A Fish Habitat Classification Model for the Upper and

A Fish Habitat Classification Model for the Upper and
)
')
)
)
)
")
')
')
A Fish Habitat Classification Model for the Upper and
Middle Sections of the Bay Of Quinte, Lake Ontario
)
)
')
')
C.K. Minns\ A. Bernard\ C.N. Bakelaar\ and M. Ewaschuk2
')
)
')
')
')
)
)
)
)
)
1Great Lakes Laboratory for Fisheries and Aquatic Sciences
Fisheries and Oceans Canada
Bayfield Institute, 8671.,akeshore Road, P.O. Box 5050
Burlington, Ontario L7R 4A6 CANADA
2Lower Trent Conservation Authority
441 Front Street, Trenton
Ontario K8V 6C1 CANADA
)
)
March 2006
)
)
)
)
Canadian Manuscript Report of Fisheries and
Aquatic Sciences No. 2748
)
)
)
)
~
)
)
)
)
)
)
)
)
1+1
Fisheries
and Oceans
Peches
et Oceans
Canada
1
1
1
Canadian Manuscript Report of
Fisheries and Aquatic Sciences
Manuscript reports contain scientific and technical information that contributes to existing
knowledge but which deals with national or regional problems. Distribution is restricted to institutions or
individuals located in particular regions of Canada. However, no restriction is placed on subject matter,
and the series reflects the broad interests and policies of the Department of Fisheries and Oceans, namely,
fisheries and aquatic sciences.
Manuscript reports may be cited as full publications. The correct citation appears above the
abstract of each report. Each report is abstracted in Aquatic Sciences and Fisheries Abstracts and indexed
in the Depaltment's annual index to scientific and technical publications.
Numbers 1-900 in this series were issued as Manuscript Reports (Biological Series) of the
Biological Board of Canada, and subsequent to 1937 when the name of the Board was changed by Act of
Parliament, as Manuscript Reports (Biological Series) of the Fisheries Research Board of Canada.
Numbers 901-1425 were issued as Manuscript Reports of the Fisheries Research Board of Canada.
Numbers 1426-1550 were issued as Department of Fisheries and the Environment, Fisheries and Marine
Service Manuscript Reports. The current series name was changed with report number 1551.
Manuscript reports are produced regionally but are numbered nationally. Requests for individual
reports will be filled by the issuing establishment listed on the front cover and title page. Out-of-stock
reports will be supplied for a fee by commercial agents.
1
1
1
1
I
1
1
I
I
Rapport manuscrit canadien des
sciences halieutiques et aquatiques
Les rapports manuscrits contiennent des renseignements scientifiques et techniques qui
constituent une contribution aux connaissances actuelles, mais qui traitent de probltlmes nationaux ou
regionaux. La distribution en est limitee aux organismes et aux personnes de regions particulieres du
Canada. II n'y a aucune restriction quant au sujet; de fait, la serie reflete la vaste gamme des interets et
des politiques du ministere des Peches et des Oceans, e'est-it-dire les sciences halieutiques et aquatiques.
Les rapports manuscrits peuvent etre cites comme des publications complerwa. Le titre exact
parait au-dessus du resumes de chaque rapport. Les rapports manuscrits sont resumes dans la revue
Resumes des sciences aquatiques et halieutiques,et ils sont classes dans l'index annuel des publications
scientifiques et techniques du Ministere.
Les numeros I it 900 de cette serie ont ete publies it titre de manuscrits (serie biologique) de
I'Office de biologie du Canada, et apres Ie changement de la designation de cet organisme par decret du
Parlement, en 1937, ont ete classes comme manuscrits (serie biologique) de I'Office des recherches sur
les pecheries du Canada. Les numeros 90 I it 1425 ont ete publies it titre de rapports manuscrits de
l'Office des recherches sur les pecheries du Canada. Les numcros 1426 it 1550 sont parus it titre de
rapports manuscrits du Service des peches et de la mer, ministere des Peches et de l'Environnemenl. Le
nom actuel de la serie a ete etabli lors de la parution du numero 1551.
Les rapports manuscrits sont produits it I' echelon regional, mais numerotes it I' echelon national.
Les demandes de rapports seront satisfaites par l'etablissement auteur don't Ie nom figure sur la
couverture et la page du titre. Les rapports epuises seront foumis contre retribution par des agents
commerciaux.
.1
I
1
I
I
I
I
I
I
I
I
CANADIAN MANUSCRIPT REPORT
)
OF FISHERIES AND AQUATIC SCIENCES No. 2748
)
A FISH HABITAT CLASSIFICATION MODEL FOR THE UPPER
AND MIDDLE SECTIONS OF THE BAY OF QUINTE,
LAKE ONTARIO
by
C.K. Minns I, A. BernardI, C.N. Bakelaar l , and M. Ewaschuk2
IGreat Lakes Laboratory for Fisheries and Aquatic Sciences, Fisheries and
Oceans Canada, Bayfield Institute, PO Box 5050, 867 Lakeshore Road,
Burlington, Ontario L7R 4A6
)
)
)
)
)
2Lower Trent Conservation Authority
441 Front Street, Trenton
Ontario K8V 6Cl CANADA
('
C
(,
(,
(
L
(,
(,
C
(
(
(
<:
(
(1
( I
(I
()
(I
(I
(I
(l
(I
(I
( I
(I
( I
( I
©Her Majesty Queen of Canada 2006
Cat. No. FS97-412748E ISSN 070-6473
Correct citation of this publication:
Minns, C.K., A. Bernard, C.N. Bakelaar, and M. Ewaschuk. 2006. A Fish Habitat
Classification Model for the Upper And Middle Sections of the Bay Of Quinte, Lake
Ontario. Can. MS Rpt. Fish. Aquat. Sci. 2748: vii+61p.
It
(I
(I
(I
(I
(I
(I
(I
(I
(I
(I
(I
(
(I
(I
(
ABSTRACT
A fish habitat classification model was developed and applied to the upper and
middle sections of the Bay of Quinte, Lake Ontario. Available habitat inventories were
assembled in a GIS database, bringing bathymetric, shoreline, substrate, and vegetation
data together in a series of layers. The classification model was developed in four steps.
In the first step, the Defensible Methods (DM) model developed by Minns et al. (2000)
was used to estimate suitability values in all habitat patches for a set of nine fish groups
each with three life stages. The fish groups were formed from the assemblage of fish
species present in the Bay of Quinte by combining them according to thermal and
vegetation preferences, and combinations of size and age-at-maturity. Different methods
of combining the 27 suitability indices were examined to allow designation of each
unique habitat patch to low, medium or high suitability categories for fish. The K-means
clustering technique was selected for classifying habitat patches into three suitability
categories, thereby exploiting natural breaks in the cumulative distributions of suitability
values and maintaining consistency with underlying habitat features. In the second step,
the spatially rare habitats for each fish group by life stage combination were used to
identify habitat patches that are important for particular fish groups and life stages but
which had been classified as medium or low suitability in the first classification step.
Criteria for recognizing rarity were used to reassign habitat patches rated low or medium
in step one to the high class. In the third step, local expert knowledge of important fish
habitats gathered from anglers and fishers were used to develop an expert classification.
This expert mapping of important fishing areas was compared with that obtained via
suitability and rarity ratings and then, in step four, used to upgrade some areas from low
or medium to high.
The final habitat classification model is a mixture of suitability, rarity and expert
ratings. The habitat suitability class assignments obtained in step one were not changed
appreciably by steps two and three. The combined suitability-rarity ratings showed good
agreement with the local expert ratings. Important fishing areas either overlapped suitable
areas or were close by where fisher access would be restricted by depth or vegetation
density. The final habitat classification for the Bay of Quinte provides a context for both
conservation and restoration efforts. Periodic updating of the classification system will be
needed as conditions change, e.g., as a result of climate change or as the effects of the
zebra mussel invasion on macrophytes and substrates mature, or as data on other habitat
elements becomes available, e.g., seasonal and spatially thermal habitat maps. Further
effort is needed to understand the procedures used by government agencies at different
levels to integrate the knowledge embodied in habitat maps into on-going fisheries and
fish habitat management.
)
RESUME
Un modele de classification des habitats des poissons a ete elabore et applique aux
sections superieure et moyenne de la baie de Quinte, dans Ie lac Ontario. Les inventaires
existants des habitats ont ete verses dans une base de donnees SIG, regroup ant ainsi en
une serie de couches des donnees sur les proprietes bathymetriques, les littoraux, les
substrats et la vegetation. L'elaboration du modele de classification s'est faite en quatre
etapes. Lors de la premiere etape, on a fait appel au modele des methodes defendables de
Minns et al. (2000) pour estimer les valeurs de convenance de toutes les parcelles
111
c:
('
(
d'habitat d'un ensemble de neuf groupes de poissons, chaque groupe etant represente par
trois stades biologiques. Les groupes de poissons ont ete constitues a partir de
I' assemblage des especes trouvees dans la baie de Quinte. Les especes ont ete combinees
selon leurs preferences en matiere de temperature et de vegetation ainsi que selon leur
taille et leur age a la maturite. Differentes methodes visant a combiner les 27 indices de
convenance ont ete examinees; on a ainsi pu attribuer a chaque parcelle d'habitat des
poissons une categorie de convenance : faible, moyenne ou elevee. La technique de
regroupement a K-moyennes a ete choisie pour classer les parcelles d'habitat dans les
trois categories de convenance, ce qui a permis d' exploiter les bris naturels dans les
distributions cumulatives des valeurs de convenance et de conserver une constance dans
les caracteristiques sous-jacentes des habitats. A la deuxieme etape, on a identifie les
parcelles d'habitat a caractere spatialement rare, jugees importantes pour certains groupes
et stades biologiques de poissons, mais qui avaient ete classees dans les categories
moyenne ou faible lors de la premiere etape. Les criteres de reconnaissance de la rarete
ont permis de promouvoir ala categorie de convenance elevee des parcelles d'habitat
designees faibles ou moyennes lors de la premiere etape. A la troisieme etape, les
connaissances d'experts locaux sur les habitats importants des poissons, recueillies aupres
de pecheurs amateurs et professionnels, ont ete utilisees pour mettre au point une
classification des experts. La cartographie des zones de peche importantes ainsi obtenue a
ete comparee a celle fournie par les classifications en fonction de la convenance et de la
rarete, puis, a la quatrieme etape, elle a servi a reclasser certaines zones, en les faisant
passer des categories faible ou moyenne a la categorie elevee.
Le modele final de classification des habitats est en fait fonde sur une
combinaison de categories de convenance et de rarete et du classement des experts. Les
classements de convenance des habitats obtenus a la premiere etape n' ont pas ete
modifies de maniere significative par les etapes deux et trois. Les classements fondes a la
fois sur la convenance et sur la rarete correspondaient bien avec les classements des
experts. Les zones de peche importantes soit chevauchaient les zones a convenance
elevee, soit se trouvaient a proximite dans des endroits OU l'acces des pecheurs etait
limite par la profondeur de I' eau ou par la densite de la vegetation. La classification finale
des habitats de la baie de Quinte fournit un contexte a la fois pour les efforts de
conservation et pour les efforts de restauration. Vne mise a jour periodique du systeme de
classification sera necessaire a me sure de I' evolution des conditions (par exemple, Ie
changement climatique et les effets des moules zebrees sur les macrophytes et les
substrats) ou de l'apparition de nouvelles donnees sur d'autres composantes des habitats
(par exemple, cartes saisonnieres ou spatiales des habitats thermiques). D'autres efforts
sont necessaires pour que les organismes des divers ordres de gouvernement integrent les
connaissances contenues dans les cartes des habitats a leurs methodes courantes de
gestion des peches et des habitats des poissons.
(~
(
(
(
(
(
('
()
( )
( )
<. )
()
(l
( 1
()
()
()
()
()
()
(
)
()
()
(I
(
(
(
(
(I
(I
(~
(
1
(I
(I
(I
(I
(I
]v
(
(
(I
)
)
)
')
)
TABLE OF CONTENTS
)
TABLE OF CONTENTS .............................................................................................................................. v
")
LIST OF T ABLES ...................................................................................................................................... VI
)
)
)
)
)
LIST OF FIGURES .................................................................................................................................... VI
LIST OF APPENDICES ........................................................................................................................... VII
INTRODUCTION ......................................................................................................................................... 1
PURPOSE AND OBJECTIVES .... ........................... .. ..... .. .... .............................................. .. .......................... . .. . 2
MATERIALS AND METHODS .................................................................................................................. 2
BAY OF QUINTE HABITAT MAPPING ................. ......... .. ...... ... ........................... ..... .. .... ........ ......................... 2
)
)
)
)
)
)
)
)
)
)
OMNR Nearshore Habitat Inventory .... .... .. ....... ............. ..... ..... ............... ..... ....... ......... .. .... .. ............. .. .. 2
Shoreline and Bathymetry .. ...... ............. .... .. ........ ..... ..... .. ..... .... ..... .. ...... ..... ..... .... .. ...... .. ... .................. .. ..3
Vegetation ..... ...... .. ....... ...... .... ......... .... .. ...... ................ .. .... ........ ...... ................... .. ......... ...... ........ .... ... .... 4
Unique Habitat Features ..... .. .. ........ ... .. ...... ........ ..... .... ..... .. ..... .. ... ......... ........................ ......... .. .... ..... ... .5
HABITAT CLASSIFICATION ApPROACH ... .. .. .. .... ......... .... .. ... .. ..... ...... .. ... .. ................... .. ... .. .. .... ...... .......... .... . 5
STEP ONE: FISH HABITAT SUITABILITY ASSESSMENT .. .. ........................................ .... .. ...... .. .. ..... ................ 6
Application of Defensible Methods ...... ...... ..... .............. ...... .. .. .. ..... ...... ..... ............ .. .... .... ... ................... 6
Location Species List .......... ....................................................................................... ....................................... 6
Selection of Fish Groups ................................................................................................................................... 6
Selection of Weights for Fish Groups and Life Stages ...................................................................................... 8
Linking Fish Habitat Categories to Species' Requirements .................................. ............................................ 8
Computation ofHabitat Suitabilities .......... ....... .... .... ..... .. ............................ .. .... ... .. ........ .. ... ................. 9
Habitat Suitability Classification ......... ...... ... .. ...................... ......... ..... ...... .... .. ..... ...... .......... ........... ...... .9
STEP Two: FISH HABITAT RARITY ASSESSMENT ...... ...... ............................ ........ .. ....... ...... ... ..... .. ........... .. 10
STEP THREE: LOCAL EXPERT KNOWLEDGE ASSESSMENT ..... .. .. .... .... .. .. .... .. .................... .......... .. ............ .. 11
Data Collection ........... ..... ....... ... ..... ... ... .. ..................................... ...... ..... ... .... ... .. ..... .... .... ..... .......... ..... 11
Fish Habitat Suitability Index Validation ............. .. ... .... ...... .............. .. .... .. .... .................. ...... ............ .. 12
STEP FOUR: OVERALL RULE-BASED CLASSIFICATION ......... ..... ....................... .... .... ...... .......... .. ............... 13
RESULTS ..................................................................................................................................................... 13
HABITAT SUITABILITY USING DEFENSIBLE METHODS: .. .. ......................................... ................................. 13
HABITAT SUITABILITY CLASSIFICATION METHODS ........ ........ ........................................... ........................ 14
HABITAT RARITY ........... .............................. ....... ..... ..... .. ....... ............................... ............... .. .... ............... 15
LOCAL EXPERT KNOWLEDGE ...................... ... ..... ... ........ ....................... ... .... ... ..... ....... ................ ....... .... ... 16
COMBINED RANKING MAPS .. ..... .... .... ...... ..... ... ..... .... ..... .. ... .... ... .... ...... ....... ..... ..... .... ... ....... ... ..... .. ... .. ... .. .. . 16
DISCUSSION ............................................................................................................................................... 17
BAY OF QUINTE HABITAT INVENTORY AND GIS DATABASE ................... .... .. ........ ... ...... ... .......... .. ........... . 17
HABITAT CLASSIFICATION MODEL ................ ... ..... ... ... .. ... .. .. .......................... ........... .... ........ ..... ............... 18
IMPLICATIONS OF RESULTS IN THE BAY OF QUINTE .. ....... .. .. ............................. ................. ....................... 19
NEXT STEPS .. ........ .... .... ........................................ ... ..... ... ...................................... ... ... ....... .. ..................... 20
ACKNOWLEDGEMENTS ........................................................................................................................ 21
REFERENCES ............................................................................................................................................ 22
APPENDICES .............................................................................................................................................38
V
c:
c:
c
c:
LIST OF TABLES
Table 1
Table 2
Table 3
Table 4
Table 5
Table 6
(,
List of weighting applied to the nine fish groups for computing a composite
suitability in the DM's estimation offish habitat suitabilities. (Life stages
were assigned equal weights of 0.3333).
Percentages of habitat area in the Bay of Quinte ranked low, medium or high
using the composite index alone, and then with the areas assigned as rare
reassigned to high, separated out for each of the 27 individual habitat
suitability indices and the composite index.
Average percentage composition of the three habitat characteristics for each
of the six K-means clusters (The cluster numbers referred to their order in
the K-means clustering).
Mean habitat suitability value in each K-means cluster group for all 27 OM
indices and the numbers of indices having the lowest pair of means (italics),
the middle pair (underlined) and the highest two pair (bold).
Percentages of habitat area within each classification (low, medium, or high)
of the clustered index that either passes (rare) or fails (-) the rarity threshold
for each of the 27 suitability indices. Shaded percentages are those areas
reclassified as high when the rarity threshold is applied.
Cross-tabulations of unique polygons by area and percentage of total area by
the composite ranks (columns) and by cluster ranks and numbers (rows), and
also with rarity and fish expert classifications. (The upper half provides
weight suitable hectares and the lower half percentages).
LIST OF FIGURES
Figure 1
A map of the Bay of Quinte, Lake Ontario, showing the study area, the
upper and middle bay sections.
Figure 2
Areas in the Bay of Quinte classified as low, medium, and high suitability
for the group of adult, cool-water fishes with age-at-maturity <= 2 years and
a low vegetation preference.
Figure 3
Areas in the Bay of Quinte classified as low, medium, and high suitability
for the group of spawning, warm-water fishes with age-at-maturity <= 3
years and a high vegetation preference.
Figure 4
Areas in the Bay of Quinte classified as low, medium and high using the
composite habitat suitability index obtained using DM.
Figure 5
Areas in the Bay of Quinte classified as low, medium and high using the Kmeans cluster analysis of all suitability indices.
Figure 6
Habitat polygons in the Bay of Quinte ranked as "rare" at least once for any
of the 27 individual fish group*life stage habitat suitability indices.
VJ
(,
(:
C.
(,
(,
(
(
(
C'
(
(
(I
Cl
(I
(I
(I
(I
(I
(I
(
(
(I
(
( 1
(I
(I
(I
(
1
(I
(I
(I
(I
(I
(I
(I
(I
(
<-
)
)
)
)
)
:1
)
Figure 7
Habitat polygons in the Bay of Quinte classified as rare when crossreferenced with the DM composite suitability class assignments of low and
medium.
Figure 8
Habitat polygons in the Bay of Quinte classified as rare when crossreferenced to the K-means clusters assigned low and medium.
Figure 9
Habitat polygons in the Bay of Quinte containing at least one expert fisher
validated fishing site.
Figure 10
Map of the Bay of Quinte indicating the areas classified as low, medium,
and high fish habitat suitability with respect to the DM composite index,
rarity, and expert fishers.
Figure 11
Map of the Bay of Quinte indicating the areas classified as low, medium and
high fish habitat suitability with respect to the K-means suitability clusters,
rarity, and expert fishers.
J
LIST OF APPENDICES
)
)
)
Appendix A Metadata documentation and data dictionary for the Bay of Quinte physical
habitat GIS database
)
)
)
)
)
)
)
)
)
)
)
)
)
Appendix B Species location list of the fish present in the Bay of Quinte
Appendix C Listing of the 9 fish species groupings for the freshwater species present in
the Bay of Quinte used in this study
Appendix D Correlation coefficients for the 27 habitat suitability indices which represent
all of the combinations with respect to thermal, age of maturity, vegetation
preference and life stage of the freshwater fish inhabiting the Bay of Quinte
(note: Life stages: A = adult, S = spawning, Y = YOY)
Appendix E Sample graph of the mean habitat suitability index values (y-axis) versus
depth (m) (z-axis) and substrate classes (x-axis) for an adult who is a
member of the cold water groups in a habitat containing only submergent
vegetation
Appendix F Mean habitat suitability index values among life stages (Adult, Spawning,
YOY) (y-axis) versus depth (m) (z-axis) and substrate classes (x-axis) for
the 9 species groups across three cover classes (no cover, submergent and
emergent vegetation)
Appendix G Graphs showing cumulative area (dashed line) and cumulative weighted
suitable area - WSA (solid line) versus Bay of Quinte habitat suitability
database indices for 1) Coldwater, 2) coolwater and 3) warmwater.
Suitability is further categorised by life stage (Adult, Spawning, and YOY),
Age of Maturity (LT2 = Less than or equal to 2 years, GT2 = Greater than 2
years, LT3 = Less than or equal to 3 years and GT3 = Greater than 3 years)
and Vegetative Preference (Low and High). The graphs depict application
of the 75 percent and 0.75 suitability cut-offs for identifying rare, highly
suitable habitat.
Vll
(
( )
()
(
INTRODUCTION
( )
In 1985, the International Joint Commission (UC) designated the Bay of Quinte as
( )
one of 43 "Areas of Concern" across the Great Lakes basin where one or more of 14
( )
beneficial ecosystem uses was impaired. Agencies in both Canada and the United States
( )
were charged by the UC with developing and implementing Remedial Action Plans
()
(RAPs) for each area. Work on the Bay of Quinte RAP began in 1985 with the formation
of a federal/provincial coordination committee. That committee was able to build on the
work of Project Quinte, a federal/provincial/university consortium of researchers who had
( )
()
( )
( )
been studying the Bay of Quinte ecosystem intensively since 1972 when plans for major
(
nutrient load reductions were first established (cf Minns et al. 1986).
(
In 1993, a Stage II RAP report, Time to Act, (Bay of Quinte RAP, 1993) was
released by the coordinating committee documenting 10 impaired beneficial water uses
out of the list of fourteen. Many of the impairments were tied to the hyper-eutrophication
that occurred in the late 1960s and early 1970s. Key impairments were those associated
with the fish community and with fish and wildlife habitat. The health of fish
(
(
(
(
( 1
(I
communities, and their fisheries, is a key indicator of ecosystem health. Further, healthy
()
fish habitats are a prerequisite for healthy fish and fisheries. The Stage II report
(
recognized that considerable alteration, fragmentation, degradation and loss of fish
(
habitat had occurred, and recommended, alongside a number of site-specific actions, the
(
development of a comprehensive fish habitat management plan. As the RAP
implementation process began emphasis was placed on improving and securing
(
(
(
ecosystem health via effective management of nutrients and contaminants. There has
(
been much progress in these areas, laying the groundwork for long-term improvements in
( 1
the health of the Bay of Quinte ecosystem. Recently, the Bay of Quinte Restoration
(
Council, the successor to the RAP coordinating committee, took on the task of
( 1
developing a fish habitat management plan for the Bay of Quinte area, using the
Department of Fisheries and Oceans (1986) policy document on the management offish
( 1
(I
( 1
habitat as its starting point. The policy is exemplified by its guiding principle of "no net
(I
loss of productive capacity offish habitats" (mirroring the policy goal ofa net gain).
(I
A key element in the development of a fish habitat management plan is
information, an assessment and analysis of the supply and quality of fish habitat available
1
(I
(
(
1
(
1
(
I
C
)
to support the productivity of fish in the ecosystem (not including tributaries). The
purpose of this report is to describe and document the development and implementation
)
')
")
)
of a fish habitat suitability model for the upper and middle sections of the Bay of Quinte.
The model draws on the considerable databases already assembled on many aspects of
the Bay of Quinte ecosystem and the methodologies already applied elsewhere in the
)
Great Lakes basin (Minns et al. 1999,2001). Habitat is classified as to its relative
)
)
)
suitability for fish although agencies managing fish habitats for productivity or speciesat-risk would prefer to know which habitats are critical, essential, or important. At
present, the state of the science precludes such precise designations (Rosenfeld and
Hatfield 2006, Morrison et al. 1999). Knowledge of the fish habitats in the Bay of Quinte
can guide and prioritize fish habitat conservation, restoration and enhancement efforts
within the Bay proper.
)
)
)
Purpose and Objectives
The overall aim of this report was to generate a scientifically defensible fish
habitat suitability classification scheme for the upper and middle areas of the Bay of
Quinte for use as a guide for future management and conservation. The classification and
assessment of the upper and middle areas of the Bay of Quinte was performed using the
)
)
)
Defensible Methods (DM) model developed by Minns et al. (2001) and with a GIS
database drawn from several sources detailing habitat characteristics (depth, vegetative
cover, and substrate). A report by Minns et al. (1999), which presented a fish habitat
classification model for areas of Severn Sound, in Georgian Bay, was used as a basis for
)
)
)
the development of fish habitat suitability maps for the upper and middle areas of the Bay
of Quinte. The work was limited to the upper and middle sections of the bay by the
availability of mapped habitat data although the Area of Concern includes the lower Bay
out to the boundaries of Adolphus Reach.
MATERIALS AND METHODS
The development of the classification model for the Bay of Quinte involves two
components: 1) assembly of a GIS database describing habitat attributes and 2)
implementation of an appropriate habitat classification model.
Bay of Quinte Habitat Mapping
OMNR Nearshore Habitat Inventory
2
A study to inventory and map fish and wildlife habitat in the nearshore zone of
the Bay of Quinte was initiated in 1991 and completed in 1993. The results of this study
included detailed mapping of substrate and aquatic vegetation that extended "from as far
inland as visible from the boat" to a depth at which substrate or aquatic vegetation was
()
( )
()
()
( )
( )
( )
not visible. Therefore the habitat information is representative of a single 'snap-shot' in
()
()
time (Smith 1993). The substrate and aquatic vegetation information collected in the
()
inventory were incorporated into the GIS habitat database used on the habitat
()
()
()
classification model.
Shoreline and Bathymetry
( )
The shoreline GIS layer was provided by the Ontario Ministry of Natural
()
Resources (OMNR) at a scale of 1: 10000. This scale was selected because it is consistent
(
with the bathymetry data provided by Canadian Hydrographic Services (CHS) and is the
()
maximum scale of the nearshore inventory data. In the GIS, the map projection data
()
()
were: Universal Transverse Mercator, Zone 18, NAD83, Central Meridian -75.0.
A bathymetry map was assembled from CHS digital chart data and field sheets,
)
( )
( )
corrected to International Great Lakes Datum 1985 (lGLD85). The shoreline was
()
incorporated by converting the line to points and set to an elevation of 74.2m above sea
(
)
level (Lake Ontario low-water datum IGLD85). A raster (3m cell size) representing
(
)
elevation was generated then reclassified to DM depth classes (0-1, 1-2,2-5,5-10, 10+
(
(
)
metres) (Appendix Figure AI). The total area covered was 190.54 km2, distributed as
follows: 0-1 m depth range had an area of28.3l km2; 1-2 m 26.67 km2; 2-5 m 83.96
)
(
( 1
km2; 5-10 m 42.37 km2, 10+ m 9.23 km2; and 1.83 km2 as dry land above the datum.
Substrate
(I
( I
Two sources of substrate data were incorporated into the habitat classification
(
model, each had a different spatial extent (nearshore and offshore) and different units.
(
The nearshore data was extracted from the OMNR habitat inventory. The inventory maps
( I
were polygons attributed with a description of the 10 dominant and sub dominant
substrate types: Bedrock, Boulder, Rubble, Gravel, Sand, Silt, Clay, Muck, Detritus, and
Marl (Smith 1993). The DefMeth substrate categories are bedrock, boulder, rubble, cobble,
(
(
( I
( I
gravel, sand, silt, clay, and hard-pan clay. These types were assigned to the Defensible
(
Methods substrate matrix where the dominant type was assigned a greater proportion than
( t
(I
3
(
1
(
1
CI
the subdominant (e.g. a polygon identified with dominant type Boulder and subdominant
Gravel, was assigned 70% boulder, 30% gravel), (Appendix Table AI).
The offshore data was collected as point samples by R. Thomas of Environment
Canada during the years 1972 and 1973 (R.L. Thomas, personal communication). These
)
)
)
)
)
samples contained percent composition of sand, silt and clay. These points were used to
generate Theissen polygons of substrate covering the offshore zone of the study as well
as to fill in nearshore areas not mapped by the OMNR habitat inventory. The two
datasets were spatially joined together, giving priority to the nearshore data, and thereby
forming one substrate map (Appendix Figure A2).
Vegetation
DM requires spatial information on vegetation cover in three categories:
emergent, submergent, and no cover. There were two sources of aquatic vegetation
information used in the habitat suitability model: the first being the OMNR habitat
inventory data, the second a submergent vegetation model developed for the Bay of
Quinte (Seifried 2002) using results from repeated transect surveys conducted between
1972 and 2000. The OMNR habitat inventory data were stored as spatial polygon files
with class and density attributes (Very sparse 5-20%, Sparse 20-40%, Moderate 40-60%,
Dense 60-80%, and Very Dense 80-100%). Of the 6 vegetation classes only values for
emergent and submergent vegetation data were used.
Emergent vegetation values were derived from the OMNR habitat inventory.
There were no additional or more recent sources of emergent vegetation available.
Seifried (2002) developed models to predict submerged vegetation density in the Bay of
Quinte for three time periods including post-zebra mussel invasion (post 1995) using a
regression tree method. The results from the post-zebra mussel period were used for the
upper and middle Bay of Quinte sections to generate a 'current' representation of
vegetation. In some cases, the two data sources for submergent vegetation overlapped
)
)
spatially. When this occurred a hierarchical method was developed to combine the two
datasets. The two sources of data were spatially joined, and then the source data for each
polygon was tested and values assigned for emergent, submergent and no cover using the
following logic:
EMERGENT VEGETATION:
4
c
•
If data exists in Inventory Then use Emergent density mid-point
•
If data does not exist in Inventory Then assume no Emergent cover
SUBMERGENT VEGETATION:
•
•
•
()
If data exists in both Inventory and Seifried model Then use mid-point of
()
Seifried model
C)
If data does not exist in Seifried model Then use inventory Submergent
()
density mid-point
()
()
If data does NOT exist in Inventory or no Seifried model Then No Cover =
()
100
()
()
NO COVER:
•
()
()
()
()
()
100 - (Emergent + Submergent) = No Cover
Examples of vegetation assignments are shown in Appendix Table A2.
()
()
(
Unique Habitat Features
A topological overlay of depth, vegetation and substrate map layers resulted in a
new layer that preserved the features of all the input layers and represents a combination
)
{J
{J
of all the habitat characteristics associated with each place (Appendix Figure A4). The
()
()
area of each unique habitat combination was summed and resulting polygon attribute
(
records were used in the input data table for the Defensible Methods software. Sample
()
records for input to Defensible Methods software are shown for the zoomed in area
highlighted in Appendix Figure A4 are shown along with input header records in
(
(
(
.
)
)
,
Appendix Table A3.
(,
Habitat Classification Approach
(
The approach employed here is the same as that reported in Minns et al. (1999)
for the Severn Sound on Georgian Bay, Lake Huron. There are four steps in the approach:
•
Step One: a combined habitat suitability assessment of all habitat patches for a
target set of fish groups by life stage;
•
Step Two: an assessment and classification of rare habitat types by fish group and
life stage to ensure that step one does not misclassify habitat patches important to
one or few fish groups;
•
Step Three: assessment and classification of areas using local fisher expert
knowledge of important fishing areas; and
5
,
(
,
(
)
(I
(I
(I
(I
(I
(I
(I
(I
<I
<.1
(I
(I
•
Step Four: implementation of a combined classification model drawing on results
obtained in the first three steps.
Step One: Fish Habitat Suitability Assessment
The fish habitat suitabilities ofthe upper and middle areas of the Bay of Quinte
were predicted using the 'Defensible Method's software (DM) described by Minns et al.
(1996). DM software uses an amalgamation of literature-based databases, compiled by
Lane et al. (1996a,b,c), which detail the habitat requirements offish during their
spawning, nursery (young-of-the-year or YOY), and adult life stages. The DM software
)
uses these life history preferences to estimate habitat suitability values for fish in defined
areas of habitat.
Application of Defensible Methods
)
)
The determination of the habitat suitability in the Bay of Quinte involved several
steps that are outlined below.
Location Species List
)
)
In the DM software the location species list determines which species habitat
preference data are used to estimate suitability indices. The location species list of those
)
)
fish inhabiting the Bay of Quinte was based on a list compiled by Hurley et al. (1986).
Only species which were cited to have been found in recent collections from the bay were
included in the species list (Appendix B). Sea lamprey and splake were not included in
the species list for the purpose of fish habitat suitability mapping. Sea lamprey make only
)
)
a brief migratory use of the Bay between spawning and larval life stages in streams and
the adult/parasitic life stage in Lake Ontario proper. Splake were never stocked in any
significant numbers in Lake Ontario and certainly did not to establish a population in the
Bay of Quinte.
Selection ofFish Groups
The location species list was sorted into smaller groups, or guilds, of species.
:>
)
This is done to ensure that when determining the overall composite habitat suitability of
an area, different groups of species can be weighted differently to reflect the objectives of
the particular application ofDM. For example, an application geared solely towards
conserving commercially important species might assign a greater weighting to some
species (i.e. walleye and yellow perch) that are more valuable than others (i.e. gizzard
6
shad and common carp). It is important to note that all species within a species group are
weighted equally in the estimation of habitat suitability values from species preference
()
()
()
()
( )
data. Group suitability values are then assigned a group weight to reflect their emphasis
()
()
in the application. Different weights are also assigned to each life-stage to place an
( )
emphasis on a specific life-stage deemed to be of great importance, e.g., to ensure that
C)
enough habitat is available for rearing and feeding as well as for spawning.
( )
The Bay of Quinte species list was divided into nine groups to allow matching of
species with similar life history preferences (Appendix C). These groups were formed by
()
( )
(
)
)
for consistency in habitat preferences within groups, it was determined that the formation
(
(
(
(
of groups based primarily on species thermal preferences, followed by age-at-maturity
(
)
(
)
compiling a database from Coker et al (2001) to compare thermal preference,
reproductive guild, age and length at maturity, maximum age and length, feeding
preference and vegetative cover preference (high versus low). After considering the need
and degree of vegetative cover preference, yielded groups of species with a high degree
)
)
)
()
of life history similarity. Coldwater species were amalgamated into a single group based
()
on the uniformity of the examined traits (i.e. coldwater species prefer very little, if any,
()
()
vegetative cover and the majority of species present are piscivores). Much of the Bay of
their habitat preferences would have a minor role in determining an overall classification.
(
(
There are large quantities of coldwater fish habitat in the lower Bay and in the adjacent
()
Quinte does not provide coldwater fish habitat, is little used by cold water fish, and hence
areas of Lake Ontario.
Coolwater and warm water species were then further divided into eight groups
)
)
(
)
(
)
( I
based on their age at maturity and their preference for vegetative cover. Coolwater
(I
species were divided into two groups, one consisting of those with an age-at-maturity of
( I
less than or equal to 2 years and another contained species with an age of maturity at
(I
greater than 2. Warmwater species were divided into two groups, one with species whose
( I
age-at-maturity was less than or equal to 3 years and another for species with age-atmaturity greater than 3. Two exceptions were made however to these warmwater
groupings. Grass pickerel (Esox american us vermiculatus) and largemouth bass
(Micropterus salmoides), both with age-at-maturity of 3 years, were placed in the latter
group defying the age-at-maturity rule as their overall life-styles and sizes were more
( I
(I
( 1
(I
( 1
(I
(I
(I
7
(
1
(I
consistent with others in that group. The four thermallage-at-maturity groups were each
then divided into two groups based on species having a low or high preference for
vegetation cover.
A sensitivity analysis of the groupings was not performed for this study as
previous work in Severn Sound (Minns et al. 1999) established the main patterns of
sensitivity when forming fish groups to estimate suitability values. Habitat use patterns
varied most between thermal groupings compared to trophic and life stage groupings. The
)
;)
age-at-maturity criteria used here divides species for the most part between small and
large adult sizes. The differing preferences for vegetation cover represent a well-known
feature in fish communities.
Selection of Weights for Fish Groups and Life Stages
To calculate habitat suitability indices using the Bay of Quinte physical habitat
database, it was necessary to assign suitability weights for each species group present
(Table 1). Each life stage (adult, yoy, spawning) was assigned an equal weight of 0.333;
this assumes that the habitat required to complete each life stage is of equal importance to
a species. This is a conservative approach as there are no clear guide-points for setting
values, although Minns et al. (1996) showed that the absolute habitat supply requirements
of adult and yoy life stages are greater than those of spawning. Lower weightings were
assigned to the cold water groups in comparison to both the cool water and warm water
species. The upper and middle sections of the Bay of Quinte are mainly shallow and
warm in the summer, and coldwater species are a minor component ofthe fish
assemblages present. This does not mean those sections of the Bay are unimportant for
coldwater species as there may be transitory use for migration or spawning in colder
periods of the year. The main fish production takes in the warmer periods and is
dominated by warmwater and coolwater species.
Linking Fish Habitat Categories to Species' Requirements
The habitat categories obtained when the inventory is assembled in a GIS are not
necessarily the same as those specified for inputs to the DM software. However each
habitat patch created by the overlay of depth, substrate, and vegetation cover can be
described as a vector, or array, of percentages (0 to 100):
•
Depth: %0-1, %1-2, %2-5, %5-10, %10+ metres.
8
The three habitat features are treated orthogonally in DM such that suitabilities are
()
()
()
()
()
()
()
()
computed for unique combinations of elements drawn singly from each feature (Minns et
()
al 2001). Then the weighted sum of suitabilities is computed based on the proportions.
(
•
Substrate: %Bedrock, %Boulder, %Cobble, %Rubble, % Gravel, %Sand, %Silt,
%Clay, %Hard-pan clay, %Pelagic.
•
Vegetation: %No cover, %Emergents, %Submergents.
Hence suitability for the combination 0-1 m depth by cobble by no cover is computed and
)
(
(
its weighted contribution is %0-1 times %cobble times %no cover. These calculations are
(
repeated for all unique habitat patches in the inventory and for each fish group and life
(
stage combination.
(
Computation of Habitat Suitabilities
(
The fish habitat suitabilities for the upper and middle areas of the Bay of Quinte
were computed as in Minns et al. (1999a) and, Minns and Bakelaar (1999). In this study,
27 basic DM suitabilities (i.e. 9 groups by 3 life stages), as well as one composite
suitability, were calculated for each of the 3609 unique habitat characteristic
(
(
(
(
combinations found in the mapped polygons. Excluded from the analysis were map
(t
(
polygons belonging to the lower bay and those lacking data for one of the three thematic
(
layers.
(t
Habitat Suitability Classification
(
Two classification schemes based on DM habitat suitability values were assessed.
In a Severn Sound habitat classification study, Minns et al (1999a) used a composite DM
t
( t
(
suitability index based on a weighted sum of the constituent fish group by life stage
,
C'
suitability indices. Classification and regression trees (CART) were used to identify
(~
suitable cut-offs for dividing the suitability range into three categories; low, medium and
(
(
(
(
high. Here, a simplified 'natural breaks' version of the Severn Sound approach was
compared with a K-means clustering approach whereby patches were classified into
limited numbers of sets using the raw fish group by life stage suitabilities as inputs.
(
,
,
,
,
( ,
The aim of suitability classification was to divide each of the 27 indices, as well
<,
as the composite index, into three different habitat suitability classes; high, medium and
('I
(I
(I
(I
low. It was important that the cut-offs for these classes be assigned in a deliberate and
non-random fashion, thus ensuring the safeguarding of innate groupings among the data
9
(
I
(
1
(
I
sets. With this factor in mind, the ESRI's ArcView GIS® was employed to determine the
')
)
)
natural break-points of the suitability data. Using the legend type 'Graduated Color' and
using a classification scheme that seeks and divides the given data by its 'natural breaks'
into three groups, the suitability classes were formed.
)
The second form of analysis was performed on the OM suitability values to gain
an additional perspective of the overall fish habitat suitability (in addition to the
)
)
)
)
)
)
composite index). This was done by amalgamating all 27 of the individual indices
suitability rankings for each unique patch. Once the Arc-View software had determined
the natural breaks for each of the 27 indices, each unique patch was ranked as either a
low, medium, or high suitability (1, 2 and 3 respectively). Thus, each unique patch had
been ranked a total of27 different times. A K-means cluster analysis (Systat 10 ®) was
performed in an attempt to divide each unique patch into six classes based on these 27
)
)
)
)
)
)
)
)
individual rankings. Assignment of each cluster to a low, medium of high suitability
cluster was done by examining the means of the 27 suitability indices of habitat patches
included in the cluster. For example, in cluster number one, the mean suitability ranking
of the habitat within all of the clustered polygons for cold water fish was 1.00. These
individual means for each index were then tallied and an overall mean was determined.
This overall mean was the main determining factor as to how the overall cluster was
ranked.
Step Two: Fish Habitat Rarity Assessment
)
)
)
)
Where composite weighted indices are used to assess habitat suitabilities there is a
concern that uncommon habitat patches classified as extremely high in suitability for one
of the 27 individual indices will be classified as either medium or low with respect to the
overall habitat suitability as classified by either composite and cluster analysis. If
)
)
conservation efforts are solely focused upon those patches considered to be high by
composite and clustered suitabilities, vital habitat for one of the 27 combinations may be
overlooked and possibly destroyed. The rarity assessment pinpoints and highlights those
)
)
potentially 'overlooked' patches.
The identification of the 'rare' habitat patches among the individual 27 guild life
stages was performed via the construction of Lorenz curves depicting percent cumulative
)
area and percent weighted suitable area (WSA) versus habitat suitability for each of the
)
)
)
10
individual indices (Appendix G). WSA is the product of area and suitability. Habitat
was classified as rare if it fell into the upper 25 th percentile of cumulative area, possessed
a suitability of 0.75 or greater, and if the cumulative WSA curves enters the box from the
()
()
( )
( )
()
()
( )
left rather than from below. These patches were easily identified by use of the
()
constructed curves. The general shape of the cumulative curves themselves reveals
C)
information regarding the nature and the quality of the overall habitat with respect to the
()
needs and preferences of the fish species belonging to each guild and life stage. If the
( )
curve has an overall concave shape, the cumulative line will then pass into the rarity
( )
cumulative line passing into the rarity box from the left hand side, indicating that in
()
()
()
( )
()
general, very little of the habitat is of high suitability (Appendix G, 2 coolwater, age-at-
(
)
(
)
quadrant from below. In this situation, a larger amount of the habitat is of higher
suitability and hence there is no rareness present (Appendix G, 3 warmwater, age-atmaturity <=3, and high vegetation preference). An overall convex shape results in the
maturity >2 and low vegetation preference). In instances where the cumulative line fails
to enter the rarity box, none of the habitat is highly suitable (Appendix G, 3 warm water,
()
(I
age-at-maturity >3 and low vegetation preference). Thus, an area in the top right corner
()
of each curve had been classified as rare, highly suitable habitat (Appendix G). Those
( 1
patches falling within this area were then cross-referenced to the composite suitability
(I
index. If a 'rare' patch fell into the low or medium composite or cluster ranking, then it
<-
was classified as a rare patch. If the 'rare' patch fell within the high composite or cluster
( 1
classification it was not reclassified as rare since it was already considered an important
patch to conserve.
(
,
(
(
Step Three: Local Expert Knowledge Assessment
(
Data Collection
(
The local expert knowledge database was based on interviews conducted with
(
local fishers, recreational and commercial, by one of the co-authors of this report, M.
(
Ewaschuk. Contacts with fishers were initially made via the Lower Trent Conservation
Authority (for recreational fishers), and via the Lake Ontario Management Unit of the
Ontario Ministry of Natural Resources and the Commercial Fishing Association (for
(
(
(
,
,
commercial fishers), and then by individual referral thereafter. In addition, local bait and
(,
(,
tackle operators were contacted. Only fishers with at least 15 years experience on the Bay
(
,
<,
t,
11
{,
C'
")
)
;
of Quinte were interviewed to identify important fish catch locations with corresponding
habitat descriptions.
Each fisher was interviewed alone. They were asked to indicate on maps
significant fishing locations and to describe the sites where particular species were most
')
)
likely to be captured or with specific habitat features, e.g., spawning areas. The location
information was digitized into the same GIS used for mapping habitat suitability. The
fishing locations were assigned to the unique habitat polygons derived from the
)
combination of depth, substrate and vegetation data for the initial suitability mapping.
The fishing information was analyzed to assess the spatial relationships of important fish
habitats from a fisher perspective.
)
Fish Habitat Suitability Index Validation
While suitability mapping of habitat is an important step towards conservation of
essential habitat, comparison of the constructed suitability maps with site-specific fish
sampling data or expert knowledge can help validate the proposed classification. In the
earlier Severn Sound study, Minns et af. (1999a) compared the a priori habitat
suitabilities with fish catches and composition at a series of nearshore electro-fishing
sites. This analysis showed suitability values were positively correlated with catch-perunit-effort of numbers, biomass and species richness, especially for warm-water and coolwater fish groups in the YOY and adult life stages. Significant correlations were not
expected for coldwater species and the spawning life stages as the sample data were
unsuitable measures of those features of the fish community.
The fisher expert knowledge provides some basis for validation of habitat
suitability although the method is limited by the subjectivity of the observations. Fishers
mainly report sites where fishing success is greater. It cannot be automatically assumed
fishers are able to identify essential fish habitat; essential fish habitat and good fishing
locations are not necessarily the same. More suitable fish habitat may be less accessible
to fishers, for example, in areas where the water is too shallow or the vegetation is
impenetrable. However, it might be expected that successful fishing sites will be at, or
close to, preferred or more frequently used habitat. The agreement between fish habitat
class assignments and the occurrence of fishing sites was assessed by determining the
proportions of fisher locations by habitat class.
12
Step Four: Overall Rule-Based Classification
In the final classification model, the low and medium class assignments obtained
for habitat suitability, either by DM composite suitability or K-means clustering, were
()
()
()
()
()
()
()
replaced by high values as identified by the rarity and local expert knowledge steps. This
()
ensured balanced use of available information by incorporating DM suitability, rarity and
()
()
fishing importance.
( )
RESULTS
( )
Habitat Suitability using Defensible Methods:
Much insight into the habitat preferences of the species within this study can be
( )
learned by plotting the mean habitat suitability index values versus substrate and depth
()
()
for each type of vegetation class (emergent, submergent, and no cover). This was done
C)
for each species group over all of the three life stages (Appendix F). General habitat
()
preference trends were found among the species groups, reinforcing the premise that the
species had been properly grouped by life history preferences (Appendix C). After
examination of the graphs in Appendix F, several general statements can be made
regarding the habitat preferences among groups.
Coldwater species over the three life stages preferred a lack of cover over all
( )
Cl
()
(I
(
)
( I
depths of water. While the cold water adults preferred pelagic areas, the spawners and
(I
YOY preferred shallow depths with a gravel-sand substrate combination. Coolwater
(
species were divided into four species groups but general statements could be made
(
( I
regarding habitat associations. The two groups preferring low vegetative (LV) cover
(
obviously displayed high suitability towards areas lacking cover, while the coolwater
(
high vegetative (HV) preference groups generally preferred areas with both submergent
(
and emergent cover. Among both LV and HV coolwater groups, a higher preference was
(
assigned to rubble-gravel-sand-silt substrate combinations.
(
Warmwater HV groups as expected showed strong preference for areas with high
submergent and emergent vegetation along with gravel-sand-silt substrate combinations
over all life stages. Warmwater LV groups generally showed a strong preference for
(
(
(
( I
areas lacking cover but some partiality for both submergent and emergent vegetation was
(
found among a few of the spawning and YOY life stages. A wider variety of substrate
( I
(
13
')
)
)
')
')
preferences existed among the LV group compared to the HV groups but some
inclination towards gravel-sand-silt combinations was evident.
A correlation analysis was performed on suitability values for the 27 indices
obtained using a matrix of all unique combinations of depth, substrate and vegetation
j
categories as inputs for DM.
)
Habitat Suitability Classification Methods
The 'natural break' classification of habitat suitability proceeded as follows. After
')
)
examining the break points assigned by the Arc View GIS, all of the classes were adopted
for use in this report, save one, the Adult Coldwater suitability. The highest habitat
suitability for this index did not exceed 0.04, thus, as a result of its extremely low overall
suitability, all the habitat was classified as low suitability. For all of the remaining 26
indices examined, the cut-off points for each of the classes fell within the following
ranges: the low-medium suitability cut-off ranged from 0.17 to 0.34; and the medium-
)
)
high suitability cut-off ranged from 0.42 to 0.62. The classifications obtained for
individual indices are shown in two examples for adult, coolwater fishes with age-atmaturity <= 2 and a low vegetation preference (Figure 2) and for the spawning of
warmwater fishes with age-at-maturity <= 3 and a high vegetation preference (Figure 3).
Habitat for adult coolwater fishes occurs in deeper areas, often away from vegetation
(Figure 2) while habitat for spawning warmwater fishes is concentrated in shallower,
vegetated areas (Figure 3).
)
)
)
The composite suitability index ranged from 0-0.53 and the low-medium and
medium-high cut-offs were 0.22 and 0.37 respectively. The percent area of habitat
assigned to each composite suitability class (low, medium, and high) for each individual
index was calculated (Table 2). The percentages varied considerably showing the
Composite ranks were unable to parallel the rankings assigned for individual indices.
)
DM composite suitability classifications derived with the weights reported in Table 1
)
showed 45.9% percent offish habitat was classified as low, 28.8% as medium, and 25.4%
as high suitability fish habitat (Figure 4).
)
)
The K-means clustering was limited to 6 clusters which were assigned in pairs to
Low, Medium and High rankings (Figure 5). The 6 cluster result had non-trivial numbers
of members in each cluster and clustering with larger numbers of clusters generally
14
()
()
()
()
()
()
resulted in single or small groups of patches forming new isolated clusters. The clusters
can be interpreted both in terms of the dominant habitat characteristics in each cluster and
in their representation of the individual DM suitability indices. The mean percentage
vegetation and substrate compositions of the clusters highlight the habitat differences
( )
substrate: one little vegetation cover and some boulders (Low 3) and one with mostly
()
()
()
submergent and emergent vegetation cover. The first medium cluster (Med. 1) was
( )
among the clusters (Table 3). The two low clusters both have predominately silt and clay
mostly silt and clay with a high percentage of submergent vegetation cover while the
other (Med. 4) had a wide mix of substrates from boulders to silt and little vegetation
cover. One high cluster (High 2) was dominated by sand and gravel with little vegetation
(
( )
<. )
( ~
cover while the other (High 5) was dominated by sand, silt and gravel with high emergent
(
and submergent vegetation cover. The percentage of area in the three ranks based on
(
clustering fish habitat suitability indices are 47.7% , 27.4%, and 24.9% for low, medium,
(
and high respectively.
(
To examine the agreement between the individual DM suitability indices and the
clustered groups, we computed the mean suitability for each DM index in each cluster
)
(
(
(
(Table 4). When the means are ranked into three levels for each index, a general pattern
(
of agreement emerges with low clusters having predominantly low and medium means,
(
high clusters having predominantly high means, and medium clusters having intermediate
()
numbers of all three mean groups. The groupings also show how the pairs of clusters in
(
the three ranks capture much of the differences between low and high vegetation
(
preferences. High-5 is dominated by indices with high vegetation preferences and High-2
with low vegetation preferences. The pattern is similar for Medium-l and Medium-4
(
(
(
I
clusters respectively. Thus the cluster groupings appear to perform better than the
(
I
composite rankings at retaining the information content of the individual suitability
( I
indices. Both the composite and clustered rankings were carried forward in the
(
developing of the final classification.
( I
(I
Habitat Rarity
Polygons were ranked as rare if they had both a suitability value of>=0.75 and
were in the upper 25 th percentile of cumulative percent area (Appendix G) for one or
more of the 27 individual suitability indices (Figure 6). Many of the polygons identified
(
(I
(
1
( I
(
15
1
1
(I
(I
(
1
as rare were already classified as high by the composite or cluster rankings. Few
polygons were reassigned from low or medium for either ranking scheme (Figures 7 and
8). With the composite rankings, a breakout of the areas assigned rare shows how little
low and medium area was changed to high when rarity was factored into the
classification (Table 2). Similarly with the cluster rankings, very little area was
reassigned from low or medium to high when rarity was added (Table 5).
Local Expert Knowledge
)
)
Fishing success information was obtained from 30 individuals who referenced
742 specific locations in the upper and middle Bay of Quinte. The locations either
referred to specific points or areas. The locations were assigned to corresponding
polygons obtained when the depth, substrate, and vegetation maps layers were overlaid.
)
A local expert knowledge map was constructed to depict those polygons containing at
least one confirmed observation of a fishing site (Figure 9). The expert fishing polygons
)
)
)
)
)
)
)
)
)
)
accounted for 6.9,5.0, and 5.1 percent of the low, medium and high composite rank areas
respectively (Table 6). Within the clustered rank areas, the respective percentages were
7.0,5.1, and 4.9. Fishing areas do not strongly coincide with areas ranked as having high
habitat suitability. However, the expert patches that did not correspond with high areas
were usually adjacent to high or medium areas.
Combined Ranking Maps
When the suitability, rarity and fisher expert assignments are combined for
composite and cluster rankings similar classification maps are obtained (Figures 10 and
11). The agreement between the two maps and the effects of combining suitability ranks
with rarity and expert assignments can be assessed using cross-tabulations of area and
percentage of area (Table 6). The agreement between the maps suggests the two
approaches taken to suitability classification by and large produce similar results. There
was a 92.3% overlap between the low-medium-high rankings for the composite and
cluster schemes. Rarity accounted for 23.2% of the area though much was assigned to a
high rank before the rarity criteria was applied. In the composite ranking scheme, 45.8%
of the area was ranked low with 28.8% and 25.3% assigned to medium and high
respectively. In the clustered scheme, the percentages were 47.7, 27.4, and 24.9
respectively. Given the ability of the cluster rankings to retain more of the individual
16
index information the final classification based on clusters, rarity and expert assignments
was preferred for an overall habitat suitability rating of the upper and middle areas of the
Bay of Quinte (Figure 11).
DISCUSSION
Bay of Quinte Habitat Inventory and GIS Database
In Canada, Ontario and the Great Lakes region the information infrastructure for
assembling GIS databases of aquatic habitats is still being developed. The geographic
database assembled for the upper and middle Bay sections of the Bay of Quinte has some
()
()
()
()
()
()
()
()
()
()
( )
( )
( )
of the problems that were evident in the previous study in Severn Sound (Minns et al.
(J
1999) but there were also some advances made. These problems and advances are
( )
outlined below.
()
The development of fish habitat GIS databases in the Bay of Quinte was
hampered by the lack of seamless, fine resolution elevation models especially in the zone
()
()
( )
between the 100-year high water mark and approximately 2 metres below the navigation
<-
chart datum. The Canadian Hydrographic Survey does not sample in depths less than 2m
(
for safety reasons therefore these areas are poorly mapped. Unfortunately, these poorly
mapped areas are in the nearshore zone which contains very important fish habitat.
(
(
Further, nearshore areas with extensive vegetation are extremely difficult to map by
(
conventional methods. Again, coastal wetland areas are poorly mapped despite their
(
(
importance for fish.
(
The substrate maps used in this study were derived from a combination of near-
(
shore visual mapping and extrapolation of low density, offshore grid sampling. Continued
(
advancements in substrate and vegetation mapping using combinations of electro-
(
acoustic and remote sensing technologies are improving the mapped representations of
(
these data but such methods are far from routine.
(
Vegetation mapping is especially difficult to complete in habitat mapping as
cover, density and composition are continually changing both within seasons and across
(
<(
years. In the Bay of Quinte much of the site level vegetation mapping predated the
(
invasion of zebra mussels, which triggered a large-scale expansion of macrophyte
(I
coverage. Further, on-going changes in the water level regime impact the location and
(
extent of wetland vegetation types. In this study, the results of a macrophyte modelling
(
(
(
17
( 1
<.1
study in the Bay of Quinte (Seifried 2002) were used to predict macrophyte spatial cover
for the post-zebra mussel regime. It was judged more important to show the situation
more as it is now rather than as it was in the early 1990s when most field surveys were
done.
Mapped coverage information was limited to the upper and middle Bay areas as
there were no nearshore inventory survey data or offshore substrate mapping for the
)
)
)
lower Bay area. Despite these limitations the habitat GIS database provided an acceptable
basis for assessing habitats and classifying fish habitats on a broad scale in the Bay of
Quinte. Extension of fish habitat mapping up into the tributaries will require additional
surveys of instream and riparian habitat features, barriers and other obstacles to
)
)
)
connectivity, and the fish communities present.
The main advance in the Bay of Quinte over the Severn Sound study was the
extension of the study area to the offshore zone. In the Severn Sound study, the habitat
mapping was limited to a corridor along the shoreline with depths less than 1.5 metres. In
)
)
)
)
)
this study, onshore and offshore datasets were combined with available depth maps to
create complete coverage for the upper and middle Bay areas. Complete coverage is
essential as fish species make use of many habitat areas through their life cycle and
throughout the year. Limiting coverage to inshore areas can create false impressions
about the importance of various habitat features to particular fish guilds and/or life stages.
Habitat Classification Model
)
The classification model used built upon the approach presented for Severn Sound
(Minns et al. 1999) and several additions or improvements were made. The number of
fish groups considered in Quinte was greater than in the Severn Sound study. First, this
reflected the greater number of fish species present overall in the Bay of Quinte. Second,
there was recognition from the earlier study that more attention needed to be given to the
)
life histories and habitat preferences of the species present if useful habitat classifications
were to be developed. A habitat classification scheme has to be practical from a
management viewpoint, implying fewer rather than more habitat classes. At the same
time, every effort must be made to accommodate the specific needs of all fish species
present. Increasing the number of fish groups ensured that more specialized habitat needs
were considered in the overall classification as both suitability and rarity were assessed.
18
More extensive use of local expert knowledge was made in this study than in the
Severn Sound case. Minns et al. (1999) provided for an expert information layer but the
database was insufficient to implement it in Severn Sound. In the Bay of Quinte, the
()
()
()
()
(
)
()
( )
fisher interview activities produced a useful assessment of fish habitats from a fishing
()
success perspective. Additional efforts to validate the suitability maps through systematic
( )
( )
fish communities surveys, such as existing electro-fishing, seining, trawling, and
commercial catch data, and through further gathering of fishers' knowledge will increase
the acceptance of habitat maps and enhance their utility for agencies and fishers
( )
( )
( )
Finally, the repertoire of methods for developing habitat classifications and
( )
assigning habitat patches to classes was further expanded. In the Severn Sound study,
( )
efforts were focused on the use of CART regressions to build overall groupings with
( )
regard to the underlying data on bathymetry, substrate and vegetation. Here, two
( )
additional approaches were explored in parallel: a) natural breaks in frequency
()
distributions as analyzed in Arc-View GIS software and b) K-means cluster analysis of
the fish groups by life stage suitability values. Both methods showed they could produce
( )
( )
(
)
acceptable classifications. The K-means cluster analysis results were preferred as they
( )
appear to better capture the distributional discontinuities in the underlying datasets, both
(
habitat characteristics and suitability indices. This is an important consideration as
(
observers are able to identify most habitat discontinuities in the field (e.g. changes in
(
substrate composition and vegetation coverage). Arbitrary classifications that lumped
(
features would have less acceptance operationally. The typical substrate and cover
compositions of the six cluster groupings (Table 3) are readily recognizable with modest
(
(
(
levels of training.
(
Implications of Results in the Bay of Quinte
(
The resulting habitat maps and the analyses of area by habitat class show that
there is much good fish habitat in the upper and middle Bay areas of the Bay of Quinte
and that there is good overall agreement between maps of habitat suitability and maps of
fishing success. The Bay of Quinte is a highly productive part of the larger Lake Ontario
ecosystem and contains considerable high quality habitat resources as reflected in the
(
(
(
(
(
(
extensive macrophyte cover and emergent wetlands. The Bay of Quinte has not had
(
excessive development that has degraded other areas elsewhere on Lake Ontario, (such as
(
(
(
19
(
( t
)
')
)
Hamilton Harbour and Toronto Harbour). However, regulation of water levels for the St.
j
)
Lawrence Seaway have greatly reduced the diversity of the wetlands and shoreline
modifications have been extensive. Given the efforts that have gone into restoring water
quality in the Bay of Quinte, local communities are well-placed to ensure that future
developments in and around the adjacent urban centres (such as Trenton, Belleville,
)
Napanee and Picton) do not further destroy or degrade important fish habitats. Of course,
those communities need to be provided with technically-sound assessments of their fish
habitat as illustrated in this habitat suitability mapping project and encouraged through
)
)
proactive area fish habitat management plans to support enforcement of the federal
fisheries act provisions for the protection of fish habitats.
Next Steps
)
)
)
)
)
)
)
)
)
There are several steps that should be pursued to build on the habitat classification
model developed here for the Bay of Quinte:
•
The information assembled in this study and the resultant analyses and maps should
be made available for use by agency (DFO, CAs) and the public via suitable media
(There are current efforts under way with DFO-Fish Habitat Management - Ontario
Great Lakes Area to make this material accessible to DFO via a website).
•
The Bay of Quinte habitat inventory and GIS database should be updated with further
surveys to improve the elevation model and to provide more up-to-date mapping of
substrate and vegetation conditions. Surveys should be extended to include the lower
)
Bay area. The habitat inventories should also be extended further into the tributaries,
)
)
)
above major barriers to fish movement to evaluate the potential benefits of barrier
removal/mitigation and to prioritize rehabilitation efforts.
•
Several conservation authorities with joint responsibility for the Bay of Quinte
habitats should develop the infrastructure for supporting and/or using the habitat
mapping and classification model on a watershed basis for future development and
)
use in planning by local and regional government agencies. Habitat and suitability
mapping tools should become part of a suite of decision-making tools used by DFO
Fish Habitat Management and their partners.
)
•
The classification scheme and its resulting maps should be integrated with fish habitat
and fisheries management plans to combine opportunities to conserve, restore and
J
)
)
20
(,
(l
enhance fish habitats in support of fishery objectives for the Bay of Quinte and
( )
beyond.
•
•
The local expert knowledge database could be expanded by increasing the numbers of
()
()
fishers interviewed, by examining particular species in greater detail, and by
(l
combining fisher observations with a variety of standardized fish sampling datasets
()
gathered by government agencies and university groups.
( )
The mapping of habitat features can be extended to consider spatial and temporal
patterns in thermal conditions given that temperature is a primary determinant of fish
•
()
( )
( )
( )
distribution and productivity in aquatic ecosystems. Additional analysis work with
()
existing fisheries survey datasets can be undertaken to provide local validation and
( )
refinement of the suitability maps.
( )
Integrated fish habitat assessments, such as the one developed here, for parts of the
()
Bay of Quinte, can be used to ensure that high quality habitats are given maximum
()
protection from the inroads of development and that other fish habitats are properly
assessed when developments are proposed. A first step in the Bay of Quinte would be
(
()
(
to make certain that none of the shorelines adjoining high quality fish habitat are
(
developed without adequate provision for habitat compensation in adjacent areas,
(
either through the creation of new habitat or by the restoration of previously degraded
(
habitats elsewhere in the Bay of Quinte. A further step requires implementation of a
(
Bay of Quinte area fish habitat management plan with all agencies assuming
responsibility for ensuring no net loss in the future and, wherever possible, securing
net gains through habitat creation and compensation to redress past losses.
)
(
(
(
(
(
(
ACKNOWLEDGEMENTS
This work was mainly funded from Fisheries and Oceans' Great Lakes Action
(
Plans, 2000-2005 and 2005-2006. The products presented here were developed using the
(
(
cooperation, support, and results of many efforts to address fish habitat issues in many
(
agencies over the last couple of decades: Ontario Ministry of Natural Resources,
( 1
especially in Napanee and Glenora Fisheries Stations; Conservation Authorities (CAs)
(
1
whose mandate includes portions of the Bay of Quinte, especially the Lower Trent CA,
(
1
(
1
Ontario Ministry of Energy and Environment especially those, like K.H. Nicholls, who
21
(I
(
1
(
I
{
I
)
supported ongoing efforts to track changes in macrophyte cover, and Fisheries and
Oceans Canada both in the Great Lakes Laboratory for Fisheries and Aquatic Sciences
and the Ontario-Great Lakes Area of Fish Habitat Management. Thanks also to Dr. Susan
Doka, Dr. Robert Randall, Paul Johanson and Mark Ferguson for their thorough reviews
and comments.
REFERENCES
)
)
Bay of Quinte RAP, 1993. Time to Act. The Bay of Qu in te Remedial Action Plan Stage 2
Report. Ontario Ministry of Environment, Kingston, Ontario. 257 pp.
Coker, G.A, C.B. Portt, and C.K Minns. 2001. Morphological and Ecological
)
)
)
)
)
)
)
)
)
)
)
)
Characteristics of Canadian Freshwater Fishes. Can. MS Rpt. Fish. Aquat. Sci
2554: iv+86p.
Department of Fisheries and Oceans, 1986. Policy for the management of fish habitat.
Department of Fisheries and Oceans, Ottawa, Ontario. 28 pp.
Lane, J.A, C.B. Portt, and C.K Minns. 1996a. Nursery habitat characteristics of Great
Lakes fishes. Can MS Rep Fish Aquat Sci. 2338:42p.
Lane, J .A, c.B. Portt, and C.K. Minns. 1996b. Adult habitat characteristics of Great
Lakes fishes. Can MS Rep Fish Aquat Sci. 2358:43p.
Lane J.A, C.B. Portt, and C.K. Minns. 1996c. Spawning habitat characteristics of Great
Lakes fishes. Can MS Rep Fish Aquat Sci. 2368:48p.
Minns, C.K, D.A Hurley, and KH. Nicholls, (Eds.) 1986. Project Quinte: point-source
)
phosphorus control and ecosystem response in the Bay of Quinte, Lake Ontario. Can.
)
Spec. Publ. Fish. Aquat. Sci. 86: 270p.
)
)
)
)
)
Minns, C.K, R.G. Randall, lE. Moore & Cairns. 1996. A model simulating the impact of
habitat supply limits on northern pike, Locus Lucius, in Hamilton Harbour, Lake
Ontario. Can J. Fish. Aquat. Sci. 53 (Suppl 1):20-34.
Minns, C.K and Bakelaar, C.N. 1999. A method for quantifYing the supply of suitable
habitat for fish stocks in Lake Erie, pages 481-496 In Munawar, M, Edsall, T, and
Munawar, I.F. (eds.). State of Lake Erie: Past, Present and Future. Backhuys Publishers,
The Netherlands. 550p.
Minns, C.K and Nairn, R.B. 1999. Defensible Methods: applications of a procedure for
assessing developments affecting littoral fish habitat on the lower Great Lakes, pages
)
:>
:>
22
(
('
e)
()
()
()
15-35 In Murphy,T.P.and Munawar, M. (eds.) Aquatic Restoration in Canada. Backhuys
Publishers, The Netherlands. 211p.
Minns, c.K., Brunette, P.c.E., Stoneman, M., Sherman, K., Craig. R., Portt, C.B., and
( )
Randall, R.G. 1999a. Development ofa fish habitat classification model for littoral areas
( )
of Severn Sound, Georgian Bay, a Great Lakes' Area of Concern. Can. MS. Rpt. Fish.
(j
Aquat. Sci. 2490: ix+86p.
( )
Minns, C.K., Doka, S.E., Bakelaar, C.N., Brunette, P.C.E., and Schertzer, W.M. 1999b.
Identifying habitats essential for pike, Esox lucius L., in the Long Point region of Lake
()
( )
(
Erie: a suitable supply approach. Pages 363-382. In L. Benaka, editor. American
)
c)
Fisheries Society Symposium 22:Fish Habitat: Essential Fish Habitat and Rehabilitation.
()
Bethesda, Maryland. 459p.
(
)
(
)
Minns, C.K., J.E. Moore, M. Stoneman, and B. Cudmore-Vokey. 2001. Defensible Methods
of Assessing Fish Habitat: Lacustrine Habitats in the Great Lakes Basin - Conceptual
()
Basis and Approach Using a Habitat Suitability Matrix (HSM) Method. Can. MS Rpt.
()
Fish. Aquat. Sci.2559:viii+70p.
Morrison, H., C.K. Minns, and 1.F. Koonce. 2001. A methodology for identifying and
()
()
()
classifying aquatic biodiversity investment areas: Application in the Great Lakes basin.
(
Aquat. Ecosystem Health and Managem. 4(1):1-12.
(I
()
Rosenfeld, 1.S. and T. Hatfield. 2006. Information needs for assessing critical habitat of
freshwater fish. Can. 1. Fish. Aquat. Sci. 63:683-698.
Seifried (Now Liesti), K.E. 2002. Submerged macrophytes in the Bay of Quinte: 1972-
)
( 1
(I
( 1
2000. Master's thesis. University of Toronto, Toronto.
Smith, A. 1993. Bay of Quinte Remedial Action Plan; Summary of the Nearshore Habitat
Inventory on the Bay of Quinte 1991-1003. OMNR. Napanee, Ontario.
( 1
( 1
( I
( 1
( 1
(I
(I
(I
(
1
(I
(
1
(I
(
23
I
(I
(I
Table 1 List of weighting applied to the nine fish groups for computing a composite
suitability in the OM's estimation offish habitat suitabilities. (Life stages were
assigned equal weights of 0.3333).
)
)
')
Species Group
Cold water
Cool water / Age of Maturity <= 2 / Low Vegetation
Cool water / Age of Maturity <=2 / High Vegetation
Cool water / Age of Maturity > 2 / Low Vegetation
Cool water / Age of Maturity > 2 / High Vegetation
Wann water / Age of Maturity <=3/ Low Vegetation
Wann water / Age of Maturity <=3/ High Vegetation
Wann water / Age of Maturity> 3 / Low Vegetation
Wann water / Age of Maturity > 3 / High Vegetation
Total
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
24
Weightings
0.0588
0.ll76
0.ll76
0.1176
0.1176
0.ll76
0.1176
0.ll76
0.ll76
1.0000
Table 2 Percentages of habitat area in the Bay of Quinte ranked low, medium or high using the composite index alone, and then with
the areas assigned as rare reassigned to high, separated out for each of the 27 individual habitat suitability indices and the
composite index.
Life
stage
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
yay
yay
yay
yay
yay
yay
yay
yay
yay
ComEosite
Habitat suitabilit~ grouEings
Thermal
Vegetation
Age at
Ereference
Ereference
maturi~
Cold
Cool
<= 2
Low
Cool
<=2
High
>2
Low
Cool
>2
High
Cool
<=3
Warm
Low
<=3
High
Warm
>3
Low
Warm
Warm
>3
High
Cold
<=2
Low
Cool
Cool
<=2
High
>2
Low
Cool
>2
Cool
High
<=3
Warm
Low
Warm
<=3
High
>3
Warm
Low
Warm
>3
High
Cold
<=2
Cool
Low
Cool
<=2
High
Cool
>2
Low
Cool
>2
High
Warm
<=3
Low
Warm
<=3
High
Warm
>3
Low
Warm
>3
High
% Area b~ comEosite onl~
Low
Medium
High
100.00
62.03
57.74
85.38
58.31
39.50
58.76
89.53
58.99
84.76
68.39
58.01
99.10
24.17
76.13
54.73
1.00
39.84
84.61
38.68
57.13
72.99
56.20
58.44
58.00
94.19
58.47
45.85
0.00
26.21
24.86
5.44
28.24
51.45
22.83
10.45
27.71
9.12
21.51
23.58
0.59
63.36
13.44
10.13
63.38
23.73
7.61
28.04
20.88
15.62
19.84
18.55
24.78
5.81
35.03
28.79
0.00
11.76
17.40
9.18
13.45
9.05
18.41
0.02
13.30
6.12
10.10
18.41
0.31
12.47
10.43
35.14
35.62
36.43
7.78
33.28
21.99
1l.39
23.96
23.01
17.22
0.00
6.50
25.36
% Area b~ comEosite and rari~
High
Rare
Low
Medium
100
62.03
57.74
85.38
58.31
39.49
58.75
89.53
58.99
84.76
68.39
58.01
99.10
24.17
76.12
54.73
1.01
39.84
84.61
38.68
57.13
72.99
56.20
58.44
58.00
94.19
58.47
45.85
0.00
26.21
24.86
5.44
28.24
51.45
22.83
10.45
27.71
9.12
21.51
23.58
0.60
63.37
13.44
10.13
63.38
23.73
7.61
28.04
20.89
15.62
19.84
18.55
24.78
5.81
35.03
28.79
0.00
6.10
7.89
3.61
12.65
8.71
7.72
0.02
13.27
4.84
4.25
6.33
0.16
5.82
10.37
26.47
22.38
36.06
7.77
32.02
21.58
5.32
23.16
7.49
7.28
0.00
6.09
25.36
25
-~~~~~~~~~-----------~~~~~--------~-
0.00
5.66
9.51
5.57
0.80
0.34
10.69
0.00
0.02
1.28
5.85
12.07
0.15
6.65
0.07
8.87
13.24
0.37
0.01
1.26
0.41
6.07
0.80
15.52
9.95
0.00
0.41
0.00
)
)
j
Table 3 Average percentage composition of the three habitat characteristics for each of the six Kmeans clusters (The cluster numbers referred to their order in the K-means clustering).
Habitat
Characteristic
Cluster Rank (number)
LowP2
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
Depth
0-lm
I-2m
2-5
5-10m
10+m
Vegetation
Emergent
Submergent
No Cover
Substrate
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
HardEan
Low ~62
Med ~12
Hi~h ~22
Hi~h ~52
36.0
14.1
25.2
14.8
3.7
57.0
18.9
23.5
0.9
0.0
43.0
26.5
29.1
1.6
0.0
50.0
20.2
19.8
8.7
1.6
38.0
17.5
26.9
15.2
2.3
6.0
23.6
9.7
0.8
0.0
8.9
21.2
69.9
22.1
35.2
42.7
8.3
74.4
17.3
7.7
21.1
71.2
2.7
11.7
85.6
22.9
56.1
21.0
3.6
11.0
5.6
5.6
1.9
4.8
30.9
34.8
0.0
0.4
1.0
3.5
3.5
3.9
8.2
41.8
37.7
0.0
0.1
0.1
0.4
0.4
0.1
8.0
53.0
37.8
0.0
0.2
8.3
14.4
14.4
18.4
21.4
13.0
9.8
0.0
0.0
0.8
2.4
2.4
20.2
62.4
7.3
4.6
0.0
0.0
0.6
0.9
0.9
11.9
64.1
15.4
6.2
0.0
)
)
)
)
)
)
)
)
Med ~42
26
Table 4 Mean habitat suitability value in each K-means cluster group for all 27 DM indices and the numbers of indices having the
lowest pair of means (italics), the middle pair (underlined) and the highest two pair (bold).
Life
Stage
Habitat Suitability Groupings
Thermal
Age
Vegtn.
Pref.
Mat.
Pref.
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Adult
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
Spawning
YOY
YOY
YOY
YOY
YOY
YOY
YOY
YOY
YOY
Cold
Cool
Cool
Cool
Cool
Warm
Warm
Warm
Warm
Cold
Cool
Cool
Cool
Cool
Warm
Warm
Warm
Warm
Cold
Cool
Cool
Cool
Cool
Warm
Warm
Warm
Warm
<=2
<=2
>2
>2
<=3
<=3
>3
>3
<=2
<=2
>2
>2
<=3
<=3
>3
>3
<=2
<=2
>2
>2
<=3
<=3
>3
>3
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
Low
High
# lows
#meds
# highs
Low (3)
0.007
0.239
0.109
0.158
0.092
0.315
0.058
0.049
0.054
0.206
0.094
0.193
0.034
0.392
0.064
0.064
0.360
0.121
0.098
0.173
0.136
0.152
0.123
0.108
0.040
0.094
0.128
18
9
o
Mean suitability value by cluster rank(#)
Low (6)
Med. (1)
Med. (4)
High (2)
High (5)
0.020
0.756
0.278
0.741
0.211
0.673
0.124
0.439
0.117
0.677
0.746
0.604
0.256
0.717
0.599
0.120
0.762
0.162
0.502
0.416
0.198
0.723
0.185
0.773
0.096
0.254
0.156
0.026
0.419
0.610
0.195
0.508
0.278
0.700
0.119
0.521
0.318
0.549
0.758
0.041
0.595
0.219
0.665
0.706
0.548
0.159
0.557
0.445
0.435
0.392
0.587
0.626
0.057
0.337
3
o
7
17
18
0.012
0.217
0.256
0.115
0.281
0.271
0.292
0.034
0.248
0.194
0.200
0.299
0.013
0.435
0.072
0.289
0.425
0.347
0.089
0.322
0.287
0.226
0.244
0.173
0.250
0.050
0.269
12
15
o
27
0.018
0.168
0.430
0.055
0.493
0.208
0.503
0.011
0.463
0.175
0.292
0.324
0.003
0.306
0.059
0.459
0.380
0.495
0.059
0.496
0.490
0.176
0.471
0.218
0.482
0.020
0.485
12
5
10
0.013
0.522
0.239
0.386
0.176
0.483
0.167
0.247
0.114
0.490
0.515
0.441
0.252
0.449
0.304
0.152
0.632
0.142
0.255
0.301
0.170
0.315
0.125
0.379
0.097
0.223
0.109
8
10
9
9
- -,
\
"'-""'r"-"',
~\
~\..-..\
...-....,
---~---
~
vv
'-' v
v
vv
v
Table 5 Percentages of habitat area within each classification (low, medium, or high) of the clustered index that either passes (rare) or
fails (-) the rarity threshold for each of the 27 suitability indices. Shaded percentages are those areas reclassified as high when
the rarity threshold is applied.
Habitat Suitability Groupings
Low (3)
Mat.
Rare
Tern!;!
Stase
Yes
Adult
Cold
na
na
Adult
Cool
<=2
Low
43.24
0.00
Adu<
Cool
<=2
High
43.24
0.00
Adu<
>2
Cool
Low
43.24
0.00
Adu<
>2
Cool
High
43.24
0.00
Adu<
<=3
Warm
Low
43.24
0.00
Adu<
<=3
Warm
High
43.24
0.00
Adu<
Warm
>3
Low
na
na
>3
High
Adu<
43.24
0.00
Warm
Spawning
Cold
43.24
0.00
Spawning
Cool
<=2
Low
43.24
0.00
Spawning
Cool
<=2
High
43.24
0.00
Spawning
Cool
>2
Low
43.24
0.00
Spawning
Cool
>2
High
43.24
0.00
Spawning
Warm
<=3
Low
43.24
0.00
Spawning
Warm
<=3
High
43.24
0.00
Spawning
Warm
>3
Low
43.20
[~
Spawning
>3
Warm
High
43.24
0.00
YOY
Cold
43.24
0.00
YOY
<=2
Cool
Low
43.24
0.00
YOY
Cool
<=2
High
43.24
0.00
YOY
>2
Cool
Low
43.24
0.00
YOY
>2
Cool
High
43.24
0.00
YOY
Warm
<=3
Low
43.24
0.00
YOY
Warm
<=3
High
43.24
0.00
YOY
Warm
>3
Low
43.24
0.00
YOY
Warm
>3
43.24
0.00
HiSh
*indicates a percentage that is less than two decimal places,
Percent Area of Habitat in each suitability cluster
High (2)
Low (6)
Medium (1)
Medium (4)
Rare
Rare
Rare
Rare
0
na
na
na
na
na
na
na
na
4.45
0.00 24.95
0.00
2.43
0.00
3.86
5.66
b
~
j
2.43
0.00
9.52
0.00
4.45
0.00 24.82
..:.........
4.45
0.00 24.82
0.00
2.43
0.00
3.95
5.57
4.45
0.00 24.82
0.00
2.43
0.00
9.52
0.00
4.45
2.36
9.24
0.28
0.00 24.95
0.00
0.07
'!
4.45
0.00 23.94
1.01~
2.43
9.52
0.00
0.00
na
na
na
na
na
na
na
na
4.45
0.00 24.95
0.00
2.43
0.00
9.52
0.00
4.45
0.00 24.95
2.35
0.00
0.07
8.31
1.21
4.45
0.00 24.95
0.00
2.28
3.82
5.70
0.15
4.45
0.00 24.95
0.00
2.43 :'0.00
9.45
0.70
4.45
0.00 24.95
0.00
2.43
0.00
9.37
0.15
4.45
0.00 24.95
0.00
2.43 i [!ill
2.89
6.63
4.45
0.00 24.95
0.00
2.43
0.00
9.45
0.07
4.45
0.00 24.95
0.00
2.43
0.00
9.52
0.00
~'Z
Q,j'ii 4.77
3.97
0.~8
24.94
0.00
2.09
4.75
4.45 ~QXQO 24.65
Q:3P 2.43 0.00 9.52 0.00
4.45
0.00 24.95
0.00
2.43
0.00
9.51
0.01
4.45
0.00 24.82
2.43
0.00
9.50
0.02
~113
4.45
0.00 24.82
!I:U 2.43 0.00 9.52 0.00
4.45
0.00 24.95
0.00
2.43
0.00
3.45
6.07
4.45
0.00 24.95
2.43
0.00
9.52
0.00
4.45
0.00 24.95
0.00
2.43
0.00
2.80
6.72
4.45
0.00 24.52
6.flO 2.43 0.00 9.52 0.00
4.45
0.00 24.95
0.00
2.43
0.00
9.52
0.00
4.45
0.00 24.82
~U
2.43 :O~O
9.52
0.00
na indicates an index with no rare habitat. Indicated.
roW
28
High (5)
Rare
na
na
15.42
0.00
6.04
9.38
15.42
0.00
14.75
0.67
15.42
0.00
5.73
9.69
na
na
15.40
0.02
15.42
0.00
15.42
0.00
3.42 12.00
15.42
0.00
11.05
4.37
15.42
0.00
8,68
6.74
7.79
7.63
15.35
0.07
15.42
0.00
14.32
1.10
15.14
0.28
15.42 *0.00
14.85
0.67
6.61
8.81
5.85
9.57
15.42 *0.00
15.14
0.28
( )
Table 6 Cross-tabulations of unique polygons by area and percentage of total area by the composite ranks
(columns) and by cluster ranks and numbers (rows), and also with rarity and fish expert
classifications. (The upper half provides weight suitable hectares and the lower half percentages).
Cluster
Rank
Area ha
Low
Med.
High.
Rare
No.
No
3
6
6
Sum
Yes
Yes
Sum
1
No
No
1
Yes
4
No
4
Sum
2
2
5
Yes
Sum
5
Yes
Sum
No
Yes
No
---------------
Percent
Low
Low
Yes
Low
No
Med.
No
Med.
Yes
High
No
High
Yes
Sum
No
Sum
Yes
()
No
Sum
Sum
Yes
Sum
Sum
2.9
23.6
7070.9 1230.1
8300.9
2.9
5.6
8.5
644.7
118.0
762.7
297.9
24.2
91.5
0.3
91.8
0.3
54.4
7810.0 1354.0
9163 .9
1305.9
355 .3
48.0
3611.1
934.6
4545 .7
272.2
- 3334.9 887.1
47.5
239.1
8.9
248.0
188.2
50.9
3.9
5.0
364.0
24.7
388.7
36.9
21.0
23.6
306.2
1.1
66.2
0.3
0.0
65.7
11.1
0.1
11.1
77.2
23.6 3757.7 902.1
497.4
4280.4
979.3
5259.7
25.3
53 .7
282.1
16.7
56.4
338.4
16.7
355.1
1421.4
51.9
0.1 1295.5
125.9
51.8
1473.3
376.5
10.0
366.6
28.3
28.3
404.8
35 .0
- 1673.1 850.6
1708.1
850.6
2558.7
452.9
16.7 3391.5 930.7
3844.4
947.4
4791.9
------------------------ -----------------------------_.
7439.7 1323.9 4234.0 951.5
732.0
76.9
12405.7 2352.3 14758.0
40.3
5.6
331.9
15.3 3156.9 907.5
3529.1
928.4
4457.5
7068.0
2.9
346.7
37.1
7454.7
4.0
1206.5
5.6
93.9
7480.0
1329.5
3
No
36.8
6.3
3
Yes
t
t
4565.9
966.8
3888.9
984.4
15934.8
3280.7
Sum
Sum
38.9
6.9
0. 1
19215.5
36.8
6.4
t
t
6
No
1.8
0.5
1.6
0.1
3.4
0.6
Yes
0.2
0.3
t
0.5
t
6
Sum Sum
38.8
6.8
1.8
0.2
40.6
7.0
Med.
1
No
17.4
4.6
1.4
0.2
18.8
4.9
Yes
0.3
1.0
1.2
t
4
No
0.1
0.1
1.6
0.2
1.9
0.1
Yes
t
0.3
0.1
t
0.3
0.1
4
Sum Sum
0.1
0.1
19.6
4.7
2.6
0.3
22.3
5.1
1.5
0.1
0.3
1.8
0.1
High.
2
No
2
Yes
0.7
6.7
0.3
7.4
0.3
0.1
1.9
0.1
2.0
0.1
5
No
5
Yes
0.2
8.7
4.4
8.9
4.4
Sum Sum
2.4
0.1
17.6
4.8
20.0
4.9
-------------------- --------------------------------------------Sum
Sum
No
38.7
6.9
22.0
5.0
3.8
0.4
64.6
12.2
0.1
4.8
0.2
1.7
16.4
4.7
18.4
Sum
Sum
Yes
Sum
( )
Sum
Sum
(I
3
Sum
Sum
Sum
Comp=
Expt=
()
()
()
()
()
23 .8
t is trace <0.1 %; - is zero
5.0
20.2
5.1
82.9
17.1
()
()
(
(
( )
( )
( )
(
(
(
(
(
(
(
43.2
(
t
(
4.0
0.5
47.7
23 .7
1.3
2.0
0.4
27.4
1.8
7.7
2.1
13.3
24.9
---,
76.8
23.2
100.0
(
(
(
(
( I
( I
( I
( I
(
I
(
I
(
I
(
(
,
,
(
1
<(I
29
( ,
t,
Figure 1 A map of the Bay of Quinte, Lake Ontario, showing the study area, the upper and
middle bay sections.
;)
)
)
Lake Ontario
)
N
)
)
t
Lake Olliario
o
)
)
)
)
)
30
4
8 km
I
e(l
(l
Figure 2 Areas in the Bay of Quinte classified as low, medium, and high suitability for the group
of adult, cool-water fishes with age-at-maturity <= 2 years and a low vegetation preference.
()
()
(. )
( )
( )
Adult_Cool
Low
Med.
"High
()
()
( )
()
( )
(
)
()
(
(
N
4000
o
4000
)
()
(
8000 Meters
)
)
(I
Figure 3 Areas in the Bay of Quinte classified as low, medium, and high suitability for the group
of spawning, warm-water fishes with age-at-maturity <= 3 years and a high vegetation
preference.
()
()
()
(
()
(.
.
( I
Spawn_Warm
Low
Med.
_High
:-=-
( I
(I
( I
( I
(I
( I
(I
(I
N
4000
o
4000
A
8000 Meters
(I
( I
(I
(I
(I
(I
31
(I
'-
)
j
)
Figure 4 Areas in the Bay of Quinte classified as low, medium and high using the composite
habitat suitability index obtained using DM.
)
)
)
)
Composite
Low
Med.
"High
)
J
)
N
4000
o
4000
A
8000 Meters
)
Figure 5 Areas in the Bay of Quinte classified as low, medium and high using the K-means
cluster analysis of all suitability indices.
)
)
,--I
Clusters
Low 3
[~=j Low 6
Med.1
Med.4
High2
HighS
N
4000
)
)
)
o
4000
8000 Meters
32
Figure 6 Habitat polygons in the Bay of Quinte ranked as "rare" at least once for any of the 27
individual fish group*life stage habitat suitability indices.
Rarity
i- -00--1 Common
_Rare
H
--
()
()
()
()
()
()
( )
()
()
()
()
()
( )
()
()
N
4000
0
4000
8000
Meters
,
()
()
( )
( )
(
)
()
()
(
)
(t
(t
(t
(t
(t
(
1
(
t
(
t
(I
(I
(I
(I
(I
(I
(I
(I
(
33
1
(
(I
(I
Figure 7 Habitat polygons in the Bay of Quinte classified as rare when cross-referenced with the
DM composite suitability class assignments of low and medium.
Rare+Comp
r------- -1 Common
_Rare
)
)
)
N
4000
o
4000
8000 Meters
)
)
Figure 8 Habitat polygons in the Bay of Quinte classified as rare when cross-referenced to the Kmeans clusters assigned low and medium.
)
Rare+Clus
[-------1 Common
_Rare
-,
4000
o
4000
A
8000 Meters
34
(
(
()
Figure 9 Habitat polygons in the Bay of Quinte containing at least one expert fisher validated
fishing site.
(l
(l
(l
Expert
I
~ Absence
_
Presence
(l
()
()
(1
(1
(1
()
()
()
(I
N
4000
0
4000
A
8000 Meters
()
( )
()
()
()
()
()
(I
( I
( I
( I
()
( )
( )
( )
(
I
(
I
(
I
(
I
(
1
(I
(I
{
1
(I
(
35
(
(
Figure 10 Map of the Bay of Quinte indicating the areas classified as low, medium, and high fish habitat suitability with respect to the
DM composite index, rarity, and expert fishers.
N
4000
o
4000
A
8000 Meters
36
Figure 11. Map of the Bay of Quinte indicating the areas classified as low, medium and high fish habitat suitability with respect to the
K-means suitability clusters, rarity, and expert fishers.
Clus+Rare+Expt
L~J Low
Med.
_High
N
4000
o
4000
A
8000 Meters
--,
37
~~~~~-~-~---~-------'-
-,.---- -,-
.-, ..-..'
--
.
APPENDICES
Appendix A Metadata tables for the Bay of Quinte physical habitat GIS database
Appendix Table Al Cross-table of Defensible methods substrate classes and the substrate categories used
in the nearshore habitat inventory, categories, etc.
Inventory
Type
Defensible Methods' Class
Bedrock
)
)
)
)
)
)
)
)
Bedrock
Boulder
Rubble
Gravel
Sand
Silt
Clay
Detritus
Muck
Marl
Boulder
Rubble
Cobble
50
50
Gravel
Sand
Silt
Clay
Hardpan Clay
lOO
100
100
lOO
lOO
20
20
-99
40
40
100
40
40
Appendix Table A2 Examples of vegetation assignments for Defensible Methods based on information
available from the OMNR Inventory and predictions of Seifried's model.
Source Information
OMNR
Inventory Description
Sub mer gent
Very Sparse (5-20%)
Emergent
Sparse (20-40%)
Emergent
Moderate (40-60%)
Submergent
Dense (60-80%)
Emergent
Very Dense (80-100%)
NO DATA
NO DATA
Defensible Methods' Values
OMNR
Vegetation
Seifried
Submergent
Model
Emergent
Submergent
NoCover
15
10
0
10
90
30
20
30
70
0
50
70
50
50
0
70
45
0
50
50
90
70
90
10
0
35
NO DATA
0
0
30
0
70
100
38
(
(
Appendix Table A3 Sample records (zoomed in area in Figure A4) from a Defensible Methods input data
file.
(
(
(
; QUINTE DEF METH
; Sample of Quinte.dat
()
()
* UnitType=Area
* Units=m2
()
()
* Order=ID,Area,AreaType,Depth,Substrate, Vegetation
* Proportions=Depth:ZO_1,Z 1_2,Z2_ 5,Z5 _1 O,Z 10+
* Proportions=Substrate:Bedrock,Boulder,Cobble,Rubble,Gravel,Sand,Silt,Clay,Hardpan,Pelagic
* Proportions=Vegetation:NoCover,Emergent,Submergent
( )
()
( )
( )
( )
182,32356.6075,UNCH,"0,0,0, 100,", "0,0,0,0,0,6,61,33,0,0"," 100,0,0"
1254,398.2930,UNCH,"0,0,100,0,","0,0,15,15,70,0,0,0,0,0","40,50,10"
1255,3411.0072,UNCH,"0,0, 100,0,0","0,70, 15, 15,0,0,0,0,0,0","40,50, 10"
1256,7.5968, UNCH, "0,0,100,0,0", "0,0,0,0,0,100,0,0,0,0", "20,50,30"
1257, 1409.7233,UNCH,"0,0, 100,0,0","0,0,35,35,30,0,0,0,0 ,0","20,50,30"
1264,1 0.3698,UNCH, "0,0,100,0,0", "0,100,0,0,0,0,0,0,0,0", "0,50,50"
1265, 7806.4303,UNCH, "0,0,100,0,0",",70,15,15,0,0,0,0,0,0" ,"0,50,50"
1266,68.4521 ,UNCH, "0,0,100,0,0", "0,0,0,0,0,20,40,40,0," ,"30,70,0"
1267, 1 176.4070,UNCH,"0,0, 100,0,0" ,"0,0,0,0,0,20,40,40,0,0" ,"0, 70,30"
1268, 131.0483,UNCH,"0,0, 100,0,0","0,0, 15, 15,70,0,0,0,0,0","0,70,30"
1271, 129.0008,UNCH,"0,0, 100,0,0", "0,0,25,25,50,0,0,0,0,0", "0,90,10"
1273,29. 1329,UNCH, "0,100,0,0,0", "0,0,0,0,0,1,32,67,0,0"," 100,0,0"
1275,6.5130,UNCH, "0,100,0,0,0", "0,0,0,0,0,1,52,47,0,0"," 100,0,0"
1339,2.0000, UNCH, "0,100,0,0,0", "0,0,0,0,0,3,51,46,0,0", "90,0,10"
()
()
()
()
(
)
(J
( )
()
( )
( ~
(
(
)
( )
(
(
(
(
(
(
(
(
<.
t
(
<.
(
39
I
(I
(I
Appendix Figure Al Ontario Base Map (OBM) 1:10000 shoreline and bathymetry in the upper and middle regions of the Bay of
Depth (m)
D
D
D
0-1
1-2
2-5
5-10
•
Lake Ontario
10+
4
8km
I
40
Appendix. Figure A2 Substrate polygons showing offshore sampling points, OMNR nearshore inventory and offshore Theissen
01 ons.
Lake Ontario
Thomas Sample Points
•
Nearshore OMNR inventory
D
Offshore Theissen polygons
a
4
("
- - --- -- ------ - - - - ~.
I
41
- -, -,
-.
-.
8 km
I
I
-.
1
.-, .-
\
-
\
.-
\
-\
-
.
\
vu
vv
u
v
Appendix Figure A3 Observed emergent and predicted submergent vegetation in the upper and middle Bay of Quinte.
r'"
.1
r
....
Bell~ville_
.
:...'
~ .,....
. ..,.,. ~
~
......
., or'' '
!
Oeseronto ..-- /
~·~t
"'.
t" .... .
....
.
.
.r'~""
'
• .f'
,...-.J
.'
Big Island
/'
"&
.'
....
I-
~'t"''"':''\
,.
to
~_~
~.
-
r.:-' ~n
.. "'--":.
/~ ~.
~
I
I
.
.
..,,!
I~
....
....
,
~. ••~
..
',
.?
Trenton '
Lake Ontario
iIoIj
t
•
NoCov>50
Emerg>50
Submerg>50
Lake Ontario
o
Mixed
42
4
8km
Appendix Figure A4 Habitat polygons, zoomed in area shows unique combinations of attribute data, see sample in Table A2
D
~
\
,-....
,-.... ,-.....
~.
~
...-... ,-.....
~,
",-....
",-.....
~
~
..-..
.-.-..
o
Habitat Polygons
- - .-
43
,-..
4
8 km
I
- -- --
Appendix B
Species location list of the fish present in the Bay of Quinte compiled from Minns et al. (1986).
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
)
Latin name
Common name
Latin name
common name
Acipenser fulvescens
Lepisosteus osseus
Amia calva
Alosa pseudoharengus
Dorosoma cepedianum
lake sturgeon
longnose gar
bowfin
alewife
gizzard shad
white sucker
northern hog sucker
bigmouth buffalo
silver redhorse
shorthead redhorse
Oncorhynchus kisutch
Oncorhynchus tshawytscha
Onchorhynchus mykiss
Salmo trutta
Salvelinus namaycush
Coregonus artedii
Coregonus clupeaformis
Osmerus mordax
Hiodon tergisus
Esox americanus
vermiculatus
Esox lucius
Esox masquinongy
Carassius auratus
Cyprinus carpio
Notemigonus crysoleucas
Notropis atherinoides
Luxilus chrysocephalus
Notropis hudsonius
Cyprinel/a spilopterus
Notropis heterodon
Luxilus cornutus
Notropis stramineus
Pimephales promelas
Pimephales notatus
Semotilus corporalis
Rhinichthys cataractae
Carpiodes cyprinus
coho salmon
Chinook salmon
rainbow trout
brown trout
lake trout
cisco
lake whitefish
rainbow smelt
mooneye
grass pickerel
Catostomus commersonii
Hypentelium nigricans
Ietiobus cyprinel/us
Moxostoma anisurum
Moxostoma
macrolepidotum
Moxostoma valenciennesi
Ietalurus nebulosus
Ietalurus punctatus
Noturus jlavus
Anguilla rostrata
Fundulus diaphanus
Lota Iota
Labidesthes sicculus
Gasterosteus aculeatus
Culaea inconstans
greater redhorse
brown bullhead
channel catfish
stonecat
American eel
banded killifish
burbot
brook silvers ide
threespine stickleback
brook stickleback
Percopsis omiscomaycus
Morone americana
Morone chrysops
Ambloplites rupestris
Lepomis gibbosus
Lepomis macrochirus
Micropterus dolomieui
Micropterus salmoides
Pomoxis nigromaculatus
Etheostoma nigrum
Etheostoma jlabellare
Perca jlavescens
Percina caprodes
Sander vitreum
Aplodinotus grunniens
Coitus bairdii
trout-perch
white perch
white bass
rock bass
pumpkinseed
bluegill
smallmouth bass
largemouth bass
black crappie
johnny darter
fantail darter
yellow perch
logperch
walleye
freshwater drum
mottled sculpin
northern pike
muskellunge
goldfish
common carp
golden shiner
emerald shiner
striped shiner
spottail shiner
spotfin shiner
blackchin shiner
common shiner
sand shiner
fathead minnow
bluntnose minnow
fallfish
longnose dace
quillback
44
Appendix C
Listing of the 9 fish species groupings for the freshwater species, present in the Bay of Quinte,
used in this study.
(1) Cold-water
lake sturgeon
CISCO
burbot
rainbow smelt
lake whitefish
trout-perch
alewife
coho salmon
rainbow trout
Chinook salmon
brown trout
lake trout
threespine
stickleback
mottled sculpin
Cool-water,
Age-at-maturity
<=2
(2) Low
vegetation cover
emerald shiner
gizzard shad
longnose dace
brook stickleback
fantail darter
Cool-water,
Age-at-maturity
>2
(4) Low
vegetation cover
mooneye
walleye
silver redhorse
fallfish
greater redhorse
Warm-water,
Age-at-maturity
<=3
(6) Low
vegetation cover
northern
hogsucker
shorthead redhorse
spotfin shiner
white perch
sand shiner
stone cat
white bass
Warm-water,
Age-at-maturity
>3
(8) Low
vegetation cover
(
(
:
(
(,
e,
freshwater drum
smallmouth bass
channel catfish
(l
(
( )
( )
C)
(3) High
vegetation cover
brook silverside
golden shiner
blackchin shiner
banded killifish
spottail shiner
log perch
rock bass
striped shiner
common shiner
johnny darter
(5) High
vegetation cover
(7) High
vegetation cover
(9) High
vegetation cover
Cl
( I
( )
American eel
yellow perch
white sucker
northern pike
black crappie
quillback
bluegill
pumpkinseed
bluntnose minnow
fathead minnow
brown bullhead
bigmouth buffalo
longnose gar
goldfish
common carp
muskellunge
grass pickerel *
bowfin
largemouth bass*
(
(
(
(
(
(
(
(
(
* Exceptions to age-at-maturity rule.
(
(
(
(
45
(
,
(
t
(
I
(
I
(
I
(
I
(
I
(I
v
vvvvuvv
v
v v -'
Appendix D: Correlation coefficients for the 27 habitat suitability indices which represent all of the combinations with respect to
thermal, age of maturity, vegetation preference and life stage of the freshwater fish inhabiting the Bay of Quinte (note: Life stages: A
= adult, S = spawning, Y = YOY).[ Correlations >= 0.5 are highlighted in bold.]
Thermal
Age at maturity
Vegetation
Coldwater
Life Sta e
Cold
A
s
y
Cool <=2
Low
A
S
Y
High
A
S
Y
Cool
>2
Low
A
A
S
-0.04
Coolwater
Less than or equal to 2
Greater than 2
Low
High
Low
High
Y
A
S
Y
A
S
Y
A
S
Y
A
S
Y
A
S
Y
A
S
Y
A
s
Y
A
s
Y
0.02
0.5\
0.02
0.14
0.13
0.05
0.11
-<l.0I
-0.03
0.00
0.08
0.00
0.12
0.16
0.01
0.11
0.07
0.04
0.07
-<l.04
-<l. 10
-0.09
0.09
0.02
0.09
0.8\
0.74
0.82
0.31
0.14
0.45
0.15
0.86
0.80
0.52
0.14
0.43
0.13
0.86
0.85
0.57
0.00
-0.05
0.04
0.80
0.41
0.83
0.04
0.02
0.10
0.77
0.68
0.39
0.14
0.42
0.1
0.94
0.73
0.62
0.05
0.49
0.06
0.85
0.97
0.74
-0.11
-0.13
-0.10
0.88
0.48
0.5\
·0.06
-<l.09
0.02
0.65
0.42
0.26
0.44
0.14
0.73
0.63
0.38
0.08
0.37
0.13
0.83
0.74
0.65
-0.01
-0.05
-0.05
0.72
0.42
0.53
0.00
-0.10
0.04
0.51
0.45
0.73
0.45
0.65
0.69
0.59
0.49
0.59
0.37
0.74
0.76
0.69
0.40
0.33
0.39
0.71
0.60
0.53
0.40
0.35
0.31
0.92
0.73
0.85
0.23
0.12
0.42
0.77
0.56
0.71
0.35
0.35
0.74
0.67
0.61
0.57
0.18
0.53
-<l.04
0.73
0.56
0.67
-0.07
0.32
-0.01
-0.07
0.32
0.85
0.54
0.71
0.11
0.14
0.66
0.85
0.77
0.70
0.00
0.53
-0.18
0.84
0.65
0.66
0.31
0.23
0.59
0.72
0.83
0.45
0.37
0.47
0.79
0.76
0.74
0.63
0.32
0.77
0.07
0.69
0.62
0.40
-0.01
-0.09
0.42
0.94
0.58
0.87
0. 13
0.11
0.52
0.86
0.77
0.82
-0.07
0.33
-0.18
0.92
0.80
0.88
0.7
0.75
S
Y
High
A
S
Y
Warm <=3
Low
Y
Y
Low
0.42
-0.01
0.83
0.94
0.63
-0.20
-0.23
-0.16
0.93
0.41
0.65
-<l.16
-<l.15
-0.04
0.15
-0.12
0.78
0.77
0.34
-0.20
-<l.22
-0.16
0.84
0.40
0.68
-0.17
-<l.20
-<l.15
0.48
0.80
0.28
0.45
0.65
0.64
0.39
0.34
0.45
0.43
0.38
0.22
0.40
0.49
0.33
0.60
0.79
0.04
0.10
0.45
0.94
0.81
0.90
-0.07
0.36
-0.19
0.94
0.87
0.84
0.33
0.43
0.50
0.66
0.60
0.56
0.52
0.32
0.62
0.04
0.52
0_60
0.37
0.15
0.05
0.38
0.65
0.54
0.75
-0.09
0.12
-0.10
0.78
0_63
0.97
0.58
-0.13
-0.16
-0.09
0.78
0.40
0.67
-<l.05
-0.09
0.10
0.75
-0.03
-0.05
-0.03
0.91
0.51
0.55
0.00
-0.03
0.01
0.05
0.38
-0.09
0.12
-0.10
0.78
0.63
0.97
0.91
0.90
-0.19
0.41
-0.30
0.93
0.87
0.67
0.8\
-0.22
0.39
-0.33
0.9\
0.87
0.55
0.21
-0.24
0.93
0.92
0.78
0.49
0.60
-0.19
-<l.21
-0.14
-0.19
A
S
0. 19
Y
High
-0.04
-0.09
0.15
A
S
Warm >3
0.56
0.22
0.8\
A
S
High
Warmwater
Less than or equal to 3
Greater than 3
Low
High
Low
High
A
S
0.28
0.24
0.07
-0.26
-0.28
-0.14
0.92
0.80
0.73
46
(
.
(
()
Appendix E
Sample graph ofthe mean habitat suitability index values (y-axis) versus depth (m) (z-axis) and
substrate classes (x-axis) for adult habitat preferences in members of the cold water groups in
habitats containing only submergent vegetation.
(
,
(
\
(
\
(
)
( .)
( )
(
I
I Cold
)
()
water - Submergent vegetation - Adult
( )
( )
( )
()
( )
1
(J
()
( )
(~
0.5
(~
(
0
a
8
@
•
~
000
-.
~
t
.~
~
( ~
0-1
1-2
00 6
li:J t;;j 0
)
(
~
( ~
~
~
(
(
(
(
..
~
~
( I
( I
( I
( I
( I
( I
( I
,
( ,
(
( f
(,
(
47
,
( I
(f
APPENDIX F: Habitat Suitability Matrices (depth vs. substrate for 9 fish groups by 3 life stages and by vegetation type)
(1) Cold-water
Cold - Submergent - Adult
Cold - Submergent - Spawning
Cold - Submergent - yay
Cold - No Cover - Adult
Cold - Emergent - Adult
Cold - Emergent - Spawning
Cold - Emergent - yay
Cold - No Cover - Spawning
Cold - No Cover - YOY
48
(2) Cool-water, age of maturity <= 2, low vegetation preference.
Cooll Less than or equal to 2 I Low Veg
ergent - Adult
Cool I Less than or equl to 2 I Low Veg _~
r o..:e,-,-n.t - Adult
0.5
0.5
Cooll Less than or equal to 2 I Low Veg
)f-.......,,,,r-'Eent - Spawning
Cooll Less than or equal to 2 I Low Veg
over - Adult
Cooll Less than or equal to 2 I Low Veg
ent - Spawning
Cooll Less than or equal to 2 I Low Veg
1
I
0.5
Cooll Less than or equal to 2 I Low Veg
-YOY
0.5
Cooll Less than or equal to 21 Low Veg
ergent - YOY
0.5
-. - - --- -- -.- - - - --- -
I
Cool I Less than or equal to 21 Low Veg
Cover - YOY
0.5
49
-. -. -.
-.
-
-. -. -.
,- - -'- --' -
-.
.-.,
.-.
\
(3) Cool-water, age at maturity <= 2, high vegetation preference.
Cool! Less than or equal to 2! High
Veg
ergent - Adult
Cool! Less than or equal to 2 ! High
Ve
mergent - Adult
Cool! Less than or equal to 2 ! High
Ve
0 Cover - Adult
Cooll Less than or equal to 2 I High
Veg - Emergent - Spawning
Cool! Less than or equal to 2! High
Veg Cover - Spawning
0 .5
Cool! Less than or equal to 2! High
Veg ergent - Spawning
0.5
Cool! Less than or equal to 2! High
Ve
bmergent - YOY
0 .5
0 .5
Cool! Less than or equal to 2 I High
Ve - mergent - YOY
0 .5
0 .5
50
(4) Cool-water, age at maturity> 2, low vegetation preference.
Cooll Greater than 21 Lov VegE
nt - Adult
-Cooll Greater than 21 LowVeg - No
- Spawning
0.5
•
g
•••••••
'
a
••••
. ,"
51
.~
-
\
(5) Cool-water, age at maturity >2, high vegetation preference.
Cool! Greater than 2! High VegSu)~LQl.!!C!!.L'- Adult
Cool! Greater than 2! High VegO::"'_II<..!.!!' -- Adult
Cool! Greater than 2! High Veg - No
Cool! Greater than 2! High VegS
- Spawning
a
Cool! Greater than 2! High Veg S
ent - yay
0.5
Cool! Greater than 2! High Veg - No
k _~""
e,,r - Adult
0 .5
Cool! Greater than 2! High Veg ......_ ""n,t. - yay
0 .5
52
(6) Wann-water, age at maturity <=3, low vegetation preference.
Warm / Less than or equal to 3/ Low
Ve
Cover - Adult
Warm / Less than or equal to 3/ Low
Veg
ergent - Adult
•
Q
••••
Warm / Less than or equal to 3/ Low
Veg rent - Spawning
.
Q
CI
Q
•••••••
Warm / Less than or equal to 3/ Low
Veg ent - Spawning
Warm / Less than or equal to 3/ Low
Veg over - Spawning
Warm / Less than or equal to 3/ Low
Ve
ergent - YOY
Warm / Less than or equal to 3/ Low
Ve
Cover-YOY
".
0.5
Warm / Less than or equal to 3/ Low
Ve
mergent - YOY
0.5
----
-------
53
--------------------~-----~
-. - -
-\
I
(7) Warm-water, age at maturity <= 3, high vegetation preference.
Warm 1 Less than or equal to 31 HIgh
Ve
mergent - Adult
Warm 1 Less than or equal to 31 HIgh
Ve
ergent - Adult
Warm 1 Less than or equal to 31 HIgh
Veg - SpawnIng
Warm 1 Less than or equal to 31 HIgh
Veg
ent - SpawnIng
Warm 1 Less than or equal to 31 HIgh
Veg
over - SpawnIng
Warm 1 Less than or equal to 31 HIgh
Ve
mergent - yay
Warm 1 Less than 0 r equal to 3 1 HIgh
V
ergent - yay
Warm 1 Less than or equal to 31 HIgh
V
Cover-yay
0.5
0 .5
---_._---
54
I
--'
(8) Wann-water, age at maturity> 3, low vegetation preference.
Warm I Greater than 3 I Low Veg Su
ent - Adult
Warm I Greater than 3 I Low Veg E
nt - Adult
Warm I Greaterthan 31 LowVeg - No
Adult
_ ---=._-
0 .5
Warm I Greater than 3 I Low Veg Sub
t - Spawning
Warm I Greater than 3 I Low Veg Em
- Spawning
Warm I Greater than 3 I Low Veg - No
- Spawning
O.S
.g• •
••••
•••••••
•
Q
•
~.
...
Warm I Greater than 3 I Low Veg S
ent - yay
0 .5
55
~--~------------------~--~-~~-------~~~----
v
u
(9) Warm-water, age at maturity> 3, high vegetation preference.
Warm 1 Greater than 3 1 High VegS
ent - Adult
Warm I Greater than 3 1 High Veg - No
- Spawning
Warm 1 Greater than 3 I High Veg S
ent - YOY
0 .5
Warm 1 Greater than 31 High Vegent - YOY
0 .5
I
J
56
APPENDIX G: Habitat Rarity Assessment Graphs
Graphs of cumulative area (dashed line) and cumulative weighted suitable area - WSA (solid line) for Bay of Quinte habitat suitability
database indices showing the application of the 75 percent area and 0.75 suitability cut-offs for identifying rare, highly suitable
patches: 1) Cold-water, 2) Cool-water and 3) Warm-water groups by life stage (adult, spawning, and yoy), age at maturity (LT2, <= 2
years; GT2 > 2; LT3 <= 3; GT3 > 3) and vegetation preference (low or high).
1) Cold-water
AduH Cold Water
-
~
Spawning Cold Water
100
w
80 I
'"
u""
60
u
_ _ _ _ 0_.
~--.----------------
I
>
,.
20
,.E
0
u
""':;
80
-.'"
Area 15.
I
Suitability
: -WSA15.
I
Suitabilitv
~~ 40 '
u
,----.
- .. -. Cumulatile
I
~U)
~
100
I
i
u
""
':;
-.
u<
I
60
~~ 40
>
20
0
0.2
0.4
0.6
1
0.8
02
Suilabiq tmices
0.4
0.6
0.8
I
--~1I1
'"
60
I
.- _--.5u~l!it~-' I
~~ 40
I
I
I
,.E
100
80
I
<>
~
u
""':;
Area 15.
Suitability
WSA15.
~U)
.
.
,.
YOY Cold Water
II
I
.
.
.
I
--,--
~U)
-
>
,.
Area '13. I
Suitability i
WSA15. !
__SlJi1.aQi~tc
20
1
,.E
!
<>
0.0
02
Suftabiflly Indices
----
i -.. - .CumUlalilel
I
I
u""
0.4
0.6
0.8
1.0
Suftabilily Incices
j
2) Cool-water
i
Adult Cool Water LT2 Low Veg
I
:; 100
I
_ .. .... ., " ;:/:-;-'vo--:!
i
.
'!
..
':; I------"'--,.,r=-l--I-·· -.
""
!;
W
I
60
u""
u
,.
'
Cumulatilel'
Area 15. j
Suitability !
.
I -WSA'I3. ,~
L SuitabilitvJ i
~U)
~~ 40
~
I
20
I
I
Spawning Cool Water LT2 Low Veg
.
-
100 , - - - - - - - - . - : .
""
':;
80 f - - - -
!;
60
u""
I
I
-::r--r---i r---:---:-c:--, l
- ,, - .Cumulatile I
Area 15. I
Suitability
WSA'I3.
Suitabilitv
~U)
..
~~ 40
>
..
YOY Cool Water LT2 Low Veg
20
=>
E
,.
E
u
u
'"
0.2
0.4
0.6
Suftabilily Indices
0.8
.
""
':;
100 , . - - - - - - - - . : . -_
80
-!!~
II
i
I
I
.
>
:;
20
'"
E
0 1'-"----.,.------'--,.----1
o
02
0.4
0.6
0.8
Suftabilily Indices
57
----------------------~----
,.
u
0.00
020
0.40
0.60
Suftabilily Indices
0.80
I'
- ,, - .CumulaUlel
AIea'l3. 'I
Suitability
WSA'I3. I
I
._...fuli~i!vJ
r
1.00
vvvv
vv
vvvv
Appendix G continued/2.
2) Cool-water continued.
-------1
Adult Cool Water LT2 High Veg
.
...
~
I
100 -,---------,----,.----,
I
~I------II
-··-·Cumulalile
u
SO
I
I
Area \S.
II _WSA\S.
Suitability
..
II
.
YOY Cool Water LT2 High Veg
100 - , - - - - - - - - - ,
:; 100 -,.-------=""""----,
m
u
~
~
u
u":
SO ~------~~_-,~-l~
_.. - . Cumulatile I
60
-<I)
1'1'
!__~itabilitL I
20
Spawning Cool Water LT2 Low Veg
~~ 40
-
,.
u
..
20
Area \s, III
Suitability I
WSA\S,
Suitability
I!
~
~
.J
I
I
I
0,2
0,4
0,6
O,S
i
E
~
..:
80
:;
60
~
u":
- .. _. Cumulalile
Area \s,
Suitability
WSA\S,
_ Suitability
-en
::~ 40
u
.~
..
20
~
E
~
0
I
L-r---r---r--'----i
-!,-,.: .;,;
.;;..,
••
<..>
I
02
I
SuHabiity Indices
I
_____________________________ J
0,4
0_6
0.0
O,S
I
Suitabifity Indices
..
100
_______ __ ...J
~
SO
r----I-------·----·--~~
..:
SO
~
.
,---;-=,~~-----;--=.......,
O.S
1.0
100 , - - - - - - - - - "
u
..:
- .. -. Cumulatii
Area \s,
Suitability
WSA\S.
Suitability
~
..
0,6
YOY Cool Water GT2 Low Veg
u
..:
0.4
Suitability In!ices
Spawning Cool Water GT2 Low Veg
100 - , - - - - - - - - -
02
20
~
SO
I----.·~-----~
- , - .. -. Cumulatile
Area \s,
Suitability
WSA\S,
Suitability I
I
m
.~
..
20
~
~
E
0 ~----~-Lr-4
o
0,2
0,4
0.6
Suilabifityfndices
O,S
O ~---.-----'---,.-----'
o
02
0,4
0,6
Suitabifity Indices
58
O,S
~
<..>
0,0
02
0_4
0.6
Suitability Indices
0.8
1.0
Appendix G continued/3.
2) Cool-water continued.
.
...
-;
-;;;
...
::.~
40
80 1--_ _ _ _.___. ,~_-,
80
_.• _. CUmulatr.e
Area IS.
Suitlilility
WSAIS.
Suitabilitv
II
I
20
'"
'"
100
I -··- ·Clrnllatilell
Area IS. I'
Suitability
WSA\S.
__ .Jl.!llt~.bilj!vJ
-."
.
!:.
100 ,..--------------~
80
60
~
I
100
~
YOY Cool Water GT2 High Veg
Spawning Cool Water GT2 High Veg
Adult Cool Water GT2 High Veg
0
'"
~--------~r-~r-~
02
0.4
0.6
::oee 60
-."
::.~
40
..
c
>
~
t>::j1-----
1--------::".
.'
..•..• ,,~
II
I
II
2Q
I
u
0.8
02
Suitabiily Indices
0.4
0.6
1- ··-· CUmuiatilel
Area\S. 'I
I
SLitability
i - WSA \S . .
Suitabilitv
I
I
,'
~
E
E
u
r--------------::-:co;;:.-~.,
0.0
0.8
0.2
0.4
0.6
0.8
1.0
Suitability Indices
Suitability Indices
3) Warm-water
Spawning Warm Water LT3 Low Veg
Adult Warm Water LT3 Low Veg
.
100
80
-;
~
100 .,--------------:,.,..
... 60
_.. _. CUmulalile
Area IS.
Suitability
WSAIS.
Suitabil tv
I
!
~
'"
80
I
::.~ 40
E
...
-----,---_._-
-."
.
.'"
.
...
-;
.
~
...
c
YOY Warm Water LT3 Low Veg
20
I
.. _.. /
u
Suitlilili
..
20
0.4
0.6
Suitabiity IncflCes
0.8
u"':
1_,
.r )'
80
I
---·--7. . . .-.. -.• "-t-
60
,
::.~
.,
.!!!
E
E
'"
:0
I
.
,-.. -.cumU'8a'le
:'
I
L_.-
!
II
40
i
/
_._!
I'
>
20
.i' ~"
I
-."
'"
:0
I
•• J.'
-
I
Area \S.
SLitability
WSA\S.
Stitaalitv
u
U
02
c
100
02
0.4
0.6
0.8
0.0
02
0.6
0.8
1.0
Suitability Indices
Suitability Indices
59
0.4
-.
-.-
- -,
- -. - -.
\
~
,
,J
Appendix G continued/4.
3) Warm-water continued.
...
.
... 100 ~-------...: so
~
...:
.
.:.~ 40
-
>
..
-;;
Area 15.
•• _ 0 ' - '
:'
-U)
~
_.. - . Comulatile
.- .. ", .~ .. ".
-:;
..u"': SO
20
II)
u-<
Suitability
Suitability
>
0 -I-L---,....--,.--.....L...----l
02
0.4
0.6
----------! ..---
60
:~ 40
'"
o
80
WSA15.
..
.. , .. -
-
j
Suitabilitv
..
I
-
I
I
WSAw' li
Suitabilit !
I
20
i
'"
E
0.4
0.6
Suitability Indices
0,0
02
80
~
60
0.8
i
0,6
0.8
1.0
YOY Warm water GT3 Low Veg
100 , . - - - - - - - -
;
0,4
SuRability Indices
100 ,....----,-=_---,---
f - - - ' - - - : - ; -I .,..L..-;:;J
-U)
~:I: 40
_.. - . Cumulatile
Area \5,
SuitOOifity
-
>
20
'"
-:;
u"':
-..,
1-
.:.~ 40
Suitability
>
-
0.4
0.6
SuRability Indices
60
0,8
Cumulatilel
I
WSA15, I
Loo __ ..?.!l~abilitV j
'"
02
.
Suitability
.
..
WSAw,
00 _
Area 15, !
60
"E
E
u
0.6
I
Suitability ,
I
0.8
"
u'"
WSA15,
"
'"
E
...
Suitability
I-
.
u
u"':
r
>
'"
.....
-
'I- oo - .cumulatilel
~-----~---+
..
Spawning Warm Water GT3 Low Veg
Area 15.
SLitabiHty
0.4
SO
Area 15, '
"
02
;
02
100 ..--- - - - - - - - ,
SuHability Incices
... ~ 40 :
.;
~
E
0.8
- .. _. Comulatile
=(1)
WSA15 .
u
80
"~oC SO
.. -.,,, .• "
20
Adult Warm Water GT3 Low Veg
...:
:
_•. _. Cumulatile
Area 15.
Suitability
• .'
Suitability Indices
.....
.
...:
100 . , . - - - - - - - - - , - - - - ,
'-u)
E
~
YOY Warm water LT3 High Veg
Spawning Warm Water LT3 High Veg
Adult Warm Water LT3 High Veg
02
0,4
0.6
SuRability In!ices
0.8
1.0
Appendix G continued/So
3) Warm-water continued.
Adult Warm Water G13 High Veg
.
~
<
~
c
~<
Spawning Warm Water GT3 High Veg
.
100
~
80
<
~
!
"E
"
'-'
I
I
60
-on
..
:~ 40
>
.!!
- .. - . Cumulatile III
Area Y.l.
Suitability
WSAY.l· 1
Suitability !
20
I
'"
E
'"
'-'
0.2
0.4
0.6
0.8
100
...-----:----I I-~~·--:-Cumulati;
80
..
c
.,,,<
60
1·· ... ··-··;
..
:~ 40
!
20
i
0
I
>
0
02
0.4
0.6
Suitability Indices
Suitability Indices
I
II
-on
~
YOY Warm Water GT3 High Veg
-
I- -
Area Y.l.
Suitability
WSAY.l.
Suitabili!y--,
.
-:;
-;;
..
.
M
<
,,<
100
80
,.
-on
:~ 40
>
~
I
- .. -' Cumulatile
AreaY.l.
Suitability ,
WSAY.l· 1
Suitability
60
20
"
E
0.8
"
'-'
0.00
020
0.40
0.60
0.80
1.00
Suitabirlly 1nd"lCes
4) Composite Suitability Index.
Corrposite
-..
M
<
~
c
~<
100
I
I
----1-
80
60
I
I
-on
I
.
:~ 40
I
I
I
.~
M
20
'"
E
"
'-'
0.00
020
0.40
0.60
0.80
1'=:1
i
Suitability
WSAY.l.
L ... _ _p.QijaQllijyJ
-
1.00
Suitability Indices
61
-~-~---~------~-----------~-~-~-
-,
.-.,
...-..
,
.
\ ,J'-
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement