Articles

Freshwater Ecoregions of the

World: A New Map of Biogeo-
graphic Units for Freshwater
Biodiversity Conservation

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

ROBIN ABELL, MICHELE L. THIEME, CARMEN REVENGA, MARK BRYER, MAURICE KOTTELAT, NINA BOGUTSKAYA,
BRIAN COAD, NICK MANDRAK, SALVADOR CONTRERAS BALDERAS, WILLIAM BUSSING, MELANIE L. J. STIASSNY,
PAUL SKELTON, GERALD R. ALLEN, PETER UNMACK, ALEXANDER NASEKA, REBECCA NG, NIKOLAI SINDORF,
JAMES ROBERTSON, ERIC ARMIJO, JONATHAN V. HIGGINS, THOMAS J. HEIBEL, ERIC WIKRAMANAYAKE, DAVID
OLSON, HUGO L. LÓPEZ, ROBERTO E. REIS, JOHN G. LUNDBERG, MARK H. SABAJ PÉREZ, AND PAULO PETRY

We present a new map depicting the first global biogeographic regionalization of Earth’s freshwater systems. This map of freshwater ecoregions is
based on the distributions and compositions of freshwater fish species and incorporates major ecological and evolutionary patterns. Covering
virtually all freshwater habitats on Earth, this ecoregion map, together with associated species data, is a useful tool for underpinning global and
regional conservation planning efforts (particularly to identify outstanding and imperiled freshwater systems); for serving as a logical framework
for large-scale conservation strategies; and for providing a global-scale knowledge base for increasing freshwater biogeographic literacy. Preliminary
data for fish species compiled by ecoregion reveal some previously unrecognized areas of high biodiversity, highlighting the benefit of looking at the
world’s freshwaters through a new framework.

Keywords: freshwater, ecoregions, biogeography, fish, mapping

G rowth of the human population, rising consumption,

and rapid globalization have caused widespread de-
gradation and disruption of natural systems, especially in
ecosystems and the diverse communities of species found in
lakes, rivers, and wetlands may be the most endangered of all
(MEA 2005).
the freshwater realm. Freshwater ecosystems have lost a greater These stressed systems support an extraordinarily high
proportion of their species and habitat than ecosystems on proportion of the world’s biodiversity. In terms of area, fresh-
land or in the oceans, and they face increasing threats from water ecosystems occupy only 0.8% of Earth’s surface, but they
dams, water withdrawals, pollution, invasive species, and are estimated to harbor at least 100,000 species, or nearly 6%
overharvesting (MEA 2005, Revenga et al. 2005). Freshwater of all described species (Dudgeon et al. 2006). Each year,

Robin Abell (e-mail: robin.abell@wwfus.org), Michele L. Thieme, Rebecca Ng, Nikolai Sindorf, and Eric Wikramanayake are with WWF in Washington, DC. Carmen
Revenga, Mark Bryer (Bethesda), James Robertson, Eric Armijo (Bolivia), Jonathan V. Higgins (Chicago), Thomas J. Heibel, and Paulo Petry (Boston) are with the
Nature Conservancy, headquartered in Arlington, Virginia. Paulo Petry is also an associate in ichthyology at the Museum of Comparative Zoology at Harvard
University in Massachusetts. Maurice Kottelat is an independent consultant in Switzerland and an honorary research associate at the Raffles Museum of Biodiversity
Research at the National University of Singapore. Nina Bogutskaya and Alexander Naseka are senior researchers at the Zoological Institute of the Russian Academy
of Sciences in St. Petersburg. Brian Coad is a research scientist at the Canadian Museum of Nature in Ottawa. Nick Mandrak is a research scientist at the Great Lakes
Laboratory for Fisheries and Aquatic Sciences, Fisheries and Oceans Canada, Burlington, Canada. Salvador Contreras Balderas is professor emeritus of the
Universidad Autónoma de Nuevo Leon in Monterey, Mexico. William Bussing is professor emeritus at the Universidad de Costa Rica. Melanie L. J. Stiassny is the
Axelrod Research Curator of Ichthyology at the American Museum of Natural History and an adjunct professor at Columbia University in New York City. Paul
Skelton is managing director of the South African Institute for Aquatic Biodiversity and professor at Rhodes University in Grahamstown, South Africa. Gerald R.
Allen is a research associate at Western Australian Museum in Perth. Peter Unmack is a postdoctoral associate in the Department of Integrative Biology at Brigham
Young University in Utah. David Olson is director of science and stewardship at Irvine Ranch Conservancy in California. Hugo L. López is head of the vertebrate
zoology department at the Museo de La Plata, assistant professor in the Facultad de Ciencias Naturales y Museo, and researcher at CIC (Buenos Aires) in Argentina.
Roberto E. Reis is a professor at Católica do Rio Grande do Sul in Porto Alegre, Brazil. John G. Lundberg is chair and curator of ichthyology, and Mark H. Sabaj Pérez
is collection manager, at the Academy of Natural Sciences in Philadelphia. © 2008 American Institute of Biological Sciences.

www.biosciencemag.org May 2008 / Vol. 58 No. 5 • BioScience 403

Articles

new freshwater species are described. For South America nant in shaping the broadscale distributions of freshwater
alone, about 465 new freshwater fish species have been de- species. As Tonn (1990) described, the species occurring in
scribed in the last five years (Eschmeyer 2006), a figure that a given river reach, lake, spring, or wetland will be a function
corresponds to a new species every four days. The presence of a hierarchy of continental-scale filters (including moun-
of species confined to small ranges is also unusually high in tain building, speciation, and glaciation) that have defined large
freshwater ecosystems; for example, 632 animal species have biogeographic patterns; regional-scale filters (such as broad
been recorded as endemic to Lake Tanganyika (Groombridge climatic and physiographic patterns, and dispersal barriers
and Jenkins 1998). such as regional catchments); and subregional and finer-
Despite this combination of extraordinary richness, high scale habitat filters (e.g., distinct physiographic types and
endemism, and exceptional threat, few broadscale conserva- macrohabitats) acting on the regional species pool. Freshwater
tion planning efforts have targeted freshwater systems and their ecoregions capture the patterns generated primarily by con-
dependent species. This relative inattention derives in part tinental- and regional-scale filters.

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

from an acute lack of comprehensive, synthesized data on the Of these filters, dispersal barriers in the form of catch-
distributions of freshwater species (Revenga and Kura 2003). ment divides (also called watersheds) are distinctive to fresh-
The most exhaustive recent global inventory of freshwater taxa waters. Unlike terrestrial species or those with aerial or
acknowledges serious survey gaps and assigns species distri- wind-dispersed life stages, obligate freshwater species—those
butions only to the level of continent (Lévêque et al. 2005). confined to the freshwater environment and unable to move
Such inventories are valuable for highlighting research pri- via land, air, or sea—generally cannot disperse from one un-
orities and providing a global picture of how taxonomic di- connected catchment to another. Furthermore, all species
versity compares across continents, but they have limited dependent on freshwater systems, whether or not they are con-
utility for conservation planning efforts, for which the largest fined to the aquatic environment, are to some extent affected
planning unit is often the river basin or ecoregion. by the hydrological and linked ecological processes of the
catchments where they live. As a result, catchments strongly
A global freshwater regionalization influence broad freshwater biogeographic patterns in most re-
Ecoregions are a widely recognized and applied geospatial unit gions. There are exceptions, however. Tectonic movements
for conservation planning, developed to represent the patterns have in some cases separated once-joined catchments, al-
of environmental and ecological variables known to influence lowing for further speciation. Also, natural drainage evolution
the distribution of biodiversity features at broad scales (Groves over geological time includes river piracy, which severs con-
et al. 2002). Building on the work of Dinerstein and col- nections and provides new interdrainage links that reconform
leagues (1995), we define a freshwater ecoregion as a large area systems. The freshwater ecoregions of the world presented here
encompassing one or more freshwater systems with a distinct reflect both the hydrological underpinning of freshwater fish
assemblage of natural freshwater communities and species. species distributions as well as historical shifts in landmasses
The freshwater species, dynamics, and environmental con- and consequent evolutionary processes.
ditions within a given ecoregion are more similar to each other
than to those of surrounding ecoregions, and together form Ecoregion delineation and species list compilation
a conservation unit. Ecoregion boundaries are not necessar- No global biogeographic framework for freshwater species was
ily determined by the turnover of species ranges (McDonald available as the foundation for our map. The applicability of
et al. 2005) but are intended to describe broad patterns of Wallace’s (1876) and Udvardy’s (1975) zoogeographic realms
species composition and associated ecological and evolu- to most freshwater taxa is unresolved (Berra 2001, Vinson and
tionary processes. Hawkins 2003), and these divisions are too large for conser-
Ecoregion delineation benefits from the best available data vation planning endeavors. Several examinations of global
describing species and systems ecology, but can proceed with freshwater biogeography (e.g., Banarescu 1990) provided in-
imperfect information (Wikramanayake et al. 2002). Global formation at somewhat finer scales but could not be clearly
ecoregion frameworks have already been developed for the translated into seamless ecoregion delineations. Where
terrestrial and, more recently, marine realms, both of which appropriate, we adapted previous continental efforts. For
are characterized by their own data limitations (Olson et al. North America, Africa, and Madagascar, we updated region-
2001, Spalding et al. 2007). In this article we demonstrate how alizations outlined in two previously published volumes
the ecoregion concept has been applied to freshwater systems, (Abell et al. 2000, Thieme et al. 2005), but we excluded a
and present the first global map of freshwater ecoregions— prior delineation for Latin America and the Caribbean (Ol-
a starting point for conservation planning anywhere on Earth. son et al. 1998) because the approach differed markedly from
Ecoregions have typically been delineated to represent our current methodology, and data have improved substan-
patterns of potential vegetation (Olson et al. 2001) and have tially since its development (e.g., Reis et al. 2003). We exam-
at times been used to characterize regional differences in ined but chose to exclude the 25 European regions of Illies’s
water quality as well (Omernik 1987). Terrestrial ecoregions impressive Limnofauna Europaea (1978) because the ap-
are delineated largely on the basis of climate, physiography, proach for delineating those regions differed considerably from
and vegetation types, but different features are often domi- ours: those regions were based on the distributions of 75

404 BioScience • May 2008 / Vol. 58 No. 5 www.biosciencemag.org

 Articles

different taxonomic groups and were drawn without reference The available data for describing fish biogeography vary
to catchments. Moreover, neither ecological nor evolutionary widely. In the United States, it is possible to map presence/
processes figured in those delineations. A complete list of all absence data for all freshwater fish species to subbasins aver-
references and experts consulted in the process of delineat- aging about 2025 square kilometers (km2) in size (NatureServe
ing ecoregions is available online (www.feow.org). 2006). But for many of the world’s species, occurrence data
We assembled our global map of freshwater ecoregions are limited to a small number of irregularly surveyed systems.
using the best available regional information describing fresh- Large parts of the massive Congo basin remain unsampled,
water biogeography, defined broadly to include the influ- for instance, with most sampling occurring near major towns
ences of phylogenetic history, palaeogeography, and ecology and most taxonomic studies of the region dating from the
(Banarescu 1990). We restricted our analyses to information 1960s. Problems with taxonomy and species concepts ham-
describing freshwater fish species distributions, with a few per broadscale analyses even where systems have been rea-
exceptions for extremely data-poor regions and inland seas, sonably well sampled (Lundberg et al. 2000). Although

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

where some invertebrates and brackish-water fish were addressing many of these problems is beyond the scope of this
considered, respectively. We focused on freshwater fish for project, in our analyses we have attempted to minimize
several reasons. On a global scale, fish are the best-studied nomenclatural errors by normalizing species names with
Eschmeyer’s Catalog of Fishes (2006; www.calacademy.org/
obligate aquatic taxa. Detailed information exists for other
research/ichthyology/catalog/).
freshwater taxa in regions like North America and Europe, but
Freshwater fish patterns were analyzed separately for
the consideration of such groups in a global analysis would
different regions of the world to account for data variability.
be difficult, given the wide variation in available data (Balian
The geographic scope of major information sources largely
et al. 2008). Freshwater dispersant fish species—those unable
defined those regions (table 1). Information sources were
to cross saltwater barriers—are better zoogeographic indicators
typically taxonomic works, some of which included bio-
than freshwater invertebrates, which can often disperse over geographical analyses. Leading ichthyologists delineated
land, survive in humid atmospheres outside water, or be ecoregions primarily by examining the distributions of en-
transported between freshwaters (Banarescu 1990). Finally, demic species, genera, and families against the backdrop of
the distributions of obligate aquatic invertebrate groups in an area’s dominant habitat features and the presence of eco-
general respond to ecological processes at localized scales logical (e.g., large concentrations of long-distance migratory
that are too small to be meaningful for ecoregion delineation species) and evolutionary (e.g., species flocks) phenomena.
(Wasson et al. 2002). Therefore, fish serve as proxies for the More than 130 ichthyologists and freshwater biogeographers
distinctiveness of biotic assemblages. We recognize that analy- contributed to the global map by either delineating or re-
ses of other taxonomic groups would almost certainly reveal viewing ecoregions.
different patterns for some regions, and that our results are Data gaps and biogeographic drivers resulted in the use of
scale dependent (Paavola et al. 2006). Our near-exclusive slightly different criteria among and even within some regions
focus on fish is a departure from earlier continental eco- (table 2, box 1). Where fish species data were reasonably
regionalization exercises (Abell et al. 2000, Thieme et al. comprehensive and available at subbasin or finer scales, we
2005), and we have updated the ecoregion delineations attributed species distributions to catchments to facilitate
accordingly. evaluation of biogeographic patterns in a bottom-up

Table 1. Regional information sources used for ecoregion delineations.

Region Primary information source

Africa Roberts 1975, Skelton 1994, Lévêque 1997, Thieme et al. 2005
Middle East No regional information sources available.
Former USSR No regional information sources available.
Remainder of Eurasia For Europe: Kottelat and Freyhof 2007; no regionwide information sources for Asia.
Australasia McDowall 1990, Allen 1991, Unmack 2001, Allen et al. 2002
Oceania Keith et al. 2002
Canada Scott and Crossman 1998
United States Maxwell et al. 1995, Abell et al. 2000
Mexico Contreras-Balderas 2000, Miller et al. 2005
Central America Bussing 1976, CLOFFSCA (Reis et al. 2003)
Caribbean Rauchenberger 1988, Burgess and Franz 1989
South America CLOFFSCA (Reis et al. 2003), Menni 2003

Note: In many cases, these same sources were used to compile species lists. A full bibliography with additional publications,
which along with unpublished data often constituted the greater part of inputs to ecoregion delineations and species lists, is
available at the Web site www.feow.org. Every region also benefited from expert input; individual contributors are listed in the
acknowledgments section and at the Web site. Regions in some cases correspond to politically rather than biophysically
defined units to take advantage of existing information sources and expertise.

www.biosciencemag.org May 2008 / Vol. 58 No. 5 • BioScience 405

Articles

approach. For example, a new high-resolution hydrographic These and other examples demonstrate that historical geo-
dataset (HydroSHEDS; www.wwfus.org/freshwater/hydrosheds. graphic events and current hydrology may have conflicting
cfm) for South America provided fine-scale catchment maps effects on the fish fauna of a particular region and thereby
that, in conjunction with newly synthesized species data (Reis argue for different boundaries. The decision to weigh some
et al. 2003), aided in the assessment of biogeography. In effects more strongly than others was made on a case-by-case
regions without extensive species data, or where major basins basis, and it is acknowledged that additional data may favor
support highly similar faunas as a result of recent glaciation, alternative delineations.
a top-down analysis used qualitative expert knowledge of With the exception of islands, individual freshwater eco-
distinctive species and assemblages to map major bio- regions typically cover tens of thousands to hundreds of
geographic patterns (table 2). Ecoregional boundaries result- thousands of square kilometers (Maxwell et al. 1995). Eco-
ing from either approach, therefore, largely coincide with
region size varies in large part because of landscape history.
catchment boundaries.
Regions with depauperate faunas resulting from recent glacia-

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

Whereas overall there is correspondence between catch-
tion events tend to have large ecoregion sizes, as do those dom-
ments and ecoregion boundaries, unconnected neighboring
inated by very large river systems (e.g., much of South
catchments were in some cases grouped together, where
strong biogeographic evidence indicates that landscape or America). Regions with recent tectonic activity or smaller,
other features overrode contemporary hydrographic integrity. more isolated freshwater systems often are divided into smaller
For example, owing to historic drainage evolution and ecoregions. For example, central Mexico has experienced
similarities in fauna, Africa’s southern temperate highveld intermittent isolation and exchange between basins owing to
combines headwaters of coastal basins that drain to the active mountain-building processes leading to small, frag-
Indian Ocean with those of the Atlantic-draining Orange mented systems with distinct faunas. We acknowledge that
basin. Considerable faunal exchange of the headwaters of data quality may also influence the size of ecoregions; for in-
the Orange River system with that of the coastal systems may stance, the entire Amazon is currently divided into only 13
have occurred as the coastal rivers eroded their basins at a faster ecoregions, but better data on species occurrences within
rate than the adjacent Orange tributaries (Skelton et al. 1995). major subbasins would most likely support finer delineations.

Table 2. Basic ecoregion delineation approaches for individual regions.

Region Delineation approach

Africa Using Roberts (1975) as a starting point, ecoregions were delineated using a top-down qualitative assessment that
incorporated expert knowledge and divisions of major river basins. In a few cases where basin divides do not circumscribe
species distributions or where basins contain internal barriers to dispersal, ecoregions straddle or divide basins.
Middle East Species lists were generated for whole drainage basins, which were then either combined with smaller catchments that
were very similar faunistically (minor desert basins, for example) or subdivided on the basis of different ecologies (e.g., the
Tigris-Euphrates with lowland marshes and upland streams).
Former USSR A species/genera/family presence/absence matrix was compiled for a hierarchy of hydrographic units, and cluster analysis
and ordination techniques (Primer v.6 statistics software) were employed to assess biotic similarities among hydrographic
units and to identify major faunal breaks.
Remainder of Eurasia For Southeast Asia and southern Europe, a bottom-up approach employing both published and unpublished field data and
expert assessment was used. East Asian, northern European, and eastern European ecoregions were delineated through a
top-down process using major basins as a starting point and incorporating traditionally recognized zoogeographic patterns
where appropriate.
Australasia For Australia, ecoregions were adapted from Allen and colleagues’ (2002) and Unmack’s (1991) “freshwater fish biogeo-
graphic provinces”; provinces were derived through similarity analyses, parsimony analysis, and drainage-based plots of
species ranges. For New Guinea, “subprovinces” of Allen (1991) were modified (primarily combined) on the basis of expert
input. For New Zealand and other islands and island groups, islands were placed in ecoregions on the basis of expert
input.
Oceania Islands and island groups were placed in ecoregions on the basis of distinctive (endemic or near-endemic) fish faunas.
Canada Separate cluster analyses were conducted on fish occurrence in the secondary watersheds in each of the nine primary
watersheds in Canada.
United States The “subregions” of Maxwell and colleagues (1995) were adopted, with relatively small modifications made following input
by regional specialists, especially the Endangered Species Committee of the American Fisheries Society.
Mexico Ecoregion delineations were based on qualitative similarity/dissimilarity assessments of major basins, using the standard
administrative hydrographical regions of the Mexican federal government. Subregions within major basins were recognized
as separate ecoregions when the fish fauna was sufficiently distinctive.
Central America Fish provinces from Bussing (1976) were revised and subdivided on the basis of the application of the similarity index to
subbasin fish presence/absence data.
Caribbean Ecoregions from Olson and colleagues (1998) were modified on the basis of similarity analyses of island-by-island species
lists and expert input.
South America Ecoregion delineations were based on qualitative similarity/dissimilarity assessments of catchments, resulting in aggrega-
tion/disaggregation. See box 1 for additional information.

Note: Some of the variations resulting from differences in data quality and biogeographic drivers across and within regions are noted. For some regions,
subecoregions (described at www.feow.org) were delineated to capture finer-scale patterns than could be represented by ecoregions.

406 BioScience • May 2008 / Vol. 58 No. 5 www.biosciencemag.org

 Articles

undescribed species. Confirmed extinct species (Ian J. Har-

Box 1. Example of criteria applied to ecoregion rison, American Museum of Natural History, New York, per-
delineation: South America.
sonal communication, 29 March 2007) were excluded, but
extirpated species were included to acknowledge restoration
The delineation process for South America followed a step- opportunities. Endemic species, defined as those occurring
wise process of subdivision of the continent’s major drainage
only in a single ecoregion, were identified first by experts and
systems. Delineation started with the historically recognized
major ichthyographic provinces exemplified in Gery (1969)
cross-checked using a species database constructed for this
and Ringuelet (1975) and proceeded with subdivision at finer project, which includes more than 14,500 described fish
scales using regionalized data on fish distributions. species. Species were coded as freshwater, brackish, or marine
using data from FishBase (www.fishbase.org), and species
The criteria for determining the merit of delineating an eco- with only brackish or marine designations were omitted
region were not uniform across the continent as a result of
from the richness and endemism totals reported here.
localized faunistic differences. In some areas, delineations

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

were based on family-level data, whereas in others, faunistic
turnover at lower taxonomic levels was the criterion. For Freshwater ecoregional map and species results
instance, astroblepid catfishes are distinct components of high- Our map of freshwater ecoregions contains 426 units, covering
elevation freshwaters along the Andes forefront, and that fami- nearly all nonmarine parts of the globe, exclusive of Antarc-
ly’s distribution was critical to informing the delineation of the tica, Greenland, and some small islands (figure 1; a full leg-
high Andean ecoregions. On the other side of the continent end is available at www.feow.org). There is large variation in
along the Atlantic coast, we used the presence or absence of the area of individual ecoregions. Large ecoregions, such as
endemic assemblages of the genus Trichomycterus, several gen- the dry Sahel (4,539,429 km2), tend to be found in more de-
era of the subfamily Neoplecostomatinae, and the presence or pauperate desert and polar regions exhibiting low species
absence of annual killifish genera and species to distinguish turnover. Smaller ecoregions are typically found in noncon-
distinct drainage complexes from one another.
tinental settings where systems are by nature smaller and
In the piedmont zones and in contact areas between lowlands species turnover is higher, as in the Indo-Malay region. The
and geologic shield areas, we used indicator groups to deter- smallest ecoregion, at 23 km2, is Cocos Island (Costa Rica);
mine where along the elevation/slope gradient the fauna was the average ecoregion size is 311,605 km2. Ecoregions ranged
changing. The distribution of lowland forms was matched from those encompassing only 1 country to those straddling
with forms found in higher-gradient systems to establish 16 countries (central and western Europe ecoregion).
where one group was dropping out and the other started
In total, we assigned more than 13,400 described freshwater
occurring. This transition zone was then established as the
operational boundary between connecting ecoregions.
fish species to ecoregions, of which more than 6900 were
assigned to single ecoregions (i.e., endemic). Examination of
For areas like Patagonia, the Titicaca altiplano, and the Mara- the fish species data synthesized by ecoregion confirms some
caibo basin, the uniqueness of the fauna, often occurring with- well-known patterns and highlights others unknown to many
in clearly defined geographic areas, permitted reasonably conservationists, managers, and policymakers working at
straightforward delineations. In the larger river basin systems
regional or global scales (figures 2a–2d). In agreement with
where there are no clear boundaries, the ecoregional limits are
the best approximation, given the current data.
previous global assessments (Groombridge and Jenkins 1998,
Revenga et al. 1998), our analysis identifies as outstanding for
both fish richness and endemism systems that include large
The process of delineating ecoregions required compiling portions of Africa’s Congo basin, the southern Gulf of Guinea
and synthesizing information on the distributions of fish drainages, and Lakes Malawi, Tanganyika, and Victoria; Asia’s
species. A logical and practical extension of the delineations Zhu Jiang (Pearl River) basin and neighboring systems; and
was the compilation of fish species lists for each ecoregion. large portions of South America’s Amazon and Orinoco
For the United States, NatureServe provided presence/absence basins. Areas confirmed for globally high richness include
data for individual species, coded to eight-digit hydrologic unit Asia’s Brahmaputra, Ganges, and Yangtze basins, as well as large
codes (HUCs); these HUC occurrences were then translated portions of the Mekong, Chao Phraya, and Sitang and Ir-
to ecoregions, and the data were manually cleaned of erro- rawaddy; Africa’s lower Guinea; and South America’s Paraná
neous occurrences derived from introductions and prob- and Orinoco. When richness is adjusted for ecoregion area,
lematic records. These species lists were then merged with additional systems such as the Tennessee, Cumberland, Mo-
those from Canada and Mexico for transnational ecoregions. bile Bay, Apalachicola, and Ozark highlands in the southeastern
For all other ecoregions, data came from the published liter- United States; portions of Africa’s Niger River Basin; the
ature, as well as from gray literature and unpublished sources islands of New Caledonia, Vanuatu, and Fiji; China’s Hainan
(see table 1; a full bibliography is available at www.feow.org). Island; and large parts of Sumatra and Borneo, among many
In all cases, experts served as gatekeepers of these data to en- other areas, are also especially noteworthy.
sure that lists were based on the best available information, Numerous systems previously identified as highly endemic
both in terms of distributions and nomenclature. Introduced for fish were confirmed, as measured by either numbers of en-
species were removed from the tallies presented here, as were demic species or percentage endemism. A subset includes

www.biosciencemag.org May 2008 / Vol. 58 No. 5 • BioScience 407

 Articles

408 BioScience • May 2008 / Vol. 58 No. 5

Figure 1. Map of freshwater ecoregions of the world, in which 426 ecoregions are delineated. An interactive version of this map that includes additional information is
available at www.feow.org.

North America 123 Oregon and Northern California Coastal 147 Ozark Highlands 170 Sierra Madre del Sur 216 Windward and Leeward Islands
101 Alaskan Coastal 124 Oregon Lakes 148 Upper Mississippi 171 Papaloapan 217 Cocos Island (Costa Rica)
102 Upper Yukon 125 Sacramento–San Joaquin 149 Lower Mississippi 172 Coatzacoalcos
103 Alaska and Canada Pacific Coastal 126 Lahontan 150 Teays–Old Ohio 173 Grijalva–Usumacinta South America
104 Upper Mackenzie 127 Bonneville 151 Cumberland 174 Upper Usumacinta 301 North Andean Pacific Slopes–Rio Atrato
105 Lower Mackenzie 128 Death Valley 152 Tennessee 175 Yucatan 302 Magdalena–Sinu
106 Central Arctic Coastal 129 Vegas–Virgin 153 Mobile Bay 176 Bermuda 303 Maracaibo
107 Upper Saskatchewan 130 Colorado 154 West Florida Gulf 304 South America Caribbean
108 Middle Saskatchewan 131 Gila 155 Apalachicola Central America Drainages–Trinidad
109 English–Winnipeg Lakes 132 Upper Rio Grande–Bravo 156 Florida Peninsula 201 Chiapas–Fonseca 305 Orinoco High Andes
110 Southern Hudson Bay 133 Pecos 157 Appalachian Piedmont 202 Quintana Roo–Motagua 306 Orinoco Piedmont
111 Western Hudson Bay 134 Rio Conchos 158 Chesapeake Bay 203 Mosquitia 307 Orinoco Llanos
112 Canadian Arctic Archipelago 135 Lower Rio Grande–Bravo 159 Southern California Coastal–Baja 204 Estero Real–Tempisque 308 Orinoco Guiana Shield
113 Eastern Hudson Bay–Ungava 136 Cuatro Cienegas California 205 San Juan (Nicaragua and Costa Rica) 309 Orinoco Delta and Coastal Drainages
114 Gulf of St. Lawrence Coastal Drainages 137 Rio Salado 160 Sonora 206 Chiriqui 310 Essequibo
115 Canadian Atlantic Islands 138 Rio San Juan (Mexico) 161 Guzman–Samalayuca 207 Isthmus Caribbean 311 Guianas
116 Laurentian Great Lakes 139 West Texas Gulf 162 Sinaloa 208 Santa Maria 312 Amazonas High Andes
117 St. Lawrence 140 East Texas Gulf 163 Mayran–Viesca 209 Chagres 313 Western Amazon Piedmont
118 Northeast US and Southeast Canada 141 Sabine–Galveston 164 Rio Santiago 210 Rio Tuira 314 Rio Negro
Atlantic Drainages 142 Upper Missouri 165 Lerma–Chapala 211 Cuba–Cayman Islands 315 Amazonas Guiana Shield
119 Scotia–Fundy 143 Middle Missouri 166 Llanos El Salado 212 Bahama Archipelago 316 Amazonas Lowlands
120 Columbia Glaciated 144 US Southern Plains 167 Panuco 213 Jamaica 317 Ucayali–Urubamba Piedmont
121 Columbia Unglaciated 145 Ouachita Highlands 168 Ameca–Manantlan 214 Hispaniola 318 Mamore–Madre de Dios Piedmont

www.biosciencemag.org
122 Upper Snake 146 Central Prairie 169 Rio Balsas 215 Puerto Rico–Virgin Islands 319 Guapore–Itenez

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

 320 Tapajos–Juruena 447 Namak 575 Southern Temperate Highveld 723 Inle Lake
321 Madeira Brazilian Shield 448 Kavir and Lut Deserts 576 Zambezian Lowveld 724 Upper Lancang (Mekong)
322 Xingu 449 Esfahan 577 Amatolo–Winterberg Highlands 725 Er Hai
323 Amazonas Estuary and Coastal 450 Turan Plain 578 Cape Fold 726 Lower Lancang (Mekong)
Drainages 451 Northern Hormuz Drainages 579 Western Madagascar 727 Khorat Plateau (Mekong)
324 Tocantins–Araguaia 452 Caspian Marine 580 Northwestern Madagascar 728 Kratie–Stung Treng (Mekong)
325 Parnaiba 453 Volga Delta–Northern Caspian Drainages 581 Madagascar Eastern Highlands 729 Mekong Delta
326 Northeastern Caatinga and Coastal 582 Southern Madagascar 730 Southern Annam
Drainages Africa and Madagascar 583 Madagascar Eastern Lowlands 731 Eastern Gulf of Thailand Drainages
327 S. Francisco 501 Atlantic Northwest Africa 584 Comoros–Mayotte 732 Chao Phraya
328 Northeastern Mata Atlantica 502 Mediterranean Northwest Africa 585 Seychelles 733 Mae Khlong
329 Paraiba do Sul 503 Sahara 586 Mascarenes 734 Malay Peninsula Eastern Slope
330 Ribeira de Iguape 504 Dry Sahel 587 S. Tome and Principe–Annobon 735 Northern Central Sumatra–Western

www.biosciencemag.org
331 Southeastern Mata Atlantica 505 Lower Niger–Benue Malaysia
332 Lower Uruguay 506 Niger Delta Northern Asia 736 Aceh
333 Upper Uruguay 507 Upper Niger 601 Irgyz–Turgai 737 Indian Ocean Slope of Sumatra and Java
334 Laguna dos Patos 508 Inner Niger Delta 602 Ob 738 Southern Central Sumatra
335 Tramandai–Mampituba 509 Senegal–Gambia 603 Upper Irtysh 739 Southern Sumatra–Western Java
336 Central Andean Pacific Slopes 510 Fouta–Djalon 604 Chuya 740 Central and Eastern Java
337 Titicaca 511 Northern Upper Guinea 605 Yenisei 741 Kapuas
338 Atacama 512 Southern Upper Guinea 606 Lake Baikal 742 Northwestern Borneo
339 Mar Chiquita–Salinas Grandes 513 Mount Nimba 607 Taimyr 743 Borneo Highlands
340 Cuyan–Desaguadero 514 Eburneo 608 Lena 744 Northeastern Borneo
341 South Andean Pacific Slopes 515 Ashanti 609 Kolyma 745 Eastern Borneo
342 Chaco 516 Volta 610 Anadyr 746 Southeastern Borneo
343 Paraguay 517 Bight Drainages 611 East Chukotka 747 Malukku
344 Upper Parana 518 Northern Gulf of Guinea Drainages–Bioko 612 Koryakia 748 Lesser Sunda Islands
345 Lower Parana 519 Western Equatorial Crater Lakes 613 Kamchatka and Northern Kurils 749 Sulawesi
346 Iguassu 520 Lake Chad 614 Okhotsk Coast 750 Malili Lakes
347 Bonaerensean Drainages 521 Lake Victoria Basin 615 Coastal Amur 751 Lake Poso
348 Patagonia 522 Upper Nile 616 Lower Amur 752 Mindanao
349 Valdivian Lakes 523 Lower Nile 617 Middle Amur 753 Lake Lanao
350 Galapagos Islands 524 Nile Delta 618 Argun 755 Northern Philippine Islands
351 Juan Fernandez Island 525 Ethiopian Highlands 619 Shilka (Amur) 756 Palawan–Busuanga–Mindoro
352 Fluminense 526 Lake Tana 620 Songhua Jiang 757 Western Taiwan
527 Western Red Sea Drainages 621 Inner Mongolia Endorheic Basins 758 Eastern Taiwan
Europe and Middle East 528 Northern Eastern Rift 622 Western Mongolia 759 Hainan
401 Iceland–Jan Mayen 529 Horn of Africa 623 Dzungaria 760 Northern Annam
402 Northern British Isles 530 Lake Turkana 624 Balkash–Alakul 761 Song Hong
403 Cantabric Coast–Languedoc 531 Shebelle–Juba 625 Tarim 762 Yunnan Lakes
404 Central and Western Europe 532 Ogooue–Nyanga–Kouilou–Niari 626 Lower and Middle Syr Darya 763 Xi Yiang
405 Norwegian Sea Drainages 533 Southern Gulf of Guinea Drainages 627 Lake Issyk Kul–Upper Chu 764 Upper Yangtze
406 Northern Baltic Drainages 534 Sangha 628 Northern Central Asian Highlands 765 Middle Yangtze
407 Barents Sea Drainages 535 Sudanic Congo–Oubangi 629 Aral Sea Drainages 766 Lower Yangtze
408 Southern Baltic Lowlands 536 Uele 630 Middle Amu Darya 767 Coastal Fujian–Zeijang
409 Lake Onega–Lake Ladoga 537 Cuvette Centrale 631 Upper Amu Darya 768 Andaman Islands
410 Volga–Ural 538 Tumba 632 Qaidan 769 Nicobar Islands
411 Western Caspian Drainages 539 Upper Congo Rapids 633 Upper Huang He
412 Western Iberia 540 Upper Congo 634 Upper Huang He Corridor Australia and Pacific
413 Southern Iberia 541 Albertine Highlands 635 Huang He Great Bend 801 Southwestern Australia
414 Eastern Iberia 542 Lake Tanganyika 636 Lower Huang He 802 Pilbara
415 Gulf of Venice Drainages 543 Malagarasi–Moyowosi 637 Liao He 803 Kimberley
416 Italian Peninsula and Islands 544 Bangweulu–Mweru 638 Eastern Yellow Sea Drainages 804 Paleo
417 Upper Danube 545 Upper Lualaba 639 Southeastern Korean Peninsula 805 Arafura–Carpentaria
418 Dniester–Lower Danube 546 Kasai 640 Hamgyong–Sanmaek 806 Lake Eyre Basin
419 Dalmatia 547 Mai Ndombe 641 Sakhalin, Hokkaido, and Sikhote– 807 Eastern Coastal Australia
420 Southeast Adriatic Drainages 548 Malebo Pool Alin Coast 808 Murray–Darling
421 Ionian Drainages 549 Lower Congo Rapids 642 Honshu–Shikoku–Kyushu 809 Bass Strait Drainages
422 Vardar 550 Lower Congo 643 Biwa Ko 810 Southern Tasmania
423 Thrace 551 Cuanza 811 New Zealand
424 Aegean Drainages 552 Namib Southern Asia 812 Vogelkop–Bomberai
425 Dnieper–South Bug 553 Etosha 701 Baluchistan 813 New Guinea North Coast
426 Crimea Peninsula 554 Karstveld Sink Holes 702 Helmand–Sistan 814 New Guinea Central Mountains
427 Don 555 Zambezian Headwaters 703 Lower and Middle Indus 815 Southwest New Guinea–Trans-Fly Lowland
428 Kuban 556 Upper Zambezi Floodplains 704 Yaghistan 816 Papuan Peninsula
429 Western Anatolia 557 Kafue 705 Indus Himalayan Foothills 817 Bismarck Archipelago
430 Northern Anatolia 558 Middle Zambezi–Luangwa 706 Upper Indus 818 Solomon Islands
431 Central Anatolia 559 Lake Malawi 707 Tibetan Plateau Endorheic Drainages 819 Vanuatu
432 Southern Anatolia 560 Zambezian Highveld 708 Namuda–Tapi 820 New Caledonia
433 Western Transcaucasia 561 Lower Zambezi 709 Ganges Delta and Plain 821 Fiji
434 Kura–South Caspian Drainages 562 Mulanje 710 Ganges Himalayan Foothills 822 Wallis–Futuna
435 Sinai 563 Eastern Zimbabwe Highlands 711 Upper Brahmaputra 823 Samoas
436 Coastal Levant 564 Coastal East Africa 712 Middle Brahmaputra 824 Society Islands
437 Orontes 565 Lake Rukwa 713 Northern Deccan Plateau 825 Tubuai Islands
438 Jordan River 566 Southern Eastern Rift 714 Southern Deccan Plateau 826 Marquesas Islands
439 Southwestern Arabian Coast 567 Tana, Athi, and Coastal Drainages 715 Western Ghats 827 Rapa
440 Arabian Interior 568 Pangani 716 Southeastern Ghats 828 Hawaiian Islands
441 Lower Tigris and Euphrates 569 Okavango 717 Sri Lanka Dry Zone 829 East Caroline Islands
442 Upper Tigris and Euphrates 570 Kalahari 718 Sri Lanka Wet Zone 830 West Caroline Islands
443 Oman Mountains 571 Southern Kalahari 719 Chin Hills–Arakan Coast
444 Lake Van 572 Western Orange 720 Sitang–Irawaddy
445 Orumiyeh 573 Karoo 721 Upper Salween

May 2008 / Vol. 58 No. 5 • BioScience 409

446 Caspian Highlands 574 Drakensberg–Maloti Highlands 722 Lower and Middle Salween
Articles

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

Articles

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

Figure 2. Preliminary freshwater fish species data for ecoregions: (a) species richness, (b) number of endemic species, (c)
percentage endemism, and (d) species per ecoregion area. Numbers may be adjusted on the basis of an ongoing process to
correct nomenclatural errors. Natural breaks (Jenk’s optimization) was the classification method used for panels (a)–(c).
This method identifies breakpoints between classes using a statistical formula that identifies groupings and patterns
inherent in the data.

highland lakes in Cameroon along with Africa’s Lake Tana; the Tocantins-Araguaia systems. The Tocantins-Araguaia, as
northwestern and eastern Madagascar; freshwaters from well as the highly endemic São Francisco, were defined as units
Turkey’s central Anatolia region, the northern British Isles, the of analysis in Revenga and colleagues (1998), but fish data were
Philippines, Sri Lanka, India’s western Ghats, the southwest- unavailable for those basins when that study was done.
ern Balkans, and northwest Mediterranean; southwestern Systems never before analyzed globally but recognized in
Australia and nearly the entire island of New Guinea; Eurasian our results as exceptionally rich for fish include those of the
lakes, including Baikal, Inle, and Sulawesi’s Lake Poso and Malay Peninsula’s eastern slope and Japan. A large number of
Malili system; Death Valley in the United States and Mexico’s ecoregions are identified for the first time for highly endemic
Pánuco system; and South America’s Iguaçu River, Lake Tit- faunas, measured as percentage endemism. Newly identified
icaca, and the freshwaters of both the Mata Atlántica and the ecoregions with at least 50% endemism include Africa’s
continent’s northwestern Pacific coast. Additionally, newly Cuanza, Australia’s Lake Eyre Basin, Mexico’s Mayrán-Viesca,
available data show that some systems previously recognized and New Zealand, as well as a large number of highly depau-
for high endemism, such as those of South America’s Guianas, perate ecoregions such as Africa’s karstveld sink holes, Turkey’s
also exhibit exceptional richness. Lake Van, the Oman Mountains, western Mongolia, and
Because our ecoregions cover all nonmarine waters, and Hawaii.
because they often exist as subdivisions of major river basins, Each of the biodiversity analyses that we offer here em-
our results also highlight a number of smaller systems for the phasizes different sets of ecoregions, suggesting that a single
first time in global analyses. Using finer-resolution data allowed measure of species diversity might overlook ecoregions of
us to identify the high richness of the Congo’s Malebo Pool important biodiversity value. In a comparative analysis of
and Kasai basin. Cuba and Hispaniola stand out for biodiversity value, ecoregions are probably best evaluated
endemism, along with the Amazon’s western piedmont and against others within the same region, with similar historical

410 BioScience • May 2008 / Vol. 58 No. 5 www.biosciencemag.org

 Articles

and environmental characteristics, and of similar size to ac- Caveats and limitations
count for the typically positive relationship between river Ecoregions are delineated based on the best available infor-
discharge and fish species richness (Oberdorff et al. 1995). mation, but data describing freshwater species and ecologi-
Nonetheless, some systems, such as the Amazon and many of cal processes are characterized by marked gaps and variation
Africa’s Rift Valley lakes, stand out by nearly any measure of in quality and consistency. Data quality is generally consid-
fish biodiversity and are indisputable global conservation ered high for North America, Australia, New Zealand, Japan,
priorities. western Europe, and Russia; moderate for Central America,
the southern cone of South America, southern and western
Conservation applications Africa, Oceania, and the Middle East; and poor for much of
The ecoregion map and associated species data summarized southeastern Asia, central and eastern Africa, and South
here have a number of conservation applications. At global America north of the Paraná River basin.
and regional scales the ecoregion map can be used to distin- Freshwater ecoregions are not homogeneous units. Within

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

guish distinct units of freshwater biodiversity to be represented individual ecoregions there will be turnover of species along
in conservation efforts. The Convention on Wetlands, for longitudinal gradients of river systems and across different
instance, requires that sites nominated as wetlands of inter- habitats such as flowing and standing-water systems. The
national importance—with wetlands defined to include all inclusion of multiple macrohabitat types within a given fresh-
freshwaters—be evaluated against a “biogeographic region- water ecoregion is a marked departure from terrestrial ecore-
alization” criterion (Ramsar Bureau 2006). Lack of a global gions, which typically encompass a single vegetation-defined
biogeographic scheme has stalled the application of this cri- biome (e.g., deciduous forests, evergreen forests, or scrub;
terion, but our global map and database may provide a nec- Wikramanayake et al. 2002).
essary framework for identifying broadscale gaps in protection. Ecoregions are imperfect units for highlighting certain
Similarly, progress toward the establishment of representative highly distinct and highly localized assemblages occurring at
networks of freshwater protected areas, as called for by the third subecoregion scales. Examples include many peat swamps or
IUCN World Conservation Congress, the fifth World Parks subterranean systems. Underground systems such as caves and
Congress, and the seventh Meeting of the Conference of the karsts may require their own planning framework, as ground-
Parties to the Convention on Biological Diversity, can now be water catchments may not correspond with the surface-water
measured using ecoregions as a proxy for finer-scale global catchments upon which our ecoregions are built.
species or habitat distribution data. At a regional level, the For reasons of practicality and scale, our ecoregion frame-
freshwater ecoregion map may be used as supplementary in- work does not take into account the distributions of freshwater
formation for implementation of the European Union’s Wa- species such as invertebrates, reptiles, and amphibians. This
ter Framework Directive (2000/60/EC), which requires a is a limitation of the ecoregional approach presented here,
characterization of surface water bodies and currently uses which is especially problematic for places such as isolated
regions defined by Illies (1978). islands where freshwater fish provide little information to
A primary use of ecoregions is as conservation planning inform biogeographic delineation. We hope this taxonomic
units (Higgins 2003). Our attribution of freshwater fish omission will serve as motivation for generating and syn-
species data to ecoregions is an important first step for data- thesizing global data for other taxonomic groups to provide
poor regions. Organizations or agencies with regional man- complementary information for conservation planners, par-
dates may choose to compare biodiversity values across ticularly when working at subecoregional scales. We recognize
ecoregions in the process of setting continental priorities that improved information in the future may warrant map re-
(Abell et al. 2000, Thieme et al. 2005). At the basin scale, visions, and we highlight areas of greatest data uncertainty in
ecoregions can help to introduce biodiversity information into part to encourage enhanced research investment in those
water-resource or integrated-basin management activities places. We believe that the critical state of freshwater systems
(Gilman et al. 2004). Where major basins are divided among and species argues against waiting for ideal biodiversity data
multiple freshwater ecoregions, whole-basin exercises can to be developed before generating urgently needed conser-
use ecoregions as stratification units to ensure adequate rep- vation tools like the ecoregion map.
resentation of distinct biotas. Where unconnected drainages Shifting transition zones for species are common, and we
are combined into a single freshwater ecoregion, planners may recommend that ecoregions be viewed as logical units for more
choose to consider a counterintuitive planning unit to in- detailed analyses and strategies. Ecoregions are intended to de-
corporate biogeographic patterns. Freshwater ecoregions de- pict the estimated original extent of natural communities
fined in previous exercises have already been put to use by the before major alterations caused by recent human activities,
Nature Conservancy and WWF in numerous conservation but original distributions can be difficult to reconstruct. As
planning efforts across North America (e.g., Upper Mississippi; new species are described, our understanding of distribution
Weitzell et al. 2003), South America (e.g., the Pantanal; de patterns may also change. Ecoregional delineation is an iter-
Jesus 2003), and Africa (e.g., the Congo basin; Kamdem- ative process, and changes to ecoregion boundaries should be
Toham et al. 2003). incorporated as new information becomes available.

www.biosciencemag.org May 2008 / Vol. 58 No. 5 • BioScience 411

Articles

There is no definitive, error-free data source for classifying and associated species data begin to improve access to pre-
fish species as freshwater, brackish, or marine. We chose to use viously dispersed and difficult to access freshwater biodiver-
the global FishBase habitat assignments, which are derived sity information. We hope that this set of products catalyzes
from the literature, to ensure that any given species in our data- additional work toward a better understanding of freshwater
base would be classified consistently wherever it occurred. We species distributions and—of equal if not more importance—
recognize that errors of omission or commission may derive leads to a ramping up of freshwater conservation activity
from inaccuracies in the FishBase assignments as well as and success.
from the habitat plasticity of some species. All species in-
formation provided to us by experts, regardless of habitat Acknowledgments
assignment, is retained in our database for future analyses. The authors would like to thank the dozens of scientists who
The preliminary richness and endemism numbers pre- contributed to development of the ecoregion map and syn-
sented here are in some cases markedly different from exist- thesis of fish species data: E. K. Abbam, Vinicius Abilhoa,

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

ing estimates in the literature. For example, our tally for Lake Angelo Agostinho, James Albert, Hector Samuel Vera
Malawi contains 431 described fish species, but other estimates Alcazar, Claudio Baigun, Eldredge Bermingham, Tim Berra,
run as high as 800 or more (Thieme et al. 2005). Our omis- Vinicius Bertaco, Richard Biggins, Flavio Bockmann, Paulo
sion of undescribed species, as well as the conservative Buckup, Noel Burkhead, Brooks Burr, Mary Burridge,
approach taken by experts in using only robust species Lauren Chapman, Lindsay Chatterton, Barry Chernoff, Lynda
occurrence data, account for many of these lower-than- Corkum, Ian Cowx, William Crampton, Alain Crivelli,
expected numbers. Numbers of endemics may in some cases Carolina Joana da Silva, Tim Davenport, Luc De Vos,
be higher than expected because endemics were identified Ignacio Doadrio, Carlos DoNascimiento, Luis Fernando
strictly through a database query for unique occurrences, Duboc, Brian Dyer, Carlo Echiverri, Jean Marc Elouard,
and many species lists are undoubtedly incomplete or use Joerg Freyhof, Christopher Frissell, German Galvis, Angus
synonyms. We anticipate that many tallies will change with Gascoigne, Abebe Getahun, A. Gopalakrishnan, Michael
further refinement of species lists but that the broad patterns Goulding, Jon Harding, Tan Heok Hui, Liu Huanzhang,
presented here will hold. Leonardo Ingenito, Michel Jégu, Howard Jelks, Aaron Jenk-
ins, Wolfgang Junk, Ad Konings, Friedhelm Krupp, Philippe
Conclusions Lalèyè, Carlos Alcala Lasso, Christian Lévêque, Flávio C. T.
The newly available species data attributed to ecoregions has Lima, Cas Lindsey, Jorge Liotta, Marcelo Loureiro, Carlos
important implications for prioritizing conservation invest- Lucena, Margarete Lucena, Paulo Henrique Lucinda, Anto-
ments. As one illustration, in 2005 the Global Environment nio Machado-Allison, Christopher Magadza, Luis Malabarba,
Facility (GEF), which spends more than $1 billion each year Mabel Maldonado, Maria Cristina Dreher Mansur, Larry
on environmental projects, adopted a new resource allocation Master, Don McAllister, Robert McDowall, J. D. McPhail,
framework. Terrestrial ecoregion maps and biodiversity data Geraldo Mendes dos Santos, Naércio A. Menezes, Roberto
were notable inputs to the framework, but parallel fresh- Carlos Menni, Jose Ivan Mojica, Peter Moyle, Thierry Ober-
water information to help guide investments was lacking. dorff, Javier Maldonado Ocampo, Mike K. Oliver, Hernan
The GEF framework fortunately leaves open the possibility Ortega, Mark Oswood, Vadim E. Panov, Carla Simone
of incorporating freshwater ecoregions and biodiversity data Pavanelli, Christine Poellabauer, David Propst, Edson Pereira,
at a later date (GEF 2005). Saul Prada, Francisco Provenzano, Gordon McGregor Reid,
In addition to providing data for scientific and conserva- Anthony J. Ribbink, Francisco Antonio Rodrigues Barbosa,
tion purposes, we aim to give the largest possible number of Ricardo S. Rosa, Norma J. Salcedo-Maúrtua, Jansen Alfredo
people access to the ecoregion-level information collected in Sampaio Zuanon, Robert Schelly, Michael Schindel, Uli
association with the global map. The information will be Schliewen, Juan Jacobo Schmitter Soto, Martin Schneider-
freely available on the Internet (www.feow.org) as well as in Jacoby, Uwe Horst Schulz, Lothar Seegers, Ole Seehausen, Scott
brochures, posters, and other publications. The freshwater Smith, John S. Sparks, Don Stewart, Donald Taphorn, Christo-
ecoregion map covers virtually all land surfaces on Earth, so pher Taylor, Guy Teugels, Louis Tsague, Denis Tweddle, Paul
people around the globe will have the opportunity to learn Van Damme, D. Thys van den Audenaerde, Stephen J. Walsh,
about the freshwater systems where they live. Claude Weber, Robin Welcomme, James D. Williams, Phillip
For most policymakers, water resource managers, and Willink, and Stamatis Zogaris. Additionally, William Esch-
even conservationists, freshwater biodiversity is more of an meyer and Stan Blum provided critical support toward im-
afterthought than a central consideration of their work. The proving our fish species database. The importance of biological
freshwater ecosystem services that support the lives and liveli- collections and the work of taxonomists are basic to all bio-
hoods of countless people worldwide are a far larger concern. geographic mapping projects, and so we acknowledge and
Yet freshwater biodiversity and ecosystem services are linked highlight the fundamental contribution of collections and tax-
through ecological integrity, and better-informed efforts to onomy to this effort and to conservation generally. Institutions
conserve freshwater biodiversity should benefit human com- and organizations that have generously provided data and
munities as well. The freshwater ecoregions of the world map assistance include the American Fisheries Society’s Endangered

412 BioScience • May 2008 / Vol. 58 No. 5 www.biosciencemag.org

 Articles
Gery J. 1969. The fresh-water fishes of South America. Pages 828–848 in Fitkau
Species Committee, the American Museum of Natural
EJ, ed. Biogeography and Ecology in South America. The Hague (Nether-
History, Belgium’s Royal Museum for Central Africa, the lands): W. Junk.
California Academy of Sciences, FishBase, Fundación La Salle Gilman RT, Abell RA, Williams CE. 2004. How can conservation biology
de Ciencias Naturales, Instituto Nacional de Pesquisas da inform the practice of integrated river basin management? Journal of River
Amazônia, IUCN, Museu de Ciências e Tecnologia PUCRS, Basin Management 2: 135–148.
Groombridge B, Jenkins M. 1998. Freshwater Biodiversity: A Preliminary
Museo de Zoologia de la Universidad Central de Venezuela,
Global Assessment. Cambridge (UK): World Conservation Monitoring
Museo de Zoologia de la Universidad Nacional de los Llanos Centre.
Occidentales, Museu Nacional do Rio de Janeiro, Nature- Groves CR, Jensen DB, Valutis LL, Redford KH, Shaffer ML, Scott JM, Baum-
Serve, South African Institute for Aquatic Biodiversity, and gartner JV, Higgins JV, Beck MW, Anderson MG. 2002. Planning for
the Zoological Museum of the University of Copenhagen. Eze- biodiversity conservation: Putting conservation science into practice.
BioScience 52: 499–512.
quiel Zamora, George Ledec, Douglas Graham, and Gonzalo
Higgins JV. 2003. Maintaining the ebbs and flows of the landscape:
Castro were instrumental in the earliest stages of this project. Conservation planning for freshwater ecosystems. Pages 291–318 in

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

We also thank Nasser Olwero for his development of the Groves C, ed. Drafting a Conservation Blueprint: A Practitioner’s Guide
FEOW Web site; Eric Dinerstein for his guidance and review to Planning for Biodiversity. Washington (DC): Nature Conservancy
of an earlier manuscript; and many additional former and and Island Press.
Illies J, ed. 1978. Limnofauna Europaea. New York: Gustav Fischer.
current WWF and Nature Conservancy staff acknowledged Kamdem-Toham A, et al. 2003. Biological Priorities for Conservation in
on the FEOW Web site, including but not limited to Jamie Pit- the Guinean-Congolian Forest and Freshwater Region. Libreville (Gabon):
tock, Allison Pease, Brian Blankespoor, and Tucker Gilman. WWF-CARPO.
This work was supported in part by grants to WWF from the Keith P, Vigneux E, Marquet G. 2002. Atlas des Poissons et des Crustacés
Coca-Cola Company and JohnsonDiversey Inc. Additional d’Eau Douce de Polynésie Française. Paris: Muséum National d’Histoire
Naturelle.
support was generously provided to the Nature Conservancy Kottelat M, Freyhof J. 2007. Handbook of European Freshwater Fishes.
by Bill Barclay, Ofelia Miramontes, and John Mordgridge. Switzerland: Steven Simpson Books.
Work in South America was supported in part by the US Lévêque C, ed. 1997. Biodiversity Dynamics and Conservation: The Fresh-
Agency for International Development through award num- water Fish of Tropical Africa. Cambridge (UK): Cambridge University
ber EDG-A-00-01-0023-00 for the Parks in Peril Program. Press.
Lévêque C, Balian EV, Martens K. 2005. An assessment of animal species
diversity in continental waters. Hydrobiologia 542: 39–67.
References cited Lundberg JG, Kottelat M, Smith GR, Stiassny MLJ, Gill AC. 2000. So many
Abell RA, et al. 2000. Freshwater Ecoregions of North America: A Conser- fishes, so little time: An overview of recent ichthyological discovery in
vation Assessment. Washington (DC): Island Press. continental waters. Annals of the Missouri Botanical Garden 87: 26–62.
Allen GR. 1991. Field Guide to the Freshwater Fishes of New Guinea. Madang Maxwell JR, Edwards CJ, Jensen ME, Paustian SJ, Parrott H, Hill DM. 1995.
(Papua New Guinea): Christensen Research Institute. A Hierarchical Framework of Aquatic Ecological Units in North Amer-
Allen GR, Midgley SH, Allen M. 2002. Field Guide to the Freshwater Fishes ica (Nearctic Zone). St. Paul (MN): USDA Forest Service, North Central
of Australia. Perth (Australia): Western Australian Museum. Forest Experiment Station. General Technical Report NC-176.
Balian EV, Lévêque C, Segers H, Martens K, eds. 2008. Freshwater Animal McDonald R, McKnight M, Weiss D, Selig E, O’Connor M, Violin C,
Diversity Assessment. Dordrecht (Netherlands): Springer. Moody A. 2005. Species compositional similarity and ecoregions: Do eco-
Banarescu P. 1990. Zoogeography of Fresh Waters, vol. 1: General Distribu- region boundaries represent zones of high species turnover? Biological
tion and Dispersal of Freshwater Animals. Weisbaden (Germany): AULA. Conservation 126: 24–40.
Berra TM. 2001. Freshwater Fish Distribution. San Diego: Academic Press. McDowall RM. 1990. New Zealand Freshwater Fishes: A Natural History and
Burgess GH, Franz R. 1989. Zoogeography of the Antillean freshwater fish Guide. Auckland (New Zealand): Hinemann Reed.
fauna. Pages 263–304 in Woods CA, Sergile FE, eds. Biogeography of the [MEA] Millennium Ecosystem Assessment. 2005. Ecosystems and Human
West Indies: Patterns and Perspectives. Boca Raton (FL): CRC. Well-being: Synthesis. Washington (DC): Island Press.
Bussing WA. 1976. Geographic distribution of the San Juan ichthyofauna of Menni RC. 2003. Peces y ambientes en la Argentina continental. Mono-
Central America with remarks on its origin and ecology. Pages 157–175 graph of the Museo Argentino de Ciencias Naturales 5: 1–316.
in Thorson TB, ed. Investigations of Nicaraguan Lakes. Lincoln: University Miller RR, Minckley WL, Norris S. 2005. Fishes of México. Chicago: University
of Nebraska. of Chicago Press.
Contreras-Balderas S. 2000. Biogeografía mexicana de peces continentales. NatureServe. 2006. NatureServe Explorer: An Online Encyclopedia of Life.
Mexicoa 2: 80–84. Version 6.1. NatureServe. (17 March 2008; www.natureserve.org/explorer)
de Jesus F, coord. 2003. Classification of Aquatic Ecosystems of the Pantanal Oberdorff T, Guegan JF, Hugueny B. 1995. Global scale patterns of fish
and the Upper Paraguay Watershed. Brasilia (Brazil): Nature Conservancy. species richness in rivers. Ecography 18: 345–352.
Dinerstein E, Olson DM, Graham DJ, Webster AL, Primm SA, Bookbinder Olson DM, Dinerstein E, Canevari P, Davidson I, Castro G, Morisset V,
MP, Ledec G. 1995. A Conservation Assessment of the Terrestrial Eco- Abell R, Toledo E. 1998. Freshwater Biodiversity of Latin America and the
regions of Latin America and the Carribean. Washington (DC): World Caribbean: A Conservation Assessment. Washington (DC): Biodiversity
Bank. Support Program.
Dudgeon D, et al. 2006. Freshwater biodiversity: Importance, threats, status Olson DM, et al. 2001. Terrestrial ecoregions of the world: A new map of
and conservation challenges. Biological Reviews 81: 163–182. life on Earth. BioScience 51: 933–938.
Eschmeyer WN. 2006. The Catalog of Fishes On-line (updated 7 November Omernik JM. 1987. Ecoregions of the conterminous United States. Annals
2006). California Academy of Sciences. (14 March 2008; www.calacademy. of the Association of American Geographers 77: 118–125.
org/research/ichthyology/catalog/fishcatsearch.html) Paavola R, Muotka T, Virtanen R, Heino J, Jackson D, Maki-Petays A. 2006.
[GEF] Global Environment Facility. 2005. Technical Paper on the GEF Spatial scale affects community concordance among fishes, benthic
Resource Allocation Framework. (4 April 2008; www.gefweb.org/ macroinvertebrates, and bryophytes in streams. Ecological Applications
Operational_Policies/Resource_Allocation_Framework.html) 16: 368–379.

www.biosciencemag.org May 2008 / Vol. 58 No. 5 • BioScience 413

Articles
Ramsar Bureau. 2006. Strategic Framework for the List of Wetlands of Thieme ML, Abell R, Stiassny MLJ, Skelton P, Lehner B, Teugels GG, Diner-
International Importance of the Convention on Wetlands (Ramsar, stein E, Kamdem-Toham A, Burgess N, Olson D. 2005. Freshwater Eco-
Iran, 1971), edition 2006. (17 March 2008; www.ramsar.org/key_guide_ regions of Africa and Madagascar: A Conservation Assessment.
list2006_e.htm) Washington (DC): Island Press.
Rauchenberger M. 1988. Historical biogeography of Poeciliid fishes in the Tonn WM. 1990. Climate change and fish communities: A conceptual frame-
Caribbean. Systematic Zoology 37: 356–365. work. Transactions of the American Fisheries Society 119: 337–352.
Reis R, Kullander S, Ferraris C, eds. 2003. Check List of the Freshwater Udvardy MDF. 1975. A Classification of the Biogeographic Provinces of the
Fishes of South and Central America. Porto Alegre (Brazil): EDIPUCRS. World. Gland (Switzerland): IUCN. Occasional Paper no. 18.
Revenga C, Kura Y. 2003. Status and trends of biodiversity of inland water
Unmack PJ. 2001. Biogeography of Australian freshwater fishes. Journal of
ecosystems. Montreal (Canada): Secretary of the Convention on Biological
Biogeography 28: 1053–1089.
Diversity. Technical Series no. 11.
Vinson MR, Hawkins CP. 2003. Broad-scale geographical patterns in local
Revenga C, Murray S, Abramovitz J, Hammond A. 1998. Watersheds of the
stream insect genera richness. Ecography 26: 751–767.
World: Ecological Value and Vulnerability. Washington (DC): World
Resources Institute. Wallace AR. 1876. The Geographical Distribution of Animals. New York:
Harper.

Downloaded from https://academic.oup.com/bioscience/article/58/5/403/234819 by guest on 18 July 2023

Revenga C, Campbell I, Abell R, de Villiers P, Bryer M. 2005. Prospects for
monitoring freshwater ecosystems towards the 2010 targets. Philosoph- Wasson J-G, Barrera S, Barrere B, Binet D, Collomb D, Gonzales I, Gourdin
ical Transactions of the Royal Society B 360: 397–413. F, Guyot J-L, Rocabado G. 2002. Hydro-ecoregions of the Bolivian
Ringuelet R. 1975. Zoogeografía y ecología de los peces de aguas continen- Amazon: A geographical framework for the functioning of river eco-
tales de la Argentina y consideraciones sobre las áreas ictiológicas de systems. Pages 69–91 in McClain ME, ed. The Ecohydrology of South
América del Sur. Ecosur 2: 1–122. American Rivers and Wetlands. Wallingford (UK): International
Roberts TR. 1975. Geographical distribution of African freshwater fishes. Association of Hydrological Sciences.
Zoological Journal of the Linnean Society 57: 249–319. Weitzell RE, Khoury ML, Gagnon P, Schreurs B, Grossman D, Higgins J. 2003.
Scott WB, Crossman EJ. 1998. Freshwater Fishes of Canada. Oakville Conservation Priorities for Freshwater Biodiversity in the Upper
(Canada): Galt House. Mississippi Basin. Arlington (VA): NatureServe and The Nature
Skelton PH. 1994. Diversity and distribution of freshwater fishes in East and Conservancy.
Southern Africa. Annals of the Royal Central Africa Museum (Zoology) Wikramanayake E, Dinerstein E, Loucks C, Olson D, Morrison J, Lamoreux
275: 95–131. J, McKnight M, Hedao P. 2002. Ecoregions in ascendance: Reply to
Skelton PH, Cambray JA, Lombard A, Benn GA. 1995. Patterns of distribu- Jepson and Whittaker. Conservation Biology 16: 238–243.
tion and conservation status of freshwater fishes in South Africa. South
African Journal of Zoology 30: 71–81.
Spalding MD, et al. 2007. Marine ecoregions of the world: A bioregionalization doi:10.1641/B580507
of coast and shelf areas. BioScience 57: 573–583. Include this information when citing this material.

414 BioScience • May 2008 / Vol. 58 No. 5 www.biosciencemag.org