Gap Light Index Research Articles

Complex canopy structures control tree transpiration: A study based on 3D modelling in a tropical rainforest

AbstractTropical rainforests are rich in tree species and comprise complex canopy structures. Transpiration by forest trees is a major hydrological flux which contributes to climate regulation. We explored the role of forest canopy structure on tree transpiration in a tropical rainforest on Sumatra, Indonesia. Drone‐based photogrammetry and the structure from motion technique were used to compute high‐resolution 3D point clouds and derive forest and tree structural variables. Transpiration of differently sized and vertically positioned trees was assessed with sap flux measurements. Per‐tree transpiration rates increased linearly with several variables related to crown dimension and were further enhanced by top‐heavy crown shape, dense leafage and increasing canopy openness as assessed with the gap light index. Under given environmental conditions, the two variables crown volume and top‐heaviness explained 74% of the observed variance in per‐tree transpiration. Transpiration rates per unit crown dimension were highest for small crowns, decreased non‐linearly with increasing crown dimension and were little affected by other analysed structural variables. Our study underlines the potential of 3D point cloud analyses for accurately determining the structure of complex forest canopies and thus better understanding and predicting transpiration rates of differently positioned trees.

Comparative growth and physiological performance of American chestnuts, oaks, hickories, and sugar maple across a silvicultural gradient in overstory retention

The development of a blight-tolerant transgenic American chestnut (Castanea dentata (Marsh.) Borkh.) would provide the potential for reintroducing this species into forests it historically dominated. However, we lack a silvicultural understanding of the environmental conditions most favorable to the growth and survival of American chestnut relative to species with which it historically co-existed. We examined effects of shelterwoods with variable residual densities on the growth and physiology of blight-tolerant chestnut seedlings relative to sugar maple (Acer saccharum Marsh.), and several oak (Quercus spp.) and hickory (Carya spp.) species across a broad range of light availabilities (gap light index; GLI; of 4–50%). To assess the hypothesis that American chestnuts are more physiologically plastic and responsive to light than oaks and hickories, we measured seedling relative diameter and height growth rates and photosynthetic light-response curves across a GLI gradient. The diameter and height growth of chestnut increased with GLI, while oaks and hickories did not. All species varied aspects of their photosynthetic physiology across the GLI gradient in an approximately equivalent manner; the dark respiration rate and light compensation point increased equivalently across species in response to increases in light availability. A multivariate analysis identified three major axes of variation across ten measured traits, largely reflecting differences in species averages and phenotypic plasticity across the GLI gradient. American chestnut seedlings were spread across these multivariate dimensions more than other species, consistent with the hypothesis of higher phenotypic plasticity in this focal species. These results suggest that low residual shelterwood densities may better promote the establishment of underplanted chestnut seedlings relative to associated tree species provided it does not result in detrimental increases in the productivity of competing understory species.

Vertical light transmission profiles in structured mixed deciduous forest canopies assessed by UAV-based hemispherical photography and photogrammetric vegetation height models

Light availability below and within forest canopies governs many biological processes. Its estimation, however, is often time consuming, especially in complex canopy structures where vertical profiles of light transmission need to be known to estimate plant performance, habitat suitability or biophysical properties across the canopy. To overcome the technical limitations of assessing vertical canopy space by ground-based systems like ropes, towers or cranes, we developed a prototype to take hemispherical photographs (HP) from an unmanned aerial vehicle (UAV). With the new airborne system, we aimed to assess the differences of vertical light availability profiles from bottom to the top of the canopy in even-aged, uneven-aged and coppice-with-standards forests. Measured vertical light profiles typically followed a sigmoid shape that differed markedly depending on canopy gap size and forest structure. The results showed important variations in the transmission of sunlight at different heights within forest canopies, highlighting the importance of vertical canopy stratification when considering radiation transfer, plant area index (PAI), leaf area index (LAI) and light availability. To further cover cases where branches hinder UAV access to dense canopy spaces, we present a workflow to estimate light availability from photogrammetrically reconstructed high-resolution vegetation height models (VHM) and validate the outputs against HP. We conclude that VHMs based on high-resolution aerial photography, photogrammetry and a structure-from-motion algorithm (SfM) deliver a solid foundation to derive a wide range of canopy metrics. With an appropriate VHM point density, direct or beam light index (BLI), diffuse light index (DLI), gap or global light index (GLI) and PAI in mixed broadleaved forests can be reliably calculated. The novel development of UAV-based HP and the workflow to calculate light regimes from VHMs thus open a wide range of new applications in forest and agricultural research and management.

Temperate rain forest species partition fine-scale gradients in light availability based on their leaf mass per area (LMA).

Ecologists are increasingly using plant functional traits to predict community assembly, but few studies have linked functional traits to species' responses to fine-scale resource gradients. In this study, it was tested whether saplings of woody species partition fine-scale gradients in light availability based on their leaf mass per area (LMA) in three temperate rain forests and one Mediterranean forest in southern Chile. LMA was measured under field conditions of all woody species contained in approx. 60 plots of 2 m2 in each site, and light availability, computed as the gap light index (GLI), was determined. For each site, species' pairwise differences in mean LMA (Δ LMA) and abundance-weighted mean GLI (Δ light response) of 2 m2 plots were calculated and it was tested whether they were positively related using Mantel tests, i.e. if species with different LMA values differed in their response to light availability. Additionally linear models were fitted to the relationship between plot-level mean LMA and GLI across plots for each site. A positive and significant relationship was found between species' pairwise differences in mean LMA and differences in light response across species for all temperate rain forests, but not for the Mediterranean forest. The results also indicated a significant positive interspecific link between LMA and light availability for all forests. This is in contrast to what is traditionally reported and to expectations from the leaf economics spectrum. In environments subjected to light limitation, interspecific differences in a leaf trait (LMA) can explain the fine-scale partitioning of light availability gradients by woody plant species. This niche partitioning potentially facilitates species coexistence at the within-community level. The high frequency of evergreen shade-intolerant species in these forests may explain the positive correlation between light availability and LMA.

Early growth of hinoki (Chamaecyparis obtusa) trees under different topography and edge aspects at a strip-clearcut site in Kyushu, Southern Japan

The positive and negative effects of the residual stand edge at a strip-clearcut site were examined on the initial growth of hinoki (Chamaecyparis obtusa) planted in areas with diverse topography. On the south edge, the gap light index was lower than on the north edge and strip center; however, the vapor pressure deficit was also lower than on the north edge and strip center, which resulted in lower physiological stress of hinoki on the south edge. Tree size on the south edge did not exceed that on the north edge and strip center. These results indicated that low light conditions due to residual trees negatively affected growth on the south edge even under the positive effect of microclimate alleviation. In valley, tree size in the second year after planting was smaller than on ridge and slope; however, tree growth during the following 2 years was higher in valley. Surface soil in valley was thinner with rockier substrates than on ridge and slope; that is, soil sedimentation type and substrates influenced the initial growth of hinoki with undeveloped roots, and after that, topography started to influence growth, probably because hinoki roots penetrated into deeper soil, which is influenced by the water gathering capacity of the valley. The initial growth of hinoki at the strip-clearcut site was predominantly affected by light rather than microclimate alleviation. This alleviation is expected to disappear when hinoki trees planted in the strip center grow enough to give additional shade to slow-growing edge hinoki.

The relationship of stand structure with canopy transmittance: Simple models and practical methods for managing understory light conditions in eastern white pine (Pinus strobus L.)-dominated forests

The relationship of stand structural features with understory light levels, estimated by gap light index (GLI), was investigated in 22 second-growth eastern white (Pinus strobus L.) and red pine (Pinus resinosa Ait.)-dominated stands in central Ontario that encompassed a broad range in density and basal area. Simple, empirical light models were developed to quantify the influence of several stand structural variables on canopy transmittance as estimated by GLI. Models were also derived to facilitate the operational identification of residual basal area, density, and percent canopy closure associated with target understory light levels that optimize the growth of white pine regeneration and its protection from weevil and blister rust when using the uniform shelterwood silvicultural system. Regression models indicated significant negative, nonlinear relationships of GLI with density, basal area, a stand density index, total crown area, and foliar biomass, while GLI was linearly related to percent canopy closure. Application of these models to identify density, basal area, and canopy closure values associated with target light levels for the regeneration and removal cuts of uniform shelterwoods demonstrates the use of this information to help guide management of white pine–red pine forests.

Silvicultural treatments for converting loblolly pine to longleaf pine dominance: Effects on resource availability and their relationships with planted longleaf pine seedlings

Throughout the southeastern United States, land managers are currently interested in converting loblolly pine (Pinus taeda L.) plantations to species rich longleaf pine (Pinus palustris Mill.) ecosystems. In a 3-year study on moderately well- to well-drained soils of the Lower Coastal Plain in North Carolina, we examined the effects of four canopy and three cultural treatments on plant resources and quantified relationships between plant resources and longleaf pine seedling survival and growth. Canopy treatments consisted of four levels of timber harvest applied to loblolly pine stands: Control (uncut, mean basal area of 16.2m2/ha), MedBA (single-tree selection to a mean residual basal area of 9.0m2/ha), LowBA (single-tree selection to a mean residual basal area of 6.4m2/ha), and Clearcut (complete canopy removal). Within each canopy treatment, we applied three cultural treatments designed to benefit the early growth of planted seedlings: no treatment (NT), herbicide (H), and herbicide plus fertilization (H+F). Gap light index (GLI) significantly differed among canopy treatments and nonlinearly increased with decreasing basal area. The H treatment resulted in higher temperatures at 10cm in the soil. Canopy thinning increased foliar calcium (Ca) concentration. The annual root collar diameter (RCD) increment of planted longleaf pine seedlings was positively correlated with GLI, and foliar phosphorus (P) and Ca concentrations but was negatively correlated with soil moisture. Our results confirm that light is an important factor controlling the growth of longleaf pine seedlings.

Managing understory light conditions in boreal mixedwoods through variation in the intensity and spatial pattern of harvest: A modelling approach

In the context of partial harvesting, adequately managing post-harvest light conditions are essential to obtain a desired composition of tree species regeneration. The objective of this study was to determine how varying the intensity and spatial pattern of harvest would affect understory light conditions in boreal mixedwood stands of northwestern Quebec using the spatially explicit SORTIE-ND light model. The model was evaluated based on comparisons of observed and predicted light levels in both mapped and un-mapped plots. In mapped plots, reasonably accurate predictions of the overall variation in light levels were obtained, but predictions tended to lack spatial precision. In un-mapped plots, SORTIE-ND accurately predicted stand-level mean GLI (Gap Light Index) under a range of harvest intensities. The model was then used to simulate nine silvicultural treatments based on combinations of three intensities of overstory removal (30%, 45% and 60% of basal area) and three harvest patterns (uniform, narrow strips, large gaps). Simulations showed that increasing overstory removal had less impact on light conditions with uniform harvests, and a more marked effect with more aggregated harvest patterns. Whatever the harvest intensity, uniform cuts almost never created high light conditions (GLI > 50%). Gap cuts, on the other hand, resulted in up to 40% of microsites receiving GLI > 50%. Our results suggest that either a 30% strip or gap cut or a 45–60% uniform partial harvest could be used to accelerate the transition from an aspen dominated composition to a mixedwood stand because both types of cut generate the greatest proportion of moderately low light levels (e.g., 15–40% GLI). These light levels tend to favour an accelerated growth response among shade-tolerant conifers, while preventing excessive recruitment of shade-intolerant species. A better understanding of how spatial patterns of harvest interact with tree removal intensity to affect understory light conditions can provide opportunities for designing silvicultural prescriptions that are tailored to species’ traits and better suited to meet a variety of management objectives.

Calculation method for sunshine duration in canopy gaps and its application in analyzing gap light regimes

Understory light is essential to the growth and survival of plants, yet the light varies temporally and spatially within forest gaps. Measurement of understory light levels using light sensors is both time and labor intensive. Sunshine duration (SD) has a strong correlation with solar radiation and has been the variable most widely used for estimating solar radiation. The power of SD-based methods for estimating solar radiation lies in its ability to quickly estimate light levels. Although several calculation methods for SD within canopy gaps are available, all the methods oversimplify canopy gaps by classifying them as cylindrical or ellipsoid and thus have a relatively low level of accuracy. In this study we developed a calculation method for SD at any given point within natural canopy gaps and we used SD to estimate solar radiation within 12 canopy gaps on Mt. Taibai in China based on the Angström-Prescott model. We then evaluated the SD-based solar radiation by the total and direct solar radiation derived from a gap light index based on hemispherical photographs (HP). Finally, we analyzed the spatial-temporal characteristics of light levels within these 12 gaps by using the solar radiation derived from SD at hundreds of simulated points in each gap. Our results showed that (1) SD-based solar radiation was not statistically different from HP-based direct solar radiation and had a strong linear correlation with HP-based total solar radiation; (2) growing-season daily mean solar radiation within the 12 gaps varied from 0.08 to 13.28MJm−2day−1 with an average of 4.13MJm−2day−1; (3) solar radiation had a positive correlation with the ratio of the square root of the canopy gap area to the mean canopy height. This relationship was significant but solar radiation had no correlation with the canopy gap area; (4) among most gaps solar radiation was greatest in May while potential SD was longest in June 2008. From these results we can conclude that the SD-based method for estimating solar radiation developed in this study can quickly and accurately estimate light levels at any specified point within canopy gaps. SD-based solar radiation appears to be a good choice for studies on the spatial-temporal characteristics of gap light levels.

Improving gap light index (GLI) to quickly calculate gap coordinates

Understory light is essential to the establishment and growth of understory plants and varies temporally and spatially within gaps. The previously defined gap light index (GLI) is a good model for assessing understory light levels, but it is time-consuming to determine gap coordinates, which are crucial to GLI, for numerous points within a gap. This paper introduces the geometric calculation (GeoCalc) of gap coordinates. GeoCalc quickly obtains gap coordinates for any specified point within a canopy gap and takes into account the tridimensional profile of the gap and the slope and aspect of the ground. The GeoCalc-based GLI was validated by the GLI derived from hemispherical photographs taken at 93 sampling points within seven natural gaps. Our results demonstrate that GeoCalc-based GLI was strongly positively correlated and not significantly differed from the GLI derived from hemispherical photographs. Next, to analyze gap light regimes and the effects of gap size, canopy height, and topography, three natural gaps of various size were selected and simulated as nine gaps with 1 and 1.5 times canopy height or on the opposite slope. Finally, we have summarized characteristics of GeoCalc-based GLI and its application.

Effects of light and microsite conditions on tree size of 6-year-old Cryptomeria japonica planted in a group selection opening

We examined the extent to which direct and indirect measures of light and microsite conditions could explain variation in tree height and diameter at the base of 6-year-old Cryptomeria japonica trees planted in a group selection opening of about 0.32 ha on a steep slope at Shiiba, Miyazaki Prefecture, southern Japan. We first used the gap light index (GLI) and soil thickness (ST) as directly measured indices. For an indirect measure of light, we used a between-cohort competition index (BCI) estimated from the position and total height of residual trees. For indirect measures of microsite, we examined topographic indices (slope, plan and profile curvature, average slope gradient, and relative elevation) derived from digital elevation models (DEMs) with different resolutions ranging from 2 to 10m. The multiple linear regression using GLI and ST explained about 45% of variation in tree size, while simple regression using only GLI explained about 35%. The contribution of ST was about half of GLI. The multiple regressions using BCI and the topographic indices did not explain any more variation than using BCI alone (R2 of about 0.26). We conclude that microsite conditions with shallower soil and steeper slope have negative effects on tree growth in group selection openings, although the relative importance is smaller than light conditions. More comprehensive studies considering several openings with more heterogeneous topography including different species are needed to generalize our growth prediction using the indirect measures, which are useful for practical forest management.

Decomposition of green foliage in eastern hemlock forests of southern New England impacted by hemlock woolly adelgid infestations

This study examined the impacts of hemlock woolly adelgid (Adelges tsugae Annand) (HWA), a small, invasive insect, on foliar chemistry, forest floor microclimate, and subsequent green foliage decomposition in eastern hemlock (Tsuga canadensis (L.) Carrière) forests. We investigated the direct effects of HWA feeding and indirect changes in microclimate on foliar decomposition by incubating HWA-infested and uninfested foliage across eight eastern hemlock dominated stands in southern New England that had different histories of HWA infestation. Infested stands had much poorer average crown health (3.4 versus 1.4 crown loss ratings), higher percent open sky (10.9 ± 2.4 vs. 5.3 ± 0.5 gap light index), and lower organic soil moisture (0.83 ± 0.02 g·g–1 vs. 1.06 ± 0.05 g·g–1) than uninfested stands. There were no significant differences in percent C, percent N, or percent lignin for the excised foliage at the start of the study. However, after 120 days, decomposing foliage from infested trees had significantly higher N concentrations (1.83% ± 0.05% vs. 1.69% ± 0.02 %) and lower C/N ratios (29.9 ± 0.8 vs. 31.6 ± 0.2) than uninfested foliage, suggesting that HWA herbivory resulted in alterations of litter chemistry as decomposition progressed. Mass loss of common uninfested foliage was lower in uninfested hemlock stands than in infested ones (30.9% ± 0.7% vs. 34.2% ± 0.1%). These rates of mass loss were significantly correlated with microclimate factors and indicate that organic soil moisture levels are controlling decomposition in HWA-infested forests. Infestation by HWA causes virtually no direct feeding damage to foliage, but it does lead to several indirect and significant ecological and functional changes over the 10- to 20-year course of infestation and stand decline.

Methods for estimating understory light conditions using crown projection maps and topographic data

AbstractFour types of indices of canopy characteristics were compared to estimate understory light conditions using crown projection maps and topographic data. Crown area (CA) was formulated from the crowns in a focal sub‐block (20 m×20 m). Canopy shade index (CSI) was formulated from the crowns in a focal sub‐block and the surrounding sub‐blocks with topographic effect. Site shade index (SSI) incorporated the shade of the slopes into CSI and represented the total physical closure of the site. Site light index (SLI) incorporated the azimuth of the site into SSI and evaluated the light condition of the understory. We tested the methods using the data on steep old‐growth Cryptomeria japonica forests on Yakushima Island and the three‐dimensional canopy structures of the forests were digitized on a geographic information system. The results were validated by the basic indices obtained by hemispherical photographs: canopy openness (CO) and gap light index (GLI). The three‐dimensional structure of the surrounding canopies and topography had significant effects on the understory light conditions. CA had no significant correlation with CO, whereas CSI and SSI had significant negative correlations with CO. Although SLI had the highest negative correlation with GLI (R2=0.69), CSI had practical utility in the estimation of GLI (R2= 0.65). Our method is applicable to past crown projection maps that have been recorded at many forest research sites, thus making it possible to reconstruct past understory light conditions and to compare them with present conditions.

Simulating the impact of small-scale extrinsic disturbances over forest species volumetric light environment

Small-scale disturbances (SSD) creating canopy gaps are fundamental to successional dynamics in temperate forests. As gap-oriented management becomes very popular, spatial aspects of gap dynamics, especially the detailed impact of disturbances on the light environment for different species, remain understudied. The aim of this study is to evaluate this effect using the individual-based model SORTIE. Using different initial conditions, 10 simulated data sets, each representing temperate forests, were artificially disturbed using four disturbance sizes. For each 3D location of the simulation space, light availability was computed using the gap light index to create volumetric light data sets. The growth functions of the nine tree species incorporated in the simulation were mapped to each light data set, generating species-dependent 3D cubes illustrating the effect of small-scale disturbances over the different species according to their autecologic relationship to light. The general impact of the simulated SSD was assessed (1) by extracting the 3D boundaries associated to the absolute spatial influence of each replicated SSD and (2) by analyzing the variation of light inside and outside these boundaries, at different height levels. Results were compared for each disturbance size. The species response to different disturbance sizes was evaluated globally and also as a function of height levels under the canopy. This study revealed that the impact of different SSD schemes is highly variable among replicates. Nonetheless, results revealed that small size disturbances exhibit more heterogeneous impact. A threshold effect was detected around a disturbance size of 1000 m 2 suggesting a relative SSD impact that decreases for large SSD sizes. It was also found that species relationship is consistent between different disturbance schemes.

Light Transmittance Estimates in a Longleaf Pine Woodland

Abstract While the importance of canopy structure in open woodlands and savannas on regulating the flow of energy and matter is well known, few studies have investigated how variation in overstory abundance influences canopy light transmission and the extent that estimates vary in their ability to characterize the light environment in these ecosystems. Canopy light transmittance (% photosynthetic photon flux density, or %PPFD) was measured with gallium arsenide phosphide (GaAsP) photodiodes and was monitored throughout the growing season in an open-canopy longleaf pine (Pinus palustris Mill.) woodland across an overstory abundance gradient. Several estimates of canopy light transmittance were also measured, including 10 minute averages of %PPFD under clear and overcast sky conditions during summer, fall, and winter, as well as light estimates derived from hemispherical photographs (gap fraction, GF: gap light index, GLI; and weighted canopy openness, WCO). Linear and curvilinear regression were used to analyze the relationship between (1) light measurements and canopy structure and (2) light estimates and growing-season % PPFD measured by light diodes. Of all light measurements in this study, ten-minute average % PPFD measurements during clear days were the most variable: only a small proportion in the 10-minute average of light reaching the understory was correlated to either overstory structure or growing-season canopy transmittance. While measures on overcast days improved the correlations, they tended to overestimate transmittance. Gap fraction (GF) and gap light index (GLI) were strongly correlated with % PPFD and overstory structure, and the relationship between GLI and % PPFD was improved by using a direct light setting of 60%. In addition to strong correlation, hemispherical photographic estimates of light fell along a 1:1 line with % growing-season PPFD; thus, estimates were relatively unbiased. Measures of light in this ecosystem need to incorporate the large spatial and temporal variation in light transmittance that is characteristic of this ecosystem if investigators are to understand the relationship between open-canopy woodlands and their light environments.

SPATIAL DISTRIBUTION OF OVERSTORY RETENTION INFLUENCES RESOURCES AND GROWTH OF LONGLEAF PINE SEEDLINGS

Increasingly, overstory retention is being used in forests traditionally managed for single-cohort structure. One rationale for retention is that residual stand structure better resembles the complex structure of forests after natural disturbance, helping to perpetuate ecosystem functions dependent on that structure. The benefits of retention come at the cost of reduced survival and growth of regeneration because of competition with residual trees. We argue that inhibition of regeneration depends not only on the number and size of residual trees, but also on their spatial arrangement, which ranges from dispersed to aggregated. We use a model of competition at the scale of seedlings to hypothesize that maximum stand-level resource availability, seedling growth, and seedling survival occur with aggregate retention, rather than dispersed retention, even with constant residual basal area. We test our hypotheses with a silvicultural experiment in longleaf pine (Pinus palustris) in Georgia, USA. Replicated treatments included an uncut control, dispersed retention, small-aggregate retention, and large-aggregate retention. We measured light, soil nitrogen, soil moisture, and growth of longleaf pine seedlings across the full range of overstory conditions in each treatment. Postharvest basal areas in the cut treatments were similar. Gap light index increased from the control to large-aggregate retention, as did nitrogen availability, measured on exchange resins. Nitrogen mineralization did not differ among treatments, nor did soil moisture or temperature. Seedling biomass increment increased significantly from the control to large-aggregate retention. Survival did not differ among treatments. We argue that these results are a consequence of exponential relationships between overstory competition intensity, resource availability, and seedling growth. Given this relationship, resources and seedling growth are low across a wide range of decreasing overstory competitor abundance but increase exponentially only at very low competitor abundance. This seedling-scale model translates into maximum stand scale resource availability and seedling growth with large-aggregate retention, compared to dispersed retention, because the probability of a seedling occupying a site free of overstory competition is greater with the former. Our research shows that one can improve competitive environments for regeneration by manipulating spatial distribution of residual trees without sacrificing the ecological benefits of overstory retention. Corresponding Editor: D. L. Peterson

Factors limiting regeneration of Quercus alba and Cornus florida in formerly cultivated coastal plain sites, South Carolina

To determine the extent that resources, conditions and herbivory limit regeneration of Quercus alba L. and Cornus florida L. in formerly cultivated coastal plain uplands, we planted seedlings of the two species in two pine and one pine–hardwood forest understory and three adjacent clearcuts. Soil carbon and moisture, available nitrogen and phosphorus, and gap light index (GLI) were measured next to each seedling. Over two growing seasons, stem and leaf herbivory were estimated and survival was recorded. At the end of 2 years, all surviving stems were harvested to determine total leaf area and 2-year biomass growth. Survival to the end of the study was not significantly different between clearcuts (62.9%) and understories (64.6%). However, clearcuts led to significantly greater biomass growth and leaf area (ANOVA; P<0.05) for both Q. alba and C. florida. Soil moisture and available nutrients were also greater in the clearcuts ( P<0.05). Using separate multiple linear (growth) or logistic (survival) regressions for each combination of three sites, two cutting treatments and two species, we found that soil moisture significantly affected survival in 12.5% and biomass growth in 8.3% of the regressions. Light availability significantly impacted biomass growth in 16.7% of the regressions. Stem and leaf herbivory had very little impact on survival (8.3%), but when combined, these two factors significantly impacted leaf area or biomass growth in 33.3% of the regressions. Seedling responses were highly variable, and no regression model accounted for more than 70.0% of this variation. In our study, stand-scale variation in seedling responses (especially the difference between clearcut and understory) was much greater than within-stand variation. Of the within-stand factors measured, herbivory was clearly the most important. To establish these species in mesic upland coastal plain sites, we recommend planting immediately after clearcutting.

The effect of spatially variable overstory on the understory light environment of an open-canopied longleaf pine forest

Spatial aggregation of forest structure strongly regulates understory light and its spatial variation in longleaf pine (Pinus palustris Mill.) forest ecosystems. Previous studies have demonstrated that light availability strongly influences longleaf pine seedling growth. In this study, the relationship between spatial structure of a longleaf pine forest and spatial pattern of understory light availability were investigated by comparing three retention harvest treatments: single-tree, small-group, large-group, and an uncut control. The harvests retained similar residual basal area but the spatial patterns of the residual trees differed. Hemispherical photographs were taken at 300 stations to calculate gap light index (GLI), an estimate of understory light availability. Stand-level mean, variation, and spatial distribution of GLI were determined for each treatment. By aggregating residual trees, stand mean GLI increased by 20%, as well as its spatial variation. Spatial autocorrelation of GLI increased as the size of the canopy gaps increased and the gaps were better defined; thus, the predictability of GLI was enhanced. The ranges of detrended semivariograms were increased from the control to the large-group harvest indicating the spatial patterns of understory GLI became coarser textured. Our results demonstrated that aggregated canopy structure of longleaf pine forest will facilitate longleaf pine seedling regeneration.

Variation in canopy openness and light transmission following selection cutting in northern hardwood stands: an assessment based on hemispherical photographs

The objective of this study was to determine how canopy openness (CO) and light transmission are affected by selection cutting, and how they vary over time following harvesting in northern hardwood stands. We sampled five sugar maple—yellow birch—beech ( Acer saccharum— Betula alleghaniensis— Fagus grandifolia) stands in Québec (Canada). The stands had been logged, using the selection system, at different times (2–14 years) before the study, and were used as a chronosequence. We also sampled portions of each stand which had been kept as uncut controls. Ten 1 ha plots were sampled (five cuts and five paired controls). We took 20 hemispherical photographs per plot, at 5 m above-ground, which was above most understory vegetation. The CO, light transmission (gap light index (GLI) sensu Canham (1988)), sunflecks characteristics, and angular distribution of openings from the zenith were calculated for each photograph. Selection cutting increased CO, especially within 60° of the zenith. The greater CO in the cuts allowed a greater light transmission (GLI), longer sunflecks, and a longer cumulative daily sunflecks duration (CDSD). The differences observed between the cuts and the control plots in terms of CO, GLI, and CDSD were greatest in the more recent cuts, and decreased as a function of time since logging. The relationships were best described by negative logarithmic (CO) and negative exponential (GLI, CDSD) models. In the youngest cut (2 years old), the CO, GLI, and CDSD were on average 2.3–2.5 times higher than in the control, while in the oldest cut (14 years old), the same variables were 1.6–1.7 times higher than in the control. The results of this study emphasize the importance of taking into account the temporal variation in canopy openness and light transmission after canopy disturbances such as selection cutting because that variation will likely have an important influence on regeneration dynamics.

Hurricane Impacts on a Mangrove Forest in the Dominican Republic: Damage Patterns and Early Recovery1

On 22 September 1998, Hurricane Georges passed over the Dominican Republic causing extensive damage to a 4700 ha mangrove forest that has been the site of a detailed study of vegetation and ecosystem dynamics since 1994. We resurveyed the vegetation in permanent plots at 7 and 18 months after the hurricane to document structural damage of the forest and evaluate early recovery patterns. The intensity of damage was patchy across the landscape. Mortality (≥5 cm DBH) ranged from 14 to 100 percent (by density) among the 23 different plots and averaged 47.7 percent across all plots. Reductions in total basal area ranged from 9 to 100 percent, averaging 42.4 percent. Mortality increased by 9 percent between surveys at 7 and 18 months post-hurricane. Interspecific differences in susceptibility to wind damage appeared to be a primary factor contributing to spatial patterns in mortality. Laguncularia racemosa experienced much less mortality (26%) than either Rhizophora mangle (50%) or Avicennia germinans (64%), and plot-level mortality was strongly associated with differences in species composition. There were no clear relationships between canopy height and tree damage at this site. Over 80 percent of the of the surviving R. mangle trees exhibited less than 50 percent crown damage, whereas ca 60 percent of the L. racemosa survivors suffered almost complete (75–100%) crown loss. By 18 months after the hurricane, the percentage of L. racemosa trees in the 75 to 100 percent damage class was reduced to 20 percent; in contrast, the health of many R. mangle individuals appeared to be declining, as the percentage of trees in the 50 to 100 percent damage class increased from 16 to 36 percent. Understory light levels, as measured by the gap light index, increased from an average value of 3 percent in the pre-hurricane forest to 51 percent at 7 months after the hurricane and decreased slightly to 47 percent at 18 months. Few saplings (>1 m tall and <5 cm DBH) survived the hurricane; 72 percent of the tagged individuals in transect-based plots and 66 percent of saplings in pre-hurricane canopy gaps were killed. Seedling and sapling populations of all three species appear to be recovering rapidly although their densities still are lower than in the pre-hurricane forest. It is too early to predict the trajectory of forest recovery, and continued monitoring of the spatial and temporal patterns of forest development is needed to improve our understanding of the role that large-scale disturbance events play on the dynamics of mangrove forest ecosystems.