An ecologically meaningful tool that uses bird surveys to measure forest health
An Ecological Index of Forest Health Based on Breeding Birds
By: Erin E. Gnass Giese, Dr. Robert Howe, Dr. Amy Wolf, Nicholas Miller, and Nicholas Walton
Index of Ecological Condition
The Index of Ecological Condition (IEC) is an easy-to-use, objective tool developed originally by Howe et al. (2007a,b) and applied by Gnass (2012) to northern mesic forests of the western Great Lakes region. Calculation of an IEC requires three important steps, two of which have already been completed for you.
1) Identify one or more “reference” gradients of environmental condition (Cenv), scaled from maximally stressed (Cenv = 0) to minimally stressed (Cenv = 10). These gradients help us identify bird species that are sensitive to environmental stressors and, to a limited extent, tell us what the indicator species actually indicate. We have used a gradient of human disturbance based on geographic information system (GIS) variables, such as land cover (e.g., percent developed lands), habitat fragmentation metrics (e.g., edge density), and other publicly-available variables that characterize the “human footprint” on the landscape (e.g., housing density) (Gnass 2012).
2) Determine sensitivity of species to environmental stress. Next, we use a modified normal distribution function (Bluman 2008) to model bird species’ responses to the environmental condition gradient. Bird data for this process were collected previously from sites (n = 917) located in a variety of landscapes ranging from high environmental stress (i.e., relatively poor condition) to low stress (i.e., relatively pristine condition) in Wisconsin and Michigan (Gnass 2012). The best-fit curve is described by three parameters (μ, σ, h) determined by computer iteration. These species’ curves are called biotic response (BR) functions. In the diagrams below (Figure 1), the x-axis corresponds to the environmental reference gradient (where low Cenv represents high stress and high Cenv represents low stress); the y-axis is the probability of a species being detected (Howe et al. 2007b, Gnass 2012). The shape of each curve can vary widely given that some species are more sensitive to disturbance than others (Howe et al. 2007a,b). For example, a fairly tolerant species such as European Starling (Sturnus vulgaris) would be more likely observed in highly degraded forests, where environmental condition values are close to 0 (Figure 1) (Gnass 2012). An intolerant, more sensitive species like Ovenbird (Seiurus aurocapilla) would more likely occur in minimally degraded forests with condition values near 10 (Figure 1) (Gnass 2012).
Figure 1. Biotic response (BR) functions of European Starling and Ovenbird (Gnass 2012). The y-axis represents probabilities of detecting species across values of condition (Cenv). The x-axis represents the reference gradient of environmental condition (Cenv) that ranges from highly degraded (Cenv = 0) to minimally degraded forests (Cenv = 10).
Following the methods of Howe et al. (2007b), we then select an assemblage of bird species for estimating the bird-based ecological condition (or IEC score) of new sites. We chose 38 breeding bird species (Table 1) that were sensitive (positively or negatively) to the environmental reference gradient and were representative of northern mesic forests in the western Great Lakes region. In order to distinguish heavily disturbed sites from healthy forested areas, however, we included a few species that are not characteristic of northern mesic forests in this region, including invasive species (e.g., European Starling) and species associated with human-disturbed landscapes (e.g., Common Grackle [Quiscalus quiscula]); Guth 1978).
3) Calculate IEC scores for new sites based on breeding bird assemblages. This step requires input from users. We have developed a computer algorithm in MS Excel (Solver add-in) to estimate the Index of Ecological Condition (or IEC score) of a new site (Howe et al. 2007b, Gnass 2012) based on breeding bird occurrences in 10-minute point unlimited-radius counts. MS Excel’s Solver tool quickly estimates and finds the IEC (ranging from 0-10) that best fits the observed data, given the previously determined species’ responses. IEC scores close to zero describe forested areas that are highly disturbed, fragmented, or associated with a heavy “human footprint”; whereas IEC scores close to 10 describe forested areas that are very healthy based on the birds found there. Results are more than just re-statements of the original reference gradient, since birds that are sensitive to the measured “human footprint” are also likely to be sensitive to other environmental stressors that we have not identified and quantified.
Table 1. List of 38 indicator breeding bird species selected for use in the Index of Ecological Condition (IEC) model for northern mesic forests in northern Wisconsin (Gnass 2012). Click on the common name of a species to view its biotic response (BR) function. Click on each species’ account provided by the Birds of North America (BNA) and Cornell University to learn more about each species.
The IEC calculator (see IEC Calculator Download and Instructions) may be used to calculate the ecological condition (or IEC score) of a single forested site, a group of sites, or a large management area where bird point count data are available. Our forest IEC model (Gnass 2012) was originally developed for applications in northern Wisconsin but it is appropriate (perhaps with slight modification of included species) for forest landscapes of northern Minnesota, the Upper Peninsula of Michigan, northern Wisconsin, and possibly other areas of the northern Great Lakes region (Bird Conservation Region 12). These calculated bird-based IEC scores can separate maximally degraded forests from minimally degraded forests and, because birds integrate many environmental factors into their habitat selection and survival at a given location, they can provide more information about a site than GIS landscape variables alone (Gnass 2012). In other words, the IEC calculator is an objective, ecologically meaningful tool for guiding ecologically sustainable forest management and conservation practices and can be used in the future to monitor the outcomes of these activities.