Leaf area measurements based on hemispheric photographs and leaf-litter collection in a deciduous forest during autumn leaf-fall

摘要: Abstract Predicted relationships between leaf area index ( LAI ) and gap frequency using the Poisson, binomial Markov theoretical models of canopy geometry (Nilson, 1971) were used to estimate from hemispherical photographs taken during autumn leaf-fall period in a maple-aspen forest southern Ontario, Canada. Both require specification parameter describe clumpiness distributions. For model, value given by Baldocchi et al. (1985) for an oak-hickory Tennessee was used, while means estimating required developed based on conditional probabilities deduced hemispheric photographs. estimates derived compared with values obtained leaf-litter collection as went fully leafed leafless range 0–∼5.1). Comparisons 6 days over this period. Estimates at bottom nine locations within 15 × 15-m plot. Analysis performed video camera commercial image analyser. Poisson model-derived results appreciably less than latter greater ∼2, favourably >2. The apparent failure model relatively larger 's attributed expected distributions deciduous forests. ∼1, all exceeded values, reflecting influence tree branches frequencies recorded An 0.5 woody element mean estimates. difference showed little trend decreasing area, but corresponding tended increase decreased, probably change true full decreased. density profiles also obtained. These field-of-view seven levels tower forest. Only equivalent one photograph each level any measurement day could be obtained, which increased inherent uncertainty relative entire estimates, With subjective smoothing produced application physically consistent appeared reasonable few measurements such that have been reported.