Ice Storm Damage to Virginia Coastal Plain Forests dur- ing the Christmas 1998 Ice Storm

On December 23-25, 1998, a major ice storm struck southeastern Virginia The storm-deposited glaze ice felled trees and limbs, causing a power outage and highway blockage. Between Februmy and April, 1999, we recorded occurrence, severity, and type of damage to trees over 2.5 cm dbh in nine mostly gently sloping plots in Matoaka Woods at the College of William and Macy. Frequency and severity of damage varied with species and with size of trees. Canopy damage occurred in 75% of large Fagus grandifolia trees, but in only 6% of small Sassafras a/bidum stems. As a group, small (2.5 to 15 cm dbh) trees were less likely to be damaged than large ( 15 cm dbh) trees, but about as likely to be severely damaged. Damage type also varied among the species and size. Despite severe damage to public utilities, damage within the forest was not great. Since few trees lost their entire crown, canopy gap sizes were small, and it not clear that much change in forest composition will result from this storm. However, increased density of ground litter will contribute to greater mineral release, and this plus small gaps may promote growth of already present seedlings and saplings. INTRODUCTION On December 23, 24, and 25, 1998, a major ice storm affected southeastern Virginia. Precipitation in the form of sleet and freezing rain accumulated to 1-3 cm of ice across the region, with Williamsburg reporting 3 cm of precipitation for the three-day period. In the City of Williamsburg and surrounding counties, 400,000 customers lost power for three to ten days following the storm. Many roads, including portions of Interstate 64 near Lightfoot, VA, were rendered impassable by fallen branches and trees (NCDC 1998a,b). The storm's impact on the community was certainly severe, and much of the infrastructure damage was caused by ice-felled branches and trees along roadsides and on forest margins. Based on the degree of damage readily observable from the roads, we felt that this storm presented an ideal opportunity to detennine the effects of ice accumulation on local forests. The great damage to roadside and forest margin trees, however, was due to their peculiar location. Without adjacent vegetation of comparable height to support their accumulated weight in ice, and with either asymmetric or fuller crowns due to lack of competition for light, individuals in the open would likely be more susceptible to damage than those in the forest. Nevertheless, preliminaiy investigation-of our potential study sites indicated that, although the damage within the forest was not as heavy as on its margins, it did appear significantenough t9 provide data for a meaningful study on the dominant tree species of the area. We surmised that the College Woods (also called Matoaka Woods), a forested area owned by the College of William and Macy, was an ideal place for a small-scale 4 VIRGINIA JOURNAL OF SCIENCE investigation into the susceptibility to ice of several major tree species on the Coastal Plain of Virginia. Matoaka Woods is made up of a variety of small, homogenous stands dominated by canopy species such as tulip poplar (Liriodendron tulipifera), oaks (Quercus spp.), beech (Fagus grandifolia), and loblolly pine (Pinus taeda). The mosaic pattern of the woods (fanned and forested patches were abandoned or last timbered at various times for various reasom) has allowed for a diversity of species, and also has ensured equal representation of a broad spectrum of size classes. In this study, our primary goal was to swvey the amount and type of damage to each of the more abundant tree species in Matoaka Woods. Of secondmy interest was the comparison of damage among different size individuals of the same species. METIIODS Our field swvey was conducted in the Matoaka Woods of the College of William and Mary between February 3 and April 7, 1999. No further forest-ravaging natural phenomena occurred between the end of the Christmas storm and the completion of our swvey. Sampling sites were chosen based on the constituent species and apparent age of the dominant individuals: younger and older stands dominated by oak species~ tulip poplar, loblolly pine, and beech were sought out with the hopes of comparing damage between different aged canopy trees of the same species or genus, as well as among the different species. The sampling sites were widely spread throughout the woods. We chose to follow Seischab et al. (1993) in our methodology. We marked a 20x40-meter plot at each sampling site. Each of these was broken into four 10x20 meter subplots for ease in sampling. In each subplot, trees larger than 2.5 cm dbh were identified by species and were placed in one of two size categories: between 2.5 and 15 cm dbh and over 15 cm dbh. In general, trees in the smaller size class were subcanopy, and those in the larger size class were in the canopy. Though we took measures to avoid bias toward areas likely to be heavily damaged (such as steep slopes above ravines; W arrillow and Mou 1999), beech-dominated stands could not be found in the more level portions of the woods. Thus, in order to sample beech, it was necessary to place two plots on slopes. Effects on the results due to this difference in topography will be discussed later. Each tree swveyed was placed in a damage class between O and 7 based on percent canopy loss due to ice damage. A rating of O corresponded to no perceptible damage, 1 to~ 5 %canopy loss, 2 to 6-10%canopy loss, 3 to 11-25%canopy loss, 4 to 26-50% canopy loss, 5 to 51-75% canopy loss, and 6 to 76-99% canopy loss. A rating of 7 was given where damage was so severe that mortality was likely. Though we quantified canopy damage as an estimate of percent of canopy lost, the accuracy of our estimates was necessarily subject to error, for we were not able to observe the leafed out canopies , of deciduous trees, nor had we previously documented canopy sizes for any of the trees surveyed. However, eveiy effort was made to be consistent. We recorded the nature of the damage to each tree, noting whether each damaged tree was uprooted (symbolized by oin the tables), had its main stem broken (symbolized by I\), had its main stem bent or bowed (C), had one or more branches completely broken from the tree ( o ), had one or more branches broken but still attached to the tree (/\). We also noted whether the damage, of whatever type, was direct (as a result of ice accumulation on the tree in question) or secondaiy (a result of ice-laden branches, ICE STORM DAMAGE TO VIRGINIA FORESTS 5 TABLE 1. Field data for individuals ~ 1' cm dbh. See text for desaiption of damage classes and types. Sample Damage class Damage type S~ies size 0 1 2 3 4 5 6 7 o-1\ ~ " 0 s Pinrutaeda 53 29 2 7 3 3 1 2 6 7 2 12 Liriodendron tulipifera 47 27 5 6 2 3 2 2 15 Quercus alba 35 21 6 1 3 3 1 5 4 Fagua grandifolia 28 7 5 4 4 3 4 4 16 1 Oxydendron arboreum 23 13 4 4 1 2 5 2 Liquidambar styraciflua 15 12 1 2 Quercus velutina 11 6 2 2 1 2 1 Acerrubrum 9 4 2 2 2 4 Carya glabra 8 6 2 2 Quercus falcata 6 3 1 2 Nyssa sylvatica 5 3 2 1 Quercus rubra 4 4 Jlexopaca 3 1 2 Quercus coccinea 3 2 1 Comus jlorida 2 2 Carya tomentosa 1 1 Fraxinius americana 1 1 Prunus serotina 1 canopies, or entire trees falling on individuals below). Recently fallen live branches~ 2.5 cm at the broken base (butt end) found in the plots were tallied by species and size; any above 10 cm diameter at the base were further roted. We did not attempt to quantify the deadwood since it was impossible to distinguish dead material felled by this stonn from that previously on the ground By performing our investigation in the winter and early spring immediately following the ice storm, we were able to easily determine the most recent open wounds and fallen branches, for the infection and decay dependent on warm temperatures had not begun. We also avoided the possibility of additional damage from other natural disasters (such as windstonns, including the hurricane that struck the study area the following summer). Because no new growth had begun on bent or wounded stems, we could distinguish fresh bending from older bending or breaking, since trees previously damaged had redirected their foliage or sprouted new stems during the last growth season The lack of intervening foliage in the understory made it easier to examine damage to canopy trees, but, as mentioned previously, percent canopy loss was harder to estimate accurately without foliage. RESULTS We found no significant differences in damage between older stands and younger stands with the same dominant species. Because or' this finding, descriptions of individual plots have not been included, and all data from each species have been merged to reflect interspecific differences and differences between the canopy and understory classes. The amount and type of damage incurred by the 27 species we encountered during our survey is shown in Tables 1 (individuals ~ 15 cm dbh) and 2 (individuals< 15 cm dbh). 6 VIRGINIA JOURNAL OF SCIENCE TABLE 2. Field data for individuals < 1 Scrn dbh. See text for description of damage classes and types. Sample Damage class Damage type S~ies size 0 2 3 4 5 6 7 oI\ ~ I\ 0 s Liriodendron tulipi.fera 146 109 12 5 1 1 4 4 9 12 9 3 5 8 Comus florida 132 106 8 2 6 5 2 3 8 1 Cl!l2 Acerrubrum 75 46 6 5 6 2 2 5 3 4 6 10 4 5 Oxydendron arborewn 54 27 4 4 4 2 6 7 3 2 11 4 4 12 Ilexopaca 49 31 3 4 5 5 7 3 6 8 Liquidambar styraciflua 44 36 3 1 2 4 1 1 3 Fagus grandifolia 40 36 3 2 1 Nyssa sylvatica 37 31 3 2 2 1 Sassafras albidum 17 16 1 Carya glabra 14 9 2 2 2 Quercus alba 6 5 1 Pinus taeda 5 1 2 2 2 2 Carya tomentosa 3 3 Castanea dentata 2 2 Cercis canadensis 5 3 1 Quercus velutina 5 4 1 Juniperus virginiana 4 2 2 Vitis rotundif olia 2 2 Diospyros virginiana 1 1