,
                         7. P. Dasgupta, Human Well-Being and the Natural En-
assumptions are biased against conserva-                                       323 (1999).
                           vironment (Oxford Univ. Press, Oxford, 2001).
tion, we consider our 100 :1 ratio as a low    8. P. A. L. D. Nunes, J. C. J. van den Bergh, Ecol. Econ. 39,  36. California Department of Fish and Game, NOAA’s
                           203 (2001).                           Channel Islands National Marine Sanctuary, A Rec-
estimate of the likely benefits of effective
                         9. G. C. Daily et al., Science 289, 395 (2000).           ommendation for Marine Protected Areas in the
conservation.                                                     Channel Islands National Marine Sanctuary (California
                         10. For further details, see supporting online material.
                                                           Department of Fish and Game, Santa Barbara, CA,
                           Many of the numbers reported here are unavoidably
Development and Wild Nature                                              2001).
                           imprecise. To enable readers to follow our working,
                           we generally present numbers used in calculations to     37. C. M. Roberts, J. A. Bohnsack, F. Gell, J. P. Hawkins, R.
In advocating greatly increased funding for
                           three significant figures, but then round off the final       Goodridge, Science 294, 1920 (2001).
the maintenance of natural ecosystems, we      results in accord with their precision.           38. UN Population Division, World Population Prospects:
                         11. K. Kumari, thesis, University of East Anglia, Norwich,
are not arguing against development. Given                                      The 2000 Revision (UN Department of Economic and
                           UK (1994).                            Social Affairs, New York, 2001).
forecast increases in the human population
                         12. G. Yaron, J. Environ. Planning Manage. 44, 85 (2001).    39. UN Development Programme (UNDP), Human Devel-
of more than three billion by 2050 (38) and   13. S. Sathirathai, Economic Valuation of Mangroves and the     opment Report, 2001 (UNDP, New York, 2001).
the fact that some 1.2 billion people still     Roles of Local Communities in the Conservation of Nat-    40. See the text of the Convention on Biological Diversity
                           ural Resources: Case Study of Surat Thani, South of
live on less than 1 US$/day (39), develop-                                      at www.biodiv.org.
                           Thailand (unpublished report, Economy and Environ-      41. This review is the result of a workshop convened by
ment is clearly essential. However, current     ment Program for Southeast Asia, Singapore, 1998).        the Royal Society for the Protection of Birds (RSPB)
development trajectories are self-evidently   14. W. van Vuuren, P. Roy, Ecol. Econ. 8, 289 (1993).        and sponsored by the RSPB and the UK Government’s
                         15. A. T. White, H. P. Vogt, T. Arin, Mar. Pollut. Bull. 40,
not delivering human benefits in the way                                       Department for Environment, Food, and Rural Affairs.
                           598 (2000).                           We thank N. Hockley, P. Gravestock, J. Scharlemann,
that they should: income disparity world-    16. P. W. Taylor, Environ. Ethics 3, 197 (1981).           and C. Tiley for help with research, and M. Avery, R.
wide is increasing and most countries are    17. D. Ehrenfeld, in Biodiversity, E. O. Wilson, Ed. (Nation-    Cowling, G. Daily, A. Gammell, D. Gibbons, J. Mc-
                           al Academy Press, Washington, DC, 1988), pp. 212–
not on track to meet the United Nations’                                       Neely, and C. Roberts for stimulating discussions.
                           216.
goals for human development and poverty                                      Supporting Online Material
                         18. B. Norton, Environ. Values 1, 97 (1992).
eradication by 2015 (39). Our findings                                      www.sciencemag.org/cgi/content/full/1073947/DC1
                         19. J. Loh et al., WWF Living Planet Report 2000 ( World
                           Wildlife Fund, Gland, Switzerland, 2000).          SOM Text
show one compelling reason why this is the




                                                                                         953
                    www.sciencemag.org SCIENCE VOL 297 9 AUGUST 2002