Successional changes in plant species diversity and composition after clearcutting a Southern Appalachian watershed

Forest Ecology and Management 92 (1997) 67­85 Successional changes in plant species diversity and composition after clearcutting a Southern Appalachian watershed Katherine J. Elliott ***, Lindsay R. Boring b blC , Wayne T. Swank % Bruce R. Haines d * USDA For. Sen:, SRS, Coweeta Hydralogic Laboratory, Otto, NC 28763, USA Joseph W. Jones Ecological Research Center, Ichawway, Newton, CA 31770, USA ' School of Forest Resources, University of Georgia, Athens, CA 30602, USA d Botany Department, University of Georgia, Athens, CA 30602, USA . Accepted 1 October 1996 and Management Aims and scope. Forest Ecology and Management publishes scientific articles concerned with forest management and conser­ vation, and in particular the application of biological, ecological and social knowledge to the management of man­made and nat­ ural forests. The scope of the journal includes aO forest ecosystems of the world. A referseing process ensures the quality and international interest of the manuscripts accepted for publication. The journal aims to encourage communication between scien­ tists in disparate fields who share a common interest in ecology and natural­resource management, and to bridge the gap between research workers and forest managers in the field to the benefit of both. The journal should be of interest to research workers, managers and policy makers in forestry, natural resources, ecological conservation and related fields. FOUNDING EDITOR Laurence L Roche, Murroe, Ireland EDITORS­IN­CHIEF For the Americas, Australia, New Zealand and the Pacific: R.F. Ffiher Department of Forest Science Texas A&M University College Station, TX 77843­2135, USA EDITORIAL ADVISORY BOARD G. Abrahamsen, Agricultural University of Norway, As, Norway M.R. Ahuja, BFH, Institute of Forest Genetics, Grosshansdorf, Germany R. Alfaro, Canadian Forestry Service, Victoria, B.C., Canada F. Andersson, Swedish University of Agricultural Sciences, Uppsala, Sweden P. Attiwi, University of Melbourne, Parkvflle, Vic., Australia J.C. Caivo, Institute of Technology, Cartago, Costa Rica J.D. Deans, Institute of Terrestrial Ecology, Penicuik, Midlothian, UK R.M. DeGraaf, USDA Forest Service, University of Massachusetts, Arnherst, MA, USA S. Diamandis, Forest Research Institute, ThessalonM, Greece D.P. Dykstra, C1FOR, Jakarta, Indonesia EP. Farreil, University College Dublin, Dublin, Ireland P.H. Freer­Smith, Forestry Commission, Famnam, UK 0. Garcia, ENGREF, Nancy, France . RA Goyer, Louisiana State University, Baton Rouge, LA, USA J.B. Hail, University College of North Wales, Bangor, UK F. Houllier, French Institute of Pondicherry, Pondlcherry, India B.M. Kumar, Kerala Agricultural University, Kerala, India J.P. "Lassoie, Cornell University, Ithaca, NY, USA For the rest of the world: G.M.J. Mohren Forest Production Ecology Group Department of Vegetation Ecology DLO­Institute for Forestry and Nature Research P.O. Box 23 6700 AA Wageningen, The Netherlands J.N. Long, Utah State University, Logan, UT, USA A.E Lugo, International Institute of Tropical Forestry, Rip Pfedras, PR, USA J A Magnembe, SADCC/ICRAF Agrcforestry Project, Zomba, Malawi . F. Makeschin, Institutfur Bodenkunde und Standortslehre, Tharandt, Germany D.C. Malcolm, University of Edinburgh, Edinburgh, UK E Malkonen, Finnish Forest Research institute, Vantaa, Finland MAR. Nahuz, Institute de Pesquisas Tecnologicas, Sao Pauto, SP, Brazil .R. PiMnen, European Forestry Institute, Joensuu, Finland R.F. Powers, Pacific Southwest Research Station, Redding, CA.USA T. Pukkala, University of Joensuu, Joensuu, Finland L Rasmussen, R1SO, Roskilde, Denmark D.D. Reed, Michigan Technological University, Houghton, Ml, USA a Ruark, USDA Forest Service, Washington, DC, USA R. Sands, University of Canterbury, Christchurch, NZ J A Stanturf, Stoneville, MS, USA O. SziWai, University of British Columbia, Vancouver, B.C., Canada K. von Weissenberg, University of Helsinki, Helsinki, Finland Publication information: Forest Ecology and Management (ISSN 0378­1127). For 1997 volumes 90­89 are scheduled for pub­ lication. Subscription prices are available upon request from the Publisher. Subscriptions are accepted on a prepaid basis only and are entered on a calendar year basis. Issues are sent by surface man except to thefollowingcountries where air delivery via SAL mail is ensured: Argentina, Australia, Brazil, Canada, Hong Kong, India, Israel, Japan, Malaysia, Mexico, New Zealand, Pakistan, PR China, Singapore, South Africa, South Korea, Taiwan, Thailand, USA For all other countries airmail rates are available on request Claims for missing issues should be made within six months of our publication (mailing) date. Orders, claims, and product enquiries: please contact the Customer Support Department at the Regional Sales Office near­ est you: New York: EJsevier Science, P.O. Box 945, New York, NY 10159­0945, USA; TeL (+1)212­633­3730, [Ton free number for North American customers: 1­888­4ES­INFO (437­4636)], Fax (+1)212­633­3680, E­mail usinfo­f@eisevier.com Amsterdam: EJsevier Science, P.O. Box 211,1000 AE Amsterdam, The Netherlands; Tel. (+31)20­485­3757, Fax (+31)20­485­ 3432, E­mail nilnfo­f@elsevler.nl Tokyo: Hsevier Science, 9­15, HIgashi­Azabu 1­chome, Minato­ku, Tokyo 106, Japan; Tel. (+81)3­5561­5033, Fax (+81)3­ 5561­5047, E­mail kyf04035@niftyserve.or.jp Singapore: Bsevier Science, No. 1 Temasek Avenue, #17­01 Millenla Tower, Singapore 039192; Tel. (+65)434­3727, Fax (+65)337­2230, E­mail asiainfo@elsevier.com.sg Forest Ecology and Management Forest Ecology and Management 92 (1997) 67­85 Successional changes in plant species diversity and composition after clearcutting a Southern Appalachian watershed Katherine J. Elliott *­*, Lindsay R. Boring bf , Wayne T. Swank a, Bruce R. Haines d * USDA For. Sen., SRS.Caweeta Hydrologic Laboratory, Otto, NC 28763, USA Joseph W. Jones Ecological Research Center, Ichauway, Newton, CA 31770, USA • " School of Forest Resources, University of Georgia, Athens, GA 30602, USA . 4 Botany Department, Uniaersity cf Georgia, Athens, CA 30602, USA b '­• • ' • • Accepted 1 October 1996 Abstract Watershed 7, a southwest­facing watershed in the Coweeta Basin, western North Carolina, USA, was clearcut in 1977. Twenty­four permanent plots were inventoried in 1974 before cutting and in 1977,1979,1984, and 1993 after clearcutting. This study evaluates changes in species diversity during early succession after clearcutting and differences in overstory tree and ground flora response to disturbance by clearcutting and their interaction with previous disturbances and subsequent stand development To quantify species diversity, we computed Shannon­Weaver's index of diversity (HO and Pielou's evenness index (/')­ Woody species diversity remained relatively stable; however, woody species richness increased in the cove­hardwoods and hardwood­pines, but remained relatively constant in the mixed­oak hardwoods. Although revegetation was rapid, forest composition has changed through succession. Opportunistic species, such as Uriodendron tulipifera, Robinia pseudoacacia, and Acer rubnan, increased in abundance, whereas Quercus velutina, Carya spp., and Q. rubra decreased. Ground flora diversity declined in the cove­hardwoods and mixed­oak hardwoods communities, but the decrease in the hardwood­pines was not significant The abundance (g biomass m~2) of ground flora was much lower in 1993 man in 1984; 79% less in the cove­hardwoods, 90% less in the mixed­oak hardwoods, and 79% less in the hardwood­pines. Watershed 7 is apparently in a transition state between early and late successional species abundance. Early successional, shade­intolerant species, such as Erechtites, Solidago, Eupatorium, Panicum, and Aster, have declined, whereas late successional, shade­tolerant species, such as Viola, Calami, Sanguinaria, Uvularia, and Veratrum are not yet well established. ­.• . ' • ' ' ' •' •* . . * " • • • i Keywords: Stand dynamics; Herbaceous and woody flora; Disturbance 1. Introduction Maintenance of species diversity has become, an important topic in forest management studies (Norse . • * Corresponding anthor. et al., 1986; Hunter, 1990; Burton et al., 1992), with special emphasis on understanding the role of van­ ous species in. the recovery of forest structure and processes (Boring et al., 1981; Schoonmaker and McKee, 1988; McMinn, 1991; Huston, 1994). The effects of human­induced disturbances in forested ecosystems on forest regeneration, structure, produc­ 0378­1127/97/$17.00 Copyright © 1997 Published by Hsevier Science B.V. AS. lights reserved. PU 80378­1127(96)03947­3 68 KJ. Elliott et aL /Forest Ecology and Management 92 (1997) 67­35 tivity, and diversity vary with frequency, intensity, and scale of disturbance (Runkle, 1985; Petraitis et aL, 1989; Huston, 1994). Although no single general­ ization prevails for describing changes in species richness and diversity through succession, eastern forest systems tend to increase in both measures after forest harvesting men decline as forests mature (Bi­ cknell, 1979; ffibbs, 1983; Reiners, 1992; Roberts, 1992; Wang and Nyland, 1993). Huston and Smith (1987) described succession as a sequential change in the relative abundances of the dominant species hi a community. During early suc­ cession, physiological characteristics, such as stress tolerance, rapid growth rate, or high nutrient acquisi­ tion, may influence species "abundance. Later hi suc­ cession, size and shade tolerance may emerge as the physiological characteristics mat affect species abun­ dance. Species are also capable of changing men* competitive ability when conditions change, but they are unable to adapt to all successions! stages or environmental conditions (Huston and Smith, 1987). This description implies that certain species or groups of species will lose dominance unless a disturbance or environmental change interferes. Thus, some suc­ cessions! stages may have more species, as well as different sets of species, man others (Hunter, 1990). However, forests are always changing owing to natu­ ral disturbances, such as wind, fire, drought, or single or multiple tree mortality, mat may create canopy gaps with earlier stages of succession. There­ fore, competitive equilibrium or steady state rarely occurs (Huston, 1979). Many generalizations for successions! change have been inferred from analysis of chronosequences of stands representing different ages (Peet and Loucks, 1977; Finegan, 1984; Roberts and Chris­ tensen, 1988). JHowever, variation among forest stands along a chronosequence can arise from inter­ acting sources including historical factors (e.g. dis­ turbance, variations in seed rain), site environment (e.g. climate, slope, aspect, and soil variables), and autogenic successions! change. The most direct and unambiguous method of documenting succession in­ volves measuring changes in a single site through. time. Because the time scale is long,' few studies have used this approach (e.g. Peet and Christensen, 1980; Hibbs, 1983; Hartnett and Krofta, 1989; Rein­ ers, 1992; Fain et aL, 1994). For the past several decades, experimental clearcuts have provided an opportunity to examine how these large­scale forest disturbances influence various processes, such as stream hydrology (Swank and Helvey, 1970; Likens et aL, 1977; Swank et aL, 1988), soil erosion (Hewlett, 1979; Van Lear et al., 1985), nutrient cycling (Johnson and Swank, 1973; Bormann et al., 1974, Bormann et al., 1977; Likens et al., 1977; Gholz et aL, 1985; Boring et al., 1988; Waide et al., 1988; Reiners, 1992), and vegetation diversity and successional patterns (Parker and Swank, 1982; Gholz et al., 1985; Hornbeck et al., 1987; Boring et aL, 1988; Reiners, 1992; Gove et al., 1992; Elliott and Swank, 1994a). In a regeneration project conducted in a clearcut watershed in the Coweeta Basin, southwestern North Carolina, studies were conducted 1, 3, and 8 years after disturbance (Boring, 1979; Boring et al., 1981, Boring et al., 1988; Boring and Swank, 1986). These studies ex­ amined the role of dominant early successional species in forest recovery and ecosystem processes, but did not address longer­term species patterns, diversity, and richness. In this study, we analyze successional patterns in composition and diversity of herbaceous and woody species hi the same clearcut watershed to age 17 years. Our objectives were to describe changes in species diversity during early succession after clearcutting, and evaluate differ­ ences in overstory and ground flora vegetation re­ sponse to disturbance by clearcutting. 2. Methods 2.1. Site description The study site, a 59 ha watershed (WS7), is lo­ cated hi the Coweeta Hydrologic Laboratory (35°04/r30ffN, 83°26'W) near Franklin, NC. The Coweeta Basin is in the Nantahals Mountains—part of the Blue Ridge province hi the Southern Ap­ palachians. Watershed 7 has a south­facing aspect and ranges hi elevation from 720 to 1065m. Slopes range from 23 to 81%. Parent rocks of schist and gneiss have weathered to form deep soils with rock outcrops present on steep slopes at high elevations (Hatcher, 1974). At lower elevations, the dominant soil series is the Tusquitee, a member of the fine­ KJ. Elliott et aL/Forest Ecology and Management 92 (1997) 67­85 loamy, mixed, mesic family of Humic Hapludults. The ridge and slope soils are dominated by the Chandler series, a member of the coarse­loamy, mi­ caceous, mesic family of Typic Dystrochrepts (Thomas, 1996). The mean annual temperature is 13°C, and average temperatures are 6.7°C in the dormant season and 1830C in the growing season. Mean annual precipitation is 183cm (Swift et al., 1988). The land­use history in the Coweeta Basin in­ cludes selective logging, woodland grazing, and burning. Before 1842, Cherokees burned semi­an­ nually to improve forage for livestock. Between 1842 and 1900, European settlers moving into the area also burned and grazed the basin. A few hectares in WS7 were probably cultivated around 1901. Be­ 69 tween 1900 and 1923, logging operations occuired over the entire basin, but catting was heaviest on the lower slopes, valleys, and accessible coves. Since 1924, human disturbances have been restricted to experimental studies (see Douglass and Hoover (1988) for a complete description of the history of the Coweeta Basin). In a woodland grazing experi­ ment in WS7 between 1941 and 1952, six head of cattle were used to assess the impact of woodland grazing on a portion of die watershed. Short­range effects were limited primarily to soil compaction and overgrazing in the cove area adjacent to the stream (Johnson, 1952; Williams, 1954). Watershed 7 was clearcut in 1977 as part of an interdisciplinary study of the physical, chemical, and biological effects on both terrestrial and aquatic C o w e e t a W a t e r s h e d 7 Plots 100 f t Contours S t reams 140Q Fig. 1. Topographic map of plot locations including streams in Watershed 7, Coweeta Basin, western North Carolina, USA. 70 KJ. ElUott etaL/Forest Ecology and Management 92 (1997) 67­83 components of the ecosystem (Swank and Caskey, 1982). Harvesting, begun in January 1977, was com­ pleted in June. Tractor skidding was used on slopes less than 20% (about 9 ha), and yarding with a mobile cable system on the remaining area. In the cutting operation, marketable timber was removed by cable logging. Most of the ridgetops and xeric slopes were cut, but were not cable logged because the volume of marketable timber was insufficient All stems of 2Jem or more dbh (diameter at breast height) were cut and logging debris was left in place with no further site preparation. This, harvest tech­ nique minimizes soil compaction and other structural disturbances of the forest floor and plant roots. were dropped and 11 plots from the remaining 124 were added to total 24 permanently marked plots. Sample sizes were increased to reflect relative areal coverage of each community within the watershed. Seven plots represented the cove­hardwoods, five the mixed­oak hardwoods, and 12 die xeric hardwood­ pines (Fig. 2). The 24 plots were remeasured hi subsequent years (1979, 1984 and 1993) to observe the changes in vegetation composition through suc­ cession. Two quadrats were located in opposite corners of each 0.08 ha plot Hardwood sprouts were sampled in 7 m X 7 m subplots and seedlings were sampled hi 3 m X 3 m subplots; values were pooled for each pair. To understand the mode of reproduction, each 2.2. Sampling procedures woody stem was classified into one of two cate­ gories: sprout, if it originated from a previously Before clearcutting, vegetation was inventoried established stump or root system; seedling, if it from 142 plots of 20 m X 40m systematically located originated from seed since clearcutting or was a over WS7. Based on previous studies (Williams, single stem from advance regeneration. This differ­ 1954; Day et aL, 1988), three community types were entiation may overestimate seed origin reproduction, identified in WS7: (1) cove­hardwoods found at particularly because root sprouts are difficult to dis­ lower elevations and along ravines at intermediate tinguish from seed origin without partial exposure of elevations; (2) mixed­oak hardwoods on mesic the root systems and because many established small southeast­facing and north­facing slopes at interme­ root systems may send up single sprouts. Robinia diate elevations; (3) hardwood­pines on xeric south­ pseudoacada sprouts were distinguished by their west­ and south­facing slopes at intermediate to up­ attachment to lateral roots. per elevations and ridgetops. Plots were classified At the end of each growing season, densities of into community types based on detrended correspon­ sprouts and seedlings were recorded separately by dence analysis (DCA) (Gauch, 1982) mat used pre­ species and diameter class on each sample quadrat cut wood vegetation data from 1974 for 24 perma­ Diameter classes were designated by 0.5 cm intervals nently marked plots. The cove­hardwoods commu­ up to a maximum of 3 cm in the first year (1977) and nity had high numbers of Rhododendron maximum, by 1.0cm intervals up to a maximum of 8cm for Hamametis virginiana, Unodendron tulipifera and years 1979 and 1984. Different species were mea­ Betula lento, with some mesic species such as Tilia sured at 3 and 40cm from ground level depending heterophylla and Aesculus octandra. The mixed­oak on the species' potential growth rates. The 3.cm hardwoods community had high numbers of diverse . measurement gave the best fit for coupling bioraass oak species, Liriodendron tulipifera, Comus florida regression equations for slow­growing species; 40cm and no or low densities of understory ericaceous was best for fast­growing species (Boring et al., shrubs, such as Rhododendron maximum or Kalmia 1981). In 1993, woody stems with a dbh of 1.0 cm or more were measured to the nearest 0.1 cm at 137m latifolia. The hardwood­pines community had high from ground level. Stems with less than 1.0cm dbh numbers of Quercus prinus, Q. coccinea and K. latifolia, and scattered Pinus rigida and Prunus were measured to the nearest 0.1 cm at 3 and 40cm from ground level. serotina (Rg. 1). • * Mid­point values of each diameter class multi­ After cutting in 1977, 18 of the 142 plots were plied by the number of stems in that class were used sampled for regrowth: eight in the cove­hardwoods, to calculate basal area for years 1977­1984. Basal five in the mixed­oak hardwoods, and five in the area of saplings estimated .from diameters at 3 cm hardwood­pines. In 1978, five plots from the 18 KJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­85 from the base would overestimate basal area. Be­ cause sapling basal areas (years 1977, 1979, and 1984) were estimated entirely from diameters mea­ sured at 3 and 40cm from the base, the values were exaggerated compared with tree basal areas of stems of more than 1cm dbh measured at conventional breast height (1.37m). After clearcutting, all herbaceous vegetation was harvested in August each year (1977­1993) from one randomly placed 1.0m2 subplot within each quadrat Vegetation was separated by species and oven­dried to constant weight at 70°C. All species identification followed nomenclature consistent with Radford et al. (1968). A woodland grazing experiment conducted from 1941 to 1952 (Williams', 1954) furnished the only 71 data available on herbaceous species presence and abundance in WS7 before clearcutting. Williams also provided insights into the impact of traditional land­ use activity on species diversity. Data collected in 1952 from 17 ungrazed, fenced plots containing two 4.04m2 subplots provided information on understory plants 30 years after recovery from the selective log­ ging that occurred from 1900 to 1923 and before the large­scale clearcutting in 1977. This data allowed a qualitative comparison of herbaceous species pres­ ence and abundance before and after clearcutting. 2.3. Data analysis To evaluate species diversity, Shannon­Weaver's index of diversity (#0 (Shannon and Weaver, 1949) UBTUL TILHET CORFLO • 54 • 55 QUERDB AESOCT BETLEN • 59 BHOMA3 NYSSYL ROBPSE 13 • 3• 127 AOtmTBtTS CABSPP ,20 QDECOC • 12 PRIJSEH SASALB KALLAT HAMVIR FAGGRA Fig. 2. Detrended correspondence analysis of the 24 permanent plots along the first two ordination axes with location of species along ordinations. B, Cove­hardwoods; 9, mixed­oak hardwoods; A, hardwood­pines. Species codes: URTUL, Imodendron tulipifertr, ACERUB, Acer rubruat, KALLAT, Kabrda latifolia; RHOMAX, Rhododendron maximum; QUECOC,. Quercus cocciaea; QUEPRI, Quercus prims; QUERUB, Quercus rubra; CORFLO, Comus florida; AESOCT, Aesadus octandnr, HAMVIR, Hamamelis uirginiana; BETLEN, Betula lento; FAGGRA, Fagus grandifblia; NYSSTfL, Nyssa sylcatica; ROBPSE, Robiniapseudoacada; CARSPP, Carya spp.; TILHET, Tilia heterophyUa; PINRIG, Pinus ritfda; PRUSER, Primus serotina. 72 KJ. Elliott etaL/Forest Ecology and Management 92 (1997) 67­85 and Kelou's evenness index (/') (Pielou, 1966) were computed. Shannon­Weaver's index is a simple quantitative expression that incorporates both species richness and the evenness of species abundance. Because the calculated value of H' alone does not show the degree to which each factor contributes to diversity, a separate measure of evenness (/') was calculated. Diversity was calculated on the basis of stem basal area per hectare for woody species and biomass per square meter for herbaceous species: H' = —Spiln. p­t, where pt is the proportion of total basal area of species i. Species evenness was calcu­ lated as J' = H'/H'wj., where fl^ax 'K *e maximum level of diversity possible within a given population, which equals ln(number of,species). We used pair­ wise f­tests (Magurran, 1988) to examine the differ­ ences in diversity between sampling years from 1974 to 1993. No statistical tests were performed for 1952 because ground flora measurements were based on density rather than biomass. 3. Results 3.1. Changes in woody species In each community, more man IS tree species regenerated after clearcutting. Shade­tolerant species were Acer rubrum, Nyssa sylvatica, Fagfis grandi­ folia, Cornus florida, Tsuga canadensis, Oxyden­ drum arboreum, Amelanchier arborea, and HamameUs virginiana. Species with intermediate shade tolerance included Carya spp., Fraxmus amer­ icana, Quercus prinus, Q. rubra, and Q. vehuina. Shade­intolerant species included Liriodendron tulipifera, Betula lento, Robinia pseudoacacia, and Q. cocdnea (Bums and Honkala, 1990). Species regenerating inirequently were Tilia heterophyUa, Diospyros virginiana. Sassafras albidum, Symplocos tinctoria, Pnmus serotina, and A pensylvanicum. In 1974, Carya spp., Q. rubra, and L tulipifera were the three most abundant tree species in the cove­hardwoods community (Table 1). After clearcutting in 1977, C. florida, A. rubrum, and L tulipifera became the most dominant tree species/ The woody vine, Vitis spp., was more dominant than L. tuliptfera. By 1984, Vitis spp. began to lose its dominant position in the community, and R. maxi­ mum became the most abundant species. By 1993, L. tulipifera was the leading dominant species at 22% of the total basal area. In the mixed­oak hardwoods community before cutting, Q. vehaina, L. tulipifera, and Carya spp. were the most abundant species, occupying 56% of the total basal area. After cutting in 1977, C florida, Vitis spp., and L tulipifera made up 64% of the basal area. C. florida remained the leading dominant to 1984. L. tulipifera regained its dominance by 1993 with 44% of the total basal area, and R. pseudoacacia became increasingly more important in the community. Q. velutina did not regain its domi­ nant position and made up less than 1% of the total basal area from 1977 through 1993. In the hardwood­pines community before cutting,. Q. prinus, K. latifolia, and Q. cocdnea were the three most abundant species with 63% of the basal area. Vitis spp. became important after disturbance but declined rapidly. K. latifolia remained dominant after cutting and increased in importance through succession. A. rubrum increased in importance 2 years after cutting and remained dominant to 1993 (Table 1). C. florida and R. pseudoacacia were more abundant immediately after clearcutting, but began to decline by 1979. With 60% of the total basal area, K. latifolia, Q. prinus, and A. rubrum were the leading dominants in 1993. Some woody species were present after clearcut­ ting, but not recorded in the overstory woody mea­ surements before clearcutting. Most of these species were shrubs and vines. Because only stems with dbh of 2Jem or more were recorded in 1974, small stemmed shrubs and vines, such as C. florida, E. americanus, P. pubera, and Vitis spp., were proba­ bly not recorded because they were small, not ab­ sent la 1974, stem density in the hardwood­pines com­ munity was more than two times greater than.in the other two communities (Table 2). This higher density was attributed primarily to K. latifolia, which grows on the upper slopes and ridges of the watershed. K. latifolia contributed 62% of the density and 19% of the basal area in the hardwood­pines (Table 1). Without K. latifolia, stem density in the hardwood­ pines community would have been 1523 stems, and basal area 22.25 m2 ha"1, which would have been lower than the other two communities. SJ. Ettiott et aL/Forest Ecology and Management 92 (1997) 67­85 Density increased substantially in all.communities following harvest, with 24­46 times more stems per hectare in 1977 than in the pre­cut forest By 1993, densities were stfll 6­9 times greater than in the pre­cut forest The 17­year­old forest (1993) of WS7 Table 1 Leading dominant woody species (more than 2% of basal area in • any year) ordered by sequence of maximum percentage contribu­ tion to basal area in 1974 Year Species Pre­cnt Post­cut 1977 1979 1984 1993 1974 Cove­hardwoods 1.4 12 1.9 0.9 Carya spp. ­ 18.0 62 45 4.0 5.7 Quercus rubra i5.r 10.7 11.1 6.4 21.6 Liriodendron tulipifera ­12.0 Betulalenta 7.8 03 5.0 73 1ZO 6.6 9.4 20.9 112 Rhododendron maximum 7.8 TUia heterophytta 6.8 03 0.8 03 0.6 0.7 63 02 0.4 03 Quercus prinus 9.8 5.7 Acerrubrum 6.0 145 11.7 5.0 0.1 0.1 0.4 0.01 Quercus alba 3.8 0.0 0.1 0.0 0.04 Aescuhts octandra 18.4 15.1 113 6.6 Comusflorida ZS 22 13 23 73 Tsuga canadensis 23 Fagus grandifolia 15 Z4 1.7 1.0 1.4 Vitisspp. 0.0 1ZO 1Z4 6.1 1.1 0.4 3.7 5.8 6.6 35 Hamamelis oirginiana 02 33 22 13 3.8 Fraxinus americana 5.1 4.1 4.6 13 Kalmia latifoiia 02 Amelanchier arborea 0.0 23 1.6 1.0 03 0.6 22 23 13 0.02 Nyssa sylvatica 0.0 1.8 32 13 93 Robinia pseudoacacia 1.7 13 0.1 Z7 Oxydendrum arboreum 05 Total 97.9 955 96.1 89.0 96.8 Mixed­oak hardwoods 20.9 Quercus velutina 0.1 03 03 0.6 14.7 13.1 7.1 43.7 Liriodendron tulipifera 183 62 12 1.4 03 Carya spp. 16.7 123 33 1.0 23 Zl Quercus prinus 63 43 8.8 82 83 Acerrubrum 272 29.8 31.8 123 Comusflorida "^­ '^ 83 Nyssa sylvatica 5.7 Z4 13.1 4.1 12 33 0.4 13 ZO 13 Oxydendrum arboreum . Z4 Quercus rubra 5.1 32 3.1 3.6 6.4 93 10.1 213 Robinia pseudoacacia 1.9 6.0 1.0 Vitisspp. 0.0 21.9 11.0 Castanea dentata 0.6 33 0.6 Z8 0.0 0.0 0.0 15 3.9 02 Kalmia latifoiia 0.6 62 6.0 13­ 0.0 Sassafras albidum 0.04 0.0 13 32 0.1 Rhododendron calendulaceum Total 993 98.1 98.0 93.1 97.6 Table 1 (continued) Species Hardwood­pines Quercus prinus Kalmia latifoiia Quercus coccinea Acerrubrum Oxydendrum arboreum Quercus aelutina Nyssa sylvatica Carya spp* Quercus alba Comusflorida Pinusrigida Robinia pseudoacacia Liriodendron tulipifera Castanea dentata Vifespp. Quercus rubra Symplocos tinctoria Sassafras albidum Rhododendron m&xiinum Pyntlaria pubera Total­ 73 Year Pre­cut Post­cut 1977 1979 1984 1974 26.0 18.8 18.0 6.8 53 4.1 32 Z7 2.6 2.6 2.2 1.8 1.8 1.0 0.0 0.7 0.0 02 0.8 0.0 98.8 6.0 20.5 0.9 6.9 5.9 0.0 5.0 1.2 0.0 9.9 0.0 8.0 1.1 5.6 21.0 2.4 1.6 IS 0.0 0.0 97.5 6.4 37.1 5.7 119 25 03 5.9' 1.7 0.6 4.0 0.0 23. 1.2 3.6 3.9 0.1 Zl 22 1.8 a? 94.9 5.6 34.7 4.8 11.7 1.0 0.2 4.8 1.4 0.4 5.1 0.02 23 1.7 • 7 33 Z5 0.6 22 2.0 63 3.1 95.7 1993 222 233 8.0 142 23 02 2.6 1.6 0.7 1.7 0.1 3.4 62. 2.8 02 0.88 03 1.4 4.8 0.1 97.1 Sample years begin in 1974 (before clearcutting) through succes­ sional time (after clearcutting in 1977) for three communities in WS7, Coweeta Basin. In 1977, number of sample plots was eight in the cove­hardwoods, five in the mixed­oak hardwoods, and five in the hardwood­pines. In 1974,1979,1984, and 1993, number of sample plots was seven hi the cove­hardwoods community, five in die mixed­oak hardwoods, and 12 in the hardwood­pines. In 1974 and 1993, woody stems with a dbh of 1.0cm or more .were measured at 137cm from the base, and stems with a dbh of less than 1.0cm were measured at 3 and 40cm from the base (Boring et aL, 1981). In 1977, 1979 and 1984, all woody stems were measured at 3 and 40cm from the base (Boring et aL, 1981). Species nomenclature follows Radford et aL (1968). is still aggrading, with most woody stems in smaller size classes (Fig. 3). In the cove­hardwoods, 58% of the density and 12% of the basal area were from stems of less than 2.5cm dbh; in the mixed­oak hardwoods, 35% of the density 'and 3% of the basal area were from stems of less than 25 cm dbh; in the hardwood­pines, 89% of me density and 33% of the basal area were from stems of less than 2Jem dbh. Densities of stems with dbh greater than 5.0cm were 2609ha­1, 2405ha­1, and 1811 ha"1, occupying 1824 m? ha'1, 1957 m2 ha"1, and 11.47m2 ha'1 74 KJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­85 basal area in the cove­hardwoods, mixed­oak hard­ woods, and hardwood­pines, respectively. Although the number of woody species present increased in the cove­hardwoods and hardwood­pines communities after clearcutting, differences in diver­ sity were not significant (H'; PrsO.lO* based on pairwise f­statistics) among years. In the mixed­oak hardwoods, the number of species decreased imme­ diately after clearcutting (Table 2), but the difference in H' was not significant However, a significant decline in H' (based on a pairwise r­statistic; Magurran (1988) (^,33, Ps0.05 = 2.059)) did occur in the mixed­oak hardwoods community from 1984 to 1993. This decline was attributed primarily to the increased dominance of two species, L. tulipifera and R. pseudoacada, which occupied 65% of the total basal area in 1993, and the reduced basal area of Q. prinus. 3.2. Origin of woody species reproduction In the cove­hardwoods, 58% of stems originated from seedlings in 1977; however, by 1979 seedling and sprout reproduction were about equal, probably a result of heavy mortality of seedlings. In the mixed­ oak hardwoods, seedling and sprout reproduction were about equal in 1977, but by 1979 sprouts accounted for 69% of the reproduction. In the hard­ wood­pines type, sprout reproduction was higher than seedling reproduction, with 64% and 82% of the stems originating from sprouts in 1977 and 1979, respectively. Community type often affected a species' primary mode of reproduction. Species mat regenerated pri­ marily by sprouting in all three communities were Castanea dentata, C. florida, N. sylvatica, 0. ar­ boreum, and R. pseudoacada. Carya spp. and Q. prinus regenerated primarily by sprouting in the mixed­oak hardwoods and hardwood­pines types. Rhododendron maximum and IL kttifolia regener­ ated almost entirely by sprouting in me cove­ hardwoods and hardwood­pines. A. rubrum regener­ ated primarily by seed germination in the cove­ hardwoods, seedling and sprout reproduction were about equal in the mixed­oak hardwoods, and the dominant mode of regeneration in the hardwood­ Table 2 . Average density, basal area, diversity (H'; based on basal area) and evenness (j") of Coweeta Basin, for sample years 1974,1977,1979,1984, and 1993 Community F S Year G Density ­ Basal area (m2ha­') 14 Cove­hardwoods 1974* 12 13 23.67 1566 1977 20 23 28 72507 4.57 20 24 32 65979 1979 751 1984 23 29 36 70294 13.68 1993 21 27 36 13267 24.85 Mixed­oak hardwoods 1974" 16 21 26 1762 24.87 14 17 20 76236 1977 526 11 14 19 55685 734 1979 ­..,_ 1984 15 18 22 43593 9.21 «'• . ­^•^ 1993 13 14 22 9993 23.78 Hardwood­pines 19 1974" 13 16 3970 27.47 19 93416 1977 13 15 6.05 15 19 25 93551 9.23 1979 20 24 30 98189 1984 1633 1993 19 25 36 35573 21X50 woody species for three commnnities in WS7, H' Variance of Z52 Z64 Z73 Z75 Z57 Z13 Z12 Z22 Z47 1.76 Z28 Z41 237 Z49 235 0.017 0.134 0.017 0.032 0.022 0.015 0.017 0.044 0.058 0.050 0.035 0.161 0.099 0.066 0.038 dof H' ±0.056 ±0.138 ±0.046 ±0.055 ±0.050 ±0.063 ±0.067 ±0.098 ±0.101 ±0.099 ±0.077 ±0.030 ±0.127 . ±0.091 ±0.062 f 0.804 . 0.784 0.781 0.756 0.717 0.752 0.748 0.741 0.777 0.569 0.708 0.804 0.727 0.718 0.637 In 1977,1979, and 1984, basal area was calculated from diameter measurements at 3 or 40cm above ground line depending on species' potential growth rates (Boring et aL, 1981). In 1993, diameters of trees with dbh of 1.0cm or more were measured at 137m above ground line and samplings with dbh of less than 1.0cm were measured as in 1977,1979, and 1984. • . F, Total number of families present in each commnnity, G, total number of genera in each community, S, total number of species present in each community. * Sampling in 1974 included woody species with dbh of Z5 cm or more; diameter was measured at 137m above ground line. KJ. Elliott et aL /Forest Ecology and Management 92 (1997) 67­83 pines was sprouting. A. arborea, Q. coccinea, and L. tulipifera regenerated primarily from seed in aJH three communities. B. lento, absent in the mixed­oak a)} 75 hardwoods and hardwood­pines communities, regen­ erated predominantly from seed hi the cove­ hardwoods. Q. velutina, a dominant species in the 12000 ­ 10000 ­ 1974 C~1 1993 <3> 8000 ­ | 6000 ­ 11, 4000 ­ '4? 2000 ­ g ­I 0 40^ 300 .200 ­ 100 • 0 m r—i 1 n » LO­rS 2J4­7^3 7.7­117 yt&VlA 173­12J ZU­2SJ) 2S.1­33.1 33J­3S.I 3&2­C.7 43.7+ b) 12000 10000 1? 8000 | 6000 It 4000 •t 2000 t •° 400­ 300 ­ 200 ­ 100 ­ 0 J3^| 1974 I 1993 1 n.n _ M ­— gj . ­ H H!~lM—. B H _ _ _ m UJ­2J 2J4­7^5 7.7­12.7 1Z3­17J 17J­213 ILO­23J) 13.1­33.1 33J­3W 3&Z­43.7 43.7+ 0 M C) . ^ '§ ^• g , .1? Q 24000 ­ 20000 16000 .. 12000 ­ 8000 ­ 4000 ­ 408hr— 30C ­' 20C ­ IOCI ­ ri L. S^a 1974 1 1 1993 1 I 1 _ „ 1JMJ m m\~\K3 7.7­U.7 ttS­17^ sa 23J5­JSJ) K.l­33.1 312­38.1 3&2­Q.7 43.7+ Diameter size class (era) Fig. 3. Size class distributions of stems (1.0cm or more dbh) in three comrannities in Watershed 7, Coweeta Basin for years 1974 and 1993: (a) cove­hardwoods; (b) mixed­oak hardwoods; (c) hardwood­pines. Diameters were measured at 137 m above ground line. ' ' 7,6 KJ. Ettiott aaL/ Forest Ecology and Management 92 (1997) 67­85 Table 3 Leading dominant gronnd flora species (2% or more of total biomass in any year) ordered by sequence of maximum percentage contribution to biomass through successional time, after clearcutting in 1977, for three communities in WS7, Coweeta Basin Biomass (%) Density Species 1952 1984 . 1977 1979 199.3 Cove­hardwoods 02 1Z4 192 3.4 03 Parthenocissus quinquejblia 4.1 5.6 11.8 83 0.8 Asters (.divaricatus, acwninatas, and undulatus) Z4 2Z5 0.1 03 113 Viola cucuUata 0.0 0.0 0.0 7.7 02 Erechtitis hieracifolia 1.6 19 72 11.1 6.8 Solidago spp. (mostly odora and curtisii) 6,5 0.0 0.0 19.0 1.1 Pardcum spp. 0.0 0.0 0.0 4.4 0.0 Acafypha rhomboidea 78.4 3.4 3.5 Zl 192 Smilax rotundifalia 0.0 3.0 32 62 0.9 PotentiHa canadensis 0.0 1.6 0.0 0.0 3.0 Eupatorium rugosum 0.0 4.7 a? 0.7 22 Botrychium virgaaanum 0.0 3.4 14.6 0.0 ZO Poiystichum acrostichoides ' 0.0 1.8 0.0 0.0 1.7 TiareOa cor&jotia 30.0 46.6 0.8 0.1 0.0 Kabus spp. (mostly attegheniensis) 0.0 7.7 0.0 0.5 0.7 Houstorda purpursa 0.0 0.0 02 1.6 18.0 Desmadium.nudiflorum 0.01 0.0 3.4 0.0 02 Monarda clinopodia 0.0 0.0 0.0 32 0.7 Camcifuga racemosa 0.0 0.0 3.1 0.0 0.0 Poaspp. 0.0 0.0 0.0 13.1 0.0 Cattum circaezans 0.04 0.0 0.0 0.0 52 Sangutnaria canadensis 0.04 3.6 13.4 US 1.1 Unidentifiable 96.5 973 97.8 913 97.9 Total Mixed­oak hardwoods 8.5 6.4 12 54.0 4.6 Solidago spp. (mostly odora and curtisii) 103 6.7 Z9 0.0 15.7 Eupatorium rugosum Z9 0.0 0.0 Z9 6.7 Viola cucuUata ^ 1.7 42 6.5 12 4.7 Asters (.dioancatus, undulatus, and acianuiatus) 0.0 0.0 0.0 0.0 4.7 Galumlatifoliian 0.0 0.1 4.1 0.4 0.0 Botrychium virginianum 0.0 0.0 23 0.0 03 Potentitta canadensis 0.0 0.0 4.4 82 0.0 Pardcum spp. 0.0 30.4 37.1 0.0 0.0 Rubus spp. (mostly attegheniensis) 0.0 0.0 27.4 0.0 0.0 Clematis oirginiana 0.0 0.0 0.0 4.9 0.0 Monarda clinopodia 65.1 5.4 0.0 0.0 43 Parthenocissus quinquefotia 0.0 02 0.0 ZO 1.1 Desmodumnudiflorum 0.0 11.5 0.0 0.6 0.0 Cinticifuga racemosa ' ^^ 0.0 03 0.0 0.0 202 Vaccattumoadnans 0.0 0.0 283 0.0 0.0 Poiystichum acrostichoides 6.6 0.0 1.1 0.0 83 Smilax glauca 0.0 0.0 0.0 8.0 0.0 Epigaea repens 0.0 0.0 . 0.0 0.0 32 Gumaphilamaculata 0.0 0.0 0.0 Z5 0.0 Houstorda purpurea 0.0 0.0 0.0 0.0 62 Prenarahes spp. 0.0 0.0 0.0 0.0 ZO * . RtieUiacUiosa 0.0 0.0 0.0 0.0 23 Uvularia pudica 0.0 0.0 0.0 0.0 10.0 Vaccinium stamineum KJ. ETUott etaL/ Forest Ecology and Management 92(2997) 67­85 Table 3 (continued) Species Vtir aifutn parvi/lonsm Unidentifiable Total Hardwood­pines Solidago spp. (mostly odora and curtisii) Parthenocissus quinquefoUa Smilax rotundifoUa Eupatorium rugosum Vaccinium vacillans Asters (dwaricatus, wndulatus, and acutninatus) Paniaon spp. Potentffla canadensis Viola atcuttata Rubus spp. (mostly aUegheniensis) Helunahus microcephalus Epigaea repens Galaxaphytta Coreopsis major SfnUox slauca Pteridiam aquilinum Chimaphila maadata Unidentifiable Total Density ' 1952 23 0.6 79.6 13 0.0 0.1 0.2 33.2 0.4 ZO 0.7 0.7 0.1 0.0 6.0 19.0 0.0 1Z1 02 43 0.9 813 77 Biomass (%) 1977 0.0 7.0 96.9 1979 0.0 0.9 96.6 1984 0.0 0.6 963 1993 0.0 03 95.4 41.1 16.2 92 7.8 3.7 ZS Z6 Z5 Z2 0.8 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.8 95.7 .0.04 0.0 19.6 0.04 13.9 0.4 23.0 4.6 03 13.4 6.6 3.0 Z2 Z2 0.4 0.0 0.0 Z4 96.1 0.9 0.05 49.8 0.0 0.0 0.2 0.5 0.0 0.66 23.4 0.0 1.6 6.5 02 0.6 73 0.0 Z8 94.1 0.6 0.0 33.5 0.8 2Z7 0.0 0.1 0.1 13.1 4.6 0.5 0.4 14.8 0.0 13 0.0 0.1 0.01 963 . In 1952 (before clearcutting), dominant species were based on 2% or more of total density. la 1952, number of sample quadrats was four for the cove­hardwoods community, 14 for the mixed­oak hardwoods, and 16 for the hardwood­pines, with a sample area of 4.0 m2 per quadrat. In 1977, number of sample quadrats was eight in the cove­hardwoods, five in the mixed­oak hardwoods, and five in the hardwood­pines, with a sample area of two (1.0m2) quadrats. In 1979, 1984, and 1993, number of sample quadrats was seven for the cove­hardwoods community, five for the mixed­oak hardwoods, and 12 for the hardwood­pines, with a sample of two (1.0m2) quadrats. Species nomenclature follows Radford et aL (1968). ' • mixed­oak hardwoods community before cutting, re­ produced only by seed germination. 3,3. Changes in woody + herbaceous ground flora La 1952, the three most abundant herbaceous taxa Vfexo­.Viola eucuUata, Desmodium nudiflorum, and Gattum circaezans in the cove­hardwoods. After cut­ ting hi 1977, Parthenocissus qidnquefoUa, V, cucul­ lata, and species within the Asteraceae family were the most abundant In the mixed­oak hardwoods, Vaccinium spp. (vacillans and stamineum), Smilax glauca, and Epigaea repens were the most abundant species hi 1952, One year after clearcutting, *Sol­ idago spp., Eupatorium rugosum, V. cucullata, and Aster spp. were the most dominant In the hard­ wood­pines, Vaccinium spp., Galax aphyUa, and S. glauca made up 64% of the density hi 1952. In 1977, Solidago spp., P. quinquefolia, S. rotundifolia, and E. rugosum were the most abundant species and accounted for 74% of the total biomass. G. aphyUa began to recover by 1993 hi the hardwood­pines community (Table 3). la 1979 and 1984, Rubus spp. was the most abundant species hi all three communi­ ties. However, by 1993, it had declined to less than 1.0% of the total ground flora biomass in the cove­ hardwoods, 0% hi the mixed­oak hardwoods, and 5% in the hardwood­pines, and P. quinquefolia dom­ inated in the mixed­oak hardwoods. Changes in ground flora through succession were attributed to species mat established or disappeared 'after disturbance or species that were short­lived or transitory. Species established after clearcutting in­ cluded Pofystichum • acrostichoides in the cove­ 78 KJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­85 hardwoods, and E. rugosum and P. quiTiquefolia in flie mixed­oak hardwoods. Short­lived or transitory species included Erechtites hieracifolia, ' Acalypha rhomboidea, E. rugosum, Tiarella cordifolia, and Rubus spp. in the cove­hardwoods; Botrychium vir­ giniand, G. latifolia, Rubus spp., Clematis virgini­ ana, Monarda clinopodia, Cimidfiiga racemosa, and P. acrostichoides in the mixed­oak; P. quinquefolia, Helianthus microcephalus, and Coreopsis major in the hardwood­pines (Table 3). Less common species in 1952 not found after clearcutting included Agrimonia parviflora, Aris­ tolochia macrophylla, Tripkora trianthophora, and Veratrum parviflorum in the cove­hardwoods; An­ gelica venenosa, Clintonia^umbellata, Erigeron pul­ chellus, Lilium michauxii, Linum striatum, Lyonia ligustrina, Persicaria hydropiperoides, Phryma lep­ tostachy, Pilea pumila, T. cordifolia, and T. tri­ anthophora in the mixed­oak hardwoods; Campan­ ula divaricata, Erythrordum americanum, Habenaria ciliaris, Menziesia. pilosa, Polygonatum biflonan, Smilacina racemosa, and Trillium spp. in the hard­ wood­pines. . Ground flora biomass peaked hi 1979 hi each community type then declined substantially by 1993. Currently, ground flora biomass in the cove­ hardwoods, mixed­oak hardwoods, and hardwood­ pines is only 8.0%, 2.0%, and 8.0%, respectively, of the peak biomass in 1979 (Table 4). In contrast to the pattern for woody species, many more ground flora species were present in all com­ munities hi 1952 man hi the years after clearcutling (Table 4). In addition to the species level changes within communities, family distributions (including woody species) have also changed since clearcuttmg (Tables 2 and 4). In the cove­hardwoods, the number of families present increased from 24 hi 1952 to 29 hi 1977 and 1979, and to more than 30 hi following years. However, hi the mixed­oak hardwoods, there were many more families represented before clearcutting; 39 families were present hi 1952, re­ duced to only 21 families hi 1977, then increased to 22 by 1993. In the hardwood­pines, 31 families were present hi 1952 compared with only 22 families immediately after cutting, which then increased to 34 families by 1984 (Tables 2 and 4). Although most families were represented by only one or two genera, families that were well represented by several genera were the Asteraceae, Ericaceae, Rosaceae, Fabaceae, and Liliaceae. For example, hi the hardwood­pines, 11 genera were found within the Liliaceae family hi 1952. After clearcutting in 1977, only Smilax re­ mained, in 1979 IJlium was found, and hi 1984 Medeola and Uvularia were sampled. Families mat Table 4 Average abundance (number of plants m~2 in 1952; g mass m~2 in all other years), diversity (H', Shannon­Weaver's index), and evenness (/', Pielon's index) of ground flora species for three community; types in WS7, Coweeta Basin Community Year F G S Abundance (m~2) H' Variance of H' Q of H' f 1952' 17 27 16.5 12 152 0.054 ±0.092 0.765 Cove­hardwoods 1977 12 17 19 33 J ­ 0.846 ~2.49 0.014 ±0.057 1979 12 22 97.8 0.689 20 2.19 0.010 ±0.043 1984 16 19 21 ±0.092 0.608 37.6 US 0.040 14 8.0 0274 1993 19 20 ±0.174 . 0.82 0.13T ! 39 11.4 0.807 Mixed­oak hardwoods­^ 1952' 23 49 3.14 ±0.055 0.038 1977 7 9 . 10 20.3 0.673 US 0.057 ±0.168 12 17 18 84.9 Z04 0.706 1979 ±0.056 0.013 1984 0.674 8 12 20.8 13 1.73 ±0.116 0.040 1993 12 16 2.1 132 0.476 16 ±0.047 0.008 Hardwood­pines 1952' 18 42 45 13.2 0.630 X40 0.107 ±0.099 1977 10 15 16 0.718 43.0 1.99 ±0.089 0.028 12 22 25 46.9 0.708 1979 228 0.018 ±0.056 24 0.519 1984 16 21 17.5 ±0.125 1.65 0.085 25 27 1993 16 3.7 1.90 ±0.152 0.576 0.141 F, Total number of families present in each community; G, total number of genera present in each community; S, total number of species present in each community. KJ. Elliott et aL /Forest Ecology and Management 92 (1997) 67­55 Table 5 . T­Stadstics for ground flora species diversity .CEP), for pairwise comparisons among years within each community in WS7, Coweeta Basin Comparison t value df P value Community 1.905 84 0.10 Cove­hardwoods 1977 vs. 1979 1977 vs. 1984 2.744 61 0.01 1977 vs. 1993 4.293 10 0.002 1.525 58 us 1979 vs. 1984 1979 vs. 1993 3.575 9 0.01 1984 vs. 1993 2.448 13 0.05 Mixed­oak hardwoods 1977 vs. 1979 Z872 30 0.01 1977 vs. 1984 ­0.578 40 ns 1977 vs. 1993 0.902 22 us 1337 35 ns 1979 vs. 1984 1979 vs. 1993 5.006 14 0.0001 1.860 22 0.10 1984, vs. 1993 1977 vs. 1979 ­L352 84 ns Hardwood­pines 197 • 7 vs. 1984 1.011 30 ns ' 1977 vs. 1993 0.219 5 ns 1.934 25 0.10 1979 vs. 1984 1979 vs. 1993 0.933 5 ns 1984 vs. 1993 0.527 9 05 r­Statistics and calculations for requisite degrees of freedom follow Magurran (1988); ns, not significant. were shared by woody and ground flora species were the Ericaceae, Fabaceae, and Rosaceae. In the cove­hardwoods community, E' was sig­ nificantly higher in 1977 than in all subsequent years. The difference in H' between 1979 and 1984 was not significant Diversity declined significantly in 1993 (Tables 4 and 5). Two species (5. rotundifb­ lia and P. acrostichoides) representing 93% of the total biomass accounted for the low /' in 1993 (Table 4). In the mixed­oak hardwoods community, H' increased significantly from 1977 to 1979, began to decline in 1984, and was significantly lower by 1993. H' in 1993 was significantly lower than in 1979 or 1984^CTables 4 and 5). Differences in H' were not significant between 1977 and 1984 or 1993. In the hardwood­pines, me difference in H' was significant between 1979 and 1984, but differences were not significant between other years (Table 5). Although no statistical tests were performed between 1952 and post­clearcut years because abundance measures differed, H' based on density was higher in 1952 in the mixed­oak hardwoods and hardwood­ pines communities than H' based on biomass in Hie years after clearcutting. 79 4. Discussion 4.1. Woody species' responses ' The diversity of woody species was relatively stable in WS7; however, tree .species richness in­ creased through succession. This trend in diversity, similar to that found in other eastern hardwood forests (Reiners, 1992; Wang and Nyland, 1993), was also found after clearcutting hi a nearby water­ shed within the Coweeta Basin (Elliott and Swank, 1994a). These succession^ changes are somewhat different from those found in northeastern deciduous forests. J?or example, Gove et al. (1992) showed a decline in tree diversity lOyears after clearc'utting whereas Reiners (1992) found a gradual decline in diversity and an increase in­richness after clearcut­ ting and herbiciding. Two years­ after clearcutting, Reiners' data (Reiners, 1992) suggested a trend in secondary succession with a mixed component of 'relay floristics' and 'initial composition'. Although most species in his undistributed reference forest eventually regenerated in the clearcutting site, most woody biomass in the latter was produced by two species uncommon in the former forest (P. pensyl­ vanica and B. papyrifera). Phillips and Shure (1990) found mat species composition changed after clearcutting small (2.0 ha size patch) mesic, mixed­ hardwood sites in the Southern Appalachians. In their study, L. tulipifera remained dominant 2 years after cutting whereas Q. rubra and Carya spp. de­ clined in relative biomass, and R. pseudoacacia, C. florida, and A. rubntm increased. Beck and Hooper (1986) found mat clearcutting a mixed­hardwood forest dominated mostly by oak resulted in a 20­ year­old stand dominated by L tulipifera, R. pseu­ doacacia, and A, rubrum. In our study, C. florida and R. pseudoacacia also increased in relative domi­ nance. However, 17 years after cutting (1993), C. florida began to decline in dominance. The substan­ tial ­decline in C florida from 1984 to 1993 was probably attributed to disease. Dogwood • anthrac­ nose, caused by Discula destructiva Redlin., had an average incidence of infection of 87% in C, florida for 1990 in the Coweeta Basin (Chellemi et al., 1992).'In contrast, & pseudoacacia continued to increase. Q. nibra also decreased in our • cove­ hardwood plots and A. rubnun, important 2 years 80 KJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­85 after cutting, had returned to pre­cut levels in the community. Hardwood forests in the Southern Appalachians revegetate quickly after disturbance because many species reproduce and grow rapidly. Although reveg­ etation was relatively rapid in WS7, the composition of the forest changed. For example, Carya spp., the leading dominant in. the cove­hardwoods community before harvest, currently makes up less man 1% of the total basal area in these communities. Species such as Carya spp. will probably not become a significant component of the stand for many decades because they disperse seed and grow slowly: Mean­ while, opportunistic species such as L. tulipifera, R. pseudoacada, and A. rubrum have increased. Be­ cause JL tulipifera and RL pseudoacada sprout quickly and grow faster than other species, .they attain early dominance. Acer rubrum, although a shade­tolerant species, produced 5000­9000 seedlings ha"1, the first year after cutting in the cove­hardwoods and mixed­oak hardwoods, and over 5000 seedlingsha" • in the­ hardwood­pines 2years after cutting. Acer rubrum was also one of the most prolific sprouting species, with 1800­6300, sprouts ha"1, depending on community type. Its abil­ ity to establish by both sexual and asexual reproduc­ tion may explain its successful regeneration follow­ ing disturbance. : . ­ : • • • . • • .. ­ Sprouts play a major role in the' revegetation process of these hardwood forests. The revegetation process on WS7 was similar to that in other eastern hardwood forests, where sprouts and suckers domi­ nate vegetation after clearcutting (Ross et aL, 1986; Phillips and Shure, 1990; White, 1991; Crow et aL, 1991; Brown, 1994). In the first year after clearcut­ ting, seedling and sprout reproduction was about equal, except in the hardwood­pines where sprout reproduction was higher. By 1979, the proportion of stems originating from: sprouts increased in all com­ munities. In the hardwood­pines, the high percentage of stems originating from sprouts (81%) probably occurred because seed propagules were scarce:and the xeric forest ­floor microclimate along the south­ west­facing slopes1 and ridges (Swank and Vose,* 1988) produced a high mortality rate of seedlings. In Southern Appalachian forests, mode of repro­ duction alone does not guarantee success., Comparing .two species that reproduced primarily' by seed, L, tulipifera and Q. velutina, in the cove­hardwoods and mixed­oak hardwoods communities provides a striking contrast Q. velutina, a leading dominant in the mixed­oak hardwoods before clearcutting, repro­ duced only from seed germination or advance seedling growth. Although stumps of Q. velutina sprout less frequently than Q. rubra, Q. prinus, and Q. cocdnea, the majority of the reproduction after harvest is usually from stump sprouting (Burns and Honkala, 1990). Because seedlings established after harvest grow too slowly to complete with sprouts of other tree species and other vegetation, they usually die after a few years (Bums and Honkala, 1990). In our study, the low basal area for this species after disturbance may be the result of a combination of factors, including low dispersal of seed hi the large" opening, low survival of seedlings, slow growth of seedlings, and lack of sprouting. Before cutting, 30% of the Q. velutina\stsias were of greater man 23 cm dbh, which probably limited sprouting; likewise, the high percentage (53%) of Q. velutina stems of less than 5 cm dbh also limited sprouting. Stump sprout­ ing from large stumps of old trees is less man from small stumps of young trees (Kays et al., 1988; Kays andCanham, 1992), : In contrast, L. tulipifera established successfully in both the cove­hardwoods and mixed­oak hard­ woods communities after clearcutting. In 1993, it was the leading dominant species, occupying 22% and 44% of the total basal area in the cove­hardwoods and mixed­oak hardwoods types, respectively. A combination of factors, including prolific seed pro­ duction, extended seed viability in the forest floor, survival of new germinants, relatively fast growth, and some stump sprouting, are responsible for mis success. L. tulipifera was a copious seeder, with 8000­10000 seedlingsha­1 produced during the first year after cutting,'whereas Q. velutina seedlings totaled 300­700 seedlingsha"1, with many present before, harvesting. Early and copious production of light, wind­dis­ persed seeds is generally correlated with the ability to respond to large disturbances (Canham and Marks, 1985). Small­seeded and less shade­tolerant species such as L. tulipifera and B. lento, exhibit minimal delay between dispersal and germination and often release seeds from autumn until spring (Canham and Marks, 1985). However, B. fento seedling produc­ SJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­55 tion was low in the cove­hardwoods until 2 years after clearcutting, when the species produced 2774 seedlings ha"1. The stem exclusion stage of stand development (Oliver and Larson, 1990) was most dramatic during the 9 year period between 1984 and 1993. Although WS7 is a young forest with most stems in small size classes, me stem exclusion stage has begun, as indi­ cated by the decrease in stem density by 81%, 77%, and 64% for the cove­hardwoods, mixed­oak hard­ woods, and hardwood­pines, respectively. Density of stems of 5.0cm or more is less than in a nearby 30­year­old clearcut watershed, whereas basal area of stems with dbh of 5.0cm or more is much higher man values reported for this same watershed (Elliott and Swank, 1994a). 4.2. Ground flora responses In general, ground flora diversity declined from 1977 to 1993 in me cove­hardwoods and mixed­oak hardwoods communities, but did not decrease signif­ icantly in hardwood­pines. In every community, more species were present in 1952 than in the years after clearcutting. This pattern parallels results reported by Gove et al. (1992), where diversity of all plant species (overstory and ground flora combined) de­ creased 10 years after clearcutting in New Hamp­ shire. Nixon and Brooks (1991) found mat herba­ ceous species diversity peaked in Year 3 after clearcutting a deciduous forest in east Texas men subsided to Year 9. The abundance Cue. g biomassm"2) of ground flora was also lower in 1993 compared with 1984: . 79% less in the cove­hardwoods, 90% less in the mixed­oak hardwoods, and 79% less in the hard­ wood­pines. With, growth of overstory trees and canopy closure^ the number of early successional, shade­intolerant species, such as Erechtites, Sol­ idago, Eupatorium, Panicum, and Aster, has de­ clined. Late successional, shade­tolerant species, such as Viola, Galium, Sanguinaria, Uvularia, and Vera­ tnan, are not well established in the watershed, even though they are common hi other areas within­the Coweeta Basin. Watershed 7 is apparently in a tran­ sition state between early and late successional species abundance. The timing of measurements pre­ 81 vented examining the response of spring ephemerals, such as Trillium, Anemone, and Claytonia, after clearcutting. Because spring ephemerals respond to changes in temperature and light (Coffins et al., 1985), clearcutting may have triggered changes in seasonal phenology, growth, and reproductive poten­ tial of these species. • Total numbers of species in each community were lower after clearcutting in 1977 man in 1952. How­ ever, to quantitatively compare species richness and diversity in 1952 with years after harvest is difficult because data on ground flora immediately before 1977 are lacking, and plot sizes and locations differ. The 25 years of succession between 1952 and 1977, and cumulative effects of land­use history G.e. graz­ ing and fire suppression), prevent interpretation of the effects of clearcutting alone. Ground flora species diversity and richness in WS7 were lower in the cove­hardwoods and mixed­ oak hardwoods and higher hi the hardwood­pines when compared with a nearby 30­year­old clearcut watershed (WS13) with the same community types (KJ. Elliott, personal observation, 1991). The lower ground flora H' in two of the community types in WS7 may be the result of several factors, including (1) me larger spatial scale of disturbance in WS7 (57ha cut in WS7 vs. 16ha cut in WS13>, (2) southwest­facing aspect of WS7, which receives higher solar radiation man the east­facing aspect of WS13; (3) total tree removal in WS7 whereas trees were cut and left in place in WS13. 4.3. Influences of complex disturbances Large forest openings significantly change the forest floor microclimate for all residual biota, in­ cluding woody seedlings and late successional herba­ ceous species (Phillips and Shure, 1990). Other in­ vestigators at Coweeta have found mat clearcutting on WS7 increased mean monthly temperatures at the litter­soil boundary for the period May­October by 8­ll0C the first year after cutting, reduced forest floor litter moisture, increased soil moisture (Swank and Vose, 1988), altered microarthropod activity in the litter (Seastedt and Crossley, 1981; Seastedt et aL, 1983), and reduced first­year decomposition of woody litter, especially on xeric south­facing slopes 82 KJ. Elliott etaL/Forest Ecology and Management 92 (1997) 67­85 (Abbott and Crossley, 1982). The increase in woody leaf area index by the third year after clearcutting resulted in forest floor shading, amelioration of the altered forest floor­microclimate, and dampening of environmental effects of forest floor biota and their processes. Although seedling and ground flora may have been affected by high mortality immediately after clearcutting, canopy closure within 3 years al­ lowed a subsequent rapid recovery of structural and functional forest processes. . Both anthropogenic (e.g. chestnut blight, fire ex­ clusion, and cattle grazing) and natural disturbances (e.g.. drought) shaped forest composition in WS7 before clearcutting. The composition of Southern Appalachian forests has been significantly altered by the loss of American chestnut (C. dentata) (Woods and Shanks, 1959; Arends, 1981; Day et al., 1988; Busing, 1989). Chestnut blight had a major impact in the Coweeta Basin, because chestnut made up an estimated 35­40% of me basal area of some forest stands (Day et al., 1988). Fire exclusioa in the Southern Appalachians has favored the expansion of evergreen shrubs (Day and Monk, 1974; Monk et aL, 1985; Lipscomb and Nilsen, 1990) and has reduced regeneration success of many Quercus species (Phil­ lips and Murdy,. 1985; Van Lear, 1991). Rhododen­ dron often dominates, understory canopy layers in riparian stands, and adversely affects development . and richness.of herbaceous and understory strata (Baker, 1994; Hedman and Van Lear,­1995). Heavy cattle grazing can also have a dramatic effect on species richness and diversity. For example, Williams (1954) found a loss of 31 species in me cove­ hardwoods community of WS7 during a 12year period (1940­1952) of heavy grazing; however, the mixed­oak and hardwood­pines types showed little to no loss of species on slopes and ridges, where . cattle were less likely to travel..In. addition, severe droughts have caused substantial tree mortality in the Southern USA (Hursh and Haasis, 1931; Tainter et al., 1984; Stringer et aL, 1989; Starkey et al., 1989; Smith, 1991; Clinton et al., 1993; Elliott and Swank, 1994b). The combined impacts of these sequential and simultaneous disturbances on plant diversity be­ fore clearcutting in 1977 would be impossible to sort­ out, yet their cumulative effects are probably no­ table. . 5. Conclusion The response of plant communities to clearcutting varied in a Southern Appalachian watershed. Woody species richness increased in the cove­hardwoods and hardwood­pines immediately after clearcutting and through 17 years of succession but remained relatively constant in the mixed­oak hardwoods com­ munity. Woody species diversity decreased in the mixed­oak hardwoods but remained relatively con­ stant in the cove­hardwood and hardwood­pines communities. L tulipifera increased in dominance in all three communities. In addition, R. maximum in­ creased in the cove­hardwoods, R, pseudoacada increased in the mixed­oak hardwoods, and K. latifo­ Ua and A. rubrum increased in the hardwood­pines. Carya spp. declined in dominance after clearcutting in the cove­hardwoods, Q. velutina and Carya spp. declined in the mixed­oak hardwoods, and Q. coc­ cinea and Q. velutina declined in the hardwood­ pines. Ground flora was in a transitional state between early and late successional species, 17 years after clearcutting. Early successional Aster, Solidago, and Eupatorium species have declined in abundance be­ cause woody species have grown rapidly and the canopy has closed. Late successional species have not become abundantly established, which has caused a significant decline in ground flora diversity in the cove­hardwoods and mixed­oak hardwoods. Total number of plant species present (woody + ground flora) increased hi all three communities during the first 3 years after cutting. Total species remained relatively constant in the cove­hardwoods and mixed­oak hardwoods from 1979 to 1993; however, total species continued to increase to 1993 in the hardwood­pines. •' , ­ . . . Qearcutting favors shade­intolerant pioneering species, such as L. tulipifera and R. pseudoacada, and shade­tolerant understory species such as R. maximum and K. latifolia. The positive responses to clearcutting by these two markedly different groups of plants strongly radicates mat retention of species of Quercus and other hard­mast producing species that have critical ecosystem functions will require additional management measures. In addition to the altered microclimatic influences KJ. Elliott etaL/ Forest Ecology and Management 92 (1997) 67­85 of clearcutting, past disturbances such as selective logging, chestnut blight, fire suppression, and wood­ land grazing have also shaped the current conditions in WS7. Although separating the cumulative effects on vegetation dynamics is difficult, this complex of disturbances is typical of conditions throughout much of the Southern Appalachians. The cumulative vege­ tation responses to clearcutting and other distur­ bances found here are indicative of regional re­ sponses of forests since the early twentieth century. Other influences of regional atmospheric pollution and climate change may also have an undefined influence on species richness and community com­ position. Acknowledgements We thank Alan Livingston, Patsy Clinton, Deidre Hewitt, Martin Nelson, Linda Chafin, Jim Graves, and numerous others for their help in field data collection. Dan Pittillo and Lee Reynolds assisted with plant identification. Drs. David H. van Lear, L. Katherine Krrkman and two anonymous reviewers provided helpful comments on mis manuscript References Abbott, D.T. and Crossley, Jr., DA, 1982. Woody litter decom­ position following cleatciilihig. Ecology, 63: 35­4Z Arends, E, 1981. Vegetation patterns a naif century following the chestnut blight in the Great Smoky Mountains National Park. M5. Thesis, University of Tennessee, Knoxville. Baker, T.T, 1994. The influence of Rhododendron maximum on species richness hi the riparian zone of Wine Spring Creek. M.S. Thesis, Clemson University, Clemson, SC Beck, D.E. and Hooper, RJvL, 1986. Development of a Southern Appalachian hardwood stand after clearcutting. S. J. AppL For., 10: 168­m;. BickneH, SJUL, 1979. Pattern and process of plant succession in a revegetating northern hardwood ecosystem. PhD. Disserta­ tion, Yale University, New Haven, CT. Boring, LJt, 1979. Early forest regeneration and nutrient conser­ vation on a clearest Southern Appalachian watershed. M.S. Thesis, University of Georgia, Athens. Boring, LR. and Swank, W.T, 1986. Hardwood biomass and net primary production following clearcmting • in the Ccweeta Basin. In: R.T. Brooks, Jr. (Editor), Proc. 1986 Southern Forest Biomass Workshop, 16­19 June 1986, Knoxville, TN. Tennessee Valley Authority, Morris, pp. 43­50: Boring, LJR, Monk, CD. and Swank, W.T., 1981. Early regener­ 83 ation of a clear­cut Southern Appalachian forest Ecology, 62: 1244­1253. Boring, L.R., Swank, W.T. and Monk, CD.. 1988. Dynamics of early successional forest structure and processes in the Coweeta Basin. In: W.T. Swank and DA Crossley, Jr. (Editors), Forest Hydrology and Ecology at Coweeta, Ecological Studies 66. Springer, New York, pp. 161­180. Bormann, FJL, IJkens, GJL, Siccama, T.G, Fierce, R.S. and Eaton, J.S, 1974. The export of nutrients and recovery of stable conditions following deforestation at Hubbard Brook. EcoL Monogr., 44: 255­277. Bormann, FJEL, Likens, G.E. and Melillo, JM, 1977. Nitrogen budget for an aggrading northern hardwood forest ecosystem. Science, 196: 981­983. Brown, D., 1994. The development of woody vegetation in the first 6 years following clearcutting of a hardwood forest for a utility right­of­way. For. Ecol. Manage., 65: 171­181. Burns, RM. and Honkala, BJL, 1990. SQvics of North America, VoL 2, Hardwoods. USDA For. Serv. Handb. 654, Washing­ ton, DC 877 pp. Burton, PJ., BaUsky, AC, Coward, LJP., Camming* S.G. and Kneeshaw, D£>., 1992. The value of managing for biodiver­ sity. For. Chron., 68: 225­237. • Busing, R.T., 1989. A half century of change in a Great Smoky Mountains cove forest. Bull. Torrey Bot Club, 116:283­288. Omham, CD. and Marks, Pi., 1985. The response of woody plants to disturbance: patterns of establishment and growth. In: S.TA Pickett and P.S. White (Editors), The Ecology of Natural Disturbance and Patch Dynamics. Academic Press, New York, pp. 197­216. Cheltemi, D.O., Britton, K.O. and Swank, W.T, 1992. Influence of site factors on Dogwood Anthracnose in the Nantahala Mountain Range of western North Carolina. Plant Dis, 76: 915­918. Clinton, BJX, Boring, LR. and Swank, W.T., 1993. Characteris­ tics of drought­induced canopy gaps in oak forests of the Coweeta Basin. Ecology, 74:1551­1558. • Collins, B.S, Dunne, ELP. and Pickett, S.TA, 1985. Response of forest herbs to canopy gaps. In: S.TA Pickett and P.S. White (Editors), The Ecology' of Natural Disturbance and Patch Dynamics. Academic Press, New York, pp. 217­234. Crow, T.R, Mroz, GJX and Gale, MJR., 1991. Regrowth and nutrient accumulations following whole­tree harvesting of a maple­oak forest Can. J. For. Res., 21:1305­1315. Day, FJ. and Monk, CD, 1974. Vegetation patterns'on a South­ ern Appalachian watershed. Ecology, 55: 1064­1074. Day, FJ>, Phillips, D.L. and Monk, CIX, 1988. Forest communi­ ties and patterns. In: W.T. Swank and DA Crossley, Jr. (Editors), Forest­Hydrology and Ecology at Coweeta, Ecologi­ cal Studies 66. Springer, New York, pp. 141­150. Douglass, J.E. and Hoover, MD, 1988. History of Coweeta. In: W.T. Swank and DA Crossley, Jr. (Editors), Forest Hydrol­ ogy and Ecology at Coweeta, Ecological Studies 66. Springer, New York, pp. 17­34. .' Elliott, KJ. and Swank, W.T, 1994a. Changes in tree species diversity after successive clearcuts in the Southern Appalachi­ ans. Vegetatio, 115: 11­18. \ 84 KJ. Elliott et aL/Forest Ecology and Management 92 (1997) 67­85 Elliott, KJ. and Swank, W.T., 199*. Impacts of drought on tree Factors affecting natural regeneration of Piedmont hardwoods. mortality and growth in a mixed hardwood forest J. Veg. ScL, S. J. AppL For., 12: 98­102. .5:229­236. ' . • . .­ Dkensi G^, Bormann, FJL, Pierce," R.S., Eaton, J.S. and John­ Fain, JJ., VoBc, TJV. and Fahey, T­L, 1994. Fifty years of change' son, NJM, 1977. Biogeochemistry of a Forested' Ecosystem in an upland forest in south­central New York general pat­ '• ' • Springer, New York, 146 pp. terns. BulLTorreyBot.Clnb, 121: 130­139. ,. : Lipscomb, M.V. and Nilsen, E.T., 1990. Environmental and Finegan, B., 1984. Forest succession. Nature, 312:109­1­14 physiological factors influencing the natural distribution, of Gauch, Jr.. H.G., 198Z MuMvariate Analysis in Community evergreen and deciduous ericaceous shrubs on northeast and . Ecology. Cambridge University Press, Cambridge, 298 pp. southwest slopes of the S. Appalachian Mountains. L Irradi­ : Gholz, H.I.., Hawk, GJ&, Campbell, A. and Cromack, K, 1985. ancetolerance.Am. J. Bot, 77:108­115. Magunaa, AJL, 1988. Ecological Diversity and its Measure. Early vegetation recovery and element cycles on a clear­cot Princeton University Press, Princeton, NJ, 179 pp. watershed in western Oregon. Can. J. For. Res., 18: 1427­ 1436. , . ­ . . . ­ " ­ . ­ . . . McMhm, J.W., 1991. Biological diversiry research: an analysis. US For. Serv. Soumeast For. Ecp. Stn. Gen. Tech. Rep. Gove, J.H., Martin, CW., Patfl, GJP., Solomon, D.S. and Horn­ beck, J.W., 1992. Plant species diversity on even­aged bar­ SE­71.' . . ' . " ' ' , ' vests at the Hubbari Brook Experimental Forest: 10­year Monk, CD., McGinty, D.T. and Day, Ff^ 1985. the ecological importance of Salmia latifolia and Rhododendron maximum resnlts. Can. J. For. Res^ 22:1800­1806. .. in the deciduous forest of the Southern Appalachians. BulL Harmed, D.C and Krofta, DJ*L, 1989. Fifty­five years.of post­fire succession in a southern mixed hardwood forest. BuIL Torrey Torrey Bot dub, 112:193­197. Nixon, E^. and Brooks, AJL, 1991. Species diversity following Bot Qub, 116:107­113. clearcntdng in eastern Texas. Tex. J. ScL, 43: 399­403. Hatcher, RJX 1974. An Introduction to the Blue Ridge Tectonic Norse, EA­, Rosenbaum, KJ_ Wflcove, D.S, "Wilcox, BA, History of Norm­east Georgia, Guidebook 13­A. Georgia Geo­ logical Survey, Georgia Department of Natural Resources, ' Ropme, Wi, Johnston, D.W. and Stout, 'MJ­, 1986. Con­ serving Biological Diversity in Our National Forests. 'Wilder­ Atlanta* • ' . . ­ ' ; . • ­• ' ness Society, Washington, DC • : : Hedman, CW. and van Lear, DJBL, 1995. Vegetative structnre Oliver, CD. and Larson, B.C,. 1990. Forest Stand Dynamics. and composition of Sonthem Appalachian riparian forests. McGraw­Hill, New York, 467 pp. '' , . BnlL Torrey Bot dub, 122:134­144. Parker, G.R. and Swank, W.T., 1982. Tree species response to Hewlett, JJX.. 1979. Forest water quality, .an experiment in har­ clear­cutting a Sonthem Appalachian watershed. Am. MML vesting and regenerating Piedmont forests. Ga. For. Res. Pap. : Hrbbs, DJL, 1983. Forty years of forest succession hi central New Nat, 108:304­310. ' ' ' ' Peet, RX. and Christensen, NI», 1980. Succession: a population England.Ecology; 64:1394­1401. . . ­. ., :.. process. Vegetatio, 43:131­140. Hornbeck, J.W., Martin, CW­ Pierce, R3.,. Bormann, F.R, Peet, RJL and Loncks, OJ­, 1977. A gradient analysis of south­ Likens,. G.E. and Eaton, J.S., 1987. The northern hardwood ern Wisconsin forests. Ecology, 58:486­499. forest ecosystem: ten years of recovery from clear­cutting. US Petraitis, P.S^ T^tham, RE. and Niesenbaum, RA, 1989. The For. Serv. Northeast For. Exp. Stn. Res. Pap. NE­RP­596. maintenance of species diversity by disturbance, Q. Rev. BioL, Hunter, Jr., MX., 1990. Wildlife, Forests, and Forestry: Principles (54:393­418. ; . . of Managing Forests for Biological Diversity. Prentice­Hall, PhQlips, DJL and Murdy, WJL, 1985. Effects of rhododendron Englewood difis. NJ, 370 pp. . . . .­...­' (.Rhododendron maximum L.) on regeneration of Sonthem Hnrsh, CR. and Haasis, F.W., 1931. Effects of 1925 summer Appalachian hardwoods. For. ScL, 31:226­233. ' : . drought on Southern Appalachian hardwoods. Ecology, 12: Phfflips, Dl. and Shore, DJ., 1990. Patch­size effects on early 380­386. . . . . succession in Southern Appalachian forests. Ecology, 71: Huston, MA­, 1979. A general hypothesis of species diversity. 204­212. . ' Am. Nat. 113: 80­101. Pielou, E.C, 1966. The measurement of diversity in different Huston, MA, 1994. Biological Diversity. The Coexistence of types of biological coHecdons. J. Theor. BioL, 13:131­144. Species on Changing Landscapes. Cambridge University Press, Radford, Ai, Abies, H£. and Befl, CJL, 1968. Manual of me Cambridge,"681 ppr;­>. . ­. . . . . . . . , Vascular Flora of the Carolinas. University of Norm Carolina Huston, MA and Smith, TAL, 1987. Plant succession: life Press, Chapel Hill, 1183 pp. history and competition. Am. Nat. 130:168­198. Reiners, WA, 1992. Twenty years of ecosystem reorganization Johnson, EA, 1952. ESect of farm woodland grazing on. water­ . following experimental deforestation and regrowth suppres­ shed values in the Southeast. J. For., 50:109­113.., . . sion. EcoL Monogr^ 62:503­523, : Johnson, Pi. and Swank, W.T, 1973. Studies of cation budgets Roberts, MJL, 1992. Stand development and overstory­under­ in the Appalachians of four experimental watersheds with story interactions in an aspen—northern hardwoods stand, For. contrasting vegetation. Ecology, 54:70­80.,. ... . , EcoL Manage, 54:157­174. '­ Kays, J.S. and Canham, CD^ 1992. Effects of time and frequency Roberts, MJR. and Christensen, NX, 1988. Vegetation variation of catting on hardwood root reserves and sprout growth. For. among mesic successional forest stands m northern­ lower ScL, 37:524­539. ' ; ..­• . :­ Michigan. Can. J. Bot, 66:1080­1090, Kays, J.S., Smith, D.W., Zedaker, SM. and Kreh, RJB^.1988. KJ. EKottetaL/Forest Ecology and Management 92 (1997) (57­55 85 Ross, M.S., Sharik, Ti. and Smith, D.W., 1986. Oak regeneration Swank, W.T., Swift, Jr., L.W. and Douglass, J.E., 1988. Stream­ after clear felling in southwest Virginia. For. Sci, 32:157­169. flow changes associated with forest cutting, species conver­ Rankle, JJL, 1985. Disturbance regimes in temperate forests. Ire sions, and natural disturbances. In: W.T. Swank and D.A. S.T.A. Pickett and P.S. White (Editors), The Ecology of Crossley, Jr. (Editors), Forest Hydrology and Ecology at Natural Distnrbance and Patch Dynamics. Academic Press, Coweeta, Ecological Studies 66. Springer, New York, pp. 297­312. New York, pp. 17­34. Schoonmaker, P. and McKee, A, 1988. Species composition and Swift, Jr., L.W., Cunningham, G.B. and Douglass, J^, 1988. diversity during secondary succession of coniferous forests in Climatology and hydrology. Ik W.T. Swank and DA. Cross­ the Western Cascade Mountains of Oregon. For. Sci, 34: ley, Jr. (Editors), Forest Hydrology and Ecology at Coweeta, Ecological Studies 66. Springer, New York, pp. 35­56. 960­979. Seastedt, T.R. and Crossley, Jr., DJV, 1981. Microarthropod Tainter, FJL, Fraedrich, S.W. and Benson, DAi, 1984. The effect response following cable logging and clearcutting in the of climate on growth, decline, and death of northern red oaks in the western North Carolina Nantahala mountains. Castanea, Sonthem Appalachians. Ecology, 62: 126­135. 49:127­137. Seastedt, TIL, Crossley, Jr., D­A, Meentemeyer, V. and Waide, J.B., 1983. A two­year study of leaf litter decomposition as Thomas, D­L, 1996. Soil Survey of Macon County, North Car­ related to macrociimatic factors and nricroarthropod abun­ olina. USDA Natural Resources Conservation Service, Wash­ . dance in the Southern Appalachians. Holarct EcoL, 6:11—16. ington, DC, 322 pp. Shannon, CE. and Weaver, W,, 1949. The Mathematical Theory Van Lear, DJL, 1991. Fire and oak regeneration in the Sonthem of Communication. University of Illinois Press, Urbana. Appalachians. US For. Serv. Southeast For. Exp. Sta. Gen. Smith, RJ*., 1991. Species composition, stand structure, and Tech. Rep, SE­69:15­21. woody detrital dynamics associated with pine mortality in the Van Lear, D.H^ Douglass, JJE^ Cox, S.K. and Augspurger, MX, Sonthem Appalachians. M.S. Thesis, University of Georgia, 1985. Sediment and nutrient export in runoff from burned and harvested pine watersheds in the South Carolina piedmont J. Athens. Environ. QuaL, 14:169­174. Starkey, DJ^, Oak, S.W., Ryan, G.W., Tainter, F.R. Redmond, C and Brown, HJX 1989. Evaluation of oak decline area in Waide, J.B., Caskey, W.H., Todd, RX. and Boring, LR^ 1988. tiie South. Forest Protection Rep. R8­TR 17, USDA, For. Changes in soil nitrogen pools and transformations following forest clearcutting. Im W.T. Swank and D.A. Crossley, Jr. Serv., Washington, DC. Stringer, J.W, KJmmerer, T.W., Overstreet, J.C and Dunn, J.P., (Editors), Forest Hydrology and Ecology at Coweeta, Ecologi­ cal Studies 66. Springer, New York, pp. 221­232. • 1989. Oak mortality in eastern Kentucky. S. J. AppL For., 13: Wang, Z, and Nyland, RIX, 1993. Tree species richness increased 86­91. Swank, W.T. and Caskey, WiL, 1982. Nitrate depletion in a by clearcutting of northern hardwoods in central New York. second­order mountain stream. J. Environ. QuaL, 11:581­584. For. EcoL Manage^ 57: 71­84. Swank, W.T. and Helvey, JD., 1970. Reduction of stteamflow . White, A.S., 1991. The importance of different forms of regenera­ increases following regrowth of cleatcut hardwood forests. In: tion to secondary succession in a Maine hardwood forest BuH. D.L. CorreH (Editor), Watershed Research in North America." Torrey Bot dub, 118: 303­311. Williams, J.G., 1954. A study of the effect of grazing upon Chesapeake Bay Center for Environmental Studies, Edgewa­ changes in vegetation on a watershed in the Southern Ap­ ter, MD, pp. 345­364. Swank, W.T. and Vose, JAL, 1988. Effects of cutting practices on palachian mountains. M.S. Thesis, Michigan State College of microenvironment in relation to hardwood regeneration. Iru Agriculture and Applied Science, East Lansing. H.C. Clay, A.W. Perkey and WJE. Edd, Jr. (Editors), Guide­ Woods, F.W. and Shanks, Ri, 1959. Natural replacement of lines for Regenerating Appalachian Hardwood Stands: Proc. chestnut by other speciesrEcology, 40: 349­361. Workshop, Morgantown, WV, 24­26 May 1988. Society of American Foresters Publication 88­03, West Virginia Univer­ sity Books, Morgantown, pp. 71­88. Forest Ecology and Management Submission of manuscripts. . Manuscripts should be submitted in triplicate. Authors from the Americas, Australia, New Zealand and the Pacific are requested to send their manuscripts to Forest Ecology and Management, Dr. Richard F. Fisher, Department of Forest Science, Texas A&M University, College Station, TX77843­2135, USA; all other authors are requested to send their manuscripts to the Editorial Office, Forest Ecology and Management, P.O. Box 181,1000 AD Amsterdam, The Netherlands. Authors in Japan please note: Upon request Elsevier Science Japan win provide authors with a list of people who can check and improve the English of their paper Before submission). Please contact our Tokyo office: Elsevier Science Japan, 1­9­15 HIgashi­Azabu, Minato­ku, Tokyo 106, Japan; tel. (03)­5561­5032; fax (03) 5561­5045. Electronic manuscripts: Electronic manuscripts have the advantage that there is no need for the rekeying of text, thereby avoiding the possibility of introducing errors and resulting in reliable and fast delivery of proofs. For the Initial submission of manuscripts for consideration, hardcopies are sufficient For the processing of accepted papers, electronic versions are preferred. After final acceptance, your disk plus two, final and exactly matching printed versions should be submitted together. Double density (DD) or high density (HD) diskettes (3.5 or 5.25 inch) are acceptable. It is important that the file saved is in the native format of the wordprocessor program used. Label the disk with the name of the computer and wordprocsssing package used, your name, and the name of the file on the disk. Further information may be obtained from the Publisher. Enquiries concerning manuscripts and proofs: questions arising after acceptance of the manuscript especially those relat­ ing to proofs, should be directed to Elsevier Editorial Services, Mayfield House, 256 Banbury Road, Oxford 0X2 TDK, UK, tel. (+44) (0)1865 314900, fax (+44) (0)1865 314990. . Advertising information: Advertising orders and enquiries may be sent to: International: Elsevier Science, Advertising Department The Boulevard, Langford Lane, Wellington, Oxford, OX5 1GB, UK; tel. (+44)(0) 1865 843565; fax (+44)(0) 1865 843976. USA and Canada: Weston Media Associates, Dan Upner, P.O. Box 1110, Greens Farms, CT 06436­1110, USA; tel. (+1) (203) 261 2500; fax (+1) (203) 261 0101. Japan: Elsevier Science Japan, Marketing Services, 1­9­15 Higashi­Azabu, Minato­ku, Tokyo 106, Japan; tel. (+81)­3­5561­5033; fax (+81)­3­5561­5047. US mailing info: Forest Ecology and Management (0378­1127) Is published monthly by Elsevier Science B.V. (Molenwerf 1, Postbus 211,1000 AE Amsterdam). Annual subscription price in the USA is US$ 2253.00 (valid in North, Central and South America), including air speed delivery. Second class postage rate is paid at Jamaica, NY 11431. USA POSTMASTERS: Send address changes to Forest Ecology and Management Publications Expediting, Inc., 200 Meacham Avenue, Elmorrt, NY 11003. AIRFREIGHT AND MAILING in the USA by Publications Expediting Inc., 200 Meacham Avenue, Elmont NY 11003. Forest Ecology and Management has no page charges For a full and complete Guide for Authors, please refer to Forest Ecology and Management, Vol. 90 Nos. 2­3, pp. 89­92. The instructions can also be found on the Word Wide Web: access under http://www.elsevier.nl or http://www.elsevier.com Copyright© 1997, BsevlerScIenc0B.V. All rights reserved. 0378­1127/97/517.00 © The paper used in this publication meets the requirements of ANSI/NISO Z39.48­1992 (Permanence of Paper). Printed in The Netherlands LANDSCAPE AND URBAN PLANNING An International Journal of Landscape Ecology, Landscape Planning and Landscape design Editor­in­Chief: J.E Rodielc, College of Architecture, Texas A&M University, College Station, TX 77843-3137, USA the premise that research AIMS AND SCOPE linked to practice will A journal concerned with ultimately improve the human conceptual; scientific and made landscape. design approaches to land use. By emphasizing ' ecological understanding and Editorial Advisory Board: a multi­disciplinary approach LD. Bishop, Paring Vi^Ajs&affa, . to analysis and planning and EG. Bolen, Wilmington, NC, USA, design, it attempts to draw ID. Brans, Schomdaf, Germany, attention to the interrelated JJ3. Byrom, Bftnbvrgh, UK, nature of problems posed by T.C.Dantel, Tucson, 42; USA nature and human use of RJUDeGraaf, Annast AM, USA, land. In addition, papers S. Gonzalez Kmso, Madrid, Spain, . dealing with ecological U. Hough, Etofafcofesi ON, Canada, processes and interactions P. Jacobs, Montreal, PQ, Canada, within urban areas, and D.S. Jones, Mefcoums, Wo, Agrafe, between these areas and the H. Lavery, Mton, QSt Australia, surrounding natural systems WJI Marsh, Fin; MJ.US4, which support them, will be DJ­Utcnefl, Cafes, 77, US4, D.G. Morrison, Athens, GA, USA, considered. Papers in which 41 Nassauer, St Paul, MN, USA, specific problems are It Nsllschar, Gusfcfc, ON, Canada, examined are welcome. D£. Paterson, Vancouver, BC, Canada, Topics might include but are tLtiamas, Madrid, Spain, not limited to landscape P. Shepard, Oaremont, 04, USA, ecology, landscape planning OJL Skaga, Alnarp, Sweden, and landscape design. HC Smarten, Syracuse, NY, USA, Landscape ecology examines 6. Sorte, Alnarp, Sweden, how heterogeneous f. Steams, Rhxielander, W, USA, combinations of ecosystems are structured, how they function and how they change. Landscape planning examines the various ways humans structure their land use changes. Landscape design involves the physical strategies and forms by which ELSEVIER land use change is actually directed. Landscape and SCIENCE PUBLISHERS Urban Pfenning is based on RC. Szara, Washington, DC, USA, J.W. Thomas, la Grands, OR, USA, PJ/Trowbrldge, ttoca, NY, USA, T.RD. Turner, London, WC, IU.Vroom, Wagenhgen, Th9 Netherlands, W.V.Wendfer, CoSege Station, TX, USA, B.­E Yang, Seoul, Korea, E.H.2ute, Tucson, AZ, USA ABSTRACTED/INDEXED IN Applied Ecology Abstracts, Biological Abstracts, Current Contents B & S, Environmental Periodicals Bibliography, Geobasa, Geographical Abstracts, LandSearch. 1934 SUBSCRIPTION DATA Volumes 23­30 (In 9 Issues) Subscription prica: Dfl. 1080,00 (US $584.00) . tad Postage ISSN 0169­2046 EkeviirScIenesPuhfehors, P.O. Box 211, 1000 AE Amsterdam, Fax: (020) 5833­203 Customers In tin USA and Canada: BsaviwSdenesPuMsteis P.O. Box 945, Madison Square Station, N«wYorit,NY1016M757,USA Fax: (212) 633^3880