Lower Seedling Density Research Articles

Early regeneration results following aggregated retention harvesting of wet eucalypt forests in Tasmania, Australia

Aggregated retention (ARN) harvesting has been adopted as an alternative to clearfelling in most of Tasmania’s wet eucalypt forests, although the effects of this silvicultural system on regeneration establishment and growth were still uncertain when this decision was made. To address this knowledge gap, eucalypt regeneration was assessed in 31 operational ARN coupes and 31 paired clearfell, burn and sow (CBS) coupes established from 2007 to 2010. There was no difference in eucalypt seedling stocking, density or height between ARN and CBS coupes at 1year of age. At 3years of age, seedling density and height did not vary with silvicultural system, and stocking was only 5% lower in ARN coupes when two outliers were removed. This early regeneration success in the ARN coupes is attributed to the high availability of receptive seedbed achieved in the broadcast burns on these coupes, the adoption of aerial sowing as a standard operating procedure, and to the absence of any increase in browsing pressure or edge-related growth suppression. Although the proportion of the harvested area within one tree height of standing forest (i.e., under ‘forest influence’) was 34% higher in ARN coupes compared to CBS coupes, there was no evidence of suppression of regeneration by the standing forest edge at 3years of age. Seedling height and density were strongly related to the state of the seedbed, and increased with increasing burn intensity. Seedling density increased by an average of 1700sph in both ARN and CBS coupes between the ages of 1 and 3years, but despite this ongoing recruitment, 24% of coupes had low seedling densities at age 3years and a review of sowing rates is recommended. This study confirms that the creation of burnt seedbed is essential for good early regeneration in wet eucalypt forests, and indicates that regeneration can be just as successful after aggregated retention harvesting as after clearfelling, so long as stands are managed appropriately.

Differential responses of Abies fabri and Rhododendron calophytum at two sites with contrasting pollution deposition and available calcium in southwestern China

The correlation between forest decline and calcium (Ca) depletion under long-term acid deposition remains elusive in China due to the high level of Ca deposition. We compared two species (Abies fabri and Rhododendron calophytum) for their growth pattern and base elements concentration in both polluted (Mt. Emei) and unpolluted (Mt. Gongga) sites in Sichuan, southwestern China. A. fabri grown at Mt. Emei had poorer crown condition, slower radial growth rate, and lower seedling density under long-term acid deposition, which correlated closely with the reduced Ca concentration in foliage and tree-ring, in comparison with those at Mt. Gongga. In contrast, R. calophytum showed a stable Ca level and thus displayed normal growth between the two sites. The differential capability of these two species to acclimate to poor Ca environment is one of the keys to understanding the long-term ecological effect of changing atmospheric acid and Ca deposition in the subalpine forest in southwestern China.

Impact of land use types on population structure and extent of bark and foliage harvest of Afzelia africana and Pterocarpus erinaceus in Eastern Burkina Faso

In the West African Sudanian regions, people depend on natural products, especially on highly valued species as source of income, fuel wood, food, medicine, fodder for livestock etc. However, land-use management coupled with unsustainable uses of highly valued trees might jeopardize the long-term viability of some species’ populations. Thus, we compared the population structures of two trees, Afzelia africana and Pterocarpus erinaceus and the extent of bark and foliage harvesting within two contrasting land-use types using a random stratified design with 45 replication plots for each species. For both species, population structures were stable in the protected area whereas they showed a declining structure in the agroforestry parklands with lower densities of seedlings and adults as well as a total lack of saplings and young mature trees. In addition, both specieswere over-exploited. More individuals of A. africana and P. erinaceus were harvested with a weak to severe intensity in the parklands, while only few individuals were harvested in the protected area, with a higher proportion of weak to medium intensity. To ensure conservation of these highly valued species, participatory introduction of juveniles and sensitization for seedling protection are required in the agroforestry parklands. Key words: Population structure, W National Park, agroforestry parklands, pruning, debarking, West Africa

What limits the distribution and abundance of the native conifer Callitris glaucophylla (Cupressaceae) in the West MacDonnell Ranges, central Australia?

The conifer Callitris glaucophylla J.Thompson & L.A.S.Johnson (Cupressaceae) is a fire-sensitive obligate seeder with a heavily fragmented distribution across the Australian continent. We undertook a broad-scale biophysical survey and analysed the population structure of 21 populations in the West MacDonnell Ranges of central Australia. C. glaucophylla had a patchy distribution associated with steep, rocky metamorphic areas with limited evidence of fire. Variation in population structures was clearly related to recent fire history. Nearly half of ‘adult’ C. glaucophylla trees (>5-cm stem diameter) from the sampled stands were dead, with the proportion at individual sites related to evidence of fire. Fire scars were evident on 48% of all live trees we measured. The overall density of live adult trees conformed to a negative exponential size-class distribution, consistent with a regionally stable population structure. However, we found higher sapling densities and lower seedlings densities than expected by this distribution. This regional peak in the sapling size class reflects a pulse of recruitment, possibly associated with a wet period in the 1970s. Low seedling densities are probably due to subsequent drought. We conclude that fire controls the distribution of Callitris on the West MacDonnell Ranges, and the timing of recruitment depends on rainfall patterns.

Growth and survival of Abies sachalinensis seedlings for three years after selection harvesting in northern Hokkaido, Japan

Previous studies have indicated that recruitment of Abies sachalinensis, a representative conifer species of northern Japan, decreased following single-tree selection harvesting in stands with dense dwarf bamboo understory. We tested the hypothesis that growth and survival of A. sachalinensis seedlings are reduced by canopy opening in that type of stand. A 0.75 ha study plot was examined, and all the seedlings (defined as trees with height 0.5–2 m) were identified and their shoot extensions measured for three years after single-tree selection harvesting (26% intensity in terms of basal area). The leader extensions of A. sachalinensis seedlings that experienced canopy opening were greatly improved. However, a negative effect on survival was also apparent; nearly 40% of seedlings died at the sites that experienced canopy opening. These results were supported by generalized linear models that examined variations of local harvesting intensity for individual seedlings. Despite the ability of A. sachalinensis to respond rapidly to exposure, some physiological stresses may have appeared, and presumably were amplified by co-occurring dwarf bamboos. With regard to the low seedling density (156 stems ha−1), common in this type of stands, local harvesting intensity in the selection system should be reduced to maintain survival of advanced regeneration, thus sustaining stand structure and composition.

An examination of stocking and early growth in the Warra silvicultural systems trial confirms the importance of a burnt seedbed for vigorous regeneration in Eucalyptus obliqua forest

Clearfelling of wet eucalypt forest followed by high intensity burning and aerial sowing, a silvicultural system designed to mimic the natural dynamic of sporadic regeneration following cataclysmic disturbance, has attracted criticism for not maintaining the structural diversity that is associated with natural disturbance. A silvicultural systems trial was established at the Warra Long-Term Ecological Research site in southern Tasmania to explore alternatives to clearfelling in tall wet eucalypt forest. Stocking, density and growth of the seedling regeneration were monitored for up to 3 years after harvesting and regeneration treatments were applied from 1998 to 2007. The treatments were clearfell with understorey islands, a patchfell, stripfell, dispersed retention, aggregated retention, and single-tree/small-group selection. High intensity burning, low intensity burning and no burning were variously applied as part of these treatments. The nature of the seedbed in each coupe was related to the harvesting and regeneration treatment. Where high intensity burns were applied there was a higher proportion of burnt seedbed available than in coupes where low intensity burns were applied. The highest seedling densities and fastest early seedling growth rates occurred on the hottest burnt seedbeds. The lowest seedling densities occurred on unburnt and undisturbed seedbeds and the slowest early growth rates occurred on unburnt and compacted seedbeds. Treatments that created the most burnt seedbed had the highest seedling densities and the fastest seedling growth. Aggregated retention is considered the most promising alternative to clearfelling. Because high intensity burns as applied to clearfell burn and sow coupes cannot be conducted in aggregated retention coupes as they would probably burn the aggregates, the lower proportion of burnt seedbed will, on average, result in lower seedling density and growth, and may compromise longer term productivity compared to clearfelled and high intensity-burnt coupes. If aggregated retention is to be successfully applied, as measured by the density and height growth of the regeneration, finding ways of successfully and consistently burning such coupes post-harvesting will be essential.

Mountain Birch Seedlings in the Treeline Ecotone, Subarctic Finland: Variation in Above- and Below-Ground Growth Depending on Microtopography

The treeline ecotone in northern Finnish Lapland is characterized by a mosaic of sites with highly varying environmental conditions. Density, age structure, growth, and root systems of mountain birch seedlings (Betula pubescens ssp. czerepanovii [Orlova] Hämet-Ahti) were studied in different microsite classes (deflation, lichen heath, dwarf shrub heath, hummock, willow shrub, sedge mire). On wind-exposed convex topography characterized by shortage of moisture and nutrients, seedling establishment is impeded, as is indicated by low seedling densities and lack of very young seedlings as well as by high rooting depths and root/shoot ratios of the few individuals. In sedge mires, birch seedlings occasionally occur in great numbers but die off at an early stage. Extremely shallow root systems point to anoxia as the main reason. Severe shoot damage is common to almost all saplings. This is likely due to grazing by reindeer (Rangifer t. tarandus). As a result, height growth is suppressed even on sites with otherwise relatively suitable conditions (e.g. willow shrub sites). Future effects of climate change which could locally improve conditions for germination and establishment of birch young growth might be overridden by the effects of high reindeer density.

Regeneration of Pinus halepensis stands after partial cutting in southern France: Impacts of different ground vegetation, soil and logging slash treatments

This study investigated the effects of different soil, ground vegetation and logging slash treatments on Aleppo pine recruitment in a mature stand after partial cut in southern France. An experimental design was assessed using five treatments: (1) chopping of the ground vegetation, (2) chopping followed by soil scarification in one or (3) in two directions, (4) controlled fire, (5) control—and two logging slash treatments: (1) presence or (2) absence of slash cover before the soil and ground vegetation treatments. Pine seedling emergence and survival were recorded on 580 one-square-meter plots at the end of spring and autumn from autumn 2005 to autumn 2007. Description of the soil surface conditions in cover classes was also established in each plot at the end of the year. Seedling emergence took place predominantly in autumn except for the last year of the experiment due to a severe drought period. Soil scarification treatments and controlled fire with slash present (i.e. intense fire) showed the highest seedling densities and bare soil cover was also the highest in these treatments. By contrast, a lower seedling density was recorded in the controlled fire without slash (i.e. moderate fire) where grass cover largely developed. Chopping also proved to be unfavorable to seedling emergence and survival. Seedling emergence was enhanced by the presence of logging slash in the moderate soil scarification treatment probably due to the additional seed source provided by the breakdown of the cones in the slash. The study showed the importance of mechanical treatments – in particular scarification – to favor pine recruitment. However a longer period of data recording is requested to analyze the response of pine seedlings to developing ground vegetation development in the coming years.

Forest Structure Under Different Management Regimes in the Western Lowlands of Nepal

ABSTRACT For landscape level conservation and rural sustenance, forests outside of protected areas are becoming increasingly important. Since most outside forests in the lowlands of Nepal are under government control and utilized as open access resources, their feasibility to achieve these objectives is questionable. We compared forests inside (with three decades of strict protection) and outside of protected areas (under government control and proposed as community forests) to assess the ecological sustainability of outside forests and to provide baseline data on structure of outside forests so that effectiveness of community forestry (after implementation) can be evaluated in the future. Our results indicate that outside forests were in a degraded condition with low tree, sapling and seedling densities and lower species diversity. Trees in lower dbh (diameter at breast height) classes were absent in these forests. Although the canopy layer was not significantly different, the ordination of plots and species showed that species composition of understory and the ground layer were different than protected forests. The present condition of outside forests is not only ecologically unsustainable but also cannot fulfill the demands for forest products of local people. The inefficiency of government to monitor against exploitation has been responsible for present conditions. We recommend that community forestry programs, which are successful in improving conditions of degraded forests in the mountains of Nepal, should be widely implemented in lowland areas (Terai) and proposed community forests should be handed over to local communities as soon as it is practicable.

Delayed Glyphosate Application Doesn't Reduce Plant Density of Glyphosate-Tolerant Alfalfa

Commercial seed of glyphosate-tolerant alfalfa contains small percentages of seeds, called nulls, that do not posses the herbicide trait. Differential competitiveness of glyphosate-tolerant seedlings compared with null seedlings may be detrimental to alfalfa stand density if delayed glyphosate application allows the null and tolerant seedlings to compete with each other. Our objective was to determine if delayed glyphosate application affects plant density of glyphosate-tolerant alfalfa stands. Commercially available glyphosate-tolerant alfalfa was seeded at 17 and 29 kg/ha, plots were treated with glyphosate one month (3- to 6-trifoliate leaf stage) or one-year after planting and plant density was monitored throughout the study. Regardless of seedling rate an average of 6.5 and 5.5% of the alfalfa plants died when glyphosate was applied one month and one year after planting, respectively, but plant density remained unchanged during the same time periods when glyphosate was not applied. Planting alfalfa at 17 kg/ha resulted in lower seedling density, but also lower plant mortality during the seeding year than planting at 29 kg/ha. We conclude that delaying glyphosate application to glyphosate-tolerant alfalfa until one year after planting does not alter the percent of glyphosate-tolerant and non-tolerant plants.

Seedling dynamics of Acer mono and Fagus crenata: an environmental filter limiting their adult distributions

The difference in distribution of tree populations between riparian sites and terraces might be explained by differential limitation of cohort establishment in the early stages of a tree’s life history. To test this hypothesis, the early population dynamics of the riparian specialist Acer mono Maxim. var. marmoratum Hara f. dissectum Rehder and the generalist Fagus crenata Blume were compared at a riparian site and an adjacent terrace in a temperate forest in northern Japan. Survival and growth of seedling cohorts were analyzed for seven years (1994–2000) by studying the topography, light conditions, and initial density of conspecific seedlings. At one year after germination, A. mono cohorts survived best at the riparian site with a lower seedling density (40%), but were less successful at the terrace (18%) and at the riparian site with a higher seedling density (25%). Survival of F. crenata cohorts was not affected by environmental factors (ca. 14% one year after germination). The critical difference in A. mono survival occurred during the first growing season; cohorts at the riparian site experienced higher mortality (0.02 day–1) before July, whereas those at the terrace decreased rapidly in August (0.03 day–1) when there was little rainfall. The height of seedlings of both species increased with age (becoming ≥ 20 cm during the study period), except for A. mono seedlings at the terrace which remained at the same height as at emergence for seven years (ca. 6 cm). Interspecific comparison revealed that A. mono survived better than F. crenata at the riparian site although their growth was equivalent, and that F. crenata grew better than A. mono on the terrace although their survival rates were equivalent. The seedling performance of the two species corresponds well with the adult distribution in the forest.

Seedling Reproduction Established with Soil Scarification within an Oak Overwood after Overstory Removal

Abstract We report on the density and height of seedling reproduction established with and without soil scarification in a mixed-oak two-stage shelterwood after overstory removal. In 1993, shallow soil scarification was conducted to bury recently disseminated northern red oak (Quercus rubra L.) acorns collected from another site and disrupt red maple (Acer rubrum L.) seedling reproduction that dominated the understory. The shelterwood overstory was removed in the winter of 1997/1998. The following spring, seedling density decreased in response to harvest activities. However, one growing season after overstory removal, seedling density for most species or species groups had recovered to preharvest levels. At that time (5 years after soil scarification), scarified plots had 16 times more northern red oak (35,119 ha−1), 50% fewer (8,076 ha−1) red maple, and 37% more total (55,432 ha−1) seedlings compared to undisturbed control plots. Other species of oak, which generally lacked mast at the time of scarification, had significantly reduced density (59% less, at 4,405 ha−1), suggesting that scarification treatments should coincide with viable mast or net loss of oak seedlings may result. Five growing seasons after treatment, the greatest proportions of seedlings were in the tallest two height classes (33–122 cm and >122 cm) and were northern red oak. In summary, 5 years after incorporating viable mast into the soil with shallow scarification within a mixed-oak shelterwood and 1 year after overstory removal, scarified areas had more advance reproduction dominated by large and vigorously growing oak seedlings, while undisturbed areas had lower seedling density of primarily large red maple.

Effects of flow pattern on riparian seedling recruitment on sandbars in the Wisconsin River, Wisconsin, USA

In riparian ecosystems, river flow is the dominant driver influencing ecological process and pattern, including the recruitment of riparian tree species. In a four-year field study (1997–2000) of seedling recruitment on sandbars in the Wisconsin River, I evaluated the hypothesis that the timing of seed dispersal and river flow interact to determine the annual composition of pioneer tree seedling cohorts. In the final three years of the study (1998–2000), growing season flow pulses strongly influenced the species composition and density of new seedling cohorts. Mortality rates of new seedlings exceeded 90% during the 1998 and 1999 flow pulses, and species composition shifted to dominance by Acer saccharinum following the 1999 flow pulse. Different species dominated the cohort of each year: in 1997, 83% of new seedlings were Betula nigra; in 1998, 63% were Salix spp. (S. nigra 39%, S. exigua ssp. interior 24%); in 1999, 89% were Acer saccharinum; and in 2000, 71% were S. exigua and 23% were Betula nigra. Cohorts retained their initial differences in species composition for at least one year after establishment. The three years (1998–2000) with summer flow pulses produced very low end-of-the-growing-season densities of new seedlings (<0.1 seedlings/m2 on random plots), and subsequent mortality in the first year after establishment further reduced seedling numbers by 53–94% on monitored plots. Thus, although summer flow patterns during the first year of growth strongly influenced species composition of seedling cohorts, low seedling densities and high overwinter mortality likely reduce the long-term impacts of initial cohort differences on successional trajectories.

河川氾濫原における土壌シードバンクの分布特性と水流の影響

In order to clarify the distribution pattern of soil seed bank and dispersion of seeds by river water flow, germination test were made by applying the surface soil collected from four sites (main channel side, bank marsh, remnant channel, bank lowland) in floodplain of River Arakawa. The abundance of seed banks in back marsh site is relatively less than any other sites. The seed banks of main channel side site had high seedling density of the species that might be dispersed by water (hydrochory) such as Veronica anagallis-aquatica, Cyperus glomeratus, which did not have special feature adapt to water dispersal. In contrast, the seed banks of the back marsh site apart from main channel had low seedling density of those species. Therefore, the difference of seedling density of the hydrochorous species can be explained by the distance from main river channel. The similarity between the floristic composition of actual vegetation and seed bank was higher in main channel side site than any other sites. It might be reflected that high disturbance by flood had been occurred in main channel side. The spatial heterogeneity of seed bank composition in lowland riparian habitat can be partially attributed to water dispersal.

Effect of Nitrogen Topdressing at the Early Growth Stage on Tillering and Yield of Direct Seeded Rice Cultivar Dontokoi under Low Seedling Density

Effect of Nitrogen Topdressing at the Early Growth Stage on Tillering and Yield of Direct Seeded Rice Cultivar Dontokoi under Low Seedling Density

Flowering, seed dispersal, seed predation and seedling recruitment in two pyrogenic flowering resprouters

A few resprouting plants in fire-prone environments have no local seed bank (soil or canopy) when a fire occurs. These species rely on post-fire flowering and the production of non-dormant seeds to exploit favourable post-fire establishment and growth conditions. For two such pyrogenic flowering species (Doryanthes excelsa Correa and Telopea speciosissima (Smith) R.Br.), we examined the timing of seed release, patterns of fruit production, seed dispersal, seed predation and seedling establishment following a fire in the Sydney region of south-eastern Australia. Both species took some 19 months after the fire to flower and the first seeds were released 2 years after the fire. D. excelsa flowered and fruited only once after the fire. For T. speciosissima, plants also flowered at least once more in the subsequent 5 years, but produced seed in only the first three post-fire flowering years. Fruit production differed between species, with fruiting individuals of D. excelsa producing fewer infructescences, similar numbers of follicles, but many more seeds per follicle than fruiting individuals of T. speciosissima. Ultimately, D. excelsa produced approximately six times as many seeds per m2 and four times as many seeds per adult in one flowering season than T. speciosissima did after four flowering (three successful fruiting) seasons. Seeds were passively dispersed from fruits borne 3–4 m (D. excelsa) or 1–2 m (T. speciosissima) above the ground. Most seeds were found within 5 m (D. excelsa) or 3 m (T. speciosissima) of parent plants. The primary seed shadow of both species was a poor predictor of the distribution of seedlings, with more seedlings occurring further from the adults than expected from the distribution of seeds. Seed loss to predators was high in both species in exclusion experiments where mammals had access to clumps of seeds (77–88%). It was variable and generally lower (8–65%) in experiments where seeds were not locally clumped. However, for T. speciosissima, at one site, some 65% of seeds were lost to mammals and invertebrates in these latter experiments. At this site, these losses appeared to influence subsequent recruitment levels, as very low seedling densities were observed. For both species, germination of seedlings first occurred some 2.5–3 years after the passage of the fire. The percentage of seeds produced to seedlings successfully established was low in D. excelsa (2–3%) and more variable across sites and years in T. speciosissima (0–18%). Resultant post-fire seedling densities of D. excelsa (two sites) and T. speciosissima at one site were similar, but they were much lower at the T. speciosissima site that had high levels of seed predation. Both D. excelsa and T. speciosissima are amongst the slowest woody resprouting species to recruit seedlings after fire in south-eastern Australia and lag years behind species with soil or canopy seed banks.

Density dependence and population differentiation of genetic architecture in Impatiens capensis in natural environments.

We identified environment-dependent constraints on the evolution of plasticity to density under natural conditions in two natural populations of Impatiens capensis. We also examined the expression of population divergence in genetic variance-covariance matrices in these natural environments. Inbred lines, originally collected from a sunny site with high seedling densities and a woodland site with low seedling densities, were planted in both original sites at natural high densities and at low density. Morphological and life-history characters were measured. More genetic variation for plastic responses to density was expressed in the sun site than in the woodland site, so the evolutionary potential of plasticity was greater in the sun site. Strong genetic correlations between the same character expressed at different densities and correlations among different characters could constrain the evolution of plasticity in both sites. Genetically based trade-offs in meristem allocation to vegetative growth and reproduction were apparent only in the high-resource environment with no overhead canopy and no intraspecific competition. Therefore, genetic constraints on the evolution of plasticity depended on the site and density in which plants were grown, and correlated responses to selection on plastic characters are also expected to differ between sites and densities. Population differentiation in genetic variance-covariance matrices was detected, but matrix structural differences, as opposed to proportional differences, were detected between populations only in the sun site at natural high density. Thus, population divergence in genetic architecture can occur rapidly and on a fine spatial scale, but the expression of such divergence may depend on the environment.

Rapid photosynthetic acclimation of Shorea johorensis seedlings after logging disturbance in Central Kalimantan.

This study examined the photosynthetic acclimation of pre-existing Shorea johorensis (Dipterocarpaceae) seedlings to the change in conditions that occurs at the time of logging in Central Kalimantan, Indonesia. The hypothesis was that the seedlings would be unable to acclimate beyond partially open conditions after canopy disturbance caused by logging, therefore limiting the potential for regeneration in the most open areas. Bleaching and reductions in the predawn ratio of variable to maximum fluorescence (F v /F m) indicated chronic photoinhibition and damage to the previously shade-adapted leaves of seedlings in an area logged 2 weeks earlier. The majority of seedlings in partially open and open environments of an area logged 3 months earlier were already growing fast. Leaves that had developed in the new environment showed only small reductions in predawn F v /F m and large increases in the light saturated rate of photosynthesis (A max) per unit area when compared to shaded seedlings. Leaves in the most open environments had higher but more variable nitrogen concentrations, A max per unit area and A max per unit mass when compared to seedlings in partially open environments. Increases in dark respiration were disproportionately large compared to increases in A max, and may have been the result of increased investment in photoprotective mechanisms. The response of stomatal conductance to the vapour pressure deficit and leaf temperature was examined, but it suggested only a 10% reduction in daily leaf level carbon gain in open environments. The ratio of leaf area to fine root mass was highest in shade-suppressed and newly exposed seedlings, suggesting a potential hydraulic limitation to transpiration during acclimation. However, rainfall during this period was high and leaf water potentials did not differ between disturbed and undisturbed environments. S. johorensis seedlings were capable of significant acclimation to conditions more extreme than partial canopy opening. Low seedling density after logging during the wet season cannot be explained by a limited potential for photosynthetic acclimation.

THE FERN UNDERSTORY AS AN ECOLOGICAL FILTER: GROWTH AND SURVIVAL OF CANOPY-TREE SEEDLINGS

We investigated the role of the fern understory of closed-canopy forests as an ecological filter shaping the density, species composition, size structure, and spatial distribution of the seedling bank. In New England deciduous forests we tested the hypothesis that the understory stratum is a selective filter that differentially influences growth and survival of tree-seedling species by comparing performance of Acer rubrum, Betula alleghaniensis, and Quercus rubra seedlings in plots where the fern understory was undisturbed, partially removed, or completely removed. We related seedling growth and survival to microenvironmental characteristics of experimental plots in order to further explore mechanisms responsible for the filtering capacity of the fern understory. The presence of a fern understory reduced growth and survival of all seedling species, but the magnitude of the effect differed among species. Mortality resulting from resource limitation in Quercus below the fern understory was balanced by mortality resulting from insect herbivory in fern-free areas. Relative biomass growth rates of all species were negatively influenced by the presence of fern cover, whereas relative height growth rates of Acer and Quercus were uniformly low and insensitive to the presence of fern cover. Growth and survival rates indicate that only Quercus seedlings can emerge from the fern stratum in the absence of understory or overstory canopy disturbance. A trade-off between persistence in low light and maximum growth in understory light levels was observed among species. The relative growth rate of Betula in terms of biomass and height was more responsive to light levels than were relative growth rates of Acer or Quercus, and the growth rate of Betula was higher than that of Acer and Quercus in all light levels. However, survival of Betula below the fern stratum was lower than survival of Acer and Quercus. The fern understory has the capacity to selectively filter tree seedlings as they grow up through it because seedling species respond differentially to the presence of fern cover. The selective filtering of tree seedlings by the fern understory results in a seedling spatial structure that reflects the spatial heterogeneity of the fern stratum. The seedling pool below the fern stratum has a lower seedling density and different species composition and size structure than the seedling pool in fern-free areas.

The Fern Understory as an Ecological Filter: Growth and Survival of Canopy-Tree Seedlings

We investigated the role of the fern understory of closed-canopy forests as an ecological filter shaping the density, species composition, size structure, and spatial distribution of the seedling bank. In New England deciduous forests we tested the hypothesis that the understory stratum is a selective filter that differentially influences growth and survival of tree-seedling species by comparing performance of Acer rubrum, Betula alleghaniensis, and Quercus rubra seedlings in plots where the fern understory was undisturbed, partially removed, or completely removed. We related seedling growth and survival to microenvironmental characteristics of experimental plots in order to further explore mechanisms responsible for the filtering capacity of the fern understory. The presence of a fern understory reduced growth and survival of all seedling species, but the magnitude of the effect differed among species. Mortality resulting from resource limitation in Quercus below the fern understory was balanced by mortality resulting from insect herbivory in fern-free areas. Relative biomass growth rates of all species were negatively influenced by the presence of fern cover, whereas relative height growth rates of Acer and Quercus were uniformly low and insensitive to the presence of fern cover. Growth and survival rates indicate that only Quercus seedlings can emerge from the fern stratum in the absence of understory or overstory canopy disturbance. A trade-off between persistence in low light and maximum growth in understory light levels was observed among species. The relative growth rate of Betula in terms of biomass and height was more responsive to light levels than were relative growth rates of Acer or Quercus, and the growth rate of Betula was higher than that of Acer and Quercus in all light levels. However, survival of Betula below the fern stratum was lower than survival of Acer and Quercus. The fern understory has the capacity to selectively filter tree seedlings as they grow up through it because seedling species respond differentially to the presence of fern cover. The selective filtering of tree seedlings by the fern understory results in a seedling spatial structure that reflects the spatial heterogeneity of the fern stratum. The seedling pool below the fern stratum has a lower seedling density and different species composition and size structure than the seedling pool in fern-free areas.