Date Awarded


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Virginia Institute of Marine Science


Carl Hershner


Woody and herbaceous vegetation were sampled in 23 tidal swamps along a tidal freshwater tributary of lower Chesapeake Bay. Four vegetative categories were ordinated with Detrended Correspondence Analysis (DECORANA). Species distribution patterns of each strata were compared with respect to edaphic factors, a wetness index, and mean water table depth. Woody species are restricted to hummocks (topographic highs). Hummocks drain as quickly as the tide drops and so are partially inundated for only short periods each day. Although low in canopy diversity, tidal swamps are floristically rich in herbaceous and woody understory species, ranking them among the most speciose in temperate North America. Canopy composition is related to the wetness of a site as determined by the percent of the forest floor covered by hollows (low inter-hummock depressions) and by mean water table depth. Fraxinus spp. and Nyssa biflora dominated swamps are best developed in wetter sites, which contain higher calcium (Ca) and organic matter (Om) levels and where the mean water table depth is about &-&17 cm. In contrast, Acer rubrum-Liquidambar styraciflua-Nyssa biflora dominated swamps occur at less wet sites where mean water table depth is deeper than 20 cm. Although DECORANA separated canopy and herbaceous strata similarly, the woody subcanopy (shrubs and small understory tree species) did not separate into the same two communities. to determine whether this pattern might be indicative of forests in general, distributional data of canopy and subcanopy species were also compared using similarly collected data from a southern Appalachian forest. Sapling (juvenile canopy species) distribution patterns were also compared in both systems. Separate ordinations were performed on canopy, sapling, and subcanopy species. Canopy trees and saplings showed a similar pattern of distribution, suggesting that resource requirements of saplings and canopy-statured adults are similar. In contrast, the subcanopy species of neither ecosystem showed any discernable distributional relationship to the canopy or sapling layers, suggesting that subcanopy life-forms may partition different resources than canopy species in temperate forests. If so, the common practise of combining sapling and subcanopy species in structural analyses may hinder our understanding of subcanopy structural patterns in forests.



© The Author