A major ice storm in January 1998 provided an opportunity to study the effects of a rare, intense disturbance on the structure of the northern hardwood forest canopy. Canopy damage was assessed using visual damage classes within watersheds of different ages at the Hubbard Brook Experimental Forest (HBEF) and changes in leaf area index in two of these watersheds. Ice thickness was measured, and ice loads of trees were estimated using regression equations. In the 60- to 120-year-old forests (mean basal area 26 m2·ha–1), damage was greatest in trees >30 cm diameter at breast height and at elevations above 600 m. Of the dominant tree species, beech (Fagus grandifolia Ehrh.) was the most damaged, sugar maple (Acer saccharum Marsh.) was the most resistant, and yellow birch (Betula alleghaniensis Britt.) was intermediate. Trees with advanced beech bark disease experienced heavier ice damage. Little damage occurred in the 14-year-old forest, while the 24- to 28-year-old forest experienced intense damage. In the young stands of this forest, damage was greatest between 600 and 750 m, in trees on steep slopes and near streams, and among pin cherry (Prunus pensylvanica L.). Recovery of the canopy was tracked over three growing seasons, and root growth was monitored 1 year after the storm. Because of the high density of advance regeneration from beech bark disease and root sprouting potential in ice-damaged beech, HBEF will likely see an increase in beech abundance in older forests as a result of the storm. There will also be a more rapid change from pioneer species to mature northern hardwoods in the younger forests. These predictions illustrate the ability of rare disturbances to increase heterogeneity of forest structure and composition in this ecosystem, especially through interactions with other disturbances.

Several methods exist for measuring forest canopies following disturbance, and the biases and differences among them are unclear. We compared techniques for measuring the northern hardwood forest's canopy structure at the Hubbard Brook Experimental Forest, New Hampshire, following the severe ice storm of January 5–10, 1998. Methods included leaf area index (LAI) using LI-COR's LAI-2000, visual damage assessments based on tree branch loss, radiation estimates from hemispherical photographs, and LAI determined from litterfall. LAI-2000 measurements were not significantly related to visual damage class estimates, but were strongly correlated with radiation estimates from hemispherical photographs and average LAI values from litterfall. LAI from the LAI-2000 and litterfall differed on a point-by-point basis, but were similar at the stand scale. The LAI-2000 has the highest precision for large-scale measurements. Visual damage estimates appear adequate for assessing large-scale patterns of disturbance intensity in the northern hardwood forest, but the LAI-2000 is more accurate at quantifying canopy structure at large plot or stand scales. Hemispherical photographs may also accomplish this, but are better suited to characterizing the distribution of canopy gaps and light availability patterns over time. Litterfall provides accurate and precise measurements of small-scale LAI patterns in deciduous forests and reveals species-specific patterns.

We studied the effects of host plant hybridization on the survival and mortality of the leaf-mining moth Phyllonorycter salicifoliella on hybrid and parental willow plants in the field and in a common garden experiment. P. salicifoliella survival differed significantly among three willow taxa in the field in 1994 but not in the field in 1995 or in the common garden. Parasitism by eulophid wasps differed significantly among taxa in 1994 and appeared to account for the variation in their survival. In the field in 1995, host feeding predation varied significant among taxa. The theory of tritrophic interactions predicts that plant genotype can affect natural enemy impact, and this study supports this prediction. Significant variation in survival and eulophid parasitism was also found among genotypes within taxa in the field in both years and in the common garden experiment. The common garden results show that genetic differences in plants affect the herbivore-parasitoid interaction. Variation among years in the patterns of survival and causes of mortality among field plants suggest that genotype by environment interactions may be important.