iSEA Survey of recreational fishing in British Columbia -
survey description and methodology
This document describes the data sources, survey design, and analytical methods used to produce estimates of recreational fishing effort and expenditures in British Columbia based on the Internet Socioeconomic analysis (iSEA) Tidal Water and Freshwater Recreational Fishing Surveys. It outlines how survey samples are drawn from licensing databases, how raw survey responses are cleaned and integrated with administrative data, and how fishing effort and expenditures are distributed across time, species, locations, and trip types. The document also describes the weighting procedures used to generate population‑level estimates and the metrics applied to assess the reliability of reported results. The objective is to provide transparency and guidance for interpreting and using the published iSEA recreational fishing data and reports.
Data description
The primary data sources described in this document are the iSEA Survey of Tidal Water Recreational Fishing in British Columbia (“Tidal Water Survey”) and the iSEA Survey of Freshwater Recreational Fishing in British Columbia (“Freshwater Survey”). The survey questions are organized into two sections – fishing effort (days), and fishing expenditures. Detailed descriptions of each survey are outlined below.
Tidal water survey
The Tidal Water Survey is administered by the Fisheries and Oceans Canada (DFO) Pacific Economic Analysis Unit and collects socioeconomic information on tidal recreational fishing activity in British Columbia (BC). To date, the Tidal Water Survey has collected recreational effort and expenditure data for the 2020, 2021, 2022, 2024 and 2025 fishing seasons. A similar pilot survey was conducted for the 2019 fishing season, however this survey was focused on collecting effort data only and excludes any information on fishing expenditures.
Each year, a stratified Footnote 1 random sample of tidal water recreational licence holders is selected from the DFO National Recreational Licensing System (NRLS) database to complete the survey. Only individuals who provide an email address when they purchased a licence are sampled. About 30% of licences are purchased from Independent Access Providers (IAP) each year, the remainder are purchased online. When purchasing a licence through an IAP, it is optional to provide an email address. As a result, about 80% of licence holders provide DFO with an email and the remaining 20% of licence holders cannot be sampled. Additionally, the sample excludes Juvenile and Family Fishing Weekend licences, due to limitations related to surveying minors in BC. Juvenile and Family Weekend licences represent around 13% of all licence sales. Each year, approximately 30% of sampled individuals provide complete responses to the survey. Detailed sample and response metrics for each year are provided in Table 1.
Table 1. Tidal water survey sample and response metrics.
| Fishing season | |||||
|---|---|---|---|---|---|
| 2025 - 2026 | 2023 - 2024 | 2022 - 2023 | 2021 - 2022 | 2020 - 2021 | |
| Total licence population Footnote 2 | 351,177 | 313,223 | 295,593 | 272,323 | 238,600 |
| Initial sample size | 50,000 | 45,000 | 40,000 | 20,000 | 20,000 |
| Complete responses | 12,754 | 12,030 | 11,545 | 5,623 | 6,461 |
| Response rate | 26% | 27% | 29% | 29% | 33% |
| % of licence population | 4% | 4% | 4% | 2% | 3% |
Freshwater survey
This Freshwater Survey is jointly administered by the Province of British Columbia, Ministry of Water Lands and Resource Stewardship, and the DFO Pacific Region Economic Analysis Unit, and collects socioeconomic information on freshwater recreational fishing activity in British Columbia (BC). To date, the Freshwater Survey has collected recreational effort and expenditure data for the 2022, 2023, 2024 and 2025 fishing seasons.
Each year, a stratified Footnote 1 random sample of 100,000 freshwater recreational licence holders is selected from the Province of British Columbia’s recreational licensing database to complete the survey. Only individuals who provide an email address and consent to email solicitation when they purchased a licence are sampled. Each year, approximately 33% of individuals indicate that they wish to not be contacted for surveys or other matters and are excluded from the sample. Additionally, the sample excludes Family Fishing Weekend licences. Fishers below the age of 16 are not required to purchase a licence to fish in freshwater in British Columbia, and are therefore implicitly excluded from the sample. Each year, approximately 20% of sampled individuals provide complete responses to the survey. Detailed sample and response metrics for each year are provided in Table 2.
Table 2. Freshwater survey sample and response metrics.
| Fishing season | ||||
|---|---|---|---|---|
| 2025 - 2026 | 2024 - 2025 | 2023 - 2024 | 2022 - 2023 | |
| Total licence population Footnote 2 | 381,168 | 357,232 | 352,019 | 346,821 |
| Initial sample size | 100,000 | 100,000 | 100,000 | 100,000 |
| Complete responses | 18,306 | 17,556 | 20,880 | 18,384 |
| Response rate | 18% | 18% | 21% | 18% |
| % of licence population | 5% | 5% | 6% | 5% |
Data cleaning and preparation methodology
Following the closure of data collection, raw survey response files are subjected to a structured, multi‑stage data‑cleaning process to produce a transformed dataset which can be used for analyses. The objective of this process is to resolve inconsistencies and issues inherent in self‑reported survey data, and prepare variables for population‑level estimation. With a few exceptions, these steps are applied consistently across the Tidal Water and Freshwater Surveys. All data‑cleaning and transformation steps are programmed to ensure consistent implementation across survey years. The detailed methodology for producing the cleaned dataset is outlined below.
Data format correcting
- Variable names and labels are standardized across survey years.
- Invalid placeholder values (eg. “999” used to represent missing values) are recoded as missing.
- Character encoding issues (e.g. language formatting inconsistencies in text fields) are corrected to ensure consistent string matching.
Ensuring record completeness and response validity
- Survey responses are screened for completeness using a standard exclusion protocol applied consistently across years. Records were removed if they failed to meet minimum data requirements necessary for analysis, such as:
- Missing responses for key fishing effort questions
- Missing identifiers required for weighting
- Incomplete surveys where respondents exited before providing effort or expenditure data
Harmonization of residency and licence attributes
- Residency and licence attributes are aligned across survey and administrative data sources:
- Residency status is derived using respondents’ province/state and country of residence from their licence records rather than relying solely on self‑reported residency fields.
- Open‑text province and country entries is standardized to account for multiple spelling and formatting variations (e.g. “BC”, “B.C.”, “British Columbia”).
- Licence characteristics (licence type and salmon stamp status) are merged from licensing databases using respondent identifiers.
- This approach ensures consistency with administrative data and reduced classification error.
Geographic data cleaning and waterbody assignment (freshwater survey only)
- In the Freshwater Survey, respondents report their fishing locations using one of two methods:
- Selecting a named waterbody from a list
- Clicking any point on an interactive map
- For map‑based selections (method 2), only GPS coordinates and the corresponding Wildlife Management Units (WMU) (no waterbody names) are recorded. To assign the point to the appropriate waterbody, a matching procedure is applied externally using GIS software as follows:
- Survey points are overlaid with official waterbody shapefiles.
- Waterbody names are assigned using a hierarchical distance based matching process (direct overlap, within 1 km, and within 10 km).
- Points beyond 10 km from any mapped waterbody are classified as “Unspecified” and excluded from spatial analysis due to high uncertainty.
- Matched waterbody names and associated management units are then merged into the survey dataset.
Respondent weight calculation
- Expansion of survey results to reflect population-level effort and expenditure estimates requires calculation of survey weights to be applied during analysis. To calculate survey weights, the following process is completed:
- Respondent records are stratified and counted by residency, licence type, and salmon stamp status.
- The number of licence holders in the population are also counted, stratified by the same attributes.
- Respondent and population counts are merged to calculate the respondent weight, as population count divided by respondent count.
- The weight effectively reflects the number of individuals in the population that each respondent “represents” based on residency, licence type and salmon stamp attributes.
- Separate weighting variables are prepared for fishing effort analyses and expenditure analyses to account for partial survey completion (fewer complete responses to the expenditure section compared to the effort section).
- These weights are included in the cleaned dataset for use in analysis.
Expenditures data cleaning and outlier treatment
- Fishing expenditures require additional processing and data cleaning:
- Major purchase spending attributable to fishing (“Major Attributable Expenditures”) is calculated as: spending on each major purchase item (e.g. on boats, fishing rods, etc.) multiplied by the reported percentage of that item which is attributable to recreational fishing in tidal/freshwater.
- Total Major (Attributable), Direct and Package expenditures are calculated by summing across detailed spending categories.
- Extremely large expenditure values are identified as statistical outliers. Rather than excluding these observations, outlier responses are retained but assigned a respondent weight of “1” to prevent disproportionate influence on population estimates.
- Expenditure weights are re-calculated for the smaller (outlier removed) expenditure sample using the weighting process described above.
Distribution of fishing effort by month, species, location, and trip type
- The survey collects the following fishing effort information from respondents:
- Fishing days by month
- Fishing days by species (annual)
- Species fished by month
- Locations (tidal subareas or freshwater waterbodies) fished by month
- Fishing days by trip type (annual)
- Fishing days by month is treated as the primary (“true”) measure of total effort. The remaining effort measures are reconciled to produce consistent estimates, such that when fishing days is aggregated by any combination of month, species, location, or trip type, they sum to the same total.
- The reconciliation and distribution process proceeds as follows:
- Allocation of species effort within months: For each respondent, annual fishing days reported by species are first allocated across months in which each species was reported as being fished. Where a species was reported in multiple months, effort is initially distributed evenly across those months. Species month proportions are then calculated to represent the relative share of fishing effort directed toward each species within a given month.
- Calculation of species days by month: The species month proportions are applied to the reported fishing days by month to estimate fishing days by month and species. This ensures that total fishing days for each month remain consistent with the respondent’s reported monthly totals.
- Allocation of fishing effort across locations: For each respondent and month, fishing effort is distributed evenly across all reported fishing locations (tidal subareas or freshwater waterbodies) for that month. Location proportions are calculated based on the number of distinct locations selected, and applied to the species month fishing day estimates.
- Allocation of fishing effort across trip types: Fishing days reported annually by trip type are converted into proportions of total trip type days for each respondent. These proportions are applied to the month species location fishing days to distribute effort across trip types while preserving total reported effort.
- Creation of distributed effort records: Combining the above steps produces unweighted fishing day estimates by respondent, month, species, location, and trip type. Internal checks are undertaken to confirm that, for each respondent and month, distributed fishing days sum back to the reported monthly totals.
- Weighting for population level estimates: Respondent level survey weights calculated previously in the data cleaning stage are applied to the distributed fishing day estimates to generate population level fishing effort. Both weighted and unweighted estimates are retained in the final datasets to support analysis, validation, and calculation of reliability (coefficient of variation) metrics.
Distribution of fishing expenditures by month, species, location, and trip type
- Fishing expenditures are reported by respondents at an aggregate level (annual totals) and are not directly attributed to specific months, species, locations, or trip types at the time of data collection. To enable analysis consistent with distributed fishing effort estimates, reported expenditures are allocated across these dimensions using respondents’ distributed fishing days.
- The expenditure distribution process proceeds as follows:
- Preparation of expenditure totals: Annual fishing expenditures are calculated for each respondent, including:
- Major purchases attributable to recreational fishing
- Direct trip related expenditures
- Package expenditures
- Calculation of effort based allocation proportions: From the fishing effort distribution process, each respondent’s (unweighted) fishing days are distributed by month, species, location, and trip type. For each respondent, the proportion of total fishing days represented by each combination of these categories is calculated.
- Allocation of expenditures across month, species, location, and trip-type categories: Reported annual expenditures are multiplied by the corresponding fishing day proportions to estimate expenditures by month, species, location, and trip type. This ensures that expenditure allocations are directly proportional to fishing effort.
- Consistency of totals: Allocated expenditures are checked to ensure that, for each respondent, expenditures distributed across all categories sum back to the respondent’s total reported expenditure.
- Generation of distributed expenditure records: This process produces unweighted expenditure estimates by respondent, month, species, location (tidal subarea or freshwater waterbody), and trip type. Both total and category specific expenditure variables are retained for analysis.
- Weighting for population level estimates: expenditure survey weights calculated previously in the data cleaning stage are applied to the distributed expenditure estimates to generate population level expenditures. Both weighted and unweighted estimates are retained in the final datasets to support analysis, validation, and calculation of reliability metrics.
- Integration with effort data: Distributed expenditure outputs are structured to align directly with the distributed fishing days dataset, enabling integrated analysis of fishing effort and economic activity across consistent dimensions. Although for some respondents who did not complete the expenditure section, data will be missing for expenditure estimates.
- Preparation of expenditure totals: Annual fishing expenditures are calculated for each respondent, including:
Additional notes on methodology
- Stratification takes place both at the sampling stage and the weighting stage. Detailed descriptions of strata used for each stage are as follows:
- Sampling: The licence population is stratified into four groups by simplified licence type (annual or term) and simplified residency (BC resident or non-BC resident) prior to sampling.
- Weighting: In total there are 60 strata used in the Tidal Water Survey weighting process and 36 strata in the freshwater weighting process. The following categories are used in the stratification process for establishing respondent weights.
- Detailed Residency: BC Resident, Alberta Resident, Washington Resident, Other Canadian Resident, Other USA Resident, Other International Resident.
- Detailed Licence Type: Annual Adult, Annual Senior, 1-day, 3-day, 5-day (tidal) or Annual, 1-day, 8-day (freshwater)
- Salmon Stamp Status: binary metric indicating whether or not the respondent purchased a salmon stamp along with their licence
- The analysis relies on some fisher and licence attributes derived from the licensing data. The licence data must be requested from the DFO NRLS (tidal), or Province of BC Recreational Wilding Information and Licensing Data (WILD) system (freshwater). Variables required from the licensing data include:
- Salmon Stamp Status
- Province or State of Residency
- Country of Residency
- Licence Residency (Resident = BC Resident; Non-resident = Other Canadian or International)
- Licence type (e.g. Annual, 1-day term, 3-day term etc.)
- GDP, Household Income, and Jobs statistics are calculated using multipliers from British Columbia’s Fisheries and Aquaculture Sector, 2022 Edition Report, developed by Lillian Hallin (2024).
- Regions in the Tidal Water Survey are defined to align with the DFO’s Survey of Recreational Fishing in Canada: Province of British Columbia tidal waters region map. This survey was last conducted in 2015. Pacific Fisheries Management Areas do not align perfectly to the survey map. Nearest visual matches were used. Regions are defined as follows.
- Barkley Sound: PFMAs 20, 21, 22, 23, 121, 123
- Central Coast: PFMAs 6, 7, 8, 9, 10, 106, 107, 108, 109, 110
- Haida Gwaii: PFMAs 1, 2, 101, 102, 130, 142
- Johnston Strait: PFMAs 11, 12, 13, 111
- North Coast: PFMAs 3, 4, 5, 103, 104, 105
- Strait of Georgia: PFMAs 14, 15, 16, 17, 18, 19, 28, 29
- West Coast Vancouver Island: PFMAs 24, 25, 26, 27, 124, 125, 126, 127
Survey results reliability
A Coefficient of Variation (CV) is calculated with each statistical result (when possible) to describe how consistent (or reliable) survey observations are. Survey observations are weighted and inflated to the populations level. Prior to the use of weighted results, the reliability of the survey results is tested by calculating the CV.
The CV looks at all the observations at a given month, species, location, and trip type aggregation level (eg. all the angling days reported for Chinook in August). It calculates how much those observations vary from their own mean. A low CV indicates that the observations are fairly consistent around their mean and reliable when weighted to the population level. A high CV indicates a lot of variation in the observations such that results may not be reliable when weighted to the population level. Based on expert adviceFootnote 3, results with CVs under 35% are considered acceptable to report for the purposes of this analysis, while results with CVs above 35% are not reported.
For the purposes of this analysis, the following guidelines for use of results are as followsFootnote 3:
- CV Less than 5.00%: Data can be used without condition
- CV 5.01% to 15.00%: Data can be used with caution
- CV 15.01% to 35.00%: Data should be used with caution
- CV Greater than 35.00%: Data should be used with extreme caution
CVs are calculated based on the transformed, unweighted data, produced from the data cleaning and distribution process described above in the methodology section. The underlying assumption of the transformed data used for CV calculations is that the calculated days and expenditures are representative of each individual respondent's activities.
Concepts, terms and definitions used in iSEA recreational survey reports
- Active fisher:
- An individual who fished in the corresponding fishing season; the identification of recreational fishers was based on the information from licensing databases and active fishers were estimated based on the survey respondents.
- BC resident:
- An individual who lived within British Columbia during the survey year.
- Direct expenditures:
- Expenditures on goods and services (food, camping, accommodation, transportation, supplies, etc.) incurred during fishing trips or excursions.
- Fishing day:
- A day or part of a day on which an individual fished for recreation.
- International fisher:
- An individual who lived outside of Canada during the survey year.
- Major purchases:
- The purchase of durable goods and investments in support of recreational fishing activities. These include boats, motors, special vehicles, camping gear and other durable goods, as well as land and building purchases.
Major purchases are broken down into two categories: (1) major purchases attributable in whole or in part to recreational fishing, defined as the total purchase price of the durable goods mentioned above; and (2) major purchases wholly attributable to recreational fishing, calculated based on the percentage of the purchase price which the respondent feels was directly attributable to recreational fishing. For the purposes of this report, only major purchases wholly attributable to recreational fishing are presented. - Other Canadian:
- An individual who lived within Canada, but outside British Columbia during the survey year.
- Package expenditures:
- Monies paid to a lodge or outfitter for a complete range of services such as lodging, food, transportation, use of fishing equipment, etc. Four types of packages were defined for the purposes of this survey:
- Charter boat package:
- A package which includes fishing in a boat and may include other goods or services such as a guide, food, or fishing supplies, but excludes accommodation and airfare.
- Fly-in package:
- A package which includes airfare and may include other goods or services such as accommodations, a guided charter, food, or fishing supplies
- Lodge package:
- A package which includes accommodation and may include other goods or services such as a guided charter, food, or fishing supplies, but excludes airfare.
- Other:
- A package which does not include fishing in a boat, accommodations, or airfare, but may include other goods or services such as a guide, food, or fishing supplies
- Percent change:
- calculated as 𝑟 = ((𝑝𝑛 − 𝑝0)/𝑝0) ∗ 100 where 𝑟 is average annual percent change, 𝑝𝑛 and 𝑝0 are the final and initial observations in the period, respectively.
- Recreational fishing:
- refers to non-commercial fishing; recreational angling and sport fishing are covered in the definition used by the survey. Note that food, social and ceremonial fishing and subsistence fishing are not covered by the survey and are, therefore, not included in any of the estimates presented in this report.
- Visitor:
- An individual who’s primary residence is outside of BC (includes Other Canadian and International fishers).
Footnotes
Related links
- Freshwater recreational fisheries dashboards
- Tidal water recreational fisheries dashboards
- Results of the Pacific tidal recreational fishing survey, 2023
- Results of the British Columbia freshwater recreational fishing survey, 2024
- Internet Socio-Economic (iSEA) survey of freshwater recreational fishing in B.C.
- Internet Socio-Economic (iSEA) survey of tidal water recreational fishing in B.C.
Contact us
Email: iSEA@dfo-mpo.gc.ca
- Date modified: