10th Annual Harvard Forest Summer Research Program

13 August 2002

Living on the Edge: A Closer Look at Within-stem Respiratory Variation in Angiosperms and Conifers

Brent Berger
Ray parenchyma cells can live anywhere from two to 200 years, but does their metabolic activity actually decline with age? Previously, respiration was expressed on a sapwood volume basis, and these measurements were used to scale respiratory rates for entire trees. The focus of this project was to explain whether decreased respiration on a per tissue basis translates into a lowered respiration rate on a per cell basis. Cores were extracted from five species representing three different xylem structures: conifers (Tsuga canadensis (L.) Carr., Pinus strobus L.), ring-porous angiosperms (Quercus rubra L., Fraxinus americana L.) and diffuse-porous angiosperms (Acer rubrum L.). Inner and outer sapwood segments were incubated in vials at 20°C and rates of O2 consumption were measured over 36 hours using a fiber optic oxygen-sensing probe. Segments were then fixed for nuclear staining and quantification of ray volume. Fluorescent nuclear staining (DAPI) and image analysis were used to determine percent live ray volume and percent ray volume, respectively. Although angiosperms had a greater proportion of ray volume and live ray volume, respiratory analyses showed that conifer cells respired at a higher rate than cells of angiosperms, thereby creating little difference on a per tissue basis. Accounting for the amount of living ray tissue generally eliminated the difference in respiration rate between inner and outer sapwood, but Fraxinus americana still showed lower per cell respiration in the inner sapwood. These results suggest that older ray parenchyma cells can respire at the same rate as those of the outer sapwood.

The Effects of Historic Land Uses on Methane Absorption Rates

Amanda Brown
Methane is considered the second most important greenhouse gas produced. Approximately 8-20% of atmospheric methane is consumed by soil micro-organisms. Recent studies suggest land use practices have an effect on soil methane consumption. The objective of this study was to examine the rate of methane oxidation from different land use histories. Five sites were selected: 1) a section of re-grown forest RF, 2) pine stand PS, 3) re-grown farmland PF, 4) active farmland CO, and 5) active cow pasture CW. Six samples per site of A-horizon soil were collected, sieved and 25g were placed in mason jars. Methane oxidation rates were measured at regular intervals over a 48-hour period first at ambient water holding capacity (15-30%). Soils were then dried to simulate severe drought conditions (<1% WHC) and consumption rates measured. Soils were then re-wetted to 50% WHC to observe the recovery of oxidation rates after drought. The RF, PS and PF showed the highest consumption rates at ambient conditions (14.5-21.5 µg kg^-1 hr^-1), while the CW and CO showed no consumption. Under drought conditions all sites showed no methane consumption. Upon re-wetting to 50% WHC the consumption rates for the RF, PS and PF did not return to pre-drought levels (1.7-2.2 µg kg^-1 hr^-1). Land use does affect methane consumption, with older less disturbed forested sites having greater oxidation rates than the active pastures. Drought halts methane consumption, which only slightly returned in forest soils after re-wetting.

Ant Community Response to Hemlock Decline

Brandon Burke
Eastern hemlock (Tsuga canadensis) decline due to infestation by the hemlock woolly adelgid (Adelges tsugae; HWA) is a major concern to ecologists in the eastern United States. The introduced pest alters stand dynamics and the structure and composition of animal communities. Avifauna communities that have been studied were shown to change as the forest declines (Tingley et al. in press). This project focuses on the invertebrate community, specifically ants. Four sites were studied from June to July 2002, comprising two HWA devastated sites in southern Connecticut representing nearly 100 % hemlock mortality, and two uninfested control sites at the Harvard Forest. Ants were quantified by using a combination of sampling techniques, including pitfall traps, baiting, leaf litter extractions, and general searching. A total of 7 ant species from 6 genera were collected from control sites and 6 ant species from 5 genera were collected from HWA devastated sites (Table 1). As predicted, the ant community in control and HWA sites differed significantly in species composition and abundance (Table 2). However, there was a highly significant difference in species composition between the two control sites, Prospect Hill and Slab City, which may suggest that temporal scale or land-use history may play a role in shaping ant assemblages. Higher ant abundance in HWA devastated sites may lead to higher nitrogen cycling as incubated soil cores infested with Aphaenogaster rudis had increased net nitrogen mineralization than cores without A. rudis. Ultimately, patterns of ant species composition in relation to disturbance may be determined by the adaptability of ants to new microhabitat and microclimatological variables associated with successional deciduous species replacing hemlock.

Tingley, M.W., Orwig, D.A., Field, R., and G. Motzkin. 2002. Avian Response to Removal of a Forest Dominant: Consequences of Hemlock Woolly Adelgid Infestations. Journal of Biogeography (in press).

Table 1. Species List for each site. Control sites were Prospect Hill and Slab City. HWA devastated sites were Burnham Brook and Devil's Hopyard.

Table 2. Comparisons of jaccard coefficient values between different sites. 1 = no difference, 0 = absolute.

Examination of Structure, Composition, and Understory Vegetation of Hemlock (Tsuga canadensis) Stands In Massachusetts

Matthew Burr
The hemlock woolly adelgid (HWA) first appeared in the eastern United States in the 1950s and entered southern New England in 1985. Studies in Connecticut, conducted prior to 2002, show that its presence in Connecticut, as elsewhere, has resulted in extensive hemlock mortality. In the summer of 2002, hemlock stands in Massachusetts were identified by aerial photography and sampled to record their structure, composition, understory vegetation, and physical characteristics prior to heavy HWA infestation and further document the spatial and temporal patterns of the HWA in New England. We sampled 30 stands larger than 3 hectares. The average overstory importance value for hemlock was 58 with a range of 24 to 92. The average basal area of sampled stands was 56 m²/ha. Average diameter at breast height was 23.0 cm for hemlocks versus 24.3 cm for other species, while average crown class was 1.5 for hemlocks (1=overtopped, 4=dominant) versus 2.1 for other species. These results indicate hemlock's importance in stand composition despite its smaller size and subordinate crown class compared to other species. Sampling began in the northwest portion of central Massachusetts. HWA occurred in 43% of the sites but exhibited low average density. Average hemlock mortality was only 6.4% with little caused by HWA to date. Hemlock was the most abundant sapling, followed by black birch (Betula lenta). Stand understory vegetation was generally limited with Gaultheria procumbens and Mitchella repens among the more common species.

Hungry Caterpillars and Thirsty Leaves: Examining the Effects of Herbivory on Leaf Hydraulics

Maggie Dietrich
In a leaf, the vasculature delivers water and nutrients to the mesophyll, which captures energy necessary for a plant's survival and growth. Preliminary studies of leaf vasculature have suggested specialized properties, such as the possibility of strong tolerance of vein damage. In general, leaves might be buffered against the disruption of fine veins, such as that caused by insect herbivores. Moreover, it has often been hypothesized that leaves can tolerate disruption of single major veins, and that leaves with different venation architecture (such as palmately vs. pinnately veined leaves) might differ in their tolerance. To test these hypotheses, we experimentally hole-punched and cut veins in leaves of eight New England tree species. Two to four weeks later, we measured stomatal conductance (g) on various locations on each leaf, and we determined the leaf hydraulic conductance (K_leaf) using an in vivo method. For Quercus rubra and Acer rubrum, our preliminary analyses indicate that hole punching had no significant impact on g, either near to or distant from the hole. Hole punching also had no impact on K_leaf . Acer and Quercus differed strongly in their response to the cutting of major veins. In pinnately-veined Quercus, cutting the primary vein (i.e., the midrib) 1 cm from its base led to a dramatic reduction of distal g (by up to 95%). However, in palmately-veined Acer, cutting the middle primary vein 1 cm from its base led to a much smaller decline of distal g (by up to 40%). The hole-punching studies demonstrate that the modular network design of the fine vein system confers immunity to fine vein damage. The vein-cutting studies show that species differ considerably in their tolerance of major vein damage, according to the architecture of their major vein systems.

Hemlock Seedling Bank Dynamics at the Harvard Forest

Jaqueline Annette Guzman
We studied the seedling bank dynamics of eastern hemlock (Tsuga canadensis) at two contrasting sites at the Harvard Forest. The objectives of the study were to document the age structure of the hemlock seedling bank, compare growth rates between sites, and determine whether recruitment into the seedling bank is episodic or continuous. All seedlings < 1.3 m in height within a 25m²-30m² area were collected from two sites at Harvard Forest (n = 70 per site). One site was old hemlock forest and the other was 50-year old mixed hardwood. Height, basal diameter, and age were determined for each seedling. I found that average ages differ between sites. At the hemlock-dominated site it was 30, while at the mixed hardwood it was 14. The oldest individual was 46 years old. Both sites displayed an exponential growth rate, suggesting that seedlings slowly and continuously grow into larger size classes. Analysis of age structure displays that recruitment into the seedling bank is continuous at both sites with peaks. The dynamics of hemlock seedlings support the idea that hemlocks use a seedling bank as a regeneration strategy. Therefore, it would be beneficial to further study hemlock seedling bank dynamics to aid in the success of future restoration projects, since the introduction of the hemlock woolly adelgid poses the threat of wiping out hemlock forests.

A Comparison of CO₂ Release by Soil Following Repeated Drying and Rewetting Events

Jeffrey Law
Production of carbon dioxide in soil is affected by wetting and drying cycles that occur throughout the year in many areas. These cycles result in a pulse of carbon dioxide being released from the soil surface after the wet up. In this study a comparison was made between the O-horizon and the A-horizon to discover how much these layers can potentially contribute to observed pulses of soil respiration. Five samples of A-horizon and three samples of the O-horizon were obtained from a mixed deciduous forest at Harvard Forest in Massachusetts. The samples were kept relatively undisturbed and were intended to act as microcosms of the soil at large. Respiration of each sample was measured at field moisture using an infrared gas analyzer. It was then dried and the flux was measured again. Four different amounts of water were applied to the samples in order to simulate rain events. They consisted of a 1mm, 2mm, 4mm, and 8mm rain simulation. After each wet up, respiration was measured immediately after the water was added and 1, 2, 4, 6, and 24 hours later. The results show that on a per weight basis the O-horizon has a higher overall flux. When the data are converted to a per unit area comparison the O-horizon and A-horizon have similar pulses. Both the O-horizon and the A-horizon can contribute significantly to soil respiration although their relative contribution in the field depends on the extent to which each is wetted under natural conditions.

Using Vegetation and Hydroperiod Data to Explain Patterns in the Distribution of Malacostraca Crustaceans

Kate Musgrove
Vernal pools are isolated, temporary aquatic habitats that support a uniquely characterized community of vertebrates and invertebrates. Many of the organisms in these pools have life cycles that are dependent on the periods of desiccation provided by the regular drying of the pools, as in the fairy shrimp (Eubranchipus sp.), which cannot live in permanent water habitats. Class Malacostraca, including Isopods (Crangonyx sp.) and Amphipods (Caecidotea sp.), presumed to be permanent-water organisms due to a lack of appropriate aestivation or other dormancy life history strategies, have been observed in vernal pools. Batzer and Sion (1999), in an upstate New York study of vernal pools, made observation of malacostracans in pools located in a landscape of undisturbed old growth maple forest, as opposed to pools in a landscape that had been cleared for agriculture or otherwise disturbed. In my study this summer, previously collected data sets (1990-1999) of pools throughout Massachusetts were analyzed to determine the hydroperiod of pools with consistently sampled malacostracans. The pools were then mapped using GIS, and, using a historical vegetation data layer, categorized according to whether or not they were forested in the 1830's. The relationship between Malocostracan presence in short hydroperiod pools with undisturbed 1830's forestation was strong (p value <.005), while the correlation of malacostracan presence as dependent on long or short hydroperiod pools did not prove statistically significant. Further analysis, including exploration of the distribution patterns of obligate vernal pools species, such as fairy shrimp and clam shrimp (Lynceus brachyrus), determined that fairy shrimp occurred frequently, without malacostracans, in historically unforested pools. However, their species distribution could not be statistically related to forestation data, as they were also consistently present in historically forested pools.

Batzer, D., K. Sion. 1999. Autumnal woodland pools of western New York: temporary habitats that support permanent water Invertebrates. Pp. 319-332 In: Darald P. Batzer, Russell B. Rader and Scott A. Wissinger (eds.), Invertebrates in freshwater wetlands of North America. John Wiley and Sons, New York.

Massachusetts Timber Harvesting: A Close Look at the Effects of Socioeconomic and Forest Characteristics

Katy Nicholson
In Massachusetts, timber harvesting happens in greater amounts in the West than in the East, despite high amounts of forest statewide. As the population increases in the West, it is probable that harvest will cease in some towns. This study found that socioeconomic and forest characteristics affect rates of timber harvesting in Massachusetts towns. Data were collected, mapped, and compiled for each of Massachusetts' 351 towns. The dependent variable, percent of a town's forest harvested in the past five years, was derived from Harvard Forest Archives' timber cutting plan records. The study found that the independent variables describing levels of development (percent of town developed and road density) and forest characteristics (core area ratio and percent forest) each explain a significant portion of the variation within the dependent variable (Table 1). Variables describing industry influence, population density, and levels of affluence (education and median income) also explain a significant portion of the variation within the dependent variable, but to a lesser extent. Population density change and the presence of logging laws do not explain a significant portion of variation. More data could be collected to better determine what affects timber-harvesting rates. Modeling the results would be useful in predicting how much timber harvest will happen in towns based on these factors and estimate when the amount harvested will approach zero. The results from studies such as these can provide a better understanding of how human actions impact the harvest potential of forests.

Table 1. Independent variables that most significantly explain variation in the dependent variable (percent of town's forest harvested in past five years) using bivariate linear regression.

The Yearly Nitrogen Budget At Old And New Soil-Warming Sites: A Comparison Of Nitrogen Allocation

Nikki Nowinski
The soil-warming projects at Harvard Forest are assessing the effects of increased temperature on C and N cycling. The 11-year experiment on the Prospect Hill tract utilizes eighteen 6x6m plots, while the new experiment on the Slab City tract utilizes two 30 x 30 m plots. The "mega-plots" will allow the study of the effects of soil-warming on nutrient storage and tree growth. I compiled the N-budget and yearly N storage patterns for the new site, and the yearly N storage patterns for the old one. The N-budget was compiled using woody increment measurements, herb biomass estimates, data from both the old site and Harvard Forest's chronic N addition experiment, and fine-root and understory data from the literature. Annual biomass and N addition by different tree species were calculated for both sites using tree ring measurements, allometric equations, and %N estimates from the literature. The N-budget showed that N distribution in the "mega-plots" was nearly identical. Trees stored additional N annually, however, they became lesser sinks over time. Consequently, annual soil N accumulation increased and was significantly correlated with time (p=.0001, df=9), and not temperature, precipitation, or both. The old site showed similar yearly patterns but 40% less overall production, probably due to stand age differences. Oaks were more productive than other species at the both the old site and the new site. Growth rate comparisons suggest that conclusions from the old experiment may be used to make predictions for the new experiment and the N-budget implies that C and N distribution is similar in the two new plots prior to treatment.

Stand Response to Inundation

Sarah Pears
A red maple stand, recently flooded after the construction of a beaver dam, was examined in order to understand the effect of inundation on stand composition and the local carbon cycle. Tree diameters were measured before and after the inundation in flooded and unflooded stands. The health of each tree was classified as 'dead', 'moribund' (trees not yet dead, but that have not fully leafed out or show other signs of distress), or 'live'. Water levels near the flooded trees were measured after inundation. Water depth is strongly correlated with tree health, indicating that inundation is having a detrimental effect on stand health. The rate at which the stand is taking up carbon in the form of live biomass has decreased from 39.1 Mg C/ha in 2000 to 25.1 Mg C/ha in 2002. The amount of live biomass in the swamp is also decreasing as trees die and are converted to dead biomass that slowly releases carbon to the atmosphere. This conversion of live biomass to dead biomass will likely change the carbon balance of the swamp. Future investigation of the decomposition of this dead biomass, including monitoring of carbon dioxide and methane effluxes, is necessary to determine the long-term impact of inundation on the carbon cycle of the stand.

The Effects of Elevated CO₂ on the Photosynthetic Rates of Competing Common Ragweed Individuals

Jen Petzold
Atmospheric levels of CO₂ have increased 29% since pre-industrial times and are expected to double in the next century. This rise in CO₂ may affect plant competition, and previous studies with birch seedlings indicate that size differences between subordinates (smaller plants) and dominants (larger plants) decrease under elevated levels of CO₂. The objectives of this study were to examine 1) the effects of elevated CO₂ on the photosynthetic rates of subordinates and dominants, and 2) the relationship between plant architecture and photosynthetic rates. Seedlings of Ambrosia artemisiifolia were planted in open top chambers and exposed to either elevated (720 ppm) or ambient (360 ppm) levels of CO₂. Maximum photosynthetic rates (p-max) were measured using a Licor 6400 system. P-max values were obtained from the topmost and bottommost leaves of each plant from four stands-two stands per each CO₂ level. It was found that increased levels of CO₂ significantly decreased the photosynthetic rates of dominants but did not alter those of subordinates. The difference in p-max values between dominants and subordinates decreased in high levels of CO₂. There was also a positive correlation between size and p-max in topmost leaves, and top and bottom leaves differed in their sensitivity to elevated levels of CO₂. Thus, both physiology and architecture have a role in the response of plants to elevated CO₂, and may contribute to changes in size differences between subordinates and dominants in competing stands.

A Study of Hemlock Stand Characteristics and Hemlock Woolly Adelgid Distribution In Central Massachusetts

Nicholas Povak
The hemlock woolly adelgid (HWA) is an aphid-like insect that dispersed into southern Connecticut in 1985, and has since migrated north into Massachusetts. These insects feed on eastern hemlock (Tsuga canadensis), reducing the tree's food source and resulting in hemlock mortality. This study was implemented in Massachusetts in order to: 1) track the migration patterns of the adelgid; 2) obtain baseline data of hemlock stands in central Massachusetts; and 3) examine the role of stand level factors such as slope, aspect, elevation and soil type on the distribution of HWA. Hemlock distribution in central Massachusetts was delineated from 1:40,000 color infrared photos and transferred to topographic maps. Thirty hemlock stands located in the northwest corner of the study area were initially sampled based on size, access, and amount of hemlock. HWA density, stand composition, and stand features were measured along a fixed transect using both variable-radius points and a 20 x 20 m plot. The majority of these stands were found on west facing aspects, and averaged 25% slope, with a humus depth of 4.5 cm. Thirteen out of our 30 stands (43.3%) had HWA present, with nine of the thirteen infested sites (69.2%) being located in our two most southerly maps. We found adelgid densities were highest in low elevation sites, although most stands were at an early stage of HWA infestation. Thirty percent of the stands had logging due to HWA. Preliminary results suggest that adelgid has migrated to the northern border of Massachusetts with heavier infestations in lower elevation sites.

Diurnal Cycles in Soil Respiration and the Role of Soil Temperature and Soil Moisture

Greg Santoni
Soil respiration, comprised of both root respiration and microbial decomposition, is an important component of the global carbon budget. Thought to be of principal importance in driving soil respiration are soil temperature and soil moisture. The goal of this experiment was to determine if a diurnal cycle exists for soil respiration, and to characterize the relationship between any observed trends in soil respiration in terms of soil moisture and soil temperature. For three days in July 2002, measurements of soil respiration, soil and air temperature, and soil moisture were taken every three hours for 27 hours at 4 plots. Each plot contains five "collars" consisting of 10" diameter ring of PVC imbedded into the ground. Two 30cm Time Domain Reflectometry (TDR) probes implanted vertically into the ground adjacent to two collars in the plot. Soil respiration was measured using a static chamber and an Infrared Gas Analyzer (IRGA). Soil respiration was found to exhibit a sinusoidal diurnal cycle with amplitude of approximately 0.6 over an average flux of 4.4 umolCO²/m²/sec for all plots. Minimum value occurred at 06:00 and maximum at approximately 18:00 EST. Respiration lagged the diurnal pattern of soil temperature by approximately three hours. The variability of soil moisture (10 - 50 % volumetric water content) demonstrated a significant correlation between soil moisture and soil respiration, but temperature may interact with this pattern. Soil respiration was greatest near the median of soil moisture values and fell off as soil moisture approached either extreme in the range. Surprisingly, only half the sample locations exhibited a modest relationship between soil temperature and respiration. The remainder exhibited minimal correlation, indicating that soil temperature may interact with other variables such as organic matter or allocation of photosynthates. Soil moisture and soil temperature can account for only part of the diurnal cycle. Respiration may also be closely linked to the metabolism of photosynthates by plant roots, a process that may be linked to periods of increased photosynthesis.

The Role of Phytochromes in Germination with Respect to Environmental Cues in Arabidopsis thaliana

Joseph Seggio
Seasonal cues can influence the timing of germination. For example a period of cold can be a seasonal cue, and the seed will germinate after that period. The purpose of this inquiry is to investigate the germination rate and percentage of phytochrome deficient mutants of Arabidopsis thaliana, when exposed to different environmental conditions. Twelve different phytochrome mutants were used, (either deficient in phytochrome A, phytochrome B, both A and B, all five or none), and they were treated with the following conditions: dark/cold, dark/warm, dark/room temperature and light/room temperature. The seeds placed into the dark were given a pre-incubation treatment, in their respective stratifications. The plants were censused twice for the number of dark germinants, (extended hypocotyls), normal germinants, (normal cotyledon), and ungerminated and dead plants were recorded. Dark germination, total germination and germination rate were compared across treatments for each genotype. The germination rate of different genotypes responded differently to pre-incubation. The warm stratification caused less germination in the dark and slower rates of germination than the other pre-incubated stratifications. There was no effect of pre-incubation on total germination, but the genotypes differed significantly from each other. Er-1 had a higher germination percentage than all of the photoreceptor mutants, except for PhyB1. Phytochrome A and B were found to play important roles in germination. Stratification did not influence the total germination percentage of the mutants, showing that phytochromes might not be able to sense the environment.

Quantifying Patterns of Sectoriality in Selected Deciduous New England Trees

Sigrid Smith
Tree responses to environmental heterogeneity are poorly understood. We feel that vasculature determines how species respond to this variability. A tree's vascular architecture can restrict the movement of water and other substances within a specific sector of the plant: this is known as sectoriality. We hypothesize that species differ in sectoriality, and that the degree of sectoriality is correlated to wood structure. Specifically, we predict that ring-porous species will exhibit greater sectoriality than diffuse-porous species, because rapid vertical flow is a characteristic of ring-porous species. We tested this in 9 species by pushing 10 mmol KCl at a known pressure from a nitrogen tank through one petiole of a growing shoot and measuring flow rates to leaves with and without direct vascular connections to that leaf. Conductance and resistance of these pathways were normalized by leaf area. We found leaves with direct vascular connections in ring-porous species had significantly lower resistance than leaves not directly connected to the leaf pushed into, whereas the diffuse-porous species did not show this strong relationship. Distance the liquid traveled to reach each petiole also appeared to affect these flow rates, especially in Betula lenta, which exhibited increased resistance to distant leaves. In the future, we hope to extend this method of quantifying flow to larger scales in trees (branch to branch and root to shoot), and to further investigate the effects this sectoriality may have on patchy resource exploitation and patchy herbivory responses.

Plumbing Pathways and Leaky Pipes: Water Distribution and Flow in Leaves of Quercus rubra and Acer saccharum

Christopher Streeter
Water distribution throughout the leaf plays a central role in leaf function. How the veins function in distribution has remained virtually undetermined. In leaves of Quercus rubra and Acer saccharum, we examined whether veins were organized as an irrigation system, or pressure manifold, providing constrictions between different vein orders in order to deliver water equitably throughout the leaf. Pushing water into excised leaves at a known delivery pressure and flow rate, we measured hydraulic resistances of intact leaves, and also for leaves cut at each vein order. We also measured the temperature response of hydraulic resistance in each case (from 10 to 25°C) to investigate the possibility of leakage through membranes. Our preliminary analyses show important differences between Quercus and Acer leaves. In Quercus we found two major constrictions: between the tertiary and fine veins (46% ± 2 SE), and across the fine veins and bundle sheath (31% ± 6 SE). In Acer we found strong constriction between primary and secondary orders (~20%), as well as across the fine veins and bundle sheath (~37%). Intact leaves showed a strong non-viscosity response, which indicates that water moving out of the veins crosses a membrane, analogous to the endodermis of roots. Our findings suggest that the leaf provides constrictions within the vein architecture, allowing even distribution of water throughout the mesophyll. The majority of leaf hydraulic resistance occurs between vein orders. Thus, venation properties play an important role in defining leaf water use and thus whole-plant water use and carbon gain.

A Genetic Basis for Germination Responses to Seasonal Cues in Arabidopsis thaliana

Shaun Takao
Seasonal seed dormancy is an important mechanism that can delay germination until environmental conditions are best suited for germination. This experiment investigates the genetic basis behind seasonal seed dormancy and subsequent germination responses to parental photoperiod and progeny stratification in Arabidopsis thaliana. Seeds representing 120 genotypes were collected from Cal x Tac recombinant inbred lines whose parents had been exposed to either a long day or short day photoperiod. These seeds were then allowed to germinate under a warm (31° C) environment or a cold (4° C) environment and censused for percent germination taking into account seeds that germinated in the dark as well and those that germinated in the light. Total germination was largely unaffected by maternal photoperiod in both warm and cold treatments. However, maternal photoperiod significantly affected dark germination in both the cold and warm progeny environments. Dark germination and total germination were both significantly influenced by cold and warm stratification treatments in long days as well as short days. In particular, total germination showed a marked increase in percent germination under warm conditions as compared to cold conditions for both maternal photoperiods. In each of the environmental conditions being investigated, there was a significant amount of genetic variation with respect to expression of germination. Thus, particular genes are influencing seed dormancy and germination. Furthermore, the high genetic correlations across environments show that the same genes are largely influencing germination across environments, although the possibility remains that some genes are uniquely affecting germination in particular environments. A preliminary Quantitative Trait Locus analysis has identified several of these genes associated with germination.

Light and Architecture of Ragweed (Ambrosia artemisiifolia) in Elevated CO₂ Environments

Jimmy Tran
Future estimates of global climate change predict a doubling of atmospheric carbon dioxide. Competition for resources in this changing environment will be a fundamental force behind the evolutionary success of plant species. To help better understand how this change will affect plant competition and the evolutionary survival of plants, I investigated how CO₂ enriched environments would influence light competition. Six open top chambers were used to grow stands of ragweed (Ambrosia artemisiifolia) in both ambient (360 ppm) and twice ambient (720 ppm) conditions of CO₂. A LICOR 1600 was used to take weekly light profile measurements every 5cm throughout the vertical height of 5 random plants in each stand, on representative sunny days. Light data for both ambient and elevated CO₂ chambers were averaged over time, and averaged over profile heights. Weekly measurements of total branch length and total plant height were made to track seedling growth and plant architecture. Initially, ambient CO₂ chambers received less light than elevated CO₂ chambers. Over time, however, ambient CO₂ chambers received more light than elevated CO₂ chambers. This difference in light availability was apparently due to shading caused by architectural changes in the plants. Elevated CO₂ chambers consistently had smaller size differences between subordinate and dominant plants than in ambient chambers. Subordinates were able to "catch up" to dominant plants in high CO₂ conditions, while ambient CO₂-grown stands maintained greater size differences. Measurements made over the remaining growing season will give a better understanding of how subordinates are able to catch up to dominants in low light environments.

Not Simply Gone with the Wind: Maternal Influence on Seed Dispersal of Arabidopsis thaliana

Naomi Wender
Though seed dispersal is a progeny characteristic, it is primarily maternally determined. Maternal influence on dispersal is twofold: maternal architectural traits, such as branching and fruit placement strongly govern dispersal, as does the surrounding maternal environment, such as sibling and neighbor density. I examined seed dispersal in Arabidopsis thaliana in both high and low density to examine environmentally regulated maternal morphology, the effect of maternal phenotypic traits on seed dispersal, and genetic variance across density treatments. Data were collected on dispersal patterns of 18 different ecotypes in both high and low-density treatments. Dispersal distance, standard deviation of dispersal distance, and progeny density (seeds/.25 cm²) were used as measurements of dispersal. I found that maternal phenotypic characters had a significant effect on all three dispersal measurements when plants were grown in high, but not low, density (Table 1). Density indirectly affected dispersal distance due to the plasticity of the ecotypes to the environment, and directly affected progeny density via neighbor interference. Genetically, almost all maternal traits varied significantly by ecotype, density, and plasticity of ecotype to density. Though there were significant differences in maternal phenotypic traits between the two densities, a maximum likelihood statistical analysis found no difference in the amount of genetic variance expressed within each environment. Because maternal characters affect dispersal in high density, these results suggest that high density plants have more potential for the evolution of dispersal and that perhaps there is a limit to the evolution of dispersal; once a certain dispersal efficiency is reached, it is very difficult for the trait to evolve any farther.

Differences in Aquatic Insect Communities in Decodon verticillatus (L.) Elliott Pools

Eugenia Yazzie
The similarities and differences of aquatic insects were compared among vernal pools in Cape Cod dominated by Decodon verticillatus (L.) Elliott. I identified the insects to order and family to see what structure there was within these similar environments. Three samples were taken from each pool during June 2002 and temperature and water depth were recorded. The pools were then compared to the data collected in April 2002, and April and May 1998. Pools were similar in the dominant vegetation of Decodon verticillatus (L.) Elliott, Sphagnum, and temperature. Pools differed in water depth, canopy, and open water. Between years hydroperiod varied: 2002 was dry and 1998 was wet. Orders present were the same in E6 and E11 in June 2002 and in E7 and E11 in April 2002, and April and May of 1998. Frequency of orders was the same in E7 and E11 in May 1998 and in all three pools in April 1998. Within-year frequencies changed. The order and family of insects between years showed significant differences within pools. Site E6 was always the most diverse. Between pool differences can be explained by differences in open water and canopy cover. Within a given year changes are due to insect life cycles. Between-year hydroperiod plays a significant role in insect community diversity.