Scale dependence of vegetation‐environment correlations: A case study of a North Carolina piedmont woodland

摘要: . Vegetation and its correlation with environment has been traditionally studied at a single scale of observation. If different ecological processes are dominant spatial temporal scales, the results obtained from such observations will be specific to observation employed lack generality. Consequently, it is important assess whether that determine community structure function similar or whether, how rapidly, under what circumstances change Indeed, early work by Greig-Smith associates (Greig-Smith 1952; Austin & 1968; see 1979; Kershaw Looney 1985; Nicholls 1988) suggested plant-plant interactions typically small but physical dominates large scales. Using gridded mapped 6.6 ha portion Duke Forest on North Carolina piedmont for case study, we examined importance in vegetation studies testing four hypotheses. First, hypothesized between composition should increase increasing grain (quadrat) size. Our support this hypothesis. Second, environmental factors most highly correlated species all sizes within 6.6-ha study area, among strongly over much larger extent ca. 3500 Forest. data not consistent either Third, smaller (< 256 m2), tree canopy contribute significantly pattern under-story. do Finally, predicted increased sampling, increase. suggest opposite may true. This confirms analyses can depend greatly samples employed. Whenever possible, sampling include variety carefully selected sample so as ensure robust. Application methods used here types could lead better understanding dominate scales.