Forest Investment Account (FIA) - Forest Science Program
FIA Project Y092251

    Effects of habitat composition on the fitness of a mature forest indicator; do thresholds exist?
Project lead: Mahon, Todd (Bulkley Valley Centre for Natural Resources Research and Management)
Author: Mahon, Todd
Subject: Forest Investment Account (FIA), British Columbia
Series: Forest Investment Account (FIA) - Forest Science Program
Industrial forest development can profoundly change the composition and spatial pattern of habitats across landscapes relative to historic ranges of natural disturbances. In central British Columbia landscape pattern is departing from the historic range of variability (Steventon 2003). In the longer term, seral stage composition is also predicted to depart significantly from historic means (Eadie 2003).

Empirical data quantifying minimum or optimal conditions for species at the landscape level are generally lacking and are recognized as a key knowledge gap (Boutin and Hebert 2002; DeStefano 2002). In an attempt to maintain biodiversity and wildlife habitat, without knowing the exact requirements for most species in a community, forest managers often utilize a variety of management practices that attempt to mimic natural process and conditions within a ‘coarse-filter’ management approach. Minimum thresholds or desired ranges for landscape composition (e.g. seral stages) and landscape pattern (e.g. patch size distribution) are often key components of coarse-filter strategies within sustainable forest management (SFM) plans. The rationale for landscape targets in SFM plans are generally based on principles of landscape ecology, however, there is little empirical evidence to support much of this theory (MacGarigal and Cushman 2002).

Most aspects of landscape ecology theory are based on two fundamental attributes, habitat amount and habitat configuration. Early papers emphasized habitat amount, but through the mid 1980s and 1990s much of the emphasis shifted to examining the effects of spatial configuration under the broad term of fragmentation (Fahrig 2002). More recently, research and literature reviews have re-emphasized the overriding importance of habitat amount relative to habitat configuration (Boutin and Hebert 2002; Fahrig 2003).

Relationships between habitat amount and the fitness or abundance of species often do not follow a linear relationship and the idea of threshold responses is pervasive in the literature (review by Dykstra 2004). Threshold responses are often described by a logistic curve that involves a dramatic change in the rate of ecological response to habitat change. Initially there may be little or no response to decreasing habitat, but once habitat loss reaches a critical region (i.e. the threshold) an ecological property (e.g. breeding rates or population size) can be drastically affected (May 1977). Habitat thresholds are widely applied in ecological modeling. A prominent example is the “extinction threshold” (Lande 1987) that is often a key component of minimum viable population analysis (e.g. Thompson and Harestad 1984).

Despite the development of fundamental principles of landscape ecology at least two decades ago (e.g. Harris 1984), and the recognition of the importance of empirical studies at that time, there have been few large scale, long-term, experimental studies to assess these principles (MacGarigal and Cushman 2002; Fahrig 2003). The number and success of field studies in this area have been limited by the logistical difficulty of conducting good landscape ecology experiments and differentiating between habitat loss and fragmentation effects (D’Eon 2002; MacGarigal and Cushman 2002; Schmiegelow and Monkkonen 2002).

The goal of this research is to provide empirical data to evaluate responses to habitat loss, with emphasis on potential threshold responses, that can be used to assess landscape ecology theory and refine SFM strategies. To do this I will quantify functional habitat relationships of a wide-ranging, mature forest-dependent species, the Northern Goshawk (Accipiter gentilis). This research is uniquely positioned to overcome some of the primary limitations of previous landscape studies by 1) examining a system undergoing profound landscape habitat changes resulting from epidemic mountain pine beetle attack and salvage logging (in the context of a natural landscape experiment), 2) studying a species that is a habitat specialist but that is relatively insensitive to habitat configuration, and 3) quantifying habitat use and prey availability to explain relationships observed between habitat composition and fitness.

This research will build on an existing research program being undertaken by the federal Sustainable Forest Management Network, the University of Alberta and the Morice and Lakes IFPA. FSP funding in 2007/08 allowed us to extend field research by an additional year to obtain optimal sample sizes. In 2008/09 we will analyze our results and work with government and industry to develop effective policy and best management practices that integrates those results. We will also conduct several extension activities to disseminate our research results and SFM strategies to the forest industry, government, and academia.
Related projects:  FSP_Y081251

Technical Report (0.8Mb)

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