ArticlePDF Available

A Three-Dimensional Mapping of the Ocean Based on Environmental Data

March 2017
Oceanography 30(1):90-103

DOI:10.5670/oceanog.2017.116

Authors:

Roger Sayre

United States Geological Survey

Dawn J Wright

Environmental Systems Research Institute (ESRI)

Sean Breyer

Environmental Systems Research Institute (ESRI)

Kevin Butler

Environmental Systems Research Institute (ESRI)

Show all 17 authorsHide

Author Posting. © The Oceanography Society, 2017. This article is posted here by permission of The Oceanography Society for personal use, not for redistribution. The definitive version was published in Oceanography 30, no. 1 (2017): 90–103, doi:10.5670/oceanog.2017.116.

Traditional oceanographic notions of vertical zonation in the ocean. Modified from Pinet (2009).

…

. Existing maps of ocean regions.

…

A vertical column of the ocean mesh framework (illustrative and not to scale), produced from World Ocean Atlas 2013 data extracted into a set of 52,487,233 points at ¼° × ¼° ( ~ 27 km × 27 km at the equator) horizontal resolution and variable depth z ranging from 5 m intervals near the surface to 100 m intervals near the deep seafloor. After constructing the mesh, points were attributed with 57-year average values for temperature, salinity, dissolved oxygen, nitrate, phosphate, and silicate.

…

EMU distributions by depth. The two-dimensional global area (km 2 ) at any depth is shown for the 22 EMUs that comprise 99% of the ocean volume. The horizontal boundary lines separating the depth zone classes are as described in the Coastal and Marine Ecosystem Classification Standard (CMECS), the Federal Geographic Data Committee (FGDC) standard for the United States (FGDC, 2012). The EMU number labels (see Appendices 1 and 2 in the online supplementary materials for names, maps, and descriptions of the EMUs) are placed at the median unit-middle depth for each EMU. Although the EMUs are not uniformly distributed into the CMECS depth zones, strong vertical separation is evident, with many small EMUs in the upper water column and fewer larger EMUs in the middle and lower water columns. Pink colors indicate warmer EMUs, and blue colors indicate colder EMUs.

…

Relationship between surface-occurring EMU distributions (colors) and marine biogeographic realms (numbered, outlined polygons), from recent work of author Costello. Spatial congruence between biogeographic realms and surfaceoccurring EMUs is apparent for some realms (e.g., 5, 7, 26, 30) but not for others (e.g., 18, 21, 22).

…

Figures - uploaded by Dawn J Wright

Content may be subject to copyright.

Content uploaded by Dawn J Wright

Content may be subject to copyright.

CITATION

Sayre, R.G., D.J. Wright, S.P. Breyer, K.A. Butler, K. Van Graafeiland,

M.J. Costello, P.T. Harris, K.L. Goodin, J.M. Guinotte, Z. Basher, M.T. Kavanaugh,

P.N. Halpin, M.E. Monaco, N. Cressie, P. Aniello, C.E. Frye, and D. Stephens.

2017. A three- dimensional mapping of the ocean based on environmental data.

Oceanography 30(1):90–103, https://doi.org/10.5670/oceanog.2017.116.

DOI

https://doi.org/10.5670/oceanog.2017.116

This article has been published in Oceanography, Volume 30, Number 1, a quarterly

USAGE

Permission is granted to copy this article for use in teaching and research.

Republication, systematic reproduction, or collective redistribution of any portion of

this article by photocopy machine, reposting, or other means is permitted only with the

approval of TheOceanography Society. Send all correspondence to: info@tos.org or

The Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA.

Oceanography

THE OFFICIAL MAGAZINE OF THE OCEANOGRAPHY SOCIETY

DOWNLOADED FROM HTTP://TOS.ORG/OCEANOGRAPHY

Oceanography | Vol.30, No.1

REGULAR ISSUE FEATURE

A Three-Dimensional Mapping of

the Ocean Based on Environmental Data

By Roger G. Sayre, Dawn J. Wright, Sean P. Breyer, Kevin A. Butler, Keith Van Graafeiland,

Mark J. Costello, Peter T. Harris, Kathleen L. Goodin, John M. Guinotte, Zeenatul Basher, Maria T. Kavanaugh,

Patrick N. Halpin, Mark E. Monaco, Noel Cressie, Peter Aniello, Charles E. Frye, and Drew Stephens

Oceanography | Vol.30, No.1

Oceanography | March 2017 91

FACING PAGE. Three-dimensional visualization

of the ecological marine units (EMUs) for the

Banda Sea. EMUs are depicted as bands on cyl-

inders, and pink colors indicate warmer EMUs,

where blue colors represent colder EMUs. On

land, the global ecological land units (ELUs) of

Sayre etal. (2014) are shown.

INTRODUCTION

Ecosystems are a central focus of many

current research and policy questions,

including: (1) What are the impacts of

climate change on ecosystems? (2) Which

ecosystems are vulnerable to climate and

other perturbations (e.g., invasive spe-

cies, land and sea use)? (3) Which ecosys-

tems should be targeted for conservation?

(4) What are the economic and social val-

ues of ecosystem goods and services?

and (5) What role do ecosystems play in

global food and environmental security

(Liu et al., 2007, 2015)? Fundamental

knowledge of the types and locations of

global ecosystems is necessary to address

these questions, yet that knowledge is

generally lacking.

e development of a new global eco-

systems map, including terrestrial, fresh-

water, and marine domains, was therefore

commissioned by the intergovernmen-

tal Group on Earth Observations (GEO,

https://www.earthobservations.org), a

consortium of over 100 nations seeking

to advance Earth observation approaches

for addressing societal challenges related

to hazards, food, water, energy, and the

environment. e new global ecosys-

tems maps were to be derived from data

rather than from expert opinion or socio-

political considerations, and they were to

be based on the physical environmental

features that are understood to inuence

the distribution of species.

Ecosystems are geographically iden-

tiable areas where the interactions

of organisms with their physical envi-

ronments produce dierences in biotic

diversity, trophic structure, and ows

of energy and materials between living

and nonliving components of the system

(Odum, 1971). On land, variations in cli-

mate, landform, and substrate establish

the environmental potential that controls

primary production and species distribu-

tions, acknowledging that evolutionary

history is also an important element of

biogeography (Bailey, 1996, 2014; Kottek

etal., 2006; Holt etal., 2013). Responding

to the GEO commission request for a

standardized, robust, and practical global

map of terrestrial ecosystems, a new map

of global ecological land units (ELUs)

was developed from an integration of

climate, landform, lithology, and land

cover (Sayre etal., 2014). We now pres-

ent a similar environmental stratication

approach for extending the global ecolog-

ical units map into the ocean through the

delineation of global ecological marine

units (EMUs).

ere are notable dierences between

mapping terrestrial and mapping marine

ecological units. First, the terrestrial

ELUs were mapped as two-dimensional

(2D) entities using a raster data surface.

Marine ecosystems, however, are funda-

mentally understood as both 2D (e.g.,sea

surface and seaoor) and three-dimen-

sional (3D; e.g., water column) entities

(e.g., Li and Gold, 2004; Wright et al.,

2007). EMUs would ideally need to be

mapped using 3D data points represent-

ing a volumetric mesh (e.g., Heinzer

etal., 2012; Reygondeau etal., 2017) and

visualized as 2D and 3D objects. Second,

the characteristics of the physical envi-

ronment that inuence the distribution

of species and ecosystems are dierent

between terrestrial and marine environ-

ments. While the ELUs were identied

as distinct combinations of bioclimate,

landform, lithology, and vegetation, those

ABSTRACT. e existence, sources, distribution, circulation, and physicochemical

nature of macroscale oceanic water bodies have long been a focus of oceanographic

inquiry. Building on that work, this paper describes an objectively derived and globally

comprehensive set of 37 distinct volumetric region units, called ecological marine

units (EMUs). ey are constructed on a regularly spaced ocean point-mesh grid,

from sea surface to seaoor, and attributed with data from the 2013 World Ocean Atlas

version 2. e point attribute data are the means of the decadal averages from a 57-year

climatology of six physical and chemical environment parameters (temperature,

salinity, dissolved oxygen, nitrate, phosphate, and silicate). e database includes over

52 million points that depict the global ocean in x, y, and z dimensions. e point

data were statistically clustered to dene the 37 EMUs, which represent physically

and chemically distinct water volumes based on spatial variation in the six marine

environmental characteristics used. e aspatial clustering to produce the 37 EMUs

did not include point location or depth as a determinant, yet strong geographic and

vertical separation was observed. Twenty-two of the 37 EMUs are globally or regionally

extensive, and account for 99% of the ocean volume, while the remaining 15 are smaller

and shallower, and occur around coastal features. We assessed the vertical distribution

of EMUs in the water column and placed them into classical depth zones representing

epipelagic (0 m to 200 m), mesopelagic (200 m to 1,000m), bathypelagic (1,000 m

to 4,000 m) and abyssopelagic (>4,000 m) layers. e mapping and characterization

of the EMUs represent a new spatial framework for organizing and understanding

the physical, chemical, and ultimately biological properties and processes of oceanic

water bodies. e EMUs are an initial objective partitioning of the ocean using long-

term historical average data, and could be extended in the future by adding new

classication variables and by introducing functionality to develop time-specic EMU

distribution maps. e EMUs are an open-access resource, and as both a standardized

geographic framework and a baseline physicochemical characterization of the oceanic

environment, they are intended to be useful for disturbance assessments, ecosystem

accounting exercises, conservation priority setting, and marine protected area network

design, along with other research and management applications.

Oceanography | Vol.30, No.1

elements of terrestrial ecosystem struc-

ture do not apply as such in the ocean,

with the exception of seaoor land-

forms. e abiotic controls on the distri-

bution of marine biota (e.g.,temperature,

as in Beaugrand etal., 2013), and nutri-

ents, as in Longhurst, 2007) were identi-

ed as analogs to the terrestrial environ-

mental characteristics. ird, the ocean

is a uid in motion, and ocean ecosys-

tem conditions can be inuenced by pro-

cesses far from their location via ocean

circulation. Marine ecosystems are there-

fore generally understood as more spa-

tiotemporally dynamic than their terres-

trial counterparts.

Considerable research in quantita-

tive water mass analysis has produced a

robust and standardized terminology for

describing water masses with respect to

their origins, properties, and locations

(Tomczak, 1999). e ocean’s hydro-

graphic structure has been described

with evolving complexity, beginning with

the seminal work of Sverdrup (1942) and

including the early delineation of global

water masses by Emery and Meincke

(1986) based on temperature and salinity

properties. Expanding on the temperature

and salinity relationships, more recent

quantitative water mass characterizations

are multidimensional and feature the use

of tracer distributions to identify water

mass origins. For example, Gebbie and

Huybers (2011) comprehensively iden-

tied the surface origin of points in the

ocean interior at 33 depth levels using cli-

matological and isotope ratio data and

tracer path analysis.

In addition to origin-based quantita-

tive water mass analysis, another com-

mon approach to identifying oceanic

water bodies involves subdivision of

ocean regions or volumes based on dif-

ferences in their physical, chemical, and

biological properties (Sherman et al.,

2005; Longhurst, 2007; Spalding et al.,

2007; Reygondeau et al., 2017). While

various marine oceanic region maps exist

(Table 1), few are global in extent, are

representative of the entire water column

in three dimensions, and were derived

from quantitative analysis of data. ree-

dimensional, globally comprehensive

subdivisions of the ocean are particularly

lacking. e quantitative, 3D analysis of

Gebbie and Huybers (2010, 2011) iden-

tied seven surface water source regions

of global ocean water, but did not map

dierent regions by depth. Reygondeau

et al. (2017) objectively subdivided the

Mediterranean Sea into 63 biogeochem-

ical regions in a true 3D analysis using

data and biologically meaningful crite-

ria to separate the water column verti-

cally into epipelagic, mesopelagic, bathy-

pelagic, and seaoor zones. However,

there has not been a purely quantita-

tive, unsupervised approach to parti-

tioning the entire global ocean water

column from aspatial statistical cluster-

ing of global ocean data, even though

an unprecedented amount of global

marine environmental data is now avail-

able (e.g.,the 2013 World Ocean Atlas of

Locarnini etal., 2013; Zweng etal., 2013;

Garcia etal., 2014a, 2014b).

We approached the challenge of aggre-

gating comprehensive marine environ-

mental data through statistical clustering,

building on the eorts of previous authors.

For example, Harris and Whiteway

(2009) used a multivariate statistical

method with six biophysical variables

(depth, seabed slope, sediment thickness,

primary production, bottom- water dis-

solved oxygen, and bottom temperature)

TABLE 1. Existing maps of ocean regions.

NAME GEOGRAPHIC

SCOPE

BASIS

Countries’ Exclusive Economic Zones (EEZs)

(United Nations, 1982) Global, Coastal Political

Large Marine Ecosystems (LMEs) (Sherman etal., 2005) Global, Coastal Management areas

Marine Ecoregions of the World (MEOWs)

(Spalding etal., 2007) Global, Coastal Expert-derived biogeography (realms, provinces) and

management units (ecoregions)

Fisheries and Agricultural Organization (FAO) Major

Fishing Areas (FAO, 2016) Global Rectangular ﬁshery statistical assessment regions

International Council for the Exploration of the Sea

Ecoregions (ICES, 2004)

Regional

(Northeast Atlantic) Large ecosystem and ﬁshery management areas

International Hydrographic Organization Seas and Oceans

(IHO, 2002) Global Geographically named areas

Ecoregions of the Oceans and Continents (Bailey, 2014) Global Expert recommended regions

Global Open Ocean and Deep Seabed (GOODS)

Biogeographic Characterization (UNESCO, 2009)

Global, Benthic and

Pelagic Expert recommended regions

Deep-Sea Provinces (Watling etal., 2013) Global, Benthic Expert-derived revision of GOODS based on literature review

Biogeochemical Provinces (Longhurst, 2007) Global Satellite ocean color

Seaﬂoor Map (GSFM) (Harris etal., 2014) Global Expert geomorphological feature extraction using 30 arc-

second bathymetry data

Deep-Sea Seascapes Map (Harris and Whiteway, 2009) Global Multivariate analysis of seabed morphology and sediments

Oceanography | March 2017 93

to objectively classify the entire ocean

oor into 53,713 separate polygons com-

prising 11 dierent categories. e 11 cat-

egories had mean polygon sizes ranging

from 1,000 km2 to 22,000 km2 and were

restricted to the seaoor. Reygondeau

et al. (2017) statistically clustered data

from the Mediterranean Sea into dis-

tinct biogeochemical regions within bio-

logically meaningful, predetermined

depth zones. To our knowledge, the pres-

ent study is the rst to objectively clas-

sify the entire global ocean water col-

umn simultaneously across all depths

into discrete regions based on compre-

hensive statistical clustering of physi-

cal and chemical environmental data

from all points in the World Ocean Atlas

(WOA)-derived ocean mesh.

Most oceanography and marine biol-

ogy textbooks include diagrams that

divide the ocean into depth zones

(e.g.,Figure1). Shallower, sunlit depths at

or near the surface are usually presented

as the photic layer, which extends to the

general limit of light penetration (99% of

incident light) at a depth of about 200 m

(Stal, 2016). Beneath this depth, photo-

synthesis is largely lacking (Costello and

Breyer, 2017). is same depth zone to

200 m is also commonly referred to as

the epipelagic zone. Beneath this zone

at depths commonly understood as

between 200 m and 1,000 m is the meso-

pelagic zone, where organismal respi-

ration is higher relative to deeper areas

(Costello and Breyer, 2017). Although

the 1,000 m depth boundary is arbi-

trarily dened, Proud etal. (2017) objec-

tively subdivided the mesopelagic region

into distinct subzones using organismal

echolocation data, and Reygondeau etal.

(2017) use ux of particulate organic car-

bon to determine biologically meaning-

ful boundaries for the mesopelagic layer.

Deeper, darker, and colder zones are usu-

ally presented as bathyl, abyssal, and

hadal zones. Although depth boundaries

for these regions are largely arbitrary and

can vary from text to text, they are meant

to describe the biogeochemical varia-

tion that is correlated with depth. It is

possible that as depth changes, variation

in temperature and chemical composi-

tion creates distinct ecological zones rep-

resented by dierent ecological commu-

nities. However, this concept has never

been objectively tested using data at a

global scale. With few exceptions (Oliver

and Irwin, 2008; Hardman-Mountford

etal., 2009; Harris and Whiteway, 2009;

Kavanaugh et al., 2014; Schoch et al.,

2014; Reygondeau et al., 2017), most

existing marine maps and zonation sys-

tems are derived from supervised clas-

sication and thus are inuenced by the

perspectives of their authors (Costello,

2009). To test whether recognizable

boundaries exist vertically and horizon-

tally in the global ocean, we clustered 3D

ocean cells into groups using an unsuper-

vised classication of physical and chem-

ical environmental variables.

METHODS

We used the complete set of variables

from the 2013 World Ocean Atlas data set,

version 2 (Locarnini etal., 2013; Zweng

et al., 2013; Garcia et al., 2014a, 2014b)

as our source of physical and chemi-

cal environmental data for dening the

ocean mesh and subsequently modeling

the ecological marine units. e WOA

data set is a compendium of data from a

variety of ocean research and monitoring

programs over the past ve decades. It is

an authoritative 57-year climatology that

contains over 52 million points, hereaer

referred to as the ocean mesh. Each point

is attributed with values for temperature,

salinity, dissolved oxygen, nitrate, phos-

phate, and silicate, and all WOA values

are corrected for the eect of pressure on

each variable. e WOA has a horizon-

tal spatial resolution of ¼° × ¼° for tem-

perature and salinity, and 1° × 1° for oxy-

gen, nitrate, phosphate, and silicate. In

the vertical dimension, points are located

at variable depth intervals, ranging from

5 m increments near the surface to 100 m

increments at depth. A total of 102 depth

zones extend to 5,500 m. e depth inter-

vals are as follows: 5 m (from 0 m to

100 m), 25 m (from 100 m to 500 m),

50 m (from 500 m to 2,000 m), and 100 m

(from 2,000 m to 5,500 m). e deepest

points for which data are available do not

necessarily represent the actual depth of

the water column because the 5,500 m

lower limit of the WOA data is approx-

imately half of the maximum depth of

the ocean (Jamieson, 2011). However,

the 5,500 m lower limit does substan-

tially exceed the mean depth (3,682.2 m)

FIGURE1. Traditional oceanographic notions of vertical zonation in the ocean. Modiﬁed

from Pinet (2009).

Oceanography | Vol.30, No.1

of the ocean as reported in the review by

Charette and Smith (2010). e WOA

data are water-column variables, but sea-

oor geomorphology may also be signif-

icant in inuencing both these variables

and species ecology. e rst global digi-

tal map of seaoor geomorphic features is

now available (Harris etal., 2014).

Temporally, the WOA archive is avail-

able in seasonal, annual, and decadal res-

olutions. Seasonal data are not available

for all points in the mesh, many of which

may not have been visited regularly over

the 57-year period. Moreover, data from

polar regions, typically collected only

during warmer summer months when

access to ice-bound regions is easier, may

under-report true salinity values. Decadal

values of the WOA represent the average

of the annual mean values for the param-

eters, themselves derived from the sea-

sonal data. We used the 57-year record of

the parameters, which are provided in the

WOA database as archival means, derived

from the decadal averages. e modeled

EMUs therefore represent average distri-

butions of the volumetric regions over the

past 57 years.

We constructed an ocean point mesh as

a 3D spatial data structure that holds the

WOA data in its highest available spatial

resolution of ¼° × ¼° (~27 km × 27 km at

the equator) in the horizontal dimension.

While the temperature and salinity data

are available at this resolution, the other

four variables (dissolved oxygen, nitrate,

phosphate, and silicate) have a coarser

native resolution (1° × 1°) and were there-

fore downscaled to the ¼° resolution to

reconcile all data to a common work-

ing horizontal resolution. is down-

sampling was accomplished by subdivid-

ing the 1° × 1° by depth-interval rectan-

gular box cuboid into sixteen ¼° × ¼°

by depth-interval cuboids and assign-

ing the original attribute values of the

parent cuboid’s centroid to the cen-

ter points of all of the ¼° subdivisions.

In this piecewise-constant re-meshing,

we assume that the attributes of the par-

ent cuboid are uniform throughout the

cuboid’s volume. is is similar to the

universal assumption in vector-based

GIS that the attributes of a vector poly-

gon are uniform throughout the poly-

gon’s extent. Statistical and nonstatistical

downscaling of coarser-resolution data

such as global climate model (GCM)

data to ner- resolution data is a com-

mon practice in global change modeling

(Hall, 2014), and it is also the basis for

pansharpening of multi resolution imag-

ery (Vivone etal., 2015).

e data matrix can be conceptualized

as columnar stacks of cells whose cen-

troids dene the point mesh (Figure2).

In areas where the deepest (5,500 m)

WOA data points did not reach the sea-

oor, the bottom of the mesh was sim-

ply extended downward to the seaoor

for visualization, without interpolating

additional data points. e mesh spac-

ing matched the WOA data matrix and

allowed for the structuring and sym-

bolization of data as columnar volumes

(or other shapes) that can be queried by

ranges of values, and can be spatially ana-

lyzed via proximity algorithms and multi-

variate statistical clustering. We rst con-

structed an “empty” ocean mesh using

the 52,487,233 WOA point locations,

and then attached the WOA attribute

data to those points. e water column

was bounded at the top by the sea sur-

face, and at the bottom by the seaoor, as

dened in the geomorphic map of Harris

etal. (2014). e set of cells intersecting

or nearest to the global shoreline (or ice

masses) dened the horizontal extent of

the water column.

We statistically clustered the points in

the mesh in order to identify environ-

mentally distinct regions in the water col-

umn. e clustering was blind to both

the depth of the point and the thickness

of the depth interval at that point’s verti-

cal position in the water column. e “big

data” nature of the clustering of the entire

ocean volume required sophisticated spa-

tial data processing and functionality.

e clustering was implemented using

SAS and all other SAS Institute Inc. prod-

uct or service names are registered trade-

marks or trademarks of SAS Institute

Inc., Cary, NC, USA). e ArcGIS plat-

form was utilized for subsequent geo-

spatial assessment and visualization of

FIGURE 2. A vertical column of the ocean

mesh framework (illustrative and not to scale),

produced from World Ocean Atlas 2013 data

extracted into a set of 52,487,233 points

at ¼° × ¼° (~27 km × 27 km at the equator)

horizontal resolution and variable depth

z ranging from 5 m intervals near the sur-

face to 100 m intervals near the deep sea-

ﬂoor. After constructing the mesh, points

were attributed with 57-year average val-

ues for temperature, salinity, dissolved

oxygen, nitrate, phosphate, and silicate.

Oceanography | March 2017 95

the clusters. We utilized a k-means clus-

tering algorithm to identify the physi-

cal and chemical structure of the water

column. e k-means algorithm deter-

mines k centroids in the data and clusters

points by assigning them to the nearest

centroid. Of hundreds of clustering algo-

rithms available, the k-means approach is

the most widely used due to its simplic-

ity, versatility, extensibility, data handling

ability, and generally robust performance

(Jain, 2010), although it is sensitive to ini-

tial placement of cluster centers (Celebi

etal., 2013). While concurrent implemen-

tation and integration of complementary

clustering approaches has been advocated

for ocean partitioning (Oliver etal., 2004;

Reygondeau et al., 2017), this multi-

algorithm approach was outside the

scope of our globally comprehensive and

data intensive analysis.

Our statistical approach was proto-

typed on a subset (97,329 points) of the

global point mesh representing the ocean

volume o the US West Coast out to the

Exclusive Economic Zone (EEZ). e

successful identication of known hydro-

graphic features (e.g., the Mendocino

Ridge and Fracture Zone o the northern

California coast) in the prototype exer-

cise provided initial assurances that the

clustering approach would be sensitive to

environmental gradients, and that scal-

ing up to global clustering was warranted.

We therefore implemented the clustering

globally on all cells (>52 million points),

with all variables included.

All the WOA variables were standard-

ized to a mean of zero and a standard

deviation of one to establish a common

basis for comparison between variables

of disparate units and value ranges and

to promote relative equal weightings

of the inputs to the clustering (Milligan

and Cooper, 1988). Aer standardiza-

tion, a Pearson’s correlation analysis of

the six inputs was implemented to iden-

tify colinearity among variables. To

determine the optimal number of clus-

ters that would best represent the col-

lective variation in the input data, clus-

tering of all the WOA points with all

six variables was executed in repeated

sequential runs, where the number of

clusters produced was incremented by

one with each run, starting with two

clusters, and ending with 100 clus-

ters. e optimum cluster number was

determined by inspection of the behav-

ior of the pseudo F-statistic (Calinski

and Harabasz, 1974; Milligan and

Cooper, 1985) across the iterations. e

pseudo-F statistic is the ratio of between-

cluster variance to within- cluster vari-

ance. Larger values of pseudo-F indicate

“tight” (i.e., low within- cluster variance)

and “well separated” (i.e., high between-

cluster variance) clusters. A plot of this

statistic against the number of clusters

should show local peaks of the pseudo-F

value at potential cluster- number optima.

We did not extend the clustering beyond

100 clusters because there was a clear

overall decline in pseudo-F values as the

number of clusters increased. Local peaks

representing relatively high pseudo-F val-

ues were found at 17, 28, 37, and 50 clus-

ters (Figure 3). We explored summary

statistics and the horizontal and verti-

cal spatial distributions at each of the

four local peaks. At 37 clusters, a strong

peak was observed prior to a relatively

sustained decline in the pseudo-F curve

(Figure 3), which we interpreted as a

point where additional clustering is less

likely to reduce the within- cluster vari-

ation. us, the 37-cluster solution was

the basis for our partitioning of the water

column, and resulted in the 37 EMUs

described below.

Following the depth-blind statisti-

cal clustering, basic descriptive statistics

(mean, minimum, maximum, and stan-

dard deviation) were produced for the six

deterministic parameters (temperature,

salinity, dissolved oxygen, nitrate, phos-

phate, and silicate) for each EMU. e

unit-middle depth for each EMU was

also calculated as the median depth of all

the points allocated into an EMU.

We then labeled the clusters using the

naming criteria (Table 2) of the Coastal

and Marine Ecosystem Classication

Standard (CMECS), a Federal Geographic

Data Committee standard for the United

States (FGDC, 2012). e CMECS labels

for the EMUs begin with their depth

zone assignments based on their median

depths, followed by a concatenation of

the CMECS descriptors for tempera-

ture, salinity, and dissolved oxygen. e

CMECS framework does not include

FIGURE3. A plot of the pseudo-F statistic (y axis) against the requested num-

ber of clusters (x axis) in successive iterations from 2 to 100 clusters, incre-

mented by one for each successive iteration. The red vertical line at 37 clus-

ters shows a strong peak prior to a relatively sustained decline in the curve of

the pseudo-F statistic, which we interpret as a stopping-point where additional

clustering does not signiﬁcantly reduce within-cluster heterogeneity (Calinski

and Harabasz, 1974; Milligan and Cooper, 1985). We therefore chose the

37-cluster solution to represent the number and distributions of global EMUs.

Pseudo-F (x 106)

Cluster #

10 50 9030 7020 6040 8015 55 9535 755 45 8525 65

Oceanography | Vol.30, No.1

standard names and value ranges for

nitrate, phosphate, and silicate. For these

three variables, labels corresponding to

high, medium, and low nutrient concen-

trations in a relative sense were deter-

mined from assessment of the observed

distribution of values. ese three nutri-

ent descriptors were added to the four

CMECS descriptors for a total of seven

descriptors in the CMECS names. e

sequence of presentation of the descrip-

tors used in the CMECS names is: depth,

temperature, salinity, dissolved oxygen,

nitrate, phosphate, and silicate. While

depth was not used as a clustering vari-

able, it is a key descriptor in the CMECS

classication and an important variable

for considering ocean depth zones, and

was therefore included in the CMECS

labeling. As an illustrative example of

the CMECS nomenclature, an EMU

cluster might be named Bathypelagic,

Very Cold, Euhaline, Severely Hypoxic,

High Nitrate, Medium Phosphate, and

Medium Silicate.

We then developed a separate and par-

allel label, the EMU name (Table 2), to

simplify the CMECS terminology. e

EMU equivalent of the CMECS clus-

ter described above is Deep, Very Cold,

Normal Salinity, Very Low Oxygen, High

Nitrate, Medium Phosphate, and Medium

Silicate. e CMECS and simplied EMU

names reect the properties of the EMU,

not its location in the ocean. Naming

the EMUs based on their chemical and

physical properties is both accurate and

“classication neutral” in the sense that

the label is purely descriptive in a compo-

sitional sense (Sayre etal., 2014).

Finally, in addition to the CMECS

and EMU compositional names, we

also developed a set of EMU volumet-

ric region names that describe both their

geographic distributions in the ocean and

their vertical positions in the water col-

umn. In the eld of quantitative water

mass analysis, water masses are generally

associated with or dened by their for-

mation regions (Tomczak, 1999; Emery,

2001) or surface water sources (Gebbie

and Huybers, 2010, 2011). Because we

TABLE 2. Depth and physicochemical properties (column 1) and corresponding depth zones

and regime units of the Coastal and Marine Ecosystem Classiﬁcation Standard (CMECS; FGDC,

2012; column 2) and of Ecological Marine Units (EMUs; column 3). CMECS regimes do not exist

for nitrate, phosphate, and silicate. They were therefore adopted from EMU terms developed

for these three variables, based on assessment and classiﬁcation of the approximately 52 mil-

lion observations for each nutrient into three relative classes (high, medium, and low).

DEPTH m CMECS MARINE OCEANIC

WATER COLUMN LAYER

EMU WATER

COLUMN LAYER

0 to <200 Epipelagic Shallow

200 to <1,000 Mesopelagic Moderate Depth

1,000 to <4,000 Bathypelagic Deep

≥4,000 Abyssopelagic Very Deep

TEMPERATURE

°C

CMECS TEMPERATURE REGIME EMU TEMPERATURE

REGIME

20 to <30 Warm to Very Warm Warm

10 to <20 Moderate to Cool Cool

5 to <10 Cold Cold

0 to <5 Very Cold Very Cold

≤0 Frozen/Superchilled Superchilled

SALINITY

dimensionless

CMECS SALINITY REGIME EMU SALINITY

REGIME

>30 Euhaline to Hyperhaline Normal Salinity

18 to 30 Lower to Upper Polyhaline Low Salinity

<18 Oligohaline to Mesohaline Very Low Salinity

DISSOLVED

OXYGEN ml/L

CMECS OXYGEN REGIME EMU OXYGEN

REGIME

≥8 Highly Oxic to Very Oxic High Oxygen

4 to <8 Oxic Moderate Oxygen

2 to <4 Hypoxic Low Oxygen

0.1 to <2 Severely Hypoxic Very Low Oxygen

<0 .1 Anoxic No Oxygen

NITRATE

M

CMECS NITROGEN REGIME

(adopted from next column)

EMU NITROGEN

REGIME

>30 High Nitrate High Nitrate

10 to 30 Medium Nitrate Medium Nitrate

<10 Low Nitrate Low Nitrate

PHOSPHATE

M

CMECS PHOSPHATE REGIME

adopted from next column)

EMU PHOSPHATE

REGIME

>5 High Phosphate High Phosphate

2.5 to 5 Medium Phosphate Medium Phosphate

<2.5 Low Phosphate Low Phosphate

SILICATE

M

CMECS SILICATE REGIME

(adopted from next column)

EMU SILICATE

REGIME

>100 High Silicate High Silicate

50 to 100 Medium Silicate Medium Silicate

<50 Low Silicate Low Silicate

Oceanography | March 2017 97

have not identied the source geogra-

phies for our EMUs, we avoid calling

them water masses herein. We instead

describe the total volumetric distribu-

tion of an EMU as a volumetric region

rather than as a water mass. Each EMU

volumetric region name contains both

a geographic descriptor and a CMECS

depth zone class (epipelagic, meso-

pelagic, bathypelagic, and abyssopelagic)

based on the median depth of the EMU.

Examples of EMU volumetric regions

include Antarctic and Subantarctic

Bathypelagic, Mediterranean and Red

Seas Mesopelagic, and Arctic and

Labrador Sea Epipelagic.

RESULTS

e strength of the relationship between

the standardized variables and the result-

ing EMU congurations was either strong

(>90 to <95% condence) or statistically

signicant (≥ 95% condence) for each

of the six input variables: temperature

(R2 = 0.95), salinity (R2 = 0.97), dissolved

oxygen (R2 = 0.91), nitrate (R2 = 0.96),

phosphate (R2 = 0.96), and silicate

(R2 = 0.94). A strong correlation (>.8)

among the three nutrient inputs (nitrate,

phosphate, silicate) was observed, sug-

gesting that they may have had a slightly

disproportionate inuence on the cluster-

ing. However, we did not remove any of the

nutrient variables and then re-cluster in

order to ensure that regional variation in

any of the six input variables could inu-

ence the clustering outcome. e ratio

of between-cluster variance to within-

cluster variance (R2/(1 − R2) was: tem-

perature, 17.47; salinity, 12.89; dissolved

oxygen, 10.72; nitrate, 24.42; phosphate,

22.11; and silicate, 15.12. ese results

indicate that the six parameters contrib-

uted strongly and approximately equally

to the identication of the clusters.

Clustering the entire set of points in

the global mesh yielded 37 mutually

exclusive clusters (EMUs) that are volu-

metric regions of relative compositional

homogeneity. For a listing of EMUs

described using CMECS and EMU ter-

minology, and the names of the EMU

volumetric regions, see Appendix 1

(available in the online supplementary

materials). Maps of the EMUs, along

with descriptive statistics on their envi-

ronmental characteristics, are found in

Appendix2 (also available in the online

supplementary materials).

e EMUs are presented in a num-

ber of 2D slices at several depths in

Figure 4. A plot of EMU area against

depth, shown in Figure 5, characterizes

the vertical position of the EMUs in the

water column and the depth at which

the maximum and minimum horizon-

tal distributions occur can be easily visu-

ally interpreted. is graph also shows

the vertical distribution of the EMUs

with respect to the CMECS boundar-

ies for epipelagic (0 m to 200 m), meso-

pelagic (200 m to 1,000m), bathypelagic

(1,000m to 4000 m), and abyssopelagic

(>4,000 m) zones. e number labels in

each EMU represent the EMU number

for cross-referencing to the EMU names

and maps in Appendices 1 and 2. ese

EMU number labels have been placed

vertically at a depth corresponding to the

median unit-middle of the EMU.

e global maps of the EMUs listed in

Appendix 2 show the maximum global

horizontal extent of the clusters look-

ing vertically from above, as well as the

FIGURE4. The global distribution of EMUs at eight depth intervals. EMUs represent physically and

chemically distinct volumetric regions based on combined temperature, salinity, oxygen, and nutri-

ent gradients. While a total of 37 EMUs were statistically determined, a number of them are small,

localized, and shallow, and are not discernible in these depth-layer maps. Black indicates regions

shallower than the depth at that layer. Major hydrographic features like Northern and Southern

Hemisphere gyre systems and coastal upwelling-based westward ﬂow of water from western con-

tinental margins are evident, particularly at shallower depths (upper left and right panels). Colors

reﬂect mean EMU temperatures, with pink colors representing warmer EMUs and blue colors rep-

resenting colder EMUs.

Oceanography | Vol.30, No.1

thickness of the EMU at any location.

Interacting visually with true volumes in

3D, especially with such a large resource,

is a soware challenge, but it is possible

with commercially available virtual globe-

based visualization soware, as shown in

Figure6. Although the EMUs are continu-

ous data surfaces, Figure6 shows that they

are more easily visualized as stacked bands

on horizontally separated cylinders.

e size and complexity of the EMU

resource, a de facto example of big data

with over 52 million attributed points

in three dimensions, potentially pres-

ents barriers to its ecient use. To help

mitigate this challenge, we developed

an open-access web application, EMU

Explorer (http://livingatlas.arcgis.com/

emu) that permits real-time query and

visualization of both the points and the

EMUs by anyone with Internet access.

e application shows the vertical prole

of the water column from sea surface to

seaoor at any user-selected surface loca-

tion, and it returns the values for the

physical and chemical properties of the

points in the column. It also identies the

EMUs associated with any point in the

vertical prole and provides the EMUs’

descriptive statistics.

DISCUSSION

EMU Geographic Distributions

Twenty-two of the EMUs are large, with

essentially global or large regional dis-

tributions, while the 15 others are small,

shallow, and coastal, and collectively rep-

resent only about 1% of the ocean vol-

ume. ey generally have lower salini-

ties than the other EMUs, and are found

where mixing of fresh and saline waters is

occurring (e.g.,the Baltic Sea and north-

ern, ice-occurring regions). While suit-

able for global-scale stratication of the

open ocean into large volumetric regions,

the ¼° horizontal spatial resolution of

the ocean mesh may not be sucient to

completely resolve the ner resolution,

ecologically meaningful coastal systems.

We therefore consider these very small,

yet statistically derived coastal clusters as

likely indicators of coastal and estuarine

EMUs that need to be further claried.

Latitudinal distribution patterns

of EMUs are observed (Figure 4 and

Appendix 2), with EMUs occurring in

latitudinal biomes that include polar

and subpolar regions (e.g., EMUs 14,

19, 23, 25 and 31), temperate regions

(e.g., EMUs 3, 8, and 37), subtropi-

cal and tropical regions (e.g.,EMUs11,

24, 26, 33), and equatorial regions

(e.g., EMUs 10, 18). Bimodal latitudinal

distributions (Northern and Southern

Hemispheres) are observed (e.g.,EMUs8,

11, 21, and 24). Some EMUs are associ-

ated with physiographic features, such as

the Mendocino Ridge and Fracture Zone,

situated near the southern boundary of

EMU30. Parts of some EMUs are located

in the discharge regions of major rivers

like the Amazon and Congo (EMUs 18

and 24). e Mediterranean and Red Seas

were clustered into a single unit (EMU9),

consistent with Longhurst’s (2007) identi-

cation of a single biogeochemical prov-

ince for the Mediterranean. Reygondeau

et al. (2017), however, objectively iden-

tied 63 three-dimensional, manage-

ment-appropriate subdivisions of the

Mediterranean, maintaining that while

global analyses are useful for macro-

scale comparisons of ocean regions, local

management strategies and policies will

require appropriately scaled geographic

assessment and accounting units.

One very large, deep, circumglobal

cluster (EMU 13) is observed in the

Pacic and Indian Oceans but is all but

absent in the Atlantic, consistent with

the recognition of a Circumpolar Deep

Water (CDW) mass by Emery and

Meincke (1986). Likewise, EMU 29 is

similar to the Arctic Deep Water (ADW)

and North Atlantic Deep Water (NADW)

units of those authors. Although some

FIGURE 5. EMU distributions by

depth. The two-dimensional global

area (km2) at any depth is shown

for the 22 EMUs that comprise

99% of the ocean volume. The hor-

izontal boundary lines separat-

ing the depth zone classes are as

described in the Coastal and Marine

Ecosystem Classiﬁcation Standard

(CMECS), the Federal Geographic

Data Committee (FGDC) stan-

dard for the United States (FGDC,

2012). The EMU number labels

(see Appendices 1 and 2 in the

online supplementary materials for

names, maps, and descriptions of

the EMUs) are placed at the median

unit-middle depth for each EMU.

Although the EMUs are not uni-

formly distributed into the CMECS

depth zones, strong vertical sepa-

ration is evident, with many small

EMUs in the upper water column

and fewer larger EMUs in the mid-

dle and lower water columns. Pink

colors indicate warmer EMUs, and

blue colors indicate colder EMUs.

30 31

21 24

926

336

Surface Area (106 km2)

Depth (m)

ABYSSO

PELAGIC

BATHYPELAGIC MESOPELAGIC EPIPELAGIC

100200 30

0400

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

3,000

4,000

5,000

5,500

Oceanography | March 2017 99

of the EMUs are similar to water masses

described by Emery and Meincke, in

other instances there is less resemblance,

which is to be expected, given that the

EMUs were derived from six composi-

tional properties rather than from the

temperature- and salinity-derived units

of Emery and Meincke.

e surface-occurring EMUs (upper

le panel of Figure4) can be compared

with the surface-derived biogeochemical

provinces (BGCPs) of Longhurst (2007).

As mentioned above, the Mediterranean

Sea was identied as a single unit (EMU9)

without subdivision in both classica-

tions. e Mediterranean and Red Seas

EMU was placed in the mesopelagic class

because its median unit-middle depth

(302 m) is between 200 m and 1,000 m, but

its vertical distribution is throughout the

water column. Other similarities between

the EMUs and Longhurst’s BGCPs are

apparent. Both classication systems

identify obvious latitudinal banding sep-

arating the Antarctic, Subantarctic, and

Southern Hemisphere tropics and equa-

torial regions. EMU 18 (North Pacic

Subtropical and Equatorial Indian

Epipelagic) closely approximates the dis-

tributions of Longhurst’s North Pacic

Equatorial Countercurrent Province and

Pacic Equatorial Divergence Province.

Several of Longhurst’s provinces in the

Arctic and Subarctic regions (e.g.,Boreal

Polar Province, Atlantic Arctic Province,

Atlantic Subarctic Province, North

Pacic Epicontinental Sea Province) cor-

respond visually with EMUs 5 (Arctic

Epipelagic), 23 (Arctic and Labrador Sea

Epipelagic), and 30 (North Pacic and

Beaufort Sea Epipelagic). e latitudinal

demarcation between Longhurst’s two

Indian Ocean provinces (Indian South

Subtropical Gyre Province and Indian

Monsoon Gyres Province) is a strong lat-

itudinal delineation between EMUs 11

(Northern Subtropical and Southern

Subtropical Epipelagic), 18 (North

Pacic Subtropical and Equatorial Indian

Epipelagic), and 24 (Tropical Pacic,

Tropical Indian, and Equatorial Atlantic

Epipelagic). e Longhurst system does

not include bimodal distributions of

provinces in both Northern and Southern

Hemispheres as was obtained with some

EMUs (e.g.,EMU8, Subantarctic, North

Atlantic, and North Pacic Epipelagic).

Overall, considerable visual correspon-

dence is observed between Longhurst’s

BGCPs and the surface-occurring

EMUs, and a more quantitative compar-

ison is merited.

Another set of ocean-surface regions

that can be compared to the surface-

occurring EMUs is that of Gebbie and

Huybers (2011), who identied seven

global surface water masses: Antarctic,

North Atlantic, Subantarctic, North

Pacic, Arctic, Mediterranean, and

Tropics, representing the formation

regions of ocean waters. e boundar-

ies separating these seven source regions

are also present in the surface-occurring

EMUs. It is evident that an aggregation of

EMUs into the Gebbie and Huybers for-

mation regions would be very “clean,”

resulting in minimal splitting of EMUs

across formation region boundaries.

Neither the Large Marine Ecosystems

of Sherman etal. (2005) nor the Marine

Ecoregions of the World of Spalding

etal. (2007) address open-ocean pelagic

ecosystems, so comparisons between

them and the geographic extent of the

EMUs are not feasible. e interpretive,

expert-derived subdivisions of the Global

Open Ocean and Deep Seabed (GOODS)

Biogeographic Classication (UNESCO,

FIGURE6. Example of the visualization approach taken to represent the EMUs in three dimensions,

mapped over space. The region shown is o the eastern coast of Japan. Although the EMUs are

mapped as a continuous surface, representing them in three dimensions is facilitated by the use of

stacked cylinders, where each color band on a cylinder is an EMU. In the coastal zone, EMUs are

single or few, and shallow, whereas oshore there are more and deeper EMUs. Surface tempera-

ture gradients are also apparent between the pink (warmer) EMUs in the south, and the blue (colder)

EMUs in the north.

Oceanography | Vol.30, No.1

100

30 realms were obtained from 2D clus-

tering of occurrence records represent-

ing over 65,000 species from the Ocean

Biogeographic Information System

(OBIS, http://www.iobis.org) database,

and they reect global patterns of spe-

cies endemicity. Initially, we note corre-

spondence between realms and EMUs

(e.g.,realms 2, 5, 7, and 30) in some areas,

but in other cases a relationship is less

apparent (e.g.,realms 18, 21, and 22). In

another global assessment of 11,567 spe-

cies occurrences representing 13 tax-

onomic groups, Tittensor et al. (2010)

identied concentrations of coastal and

open-ocean species in the eastern Pacic

and in mid-latitudinal belts, respectively.

e mapping of the EMUs will allow

improved characterization of the hor-

izontal and vertical distributions and

the chemical and physical natures of

species-rich regions.

Limitations and Future Work

We recognize limitations in our work

related to both temporal scaling dimen-

sions and parameters selected for the

clustering. e WOA data oer several

native temporal resolutions (seasonal,

annual, and decadal) that we did not

exploit. As our aim was to use long-term

historical average values for the point

locations, we used the 57-year mean val-

ues for the six parameters to map EMU

2009) are similarly dicult to compare

with the quantitative, statistically derived

EMUs presented here.

EMU Depth Distributions

e number of EMUs is highest at or

near the surface, and decreases with

depth (Figures 4 and 5). Although we

used the CMECS criteria and the median

unit-middle value to classify the EMUs

into epipelagic, mesopelagic, bathype-

lagic, or abyssopelagic zones, Figure 5

shows that those depth class assign-

ments, while informative, are also imper-

fect. Many EMUs are distributed across

depth zone boundaries, with some hav-

ing most of their distribution in one zone,

but with vertical extensions into upper or

lower depth zones as well. EMUs vary

considerably in water column position,

thickness, and horizontal area at varying

depths. No EMUs were classied as abys-

sal as there were no EMUs with a median

unit- middle depth >4,000 m. However,

the distributions of many EMUs extend

beyond the 4,000 m bathypelagic/

abyssopelagic boundary. Acknowledging

some overlap, the EMUs appear to be bet-

ter separated, visually, at depths approx-

imately corresponding to water column

positions from 0 m to 200 m (upper

water column), 200m to 2,000 m (mid-

dle water column), and >2,000 m (lower

water column). Although the EMUs are

currently classied into CMECS depth

zones using standardized criteria, an

attempt to statistically separate EMUs

into depth classes and associated quanti-

tative determination of the depth bound-

aries for those groupings appears war-

ranted. In addition, the EMUs should

be assessed for their depth-based rela-

tionship to physical properties like light

attenuation limits (Stal, 2016) and biolog-

ically mediated phenomena like respira-

tion (Costello and Breyer, 2017) and car-

bon ux (Reygondeau et al., 2017). We

suggest that the depth boundaries of the

vertical zones and the environmental fac-

tors controlling the vertical separation of

the pelagic ocean need additional analy-

sis and further clarication.

Biogeography and EMUs

Biogeographic regions are delineated

from an analysis of species distribu-

tion data. e number and types of tax-

onomic groups represented (Fontaine

etal., 2015), as well as the relative focus

on endemism (Briggs and Bowen, 2012),

can vary widely. Although not quan-

titatively assessed, we made a prelimi-

nary visual comparison (Figure7) of the

EMUs’ spatial distributions and the dis-

tributions of 30 marine biogeographic

realms developed from statistical clus-

tering of species-distribution data, based

on recent work of author Costello. e

FIGURE7. Relationship between

surface-occurring EMU distri-

butions (colors) and marine bio-

geographic realms (numbered,

outlined polygons), from recent

work of author Costello. Spatial

congruence between biogeo-

graphic realms and surface-

occurring EMUs is apparent for

some realms (e.g., 5, 7, 26, 30)

but not for others (e.g.,18, 21, 22).

16 15

14 20 9

Oceanography | March 2017 101

extents and locations, and this approach

has been successful in mapping ocean

regions. However, it prohibits assess-

ment of temporal variability and trends.

Recent work shows that it would be pos-

sible to construct temporally sequenced

EMU distribution maps (e.g.,Oliver and

Irwin, 2008; Reygondeau et al., 2013;

Kavanaugh etal., 2014). We now have a

framework for that assessment and are

planning the development of seasonal,

annual, and decadal characterizations of

oceanic water masses. However, the com-

putational requirements for six variables

increase by orders of magnitude when

contemplating temporal variations for

over 52 million points. is is currently a

big-data challenge (Gallagher etal., 2015;

Alder and Hostetler, 2015; Coro et al.,

2016; Wright, 2016), but as spatial pro-

cessing technologies evolve, these kinds

of analyses will be rendered less compu-

tationally intense than they are at present.

e clustering of oceanic data to derive

EMUs was based on the six variables in

the WOA data set. e addition of other

variables would likely inuence the oce-

anic partitioning we present here. e

inclusion of data on particulate organic

carbon (POC), carbonate contents, and

ocean current patterns might inuence

the clustering results. POC plays a crucial

role in the marine and global carbon cycle

and is a primary component of oceano-

graphic food webs (Buesseler etal., 2007).

POC ux was one of the parameters used

to subdivide the Mediterranean Sea into

63 biogeochemical regions (Reygondeau

etal., 2017), where it was used to quan-

titatively separate the mesopelagic layer

from the bathypelagic layer. Variability

in the vertical ux of POC is import-

ant for understanding the main path-

ways by which organic carbon is formed

in ocean surface waters via photosynthe-

sis and then transferred to the deep ocean

where it may be sequestered (Lutz et al.,

2007). Similarly, the carbonate chemis-

try of the ocean is ecologically import-

ant, as the persistence of ocean acidi-

cation is likely to have implications for

many surface and pelagic ecosystems and

communities (Sherman, 2014; Wallace

etal., 2014; resher etal., 2015). Finally,

variables associated with ocean currents,

such as ow direction and magnitude,

may substantially inuence EMU char-

acteristics and distributions. We plan to

pursue the addition of these attributes to

the ocean mesh and to study their eects

on EMU distributions in future statistical

clustering analyses.

We also plan to enrich the EMU

resource by combining the EMU data

with other data layers. is will result in

the creation of new 2D layers for the sea

surface and the seaoor. For example,

for the seaoor, we have combined the

bottom-occurring EMUs with the sea-

oor physiographic regions and features

of Harris etal. (2014) in order to evalu-

ate the inuence of seaoor geomorphol-

ogy on the water-column structure above

it. We also have combined a 13-year aver-

age ocean color value data set (chloro-

phyll a from the NASA Aqua-MODIS

sensor) to our surface-occurring EMUs,

and plankton abundance characteris-

tics are now available for the surface data

points and the surface-occurring EMUs.

We intend to continue adding associa-

tive attributes from other globally avail-

able resources, and we are exploring the

relationship between EMUs and estab-

lished temporally dynamic climatological

classications (Oliver and Irwin, 2008;

Kavanaugh etal., 2016).

While we are calling these 37 volu-

metric regions EMUs, we acknowledge

that their true ecological character has

not yet been established. eir deriva-

tion from entirely physicochemical data,

and their similarity to widely recognized

global surface waters, lends validation to

the EMUs as physically and chemically

distinct volumetric regions. We call them

ecological in the general sense that depth,

temperature, salinity, oxygen, and nutri-

ents are known to be important in struc-

turing biotic distributions (Longhurst,

2007; Oliver and Irwin, 2008), and

because microbial processes shape nutri-

ent and oxygen distributions through-

out the water column (Kavanaugh etal.,

2016), but we have not documented the

relationship between environmental vari-

ation and species diversity. As a rst step,

we are currently undertaking a more

quantitative assessment of this relation-

ship between physically and chemically

distinct regions in the ocean and spe-

cies biogeography, the results of which

will facilitate a deeper understanding of

the true ecological nature of the EMUs.

For example, we are exploring the cross-

indexing of OBIS species records and

EMUs, and we expect that subsequent

versions of EMUs will not only contain

species records as attributes but may also

change geographically to be more reec-

tive of marine organism distributions.

Finally, we recognize the opportu-

nity to use ner-resolution data to cre-

ate a more rened mapping of EMUs at

regional and local scales, as was demon-

strated by Reygondeau et al. (2017).

Moreover, we plan to elucidate the coastal

and estuarine units in greater detail in an

independent development of a set of eco-

logical coastal units (ECUs), which will

be undertaken along the entire global

shoreline. We are working on the devel-

opment of a new global shoreline vector

extracted from satellite imagery (30 m

spatial resolution) and attributed with

environmental characteristics as the spa-

tial framework for the planned develop-

ment of a global set of ECUs.

CONCLUSIONS

e present EMU mapping eort is an

objective partitioning of the global ocean

into environmentally distinct volumet-

ric region units using an aspatial cluster-

ing exercise where clusters were not con-

strained into particular ocean regions

and the cluster sites were selected by

homogeneity in physical and chemical

parameters only, blind to both depth and

location. We have developed a new clas-

sication scheme for 37 compositionally

varying marine volumetric regions. eir

properties are listed using both standard-

ized CMECS descriptors and new EMU

terminology. EMU volumetric region

names are compiled by combining a

Oceanography | Vol.30, No.1

102

geographic descriptor and a depth zone

term (Appendices 1 and 2).

e aim of this work was to produce a

new global characterization and detailed

data set of marine environments as a

resource for biogeographic assessments,

impact studies, biodiversity priority set-

ting, and ecosystem accounting, man-

agement, and research. e mapping of

global ecological land units (ELUs), spon-

sored by the GEO commission, has now

been extended to the ocean. We parti-

tioned the global ocean water column

into 37 physically and chemically dis-

tinct volumetric regions using available

data from the 57-year World Ocean Atlas

data set. e global map of EMUs is an

initial, unsupervised, statistical classi-

cation approach to mapping ocean envi-

ronmental structure in three dimen-

sions. Based on this methodology, we can

re-cluster the ocean using additional or

dierent deterministic variables if desir-

able, and also cluster the global ocean in

dierent time intervals ranging from sea-

sonal to annual to decadal to explore the

temporal geographies of EMUs. e exis-

tence of the EMUs has the potential to

facilitate research on the extent to which

environmental drivers control biotic dis-

tributions. e EMU data we have cre-

ated allow for characterization of the

physical and chemical environment con-

tained in marine protected areas, shing

grounds, or other marine geographies.

As an open-access resource, the EMU

data are available to scientists, managers,

and the interested public. Future work

will enrich the EMU resource by adding

additional attributes to the ocean mesh

and developing a ner resolution ecolog-

ical coastal unit (ECU) product along the

global shoreline.

SUPPLEMENTARY MATERIALS

Appendices 1 and 2 are available at https://doi.org/

10.5670/oceanog.2017.116.

REFERENCES

Alder, J.R., and S.W. Hostetler. 2015. Web based visu-

alization of large climate data sets. Environmental

Modelling & Software 68:175–180, https://doi.org/

10.1016/j.envsoft.2015.02.016.

Bailey, R.G. 1996. Ecosystem Geography. Springer-

Verlag, New York, NY, 204 pp.

Bailey, R.G. 2014. Ecoregions – The Ecosystem

Geography of the Oceans and Continents, 2nd ed.

Springer, New York, NY, 180 pp.

Beaugrand, G., I. Rombouts, and R.R. Kirby. 2013.

Towards an understanding of the pattern of bio-

diversity in the oceans. Global Ecology and

Biogeography 22:440–449, https://doi.org/10.1111/

geb.12009.

Briggs, J.C., and B.W. Bowen. 2012. A realignment

of marine biogeographic provinces with par-

ticular reference to ﬁsh distributions. Journal

of Biogeography 39:12–30, https://doi.org/

10.1111/j.1365-2699.2011.02613.x.

Buesseler, K.O., C.H. Lamborg, P.W. Boyd, P.J. Lam,

T.W. Trull, R.R. Bidigare, J.K.B. Bishop, K.L. Casciotti,

F. Dehairs, M. Elskens, and others. 2007. Revisiting

carbon ﬂux through the ocean’s twilight zone.

Science 316(5824):567–570, https://doi.org/10.1126/

science.1137959.

Calinski, T., and J. Harabasz. 1974. A dendrite

method for cluster analysis. Communications

in Statistics 3(1):1–27, https://doi.org/10.1080/

03610927408827101.

Celebi, M., H. Kingravi, and P. Vela. 2013. A compara-

tive study of ecient initialization methods for the

k-means clustering algorithm. Expert Systems with

Applications 40(1):210–215, https://doi.org/10.1016/

j.eswa.2012.07.021.

Charette, M.A., and W.H.F. Smith. 2010. The volume

of Earth’s ocean. Oceanography 23(2):112–114,

https://doi.org/10.5670/oceanog.2010.51.

Coro, G., C. Magliozzi, A. Ellenbroek, K. Kaschner, and

P. Pagano. 2016. Automatic classiﬁcation of climate

change eects on marine species distributions in

2050 using the AquaMaps model. Environmental

and Ecological Statistics 23:155–180,

https://doi.org/10.1007/s10651-015-0333-8.

Costello, M.J. 2009. Distinguishing marine hab-

itat classiﬁcation concepts for ecological

data management. Marine Ecology Progress

Series 397:253–268, https://doi.org/10.3354/

meps08317.

Costello, M.J., and S. Breyer. 2017. Ocean depths: The

mesopelagic and implications for global warming.

Current Biology 27:113–119, https://doi.org/10.1016/

j.cub.2016.11.042.

Emery, W.J. 2001. Water types and water

masses. Pp. 1,556–1,567 in Encyclopedia of

Ocean Sciences. J.H. Steele, S.A. Thorpe, and

K.K. Turekian, eds, Academic Press, San Diego, CA,

3,399 pp.

Emery, W.J., and J. Meincke. 1986. Global water

masses: Summary and review. Oceanologica

Acta 9(4):383–391.

FAO (Food and Agriculture Organization). 2016.

Coordinating Working Party Handbook of Fishery

Statistical Standards, Section H: Fishing Areas

for Statistical Purposes. FAO Fisheries and

Aquaculture Department, http://www.fao.org/

ﬁshery/cwp/handbook/h/en.

FGDC (Federal Geographic Data Committee). 2012.

Coastal and Marine Ecological Classiﬁcation

Standard Version 4.0. Report FGDC-STE-18-2012,

Federal Geographic Data Committee Secretariat,

US Geological Survey, Reston, Virginia, 343 pp.

Fontaine, A., R. Devillers, P.R. Peres-Neto, and

L.E. Johnson. 2015. Delineating marine eco-

logical units: A novel approach for deciding

which taxonomic group to use and which tax-

onomic resolution to choose. Diversity and

Distributions 21(10):1,167–1,180, https://doi.org/10.1111/

ddi.12361.

Gallagher, J., J. Orcutt, P. Simpson, D. Wright,

J. Pearlman, and L. Raymond. 2015. Facilitating

open exchange of data and information. Earth

Science Informatics 8:721–739, https://doi.org/

10.1007/s12145-014-0202-2.

Garcia, H.E., R.A. Locarnini, T.P. Boyer, J.I. Antonov,

O.K. Baranova, M.M. Zweng, J.R. Reagan, and

D.R. Johnson. 2014a. World Ocean Atlas 2013

version 2 (WOA13 V2), Volume 3: Dissolved

Oxygen, Apparent Oxygen Utilization, and Oxygen

Saturation. S. Levitus, ed., and A. Mishonov,

technical ed., NOAA Atlas NESDIS 75, NOAA

National Centers for Environmental Information,

Silver Spring, MD, 27 pp.

Garcia, H.E., R.A. Locarnini, T.P. Boyer, J.I. Antonov,

O.K. Baranova, M.M. Zweng, J.R. Reagan, and

D.R. Johnson. 2014b. World Ocean Atlas 2013

version 2 (WOA13 V2), Volume 4: Dissolved

Inorganic Nutrients (phosphate, nitrate, silicate).

S. Levitus, ed., and A. Mishonov, technical ed.,

NOAA Atlas NESDIS 76, NOAA National Centers

for Environmental Information, Silver Spring, MD,

25 pp.

Gebbie, G., and P. Huybers. 2010. Total matrix inter-

comparison: A method for determining the geom-

etry of water-mass pathways. Journal of Physical

Oceanography 40:1,710–1,728, https://doi.org/

10.1175/2010JPO4272.1.

Gebbie, G., and P. Huybers. 2011. How is the ocean

ﬁlled? Geophysical Research Letters 138, L06604,

https://doi.org/10.1029/2011GL046769.

Hall, A. 2014. Projecting regional change. Science

346(6216):1,461–1,462, https://doi.org/10.1126/

science.aaa0629.

Hardman-Mountford, N.J., T. Hirata, K. Richardson,

and J. Aiken. 2009. An objective methodology

for the classiﬁcation of ecological pattern into

biomes and provinces for the pelagic ocean.

Remote Sensing of Environment 112:3,341–3,352,

https://doi.org/10.1016/j.rse.2008.02.016.

Harris, P.T., M. Macmillan-Lawler, J. Rupp, and

E.K. Baker. 2014. Geomorphology of the oceans.

Marine Geology 352:4–24, https://doi.org/10.1016/

j.margeo.2014.01.011.

Harris, P.T., and T. Whiteway. 2009. High seas

marine protected areas: Benthic environmen-

tal conservation priorities from a GIS analysis of

global ocean biophysical data. Ocean & Coastal

Management 52:22–38, https://doi.org/10.1016/

j.ocecoaman.2008.09.009.

Heinzer, T.J., M.D. Williams, E.C. Dogrul,

T.N. Kadir, C.F. Brush, and F.I. Chung. 2012.

Implementation of a feature-constraint mesh gen-

eration algorithm within a GIS. Computers and

Geosciences 49:46–52, https://doi.org/10.1016/

j.cageo.2012.06.004.

Holt, B.G., J. Lessard, M.K. Borregaard, S.A. Fritz,

M.B. Araújo, D. Dimitrov, P. Fabre, C.H. Graham,

G.R. Graves, K.A. Jønsson, and others. 2013.

An update of Wallace’s zoogeographic

regions of the world. Science 339(6115):74–78,

https://doi.org/10.1126/science.1228282.

ICES (International Council for Exploration of the

Seas). 2004. Report of the Regional Ecosystem

Study Group for the North Sea. April 5–7, 2004,

ICES CM 2004/ACE:06, International Council for

Exploration of the Seas, Copenhagen, Denmark.

IHO (International Hydrographic Organization).

2002. International Hydrographic Organization

Publication S-23: Limits of Oceans and Seas, draft

4th ed. International Hydrographic Organization,

Paris, Monaco, https://www.iho.int/mtg_docs/

com_wg/S-23WG/S-23WG_Misc/Draft_2002/

Draft_2002.htm.

Jain, A.K. 2010. Data clustering: 50 years

beyond k-means. Pattern Recognition

Letters 31(8):651–666, https://doi.org/10.1016/

j.patrec.2009.09.011.

Jamieson, A.J. 2011. Ecology of deep oceans:

Hadal trenches. eLS, https://doi.org/10.1002/

9780470015902.a0023606.

Kavanaugh, M.T., B. Hales, M. Saraceno, Y.H. Spitz,

A.E. White, and R.M. Letelier. 2014. Hierarchical

and dynamic seascapes: A quantitative frame-

work for scaling pelagic biogeochemistry and ecol-

ogy. Progress in Oceanography 120:291–304,

https://doi.org/10.1016/j.pocean.2013.10.013.

Kavanaugh, M.T., M.J. Oliver, F.P. Chavez, R.M. Letelier,

F.E. Muller-Karger, and S.C. Doney. 2016.

Seascapes as a new vernacular for pelagic ocean

monitoring, management and conservation. ICES

Journal of Marine Science 73(7):1,839–1,850,

https://doi.org/10.1093/icesjms/fsw086.

Oceanography | March 2017 103

Kottek, M., J. Grieser, C. Beck, B. Rudolf, and F. Rubel.

2006. World map of the Köppen-Geiger cli-

mate classiﬁcation updated. Meteorologische

Zietschrift 15(3):259–263.

Li, Z., and C. Gold. 2004. Multi-dimensional geo-

spatial technology for geosciences. Computers

and Geosciences 30(4):321–323, https://doi.org/

10.1016/j.cageo.2003.09.010.

Liu, J., T. Dietz, S.R. Carpenter, M. Alberti, C. Folke,

E. Moran, A.N. Pell, P. Deadman, T. Kratz,

J. Lubchenco, and others. 2007. Complexity

of coupled human and natural systems.

Science 317(5844):1,513–1,516, https://doi.org/

10.1126/science.1144004.

Liu, J., H.A. Mooney, V. Hull, S.J. Davis, J. Gaskell,

T. Hertel, and J. Lubchenco, K.C. Seto, P. Gleick,

C. Kremen, and S. Li. 2015. Systems integra-

tion for global sustainability. Science 347(6225),

https://doi.org/10.1126/science.1258832.

Locarnini, R.A., A.V. Mishonov, J.I. Antonov, T.P. Boyer,

H.E. Garcia, O.K. Baranova, M.M. Zweng, C.R. Paver,

J.R. Reagan, D.R. Johnson, and others. 2013. World

Ocean Atlas 2013 version 2 (WOA13 V2), Volume 1:

Temperature. S. Levitus, ed., and A. Mishonov,

technical ed., NOAA Atlas NESDIS 73, NOAA

National Centers for Environmental Information,

Silver Spring, MD, 40 pp.

Longhurst, A. 2007. Ecological Geography of the Sea,

2nd ed. Academic Press, San Diego, CA, 542 pp.,

https://doi.org/10.1016/B978-012455521-1/50000-0.

Lutz, M.J., K. Caldeira, R.B. Dunbar, and

M.J. Behrenfeld. 2007. Seasonal rhythms of net

primary production and particulate organic car-

bon ﬂux to depth describe the eciency of bio-

logical pump in the global ocean. Journal of

Geophysical Research 112, C10011, https://doi.org/

10.1029/2006JC003706.

Milligan, G.W., and M.C. Cooper. 1985. An examina-

tion of procedures for determining the number of

clusters in a data set. Psychometrika 50(2):159–179,

https://doi.org/10.1007/bf02294245.

Milligan, G.W., and M.C. Cooper. 1988. A study of stan-

dardization of variables in cluster analysis. Journal

of Classiﬁcation 5:181–204, https://doi.org/10.1007/

BF01897163.

Odum, E.P. 1971. Fundamentals of Ecology, 3rd ed.

W.B. Saunders, Philadelphia, PA, 574 pp.

Oliver, M.J., S. Glenn, J.T. Kohut, A.J. Irwin,

O.M. Schoﬁeld, M. Moline, and W.P. Bissett. 2004.

Bioinformatic approaches for objective detec-

tion of water masses on continental shelves.

Journal of Geophysical Research 109, C07S04,

https://doi.org/10.1029/2003JC002072.

Oliver, M.J., and A.J. Irwin. 2008. Objective global

ocean biogeographic provinces. Geophysical

Research Letters 35, L15601, https://doi.org/

10.1029/2008GL034238.

Pinet, P.R. 2009. Invitation to Oceanography, 5th ed.

Jones and Bartlett Publishers Inc.

Proud, R., M.J. Cox, and A.S. Brierley. 2017.

Biogeography of the global ocean’s mesopelagic

zone. Current Biology 27:113–119, https://doi.org/

10.1016/j.cub.2016.11.003.

Reygondeau, G., C. Guieu, F. Benedetti, J. Irrison,

S. Ayata, S. Gasparini, and P. Koubbi. 2017.

Biogeochemical regions of the Mediterranean

Sea: An objective multidimensional and mul-

tivariate environmental approach. Progress in

Oceanography 151:138–148, https://doi.org/10.1016/

j.pocean.2016.11.001.

Reygondeau, G., A. Longhurst, E. Martinez,

G. Beaugrand, D. Antoine, and O. Maury.

2013. Dynamic biogeochemical provinces

in the global ocean. Global Biogeochemical

Cycles 27(4):1,046–1,058, https://doi.org/10.1002/

gbc.20089.

Sayre, R., J. Dangermond, C. Frye, R. Vaughan,

P. Aniello, S. Breyer, D. Cribbs, D. Hopkins,

R. Nauman, W. Derrenbach er, and others. 2014.

A New Map of Global Ecological Land Units:

An Ecophysiographic Stratiﬁcation Approach.

Association of American Geographers,

Washington, DC, 46 pp.

Schoch, G.C., D.M. Albert, and C.S. Shanley. 2014.

An estuarine habitat classiﬁcation for a com-

plex ordal island archipelago. Estuaries and

Coasts 37(1):160–176, https://doi.org/10.1007/

s12237-013-9622-3.

Sherman, K. 2014. Adaptive management institu-

tions at the regional level: The case of Large

Marine Ecosystems. Ocean and Coastal

Management 90:38–49, https://doi.org/10.1016/

j.ocecoaman.2013.06.008.

Sherman, K., M. Sissenwine, V. Christensen,

A. Duda, G. Hempel, C. Ibe, S. Levin, D. Lluch-

Belda, G. Matishov, J. McGlade, and others.

2005. A global movement toward an ecosys-

tem approach to marine resources management.

Marine Ecology Progress Series 300:275–279.

Spalding, M.D., H.E. Fox, G.R. Allen, N. Davidson,

Z.A. Ferdaña, M. Finlayson, B.S. Halpern,

M.A. Jorge, A. Lombana, S.A. Lourie, and

others. 2007. Marine ecoregions of the world:

A bioregionalization of coastal and shelf areas.

BioScience 57(7):573–583, https://doi.org/10.1641/

B570707.

Stal, L.J. 2016. The euphotic realm. Pp. 209–225 in

The Marine Microbiome. L.J. Stal and M.S. Cretoiu,

eds, Springer International Publishing, Switzerland.

Sverdrup, H.U., M.W. Johnson, and R.H. Fleming.

1942. The Oceans: Their Physics, Chemistry and

General Biology. Prentice-Hall, Englewood Clis,

New York, 1,087 pp.

Thresher, R.E., J.M. Guinotte, R.J. Matear, and

A.J. Hobday. 2015. Options for managing

impacts of climate change on a deep-sea com-

munity. Nature Climate Change 5:635–639,

https://doi.org/10.1038/nclimate2611.

Tittensor, D.P., C. Mora, W. Jetz, H.K. Lotze, D. Ricard,

E.V. Berghe, and B. Worm. 2010. Global pat-

terns and predictors of marine biodiver-

sity across taxa. Nature 466(7310):1,098–1,101,

https://doi.org/10.1038/nature09329.

Tomczak, M. 1999. Some historical, theoretical, and

applied aspects of quantitative water mass anal-

ysis. Journal of Marine Research 57:275–303,

https://doi.org/10.1357/002224099321618227.

UN (United Nations). 1982. Article 57: Breadth of

the Exclusive Economic Zone. United Nations

Convention on the Law of the Sea of 10 December

1982, United Nations Secretary General,

Paris, France.

UNESCO (United Nations Educational Scientiﬁc and

Cultural Organization). 2009. Global Open Oceans

and Deep Seabed (GOODS): Biogeographic

Classiﬁcation. UNESCO-IOC Technical Series 84,

United Nations Educational and Scientiﬁc Council

Intergovernmental Oceanographic Commission,

Paris, France, 96 pp.

Vivone, G., L. Alparone, J. Channussot, M. Dalla Mura,

A. Garzelli, G. Licciardi, R. Restaino, and L. Wald.

2015. A critical comparison among pansharpen-

ing algorithms. IEEE Transactions on Geoscience

and Remote Sensing 53(5):2,565–2,586,

https://doi.org/10.1109/TGRS.2014.2361734.

Wallace, R.B., H. Baumann, J.S. Grear, R.C. Aller, and

C.J. Gobler. 2014. Coastal ocean acidiﬁcation: The

other eutrophication problem. Estuarine, Coastal,

and Shelf Science 148:1–13, https://doi.org/10.1016/

j.ecss.2014.05.027.

Watling, L., J. Guinotte, M.R. Clark, and C.R. Smith.

2013. A proposed biogeography of the deep

ocean ﬂoor. Progress in Oceanography 111:91–112,

https://doi.org/10.1016/j.pocean.2012.11.003.

Wright, D.J., ed. 2016. Ocean Solutions, Earth

Solutions, 2nd ed. Esri Press, Redlands, CA, 500 pp.

Wright, D.J., M.J. Blongewicz, P.N. Halpin, and

J. Breman. 2007. Arc Marine: GIS for a Blue Planet.

Esri Press, Redlands, CA, 202 pp.

Zweng, M.M., J.R. Reagan, J.I. Antonov, R.A. Locarnini,

A.V. Mishonov, T.P. Boyer, H.E. Garcia,

O.K. Baranova, D.R. Johnson, D. Seidov, and

M.M. Biddle. 2013. World Ocean Atlas 2013

version 2 (WOA13 V2), Volume 2: Salinity.

S. Levitus, ed., and A. Mishonov, technical ed.,

NOAA Atlas NESDIS 74, NOAA National Centers

for Environmental Information, Silver Spring,

Maryland, 39 pp.

ACKNOWLEDGMENTS

We are grateful for the thoughtful reviews of three

U.S. Geological Survey scientists, Jonathan H. Smith,

Page C. Valentine, and Ilsa B. Kuner, as well as those

of three anonymous reviewers. Cressie’s research

was partially supported by a 2015–2017 Australian

Research Council Discovery Project (DP150104576).

Goodin’s research was partially supported by the

Langar Foundation. Kavanaugh’s research was par-

tially supported by the National Ocean Partnership

Program’s Marine Sanctuaries as Sentinel Sites for

a Demonstration Marine Biodiversity Observation

Network award (NNX14AP62A). Any use of trade,

product, or ﬁrm names is for descriptive pur-

poses only and does not imply endorsement by the

U.S. Government.

AUTHORS

Roger G. Sayre (rsayre@usgs.gov) is Senior

Scientist for Ecosystems, U.S. Geological Survey

(USGS), Reston, VA, USA. Dawn J. Wright is Chief

Scientist, Sean P. Breyer is Program Manager

ArcGIS Content, Kevin A. Butler is Spatial Statistics

Product Engineer, and Keith Van Graafeiland is

Product Engineer, all at Esri, Redlands, CA, USA.

Mark J. Costello is Associate Professor, Institute of

Marine Science, University of Auckland, Auckland,

New Zealand. Peter T. Harris is Managing Director,

GRID-Arendal, Arendal, Norway. Kathleen L. Goodin

is Chief of Sta, Conservation Science Division,

NatureServe, Arlington, VA, USA. John M. Guinotte

is Fish and Wildlife Biologist, U.S. Fish and Wildlife

Service, Denver, CO, USA, formerly at the Marine

Conservation Institute, Seattle, WA, USA, during this

study. Zeenatul Basher is Research Technologist,

USGS, Reston, VA, USA. Maria T. Kavanaugh is

Research Associate, Woods Hole Oceanographic

Institution, Woods Hole, MA, USA. Patrick N. Halpin

is Associate Professor of Marine Geospatial Ecology,

Nicholas School of the Environment, Duke University,

Durham, NC, USA. Mark E. Monaco is Acting Director,

Center for Coastal Monitoring and Assessment,

National Ocean Service, National Oceanic and

Atmospheric Administration, Silver Spring, MD, USA.

Noel Cressie is Distinguished Professor, National

Institute for Applied Statistics Research Australia,

University of Wollongong, Wollongong, Australia.

Peter Aniello is Geospatial Scientist, Sandia National

Laboratory, Albuquerque, NM, USA. Charles E. Frye

is Chief Cartographer, and Drew Stephens is Ocean

Industry Manager, both at Esri, Redlands, CA, USA.

ARTICLE CITATION

Sayre, R.G., D.J. Wright, S.P. Breyer, K.A. Butler,

K. Van Graafeiland, M.J. Costello, P.T. Harris,

K.L. Goodin, J.M. Guinotte, Z. Basher, M.T. Kavanaugh,

P.N. Halpin, M.E. Monaco, N. Cressie, P. Aniello,

C.E. Frye, and D. Stephens. 2017. A three- dimensional

mapping of the ocean based on environmental data.

Oceanography 30(1):90–103, https://doi.org/10.5670/

oceanog.2017.116.

Appendix(1." EMU"numbers," Coastal" and" Marine" Ecosystem" Classification"System" (CMECS)" names,"EMU" names," and" EMU"

volumetric"region" names" and"codes."CMECS"does"not"include" descriptors"for"nitrate,"phosphate," and" silicate,"so"the" labels"

for"these"are"adopted"from" EMU" names" (high," medium"and"low)"based" on" the" observed" distribution"of"values."EMUs"and"

their"number" designations" are" unique."However," volumetric" region"names" are" compiled" from"geographic" and" depth"zone"

descriptors,"and"are"not" unique."Volumetric"region"codes"begin"with"upper" case"geographic"descriptors"followed"by"lower"

case"depth"zone"descriptors"(epipelagic"(0"to"<200"m);"mesopelagic"(200"to"1,000"m);"bathypelagic"(1,000"to"4,000"m);"and"

abyssopelagic"(>4,000" m)." Median" middle-unit"depth"was"used"to" assign" each" EMU" into"a"CMECS"depth"zone."Geographic"

features"in"the"EMU"volumetric"region"names"do"not"include"references"to"isolated"occurrences"of"the"water"volume."Refer"

to" Appendix" 2" for" maps" showing" geographic" locations" of" EMU" volumetric" regions." Compare" identical" volumetric" region"

names" for" details" of" water" volume" properties." Abbreviations:" epi=epipelagic;" meso=mesopelagic;" bathy=bathypelagic;"

abysso=abyssopelagic;" A=Arctic;" AA=Antarctic;" AS=Arabian" Sea;" AT=Atlantic;" ATS=Atlantic" Subtropical;" BA=Baltic" Sea;"

BE=Beaufort"Sea;" BL=Black" Sea;" BR=Bering" Sea;" C=Caspian" Sea;"E=Equatorial;"EA=Equatorial"Atlantic;" EI=Equatorial"Indian;"

I=Indian;" L=Labrador" Sea;" M=Mediterranean" Sea;" NA=North" Atlantic;" NP=North" Pacific;" NPS=North" Pacific" Subtropical;"

NS=Northern"Subtropical;"P=Pacific;" PR=Persian" Gulf;"R=Red" Sea;" SPS=South" Pacific" Subtropical;" SS=Southern" Subtropical;"

SAA=Subantarctic;"SA=South"Atlantic;"SP=South"Pacific;"ST=Subtropical;"T=Tropical;"TA=Tropical"Atlantic;"TI=Tropical"Indian;"

TP=Tropical"Pacific."

EMU"

No."

CMECS"Name""

EMU"Name"

EMU"Volumetric"Region""

Name"and"Code""

Mesopelagic,"Cold,"Polyhaline,"Severely"

Hypoxic,"Low"Nitrate,"Medium"

Phosphate,"High"Silicate"

Moderate"Depth,"Cold,""

Low"Salinity,"Very"Low"Oxygen,"Low"

Nitrate,"Medium"Phosphate,"High"

Silicate"

Black"and"Caspian"Seas"

Mesopelagic"(BL_C_meso)"

Epipelagic,"Cold,"Polyhaline,"Hypoxic,"

Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Cold,"Low"Salinity,"Low"

Oxygen,"Low"Nitrate,"Low"Phosphate,"

Low"Silicate""

Black"and"Caspian"Seas"

Epipelagic"(BL_C_epi)"

Bathypelagic,"Very"Cold,"Euhaline,"

Severely"Hypoxic,"High"Nitrate,"Medium"

Phosphate,"High"Silicate"

Deep,"Very"Cold,"Normal"Salinity,"Very"

Low"Oxygen,"High"Nitrate,"Medium"

Phosphate,"High"Silicate"

North"Pacific"and"Arabian"

Sea"Bathypelagic"

(NP_AS_bathy)"

Mesopelagic,"Cold,"Polyhaline,"Severely"

Hypoxic,"Low"Nitrate,"High"Phosphate,"

High"Silicate"

Moderate"Depth,"Cold,"Low"Salinity,"

Very"Low"Oxygen,"Low"Nitrate,"High"

Phosphate,"High"Silicate"

Black"Sea"Mesopelagic"

(BL_meso)"

Epipelagic,"Frozen/Superchilled,"

Polyhaline,"Highly"Oxic,"Low"Nitrate,"

Low"Phosphate,"Low"Silicate"

Shallow,"Superchilled,"Low"Salinity,"

High"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Arctic"Epipelagic"(A_epi)"

Epipelagic,"Cold,"Polyhaline,"Oxic,"Low"

Nitrate,"Low"Phosphate,"Low"Silicate"

Shallow,"Cold,"Low"Salinity,"Moderate"

Oxygen,"Low"Nitrate,"Low"Phosphate,"

Low"Silicate"

Black"and"Caspian"Seas"

Epipelagic"(BL_C_epi)"

Epiplagic,"Moderate"to"Cool,"

Mesohaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Cool,"Very"Low"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Caspian"Sea"Epipelagic"

(C_epi)"

Epipelagic,"Moderate"to"Cool,"Euhaline,"

Oxic,"Medium"Nitrate,"Low"Phosphate,"

Low"Silicate"

Shallow,"Cool,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Low"Silicate"

Subantarctic,"North"

Atlantic,"and"North"Pacific"

Epipelagic"

(SAA_NA_NP_epi)"

Mesopelagic,"Moderate"to"Cool,"

Euhaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Moderate"Depth,"Cool,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Mediterranean"and"Red"

Seas"Mesopelagic"

(M_R_meso)"

10"

Mesopelagic,"Cold,"Euhaline,"Severely"

Hypoxic,"High"Nitrate,"Low"Phosphate,"

Low"Silicate"

Moderate"Depth,"Cold,"Normal"Salinity,"

Very"Low"Oxygen,"High"Nitrate,"Low"

Phosphate,"Low"Silicate"

Equatorial"Indian,"Tropical"

Atlantic,"and"Tropical"Pacific"

Mesopelagic"

(EI_TA_TP_meso)"

11"

Epipelagic,"Moderate"to"Cool,"Euhaline,"

Oxic,"Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Cool,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Northern"Subtropical"and"

Southern"Subtropical"

Epipelagic"(NS_SS_epi)"

12"

Epipelagic,"Very"Cold,"Mesohaline,"

Severely"Hypoxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Very"Cold,"Very"Low"Salinity,"

Very"Low"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Baltic"Sea"Epipelagic"

(BA_epi)"

13"

Bathypelagic,"Very"Cold,"Euhaline,"

Hypoxic,"High"Nitrate,"Medium"

Phosphate,"High"Silicate"

Deep,"Very"Cold,"Normal"Salinity,"Low"

Oxygen,"High"Nitrate,"Medium"

Phosphate,"High"Silicate"

Pacific"and"Indian"

Bathypelagic"(P_I_bathy)"

14"

Bathypelagic,"Very"Cold,"Euhaline,"Oxic,"

High"Nitrate,"Low"Phosphate,"High"

Silicate"

Deep,"Very"Cold,"Normal"Salinity,"

Moderate"Oxygen,"High"Nitrate,"Low"

Phosphate,"High"Silicate"

Antarctic"and"Subantarctic"

Bathypelagic"

(AA_SAA_bathy)"

15"

Bathypelagic,"Cold,"Polyhaline,"Anoxic,"

Low"Nitrate,"High"Phosphate,"High"

Silicate"

Deep,"Cold,"Low"Salinity,"No"Oxygen,"

Low"Nitrate,"High"Phosphate,"High"

Silicate"

Black"Sea"Bathypelagic"

(BL_bathy)"

16"

Epipelagic,"Very"Cold,"Mesohaline,"

Highly"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Very"Cold,"Very"Low"Salinity,"

High"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Baltic"Sea"Epipelagic"

(BA_epi)"

17"

Epipelagic,"Cold,"Mesohaline,"Oxic,"Low"

Nitrate,"Low"Phosphate,"Low"Silicate"

Shallow,"Cold,"Very"Low"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Baltic"Sea"Epipelagic"

(BA_epi)"

18"

Epipelagic,"Warm"to"Very"Warm,"

Euhaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Warm,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate""

North"Pacific"Subtropical"

and"Equatorial"Indian"

Epipelagic"(NPS_EI_epi)"

19"

Epipelagic,"Cold,"Euhaline,"Oxic,"

Medium"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Cold,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Low"Silicate"

Subantarctic"and"North"

Pacific"Subtropical"

Epipelagic"(SAA_NPS_epi)"

20"

Epipelagic,"Very"Cold,"Mesohaline,"Oxic,"

Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Very"Cold,"Very"Low"Salinity,"

Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Baltic"Sea"Epipelagic"

(BA_epi)"

21"

Epipelagic,"Warm"to"Very"Warm,"

Euhaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Warm,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Atlantic"Subtropical"and"

South"Pacific"Subtropical"

Epipelagic"(ATS_SPS_epi)"

22"

Epipelagic,"Cold,"Mesohaline,"Hypoxic,"

Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Cold,"Very"Low"Salinity,"Low"

Oxygen,"Low"Nitrate,"Low"Phosphate,"

Low"Silicate"

Baltic"and"Black"Seas"

Epipelagic"(BA_BL_epi)"

23"

Epipelagic,"Frozen/Superchilled,"

Euhaline,"Highly"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Superchilled,"Normal"Salinity,"

High"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Arctic"and"Labrador"Sea"

Epipelagic"(A_L_epi)"

24"

Epipelagic,"Warm"to"Very"Warm,"

Euhaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Warm,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Tropical"Pacific,"Tropical"

Indian,"and"Equatorial"

Atlantic"Epipelagic"

(TP_TI_EA_epi)"

25"

Epipelagic,"Frozen/Superchilled,"

Euhaline,"Highly"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Superchilled,"Normal"Salinity,"

High"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Arctic"and"Labrador"Sea"

Epipelagic"(A_L_epi)"

26"

Mesopelagic,"Moderate"to"Cool,"

Euhaline,"Hypoxic,"Medium"Nitrate,"Low"

Phosphate,"Low"Silicate"

Moderate"Depth,"Cool,"Normal"Salinity,"

Low"Oxygen,"Medium"Nitrate,"Low"

Phosphate,"Low"Silicate"

Tropical"and"Subtropical"

Mesopelagic"(T_ST_meso)"

27"

Epipelagic,"Very"Cold,"Polyhaline,"Oxic,"

Low"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Very"Cold,"Low"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Caspian"Sea"Epipelagic"

(C_epi)"

28"

Epipelagic,"Moderate"to"Cool,"

Mesohaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Cool,"Very"Low"Salinity,"Low"

Nitrate,"Low"Phosphate,"Low"Silicate"

Caspian"Sea"Epipelagic"

(C_epi)"

29"

Bathypelagic,"Very"Cold,"Euhaline,"Oxic,"

Medium"Nitrate,"Low"Phosphate,"Low"

Silicate"

Deep,"Very"Cold,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Low"Silicate"

Arctic"and"North"Atlantic"

Bathypelagic"(A_NA_bathy)"

30"

Epipelagic,"Very"Cold,"Euhaline,"Oxic,"

Medium"Nitrate,"Low"Phosphate,"Low"

Silicate"

Shallow,"Very"Cold,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Low"Silicate"

North"Pacific"and"Beaufort"

Sea"Epipelagic"(NP_BE_epi)"

31"

Epipelagic,"Frozen/Superchilled,"

Euhaline,"Oxic,"Medium"Nitrate,"Low"

Phosphate,"Medium"Silicate"

Shallow,"Superchilled,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Medium"Silicate"

Antarctic"and"Bering"Sea"

Epipelagic"(AA_BR_epi)"

32"

Epipelagic,"Warm"to"Very"Warm,"

Mesohaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Warm,"Very"Low"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Caspian"Sea"Epipelagic"

(C_epi)"

33"

Mesopelagic,"Very"Cold,"Euhaline,"

Severely"Hypoxic,"High"Nitrate,"Medium"

Phosphate,"Medium"Silicate"

Moderate"Depth,"Very"Cold,"Normal"

Salinity,"Very"Low"Oxygen,"High"Nitrate,"

Medium"Phosphate,"Medium"Silicate"

Tropical"Pacific"and"Tropical"

Indian"Mesopelagic"

(TP_TI_meso)"

34"

Epipelagic,"Very"Cold,"Polyhaline,"

Anoxic,"Low"Nitrate,"High"Phosphate,"

High"Silicate"

Shallow,"Very"Cold,"Low"Salinity,"No"

Oxygen,"Low"Nitrate,"High"Phosphate,"

High"Silicate"

Black"Sea"Epipelagic"

(BL_epi)"

35"

Epipelagic,"Frozen/Superchilled,"

Euhaline,"Oxic,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Shallow,"Superchilled,"Normal"Salinity,"

Moderate"Oxygen,"Low"Nitrate,"Low"

Phosphate,"Low"Silicate"

Arctic"and"Labrador"Sea"

Epipelagic"(A_L_epi)"

36"

Bathypelagic,"Very"Cold,"Euhaline,"Oxic,"

Medium"Nitrate,"Low"Phosphate,"Low"

Silicate"

Deep,"Very"Cold,"Normal"Salinity,"

Moderate"Oxygen,"Medium"Nitrate,"

Low"Phosphate,"Low"Silicate"

Atlantic,"Subantarctic,"and"

North"Pacific"Subtropical"

Bathypelagic"

(AT_SAA_NPS_bathy)"

37"

Bathypelagic,"Very"Cold,"Euhaline,"Oxic,"

High"Nitrate,"Low"Phosphate,"Medium"

Silicate"

Deep,"Very"Cold,"Normal"Salinity,"

Moderate"Oxygen,"High"Nitrate,"Low"

Phosphate,"Medium"Silicate"

Subantarctic,"South"Atlantic"

and"North"Pacific"

Bathypelagic"

(AA_SAT_NP_bathy)"

Appendix 2. Summary descriptions and maps of the 37 EMUs.

Ecological Marine Unit 1 Summary

Technical Name

Mesopelagic, Cold, Polyhaline, Severely Hypoxic, Low Nitrate, Medium Phosphate, High Silicate

Common Name

Moderate Depth, Cold, Low Salinity, Very Low Oxygen, Low Nitrate, Medium Phosphate, High Silicate

EMU Volumetric Region Name and Code

Black and Caspian Seas Mesopelagic (BL_C_meso)

EMU 1 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-0.65

8.3

14.72

1.62

Salinity (unitless)

20.19

21.42

22.32

0.34

Dissolved Oxygen (µmol/l)

0.2

0.83

0.15

Nitrate (µmol/l)

0.83

4.38

0.74

Phosphate (µmol/l)

2.47

4.27

5.22

0.56

Silicate (µmol/l)

58.88

109.61

196.55

25.15

EMU Water Volume (Km3)

70,488.07

Percent of EMU to Global

0.01%

Unit Middle Median (m)

220.87

Thickness Mean (m)

148.13

Ecological Marine Unit 2 Summary

Technical Name

Epipelagic, Cold, Polyhaline, Hypoxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cold, Low Salinity, Low Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Black and Caspian Seas Epipelagic (BL_C_epi)

EMU 2 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-0.44

7.27

9.56

2.04

Salinity (unitless)

18.43

19.82

21.61

0.71

Dissolved Oxygen (µmol/l)

0.52

2.68

4.67

1.13

Nitrate (µmol/l)

2.85

4.22

1.04

Phosphate (µmol/l)

0.18

0.82

2.5

0.35

Silicate (µmol/l)

4.28

33.65

64.21

13.16

EMU Water Volume (Km3)

29,302.16

Percent of EMU to Global

0.00%

Unit Middle Median (m)

85.28

Thickness Mean (m)

52.21

Ecological Marine Unit 3 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Severely Hypoxic, High Nitrate, Medium Phosphate, High Silicate

Common Name

Deep, Very Cold, Normal Salinity, Very Low Oxygen, High Nitrate, Medium Phosphate, High Silicate

EMU Volumetric Region Name and Code

North Pacific and Arabian Sea Bathypelagic (NP_AS_bathy)

EMU 3 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

0.47

2.6

5.16

0.62

Salinity (unitless)

33.52

34.54

35.03

0.12

Dissolved Oxygen (µmol/l)

0.21

1.47

2.98

0.62

Nitrate (µmol/l)

26.67

42.34

56.12

1.82

Phosphate (µmol/l)

2.27

3.03

4.1

0.15

Silicate (µmol/l)

106.22

149.53

247.19

19.52

EMU Water Volume (Km3)

96,803,913.19

Percent of EMU to Global

7.09%

Unit Middle Median (m)

1,598.10

Thickness Mean (m)

1,253.52

Ecological Marine Unit 4 Summary

Technical Name

Mesopelagic, Cold, Polyhaline, Severely Hypoxic, Low Nitrate, High Phosphate, High Silicate

Common Name

Moderate Depth, Cold, Low Salinity, Very Low Oxygen, Low Nitrate, High Phosphate, High Silicate

EMU Volumetric Region Name and Code

Black Sea Mesopelagic (BL_meso)

EMU 4 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

4.62

8.89

14.93

0.15

Salinity (unitless)

21.72

22.03

22.4

0.16

Dissolved Oxygen (µmol/l)

0.1

0.7

0.11

Nitrate (µmol/l) 0 0.11 1.45 0.26

Phosphate (µmol/l)

5.13

6.18

7.27

0.54

Silicate (µmol/l)

111

160.29

231.35

30.75

EMU Water Volume (Km3)

137,671.34

Percent of EMU to Global

0.01%

Unit Middle Median (m)

498.16

Thickness Mean (m)

445.63

Ecological Marine Unit 5 Summary

Technical Name

Epipelagic, Frozen/Superchilled, Polyhaline, Highly Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Superchilled, Low Salinity, High Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Arctic Epipelagic (A_epi)

EMU 5 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.62

-0.96

12.1

1.05

Salinity (unitless)

25.99

28.59

29.63

0.78

Dissolved Oxygen (µmol/l)

5.03

8.71

9.89

0.48

Nitrate (µmol/l)

1.16

15.24

1.08

Phosphate (µmol/l)

0.79

2.24

0.24

Silicate (µmol/l)

0.31

9.36

32.87

5.37

EMU Water Volume (Km3)

79,011.98

Percent of EMU to Global

0.01%

Unit Middle Median (m)

12.32

Thickness Mea n (m)

18.47

Ecological Marine Unit 6 Summary

Technical Name

Epipelagic, Cold, Polyhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cold, Low Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Black and Caspian Seas Epipelagic (BL_C_epi)

EMU 6 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-0.38

9.24

19.38

4.06

Salinity (unitless)

16.22

18.35

20.65

0.45

Dissolved Oxygen (µmol/l)

4.13

6.65

9.57

1.05

Nitrate (µmol/l)

0.05

1.13

5.8

0.79

Phosphate (µmol/l)

0.1

0.28

1.11

0.15

Silicate (µmol/l)

4.04

13.86

56.85

6.79

EMU Water Volume (Km3)

28,710.79

Percent of EMU to Global

0.00%

Unit Middle Median (m)

30.45

Thickness Mean (m)

26.94

Ecological Marine Unit 7 Summary

Technical Name

Epiplagic, Moderate to Cool, Mesohaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cool, Very Low Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Caspian Sea Epipelagic (C_epi)

EMU 7 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

4.81

15.62

20.91

3.05

Salinity (unitless)

7.81

9.4

10.96

0.86

Dissolved Oxygen (µmol/l)

5.05

6.8

7.93

0.44

Nitrate (µmol/l)

0.02

0.25

2.78

0.45

Phosphate (µmol/l)

0.05

0.17

0.7

0.11

Silicate (µmol/l)

3.38

6.64

29.29

5.22

EMU Water Volume (Km3)

1,368.36

Percent of EMU to Global

0.00%

Unit Middle Median (m)

13.85

Thickness Mean (m)

18.74

Ecological Marine Unit 8 Summary

Technical Name

Epipelagic, Moderate to Cool, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cool, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Subantarctic, North Atlantic, and North Pacific Epipelagic (SAA_NA_NP_epi)

EMU 8 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

3.39

11.05

18.44

1.96

Salinity (unitless)

32.57

34.59

35.65

0.45

Dissolved Oxygen (µmol/l)

3.75

5.73

7.29

0.61

Nitrate (µmol/l) 0.01 10.62 22.7 4.14

Phosphate (µmol/l)

0.11

0.84

1.83

0.23

Silicate (µmol/l)

0.41

5.96

34.17

4.18

EMU Water Volume (Km3)

26,219,376.97

Percent of EMU to Global

1.92%

Unit Middle Median (m)

151.24

Thickness Mean (m)

219.21

Ecological Marine Unit 9 Summary

Technical Name

Mesopelagic, Moderate to Cool, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Moderate Depth, Cool, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Mediterranean and Red Seas Mesopelagic (M_R_meso)

EMU 9 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

12.52

15.28

29.53

2.86

Salinity (unitless)

37.34

38.62

40.77

0.41

Dissolved Oxygen (µmol/l)

2.04

4.69

5.98

0.53

Nitrate (µmol/l) 0 3.75 15.02 2.84

Phosphate (µmol/l)

0.02

0.19

0.77

0.11

Silicate (µmol/l)

0.31

4.76

26.46

2.78

EMU Water Volume (Km3)

3,998,804.78

Percent of EMU to Global

0.29%

Unit Middle Median (m)

301.77

Thickness Mean (m)

1,481.43

Ecological Marine Unit 10 Summary

Technical Name

Mesopelagic, Cold, Euhaline, Severely Hypoxic, High Nitrate, Low Phosphate, Low Silicate

Common Name

Moderate Depth, Cold, Normal Salinity, Very Low Oxygen, High Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Equatorial Indian, Tropical Atlantic, and Tropical Pacific Mesopelagic (EI_TA_TP_meso)

EMU 10 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

4.66

9.83

24.2

2.26

Salinity (unitless)

33.63

34.78

36.94

0.3

Dissolved Oxygen (µmol/l)

0.03

1.56

3.55

0.87

Nitrate (µmol/l)

9.87

30.84

43.71

4.2

Phosphate (µmol/l)

1.26

2.28

3.36

0.31

Silicate (µmol/l)

6.75

31.96

96.02

12.38

EMU Water Volume (Km3)

45,669,990.26

Percent of EMU to Global

3.34%

Unit Middle Median (m)

390.89

Thickness Mean (m)

339.64

Ecological Marine Unit 11 Summary

Technical Name

Epipelagic, Moderate to Cool, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cool, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Northern Subtropical and Southern Subtropical Epipelagic (NS_SS_epi)

EMU 11 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

10.07

16.46

21.67

2.26

Salinity (unitless)

33.93

35.33

36.46

0.41

Dissolved Oxygen (µmol/l)

3.25

5.23

6.46

0.46

Nitrate (µmol/l)

0.01

3.66

19.98

3.27

Phosphate (µmol/l)

0.05

0.39

1.4

0.19

Silicate (µmol/l)

0.27

3.33

20.34

1.86

EMU Water Volume (Km3)

22,994,027.00

Percent of EMU to Global

1.68%

Unit Middle Median (m)

101.22

Thickness Mean (m)

172.49

Ecological Marine Unit 12 Summary

Technical Name

Epipelagic, Very Cold, Mesohaline, Severely Hypoxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Very Cold, Very Low Salinity, Very Low Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Baltic Sea Epipelagic (BA_epi)

EMU 12 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

0.72

4.73

14.05

1.36

Salinity (unitless)

8.98

10.22

12.03

0.64

Dissolved Oxygen (µmol/l)

0.29

1.62

4.36

0.64

Nitrate (µmol/l)

3.15

5.85

7.72

0.82

Phosphate (µmol/l)

1.2

2.39

3.85

0.39

Silicate (µmol/l)

28.7

43.6

60.68

4.95

EMU Water Volume (Km3)

3,865.12

Percent of EMU to Global

0.00%

Unit Middle Median (m)

88.50

Thickness Mean (m)

33.47

Ecological Marine Unit 13 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Hypoxic, High Nitrate, Medium Phosphate, High Silicate

Common Name

Deep, Very Cold, Normal Salinity, Low Oxygen, High Nitrate, Medium Phosphate, High Silicate

EMU Volumetric Region Name and Code

Pacific and Indian Bathypelagic (P_I_bathy)

EMU 13 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-0.38

1.93

5.54

0.51

Salinity (unitless)

33.43

34.67

34.93

0.05

Dissolved Oxygen (µmol/l)

1.69

3.26

4.33

0.43

Nitrate (µmol/l)

25.26

37.03

48.49

1.57

Phosphate (µmol/l)

0.53

2.6

3.36

0.12

Silicate (µmol/l)

88.01

138.03

189.63

19.05

EMU Water Volume (Km3)

347,060,603.65

Percent of EMU to Global

25.40%

Unit Middle Median (m)

2,871.85

Thickness Mean (m)

2,010.79

Ecological Marine Unit 14 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Oxic, High Nitrate, Low Phosphate, High Silicate

Common Name

Deep, Very Cold, Normal Salinity, Moderate Oxygen, High Nitrate, Low Phosphate, High Silicate

EMU Volumetric Region Name and Code

Antarctic and Subantarctic Bathypelagic (AA_SAA_bathy)

EMU 14 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-2.05

0.88

3.31

0.69

Salinity (unitless)

33.27

34.7

34.9

0.04

Dissolved Oxygen (µmol/l)

2.96

4.74

6.86

0.41

Nitrate (µmol/l)

16.96

32.71

40.33

1.21

Phosphate (µmol/l)

1.49

2.27

3.02

0.09

Silicate (µmol/l)

72.25

115.2

167.23

13.89

EMU Water Volume (Km3)

277,552,060.90

Percent of EMU to Global

20.31%

Unit Middle Median (m)

2,595.55

Thickness Mean (m)

2,096.30

Ecological Marine Unit 15 Summary

Technical Name

Bathypelagic, Cold, Polyhaline, Anoxic, Low Nitrate, High Phosphate, High Silicate

Common Name

Deep, Cold, Low Salinity, No Oxygen, Low Nitrate, High Phosphate, High Silicate

EMU Volumetric Region Name and Code

Black Sea Bathypelagic (BL_bathy)

EMU 15 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

7.18

15.16

0.17

Salinity (unitless)

21.98

22.29

23.44

0.06

Dissolved Oxygen (µmol/l)

0.07

0.01

Nitrate (µmol/l)

0.03

1.45

0.2

Phosphate (µmol/l)

6.84

8.02

9.76

0.55

Silicate (µmol/l)

91.49

192.79

258.23

25.59

EMU Water Volume (Km3)

277,932.33

Percent of EMU to Global

0.02%

Unit Middle Median (m)

1,297.92

Thickness Mean (m)

962.14

Ecological Marine Unit 16 Summary

Technical Name

Epipelagic, Very Cold, Mesohaline, Highly Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Very Cold, Very Low Salinity, High Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Baltic Sea Epipelagic (BA_epi)

EMU 16 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

1.76

4.22

11.99

1.73

Salinity (unitless)

5.01

5.61

6.49

0.45

Dissolved Oxygen (µmol/l)

6.43

8.41

8.95

0.49

Nitrate (µmol/l) 0.07 3.35 7.51 1.85

Phosphate (µmol/l)

0.05

0.18

0.67

0.12

Silicate (µmol/l)

9.06

18.46

29.67

6.17

EMU Water V olume (Km3)

7,665.87

Percent of EMU to Global

0.00%

Unit Middle Median (m)

25.20

Thickness Mean (m)

49.24

Ecological Marine Unit 17 Summary

Technical Name

Epipelagic, Cold, Mesohaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cold, Very Low Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Baltic Sea Epipelagic (BA_epi)

EMU 17 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

1.34

5.84

14.24

2.06

Salinity (unitless)

6.33

7.32

8.91

0.42

Dissolved Oxygen (µmol/l)

6.11

7.94

8.51

0.37

Nitrate (µmol/l)

0.12

1.97

5.68

0.72

Phosphate (µmol/l)

0.14

0.42

1.01

0.14

Silicate (µmol/l)

2.01

12.13

29.9

2.61

EMU Water Volume (Km3)

10,110.28

Percent of EMU to Global

0.00%

Unit Middle Median (m)

25.50

Thickness Mean (m)

40.59

Ecological Marine Unit 18 Summary

Technical Name

Epipelagic, Warm to Very Warm, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Warm, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

North Pacific Subtropical and Equatorial Indian Epipelagic (NPS_EI_epi)

EMU 18 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

16.42

26.15

30.33

2.64

Salinity (unitless)

31.89

34.43

35.01

0.43

Dissolved Oxygen (µmol/l)

1.5

4.53

5.83

0.41

Nitrate (µmol/l)

1.39

21.45

2.17

Phosphate (µmol/l)

0.03

0.27

1.37

0.18

Silicate (µmol/l)

0.38

3.73

28.85

2.52

EMU Water Volume (Km3)

6,075,341.37

Percent of EMU to Global

0.44%

Unit Middle Median (m)

39.45

Thickness Mean (m)

75.65

Ecological Marine Unit 19 Summary

Technical Name

Epipelagic, Cold, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cold, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Subantarctic and North Pacific Subtropical Epipelagic (SAA_NPS_epi)

EMU 19 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.52

5.5

13.74

1.85

Salinity (unitless)

33.11

34.09

34.7

0.2

Dissolved Oxygen (µmol/l)

4.21

6.64

7.74

0.51

Nitrate (µmol/l)

9.3

20.64

32.16

3.51

Phosphate (µmol/l)

0.49

1.49

2.2

0.2

Silicate (µmol/l)

10.46

54.73

6.04

EMU Water Volume (Km3)

14,030,918.23

Percent of EMU to Global

1.03%

Unit Middle Median (m)

114.29

Thickness Mean (m)

257.08

Ecological Marine Unit 20 Summary

Technical Name

Epipelagic, Very Cold, Mesohaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Very Cold, Very Low Salinity, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Baltic Sea Epipelagic (BA_epi)

EMU 20 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

1.11

4.04

8.88

1.23

Salinity (unitless)

7.47

8.79

0.54

Dissolved Oxygen (µmol/l)

2.69

4.78

7.2

1.06

Nitrate (µmol/l)

2.33

4.54

6.03

0.89

Phosphate (µmol/l)

0.54

1.3

2.11

0.37

Silicate (µmol/l)

14.59

26.88

40.52

5.5

EMU Water Volume (Km3)

1,747.60

Percent of EMU to Global

0.00%

Unit Middle Median (m)

65.63

Thickness Mean (m)

12.41

Ecological Marine Unit 21 Summary

Technical Name

Epipelagic, Warm to Very Warm, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Warm, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Atlantic Subtropical and South Pacific Subtropical Epipelagic (ATS_SPS_epi)

EMU 21 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

12.67

22.53

29.73

Salinity (unitless)

35.77

36.48

38.13

0.36

Dissolved Oxygen (µmol/l)

2.43

4.8

5.88

0.4

Nitrate (µmol/l)

0.01

1.56

16.5

2.34

Phosphate (µmol/l)

0.01

0.2

1.4

0.15

Silicate (µmol/l)

0.39

1.72

18.82

1.14

EMU Water Volume (Km3)

10,810,147.83

Percent of EMU to Global

0.79%

Unit Middle Median (m)

63.96

Thickness Mean (m)

182.66

Ecological Marine Unit 22 Summary

Technical Name

Epipelagic, Cold, Mesohaline, Hypoxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cold, Very Low Salinity, Low Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Baltic and Black Seas Epipelagic (BA_BL_epi)

EMU 22 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

0.07

8.15

15.55

4.89

Salinity (unitless)

13.37

15.87

17.01

0.84

Dissolved Oxygen (µmol/l)

2.32

2.98

3.79

0.55

Nitrate (µmol/l)

6.13

6.62

6.97

0.35

Phosphate (µmol/l)

1.3

1.6

1.86

0.21

Silicate (µmol/l)

31.11

37.82

43.72

4.62

EMU Water Volume (Km3)

688.12

Percent of EMU to Global

0.00%

Unit Middle Median (m)

7.67

Thickness Mean (m)

10.43

Ecological Marine Unit 23 Summary

Technical Name

Epipelagic, Frozen/Superchilled, Euhaline, Highly Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Superchilled, Normal Salinity, High Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Arctic and Labrador Sea Epipelagic (A_L_epi)

EMU 23 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.88

-1.13

14.03

1.31

Salinity (unitless)

30.92

32.22

33.24

0.45

Dissolved Oxygen (µmol/l)

4.5

8.02

10.38

0.51

Nitrate (µmol/l) 0 4.65 17.03 2.63

Phosphate (µmol/l)

0.05

0.96

2.49

0.31

Silicate (µmol/l)

0.61

11.44

45.16

6.19

EMU Water Volume (Km3)

392,386.99

Percent of EMU to Global

0.03%

Unit Middle Median (m)

40.18

Thickness Mean (m)

33.21

Ecological Marine Unit 24 Summary

Technical Name

Epipelagic, Warm to Very Warm, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Warm, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Tropical Pacific, Tropical Indian, and Equatorial Atlantic Epipelagic (TP_TI_EA_epi)

EMU 24 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

18.78

24.77

29.54

2.52

Salinity (unitless)

34.73

35.39

36.26

0.3

Dissolved Oxygen (µmol/l)

1.76

4.58

5.51

0.43

Nitrate (µmol/l)

2.05

15.01

2.67

Phosphate (µmol/l)

0.01

0.31

1.4

0.23

Silicate (µmol/l)

0.25

2.95

17.58

1.94

EMU Water Volume (Km3)

11,637,344.73

Percent of EMU to Global

0.85%

Unit Middle Median (m)

50.00

Thickness Mean (m)

94.99

Ecological Marine Unit 25 Summary

Technical Name

Epipelagic, Frozen/Superchilled, Euhaline, Highly Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Superchilled, Normal Salinity, High Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Arctic and Labrador Sea Epipelagic (A_L_epi)

EMU 25 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.76

-1.27

13.84

1.11

Salinity (unitless)

29.54

30.7

31.64

0.48

Dissolved Oxygen (µmol/l)

5.08

8.65

10.39

0.43

Nitrate (µmol/l) 0.03 1.91 16.5 1.35

Phosphate (µmol/l)

0.01

0.88

2.21

0.23

Silicate (µmol/l)

0.61

9.56

37.57

3.75

EMU Water Volume (Km3)

221,753.31

Percent of EMU to Global

0.02%

Unit Middle Median (m)

25.28

Thickness Mean (m)

29.37

Ecological Marine Unit 26 Summary

Technical Name

Mesopelagic, Moderate to Cool, Euhaline, Hypoxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Moderate Depth, Cool, Normal Salinity, Low Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Tropical and Subtropical Mesopelagic (T_ST_meso)

EMU 26 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

6.8

14.27

26.44

3.62

Salinity (unitless)

33.59

35.14

36.91

0.46

Dissolved Oxygen (µmol/l)

0.47

3.21

4.99

0.75

Nitrate (µmol/l)

1.8

17.48

32.24

4.37

Phosphate (µmol/l)

0.47

1.27

2.43

0.28

Silicate (µmol/l)

1.61

12.37

60.31

6.41

EMU Water Volume (Km3)

26,333,737.98

Percent of EMU to Global

1.93%

Unit Middle Median (m)

242.26

Thickness Mean (m)

159.79

Ecological Marine Unit 27 Summary

Technical Name

Epipelagic, Very Cold, Polyhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Very Cold, Low Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Caspian Sea Epipelagic (C_epi)

EMU 27 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.26

2.95

19.23

4.03

Salinity (unitless)

20.42

22.6

25.32

1.29

Dissolved Oxygen (µmol/l)

4.63

7.59

9.59

1.38

Nitrate (µmol/l)

0.92

6.53

1.19

Phosphate (µmol/l)

0.06

0.46

1.52

0.32

Silicate (µmol/l)

1.62

15.62

49.03

9.33

EMU Water Volume (Km3)

21,833.01

Percent of EMU to Global

0.00%

Unit Middle Median (m)

26.81

Thickness Mean (m)

15.74

Ecological Marine Unit 28 Summary

Technical Name

Epipelagic, Moderate to Cool, Mesohaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Cool, Very Low Salinity, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Caspian Sea Epipelagic (C_epi)

EMU 28 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

0.19

12.67

19.84

3.43

Salinity (unitless)

10.96

12.76

14.27

0.36

Dissolved Oxygen (µmol/l)

3.79

6.72

9.34

0.6

Nitrate (µmol/l)

0.19

6.13

0.69

Phosphate (µmol/l)

0.06

0.2

1.3

0.1

Silicate (µmol/l)

2.1

6.49

56.85

7.02

EMU Water Volume (Km3)

14,208.45

Percent of EMU to Global

0.00%

Unit Middle Median (m)

19.60

Thickness Mean (m)

33.56

Ecological Marine Unit 29 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Deep, Very Cold, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Arctic and North Atlantic Bathypelagic (A_NA_bathy)

EMU 29 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-2

0.69

7.4

1.72

Salinity (unitless)

33.93

34.89

35.38

0.12

Dissolved Oxygen (µmol/l)

4.83

6.71

8.31

0.3

Nitrate (µmol/l)

1.69

13.83

22.58

2.17

Phosphate (µmol/l)

0.1

0.96

1.6

0.13

Silicate (µmol/l)

0.29

9.87

41.33

3.28

EMU Water Volume (Km3)

39,402,422.32

Percent of EMU to Global

2.88%

Unit Middle Median (m)

1,127.70

Thickness Mean (m)

2,168.50

Ecological Marine Unit 30 Summary

Technical Name

Epipelagic, Very Cold, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Very Cold, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

North Pacific and Beaufort Sea Epipelagic (NP_BE_epi)

EMU 30 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.79

3.88

18.56

4.28

Salinity (unitless)

30.95

32.82

33.78

0.33

Dissolved Oxygen (µmol/l)

4.2

6.76

8.31

0.5

Nitrate (µmol/l)

13.11

29.71

5.02

Phosphate (µmol/l)

0.25

1.4

2.7

0.36

Silicate (µmol/l)

0.61

78.33

10.61

EMU Water Volume (Km3)

1,688,982.45

Percent of EMU to Global

0.12%

Unit Middle Median (m)

66.47

Thickness Mean (m)

75.72

Ecological Marine Unit 31 Summary

Technical Name

Epipelagic, Frozen/Superchilled, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Medium Silicate

Common Name

Shallow, Superchilled, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Medium

Silicate

EMU Volumetric Region Name and Code

Antarctic and Bering Sea Epipelagic (AA_BR_epi)

EMU 31 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-2.05

-0.27

5.81

1.39

Salinity (unitless)

32.12

34.05

34.92

0.27

Dissolved Oxygen (µmol/l)

4.98

7.3

8.98

0.52

Nitrate (µmol/l)

12.71

27.53

39.64

2.21

Phosphate (µmol/l)

0.99

1.94

3.08

0.15

Silicate (µmol/l) 8.21 50.42 103.82 18.01

EMU Water Volume (Km3)

7,868,077.67

Percent of EMU to Global

0.58%

Unit Middle Median (m)

65.30

Thickness Mean (m)

170.49

Ecological Marine Unit 32 Summary

Technical Name

Epipelagic, Warm to Very Warm, Mesohaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Warm, Very Low Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Caspian Sea Epipelagic (C_epi)

EMU 32 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

12.46

21.59

26.49

2.45

Salinity (unitless)

5.18

6.66

7.9

0.58

Dissolved Oxygen (µmol/l)

4.94

6.79

7.62

0.56

Nitrate (µmol/l)

0.02

0.15

0.45

0.09

Phosphate (µmol/l)

0.05

0.11

0.3

0.05

Silicate (µmol/l)

2.13

4.67

13.41

1.37

EMU Water Volume (Km3)

1,512.82

Percent of EMU to Global

0.00%

Unit Middle Median (m)

15.56

Thickness Mean (m)

27.92

Ecological Marine Unit 33 Summary

Technical Name

Mesopelagic, Very Cold, Euhaline, Severely Hypoxic, High Nitrate, Medium Phosphate, Medium Silicate

Common Name

Moderate Depth, Very Cold, Normal Salinity, Very Low Oxygen, High Nitrate, Medium Phosphate,

Medium Silicate

EMU Volumetric Region Name and Code

Tropical Pacific and Tropical Indian Mesopelagic (TP_TI_meso)

EMU 33 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

0.74

4.99

10.44

1.33

Salinity (unitless)

33.39

34.53

35.69

0.26

Dissolved Oxygen (µmol/l)

0.04

1.65

3.38

0.71

Nitrate (µmol/l)

39.25

49.67

2.77

Phosphate (µmol/l)

1.84

2.87

3.88

0.2

Silicate (µmol/l)

25.97

87.61

127.58

18.54

EMU Water Volume (Km3)

87,570,377.64

Percent of EMU to Global

6.41%

Unit Middle Median (m)

935.40

Thickness Mean (m)

603.19

Ecological Marine Unit 34 Summary

Technical Name

Epipelagic, Very Cold, Polyhaline, Anoxic, Low Nitrate, High Phosphate, High Silicate

Common Name

Shallow, Very Cold, Low Salinity, No Oxygen, Low Nitrate, High Phosphate, High Silicate

EMU Volumetric Region Name and Code

Black Sea Epipelagic (BL_epi)

EMU 34 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.1

4.79

14.49

4.24

Salinity (unitless)

22.32

23.58

25.56

1.2

Dissolved Oxygen (µmol/l)

Nitrate (µmol/l) 0 0 0 0

Phosphate (µmol/l)

9.76

10.29

11.62

0.4

Silicate (µmol/l)

158.85

194.1

218.22

14.06

EMU Water Volume (Km3)

24,928.19

Percent of EMU to Global

0.00%

Unit Middle Median (m)

157.95

Thickness Mean (m)

52.39

Ecological Marine Unit 35 Summary

Technical Name

Epipelagic, Frozen/Superchilled, Euhaline, Oxic, Low Nitrate, Low Phosphate, Low Silicate

Common Name

Shallow, Superchilled, Normal Salinity, Moderate Oxygen, Low Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Arctic and Labrador Sea Epipelagic (A_L_epi)

EMU 35 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-2.04

-0.9

10.37

1.4

Salinity (unitless)

32.88

33.89

34.88

0.43

Dissolved Oxygen (µmol/l)

5.08

7.44

9.17

0.53

Nitrate (µmol/l)

0.22

6.88

24.6

2.45

Phosphate (µmol/l)

0.09

0.74

1.49

0.23

Silicate (µmol/l)

0.44

8.03

74.12

4.71

EMU Water Volume (Km3)

889,680.67

Percent of EMU to Global

0.07%

Unit Middle Median (m)

60.08

Thickness Mean (m)

80.45

Ecological Marine Unit 36 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Oxic, Medium Nitrate, Low Phosphate, Low Silicate

Common Name

Deep, Very Cold, Normal Salinity, Moderate Oxygen, Medium Nitrate, Low Phosphate, Low Silicate

EMU Volumetric Region Name and Code

Atlantic, Subantarctic, and North Pacific Subtropical Bathypelagic (AT_SAA_NPS_bathy)

EMU 36 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-0.99

4.58

12.05

2.12

Salinity (unitless)

33.23

34.76

35.86

0.3

Dissolved Oxygen (µmol/l)

3.21

5.27

6.36

0.53

Nitrate (µmol/l)

12.16

22.74

35.69

Phosphate (µmol/l)

0.73

1.53

2.23

0.22

Silicate (µmol/l)

27.13

80.48

13.82

EMU Water Volume (Km3)

185,999,904.33

Percent of EMU to Global

13.61%

Unit Middle Median (m)

1,580.53

Thickness Mean (m)

957.82

Ecological Marine Unit 37 Summary

Technical Name

Bathypelagic, Very Cold, Euhaline, Oxic, High Nitrate, Low Phosphate, Medium Silicate

Common Name

Deep, Very Cold, Normal Salinity, Moderate Oxygen, High Nitrate, Low Phosphate, Medium Silicate

EMU Volumetric Region Name and Code

Subantarctic, South Atlantic and North Pacific Bathypelagic (AA_SAT_NP_bathy)

EMU 37 Summary Statistics

Minimum

Mean

Max

Standard Dev.

Temperature (°C)

-1.97

3.13

10.51

1.23

Salinity (unitless)

32.67

34.52

35.12

0.2

Dissolved Oxygen (µmol/l)

2.39

4.25

6.02

0.52

Nitrate (µmol/l) 10.54 32.7 41.95 2.29

Phosphate (µmol/l)

1.45

2.26

3.07

0.17

Silicate (µmol/l)

62.6

101.91

18.9

EMU Water Volume (Km3)

152,103,923.07

Percent of EMU to Global

11.13%

Unit Middle Median (m)

1,139.29

Thickness Mean (m)

779.85

BenthicNet: A global compilation of seafloor images for deep learning applications

Article

Full-text available

Feb 2025

Advances in underwater imaging enable collection of extensive seafloor image datasets necessary for monitoring important benthic ecosystems. The ability to collect seafloor imagery has outpaced our capacity to analyze it, hindering mobilization of this crucial environmental information. Machine learning approaches provide opportunities to increase the efficiency with which seafloor imagery is analyzed, yet large and consistent datasets to support development of such approaches are scarce. Here we present BenthicNet: a global compilation of seafloor imagery designed to support the training and evaluation of large-scale image recognition models. An initial set of over 11.4 million images was collected and curated to represent a diversity of seafloor environments using a representative subset of 1.3 million images. These are accompanied by 3.1 million annotations translated to the CATAMI scheme, which span 190,000 of the images. A large deep learning model was trained on this compilation and preliminary results suggest it has utility for automating large and small-scale image analysis tasks. The compilation and model are made openly available for reuse.

A study of benthic habitat classification schemes in the northern equatorial Pacific based on a Gaussian mixture model

Article

Full-text available

Jan 2025
J ENVIRON MANAGE

Due to the complexity and extreme environment of the deep sea, related research is limited by the level of detection technology, and the overall understanding of this topic is limited. At present, the increasing number of disturbances in the deep sea, such as those caused by deep-sea mining, necessitates the urgent formulation of scientifically informed deep-sea environmental management decisions. In response to this challenge, a data-driven, large-scale, and probability-based benthic habitat classification scheme is proposed for the northern equatorial Pacific. The dominant environmental factors were identified as topographic features (depth and slope), nutrients (particulate organic carbon flux entering the seabed and phosphate), and water mass properties (dissolved oxygen and pH), and the importance assessment indicated that nutrient played a dominant role, followed by water mass properties, with topographic features having a relatively minor influence. The multiple classification evaluation parameters determined 5 clusters of benthic habitats in the northern equatorial Pacific using the Gaussian mixture model (GMM). Furthermore, this paper provides the first discussion of ecotones at the edges of habitats and suggests that ecotones are more cost-effective for environmental management. In conclusion, the benthic habitat classification scheme could provide a consistent unit for habitat mapping and representative assessment of marine protected area network design, as well as a scientific basis for the protection and management of the international seabed area.

Biodiversity Informatics: Building a Lifeboat for High Functionality Data to Decision Pipeline

Book

Jan 2024

The uplift rate of Sulawesi East Arm and the activity of Batui Fault using a tectonic geomorphology approach in the Luwuk Area, Sulawesi

Article

Full-text available

Feb 2025

Sulawesi is a tectonically active area with convergent triple-junction amidst Eurasian, Indo-Australian and Pacific Plate. The study area is located in the suture zone of sedimentary Banggai-Sula Microcontinent and East Sulawesi Ophiolite. This research is aimed to determine fault activity and rate of tectonic uplift of Luwuk mountainous area. Field geological mapping had been done to determine lithology and geological structure along with morphometric analysis to measure definitive result of tectonic activities. Geological structure features which encountered in this area are Batui Thrust, Pasini Thrust, Lobu Balongan Fault Zone, and Lambangan Anticline. The sedimentary rocks have average N315oE dip direction with angle of 41o. Based on tectonic geomorphology and morphometric analyses are resulting the most active tectonic areas passed by the Batui and Pasini Thrust which have NE-SW direction and Lobu Balolang Fault Zone relative to others. The Batui Thrust is considered as active fault intersected the East Sulawesi Ophiolite and Banggai-Sula Microcontinent which segmented with wrench faults. Number of thrust faults parallel to the Batui Thrust generate Batui Thrust Belt which propagate from the ultramafic rocks to the sedimentary rocks within the suture zone. As a result, the presence of active fault is uplifting the East Arm of Sulawesi with 0.408 ± 0.008 mm and 0.213 ± 0.046 mm uplift rate per year by the presence of high altitude Celebes Molasses. These rates are getting higher to the west. Keywords: Batui Thrust, East Sulawesi Ophiolite, Luwuk, morphometry, tectonic geomorphology, uplift rate

Future directions for deep ocean climate science and evidence-based decision making

Article

Full-text available

Oct 2024

Introduction A defining aspect of the Intergovernmental Panel on Climate Change (IPCC) assessment reports (AR) is a formal uncertainty language framework that emphasizes higher certainty issues across the reports, especially in the executive summaries and short summaries for policymakers. As a result, potentially significant risks involving understudied components of the climate system are shielded from view. Methods Here we seek to address this in the latest, sixth assessment report (AR6) for one such component—the deep ocean—by summarizing major uncertainties (based on discussions of low confidence issues or gaps) regarding its role in our changing climate system. The goal is to identify key research priorities to improve IPCC confidence levels in deep ocean systems and facilitate the dissemination of IPCC results regarding potentially high impact deep ocean processes to decision-makers. This will accelerate improvement of global climate projections and aid in informing efforts to mitigate climate change impacts. An analysis of 3,000 pages across the six selected AR6 reports revealed 219 major science gaps related to the deep ocean. These were categorized by climate stressor and nature of impacts. Results Half of these are biological science gaps, primarily surrounding our understanding of changes in ocean ecosystems, fisheries, and primary productivity. The remaining science gaps are related to uncertainties in the physical (32%) and biogeochemical (15%) ocean states and processes. Model deficiencies are the leading cited cause of low certainty in the physical ocean and ice states, whereas causes of biological uncertainties are most often attributed to limited studies and observations or conflicting results. Discussion Key areas for coordinated effort within the deep ocean observing and modeling community have emerged, which will improve confidence in the deep ocean state and its ongoing changes for the next assessment report. This list of key “known unknowns” includes meridional overturning circulation, ocean deoxygenation and acidification, primary production, food supply and the ocean carbon cycle, climate change impacts on ocean ecosystems and fisheries, and ocean-based climate interventions. From these findings, we offer recommendations for AR7 to avoid omitting low confidence-high risk changes in the climate system.

Future Directions for Deep Ocean Climate Science and Evidence-Based Decision Making

Preprint

Jul 2024

CONCEPTUAL FOUNDATIONS FOR THE USE OF GIS TECHNOLOGIES FOR THE STUDY OF COASTAL ECOSYSTEMS OF THE AZOV-BLACK SEA BASIN IN THE FRAMEWORK OF BLUE GROWTH

Article

Dec 2023

A. V. SRYBERKO

Topicality. The paper discusses the possibility of using modern GIS technologies for studying coastal ecosystems of the Azov-Black Sea basin within the framework of blue growth. Today, GIS technologies greatly facilitate the study of the impact of natural and anthropogenic factors on the natural environment. The development of these technologies has increased greatly in recent decades. Now they can be used as a separate and independent tool for research in various areas of production. Based on the results obtained on the basis of GIS technologies, it is possible to draw scientific conclusions or create and implement management decisions to maintain sustainable or resilient development of the state of the environment under significant environmental pressure. Also conduct an economic, environmental, social assessment of the consequences of anthropogenic and natural impacts on marine and coastal ecosystems and the terrestrial ecosystem.Aim and tasks. Scientific substantiation of the possibilities of using modern GIS technologies regarding the sustainable development of the coastal ecosystem of the Azov-Black Sea basin within the framework of blue growth was the aim of the study.Materials and Methods. The studies were based on coastal mapping methods. The classification of environmental parameters based on United States Geological Survey (USGS) data was the initial information. The ecological coastal classification (Ecological Coastal Units (ECU)) was developed by the US Geological Survey (USGS) in collaboration with Esri and the Marine Biodiversity Observation Network (MBON). These data were developed as part of a Group on Earth Observation (GEO) initiative called GEO Ecosystems (GEO ECO). They are linked to the GEO ECO task of developing global data on coastal ecosystems.Research results. The state of the coastal strip of the seas of Ukraine according to the classification of environmental parameters based on data from the United States Geological Survey (USGS) was described in this paper. It was stated that the ecological status of coastlines is classified into five ocean parameters, two land parameters and three coastal parameters. The research results showed that the ecological situation of the coastal strips of the seas of Ukraine generally corresponds to the second, third, sixth and fifteenth classes according to the ecological coastal classification.Conclusion. It is presented that the coastal strip classification method discussed in this paper based on US Geological Survey (USGS) data can serve as a very powerful tool in the process of managing coastal zones of the Azov-Black Sea basin. It has been established that GIS technologies at the current stage of development have very great potential for studying coastal ecosystems of the Azov-Black Sea basin within the framework of blue growth.

Reprint: The Cells of Ecosystem Functioning: Towards a holistic vision of marine space

Chapter

Sep 2024

Big data in Earth science: Emerging practice and promise

Article

Mar 2024
SCIENCE

Improvements in the number and resolution of Earth- and satellite-based sensors coupled with finer-resolution models have resulted in an explosion in the volume of Earth science data. This data-rich environment is changing the practice of Earth science, extending it beyond discovery and applied science to new realms. This Review highlights recent big data applications in three subdisciplines—hydrology, oceanography, and atmospheric science. We illustrate how big data relate to contemporary challenges in science: replicability and reproducibility and the transition from raw data to information products. Big data provide unprecedented opportunities to enhance our understanding of Earth’s complex patterns and interactions. The emergence of digital twins enables us to learn from the past, understand the current state, and improve the accuracy of future predictions.

Visualization of 3D Hyperspectral Soil Mapping Data via Autoencoder-based Clustering

Conference Paper

Dec 2023

Seascapes as a new vernacular for pelagic ocean monitoring, management and conservation

Article

Full-text available

Jul 2016

For terrestrial and marine benthic ecologists, landscape ecology provides a framework to address issues of complexity, patchiness, and scale—providing theory and context for ecosystem based management in a changing climate. Marine pelagic ecosystems are likewise changing in response to warming, changing chemistry, and resource exploitation. However, unlike spatial landscapes that migrate slowly with time, pelagic seascapes are embedded in a turbulent, advective ocean. Adaptations from landscape ecology to marine pelagic ecosystem management must consider the nature and scale of biophysical interactions associated with organisms ranging from microbes to whales, a hierarchical organization shaped by physical processes, and our limited capacity to observe and monitor these phenomena across global oceans. High frequency, multiscale, and synoptic characterization of the 4-D variability of seascapes are now available through improved classification methods, a maturing array of satellite remote sensing products, advances in autonomous sampling of multiple levels of biological complexity, and emergence of observational networks. Merging of oceanographic and ecological paradigms will be necessary to observe, manage, and conserve species embedded in a dynamic seascape mosaic, where the boundaries, extent, and location of features change with time.

Automatic classification of climate change effects on marine species distributions in 2050 using the AquaMaps model

Article

Full-text available

Mar 2016
ENVIRON ECOL STAT

Habitat modifications driven by human impact and climate change may influence species distribution, particularly in aquatic environments. Niche-based models are commonly used to evaluate the availability and suitability of habitat and assess the consequences of future climate scenarios on a species range and shifting edges of its distribution. Together with knowledge on biology and ecology, niche models also allow evaluating the potential of species to react to expected changes. The availability of projections of future climate scenarios allows comparing current and future niche distributions, assessing a species’ habitat suitability modification and shift, and consequently estimating potential species’ reaction. In this study, differences between the distribution maps of 406 marine species, which were produced by the AquaMaps niche models on current and future (year 2050) scenarios, were estimated and evaluated. Discrepancy measurements were used to identify a discrete number of categories, which represent different responses to climate change. Clustering analysis was then used to automatically detect these categories, demonstrating their reliability compared to human supervised classification. Finally, the distribution of characteristics like extinction risk (based on IUCN categories), taxonomic groups, population trends and habitat suitability change over the clustering categories was evaluated. In this assessment, direct human impact was neglected, in order to focus only on the consequences of environmental changes. Furthermore, in the comparison between two climate snapshots, the intermediate phases were assumed to be implicitly included into the model of the 2050 climate scenario.

Dissolved Oxygen, Apparent Oxygen Utilization, and Oxygen Saturation

Technical Report

Full-text available

Dec 2013

This atlas consists of a description of data analysis procedures and horizontal maps of climatological distribution fields of dissolved oxygen, apparent oxygen utilization (AOU), and dissolved oxygen saturation at selected standard depth levels of the world ocean on a one-degree latitude-longitude grid. The aim of the maps is to illustrate large-scale characteristics of the distribution of dissolved oxygen. The oceanographic data fields used to generate these climatological maps were computed by objective analysis of all scientifically quality-controlled historical dissolved oxygen data in the World Ocean Database 2013. Maps are presented for climatological composite periods (annual, seasonal, monthly, seasonal and monthly difference fields from the annual mean field, and the number of observations) at 102 standard depths.

Corrigendum: Options for managing impacts of climate change on a deep-sea community

Article

Full-text available

Jul 2015

The deep sea hosts some of the world's largest, oldest, and most sensitive ecosystems. Climate change and ocean acidification are likely to have severe implications for many deep-sea ecosystems and communities, but what, if anything, can be done to mitigate these threats is poorly understood. To begin to bridge this gap, we convened a stakeholder workshop to assess and prioritize options for conserving legislatively protected deep-sea coral reefs off southeast Australia that, without management intervention, are likely to be severely degraded within decades as a result of climate change. Seventeen possible options were explored that span biological, engineering and regulatory domains and that differed widely in their perceived costs, benefits, time to implementation, and risks. In the short term, the highest priority identified is the need to urgently locate and protect sites globally that are, or will become, refugia areas for the coral and its associated community as climate change progresses.

Video-of-Ocean-Solutions

Data

Full-text available

Aug 2015

Dawn J Wright

Ecological Geography of the Sea

Article

Alan Longhurst

Ocean Depths: The Mesopelagic and Implications for Global Warming

Article

Jan 2017
CURR BIOL

The mesopelagic or ‘twilight zone’ of the oceans occurs too deep for photosynthesis, but is a major part of the world’s carbon cycle. Depth boundaries for the mesopelagic have now been shown on a global scale using the distribution of pelagic animals detected by compiling echo-soundings from ships around the world, and been used to predict the effect of global warming on regional fish production.

Biogeography of the Global Ocean’s Mesopelagic Zone

Article

Dec 2016

The global ocean's near surface can be partitioned into distinct provinces on the basis of regional primary productivity and oceanography [1]. This ecological geography provides a valuable framework for understanding spatial variability in ecosystem function but has relevance only partway into the epipelagic zone (the top 200 m). The mesopelagic (200-1,000 m) makes up approximately 20% of the global ocean volume, plays important roles in biogeochemical cycling [2], and holds potentially huge fish resources [3-5]. It is, however, hidden from satellite observation, and a lack of globally consistent data has prevented development of a global-scale understanding. Acoustic deep scattering layers (DSLs) are prominent features of the mesopelagic. These vertically narrow (tens to hundreds of m) but horizontally extensive (continuous for tens to thousands of km) layers comprise fish and zooplankton and are readily detectable using echosounders. We have compiled a database of DSL characteristics globally. We show that DSL depth and acoustic backscattering intensity (a measure of biomass) can be modeled accurately using just surface primary productivity, temperature, and wind stress. Spatial variability in these environmental factors leads to a natural partition of the mesopelagic into ten distinct classes. These classes demark a more complex biogeography than the latitudinally banded schemes proposed before [6, 7]. Knowledge of how environmental factors influence the mesopelagic enables future change to be explored: we predict that by 2100 there will be widespread homogenization of mesopelagic communities and that mesopelagic biomass could increase by approximately 17%. The biomass increase requires increased trophic efficiency, which could arise because of ocean warming and DSL shallowing.

Delineating marine ecological units: A novel approach for deciding which taxonomic group to use and which taxonomic resolution to choose

Article

Aug 2015
DIVERS DISTRIB

Aim Ecological maps are increasingly used to support marine management and conservation. However, the biological datasets used to produce these maps are typically limited to taxonomic groups identified to the specific taxonomic levels available. Ecological units should, however, reflect the broader marine ecosystem, independent of the datasets used. This study assessed the influence of taxonomic groups and taxonomic resolution on the process of ecological mapping. Location Estuary and Gulf of St Lawrence (EGSL), Canada. Methods A dataset of more than 200 taxa of benthic macrofauna was used to create a set of biological matrices corresponding to different taxonomic groups (i.e. vertebrates, invertebrates, arthropods, echinoderms, molluscs, all taxa) and different taxonomic levels from species to class. Multivariate regression trees (MRTs) were used to identify environmental drivers of taxa distribution and to create ecological maps. Similarity between maps was assessed using pairwise comparisons. First, the relationships between the two classification legends were assessed using association plots on the partitions in the corresponding trees. Then, the spatial agreement of ecological units believed to represent the same habitat types was quantified and mapped. Results The comparison across different taxonomic groups showed a substantial level of similarity between ecological maps, indicating that ecological units defined for a specific taxonomic group can be considered to some extent as representative of the entire benthic macrofauna. Moreover, little information was lost when working at the family rather than species level, and common patterns of community distribution could still be distinguished at the class level. Main conclusions Using a novel spatially explicit approach for comparing ecological maps, this study demonstrates that datasets limited by taxonomic breadth or resolution can perform nearly as well as more extensive datasets. These simplifications should improve our ability to manage marine ecosystems.

Ecology of Deep Oceans: Hadal Trenches

Chapter

Oct 2011

Alan J. Jamieson

The majority of the Earth is covered by ocean which itself averages greater than 3500 m deep. It is comprised mainly of the vast deep ocean plains, which span a depth range of 2000–6000 m. The maximum depth of the ocean however is nearly 11 000 m deep. These extremely deep areas are a result of tectonic plate convergence where an oceanic plate is subducted beneath a neighbouring continental plate. This subduction forms extraordinarily deep trenches which comprise what is known as the ‘hadal zone’. The hadal zone (6000–11 000 m) is therefore encompassed by clusters of disjunct and often extremely isolated deep trenches and troughs. The morphology of these habitats had provided a setting where environmental conditions (hydrostatic pressure and food supply) differ greatly from the majority of the deep‐sea and has resulted in high levels of species endemism at the greatest depths. Key Concepts The deepest biozone in the oceans is the hadal zone (6000–11 000 m deep). The hadal zone is comprised mainly of deep trenches formed by tectonic convergence. The trench environment is characterised by high hydrostatic pressure, low temperatures, the absence of light and a limited food supply. Trench topography provides a unique setting with regard to food supply input because organic matter (food) is accumulated at the deepest trench axis. High levels of species endemism are observed in trenches as a result of their geographical isolation. Physiological adaptation to high pressure and low food supply is an essential prerequisite for survival in the trenches. Most metazoan taxa are represented in these deep trenches. Until recently, the technical challenges of studying the hadal zone were responsible for a dearth of information regarding the ecology at these depths. Research into the ecology of the hadal zone is in its infancy relative to shallower biozones.

A Three-Dimensional Mapping of the Ocean Based on Environmental Data

Abstract and Figures

Recommended publications

How the Oceanographic Community Created a National Oceanographic Partnership Program

Stratifying ocean sampling globally and with depth to account for environmental variability

Mapping near surface global marine ecosystems through cluster analysis of environmental data

Modeling global hammond landform regions from 250-m elevation data: Karagulle

Marine Ecosystems of the World