Comparison of hardwood stand structure after partial harvesting using intensive canopy maps and geostatistical techniques

Abstract

Intensive canopy mapping and geostatistical techniques were used to evaluate spatial heterogeneity in stand structure after diameter-limit harvests and a shelterwood seed cut in an Appalachian hardwood stand. The size, density, and location of tree crowns and stems were mapped to provide baseline data for predetermined locations within treatment blocks. Spatial data were collected from the same areas. Variogram models were used to compare spatial dependence in canopy cover and regeneration among the harvests. Kriged estimates of canopy cover were obtained and compared to intensive canopy maps. Canopy cover was spatially dependent on a range of 11.5 m in the shelterwood seed cut, 7.5 m in the 30 cm, and 8.1 m in the 41 cm diameter-limit harvests. Spatial dependence of regeneration extended to a range of 9.7 m in the shelterwood seed cut and 7.3 m in the 41 cm diameter-limit harvest. There was little dependence in regeneration presence within the 30 cm diameter-limit harvest at the scale sampled. Shade-tolerant regeneration had longer ranges of spatial dependence in the shelterwood seed cut. Shade-intolerants were autocorrelated at greater distances within the 41 cm diameter-limit harvest. Tolerant regeneration was positively correlated with the presence of canopy to a distance of 5–10 m. Likewise, the presence of shade-intolerant regeneration was negatively correlated with canopy cover at the same distance. Kriged and intensive canopy maps appeared to match well; however, when sampled, only the kriged maps of the 30 cm diameter-limit harvest corresponded with the intensive canopy maps to a significant extent. Geostatistics provided an excellent means of depicting spatial dependence in these harvests. Future refinements of these techniques could allow larger areas to be sampled with less error, as well as with a large reduction in both cost and effort when compared to intensive canopy mapping.