Defensible Methods Of Assessing Fish Habitat: Lacustrine

Defensible Methods Of Assessing Fish Habitat: Lacustrine
Defensible Methods Of Assessing Fish Habitat: Lacustrine
Habitats In The Great Lakes Basin - Conceptual Basis And
Approach Using A Habitat Suitability Matrix (HSM) Method
Charles K. Minns1, James E. Moore2, Mike Stoneman1, and
Becky Cudmore-Vokey1
1
Great Lakes Laboratory for Fisheries and Aquatic Sciences
Fisheries and Oceans Canada
Bayfield Institute, 867 Lakeshore Road, P.O. Box 5050
Burlington, Ontario L7R 4A6 CANADA
2
JEMSys Software Systems Inc., 22 Marion Crescent
Dundas, ON L9H 1J1
March 2001
Canadian Manuscript Report of Fisheries and
Aquatic Sciences 2559
Fisheries
and Oceans
Pêches
et Océans
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 Department’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.
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 problèmes nationaux ou
régionaux. La distribution en est limitée aux organismes et aux personnes de régions particulières du
Canada. Il n’y a aucune restriction quant au sujet; de fait, la série reflète la vaste gamme des intérêts et
des politiques du ministére des Pêches et des Océans, e’est-à-dire les sciences halieutiques et aquatiques.
Les rapports manuscrits peuvent être cités comme des publications complèrwa. Le titre exact
paraît au-dessus du résumés de chaque rapport. Les rapports manuscrits sont résumés dans la revue
Résumés des sciences aquatiques et halieutiques,et ils sont classés dans l’index annuel des publications
scientifiques et techniques du Ministére.
Les numéros 1 à 900 de cette série ont été publiés à titre de manuscrits (série biologique) de
l’Office de biologie du Canada, et aprés le changement de la désignation de cet organisme par décret du
Parlement, en 1937, ont été classés comme manuscrits (série biologique) de l’Office des recherches sur
les pêcheries du Canada. Les numéros 901 à 1425 ont été publiés à titre de rapports manuscrits de
l’Office des recherches sur les pêcheries du Canada. Les numéros 1426 à 1550 sont parus à titre de
rapports manuscrits du Service des pêches et de la mer, ministère des Pêches et de l’Environnement. Le
nom actuel de la série a été établi lors de la parution du numéro 1551.
Les rapports manuscrits sont produits à l’échelon régional, mais numérotés à l’échelon national.
Les demandes de rapports seront satisfaites par l’établissement auteur don’t le nom figure sur la
couverture et la page du titre. Les rapports épuisés seront fournis contre rétribution par des agents
commerciaux.
Canadian Manuscript Report
of Fisheries and Aquatic Sciences 2559
March 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
by
Charles K. Minns1, James E. Moore2, Mike Stoneman1,
and Becky Cudmore-Vokey1
1
Great Lakes Laboratory for Fisheries and Aquatic Sciences
Fisheries and Oceans Canada
Bayfield Institute, 867 Lakeshore Road, P.O. Box 5050
Burlington, Ontario L7R 4A6 CANADA
2
JEMSys Software Systems Inc., 22 Marion Crescent
Dundas, ON L9H 1J1
Minister of Supply and Services Canada 2001
Cat. No. FS97-4/2559 ISSN 070-6473
Correct citation of this publications:
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.
ii
ABSTRACT
This report presents a detailed description of the structure and application of a quantitative fish
habitat assessment tool for use in evaluating development proposals affecting lacustrine fish
habitats in the Great Lakes and beyond. The scientific and regulatory context for the
development of the tool are introduced along with a formal quantitative framework developed by
Minns (1997) as a basis for assessing net change of productive capacity of fish habitats.
Suitability indices are used as practical and pragmatic measures of habitat productivity and
weighted suitable areas (area times suitability) are derived as surrogate measure of total
productivity of the habitats affected at a development site. A habitat suitability matrix (HSM)
model is described for estimating habitat suitability indices along with the rules and criteria that
must be applied to allow evaluation of fish habitats. The HSM model uses pooled matrices
representing the aggregate habitat preferences of many species by life stage to ensure that the
needs of all fishes occurring in an ecosystem or ecoregion are considered. Species requirements
are grouped by life stage and ecological groupings according to trophic level and thermal
preference. Habitat scenario data sets describe the pre- and post-development conditions at a
proposed site as series of habitat patches with characteristics of depth range, substrate type, and
cover type. The HSM model generates suitability values for specific combinations of depth,
substrate and cover that can be assigned to individual habitat patches. The rules and criteria for
preparing the habitat scenario data sets and computing suitability matrices are described in detail
with examples. The basis for using this approach in a regulatory context is described as the HSM
approach to defensible methods of assessing fish habitat and is intended as an aid to the existing
decision process, not as a replacement. Case studies are described as an aid for potential users of
this approach. The whole approach has been implemented as an internet website for use by
consultants working for developers and by regulatory agents. The scope for application of the
approach is outlined along with a discussion of the limits for wise use of the approach.
Implementation of the HSM model as a basis for assessing fish habitat represents a first step
away from subjective and toward objective assessment tools. It is expected that further research
will provide the basis for iterative improvements and enhancements.
RÉSUMÉ
Le présent rapport décrit de façon détaillée la structure et l’application d’un outil d’évaluation
quantitative de l’habitat du poisson qui sera utilisé pour évaluer les projets d’aménagement
affectant les habitats lacustres dans les Grands Lacs et ailleurs. On y présente le contexte
scientifique et réglementaire de l’élaboration de cet outil, ainsi qu’un cadre de travail quantitatif
officiel élaboré par Minns (1997) comme base d’évaluation de la variation nette de la capacité de
production des habitats du poisson. Des indices de qualité sont utilisés comme mesures pratiques
et pragmatiques de la productivité de l’habitat, et des aires appropriées pondérées (aire multipliée
par qualité) servent de substitut à la mesure de la productivité totale des habitats touchés dans un
site soumis à l’aménagement. On décrit un modèle de matrice de qualité de l’habitat (MQH) qui
sert à estimer les indices de qualité des habitats, ainsi que les règles et les critères qui doivent
être appliqués pour évaluer les habitats du poisson. Dans le modèle MQH, on utilise des matrices
regroupées représentant l’ensemble des habitats privilégiés de nombreuses espèces, par stade de
développement, de façon que les besoins de tous les poissons présents dans un écosystème ou
iii
dans une écorégion soient pris en compte. Les besoins des espèces sont groupés selon le stade de
développement et le groupement écologique en fonction du niveau trophique et des préférences
thermiques. Les ensembles de données sur les scénarios d’habitat décrivent les conditions avant
et après l’aménagement d’un site proposé sous forme d’une série de parcelles d’habitat
caractérisées par la profondeur de l’eau, le type de substrat et le type de couvert. Le modèle
MQH produit des valeurs de qualité pour des combinaisons spécifiques de profondeur, de
substrat et de couvert qui peuvent être attribuées à chaque parcelle d’habitat. Les règles et les
critères pour établir les ensembles de données pour les scénarios sur l’habitat et calculer les
matrices de qualité sont décrits en détail et accompagnés d’exemples. Le fondement de
l’utilisation de cette approche dans un contexte réglementaire est décrit (approche MQH); cette
approche fournit des méthodes valables d’évaluation de l’habitat du poisson; elle est conçue
comme une aide au processus décisionnel existant, et ne le remplace pas. Les études de cas sont
décrites pour aider les utilisateurs potentiels de cette approche. L’ensemble de l’approche
constitue maintenant un site web à l’intention des conseillers collaborant avec des promoteurs et
des agents chargés de la réglementation. On présente la portée de l’application de l’approche,
ainsi qu’une analyse des limites en vue d’une utilisation rationnelle de l’approche. La mise en
œuvre du modèle MQH comme base d’évaluation de l’habitat du poisson représente la première
étape du passage d’outils d’évaluation subjectifs vers des outils objectifs. Les travaux de
recherche complémentaires serviront de base à la poursuite de l’amélioration et de l’actualisation
de la méthode.
iv
TABLE OF CONTENTS
ABSTRACT....................................................................................................................................iii
RÉSUMÉ........................................................................................................................................iii
LIST OF FIGURES .......................................................................................................................vii
LIST OF TABLES .........................................................................................................................vii
LIST OF APPENDICES ...............................................................................................................viii
1.0 INTRODUCTION .....................................................................................................................1
1.1 Legislative and Policy Background ........................................................................................2
1.2 A Quantitative Framework for Measuring Net Change .........................................................3
1.3 Previous Quantitative Assessment Schemes ..........................................................................4
1.4 Content of this Report.............................................................................................................4
2.0 FRAMEWORK FOR ASSESSING FISH HABITAT..............................................................4
2.0.1 Principles..........................................................................................................................5
2.0.2 Assumptions.....................................................................................................................6
2.1 Overview of Habitat Suitability Matrix Method ....................................................................7
2.2 Definition................................................................................................................................8
2.2.1 Location Species List.......................................................................................................8
2.2.2 Assembling Fish Species Groups.....................................................................................9
2.2.3 Fish Habitat Criteria.........................................................................................................9
2.2.4 Group and Criteria Weights ...........................................................................................10
2.3 Habitat Scenarios ..................................................................................................................11
2.3.1 Data Specifications and Format .....................................................................................11
2.3.2 Scenario Options ............................................................................................................11
2.3.3 Area Designations ..........................................................................................................12
2.3.4 Habitat Characterization Options...................................................................................12
2.4 Computation.........................................................................................................................14
2.4.1 Individual Species Habitat Suitability............................................................................14
2.4.2 Group Habitat Suitabilities.............................................................................................15
2.4.3 Weighted Suitable Area (WSA).....................................................................................15
2.4.5 Summation.....................................................................................................................15
2.4.6 Cover Preferences ..........................................................................................................15
2.5 Results...................................................................................................................................16
2.5.1 Habitat Criteria...............................................................................................................16
2.5.2 Species and Group Summaries.......................................................................................17
2.5.3 Habitat Supply and Transformation...............................................................................17
2.5.4 Integrated Assessment ....................................................................................................17
2.6 An Overview of Habitat Suitability Index Values................................................................17
3.0 INFORMATION GATHERING AND DATA ASSEMBLY.................................................18
3.1 Information Gathering ..........................................................................................................18
3.1.1 Position, Dimensions, and Context ................................................................................19
3.1.2 Physical Habitat Characteristics.....................................................................................19
3.1.3 Other Habitat Conditions ...............................................................................................21
3.1.4 Fish Community and Fisheries.......................................................................................22
3.2 Data Assembly for an Application.......................................................................................22
3.2.1 Physical Habitat Assessment Scenarios .........................................................................23
3.2.2 Cross-link Scenario and Database Habitat Classes........................................................23
v
3.2.3 Fish Community, Ecosystem Type, and Fishery Objectives .........................................24
3.2.4 Location Spp., Fish Group and Life Stage Weights.......................................................24
3.2.5 Assessment and Analysis Steps......................................................................................25
4.0 CASE STUDIES ......................................................................................................................25
4.1 Case 1: Brant Inn Node, Burlington, Ontario .......................................................................26
4.1.1 Background ....................................................................................................................26
4.1.2 Process............................................................................................................................26
4.1.3 Application.....................................................................................................................27
4.1.4 Results ............................................................................................................................27
4.1.5 Outcome .........................................................................................................................28
4.1.6 Commentary...................................................................................................................28
4.2 Case 2: Wallik Property, Oakville, Ontario ..........................................................................28
4.2.1 Background ....................................................................................................................28
4.2.2 Process............................................................................................................................29
4.2.3 Application.....................................................................................................................29
4.2.4 Results ............................................................................................................................30
4.2.5 Outcome .........................................................................................................................30
4.2.6 Commentary...................................................................................................................30
5.0 DISCUSSION..........................................................................................................................30
5.1 Conceptual Foundations .......................................................................................................30
5.1.1 Habitat Suitability Index Models ...................................................................................30
5.1.2 Balancing Losses and Gains...........................................................................................31
5.1.3 Use of Weights in Decision-making ..............................................................................32
5.1.4 Pragmatism versus Scientific Rigour .............................................................................33
5.2 Scientific Defensibility Requirements..................................................................................34
5.3 Proponent-Regulator Interactions .........................................................................................34
5.4 Future Plans and Extensions .................................................................................................35
6.0 ACKNOWLEDGEMENTS .....................................................................................................35
7.0 REFERENCES ........................................................................................................................36
vi
LIST OF FIGURES
Figure 2.1.1 A flow-chart showing the overall structure of HSM and the activities in the
Definition, Analysis, and Reports components..............................................................................40
Figure 2.4.1 A representation of the habitat suitability matrix (r) for a single species-life stage
combination....................................................................................................................................44
Figure 2.6.1 Mean habitat suitability index values among life stages (y-axis) versus depth (zaxis) and substrate classes (x-axis) for three thermal groupings of fish species (warm, cool and
cold) across three cover classes (no cover, submergent, and emergent vegetation)......................54
Figure 3.2.1 The framework for completing an application of the Habitat Suitability Matrix
method. The numbers provide the logical order in which the steps are completed with subordering indicated with lowercase letters.......................................................................................56
Figure 3.2.2 The fish habitat area categories illustrated for the construction of a solid pier on the
shore of a lake. ...............................................................................................................................57
Figure 4.1.1 Illustration of case study 1, Brant Inn Node, A) prior to project and B) components
of the preferred development option. .............................................................................................58
Figure 4.1.2 Bar-charts of A) pre-, post-, and net weighted suitable area (WSA) and B) area
affected by type for the original 6 development options (option 0 was to do nothing) at Brant Inn
Node on the Burlington waterfront of Lake Ontario. ....................................................................59
Figure 4.2.1 Illustration of the A) pre-development condition and B) post-development condition
of case study 2, the Wallik property. .............................................................................................61
Figure A1 Plots showing the relationship between the set of individual fish metrics (number of
species, log numbers, log biomass and adjusted IBI) and the standard factor score obtained with
the first principal component. ........................................................................................................68
Figure A2 Plot of the fish community metric based on the PCA versus log10 maximum effective
fetch, km, for 501 electrofishing samples. The edge bar diagrams show the frequency
distributions for the two data variables and the dashed lines show the mid-points for the two
clusters and the fetch cut-off for the two clusters. .........................................................................69
LIST OF TABLES
Table 2.3.1 Sample hypothetical pre- and post-development scenario data files for use in an
application of HSM........................................................................................................................41
Table 2.3.2 Area categories for assessing net change of productivity of fish habitat after Minns
(1995a, 1997). All areas in HSM scenarios should be assigned to one of these categories. .........42
Table 2.3.3 Alternate methods for assigning percentages to vegetation cover categories
contingent on the type of information available. ...........................................................................43
Table 2.4.1 How species level suitability matrices (r) are pooled into fish group matrices........45
Table 2.4.2 Assignment of weights among fish group and life stage suitability matrices (r). The
weights are proportions that sum to 1 on each axis and the sum of their cross-products also
equals 1. .........................................................................................................................................45
Table 2.4.3 Inferred no cover preference levels based on the cross-matrix of emergent and
submergent preferences for use when computing suitability matrices of Great Lakes basin
lacustrine fishes..............................................................................................................................46
Table 2.5.1 Sample input records of a hypothetical scenario pair for an exposed Lake Ontario
shoreline site used as the basis for demonstrating sample output..................................................47
Table 2.5.2 Summary of habitat criteria provided in a standard report for an application of HSM:
A) Lake habitat types and weights, and B) Lake habitat classification. ........................................48
vii
Table 2.5.3 Lake habitat cross-classifications for linking reporting and scenario attribute classes
to the internal classes for all three life stages in A) Depth zones, B) Substrates, and C)
Vegetation cover. ...........................................................................................................................49
Table 2.5.4 Fish group names, weights, numbers of species overall and by life stage, and species
lists for an application of HSM drawing from the Lake Ontario fish taxon. .................................50
Table 2.5.5 Sample habitat supply area summary, m2 , by area type and habitat category for adult
and spawning+YOY habitat classes. Using the input matrices, scenario areas are apportioned
among the internal habitat classes..................................................................................................51
Table 2.5.6 Weighted suitable areas (WSA) for the demonstration scenarios showing the format
and organization of the report table. WSA, m2 , are the equivalent of the habitat area if habitat
suitability were at the maximum....................................................................................................52
Table 2.6.1 Average suitability index values by thermal group of fish species for 150 unique
combinations of depth, substrate, and vegetation cover classes, computed using HSM. ..............53
Table 2.6.2 Pearson correlations and their Bonferroni-adjusted significance levels among 18
habitat suitability indices representing all combinations of thermal, life stage, and trophic level
groupings of the freshwater fishes occurring in lakes in the Great Lakes basin............................55
Table 4.1.2 The weighted suitable area (WSA) results for the final scenario of case study 1, the
Brant Inn Node, application of HSM, showing pre- and post-development, net change, and
percent change in WSA by life stage, fish groups and pooled groups, along with group weights.
.......................................................................................................................................................60
Table 4.2.1 The weighted suitable area (WSA) results for case study 2, the Wallik property,
application of HSM, showing pre- and post- development, net change, and percent change in
WSA by life stage, fish groups and pooled groups, along with group weights.............................62
Table 5.1.1 Derivation of eight principles of HSM method by comparison with the fundamental
principles of cost-benefit analysis, after Griffin (1998).................................................................63
Table A1 Results of principle components analysis to obtain a factor representative of several
fish community metrics using 501 electrofishing samples from various shore locations in the
lower Great Lakes. .........................................................................................................................67
Table A2 Results of the analyses of variance for the K-means clustering. ...................................67
Table A3 The results of the K-means clustering of sample data (Figure A2) using log10
maximum effective fetch and the standardized fish metric. ..........................................................67
Table A4 Rules for assigning Condition Index values in analyses of breakwall projects on highly
exposed shorelines of large lakes...................................................................................................67
LIST OF APPENDICES
Appendix A: Adjusting analysis for changes in wave exposure on the Great Lakes. ...................64
Appendix B: Pre- and post-development scenario habitat data sets for the two case studies
presented in section 4. Area types are described in Table 2.3.2. ...................................................70
viii
1.0 INTRODUCTION
Human development activities in aquatic ecosystems can alter, disrupt, or destroy fish
habitat. Assessing the induced changes in fish habitat features and then detecting responses in
fish communities are scientifically complex problems (Minns et al. 1996). There are too many
separate development activities being proposed every day, too few people in habitat management
agencies for every proposed activity, and too much time would be required to complete a
thorough assessment in every instance. Quantifying changes in habitats and the effect of those
changes on fish populations is costly and time-consuming.
Assessment of proposed development activities also poses a major conundrum. Complete
studies of human development, or ecosystem restoration activities, requires both pre- and postproject data. This comparative approach cannot be followed for most projects as the decision to
allow the activity must always precede the post-development assessment and, given human
nature, projects are rarely undone even when post-project assessment shows damage has
occurred. As an alternative, experimental studies of classes of development or restoration
activities could be a useful guide. Unfortunately, few such experiments have been performed.
There have also been few efforts applied to scientifically evaluate the impact of earlier
development activities as a guide to assessing proposed changes (Cudmore-Vokey et al. 2000).
In the absence of experimental studies or the opportunity to gather comparative evidence,
assessment methods must be sought that make best use of the available scientific evidence. Such
methods will provide the most defensible estimates of impact and hence as sound a basis for
decision-making as is achievable with existing science. Of necessity, these methods will have to
rely on predictive models. Such models are amenable to scientific testing and models can be
modified or replaced as new knowledge arises. These habitat assessment and evaluation tools can
be collectively described as ‘scientifically defensible methods.’
The purpose of this report is to present a conceptual basis of the ‘Habitat Suitability
Matrix’ (HSM) method and software used to assess development projects affecting lacustrine
fish habitats in the Great Lakes basin. HSM is based on a prototype described by Minns et al.
(1995a) for coastal headlands and islands in the lower Great Lakes. The approach has been
implemented as an interactive software application which will soon be available (estimated date,
April 2001) to the public on the internet (see http://www.bio-software.com). This report
describes in detail the approach taken to assessing fish habitat in lacustrine ecosystems using
scientifically defensible methods. The report also gives a step by step rationale and guide for
potential users of the software. However, it is not intended as a software user guide.
The scope for application of this current version of the approach is as follows:
• It is mainly intended to be used to assess development projects occurring in the Great
Lakes and other large inland lakes in the Great Lakes basin.
• With due attention to the fish species being considered and the landscape contexts, this
version should also be applicable to other large lakes and smaller inland lakes throughout
Ontario.
• There is also a similar database for fish species occurring in lacustrine habitats in
Newfoundland and Labrador. It is not described here but the approaches available for
Great Lakes basin applications is transferable.
• It is primarily intended for the assessment of development activities involving physical
alteration of lacustrine habitats such as shoreworks and coastal protection; harbours,
piers, and docks; or creation of promontories and islands.
1
•
This current version is not intended for use for assessing development activities affecting
streams and rivers – planned later versions will address these activities.
• Completion of an HSM-based assessment does not, by itself, constitute completion of the
review, assessment, and decision processes undertaken by Fish Habitat Management staff
or their delegated authorities, but can constructively contribute to the fulfilment of those
processes as required under the Fisheries Act and in compliance with the federal Policy
for the Management of Fish Habitat.
This introduction continues with a review of relevant legislation and policy, a brief
description of the net change equations under-pinning the approach, a review of literature
regarding other approaches to assessment and the origins of some aspects of HSM approach to
developing a scientifically-defensible methodology, a summary statement of the principles and
assumptions that are the foundation of the approach, and finally, an outline of the contents of
succeeding sections.
1.1 Legislative and Policy Background
The federal government of Canada has specific legislation, the Fisheries Act, and policy,
The Policy for the Management of Fish Habitat, that provide the primary reference points for
assessing proposed changes to fish habitats and for deciding if they are permissible. These
instruments are administered by Fisheries and Oceans Canada. Both implicitly have quantitative
elements indicating the potential for numerical assessment criteria to be used in regulatory
decision-making.
The original Fisheries Act dates from Confederation in the mid-19th century. The Act
provided a means for managing and protecting Canada’s fishery resources and covers all fishing
areas, seas and inland waters. The Act is binding on all levels of government in Canada and
supersedes provincial legislation. Habitat and pollution provisions were added to the Act in the
1970s. Section 35 deals specifically with habitat protection. Section 35(1) is a general
prohibition forbidding the ‘harmful alteration, disruption or destruction’ (HADD) of fish
habitats. Under Section 35(2) the Minister of Fisheries and Oceans can authorize a HADD
allowing a project to avoid violating 35(1). Securing an authorization is not mandatory but large
penalties can be imposed on those causing a HADD without prior ministerial authorization.
Section 35 is implicitly quantitative as the decision that an activity affecting fish habitat is
harmful implies there are thresholds or gradients of negative impacts on fish productivity.
In 1986, DFO issued its Policy for the Management of Fish Habitat. The Policy
recognized the essential role of habitats in maintenance of healthy, self-sustaining, productive
fisheries and reaffirmed the government’s commitment to managing and protecting them. The
guiding principle of the Policy is ‘no net loss of the productive capacity of habitats’. Consistent
with the Fisheries Act, productive capacity is defined as ‘the maximum natural capability of
habitats to produce healthy fish, safe for human consumption, or to support or produce aquatic
organisms upon which fish depend’. Three policy goals provide the basis for achieving net gain:
conservation of existing habitats, restoration of damaged ones, and development (or creation) of
new ones. The Policy also addresses the need for integrated management of fish habitats to
support fish productivity and fisheries in the context of other ecosystem uses and requirements.
The policy is also implicitly quantitative as determining whether or not changes to fish habitats
achieve a net gain or loss requires the comparison of measures of productivity estimated with
and without the proposed change.
In practice, the habitat provisions in Section 35 of the Act and the Policy are linked. First,
proponents are directed to avoid unacceptable HADDs. Next, proponents are encouraged to seek
2
mitigation actions that prevent any HADD and hence avoid the requirement for an authorization.
Finally, if mitigation fails or is insufficient to prevent a HADD, an authorization is required and
proponents are then obligated to develop a set of compensatory actions that will result in an
overall net gain, or at least no net loss (NNL). There are several types of potential compensation:
combinations of habitat creation or productivity enhancement, like-for-like or unlike, within the
same or different ecological unit, and use of rare measures, such as artificial production. Likefor-like habitat compensation within the same ecological unit is the most preferred type (DFO
2000).
1.2 A Quantitative Framework for Measuring Net Change
Minns (1995a, 1997) presented a quantitative basis for assessing net gain or loss of
natural productivity of fish habitats. The equations given in those papers provide the basic
framework used to compute net change with the HSM approach. Net change is computed as
follows (equation 15 in Minns 1997):
∆PNOW
= [pMOD - pNOW].AMOD - pMAX.ALOSS + [pCOM - pNOW].ACOM
where,
∆PNOW
= net change of natural productivity of fish habitat
= Area of habitat lost due to development activity
ALOSS
AMOD
= Area modified, directly and indirectly, as a result of the development
activity
ACOM
= Area created or modified elsewhere to compensate for the development
activity
= Maximum potential unit area productivity rate (or productive capacity)
pMAX
pNOW
= Present unit area productivity rate
pMOD
= Modified unit area productivity rate in affected areas
pCOM
= Compensation unit area productivity rate in affected areas.
This equation can be generalized for a habitat area consisting of a mosaic of habitat units with
varying attributes and unit area productivity rates. The maximum productivity rate, PMAX, is set to
a value of 1 and all others are proportions of that reference value. Where multiple habitat units
are involved, the highest maximum productivity rate is set to one and all other rates, maximum
or otherwise, are expressed as proportions. There is no single variable that measures the
productivity rate of a unit of fish habitat. A variety of metrics might be used to infer the overall
productivity per unit area, e.g., fish biomass, fish productivity, benthic biomass, species richness
of the macrophyte or zooplankton communities, or an index of biotic integrity. To compute a
proportional productivity index for various sites or habitat types, data on one or preferably
several metrics combined would be rescaled using the representative portion of the range. For
example, Minns et al (1994) used the 95th percentile of various metrics to define the upper
bounds for indices used in developing an index of biotic integrity.
The division of areas affected by a development project into loss, modified, and
compensation categories are implemented in the HSM approach and provides a straightforward
means of appreciating the scope of impacts attributable to a development activity.
Successful implementation of a scientifically-defensible method like HSM hinges on
developing a method of estimating unit area productivity rates, or an equivalent, surrogate metric
such as habitat suitability indices. Pragmatic derivation of indices acting as surrogates for habitat
productivity will have to continue as the norm until more explicit, defensible methods of
accountng for productive capacity have emerged.
3
Various properties of this mathematical scheme were described by Minns (1995a, 1997).
The key property with respect to assessing development activities are the distinctions among
absolute loss where habitat is destroyed, absolute gain where habitat is created, and relative
changes in modified or enhanced habitats. This property provides an a posteriori rationale for the
compensation multiplier rules of thumb that have often been developed in the past. For example,
if a development destroys one unit of habitat with a proportional productivity rate of 1,
improvement of productivity rate from 0.4 to 0.6 will require that five units of habitat be
modified to achieve no net loss. In simple terms, the proponent is charged for losses at the
maximum productivity rate and for modification at the difference of future and current rates, and
is credited for improvements due to compensation at the difference of future and current rates.
1.3 Previous Quantitative Assessment Schemes
There are relatively few quantitative schemes for use in the assessment and management
of development activities affecting fish habitat. The Habitat Evaluation Procedures (HEP) and
associated Habitat Suitability Index (HSI) modelling approaches developed by the U.S. Fish and
Wildlife Service (USFWS 1981, Terrell et al. 1982, Terrell 1984) in the late 1970s and early
1980s contributed significantly to the conceptual evolution of the HSM approach. The siteassessment scheme developed by Christie (1982) for Ontario Hydro to assist with the evaluation
and selection of sites suitable for nuclear power plants on the Great Lakes provided much
stimulation for the design of model described here. Concepts embodied in the gap analysis
process (GAP) for terrestrial ecosystems (Scott et al. 1993, Jennings 2000) provided additional
stimulation.
Christie (1982) compiled an extensive database characterizing various life history and
habitat preference characteristics of Great Lakes fishes and used it to examine the database’s
aggregate power to predict presence-absence of particular species at specific locations. Overall,
his scheme performed quite well when applied to sites where fish community survey data were
available. The scheme was a descriptive rule-based approach with minimal computation needs.
While it showed the potential power of life history and habitat attributes of species, no attempt
was made to directly assess the suitability of specific habitat areas. Grandmottet (1983) and
Degiorgi and Grandmottet (1993) laid the groundwork for a life-history habitat attribute-based
approach to assessing fish habitat which is providing much stimulus for stream habitat
classification and assessment in France. These studies showed how species-based assessment of
preference can provide a means for discriminating among habitat types in freshwater ecosystems.
1.4 Content of this Report
This report builds on the preliminary work reported by Minns et al. (1995a). In section 2,
the framework for this defensible method, a Habitat Suitability Matrix (HSM) method, is
presented with detailed descriptions and examples for each step. In section 3, preliminary
guidelines are presented for the collection and organization of all information required for
successful application of the approach. In section 4, a pair of case studies is presented illustrating
the application of the HSM-based assessment model to development projects affecting fish
habitat. In section 5, a review and analysis of conceptual foundations of the methodology is
presented, identifying current limitations of the approach and potential future scope for
extension, elaboration, and application of the framework.
2.0 FRAMEWORK FOR ASSESSING FISH HABITAT
Minns et al. (1995a) presented a prototype for the present approach to assessing the
potential impact of human development activities on fish habitat in lakes. The essence of the
approach is the idea that the habitat preferences of individual fish species and life stages can
4
be quantified and aggregated into habitat suitability indices that in turn can be used as
surrogate measures of fish habitat productivity. The derivation and application of habitat
suitability indices has been embedded in an assessment framework, known as Defensible
Methods of Assessing Fish Habitat, and implemented in an software product known as the
Habitat Suitability Matrix (HSM) method.
The implementation of the framework is founded on and guided by a basic set of
principles and assumptions. These principles and assumptions need to be recognized from the
outset.
2.0.1 Principles
• Use assemblages not single species - Development sites should be assessed in terms of the
habitat requirements of all fish species potentially using the area. Assessments based on the
requirements of single species are unlikely to represent the needs of the whole assemblage.
Experience from ecosystem science indicates that groups of species have linked
requirements. For example, a top predator has habitat and prey needs. The habitat needs of
the prey must be considered in efforts to sustain predator populations. This principle is
consistent with the ecosystem approach which is widely espoused for the management of
freshwater and other ecosystems in North America and elsewhere (Allen et al. 1993; Minns
1995b).
• Use surrogate physical/chemical indicators of fish productivity - Direct measures of fish
productivity are costly and time-consuming to obtain. In the context of site-level
development assessments, a measure of fish productivity might be possible for the predevelopment scenario. However a prediction for the post-development scenario will still be
needed as authorization to proceed must inevitably precede project implementation and
attainment of post-development conditions. This is not to imply that all biological evidence
should be ignored. However, site-specific assessments of the biological community
undertaken in support of applications for development approval rarely include adequate data.
Assessments consisting of a few site visits with limited sampling cannot possibly provide an
adequate appraisal of year-round and diurnal use of habitat by the complete fish assemblage
and the other organisms whose productivity supports that of fish.
• The science is incomplete - Fish habitat science is incomplete and will remain so for the
foreseeable future. Knowledge of the details of the life histories, habitat requirements,
trophic interactions of many fish, and other biota present in freshwater ecosystems, is poor or
extremely limited. There are on-going efforts to expand fish habitat science though there are
many types of ecosystems and development activities to be considered (cf Lester et al. 1997;
Levings et al. 1997). Decisions will still be made despite the incompleteness of the science.
It is impractical to delay decision-making for any development activity because of a lack of
knowledge as long as there is no societal commitment to funds the responsible agencies to
fill those information and knowledge gaps. Existing conceptual understanding and data must
be used in a scientific manner to provide the best possible guidance and assessment.
• Precautionary Principle - Where uncertainty is very high for certain classes of development
or classes of habitat, the Precautionary Principle should apply (Garcia 1994; Richards and
Maguire 1998). The default conclusion should be that the proposed activities will cause an
unacceptable net loss of productivity of fish habitat and that the project should not proceed in
the absence of further studies to reduce the uncertainty to acceptable levels. Application of
this principle requires a careful exercise of judgement taking into account the size and/or
significance of the fish habitat productivity at risk.
5
•
Quantitative process - Habitat assessment is a quantitative process involving an integrated
consideration of the quantity and quality of constituent habitats. Assessment of net gain or
loss as described in the Policy is only possible if a quantitative approach is used (Minns
1995a, 1997). The steps involved in the development of quantification force the assessment
process to become transparent and assumptions are made explicit.
• Affected areas only - Only habitat areas affected, directly or indirectly, by a proposed
development should be included in a site assessment. If larger frames of reference are
adopted for an assessment, it is easy to generate an analysis where the relative or proportional
impact of any site-specific development will appear to be minimal or trivial. This situation
might be less of a problem if area fish habitat management plans as envisaged in the Policy
were in place. Such plans would make explicit links between fish community and fishery
objectives and identify the quantities and qualities of habitats needed to sustain those
objectives. With such a plan as a reference point for evaluation of site-specific assessments,
decision-makers would perhaps have more leeway and flexibility when considering the
significance of expected impacts at particular sites. Meanwhile, in the absence of such plans,
the only prudent course is to limit the assessment to affected areas, thereby conforming to the
Precautionary Principle.
• Equal treatment of pre- and post-development scenarios - The same criteria (habitat features,
fish assemblage, weights, etc.) should be used for both pre- and post- development scenarios
at a site. Each site-specific assessment should be judged in the same context with the same
fish community and fishery objectives as a reference point. If, for instance, a proponent were
able to assess the pre-development scenario with one set of objectives and the postdevelopment one with another set, this would create distortions where habitat favourable for
one group of species would be replaced with habitat suitable for another group of species.
Certainly, there should be some leeway for fish community and fishery objectives to evolve
over time in response to increased understanding, changing fishery priorities, etc., but sitespecific development activities should not be used as an instrument of such an evolution.
Such an approach would be irresponsible and would likely encourage opportunism.
2.0.2 Assumptions
Given the embryonic status of quantitative fish habitat science, a number of assumptions
are necessary to facilitate the development of a quantitative assessment scheme. None of these
assumptions has been adequately tested or challenged but they represent a minimum set. While
for particular life stages of some fish species there may be evidence that violates these
assumptions, for the majority of species our knowledge base is minimal. As the purpose here is
to provide some improvement over the subjectivity of decision-making now in place, these
assumptions are a minimum requirement to allow effective scientific use of what little
information is available. It is expected, hoped even, that future scientific research will provide
evidence that will confirm, enhance or supplant this initial effort. The key assumptions are as
follows:
• Suitability as a surrogate for productivity – The approach taken here assumes there is a
simple one-to-one correspondence between suitability indices, scaled between 0 and 1, and
the fish productivity of habitats. Alternative assumptions would require that non-linearities in
habitat-productivity relationships and life-stage bottlenecks be fully known (cf Minns et al.
1996). Until such non-linearities are better understood, the linear assumption is the prudent
course.
6
•
Suitability criteria - The analysis of suitability as a measure of maximum potential natural
productivity, or productive capacity, only considers three habitat features: depth, substrate,
and vegetation cover. While other features of cover have been noted (Lane et al. 1996a,b,c)
as being important for various species and life-stages, there is little evidence that habitat
mapping is ever of sufficient detail in lacustrine environments to justify their inclusion in this
version. However, work is currently being done to expand cover preferences to include nonvegetated cover. Suitability indices based on physical habitat features are most likely
reflective of the maximum potential productivity or productive capacity and hence net gains
will be largely achieved by shifting habitat conditions toward those with higher intrinsic
productivity values.
• Condition index: - Suitability indices may be modified to reflect current, realized natural
productivity through the use of a condition index. This condition index is scaled between 0
and 1, and reflects an assessment of the impact of other non-physical habitat variables such
as nutrient, contaminant, and acidity status in water and substrates, the presence of health
impairments in the fish due to human activities, current patterns and velocities, and thermal
habitat. For example, if water quality at an urban habitat is currently impaired by poorly
managed stormwater flow, a condition that will be fixed as part of a proposed habitat
alteration, the condition factor for affected habitat might be set below 1 in the scenario
describing present conditions and set at 1 for the post-development scenario. Another
example describes the use of the condition index in relation to altered shoreline exposure
later in the report, section 3.1.2, and in Appendix A.
• Mathematically simple - All habitat features are treated independently in the calculation of
suitability. Simple addition and multiplication are the means by which elements are linked.
• Relative scaling - All suitabilities can be expressed on a relative scale of zero to one (0-1)
allowing weighted suitable areas to be computed. This in turn allows suitable areas to be
computed as the product of area, suitability, and condition then summed across all affected
areas in scenarios.
2.1 Overview of Habitat Suitability Matrix Method
The assessment framework has three main components: Definition, Analysis, and
Reports (Figure 2.1.1). Users provide the necessary definition information, run an analysis, and
then select from among report options to obtain a set of results to assess expected net gain or loss
of productivity of fish habitat. Habitat scenarios for pre- and post-development contexts are
prepared as part of the Definition component and must meet a set of basic requirements before
the Analysis component can proceed. A number of decisions about the membership of the fish
assemblage being assessed and the linked fishery-habitat management objectives at the site and
in adjacent areas must also be settled during completion of the Definition component.
The Definition component has four modules: Location, Species Groups, Scenarios, and
Habitat Variables (additional modules may be added in the future covering topics such as project
documentation and statistics on the history of the definition). The Analysis component is where
the analytical model is specified, the integration of inputs and reference databases, and the
calculation of results is performed. The actual mechanics of the analysis are not seen by or
accessible to the user of the software but all the details of how the calculations are performed is
described in detail below. The Reports component provides several options for exploring the
many layers of calculations and intermediate results that together produce the integrated results
for an assessment of net gain or loss. Data in results tables can be copied via a clipboard to
spreadsheets or they can be exported as files. In addition, results tables currently opened in the
7
Reports component can be printed. A standard print-out gives a detailed statement of the
decisions made in setting up the definition and a subset of results judged to be a minimum set for
appraising a development/restoration referral.
2.2 Definition
Several linked steps make up the Definition component. Before the formal definition
phase can proceed, preparatory discussions among proponents, regulatory agencies, and other
stakeholders are advised to build a consensus on the fishery and fish habitat objectives as a
context for the assessment. The formal steps include:
(a) Location where a biogeographically appropriate species list is selected;
(b) Groups where the species are assembled into groups using agreed criteria involving both
ecological or user-defined principles. Group weights are assigned in this step indicating
the relative importance or significance attached to different groups of fish species;
(c) Scenarios where a series of one or more physical habitat scenarios for an area assessment
are brought into the definition; and
(d) Fish Habitat Features are first used to select and weight the fish habitat features to be used
in the assessment. Later, after scenarios have been imported, input, internal and reporting
habitat categories must be reconciled.
Choices made in the Definition component strongly influence the results of the analysis and must
be accompanied by supporting justification and documentation as determined by regulatory
authorities. The process leading to these choices must be clearly documented in the final
assessment report.
2.2.1 Location Species List
Selection of a species list for the site is an important step in the Definition component.
The species assemblage chosen shapes the context of the analysis. If the habitat preferences of
coldwater piscivores are the only ones considered, habitats important to warmwater and
coolwater species will be devalued in the analysis. Hence, the species list should not be based
solely on the detection of species at or in the immediate vicinity of the proposed development
site although any species so reported should be included. Surveys with limited spatial or
temporal extent are unlikely to secure a complete list. Pre-existing compilations provide a more
reliable basis for building a list. Complementary approaches are recommended for Great Lakes
sites and inland lakes in the Great Lakes basin or the rest of Ontario.
On the Great Lakes, the complete species list reported should be used. There is no need to
worry about including species which spend all or some life stages confined to lotic habitats. The
species list is filtered by life stage and any confined exclusively to streams and rivers are
excluded from the analysis. Dr. E.J. Crossman and his M.Sc. student, Becky Cudmore, at the
Royal Ontario Museum (personal communication) have been compiling a complete species
listing for all of the Great Lakes (Cudmore-Vokey and Crossman 2000).
In individual inland lakes, species lists tend to be shorter than the list reported for their
secondary or tertiary watershed as a whole (Minns 1989). Various evidence points to relatively
high levels of species turnover in inland lakes and streams, particularly among the smaller
species (Magnuson et al. 1994, He and Kitchell 1990, Robinson and Tonn 1989, Jackson and
Harvey 1997). Therefore it is recommended, as a starting point, that tertiary watershed species
lists be used for habitat evaluation in all lakes occurring in each watershed. The species on a
watershed list can be pared away where particular circumstances justify it. For example, while
lake trout may be present in a watershed, only lakes with an oxygenated hypolimnion during the
summer will be capable of supporting them.
8
There are several sources of reported watershed species lists (Minns 1989, Minns and
Moore 1995, Doka et al. 1997) and reliable maps of most fish species distribution are provided
for Ontario in Mandrak and Crossman (1992). A comprehensive national tertiary watershed
database is being prepared as a standardized source for HSM applications (Mandrak and Minns,
unpublished material).
The creation and editing of a location species list is undertaken in a separate module of
the software application. A library of species found in the Great Lakes, other large lakes, tertiary
watersheds, and other significant categories will be available as part of the software application
with a georeferencing module to allow users to determine the watershed code from latitude and
longitude coordinates.
Users need to ensure that the reviewing agencies agree with the species list selected for a
location. Arbitrary lists and lists containing very few species will produce spurious or
questionable results.
2.2.2 Assembling Fish Species Groups
There are many bases on which fish species might be grouped. Species might be grouped
on the basis of their usefulness to people (sport vs. commercial, game vs. pan vs. forage, etc.).
However, since the objectives of fish habitat management are closely aligned to those of
ecosystem management, ecological attributes probably have more relevance. Fish are often
grouped according to their thermal preferences (warmwater, coolwater, and coldwater) and food
habits (piscivore, omnivore, detritivore, herbivore, invertivore, etc.). The spawning guild (Balon
1975) of the species may play a major role and might be used as a grouping criteria. Taxonomy
might also be used to form groups although they will often align with other attributes. These
attributes have been tabulated along with other life style characters to assist the formation of fish
groups.
The presence-absence of piscivory is one of the most important attributes given the role
of top-down role predators have in shaping the particle-size spectrum of biomass and production.
Different sites will have different thermal regimes and the fish groups will play an important role
once species’ group weights are applied. Many exposed sites on the north shore of Lake Ontario
are affected by upwellings during the summer, bringing coldwater species inshore. Those species
may be an important part of a nearshore fishery and may be weighed more heavily in the analysis
of habitat alterations. Elsewhere, warmwater species may be more favoured.
Additional criteria for the creation of species groups have been established: taxonomic
groups at the genus and family level, and life history attributes including size and age at maturity
and at maximum sizes. These may be used to divide the species list into a workable set of groups
reflecting the fisheries objectives at that location. Care must be taken to avoid creating groups
with the same species occurring more than once as exclusivity cannot be guaranteed and the
results are compromised.
Users may choose to create arbitrary fish groups but need to make sure the reviewing
agencies agree with the group assignments. If arbitrary groupings of fishes are created, there
should be further investigating by the reviewing agency to ensure the groups do not induce a
flawed analysis. For example, mixing coldwater and warmwater species in the same group might
tend to undervalue habitats preferred by coldwater species.
2.2.3 Fish Habitat Criteria
In the prototype version of HSM method (Minns et al. 1995), there were habitat
suitability criteria based on:
(a) the spawning and adult habitat requirements compiled by Christie (1982);
9
(b) a model predicting aggregate fish attributes on the basis of vegetation cover (Randall et
al. 1996) for transects sampled at the 1.5 metre contour, and
(c) suitabilities extrapolated to other depths using results presented by Keast and Harker
(1977).
Some of these criteria proved difficult to expand to a basis for a fuller implementation of HSM.
Christie (1982) had compiled a large amount of habitat use information for species occurring in
the Great Lakes but often the categories for various habitat features were too coarse for practical
use in shallow nearshore waters. It is expected that work, such as that by Randall et al. (1996),
will eventually provide alternate models linking fish assemblage attributes to habitat features in
the future once the range of habitat types is more comprehensive and the means of generating
suitabilities from predictive modelling is established. Systematic comparative evidence of how
fish assemblage attributes vary with depth in lakes is not very common. Such evidence will be
needed if more sophisticated suitability models are to be developed.
To overcome the limitations encountered in the prototype, a standardized process of
assembling a summary of habitat preferences by life stage and species was established. The
habitat features were limited to depth, substrate, and cover (mainly vegetation). This process
relied on review and summary of available literature followed by expert peer-review of the
results and publication. At present there are three habitat criteria recommended for HSM-based
analyses in Great Lakes basin lakes: spawning, young-of-the-year (YOY), and adult habitat
requirements based on the data sets compiled by Lane et al. (1996a,b,c). The categories used for
adults differed from those used for spawning and YOY stages in the Lane et al. reports. Those
differences have now been eliminated and a revisory for the Lane et al. databases is being
prepared. These qualitative indications of habitat preference provide a basis for a quantitative
suitability measure when combined for groups of species. For each species and life stage,
suitability matrices are computed for combinations of depth, substrate, and cover categories.
Databases for stream habitats in the Great Lakes basin and lake habitats in Newfoundland
and Labrador has been completed. Extensions of HSM method to those contexts are under
development and will be reported elsewhere. Lacustrine databases for British Columbia, the
Prairies, and the North (NWT and Nunavut) are under development and will be added later,
widening the geographic scope of applications.
2.2.4 Group and Criteria Weights
For the analysis to produce an integrated assessment of net gain or loss, weights must be
applied to both the fish groups and the life stage habitat criteria. The weights are proportions
whose sum across groups or criteria must equal one. For example, if the species list is divided
into six groups according to thermal preference and the presence-absence of piscivory, each
group could be assigned the same weight of 1/6th. This indicates that all three thermal groups are
given equal weight and that piscivores are weighted equally to non-piscivores. These weights
eliminate any differences in the number of species contributing to each group. Thus, if 5
piscivores form one group and 25 non-piscivores form another group, one composite suitability
matrix is compiled for each of the groups and the external user-supplied weights determine their
contribution, not the number of species. Setting the values of the weights requires a reasoned
assessment of the ecosystem conditions, the fish assemblage and fisheries, and fishery objectives
in consultation with the appropriate management agencies and the process is described in more
detail in section 3.2.4.
10
2.3 Habitat Scenarios
2.3.1 Data Specifications and Format
The scenario data files for HSM applications have a simple text-based, comma-delimited
format. Users may derive their scenario information from more complicated application software
but it is required that these applications export their results in the standard format. The data files
contain three main types of record: titles, directives, and habitat units (Table 2.3.1). The title
records allow the user to keep track of different scenarios as various options are considered. The
first title record provides the name used in HSM to track a scenario. The directives records
provide information about the habitat unit type, i.e. area, volume or length. Also, under
directives, the units record gives the measurement scale, e.g. metres squared. The order record
provides the order of variables on the habitat unit records. The unit identifier allows the user to
keep track of habitat units on maps and to compare the same unit in pre- and post-development
scenarios. The area associated with a particular ID should preferably be the same in both cases
when dealing with habitat units that are modified as this will facilitate more detailed
comparisons.
2.3.2 Scenario Options
The HSM method can be applied in two distinct ways: (1) to compare pre- and postdevelopment scenarios and compute net gain or loss of suitable habitat, and (2) to provide an
estimate of suitable habitat supply or to map relative habitat suitability for a larger part or all of a
lake ecosystem. The first way is the primary one for site-specific applications, while the second
is linked to larger-scale efforts to develop area habitat management plans for fish habitat. In both
instances, it is necessary to delimit the areas of habitat affected or involved and to break down
the areas into a series of discrete units or patches each with a distinct set of habitat
characteristics. There is no a priori minimum patch area for these analyses and those using this
approach must be guided by the resolution of the habitat mapping information available and the
preciseness of the habitat classification criteria.
To compute the net gain or loss of habitat, one scenario must be designated PRE and
another POST. Since the habitat area included in the PRE scenario is primarily defined by the
modifications shown in the POST scenario, these scenarios will usually form pairs. In the
planning phase several alternate schemes may be considered. Users will have to ensure that they
have a way of documenting the differences among several PRE/POST pairs. Sometimes triplet
scenarios will exist, consisting of the baseline pre-development case, the post-development case,
and the post-development case with added compensation. The first scenario entered into the
software package will be treated as the pre-development or baseline scenario for net change
calculations. As application of this approach is intended to be a complementary precursor to
authorization of development activity, the opportunity to obtain a map of pre development
habitats always exists. The resolution and precision of the mapping required is not prescribed
here but will be influenced by regulatory consideration of the potential extent of the impacts, the
nature and significance of the fishery resource affected, the complexity of the situation, and the
availability of prior mapping.
For a whole system supply analysis only one scenario will be used. The results can be
used to refine fishery and fish community management plans, to identify restoration and
conservation needs, to define upper bounds for human development, etc. A detailed example of a
whole system analysis is described for Severn Sound on Georgian Bay (Minns et al 1999).
11
2.3.3 Area Designations
In Minns’ (1995a, 1997) net change accounting equations for fish habitat, a number of
different area types are defined. These area types are used to classify habitat patches in scenarios
(Table 2.3.2) and provide a convenient basis for deriving an overview of a proposed
development activity and associated loss, modification, and compensation areas.
By definition pre-development (PRE) scenarios cannot contain COMC areas and postdevelopment (POST) scenarios cannot contain LOSS areas. MODD and MODI areas must be
included in both PRE and POST scenarios, using the development dimensions to delimit them in
the PRE. The MOD areas will have different habitat characteristics in the PRE and POST
scenarios. When COMM areas are added to a POST scenario, those areas with their PRE habitat
conditions are also added, though again the habitat characteristics will change. UNCH areas may
be included to provide a complete habitat picture for the habitat patches involved but the area
and habitat characteristics must be identical in both PRE and POST scenarios. UNCH patches
make no difference to the PRE-POST net change calculations.
2.3.4 Habitat Characterization Options
The definitions for the habitat scenarios must include a designation of which type of
depth, substrate, and cover assignment method will be used. There are two methods of
assignment: A) the habitat features are specified as proportional vectors using a defined set of
categories, normally those fixed ones present in the internal habitat requirements database, or B)
habitat features are given arbitrary names or descriptors in the external scenario data file and the
user links each name to a user-specified proportional vector. If an arbitrary set of categories is
used in method A, the user will have to specify the a proportional vector elsewhere in the
definition process. Where detailed analysis of habitat characteristics is available, method A
might be appropriate. More often, habitat units are described in fairly general terms and only an
approximate knowledge of the proportional composition is available, making method B more
suitable. In general, method A is more appropriate where detailed habitat mapping has occurred
and the units of habitat are relatively small. In coarser assessments, habitat units are relatively
larger and method B is more appropriate.
The proportional vectors for the three habitat features have the following interval
boundaries:
Depth – 0, 1,2,5, and 10 metres
Substrate – bedrock, boulder, cobble, rubble, gravel, sand, silt, clay,
hard-pan clay, and pelagic
Cover - no cover, submerged vegetation, and emergent vegetation.
In method A, a substrate type containing 40 % cobble, 30 % rubble, and 30 % gravel will be
represented as follows: 0,0,40,30,30,0,0,0,0,0 ensuring that the percentages sum to 100. In
method B, the same substrate might be named CORUGR1 (for cobble-rubble-gravel mix 1) and
in the habitat features the proportions would be entered manually.
There are a variety of ways the habitat characteristic information might be obtained. The
simplest approach is to develop maps for each habitat feature and then overlay the maps to
identify habitat units with distinct characteristics. For each habitat feature, the following
guidance is offered:
Depth: Most information about depth will tend to be derived from existing contour maps and
hence depths will most often be reported as a range. Although detailed habitat surveys may result
in patches with specific depths assigned. Whether reported by range or specific value, the depth
categories will be have to be cross-linked to the depth ranges used to compute the habitat
12
suitability matrices. The simplest method of cross-mapping ranges is to use the depth intervals to
compute proportions. For example, if an input depth range is 0-5 metres and must be mapped
onto 0-1, 1-2, and 2-5 metre ranges, assume 20 percent is assigned to each of the first two ranges
and 60 percent to the last, representing depth intervals of 1,1, and 3 metres respectively.
Substrate: There are two basic methods of assigning substrate values to a habitat patch:
categorical and exact proportions. In the categorical method, each patch is assigned one of a
finite set of substrate categories based on various published schemes for designating substrate by
particle size (Valere 1996) and corresponding to the categories used in the requirements
database. For example, a patch may be designated bedrock, sand, or clay and for the purposes of
calculating suitability, the patch is assumed to have 100 percent of that substrate type. Some
categorical labelling systems correspond to fixed ranges of proportions method, e.g., Folk
(1954). The centroids of each category in a Folk-like scheme can be determined and proportional
composition determined. For example, patches may be described as sandy-gravel, silty-sand, or
mud. Each label is consistent with a fixed range of proportions for gravel, sand, silt, and clay.
When calculating habitat suitability, the substrate proportions are used to compute a weighted
mean of substrate-specific suitabilities. In the exact proportions method, each patch has a vector
associated with it specifying the proportions of each substrate category present. Again, the
proportions are used to compute a weighted mean of suitabilities.
To ensure a correct matching of user-supplied substrate categories with those used for
assigning suitability values, a proportional assignment matrix must be specified for the
categorical and fixed proportion methods. Otherwise the input habitat scenarios must contain the
proportions of substrates for all nine substrate types for every habitat patch. Anthropogenic
substrates represent a challenge and will generate more debate. For example, if very large
boulders are being used to construct a robust coastal feature, it is not equivalent to the boulder
category in the internal requirements database. A more appropriate representation might be 65 %
bedrock, 25 % boulder for the main characteristic plus 5 % rubble and 5 % sand to represent the
patchy accumulations of finer substrates collecting in and around the interstices.
Cover: There are three ways of specifying cover in a habitat patch: (1) Categorical - no cover,
emergent or submergent; (2) Proportional additive - exact proportions of area are provided and
sum to 100 percent; (3) Proportional non-additive - proportions of emergent and submergent may
overlap (some observers distinguish between vegetation in the water column and that which
sticks through the water surface) (Table 2.3.3). In the categorical scheme, the user must have a
defined basis for assigning one of three categories of cover and the weight is assumed to be 100
percent of that category assigned. In the proportional additive scheme, proportions are assigned
to the three categories and sum to 100 percent. The proportions are used as weights for
combining suitability values. In the proportional non-additive scheme, the no cover value is
computed as 100 - Max(percent emergent cover, percent submergent cover). The three
proportions are then summed and used to re-weight the proportions so they add to 100 percent.
For the purposes of computing habitat suitabilities, habitat areas are orthogonally
disaggregated into homogeneous sub-areas according to the percentage vectors of depth,
substrate, and cover. For example, a habitat area of 100 m2 assigned the following attributes:
depth – 50% 0-1m, 50% 1-2m; substrate – 25% gravel, 50% sand, and 25% silt; and cover – 30%
submergent and 70% nocover, would be disaggregated by the software as follows:
13
Area, m2
3.75
7.50
3.75
8.75
17.50
8.75
3.75
7.50
3.75
8.75
17.50
8.75
Depth zone
0-1
0-1
0-1
0-1
0-1
0-1
1-2
1-2
1-2
1-2
1-2
1-2
Substrate
Gravel
Sand
Silt
Gravel
Sand
Silt
Gravel
Sand
Silt
Gravel
Sand
Silt
Cover
Submergent
Submergent
Submergent
Nocover
Nocover
Nocover
Submergent
Submergent
Submergent
Nocover
Nocover
Nocover
These disaggregated data can be used to assemble habitat profiles using the reference classes,
making comparisons of scenario information simpler. Corresponding HSM values are then used
to compute the weighted suitable area (WSA) for each habitat patch.
2.4 Computation
2.4.1 Individual Species Habitat Suitability
The habitat requirements database for spawning, YOY, and adult life stages contain
preferences for depth ranges, substrate categories, and cover. Preferences are rated as nil, low,
medium, and high (0,1,2, or 3) for substrate and cover. Depth range preferences often vary
seasonally with thermal patterns and life history progressions. Depth preferences were rated by
the number of seasons involved (1,2, or 4) by Lane et al (1996a,b,c) but were represented as
either nil or high when used in database used here. Each habitat feature is treated independently
as there are few species for which non-orthogonal patterns have been described.
The initial suitability matrix (Figure 2.4.1) combining depth, substrate and cover
categories is computed as follows (each matrix consist of a cube of cells representing all
combinations of the three sets of categories):
1) Preference for a single category of a habitat feature, P?? where ?? represents the habitat
feature and category, has an integer value as described above. If preferences are initially
expressed using a 0 to 1 numerical scale, the later computational outcomes are the same.
2) The preference for a combination of depth (PZi), substrate (PSj) and cover (PCk) categories is
computed as the product of the constituent preferences, PZSCijk = PZi.PSj.PCk . For example, a
species with a high preference (3) for sand, a medium preference (2) for submerged
vegetation, in 0-1 metres of water in the spring (1) has a weight for that combination of 3
times 2 times 1 equals 6. Values for all cubes in the matrix can be assigned in the same way.
If any one category has a zero, the matrix combination value is zero.
3) The combined preferences are summed over all permutations of depth, substrate, and cover
categories to give:
PZSC*** = ΣΣΣ PZi.PSj.PCk.
4) Then proportional suitabilities are computed for each cell in the 3-way matrix of categories
as the ratio of the cell preference to the preference sum:
SZSCijk = PZSCijk/PZSC***.
This ensures that the contribution of each matrix for a particular species-life stage to
subsequent species and life stage grouping is equal to total weight of 1. SZSCijk represents the
proportional suitability of a matrix cell.
14
2.4.2 Group Habitat Suitabilities
Having computed separate suitability matrices for all the species included in the Location
list by life stage, those matrices are pooled by species and life stage grouping (Tables 2.4.1 and
2.4.2, respectively). [A combination of life stages, and trophic and thermal preferences have been
the basic approach adopted for forming groups throughout the development of the HSM
approach although other criteria for forming groups can be used. Whatever grouping criteria are
users will have to provide a rationale for their choices.]Matrices are summed cell by cell and
then the cell sums are standardized to a scale of 0 to 1 by dividing each by the maximum cell
sum. Thus the group suitability is:
GSZSCijk = (Σ SZSCijk ) / Max(Σ SZSCijk ).
The life stage composite suitability matrices are generated by applying the fish species
group weights and computing a weighted sum of the group suitability matrices. The overall
composite suitability matrix is then computed from a weighted sum of the life stage matrices
using the life stage weights.
2.4.3 Weighted Suitable Area (WSA)
Once the composite suitability matrices are prepared, the categorical characteristics of a
habitat patch can be used to compute a patch suitability. If the habitat patch is described by a
simple combination of the intrinsic habitat categories, e.g., depth - 1-2 metres, substrate - sand,
and no cover, the appropriate suitability can be looked up in the matrix and assigned. If a habitat
patch is described by a set of orthogonal vectors assigning proportions to depth, substrate, and
cover category vectors, e.g., depth - 0-1 metres (0.5), 1-2 metres (0.5) ; substrate – gravel
(0.25), sand (0.5), silt (0.25); and cover – nocover (0.5), submerged (0.5), the disaggregated
homogeneous habitat sub-areas (see section 2.4.2 for details) are assigned suitabilities, and then
pooled across the sub-areas.
When habitat suitability indices have been assigned to all habitat patches by fish group
and life stage criteria, weighted suitable area (WSA) is calculated by multiplying the patch area
by each of its corresponding suitability values. A single habitat patch may have a range of
suitability values attached to it, representing different fish species or groups and life stages.
Suitability for warmwater non-piscivores spawning may be high while suitability for coldwater
adult piscivores may be low.
A fundamental assumption of this type of suitability methodology is the area equivalence
of weighted suitable areas. For a given fish group- life stage criteria combination, 100 hectares
with a suitability of 0.25 is assumed to be equivalent to 50 hectares with a suitability of 0.5.
2.4.5 Summation
After the WSA for each habitat patch has been computed, the sums by group and life
stage criteria can be calculated for each scenario. The aggregate WSA values can be compiled
for each fish group and life stage along with the composite, aggregate values generated via the
weighted sums.
2.4.6 Cover Preferences
The mode of use of cover preference information was different than that for depth and
substrate. With depth and substrate categories, all species and life stages have some preference
recorded for at least one category of each habitat feature. Some species show no preference for
submergent or emergent vegetation but there is no literature basis for inferring preference for no
cover. A no cover category is needed to allow multi-factor suitabilities to be computed. To
overcome this difficulty, an implied no cover category is computed based on the strength of other
cover associations (Table 2.4.3).
15
2.5 Results
Users of the HSM method software are provided with a wide variety of options for
examining the results of an assessment. The results may be viewed as windows interactively or
printed in standard formats. As there are many layers to the calculations and many different ways
of viewing the analysis, the approach with the software has been to provide access to all steps.
Being able to examine results of intermediate steps in the calculations should help users
understand and interpret the outcomes.
When printing the results of any analysis, the user can choose between a standard report
or a custom selection of material drawn from the views generated in the interactive assessment.
The standard report provides a minimum set of results that are used to judge net gain or loss of
habitat productivity, documenting the various inputs provided and decisions made in setting up
the analysis.
Additional result formats give the user the means of appreciating the detail underlying
the more aggregated results. For example, the user may want to assess the contribution that
centrarchid YOY requirements made to the WSA results obtained for non-piscivorous
warmwater YOYs.
The user can assemble results based on habitat criteria, species (singly or in groups), or
area type as well as the integrated assessment. The choices are made from a results options menu
and then the results tables are generated for inspection.
The output may be grouped into a linked series of components: habitat criteria, species
and group summaries, model specification, habitat supply and transformation, integrated
assessment. Sample outputs have been generated using a pair of hypothetical scenarios for an
exposed site on the Lake Ontario shoreline (Table 2.5.1).
2.5.1 Habitat Criteria
There are several types of information documenting the habitat criteria: a) habitat types
and weights used, b) lake habitat classification, c) lake habitat classification for reporting, d)
lake habitat input-output mapping for output, and e) lake habitat input-output mapping for input.
• Habitat types and weights: This is a listing of the life-stage habitat types considered in the
analysis and the weights used when aggregating the weighted suitable areas (Table 2.5.2A).
• Lake habitat classification: This is a listing of the habitat features and classes present in the
habitat preferences database for all the habitat types included in the assessment (Table
2.5.2B).
• Lake habitat classification for reporting: As above, this is a listing of the habitat features and
categories considered in the analysis. These lists may be somewhat truncated compared to
those in habitat types and weights and lake habitat classification if certain categories are
considered unnecessary for a particular analysis.
• Lake habitat input-output mapping for output: The categories used to output results depend
on the form of the categories used in the scenarios. If proportional vectors of categories were
used, then those categories are used in the output. If vectors are not used in scenarios, the
categories in the requirements database are used by default but can be reduced by combining
them when the habitat types are specified in the definition phase. While two categories may
be combined for output purposes, the suitability calculations for the analysis are performed
for each category separately.
• Lake habitat input-output mapping for input: As the habitat feature categories used in the
development of scenarios for analysis using HSM do not always correspond to those present
in the habitat requirements database, it is necessary to map input categories onto database
16
categories. For instance, a substrate may be described as gravelly-cobble in a input scenario.
This substrate can be mapped to the database categories by assigning proportions of gravel
and cobble (Table 2.5.3). Those proportions have to be agreed among the experts and
assessors acting for the proponent and the regulatory agencies. This step provides the user
with some freedom with respect to the specification of habitat conditions in scenarios. If this
user-specification approach is adopted, those responsible on the regulatory side for assessing
analysis should make sure that the mappings are reasonable.
2.5.2 Species and Group Summaries
The species and group summary contains the group names, their weights, a count of the
number of species considered overall and by habitat type, and the list of species associated with
each group (Table 2.5.4).
2.5.3 Habitat Supply and Transformation
From the input scenarios, a series of habitat supply summaries are created (Table 2.5.5).
By the area types listed earlier (Table 2.3.2) such as LOSS, MODD, etc., the equivalent areas are
compiled for each category within each habitat feature. Each habitat feature is summarized
independently, ignoring the combinations of categories present in any particular habitat patch.
The categories used are drawn by default from the habitat preferences database if none are
specified. However, they are better thought of as being provided by the user since they can be:
a) changed from the defaults by renaming, adding to, or deleting, or b) fixed via the proportions
directive when the scenario data is in the form of vectors of proportions. Area values are reported
for all scenarios and for all habitat types, life stages are considered. Additional columns of
results are generated, measuring the difference between the first scenario and other scenarios in
that analysis. The net change columns provide a direct indication of where the greatest habitat
changes will occur.
2.5.4 Integrated Assessment
The final table of results in the standard report gives an integrated assessment of the
scenarios expressed as weighted suitable areas, areas times suitability values (Table 2.5.6). The
WSA values are given by habitat type and fish group along with the combined and overall
values. The overall sum of WSA represents the combination of all the constituents estimated
obtained after the group and habitat type weights are applied.
Given that the many steps in the process necessary to define an analysis can be justified,
the integrated assessment of WSA provides the simplest view of the outcome of different
development options. The WSA values are assumed to represent a relative measure of natural
fish productivity and a fair basis for comparison of pre- vs. post-development scenarios. The
results of one analysis cannot be directly compared with the results of another unless the same
assumptions are applied throughout. It is not practical to impose a fixed assessment regime for
all users to produce uniform results for all applications. Each ecosystem, its fish assemblage, and
fisheries has particular characteristics and hence flexibility in weights and criteria are necessary.
The common element across all applications is that consistent rules are applied to the
comparision of pre-/post-development scenarios.
Additional, more specific results tables can be generated at almost any stage in the
reporting and selectively printed.
2.6 An Overview of Habitat Suitability Index Values
As an illustration, representative habitat suitability index values were computed using the
complete Great Lakes basin fish habitat requirements database. All fish species were included
and they were divided into eighteen groups based on thermal preference (warm, cool, or cold),
17
trophic level (piscivore or not), and life stage (spawning, YOY, and adult). Then suitability
values were computed using the HSM method for all unique combinations of habitat classes for
depth, substrate, and vegetation. This process produced a matrix of eighteen suitability values for
each of 150 habitat combinations. As the suitability indices generally show less variability within
thermal groups, an arithmetic mean suitability was computed within each thermal group to
provide a general guide to expected suitability values (Table 2.6.1).
The habitat suitability indices computed using the HSM model vary as expected from the
source data in relation to depth, substrate, and vegetation cover (Figure 2.6.1). The charts were
prepared by computing mean values across the three life stages, adult, yoy, and spawning to
derive representative values. The charts for warmwater fishes show that higher suitabilities are
centred on the gravel-sand portion of the substrate spectrum in shallower depth increasing from
no cover through submergent cover to emergents (Figure 2.6.1A, B, and C).The charts for
coolwater fishes show a similar pattern with higher preference for greater depths and lower
peaks in submergent and emergent vegetation (Figure 2.6.1D, E, and F). The charts for coldwater
fishes show a markedly different pattern (Figure 2.6.1G, H, and I). Coldwater fishes show little
preference for submergent and emergent vegetation but the highest values occur again in shallow
depth over gravel and sand reflecting seasonal spawning and fry habitat preferences. Overall
coldwater fishes prefer no cover with a broad range of preference over substrates, narrowing to
rubble, gravel and sand in shallow waters.
A simple correlation analysis among the 18 habitat suitability indices was performed
using a tabulation of all unique combinations of depth, substrate, and vegetation categories to
show the level of redundancy built into the approach especially among warmwater and coolwater
groupings whose habitats are most often affected by nearshore development activities. The
correlation matrix shows (Table 2.6.2) that all correlations among warmwater and coolwater
indices are positive and significant at P=0.01, Bonferroni-corrected. There are few significant
correlations between coldwater and either warmwater or coolwater ones. Among the coldwater
indices, one third are positive and significant at P=0.01. The irregular pattern of correlation
among coldwater indices is consistent with their more varied life-history patterns of habitat use.
It should also be noted that coldwater species are more likely to have a life stage that is confined
to streams and rivers, rather than lakes.
3.0 INFORMATION GATHERING AND DATA ASSEMBLY
3.1 Information Gathering
Unlike the situation with stream fish habitat where there are a number of recommended
survey and assessment protocols for fish and their habitat (e.g. Stanfield et al. 1997), there are no
universally established protocols for lacustrine environments in Canada. Several agencies and
groups have prepared local guidelines and they should be used where available. There is a DFO
research program under way with researchers in several DFO regions, as well as from Ontario
Ministry of Natural Resources, engaged in studies to compare and test various survey methods in
freshwater with a view to recommending guidelines. The guidelines will have to be flexible to
allow the potentially enormous ranges and scales of site-specific conditions and development
activities. It is not practical to dictate a single standard for lacustrine assessment from the
vantage point of this tool as regulatory agencies have a range of factors and issues to consider
when deciding what types and amounts of information to require for any particular development
application.
Besides the actual details of survey design and execution, protocols should also address
how the data collected can or should be used to aid decision-making. Too often assessment
18
protocols show how to collect and summarize data but not how to apply it to real-world contexts.
The approach followed here is pragmatic and recognizes there may be several levels of
protocols. The highest level of protocol may only be used in concentrated research studies or
referral situations where unique, highly valued, highly sensitive fisheries resources are
threatened by the nature and/or scope of the proposed development activities. An intermediate
level might be used where a number of sites must be examined or where the extent of the area
under review is very large. The lowest level might rely on a relatively quick, one-time inspection
and description of the site with little actual sampling. The potential cost of applying different
levels of protocol will always be weighed against the cost-benefit economics of the proposed
development.
At this early stage in the evolution of complete survey and assessment guidelines for lake
habitat, it is more appropriate to try to identify some benchmark characteristics for data
gathering, allowing both proponent and regulator some leeway, then judging the adequacy of the
data presented in support of a referral.
3.1.1 Position, Dimensions, and Context
This section refers to information normally expected by the regulator when reviewing an
application and does not include any information specifically required for an HSM analysis.
• Site Position: It is essential that the exact location of the proposed development be known.
By convention, an established hierarchy of government and land-ownership boundaries have
been used to describe location. Now with the advent and widespread application of global
positioning systems (GPS), it is important that exact geographical co-ordinates be obtained
for a proposed development.
• Project Dimensions: The areal extent of the development activities as they affect fish habitat
should be known. Other on land areas should also be reported, especially where the landward
activities may indirectly impinge on adjacent fish habitat.
• Lake Dimensions: It is important to know something about the lake as a whole where a
development is proposed. This information becomes more important for smaller lakes or if
the project dimensions represent a larger portion of the lake. In larger lakes, it may be more
appropriate to know the size of the bay or arm of the lake where the development will take
place.
• Context: Knowing the above information for each referral becomes useful on an aggregate
basis as the cumulative impacts of multiple developments on particular lake systems. ( A
referral is the process by which proponents submit descriptions of planned developments
with details about the potential lost and modification of fish habitat to DFO and other
regulatory agencies seeking s determination under section 35 of the Fisheries Act with regard
to HADDs in fish habitat.) The overall levels of previous and anticipated development on a
lake should be known in general. However it is difficult to make use of such knowledge in a
particular referral unless there is an explicit habitat management plan for the whole lake with
clear habitat conservation targets. Rareness and uniqueness of habitats important to elements
of the fish community should factor into the overall decision-making process, even if they
cannot be explicitly incorporated in the application of HSM methods.
3.1.2 Physical Habitat Characteristics
• Depth: Bathymetric data should be gathered with a resolution of at least one metre vertically
as this is the finest level of resolution in the depth portion of the habitat requirements
database. Variations in water level should be factored into the analysis where appropriate.
The traditional high water mark may not be inundated very often. Hence, it usually makes
19
•
•
sense for proponent and regulator to agree on a reference datum, a norm, for the purposes of
establishing which areas are fish habitat and which are not in an application. If an area is
routinely inundated on a seasonal basis, it should be considered fish habitat especially as
some life stages may make corresponding seasonal use of the habitat. If variation in water
level is a significant issue, it may be appropriate to incorporate the variability by determining
the percentage of time when each elevation is inundated and using that as a Condition Index
in the evaluation of the fish habitat. However, a time budget can be misleading. For example,
if an area is inundated for 20 percent of the year but routinely used for spawning and YOY
rearing, the should be 100 not 20 percent. Attempts to derive a detailed, rigorous method for
handling water level variation are likely to lead to onerous requirements that accomplish little
in most instances.
Substrate: In ideal circumstances, it would be preferable that substrate always be recorded on
a uniform patch basis as a statement of percent composition using the particle size categories
previously described. Such an approach could engender substantial survey and analytical
costs. The resulting substrate data would likely have a level of precision far greater than that
present in the fish habitat requirements database. This is not usually justifiable. Therefore, it
is more reasonable that substrate areas should be delimited and estimated using visual
descriptive methods. Obviously due attention should be given to ensuring that an adequate
level of survey effort is directed to mapping substrate areas. A combination of visual survey
and/or echo-sounding, and point grab samples for inspection with good positioning
information, should allow a fairly accurate map to be generated. Often surveyors will record
the presence of ‘mud’ or ‘loon shit’ as substrate types. These may be interpreted in HSM as
being mixtures of silt and clay possibly with a minor contribution of sand. Man-made
substrates such as rocky revetments and rip-rap are usually represented as mixtures of
bedrock and boulders with small amounts of finer materials (cobble, rubble, gravel, etc.)
depending on the amounts of interstitial space and surface pockets capable of accumulating
and retaining them.
Cover: The literature-based review of cover habitat requirements in lakes mainly produced
information on use of submerged and emergent vegetation. However, in some instances,
coarser substrates laying over finer ones, e.g. boulders and cobbles laying over a mixture of
sand and gravel, can also provide some cover. With vegetated areas of habitat, the general
approach is to take larger areas with particular depth and substrate characteristics and assign
overall levels of vegetation cover, e.g., 20 percent submerged, 40 emergent and 40 with no
cover, rather than generate numerous small areas with or without 100 percent cover. Given
the method by which the habitat characteristics in percentage vectors are analyzed, the
greater detail is unnecessary. In this version, information on other types of cover, such as
larger detritus, woody debris, and overhanging vegetation, was rarely included. For nonvegetation types of cover, the approach has been to allow some leeway in interpreting them
as having a percentage vegetation cover equivalence, weighed more heavily on submerged
vegetation. Thus in a shallow shore area with 25 percent coarse woody debris cover, 15
percent submerged cover might be used in the HSM scenario. The lesser percentage takes
some account of the lesser surface:volume ratios in woody debris and the space it occupies.
Care must be taken not to abuse this stretching of cover definitions. [A new literature review
is being undertaken to try to expand the range of cover types considered in the Great Lakes
lacustrine habitat requirements database (Vokey-Cudmore and Minns, DFO Unpublished
information).]
20
•
Shoreline Exposure on the Great Lakes and other large inland lakes: As a result of
cumulative experience with applications of the HSM method in the Great Lakes region, a
situation has been recognized requiring a use of the Condition Index. On highly exposed
shores, developers often propose to build breakwalls to accommodate marina and other
recreational developments. In the lee of the breakwall, the wave exposure of habitats will be
reduced. In the sheltered area indirectly affected by the breakwall, the depth, substrate and
cover characteristics may be unaltered, especially as marina developers strive to balance
shelter from waves with sufficient circulation to avoid the creation of stagnant areas. It is
known from experience that fish abundance will increase in these sheltered areas although
the composition of this fish assemblage may differ from that found in adjacent naturally
sheltered, high productivity habitats (Randall and Minns 2000 in review). An analysis of an
existing littoral zone fish-habitat database showed that a composite index of fish assemblage
species richness, number and biomass per unit effort, and the Index of Biological Integrity
(IBI) was higher on average in areas with a maximum effective fetch of 10 km than in areas
with a greater fetch value, e.g. 1.0 versus 0.64 (see Appendix A for details of the derivation).
This result provides a basis for using the Condition Index to reflect this outcome. In general
the rule would be: If maximum effective fetch > 10 km set Condition Index (CI) = 0.64,
otherwise CI = 1.0 (Table A3 provides a breakdown of Condition Index assignments). In
practice this would mean that areas indirectly modified (MODI) in the lee of the breakwall or
directly modified areas (MODD or COMM) on the leeward side of the breakwall would be
assigned a CI value of 0.64 in the PRE-scenario and a value of 1.00 in the POST-scenario.
LOSS areas in the PRE- scenarios would be given a value of 1.00 as they must charged at the
maximum productivity. MODD areas on the windward side of the breakwall will not
experience any change in their exposure status and thus the 0.64 value would be used in
PRE- and POST-scenarios. COMC areas are presumed to provide good quality habitat and
are assigned a Condition Index value of 1.0 in POST-scenarios. * It should be noted that this
alteration is provided as an interim adjustment. The research necessary to distinguish
between increased productivity in such habitats from a purely attractant-concentration effect
has not been done. In the comparable situation of reefs, marine and freshwater, longstanding research has failed to distinguish between the two possibilities though a rising
consensus suggests that the attraction effect may be greater, thereby indicating that no
productivity increase can be inferred in such situations.*
3.1.3 Other Habitat Conditions
• Water Quality: Where the development will have no impact on prevailing levels of common
water quality parameters such as nutrients, chlorophyll, temperature, conductivity, and
turbidity, and where the overall state of water quality is judged to be satisfactory, no special
consideration is needed. If activities associated with the development will lead to changes,
positive or negative, in overall water quality, there may be a need to explicitly factor the
changes into the HSM assessment. For example, if, as part of a shoreline project, stormwater
flows would be managed to reduce sediment loads and hence local turbidity levels, it might
be appropriate to make allowance for any deleterious levels at present and the improvement
afterwards. This could be accomplished by setting the Condition Index to less than 1.0 in the
pre-development scenario and to 1.0 in the post-development scenario. This type of
productivity adjustments must be used with care and do not apply to LOSS areas. One
situation to clearly avoid are unreasonable claims that water quality conditions are extremely
poor now and will ‘miraculously’ improve after the development occurs. It is also worth
21
remembering that applying Condition Indices to modified and compensation areas in pre- and
post-scenarios may increase the compensation area required to offset losses which are
assigned maximum unit productivity values. Uses of the condition index will be closely
scrutinized by regulators as there is potential for misuse and deliberate biasing of
assessments.
• Contaminants/Toxic Chemicals/Impaired Fish Health: The presence of actionable levels of
contaminants in fish flesh or fish food or of toxic chemicals in water or sediment, or the
presence of higher levels of chemical-induced health impairments in fish and the biota they
depend on, point to potential impairments of fish productivity and/or fishery products. The
presence and potential impact of these factors can be incorporated into a HSM assessment.
This can be achieved using the Condition Index. This should definitely be done when the
proposed development will change the chemical situation. In situations where the proponent
is not accountable for the chemically-induced phenomenon and the activities won’t change
them, their inclusion is probably unnecessary. In some Great Lakes’ Areas of Concern,
contaminant clean-ups and fish habitat improvement have been applied simultaneously to a
site. In such instances, the proponent should be allowed the benefit of the improved
usefulness of the fish habitat productivity in the HSM analysis.
3.1.4 Fish Community and Fisheries
• Prior Research and Assessment: Every effort should be made to assemble prior fish
community data collected at or near the site as well as information about the lake or lake
section as a whole. In some instances, there may be specific evidence of the uniqueness of
the site as fish habitat that can be weighed into the decision process.
• Fishery Exploitation and Management: Evidence of the size and importance of local fisheries
provide valuable supplementary information that can be used as a guide in the HSM
definition phase. This information aids with the species list for the location and the formation
and weighting of the fish groups.
• New Fish Assessments: Unless there are sufficient time and resources to conduct a thorough
assessment of the fish community, on-site, locally, and possibly regionally, using
standardized, benchmarked methods appropriate to the lacustrine habitats present, it is
unlikely that new data derived from limited time, effort, and gear, can provide major new
insights. Unlike in many regions in Canada and elsewhere where there are well-studied
stream and river habitats, there are insufficient reference or benchmark fish community data
gathered with standardized methods in lacustrine habitats to allow clear interpretation of new
observations. Indeed, it is more likely that small amounts of incomplete data can confuse and
obscure the main habitat management issues. The habitat policy is concerned with
conservation, restoration, and creation of maximum potential habitat productivity for fish.
Whether or not a few specimens of one or other species is caught at the site on one occasion
is unlikely to reveal much about its productivity value. Clearly there are exceptions to these
generalizations. If the sampling reveals a rare spawning location or the presence of a new or
rare species, the evidence must be weighed into the assessment process, although it is
unlikely to directly impact the HSM part of the assessment.
3.2 Data Assembly for an Application
There are several steps involved in the assembly of the input data before an application
using HSM can be completed. The sequence of steps including analysis and assessment is
outlined in a framework (Figure 3.2.1). The numbered steps cover the main sequence from a user
perspective. The documentation should consist of numerical data, maps, and supporting text
22
rationales. Habitat characteristics and ranges should be developed with due attention to the
classes used in the habitat requirements database. Development activities affecting habitat and
compensation activities should be assessed in terms of their appropriateness, feasibility, and
sustainability. It is vital that users thoroughly document their evidence and decision processes as
they assemble their data files and definition criteria.
3.2.1 Physical Habitat Assessment Scenarios
• Development Activity Footprint: The first step is to define the total development footprint as
it affects fish habitat. This footprint area is divided into the three area types: LOSS, MODD,
and MODI. The first two should represent the direct construction effects while the third may
arise as a result of new structures altering current, flow, sediment transport, and other
conditions in adjacent areas (Figure 3.2.2). Determining MODI might involve some
modelling or simulation (item 1b in Figure 3.2.1).
• Existing Habitat: Once the affected habitat areas have been identified, the existing habitat
characteristics including depth contours, substrate and cover types should be mapped. Next
delimit separate patches of existing habitat with relatively uniform characteristics. Note
should also be made of other factors such as water uses and water quality impinging on the
habitat, and of adjacent terrestrial habitat conditions.
• Intersect Footprint and Habitat Areas: Using the footprint and existing habitat maps, intersect
the areas to define a set of areas with defined area type and habitat characteristics. This
provides the basis for the pre-development scenario.
• Post-development Habitat: In the modified areas, estimate what the habitat characteristics
will be after the development has taken place and stabilized. This may require some further
sub-division of areas to capture various transitions of depth, substrate, etc. This process
should be well documented.
• Pre- and Post-development Scenarios: Now both scenario files can be prepared. Ideally, nonLOSS areas should be uniquely identifiable in both PRE- and POST-scenarios files to make
comparisons of on-site habitat changes easier to track and assess. LOSS areas only appear in
PRE-scenarios.
• Compensation Options: Compensation areas should be defined separately, but may either be
included in the initial pair of scenarios or introduced into a second pair of scenarios to help
highlight the benefits of the compensation measures.
3.2.2 Cross-link Scenario and Database Habitat Classes
When setting up an application of HSM, there are two options for specifying the linkage
between the habitat classes used to define the scenarios and the habitat classes used in the habitat
requirements database used by HSM to estimate suitabilities.
In the first option, depth, substrate, and cover characteristics of habitat areas are given an
account using vectors describing the percentage composition of all the classes within each
characteristic. For example, a substrate characteristic for one area may be described as 25 percent
gravel, 50 percent sand, 15 percent silt and 10 percent clay. As all substrate classes (bedrock,
boulder, rubble, cobble, gravel, sand, silt, clay, hardpan clay, and pelagic) may be reported, the
substrate vector will be entered as “0,0,0,0,25,50,15,10,0,0”. If this approach is taken, the
scenario and area reporting classes are automatically linked in the analysis phase.
In the second option, the user may create their own labels for habitat classes, e.g. shallow,
deep, gravelly-sand, bedrock-boulder, armour1, armour2, veggy, sub25emer10, etc. Then during
the creation of the application definition, the user may have to provide the input necessary to link
scenario and area reporting classes. Since different life stages of fish have slightly different set of
23
habitat classes, this linkage information is required for all life stages and habitat characteristics.
For instance, gravelly-sand may be mapped into the database as 25 percent gravel and 75 percent
sand, or armour1 as 50 percent bedrock, 30 percent boulder and 20 percent rubble, or shallow as
60 percent 0 to 1 metres and 40 percent 1 to 2 metres.
In either option, it is important that both the users, project proponents and regulatory
assessors, pay attention to the choices made to characterize habitats as the choice can greatly
influence the suitability values assigned to different habitat areas. Before attempting to use the
software to create a definition and run an analysis, the user should prepare the class mapping
information and gather relevant supporting evidence and rationale.
3.2.3 Fish Community, Ecosystem Type, and Fishery Objectives
Documentation and evidence of the fish community, ecosystem characteristics, and
fishery objectives for the lake, or region of the lake should be gathered as a backdrop to the
specification of species to be considered and the various weights required. This step will often
involve consultation with other agencies and individuals, as well as local and regional planning
documents.
The fish community list should be complete. In the case of smaller inland lakes, tertiary
watershed scale list are preferable to incomplete lists for specific locations. In the case of large
lakes, the entire pecies list should suffice. Longer, diverse lists will tend to produce more stable,
robust suitability values while short lists could produce misleading results. Given the potential
for species turnover, the assessment should be weighed toward the long-term mean potential of
any lake to support fish. Besides, the Fisheries Act mandate embraces all fish as potential
contributors to fisheries productivity.
A consideration of the lake ecosystem characteristics should have some bearing on the
refinement of the species list. For example if a lake is too shallow to support a hypolimnion
during the summer, there is little point including lake trout or related species in the analysis.
Equally, if the lake is in an area not included in the distribution of a species, that species should
be excluded. As more evidence emerges on the general patterns evident in the occurrence of fish
species in lakes, it should be possible to provide refined guidance of the assembly of a fish
species list. [Work is underway to create a reference library of fish species occurrences by
tertiary watershed for all of Canada as well as for larger waterbodies.]
The fishery management objectives, as defined by the appropriate management agency,
should be taken as a starting point for identifying preferences and weights among the fish
community. Considerations regarding rarer species, trophic structure, and the potential
productivity contribution of various species should be considered. Often fishery management
plans are directed at exploitation with secondary consideration of community structure and
sustainability in the longer-term. Generally the fish community objectives for the site should
preferentially take a longer-term, ecosystem view rather than a narrowly defined fish exploitation
view.
3.2.4 Location Spp., Fish Group and Life Stage Weights
To complete the definition phase of the HSM application, the user must select a species
list, assemble the species into groups, and assign weights for fish groups and life stages. The
location species list is based on information gathered in the previous step.
Fish groups can be assembled using thermal preferences (warm, cool, cold) and trophic
position (piscivore vs. non-piscivore). This divides the potential species list into six groups. This
set of groupings is recommended as a default choice. Other criteria might be used to group the
24
fish species but clearly ecological criteria are preferable to others that might be based on
potential use of the fish or their origin.
The fish group weights are set to be equal as the default option. The weights can be reset
to recognize the fishery and/or ecosystem attributes shaping fish productivity in the lake. In large
lakes, the weights may be directed to local priorities rather than global ones. For instance, in
Lake Ontario, the offshore, coldwater pelagic fishery is the main one but inshore, given past
habitat losses and alterations and a desire to restore stream-nearshore-open lake linkages for fish
communities, weights showing preference for warmwater and coolwater species may be more
appropriate.
In selecting the fish group weights, it should be recognized that small weight differences
will not make much difference to the final analysis results. Analyses of the effect of various
increments was reported in Minns et al 1999 (SEVERN SOUND RPT). Generally increments of
10 percent or more are needed to induce significant outcome changes. Since non-piscivorous
species outnumber piscivorous ones, the fish group weighting process allows the opportunity to
place greater emphasis on piscivores which obviously play a greater role in exploitation of
fishery resources.
The life stage weights are set independently of the fish group weights and apply equally
to all fish groups on the assumption that all life stages are important to all species if viable stocks
are to be sustained. The default option assumes that all three life stages are equally important.
There are various opinions and limited definitive evidence to guide the choice of weights here. It
is worth noting though that HSM results are less sensitive to differences among life stage
weights than they are to thermal preference and trophic level weights for fish groups (Minns et al
1999).
3.2.5 Assessment and Analysis Steps
Once all the data for scenarios and species lists have been assembled and weights
assigned, the analysis can proceed. The first results inspected are the overall ones showing
weighted suitable areas for each scenario with a breakdown by fish group and life stage. The
estimates obtained here will largely determine the next steps in the referral process. There are
always winners and losers when WSA is examined across the fish groups. Multiple scenario
pairs can be examined simultaneously as alternate options are evaluated and preferred options
refined. The second results inspected are usually the area summaries where the overall changes
in habitat are summarized. Many constituent results tables are available for inspection where the
user seeks clarification of the overall results.
Once the iterative analysis of scenario pairs is complete and a preferred option identified,
the detailed report of the analysis and all key assumptions can be printed and added to the other
referral documentation, detailing the proposed development activities, current and expected
habitat conditions, and mitigation and compensation measures.
4.0 CASE STUDIES
Both case studies provided here originate from Halton Region on Lake Ontario. They are
described with more contextual detail in Minns and Nairn (1999). Greater emphasis has been
placed on the implementation of the HSM computations in this presentation of the case studies.
The first case study comes from Burlington where the municipality is engaged in a long-term
program to develop it shoreline areas as a recreational resources for local citizens and tourists.
The second case study involves a private property owner in Oakville whose shore protection
structures were in urgent need of replacement or restoration. HSM was applied in both cases with
the assistance of staff in the Fish Habitat Management (FHM), Ontario, group with Fisheries and
25
Oceans. In both cases, the interaction enjoyed the support and cooperation of the consultants
working for the proponents, local municipal planning offices, and the Halton Regional
Conservation Authority who have responsibility for managing the coastal zone.
4.1 Case 1: Brant Inn Node, Burlington, Ontario
4.1.1 Background
The Brant Inn Node waterfront occupies about 300 m of frontage east to west from a
formal park in downtown Burlington to the Burlington Beach park. Historically, this shoreline
featured a transition from the sandy bay mouth barrier beach to the eroding shale bluffs to the
east. Over the last century this shoreline was significantly altered, initially when the opening to
Burlington Bay was shifted from near the site to the present location, a canal further to the south,
and then by ongoing lake-filling to support a rail line, and later to protect the land base of the old
Brant Inn. Prior to this project, the frontage was protected by concrete sea-walls, armour stone,
and concrete rubble in varying states of disrepair, illustrated in Figure 4.1.1A. The near shore
area at the site consisted of two separate zones with exposed shale bedrock in the small
embayment and further to the east, sandy substrate from the old Brant Inn Pier towards the south
and Burlington Beach. Due to it’s position in the north-west corner of the lake, the only wave
exposure is towards the east.
The City of Burlington, and its partners (the Halton Region Conservation Authority and
the Region of Halton), plan to develop the Brant Inn Node to create a highly animated point of
interest at the meeting place of two distinctly different park environments. The aim was that the
new development would meet a range of recreational and aesthetic objectives. W.F. Baird &
Associates were retained to develop a series of concepts to regenerate the shoreline and
waterfront area, to undertake a process to select a preferred alternative and to complete final
design for this option.
Seven potential options were developed that ranged from a minimal footprint approach
where the existing shoreline position was maintained to options involving significant amounts of
lake-filling. Based on the vision statement and a series of objectives outlined by City staff, a
group of evaluation criteria were prepared to assess the various options in order to select a
preferred option. Through the public consultation process and through input from City staff, the
evaluation criteria were ranked in order of importance as listed below (from most to least
important): 1- Public access; 2 - Visual impact, 3 - Water quality, 4 - Aquatic habitat, 5 - Open
space, 6 - Terrestrial habitat, 7 - Year-round use, 8 - Impact on surrounding land use, 9 - Heritage
resources, 10 - Costs, 11 - Day use boating, 12 - Potential for construction phasing, and 13 Recreational fishing.
4.1.2 Process
The City of Burlington along with its partners approached Fisheries and Oceans FHM to
obtain guidance on how to proceed from the initial set of seven options, including the option of
doing nothing (option 0 on Figure 4.1.2A). FHM suggested that the proponents use the HSM
approach both to screen the initial redevelopment options and then, once the option set had been
narrowed, to refine and detail a preferred option showing a net gain of productivity of fish
habitat whilst meeting the City's objectives.
Based on consultations with Ontario Ministry of Natural Resources and DFO staff, the
fish community objectives for the HSM assessment were weighed equally among the three
thermal and two piscivore/non-piscivore groupings of the Lake Ontario fish community. The
main lake fishery is dominated by stocked coldwater species. Historical development along the
Ontario shoreline in the western portion of the lake removed a lot of littoral habitat features that
26
supported warmwater and coolwater species and helped link riverine and open lake habitats. The
overall goal is to restore a more balanced mix of habitat use and productivity in the littoral
regions of the lake. As experts were sure that coldwater habitat features were abundant in the
lake, the proponents were encouraged to include habitat features enhancing productivity for
warmwater and coolwater species. Wetland habitat creation was also encouraged, albeit on a
small scale, as historical coastal wetland losses on Lake Ontario have been very extensive.
4.1.3 Application
A consultant team, consisting of coastal engineers (W.F. Baird & Associates),
environmental scientists (Gartner Lee) and landscape architects (Fundamental Design) proceeded
to rate each of the six options against the City’s thirteen evaluation criteria. Meanwhile, an
application of the HSM approach proceeded to determine the impact in terms of the net change
in Weighted Suitable Area (WSA) between pre- and post-development scenarios. Only two of
the seven options came close to achieving a net gain and minimizing the area affected Figure
4.1.2B). Several refinements of the options showing a net gain were evaluated until a final option
meeting Fisheries and Oceans’ requirements and the City’s objectives was obtained. The
preferred option, which was ranked highest regarding aquatic habitat (highest WSA gain),
features the following components: a) A wetland (0.10 hectares in area) excavated from the
existing land base, connected to the lake, and protected by a revetment and a shoal in the
entrance; b) The reconstruction of the Old Brant Inn Pier to create an area for public gatherings
and day use boat mooring; c) Lake-filling of the small embayment (0.27 hectares in area) to
create more land base next to the main access point for pedestrians and vehicles; and d) A cobble
beach between two rubble-mound headlands (Figure 4.1.1B).
The HSM assessment of the preferred option consisted of defining areas lost and areas
directly modified and was used to refine the design to maximize the WSA gain. To define the
scenarios pairs (pre- and post-) for each option, the areas bound by defined homogeneous
features with respect to depth ranges, substrate types, and cover characteristics were estimated
from design plans. Areas were also sorted to their pre- versus post-status according to the scheme
described in Table 2.3.2. The areas were defined by W.F. Baird & Associates for the design
team and the HSM model was run by DFO staff. In simple terms, the lake-fill area was a loss,
consisting of exposed bedrock (with low habitat quality), but was more than balanced by creating
a smaller area of higher quality wetland habitat in addition to improved habitat associated with
the cobble beach and the rubble-mound structures. Using the new algorithms relating the
survivorship of macrophytes to threshold wave conditions (W.F. Baird & Associates 1996), the
width of the opening to the wetland was maximized and the crest elevation of the revetment was
minimized.
4.1.4 Results
Meeting the park and recreational objectives of the City of Burlington required loss of a
large area of habitat, 2690 m2, and modification of a further 2203 m2 (Appendix B). As the
shoreline is subject to substantial wave energy, a large area had to be armoured. The pre-scenario
shore was dominated by bedrock with scattered areas overlain with sand. The modification of a
large section of shoreline to generate a cobble beach and creation onshore of a small wetland
(1000 m2) with both emergent and submergent vegetation provide the basis for offsetting the
direct losses.
In the suitability and weighted suitable area (WSA) calculations equal weights were used
for six groups of fish based on thermal preference and piscivore/non-piscivore combinations for
the three life stages. The overall WSA results indicated a shift from 598.2 to 744.4 equivalent
27
area units (actual areas are scaled by dimensionless suitability values with a range 0-1; an
equivalent area unit is the area the WSA would represent if the suitability were 1), an increase of
146.2 units or 24.4 percent (Table 4.1.2). The WSA changes were not evenly distributed among
the life stages. The YOY WSA shows a loss of 76.0 units or 12.5 percent and was due mainly to
the reduction of habitat with sand substrate. The adult and spawning WSA showed large gains.
Among the fish groups, the coolwater non-piscivores and both coldwater groups show losses of
both adult and YOY WSA (Table 4.1.2), ranging from 19.8 to 728.2 units. Coldwater piscivore
habitat shows the greatest percentage losses. Warmwater groups and coolwater piscivores show
WSA gains for all three life stages.
The Brant Inn Node results demonstrated the benefits to be gained via the creation of
habitats like a wetland, which have very high suitability values. Many species make use of
emergent and submergent vegetation during one or more life stages (Lane et al. 1996 a,b,c). The
loss of large sandy habitats and their replacement in modified areas with cobble and armourstone would by itself have produced net losses as more species make use of fine-textured
substrates than coarse ones. The overall net gain was achievable only with the creation of a small
wetland by excavation onshore.
4.1.5 Outcome
The final design of the preferred concept was submitted for review under the Canadian
Environmental Assessment Act (CEAA) at the beginning of August 1997. CEAA requirements
have no direct link with fish habitat management. Only fish habitat managers, e.g. DFO, are
concerned with fish habitat but the decision to authorize a HADD triggers CEAA under federal
legislation. With the benefit of having already completed a HSM assessment indicating a gain in
WSA, an Authorization was provided under Fisheries Act later in October of 1997 and approvals
were also received in October from Coast Guard, Halton Region Conservation Authority and the
Ontario Ministry of Natural Resources. Construction on the project commenced in November
1997.
4.1.6 Commentary
This case study illustrated the three stages in the progress from assessment to
authorization when the HSM analysis played a role. First, it helped narrow the range of initial
options by constraining the set to those showing a net gain or a potential, through modification,
to attain a net gain of habitat productivity. Second, it helped with the iterative process of refining
the preferred option as the planners and their consultants sought to maximize the attainment of
their park development objectives while retaining the ability to meet the no net loss goal when
requesting an authorization. Third, the completion of analysis and documentation of the final,
preferred option meant that Fisheries and Oceans regulatory staff had a very good appreciation of
the fish habitat implications of the proposed development activity when issuing the
authorization.
4.2 Case 2: Wallik Property, Oakville, Ontario
4.2.1 Background
The owner of a property in Oakville with 29 m of frontage on Lake Ontario required
restoration of the shoreline protection. The shoreline consisted mostly of an eroding shale bluff
6 m in height. The Shoreline Policy of the Province of Ontario requires that erosion hazards be
addressed through shoreline stabilization (where necessary) prior to the municipal approval of
site plans under the Planning Act.
Historically, this shoreline would have consisted of an eroding shale bluff with shingle
beaches at the toe of the bluffs supplied by the natural process of bluff and lake bed erosion.
28
During this century, more than 75% of the Halton shoreline of Lake Ontario was protected with
some form of sea-wall or revetment (personal communication, Teresa Labuda, Halton Region
Conservation Authority). Therefore, much of the natural source of shingle supply was
eliminated and any beach or near shore deposits of shingle were largely eliminated (resulting in
narrow or non-existent beaches and a lake bed swept clean of loose shingle deposits).
Prior to the development of design alternatives, the owner of the property indicated his
objectives for shoreline treatment: a) Improve access to the water’s edge from the top of the
bluff; b) Encourage beach development for access to the water’s edge; c) Provide cost effective
long term shoreline stabilization; d) Protect the crest of the bluff with bioengineering techniques,
where possible, rather than with a large coastal structure; e) Minimize the crest elevation and
visual impact of the shoreline protection; f) To the extent possible, naturalize the shoreline; and
g) Minimize the potential for future maintenance of the shoreline treatment.
The coastal engineers, W.F. Baird & Associates, developed five conceptual designs for
their client: (1) Do nothing and monitor the future erosion; (2) Construct offshore breakwaters
and place imported beach material; (3) Construct a rubble-mound headland and allow beach to
develop from natural supply; (4) Construct a rubble-mound scour pad at the toe of the bluff; and
(5) Construct an armour-stone wall. A qualitative assessment of the options was made based on
the application of the client’s objectives and the nine principles for an ecosystem approach to
shoreline treatment promoted by the Waterfront Regeneration Trust (see WRT 1996): clean,
green, usable, diverse, open, accessible, connected, affordable and attractive. The construction
of offshore breakwaters and the placement of imported beach material was considered the only
option meeting all of the client’s objectives and most of the nine principles.
4.2.2 Process
Initially, the innovative, ecosystem approach to shoreline treatment selected by the client
encountered significant resistance from the approving agencies. There were concerns about the
impact to fish habitat under the Fisheries Act administered by the Ontario Ministry of Natural
Resources (OMNR) and the federal Department of Fisheries and Oceans (DFO), issues about
infringement on the lake bed under the Public Lands administered by OMNR and concerns
regarding navigation under the Navigable Waters Protection Act administered by the Coast
Guard. Recognizing the importance of the precedent in promoting an innovative approach to
shoreline treatment for this heavily developed section of shoreline, the Halton Region
Conservation Authority (HRCA) brokered a meeting on 5 June 1997 between representatives of
all the concerned agencies, the client and his designers. The catalyst for a consensus to approve
the innovative shore treatment approach in an expedient manner was the application of the HSM
approach to assessing fish habitat.
4.2.3 Application
The basis for the application of the HSM approach was the comparison of the pre- and
post-conditions illustrated in Figures 4.2.1A and 4.2.1B. The preferred alternative would alter
several physical characteristics of the existing habitat conditions along the Wallik shoreline. The
offshore breakwaters would be constructed directly on the exposed shale lake bed. This would
result in some lost lake bed area as well as modification of exposed shale to armour stone with
interstitial spaces. Around the breakwaters, 8 to 20 cm cobbles would be placed over the
existing exposed lake bed. Therefore, the impacts consisted of areas lost and directly modified
areas, for this small project the indirect impacts were assumed to be insignificant. W.F. Baird &
Associates defined the pre- and post-project physical conditions by areas with defined habitat
29
characteristics and DFO staff applied the HSM software to define WSA for the pre- and postdevelopment scenarios.
4.2.4 Results
To meet the shore protection objectives of both the property owner and Halton Region
Conservation Authority in a highly exposed location, there was no scope for creating sheltered
fine texture substrates with cover. The area losses were kept to a minimum (77 m2) and the main
modifications produced cobble beach overtop the bedrock retained with appropriately placed
boulders (250 m2) (Appendix B).
Replacing bedrock with cobble and some boulder substrates generated the increased
WSA, rising from 7.6 to 37.4, an increase of 392.1 percent (Table 4.2.1). Among the life stages,
the largest WSA gain was for spawning while the largest percentage gain was for adult. In the
pre-development scenario, the WSA was zero for 9 fish group by life stage combinations. The
greatest WSA gains were obtained for spawning coldwater non-piscivores (77.5 units) and adult
coolwater non-piscivores (77.3 units). The only losses were spawning warmwater piscivores (6.4 units) and YOY coolwater non-piscivores (-0.1 units).
These results for the Oakville site are easily interpretable. Bedrock is a substrate favoured
by few species (Lane et al. 1996 a,b,c). As a results, relatively low suitabilities would be applied
to bedrock and replacement with other substrates, even relatively coarse ones such as boulder
and cobble would result in net gains.
4.2.5 Outcome
The preferred alternative received an Authorization for Works or Undertakings Affecting
Fish Habitat from DFO and an authorization to proceed by Coast Guard both on 6 October 1997.
HRCA provided authorization to proceed also on behalf of OMNR and the Public Lands Act in a
letter dated 24 October 1996. Construction commenced in November 1997.
4.2.6 Commentary
Prior to the intervention of the HSM application, the parties had tried to reach a
consensus without success. The results of the analysis of the preferred option provided all parties
with a fuller basis for discussion and shared understanding.
5.0 DISCUSSION
5.1 Conceptual Foundations
There are four main foundation elements in the Habitat Suitability Matrix (HSM)
approach: a) the use of comparative assessments of habitat preferences by species and life stages
to derive numerical habitat suitability index models, b) the balanced assessment of habitat
productivity losses and gains to meet the Policy’s guiding principle of ‘no net loss’, c) the use of
weights to integrate the differing requirements of various elements of the fish community, and d)
the use of a more pragmatic perspective to the application of this numerical assessment tool in
the workaday environment of fish habitat management as compared to the more rigorous work of
scientific research.
5.1.1 Habitat Suitability Index Models
The habitat suitability indices used here as surrogates for fish habitat productivity are
based on simple measures of preference and predicated on the key assumption that use of habitat
is an indication of that habitat’s contribution to fish productivity. The use of simple preference
measures is dictated by the lack of detailed quantitative assessments of the importance of various
habitat for most species and life stages. It is argued that the accumulation of this preference data
across numbers of species and life stages leads to more reliable, aggregate indicators of habitat
productivity. This approach is consistent with the general tendency present in much of
30
conservation biology to infer that sites supporting greater numbers and diversities of various taxa
have greater significance to the persistence of that biodiversity than sites supporting lesser
amounts.
The simple species-life stage suitability models used here can be viewed as an
intermediate point on the gradient between no knowledge and understanding of habitat
requirements and complete knowledge and understanding; the former being a closer
representation of the current situation for many species and the later being an ideal, unattainable
situation. In the minimum model where it is assumed that nothing is known about habitat
requirements, the null hypothesis that all habitat classes are equally preferred, and hence have
equal suitability, applies under the rule of the Precautionary Principle. In this case, application of
HSM leads logically to the outcome that all losses have to be matched by the creation of new
compensatory habitats since all suitabilities are the same and no differential productivity changes
are possible. Increments in the amount known of the differences in habitat preferences leads to
differences in habitat suitabilities and hence increases the flexibility in attaining no net loss.
Much of the basic framework for the development of habitat suitability index models and
their application was laid down during the establishment of the Habitat Evaluation Procedures
(HEP) in the late 1970s and early 1980s (USFWS 1981). Terrell et al. (1982) described the
specific information needs for lacustrine and riverine HSI modelling for HEP. Most HSI models
were developed as consensus models assembled according to groups of species via workshops of
ecosystem experts aided by facilitators and modellers. In HSM, systematic literature review has
been used to replace the consensus workshop modelling. In aquatic applications of HEP,
especially lacustrine ones, whole systems were the unit of assessment and hence many suitability
models were assembled using whole lake characteristics such as area, mean depth, and water
quality measures, as predictors for fish performance. In HSM, the assessments are focused on
sub-areas in lakes and hence the fact that most of the HSI models could not be transposed or
reused though the general lake level information is highly relevant to establishing the overall
context.
5.1.2 Balancing Losses and Gains
The approach to fish habitat management embodied in both the Fisheries Act and the
Policy represents a Kantian rather than a utilitarian (Sagoff 1996) framework for environmental
management. The Fisheries Act takes a rule-based approach based on non-monetary
considerations of the necessity of fish habitat and its productivity for fish to persist and thrive,
and hence sustainably support fisheries. The methodology assembled in the HSM method is
based on that approach and framework. Only fish habitat is considered in the analyses.
In many respects, HSM represents a constrained non-monetary form of cost-benefit
analysis (CBA). CBA in a general context clearly represents a utilitarian approach wherein all
elements are assigned a monetary value. The replacement of goods and services with money
renders all things interchangeable. This utilitarian approach is not the course followed in the
Fisheries Act and the Policy. HSM uses weighted suitable areas as a surrogate currency for fish
habitat productivity and within the confines of the analysis the monetary costs of alternate habitat
scenarios are not considered. Indeed there is no mandate within the Fisheries Act provisions or
the Policy for considering monetary issues. However, the HSM does reflect many of the
principles identified by Griffin (1998) as being fundamental to CBA. Where a principle is not
matched, the future potential for making it match is identified. Griffin laid out eight principles
and for most of them a non-monetary equivalent is presented in HSM (Table 5.1.1) and
expanded upon as follows:
31
•
Benefits exceed costs – a non-negative outcome from the weighted suitable area analysis is
necessary before a proposal is deemed acceptable according to HSM.
• Differences – the approach requires a balanced comparison of PRE- and POST-scenarios
with the welfare of the fish being the only consideration.
• Measurement units – unlike in CBA where costs and benefits are valued with mixes of
production and consumption costs, fish habitat productivity values are used consistently
throughout.
• Producer and consumer net benefits – the net productivity changes treat fish as producer and
consumer with human intervention as an externality. Unlike CBA, supply and demand
factors with respect to fish habitat productivity are not currently considered in HSM.
However, the need to develop ways of considering such factors as reflected in density
dependence mechanisms is recognized here and in Minns et al. (1996). The current approach
is consistent with conservative, precautionary methodology.
• Zero-sum transfers – If loss of one area of fish habitat appears to be exactly matched by
creation close by of an equivalent area, both areas are considered in any analysis as no
change in habitat induced by a project and can be ignored. However, inclusion of habitat
areas unaffected by the project is not recommended as the productivity values associated
with them may distort the interpretation of the analysis.
• Temporal discounting – Temporal discounting of habitat productivity changes is not used in
the current version of HSM. Pre- and post-development scenarios are assessed as if they will
persist forever. There are circumstances when discounting might be required such as when
temporary changes in habitat productivity occur over and above those described in the preand post-scenarios. For example, a mine might be developed, operate for 20 years, and then
be decommissioned. The PRE-scenario would be based on the pre-development condition
and the POST condition after decommission. In the intervening period, some habitat
productivity may be unavailable due to temporary uses of the fish habitat. To bring the
between-scenario into the overall analysis would require some form of discounting. At
present, no basis for deriving discount rates for temporary losses of habitat productivity has
been identified.
• Unmeasured changes – Issues such as the cultural concerns of area residents such as First
Nations, the existing or potential fisheries activities, or the presence of threatened species or
populations are not explicitly factors in a HSM analysis but are taken up in the decision
process followed by those managing fish habitat under the Fisheries Act.
5.1.3 Use of Weights in Decision-making
The use of weights in HSM recognizes that there are many sets of habitat requirements
and fishery objectives that must be factored into the analysis of net change of fish habitat
productivity. The weights are necessary to resolve potential conflicts among elements of the fish
community as habitat alteration inevitably involves a balancing among specific winners and
losers. The basic method for deriving weights here is based on a consensus using ecological and
fishery criteria. If the proponents and regulators have difficulty reaching that consensus, the
more structured methods could be employed to bridge any impasse. The basic approach adopted
here for HSM reflects methods that are well-established in other decision-making contexts where
numerical methods are used.
Healey (1984) provided one of the early examples of multi-objective, multi-player
decision-making in a fisheries context. He applied weights to sets of criteria for two fisheries,
Skeena River salmon and New England herring, where the overall objective was to define an
32
‘optimum yield’. Several interests were represented in each case study and there was a desire to
balance conservation, economic, and social objectives. Healey’s approach was based on the
principles laid out by Keeney and Raiffa (1977) in their book on decision-making, a keystone
reference for most subsequent decision-making research. Minns et al. (1993a,b) applied Saaty’s
(1982) analytical hierarchy ranking technique for deriving weights in decision-making in an
application of Kozlowski’s (1986) threshold approach to fish habitat assessment in Hamilton
Harbour. Maguire (1986) gives another biological example concerning the management of
endangered species. Munda et al. (1994) and Snell (1994) provide overviews of the range of
environmental decision-making aids where multiple objectives must be reconciled.
5.1.4 Pragmatism versus Scientific Rigour
Wilk (1985) in a discussion of the role of science in decision-making regarding control of
acidic precipitation in North America quoted Kant’s advice: “It is often necessary to make a
decision on the basis of knowledge sufficient for action, but insufficient to satisfy the intellect.”
Whenever there is an attempt to build links between science and management, issues arise about
which codes of practice and traditions should prevail. Viewed as a product of science, it might be
argued that the usual rules should apply, e.g., testable hypotheses, experimentation, replication,
statistical significance. These are stringent requirements and cannot be applied on a day to day
basis when applied in a practical management context. A healthy dose of pragmatism is
necessary if tools grounded in science are to be developed and brought into routine use.
HSM represents a short step, comparable to Rigler’s (1982) characterization of the
current state of empirical ecology, on the road from “we can predict nothing” to “we can predict
everything.” While the tools and methodologies must be well-grounded in the available science,
their application must be viewed as a small step from a position where decisions are made in a
context of almost complete ignorance without the benefit of any systematic, quantitative
assessment toward an unachievable position where decisions are made in a context of complete
knowledge with the benefit of a statistically verifiable quantitative assessment. HSM is an
empirical approach with no pretensions of providing an explanatory model of how fish
productivity derives from the productivity potential of habitats. It is designed for its intended use
as a problem-solving tool from the outset.
Funtowicz and Ravetz (1994) recognized three types of problem-solving science: applied,
professional consultancy, and post-normal, dependent on the interplay of the level of uncertainty
and the perception of the decisions stakes. Applied science operates where the uncertainties and
the stakes are low and represents the normal view of science where well-defined puzzles are
unravelled without critical external decision pressures. Post-normal science is presented as the
arena where the really big issues like global climate change and sustainability are tackled with
big human stakes and immense complexity and uncertainty. HSM is designed for use in the
intermediate science realm of professional consultancy. There the problem-solving involves
direct client service with unique situations involved. For the participants, the stakes are
comparatively high but not of global concern. The situations may be similar to others but it is the
particular set of circumstances in time and place that must be resolved. Funtowicz and Ravetz
(1994) proposed that professional consultancy science has four elements: purpose, people,
process, and product. Here the purpose is to provide an objective method for appraising
development activities for their compliance with a no net loss objective, the people are the
proponents and the regulators, the process involves both the use of HSM to structure the
assessment and the negotiations among the parties, and the product is the decision with
supporting documentation.
33
5.2 Scientific Defensibility Requirements
As Oreskes et al. (1994) cleared stated, “Verification and validation of numerical models
of natural systems is impossible.” The models are always incomplete representations of the
natural systems and a variety of models can potentially produce similar predictions. The habitat
suitability index model at the heart of HSM creates a numerical reflection of the observational
data compiled in the habitat requirements databases (Lane et al. 1996 a,b,c). Thus the model
provides its own initial level of confidence. As the model used is empirical, only empirical
confirmation of the model’s predictions is possible. Such confirmation requires one or both of
the following:
• Comparative surveys across a wide range of habitat types in a several lakes, gathering
both habitat and fish community metrics. Significant correlations between suitability
indices computed using the habitat data and a range of fish community measures would
provide a higher level of confidence.
• Replicated experimental studies of actual habitat alterations where both habitat and fish
are assessed before and after the alterations, showing results consistent with HSM
predictions would provide an even higher level of confidence.
Data collection for the first type of confirmation study has been conducted in the littoral
zone at several locations in the Great Lakes (Ontario, Erie, and Georgian Bay in Lake Huron)
with a range of habitat types at each location (Minns et al.1994, Randall et al.1996, Valere
1996). Analysis and modelling work will be reported in a separate publication (Minns et al. in
prep).
5.3 Proponent-Regulator Interactions
Traditionally these interactions have been confrontational with both sides using nonquantitative rationales to support their positions. Using a scientifically-defensible tool like
HSM does not replace the proponent-regulator interactions, but has the potential to place
discussions on a sounder, more objective footing. Assumptions are laid out explicitly via the
definition and the scenario files. The information supporting them has to be extensively
documented and presented. Once both parties have agreed to use a quantitative method such as
HSM, much of the potential for contention has been neutralized. Rather than disputing the
relative merits of various habitats and actions subjectively, the parties must come to agreement
about the method of valuing habitat units within the context of the fisheries objectives for that
general area. Use of quantitative methods places an onus on both parties to present arguments
and information that can withstand scrutiny.
The proponent is provided with an assessment tool that can be used to reduce the options
for a development to those likely to produce a net gain of productive capacity and to refine the
design of habitat modifications and enhancements. The proponent must provide a reasonable
assessment of the existing habitats, an assessment that can withstand an audit. Use of a method
like HSM should also decrease the volume of documentation required to describe the project and
proposed changes to fish habitats. Much time can be saved as the proponent will try to avoid
bringing forward proposals that clearly fail to satisfy the basic requirement for no net loss.
The regulator will still have to examine and judge the broader aspects and implications of
a proposed development, where they lie beyond the current capability of methods like HSM.
The regulator should be able to spend less time assessing the proposal. The regulatory agency
benefits as it is now able to obtain an objective assessment of the net impact of each project
when methods like HSM can be used.
34
Both parties can save much time and effort as the merits of projects can be assessed more
quickly.
5.4 Future Plans and Extensions
At the time of writing, a number of future expansions of the HSM approach are being
investigated:
Further regional lake habitat requirements databases are being assembled for British
Columbia and Yukon, and for the North encompassing NWT and Nunavut. These databases
will allow lake-based applications of HSM to be implemented in those areas.
A national location species list database is being completed and will be added to the software
package. This database will allow the user to obtain a working freshwater fish species list for
any tertiary watershed area in Canada.
The Great Lakes lacustrine habitat requirements database is currently reassessed. Since the
original databases were assembled in the mid-1990s, there has been an increase in the
appearance of new fish habitat literature. The review will be used to update and revise the
existing database. In addition, it has been apparent that vegetation cover was adequately
assessed in the original literature survey. A more concerted effort to appraise the use of other
forms of cover is under way and it is expected that the results will allow the range of
categories to be expanded in a future revision of the database. Meanwhile, the current
practice of inferring a vegetation cover equivalency for other types of cover will continue.
An exploratory investigation of the potential to develop a stream-based application of HSM
is under way in the Great Lakes basin with the cooperation of the Ontario unit of Fish Habitat
Management. A stream habitat requirements database already exists for that area (Portt et al.
1999). Certain types of developments affecting streams such as highway culverts and stream
realignments, may be highly suited to an HSM-based assessment.
In each of the cases described above, complementary efforts to verification and calibration
using field data sets will be undertaken to provide the supporting evidence necessary to
justify the use of any model like HSM.
6.0 ACKNOWLEDGEMENTS
Many biologists of Fish Habitat Management in Ontario have given considerable
assistance throughout the development and testing of this tool. Ms. Christine Stoneman, who has
now moved on to work for FHM in Alberta, provided help, input, and feedback on a number of
case study referrals. Mr Serge Metikosh, who sadly has since moved on to private sector
consulting opportunities in Alberta, provided much encouragement, many opportunities, and
financial assistance to this project and became a vocal advocate for it within and beyond DFO.
We are truly grateful for Serge’s help.
The software programmers, Bio-Software in Hamilton Ontario, who are the developers
and owners of the software code implementing this approach have provided sterling support
throughout the project along with much useful feedback both from a programming and from a
fisheries-fish habitat perspective, drawing on their experiences in previous incarnations as
biologists. Hopefully their efforts will be rewarded with some commercial success based on this
tool.
Mr. Lorne Greig and Dr. Don Meisner of ESSA Technologies provided considerable
assistance at the outset of the process that led to these developments. Their fisheries backgrounds
and experience in the delivery of adaptive management workshops proved invaluable.
Dr. Rob Nairn of W.F. Baird and Associates gave much encouragement, time and
thought to the coastal process and engineering considerations laying behind the design of new
35
physical habitats. Dr. Nairn was also a willing guinea pig with some of the early applications of
HSM to referrals occurring in Burlington.
7.0 REFERENCES
Allen, T.F.H., B.L. Bandurski, and A.W. King. 1993. The ecosystem approach: theory and
ecosystem integrity. Report of the Great Lakes Science Advisory Board, Internat. Joint
Comm., Washington D.C. and Windsor, Ont. 64p.
Balon, E.K. 1975. Reproductive guilds of fishes: a proposal and definition. J. Fish. Res. Board.
Can. 32:821-864.
Christie, A.E. 1982. Fish species of Ontario: a summary of some characteristics and their validity
to predict presence at a given site. Ontario Hydro Technical Rpt. 82-287.
Cudmore-Vokey, B.C. and E.J. Crossman. 2000. Checklists of the fish fauna of the Laurentian
Great Lakes and their connecting channels. Can. MS Rep. Fish. Aquat. Sci. 2550:p.39.
Cudmore-Vokey, B.C., M. Lange, and C.K. Minns. 2000. Database documentation and critical
review of national habitat compensation literature. Can. MS Rep. Fish. Aquat. Sci.
2526:p.35.
Degiorgi, F. and J.P. Grandmottet. 1993. Relations entre la topographie aquatique et
l’organisation spatiale de l’ictyofaune lacustre: définition des modalités spatiales d’une
stratégie de prélèvement reproductible. Bull. Fr. Pêche Piscic. 329:199-220.
Department of Fisheries and Oceans. 2000. Habitat compensation directive. In preparation.
Doka, S.E., M.L. Mallory, D.K. McNicol, and C.K. Minns. 1997. Species richness and species
occurrence of five taxonomic groups in relation to pH and other lake characteristics in
southeastern Canada. Can. Tech. Rpt. Fish. Aquat. Sci. 2179:62p.
Folk, R.L. 1954. The distinction between grain size and mineral composition in sedimentary rock
nomenclature. J. Geol. 62:344-359.
Funtowicz, S.O. and J.R. Ravetz. 1994. Uncertainty, complexity, and post-normal science.
Environm. Toxicol. Chem. 13(12):1881-1885.
Garcia, S.M. 1994. The precautionary principle: its implication to capture fisheries management.
Ocean and Coastal Management 22:99-125.
Grandmottet, J.P. 1983. Principales exigencies des teleosteens dulcicoles vis á vis de l’habitat
aquatique. Annales Sci. Univ. Tranche-Comté Besançon (4 Ser.) (Biol. Anim.), 4:3-32.
Griffin, R.C. 1998. The fundamental principles of cost-benefit analysis. Water Resources Res.
34(8):2063-2071.
He, X. and J.F. Kitchell. 1990. Direct and indirect effects of predation on a fish community: a
whole-lake experiment. Trans. Amer. Fish. Soc. 119:825-835.
Healey, M.C. 1984. Multiattribute analysis and the concept of optimum yield. Can. J. Fish.
Aquat. Sci. 41:1393-1406.
Jackson, D.A. and H.H. Harvey. 1997. Qualitative and quantitative sampling of lake fish
communities. Can. J. Fish. Aquat. Sci. 54:2807-2813.
Jennings, M.D. 2000. Gap analysis: concepts, methods, and recent results. Landscape Ecology
15:5-20.
Keast, A and J. Harker. 1977. Fish distribution and benthic incertebrate biomass relative to
depth in an Ontario lake. Environ. Biol. Fishes 2(1):53-70.
Keeney, R.L. and H. Raiffa. 1977. Decisions with multiple objectives. John Wiley and Sons,
New York, NY. 569p.
Kozlowski, J. 1986. Threshold Approach in Urban, Regional, and Environmental Planning: theory
and practice. Univ. Queensland Press, Queensland, Australia. 262p.
36
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.
Lester, N.P., K.J. Cornelisse, L. Grieg, C.K. Minns, and M.L. Jones. 1997. Proceedings of the
1997 Science for Fish Habitat Management Workshop. Can. MS Rpt. Fish. Aquat. Sci.
2439:vi+36p.
Levings, C.D., C.K. Minns, F. Aitkens (Editors). 1997. Proceedings of the DFO Workshop on
Research Priorities for Assessing Productive Capacity for Fish Habitat Management and
Impact Assessment, Sidney, B.C., May 13-15, 1996. Can. Tech. Rpt. Fish. Aquat. Sci.
2147:v+122p.
Magnuson, J.J., B.J. Benson, and A.S. McLain. 1994. Insights on species richness and turnover
from long-term ecological research: fishes in north temperate lakes. Amer. Zool. 34:437451.
Maguire, L.A. 1986. Using decision analysis to manage endangered species populations. J.
Environm. Managem. 22:345-360.
Mandrak, N.E. and E.J. Crossman. 1992. A Checklist of Ontario Freshwater Fishes annotated
with distribution maps. Royal Ontario Museum, Toronto, Ontario. 176p.
Minns, C.K., M. Stoneman, R.G. Randall, and B. Cudmore-Vokey. in preparation. An a priori
multi-species fish habitat suitability matrix (HSM) index model for lacustrine fishes in
the Great Lakes basin: formulation and evaluation.
Minns, C.K. 1989. Factors affecting fish species richness observed in Ontario's inland lakes. Trans.
Amer. Fish. Soc. 118:533-545.
Minns, C.K. 1995a. Calculating net change of productivity of fish habitats. Can. MS Rep. Fish.
Aquat. Sci. 2282:37p.
Minns, C.K. 1995b. Approaches to assessing and managing cumulative ecosystem change, with
the Bay of Quinte as a case study: an essay. J. Aquat. Ecosystem Health 4:1-24.
Minns, C.K. 1997. Quantifying “no net loss” of productivity of fish habitats in Canada. Can. J.
Fish. Aquat. Sci. 54:2463-2473.
Minns, C.K., V.W. Cairns, R.G. Randall, and J.E. Moore. 1993a. Chapter 14. UET: A tool for fish
habitat management? pages 236-245 in Kozlowski, J. and Hill, G. (eds), Towards planning
for sustainable development.
Minns, C.K., V.W. Cairns, R.G. Randall, and J.E. Moore. 1993b. Chapter 15. UET: A practical
application in Canadian lakes. pages 246-267 in Kozlowski, J. and Hill, G. (eds), Towards
planning for sustainable development.
Minns, C.K., V.W. Cairns, R.G. Randall, and J.E. Moore. 1994. An index of biotic integrity (IBI)
for fish assemblages in the littoral zones of Great Lakes’ Areas of Concern. Can. J. Fish.
Aquat. Sci. 51:1804-1822.
Minns, C.K., J.D. Meisner, J.E. Moore, L.A. Greig, and R.G. Randall. 1995. Defensible Methods
for Pre- and Post-Development Assessment of Fish Habitat in the Great Lakes. I. A
Prototype Methodology for Headlands and Offshore Structures. Can. MS Rpt. Fish.
Aquat. Sci. 2328:xiii+65p.
37
Minns, C.K. and J.E. Moore. 1995. Factors limiting the distributions of Ontario’s freshwater
fishes: the role of climate and other variables, and the potential impacts of climate
change. Can. Spec. Publ. Fish. Aquat. Sci. 121:137-160.
Minns, C.K., J.R.M. Kelso, and R.G. Randall. 1996. Detecting the response of fish to habitat
alterations in freshwater ecosystems. Can. J. Fish. Aquat. Sci. 53(Suppl.1):403-414.
Minns, C.K. and R.B. Nairn. 1999. Definsible Methods: applications of a procedure for
assessing developments affecting littoral fish habitat on the lower Great Lakes. In
Aquatic Restoration in Canada. T.P Murphy and M. Munawar (eds.). pages 15-35.
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. 1999. Development of a fish habitat classification model for littoral areas of Severn
Sound, Georgian Bay, a Great Lakes’ Area of Concern. Can. MS. Rpt. Fish. Aquat. Sci.
2490: ix+86p.
Munda, G., P. Nijkamp, and P. Rietveld. 1994. Qualitative multicriteria evaluation for
environmental management. Ecol. Economics 10:97-112.
Oreskes, N., K. Shrader-Frechette, and K. Belitz. 1994. Verification, validation, and
confirmation of numerical models in the earth sciences. Science 263:641-646.
Portt, C.B., G. Coker, and C.K. Minns. 1999. Riverine habitat characteristics of fishes of the
Great Lakes watershed. Can. MS Rpt. Fish. Aquat. Sci. 5481:68p.
Randall, R.G., C.K. Minns, V.W. Cairns, & J.E. Moore. 1996. The relationship between an index of
fish production and submerged macrophytes and other habitat variables at three littoral areas
in the Great Lakes. Can. J. Fish. Aquat. Sci. 53(Suppl 1):35-44.
Randall, R.G. and C.K. Minns. 2000. Use of fish production per unit biomass ratios for measuring
the productive capacity of fish habitats. Can. J. Fish. Aquat. Sci. 57:1654-1667.
Richards, L.J. and J-J. Maguire. 1998. Recent international agreements and the precautionary
approach: new directions for fishery management science. Can. J. Fish. Aquat. Sci.
55:1545-1552.
Rigler, F.H. 1982. Recognition of the possible: an advantage of empiricism in ecology. Can. J.
Fish. Aquat. Sci. 39:1323-1331.
Robinson, C.L.K. and W.M. Tonn. 1989. Influence of environmental factors and piscivory in
structuring fish assemblages of small Alberta lakes. Can. J. Fish. Aquat. Sci. 46:81-89.
Royal Commission on the Future of the Toronto Waterfront. 1992. Regeneration: Toronto’s
waterfront and the sustainable city: final report. 530p.
Saaty, T.L. 1982. Decision making for leaders. Wadsworth, Belmont, California. 291p.
Sagoff, M. 1996. On the value of endangered and other species. Environm. Managem.
20(6):897-911.
Scheffer, M., M.R. de Redelijkhied, and F. Noppert. 1992. Distribution and dynamics of
submerged vegetation in a chain of shallow eutrophic lakes. Aquat. Bot. 42:199-216.
Scott, J.M.; F. Davis; B. Csuti; R. Noss; B. Butterfield; C. Groves; H. Anderson; S. Caicco; F.
D’Erchia; T. Edward Jr; J. Ulliman; R. G. Wright. 1993. GAP Analysis: A geographic
approach to protection of biological diversity. Wildl. Monogr. 123:1-41.
Snell, M.J. 1994. Decision aid for planning in the real world. J. Inst. Water Environm.
Management 8(4):362-370.
Stanfield, L., M. Jones, M. Stoneman, B. Kilgour, J. Parish, and G. Wichert. 1997. Stream
assessment protocol for Southern Ontario. Ontario Ministry of Natural Resources, Great
Lakes Salmonid Unit. Picton, Ontario.
38
Terrell, J.W., T.E. McMahon, P.D. Inskip, R.F. Raleigh, and K.L. Williamson. 1982. Habitat
suitability models: Appendix A. Guidelines for riverine and lacustrine applications of fish
HSI models with the habitat evaluation procedures. FWS/OBS-82/10.A, Sept.65p.
Terrell, J.W. editor. 1984. Proceedings of a workshop on fish habitat suitability index models.
Biological Report 85(6), U.S. Fish and Wildlife Service.
Thomas, M.L.H. 1986. A physically-derived exposure index for marine shorelines. Ophelia
25(1)1-13.
U.S. Fish and Wildlife Service. 1981. Standards for the development of habitat suitability index
models for use in the Habitat Evaluation Procedures, USDI Fish and Wildlife Service.
Division of Ecological Services. ESM 103.
Valere, B.G. 1996. Productive capacity of littoral habitats in the Great Lakes: field sampling
procedures (1988 – 1995). Can. MS Rpt. Fish. Aquat. Sci. 2384:50p.
Wilk, L.J. 1985. Responsibilities of scientists: examination of the acid precipitation problem. Sci.
Total Environm. 44:293-299.
39
Figure 2.1.1 A flow-chart showing the overall structure of HSM and the activities in the
Definition, Analysis, and Reports components.
Component
Activity
Description
Definition
Location
- Fish species list identified for general area, e.g., Lake
Ontario, Georgian Bay, or Hamilton Harbour.
- Species list divided into functional groups and group
weights assigned based on fish community objectives.
- Habitat requirements to be assessed - defaults to
spawning, YOY, and adults needs.
- Physical habitat assessment data sets are assembled for
pairs of pre- and post-development options.
- Depth, substrate and cover type used in scenarios
are ‘mapped’ to those in the requirements database.
- Suitability values are computed for species, groups,
life stages, and group-weighted aggregations.
Groups
Habitat Elements
Scenarios
I/O Mapping
Analysis
Suitabilities
...
Inventories
- Unrated habitat supply is inventoried by area type and
habitat category for each scenario.
...
Weighted Suitable
Areas (WSA)
Reports
Suitabilities
- By data set, area is multiplied by suitability and summed
for species, groups, and aggregates by scenario, area, and
habitat category.
- Tables of species, group, and aggregate suitabilities
are prepared for display and interpretation.
...
Inventories
- Tables of habitat by area type and habitat category are
prepared.
...
Weighted Suitable
Areas (WSA)
- Tables of WSA are prepared at several levels of
aggregation. Results can be printed and transferred to
spreadsheet software.
[...= intermediate steps]
40
Table 2.3.1 Sample hypothetical pre- and post-development scenario data files for use in an
application of HSM.
[The first record contains the title for use in HSM. It and succeeding comment records begin
with a semi-colon. Directive records describe the data records that follow and begin with an
asterisk. Blank records are ignored and can be used to structure the file.]
Pre-development habitat scenario
Post-development habitat scenario
; OPTNPRE
; Option Pre-development
; OPTNPOST
; Option Post-development
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
01, 200, LOSS, 0.0, SAND, NOVEG
02, 1800, LOSS, 0.0, BEDROCK, NOVEG
03, 300, LOSS, 0.5, BEDROCK, NOVEG
04, 120, LOSS, 0.5, SAND, NOVEG
05, 270, LOSS, 0.5, SAND, NOVEG
06, 1260, MODD, 0.5, BEDROCK, NOVEG
07, 140, MODD, 0.5, SAND, NOVEG
08, 320, MODD, 0.5, BEDROCK, NOVEG
09, 115, MODD, 0.5, SAND, NOVEG
10, 318, MODD, 0.5, SAND, NOVEG
11, 50, MODD, 1.0, SAND, NOVEG
06, 1260, MODD, 0.5, COBBLE, NOVEG
07, 140, MODD, 0.5, COBBLE, NOVEG
08, 320, MODD, 0.5, ARMOUR, NOVEG
09, 115, MODD, 0.5, ARMOUR, NOVEG
10, 318, MODD, 0.5, ARMOUR, NOVEG
11, 50, MODD, 0.5, ARMCOB, NOVEG
12, 1000, COMC, 1.0, SILTYSAND, VEGGY
41
Table 2.3.2 Area categories for assessing net change of productivity of fish habitat after
Minns (1995a, 1997). All areas in HSM scenarios should be assigned to one of these
categories.
Code
LOSS
Name
Loss
MODD
Modified-Direct
MODI
Modified-Indirect
COMM
Compensation-Modified
COMC
Compensation-Created
UNCH
Unchanged
Description
Destruction of habitat caused by the
development activity, e.g., infilling.
Habitat directly modified by the
development, e.g., changes in depth and
substrate due to the addition of
construction materials.
Habitat indirectly modified as a result of
the construction, e.g., the development
creates a barrier which alters the sediment
retention characteristics and reduces the
wave/wind exposure of the site.
Existing habitat deliberately modified
outside the actual development site with
the intention of increasing fish
productivity to compensate for loss and/or
decrease at the development site.
Habitat created where none existed with
the intention of increasing fish
productivity to compensate for loss and/or
decrease at the development site.
Habitat remaining unaltered by the
development but included in the
accounting of both PRE and POST
scenarios or for total supply analysis.
42
Table 2.3.3 Alternate methods for assigning percentages to vegetation cover categories
contingent on the type of information available.
Method
Categorical
(P/A)
No Cover
Emergent
Submergent
Proportion
Additive
Proportion
Non-additive
Re-scaled
Cover Category Percentages
No Cover
Emergent (E)
Submergent (S)
Total
100
100
0
0
100 - (%E+%S)
0
100
0
%E
0
0
100
%S
100 - Max(%E,
%S)
%E
%S
[100 Max(%E,%S)/T
PN
100*%E/TPN
100*%E/TPN
43
100
100 Max(%E,%S) +
$E + %S = TPN
100
Figure 2.4.1 A representation of the habitat suitability matrix (❒) for a single species-life stage combination.
44
Table 2.4.1 How species level suitability matrices (❒) are pooled into fish group matrices.
Species
Weights
1
2
3
1
1
1
A
B
-
❒
❒
nA
4
1
5
1
.
1
.
1
.
1
N
1
No. spp.
Group matrices
❒
Fish Groups
C
-
..
-
❒
❒
-
nB
❒
-
❒
❒
-
nC
❒
❒
❒
M
-
❒
❒
nM
❒
Table 2.4.2 Assignment of weights among fish group and life stage suitability matrices (❒). The
weights are proportions that sum to 1 on each axis and the sum of their cross-products also
equals 1.
Fish Groups
Weights
Spawning
WSP
Life Stages
YOY
WYOY
A
B
C
.
M
Sum
WA
WB
WC
.
WM
1
❒
❒
❒
❒
❒
❒
❒
❒
❒
❒
Adult
WAD
❒
❒
❒
❒
❒
Sum
1
1
45
Table 2.4.3 Inferred no cover preference levels based on the cross-matrix of emergent
and submergent preferences for use when computing suitability matrices of Great
Lakes basin lacustrine fishes.
Emergents
Nil
Low
Medium
High
Nil
High
Medium
Low
Low
Submergents
Low
Medium
Medium
Low
Low
Low
Low
Nil
Nil
Nil
46
High
Low
Nil
Nil
Nil
Table 2.5.1 Sample input records of a hypothetical scenario pair for an exposed Lake Ontario
shoreline site used as the basis for demonstrating sample output.
***
; PREOPTN
; Pre-development Option
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
01, 2100, LOSS, 0.0, SAND, NOVEG
02, 590, LOSS, 0.0, BEDROCK, NOVEG
03, 1580, MODD, 0.5, BEDROCK, NOVEG
04, 623, MODD, 0.5, SAND, NOVEG
***
; PSTOPTN
; Post-development Option
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
01, 50, MODD, 0.5, ARMCOB, NOVEG
02, 1400, MODD, 0.5, COBBLE, NOVEG
03, 753, MODD, 0.5, ARMOUR, NOVEG
04, 270, COMC, 1.0,SILTYSAND, VEGGY
***
47
Table 2.5.2 Summary of habitat criteria provided in a standard report for an application of HSM: A) Lake habitat types and weights,
and B) Lake habitat classification.
A) Lake habitat types and weights:
Name
Adult
Spawning
YOY
Weight
0.33
0.33
0.33
Description
Adult habitat
Spawning habitat
Nursery habitat
B) Lake habitat classification:
Type
Adult
Spawning
YOY
Variable
Depthzone
Substrate
Vegetation
Depthzone
Substrate
Vegetation
Depthzone
Substrate
Vegetation
1
Z0_1
Bedrock
Submergent
Z0_1
Bedrock
Submergent
Z0_1
Bedrock
Submergent
2
Z1_2
Boulder
Emergent
Z1_2
Boulder
Emergent
Z1_2
Boulder
Emergent
3
Z2_5
Cobble
NoCover
Z2_5
Cobble
NoCover
Z2_5
Cobble
NoCover
4
Z5-10
Rubble
5
Z10+
Gravel
6
7
8
9
10
11
Sand
Silt
Clay
Hardpan
Pelagic
No_use
Z5-10
Rubble
Z10+
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
No_use
Z5-10
Rubble
Z10+
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
No_use
48
Table 2.5.3 Lake habitat cross-classifications for linking reporting and scenario attribute classes to the internal classes for all three life
stages in A) Depth zones, B) Substrates, and C) Vegetation cover.
A) Depth zones
Reporting:
Z0_1
Z1_2
Z2_5
C) Vegetation cover
0_1
X
1_2
2_5
5_10
10+
X
X
Z5_10
Z10+
reporting:
Submergent
Emergent
NoCover
Submergent
X
Emergent
NoCover
Scenario:
NOVEG
Submergent
Emergent
NoCover
1.0
VEGGY
0.25
0.25
0.5
Clay
Hardpan
Pelagic
No_use
X
X
X
X
Scenario:
0.0
0_1
1.0
0.5
1.0
1.0
1.0
1_2
2_5
5_10
10+
B) Substrates
reporting:
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
No_use
Bedrock
X
Scneario:
ARMCOB
ARMOUR
BEDROCK
COBBLE
SAND
SILTYSAND
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
X
X
X
X
X
X
X
X
X
X
0.5
1.0
Boulder
0.7
0.5
Cobble
0.3
Rubble
Gravel
Sand
Silt
1.0
0.75
0.25
1.0
49
Clay
Hardpan
Pelagic
No_use
Table 2.5.4 Fish group names, weights, numbers of species overall and by life stage, and species
lists for an application of HSM drawing from the Lake Ontario fish taxon.
Group ID
Group name
Weight
Total spp.
Adult spp.
Spawning spp.
YOY spp.
Species
WARMP
Warmwater
Piscivores
0.167
4
4
4
4
bowfin
white bass
smallmouth bass
largemouth bass
WARMNP
Warmwater
Non-piscivores
0.167
29
28
28
27
gizzard shad
central
mudminnow
quillback
northern hog
sucker
goldfish
common carp
rosyface shiner
spotfin shiner
sand shiner
COOLP
Coolwater
Piscivores
0.167
7
6
5
5
longnose gar
northern pike
COOLNP
Coolwater
Non-piscivores
0.167
42
40
39
40
alewife
rainbow smelt
COLDP
Coldwater
Piscivores
0.167
9
9
5
7
chum salmon
coho salmon
COLDNP
Coldwater
Non-piscivores
0.167
14
14
14
14
lake sturgeon
lake whitefish
muskellunge
grass pickerel
mooneye
white sucker
chinook salmon
rainbow trout
lake herring
bloater
tiger muskellunge
American eel
walleye
bigmouth buffalo
silver redhorse
shorthead redhorse
greater redhorse
northern redbelly
dace
finescale dace
redside dace
lake chub
brassy minnow
eastern silvery
minnow
river chub
golden shiner
pugnose shiner
emerald shiner
bridle shiner
common shiner
blackchin shiner
blacknose shiner
spottail shiner
blacknose dace
longnose dace
Atlantic salmon
brown trout
brook trout
lake trout
burbot
kiyi
shortnose cisco
shortjaw cisco
round whitefish
longnose sucker
bluntnose minnow
fathead minnow
central stoneroller
yellow bullhead
brown bullhead
channel catfish
stonecat
tadpole madtom
white perch
rock bass
pumpkinseed
bluegill
white crappie
black crappie
warmouth
orangespotted
sunfish
freshwater drum
carpXgoldfish
rudd
creek chub
fallfish
banded killifish
brook stickleback
threespine
stickleback
ninespine
stickleback
yellow perch
eastern sand darter
rainbow darter
Iowa darter
fantail darter
least darter
johnny darter
logperch
channel darter
tessellated darter
brook silverside
50
pearl dace
trout-perch
mottled sculpin
slimy sculpin
deepwater sculpin
Table 2.5.5 Sample habitat supply area summary, m2, by area type and habitat category for adult
and spawning+YOY habitat classes. Using the input matrices, scenario areas are apportioned
among the internal habitat classes.
Area type
Variable
LOSS
Depth zones
Substrate
Vegetation cover
MODD
Depth zones
Substrate
Vegetation cover
COMC
Depth zones
Substrate
Vegetation cover
Class
0_1
1_2
2_5
5_10
10+
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hard-pan
Pelagic
No-use
Submergent
Emergent
NoCover
0_1
1_2
2_5
5_10
10+
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hard-pan
Pelagic
No-use
Submergent
Emergent
NoCover
0_1
1_2
2_5
5_10
10+
Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hard-pan
Pelagic
No-use
Submergent
Emergent
NoCover
51
Adult, Spawning and YOY
PRE area
2690.0
2100.0
590.0
2690.0
2203.0
1580.0
623.0
2203.0
-
POST area
2203.0
376.5
411.5
1415.0
2203.0
1000.0
750.0
250.0
250.0
250.0
500.0
Table 2.5.6 Weighted suitable areas (WSA) for the demonstration scenarios showing the format
and organization of the report table. WSA, m2, are the equivalent of the habitat area if habitat
suitability were at the maximum.
Type
Adult
Group ID
Warmwater Piscivores
Warmwater Non-piscivores
Coolwater Piscivores
Coolwater Non-piscivores
Coldwater Piscivores
Coldwater Non-piscivores
Weights
0.17
0.17
0.17
0.17
0.17
0.17
Pre-scenario
695.3
705
75.6
1220.9
32.8
97.8
Post-scenario
724.4
956.3
468.8
1120.5
13.5
65.2
Spawning
Warmwater Piscivores
Warmwater Non-piscivores
Coolwater Piscivores
Coolwater Non-piscivores
Coldwater Piscivores
Coldwater Non-piscivores
0.17
0.17
0.17
0.17
0.17
0.17
872.8
1048.2
318.7
888.2
432.4
729.9
674.4
1082.7
339.1
843.8
329.9
837.9
YOY
Warmwater Piscivores
Warmwater Non-piscivores
Coolwater Piscivores
Coolwater Non-piscivores
Coldwater Piscivores
Coldwater Non-piscivores
0.17
0.17
0.17
0.17
0.17
0.17
729.5
336.3
186.5
1044.8
1182.3
178.1
740.6
591.6
248.3
725.3
513.9
192.8
Adult
Spawning
YOY
0.33
0.33
0.33
471.2
715
609.6
558.1
684.7
502.1
598.6
581.6
Weighted Sum
Overall Sum
52
Table 2.6.1 Average suitability index values by thermal group of fish species for 150 unique
combinations of depth, substrate, and vegetation cover classes, computed using HSM.
Depth Substrate
Z0_1 Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
Z1_2 Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
Z2_5 Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
Z5-10 Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
Z10+ Bedrock
Boulder
Cobble
Rubble
Gravel
Sand
Silt
Clay
Hardpan
Pelagic
Warmwater
No Cover
0.047
0.057
0.114
0.298
0.449
0.461
0.264
0.061
0.018
0.073
0.047
0.057
0.088
0.238
0.328
0.360
0.264
0.061
0.018
0.073
0.025
0.063
0.069
0.151
0.191
0.193
0.154
0.023
0.031
0.073
0.024
0.062
0.065
0.115
0.140
0.142
0.118
0.010
0.024
0.073
0.024
0.031
0.035
0.037
0.037
0.036
0.025
0.010
0.006
0.005
Emergent Submergent
0.036
0.033
0.028
0.025
0.041
0.033
0.171
0.122
0.472
0.318
0.958
0.710
0.903
0.687
0.151
0.143
0.032
0.018
0.000
0.000
0.036
0.033
0.028
0.025
0.041
0.033
0.171
0.122
0.472
0.318
0.932
0.685
0.877
0.661
0.151
0.143
0.032
0.018
0.000
0.000
0.016
0.013
0.045
0.041
0.052
0.047
0.095
0.060
0.272
0.141
0.348
0.220
0.331
0.200
0.083
0.054
0.045
0.031
0.000
0.000
0.011
0.011
0.037
0.037
0.040
0.040
0.041
0.041
0.129
0.049
0.109
0.028
0.094
0.014
0.011
0.008
0.024
0.024
0.000
0.000
0.011
0.011
0.010
0.010
0.014
0.014
0.014
0.014
0.029
0.023
0.026
0.019
0.019
0.014
0.011
0.008
0.006
0.006
0.000
0.000
Coolwater
No Cover
0.012
0.060
0.092
0.262
0.463
0.480
0.326
0.038
0.012
0.017
0.012
0.060
0.059
0.147
0.350
0.386
0.295
0.011
0.012
0.017
0.003
0.017
0.017
0.042
0.116
0.116
0.077
0.008
0.012
0.035
0.003
0.006
0.007
0.025
0.080
0.076
0.033
0.005
0.009
0.052
0.003
0.005
0.005
0.017
0.067
0.060
0.018
0.005
0.008
0.052
53
Emergent Submergent
0.004
0.004
0.018
0.016
0.030
0.021
0.095
0.083
0.398
0.268
0.647
0.527
0.781
0.569
0.103
0.142
0.012
0.012
0.000
0.000
0.004
0.004
0.018
0.016
0.027
0.018
0.071
0.058
0.312
0.181
0.513
0.331
0.656
0.382
0.103
0.080
0.012
0.012
0.000
0.000
0.004
0.004
0.013
0.013
0.020
0.015
0.046
0.026
0.177
0.098
0.235
0.159
0.302
0.185
0.060
0.041
0.012
0.012
0.000
0.000
0.004
0.004
0.007
0.007
0.009
0.008
0.029
0.015
0.079
0.047
0.108
0.086
0.132
0.113
0.005
0.002
0.009
0.009
0.000
0.000
0.004
0.004
0.005
0.005
0.007
0.006
0.027
0.012
0.065
0.034
0.070
0.053
0.095
0.081
0.005
0.002
0.008
0.008
0.000
0.000
Coldwater
No Cover
0.023
0.137
0.135
0.356
0.456
0.378
0.071
0.015
0.028
0.030
0.023
0.104
0.068
0.212
0.342
0.361
0.076
0.015
0.028
0.030
0.023
0.133
0.079
0.172
0.304
0.350
0.092
0.014
0.029
0.147
0.174
0.212
0.220
0.244
0.194
0.215
0.303
0.512
0.148
0.401
0.174
0.209
0.218
0.245
0.199
0.218
0.309
0.498
0.148
0.439
Emergent Submergent
0.000
0.000
0.006
0.004
0.007
0.004
0.015
0.009
0.061
0.064
0.074
0.063
0.024
0.013
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.006
0.004
0.007
0.004
0.015
0.009
0.026
0.013
0.062
0.015
0.024
0.013
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.002
0.000
0.003
0.000
0.011
0.011
0.024
0.013
0.025
0.015
0.024
0.009
0.003
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.006
0.004
0.007
0.004
0.013
0.014
0.011
0.015
0.012
0.017
0.011
0.009
0.003
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.007
0.006
0.008
0.006
0.012
0.014
0.014
0.014
0.016
0.016
0.015
0.009
0.006
0.003
0.000
0.000
0.000
0.000
Figure 2.6.1 Mean habitat suitability index values among life stages (y-axis) versus depth (z-axis) and substrate classes (x-axis) for
three thermal groupings of fish species (warm, cool and cold) across three cover classes (no cover, submergent, and emergent
vegetation).
B) Warm - Submergent
C) Warm - Emergent
1.0
54
Pe
C
l
Si
Sa
r
u
G
C
o
Pe
C
l
Si
r
Sa
G
C
o
u
0_1
1_2
2_5
5_10
10+
Pe
C
l
Si
Sa
r
u
G
C
o
0.0
H
p
Pe
C
l
H
p
Si
Sa
r
u
G
C
o
R
Pe
C
l
H
p
Si
Sa
r
u
G
C
o
R
Be
Bo
0_1
1_2
2_5
5_10
10+
0.0
Be
0_1
1_2
2_5
5_10
10+
0.5
R
0.5
0.0
R
1.0
Bo
0.5
Be
I) Cold - Emergent
1.0
H
p
Pe
C
l
Si
Sa
r
u
G
C
o
R
Be
H) Cold - Submergent
1.0
0_1
1_2
2_5
5_10
10+
0.0
Bo
0_1
1_2
2_5
5_10
10+
0.0
H
p
Pe
C
l
H
p
Si
Sa
r
u
G
C
o
R
Be
Bo
G) Cold - No Cover
0.5
Bo
0_1
1_2
2_5
5_10
10+
R
1.0
0.5
0.0
Be
F) Cool - Emergent
1.0
0.5
0_1
1_2
2_5
5_10
10+
0.0
H
p
Pe
C
l
Si
Sa
r
u
G
C
o
R
E) Cool - Submergent
1.0
H
p
Pe
C
l
H
p
Si
Sa
r
u
G
C
o
R
Be
Bo
D) Cool - No Cover
0_1
1_2
2_5
5_10
10+
0.0
Be
0_1
1_2
2_5
5_10
10+
0.0
0.5
Bo
0.5
Bo
0.5
1.0
Be
1.0
Bo
A) Warm - No Cover
Table 2.6.2 Pearson correlations and their Bonferroni-adjusted significance levels among 18 habitat suitability indices representing all
combinations of thermal, life stage, and trophic level groupings of the freshwater fishes occurring in lakes in the Great Lakes basin.
Thermal
Warm
Life stage
Adult
Trophy
Warm
Adult
Pisc.(P)
Non-pisc.(N)
Spawning
Pisc.
Non-pisc.
YOY
Pisc.
Non-pisc.
Cool
Adult
P
Cool
Spawning
N
0.77
P
YOY
N
P
Adult
N
P
N
Adult
P
N
P
YOY
N
P
N
0.07
0.67
0.67
0.68
0.73
0.35
0.44
0.59
0.73
0.09
0.10
0.18
0.10
0.35
0.81
0.91
0.73
0.90
0.60
0.71
0.67
0.82
-0.05
0.07
0.28
0.10
0.31
0.06
0.83
0.71
0.91
0.64
0.62
0.77
0.56
0.56
0.57
-0.09
-0.01
0.06
-0.05
0.01
-0.01
0.71
0.86
0.65
0.84
0.68
0.82
0.55
0.73
-0.07
0.02
0.30
0.09
0.21
0.02
0.85
0.63
0.72
0.46
0.52
0.87
0.75
-0.05
0.07
0.16
0.13
0.26
0.10
0.78
0.74
0.72
0.56
0.76
0.74
-0.07
0.03
0.01
-0.05
0.05
0.02
-0.06
Pisc.
0.53
Pisc.
Pisc.
Pisc.
0.52
0.35
0.63
0.50
-0.10
-0.02
-0.10
-0.16
-0.10
0.41
0.75
0.57
0.85
0.11
0.18
0.43
0.18
0.48
0.08
0.56
0.39
0.44
-0.06
-0.02
-0.03
-0.03
-0.01
-0.03
0.36
0.61
-0.04
0.03
0.60
0.26
0.40
0.05
0.67
-0.04
0.04
-0.02
-0.05
0.07
0.02
-0.04
0.12
0.33
0.17
0.48
0.13
0.54
0.00
0.00
0.02
0.00
0.07
Pisc.
Non-pisc.
YOY
N
0.62
Non-pisc.
Spawning
P
Spawning
0.92
Non-pisc.
Cold
N
Adult
0.47
Non-pisc.
YOY
P
YOY
0.81
Non-pisc.
Spawning
Cold
Spawning
Pisc.
0.27
0.18
0.20
0.47
0.68
0.11
0.57
0.75
0.27
Non-pisc.
Bold - significant at P=0.01, Italics – significant at P=0.05 but >0.01.
55
Figure 3.2.1 The framework for completing an application of the Habitat
Suitability Matrix method. The numbers provide the logical order in which the
steps are completed with sub-ordering indicated with lowercase letters.
1a. Site-specific
Habitat
Inventory
1. Physical Habitat
Assessment
Scenarios
Species’ Habitat
Requirements
Classes
⇑
Species’ Habitat
Requirements
Database
2. Cross-link
Scenario &
Database Classes
3. Fish Community,
Ecosystem Type, &
Fishery Objectives
5. ‘HSM’
Assessment
4. Location Spp.,
Fish Groups &
Life Stage
Weights
Þ
Þ
6. Suitability Values
And Weighted
Suitable Areas
7. Assess
Net Gain / Loss of
Habitat Productivity
56
⇐
⇐
⇔
1b. Physical Habitat
Modelling and
Conditions
6a. Inspect
Supporting
Results Tables
Figure 3.2.2 The fish habitat area categories illustrated for the construction of a solid pier on the shore of a lake.
57
Figure 4.1.1 Illustration of case study 1, Brant Inn Node, A) prior to project and B) components of the preferred development option.
58
Figure 4.1.2 Bar-charts of A) pre-, post-, and net weighted suitable area (WSA) and B)
area affected by type for the original 6 development options (option 0 was to do nothing)
at Brant Inn Node on the Burlington waterfront of Lake Ontario.
A ) W S A p re -, p o s t-, a n d n e t
3 0 0 0 .0
P re
P ost
Net
Weighted suitable area
2 0 0 0 .0
1 0 0 0 .0
0 .0
-1 0 0 0 .0
-2 0 0 0 .0
0
1
2
3
4
5
6
O p tio n S c e n a rio s
B ) A re a b y ty p e
6000
LO SS
MODD
MODI
COMC
4000
Area m2
2000
0
-2 0 0 0
-4 0 0 0
-6 0 0 0
1
2
3
4
O p tio n S c e n a rio s
59
5
6
Table 4.1.2 The weighted suitable area (WSA) results for the final scenario of case
study 1, the Brant Inn Node, application of HSM, showing pre- and post-development,
net change, and percent change in WSA by life stage, fish groups and pooled groups,
along with group weights.
Life Stage
Group
Adult
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Spawning
YOY
Pooled:
Adult
Spawning
YOY
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
695.3
705.0
75.6
1220.9
32.8
89.7
872.8
1048.2
318.7
888.3
432.4
729.9
729.5
336.3
186.5
1044.8
1182.3
178.1
908.8
1255.7
788.0
1139.5
12.6
69.9
1228.5
1404.7
589.7
1222.8
793.8
783.9
802.1
918.3
228.8
654.2
454.1
144.0
Net
Change
213.5
550.7
712.4
-81.4
-20.2
-19.8
355.7
356.5
271.0
334.5
361.4
54.0
72.6
582.0
42.3
-390.6
-728.2
-34.1
0.33
0.33
0.33
469.9
715.0
609.6
695.7
1003.9
533.6
225.8
288.9
-76.0
48.1
40.4
-12.5
598.2
744.4
146.2
24.4
Weight
Pre-
Overall Sum
60
Post-
%
30.7
78.1
942.3
-6.7
-61.6
-22.1
40.8
34.0
85.0
37.7
83.6
7.4
10.0
173.1
22.7
-37.4
-61.6
-19.1
Figure 4.2.1 Illustration of the A) pre-development condition and B) post-development
condition of case study 2, the Wallik property.
61
Table 4.2.1 The weighted suitable area (WSA) results for case study 2, the Wallik
property, application of HSM, showing pre- and post- development, net change, and
percent change in WSA by life stage, fish groups and pooled groups, along with group
weights.
Life Stage
Group
Adult
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Warm Pisc
Warm Non-Pisc
Cool Pisc
Cool Non-Pisc
Cold Pisc
Cold Non-Pisc
Spawning
YOY
Pooled:
Adult
Spawning
YOY
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.17
0.0
2.1
0.0
0.7
0.0
2.6
59.6
1.9
5.8
0
18.7
0
29.2
0.0
0.0
16.1
0.0
0.0
38.7
58.3
12.6
78.0
1.9
5.1
33.2
56.1
6.9
76.6
84.9
77.5
40.1
9.7
0.0
16.0
65.5
11.7
Net
Change
38.7
56.2
12.6
77.3
1.9
2.5
-26.4
54.2
1.1
76.6
66.2
77.5
10.9
9.7
0.0
-0.1
65.5
11.7
0.33
0.33
0.33
0.9
14.3
7.5
32.4
55.9
23.8
31.5
41.6
16.3
3500.0
290.9
217.3
7.6
37.4
29.8
392.1
Weight
Pre-
Overall Sum
62
Post-
%
+
2676.2
+
11042.9
+
96.2
-44.3
2852.6
19.0
+
354.0
+
37.3
+
-0.6
+
+
Table 5.1.1 Derivation of eight principles of HSM method by comparison with the
fundamental principles of cost-benefit analysis, after Griffin (1998).
Principle
Wording
#
Cost-Benefit Analysis*
HSM Method**
One
Projects are deemed economically
+Analyses are deemed acceptable if
acceptable “… if the benefits to
the benefits in fish habitat
whomsoever they accrue are in
productivity exceed the costs
excess of the estimated costs…”
Two
Welfare changes pertain to
+Fish habitat productivity changes
differences between with- and
pertain to differences between prewithout-project scenarios
and post-development scenarios
Three
Cost measurement is founded on
+Measures of potential habitat
social opportunity costs
productivity provide the reference
currency for analyses
Four
Producer benefits are to be measured +/?Fish are the beneficiary in both
as producer surplus changes
instances and the benefits are the
Five
Consumer benefits are to measured as marginal changes in fish habitat
consumer surplus changes
productivity
Six
Zero-sum transfers of benefits or
-All habitats and their productivities
costs are to be ignored
affected by the project are included in
the analysis
Seven
Temporal aggregation employs
?Discounting is not considered in the
discounting
current version
Eight
Unmonetized welfare changes are to
+Other issues and factors not
be disclosed
expressed in the analysis should be
captured in the final decision-making
* Exact wording from section 7 of Griffin (1998)
** + implies agreement, - disagreement, and ? incomplete in current version
63
Appendix A: Adjusting analysis for changes in wave exposure on the Great Lakes.
Introduction
Cumulative experience with a range of development projects affecting littoral
habitats in the Great Lakes and other large Ontario inland lakes indicated an anomaly in
the case of breakwalls installed on shores with high wave exposure. The breakwalls are
usually installed as part of a marina development or to provide some relief for the
existing shoreline. They are usually designed to provide for some level of water
exchange/circulation minimizing the potential for stagnation in the lee of the breakwall.
The breakwall will substantially reduce the wave exposure and physical disturbance of
fish habitats in the lee of the structure. The basic configuration of the Habitat Suitability
Matrix (HSM) model does not reflect this effect which is generally noted to be associated
with increasing fish abundance. The HSM model only considers combinations of depth,
substrate and cover. These values may not change in indirectly modified areas (MODI) in
the lee of the breakwall and hence the weighted suitable area values for those habitat
areas would be unchanged between pre- and post-development scenarios. As a result the
proponent can face great difficulty trying to devise a package of habitat modifications
and compensation to offset the losses associated with the breakwall structure itself.
The HSM method does allow for the use of an auxiliary variable, the Condition
Index, in situations where additional habitat conditions change. The purpose of the
analyses reported here was to derive estimates from field data of the expected change in
fish community measures arising from changes in wave exposure. These estimates might
then be used to derive a simple method for estimating Condition Indices for areas in the
lee of a breakwall that are either modified indirectly or through compensation.
Materials and Methods
The data used to examine the effect of wave exposure on fish community metrics
was drawn from the Great Lakes nearshore electrofishing transect database (Valere
1966). Using a standard 100 metre transect along the 1.5 metre contour, a wide range of
shore habitats have been assessed in lakes Ontario and Erie and in Georgian Bay. The
biomass and number of fish caught are recorded by species for each transect sample. For
this analysis, data from Hamilton Harbour were excluded as previous analyses have
shown that the fish community metrics are lower there compared to other sites due to
factors other than physical habitat (Minns et al. 1994). In addition, data collected at a
series of harbours and marinas on lakes Ontario and Erie were also excluded from the
analysis to ensure that the adjustment factors for exposure were based on natural habitat
conditions rather than any factors present at harbours and marinas apart from wave
exposure changes. This left a data set of 501 transect samples.
Four measures were derived for each transect sample: species richness, biomass,
number per unit effort, and the adjusted Index of Biotic Integrity (IBI) which is indicative
of the biodiversity at each site. These are measures previously shown to be key indicators
for the fish community in the littoral zone of the Great Lakes (Randall et al. 1996, Minns
et al. 1994.). To derive a single metric representative of these four measures, a principal
components analysis (PCA) was performed and the first factor score used as a fish
community metric. The biomass and numbers measures were transformed by taking log10
(X+1) to produce distributions more closely approximating normal.
Maximum effective fetch was used as an indicator of the wave exposure at each
transect location (Thomas 1986). Fetch has been previously been shown to be an
64
important predictor of changes in fish community metrics (Randall et al. 1996). Effective
fetch (km) was computed using the formula of Scheffer et al. (1992) from raw fetch
distances measured for 16 compass bearings using a geographical information system and
the maximum at each site selected. For statistical analysis, a log10 transformation of the
fetch values was used to ensure that the fetch variable approximated normal distribution.
Results
The PCA analysis produced a single factor with an eigenvalue greater than 1,
accounting for 61.6 percent of variability (Table A1). The factor scores for the first
component were taken as composite measures for the four fish community metrics. Each
of the constituent measures was highly correlated with the metric (Figure A1).
A linear regression of the fish community factor score as a function of log10
maximum effective fetch was highly significant (P<0.01):
Factor = 0.630 – 0.159 Log10 Max. Effective Fetch (r = 0.476, F=145.9, n=501)
However, the result showed considerable variability around the regression line (Figure
A2). There appeared to be a step drop in the level of the fish metric beyond a certain level
of maximum effective fetch.
A simpler analysis using K-means clustering supported the idea of dividing the
data into two clusters with a cut-off on the fetch axis at 10 km (Tables A2 and A3). The
ANOVAs for separation of factor and fetch data were highly significant (Table A2). If
the fish metric mean of the low fetch cluster is used as a reference point, the high fetch
cluster’s fish metric mean has a ratio of 0.64, meaning fish community measures at high
fetch sites are approximately 64 percent of those at low fetch sites (based on Table A3).
Discussion and Recommendation
The results obtained here are consistent with those obtained in previous reported
analyses. They provide a basis for modifying scenarios used with the Habitat Suitability
Matrix method to assess net change at wave-exposed sites on the Great Lakes, and in
other other large Ontario lakes such as Simcoe, Nipigon, Nipissing, etc., where winddriven wave processes are a major forcing effect on exposed shorelines.
By default the Condition Index for all habitat patches affected by a marina
development are given a value of 1 (Table A4). Areas of loss (LOSS) keep this default
condition as losses are always charged at the maximum rate for net change calculations.
Areas of direct modification (MODD) and any modified compensation area (COMM) on
the windward side due to the footprint of the breakwall should be assigned a value of
0.64 in both pre- and post-scenarios to reflect the continuing presence of wave exposure
reducing the value of the habitat to fish. The slope areas on the leeward side of the
breakwall (also MODD) and areas in the lee of the breakwall but not on the footprint
itself are considered to be indirectly modified (MODI) are assigned the high exposure
value in the pre-development scenario (0.64) and the low exposure value in the postdevelopment scenario (1.00) to reflect the change in habitat suitability induced by
reducing the wave exposure. In straightforward situations, it should be simple for
proponent and regulator to agree on the extent of this MODI area. Otherwise, an exact
analysis of fetch changes will be necessary to delimit the MODI area where maximum
effective fetch shifts from > 10 km to < 10 km. Any compensation areas in the lee of the
breakwall where habitats are modified to enhance productivity (COMM), i.e. through the
addition of shallow cribs supporting submergent or emergent vegetation or of other cover
65
types, can be treated as MODI areas. As created compensation areas (COMC) are
expected to make a positive contribution to the net change balance as long as they are
positioned in the lee of the breakwall, a Condition Index of 1.00 should be assigned. The
results of analyses with and without the application of the Condition Index should be
reported as part of the habitat assessment under Section 35 of the Fisheries Act.
This approach reflects the superficial situation seen around breakwalls, although
whether breakwalls, like reefs, enhance habitat productivity or merely act as attractants
and concentrators for fish from surrounding areas is unknown. This situation prevails,
especially in marine ecosystems, despite many decades of research. Should research
subsequently reveal that the main impact is as an attractant, the rationale for making the
adjustments in Condition Index laid out for HSM analyses of breakwalls would be greatly
weakened and the indices may no longer be applicable. The contribution of habitats to the
natural productivity of fish communities must remain the primary basis for assessing net
gain or loss of productive capacity.
66
Table A1 Results of principle components analysis to obtain a factor representative of
several fish community metrics using 501 electrofishing samples from various shore
locations in the lower Great Lakes.
Factor
Eigenvalue
Coefficients
Variables
Log10 Biomass
Log10 Numbers
No. species
Adjusted IBI
1
2.464
2
0.939
3
0.333
4
0.264
0.831
0.856
0.884
0.509
Table A2 Results of the analyses of variance for the K-means clustering.
Variable
Log Max EF
Fish factor
Total
Between SS
154.31
3.33
157.64
df
1
1
2
Within SS
47.56
19.41
66.97
df
499
499
F-ratio
1619.02
85.65
Table A3 The results of the K-means clustering of sample data (Figure A2) using log10
maximum effective fetch and the standardized fish metric.
Cluster
1
2
Variable
Max. effective fetch
Fish metric
Max. effective fetch
Fish metric
Minimum
-0.42
0.00
1.00
0.00
Mean
0.32
0.58
1.70
0.37
Maximum
1.00
1.00
2.04
0.91
Std.Dev.
0.32
0.18
0.26
0.25
Table A4 Rules for assigning Condition Index values in analyses of breakwall projects on
highly exposed shorelines of large lakes.
Area type
LOSS
MODD windward
MODD leeward
MODI leeward
COMM leeward
COMM windward
COMC
Pre-development scenario
1.00
0.64
0.64
0.64
0.64
0.64
n.a
67
Post-development scenario
n.a.
0.64
1.00
1.00
1.00
0.64
1.00
n
398
103
12
7
10
6
Log10(NUMBERS)
Log10(BIOMASS)
Figure A1 Plots showing the relationship between the set of individual fish metrics
(number of species, log numbers, log biomass and adjusted IBI) and the standard factor
score obtained with the first principal component.
8
6
4
3
1
0.2 0.4 0.6 0.8 1.0
Fish Community Metric
0
0.0
1.2
16
0.2 0.4 0.6 0.8 1.0
Fish Community Metric
1.2
0.2 0.4 0.6 0.8 1.0
Fish Community Metric
1.2
100
90
80
12
Adjusted IBI
Number of Species
4
2
2
0
0.0
5
8
70
60
50
40
30
4
20
10
0
0.0
0.2 0.4 0.6 0.8 1.0
Fish Community Metric
0
0.0
1.2
68
Figure A2 Plot of the fish community metric based on the PCA versus log10 maximum
effective fetch, km, for 501 electrofishing samples. The edge bar diagrams show the
frequency distributions for the two data variables and the dashed lines show the midpoints for the two clusters and the fetch cut-off for the two clusters.
Fish Community Metric
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-1
0
1
2
3
Log 10 Maximum Effective Fetch
69
Appendix B: Pre- and post-development scenario habitat data sets for the two case studies presented in section 4. Area types
are described in Table 2.3.2.
Application:
Brant Inn Node
Wallik property
Pre-development habitat scenario
; OPTNPRE
; Option Pre-development
Post-development habitat scenario
; OPTNPOST
; Option Post-development
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
01, 200, LOSS, 0.0, SAND, NOVEG
02, 1800, LOSS, 0.0, BEDROCK, NOVEG
03, 300, LOSS, 0.5, BEDROCK, NOVEG
04, 120, LOSS, 0.5, SAND, NOVEG
05, 270, LOSS, 0.5, SAND, NOVEG
06, 1260, MODD, 0.5, BEDROCK, NOVEG
07, 140, MODD, 0.5, SAND, NOVEG
08, 320, MODD, 0.5, BEDROCK, NOVEG
09, 115, MODD, 0.5, SAND, NOVEG
10, 318, MODD, 0.5, SAND, NOVEG
11, 50, MODD, 1.0, SAND, NOVEG
06, 1260, MODD, 0.5, COBBLE, NOVEG
07, 140, MODD, 0.5, COBBLE, NOVEG
08, 320, MODD, 0.5, ARMOUR, NOVEG
09, 115, MODD, 0.5, ARMOUR, NOVEG
10, 318, MODD, 0.5, ARMOUR, NOVEG
11, 50, MODD, 0.5, ARMCOB, NOVEG
12, 1000, COMC, 1.0, SILTYSAND, VEGGY
; WALLPRE
; Case 2 Pre High Water
; WALLPST
; Case 2 Post High Water
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
* UnitType=Area
* Units=m2
* Order=ID, Area, AreaType, Depth, Substrate, Vegetation
01, 14, LOSS, 1.0, BEDROCK, NOVEG
02, 46, LOSS, 1.0, BEDROCK, NOVEG
03, 17, LOSS, 1.0, BEDROCK, NOVEG
04, 40, MODD, 1.0, BEDROCK, NOVEG
05, 130, MODD, 1.0, BEDROCK, NOVEG
06, 80, MODI, 1.0, BEDROCK, NOVEG
04, 40, MODD, 1.0, BOULDER, NOVEG
05, 130, MODD, 1.0, COBBLE, NOVEG
06, 80, MODI, 1.0, COBBLE, NOVEG
70
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