Articles

Marine Ecoregions of the World: A Bioregionalization of Coastal and Shelf Areas

MARK D. SPA L D I N G , HELEN E. FOX , GERALD R. ALLEN, NICK DAV I D S O N , Z ACH A. FERDA A , MAX FINLAY S O N , BENJAMIN S. HALPERN, MIGUEL A. JORGE, AL LOMBANA, SARA A. LOURIE, KIRSTEN D. MARTIN, E D M U N D M C M A N U S , JENNIFER MOLNAR, CHERI A. RECCHIA, AND JAMES RO B E R T S O N

The conserva tion and su s t a i n a bleuse of marine re sou rces is a high l i gh te goal on a growing nu m ber of n a tional and intern a tional policy agen d a s . d Un fo rtunately, ef fo rts to assess pro gress, as well as to stra tegi c a lly plan and pri o ri ti ze new marine conserva tion measu res, have be en hampered by the lack of a detailed, co m preh en s ive bi o ge o graphic sys tem to classify the oceans. Here we repo rt on a new global sys tem for coastal and shel f a reas: the Ma rine Eco regions of the World, or MEOW, a nested system of 12 realms, 62 provinces, and 232 eco regions. This sys tem provides considera bly bet ter s pa tial resol u tion than earl i er global systems, yet it pre serves many common el em ents and can be cro s s - referen ced to many regional bi o ge o graphic classifications. The design a tion of terre s trial eco regions has revol u ti o n i zed priority set ting and planning for terre s trial co n serva tion; we anti ci pa te similar ben efits from the use of a coh erent and cred i ble marine sys tem . Key wo rds: eco regions, m a rine bi o ge o gra p hy, mapping, m a rine prote cted areas, repre sen t a tive co n serva ti o n

apped classifications of patterns in biodivers i t y h ave long been an important tool in fields from evo luti onary studies to con s erva ti on planning (Forbes 1856, Wallace 1876, Spellerberg and Sawyer 1999, Lo u rie and Vincent 2004). The use of su ch sys tems (notably, the widely c i ted system developed by Ol s on et al. [2001]) in broadscale con s erva ti on , however, has largely been restri cted to terre strial studies (Ch a pe et al. 2003, Ha zen and An t h a m a t te n 2004, Hoe k s tra et al. 2005, Bu r gess et al. 2006, Lamoreux et al. 2006). In the marine envi ronment, existing gl obal cl a s s if i c a ti on systems remain limited in their spatial resoluti on. Some are inconsistent in their spatial covera ge or met h odo l ogical approach. The few publications that have attem pted to use biogeographic regionalizati on in gl obal marine con s erva ti on planning (e.g. , Kell eh er et al. 1995, Ol s on and Di n ers tein 2002) have been qualitative, and have ex pre s s ed concern abo ut the lack of an adequ a te gl obal classification.

In the absence of compelling gl obal covera ge , numerous regi onal cl a s s i f i c a ti on have been created to meet regi on a l s planning need s . Th i s , of course, does not satisfy the need for a gl obal system that is consistent ac ross the many marine realms and coastal zones. Bi ogeogra phic cl a s s i f i c a ti ons are essen tial for developing eco l ogi c a lly repre s en t a tive sys tems of protected areas, as requ i red by intern a ti onal agreem ents su ch as the Conven ti on on Bi o l ogical Divers i ty s Programme of Work on Pro tected Areas and the Ramsar Conven ti on on Wetlands. Marine space is sti ll grossly underrepre s ented in the gl obal protected areas network (on ly about 0.5% of the su rf ace area of the oceans is curren t ly protected ; Ch a pe et al. 2005), a fact that adds urgency to the need for tools to su pport the scaling up of ef fective, representative marine conserva ti on. The key idea u n derlying the term repre s en t a tive is the intent to pro tect a full ra n ge of bi od ivers i ty worl dwi degen e s , spec i e s , and

Mark D. Spalding (e-mail: mspalding@tnc.org), Za ch A. Ferdaa, Jennifer Mol n a r, and James Robert son are co n serva tion sci en tists in The Na tu re Co n servancys Co n serva tion Stra tegies Group, Arlington, VA 22203. Hel en E. Fox and Al Lo m bana are marine biol o gists in the Co n serva tion Sci en ce Program, Wo rld Wi l dl i fe FundUS, Washington, DC 20037. Gerald R. All en is a research asso ci a te at the We s tern Au s tralian Mu seum, Pert h , Western Australia 6986, Australia. Nick Davi d son is the depu ty se cret a ry gen eral of the Ramsar Co nven tion Secret a riat, CH-1196 Gl a n d , Swi t zerland. Max Fi n l ayson is a mem ber and fo rm er chair of Ramsars Sci en tific and Technical Revi ew Pa n el and pri n ci pal re se a rch er in wetland ecol o gy at the In tern a tional Water Managem ent In s ti tu te , Col o m b o, Sri Lanka. Ben jamin S. Ha l pern is proje ct coord i n a to for eco s ys tem - ba sed managem ent of coastal marine sys tems at the National Cen ter for Ecol o gical An a lysis and Syn t h e s i s , Santa Ba rba ra , CA 93101. r Mi g u el A. Jo rge is depu ty dire ctor of WWF In tern a ti o n a l s Gl obal Marine Pro gramme, CH-1196 Gland, Swi t zerl a n d . Sara A. Lou rie is a re se a rch asso ci a te at the Red path Mu seum, Mc G i ll Un ivers i ty, Mo n treal, Q u ebec H3A 2K6, Canada. Kirsten D.Martin was a marine pro gram officer with IUCN (World Co n serva tion Union) wh en this arti cle was prepa red and is curren t ly wo rking as a freelance co n sultant for the Cen sus of Marine Li fe In i ti a tive , 1205 Geneva , Swi t zerl a n d . Ed mund McMa nu s is a senior pro gram of f i cer in the UNEP (Un i ted Nations Envi ro n m ent Pro gramme) Wo rld Co n servation Mo n i to ring Cen tre, C a m b ri d ge CB3 0DL, Un i ted Kingdom. C h eri A. Re cchia is marine pro gram dire ctor at the Wi l dl i fe Co n serva tion Society, New Yo rk , NY 10461. 2007 American In s ti tu te of Biological Sciences.

www.biosciencemag.org

July/August 2007 / Vol. 57 No. 7 BioScience 573

Articles
high er taxa, along with the com mu n i ties, evo luti on a ry pattern s , and eco l ogical processes that sustain this diversity. Biogeogra phic classifications provi de a crucial foundation for the assessment of representativeness (Ol s on and Di n ers tei n 2002, Lourie and Vincent 2004). The growing commitment by governments and the United Nations (UN; e.g., the UN Law of the Sea, the UN Fish Stock s Agreem ent) to implem ent com preh en s ive arra n gem ents for ocean govern a n ce provi des an additi onal arena in wh i ch m a rine biogeogra phic cl a s s i f i c a ti ons are needed. Bi ogeogra phic regi ons are natural fra m eworks for marine zon i n g, wh i ch is a tool increasingly used by regi onal fisheries managem ent organizations. In this article, we pre s ent a new bi ogeogra phic classification for the worlds coastal and shel f areas, wh i ch draws heavi ly on the ex i s ting gl obal and regi onal litera tu re . We bel i eve that this cl a s s i f i c a ti onwi ll be of c ri tical importance in su pporting analyses of patterns in marine bi od ivers i ty, in understanding proce s s e s , and, perhaps most important, in directing futu re ef forts in marine resource managem ent and con s erva ti on . S h erman et al. 2005). Un l i ke the sys tems of Bri ggs and Lon ghu rs t , LMEs repre s ent an ex pert - derived system wi t hout a ri gorous, rep l i c a ble core definiti on . LMEs are rel atively large regi ons on the order of 200,000 km2 or greater, ch a racterized by distinct: (1) bathym etry, (2) hydrography, (3) produ ctivi ty, and (4) troph i c a lly depen dent populati on s (www.lme.noa a . gov/Portal/). LMEs are largely conceived as units for the practical app l i c a ti on of transboundary managem ent issues (fish and fisheri e s , po lluti on, habitat re s torati on , produ ctivi ty, socioeconomics, and govern a n ce ) . Th e LME sys tem focuses on produ ctivi ty and oce a n ogra ph i c processes, and in its pre s ent form omits substantial areas of islands in the Pacific and the Indian oceans. These and other gl obal sys tems con ti nue to play an important role in devel oping our understanding of m a rine bi ogeography and in practical issues of natural re s o u rce management. However, improvements are clearly possible and de s i ra bl e . An ideal system would be hiera rchical and nested , and would all ow for mu l tiscale analyses. E ach level of the h i erarchy would be relevant for con s erva ti on planning or m a n a gem ent interven ti on s , from the gl obal to the local, a lthough it is beyond the scope of the pre s ent ef fort to classify i n d ividual habitats or small er features, su ch as indivi dual estu a ries or seagrass meadows . We focus here on coastal and shel f waters , com bining benthic and shel f pelagic (neri tic) biotas. These waters represent the areas in wh i ch most marine bi od ivers i ty is confined, wh ere human interest and atten ti on are greatest, and wh ere there is often a complex syner gy of t h reats far gre a ter than in offshore waters (UNEP 2006). From a bi od ivers i ty pers pective, it is not simply that coastal and shel f waters have gre a ter s pecies numbers and high er produ ctivity, but also that they are biogeogra phically disti n ct from the adjacent high seas and deep benthic envi ron m ents (Ekman 1953, Hed gpeth 1957a, Bri ggs 1974). Our inten ti on was to devel op a hiera rchical system based on taxon omic con f i g u ra ti on s , influen ced by evo luti on a ry history, patterns of dispers a l , and isolati on . We drew up initial guidelines on def i n i ti ons and nom en clatu re to guide the f i rst data-ga t h ering phase, then reviewed and ref i n ed them itera tively on the basis of the ava i l a ble data. We revi ewed over 230 works in journals, NGO (nongovernmental or ga n i z a ti on) report s , govern m ent publ i c ati ons, and other source s . For each of these, we looked at the u n derlying data and at the process of identificati on and def i n i ti on of bi ogeogra phic units; we also con s i dered the objectives of the classifications. To facilitate comparisons, we used digital mapped vers i ons of many of the ex i s ting bi ogeographic units. More than 40 independent ex perts provided further advice (see the ack n owl ed gm ents section). We ref i n ed a d raft classification scheme thro u ghan assessment and revi ew process that involved a three-day workshop. In arriving at our cl a s s i f i c a ti onsch eme, we ad h ered to three principles for our classification: that it should have a strong bi ogeographic basis, offer practical uti l i ty and be ch a racterized by pars i m ony. ,
www.biosciencemag.org

A p p roaches for defining boundaries

Ob s erva ti ons of gl obal bi ogeogra phic patterns in the mari n e environment inclu de early works by Forbes (1856), Ekman (1953, f i rst publ i s h ed in German in 1935), and Hed gpeth (1957a), and more recent publications by Bri ggs (1974, 1995), Hayden and co lleagues (1984), Ba i l ey (1998), and Lon ghu rs t (1998). These aut h ors used a va ri ety of def i n i ti ons and criteria for drawing biogeogra phic divisions. For example, Bri ggs (1974, 1995) focused on a system of coastal and shelf provinces defined by their degree of en demism (> 10%). This strong taxonomic focus and clear definition have led to rel a tively wi des pre ad adopti on of Briggss sys tem, i n cluding its use by Hayden and co lleagues (1984), with minor amen d m en t s , as a part of their classification of the coastal and marine environments. Adey and Sten eck (2001) provided indepen den t veri f i c a ti on of many of Bri ggss subdivisions in a stu dy that model ed thermogeographicregions of evolution a ry stability. Another important sys tem a tic approach , aimed mainly at pel a gic system s , is the two-tier sys tem devised by Lon ghu rs t (1998), wh i ch focuses on biomes and biogeoch em i c a l provi n ce s . These subdivi s i ons were based on a det a i l edarray of oce a n ographic factors , te s ted and mod i f i edusing a large gl obal database of ch l orophyll profiles. The results repre s ent one of the most com preh en s ive parti ti onings of the pel a gi c biota, but the sch eme is of l i m i te uti l i ty in the complex sysd tems of coastal waters, a fact acknowledged by the aut h or, wh o has recommended combining his open ocean system with others for coastal and shel f w a ters (Wa t s on et al. 2003; Alan R. Longhurst, Galerie lAcademie, Ca jarc, France, pers onal comm n i c a ti on , 2 Novem ber 2004). u The sys tem of large marine ecosystems (LMEs) was devel oped over many years by a number of regi onal experts, with considera ble input from fisheries scien tist Ken Sherman (e.g. , Sherman and Al exander 1989, Hem pel and Sherman 2003,
574 BioScience July/August 2007 / Vol. 57 No. 7

Articles
A strong biogeographic basis. All spatial units were def i n ed on a broadly com p a ra ble bi ogeogra phic basis. Ex i s ting systems rely on a broad array of source inform a ti on ra n ge d i s con ti nuities, dominant habi t a t s , geomorph o l ogical fe atu res, currents, and tem pera tures, for exampleto iden ti f y a reas and boundaries. In many cases these diver gent approaches are compatible, given the close links bet ween bi od iversity and the underlying abi o tic drivers (see the com p a ri s ons bel ow). We preferred to be inform ed by compo s i te studies that combi n ed multiple diver gent taxa or mu ltiple oceanographic drivers in the deriva ti on of boundaries, as these were more likely to captu re robust or rec u rring patterns in overa ll bi od iversity. A nu m ber of s ys tems we revi ewed were broadly bi ogeographic, but with some ad justments to fit political boundaries. Wh ere it was po s s i ble to discern the bi ogeogra phic el ements from the political, these systems were sti ll used to inform the proce s s . Practical utility. We sought to devel op a nested system , operating gl ob a lly at broadly consistent spatial scales and incorporating the full spectrum of habitats found across shelves. We thus avoided very fine-re s o luti on sys tems that sep a ra ted coastal and shel f w a ters into consti tu ent habi t a t s . We ch o s e not to try to define minimum or maximum spatial areas for our bioregions, but in some cases we did seek out systems that subdivi ded very large spatial units (su ch as Bri ggss In do Polynesian Provi n ce , wh i ch covers more than 20% of the worl d s shallow shelf areas) or that amalga m a ted fine-scale units su ch as single large estuaries or sounds. Parsimony. Th ere are a nu m ber of re s pected and widely util i zed gl obal and regi onal system s , and lack of a greem ent between su ch sys tems can be probl ematic. In devel oping a new s ys tem , we sought to minimize further diver gen ce from exi s ting sys tems, yet sti ll to obtain a tru ly gl obalcl a s s i f i c a ti on system .We did this by adopting a nested hiera rchy that (a) utilized systems that are already wi dely adopted (e.g. , the Na tu re Con s ervancys sys tem in mu ch of the Am ericas and the Interim Ma rine and Coastal Regi on a l i s a ti on for Au s tralia) and (b) fitted cl o s ely within broader-scale sys tems or alongside o t h er regi onal sys tems.
evo luti on a ry history. Realms have high levels of en demism, including unique taxa at generic and family l evels in some groups. D riving factors behind the developm ent of su ch unique biotas include water tem peratu re, historical and broadscale isolati on, and the proxi m i ty of the benthos.

This arti cle, with its focus on coastal and shelf areas, doe s not consider realms in pel a gic or deep benthic envi ronments. This is an area requ i ring furt h er analysis and devel opm en t . Provinces. Ne s ted within the realms are provi n ce s :
Large areas def i n ed by the presence of disti n ct biotas that have at least some co h e s i on over evo luti on a ry time frames. Provi n ces will hold some level of en demism, pri n c i p a lly at the level of species. Although histori c a l isolation wi ll play a role, many of these disti n ct biotas have ari s en as a re sult of distinctive abi o tic fe a tu res that circumscri be their boundaries. These may include geom orphological fe a tu res (isolated island and shelf s ys tems, semiencl o s ed seas); hydrogra phic fe a tu res ( c u rrents, upwellings, ice dynamics); or geochemical influences (broadest-scale el em ents of nutrient su pp ly and salinity ) .

In eco l ogical term s , provi n ces are cohesive units likely, for example, to en compass the broader life history of many cons ti tu ent taxa, i n cluding mobile and dispers ive species. In many areas, the scale at wh i ch provi n ces may be conceived is similar to that of the det a i l edspatial units used in gl obal systems su ch as Bri ggss provi n ces, Longhu rs ts bi ogeochemical provinces, and LMEs. Ecoregions. Ecoregi ons are the smallest-scale units in the Ma rine Ecoregi ons of the World (MEOW) sys tem and are def i n ed as fo ll ows :
Areas of relatively hom ogeneous species com po s i ti on , cl e a rly disti n ct from ad jacent systems. The species compo s i tion is likely to be determined by the predom i n a n ce of a small nu m ber of eco s ys tems and/or a disti n ct su i te of oceanogra phic or topographic fe a tu re s . The dom inant bi ogeogra phic forcing agents defining the ecoregi ons vary from loc a ti on to loc a ti on but may include isolation, upwelling, nutri ent inputs, freshwater influx, tem pera tu re regimes, ice regimes, ex po su re , sediments, c u rrents, and bathym etric or coastal complex i ty.

Definitions
Af ter the review proce s s , we arrived at a set of c ri tical working def i n i ti on s . Realms. The systems largest spatial units are based on the terre s trial concept of realms, de s c ri bed by Udvardy (1975) as con ti n ent or su bcon ti n en t - s i zed areas with unifying fe atu res of geogra phy and fauna/flora/veget a ti on . From our m a rine perspective, realms are def i n ed as fo ll ows :
Very large regions of coastal, benthic, or pelagic ocean ac ross wh i ch biotas are intern a lly co h erent at high er taxon omic levels, as a re sult of a shared and unique
www.biosciencemag.org

In eco l ogical term s , these are stron gly co h e s ive units, su fficiently large to encompass ecological or life history processes for most sedentary species. Althoughsome marine ecoregions may have important levels of en dem i s m , this is not a key determinant in ecoregi on iden ti f i c a ti on, as it has been in terre s trial ecoregi on s .
July/August 2007 / Vol. 57 No. 7 BioScience 575

Articles
We su ggest that the most appropri a te outer boundary for these coastal and shel f realms, provinces, and ecoregions is the 200-meter (m) isobath, wh i ch is a widely used proxy for the s h el f ed ge and of ten corre s ponds to a dramatic ecoton e (Forbes 1856, Hed gpeth 1957b, Briggs 1974). Su ch a sharp boundary can on ly be indicative: Shelf breaks are not always cl e a r;the bathym etric location of an equivalent bi o tic tra nsition is highly va ri a bl e ; and there is con s i derable overl a p and influ ence bet ween shel f , slope , and ad jacent pel a gic bi otas. At the same time, most of thecl a s s i f i c a ti ons that we revi ewed have been heavi ly influ enced by data from nearshore and intertidal biotas, and data from deeper water typically had dec reasing influ en ce on boundary def i n i ti on s . We bel i eve that beyond 200 m, o t h er bi ogeogra phic patterns wi ll increasingly predominate, altering or hiding the patterns represented by the sys tem propo s ed here. with a stronger bi ogeographic basis than the current LME delineations. Both the Bri ggs and Hayden sys tems and the LMEs show con s i dera ble va ri a ti on in the size of t h eir spatial units; the Briggs approach of using 10% en demism disti nguishes many isolated commu n i ties around oceanic islands, but fails to disaggregate vast areas with gradual faunal changes, even wh ere the increm ental ef fects of su ch ch a n ges are very l a r geindeed (e.g., the Indo-Pacific). The large spatial units in a ll of these systems clearly encompass significant levels of internal biogeographic heterogeneity, which we were keen to disa ggrega te thro u gh a more detailed system of ecoregi on s . We found regi onal sys tems for almost all coastal and shel f w a ters , a l t h o u ghmany are de s c ri bed on ly in the gray literatu re. No t a ble excepti ons were the Russian Arctic and the continental coasts of mu ch of So uth, So utheast, and East Asia. For these areas, we rel i ed heavi ly on gl obal data sets and u n p u bl i s h ed ex pert opinion, using more focused bi ogeogra phic publ i c a ti ons (where available) for refining indivi dual boundaries. Figure 1 dep i cts the review proce s s , showing four bi ogeogra phic sch em e s : Bri ggss system of provi n ces (1974, 1995); an ex pert-derived sys tem com bining bi o tic and abi o tic fe atu res for South Am erica (Sull ivan Sealey and Bustamante 1999); the current LMEs; and a regi onal cl a s s i f i c a ti onbased on a single taxon omic grouping (decapod crustaceans; Bo s chi 2000). Despite their different ori gi n s , these systems show a re-

A global, nested system

We propose a nested system of 12 realms, 62 provinces, and 232 ecoregi ons covering all coastal and shel f w a ters of the worl d . As the MEOW sys tem is based on ex i s ting classifications, va ri a ti on and mismatch among sys tems led to ch a ll en ge s and compromises. The gl obal coastal classification of Bri ggs s and Hayden , for example, do not show great con gruen ce with the LMEs. The Briggs and rel a ted Hayden sys tems appe a red to be more cl o s ely all i ed to our need for a sys tem

Fi g u re 1. Re co n ci l i a tion of d i f f ering boundary systems for South America. The map on the left ill u s tra tes fou r bi ogeographic sys tems: (A) Bri ggss provinces, (B) Sull ivan Sea l ey and Bustamantes provinces, (C) large marine eco s ys tem s , and (D) Bo sch i s provinces. Sys tem similari ties are ex em plified in three inset maps: nort h ern Peru (inset 1), Cabo Frio (inset 2), and Chilo Island (inset 3). The map on the ri ght shows the Marine Eco regions of the Wo rld provinces (labeled) and their eco region subdivision bou n d a ries.

576 BioScience July/August 2007 / Vol. 57 No. 7

www.biosciencemag.org

Articles
markable congru en ce at a nu m ber of key bi ogeogra ph i c boundaries. Thus, it was po s s i ble to adopt a single sys tem as a primary source, and the MEOW provi n ces (figure 1, ri ght) were b a s ed almost en ti rely on Su ll ivan Se a l ey and Bu s t a m a n te (1999), while remaining well align ed with the other system s . At a finer resolution, the ecoregions for South America are derived almost en ti rely from the same publ i c a ti on (Su ll iva n Se a l ey and Bu s t a m a n te 1999), this being the only com prehensive sys tem for these coasts. Even at this scale, however, ef forts were made to loc a te indepen dent verification of boundaries, and it is re a s su ring to note that these more det a i l edsubdivisions were often su pported by data from other oce a n ographic and eco l ogical litera ture (see , e.g., S trub et al. [1998], Fern a n dez et al. [2000], Ojeda et al. [2000], and Ca mus [2001] for data concerning the Chilean coast). Al t h o u gh the boundaries in other regi ons were not as simple to resolve as those along the So uth Am erican coa s t , we app l i ed the same approach e s . The secti on that fo llows gives some information on the key sources used in drawi n g boundaries.

Marine Ecoregions of the World

Box 1 and figures 2 and 3 give a summary of the en ti re MEOW system , wh i ch covers all coastal and shelf waters shallower than 200 m. The shaded area of e ach map (figure s 2, 3) ex tends 370 kilom eters (200 nautical miles) of fs h ore (or to the 200-m isobath, where this lies further offshore),

Fi g u re 2. Final bi ogeographic framewo rk: Realms and provi n ce s . (a) Biogeographic realms with eco region boundaries outlined. (b) Provinces with eco regions outlined. Provi n ces are nu m bered and listed in table 1.

www.biosciencemag.org

July/August 2007 / Vol. 57 No. 7 BioScience 577

Articles
Box 1. Marine Ecoregions of the Wo rl d .
N u m b e rs for the provinces and ecoregions match those shown on the maps in figures 2b and 3. Realms are indicated in boldface, provinces (162) in italics, and ecoregions (1232) in roman type. Arctic 1. Arctic (no provinces identified) 1. North Greenland 2. North and East Iceland 3. East Greenland Shelf 4. West Greenland Shelf 5. Northern Grand BanksSouthern Labrador 6. Northern Labrador 7. Baffin BayDavis Strait 8. Hudson Complex 9. Lancaster Sound 10. High Arctic Archipelago 11. BeaufortAmundsenViscount MelvilleQueen Maud 12. Beaufort Seacontinental coast and shelf 13. Chukchi Sea 14. Eastern Bering Sea 15. East Siberian Sea 16. Laptev Sea 17. Kara Sea 18. North and East Barents Sea 19. White Sea Temperate Nort h e rn Atlantic 2. Northern European Seas 20. South and West Iceland 21. Faroe Plateau 22. Southern Norway 23. Northern Norway and Fi n n m a rk 24. Baltic Sea 25. North Sea 26. Celtic Seas 3. Lusitanian 27. South European Atlantic Shelf 28. Saharan Upwelling 29. Azores Canaries Madeira 4. M e d i t e rr nean Sea a 30. Adriatic Sea 31. Aegean Sea 32. Levantine Sea 33. Tunisian Plateau/Gulf of Sidra 34. Ionian Sea 35. Western Mediterranean 36. Alboran Sea 5. Cold Temperate Northwest Atlantic 37. Gulf of St. LawrenceEastern Scotian Shelf 38. Southern Grand BanksSouth Newfoundland 39. Scotian Shelf 40. Gulf of Maine/Bay of Fundy 41. V rginian i 6. Warm Temperate Northwest Atlantic 42. Carolinian 43. Northern Gulf of Mexico 7. Black Sea 44. Black Sea Temperate Nort h e rn P c i f i c a 8. Cold Temperate Northwest Pacific 45. Sea of Okhotsk 46. Kamchatka Shelf and Coast 47. Oyashio Curr e n t 48. Nort h e a s t e rn Honshu 49. Sea of Japan 50. Yellow Sea 9. Warm Temperate Northwest Pacific 51. Central Kuroshio Current 52. East China Sea 10. Cold Temperate Northeast Pacific 53. Aleutian Islands Gulf of Alaska North American Pacific Fijordland Puget Trough/Georgia Basin Oregon, Washington, Vancouver Coast and Shelf 58. Nort h e rn Californ i a 11. Warm Temperate Northeast Pacific 59. S o u t h e rnCalifornia Bight 60. Cortezian 61. Magdalena Transition Tropical Atlantic 12. Tropical Northwestern Atlantic 62. Bermuda 63. Bahamian 64. E a s t e rn Caribbean 65. Greater Antilles 66. S o u t h e rn Caribbean 67. Southwe s t e rn Caribbean 68. We s t e rn Caribbean 69. S o u t h e rnGulf of Mexico 70. Floridian 13. N o rth Brazil Shelf 71. Guianan 72. Amazonia 14. Tropical Southwestern Atlantic 73. Sao Pedro and Sao Paulo Islands 74. Fe rnando de Naronha and Atoll das Rocas 75. Nort h e a s t e rn Brazil 76. E a s t e rn Brazil 77. Trindade and Martin Vaz Islands 15. St. Helena and Ascension Islands 78. St. Helena and Ascension Islands 16. West African Transition 79. Cape Verde 80. Sahelian Upwelling 17. Gulf of Guinea 81. Gulf of Guinea West 82. Gulf of Guinea Upwelling 83. Gulf of Guinea Central 84. Gulf of Guinea Islands 85. Gulf of Guinea South 86. Angolan Western Indo-Pacific 18. Red Sea and Gulf of A d e n 87. Nort h e rn and Central Red Sea 88. S o u t h e rnRed Sea 89. Gulf of Aden 19. Somali/Arabian 90. Arabian (Pe rsian) Gulf 91. Gulf of Oman 92. We s t e rn Arabian Sea 93. Central Somali Coast 20. Western Indian Ocean 94. Nort h e rn Monsoon Current Coast 95. East African Coral Coast 96. Seychelles 97. Cargados Carajos/Tromelin Island 98. Mascarene Islands 99. Southeast Madagascar 100. We s t e rn and Northern Madagascar 101. Bight of Sofala/Swamp Coast 102. Delagoa 21. West and South Indian Shelf 103. We s t e rn India 104. South India and Sri Lanka 22. Central Indian Ocean Islands 105. Maldives 106. Chagos 54. 55. 56. 57. 23. Bay of Bengal 107. E a s t e rn India 108. Nort h e rn Bay of Bengal 24. Andaman 109. Andaman and Nicobar Islands 110. Andaman Sea Coral Coast 111. We s t e rn Sumatra Central Indo-Pacific 25. South China Sea 112. Gulf of Tonkin 113. S o u t h e rn China 114. South China Sea Oceanic Islands 26. Sunda Shelf 115. Gulf of Thailand 116. S o u t h e rn Vietnam 117. Sunda Shelf/Java Sea 118. Malacca Strait 27. Java Transitional 119. S o u t h e rn Java 120. Cocos-Keeling/Christmas Island 28. South Kuroshio 121. South K roshio u 29. Tropical Northwestern Pacific 122. Ogasawara Islands 123. Mariana Islands 124. East Caroline Islands 125. West Caroline Islands 30. Western Coral Triangle 126. Palawan/North Borneo 127. Eastern Philippines 128. Sulawesi Sea/Makassar Strait 129. Halmahera 130. Papua 131. Banda Sea 132. Lesser Sunda 133. Northeast Sulawesi 31. Eastern Coral Triangle 134. Bismarck Sea 135. Solomon Archipelago 136. Solomon Sea 137. Southeast Papua New Guinea 32. Sahul Shelf 138. Gulf of Papua 139. Arafura Sea 140. A rnhem Coast to Gulf of Carp e n t e r i a 141. Bonaparte Coast 33. N o rtheast Australian Shelf 142. Torres Strait Nort h e rn Great B a rrier Reef 143. Central and Southern Great B a rrier Reef 34. N o rthwest Australian Shelf 144. Exmouth to Broome 145. Ningaloo 35. Tropical Southwestern Pacific 146. Tonga Islands 147. Fiji Islands 148. Vanuatu 149. New Caledonia 150. Coral Sea 36. Lord Howe and Norfolk Islands 151. Lord Howe and Norfolk Islands Eastern Indo-Pacific 37. Hawaii 152. Hawaii 38. Marshall, Gilbert, and Ellis Islands 153. M a rshall Islands 154. Gilbert/Ellis Island

578 BioScience July/August 2007 / Vol. 57 No. 7

www.biosciencemag.org

Articles
Box 1. (continued)
N u m b e rs for the provinces and ecoregions match those shown on the maps in figures 2b and 3. Realms are indicated in boldface, provinces (162) in italics, and ecoregions (1232) in roman type. 39. Central Po l y n e s i a 155. Line Islands 156. Phoenix/Tokelau/Nort h e rn Cook Islands 157. Samoa Islands 40. Southeast Polynesia 158. Tuamotus 159. Rapa-Pitcairn 160. S o u t h e rn Cook/Austral Islands 161. Society Islands 41. Marquesas 162. Marquesas 42. Easter Island 163. Easter Island Tropical Eastern Pacific 43. Tropical East Pacific 164. Revillagigedos 165. Clipperton 166. Mexican Tropical Pacific 167. ChiapasNicaragua 168. Nicoya 169. Cocos Islands 170. Panama Bight 171. Guayaquil 44. Galapagos 172. Nort h e rn Galapagos Islands 173. E a s t e rn Galapagos Islands 174. We s t e rn Galapagos Islands Temperate South America 45. Warm Temperate Southeastern Pacific 175. Central Pe ru 176. Humboldtian 177. Central Chile 178. Araucanian 46. Juan Fernndez and Desventuradas 179. Juan Fernndez and Desventuradas 47. Warm Temperate Southwestern Atlantic 180. S o u t h e a s t e rn Brazil 181. Rio Grande 182. Rio de la Plata 183. Uruguay-Buenos Aires Shelf 48. Magellanic 184. North Patagonian Gulfs 185. Patagonian Shelf 186. Malvinas/Falklands 187. Channels and Fjords of S o u t h e rn Chile 188. Chiloense 49. Tristan Gough 189. Tristan Gough Temperate Southern Africa 50. Benguela 190. Namib 191. Namaqua 51. Agulhas 192. Agulhas Bank 193. Natal 52. AmsterdamSt Paul 194. AmsterdamSt Paul Temperate Australasia 53. Northern New Zealand 195. Ke rmadec Island 196. Nort h e a s t e rnNew Zealand 197. Three KingsNorth Cape 54. Southern New Zealand 198. Chatham Island 199. Central New Zealand 200. South New Zealand 201. Snares Island 55. East Central Australian Shelf 202. Tweed-Moreton 203. Manning-Hawkesbury 56. Southeast Australian Shelf 204. Cape Howe 205. Bassian 206. Western Bassian 57. Southwest Australian Shelf 207. South Australian Gulfs 208. Great Australian Bight 209. Leeuwin 58. West Central Australian Shelf 210. S h a rk Bay 211. Houtman S o u t h e rn Ocean 59. Subantarctic Islands 212. Macquarie Island 213. Heard and Macdonald Islands 214. Kerguelen Islands 215. Crozet Islands 216. Prince Edward Islands 217. Bouvet Island 218. Peter the Fi rst Island 60. Scotia Sea 219. South Sandwich Islands 220. South Georgia 221. South Orkney Islands 222. South Shetland Islands 223. Antarctic Peninsula 61. Continental High Antarctic 224. East Antarctic Wilkes Land 225. East Antarctic Enderby Land 226. East Antarctic Dronning Maud Land 227. Weddell Sea 228. Amundsen/Bellingshausen Sea 229. Ross Sea 62. Subantarctic New Zealand 230. Bounty and Antipodes Islands 231. Campbell Island 232. Auckland Island

but , as alre ady noted , we con s i der the principal focus of this classification to be the benthos above 200 m and the overlying w a ter column. Key sources inclu ded the fo llowi n g :
Bi ogeogra phic assessments in the peer-revi ewed litera tu re, including the global studies alre ady m en ti oned and many regi onal publ i c a ti ons (e.g. , Bustamante and Branch [1996] and Turpie et al. [2000] for tem pera te sout h ern Africa, Linse et al. [2006] for the So ut h ern Ocean) Ecoregi onal assessments con du cted by NGOs (e.g. , Su ll ivan Sealey and Bustamante [1999] for Latin America, WWF [2004 and unpubl i s h ed reports] for mu ch of Africa, Green and Mous [2006] for the Coral Tri a n gle provinces) Governmen t - derived or su pported systems (e.g. , Th ackway and Cre s s well [1998] for Australia, Powles et al. [2004] for Canada) Input from several of the aut h ors of this arti cle and assessments commission ed explicitly for the MEOW
www.biosciencemag.org

process (e.g. , unpubl i s h ed reports by Jerry M. Kemp in 2005 for the Mi d dle Eastern seas and by S. A. L. in 2006 for the Andaman to Java coasts); the system for the In do-Pacific oceanic islands was devel oped by one of us (G. R. A.) on the basis of many years of f i eld ex peri ence, ex pert revi ew, and net working with other scientists ac ross the regi on

These sch emes were assessed alongside other bi ogeogra ph i c litera tu re, and in some cases altera ti ons were made to better repre s ent the arguments of bi ogeogra phy, utility, and pars im ony out l i n edabove. A full listing of the sources referen ced can be found at www.nature.org/MEOW or www.worldwildlife. org/MEOW. The propo s ed realms adopt the broad lati tudinal divis i ons of po l a r, tem perate, and tropical, with su b d ivi s i on s based on ocean basin (broadly fo llowing the oceanic bi om e s of Longhu rst [1998]). In the tem pera te waters of the So uthern Hemisphere, we diverge from this approach. We consider the differences ac ross the oceans too substantial, and the con n ecti ons around the con ti n ental margins too great, to su pport ei t h er ocean basin subdivi s i ons or a single circ u mgl obal realm (equ iva l ent to Lon ghu rs ts An t a rctic We s terly Winds Bi ome), and hen ce we have adopted con ti n ental
July/August 2007 / Vol. 57 No. 7 BioScience 579

Articles

580 BioScience July/August 2007 / Vol. 57 No. 7

www.biosciencemag.org

Articles
m a r gin realms for tem pera te Australasia, sout h ern Af ri c a , and South Ameri c a . The paucity of ex i s ting literature discussing these broadest-scale bi ogeogra phic units from a gl obal perspective pre s ents a stark con trast to the terre s tri a l bi ogeogra phic litera tu re . The level of internal heterogen eity of biotas within different realms is qu i teva ri ed . For some realms, the differen ces in biota at the provincial level are substantial, including the warm tem pera te faunas on either side of the Temperate South Am erica realm and the tropical faunas on ei t h er side of the Tropical Atlantic realm. By contrast, we have subdivi ded the widely used Indo-Pacific realm into three units. This is the regi on of greatest diversity, and it covers a vast area. Across this regi on are clinal changes in taxa that lack clear breaks, but are sufficiently large that faunas at ei t h er end bear little re s embl a n ce to each other. Our In do-Pacific su b d ivi s i ons (wh i ch it might be appropri a te to con s i der as su brealms) fo ll ow less clearly def i n ed bi ogeogra phic bo u n d a ries than other realms, but these divisions produ ce spatial units that are more com p a ra ble to other realms in overa ll bi od ivers i ty, l evels of en dem i s m , and spatial are a . At broader scales, we undertook a simple spatial analysis to explore the links or possible crossovers between the MEOW s ys tem , LMEs, and Bri ggss provi n ce s . The incom p l ete covera ge of the LME sys tem is cl e a rly limiting for gl obal cons erva ti on planning: 78 of our 232 ecoregions inclu de a substantive area (greater than 10% of their total area) that is not covered by any LME. Of the remainder, s ome 49% of LMEs show good congru ence (> 90% of shelf a rea) with either single ecoregi ons or ecoregi on com bi n a ti on s . (Th e bo u n d a ryof the Arctic LME has not been mapped, and so was i gn ored in these calculati ons.) In comparison, 30 of Bri ggss 53 provi n ces (57%) show good congru ence (> 90% of shelf area) with single ecoregi ons or ecoregi on combinations. Th i s f i g u rerises to 39 (74%) if we inclu de con gru en ce at 85% of the shel f area. We also used the MEOW system to look at the covera ge of the marine and coastal net work of Ramsar sites. Con tracti n g p a rties to the Ramsar Conven tion have com m i t te to ach i eve d a coherent and comprehensive national and international network (Ramsar Conven ti on 1999), a l t h o u ghuntil now it has not been po s s i ble to assess the bi ogeogra phic coverage of m a rine and coastal Ramsar sites at the gl obal level. The results of this overl ay are pre s en ted in table 1. One value of biogeographic classifications is their use in uncovering inequ i ties and dra m a tic gaps in con s erva ti on covera ge .Al t h o u gha more thoro u gh analysis would be requ i red to determine more clearly the degree of repre s en t a ti on provi ded by the existing sel ecti on of Ramsar sites, s ome basic observa ti ons are immed i a tely apparen t . The Ramsar net work is ex ten s ive, but it is dominated by sites in the tem pera te Nort h Atlantic and shows a striking paucity of sites in, for example, the eastern Indo-Pacific and the Sout h ern Ocean. At finer hiera rchical resolution, further gaps can be identified: While 92% of realms are repre s en ted, this translates to only 73% of provinces and 52% of ecoregions, leaving some 112 ecoregions with no Ramsar representation. These gaps are wi de s pre ad , i n cluding four ecoregi ons in the tem pera te North Atlantic.

Conclusions
The MEOW cl a s s i f i c a ti onprovides a critical tool for mari n e con s erva ti on planning. It will enable gap analyses and assessments of repre s en t a tiveness in a global fra m ework . It provi des a level of detail that wi ll su pport linkage to practical conserva ti on interven ti ons at the field level. For example, t wo major intern a ti onal con s erva ti on organizations (the Na tu re Con s ervancy and WWF) use ecoregi ons as planning units. From a global standpoint, the MEOW system offers similar opportunities for the marine envi ronment. It also provi des a rational framework in wh i ch to analyze patterns and processes in coastal and shel f bi od ivers i ty. The gl obal and hiera rchical natu re of the MEOW can su pport analytical approaches that move bet ween scales. Using MEOW, gl obal information can also be used to target acti on on the gro u n d , while fiel d - l evel inform a ti on can be p l aced alongside information on ad jacent or remote locations,

Table 1. The geographic spread of m a rine and coastal Ramsar sites within the Ma rine Eco regions of the Wo rl d classification.
Ecoregions Total Ramsar sites
26 374 38 117 41 35 1 29 14 9 25 0 709

Provinces Percentage with Ramsar sites

53 84 71 68 56 40 8 73 60 60 53 0 52

Realm
Arctic Temperate Nort h e rn Atlantic Temperate Nort h e rn Pacific Tropical Atlantic We s t e rn Indo-Pacific Central Indo-Pacific E a s t e rn Indo-Pacific Tropical Eastern Pacific Temperate South America Temperate Southern Africa Temperate Australasia S o u t h e rnOcean Total

Number with Ramsar sites

10 21 12 17 14 16 1 8 9 3 9 0 120

Total number
19 25 17 25 25 40 12 11 15 5 17 21 232

Number with Ramsar sites

1 6 4 4 7 10 1 2 3 2 5 0 45

Total number
1 6 4 6 7 12 6 2 5 3 6 4 62

Percentage with Ramsar sites

100 100 100 67 100 83 17 100 60 67 83 0 73

www.biosciencemag.org

July/August 2007 / Vol. 57 No. 7 BioScience 581

Articles
providing a wider spatial pers pective. Rooted in ex i s ting regi onal system s , the base units of the MEOW alre ady underpin con s erva ti on ef forts at regi onal level s , and a strong body of m a rine ecoregi onal planning literatu re illu s tra tes how gl obal or regi onal con cerns can be converted into field-based con s erva ti on action (Banks et al. 2000, Beck and Odaya 2001, Lars en et al. 2001, Kra m er and Kra m er 2002, Ferdaa 2005). The va lue of the MEOW system ex tends beyond con s erva ti on planning. Looking afresh at the broader-scale cl a s s e s and taking adva n t a ge of the improved resoluti on of fered by the MEOW system, it is possible to revi ew wi der issues of biodiversity distri bution and evolution. At the broadest scales, the most important el ements of bi ogeographic subdivision are the b a rri ers that have sep a ra ted su b s t a n tial areas over evo luti onary timescales (Adey and Sten eck 2001). In the MEOW realms (noting the special case of the Indo-Pacific de s c ri be d a bove ) , these barri ers consist of landmasses, wi de ocean basins, and tem pera tu re grad i ents. Al t h o u ghthere is variati on in degree , the provi n ces can be s een as finer-scale units of evoluti onary isolati on . Th ey align with many of the more important factors driving recent and contem pora ry evoluti onary processes. Tem pera tu re , or latitu de, con ti nues to play an important role (sep a ra ting warm and cold tem pera te provi n ce s ) , but so does the furt h er isolation provided by deep water, n a rrow straits, or rapid changes in shelf conditions. E l s ewhere , the con n ectivi ty provi ded by ocean currents, such as the Antarctic Coastal Cu rrent and the Ca n a ries Curren t , can be seen in the classifications, and the i m port a n ce of bi o l ogical stepp i n g - s tones thro u gh va ri o u s island chains is cl e a rly illu s trated . F i n a lly, the ecoregi on s , wh i ch distinguish the MEOW sys tem, ref l ect unique eco l ogical patterns that ex tend beyond the broad drivers of evoluti onary processes. Of course, as Wa ll ace (1876) noted ,nothing like a perfect zoological division of the earth is possible. The causes that have led to the present distri bution of animal life are so varied, their action and reaction have been so complex, that anomalies and i rreg u l a ri ties are su re to exist wh i ch wi ll mar the sym m etry of any ri gid system (p. 53). Con s equen t ly, the use of bi ogeogra phic data in a gl obal classification is inevi t a bly a process of accommodation and pra gmatism. The lines we have drawn should be regarded as indicative, m a rking approx i m a te locations of relatively rapid change in dominant habitats or commu n i ty composition. Ocean bo u n d a ries shift con ti nu o u s ly with we a t h erpattern s , with seasons, and with longer or more random flu ctu a ti ons in oce a n ographic conditions. In the futu re, the impacts of cl i m a te change wi ll add to the instabi li ty of many bo u n d a ries in the ocean (Sa garin et al. 1999, Be a u grand et al. 2002, Hiscock et al. 2004). The need for a compreh ensive, det a i l ed , and glob a lly cons i s te marine bi ogeogra phy has been recogn i zed for many nt years in marine con s erva ti on. The requ i rements for representative approaches to marine protected area de s i gn a ti on in va rious national, regi onal, and gl obal planning commitments and legal fra m eworks have given ad ded urgency to this need . The MEOW system provi des a basis for planning for coastal
582 BioScience July/August 2007 / Vol. 57 No. 7

and shel f areas, and the links bet ween this sys tem and other gl obal and regional systems make it possible to adopt and use it with minimal disru pti on to ex i s ti g data sets or analytical n approach e s . The unique co ll a bora ti on of conserva ti on or ganizations in devel oping this system adds further value, and may redu ce the du p l i c a ti on of ef fort that so often undermines gl obal conserva ti on approaches (Mace et al. 2000). In short, the system propo s ed here is powerful and robu s t , and should prove to be of great va lue in con s erva ti on planning and broader bi ogeogra phic discussion. Two intern a ti onal conserva ti on agencies (the Nature Conservancy and WWF) have already begun to use this sys tem and ex pect to use it more widely in the futu re. Similarly, m em bers of the Scien tific and Technical Revi ew Pa n el of the Ramsar Conven ti on who parti c i p a ted in developing this sys tem are undertaking more det a i l ed analyses to ex p l ore its utility to support the futu re identification and designation of coastal and marine Wetlands of Intern a ti onal Im portance.

Acknowledgments
The Ma rine Ecosystems of the World system draws heavi ly on the work of others , i n cluding the hu n d reds of con tri butors to the publicati on s , gray literatu re , and work s h ops that cre a ted the many regi onal classifications. In ad d i tion, we would especially like to thank the fo ll owing people, who have provi ded advi ce or commentary: Asa Andersson, Jeff Ardron, All i s on Arnold, Paul Ba rber, Mike Beck, Ca rlo Nike Bianchi, John Bo l ton, G eor ge Bra n ch , John Bri ggs , G eor gina Bu s t amente, Rod ri go Bu s t a m en te , Jose Fa ri n a , Ser gio Floeter, Angus Gascoi gn e , Ser ge Gof a s , Charlie Griffiths, Huw Griffiths, Ra n dy Ha gen s tei n , Jon Hoe k s tra, D avid Jo h n , Peter Kareiva, Ken Kassem, Jerry Kem p, Phil Kra m er, Ka trin Linse, G i lly Llewelly n , Stephan Lut ter, Kasim Moosa, Alexis Morgan, Dag Nagoda, Ser gio Navarete, Kate Newm a n , (Bina) Maya Paul, Sian Pu ll en, Ca llum Roberts, Rod Salm, Andrew Smith, Jennifer Smith,Vassily Spiri donov,Vi ctor Springer, Juan Luis Su rez de Vivero, Ma rco Taviani, Ch a rlie Veron, Eleni Vo u l t s i adou, Mo h i deen Wafar, Ca rden Wall ace , Kathy Walls, and David Woodland.

R e f e re nces cited
Adey WH, S ten eck RS. 2001. Th erm ogeogra phy over time cre a tes bi ogeogra phic regi on s : A tem perature/space/time-integra ted model and an abu n d a n ce - wei gh ted test for benthic marine algae. Jo u rnal of Phyco l ogy 37: 677698. Ba i l ey RG. 1998. Ecoregions: The Ecosystem Geogra phy of the Oceans and Continents. New York : Springer. Banks D, Williams M, Pearce J, Springer A, Hagen s tein R, Ol s onD, ed s . 2000. Ecoregi on-based Con s erva ti on in the Bering Se a : Iden ti f ying Import a n t Areas for Biod ivers i ty Con s erva ti on . Washington (DC ) : World Wi l dl i fe Fu n d , The Na tu re Con s ervancy of Alaska. Beaugrand G, Reid PC, Ibaez F, Lindl ey JA, Edwards M. 2002. Reorganization of North At l a n tic marine copepod bi od ivers i ty and climate. Science 296: 16921694. Beck MW, Odaya M. 2001. Ecoregi onal planning in marine environments: Iden ti f ying pri ority sites for con s erva ti on in the nort h ern Gu l f of Mexico. Aqu a tic Con s erva ti on: Marine and Fre s hwater Ecosystems 11: 235242.

www.biosciencemag.org

Articles
Bo s chi E. 2000. S pecies of dec a pod cru s t aceans and their distri buti on in the Am erican marine zoogeogra phic provinces. Revista de Investigacin y De s a rro llo Pesquero 13: 7136. Bri ggs JC. 1974. Ma rine Zoogeogra phy. New York : McGraw-Hi ll . . 1995. G l obal Biogeogra phy. Am s terdam: Elsevier. Bu r gess ND, Hales JDA, Ri cketts TH, Di n ers tein E. 2006. Factoring species, n on - s pecies va lues and threats into bi od ivers i ty prioritisation ac ross the ecoregions of Af rica and its islands. Biological Con s erva ti on 127: 383401. Bu s t a m a n te RH, Bra n ch GM. 1996. Large scale patterns and trophic s tru cture of So ut h ern Af rican rocky shore s : The roles of geogra phic va riati on and wave ex po su re. Journal of Bi ogeogra phy 23: 339351. Ca mus PA. 2001. Biogeografa marina de Chile con tinental. Revista Ch i l ena de Hi s toria Natural 74: 587617. Ch a pe S, Blyth S, Fish L, Fox P, Spalding M. 2003. 2003 United Na ti ons List of Pro tected Are a s . Gland (Swi t zerland): IUCNWorld Con s ervation Union; Ca m bri d ge (Un i ted Kingdom ) : UNEP World Con s erva ti on Mon i toring Centre. Ch a pe S, Ha rrison J, Spalding M, Lys en ko I. 2005. Measu ring the ex tent and ef fectiveness of pro tected areas as an indicator for meeting gl obal bi od ivers i ty target s . Proceed i n gs of the Royal Soc i ety B 360: 443455. Ekman S. 1953. Zoogeography of the Sea. Lon don : Si d g wi ckand Jackson. Ferdaa Z. 2005. Ne a rs h ore marine con s erva ti on planning in the Pacific Northwest: Ex p l oring the uti l i ty of a siting algorithm for repre s en ting marine bi od ivers i ty. Pa ges 150172 in Wright DJ, Scholz AJ, eds. P l ace Ma t ters : Geospatial Tools for Ma rine Scien ce , Con s erva ti on, and Management in the Pacific Northwest. Corva ll i s : O regon State Univers i ty. Fern a n dez M, Jara m i llo E, Marqu et PA, Moreno CA, Nava rrete SA, Ojeda FP, Valdovinos CR,Vasquez JA. 2000. Divers i ty dynamics and bi ogeogra phy , of Chilean benthic nears h ore ecosystems: An overvi ew and guidelines for con s erva ti on. Revista Ch i l en de Hi s toria Na tu ral 73: 797830. a Forbes E. 1856. Map of the distri buti on of marine life. Pages 99102 and plate 131 in Jo h n s ton AK, ed. The Physical Atlas of Na tu ral Phen om en a . E d i n bu r gh (Scotland): Wi lliam Blackwood and Son s . Green A, Mous P. 2006. Del i n e a ting the Coral Tri a n gl e , Its Ecoregi ons and Fu n ctional Seascapes. Report Ba s ed on an Ex pert Work s h op Held at the TNC Coral Tri a n gle Cen ter, Bali, In donesia (AprilMay 2003), and on Ex pert Consultations Held in June and August 2005. Vers i on 3.1 (Febru a ry 2006). Bali (Indon e s i a ) : The Na tu re Con s erva n c y, Coral Tri a n gle Cen ter; Brisbane (Australia): G l obal Ma rine In i ti a tive , In do Pacific Re s o u rce Cen tre. Hayden BP, Ray GC, Dolan R. 1984. Classification of coastal and marine envi ron m en t s . Environ m ental Con s erva ti on 11: 199207. Hazen HD, Anthamatten PJ. 2004. Repre s en t a ti on of eco l ogical regi ons by pro tected areas at the gl obal scale. Physical Geogra phy 25: 499512. Hed gpeth JW. 1957a. Ma rine bi ogeogra phy. G eo l ogical Soc i ety of Am erica Mem oi rs 67: 359382. . 1957b. C assification of m a rine environ m ents. G eo l ogical Soc i ety l of Am erica Mem oi rs 67: 1728. Hem pel G, Sherman K, ed s . 2003. L a r ge Ma rine Eco s ys tems of the Worl d : Trends in Ex p l oitation, Pro tecti on, and Research . Am s terdam: E l s evier. Hiscock K, So ut hward A, Ti t t l ey I, Hawkins S. 2004. E f fects of ch a n ging tem pera tu re on benthic marine life in Britain and Irel a n d . Aqu a tic Con s erva ti on : Marine and Fre s hwater Eco s ys tems 14: 333362. Hoe k s tra JM, Bo u ch er TM, Ri cketts TH, Roberts C. 2005. Con f ron ting a bi ome cri s i s : G l obal dispari ties of habitat loss and pro tecti on. Eco l ogy Let ters 8: 2329. Kell eh er G, Bl e a k l eyC, Wells S, eds. 1995. A Global Repre s en t a tive System of Marine Protected Areas, vols. 24.Washington (DC ) : Great Ba rrier Reef Marine Park Authori ty,World Bank, IUCN (World Conservation Union). Kra m er PA, Kra m er PR. 2002. Ecoregional Con s erva tion Planning for the Mesoamerican Ca ri bbean Reef. Washington (DC ) : World Wildlife Fund. L a m oreux JF, Morri s on JC, Ri cketts TH, Ol s on DM, Diners tein E, McKnight MW, Shu ga rt HH. 2006. Global tests of bi od ivers i ty con cordance and the i m portance of en dem i s m . Nature 440: 212214. L a rs en T, Na goda D, An ders on JR, ed s . 2001. The Ba rents Sea Ecoregi on: A Biod ivers i ty Assessment. Oslo (Norw ay): World Wi l dl i fe Fund. Linse K, Griffiths HJ, Ba rnes DKA , Cl a rke A. 2006. Biod iversity and bi ogeography of Antarctic and sub-Antarctic mollusca. Deep Sea Research II 53: 9851008. Lon ghu rst A. 1998. Eco l ogical Geogra phy of the Se a . San Di ego : Ac adem i c Press. Lo u rie SA, Vincent ACJ. 2004. Using bi ogeogra phy to help set pri ori ties in m a rine con s erva ti on. Con s erva ti on Biology 18: 10041020. Mace GM, et al. 2000. Its time to work toget h er and stop duplicating con s erva ti on ef fort s . Na tu re 405: 393. O j eda FP, L a bra FA, Muoz A A . 2000. Biogeogra phic patterns of Chilean l i t toral fishes. Revista Ch i l en de Historia Na tu ral 73: 625641. a Ol s on DM, Di n ers tein E. 2002. The Global 200: Priori ty ecoregi ons for con s erva ti on. Annals of the Mi s s o u ri Botanical Garden 89: 199224. Ol s on DM, et al. 2001. Terre s trial ecoregi ons of the worl d : A new map of life on Eart h . BioScience 51: 933938. Powles H,Ven det te V, Si ron R, OBoyle B. 2004. Proceedings of the Canadian Marine Ecoregi ons Work s h op. Ot t awa (Ca n ada): F i s h eries and Oceans Ca n ada. Ramsar Conven ti on . 1999. S trategic Fra m ework and guidelines for the f uture devel opment of the List of Wetlands of In ternati onal Im portance. Re s o luti on VII.11 of the 7th Con feren ce of the Con tracting Parties. (1 June 2007; www.ramsar.org/res/key_res_vii_index.htm) Sa ga rin RD, Barry JP, Gilman SE, Ba x ter CH. 1999. Cl i m a te rel a ted changes in an intertidal com munity over short and long time scales. Eco l ogical Mon ogra phs 69: 465490. S h erman K, Alex a n der LM. 1989. Biomass Yi elds and Geography of Large Marine Eco s ys tems. Bo u l der (CO): We s t vi ewPress. S h erman K, et al. 2005. A gl obal movem ent toward an ecosystem approach to management of marine resources. Marine Eco l ogy Progress Series 300: 275279. S pell erberg IF, Sawyer JWD. 1999. An Introducti on to Applied Biogeogra phy. Ca m bri d ge (Un i ted Kingdom ) : Ca m bri d ge Univers i ty Press. S trub PT, Mesas JM, Mon tecino V, Rut llant J, Salinas S. 1998. Coastal ocean circulation of f we s tern So uth Ameri c a . Pa ges 273313 in Robi n s on A , Brink K, ed s . The Se a , vo l . 14A: The Global Coastal Ocean: In terdisciplinary Regi onal Studies and Syntheses. New York: John Wi l ey and Son s . Su ll ivan Se a l ey K, Bustamante G. 1999. Set ting Geogra phic Priori ties for Marine Con s erva ti on in Latin Am erica and the Caribbe a n . Arl i n g ton (VA): The Na tu re Con s ervancy. Th ackway R, Cresswell ID. 1998. In terim Ma rine and Coastal Regi on a l i s ati on for Australia: An Ecosystem-based Classification for Marine and Coastal Envi ron m en t s . Vers i on3.3. Ca n berra (Australia): E nvi ron m ent Au s tralia, Com m onwealth Dep a rtment of the Envi ron m en t . Tu rpie JK, Beck l ey LE, Ka tua SM. 2000. Bi ogeogra phy and the sel ecti on of priori ty areas for conservation of South Af rican coastal fishes. Biological Con s erva ti on 92: 5972. Udva rdy MDF. 1975. A C assification of the Bi ogeogra phic Provinces of l the World. Mor ges (Swi t zerland): Intern a tional Union for Con s erva ti on of Na tu re and Na tu ral Resources. [UNEP] United Nations Envi ronment Programme. 2006. Marine and Coastal Ecosys tems and Human Well - bei n g : A Synthesis Report Ba s ed on the F i n d i n gs of the Mi ll en n ium Eco s ys tem Assessment. Nairobi: UNEP. Wa ll ace AR. 1876. The Geogra phical Di s tributi on of Animals: With a Stu dy of the Rel a ti ons of Living and Ex ti n ct Faunas as Elucidating the Past Ch a n ges of the Earths Su rf ace. Lon don : Macmillan. Wa t s onR, Pa u ly D, Ch ri s ten s enV, Froese R, Lon ghu rst A, Platt T, Sa t hyend ranath S, S h erman K, ORei lly J, Cel one P. 2003. Ma pping fisheri e s on to marine ecosystems for regi onal, oceanic and gl obal integra tions. Pa ges 365396 in Hem pel G, Sherman K, eds. L a r ge Marine Ecosystems of the Worl d : Trends in Exploitati on , Pro tecti on , and Research . Am s terdam: E l s evier. [WWF] World Wi de Fund for Na tu re. 2004. The Eastern Af rican Marine E coregi on Vi s i on: A Large Scale Con s ervati on Approach to the Ma n a gem ent of Bi od ivers i ty. Dar es Salaam (Tanzania): WWF. doi:10.1641/B570707 In clu de this information wh en citing this materi a l .

www.biosciencemag.org

July/August 2007 / Vol. 57 No. 7 BioScience 583