Jeff S. Glitzenstein

EFFECTS OF FIRE REGIME AND HABITAT ON TREE DYNAMICS IN NORTH FLORIDA LONGLEAF PINE SAVANNAS

Frequent, low intensity fire was an important component of the natural dis- turbance regime of presettlement savannas and woodlands in the southeastern USA dom- inated by longleaf pine (Pinus palustris), and prescribed burning is now a critical part of the management of these endangered habitats. Fire season, fire frequency, and fire intensity are three potentially important, though still little understood, components of both natural and managed fire regimes. In this long-term (8-yr) study, we experimentally (through the use of prescribed burning) tested for effects of fire season (eight different times throughout the year) and fire frequency (annual vs. biennial burning), on population dynamics (re- cruitment, growth, mortality, change in density, and change in basal area (the total basal area of all stems in a plot)) and species composition of trees in two quite different types of longleaf-pine-dominated habitats (north Florida sandhills and flatwoods). Limited fire temperature and intensity data were also collected during one year to examine the rela- tionship between fire behavior (temperature and intensity) and tree mortality. Contrary to prior hypotheses, our results showed few systematic or predictable effects of season or frequency of burning on dynamics of longleaf pine. Instead, variability in the population dynamics of this species appeared to be related largely to variation in fire behavior, regardless of the season of burning. Consistent with prior hypotheses, we found that deciduous oak species (Quercus laevis, Q. margaretta, and Q. incana) were least vulnerable to dormant-season burning and most vulnerable to burning early in the growing season. This was shown particularly by seasonal trends in the effect of burning on oak mortality (both topkill and complete kill) and, to a lesser extent, on oak recruitment. Oak densities and basal areas also declined in the spring- burned plots, resulting in a shift away from oaks and towards increased dominance by longleaf pine. Detrimental effects of spring burning on oaks were partly explained by fire behavior, but there appeared also to be an important residual effect of burning season, particularly on complete kill. Though longleaf pine population dynamics did not differ markedly as a result of burning season and frequency, we did find important differences in pine dynamics between the two habitats (i.e., sandhills and flatwoods). In general, populations of longleaf pines in the sandhills appeared to be density regulated, while flatwoods pine populations were declining regardless of the level of intraspecific competition. This suggests that long-term persistence of longleaf pine, and perhaps other fire-adapted species in frequently burned longleaf-pine- dominated communities, may be determined by complex interactions between habitat factors and fire regimes.

Woody Seedling Dynamics in an East Texas Floodplain Forest

We monitored woody plant seed deposition, seedling emergence, and the survival and growth of seedlings (i.e., plants ≤0.5 m tall regardless of age) in an East Texas river floodplain forest from 1979 through 1984. In addition, we estimated the relative importance of flooding, drought, fungal attack, herbivory, proximity to a conspecific adult, and shade in causing seedling mortality. Tree species fell into two major groups on the basis of their demographic characteristics and responses to unfavorable conditions. The first group was composed of heavy—seeded species, of which water oak (Quercus nigra) was the primary example. They produced few seeds, but had high seedling survival. Seedlings of these species emerged late in the summer, thereby avoiding peak periods of flooding and damping—off mortality. Seedling survival was little affected by drought, herbivory, or proximity to a conspecific adult. The second group included most of the common tree species (e.g., ironwood, Carpinus caroliniana; sweetgum, Liq...

DISTURBANCE, SUCCESSION, AND MAINTENANCE OF SPECIES DIVERSITY IN AN EAST TEXAS FOREST'

Vegetation change following selective logging and major natural disturbance was in- vestigated through the use of stand history reconstruction and population monitoring in an east Texas pine-hardwood forest. High frequencies of release in beech (Fagus grandifolia) and white oak (Quercus alba) occurred after apparently natural disturbances in 1800-1810, 1860-1870, and 1960-1970, and after selective logging beginning 1910. Pulses of recruitment occurred after all disturbances except the most recent one. Age data suggest that recruitment after the 1910 disturbance was asynchronous among species, with loblolly pine (Pinus taeda) regenerating first, followed sequentially by white oak, red maple (Acer rubrum), and beech. At present, only magnolia (Magnolia grandiflora) is showing significant new recruitment. After the 1910 disturbance, regenerating pines grew rapidly and became dominant in the re-forming overstory. Most hardwoods originating after the 19 10 disturbance did not reach the overstory. White oak initially outgrew maple and beech and is presently larger. Over the last 20 yr beech growth has been greater than white oak growth, resulting in a decline in the average size difference between individuals of these species. Presently, mortality rates of understory populations are high for sweetgum, blackgum, and the oaks, and are low for beech and magnolia. These latter two species are also the faster growing. Ifthese trends continue, beech and magnolia will become increasingly predominant in the forest understory and eventually in the overstory as well. Map data show that saplings and small trees are abundant under pine and oak, but not under beech and magnolia. Future disturbance is therefore likely to accelerate succession to more shade-tolerant species in parts of the forest now dominated by pine and oak, but is most likely to re-initiate new regeneration, including pine and oak, in areas now dominated by beech and magnolia. These results suggest a pattern of cyclical replacement driven by disturbance, a pattern which may help preserve species diversity in southern mesic forests.

Effects of environment and land-use history on upland forests of the Cary Arboretum, Hudson Valley, New York.

GLITZENSTEIN, J. S., C. D. CANHAM, M. J. McDoNNELL AND D. R. STRENG (Institute of Ecosystem Studies, Millbrook, NY 12545). Effects of environment and land-use history on upland forests of the Cary Arboretum, Hudson Valley, New York. Bull. Torrey Bot. Club 1 17: 106-122. 1990.-Relationships of vegetation to environment and land-use history were investigated in forests of the Mary Flagler Cary Arboretum in the Hudson Valley of New York. Vegetation data were obtained from 76 I/2 ha circular plots randomly located within the forest. Environmental data collected at each plot included slope, aspect, canopy openness, soil texture and nutrients, topographic position, and presence of exposed rock; gravimetric soil moisture was determined weekly during 1985 for a subsample of 25 stands. Land-use history information came from historical records (land deeds and U.S. and N.Y.S. census records), stone-fence locations, landscape patterns in stand ages, an old aerial photograph, and soils data. Vegetation analyses identified three major community types. One group of stands, dominated by chestnut oak (Quercus prinus L.) and northern red oak (Quercus rubra L.), occurred on steep, rocky, upper slope sites never cleared for agriculture. Distinct vertical stratification of dominant canopy species in these stands is consistent with a probable history of intensive selective cutting early in this century. Both of the other major community types occurred primarily on abandoned agricultural land. Stands dominated by white oak (Quercus alba L.), black oak (Quercus velutina Lam.) and pignut hickory (Carya glabra (Mill.) Sweet) tended to occur at lower elevations on rocky, nutrient poor sites probably derived from abandoned pastures. The significantly more open canopy in these stands, less distinct vertical stratification of canopy trees, and a diverse herbaceous understory frequently including grasses and sedges, also suggests very gradual invasion of these forests onto old pasture sites. The third major vegetation type, dominated by red maple (Acer rubrum L.) and white pine (Pinus strobus L.), tended to occur on finer textured, less rocky old field sites possibly abandoned from cultivation. Comparison of current vegetation with witness tree data from early land survey records suggests that the white oak-black oakhickory type was prevalent on lower slope sites prior to forest clearing, but has declined in importance relative to the red maple type during the past 100 years of abandonment of land from agriculture.

Effects of the December 1983 tornado on forest vegetation of the Big Thicket, southeast Texas, U.S.A.

Abstract On December 10, 1983, a large tornado touched down in southwestern Tyler County in southeastern Texas, U.S.A., causing major damage to two areas of forest in the Big Thicket National Preserve (BTNP). The sizes of the disturbed areas were 31.1 ha in the Hickory Creek Unit and 10.4 ha in the Turkey Creek Unit of the BTNP. Canopy cover at Hickory Creek was reduced by 83%, live basal area declined by 66.7%, and live density declined by 33.6%. At Turkey Creek, canopy cover was reduced by 46%, live basal area declined by 37.8% and live density declined by 14.4%. Large percentages of surviving trees at both sites were badly damaged: 31.3% at Hickory Creek and 48.9% at Turkey Creek. Tornado mortality was strongly concentrated in the large size classes and, for this reason, the primary short-term effect on forest composition at both sites was to select strongly against the previously dominant canopy species.

IDENTIFYING FUNCTIONAL GROUPS OF TREES IN WEST GULF COAST FORESTS (USA): A TREE-RING APPROACH

A dendroclimatic study of important tree species in the west Gulf Coastal Plain region, USA, was conducted to evaluate how climate affects tree radial growth in this southeasternmost section of the Eastern Deciduous Forest/Southern Evergreen Forest. We established an east–west transect from western Louisiana to central Texas that crossed the western range limits of each of 16 species and developed a network of 104 annual tree ring-width chronologies from 38 sites. Of the 104 chronologies, 99 series from the genera Pinus, Quercus, and Fagus were analyzed using rotated principal components analysis (RPCA). The RPCA revealed the presence of three robust phylogenetic signals in the tree-ring patterns, which partitioned the data into the Pinus species (PISP), the oak species in the black oak subgenus Erythrobalanus (QUBO), and the oak species in the white oak subgenus Leucobalanus (QUWO). The Fagus chronologies (FAGR) also loaded most highly with the QUWO series, resulting in a combined QUWO/FAGR factor. This partitioning occurred even though tree species within each phylogenetic group came from contrasting xeric and mesic sites and, in the case of the QUWO/FAGR factor, from different genera. Only in the xeric western range limits of the transect did site location begin to override the phylogenetic groupings. Consequently, responses to climate based on genetics appeared to be more important than ecological and site characteristics in determining the tree-ring patterns of the sampled species overall. We tested this hypothesis by independently modeling the dendroclimate signals in the tree-ring chronologies using monthly precipitation and maximum temperature data. The resulting climate correlation functions were subjected to RPCA as before. As we did so, the same phylogenetic groups emerged. All of the chronologies were drought sensitive. However, the phylogenetic differences in climate response were related to differences in the timing of the peak monthly responses to climate and to the differing patterns of climate response in the months prior to the current growing season. The findings of this study indicate that there is an underlying organizing principle based on genetics that determines how certain phylogenetic groups of trees respond to climate in a way that is largely independent of the site environment. At a coarse level, these phylogenetic distinctions persist even at the most stressed sites near tree range limits, though distinctions within genera start to break down. These findings therefore suggest functional groupings of tree species, which can be used in vegetation/climate models that attempt to predict realistically how such forests will respond to future climate changes.

Real-time and time-integrated PM2.5 and CO from prescribed burns in chipped and non-chipped plots: firefighter and community exposure and health implications

In this study, smoke data were collected from two plots located on the Francis Marion National Forest in South Carolina during prescribed burns on 12 February 2003. One of the plots had been subjected to mechanical chipping, the other was not. This study is part of a larger investigation of fire behavior related to mechanical chipping, parts of which are presented elsewhere. The primary objective of the study reported herein was to measure PM2.5 and CO exposures from prescribed burn smoke from a mechanically chipped vs. non-chipped site. Ground-level time-integrated PM2.5 samplers (n=9/plot) were placed at a height of 1.5 m around the sampling plots on the downwind side separated by approximately 20 m. Elevated time-integrated PM2.5 samplers (n=4/plot) were hung atop ∼30 ft poles at positions within the interior of each of the plots. Real-time PM2.5 and CO data were collected at downwind locations on the perimeter of each plot. Time-integrated perimeter 12-h PM2.5 concentrations in the non-chipped plot (AVG 519.9 μg/m3, SD 238.8 μg/m3) were significantly higher (1-tail P-value 0.01) than those at the chipped plot (AVG 198.1 μg/m3, SD 71.6 μg/m3). Similarly, interior time-integrated 8-h PM2.5 concentrations in the non-chipped plot (AVG 773.4 μg/m3, SD 321.8 μg/m3) were moderately higher (1-tail P-value 0.06) than those at the chipped plot (AVG 460.3 μg/m3, SD 147.3 μg/m3). Real-time PM2.5 and CO data measured at a position in the chipped plot were uniformly lower than those observed at the same position in the non-chipped plot over the same time period. These results demonstrate that smoke exposures resulting from burned chipped plots are considerably lower than from burned non-chipped plots. These findings have potentially important implications for both firefighters working prescribed burnings at chipped vs. non-chipped sites, as well as nearby communities who may be impacted from smoke traveling downwind from these sights.

Effects of fire-generated gaps on growth and reproduction of golden aster (Pityopsis graminifolia)"

only slightly greater (- 10%), however, inside gaps than outside gaps in May-burned blocks four months after the May fires. We therefore conclude that golden aster potentially performs better within fire-generated gaps than within dense vegetation dominated by wiregrass, at least in years without lightning season (May) fires. We suggest, however, that such differences between gaps and dense vegetation likely have little effect on the long term performance of golden aster in pine savannas frequently burned in May.

Tornado and fire effects on tree species composition in a savanna in the Big Thicket National Preserve, southeast Texas, USA

Abstract Ordination showed that species composition in a savanna shifted toward mixed pine-hardwood types after the tornado damage in 1983. Of the twenty 250 m2 study plots, one baygall plot remained unchanged in its position in ordination space, as did five mixed pine-hardwood plots and six savanna plots. However, eight savanna plots moved from the savanna space to the mixed pine-hardwood space after the tornado. Prescribed burns in 1986 and 1991 had a modest effect in reversing this trend: only four of the eight plots returned to the savanna ordination space by 1991. The changes in species composition, summarized by movement of plots in ordination space, reflected the conflicting affects of the tornado and fire: the tornado tended to change savanna to mixed pine-hardwood by differentially removing pines and stimulating hardwood growth. Fires tended to reduce hardwood density. However, stand opening and increased fuel loads following the tornado did not result in fires intense enough to dramatically enhance savanna recovery.

Wiregrass (Aristida beyrichiana) May Limit Woody Plant Encroachment in Longleaf Pine (Pinus palustris) Ecosystems

Abstract Wiregrass (Aristida beyrichiana) is a dominant groundcover species that facilitates fire in southeastern U.S.A. pine savannas, thereby limiting woody plant cover and maintaining a herbaceous dominated understory. In December 1993 two of us planted a plot of wiregrass (Aristida beyrichiana) in the midst of fire-maintained little bluestem (Schizachyrium scoparium) longleaf pine (Pinus palustris) savanna in the outer Coastal Plain of South Carolina. The plot and the surrounding area burned three times in the following 20 y. Vegetation sampling carried out in late summer 2013 indicated wiregrass dominated the plot and the majority of little bluestem had disappeared. The wiregrass plot was comparatively open and grass dominated, whereas the surrounding formerly bluestem dominated stand had filled in with loblolly pine (Pinus taeda) saplings as well as hardwood trees and shrubs. In addition wiregrass had reproduced and established away from the original planted area, most noticeably within a soil-disturbed plow line. A subsequent prescribed fire in spring 2014 burned with higher intensity within the wiregrass plot than in the surrounding area. Our observations suggest suppression of woody plant encroachment by dense wiregrass in pine savannas even during long fire free periods, which should reduce the likelihood of transition to hardwood dominated ecosystems.