Comments on Modifiers and Attributes

Document Sample
Comments on Modifiers and Attributes Powered By Docstoc
					    Pilot Testing of the Coastal and Marine Ecological Classification Standard –
                                     Version III
                                  Project Abstracts
                                          9 April 2010


Pilot testing of any new classification system is essential to examine the logical relationships
between the system elements, identify areas of confusion, and test the applicability of the system
for mapping and data collection. Throughout its development history CMECS has been tested in
various geographies using different methods. The results have and will continue to inform the
refinement of the system. The pilots have also demonstrated how well CMECS relates (cross-
walks) to other systems in common use.

Following are short summaries of the principle pilots conducted through March 2010. In most
cases the results have been included in the April 2010 version of CMECS. The results of the
ShoreZone pilot have not yet been incorporated but will guide the next round of revisions.

       CMECS Pilot Project                           Type of Project

       Florida Bay, Florida                          Cross-walk from SCHEME
       Muir Inlet, Alaska                            Development from source data
       Long Island South Shore, New York             Cross-walk from SCHEME
       Northern Gulf of Mexico, Mississippi          Development from source data
       Redfish Bay, Texas                            Cross-walk from SCHEME
       Southeast Alaska                              Cross-walk from ShoreZone system




                                                1
                              CMECS Pilot Study Abstract
Name of Pilot Project: Florida Bay-Florida Keys, FL
Geography: Monroe County, Florida
Author: Christopher J. Madden, Kathy Goodin, Robert Solomon
Affiliation: NatureServe
Reference: Madden, Christopher J., Kathleen L. Goodin and Robert Solomon. 2008. Ecological
Classification of Florida Bay Using the Coastal Marine Ecological Classification Standard
(CMECS) Version III. NatureServe, Arlington, Virginia. 40 pp.




     Florida Bay SCHEME polygons cross-walked into the CMECS benthic cover classes.


Objective: This pilot was undertaken to test the ability of an advanced version of CMECS IIb
(Madden et al. 2006) to adequately classify a subtropical seagrass-mangrove estuarine
ecosystem. It was also a test crosswalk and conversion of previously classified scenes in a south
Florida estuary using the SCHEME classification (Madley et al. 2002) and a classified benthic
type map. The pilot project applied and further developed the CMECS concept of a multi-
component classification including the water column and surface geology using multiple data
and map sources. It also tested the ability to accommodate a locally-targeted classification and
application within the more comprehensive framework of CMECS.

Methods Overview: The classification and map were based on previously acquired imagery of
Florida Bay. Avineon Inc., of Clearwater, Florida produced an ARC/INFO shape file of

                                                2
Submersed Aquatic Vegetation (SAV) and other selected cover types within the Florida Bay
study area for the Florida Fish and Wildlife Conservation Commission (Madley et al 2006). The
classification system was adapted from the Florida SCHEME (Madley et al. 2002). Additional
information from a mapping study of bottom types of Florida Bay (Prager and Halley 1997) was
merged with the SCHEME data to resolve substrates beneath bottom cover visualized in the
photo-interpreted maps. Ancillary information was used from a reef mapping study from Jaap
and Hadlock (1990) and depth contour maps from Kourafelou et al. (2005) to create preliminary
maps of higher level CMECS Systems for Florida. Deep water Systems were distinguished on
depth contours taken from a bathymetric map of the seabed around south Florida to delineate
shallow Systems- Estuarine, Freshwater Influenced and Marine (formerly Nearshore)- from the
Neritic System at the 30 m contour and the Neritic from the Oceanic System at the shelf break
(average 200 m contour).

Version of CMECS used: Version IIb (July 2007).

Purpose of Original Studies:
Benthic mapping of Florida Bay habitats, bottom types and substrate/sediment types.

References for Original Studies:
Jaap, W. C. and P. Hadlock. 1990. Coral Reefs. pp. 574-618 in Myers, R. L., Ewel, J. S. [eds] Ecosystems of
          Florida, USA. Univ. of Central Florida Press, Orlando, FL.
Kourafalou, V.H., R.S. Balotro and T.N. Lee. 2005. The SoFLA-HYCOM (South Florida HYCOM) Regional
          Model around the Straits of Florida, Florida Bay and the Florida Keys. UM/RSMAS Tech. Rep. 2005-03,
          28 pp.
Madden, C. J., D. H. Grossman and K. L. Goodin. 2006. Coastal and Marine Systems of North America: A
          framework for a coastal and marine ecological classification standard. Version II. NatureServe. Arlington,
          VA. 48 pp.
Madley, Kevin et al. 2002. Florida System for Classification of Habitats in Estuarine and Marine Environments
          (SCHEME). Florida Fish and Wildlife Conservation Commission Florida Marine Research Institute.
Madley, K., J. Burd, N. Morton, P. Carlson and K. O‟Keife. 2006. Florida Bay Seagrass Mapping Project.
          Submitted to: SFWMD Under Agreement No. C-C20302A, FWC File Code F2391-03-F06 A. Huffman,
          Project Manager. 28 pp. plus seven appendices.
Prager, E. and R. B. Halley, 1997. Florida Bay Bottom Types Map USGS Open-File Report OFR 97-526 St.
          Petersburg, FL USGS, Center for Coastal and Regional Marine Studies
          http://sofia.usgs.gov/publications/ofr/97-526.

Technology Used for Initial Data:
Photo-interpretation of 1:24,000 scale natural color aerial photographs was used to develop and
classify a map using field reconnaissance to ground truth and resolve boundaries.

Standard crosswalked to:
SCHEME- The System for Classification of Habitats in Estuarine and Marine Environments
(Madley et al. 2002) was modified by Avineon for the source map and database.

Results/Products Overview:
The project produced a reclassified benthic cover map and a set of individual maps for each
CMECS component in a GIS which could be overlaid on a base map of Florida Bay. It also
produced a detailed report with additional analysis and a CMCES-SCHEME crosswalk.


                                                        3
Lessons Learned:
      Advantages of using CMECS:
          CMECS units coincided generally with or were easily cross-walked to source map
           classification units
          Little re-delineation of the map polygons was required
          Generally usable cross-walk with few ambiguities
          CMECS units were consistent and ecologically relevant to CMECS goals
          Additional type definition enabled by CMECS provides some resolution of water
           column units in the vertical dimension and underlying substrate
          Addition source material merged with primary classified map provided additional
           information about benthic cover units, reef types, underlying surface substrate type
           and emergent vegetation types that complemented the primary information.
           CMECS accommodated all additional information easily

       Limitations of CMECS:
          Difficult to differentiate Sub-tidal from Inter-tidal given available imagery
          Difficult to differentiate Unconsolidated shore vs. Unconsolidated bottom
            Impossible to determine bottom types when obscured by turbid water in imagery
          Difficult to determine some geoform types given available imagery and data
          Continuous and discontinuous cover modifier categories are different between
            SCHEME and CMECS and could not be resolved without precise percent cover

Recommendations for Further Testing:
This pilot demonstrated CMECS enabled higher resolution of ecological units and classified
types than the local classification systems. Protocols for implementing multiple data sources in
producing CMECS maps should be developed. CMECS could also benefit from further testing
and development of major components, specifically the emerging water column geoform
components of the classification. Development of application guidance and mapping guidance
for CMECS is needed and will be promoted and enhanced by further pilot testing.




                                                4
                              CMECS Pilot Study Abstract
Name of Pilot Project: Muir Inlet
Geography: Glacier Bay, Alaska
Author: Guy Cochrane
Affiliation: USGS Coastal and Marine Geology
Reference: Trusel, L.D., G.R. Cochrane, L.L. Etherington, R.D. Powell, and L.A. Mayer, in
press. Marine Benthic Habitat Mapping of Muir Inlet, Glacier Bay National Park, Alaska. USGS
Scientific Investigations Map. In press.




                    Location of Muir Inlet in Glacier Bay National Park, Alaska


Objective:
This study is a bottom-up approach to mapping benthic habitats (cf. Greene et al. 2007) of Muir
Inlet in Glacier Bay National Park (Figure 1). Seafloor substrate and morphology are

                                                5
characterized so that they may be correlated to and associated with a specific set of benthic
conditions favorable for local biota. These include both commercially and recreationally
important species such as king crabs (Paralithodes camtschaticus), Tanner crabs (Chionoecetes
bairdi) and Pacific halibut (Hippoglossus stenolepis) (Mondragon et al. 2007a, Mondragon et al.
2007b). Seafloor substrates and therefore potential benthic habitats were mapped with depths
ranging from just below the surface to greater than 300 m in the deepest fjord basins. The
substrate was characterized by using a combination of high-resolution multibeam bathymetry
and backscatter imagery, numerous groundtruthing sources, and knowledge of the local fjord
environment.

Methods Overview:
The multibeam bathymetric and acoustic backscatter data were compiled with additional
information in an ESRI™ Geographic Information Systems (ArcGIS™) database to create several
map products. The National Oceanic and Atmospheric Administration (NOAA) and
NatureServe‟s Coastal and Marine Ecological Classification Standard (CMECS) was used to
classify seafloor substrate, potential benthic habitats, water column properties, and seafloor
geomorphology (cf. Madden et al. 2008). An additional map of modern sediment flux
measurements using quantitative differential bathymetry is also presented. These maps serve as
the first high-resolution bathymetric surveys for Muir Inlet, as baselines for further research, and
have implications for park management.


Version of CMECS used:
Version III - July 2008

If based on a previous study
Purpose of Original Study:
N/A

Reference for Original Study:
N/A

Technology Used for Initial Data:
Benthic habitat and seafloor substrate mapping for Muir Inlet were conducted using multibeam
sonar data that were collected onboard the R/V Maurice Ewing cruise EW0408 on September 7,
2004. A Kongsberg Simrad EM-1002 multibeam echosounder was installed on the Ewing for the
purposes of high-resolution bathymetric surveying specific to the shallow to intermediate depth
of the southern Alaska study locations. The EM-1002 simultaneously collects both bathymetric
and co-registered backscatter data, providing information about seafloor morphology and
composition.
Multiple resources were utilized to aid in classifying the seafloor substrate. These groundtruthing
sources include USGS seafloor video observations (Harney et al. 2007), sediment samples and
cores (Cowan et al. 1999, Hunter et al. 1994, Jackolski et al. 2006), seismic reflection profiles
(Molnia et al. 1984, Seramur et al. 1997, Cowan et al. 1999), seafloor dive observations from
other habitat investigations (Stone et al. 2005), abundant aerial and ground-based photography.


                                                 6
Standard crosswalked from:
N/A

Results/Products Overview:
Describe and attached the products (eg. a reclassified benthic cover map; individual maps for
each component, a detailed crosswalk table, detailed report with additional analysis, etc.).

The product, a USGS Scientific Investigation Map, is in press. Complete results and products
will be found in Trusel, L.D., G.R. Cochrane, L.L. Etherington, R.D. Powell, and L.A. Mayer, in
press. Marine Benthic Habitat Mapping of Muir Inlet, Glacier Bay National Park, Alaska. USGS
Scientific Investigations Map.

Coverage for the BCC substrate classification: As expected, unconsolidated bottom types
dominate the Muir Inlet seascape, owing to the very large glacimarine sediment flux. The largest
measured area of Muir Inlet was the mesobenthic depth zone (Table 3), within which exist the
majority of the basin floors (Table 7), a primary reason for the high proportion of unconsolidated
bottom. Nearly 90% of Muir Inlet is covered by mud, far more any other bottom type. The
second most abundant characterized substrate in Muir Inlet is bedrock, which comprises about
6% of mapped substrate. Mixed sediments, which are likely to be a mix of sand- and gravel-sized
particles, account for 3.2% of the measured substrates. Boulder / rubble was characterized in
1.3% or nearly 1 km2 of Muir Inlet. The cobble / gravel subclass could only be confidently
placed over an area of 0.29 km2

Table 1. Substrate distribution by depth zone for Muir Inlet.

                  Deep Infralittoral:       Circalittoral:        Circalittoral        Mesobenthic:
                                                              (offshore): 80-200m    200-1000m water
                 5-30m water depth      30-80m water depth        water depth             depth
    CMECS        Percent Area)          Percent Area)        Percent Area)          Percent Area)
   subclass                (km2                     (km2                (km2                 (km2

Mud              0.05       0.04        3.53     2.57        28.79     20.95        56.35    41.00
Mixed
                 0          0           0.11     0.08        1.94      1.41         1.26     0.91
Sediments
Cobble /
                 0          0           0.03     0.02        0.34      0.25         0.03     0.02
Gravel
Boulder /
                 0          0           0.01     0.01        0.56      0.41         0.79     0.57
Rubble
Bedrock          0.01       0.01        0.43     0.31        3.95      2.88         1.83     1.33


Muir Inlet lies within the „continental margin‟ megageoform and is also a „fjord‟ megageoform.
For simplicity, we characterized geoforms down to the mesogeoform scale (10s meters to
kilometers in size). Additionally, the resolution of our data was a limiting factor in geoform
characterization. The primary mesogeoforms characterized within Muir Inlet are fjord wall,
floor, delta, and moraine geoforms, The high relative proportion of walls and floors is expected
and is characteristic of glacial fjord morphology. Numerous deltas exist along the fjord walls that
are both fluvial and glacifluvial in origin, with most at least partially fed by glacial meltwater.

                                                   7
Table 2. Major geoform distribution for Muir Inlet.

              Megageoform          Mesogeoform         Percentage of           Area (km2)
                                                         total area

                                        Delta               5.8                     4.6
                                        Floor              38.7                     30.7
                 Fjord
                                     Moraine               15.5                     12.3
                                        Wall               40.0                     31.7



Table 3. Major geoform distribution by depth zone for Muir Inlet.

                  Deep Infralittoral:        Circalittoral:            Circalittoral          Mesobenthic:
                                                                   (offshore): 80-200m      200-1000m water
                 5-30 m water depth       30-80m water depth           water depth               depth
    CMECS        Percent Area)            Percent Area)           Percent Area)            Percent Area)
   subclass                (km2                      (km2                    (km2                   (km2

Delta            0          0             0.37     0.29           3.68      2.92           1.80     1.43
Floor            0          0             0        0              0.69      0.55           38.0     30.13
Moraine          0          0             1.38     1.10           10.20     8.08           3.93     3.11
Wall             0.05       0.04          2.67     2.11           22.89     18.15          14.36    11.39


As with the benthic cover and geoform components, the water column component is classified
within the „estuarine‟ system. Benthic depth zones associated with oceanographic stations in
Muir Inlet are classified as „circalittoral offshore‟ (80-200 m) and „mesobenthic‟ (200-1000 m).
The upper water column within Muir Inlet (above pycnocline) is always in the „epipelagic‟ water
column depth zone (>0-200 m), while the lower water column varies between „epipelagic‟ and
„mesopelagic‟ (200-1000 m).




                                                       8
                       Maps of (A) upper and (B) bottom water column CMECS habitats.


Lessons Learned:
Glacier Bay is a good test of CMECS because of the large temporal and physical variability of
the environment at many periods and scales.


Overall, Glacier Bay had a comprehensive (both temporal and spatial) data set of attributes to
test the application of CMECS as opposed to Columbia River estuary testing done for version II
by Keefer et al. (2008). Despite the comprehensive data set in Glacier Bay, classifying many of
the attributes was very difficult. This raises the question: what happens in areas that do not have
very good or comprehensive attribute availability?




                                                    9
In Glacier Bay, CMECS is potentially very helpful for research applications in explaining spatio-
temporal variability in invertebrate populations; could help identify and locate resources
(geologic, biotic) that are present; could be used to map out no-fishing zones (crab fisheries).
Having three separate components, separate layers and different databases, is acceptable. But is
it practical in terms of quick integrated access and analysis? The specifics of the database
structure need to be better established to facilitate GIS analysis. The benthic cover component is
hierarchical (each attribute is nested in a higher level); the water column component is not
hierarchical, and attributes are independent of one another; geoform component is not
hierarchical and attributes are independent of one another (however, they do progress from large
to small scales). Therefore, the three components are put together differently. Does this cause
any database or GIS problems?
A CMECS database should include a time stamp and a mechanism for updates. Within CMECS,
data are entered as an inventory; however, one goal of mapping is to identify temporal change.
Within CMECS, is there a mechanism in the database that allows data to be updated or added
temporally?
A CMECS map would need a companion document to help interpret information created with
classification. What does a combination of classes mean ecologically? The biotope class does not
fully describe unique ecosystems present in a mapped area. A given area might function
similarly to another though it has different values for several attributes. An accompanying
document that says areas are sensitive because they have soft corals, or a habitat type is tough to
impact (resiliency, risk), etc. Is this part of CMECS or would it be better to make it part of a
report that includes a CMECS inventory of attributes?
For sites that do not have abundant data, one should not be able to classify down to the finest
degree. There should be rules about the type and resolution of data required at each resolution of
detail.
Some of problems with CMECS II identified in Keefer et al. (2008) were carried on through to
Version III. Keefer et al. (2008) suggest that the CMECS could be improved by refining
classification thresholds to better reflect ecological processes, by direct integration of temporal
variability, and by more explicitly linking physical and biological processes with habitat patterns.
Specific examples in the Keefer et al. (2008) text:
    "A single physical descriptor - temporal persistence - was largely qualitative, which may at
    times be inadequate in highly dynamic estuarine systems like the CRE." (p. 102)
    "CMECS lacks a suitable framework to capture and classify this type of predictable
    variability" (p. 102)
    "Temporal variability should be more fully integrated into CMECS, perhaps as an additional
    category, nested within each hierarchical level" (p. 103)
    "We believe energy intensity should perhaps be further classified. Although our study sites
    had significantly different benthic habitats, apparently as a function of energy inputs, all but
    one were classified as 'low energy' using the CMECS intensity scale, suggesting that this
    scale was too coarse for differentiating estuarine habitats that structure in response to subtle
    but critical energy differences." (p. 103)



                                                10
Comments on the Benthic Cover Component:
Overall, there is concern regarding the choice of abiotic versus biotic cover type at a very high
level within the CMECS system because it drives all subsequent classifications. Why have the
split? One should be able to capture both abiotic and biotic components rather than either/or. If
cover is either biotic or abiotic, one is not able to assess the relationship between dominant
biological community (e.g. deep reef, coral garden, mussel beds) and substrate. Why lose the
abiotic information biotic cover type is chosen? This approach inhibits the ability to associate
abiotic cover to benthic and demersal organisms, an important process in defining preferred
benthic habitats.
If there is no information to determine either abiotic or biotic cover, there should be a field or
variable to state this. As currently defined, if a substrate has less than or greater than 10 % cover,
it is classified either abiotic or biotic, respectively. As an example, it can be envisioned that a
substrate may have a 25 % vegetation or epifaunal cover over an abiotic cover of 75 %. Biotic
cover may be greater than 10 %, but it may not be dominant (as the cover type choice is defined).
Why is 10 % considered the dividing point? Rather than having a cut-off value (classifier) for
biotic cover (at 10 %), have modifier of percent cover.
 There are some issues and need for guidelines for implementation of CMECS in GIS (i.e. can a
polygon have both biotic and a-biotic elements?). If biotic cover is observed at a point sample,
how far out from that point can you extrapolate into a polygon, or should CMECS be restricted
to point data without inferences. How do you make a biotic map? Is the final CMECS product a
point shapefile biotic data layer over continuous abiotic polygons?
The definition of unconsolidated cover was vague. There is unconsolidated material that is hard
but not rock. Specific to regions influenced by glaciation, consolidated glacigenic sediments
(diamicts/till) are likely to exist. Within CMECS currently, it is unclear how this should be
classified (is it to be classified in the rock bottom class?). We feel there are not enough substrate
classes to distinguish important ecosystems. We suggest three: rock, coarse sediment, and fine
sediment, as these have been found to be the most important to benthic habitats (Greene et al.
2007; Cochrane 2008).
If the importance of biotic cover drives the benthic cover component classification, collecting
biotic data would be a top priority for any agency applying the CMECS system; having an
“unknown” status that high up in the hierarchy is problematic. Collection of biotic data over
wide spatial scales is an immense task, especially in Glacier Bay. If remote sensing provides
only an abiotic map, is it a CMECS map?
Habitat maps should be a useful product, such as reflecting the composition of species, for
example. What should a manager expect to find biologically based on depth, bottom, etc?
CMECS may meet these criteria in biotic cover or biotope, but its utility is uncertain based on
the biotic/abiotic classification fork. And those elements of CMECS do not generate species
composition. Biotope seems to be specific to organisms attached to substrate. Therefore, to some
degree, classes should be more detailed.
Based on our understanding of biotope, benthic biotopes identified for Glacier Bay include
Modiolus sp. mussel beds, deep coral/sponge mosaic, and Nereocystis sp. kelp beds.




                                                 11
Comments on the Water Column Component:
The greatest challenge within the WCC is how to deal with temporal change. The one attribute of
„temporal persistence‟ is not sufficient to capture patterns of temporal variability among multiple
aspects of the system. For the Glacier Bay pilot effort the average values of attributes across all
seasons and all conditions (neap and spring tide conditions) were used, averaging a highly
variable system. Averaging of data lost some of the most important sources of variability and
made the system appear homogenous, despite known strong spatial gradients along the Bay axis.
Averaging caused all of Glacier Bay to have the same energy intensity (all low); there are
extremes in energy levels throughout the Bay that are not captured. When turbidity levels were
averaged over the year (using photic depth instead of Sechhi depth), the CMECS class indicated
that glacially influenced areas of the park were defined as „clear‟ (5-20m Secchi depth). This
class and description does not reflect the highly turbid conditions within the Bay throughout
much of the year (Glacier Bay has some of the highest suspended sediment concentrations in the
world).
Possible solutions to the temporal aspect could be to map each of the seasons separately, given
the very large seasonality in water column properties (e.g. salinity, stratification, turbidity, light
levels, chlorophyll a) that drive biological patterns. Modifiers would be helpful to refine the
attributes (e.g. average was low, but highly variable). Capturing maximum values also would be
important.
For the Glacier Bay pilot project water column structure was used as a way to define vertical
zones; horizontal attributes were defined for each of the vertical depth zones (upper water
column layer, bottom water column layer). Water column depth zones did not capture important
vertical variation in the water column (i.e. dynamic surface layers). Specific guidance on how to
use vertical attributes (water column depth zones; water column structure) and how to integrate
with horizontal attributes are needed. How are these attributes combined in a database?
Several of the main attributes of the water column component were not populated because it was
unclear what was meant by the macro- and meso-hydroforms. For example, how do you define
an effluent, a small freshwater lens, or a convergence? What boundaries and scales should be
used to define these mesohydroforms? Can one only choose one mesohydroform or can one
define multiple mesohydroforms in the same space?
The lifeforms attribute was not very useful as none besides phytoplankton maximum layer are
observed in Glacier Bay. The phytoplankton maximum layer was assigned to the upper water
column, which was not very useful. No biotopes were defined within the water column. What
else would be considered a biotope (as it is currently defined) besides the example of Sargassum
mats that is given?

Comments on the Geoform Component:
More than one kind (mega/mesogeoform) of fjord is desirable (e.g. Glacier Bay is a temperate
glacial fjord). It may be useful for the classification if rules define how fjords differ by using
data inputs to help define how the fjords are different. Alternatively, the WCC could be used to
define the type of fjord, rather than explicitly having a category for fjord type given that there are
only temperate fjords in the U.S.
Spatial scales and nesting hierarchy need to be better defined. Glacier Bay is a fjord
megageoform within the Continental Margin megageoform. Based on the CMECS geoform scale

                                                  12
definitions. Defining morphologic features based on scales that overlap is problematic and
subjective. Example: megageoform is anything “kilometers to 10s of kilometers and larger”.
Mesogeoform is anything “10s of meters to kilometers”. Certain morphologic features may be
classified as either or both because there is overlap between the scales. Setting specific cut-offs
between the scales would make this process less subjective. Considering the goal of a nation-
wide classification standard where geoforms and their associated habitats may be compared,
defining specific scale boundaries would be useful.
The documentation is not clear when talking of scale of geoform. Are geoforms based on a linear
scale or area? If it is a linear scale, should the longest axis be used? For example, a very narrow
but long channel may not cover much area and may fit into two geoform classes depending on
which axis is measured if it is defined linearly. If geoforms are defined by areal extent, which
may be the best option, the scales should be re-defined. For example, mesogeoform is defined as
“10s of meters – kilometers”. If this is an area, this equates to 1.0x10-5 km2 to 1+ km2 and nearly
every morphologic feature would fit into the mesogeoform category.
Geoforms need to be expanded to include sills, which are common in glacial fjords. While
“shoal” works, sill would be more appropriate in Glacier Bay. If it is acceptable to add new
geoforms, it should be explicitly stated in the framework. The geoform „ice feature‟ needs to be
defined. The term could mean several or all things in a glacial fjord.
Bathymetric Position Index (BPI) is not captured in geoform component although it is
considered very useful. BPI should be calculated at multiple scales (small to large). Same value
may be defining a small rise or an extensive bank depending upon scale. Rules would be needed
to set radius. It would be useful to define the scales (that are useful at applicable; e.g. scales of
fish species may be important but different from other faunal community component). BPI could
also be an included as a modifier attribute.

Comments on Modifiers and Attributes:
Substrate is not included as a modifier attribute, therefore, if a biotic subclass is chosen, there is
no place for abiotic substrate information. Groundtruthing methods used by USGS and NOAA
include observations of primary substrate (>50%) and secondary substrate (>20%).
It would be useful to add an attribute for water column stratification. Stratification is both
important to species within the water column and for benthic organisms at shallow depth in
glaciated fjords and estuarine environments.
CMECS calls for Secchi depth as a measurement of turbidity. Secchi disks are not often used for
oceanographic measurements and these values should be changed to reflect optical
backscatterance and calculated concentrations of sediment within the water column.
Alternatively, euphotic depth levels could be used (i.e. where surface light levels become
minimal).
It appears that for several of the attributes, the classes need to be refined. For example, Glacier
Bay fell into the „moderate‟ category for tide range (1 – 5 m). The Bay has one of the largest
tidal fluxes in the world (not very many places that would go into the „large‟ tide range class,
except the Bay of Fundy). This would suggest that the classes need to be further refined or a
series of subclasses should be added to the classification scheme. Also, the classes within energy
intensity need to be refined to include more classes.


                                                  13
To reiterate what was highlighted in Keefer et al. 2008, we believe energy intensity should be
further classified. Although our study sites had significantly different benthic habitats (e.g.
cobble/sand versus mud), apparently as a function of energy levels, they were similarly classified
as 'low energy' using the CMECS intensity scale, suggesting that this scale was too coarse for
differentiating estuarine habitats. Similarly, the energy intensity classes were not able to
differentiate between very different upper water column energy regimes within the Bay. Part of
this problem could be due to averaging values to assign only one class for the attribute.
Sedimentation rate would be a useful addition to either the BCC, WCC, or as a modifier. High
clastic sediment fluxes to the benthos in glaciated fjords have been found to be an important
physical control on the distribution of biota (e.g. Carney et al. 1999; Wlodarska-Kowalczuk and
Pearson 2004). Sediment flux is potentially a dominant and limiting factor that defines benthic
habitats for some organisms. This is relevant in fjords and other areas with high sediment fluxes.

Recommendations for Further Testing:
The latest version of CMECS (August 2009) addressed some of the problems described above.




                                                14
                              CMECS Pilot Study Abstract
Name of Pilot Project: Long Island South Shore
Geography: Long Island, New York
Author: Mark Finkbeiner
Affiliation: NOAA Coastal Services Center
Reference: N/A




      The two Long Island South Shore pilot areas are shown in red. The western area is a
      more turbid system dominated by macroalgae, while the eastern site is
      predominantly seagrass.


Objective:
Cross-walk an existing data set into the CMECS classification in two representative areas along
the LI south shore. Portions of the Aquatic Setting, Biotic Cover (BCC), and Surficial Geology
(SGC) components were tested.

Methods Overview:
Existing SCHEME habitat polygons were overlain in the source ortho-photograhy. The
SCHEME attributes were translated into CMECS units by comparing unit definitions. No new
delineation was attempted. Every habitat polygon had to be assigned a CMECS attribute.




                                               15
Version of CMECS used (Version # and Date):
Version III - July 2008

If based on a previous study
Purpose of Original Study:
The original mapping was done to establish a baseline habitat inventory (especially seagrass) for
Long Island‟s south shore estuaries and to support future change detection efforts. The data
were incorporated into the South Shore Estuary Reserve‟s comprehensive plan.

Reference for Original Study:
http://www.cscs.noaa.gov/benthic/data/northeast/longisl.html
http://www.estuary.cog.ny.us/resource/benthic.htm

Technology Used for Initial Data:
Analog aerial photography supported by field work using underwater videography.

Standard crosswalked from:
The Florida System for Classifying Habitats in Estuarine and Marine Environments (SCHEME).
This system was designed for the types of mapping done on Long Island and is designed with
aerial imagery as the principle data source.
http://research.myfwc.com/features/view_article.asp?id=24987

Results/Products Overview:
Digital vector polygon data (ESRI shapefile) with CMECS unit attributes, usually to the subclass
level in the BCC, and the class level in SGC. The percent cover modifier was used, as was the
system level in the Aquatic Setting.
The output product could be considered a “semi-integrated” CMECS data set as one or more
components (but not all) were compiled into the same data set.

Lessons Learned:
Generally the SCHEME system cross-walks easily into the CMECS BCC and SGC components
although it does not populate either component completely. Despite the relative equivalency of
SCHEME and BCC/SGC, engagement with the user is still needed to provide background on the
CMECS definitions and to assist in the cross-walk.
The implications for other CMECS work are that the Implementation Team should work with all
pilots testers to ensure a successful outcome.

       Advantages of using CMECS: (Bulleted list with enough detail to understand the issue)
             Hierarchical structure allowed aggregation upward to more general units
                where data did not support a more detailed
             Little re-delineation required.
             Overall smooth cross-walk. Few cases were unresolvable.


       Limitations of CMECS: (Bulleted list with enough detail to understand the issue)

                                               16
              Difficult to differentiate Sub-tidal from Inter-tidal given available imagery.
              User seemed to misunderstand distinctions between Estuarine, Fresh-water
               influenced, and marine systems. Check to see that documentation is clear.
              Difficult to differentiate Unconsolidated Shore vs. Unconsolidated Bottom with
               existing imagery.
              Patchy modifier categories different between SCHEME and CMECS. Could not
               be resolved without precise percent cover values.

Recommendations for Further Testing:
Guidance for discriminating intertidal from subtidal systems would be helpful. More language
may be needed to define the estuarine system, especially in areas with smooth transitions to tidal
and freshwater rivers.
The “shore” and “bottom” class designation names in the SGC may not be needed since the
Aquatic Setting - subsystem would cover this.




                                                17
                              CMECS Pilot Study Abstract

Name of Pilot Project: Northern Gulf of Mexico
Geography: Mobile Bay, Mississippi Sound and Southeast Louisiana
Author: Becky Allee
Affiliation: NOAA Gulf Coast Services Center
Reference: N/A

Objective:
This project used CMECS to classify sediment grab samples from areas throughout the Northern
Gulf of Mexico. Two CMECS components, the Biotic Cover Component and the Surface
Geology Component were used for the classification. The classified data were then used to create
a sediment map for the project area.

Methods Overview:
Sediment grab samples were collected during survey trips to map marine debris in the Northern
Gulf of Mexico after the hurricane season of 2005. The sediment grab samples were analyzed for
grain size, composition and total organic carbon at Louisiana State University, Department of
Environmental Services. Data analyses were provided in Excel spreadsheets and a column for
CMECS coding was added to the spreadsheet. Each sample point was given a CMECS code and
was then used to develop a GIS map.

Version of CMECS used:
Version III February 2009

If based on a previous study
Purpose of Original Study:
This pilot was not based on a previous study.

Reference for Original Study:

Technology Used for Initial Data:
Sediment grab sampling in a systematic grid pattern

Standard crosswalked to:

Results/Products Overview:

Lessons Learned:

       Advantages of using CMECS: (Bulleted list with enough detail to understand the issue)
             Straight forward, easy classification of the sediment data.
             There was no questionable classification of units.


                                                18
       Limitations of CMECS: (Bulleted list with enough detail to understand the issue)
           Sub-tidal could not be differentiated from Inter-tidal using the point data.
           The limitations were with the data, not CMECS.

Recommendations for Further Testing:
The point data should be interpolated in a GIS to produce a sediment layer map.




CMECS Surface Geology Component classification of sediment grab samples in Mobile Bay,
AL (left) and off coast of Dauphin Island, AL (Right). Coding: Es = Estuarine System; NS =
Nearshore1 = Subtidal Subsystem; UB = Unconsolidated Bottom; 2 = > 50% sand; 3 = > 50%
mud; 6 = mixed sediments.




                                              19
                   CMECS Surface Geology Component classification of
                   sediment grab samples around Mississippi Sound.
                   Coding: Es = Estuarine System; 1 = Subtidal Subsystem;
                   UB = Unconsolidated Bottom; 2 = > 50% sand; 3 = >
                   50% mud; 6 = mixed sediments.




CMECS Surface Geology Component classification of sediment grab samples around
Mississippi Sound (Left, upper), Breton Sound (Left, lower) and West Bay (Right), LA. Coding:
FI = Freshwater-influenced; 1 = Subtidal Subsystem; UB = Unconsolidated Bottom; 2 = > 50%
sand; 3 = > 50% mud; 6 = mixed sediments.


                                             20
                              CMECS Pilot Study Abstract
Name of Pilot Project: Redfish Bay
Geography: Redfish Bay, Texas
Author: Mark Finkbeiner
Affiliation: NOAA Coastal Services Center
Reference: N/A




              Harbor Island, Redfish Bay State Scientific Area Texas with
              SCHEME classes overlain on ADS-40 imagery. Greens are
              seagrass meadows, yellows are unconsolidated bottom areas, and
              magenta areas are mangroves.


Objective:
Cross-walk an existing data set into the CMECS classification in the Redfish Bay State Scientific
Area. Portions of the Aquatic Setting, Biotic Cover (BCC), and Surficial Geology (SGC)
components were tested.

Methods Overview:
Existing SCHEME habitat polygons were overlain in the source ortho-photograhy. The
SCHEME attributes were translated into CMECS units by comparing unit definitions. No new
delineation was attempted. Every habitat polygon had to be assigned a CMECS attribute.




                                               21
Version of CMECS used:
Version III - July 2008

If based on a previous study
Purpose of Original Study:
 The original mapping was done to establish a time 2 habitat inventory (especially seagrass) for
Texas‟s Coastal Bend estuaries to support examine habitat change and support the state‟s
Seagrass Monitoring Program.

Reference for Original Study:
http://www.cscs.noaa.gov/benthic/data/gulf/bend.html
http://www.tpwd.state.tx.us/landwater/water/habitats/seagrass/monitoring.phtml

Technology Used for Initial Data:
Digital multi-spectral imagery (ADS-40) supported by field work using underwater videography.

Standard crosswalked from:
The Florida System for Classifying Habitats in Estuarine and Marine Environments (SCHEME).
This system was designed for the types of mapping done on coastal Texas and is well-suited for
use where aerial imagery ais the principle data source.
http://research.myfwc.com/features/view_article.asp?id=24987

Results/Products Overview:
Digital vector polygon data (ESRI shapefile) with CMECS unit attributes, usually to the subclass
level in the BCC, and the class level in SGC. The percent cover modifier was used, as was the
system level in the Aquatic Setting.
The output product could be considered a “semi-integrated” CMECS data set as one or more
components (but not all) were compiled into the same data set.

Lessons Learned:
Generally the SCHEME system cross-walks easily into the CMECS BCC and SGC components
although it does not populate either component completely. Despite the relative equivalency of
SCHEME and BCC/SGC, engagement with the user is still needed to provide background on the
CMECS definitions and to assist in the cross-walk.
The implications for other CMECS work are that the Implementation Team should work with all
pilots testers to ensure a successful outcome.

       Advantages of using CMECS: (Bulleted list with enough detail to understand the issue)
             Hierarchical structure allowed aggregation upward to more general units
                where data did not support a more detailed
             Little re-delineation required.
             Overall a smooth cross-walk. Few cases were un-resolvable.


       Limitations of CMECS: (Bulleted list with enough detail to understand the issue)

                                               22
              Issue of differentiating Sub-tidal from Inter-tidal given available imagery is less
               of an issue in coastal Texas due to the limited tidal range, but still needs to be
               evaluated in CMECS. Mangroves in this area cross the line between intertidal
               and sub-tidal (permanently flooded) making this distinction somewhat difficult.
              Differentiating whether mangroves should be assigned to the Scrub-Shrub or
               Forested wetlands class is also somewhat problematic using this imagery.
              Difficult to differentiate Unconsolidated Shore vs. Unconsolidated Bottom with
               existing imagery. This is a larger problem further south in coastal Texas where
               vast expanses of mudflats are periodically submerged and exposed based on
               overall water level and wind events.
              Patchy modifier categories different between SCHEME and CMECS. Could not
               be resolved without precise percent cover values.

Recommendations for Further Testing:
As with other pilot results, more guidance for discriminating intertidal from subtidal systems
would be helpful. Although this wasn‟t mapped, the inlet in the southeast part of the study area
may be a case where boundaries need to be considered from estuarine and freshwater influenced.
The “shore” and “bottom” class designation names in the SGC are not needed since the Aquatic
Setting - subsystem would cover this.




                                                23
                                   CMECS Pilot Study Abstract

Name of Pilot Project: ShoreZone to CMECS Cross-Walk
Geography: Southeast Alaska
Author: John Harper
Affiliation: Coastal &
Ocean Resources Inc,,                                                              CMECS - ShoreZone
                              Kruzof North
Sidney, BC                       33 km                                              Cross-Walk Sites
                               N

Objective: This project
tested the potential for
                                                                                             Bara
                                                                                                 n
migrating ShoreZone data




                                                                                                      of
                                                  Kru




                                                                                                         I
to the new Coastal and                               zo




                                                                                                         s
                                                                                            Sitka Area
Marine Ecological                                                                             53 km




                                                            fI
Classification System


                                                              sla
(CMECS; Madden et al

                                                                 nd
2009). ShoreZone is a                                                                         Sitka
                                                                                              #
relatively mature coastal
habitat mapping system
with nearly 100,000 km
of contiguous coastline
                                                       Kruzof South
mapped in Alaska, British                                 35 km
Columbia and
Washington. If such a
data cross-walk is             5      0      5 Kilometers
possible, a substantial
CMECS dataset could be       Figure 1. Location of the three test sections of shoreline used in the pilot cross-
                             walk.
created from existing
data.

To test the cross-walk approach, three pilot areas from Sitka Sound were selected for the cross-
walk. These areas represented about 122 km of shoreline, 522 alongshore units and 1,966 across-
shore components. A variety of exposures, landforms, substrate, biota and salinity regimes are
represented within these three pilot sections.


Methods Overview: Several sentences describing How the pilot was done. General data
collection, crosswalking approach.

The ShoreZone dataset for each test section were extracted from the data master SE Alaska
ShoreZone dataset for use in the pilot. The three test sections were cross-walked mapping units
of ShoreZone component data to the Biotic Cover Component (BCC), Surficial Geology
Component (SGC) and Geoform Components (GFC) of CMECS. The primary mapping unit of
ShoreZone is the alongshore unit whereas the primary mapping unit of CMECS is the across-
shore component (Fig. 2). The alongshore units of ShoreZone do not translate to CMECS. While

                                                   24
this is a fundamental difference between the
two systems, the vast majority of the
ShoreZone data could be cross-walked as
the ShoreZone across-shore components
and the CEMCS mapping unit are
essentially equivalent.

Version of CMECS used (Version # and
Date):
Version III – 24 August 2009

Madden, C. J., K. Goodin, R.J. Allee, G.
  Cicchetti, C. Moses, M. Finkbeiner, D.
  Bamford, 2009. Coastal and Marine
  Ecological Classification Standard.
  NOAA and NatureServe. 107 p


If based on a previous study
Purpose of Original Study: (also provide          Figure 2. Schematic representation of ShoreZone
link and contact info if available):              mapping system, showing an alongshore unit and the
ShoreZone is a coastal habitat mapping            associated across shore components and biobands.
system that has been widely applied through
the Pacific Northwest, including the 5,000 km of Washington Coast, the 40,000 km of the BC
coast and over 40,000 km of the Alaska coast (to date). ShoreZone is a benthic coastal habitat
mapping system that was developed in the late 1970s (Owens 1980; Howes et al 1994) and
substantially revised in the early 1990s with the addition of biological mapping attributes
(Harper et al 1994; Searing and Frith 1995). The system is biased towards a mapping system,
rather than a classification system – most of the attributes are features that can actually be seen
on coastal. low-tide imagery (e.g., a pebble-sand beach berm) or directly inferred from the
imagery (e.g., coastal stability. ShoreZone is used for spill response planning, conservation
planning, habitat capability modeling and often, the web-posted low-tide imagery is the only
publicly available imagery (because daylight low tides on the west coast only occur early in the
morning).




                                                 25
Reference for Original Study:

Harney, J.N., Morris, M.C., and Harper, J.R. 2008. ShoreZone Coastal Habitat Mapping:
Protocol for the Gulf of Alaska. Contract Report by Coastal & Ocean Resources Inc. of Sidney,
BC for the Exxon Valdez Trustee Council, Anchorage, Alaska. 157 p.
http://alaskafisheries.noaa.gov/habitat/shorezone/szintro.htm

Technology Used for Initial Data:

The data source for ShoreZone is purpose-collected, low tide oblique and photographic coastal
imagery. Approximately 200 to 300 km of minus tide, oblique imagery are typically collected in
a single days overflight with a helicopter. The georeferenced imagery is then used to classify
coastal morphology, substrate and biota using a standardized protocol (Harney et al 2008). The
classification data are linked spatially to line segments to permit a variety of mapping formats.
The representation of ShoreZone by line segments makes the dataset cartographically simple to
implement.

Both physical features and biological features are cataloged within the ShoreZone system.
Physical features include items like beach berms, cliffs, tidal deltas, each of which has an
associated substrate type (Fig. 2). Biological features refer mostly to intertidal biological
assemblages that typically appear as distinct colour/texture bands along the coast (biobands; Fig.
2). Information is cataloged in terms of alongshore units (delineated on a the digital high-water
line shoreline) and across-shore components, which are described in a searchable database but
which are not cartographically delineated on maps.

Standard Cross-walked from: (if the study involved a crosswalk of CMECS to another
classification standard, provide the name and link if available):

Harney, J.N., Morris, M.C., and Harper, J.R. 2008. ShoreZone Coastal Habitat Mapping:
Protocol for the Gulf of Alaska. Contract Report by Coastal & Ocean Resources Inc. of Sidney,
BC for the Exxon Valdez Trustee Council, Anchorage, Alaska. 157 p.
http://alaskafisheries.noaa.gov/habitat/shorezone/szintro.htm

Results/Products Overview:

Harper, J.R. and S. Ward 2010. Data Cross-walk Between the ShoreZone Coastal Habitat
Mapping System and Coastal and Marine Ecological Classification System (CMECS). Contract
Report by Coastal & Ocean Resources Inc., Sidney, BC to the National Marine Fisheries Service,
Juneau, AK, 39p. http://alaskafisheries.noaa.gov/habitat/shorezone/szintro.htm

A flat database was created to capture data using the CMECS classification protocol. ShoreZone
data were transferred from five tables or databases to the CMECS database. An estimated 75 -
80% of the ShoreZone data was easily transferred. The 20-25% that could not be transferred
relates primarily to the fundamentally different spatial mapping units of ShoreZone and CMECS.
The primary ShoreZone mapping unit is an alongshore segment or shore unit (522 in the test

                                                26
sections); there are a number of ShoreZone attributes that apply to the entire unit (e.g., shore
type, habitat type). The primary mapping unit of CMECS is a depth ribbon, including supratidal,
intertidal and shallow subtidal zones (1,966 in the test sections). While the data from the
ShoreZone across-shore components could be transferred to these CMECS depth ribbons (the
75-80%or the data), there is no summary indicator in CMECS for the across-shore suite of shore
features cataloged in CMECS Geoform, Surficial Geology or Biotic Cover layers.

The spatial representation of ShoreZone units is a line segment, delineated by segmenting the
digital high water line. All attribute data are linked to the line segments. Across-shore zonation
of forms, material and biota are preserved within ShoreZone by using an indexing system; this
indexing system explicitly links forms, materials and biota and is transferable to CMECS.
CMECS may not have envisaged the use of line segment representation for mapping units as it
was designed primarily as a classification system, but with the addition of indexing, CMECS
appears to work well with line segments for spatial units.

One of the most challenging aspects of the cross-walk is how patchiness of surficial substrate is
captured in CMECS. Small scale variability in horizontal and vertical substrates is common on
glaciated shorelines. The categorization of all substrates as either rock shore or unconsolidated
shore is problematic, especially at smaller mapping scales. For example in the 5,000 km
mapping project of Prince William Sound, 45% of the shoreline is classified as some type of
combination of rock and sediment. CMECS may wish to consider an intermediate rock-and
sediment shore category in their substrate classification.

The overall assessment for data transfer between ShoreZone and CMECS is good – 75-80% of
the data is transferable and data most commonly used in habitat capability modeling was moved
to CMECS. An additional test of CMECS as a mapping system will be the interpretation of
shoreline attributes directly from imagery into the CMECS data structure that was created for
this cross-walk.

Lessons Learned: Overall description of advantages and challenges of applying CMECS. Use
full sentences to describe them.

        Lesson Learned
1. Overall much of the ShoreZone data could be transferred into the CMECS system. We
estimate that about 75-80% of the information can be transferred. Portions of ShoreZone that can
not be transferred are summary indicator type attributes.

2. The two systems have fundamentally different mapping units – in ShoreZone line segments
are the primary mapping unit and across-shore components are a secondary mapping
subdivision. The component data were transferable to the CMECS units. Relative position of the
CMECS units within the intertidal zone is preserved with a sequential indexing system because
the CMECS unit data are presented as a single line segment.

3. A challenging aspect of the mapping in Alaska is the patchiness of intertidal substrate. While
the detailed substrate characterization of ShoreZone can be transferred to CMECS at the detail
classification level, a significant proportion (30%) of the intertidal zone is a combination of rock

                                                 27
and sediment which does not roll up conveniently into the more general levels of the CMECS
classification.

       Recommendations
1. Because CMECS is a classification system, there is little mention of the spatial units types
and mapping scales. This topic will require discussion if CMECS moves towards a mapping
system.

2. The CMECS classification is somewhat east-coast centric and a number of additions are
required for implementation in Alaska:
    add “ice” as a substrate (to accommodate glacial ice fronts).
    add permafrost into classification, possibly in process section.
    many Alaska species are missing in the biotic group and biotope descriptions.
    consideration of glaciated coasts as having unique geological stratification that is
       important to be accommodated by the Surficial Geology Component.

3. Although the GFC, SGC and BCC are developed as independent layers in the classification
system, geomorphology and substrate are likely to be used for delineating spatial units in
mapping applications. CMECS may want to provide some guidance for explicitly linking the
various components in the system.

4. ShoreZone has documented that rock and sediment combinations are quite common on the
Alaska coast. It would be quite realistic to revise the higher level Surficial Geology Component
classes to include three classes of substrate: (a) rock shores, (b) sediment shores and (c) rock and
sediment shores to reflect the relatively common occurrence of all three coastal habitats.




                                                28

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:9
posted:10/28/2011
language:English
pages:28
xiaohuicaicai xiaohuicaicai
About