Canopy leaf area index (LAI) is important for predicting stand growth response to silviculture, but it is difficult to quantify because of high variability, time constraints, and limitations of nondestructive techniques. We used an uninterrupted 17-year record of litterfall in a 60-year-old
Pinus strobus L. plantation in central Maine to quantify LAI in response to both B-line and low-density thinning and to evaluate individual-tree allometric leaf area prediction models fitted to data from 51 destructively
sampled trees. Allometric model performance was inconsistent between the tree and stand levels; the most robust model at both scales predicted leaf area from sapwood basal area and crown length. The LAI of the control treatment declined gradually from 4.5 to 4, with interannual variability
associated with disturbances to the canopy. Thinning reduced LAIs but not in proportion to the number of trees removed by thinning because LAIs were similar between the B-line and low-density treatments. At the tree level, differences between the treatments were substantial, with low-density
tree leaf area increasing nearly fivefold over the study period, twice the response of comparable B-line trees. These results demonstrate the dynamic nature of leaf area, the difficulty in predicting it accurately, and the influence of silvicultural activities.