VARIATION IN CANOPY GAP FORMATION AMONG DEVELOPMENTAL STAGES OF NORTHERN HARDWOOD STANDS

Trends in gap dynamics among pole, mature, and old-growth northern har dwood stands were investigated on eight sites in the Porcupine Mountai ns of western upper Michigan. Recent gaps (created between 1981 and 19 92) were identified using permanent plot records of tree mortality, wh ile older gaps (1940-1981) were identified using stand reconstruction techniques. Although canopy gaps were somewhat more numerous in pole a nd mature stands, gaps were <25% as large as those in old-growth stand s because of smaller gap-maker size, and the proportion of stand area turned over in gaps was only about half as large. Gap makers in younge r stands generally had mean relative diameters (ratio of gap-maker DBH to mean DBH of canopy trees) <1.0 and were disproportionately from mi nor species such as eastern hophornbeam (Ostrya virginiana (Mill.) K. Koch). Gap makers in old-growth stands had mean relative diameters >1. 5 and were predominantly from the dominant canopy species. Even in old -growth forests, most gaps were small (mean 44 m(2)) and created by si ngle trees. Based on the identity of the tallest gap tree in each gap, nearly all shade-tolerant and midtolerant species have been successfu l in capturing gaps, but gap capture rates for some species were signi ficantly different from their relative density in the upper canopy. Th e tallest gap trees of shade-tolerant species were often formerly over topped trees, averaging more than 60% of the mean canopy height and ha ving mean ages of 65-149 years. Canopy turnover times, based on gap fo rmation rates over a 50-year period, were estimated to average 128 yea rs for old-growth stands dominated by sugar maple (Acer saccharum Mars h.) and 192 years for old-growth stands dominated by hemlock (Tsuga ca nadensis (L.) Carriere). While these estimates of turnover time are su bstantially shorter than maximum tree ages observed on these sites, th ey agree closely with independent data on mean canopy residence time f or trees that die at the average gap-maker size of 51 cm DBH. The data support previous hypothetical explanations of the apparent discrepanc y between canopy turnover times of <130 years for hardwood species and the frequent occurrence of trees exceeding 250 years of age.