Submitted to: Wildlife Society Bulletin
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: July 1, 2006
Publication Date: November 1, 2006
Citation: Boyd, C.S., Hopkins, K.T., Svejcar, A.J. 2006. A photo-based monitoring technique for willow communities. Wildlife Society Bulletin 34(4):1049-1054. Interpretive Summary: Willows (Salix spp.) provide important habitat for numerous wildlife species, and help maintain the integrity of riparian ecosystems, however, repeatable techniques for monitoring willow abundance at the community scale are currently lacking. We determined between analyst variability and field sampling error associated with a photographic/computer image analysis technique for estimating the spatial area of willow clumps. Results suggest that area estimates for willow clumps were repeatable between analysts and error associated with field sampling was low. Our technique will allow managers to generate repeatable estimates of willow abundance while minimizing analyst-based error. Additionally, the use of a photographic technique provides a permanent record of vegetation status for use in long-term monitoring.
Technical Abstract: Willow (Salix spp.) and associated riparian shrubs communities provide habitat to a wide variety of wildlife species, but ground-based monitoring of willow abundance has proven difficult. The objective of this study was to evaluate the variability of collection and analysis of field data for a photo-based monitoring technique for willow communities. Variation in data collection was evaluated by photographing 5 willow clumps, 10 times each and comparing area estimates within clump. This was assumed to mimic variability associated with repeat monitoring of a given clump(s) over time. High visibility markers were set at known distances apart to provide scale references in the photographs. Camera and markers were removed and replaced between successive photographs. Scanned images of the photographs were spatially rectified using digital image processing software. The two-dimensional area of willow clumps was determined by digitizing clump boundaries within rectified images. Variability associated with image analysis was examined by asking 6 analysts to determine willow clump area for a series of 5 images and comparing the results across analysts. Sampling error for each photograph was calculated by dividing root mean square error by the mean value. Results indicate field data collection produced minimal variability; sampling error averaged 1.82% (+/- 1.05). Between analyst sampling error averaged 1.63% (+/– 0.73) across clumps, and was <3% for all clumps. Trained analysts took <5 min per image to obtain area estimates. These results indicate our technique produces quick and repeatable estimates of willow abundance, would be useful in evaluating change in abundance over time, and minimizes person to person variability.