Spatial Tree Mapping Using Photography.

Materials and Methods Study Location Sample plots were located on the J.D. Irving, Ltd., Black Brook District in northwestern New Brunswick, Canada (47°00 to 47°30 N, 67°20 to 68°10 W). The District is 190,000 ha and represents some of the most intensively managed forests in Canada. Sample plots were a subset of plots established as part of a harvesting trial. Plots were established on a systematic grid across the study area, and a subset was selected for this study to represent a range of stand conditions, densities, species compositions, and management regimes. Sampled plot characteristics are summarized in Table 1. Plot Establishment Tree Mapping Sample plots were established as variable-radius plots using a 1-m ha 1 basel area factor (BAF) angle gauge (1 m ha 1 4.356 ft ac ). The distance from plot center to each tallied tree (nearest 0.1 m) and the azimuth (nearest degree) were measured using a tape measure and compass mounted on a tripod and recorded in a field data collector. Distances were measured from plot center to the face of the tree. The dbh (to the nearest 0.01 m) of tallied trees was measured using tree calipers. Using recorded distance and tree diameter, the field data collector computed whether measured trees were in or out of the tally. This capability was useful in the field for assessing trees that were borderline based on the angle gauge. The tally sweep began at north azimuth (0°) and continued clockwise around the plot. Photo Measurements We evaluated tree distance measurements on photos using two methods: a) targets placed on trees; and b) tree dbh measured from the photograph. After measuring all trees, a target printed on card stock 21.6 27.9 cm was placed on each tree at breast height (1.3 m). In the center of each target was a 17.8 17.8-cm black square, and printed above this was an indexing tree number to relate photographic measurements to field measurements. The targets are not required to make the measurements and can be considered optional; however, it is important to have a method for indexing trees in the photos to tallied attributes. A digital camera mounted in portrait orientation on a panoramic tripod mount (Kaidan Kiwi) was placed on a tripod 1.3 m directly above the plot center. The panoramic tripod mount ensured that the camera’s focal point was directly above the rotational axis of the tripod, such that the resulting series of photographs could be stitched together with minimum distortion. The degree index on the mount was used to determine accurate camera rotation between photographs. Two camera models were used, an Olympus D550z (3 megapixels) and a Canon PowerShot SD200 (3.2 megapixels). Both cameras are entry-level “point-and-shoot” type cameras with a replacement cost between $150 and $200 each. The first plot photograph was taken oriented toward the first tree measured, and a series of 24 photographs, taken at 15° intervals around plot center, was then obtained. The 15° interval provided sufficient overlap between adjacent photographs to minimize distortion in the stitched panoramic image. Camera setup and photography averaged approximately 5 minutes per plot. Postprocessing Photographs were downloaded from the camera to a computer and organized into folders by plot. A software program called The Panorama Factory (Smoky City Design, LLC) was used to stitch Tree 1 x = 491.46p 7-degrees Tree 8 x = 2498.14p X-degrees Target measurement