PACIFIC COASTAL RESOURCES ATLAS

FOR

BRITISH COLUMBIA

By

Jacqueline Booth and Associates

Brad Mason, Fisheries and Oceans Canada

Michael Berry ICNRC / CMN

Mark Zacharias MSRM

Suzanne Richer

Table of Contents

The Pacific Coastal Resources Atlas (PCRA) for British Columbia, Canada, was developed in response to the expressed need for an easily accessible information source that can be used by anyone with an interest in Coastal Marine Resources Planning.  Government agencies, Regional and Community Governments and NGO’s are able to access the best available coastal resources datasets; all interest groups will be “working from the same page” of information.  The system allows authorized users to add new information (in a timely manner) as it becomes available using “on-line” data entry tools.

To date, both freshwater and coastal resources fishery information has been housed in a huge variety of data warehouses; some of these are digital, some are available in hardcopy form only, and much of our coastal resources knowledge still remains in the form of unwritten Traditional Ecological Knowledge.  Amalgamating these sources around a single access point (using access agreements and links to a variety of data servers) will streamline coastal planning processes.  The benefits to Governments and Communities are significant; requests for existing information from Government agencies will be greatly reduced (with associated cost savings) – community members will have a greater sense of ownership and stewardship of natural resources and will be better informed when resource use conflict decisions are being made.

A number of government and non-government web sites distribute selected coastal resources data and maps specific to a particular species or habitat.  The data presented is often in different formats and, on occasion, may not be consistent.  The Community Mapping Network (CMN) initiative is presently the only system that allows password-authorized entry of new information using web-based mapping tools.  Although there is a lot of information about British Columbia’s coastal resources in the various data warehouses, there is a much larger area of the BC coast for which no information exists!  The cost of doing detailed surveys in this huge coastal area would be phenomenal, (at a recent conference at the Institute of Ocean Sciences one participant estimated it would take 60+ years using the best available technology just to accurately map the BC coastal seafloor and inshore).  The ability to map additional coastal resources using local knowledge (including new agency-sourced information) will greatly assist BC residents in making responsible resource use decisions.

The importance of metadata used in the PCRA approach is stressed throughout the following system description.  The tools described allow for the input of new information that will come from a variety of sources with an associated variety of accuracy, (as is the case with most existing information sources).  The system allows anyone who views the maps and data reports to review the sources of the information and the relative accuracy of the observations.  As time goes on, any information that has not been adequately field-truthed will be subjected to more detailed examination and up-dated appropriately.  In the meantime, “suspected presence” (of any species or habitat) is better than no information at all as coastal resource use planning decisions are being considered.

The Pacific Coastal Resources Atlas is designed to streamline the collection and dissemination of marine habitat and fishery resource information for coastal BC.  The goal of the system is to create an easily accessible source of spatially georeferenced marine habitat and resource information.  Although the individual databases and GIS layers may reside on different computers in a range of agency data warehouses, the PCRA system is designed to make all of the information accessible to information users and providers at one specific location through the internet.

There are many uses for spatially georeferenced marine habitat and fishery information.  Some examples of the business needs for this information include:

• Habitat referrals for foreshore leases and licences;
• Oil spill contingency planning and response;
• Community shoreline planning and zoning;
• Provincial coastal planning;
• Provincial and Federal Marine Protected Area analysis;
• Fisheries research which looks for correlation of marine resources with habitat.

The Pacific Coastal Resources Atlas is available through the Community Mapping Network (CMN).  The Community Mapping Network was created to share a wealth of natural resource information and maps with communities in British Columbia, Canada.  Government and community natural resource information is integrated and made accessible through a user friendly, interactive mapping system called Autodesk Mapguide.  A series of servers are utilized to share the workload for serving province-wide topographical base maps, high resolution orthophotography, Canadian Hydrographic Service Charts and selected resource information.  Maps and natural resource information are “web-served” to assist communities and local governments with landuse planning, to promote conservation and protection of sensitive habitats and to raise awareness and respect for ecological values.

Accuracy of the data

Information provided in the PCRA must be both accurate and timely.  Currently, data included in the system has been gathered from Fisheries and Oceans Canada (DFO) staff or individuals under contract to DFO who are known to have knowledge of the spatial distribution of one or more marine resource or habitats on the BC coast.  (see below for a description of data layers, data sources and metadata).  The majority of the data came from interviews with regional DFO habitat and fishery officers and local, knowledgeable coastal residents including First Nations.  A second, equally important source of information was the databases maintained by staff at the Pacific Biological Station and Regional offices.  For example, staff at the Pacific Biological Station manage a coastwide database of herring spawn.  It is hoped that in future, trained and certified individuals or organizations may enter information on coastal habitats and fishery resource themselves using online data entry tools developed through the Community Mapping Network.

The PCRA system allows the user to access the accuracy of both the spatial location of a feature (habitat/fishery/species) and the information about that feature by providing meta data in the data records about who collected the information, how it was collected and when it was collected.  Spatial accuracy is also indicated in specific fields in the database.

All data have been fitted to a TRIM coastline with depth contours, where incorporated, utilizing CHS charts.

The PCRA system has been designed to include a wide variety of information on the distribution, abundance and relative importance of marine flora and fauna, fish habitat and fisheries.  A complete list of all the data layers and their file names is included in Appendix A of this report.  The data are grouped into the following categories:

In addition previous compilation of fisheries resource information collected by DFO and/or the Provincial Government throughout the 1990s has been incorporated into the PCRA.

• Strait of Georgia  (1992/93)

• West Coast Vancouver Island  (1994/95)

• Johnstone Strait (1994/95)

• Central Coast (1994/95)

• North Coast (1997)

• North Island Straits/Quatsino (2002)

The data that has been incorporated into the Atlas has come from a number of sources.  These include data that have already been georeferenced into a GIS during past one time data compilation projects, existing georeferenced databases that are part of an ongoing data collection system, and new data that are coming in from certified data providers.  These data sources are summarized in Table 1 below and discussed in greater detail in the following sections.

Table 1: SHIM data sources.

In the late 1980s the Environmental Emergency Services Branch of the BC Ministry of Environment, began the process of compiling marine biological information for Biological sensitivity mapping for oil spill response.   This process continued through the 1990s (now under the provincial Land Use Co-ordination Office (LUCO) and has ended up the responsibility of Decision Support Services under the Ministry of Sustainable Resource Management (MSRM) in 2002.  These data compilations are now used for coast marine planning and impact assessment in addition to oil spill contingency planning and response. 

Much of the information for the provincial marine data compilation came from the Department of Fisheries and Oceans (DFO).  In 1993, the South Coast Division of DFO initiated a system to collect, organize and report information pertinent to oil spill planning from fisheries staff (Booth and Rueggeberg 1994).  DFO recognized that a significant proportion of the existing fisheries and fish habitat information for the BC coast comes from the personal knowledge and experience of district fisheries staff or on unpublished file material in district fisheries offices.  The program was therefore designed to collect and compile this information into a single database where it can be accessed in a timely manner.   Because the funding for this project came from a special fund set up after the Nestucca Oil Spill, the initial data compilation was focused on DFO statistical areas along the west coast of Vancouver Island (Booth and Rueggeberg 1994, Booth et al.  1995, Lightly, 1996, Lightly and Hillaby 1996).  The data sources used in the DFO compilations included:

• District fisheries staff familiar with the study area; the nature of their experience and number of years associated with the study area are noted, as indicators of the comprehensiveness of the data being provided.
• Reports based at least in part on personal knowledge of district fisheries staff; these include Records of Management Strategies, annual fisheries management strategies, and the Barkley Sound Planning Strategy.
• Existing databases based on information provided by district fisheries staff or from interviews with long-time coastal residents.

Using the DFO data collection model, a standard and methodology was used for all interviews and all interviews were conducted by a qualified fisheries biologist.  Each interviewee was provided with a series of Canadian Hydrographic Service charts covering their area of interest and was then asked to identify specific fish ^[2] habitats ^[3] , fish distributions ^[4] or areas of fishing activity.  Where practical, the largest scale charts available were used (usually 1:40,000 scale) used for compilation; however in the case of fisheries that covered a broad geographical extent (e.g.  offshore groundfish trawl and salmon troll fisheries) a smaller scale chart was used.  As each area was delineated, it was assigned an identifier and the interviewee was queried to collect relevant information concerning the area or fishery.  Interviews were conducted with at most one or two interviewees at a time.  When returning to an area for updates, the interviewees were asked to verify or comment on the old datasets.  Interviewees were also asked to search their files and archives for relevant hard copy or digital information that supplemented their verbal knowledge.

After completion of the interviews the maps and data forms were entered into a Geographic Information System and verified.  If there was overlap or discrepancies in the information obtained from different sources the data were scrutinized and in most instances were merged into a single observation with two sources.  The criteria used to make such decisions were (1) to give preference to the information obtained from the individual with the most direct knowledge and experience of the specific fishery, (2) to give preference to the information from the larger scale chart, and (3) to preserve detail.  In instances where one of the data points was part of a coastwide dataset (e.g. the distribution of herring spawn, clam beds or kelp beds) the coastwide dataset was kept intact and the information from other sources was maintained but kept as a separate entity.  These discrepancies were flagged to be sorted out by further field studies.

The data included in the marine fishery resources are a series of polygon, line or point themes with common data structures.  The spatial format of the theme was decided based on the precision of the georeferencing.  For example, seal and sea lion haul outs are represented by a point, coastline with herring spawn recorded as present as a line and areas where the offshore and along shore distribution of herring spawn has been mapped are represented by a polygon theme.  The majority of layers representing fisheries are represented as polygons. 

Each polygon, point or line represents an area where a particular habitat is located or type of fishery operates.  All records include information that allows the user to judge the accuracy of the mapped data (see details in following Section on Metadata).  The layers mapping the location of fishing areas included information on the timing and intensity of use specific to that location.  In most instances, a single fishery consists of a series of locations (several polygons) with specific timing and levels of resource use indicated for each location.  Fisheries targeting the same species but with differences in harvesting methods and/or timing ^[5] were kept as separate data layers.

The marine fishery resources of the BC Coast that have been included in the DFO compilations of Fisheries and Habitat Resource Information are listed in Appendix A.  For the most part the distribution of fish and invertebrates is inferred by the location of the fishery(s) for that species.  With the exception of herring spawning, clam beds and rivers that support anadromous fish stocks, the only systematic distribution information available is harvesting data.  While we do have some understanding of the habitat preferences of harvested species, it is insufficient to allow estimation of species distributions with any confidence.  The distributions of these fisheries resources have therefore been inferred from the intensity of harvesting.  The weakness of this approach is that harvesting intensity in some areas is a reflection of quotas and management controls.

A number of attributes are used to describe each data record. These include information on the location, timing and intensity of use as well as details on species, fishing gear used and other relevant information. Not all fields are used for all data layers, only those that are relevant. List all of the data description fields; the next section describes those data fields that describe how the information was collected and its accuracy and precision.

Table 2: Fields used to describe attributes of areas represented by individual data records.

The layer used for identifying kelp beds has an additional fields to indicate if the area has been surveyed by the province in the past. This field “SURVEYED” is a logical field and can be entered by the compiler or by the database manager at a later date.

The following table (Table 3 ) lists the individual logical fields used to indicate which species are taken at the location described by a data record. This approach is only used in mixed species fisheries.

Table 3 : Species selections and other fields for individual fisheries or resources.

The layer HSpawn_Pnt is based on a DFO in house dataset which references all observed occurrences of herring spawn since 1928 referenced to points which roughly divide the coast into 1 km segments. These points have been moved so that they are georeferenced to the provincial TRIM coastline. The structure of this database reflects the data recorded by the DFO Herring Section at the Pacific Biological Station and is listed below.

Table 4: Fields in the layer for herring spawn as represented by points (HSpawn_Pnt).

A separate data layer Hspawn_Shore, was created to amalgamate all of the herring spawn into one layer referenced to shoreline segments. This layer is created by the manager of the Atlas and includes all available information on the location, timing, intensity and frequency of herring spawn. It is intended to be viewed but not updated by users of the atlas. The layer has all of the fields included in the layer Hspawn_Pnt as well as the following two additional fields:

         FOLOC_NO which links the coastal line segment to the layer HSpawn_FO_shore

         KMLOCATION which links the coastal line segment to the layer HSpawn_Pnt

The layer Groundfish_Trawl has a unique field structure as it tries to capture the nature of the trawl fishery which includes seasonal fishing to spread effort over the year and mixed species.

Table 5: Fields in the layer for Groundfish Trawl Fisheries (Groundfish_Trawl).

The layers Salmon_Troll and Salmon_Net are also unique in that they have the capability of identifying the run (natal river) of salmon being targeted. This information can be important as openings for salmon fisheries are typically designed around the movements of returning salmon. The fields unique to these layers include:

Table 6: Fields in the layers for commercial salmon fisheries (Salmon_Net, Salmon_Troll)

The seal and sea lion layer databases are designed with a unique structure to capture information on the behaviour and age structure (pups vs. adults) at individual locations. The fields for the point layers describing concentrations of seals or sea lion concentrations include:

Table 7: Fields used to describe Seal and Sea Lion concentrations.

The RI ratings for seals and sea lions were based on the maximum number of animals observed at a site. The ratings were decided based on the typical numbers seen at sites coastwide.

The activity of the seals or sea lions around the site is selected from a pick list that includes the following choices:

Metadata attached to each data record provide information that enables the user to judge the accuracy and precision of that record.  The metadata includes the following fields that are described in more detail below:

Table 8: Fields used to describe the temporal, spatial and biological accuracy and precision of the data.

Accuracy

The accuracy with which the data has been mapped is recorded as a subjected estimate in this field.  If the data are added from an existing database, or if the data are digitized from hard copy CHS charts or other maps, then the following pick list is provided for selecting the data accuracy:

The chart scale is linked to the CHS chart number that is recorded in a separate field.  A list of all CHS charts and the associated scale are included in the Atlas database.

If, however, the data are entered on line via screen digitizing at the SHIM web site (and are therefore not linked to the accuracy of a chart or GPS, the following choices are provided for recording the spatial accuracy of a mapped location:

• General Location only (1 km accuracy)

• Good alongshore accuracy (10-20 m), offshore extent unknown or to 250 m

• Good alongshore and offshore accuracy (10-20 m in both directions)

Obs_Method

The method used to georeference and describe the area is also chosen from a pick list with the following choices:

“Personal observations” are differentiated from “field work, measured directly” in that the first is an incidental observation while the latter is a dedicated field survey such as a habitat assessment or herring spawn dive survey.  “Traditional knowledge” implies that it is a collective knowledge of a group of person such as a First Nations group.  Both “second hand knowledge” and “published and compiled information” are one step removed from the original observer; the difference lies in that the former is verbal while the latter is based on a report, map or database.  If the observation method is based on “published or compiled information” the details of source of that information should be contained in the “Source” field of the data record.

Field_Method

If the field “Obs_Method” is indicated as “Field_Work” (code 1) ^[6] then the nature of the field work should be indicated. 

Field_Sampling_Method

If the field “Obs_Method” is indicated as “Field_Work” (code 1) then if field samples were collected the nature of these samples should be indicated.  If no samples were taken or if the observation is not based on a field survey then the field should be left at its default value of “N/A” (code 0). 

Photo

This is a logical field (Yes/No) that indicates if a photo has been taken of the habitat, fish or fishery being mapped at the location of a data record.  This is provided as an additional method of data verification.  Provision is made on the on line data entry site of the Atlas for submitting a digital photo over the Internet.  This photo can then be displayed when the data record is queried.

GPS

This is a logical field (Yes/No) that indicates whether or not a GPS was used to georeference the spatial extents of the data record.  It is intended that in the near future, provision will be made on the Atlas on line data entry site for submitting the spatial extents of a data record as a GPS file.  At present this information gives some indication of the spatial accuracy of a data record and the spatial extents and attributes of records can be submitted digitally in standard GIS format.

GPS_Differential

This is a logical field (Yes/No) that indicates whether or not the GPS used had differential correction either in the field or back in the office.

GPS_Model

If a GPS was used in georeferencing the spatial extents of the data record, then the make and model number, plus any user modifications, should be indicated in this field.

Source

A main source (Source_1) and up to two corroborating sources (Source_2, Source_3) can be entered for every data record.  A source can be a person, published or unpublished report or map, or a database.  The details of each source must be added prior to use in a separate data file “Sources”.  This file has the following data fields:

The pick list for the sector to which the source belongs is given below.

Data_Dte

Date of when the observations or field studies were made.  May be a range of years if it is based on a time series of data or a period over which observations were made.

Map_Dte

Date when data were first gathered by Compiler from the data source.  If the data were already in the atlas but have been updated or verified since the initial compilation, then this date should be recorded. 

Compiler

Name of person responsible for compiling the information and submitting it to the Atlas.  In order to submit information to the Atlas, a person must be certified by DFO.  The certification process is under development but will include a training program and requires a certain degree of expertise or knowledge in marine resources, GIS and data management.  Once certified an individual’s name is added to the list they are provided with a user id and password for on line data entry over the CMN/PCRA web site.  A data table with metadata on each compiler is required before a compiler can enter data.  The fields in this database include the following details that indicate to the user who the compiler is and how to contact them.

•          Name of compiler

•          Position of compiler in organization

•          Organization compiler belongs to

•          Postal Address

•          Email address

•          Phone number

Other independent datasets managed and compiled by other programs.