Plant community responses to a gradient of site preparation intensities in pine plantations in the Coastal Plain of North Carolina

Intensively managed loblolly pine (Pinus taeda) forests are common in the southeastern United States and offer opportunities for conservation of biologic diversity. Within intensively managed landscapes, stand establishment relies on combinations of mechanical and chemical site preparation and herbaceous weed control (HWC) to manage competing vegetation and increase pine production. However, few long-term studies have described relationships between intensity of stand establishment and effects on plant communities. Therefore, we examined effects of 6 treatments that varied in intensity via combinations of mechanical (wide spacing and strip shear or narrow spacing and roller chop) and chemical (application or no application) site preparation treatments with HWC (broadcast or banded) from 1 to 8 years after site preparation on plant communities in loblolly pine plantations (n = 6) in the Coastal Plain of North Carolina, USA. All treatments resulted in abundant and diverse plant communities. Chemical site preparation had short lived (?4 year) effects on the herbaceous community but long-term effects on woody plants and pine cover. Increasing management intensity by including broadcast HWC or roller chop/narrow spacing did not additively reduce woody vegetation cover or species richness. However, broadcast HWC reduced grass, vine, and forb cover in the first year post-treatment. Average Morista community similarity values ranged from 0.69 to 0.89 among treatments and plots receiving the same chemical site preparation contained the most similar plant communities. Banded HWC can be paired with wide spacing to maximize herbaceous plant growth important for many wildlife species, particularly in the first few years after site preparation. Site preparation techniques should be tailored to local site conditions, plant communities, and management objectives.     

Growth Response of Loblolly Pine (Pinus taeda L.) 3–5 Years After Plantation Establishment Using Different Management Intensities

The upward trend of intensive management in Southern pine forests is expected to continue, both in area and intensity level. Much of the Mississippi Lower Coastal Plain (LCP) is managed intensively using some combination of mechanical site preparation, chemical site preparation, and herbaceous weed control (HWC). We studied pine growth response and competition control on loblolly pine (Pinus taeda L.) plantations 3–5 years following establishment using five combinations of chemical site preparation, mechanical site preparation, and HWC. Treatments were designated a priori as 1 (least intensive) through 5 (most intensive) largely on the basis of anticipated impact on the vegetation community. We measured pine height and diameter at breast height (dbh); woody stem density; hardwood basal area (BA); coverage of herbaceous plants, understory woody plants, and pine trees; and estimated differences in pine response using age-shift calculations at age 5. Pine height and dbh were correlated with treatment intensity; treatment 5 maintained an average advantage of 1.4-m height and 2.5-cm dbh over treatment 2, the least responsive treatment. Woody stem density varied widely and was not affected by treatment; understory woody coverage was also unaffected by treatment. Coverage of herbaceous plants was reduced in treatments receiving broadcast HWC, and in treatment 2, where slower establishment of pines may have acted as a release for competing woody plants. Age-shift gains relative to treatment 2 ranged from 0.4–1.0 years. Based on year 5 measurements of hardwood BA, it is likely that treatments 1 and 2 will fall further behind treatments 3–5 as the stands mature. Greatest control of competing vegetation and maximum growth of pines was achieved with the most intensive treatment.     

Vegetation community responses to different establishment regimes in loblolly pine (Pinus taeda) plantations in southern MS,USA

Management treatments involving multiple herbicide applications are standard procedure on most industrial pine plantation sites in the southern USA, raising concerns about biodiversity impairment. Management decisions impact not only plant communities but also the habitat potential they create for wildlife. We tested the effects of five intensities of stand establishment treatments on vegetation communities in loblolly pine plantations (n = 4) to age 5 in the Middle Coastal Plain (MCP) of Mississippi. Measurements were species richness, diversity, coverage, and community composition. Treatments were combinations of mechanical site preparation (MSP), chemical site preparation (CSP), and herbaceous weed control (HWC) both banded or broadcast for 1 or 2 years using the same herbicide mixtures. Tree richness and diversity were reduced by increasing treatment intensity; tree coverage, which included crop and non-crop trees, was less in moderate-intensity treatments. Vine richness and coverage were less in more intensive treatments, but 2 diversity indices differed on whether vine diversity was likewise affected. Richness and coverage of forbs and graminoids was lessened by broadcast HWC, with effects mostly limited to the year of application. Plant communities differed in all 5 years, with CSP acting as the primary factor for years 2–5. Early seral communities were favored by CSP, but broadcast HWC suppressed resulting herbaceous plants. Though CSP may somewhat reduce stand-level plant diversity, it may increase overall biodiversity within plantation-dominated landscapes by creating early succession plant communities that enhance wildlife habitat.     

Stand restoration burning in oak–pine forests in the southern Appalachians: effects on aboveground biomass and carbon and nitrogen cycling

Understory prescribed burning is being suggested as a viable management tool for restoring degraded oak–pine forest communities in the southern Appalachians yet information is lacking on how this will affect ecosystem processes. Our objectives in this study were to evaluate the watershed scale effects of understory burning on total aboveground biomass, and the carbon and nitrogen pools in coarse woody debris (CWD), forest floor and soils. We also evaluated the effects of burning on three key biogeochemical fluxes; litterfall, soil CO₂ flux and soil net nitrogen mineralization. We found burning significantly reduced understory biomass as well as the carbon and nitrogen pools in CWD, small wood and litter. There was no significant loss of carbon and nitrogen from the fermentation, humus and soil layer probably as the result of low fire intensity. Burning resulted in a total net loss of 55 kg ha⁻¹ nitrogen from the wood and litter layers, which should be easily replaced by future atmospheric deposition. We found a small reduction in soil CO₂ flux immediately following the burn but litterfall and net nitrogen mineralization were not significantly different from controls throughout the growing season following the burn. Overall, the effects of burning on the ecosystem processes we measured were small, suggesting that prescribed burning may be an effective management tool for restoring oak–pine ecosystems in the southern Appalachians.     

Application of Choice Experiments to Determine Stakeholder Preferences for Woody Biomass Harvesting Guidelines

Biomass harvesting guidelines (BHGs) have been developed to address concerns about the sustainability of harvesting woody biomass. Assessing preferences among BHG stakeholders is important for designing operationally feasible and socially acceptable standards in different contexts. We used choice modeling to determine how foresters, loggers, and landowners perceived the relative importance of stumpage price, wildlife habitat quality, percentage of coarse woody debris (CWD) remaining, and distribution of CWD in their choices of BHG scenarios. Responses (N = 718) indicated stumpage price was nearly double the importance of wildlife habitat quality, and three times more important than debris distribution and debris remaining.     

Twenty year site preparation effects on sub-boreal lodgepole pine performance

We examined the effects of various mechanical site preparation methods and windrow burning on container-grown planted lodgepole pine (Pinus contorta var. latifolia) survival and growth for 20?years after treatment at a sub-boreal site in north-central British Columbia, Canada. Survival was uniformly high (??80%) regardless of treatment, indicating that site preparation was not necessary to establish pine on this site. Significant treatment effects on height, diameter, and stem volume were present at all assessment dates, but only the windrow burning treatment was associated with growth gains over the untreated control after two decades. Pine planted at the disk trench hinge were significantly larger than control pine only until year five. Of the mechanical treatments, only coarse mixing (by bedding plow) continued to have a significant effect on pine growth for as many as 9?years after treatment. Despite the disappearance of significant differences between mechanical treatments and the untreated control by year 20, the magnitude of stand volume increases suggests the potential for mechanical site preparation to have a beneficial effect on future timber supply. Repeated measures analysis confirmed that trends in early diameter growth differed between the untreated control and the windrow burning or coarse mixing treatments. These data are also potentially valuable for verifying growth and yield or carbon budgeting modelling tools.     

Modeling production and decay of coarse woody debris in loblolly pine plantations

Forest management can have large impacts on the production and yield of coarse woody debris (CWD) in terrestrial ecosystems, yet few modeling tools exist to inform such efforts. The goal here was to develop a set of prediction equations for use in conjunction with loblolly pine (Pinus taeda L.) modeling and inventory systems to estimate CWD yields at scales ranging from individual trees to whole plantations. Permanent field plots from a 21-year study of thinning effects on plantation growth and yield across the commercial range of the species in the southern United States were surveyed to obtain sample data on CWD volume, density, and mass. Measured CWD properties were combined with inventory records of tree mortality over the study duration to characterize CWD production, decay and yield in a series of prediction equations. The resulting equations predict CWD attributes of dead trees including dry weight (kg) and fraction of standing versus downed woody material based on the time since death (years), tree diameter at breast height (cm) and height (m) at time of death and geographic coordinates of latitude and longitude. A stand-level equation predicts total CWD yield (Mg ha⁻¹) for thinned or unthinned stands based on plantation age, stem density (trees ha⁻¹), and the average height of dominant and codominant trees (m). Piece-level equations predict dry density (kg m⁻³) or nitrogen concentration (%) of CWD pieces based on their position (standing or down), ordinal decay classes, and latitude. The tree and stand-level prediction equations are designed for use in GIS or growth and yield modeling systems. The piece-level equations are designed to be used in inventory applications that survey CWD. The equations should facilitate the accurate and facile determination of mass, carbon, and nitrogen contents of CWD in planted loblolly pine forests of the southern United States.     

Influence of fire intensity on structure and composition of jack pine stands in the boreal forest of Quebec: Live trees,understory vegetation and dead wood dynamics

North American jack pine (Pinus banksiana Lamb.) stands are generally characterized by an even-aged structure resulting from high intensity fires (HIF). However, non-lethal fires of moderate intensity (MIF), which leave behind surviving trees, have also been reported. The objectives of this study were two-fold: (1) assess the concurrent dynamics of live trees, understory vegetation and different types of coarse woody debris (CWD) during succession after HIF; and (2) document how MIF affects stand structure component dynamics compared to HIF. Stands affected by both HIF and MIF were selected. Tree characteristics and age structure, understory biomass, and CWD volume were assessed. Our results suggest that the structural succession of jack pine stands following HIF comprises three stages: young stands (<48 years), premature and mature stands (58–100 years) and old stands (>118 years). Canopy openness and jack pine density significantly decreased with time since HIF, while black spruce density and CWD volume significantly increased. The highest structural diversity was measured in the premature and mature stands. Compared to HIF, MIF increased mean jack pine basal area, decreased average stand density, delayed the replacement of jack pine by black spruce replacement in the canopy, decreased CWD volume, and significantly increased bryophytes mass. MIF increased the diversity of live trees and generally decreased CWD structural diversity. The study confirms the diversity of natural disturbance magnitude and successional processes thereby initiated. Thereafter, it appeared to be relevant for adjustment of disturbance emulating forest-management systems.     

The effect of shelterwood harvesting and site preparation on eastern red-backed salamanders in white pine stands

We studied the effects of the regeneration cut of the shelterwood system and four site preparation options on populations of eastern red-backed salamanders in 90–100-year-old white pine forests in central Ontario, Canada. We established the study in 1994 using a randomized complete block design with three replicates and five treatments: (1) no harvest, no site preparation; (2) harvest, no site preparation; (3) harvest, mechanical site preparation; (4) harvest, chemical site preparation; (5) harvest, mechanical and chemical site preparation. We applied harvest and site preparation treatments from fall 1995 to fall 1997. We collected pre-treatment data in spring and summer of 1995 and post-treatment data from 1998 to 2002. We monitored salamander abundance using a grid of 20 cover boards surveyed 10 times per year within each of the 15 treatment plots. We also quantified changes in overstory and understory cover, supply of downed woody debris, and disturbance to the forest floor. Our data suggest that shelterwood cutting and site preparation can have immediate negative effects on the abundance of red-backed salamander populations in pine forest. However, effects are relatively short lived (<5 years). Changes in abundance appeared to be related to overstory and understory cover, and forest floor disturbance.     

Effects of structural complexity enhancement on eastern red-backed salamander (Plethodon cinereus) populations in northern hardwood forests

Managing for stand structural complexity in northern hardwood forests has been proposed as a method for promoting microhabitat characteristics important to eastern red-backed salamanders (Plethodon cinereus). We evaluated the effects of alternate, structure-based silvicultural systems on red-backed salamander populations at two research sites in northwestern Vermont. Treatments included two uneven-aged approaches (single-tree selection and group-selection) and one unconventional approach, termed “structural complexity enhancement” (SCE), that promotes development of late-successional structure, including elevated levels of coarse woody debris (CWD). Treatments were applied to 2 ha units and were replicated two to four times depending on treatment. We surveyed red-backed salamanders with a natural cover search method of transects nested within vegetation plots 1 year after logging. Abundance estimates corrected for detection probability were calculated from survey data with a binomial mixture model. Abundance estimates differed between study areas and were influenced by forest structural characteristics. Model selection was conducted using Akaike Information Criteria, corrected for over-dispersed data and small sample size (QAIC_c). We found no difference in abundance as a response to treatment as a whole, suggesting that all of the uneven-aged silvicultural systems evaluated can maintain salamander populations after harvest. However, abundance was tied to specific structural habitat attributes associated with study plots within treatments. The most parsimonious model of habitat covariates included site, relative density of overstory trees, and density of more-decayed and less-decayed downed CWD. Abundance responded positively to the density of downed, well-decayed CWD and negatively to the density of poorly decayed CWD and to overstory relative density. CWD volume was not a strong predictor of salamander abundance. We conclude that structural complexity enhancement and the two uneven-aged approaches maintained important microhabitat characteristics for red-backed salamander populations in the short term. Over the long-term, given decay processes as a determinant of biological availability, forestry practices such as SCE that enhance CWD availability and recruitment may result in associated population responses.     

End-point diameter and total length coarse woody debris models for the United States

Coarse woody debris (CWD) may be defined as dead and down trees of a certain minimum size that are an important forest ecosystem component (e.g., wildlife habitat, carbon stocks, and fuels). Due to field efficiency concerns, some natural resource inventories only measure the attributes of CWD pieces at their point of intersection with a sampling transect (e.g., transect diameter) although measurements of large-end diameter, small-end diameter, and length are often required by natural resource managers. The goal of this study was to develop a system of empirical models that predict CWD dimensions (e.g., large-end diameter) based on CWD attributes measured at the point of intersection with a sample transect and ancillary data (e.g., ecological province). Results indicated that R-squared (R²) values exceeded 0.60 for most of this study's CWD large-end diameter and small-end diameter with only fair results for the length models. The mean residuals of numerous CWD models were within the measurement tolerance expected of actual field crews. Despite remaining unexplained variation, these CWD models may provide foresters with an alternative to the time-consuming activity of measuring all CWD dimensional attributes of interest during large-scale forest inventories.     

Bark Beetle Demography in a Jeffrey Pine Stand as Influenced by Mechanized Thinning and Prescribed Fire

Forest thinnings implemented with cut-to-length and whole-tree harvesting systems followed by underburning were evaluated for their effects on bark beetle prevalence in pure, uneven-aged Jeffrey pine (Pinus jeffreyi Grev. & Balf.) interspersed with isolated California white fir (Abies concolor var. lowiana [Gord.] Lemm.). Based on pitch tube counts in a stand with a moderate bark beetle population in its pine component, the Jeffrey pine beetle (Dendroctonus jeffreyi Hopkins) generally preferred larger trees before treatment implementation, but after exhibiting mixed pretreatment tendencies concerning stand density demonstrated a posttreatment proclivity toward higher density. Cut-to-length thinning followed by underburning increased the pine beetle population while whole-tree thinning unaccompanied by burning reduced it. Tree mortality was induced by the bark beetle infestation but was not its sole cause. Pitch tube abundance on white fir far exceeded that on Jeffrey pine, and the greatest influence on the fir engraver (Scolytus ventralis LeConte) population was the prevalence of its host tree. The responses presented herein to these thinning and burning practices, which are being increasingly utilized in forest restoration efforts in the western USA, provide natural resource managers insight into potential forest health outcomes when implemented in Jeffrey pine and similar dry site forest types.     

Diversity of polyporous fungi (Polyporaceae) in northern boreal forests: effects of forest site type and logging intensity

The effects of forest site type and logging intensity on polyporous fungi were studied in subxeric, mesic and herb-rich forests and spruce mires in northern Finland. The species richness of polypores did not follow the fertility gradient of the site types, but was connected with the amount and diversity of coarse woody debris (CWD). The total number of species, and the numbers of indicator and threatened species were equal in subxeric pine forests and in more fertile spruce-dominated stands. The species composition of pine-dominated forests differed conspicuously from that of spruce-dominated site types. The total number of species was not affected by logging intensity, but no virgin forest species or threatened species were found on the sites where the number of cut stumps exceeded 150 stumps ha^?1. Increasing logging intensity decreased the number of polypore observations, indicating reduced substrate availability. The results stress the importance of protecting not only fertile spruce-dominated stands, but also poorer, pine-dominated forests, and sites with high and diverse CWD content.     

Simultaneous prediction of the development of loblolly pine and woody competitors in young plantations

A system for examining the effects of hardwood density and cover of herbaceous components on mean size in newly established loblolly pine plantations was developed using a site preparation study located in the piedmont province of Georgia (USA). Multispecies density models were derived to predict the sum of crown heights for eight important hardwood species using both the current or age 1 number of rootstocks and herbaceous vegetation to account for intra- and interspecific effects. The predicted sums of crown heights for the woody species were then used as predictors of mean height, diameter, and volume for loblolly pine. A seemingly unrelated regression procedure was used to compensate for correlations in error components within each system of equations that result from using predicted crown heights as regressor variables. The effects of varying sweetgum density and andropogon cover on the inter-related components of the systems based on current and age 1 competition were examined for mean height and mean diameter of the planted loblolly pines. The cumulative impacts of associated vegetation on loblolly pine growth and the relative sensitivity of height and diameter to competing vegetation were demonstrated.     

Twenty-six-year response of ponderosa pine and Douglas-fir plantations to woody competitor density in treated stands of madrone and whiteleaf manzanita

Ponderosa pine (Pinus ponderosa Dougl.) grown in mixture with whiteleaf manzanita (Arctostaphylos viscida Parry) and Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco var. menziesii) grown in mixture with Pacific madrone (Arbutus menziesii Pursh) in southwestern Oregon showed an increase in growth with removal of competing woody cover. Both conifer species had roughly one-third the volume at plantation ages 26–27 when grown with uncontrolled competition compared to where woody competition was completely controlled at age 2. Intermediate levels of competitors usually led to intermediate levels of growth, but this was more evident with Douglas-fir than pine. When competition was reduced or removed, height/age relationships for Douglas-fir at plantation ages 23 and 27 reflected medium site quality rather than low quality as estimated from adjacent stands, indicating that these sites are potentially more productive than perceived with uncontrolled dense woody cover. These studies support the concept that competition management may allow some poor sites of ponderosa pine or Douglas-fir to be managed on the basis of a higher site potential.     

Dead wood volume to dead wood carbon: the issue of conversion factors

Requirements for emission reporting under the Kyoto protocol demand an estimate of the dead wood carbon pool in forests. The volume of dead wood consists of coarse woody debris, smaller woody debris and dead roots. The measurement of dead wood volume was included in the most recent National Forest Inventory in Switzerland. To convert dead wood volume into carbon two conversion factors are required: (a) carbon (C) concentration and (b) wood density. So far internationally accepted default values for C concentration (50%) and for wood density (density of alive trees) were used as default values to estimate dead wood carbon, since local measurements were lacking. However, in a field study at 34 sites in Switzerland, the C concentration and density of CWD from Picea abies and Fagus sylvatica of four decay classes were measured recently. The results showed that C concentration in CWD differed significantly between species but did not change due to decay class. The density of CWD decreased significantly with an increase in decay class and it also differed between species. The decrease in CWD density was more pronounced for F. sylvatica than for P. abies. We assessed correlations between climate attributes and CWD density using regression analysis. The modeled densities and measured C concentrations were then expanded with the help of CWD volume data from the NFI3. Spruce CWD and thus spruce CWD carbon is much more abundant in Swiss forests than beech CWD carbon. The majority of spruce CWD is located in the Alps and Pre-Alps. The CWD volume from P. abies was 10 times higher than that from F. sylvatica. Thus, changes in conversion factors for P. abies CWD affected the overall estimate of dead wood carbon in Swiss forests much more than changes in conversion factors for F. sylvatica CWD. Current improvements in CWD conversion factors decreased the estimated amount of spruce CWD carbon by 23.1% and that of beech by 47.6%. The estimated amount of CWD carbon in Swiss forests is decreased by 31%. Since improved estimation methods are currently not applied to smaller woody debris and dead root material, the estimated amount of dead wood carbon is only reduced by 15%. Improving conversion factors for all dead wood fractions would presumably decrease the amount of dead wood carbon by additional 16%.     

Plant species cover,frequency, and biomass: Early responses to clearcutting,burning, windrowing,discing, and bedding in Pinus elliottii flatwoods

Pinelands in a 49 ha naturally-regenerated, mature flatwoods forest in north Florida were clearcut harvested in the fall of 1978, site prepared by burning, shearing and piling, discing, and bedding, and planted to slash pine (Pinus elliottii) in 1979. Three vegetation surveys were conducted: one prior to harvest in the summers of 1977 and 1978, and two subsequent to planting in the summers of 1980 and 1981.Cover and frequency of all plant species encountered were assessed on permanent transects. Foliage biomass by species was assessed by destructive sampling of distinct subplots within permanent plots. The aim was to assess plant species cover, frequency, and biomass responses to the forest operations imposed.Pine was eliminated as a dominant genus by harvesting. Planted pines were a fast increasing, but not dominant, component of the vegetation at 2 years of age. Previously dominant shrubs were severely reduced — often by approximately two orders of magnitude. Indeed, woody species were severely reduced: woody cover from 151 to 12% of surface area at plantation age 2 years; woody biomass from 6223 to 521 kg/ha.Conversely, herbaceous species were substantially increased: herbaceous cover from 38 to 51% of surface area at 2 years; herbaceous biomass from 382 to 1439 kg/ha. Thus, a predominantly woody ecosystem was converted to a predominantly herbaceous one for 2 years following planting.There was little change in plant species richness as a result of forest operations. Plant diversity changes were mixed the 1st year but diversity substantially increased the 2nd year. Comparisons with a companion study (Conde et al., 1983) suggest that, while increases in diversity following maximum site preparation may lag increases following minimum site preparation, cover, frequency, and biomass diversity all converge to common values after 2 years despite disparity in treatments.     

Coarse woody debris dynamics in a post-fire jack pine chronosequence and its relation with site productivity

The long-term relationships between coarse woody debris (CWD) dynamics, soil characteristics and site productivity have, so far, received little attention. The objectives of the study were to describe CWD dynamics along a post-fire chronosequence (43–86 years after fire) in jack pine (Pinus banksiana Lamb.) stands, assess the importance of buried CWD in terms of soil available water holding capacity (AWHC), and investigate relationships between CWD, AWHC, nutrient retention and site productivity.

Twelve jack pine stands on sandy, mesic sites of glaciolacustrine origin were surveyed. Buried wood volume within the forest floor varied between 1 and 57 m³ ha⁻¹ (4–92% of total site CWD volume) and showed no relationship with time. Downed log mass accumulation followed a “U shaped” successional pattern with time since fire. Buried wood AWHC was negligible compared with that of the 0–20 cm mineral soil layer. The most productive sites were characterised by higher forest floor dry weight, effective CEC and water holding capacity in the mineral soil. Path analyses of relationships between organic matter content, CWD and forest floor CEC showed that CEC was conditioned by forest floor organic matter and buried wood content.     

Long-term response of planted conifers, natural regeneration, and vegetation to harvesting, scalping, and weeding on a boreal mixedwood site

This study reports 14th-year response of a boreal mixedwood stand to different harvest intensities (uncut, 50% partial cut with and without removal of residuals after 3 years, and clearcut), spot site preparation treatments (none and scalped), and chemical weeding frequencies (none, single, and multiple) in northeastern Ontario. The response variables include the survival and growth of planted white spruce (Picea glauca [Moench] Voss) and jack pine (Pinus banksiana Lamb.), height and density of natural regeneration and shrubs, and cover of shrubs and non-woody vegetation. Harvesting and weeding generally improved survival and growth of planted trees, although white spruce survival did not significantly differ among the three weeding frequencies. Harvesting tended to increase heights of hardwood (mostly trembling aspen (Populus tremuloides Michx.)) and conifer (largely balsam fir (Abies balsamea (L.) Mill.).) natural regeneration, cover and density of shrubs, and cover of herbs, lichens, and ferns. Chemical weeding reduced height, density and cover of shrubs, height and density of hardwood regeneration, and fern cover, but increased moss and lichen cover. Spot scalping did not significantly affect planted seedling, natural regeneration, or the vegetation.Maximum survival and growth of planted white spruce and jack pine were achieved using a combination of clearcutting and multiple weeding. However, partial cutting followed by a single weeding produced acceptable survival and reasonable growth of planted trees, particularly for white spruce. Partial canopy removal alone substantially reduced the amount of hardwood regeneration, relative to clearcutting, but did not adequately suppress understory shrubs. Significant improvement in seedling growth following multiple weedings was evident primarily in the complete canopy removal treatments: 50% partial cut with removal of residuals after 3 years and clearcut. While the effects of harvesting and weeding on planted crop trees found in the 5th-year assessments generally persisted at year 14, survival decreased, likely due to light competition from developing hardwood and shrubs.