Forest Investment Account (FIA) - Forest Science Program
FIA Project Y081231

    Developing Fish Habitat Models for Broad-Scale Forest Planning in the BC Southern Interior
Project lead: Porter, Marc (ESSA Technologies Ltd.)
Contributing Authors: Porter, Marc; Pickard, Darcy; Wieckowski, Katherine; Bryan, Katy
Subject: Forest Investment Account (FIA), British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
Knowledge of fish distribution in BC has evolved since McPhail & Carveth (1993) first described fish distribution and provincial fish communities using eight zoogeographic regions (subsequently defined in GIS, Porter 2004). Since that time, the province’s fish occurrence databases have been used to develop empirical models to predict distributions for 37 fish species (Porter & Frid 2004; Porter & Nelitz. 2006). These models are currently being used by the BC MOE to assist in identifying Fisheries Sensitive Watersheds (FSWs) that require additional conservation and management. A limitation of these models, however, is that they only predict fish occurrences for 3rd order watershed polygons. In addition, they do not assess the relative quality of habitats across streams or reaches.
This proposal intends to employ recent improvements to the intelligence of the province’s Watershed Atlas stream linework and newly developed provincial fish datasets to develop improved models for predicting the distribution and abundance of key fish species at finer spatial scales. The goal of this work is to improve fish distribution models and quantify species specific habitat quality in order to assist planning of forestry activities and optimize associated field inventories.
To achieve this goal, we propose the following work plan: (1) Hold model design meetings, conduct preliminary design, and elicit feedback. Several project team meetings will be convened to develop a detailed approach to improved fish habitat modelling. Background research, internal team discussions and consultation with MOE fisheries researchers will focus on identifying: (1)a pilot study area, (2)fish species of interest (salmon, trout and char would be an expected focus), (3)critical habitat features to distinguish relative habitat values and (4)habitat features that could be calculated using available data. Reach-scale indicators would be prioritized based on their scientific defensibility, feasibility and spatial extent. A detailed description of the model design will be sent to external fish habitat experts for peer review [Deliverable 1]. (2) Calculate reach-scale indicators. Based on the priorities from Task 1, selected reach-scale indicators will be calculated within the study area for relevant species. Where data are available and calculations are feasible, a GIS-based network analysis will be used to calculate reach-scale indicators such as: (1)stream thermal regime (using results from FSP project Y07-1057), (2)accessibility of habitats (based on consideration of barriers and gradients), (3)stream width (based on hydraulic geometry relations reported by Simons & Albertson), (4)stream gradient (surrogate for channel morphology), (5)connectivity to areas with know fish (based on fish sampling) and (6)water quality parameters (pH, TDS). (3) Complete statistical analysis of relationships between fish observations and reach-scale habitat features. Fish occurrences will be extracted from the province’s fish datasets. Using the reach-scale indicators from Task 2, and provincial fish observations, we will develop empirical models to identify reach-scale variables that best explain distribution of species within stream networks using classification and regression tree approaches (Boyce & McDonald 1999; Guisan & Zimmerman 2000; Guisan et al. 2002). (4) Develop rules-based system for describing relative habitat value. Using reach-scale variables and fish occurrences identified in Task 3 in conjunction with existing datasets on fish abundance (e.g., FDIS) we will then develop a rules-based system for ranking the habitat value of stream segments for each fish species of interest. Expert opinions may also be used to characterize relative value of habitat features in these streams. (5) Determine relative habitat value of stream segments. Using reach-scale indicators calculated for each stream segment in Task 2 and rules based system from Task 4, we would combine this information to characterize relative habitat value for all streams within the study area. Relative habitat value would be characterized using qualitative rankings of habitats (e.g., traffic light coding – red, yellow, green) (6) Prepare summary report and maps.
To communicate project results, a model summary report / paper [Deliverable 2] will be prepared describing the scientific basis for completing project tasks, analytical steps, and results of habitat modelling. This paper will be submitted to the BC Journal of Ecosystems and Management for publication. Digital maps [Deliverable 3] will be prepared illustrating the predicted relative habitat value of stream segments across the study area. Maps would be distributed to regional biologists to receive a qualitative evaluation of the accuracy of the model. (7) Present modelling results. An outreach meeting [Deliverable 4] will be convened in Victoria to communicate the approach and results to government managers with the Ministry of Forests and Range and Ministry of Environment. This project will improve accuracy of predicted fish species distributions and provide quantitative (where possible) or qualitative estimates of relative habitat values for streams across a pilot study area. Because information on fish communities is used to designate Fisheries Sensitive Watersheds and Temperature Sensitive Streams as required by the Forest and Range Practices Act, and fish bearing status is also used to guide riparian management, improved predictions of fish distribution and habitat capacity are essential for broad-scale forest planning. An improved understanding of fish bearing status and connectivity of habitats is also important given the potential for year-to-year variation in habitat use of small streams. Carrying capacity modeling and stream segment analysis, while of value in its own right, will also provide input into a number of regional and provincial processes currently underway in British Columbia. For example, the federal governments’ Wild Salmon Policy could use carrying capacity estimates to define reference points for the status of various stocks of salmon. Similarly, the BC Ministry of Environment is in the process of defining “Fisheries Sensitive Watersheds”, using methods which require estimates of both the conservation value and the commercial value of the various aquatic species associated with watersheds. At a broader scale, the Nature Conservancy of Canada uses the aquatic species information combined with other biodiversity data to conduct ecoregional assessments across British Columbia. The reach and stream analysis proposed in this study therefore fits well as a foundation for a multi-scale approach starting at the reach level, and progressing to streams, watersheds, major drainages, ecoregions and ecoprovinces.


Fish Habitat Modeling Report (5.7Mb)

To view PDF documents you need Adobe Acrobat Reader, available free from the Adobe Web Site.

Updated August 16, 2010 

Search for other  FIA reports or other Ministry of Forests and Range publications.

Please direct questions or comments regarding publications to