close

Вход

Забыли?

вход по аккаунту

?

Geophysical approaches to the classification delineation and monitoring of marine habitats and their communities.

код для вставкиСкачать
AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Published online in Wiley InterScience
(www.interscience.wiley.com). DOI: 10.1002/aqc.525
VIEWPOINT
Geophysical approaches to the classification, delineation and
monitoring of marine habitats and their communities
JOHN C. ROFFa,*, MARK E. TAYLORb and JOSH LAUGHRENc
a
Zoology Department, University of Guelph, Guelph, Ontario, Canada N1G 2W1
Beak International Incorporated, 14 Abacus Road, Brampton, Ontario, Canada L6T 5B7
c
World Wildlife Fund Canada, 245 Eglinton Avenue E. Suite 410, Toronto, Ontario, Canada M4P 3J1
b
ABSTRACT
1. If marine environments are to be systematically protected from the adverse effects of human
activities, then identification of the types of marine habitats and the communities they contain, and
delineation of their boundaries utilizing a consistent classification is required. Human impacts on
defined communities can then be assessed, the ‘health’ of these communities can be monitored, and
marine protected areas can be designated as appropriate.
2. Schemes to classify habitats at local and regional scales, according to their geophysical
properties, may identify different factors as determinants, and/or use them in different sequences in a
hierarchical classification. We examined the reasons for these differences in local and regional
applications of a global concept, and argue that a common set of factors could be applied in a
defined and defensible sequence to produce a common hierarchy of habitat types among geographic
regions.
3. We show how simple mapping and GIS techniques, based on readily available data, can lead to
the identification of representative habitat types over broad geographic regions. We applied a
geophysical framework first to the entire Canadian coastline and second to the Scotian Shelf of
Atlantic Canada to establish broad scale marine natural regions and ‘seascapes’, respectively. This
ecosystem level approach } which defines representative habitat types } is a fundamental
prerequisite for many purposes. It can form the basis for further analyses including: definition of
community types from habitat } community relationships; evaluation of the potential roles of focal
species in marine conservation; evaluation of candidate marine protected areas; definition of
unaffected reference areas against which the effects of human activities can be gauged; guidance for
water quality monitoring studies; management of marine resources.
Copyright # 2003 John Wiley & Sons, Ltd.
KEY WORDS: marine protected areas; marine habitats; marine communities; geophysical factors; classification;
conservation; monitoring
*Correspondence to: Prof. J.C. Roff, Zoology Department, University of Guelph, Guelph, Ontario, Canada NIG 2W1.
Present address: Prof. J.C. Roff, Biology Department, Acadia University, Wolfville, Nova Scotia, Canada BOP 1X0.
Copyright # 2003 John Wiley & Sons, Ltd.
Received 28 January 2001
Accepted 25 February 2002
78
J.C. ROFF ET AL.
INTRODUCTION
In a previous paper (Roff and Taylor, 2000), we argued that if marine environments are to be systematically
managed and protected from the adverse effects of human activities, then identification of the types of marine
habitats and the communities they contain and delineation of their boundaries within a consistent
classification is required. Once this is done, human impacts on defined communities can be assessed, the
‘health’ of these communities can be monitored, and marine protected areas can be designated as appropriate.
Appreciation of the merits of habitat classifications schemes, and how habitats can be ranked
hierarchically seems to be growing. In the marine environment there are now several examples of such
classifications at the regional or local level, for example } states, provinces or specific estuaries (e.g. Welch,
1978; Davis et al., 1994; Zacharias et al., 1998; Foster-Smith et al., 2000). However, there are few published
examples of the mapping that can result from such exercises over larger areas, or the rationale for the
hierarchy generated. There are also few examples of the potential uses to which such mapping can be put in
terms of marine conservation initiatives at the national or international level. Although marine ecologists
and conservationists appear to recognize that mapping of habitats and their biological communities in the
marine environment over larger areas is } of necessity } based on geophysical factors (see e.g. Longhurst,
1998; Roff and Taylor, 2000) they do not yet seem to have fully realized the potential strengths and uses of
the process. Nor has this process yet been incorporated into conservation plans at the continental or
international level, where emphasis has been on ill-defined ‘ecosystems’ (e.g. Sherman and Alexander, 1989)
or global biomes defined solely from pelagic characteristics (Longhurst, 1998).
In order to demonstrate the procedures, and potential applications of such mapping, we applied a
geophysical framework first to the entire Canadian coastline (out to the 200 mile limit or to the international
boundaries of the Canadian Economic Exclusion Zone) to establish broad scale marine natural regions, and
second to a specific example area } the Scotian Shelf of Atlantic Canada. Canada has the longest coastline,
the largest archipelago and the second largest continental shelf of any country in the world, and exerts
jurisdiction out to the 200 mile limit. In such a context, a geophysical classification of habitat types, as
surrogates for marine community types, represents the only practical foundation for marine conservation.
There are many potential approaches to marine conservation, corresponding to the several levels of the
ecological hierarchy (see e.g. Zacharias and Roff, 2000), but the ecosystem level approach } which defines
representative habitat types (ones typical of their surroundings at some scale, Roff and Taylor, 2000) } is a
fundamental prerequisite for many purposes. We show how simple mapping and GIS techniques based on
readily available data, can lead to the identification of broad representative habitat types over wide geographic
regions. Such a product can then form the basis of further analyses including: broad definition of community
types from habitat – community relationships; evaluation of the potential roles of focal species in marine
conservation (see Zacharias and Roff, 2001a); evaluation of candidate marine protected areas; definition of
unaffected reference areas against which the effects of human activities can be gauged; guidance for water
quality monitoring studies, and so forth. In this paper, we are primarily concerned with the structures of the
marine environment at the habitat – community level of organization. Classification and delineation of
distinctive habitats (ones atypical of their surroundings at some scale, Roff and Taylor, 2000) is treated in
Roff and Evans (2002). Evaluation of ecological integrity, which requires knowledge of the processes
occurring in selected areas (see Zacharias and Roff, 2000) also lies beyond the scope of this communication.
AN OVERALL APPROACH TO HABITAT CLASSIFICATION } THEORY AND PRAGMATICS
Marine communities are recognized and differentiated either taxonomically (e.g. pelagic fisheries, benthic
macroalgae) or according to their habitat (e.g. intertidal rocky shore, salt marshes). Given the task of
mapping marine habitat types according to geophysical properties, from a global perspective, most marine
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
79
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
ecologists would probably identify a set of factors similar to (or at least including) those in Roff and Taylor
(2000) } (see Table 1). However, schemes to classify habitats at a local (10’s to 100’s of km) or regional
(100’s to 1000’s of km) scale may identify different sets of actual factors as determinants, and use them in
different sequences in a hierarchical classification (e.g. Dethier, 1992; Connor, 1997; Zacharias et al., 1998;
Roff and Taylor, 2000). It is important to determine the reasons for these differences in local or regional
applications of a global concept, and examine whether a common scheme for habitat classification could be
developed among geographic regions. If it could, then such a generalized scheme would considerably
facilitate conservation and environmental efforts at both regional and international levels as explained
further below (see the ‘Potential applications of habitat classifications’ section).
There are a number of important considerations for the development of a generalized habitat
classification scheme. First, the potential (global) set of factors that can be used to discriminate among
habitat types must be determined by what can be mapped from available geophysical data and what can be
readily obtained by remote or in situ sensing. Fortunately, our ability to map geophysical factors } at least
locally } is improving as sensors become more sophisticated (see e.g. Foster-Smith et al., 2000).
Second, there may be some redundancy between factors or they may need to be computed in different
ways. Thus, some combinations of factors may be used as surrogates for others; for example slope and
Table 1. Geophysical factors that can be used to classify habitat types
General list
Factors used in this
classification
Predominant scale of
effect of factors
Oceanographic
Oceanographic
Ice cover
Temperature
Salinity
Water masses
(temperature and salinity signatures)
Temperature anomalies
Temperature gradients (e.g. fronts)
Light penetration
Water column stratification
Nutrient concentrations
Tidal amplitude
Tidal currents
Exposure (to wave action or desiccation)
Ice cover
Temperature
Global/regional
Global/regional/local
Water column stratification
Regional
Exposure
(to wave action or desiccation)
Regional/local
Oxygen concentration
Physiographic
Tectonic motion
Latitude
Depth (bathymetry)
Relief/bottom slope
Rate of change of slope/heterogeneity
Substrate particle size
Rock type
Physiographic
Depth (bathymetry)
Relief/ bottom slope
Regional
Regional
Substrate particle size
Local
NB: By local is meant sub-national jurisdiction or scales of 10’s to 100’s of km.
By regional is meant sub-national to national jurisdiction or scales of 100’s to 1000’s of km.
For further interpretation see Roff and Taylor (2000).
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
80
J.C. ROFF ET AL.
current speed may be used as a surrogate for substrate type. Also, water column stratification may be
predicted and modelled from the Simpson – Hunter stratification parameter (h/U3, where h is the water
depth and U the tidal velocity; see Pingree, 1978), but only where the lunar tidal component dominates. In
other areas, actual data on water column stratification may be needed (e.g. Ds t/Dz; the change in density }
Ds t } with change in depth } Dz). However, the two parameters can be cross calibrated and essentially
show the same thing (see Table 2).
Third, the actual set of factors chosen for a classification hierarchy within any region will depend upon
the natural range of variation in each one. Some factors may not be applicable within a particular region
because they show little variation. For example, outside of estuaries (which comprise a set of habitats
separate from marine ones and should be separately classified) salinity is an important determinant of
community type in the Straits of Georgia, British Columbia (Roff et al., unpublished data) and in the Baltic
Sea (Hallfors et al., 1981), but not on the east coast of Canada (Day and Roff, 2000) or in the
Mediterranean (Connor et al., 1993). This is because the first two of these regions are partially landlocked
and subject to considerable land run-off and hence salinity variation. Thus they are estuarine in character.
In the following two examples, salinity does not vary sufficiently to act as a major determinant of
community type. Thus, although a common set of factors can be envisaged for a hierarchy (at least in
temperate areas throughout the world), in any particular region } where some factor may be relatively
homogeneous (i.e. it shows low variation) } it may not aid in discriminating among habitat types. Further,
it follows that in tropical and sub-tropical regions, where there is little variation in several geophysical
factors (e.g. temperature, salinity, stratification), we may need to place greater reliance on direct mapping
of the biological communities themselves } as is indeed typically the practice in such regions.
Fourth, the sequence in which factors enter a hierarchy should ideally be determined by the same
principle that applies in the ‘Natural System of Biological Classification’, i.e. it should depend upon which
has the greatest ability to discriminate among habitat types (and by implication their communities and
array of taxa). This means that habitat types at the upper levels of a hierarchy should be more distinct from
one another, while those within categories at lower levels of a hierarchy should be more similar. By
implication and association, this also means that community types should become more similar to one
another within the lower categories of a hierarchy, just as taxa do in a taxonomic classification system. In
practice, this calls for some expert judgement and knowledge of the taxonomy, biogeography and
physiology of the whole array of marine organisms. We shall not include a full defence of the sequence of a
hierarchy here, which may be expected to evolve just as has the ‘Natural System of Biological
Classification’. However, the sequence will also be based partly upon the scale at which each factor exerts
its predominant effect. Tables 1 and 2 indicate how we rank the factors used, and the sequence in which we
judge them to enter our hierachy (see also Allen and Starr, 1982; Roff and Taylor, 2000).
Following from this reasoning, it should be clear that } at the continental or international level } a
common set of factors could be applied in a defined and defensible sequence to produce a common
hierarchy of habitat types. However, it is to be expected that some factors will not apply in some regions,
i.e. } they do not help to discriminate among habitat types. In such cases, these parts of the hierarchy are
simply empty, and we pass on to the next factor and the next level. Examples of habitat mapping from
Canada now follow.
NATIONAL AND REGIONAL EXAMPLES OF MARINE HABITAT CLASSIFICATIONS FROM
CANADA
The geophysical factors on which our proposed hierarchy is based were presented and evaluated in Roff
and Taylor (2000) and are further explained in Day and Roff (2000).
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Copyright # 2003 John Wiley & Sons, Ltd.
Temperate
(avg. temp>08C,
5 188C)
Subtropical
(avg. temp>68C
in winter,
>188C in summer)
Boreal
(avg. temp>08C)
Benthic
Abyssal/hadal
(>2000 m)
Euphotic
(0–50 m)
Dysphotic/
aphotic
(50–200 m)
Bathyl
(200–2000 m)
Benthic
Benthic
substrate
LEVEL 8
Benthic–exposure Benthic–
sediments
Cold subarctic Exposure
Mud
(568C)
(depth 550 m)
Moderate
Benthic–slopeb
Mostly sand
temperate
(20–80% sand)
(6–98C)
High slope
Partially sand
(slope >2%)
(0–20% sand)
Low slope
Partially gravel
Warm gulf
(slope 52%)
(5–50% gravel)
stream
(>98C)
Mostly gravel
(>50% gravel)
Bathypelagic
(1000–2000 m)
Abyssal/hadal
(>2000 m)
Frontal
(1005da51000)
Epipelagic
(0–200 m)
Mesopelagic
(200–1000 m)
Mixing and
wave action
LEVEL 7
Pelagic–
stratificationa
Stratified
(da>1000)
Nonstratified
(da5100)
Pelagic
Benthic
temperature
LEVEL 6b
Density anomaly, da, is calculated as 1500 (Dst/Dz) (the rate of change of density with respect to depth). Density anomaly was used as a surrogate for the stratification
parameter (h/U3) where: h=water depth, U=current velocity, see Pingree (1978).
b
Slope determined for areas >50 m in depth.
a
Atlantic
LEVEL 6
Segregation Vertical
segregation
of benthic
and pelagic
Realms
LEVEL 5
Not applied Pelagic
LEVEL 4
Marine
LEVEL 3
Sea-ice
cover
LEVEL 2
Environment Geographic Temperature
type
range
LEVEL 1
Table 2. Hierarchical classification of geophysical factors used to define habitat types of the Scotian Shelf
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
81
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
82
J.C. ROFF ET AL.
Data and mapping techniques for the entire Canadian Economic Exclusion Zone (EEZ)
The National Marine Natural Regions of Canada data set has been developed by WWF Canada. The
resulting classification system is a hierarchical framework based on ecological principles and recurrent
geophysical and oceanic factors of the marine environment (see Roff and Taylor, 2000; Day and Roff,
2000). Marine natural regions were derived by combining data layers for geographic range (i.e. the EEZ),
temperature, sea ice cover and vertical segregation (depth) using ARC/INFO (Environmental Systems
Research Institute, Inc., Redlands, CA 92373 USA). Details on the classes within each layer can be found in
Day and Roff (2000).
Data sets from several sources were compiled and integrated to develop the maps. Sea surface
temperature was used as a surrogate for water mass temperature. Canada-wide sea surface temperature
data were obtained from the Ocean Climate Lab (OCL) at the National Oceanic and Atmospheric
Administration (NOAA) National Oceanic Data Center (NODC) } (http://www.nodc.noaa.gov/OC5/
readmehr.html). The data set was based the work of Boyer and Levitus (1997). Data for sea ice and
frequency of ice occurrence were derived from Sea Ice Map of Canada (Scale 1: 12 500 000) from the
National Atlas Information Service (1993). Bathymetric data were obtained from the National Atlas
Information Service through the digital topographic database of the Geo-Access Division, Canada Centre
for Remote Sensing, Scale 1:7 500 000 (http://geogratis.cgdi.gc.ca/frames.html).
Data and mapping techniques for the Scotian Shelf
The framework used for the Scotian Shelf case study is a variant of Day and Roff (2000) and Roff and
Taylor (2000) as indicated in Table 2. Surface and bottom temperatures for the Scotian Shelf area were
compiled primarily from Petrie et al. (1996). Ice data were available from the Department of Fisheries and
Oceans, Canada, and the winter surface isotherm of 08C was used as an approximation for the seasonal ice
limit.
Depth contours at an original scale of 1:2 000 000 were supplied in digital form from the Scotian Shelf
Atlas, by the Geological Survey of Canada, Atlantic Geoscience Centre, Bedford Institute (1991). Depth
regions were mapped using the 0, 50, 200, 1000 and 2000 m contours.
Water density profiles between surface and 50 m or bottom, were calculated from summer temperature
and salinity data or density data, supplied by J. Loder (Bedford Institute of Oceanography). From SW
Nova Scotia and Gulf of Maine, this parameter was cross calibrated against the modelled stratification
parameter, h/U3 (see Pingree, 1978; Iles and Sinclair, 1982). Values of Dst/Dz 1500 between 100 and 1000
were selected to separate stratified from non-stratified waters (values 5100 indicate non-stratified water,
values >1000 indicate strongly stratified water).
Waters less than 50 m in depth are subject to wave exposure which is an important determinant of
community type. However, mapping of exposure classes would require a finer scale of analysis than was
undertaken in this case study. Therefore, all waters less than 50 m in depth were mapped as subject to
exposure, but no further analysis was done. A finer analysis of exposure would be important for any coastal
planning exercise. Mapping of slopes on the shelf was generated within ARC/INFO using the bathymetry
point data. The mapping used a triangulation method between adjacent points to determine ‘rise over run’.
The surficial bottom sediments types on the shelf have been mapped by Fader (1991) and Piper (1991).
Additional information on sediment texture (from C. Amos, Geological Survey of Canada, Dartmouth,
Nova Scotia) was also incorporated. Five classes were used instead of the national classification viz. mud,
partially sand (i.e. 0–20% sand), mostly sand (20–80% sand), partially gravel (5–50% gravel), and mostly
gravel (>50% gravel).
Pelagic seascapes were derived using ARC/INFO techniques by overlaying data layers for temperature,
vertical segregation (pelagic depth classes) and stratification. Benthic seascapes were similarly derived by
overlaying data layers for temperature, vertical segregation (benthic depth classes), benthic temperature,
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
83
exposure and benthic substrate (sediments). Pelagic and benthic seascapes were further combined to
generate a single layer representing both benthic and pelagic seascapes.
Geophysical maps of Canada’s Exclusive Economic Zone
There are major differences in the marine climate within the waters of Canada’s EEZ (Plates 1(a)–(c)).
These do not follow terrestrial climate variations nor do they follow gradients of latitude. Variations in
temperature should be considered separately within the pelagic and benthic realms, because of vertical
stratification. On the east coast of Canada, surface water masses (0–50 m) of Arctic origin penetrate as far
south as the Gulf of Maine. Planktonic species which are indicators of Arctic waters (e.g. Calanus glacialis
and Calanus hyperboreus) can be found in low numbers all along the Scotian Shelf even during the summer
months (Tremblay and Roff, 1983). Conversely, deeper waters of temperate Atlantic origin penetrate as far
north as Ungava Bay and Hudson Strait, where Calanus finmarchicus still dominates the zooplankton (Roff
and Legendre, 1986). The separate movements of upper and lower water masses, their different temperature
and salinity effects, and the differential effects on water column and benthic organisms, all highlight the
necessity of considering the pelagic and benthic realms separately within a hierarchy (see Day and Roff,
2000).
Sea ice (Plate 1(c)) is a major correlate of community type for several reasons. Intertidal ice scour
substantially lowers biodiversity (Bergeron and Bourget, 1986). Ice cover determines in part submarine
light penetration and the timing and amplitude of seasonal cycles of production. In Arctic waters it is
correlated with increased stability of the water column and reduced seasonal production. Under permanent
ice cover, primary production is extremely low (Subba Rao and Platt, 1984). Temperature, pressure and
light all vary with depth which can act as a surrogate for these factors in the vertical plane. Depth (Plate
1(d)) therefore represents a composite of several physiological, trophic and topographic factors for both the
pelagic and benthic realms.
The number of categories within each geophysical factor is to some degree arbitrary, and will be
determined by the variability within the region under consideration. For each of the factors considered here
we have selected five categories at the national level. These conform to similar divisions in previous studies
and can be rationalised in various ways (see Roff and Taylor, 2000). For example, the 0–50 m depth interval
approximates the euphotic zone in coastal waters, and the 200 m depth represents the edge of the
continental shelf. By overlaying the maps of these factors for the whole EEZ, we produce the resultant map
of Marine Natural Regions of Canada (Plate 2).
It is instructive to compare the greater homogeneity of natural regions on the west coast of Canada with
those on the east coast. British Columbia lies at the junction of the sub-tropical and sub-polar gyres, and
experiences greater spatial homogeneity and lower seasonal variation of temperature (Zacharias et al.,
1998). However, because of the partially land-locked nature of its southern coastal waters, they are
estuarine in character. Thus we should expect that the dominant marine characteristics of a region and the
relative significance of factors would vary according to both ocean basin and regional effects. For example,
marine waters on the west and east sides of an ocean basin will conform to major differences in circulation
and ocean currents, especially westward intensified currents, gyral systems and upwelling regimes.
Superimposed at the regional level, local terrestrial topographic effects } for example mountains which
encourage precipitation and fjord-like areas (e.g. coasts of British Columbia and Norway) and offshore
islands (e.g. Vancouver Island) } may constrain runoff and lead to estuarine conditions in coastal waters.
We should therefore expect, even at a given latitude, that our regional hierarchy will differ between east and
west coasts of the major continents.
An analysis of the British Columbia seascapes is presently underway, and it is expected to result in a
somewhat different hierarchy from the east coast of Canada (Roff et al., unpublished data). As an example
of a regional classification scheme, we present our study of the Scotian Shelf on the east coast of Canada.
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
84
J.C. ROFF ET AL.
Geophysical maps of The Scotian Shelf
The Scotian Shelf includes the marine waters off the Atlantic coast and adjacent to Nova Scotia extending
out to the 200 mile limit (Plates 3 and 4). It totals an area of approximately 316 000 km2. The northern limit
of the study area lies west of Cape Breton Island (i.e. the Laurentian Channel) while the southern limit
includes part of Georges Bank; the Bay of Fundy is also within the study area.
The Scotian Shelf is an area of strong contrasts in temperature, due to the inshore influence of the cold
Labrador Current and the warm offshore Gulf Stream. Outside of estuaries, salinity variations on the shelf
are only important in the uppermost Bay of Fundy and inner Gulf of St. Lawrence (Petrie et al., 1996). The
08C, 58C and 188C isotherms were chosen to distinguish the major water masses (Plate 3(a)). Marine waters
are continuous, therefore any limits set by temperatures must have some element of arbitrariness. However,
in plotting temperatures on the shelf, it became apparent that there was marked congruence between the
58C winter isotherm and the 188C summer isotherm. This is probably a good indication of the average
boundary between temperate slope waters and the sub-tropical waters of the Gulf Stream, irrespective of
season. The presence or absence of winter ice is significant for many organisms, and distinguishes cold
(boreal) from temperate waters. The approximate location of the seasonal ice limit is adequately described
by the winter 08C surface isotherm, which runs into the northern part of Cape Breton. Depth classes are
shown in Plate 3(b).
Most of the Scotian Shelf is influenced by a mixed tidal regime of low amplitude leading to a stratified
water column during the summer months (J. Loder, Bedford Institute of Oceanography, pers. comm.).
From south-west Nova Scotia into the Gulf of Maine, the area is progressively dominated by the M2 tidal
component, and resonanace in the Bay of Fundy produces exceptionally high tides. These regions are
therefore either unstratified or include frontal regions (Plate 3(c)). The vertical stratification of a water
column separates its communities both spatially and temporally, and stratified and non-stratified waters
differ both in their annual productivity regimes and in their community structures (Pingree, 1978). Thus
banks and coastal regions of high tidal amplitude will remain unstratified, and may retain or accumulate
populations of important pelagic species such as larval fish (Jeffrey and Taggart, 1999). Transitional
(frontal) areas can represent boundaries between populations or communities and may persist as regions of
high production throughout the summer months (Pingree, 1978; Iles and Sinclair, 1982). The combination
of temperature, depth classes and stratification regime results in the pelagic seascapes shown in Plates 3(d)
and 5(a).
This region becomes stratified during the summer months and is under the influence of the cold south
flowing Labrador waters, hence the bottom temperature (Plate 4(a)) remains cold over much of the shelf.
Higher temperature (and higher salinity) water may displace this colder but less saline water in the basins
along the shelf. Areas of high exposure and slope, subject to wave action or currents are shown in Plate
4(b).
Sediment types (Plate 4(c)), which result from a combination of factors including current velocity, have a
major effect on the type of benthic community that develops in a region (e.g. Barnes and Hughes, 1982).
The benthic seascapes that result from the combination of these factors are shown in Plates 4(d) and 5(b).
Finally, by combining both the pelagic and benthic seascapes we derive the combined seascapes for the
Scotian Shelf (Plate 6).
POTENTIAL APPLICATIONS OF HABITAT CLASSIFICATIONS
There are many potential uses for a system of habitat classification such as that presented here (see Table 3),
and it has several advantages over ecological classifications based only on species (see Roff and Taylor,
2000). Data on distributions of individual species are generally only available for those that are
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Permanent
Seasonal
Variable
Polyna
No Ice
Sea-Ice Cover
D. Generalized Bathymetry
B. Climatic Regions
0m - 50m (Epipelagic - Euphotic)
50m - 200m (Epipelagic - Dysphotic)
200m - 1000m (Mesopelagic - Bathyl)
1000m - 2000m (Bathypolagic - Bathyl)
> 2000m (Abyssal - Abyssal)
Depth Classes (Palagle - Benthie)
Temperature Climatic Class
Arctic
Boreal
Sub-Arctic
Sub-Tropical
Temperate
Plate 1. Geophysical maps of the Canadian Exclusive Economic Zone. (A) Extent of marine regions within Canada’s Exclusive Economic Zone, (B) Marine climatic
regions based on surface temperatures, (C) Extent and classes of sea ice cover and (D) Generalized bathymetry.
C. Sea-Ice Cover
Canada's Exclusive Economic Zone (EEZ)
A. Extent of Marine Regions
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Marine Natural Regions of Canada
Atlantic, Boreal, Absent, Abyssal - Abyssal
Atlantic, Boreal, Seasonal, Bathypelagic - Bathyl
Atlantic,Boreal, Seasonal, Epipelagic - Dysphotic
Atlantic, Boreal, Seasonal, Epipelagic - Euphotic
Atlantic, Boreal, Variable, Abyssal - Abyssal
Atlantic, Boreal, Variable, Bathypelagic - Bathyl
Atlantic, Boreal, Variable, Epipelagic - Dysphotic
Atlantic, Boreal, Variable, Epipelagic - Euphotic
Atlantic, Arctic, Seasonal, Abyssal - Abyssal
Atlantic, Arctic, Seasonal, Bathypelagic - Bathyl
Atlantic, Arctic, Seasonal, Epipelagic - Dysphotic
Atlantic, Arctic, Seasonal, Epipelagic - Euphotic
Atlantic, Sub-Arctic, Variable, Bathypelagic - Bathyl
Atlantic, Sub-Arctic, Variable, Epipelagic - Dysphotic
Atlantic, Sub-Arctic, Variable, Epipelagic - Euphotic
Atlantic, Sub-Tropical, Absent, Abyssal - Abyssal
Atlantic, Sub-Tropical, Absent, Bathypelagic - Bathyl
Atlantic, Temperate, Absent, Abyssal - Abyssal
Atlantic, Temperate, Absent, Bathypelagic - Bathyl
Atlantic, Temperate, Absent, Epipelagic - Dysphotic
Atlantic, Temperate, Absent, Epipelagic - Euphotic
Atlantic, Temperate, Variable, Bathypelagic - Bathyl
Atlantic, Temperate, Variable, Epipelagic - Dysphotic
Atlantic, Temperate, Variable, Epipelagic - Euphotic
Atlantic Basin
(Basin, Climatic/Temperature, Sea-Ice Cocer, Pelagic-Banthic Depth Class)
Plate 2. The Marine Natural Regions of Canada based on overlay of distributions of geophysical factors from Plates 1(A)–(D).
Pacific, Temperate, Absent, Abyssal - Abyssal
Pacific, Temperate, Absent, Bathypelagic - Bathyl
Pacific, Temperate, Absent, Epipelagic - Dysphotic
Epipelagic-Euphotic (0-50m) Depth Class not shown
Pacific Basin
(Basin, Climatic/Temperature, Sea-Ice Cocer, Pelagic-Banthic Depth Class)
Arctic, Arctic, Permanent, Abyssal - Abyssal
Arctic, Arctic, Permanent, Bathypelagic - Bathyl
Arctic, Arctic, Permanent, Epipelagic - Dysphotic
Arctic, Arctic, Permanent, Epipelagic - Euphotic
Arctic, Arctic, Polyna, Bathypelagic - Bathyl
Arctic, Arctic, Polyna, Epipelagic - Dysphotic
Arctic, Arctic, Polyna, Epipelagic - Euphotic
Arctic, Arctic, Seasonal, Abyssal - Abyssal
Arctic, Arctic, Seasonal, Bathypelagic - Bathyl
Arctic, Arctic, Seasonal, Epipelagic - Dysphotic
Arctic, Arctic, Seasonal, Epipelagic - Euphotic
Arctic, Sub-Arctic, Polyna, Bathypelagic - Bathyl
Arctic, Sub-Arctic, Polyna, Epipelagic - Dysphotic
Arctic, Sub-Arctic, Polyna, Epipelagic - Euphotic
Arctic, Sub-Arctic, Seasonal, Abyssal - Abyssal
Arctic, Sub-Arctic, Seasonal, Bathypelagic - Bathyl
Arctic, Sub-Arctic, Seasonal, Epipelagic - Dysphotic
Arctic, Sub-Arctic, Seasonal, Epipelagic - Euphotic
Arctic Basin
(Basin, Climatic/Temperature, Sea-Ice Cocer, Pelagic-Banthic Depth Class)
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
-64
-64
-62
-62
-60
-60
-58
-58
42
44
46
0
Kilometers
100
Frontial
Stratified
Well Mixed
200
C
Stratification Classes
200
-66
-64
-62
-60
-58
-66
-64
-62
-60
-58
100
Kilometers
200
42
44
46
0
Kilometers
100
200
D
Pelagic Seascapes
0
Epilagic / Euphotic
Epilagic / Dysphotic
Mesopelagic / Bathyl
Bathypelagic / Bathyl
Abyssal / Abyssal
Plate 3. The pelagic realm of the Scotian shelf. (A) Marine climatic regions based on surface temperatures, (B) Bathymetry showing depth classes (applies to
pelagic and benthic), (C) Stratification classes based on Dst/Dz (see text and Table 2) and (D) Pelagic Seascapes based on overlay of distributions of geophysical
factors from Plates 3(A)–(C).
-66
-66
Kilometers
100
42
42
0
44
46
B
Depth Classes
(Pelagic/Benthic)
44
46
Boreal
Temperate
Sub-Tropical
A
Climatic Zones
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
-64
-64
-62
-62
-60
-60
-58
-58
100
Kilometers
200
42
44
46
0
Kilometers
100
Mostly Sand
(20-80% sand)
Partially Sand
(0-20% sand)
Partially Gravel
(5-50% gravel)
Mostly Gravel
(>50% gravel)
Mud
200
C
Sediment Classes
0
Moderate Temperate
(6-9 °C)
Warm Gulf Stream
(>9 °C)
Cold Subarctic
(< 6 °C)
-66
-66
-64
-64
-62
-62
-60
-60
-58
-58
Kilometers
100
200
High Slope (< 2%)
Low Slope (< 2%)
Subject to Exposure
(depth < 50m)
0
42
44
46
Kilometers
100
200
D
Benthic Seascapes
0
42
44
46
B
Exposure and Slope
Classes
Plate 4. The benthic realm of the Scotian shelf. (A) Marine climatic regions based on bottom temperature zones, (B) Exposure and slope classes, (C) Sediment
classes and (D) Benthic seascapes based on overlay of distributions of geophysical factors from Plates 4(A)–(C).
-66
-66
42
44
46
A
Bottom Temperature
Zones
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
C. Seascape Heterogeneity
10
1
Number of
unique
seascapes
Plate 5. Keys and heterogeneity map. (A) Interpretive key for pelagic seascapes, (B) Interpretive key for benthic seascapes, (C) Seascape heterogeneity of
combined pelagic and benthic seascapes. Intensity of colour increases with number of unique seascapes captures in each 18 km radius circle (see text for further
explanation).
(6-9) - High Slope - Mostly Gravel
(6-9) - High Slope - Mostly Sand
(6-9) - High Slope - Mud
(6-9) - High Slope - Partially Gravel
(6-9) - Low Slope - Mostly Gravel
(6-9) - Low Slope - Mostly Sand
(6-9) - Low Slope - Mud
(6-9) - Low Slope - Partially Gravel
(6-9) - Low Slope - Partially Sand
(<6) - Exposed - Mostly Gravel
(<6) - Exposed - Mostly Sand
(<6) - Exposed - Partially Gravel
(<6) - Exposed - Partially Sand
(<6) - High Slope - Mostly Gravel
(<6) - High Slope - Mostly Sand
(<6) - High Slope - Mud
(<6) - High Slope - Partially Gravel
(<6) - Low Slope - Mostly Gravel
(<6) - Low Slope - Mostly Sand
(<6) - Low Slope - Mud
(<6) - Low Slope - Partially Gravel
(<6) - Low Slope - Partially Sand
(>9) - High Slope - Mostly Sand
(>9) - High Slope - Mud
(>9) - Low Slope - Mostly Gravel
(>9) - Low Slope - Mostly Sand
(>9) - Low Slope - Mud
(>9) - Low Slope - Partially Gravel
(Benthic Temperature in celcius)-Exposure and Slope-Sediment Type
B. Key for Benthic Seascapes
Boreal (Atlantic) - Epipelagic - Frontal
Boreal (Atlantic) - Epipelagic - No Data
Boreal (Atlantic) - Epipelagic - Stratified
Boreal (Atlantic) - Epipelagic - Well Mixed
Boreal (Atlantic) - Mesopelagic - Frontal
Boreal (Atlantic) - Mesopelagic - Stratified
Sub Tropical (Atlantic) - Abyssal - Frontal
Sub Tropical (Atlantic) - Abyssal - Stratified
Temperate (Atlantic) - Abyssal - Frontal
Temperate (Atlantic) - Abyssal - Stratified
Temperate (Atlantic) - Bathypelagic - Stratified
Temperate (Atlantic) - Epipelagic - Frontal
Temperate (Atlantic) - Epipelagic - Stratified
Temperate (Atlantic) - Epipelagic - Well Mixed
Temperate (Atlantic) - Mesopelagic - Frontal
Temperate (Atlantic) - Mesopelagic - Stratified
Temperate (Atlantic) - Mesopelagic - Well Mixed
Temperature (Ocean Basin) - Depth - Stratification
A. Key for Pelagic Seascapes
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
-64
-62
-60
-58
0
Kilometers
100
200
Plate 6. The combined seascapes of the Scotian shelf based on overlay of distributions of geophysical factors for the pelagic and benthic seascapes from
Plates 3(D) and 4(D).
-66
40
42
44
46
SCOTIAN SHELF
Seascapes
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
85
Table 3. Potential applications of habitat classification schemes
Definition of habitat – community type associations
Assessment of habitat suitability for defined purposes e.g. fisheries enhancement/aquaculture
Assessment of conflicts between actual or intended resource uses
Examining patterns of biodiversity distribution
Judging potential impact of invading species
Evaluation of candidate Representative Marine Protected Areas
Assessment of the potential role of focal species (e.g. umbrella and flagship species) in marine conservation
Guide to environmental monitoring programmes
Guide to habitat management and management practices
Guide to selection of unaffected reference areas for environmental monitoring
Framework for assessment and evaluation of ecosystem level processes
Framework for assessment of global warming effects
commercially exploited. Unfortunately, these are the very ones most impacted by human actions in terms of
distribution ranges and abundance. In addition, we may have no clear idea of how commercial species are
related to community types, i.e. whether or not they may act as composition indicators (see Zacharias and
Roff, 2001a).
Habitat – community associations
Relationships between habitat type and community type (ultimately to define biotopes) can only be verified
by direct sampling. However, habitat mapping provides guidance for sampling regimes and the probable
location of boundaries or gradients between community types. At this point it should be explicitly noted
that our seascapes (especially in the benthic realm) may not correspond to biotopes, but may rather
correspond to sets of community types that require further subdivision according to more detailed physical
factors. In this respect, further knowledge of ‘composition indicator’ species would be a major asset (see
Zacharias and Roff, 2001a).
In intertidal regions in temperate waters, the major habitats and their communities are relatively well
distinguished by physical factors. Thus, rocky shores are typically dominated by species of Fucus, Littorina,
etc. with variants from exposed to sheltered areas, while soft bottom shores are typically dominated by
species of burrowing molluscs e.g. Mya, Macoma and polychaetes e.g. Nephthys (see Connor, 1997).
Progressing into deeper waters, however, species diversity generally increases, and relationships between
major habitat types and community types can become less distinct. Thus the subtleties of habitat
characteristics will need to be more discretely defined than in the hierarchy presented here. The lowest levels
of the hierarchy we present (see also Roff and Taylor, 2000) are therefore unlikely to correspond to discrete
habitats containing discrete biocoenoses, especially in deeper waters. They probably represent some higher
aggregation of community type containing similar biocoenoses of related species composition. There is still
considerable room for research on this topic to achieve clearer definitions of habitat – community
associations (see e.g. Glenmarec, 1973; Wildish, 1977; McLusky and McIntyre, 1988).
Biocoenoses may also correspond to successional states following relatively predictable seasonal or interannual cycles (dependent on some combination of physical and biological events) or they may result from
largely unpredictable biological processes such as predation, competition, keystone species effects, or recolonization following physical disturbance. Thus we could easily describe as different biocoenoses, what is
really the same community, but with species abundances or occurences shifted along successional axes.
Such effects are still far from understood (cf. Grassle and Sanders, 1973) even in intertidal communities.
How much of the variation we see within communities is due to physical factors versus biological ones is
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
86
J.C. ROFF ET AL.
still an annoying question. In subtidal rocky communities for example there are major inter-annual changes
along the Laminaria } sea urchin (e.g. Echinus, Strongylocentrotus) axis (e.g. Scheibling et al., 1999).
For marine conservation purposes, it should be emphasized that the hierarchy presented here leads to the
identification of sets of representative habitats containing representative community types (see Roff and
Taylor, 2000). Distinctive (i.e. } locally unique) communities will only be illuminated by serendipitous
discovery or deliberate survey. There will always be surprises in the oceans! For example, no one had
predicted the extraordinary communities of totally new species associated with the deep sea vents in areas
of seafloor spreading (Ballard, 1977); they are distinct and unique. However, now that we know the
characteristics of their habitats (i.e. ocean ridges in areas of seafloor spreading, local temperature
anomalies, etc.) we can readily predict areas where such communities are likely to be found. Other
extraordinary deep sea communities e.g. coral and sponge beds should also become predictable in their
distributions as we learn more of the specifics of their habitat characteristics. Our view is therefore to
advocate persistence in investigating the subtleties of habitat characteristics, rather than to plead that all
marine communities are distinct and cannot be predicted from crude physico-chemical factors.
Habitat suitability assessment and resource utilization conflicts
Mapping of environmental variables in order to assess suitability of habitat for individual species of
commercial importance has now become commonplace (e.g. Brown et al., 1997). The habitats deemed
important for spawning, recruitment and feeding are increasingly and with more accuracy being described
in terms of their physical characteristics. We can also use maps of habitat characteristics to gauge potential
conflicts between conservation goals and resource utilization.
Examining patterns of biodiversity distribution
The global distribution of marine biodiversity has been established } by observation } in broad outline.
However, our ‘theories’ to account for these patterns generally lack explanatory power especially at
regional scales (100’s to 1000’s of km). There has been a paucity of studies to investigate relationships
between biodiversity and habitat characteristics (through geophysical factors) at such regional scales. Yet it
is precisely at the regional level that nation states exert jurisdiction over their marine resources and can best
effect conservation measures. Zacharias and Roff (2001b) showed that intertidal diversity was strongly
related to the geophysical environment, as a function of seasonal variations of temperature, salinity and
exposure to wave action. Such analyses between biodiversity and geophysical factors are vital to
systematically implemented conservation efforts at regional scales.
Judging potential impact of invading species
Increases in world-wide transportation have caused many species to significantly extend their geographic
ranges. However, although individual species may be transported from one side of an ocean to another, or
even between ocean basins, they still retain fidelity to the basic geophysical characteristics of their original
habitat. Thus, the potential range of an invading species should be predictable in its new geographic
location, and its ecological effects can then be assessed (e.g. Grosholz et al., 2000).
Evaluation of candidate Representative Marine Protected Areas
The mapping exercise described here is an indispensable preliminary stage in the process of evaluating sites
for Marine Protected Areas (MPAs). For example, we may wish to establish a set of MPAs that total some
target proportion of the continental shelf within a country’s jurisdiction. Strategies in selecting MPAs may
vary, but two suggest themselves: select a set of MPAs each of which captures the maximum habitat
heterogeneity within a region, or select a set of MPAs each of which is more homogeneous in habitat type,
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
87
but which collectively cover the target proportion of all representative habitat types. We carried out such a
mapping exercise on the Scotian Shelf to identify areas of greater and lesser habitat heterogeneity. We
superimposed a 2 2 km2 grid over the Scotian Shelf seascape maps, and drew 18 km radius cells around
each grid point. Within each cell, seascape heterogeneity was then calculated using the ‘focal variety’
function of ARC/INFO. Results are shown in Plate 5(c), where greater colour intensity represents higher
habitat heterogeneity. Candidate areas for either potential strategy now immediately suggest themselves
within a geographic region. By summing the results of such exercises, we can also evaluate how much of
each representative habitat type has been captured by a set of candidate MPAs. Thus our goal of protecting
some target proportion of all representative habitat types within a region, can be achieved.
The role of Focal Species in Marine Conservation
In marine conservation the emphasis has recently changed from species to spaces (Zacharias and Roff,
2000). However, public support still predominantly identifies with larger marine mammals, specifically with
those charismatic focal species (see Zacharias and Roff, 2001a) referred to as ‘flagships’. Unfortunately, the
trophodynamic significance of these species in the marine environment remains largely undefined (Bowen,
1997), and they may not be ecologically important (Katona and Whitehead, 1988). However, the
classification scheme proposed here now raises the possibility of directly examining the role of particular
focal species for defined marine conservation strategies. For example, knowing the distribution range (of
feeding grounds, breeding areas, etc.) of a particular focal species, we can ask questions such as: What does
this distribution range correspond to in terms of representative habitats? Can we usefully designate an
MPA based on the protection of a particular focal species or its required habitat, i.e.: does it act as an
umbrella species (see Simberloff, 1998; Zacharias and Roff, 2001a)? We might continue to ask: How many
such umbrella species would we need to protect in order to conserve a designated proportion of all types of
representative habitats in a region? Is it reasonable to propose focal species for marine conservation and if
so which ones or which combinations of focal species might we choose, and what would be the ecological
rationale? In all cases, in order to evaluate such possible strategies based on any focal species (e.g. umbrellas
or flagships) the correspondence between the distributions of individual species and habitat mapping is
essential.
Monitoring and management
Maps of habitat types can play an important role in assessing and predicting the likely impacts of human
effects on benthic communities. Thus, the broad community types that may be impacted and the
approximate proportion of each community type that may be affected within national boundaries could all
be assessed. Such maps can also direct our monitoring efforts. For example, the communities and habitats
of highest priority to be monitored are presumably those that are most sensitive to defined or expected
impacts, or most likely to be impacted. These can potentially be defined from habitat mapping.
Although we may not be able to specify the precise biocoenosis expected in a region from the local
habitat characteristics, when making comparisons among similar community types we can still assess
various criteria against some expected measure. For monitoring purposes, we may use several criteria to
assess the ‘health’ of a marine community, for example: species numbers, species diversity, various biotic
indices, species abundances, community condition indicators or community composition analyses (see e.g.
Warwick and Clarke, 1994; Zacharias and Roff, 2001a). Perhaps most usefully, we can use some ‘distance
measure’ of the difference between the observed and expected community composition (inferred from local
habitat characteristics) as an index of ‘ecosystem health’ or environmental quality.
The broad type of community expected in a region can be judged from the mapped habitat type and/or
from composition indicator species. Habitat mapping exercises will also aid in defining a suitable ‘reference’
community geographically elsewhere, in an unaffected area. Note that ‘expected’ conditions need to be
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
88
J.C. ROFF ET AL.
carefully specified from environmental parameters and seasonal regimes. Thus, high chlorophyll levels
occurring during the spring phytoplankton bloom, or in a region of known upwelling are to be expected.
However, the same chlorophyll level where it is not expected may be an indicator of pollution by nutrients.
This is the significance of observed versus expected and distance measures.
Selection of reference areas
In the process of selection of reference areas (those unaffected by environmental impacts) certain unaffected
habitat/community types may not exist within the marine jurisdictions of individual countries. They may
therefore need recourse to regions outside their own boundaries as reference areas. Unless a classification
scheme of the sort proposed here is in existence, such reference areas may not be readily found. Our
framework will assist in the identification of corresponding habitat types elsewhere. In some cases it may
even be necessary to examine corresponding areas beyond regional boundaries. As an extreme example, we
could suggest a comparison between disturbed benthic communities of the Gulf of Bothnia (Baltic Sea) and
unaffected communities in Rupert Bay (Hudson Bay, Canada)! The species assemblages may differ, but
community types and functional groups (guilds) should be similar in corresponding habitat types.
Assessment of ecosystem level processes
Monitoring, by definition, only measures ‘structures’; ecological processes are time dependent and more
difficult to measure. Assessment of ecological integrity requires an understanding of ecosystem level
processes, although these may often be deduced from appropriate structures (see Zacaharias and Roff,
2000). Processes such as recolonization of areas by recruitment of meroplanktonic larvae are vital to the
continued existence and health of marine communities. A classification of marine habitats such as that
proposed here is an indispensable framework against which to judge both the interactions of habitat and
community structure, and ecosystem level processes.
Global warming effects
The process of global warming may cause substantial local and regional changes in temperature, salinity
and ocean currents within the next 200 years or less. A marine habitat classification scheme constitutes a
powerful tool for the evaluation of climate change scenarios and for prediction of impacts in the marine
environment, including impacts on marine communities, biodiversity, and social and economic practices. If
our system of classification were based only on surveys of biological organisms, whose distributions and
ranges were continually changing under the effects of global warming, then direct sampling would be
constantly required. However, because only the upper levels of our hierarchy are expected to change in the
face of global warming (mainly due to changes in temperature and salinity regimes), habitat boundaries can
be easily re-drawn according to modeled scenarios and impacts can be predicted.
CONCLUSIONS AND FUTURE GOALS
Although, as we have shown, the classifications presented here have several applications, our interest is
primarily in marine conservation and defining frameworks and protocols for the establishment of MPAs.
We should clearly note that the procedures we describe here comprise only one possible ecological
approach to marine conservation (see Day and Roff, 2000; Zacharias and Roff, 2000 for a fuller account),
and only one step in the process of actual selection of MPAs. Nevertheless, habitat classification and
mapping is an indispensable step. However, there are several further steps in the planning processes that
must still be undertaken. Note for example, that we have not considered how MPA sites would be selected,
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
GEOPHYSICAL APPROACHES TO MARINE CONSERVATION
89
nor evaluated relationships among the members of any set of candidate MPAs, nor the concept of what
constitutes a ‘Network’ of MPAs. In order to complete the planning process and define a network of
MPAs, we must consider processes, such as those comprising ‘ecological integrity’, as well as structures at
the ecosystem level (e.g. Zacharias and Roff, 2000). Our goal is to try to define the procedures, involving the
fewest number of arbitrary decisions, that would lead to the optimal selection of regional, national and
international networks of MPAs. In future communications, we shall continue to define the procedures that
can lead to such a goal and afford sustainable protection to, at least, selected areas of the marine
environment.
ACKNOWLEDGEMENTS
Our thanks to Cory Lavoie for assistance in computer mapping and contributions to analysis of habitat
heterogeneity, and to Hussein Alidina for mapping and GIS interpretations. JCR thanks the speakers and
delegates to the SNIFFER conference in Edinburgh November 2000, for their thoughtful presentations and
contributions which helped greatly to determine the contents of this paper.
REFERENCES
Allen TFH, Starr TB. 1982. Hierarchy: Perspectives for Ecological Complexity. University of Chicago Press: Chicago.
Atlantic Geoscience Centre. 1991. Intoduction 2: bathymetry and physiographic features. In East Coast Basin Atlas
Series: Scotian Shelf. Atlantic Geoscience Centre, Geological Survey of Canada, p. 3.
Ballard RD. 1977. Notes on a major oceanographic find. Oceanus 20: 35–44.
Barnes RSK, Hughes RN. 1982. An Introduction to Marine Ecology. Blackwell Scientific: Oxford.
Bergeron P, Bourget E. 1986. Shore topography and spatial partitioning of crevice refuges by sessile epibenthos in an ice
disturbed environment. Marine Ecology Progress Series 28: 129–145.
Bowen WD. 1997. Role of marine mammals in aquatic ecosystems. Marine Ecology Progress Series 158: 267–274.
Boyer TP, Levitus S. 1997. Objective analyses of temperature and salinity for the world ocean on a 1/4 degree grid.
NOAA Atlas NESDIS 11.
Brown SK, Buja KR, Jury SH, Monaco ME, Banner A. 1997. Habitat Suitability Index Models for Casco and Sheepscot
Bays, Maine. Silver Spring, MD: National Oceanic and Atmospheric Administration, and Falmouth, ME: U.S. Fish
and Wildlife Service, 86 pp.
Connor DW. 1997. Marine Biotope Classification for Britain and Ireland. Joint Nature Conservation Review,
Peterborough, UK.
Connor DW, Hiscock K, Foster-Smith RL, Covey R. 1993. A classification system for benthic marine biotopes. In
Biology and Ecology of Shallow Coastal Waters, Eleftheriou A, Ansell AD, Smith CJ (eds), Proceedings of the 28th
European Marine Biology Symposium. Institute of Marine Biology Crete, Iraklio, Crete. Olsen and Olsen
Fredensborg; 155–165.
Davis DS, Stewart PL, Loucks RH, Browne S. 1994. Development of a biophysical classification of offshore regions for
the Nova Scotia continental shelf. Coastal Zone Canada 94, Halifax; 2149–2157.
Day J, Roff JC. 2000. Planning for Representative Marine Protected Areas: A Framework for Canada’s Oceans.
Report prepared for World Wildlife Fund Canada, Toronto.
Dethier MN. 1992. Classifying marine and estuarine natural communities: an alternative to the Cowardin system.
Natural Areas Journal 12: 90–99.
Fader GBJ. 1991. Surficial geology and physical properties 5: surficial geology. In East Coast Basin Atlas Series: Scotian
Shelf, Atlantic Geoscience Centre, Geological Survey of Canada, p. 119.
Foster-Smith RL, Davies J, Sotheran I. 2000. Broad scale remote survey and mapping of the sublittoral habitats and
biota: technical report of the Broadscale Mapping Project. Scottish Natural Heritage Research, Survey and
Monitoring Report No. 167.
Glenmarec M. 1973. The benthic communities of the European North Atlantic continental shelf. Oceanography and
Marine Biology. An Annual Review 11: 263–289.
Grassle JF, Sanders HL. 1973. Life histories and the role of disturbance. Deep-Sea Research 20: 643–659.
Grosholz ED, Ruiz GM, Dean CA, Shirley KA, Maron JL, Connors PG. 2000. The impacts of a nonindigenous marine
predator in a California bay. Ecology 81: 1206–1224.
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
90
J.C. ROFF ET AL.
Hallfors G, Niemi A, Ackefors H, Lassig J, Leppakoski E. 1981. Biological oceanography (chapter 5). In The Baltic
Sea, Voipio A (ed.) Elsevier Oceanography Series 30, Elsevier: Amsterdam.
Iles TD, Sinclair M. 1982. Atlantic herring: stock discreteness and abundance. Science 215: 627–633.
Jeffrey JS, Taggart CT. 1999. Growth variation and water mass associations of larval silver hake (Merluccius bilinearis)
on the Scotian Shelf. Canadian Journal of Fisheries and Aquatic Sciences 57: 1728–1738.
Katona S, Whitehead H. 1988. Are Cetacea ecologically important? Oceanography and Marine Biology. An Annual
Review 26: 553–568.
Longhurst A. 1998. Ecological Geography of the Sea. Academic Press: London.
McLusky DS, McIntyre AD. 1988. Characteristics of the benthic fauna. In Ecosystems of the World 27: Continental
Shelves, Postma H, Zijlstra JJ (eds). Elsevier: Amsterdam; 131–154.
National Atlas Information Service. 1993. CANADA SEA ICE The National Atlas of Canada, 5th edn. Canada Center
for Mapping, Energy, Mines and Resources: Canada.
Petrie B, Drinkwater K, Gregory D, Pettipas R, Sandstrom A. 1996. Temperature and salinity atlas for the Scotian
Shelf and Gulf of Maine. Ocean Sciences Division, Maritimes Region Department of Fisheries and Oceans and the
Bedford Institute of Oceanography. Canadian Technical Report of Hydrography and Ocean Sciences, No. 171,
398pp.
Pingree RD. 1978. Mixing and stabilisation of phytoplankton distributions on the Northwest European shelf. In Spatial
Patterns in Plankton Communities, Steele JH (ed.). Plenum Press: New York; 181–220.
Piper DJ. 1991. Surficial geology and physical properties 6: deep water surficial geology. In East Coast Basin Atlas
Series: Scotian Shelf. Atlantic Geoscience Centre: Geological Survey of Canada, p. 121.
Roff JC, Evans S. 2002. Frameworks for marine conservation } non-hierarchical approaches and distinctive habitats.
Aquatic Conservation: Marine and Freshwater Ecosystems 12: 635–648.
Roff JC, Legendre L. 1986. Physico-chemical and biological oceanography of Hudson Bay (chapter 14). In Canadian
Inland Seas, Martini IP (ed.), Elsevier Oceanography Series. Elsevier: Amsterdam.
Roff JC, Taylor ME. 2000. National frameworks for marine conservation } a hierarchical geophysical approach.
Aquatic Conservation: Marine and Freshwater Ecosystems 10: 209–223.
Scheibling RE, Hennigar AW, Balch T. 1999. Destructive grazing, epiphytism, and disease: the dynamics of sea urchin
} kelp interactions in Nova Scotia. Canadian Journal of Fisheries and Aquatic Sciences 56: 2300–2314.
Sherman K, Alexander LM. 1989. Biomass yields and geography of large marine ecosystems. American Association of
Advanced Science Selected Symposia 111: 1–493.
Simberloff D. 1998. Flagships, umbrellas, and keystones: is single-species management pass!e in the landscape era?
Biological Conservation 83: 2247–257.
Subba Rao DV, Platt T. 1984. Primary production of Arctic waters. Polar Biology 3: 191–201.
Tremblay MJ, Roff JC. 1983. Community gradients in the Scotian Shelf zooplankton. Canadian Journal of Fisheries and
Aquatic Sciences 40: 598–611.
Warwick RM, Clarke KR. 1994. Relearning the ABC: taxonomic changes and abundance/biomass relationships in
disturbed benthic communities. Marine Biology 118: 739–744.
Welch DM. 1978. Land/Water Classification. A review of water classifications and proposals for water integration into
ecological land classification. Ecological Land Classification Series, No. 5. Lands Directorate, Environment Canada:
Ottawa.
Wildish DJ. 1977. Factors controlling marine and estuarine sublittoral macrofauna. Helgoland Wissenschaft
Meeresuntersuchungen 30: 445–454.
Zacharias MA, Howes DE, Harper JR, Wainwright P. 1998. The British Columbia marine ecosystem classification:
rationale, development, and verification. Coastal Management 26: 105–124.
Zacharias MA, Roff JC. 2000. Conserving marine biodiversity: a hierarchical approach. Conservation Biology 14: 1327–
1334.
Zacharias MA, Roff JC. 2001a. Use of focal species in marine conservation and management: a review and critique.
Aquatic Conservation: Marine and Freshwater Ecosystems 11: 59–76.
Zacharias MA, Roff JC. 2001b. Explanation of patterns of intertidal diversity at regional scales. Journal of
Biogeography 28: 471–483.
Copyright # 2003 John Wiley & Sons, Ltd.
Aquatic Conserv: Mar. Freshw. Ecosyst. 13: 77–90 (2003)
Документ
Категория
Без категории
Просмотров
2
Размер файла
942 Кб
Теги
monitoring, habitat, approach, delineation, classification, communities, geophysics, marina
1/--страниц
Пожаловаться на содержимое документа