Search

• This study was conducted to determine whether elevated CO

Despite predictions that both atmospheric CO

We measured the effect of elevated CO

The response of temperate forest ecosystems to elevated atmospheric CO

The savannas (cerrado) of south-central Brazil are currently subjected to frequent anthropogenic burning, causing widespread reduction in tree density. Increasing concentrations of atmospheric CO

Summary 1. Plants can provision seeds by optimizing seed size, number and nutrient content to maximize parental fitness. According to the McGinley‐Charnov hypothesis, seed size should be determined by the ratio of carbon to nitrogen (C/N) available to the plant, with larger seed size correlating with larger C/N ratios and smaller absolute N content. 2. This hypothesis was tested by establishing monocultures of Phytolacca americana L. (Phytolaccaceae) at three population densities under ambient and elevated CO 2 environments, with two availabilities of soil N. 3. Elevated CO 2 reduced both seed size and N concentration while increasing the C/N ratio; high soil N availability produced the opposite result for N concentration and C/N ratio. Higher planting densities reduced plant biomass, but did not alter seed size. 4. In accordance with the McGinley‐Charnov hypothesis, larger seeds had both larger C/N ratios and smaller N content. However, the increase in C/N ratio caused by elevated CO 2 corresponded with smaller seeds overall: elevated CO 2 reduced seed size, although the seed size‐C/N relationship remained positive. 5. These results suggest an alternative mechanism to explain variation in seed size, and suggest that future climate conditions may alter seed quality and plant reproductive behaviour.

� Springer-Verlag 2005 Abstract Tree-ring characteristics in four species were ex- amined to address whether co-occurring mature trees of dif- ferent successional status respond differently to drought, and whether saplings of these species have a greater re- sponse to drought than mature trees. We examined saplings and mature trees of paper birch, yellow birch, red maple and sugar maple, which varied in successional status (shade- tolerance) and co-occurred at Harvard Forest, Petersham, Mass., USA. Three drought events in 1964-1966, 1981 and 1995 were identified using climate data. For mature trees, there was no significant interspecific difference in relative changes in ring-width index (RWI) during the 1964-1966 and 1995 drought events. However, the interspecific differ- ence was significant in the 1981 drought event. Response function analysis for mature trees showed that the radial growth of sugar maple was mainly controlled by spring and summer precipitation, red maple by spring and sum- mer precipitation and temperature, yellow birch by winter and summer precipitation, and spring and summer tem- perature, and paper birch by spring and summer precipita- tion and spring temperature. Saplings of sugar maple and yellow birch, but not red maple and paper birch, showed significant positive correlations between RWI and annual total precipitation. In the 1995 drought event, saplings and

To address the role of canopy-seedling feedbacks in the structure and dynamics of mixed conifer broad-leaved forests in the eastern US, we monitored seedling regeneration patterns and environmental conditions in the understorey of stands dominated by either hemlock (Tsuga canadensis) or red oak (Quercus rubra) for three years. Hemlock seedlings were favoured over other species' seedlings in hemlock stands (a true positive feedback), due to a combination of high seed inputs, high seedling emergence and relatively high seedling survival during the growing season, which allowed hemlock to remain dominant under its own canopy. Red oak stands favoured a suite of mid-successional broad-leaved species over hemlock. A more even age structure of broad-leaved species in red oak stands revealed that high seedling survival in such stands were driving this feedback. Canopy-mediated variations in both understorey light availability (1.5% for hemlock vs 3.5% for red oak) and soil pH (3.9 for hemlock vs 4.4 for red oak) were found to be the primary correlates of stand-level differences in seedling regeneration dynamics. In mixed temperate forests in the eastern US, canopy-seedling feedbacks could act to slow successional trajectories and contribute to the maintenance of a stable landscape structure over many generations.

We used sap flow as a measure of whole-tree function to examine how coniferous and broad-leaved species in mixed temperate forests differ in canopy-level transpiration and photosynthetic rates. We used heat dissipation probes to measure whole-tree sap flow in three species throughout one full year and then combined these measurements with micrometeorological monitoring and leaf-level gas exchange to determine whole-tree carbon gain. Both broad-leaved species (red oak, Quercus rubra L.; red maple, Acer rubrum L.) had two- to four-fold greater annual fluxes of water and carbon on a ground area basis than did the conifer (eastern hemlock, Tsuga canadensis (L.) Carrière), with red oak trees additionally showing 60–80% higher fluxes than red maple. Despite fixing one-third of its carbon when broad-leaved species were leafless, hemlock was not able to compensate for its low photosynthetic rates during the growing season. Productivity measures derived from annual growth rings and eddy covariance confirmed that whole-tree sap flow provided a valuable estimate of both the magnitude of current forest fluxes and differences in individual species' fluxes. Our results indicate that the predicted loss of hemlock from mixed temperate forests could potentially increase whole-forest water loss and carbon gain by two- to four-fold, provided sufficient nitrogen and water remain available to support such a change.

Summary We investigated how light and CO2 levels interact to influence growth, phenology, and the physiological processes involved in leaf senescence in red oak (Quercus rubra) seedlings. We grew plants in high and low light and in elevated and ambient CO2. At the end of three years of growth, shade plants showed greater biomass enhancement under elevated CO2 than sun plants. We attribute this difference to an increase in leaf area ratio (LAR) in shade plants relative to sun plants, as well as to an ontogenetic effect: as plants increased in size, the LAR declined concomitant with a decline in biomass enhancement under elevated CO2 Elevated CO2 prolonged the carbon gain capacity of shade-grown plants during autumnal senescence, thus increasing their functional leaf lifespan. The prolongation of carbon assimilation, however, did not account for the increased growth enhancement in shade plants under elevated CO2. Elevated CO2 did not significantly alter leaf phenology. Nitrogen concentrations in both green and senesced leaves were lower under elevated CO2 and declined more rapidly in sun leaves than in shade leaves. Similar to nitrogen concentration, the initial slope of A/Ci curves indicated that Rubisco activity declined more rapidly in sun plants than in shade plants, particularly under elevated CO2. Absolute levels of chlorophyll were affected by the interaction of CO2 and light, and chlorophyll content declined to a minimal level in sun plants sooner than in shade plants. These declines in N concentration, in the initial slope of A/Ci curves, and in chlorophyll content were consistent with declining photosynthesis, such that elevated CO2 accelerated senescence in sun plants and prolonged leaf function in shade plants. These results have implications for the carbon economy of seedlings and the regeneration of red oak under global change conditions.

Changes in forest species composition could influence ecosystem carbon uptake rates. To understand how species differed in their contributions to canopy photosynthesis, we investigated how the dominant coniferous (eastern hemlock, Tsuga canadensis (L.) Carr.) and broad-leaved (northern red oak, Quercus rubra L.; red maple, Acer rubrum L.) species in a central Massachusetts forest differed in canopy carbon uptake rates. We considered what factors influenced in situ leaf-level photosynthesis and then used a bottom-up summation approach to estimate species-specific total canopy carbon uptake rates. Variation in canopy light strongly influenced leaf-level photosynthetic rates: sunlit leaves had significantly higher rates than shaded leaves, and photosynthesis increased with canopy height. Species also differed in leaf-level photosynthetic rates, with the broad-leaved species having up to twofold higher rates than hemlock. Within hemlock, needles older than 2 years had lower photosynthesis than younger needles. Variation in leaf-level photosynthesis scaled up to influence canopy carbon uptake rates. Red oak consistently had the highest canopy photosynthetic rates, while through the season, hemlock's relative contribution to carbon flux increased and that of red maple decreased. Thus, in such mixed forests, future changes in species composition could have substantial impacts on forest carbon dynamics, particularly if red oak is the primary broad-leaved species to expand at the expense of hemlock.

Rising levels of atmospheric CO2 may alter patterns of plant biomass production. These changes will be dependent on the ability of plants to acquire sufficient nutrients to maintain enhanced growth. Species-specific differences in responsiveness to CO2 may lead to changes in plant community composition and biodiversity. Differences in species-level growth responses to CO2 may be, in a large part, driven by differences in the ability to acquire nutrients. To understand the mechanisms of how elevated CO2 leads to changes in community-level productivity, we need to study the growth responses and patterns of nutrient acquisition for each of the species that comprise the community. In this paper, we present a study of how elevated CO2 affects community-level and species-level patterns of nitrogen uptake and biomass production. As an experimental system we use experimental communities of 11 co-occurring annuals common to disturbed seasonal grasslands in south-western U.S.A. We established experimental communities with approximately even numbers of each species in three different atmospheric CO2 concentrations (375, 550, and 700 ppm). We maintained these communities for 1, 1.5, and 2 months at which times we applied a 15N tracer (15NH415NO3) to quantify the nitrogen uptake and then measured plant biomass, nitrogen content, and nitrogen uptake rates for the entire communities as well as for each species. Overall, community-level responses to elevated CO2 were consistent with the majority of other studies of individual- and multispecies assemblages, where elevated CO2 leads to enhanced biomass production early on, but this enhancement declines through time. In contrast, the responses of the individual species within the communities was highly variable, showing the full range of responses from positive to negative. Due to the large variation in size between the different species, community-level responses were generally determined by the responses of only one or a few species. Thus, while several of the smaller species showed trends of increased biomass and nitrogen uptake in elevated CO2 at the end of the experiment, community-level patterns showed a decrease in these parameters due to the significant reduction in biomass and nitrogen content in the single largest species. The relationship between enhancement of nitrogen uptake and biomass production in elevated CO2 was highly significant for both 550 ppm and 700 ppm CO2. This relationship strongly suggests that the ability of plants to increase nitrogen uptake (through changes in physiology, morphology, architecture, or mycorrhizal symbionts) may be an important determinant of which species in a community will be able to respond to increased CO2 levels with increased biomass production. The fact that the most dominant species within the community showed reduced enhancement and the smaller species showed increased enhancement suggest that through time, elevated CO2 may lead to significant changes in community composition. At the community level, nitrogen uptake rates relative to plant nitrogen content were invariable between the three different CO2 levels at each harvest. This was in contrast to significant reductions in total plant nitrogen uptake and nitrogen uptake relative to total plant biomass. These patterns support the hypothesis that plant nitrogen uptake is largely regulated by physiological activity, assuming that physiological activity is controlled by nitrogen content and thus protein and enzyme content.

The response of temperate forest ecosystems to elevated atmospheric CO2 concentrations is important because these ecosystems represent a significant component of the global carbon cycle. Two important but not well understood processes which elevated CO2 may substantially alter in these systems are regeneration and nitrogen cycling. If elevated CO2 leads to changes in species composition in regenerating forest communities then the structure and function of these ecosystems may be affected. In most temperate forests, nitrogen appears to be a limiting nutrient. If elevated CO2 leads to reductions in nitrogen cycling through increased sequestration of nitrogen in plant biomass or reductions in mineralization rates, long-term forest productivity may be constrained. To study these processes, we established mesocosms of regenerating forest communities in controlled environments maintained at either ambient (375 ppm) or elevated (700 ppm) CO2 concentrations. Mesocosms were constructed from intact monoliths of organic forest soil. We maintained these mesocosms for 2 years without any external inputs of nitrogen and allowed the plants naturally present as seeds and rhizomes to regenerate. We used 15N pool dilution techniques to quantify nitrogen fluxes within the mesocosms at the end of the 2 years. Elevated atmospheric CO2 concentration significantly affected a number of plant and soil processes in the experimental regenerating forest mesocosms. These changes included increases in total plant biomass production, plant C/N ratios, ectomycorrhizal colonization of tree fine roots, changes in tree fine root architecture, and decreases in plant NH4+ uptake rates, gross NH4+ mineralization rates, and gross NH4+ consumption rates. In addition, there was a shift in the relative biomass contribution of the two dominant regenerating tree species; the proportion of total biomass contributed by white birch (Betula papyrifera) decreased and the proportion of total biomass contributed by yellow birch (B. alleghaniensis) increased. However, elevated CO2 had no significant effect on the total amount of nitrogen in plant and soil microbial biomass. In this study we observed a suite of effects due to elevated CO2, some of which could lead to increases in potential long term growth responses to elevated CO2, other to decreases. The reduced plant NH4+ uptake rates we observed are consistent with reduced NH4+ availability due to reduced gross mineralization rates. Reduced NH4+ mineralization rates are consistent with the increases in C/N ratios we observed for leaf and fine root material. Together, these data suggest the positive increases in plant root architectural parameters and mycorrhizal colonization may not be as important as the potential negative effects of reduced nitrogen availability through decreased decomposition rates in a future atmosphere with elevated CO2.

The impact of elevated atmospheric CO2 on belowground plant growth is poorly understood relative to its effects on aboveground growth. We carried out a study of the seasonal dynamics of gross root production and death to determine how elevated CO2 affected the dynamics of net and gross root production through a full growing season. We quantified gross root production and root loss from sequential, in situ images of fine roots of t Betula papyrifera in ambient (375 ppm.) and elevated (700 ppm) CO2 atmospheres from 2 weeks following germination through leaf senescence. We found that elevated CO2 led to increases in the magnitude of cumulative gross production (ΣP) and cumulative gross loss (ΣL) of roots. However, the effect of elevated CO2 on these processes was seasonally dependent. Elevated CO2 led to greater levels of enhancement in ΣP early in the growing season, prior to maximum standing root length (NP). In contrast, elevated CO2 led to greater levels of enhancement in ΣL in the last half of the growing season, after maximum NP had been reached. This difference in the timing of when elevated CO2 affects ΣP and ΣL led to a transitory, early enhancement in NP. By the end of the growing season, there was no significant effect of elevated CO2 on NP, and ΣP was 87% greater than NP for ambient CO2 and 117% greater in elevated CO2. We conclude that static assessments of belowground productivity may greatly underestimate gross fine root productivity and turnover and this bias can be exaggerated with elevated CO2.

We examined growth and allocation responses to CO2 enrichment for three species of co-occurring temperate forest tree seedlings grown in pots of varying rooting volumes and nutrient supply. Under both current and projected future CO2 atmospheres, tree seedling growth was substantially greater with greater total nutrient supply (either due to increased nutrient addition rate or increased rooting volume) for all species. Increasing rooting volume alone, holding total nutrient supply constant, increased growth for gray and yellow birch and decreased growth for red maple. Root/shoot ratios were less and specific leaf masses were greater for plants grown in smaller pots, suggesting that the smaller pots did restrict root growth with consequences for whole-plant carbon allocation. After 12 wk of growth at light levels simulating those found in small gaps in temperate forests, each species exhibited growth, allocational and/or architectural differences due to increased CO2. Of 11 traits measured, 9 were significantly altered by CO2 regime. Gray birch responded in architectural and allocational parameters only; total carbon accumulation after 12 wk of growth was not affected by CO2 regime. Red maple and especially yellow birch grew larger in elevated CO2, and were less responsive in architectural and allocational parameters than gray birch. Increasing N concentration did not increase CO2-induced growth enhancements, except for increased leaf production in gray birch. In fact, C02-induced increases in branch production were greatest at low nutrient concentration. Pot size had no effect on CO2- induced growth responses, except that CO2-induced enhancement in branch production was greater in smaller pots. With few exceptions, conditions within pots did not influence responses to elevated CO2, despite the many growth and architectural responses manifested by these tree seedlings in response to C02, nutrient regime, and pot size.

This paper explores how the response of a temperate forest ecosystem to climate change might depend on species diversity and community change. In particular, we look at the dynamics of a model of temperate forest growth under doubled CO2. We combine a detailed, field-calibrated model of forest dynamics (Pacala et al 1993) with greenhouse data on the range of seedling biomass growth response to doubled CO2 concentrations (Bazzaz et al. 1990; Bazzaz & Miao 1993). Because total ecosystem response to climate change depends delicately on many environmental variables other than CO2, we isolate the effects of community change by comparing runs of the regular model, allowing dynamic community change, with runs of a reduced model that holds species composition static by using a single tree species with average parameters. Simulations that allowed community change instead of holding species composition constant showed a roughly 30% additional increase in total basal area over time scales of 50-150 years. Although the model omits many possible feedbacks and mechanisms associated with climate change, it suggests the large potential effects that species differences and feedbacks can have in ecosystem models and reinforces the possible importance of diversity to ecosystem function (Naeem et ai 1994; Tilman & Downing 1994) over time scales within the planning horizon for global change policy.

In this paper we present a conceptual model of integrated plant-soil interactions which illustrates the importance of identifying the primary belowground feedbacks, both positive and negative, which can simultaneously affect plant growth responses to elevated CO2. The primary negative feedbacks share the common feature of reducing the amount of nutrients available to plants. These negative feedbacks include increased litter C/N ratios, and therefore reduced mineralization rates, increased immobilization of available nutrients by a larger soil microbial pool, and increased storage of nutrients in plant biomass and detritus due to increases in net primary productivity (NPP). Most of the primary positive feedbacks share the common feature of being plant mediated feedbacks, the only exception being Zak et al.'s hypothesis that increased microbial biomass will be accompanied by increased mineralization rates. Plant nutrient uptake may be increased through alterations in root architecture, physiology, or mycorrhizal symbioses. Further, the increased C/N ratios of plant tissue mean that a given level of NPP can be achieved with a smaller supply of nitrogen.

A plant's competitive ability is determined by its capacity to acquire resources from resource pools shared with neighbors. Resource acquisition abilities depend on plant part placement in relation to available resources. Using a model system of annuals, we examined how the architecture of and allocation to plant parts was affected by different neighbor species. Species having different morphologies were grown as "targets" (target plants) in continuous canopies of each of four neighbor species in the greenhouse. Architectural and allocational traits of Abutilon theophrasti did not vary significantly with neighbor identity. Several architectural and allocation traits of Dutura stramonium and Polygonum penslvanicum plants did vary significantly among neighbor identities; however, in all but two cases, analyses of covariance revealed that these differences were due to differences in shoot biomass among neighbors. We could therefore not relate changes in architecture and allocation to the effects of specific neighbors, and therefore to competitive ability under specific conditions. In all species part masses were more variable than part lengths or areas, suggesting that allocational flexibility in these species allowed them to maximize part size at the expense of part mass to some extent. When we examined within plant biomass allocation patterns among plants of different size classes, regardless of neighbor, we found differences among species in how these patterns changed with plant size. Abutilon and Datura plants in the smallest size class showed a marked reduction in allocation to stems in the upper canopy, while producing similar numbers of nodes, when compared to plants in the largest size class. By contrast, Polygonum plants in the smallest size classes produced many fewer nodes than, but had a relative allocation pattern similar to, plants in the largest size class. These differences had consequences for plant shape, and so suggest differences among species in architectural response to shading. Any such response that allowed a plant to increase its resource—gathering ability would represent an important mechanism of competition.

The incongruency of diurnally varying resources essential to plants may detrimentally affect plants early in their development as indicated by reduced water use efficiency and carbon gain. Typical diurnal patterns of light and CO(2) availability in a midsized temperate herbaceous or forest gap were simulated in specially designed growth chambers. A sinusoidally varying CO(2) treatment (400 ppm minimum, 800 ppm maximum) approximated the diurnal cycle of CO(2) at the soil surface, while a steady-state CO(2) treatment (600 ppm) with the same average CO(2 )concentration provided a control. Crossed with these two CO(2) treatments were two light regimes, one with 3 h of high light (850 μmol·m·s) in the morning (west side of a gap), and the other with 3 h of high light in the afternoon (east side). All treatments received baseline low light (55 μmol·m·s) for 14 h during the day. Rumex crispus was selected as a model species because of its rosette leaves, which grow close to the ground where diurnal CO(2 )variation is greatest. The relative timing of diurnal variations in light and CO(2) significantly affected seedling water use efficiency, carbon gain, and morphology. Total biomass, photosynthetic rates, daily integrated carbon, water use efficiency, and leaf area were enhanced by morning exposure to high light. Seedlings that were exposed to peak values of light and CO(2) incongruently, i.e., those plants receiving intense afternoon light with diurnally varying CO(2), were detrimentally affected relative to control plants receiving intense afternoon light with steady-state CO(2). The results of this experiment indicate that the incongruent availability of required resources-such as light and CO(2)-can detrimentally affect performance relative to when resources are congruent. These contrasting resource regimes can occur on the east and west side of gaps.

Plant competition is a process that occurs through the negative effects that individual plants have on resource availability to neighboring individuals. Therefore com- petition experiments need to examine how different species change resource availability to their neighbors, and how different species respond to these changes-allocationally, ar- chitecturally, and physiologically-through time. In a greenhouse study we used a model system of annuals to examine how canopies of species having differing morphologies differed in their architectures and light-interception abilities, and how different species performed when grown in these canopies. Abutilon theophrasti, Datura stramonium, and Polygonum pensylvanicum were grown as "targets," surrounded by neighbors of a single species. Neigh- bors could be any one of the target species or Setaria faberii. Plants were grown in pots, with one target plant and four neighbor plants. Detailed measurements of neighbor canopy structure and target plant canopy architecture were made at five harvests. Species with different morphologies showed large differences in canopy structure, par- ticularly when grass and forb species were compared. Setaria, a grass, had a more open canopy than the other species (all forbs), and was a consistently weak competitor. Overall, however, the relative effects of different neighbors on target biomass varied with target species. Target biomass was poorly correlated with neighbor biomass and leaf area, but was highly correlated with a measure of target light-interception ability that took into account both target leaf deployment and neighbor light interception. Despite clear differ- ences among neighbor species in canopy structure and effect on light penetration, the results suggest no broad generalizations about the effects of different species as neighbors. Knowl- edge of morphological, physiological, and life history characteristics of both the target and neighbor species may be necessary to explain the results of their competition.

Adaptive phenotypic plasticity is the predicted evolutionary response to fine-grained fluctuation in major environmental factors, such as soil moisture in plant habitats. This study examines genotypes from two natural populations of Polygonum persicaria, one from a relatively homogeneous, moderately moist site, and one from a site in which severe drought and root flooding occur within single growth seasons. Norms of reaction (phenotypic response curves) were determined for a random sample of eight and ten cloned genotypes, respectively, from each of the populations over a controlled moisture gradient ranging from drought to flooding.

Norms of reaction for a number of growth and reproductive characters were determined for 15 randomly sampled Polygonum persicaria genotypes, from two natural populations originating in sites with very different nutrient availabilities. Under severely limiting nutrient conditions, these genotypes shared not only plastic responses such as increased root-to-shoot ratio, but a surprising constancy in such functionally essential characters as leaf area ratio, leaf nitrogen concentration, and propagule nitrogen content. Because functional homeostasis depends on flexibility in underlying characters, similar homeostatic results can be achieved through different combinations of underlying plastic and fixed responses in genetically different entities. For example, plants in each population maintained a relatively constant propagule nitrogen content under extreme low-nitrogen conditions by varying either the size or the tissue nitrogen concentration of propagules. These genotypes also tolerated excessive nutrient levels toxic to many plants, evidently by storing excess nutrients in shoots. Although development was altered under such circumstances, reproductive fitness was maintained.

An important motivation for the study of variability in size and reproductive output in plant populations is its potential relation to natural selection. However, very few data are available to assess the genetic component of fitness-related traits in competing plant populations, or the differential performance of plant genotypes at different densities. To address these issues we conducted an experiment using 25 genotypes of a colonizing herbaceous annual, Polygonum pensylvanicum. These genotypes were randomly sampled from a natural population and cloned by axillary meristern enhancement. Cloned plants were grown in a glasshouse at three densities spanning the range encountered in the natural population (from individually grown to 850 individuals/i2). The growth and fate of a total of 1400 individuals were followed over the course of a 10-wk growing period. Variability in size and reproductive output (as measured by the coefficient of variation of vegetative and reproductive dry mass) increased with density. Early plant size measures were positively correlated with subsequent relative growth rates in dense populations, but not among individually grown plants. These observations indicate the likely importance of asymmetric or "one-sided" competition in the dense populations. The proportion of variance in final size and reproduction explained by genotype was generally higher for individually grown plants than for plants grown under crowded conditions. We suggest that this may result from asymmetric competitive interactions working to amplify early size differences resulting primarily from environmental and developmental "noise." The same genotypes were not superior across all densities. Qualitative ("cross-over") interac- tions for fitness-related characters were observed in comparing genotype performance be- tween the individually grown vs. the low and high density treatments. Genotypes with an early size advantage were predictably favored in dense populations, but the genetic cor- relation between early and final performance was weaker among individually grown plants. In sum, density increased relative variation in fitness correlates such as reproductive bio- mass, but decreased the heritability of these traits. The response of selection is the product of these two opposing forces. Applying our results to some elementary quantitative genetic models suggests that the potential for natural selection would increase with population density, while the potential for genetic drift would decrease. Such patterns may be of particular evolutionary importance in colonizing annuals, whose life histories imply an alternate exposure of genotypes to high and low densities.

This study examined the consequences of differing daily time courses of light availability, while controlling for total photon flux density (PFD, measured in moles per square metre), on the physiology, architecture, and growth of seedlings of two birch species (Betula populifolia and B. alleghaniensis) that differ in shade tolerance. In an experimental garden, seedlings were grown in two sets of contrasting diurnal light regimes, "gaps" and "shadehouses," and at four levels of total integrated PFD (°12, 27, 50, and 70% of full sun.) In gaps, seedlings received relatively more heterogeneous diurnal light regimes, with midday full sun peaks lasting between 40 and 280 min on sunny days, depending on gap size. In shadehouses, seedlings received daily total and average PFDs similar to that in gaps, but received no midday direct sunlight, and an overall more uniform distribution of light. The daily time course of light availability, independent of total PFD, significantly affected growth of seedlings. For both species, seedlings in shadehouses grew significantly larger than gap seedlings. Differences between diurnal treatments were greatest at lower integrated light levels in B. populifolia, but not for the more shade—tolerant B. alleghaniensis. Diurnal light regimes also significantly influenced the degree of plasticity in a seedling's sun—shade responses. However, the magnitude of effect due to diurnal light regime depended on the particular trait investigated, the integrated light level, and the species. At similar total PFDs most characters (e.g., specific leaf mass, leaf mass ratio) exhibited more of a sun—type response in shadehouses that in gaps, however, maximum net photosynthesis and chlorophyll a/b ratios reflected more of a sun—type response in gap light regimes. For most growth, physiological, and morphological characters, seedling responses to increasing total PFD were generally more sensitive in gap than in shadehouse regimes. Our results also show that differences between species were greater when compared along the total PFD gradient in gaps vs. shadehouses. Both these results suggest that experimental studies employing uniform light regimes, which incorporate little of the temporal variability experienced by seedlings in natural gaps, may underestimate plasticity within regenerating tree species, and the potential for species niche differences and coexistence.

We studied how an enriched CO2 atmosphere, in a fully crossed design of light and nutrients, influenced 1 st-yr seedling growth in six New England deciduous forest tree species. The species, in the order of increasing shade tolerance, were gray birch (Betula populifolia), ash (Fraxinus americana L.), red maple (Acer rubrum L.), red oak (Quercus rubra L.), yellow birch (Betula alleghaniensis Britton), and striped maple (Acer pensylvan- icum). Elevated CO2 environments significantly stimulated the seedling growth of all six species. Generally this was more pronounced in low light. The greatest stimulation was found under the condition of low light and high nutrients. However, individual species responded differently to elevated CO2 levels. Among the three early-successional species, gray birch, ash, and red maple, a significant increase in seedling growth under elevated CO2 conditions was found only with high nutrients. The three late-successional species grown under elevated CO2 conditions (red oak, yellow birch, and striped maple) showed a greater percentage increase in seedling growth in low light than in high light. Thus, for the early-successional species, the degree of enhancement of seedling growth by elevated CO2 levels was more sensitive to nutrient levels, while in the late-successional species the enhancement was more sensitive to the level of light. Moreover, species with large seeds (e.g., red oak) exhibited a greater response to elevated CO2 levels under low light than species with small seeds (e.g., gray birch). The results emphasize the importance of plant species as well as other environmental resources in modifying the response of plants to elevated CO,. Considering the light and nutrient environment observed in forest gaps of various sizes, the results of the present experiment suggest seedling regeneration in New England deciduous forests may be altered in a future high CO2 environment.

The objectives of this study were (i) to test the hypothesis that changes in phenology with CO2 are a function of the effect of CO2 upon growth and (ii) to determine if CO2-induced changes in phenology can influence competitive outcome. We examined the effect of 350, 525, and 700 μL∙L−1 CO2 on Guara brachycarpa, Gailardia pulchella, Oenothera laciniata, and Lupinus texensis. Plants were grown as individuals in 150-, 500-, or 1000-mL pots and in competition in 1000-mL pots. Growth and development were monitored at twice-weekly intervals by recording the number of leaves and noting the presence or absence of stem elongation, branching, flower buds, and open flowers. Elevated CO2 affected both growth and phenology, but the direction and magnitude of effects varied with species and soil volume. Elevated CO2 did not appear to affect development through its effect on growth. Those treatments in which there were significant effects of CO2 on growth were generally different from those treatments in which CO2 affected phenology. Rather than affecting phenology by changing plant size, CO2 appeared to affect phenology by modifying the size at which plants switched from one stage to the next. The level of CO2 changed competitive outcome; the importance of Lupinus increased whereas that of Oenothera decreased with increased CO2. These changes were more closely related to the effect of CO2 on growth than its effect on phenology. Key words: time of flowering, size at flowering, competition, photoperiod, rate of development.

Plants of Plantago major were grown for two generations in different en- vironments. Seed germination of the second generation was examined for the effects of maternal environments, as well as subsequent plant growth and reproduction in the presence of inter- and intraspecific competitors. Maternal treatments were (1) plants that had received nutrient pulses in both the first and second generations (PP plants), (2) plants that had received a nutrient pulse only in the first generation (PC plants), (3) plants that had received a nutrient pulse only in the second generation (CP plants), and (4) plants that did not receive a nutrient pulse in either generation (CC plants). Under favorable conditions (with- out interaction with a leaf canopy and root system of grasses), seeds of CC plants generally germinated earlier and had a higher final germination percentage. In contrast, under a Poa canopy, seeds of PP plants germinated earlier and had a higher final germination percentage. For adult plants, there were significant interactions between maternal environment and neighbor density. The effects of maternal environments on leaf area generally were evident before flowering but disappeared after flowering. The greatest difference in leaf area occurred between CC and PP plants, and these were often different from CP and PC plants. There was a significant three-way interaction among the type of competition, maternal treatment, and neighbor density for biomass characters. In intraspecific competition with one neighbor, the spike biomass was highest in PP plants followed by CP plants, PC plants, and CC plants, with the last significantly lower than the first two. In general, the juvenile maternal effect did not necessarily result in an adult reproductive advantage. The influence of the maternal environment differed among plant characters. These results demonstrate that maternal environmental factors can affect plant phe- notypic characters for at least three generations. The expression and the extent of these maternal effects varied with the resource condition experienced by the progeny, the intensity of competitive interactions, and life history stage.

Habitat-selection concepts have rarely been explicitly used for plants, perhaps because the majority of them are immobile. For plants, habitat selection results from evolutionary adjustment of species to environmental factors so that the species functions better in some habitats than in others. Habitat choice refers to the ability of a plant to disperse, in space or time, to preferred patches. Habitat specialization means that a species performs best in a small subset of patches in a given location. The modular structure of plants and their growth response to the patterns of resource availability allows them to occupy large areas, exposing them to spatial and temporal heterogeneity. Modular structure, to a large extent, determines the options available, as well as the constraints, for plants choosing habitats. Choice, however, may be made by the habitat rather than by the plant. That is, the characteristics of the habitat determine which species of the plants that disperse into the habitat become establis...

Plantago major and P. rugelii maternal plants were grown at two background nutrient levels and received nutrient pulses given at different times during their life-span. The influences of these maternal treatments on the reproduction of progeny were examined in four different progeny environments represented by two treatments: with and without competition, and with and without a nutrient pulse. We found that the effects of maternal nutrient pulses on the two species were related to the environments of both parents and progeny. When grown without competition, progeny from parent P. major plants main- tained at the low background nutrient level and given a nutrient pulse had a higher spike biomass and a higher proportional allocation to spikes than did progeny from parents that did not receive a nutrient pulse. In contrast, the progeny of parents that were maintained at the high nutrient level showed no response to a nutrient pulse. In P. rugelii the effect of maternal nutrient pulses on the total and leaf biomass of the progeny was related to both maternal background nutrient levels and nutrient pulse treatments to the progeny. For both species, the timing of maternal nutrient pulses had no significant effect on the characters studied. In competition, for P. major, the progeny from parents given a nutrient pulse during fruit maturation consistently produced a greater amount of spike biomass relative to their neighboring plants from parents that did not receive nutrient pulses, regardless of both the maternal background nutrients and progeny pulse treatments. However, for the progeny from parents given a nutrient pulse during the vegetative and flowering stages, their com- petitive ability relative to their neighbors depended on both the timing of maternal pulse and the progeny pulse treatment. For P. rugelii, the performance (leaf biomass) of the two neighboring progeny depended on maternal background nutrient levels. These results show that maternal nutrient conditions can influence competitive inter- actions in the progeny generation. The effects may have ecological and evolutionary im- plications particularly for initial colonization of infertile sites and for the ability of natural selection to act on phenotypic variation. The strength of these maternal effects depends on both the maternal environment and the progeny environment.

We examined effects of differing levels of physical belowground space on the performance of several colonizing annual species in two experiments. Individual plants were grown in controlled greenhouse conditions under an experimental regime that inde- pendently varied nutrient addition rates and the physical space available for the deployment of roots while maintaining water supply at adequate rates. When equal amounts of nutrients were given, plants grown in greater volumes had greater vegetative growth and often higher reproductive output. Species differed in patterns of response to physical space in both vegetative and reproductive parameters. Abutilon theophrasti given larger soil volumes increased allocation to reproductive tissues relative to vegetative tissues. Setaria faberii, however, responded through a phenological shift: earlier flowering and greater reproductive output in smaller volumes. These results indicate that physical underground space, released in the formation of certain types of gaps, may influence the performance of colonizing annuals beyond providing access to other soil resources, such as nutrients and water.

Seeds of tree species commonly invading old-fields (Fraxinus americana, Gleditsia triacanthos, Morus rubra, Platanus occidentalis, and Prunus serotina) were germinated at eight temperatures from 5 C to 40 C, with six moisture levels (2% to 18% gravimetric moisture content) at each temperature. For most species, total seedling emergence and emergence rat exhibited approximate bivariate Gaussian response surfaces. (...)

Evidence from many sources shows that the concentration of atmospheric CO2 is steadily rising (61, 17). This rise is strongly correlated with the increase in global consumption of fossil fuels (104). There are also significant contributions from the clearing of forests, especially in the tropics (136, 55). Controversy continues, however, as to whether the biosphere is presently a source or a sink for carbon (see 52, 54, 56). Despite this controversy, most scientists agree that rising CO2 levels will have substantial direct and indirect effects on the biosphere (80). Because CO2 is a greenhouse gas, its increase in the atmosphere may influence the earth's energy budget. Several climatologists have used general circulation models to predict changes in mean annual global temperature (58, 108). While these models differ in detail, they all predict increased global warming and substantial shifts in precipitation patterns. Recently, some scientists (60) have questioned the predictions of these models. But regardless of changes in global temperature and other climate variables, rising CO2 can influence world ecosystems by direct effects on plant growth and development. The large body of literature on the response of crops and intensively managed forests to elevated CO2 is not treated in this review because there are

Small—scale gaps are often colonized by more than one plant, yet few individuals. We examined interactions among co—invading plants within small gaps in an old—field community as a function of the size of the gap being invaded. Several colonizing annual species were grown, as single individuals or in conspecific or heterospecific pairs, in artificially created gaps of 10, 20, and 40 cm diameter in a Poa pratensis L. background in two successive years. The identity of the neighbor influenced plant survivorship for one of the two years studied, but had no effect on the probability of becoming reproductively mature, or on levels of seed production. In contrast, vegetative growth was influenced by neighbor identity in both years. Both positive and negative plant—plant interactions were found; the presence of a neighbor did not always result in competitive reduction in the target individual. Conspecific neighbors were found to reduce vegetative biomass to a greater extent than heterospecific neighbors for one of the two years studied, primarily due to the sensitivity of one species, Setaria faberii, to conspecific neighbors. We found no evidence of a consistent competitive hierarchy present in this system. Finally, although increasing gap size had significant, positive effects on survivorship, growth, and reproduction, the size had being colonized had no effect on plant—plant interactions between co—invading individuals. In this system the number and identity of invading individuals, as well as the characteristics of the gap itself, interact to dictate the composition of the invading community and the relative performance of each individual colonist. This complexity may contribute to increased diversity in plant communities in which small—scale disturbances provide opportunities for earlier successional species to persist in localized patches within later successional vegetation.

Sometime near the middle of the next century the human population will have doubled. It will, according to projections, increase from about six billion to almost twelve billion people. We must feed and shelter these additional humans. Additionally, it is expected that, this increase in human population will result in a major increase in the amount of fossil fuel consumption above and beyond the current per capita consumption. Therefore, the emission of carbon dioxide will increase dramatically and may become a much more critical environmental issue than at present (Fig. 1). Currently, a person in the United States of America, on average, uses 22 tons of carbon per year, whereas a person in India, for example, uses only 0.7 tons of carbon per year (Fig. 2). Thus, an individual in the United States consumes almost 40 times as much as an individual in India. Progress, which is the aim of many developing nations will be accomplished only by increased per capita energy consumption. As a result, the carbon output per person, in these developing countries will dramatically increase. It may even double. For example, if we take current emission in Russia (currently between the United States and India) as a standard and assume that all developing countries at the moment, aspire for development, we reach an astonishing fact. The amount of energy used in the amount of carbon dioxide emitted in the atmosphere, would show that the emission of carbon dioxide could go up by a factor of 4 or there about. Apparently, there is no escape from the fact that the concentration of carbon dioxide in the atmosphere will significantly rise and that, in addition to direct effects on plant photosynthesis, growth allocation and phenology, there will be changed in weather patterns. CO2 is predicted to influence primary productivity on the land and the ocean (Fig. 3). Thus, the need to better understand the mechanisms of photosynthesis and photosynthate allocation.

Tropical forest ecosystems are large stores of carbon which supply millions of people with life support requirements. Currently tropical forests are undergoing massive deforestation. Here, I address the possible impact of global change conditions, including elevated CO2, temperature rise, and nitrogen deposition on forest structure and dynamics. Tropical forests may be particularly susceptible to climate change for the following reasons: (1) Phenological events (such as flowering and fruiting) are highly tuned to climatic conditions. Thus a small change in climate can have a major impact on the forest, its biological diversity and its role in the carbon cycle. (2) There are strong revolutionary interactions, such as pollination seed dispersal, with a high degree of specialization, i.e., only certain animals can effect these activities for certain species. Global change can decouple these tight coevolutionary interactions. (3) Because of high species diversity per unit area, species of the tropical rain forest must have narrow niches. Thus changes in global climate can eliminate species and therefore reduce biological diversity. (4) Deforestation and other forms of disturbance may have significant feedback on hydrology both regionally and globally. The predicted decline in the rainfall in the Amazon Basin and the intensification of the Indian monsoon can have a large effect on water availability and floods which are already devastating low-lying areas. It is concluded that tropical forests may be very sensitive to climate change. Under climatic change conditions their structure and function may greatly change, their integrity may be violated and their services to people may be greatly modified. Because they are large stores of great biological diversity, they require immediate study before it is too late. The study requires the collaboration of scientists with a wide range of backgrounds and experiences including biologists, climate modellers, atmospheric scientists, economists, human demographers and sociologists in order to carry out holistic and urgently needed work. Global climatic change brings a great challenge to science and to policy makers.

Plant species differ broadly in their responses to an elevated CO

In assessing the capacity of plants to adapt to rapidly changing global climate, we must elucidate the impacts of elevated carbon dioxide on reproduction, fitness and evolution. We investigated how elevated CO

Chronic nitrogen deposition has the potential to alter seedling shade toler- ance and growth in the temperate forests of northeastern United States, by affecting both the form and the quantity of available nitrogen. Simulated deposition treatments were applied to seedlings of four birch species that co-occur at Harvard Forest (Betula lenta, B. alleghaniensis, B. populifolia, and B. papyrifera). Seedlings were individually potted in forest soil, and grown under light treatments representative of forest understory and treefall gap light levels. In a split-plot design, N was applied at 25 and 50 kg ha- -yr-', as either nitrate, ammonium nitrate, or ammonium, within each light environment. While B. po- pulifolia and B. papyrifera, and B. lenta all showed increased biomass allocation to leaves with increased N, only B. lenta showed a significant growth response to the type of N added, and this response was conditional on rate of N application and light environment. At low light, nitrate-fed B. lenta grew best, and also at low rate of supply, nitrate treatments out-performed ammonium treatments. Greater growth under these conditions is probably the result of higher biomass allocation to leaves, and greater specific leaf area, which increased the leaf area ratio, and improved the capacity for carbon gain. Under N deposition regimes that increase soil nitrate availability, the differences in response of B. lenta and B. alleghaniensis to nitrate at low light may potentially lead to changes in the species com- position of the seedling communities in the understory. When a treefall occurs, a different species mix of seedlings will be released, with potential consequences for sapling and tree species mix.

The effect of increasing atmospheric CO

Previously we examined how limited rooting space and nutrient supply influenced plant growth under elevated atmospheric CO

Recently, it has been suggested that small pots may reduce or eliminate plant responses to enriched CO

Plants often respond to elevated atmospheric CO

Several aspects of genotype-environment interaction may act to modulate natural selection in populations that encounter variable environments. In this study the norms of reaction (phenotypic responses) of 20 cloned genotypes from two natural populations of the annual plant Polygonum persicaria were determined over a broad range of controlled light environments (8%-100% full sun). These data reveal both the extent of functionally adaptive phenotypic plasticity expressed by individual genotypes, and the patterns of diversity among genotypes for characters relevant to fitness, in response to an environmental factor that is both highly variable within populations and critical to growth and reproduction.

To determine the effects of elevated CO

We tested the prediction that plants grown in elevated CO