2015

Eutrophication is a critical threat to many coastal ecosystems; however, seagrass meadows can potentially mitigate some effects of nutrient pollution by stimulating biogeochemical processing of excess nutrients. In this study, potential and ambient nitrate reduction rates were measured in the sediment of restored seagrass meadows in order to determine the magnitude of subsurface nitrate reduction and to compare two competing nitrate reduction processes: denitrification and anaerobic ammonium oxidation (anammox). Nitrate reduction rates were measured using both jar incubations and in situ porewater incubations in combination with the isotope pairing technique. Jar incubations showed significant potential denitrification rates of 2.7 ± 0.3 nmol N/cm³ sediment/hr (mean ± SE), while anammox contributed less than 1.5% of potential nitrate reduction rates. In situ porewater incubations were conducted using the novel push-pull isotope pairing technique to capture ambient denitrification rates in the seagrass root-zone. These incubations showed significant subsurface denitrification rates of 19.4 ± 1.7 μmol N/m²/hr (mean ± SE). Potential denitrification rates measured with the push-pull technique were an order of magnitude greater than ambient rates, consistent with the jar incubation results. Overall, these results suggest that in vegetated sediments, substantial denitrification occurs within the root zone, while anammox contributes minimally to net nitrate reduction. The high potential rates suggest that an increased nutrient load to the seagrass meadow could enhance denitrification and the removal of nitrate from the ecosystem. Author affiliations: Department of Environmental Sciences, University of Virginia

Planetary boundary layer (PBL) depths represent a crucial parameter in weather forecasting, modeling of air quality, and in describing the transfer of momentum, energy, temperature, moisture and aerosols with the free troposphere aloft. PBL depths are highly dependent on local sources of heating and cooling, the diurnal cycles of the latter, as well as on surface topography and features in the underlying terrain. Over complex, mountainous terrain, the aforementioned processes interact over several spatial and time scales, further complicating the understanding of PBL depths. Since the early days of standardized meteorological measurements, the only means of successfully estimating PBL depths was by using radiosondes, while more recently, ground-based or air-borne LIDAR instruments, and wind profilers have been used. These estimates were highly localized in both space and time. However, the launch of the Cloud-Aerosol Layer and Infrared Pathfinder Satellite Observations (CALIPSO) satellite mission finally enabled the development of PBL depth climatology on a global scale. In this study, we explore the potential of using these satellite-based estimates of PBL depths to develop a smaller, more regional case study of daytime PBL depths over the highly complex topography of the Rocky Mountains. We found that the deepest PBLs occur in the summer over the southern Rockies, while during the winter the overall spatial structure of the PBL depth tends to be more homogeneous and also affected by the presence of cold air pools. Since CALIPSO overpasses this region during the early afternoon, we also investigate the time evolution of daytime PBL by comparing CALIPSO results with those based on radiosondes, which are usually launched during late afternoon over the Rockies.Author affiliations: Department of Environmental Sciences, University of Virginia

Simulation models are used to study systems that may not easily lend themselves to direct observations. One such system is the vast Siberian boreal forest, which occupies over 600 million hectares and experiences changes over timeframes on the order of decades to centuries. We developed a new spatially-explicit dynamic vegetation model SibBorK that simulates tree growth and associated forest processes in 3-dimensions on real terrain. As an individual-based gap model, SibBorK simulates the establishment, growth, and mortality of up to 90,000 trees on a 9-hectare swath of landscape. Each individual tree interacts with its neighbors through intra and interspecific competition based on species-specific tolerances to shade, drought, soil nutrition, and degree day requirements.

SibBorK was parameterized from and calibrated to regional forestry yield tables for southern taiga region of central Siberia. The model was further tested against an independent dataset from a forest inventory. The ability of the model to simulate monospecies and mixed forest stands, even-aged cohorts and mixed-age forest, was quantitatively evaluated using stand characteristics of diameter-at-breast height (DBH), height, stem density, biovolume, basal area, and species composition. The generalizability of the model was assessed through simulation of a northern taiga forest described in the literature and located outside of the calibration region. This assay of multi-dimensional tests shows that SibBorK is particularly good at predicting stand structure and species composition on poor soils, which dominate the Siberian landscape. SibBorK is robust, as it is able to predict stand characteristics in boreal forest stands without the need for fitting or re-parameterization when simulating locations outside of the initial calibration region. Future applications of SibBorK include climate sensitivity analysis to assess how boreal forest structure and composition may change with shifts in temperature and precipitation regimes.Author affiliations: Department of Environmental Sciences, University of Virginia

Over 80% of flowering plants worldwide depend on animal pollination, with insects being the most prolific pollinators, contributing billions of dollars to US agriculture every year. However, despite the importance of insects to agriculture, little is known about movement of insects across agricultural-non-agricultural continuums, making it difficult to assess the benefits or stresses agriculture imposes on pollinator populations. In this study, we examine the use of naturally occurring stable isotope signatures as a potential method to distinguish where insects in a landscape have been feeding. To ascertain if naturally occurring stable isotopes can be used as unique signatures for a given area or plant species, we conducted a study examining four approximately .5X.5 km areas with presumably different 15N levels. Areas included a plot that had been burned within a year of the onset of the study, within two years, a plot that had not been burned for at least a decade, and a conventionally fertilized agricultural plot.In each treatment area, five species of plants (Apocynum cannabinum, Solidago speciosa, Allium tricoccum, Rhamus davurica, and Physalis longifolia) were sampled and ¹⁵N differences assessed using stable isotope analysis. There were significant differences between the agricultural plot and all other plots for Apocynum cannabinum, Solidago speciosa, and Allium tricocum species (F_(1,58)=4.77, p= .031). Additionally, each species within a given plot was significantly unique (F_{(2, 58)}= 4.98, p= .023).

To ascertain if these differences were still discernable in insects feeding in fields, 18 Dogbane beetles (Chrysochus auratus Fabriciu) were collected from their specialist host Apocynum cannabinum, for each plot, and analyzed for isotopic differences: dogbane beetles in agriculture and unburned areas were significantly different than those found in the burned plots (F_{(2, 58)}= 5.68, p= .026; F_{2, 58)}= 4.22, p= .007). Results from this study indicate that if the isotopic signature for a specific plant species is known, it may be possible to determine if a specialist insect forager has been feeding primarily in unmanaged fields or in a managed cropland. Understanding the movements of insects between crops and agriculture land is crucial in order to identify best-management practices to maximize pollination. Author affiliations: Department of Environmental Sciences, University of Virginia

Shellfish aquaculture is prominent in many coastal and estuarine environments, having both ecological and economic effects. Bahía San Quintín is a reverse estuary in Baja California, Mexico where Pacific oysters (Crassostrea gigas) are cultivated. While oysters likely feed heavily on phytoplankton especially during upwelling periods, we hypothesized that other forms of organic matter such as seagrass (Zostera marina) and macroalgae (Ulva spp.) are used by the oysters, especially in the most inshore portions of the bay. We measured the carbon and hydrogen stable isotope composition of oysters and their potential food resources and applied a Bayesian mixing model to evaluate resource use. Although we did not find any strong seasonal effects due to upwelling, there was a strong spatial gradient in resource use. Phytoplankton were most important at a lower (oceanic) site (median resource use for two sampling times: 68 and 79%) and decreased up the estuary as macroalgae became more important (44 and 56%). At all sites for both sampling times seagrass was a minimal resource (3-7%) for oysters. The gradient of high phytoplankton use at the lower site to increased macroalgal use at the upper site is likely due to resource availability in the bay. Results indicate the adaptability of oysters to varying resource availability and the possibility of a higher system carrying capacity given multiple potential food sources.Author affiliations: ¹Department of Environmental Sciences, University of Virginia, ²Instituto de Investigaciones Oceanologicas, Universidad Autonoma de Baja California, Ensenada, Baja California, Mexico

Wave regeneration phenomena have been the focus of many studies in stressed conifer forests. In these systems, suppressed seedlings are released following the synchronous death of canopy trees. This cyclical pattern across the landscape is in contrast to the classic successional pattern, in which a “climax” community eventually evolves, producing stable biomass over time. These cycles occur over hundreds of years, and thus studying them in the field is difficult, if not impossible in some cases. This difficulty highlights the advantages of vegetation modeling studies. We used the individual-based gap model, the University of Virginia Forest Model Enhanced (UVAFME) to simulate forest dynamics over time at a high-elevation, subalpine forest (dominated by Engelmann spruce and subalpine fir) in southern Wyoming. UVAFME was first calibrated to the study site. Following successful model calibration, UVAFME was validated by running it up an elevation gradient to determine if it could accurately simulate changes in species composition with elevation. UVAFME was then run exclusively at the high-elevation location for periods of 3,000 years to simulate long-term forest dynamics at the site. It was found that the subalpine zone of the Rocky Mountains do in fact show wave phenomena, both at the plot scale and the landscape scale. By itself, Engelmann spruce demonstrates a natural wave frequency of 500 years, whereas subalpine fir has a natural wave frequency of 300 years. Together, both species have a wave frequency of 500 years. This output corresponds well with field data from similar subalpine sites, and from other high-elevation conifer sites. While the theory of forest succession from bare ground up to an eventual climax community has been seen in many types of landscapes, in the case of high-elevation, fairly harsh landscapes, the notion of stable biomass over time may not be accurate. Author affiliations: Department of Environmental Sciences, University of Virginia

Warming in the Alaskan Arctic has resulted in a lengthening of the growing season and changes to the distribution and composition of tundra vegetation. Past studies have mapped tundra vegetation at relatively coarse spatial scales; however, vegetation changes in the Arctic are occurring at fine spatial scales that would benefit from hyperspectral remote sensing. This research uses handheld hyperspectral remote sensing data and harvested biomass data collected during the 1999 peak-growing season at Ivotuk, Alaska (68.49°N, 155.74°W) to estimate biomass quantities at six other sites in the Alaskan Arctic. Ivotuk is located on the North Slope, and is comprised of four plant communities including moist acidic tundra (MAT), moist nonacidic tundra (MNT), mossy tussock tundra (MT), and shrub tundra (ST). Hyperspectral data were collected at biweekly intervals using narrow, ~ 1.42 nm wavebands. Only wavelengths from 400-1060 nm will be used in this analysis. Spectra will be normalized prior to analysis using continuum removal. Biomass categories at Ivotuk will then be regressed against the hyperspectral data using the partial least squares regression method. The resulting equations will be used to predict biomass quantities at six different locations in the Arctic for which we also have hyperspectral information: Howe Island, Happy Valley, Deadhorse, Sagwon moist acidic tundra (MAT), Sagwon moist non-acidic tundra (MNT), and Franklin Bluffs. The goal of this research is to develop equations that can then be used to estimate biomass at sites using only spectral data. Author affiliations: Department of Environmental Sciences, University of Virginia

Species traits, particularly those that impact fitness, can shape the evolutionary relationships among coexisting species. Trait distribution (underdispersion, overdispersion) within communities can provide evidence of key ecological interactions (e.g., competition, facilitation) that can contribute to assembly. The distribution of floral colors in a community may reflect pollinator-mediated interactions, and the phylogenetic distribution of color can also affect inferences of ecological mechanisms at play. Additionally, the scale of local habitat may influence the type or strength of ecological interactions among co-occurring species. We examined how floral color is distributed within replicated co-flowering assemblages with the use of pollinator color vision models. Incorporating these biologically relevant models into the study of floral color assembly processes is relatively new and untested for an entire co-flowering community with generalist pollinators. We modeled floral spectra of 55 co-flowering species through honeybee and syrphid fly color vision to assess color trait structure across 14 serpentine seep communities in California. We then compared our findings to null model predictions. We asked: is there evidence for nonrandom distribution of floral color in the community? Is there phylogenetic signal for floral color? If so, is there phylogenetic underdispersion or overdispersion across local communities? Is there an effect of habitat scale on these outcomes? We found that the observed color assemblage is not due to any phylogenetic history. We found a significant negative relationship between habitat scale and trait dispersion. Competitive exclusion could be a dominant interaction outcome at small scales, but it is less detectable/unimportant at larger scales. Author affiliations: ¹Department of Environmental Sciences, University of Virginia, ²University of Pittsburgh

“Nutrient bioextraction” using bivalve aquaculture is recently being considered as a tool to mitigate coastal nutrient pollution. While nutrients removed via shellfish harvest are relatively easily calculated, denitrification in sediment impacted by aquaculture remains poorly quantified. We measured net denitrification and nutrient fluxes from sediment at an off-bottom oyster farm (rack and bag cultivation of Crassostrea virginica) in a Chesapeake Bay tributary. During the two seasons studied thus far, we found slightly enhanced rates of denitrification both under and near oyster racks relative to a nearby reference site. However, rates at all sites were low, and within or near the detection limit of the analyzer. Nitrate (NO₃) fluxes were low and addition of NO₃ increased net denitrification, indicating that denitrification at the site is NO₃ limited. The marginal increase in denitrification was dwarfed by sediment ammonium (NH₄) fluxes, which were more than an order of magnitude higher than N₂ fluxes. In addition, the total “loss” of N through denitrification was trivial in comparison to that removed by harvest. The data collected thus far indicate that denitrification is not a significant pathway for nutrient bioextraction at this site. Author affiliations: ¹Department of Environmental Sciences, University of Virginia, ²Virginia Institute of Marine Science

We found that burrow- building polychaetes induce oscillating redox conditions in both the burrow water and the sediment at the burrow wall. We show that ambient temperature affects the patterns of these oscillations; specifically, increased temperature resulted in increased frequency of burrow ventilation and oxygen flux across the burrow wall. Author affiliations: Department of Environmental Sciences, University of Virginia

Turbulence is the key driver for the vertical transport of heat, momentum and moisture in the convective boundary layer. Turbulence features of thermodynamic variables as well as fluxes of heat and moisture are well investigated over flat terrain but they still remain poorly defined over complex terrain. The objective of this study is to investigate spatial variability of turbulent kinetic energy over mountainous terrain. Turbulence measurements were made using Naval Twin Otter (TO) in-situ airborne measurements performed during the first field experiment of The Mountain Terrain Atmospheric Modeling and Observations (MATERHORN) Program (at Dugway Proving Ground (Utah, USA) in Fall 2012. The in-situ airborne observations of meteorological variables were made at 10 Hz over the steep isolated mountain, Granite Peak, of a horizontal and vertical scale of about 10 km and 1 km, respectively. Dynamically and thermally induced modifications of the CBL turbulent structure around Granite Peak are investigated including the presence of organized turbulent regions. The spatial variability of TKE is demonstrated by comparing boundary layer turbulence statistics as a function of height and a function of location upwind and downwind of Granite Peak. Analysis of TKE, calculated from airborne in-situ observation, exhibits as much as 1.0 and 1.5 m²s² variability in TKE across a region ~ 3 km to the east of Granite Peak (Sagebrush) and over the east slope of Granite Peak (Eastern Slope), respectively. The results also show a distinct increase of 0.5-2.0 m²s² from ~ 3 km east to the east slope of Granite Peak, respectively. The origins of the observed turbulence spatial pattern is examined to determine the relative contribution by shearing and buoyancy mechanisms. Additionally, the nature of turbulence is explored using spectral analysis to determine the dominant scales important for turbulence generation.Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²Simpson Weather and Associates, Charlottesville, VA, USA

Surface waters are an important pathway for the transport of atmospherically deposited mercury (Hg) from terrestrial watersheds. Dissolved Hg (Hg_D) is thought to be more bioavailable than particulate Hg and has been found to be strongly correlated with dissolved organic carbon (DOC) in numerous watersheds. The ratio of Hg_D to DOC is highly variable from site to site, which we hypothesize is strongly dependent on local environmental factors such as atmospheric deposition, soil organic carbon (SOC), and quality of DOC measured by specific ultraviolet absorbance at 254nm (SUVA₂₅₄). Nineteen watersheds throughout the United States were used in this study to determine the relationship between the ratio of Hg_D:DOC, Hg wet deposition, SOC, and SUVA₂₅₄. There was no correlation found between atmospheric mercury wet deposition and Hg_D:DOC (r² = 0.04; p = 0.44) or between SUVA₂₅₄ and Hg_D:DOC (r² = 0.18; p = 0.05). SOC was able to explain about 80% of the variation in the Hg_D:DOC ratio (r² = 0.80; p < 0.01). A mathematical framework was developed to explain the observed power-law relationship between SOC and Hg_D:DOC based on soil carbon pools. The framework infers that the amount of Hg adsorbed to SOC does not increase in proportion to SOC at high SOC levels and points towards a Hg supply limitation for adsorption to soils with relatively deep carbon pools. Overall, this study identifies SOC as a first-order control on the association of HgD and DOC and indicates that globally available SOC datasets can be utilized to predict Hg transport in stream systems.Author affiliations: Department of Environmental Sciences, University of Virginia

Aside from the aesthetic problem plastics in the ocean present, there is also great concern for the environmental impacts as well. Prior to 2009, the primary environmental concern regarding plastic hinged on the idea that plastic does not readily degrade and may take hundreds to thousands of years to do so. The reasoning behind this is that plastics are predominantly composed of hydrocarbons. Hydrocarbons, by their chemical nature, tend to be highly unreactive, especially the alkanes.

The idea of hydrocarbon stability in natural environments was challenged when Saido (2009) reported detectable amounts of styrene monomers, dimers, and trimers in ocean water and sand from beaches along the Japanese coastline. These molecules are not synthesized in nature but are the result of manmade products.

Reported here, is the first evidence for purportedly rapid plastic abiotic degradation associated with popular consumable hard plastics: high density polyethylene, polycarbonate, and polyethylene terephthalate (HDPE, PC, and PETE) under typical oceanic surface temperatures (20-23 ⁰C). The rates of degradation are compared in these simulated marine environments to those observed in fresh waters and display a strong disparity.

We continue to investigate the degradation process taking advantage of laboratory-based and field-based experiments. Along with observing the physical and chemical changes that take place, we follow the changing isotope composition of plastics during abiotic degradation processes which would serve as a proxy for the length of time pieces of plastic have been circulating in surface ocean waters, thereby acting as a geochronometer. Author affiliation: Department of Environmental Sciences, University of Virginia

Spartina alterniflora, salt marsh cordgrass, is the dominant plant in coastal wetlands along the North American Atlantic coast. Salt marsh dieback has affected Spartina marshes from the Gulf of Mexico to New England. The cause of dieback is not clear; however, dieback may reduce the diversity of S. alterniflora clones within a population or alter other genetic characteristics of a population by eliminating some genotypes. Nine polymorphic microsatellite loci were used to quantify the genetic characteristics (e.g., allelic richness, diversity, polyploidy, fixation index) of the S. alterniflora populations at five salt marshes, as well as, to measure the spatial structure (size and shape of clones) of a single population in Upper Phillips Creek marsh (UPC), a marsh that experienced dieback. Over 250 individual plant samples were collected at 3 spatial scales for these experiments. Initial results show that triploid plants are relatively common at UPC and Oyster Harbor marshes, suggesting that these two populations may be experiencing stress, while polyploidy was not observed at the three other marshes. Based on the size of plant clones (< 10 m², but > 10 cm²) and the size of the dieback patches (>10 m²) in UPC marsh, it is likely that multiple clones were killed during the dieback events and may have reduced or eliminated some genotypes from the population. While analysis of the genetic results is still ongoing, when they are completed, the results will provide important information on how dieback events may impact the genetic diversity of S. alterniflora populations and their ability to adapt to environmental variation and change. Author affiliations: Department of Environmental Sciences, University of Virginia

Biological invasions may impact the multi-trophic interactions of ecosystems. Exotic species invading deciduous forest can change leaf litter inputs, which can impact detrital food webs. Microbial decomposer communities may differ between exotic and native leaves due to differences in physical and biochemical properties. Effects on this foundational trophic level could have cascading effects the food web. This study investigated the effects of invasive shrubs (Lonicera maackii and Rhamnus davurica) and an invasive tree (Ailanthus altissima) on the litter-dwelling communities of a hardwood forest in Virginia. Our hypotheses were that both the source of litter (native vs. invasive) influences the number, distribution, and composition of both the microbial and arthropod communities.

Invasive litter decomposed faster than native litter and there was less litter cover in invaded sites than native sites. The nutritional quality of the litter was also higher for invasive leaves than native leaves. We found significantly more bacteria in litter from all three invasives compared to native. Lonicera litter supported significantly more fungi than native litter. We found no difference in total arthropod abundance between invasive and native litter, but when the population of ants was excluded, there were significantly more arthropods in the litter from all three invasives compared to native litter. Arthropod richness was not affected by litter source. The abundance of trophic groups was significantly greater in litter from Ailanthus and Lonicera, compared to native litter. Additionally, the richness of trophic groups was significantly greater in the litter from all three invasive species compared to native litter. These results show that the identity of the resource matters for the litter-dwelling communities. By altering the resource base, plant invasions may have long-term effects on the multi-trophic interactions of this ecosystem. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²Blandy Experimental Farm, University of Virginia

View Presentation
The quantification of terrestrial carbon storage lies on the forefront of scientific understanding of the global carbon cycle. Existing mature and secondary successional forests act as an extremely effective means of long-term carbon sequestration. The standard method of assessing forest carbon is typically based on national or regional-scale allometric equations that are often not representative on the plot-level. Improvement of these measurements is necessary in order for collaborative multi-national carbon monitoring programs to be successful in areas with tree species that have insufficiently documented allometric relationships. On a global scale, strengthening confidence in these estimates will aid in development of landscape-scale modeling projections of carbon storage. Furthermore, the study of the structural allocation of this carbon is needed in order to fully understand the role of forests in the carbon cycle. We determine C storage from volume measurements with a high-precision Terrestrial Laser Scanner (TLS), substantially improving current standard ground validation techniques. This technology is utilized on several 30 m x 30 m plots in a Virginia temperate forest. Aboveground C is calculated on each of the study sites with commonly used allometric equations to offer a realistic comparison of field-based estimations to TLS-derived methods. In general, local application of national-scale allometrically derived height relationships result in an underestimate of actual tree height. A strong (R² = 0.85, n=18) allometric relationship between DBH and height can be created from relatively few individuals. TLS-derived biomass measurements are closely correlated to national-scale biomass equations at lower levels, but agreement at high levels is unpredictable (R² = 0.89, n=13). A local allometric equation created from these data (d) results in an excellent fit (R² = 0.88). Author affiliations: Department of Environmental Sciences, University of Virginia.

View Presentation
Plants with Crassulacean Acid Metabolism (CAM) are increasing their cover in many dryland regions around the world. Their increased dominance has been related to climate warming and atmospheric CO2 fertilization, while the effect of interspecies interactions and the role of CAM plant facilitation by trees and grasses remain poorly understood. Woody plants are known for their ability to directly facilitate CAM plants through amelioration of the abiotic environment. Mechanisms of indirect facilitation of trees on CAM plants in tree-grass-CAM associations, however, have received less attention. It is also unclear whether grasses might facilitate CAM plants in mixed tree-grass-CAM communities. For instance, the inclusion of grasses in tree-CAM associations could enhance hydraulic lift and facilitate CAM plants in their access to shallow soil moisture at the expenses of deep rooted trees. If this effect outweighs the competitive effects of grasses on CAM plants, grasses could overall facilitate CAM plants through hydraulic lift. Here we develop a process-based ecohydrological model to investigate the direct and indirect facilitation in tree-CAM-grass associations; the model quantifies transpiration of CAM plants when isolated as well as in associations with trees and/or grasses. It is found that woody plants having a high root overlap with CAM plants indirectly facilitate CAM plants by significantly reducing grass transpiration in shaded conditions. For situations of a low to moderate root overlap, facilitation may occur both directly and indirectly. Conversely, grasses are unable to indirectly facilitate CAM plants through the mechanism of hydraulic lift because the competitive effects of grasses on CAM plants outweigh the facilitation induced by hydraulic lift. Author affiliations: Department of Environmental Sciences, University of Virginia

View Presentation
Tropical peatland forest ecosystems have the capacity to store tremendous amounts of organic carbon in permanently waterlogged soils. However, in Indonesia peatlands are a major and growing source of carbon emissions due to increasing pressure on these ecosystems from the expansion of oil palm. Deforestation and drainage of tropical peat swamp forests for conversion to other uses results in a loss of carbon storage through the clearing and burning of forest vegetation as well as decomposition of peat soils and increased frequency of fires following drainage. We are using data from the Gravity Recovery and Climate Experiment (GRACE) and a global forest cover change product to investigate trends in terrestrial water storage associated with land use conversion in Indonesian peatlands between 2003 and 2013. Our initial analysis indicates that there are secular trends in GRACE terrestrial water storage observations in Indonesia that are consistent with drainage of peat for the establishment of oil palm plantations. A decreasing trend in GRACE terrestrial water storage anomaly measurements over the observation period indicates a substantial loss of ground water storage in peatlands. We combine data on peat organic carbon density, spatially explicit information on peat extent and thickness, and GRACE derived water storage trends to quantify anthropogenically driven water table changes and assess potential impacts on soil carbon storage. Our research represents the first known application of spaceborn observation of tropical peatland drainage to assess soil carbon storage loss. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²NASA Jet Propulsion Laboratory ³Earth Sciences Department, University of California, Irvine

View Presentation
Some of the most visible effects of climatic and land-use change can be observed on alluvial fans in arid regions, such as the American southwest. These changes have a major effect on flow path selection, which presents a danger to local communities. Here we use a combination of landform evolution modeling and a field campaign to Eureka Valley, Death Valley National Park, to identify triggers for shifts in channel direction. Preliminary results indicate a preference for channel avulsions to reoccupy former channels. In addition, we find that fine-grained overbank deposits are a key part of alluvial fan construction, a component that has been missing from many existing fan models. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²Space Sciences Division, NASA Ames Research Center ³Division of Geological and Planetary Science, California Institute of Tehnology

View Presentation
Bumble bees are important pollinator species for wild and economically valuable agricultural plants. While bumble bees are subject to many types of parasites, relatively little is known about sub-lethal effects caused by conopid fly (Diptera: Conopidae) parasites prior to the host’s death about 12 days after infection. In order to determine if conopid infections reduce pollen return in Bombus impatiens individuals, pollen collection was compared to the presence and stage of parasite infection and overnight trip status in commercial colonies allowed to forage in the wild. Bombus impatiens individuals were affixed with RFID tags to provide data on foraging trip length and total time spent outside the colony while video monitoring provided a visual method of scoring pollen return. Bees were then monitored daily in captivity after the foraging period until natural death occurred. They were then immediately dissected to ascertain conopid parasite status. The timing of infection was determined by using an infection date scale based off of conopid size at time of host death and video data was studied to look for changes in pollen loads. It was found that parasitism is marginally related to decreased pollen returns. It was also determined that overnighting behavior is not associated with conopid parasitism but rather trip day, and also results in reduced pollen returns. These findings contradict previous studies that suggested that overnighting and conopid parasitism were related. Finally, a mathematical model based on observed mortality and parasitism rates showed that conopid infections reduce the lifespan of bumble bees substantially beyond other sources of mortality. It is stilled believed overnighting behavior, reduced pollen loads and conopid parasitism could have possible long-term implications for overall colony success. Author affiliations: Department of Environmental Sciences, University of Virginia

View Presentation
Darkness is a defining feature of the natural, nighttime environment that allows some nocturnal animals to communicate using a language of light. Fireflies produce bioluminescent signals to entice potential mates, and in some cases, other fireflies as food. Here, we share some surprising behavioral observations about how firefly flash communication signals are used and abused in nocturnal food webs. Author affiliations: ¹St. Mary’s College of Maryland ²Department of Environmental Sciences, University of Virginia

View Presentation
Plants have evolved a variety of chemical and physical traits as mechanisms of defense against herbivores. Many plants defend themselves against herbivores by constitutively releasing foliar volatile compounds that repel insect herbivores. Herbivory can also induce emission of volatile compounds that attract enemies of herbivores. Some plants also contain trichomes, leaf hairs that act as a physical obstacle that inhibits insect movement. Previous studies have found interactions between trichomes and other plant defense traits that range from synergistic to antagonistic in mediating insect behavior. Inbreeding is a common population-level process in plants that accentuates the phenotypic effects of deleterious recessive alleles and could disrupt biosynthesis of volatile compounds. We investigated the roles of herbivory, trichome density, inbreeding effects, and their interactions in foliar volatile production in the mixed-mating herb Mimulus guttatus. We collected volatile samples from M. guttatus in a 2x2x2 factorial design with herbivory (damaged or undamaged), trichome density (high or low), and pollination (inbred or outbred) as treatments. Herbivory increased production of 6 of the most abundant compounds, suggesting that these compounds function in defense against herbivores. For one of these compounds, emission in undamaged plants was significantly higher in low-trichome than high-trichome plants, suggesting a tradeoff between these physical and chemical defense traits. Two other compounds showed a significant inbreeding effect. In particular, for both compounds production was significantly higher in outbred than inbred damaged plants, but no differences were shown between outbred and inbred undamaged plants, suggesting that inbreeding could inhibit increased production of these compounds as a defense response. Future work will examine the role of these volatile phenotypic differences in resistance of M. guttatus to its specialist herbivore Junonia coenia. These results will have broader implications for the evolution and ecology of plant-herbivore interactions and the evolution of plant mating systems. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²Department of Environmental Systems Science, ETH Zürich ³Department of Entomology, Pennsylvania State University ⁴Blandy Experimental Farm, University of Virginia

View Presentation
The organic carbon cycle of the Western Arctic Ocean is one of the most significant remaining obstacles to establishing a pan-Arctic organic carbon cycle. With modifications to sea ice extent, timing and abundances of the seasonal productivity and alteration in fresh water discharge expected from global climate change, a better estimate of the sources of organic carbon to the ocean sediments has become more pressing. However, the presence of three end-members: terrestrial organic carbon supplied by coastal erosion and riverine discharge, pelagic marine production from plankton and significant sea ice algal productivity complicates the traditional use of bulk isotope data to construct linear mixing models. Instead, in this report, we use a Bayesian analysis of the bulk data, allowing for the construction of robust mixing models and enabling the probabilistic estimation of the relative contributions of the three sources. Bulk isotope data from Shelf Basin Interaction cores, together with biomarker and compound specific isotope data from the same cores are used to quantify the various contributions and construct a model of the carbon cycle in the Western Arctic Ocean. Bulk carbon is enriched (δ13C has a mean value of -19.4±1.3‰), but the concentration of terrestrial biomarkers is also high. Based upon Bayesian analysis, 50-70% of the organic carbon is of marine origin. Terrestrial carbon made up 15%, with sea ice algal carbon accounting for the remaining 25-35%. We suggest that the discrepancy between the enriched bulk carbon and high terrestrial biomarker concentration is better explained by preferential preservation of the more recalcitrant terrestrial biomarkers than by a very high terrestrial contribution to bulk carbon. Author affiliations: Department of Environmental Sciences, University of Virginia

Reactive nitrogen is necessary for crop and animal production, but when it is lost to the environment, it creates a cascade of detrimental environmental impacts including smog, acid rain, eutrophication, climate change and stratospheric ozone depletion. The nitrogen dilemma is to maximize the food production benefits of reactive nitrogen, while minimizing losses to the environment.

In 2012, Allison Leach, Jim Galloway and others created the first nitrogen footprint to help consumers learn about the reactive nitrogen losses to the environment that result from individual’s lifestyle choices. The nitrogen lost during food production was estimated with Virtual Nitrogen Factors (VNFs) that quantify the amount of nitrogen lost to the environment per unit nitrogen consumed.

Since creating the tool, users have often suggested that alternative agricultural production systems, such as USDA certified organic products, may be less harmful than conventional production. To explore this further, we have partnered with The Organic Center to create VNFs that reflect organic production. These new factors will be added to the online N-Calculator so that consumers can see how their nitrogen footprint would change if they switched to organically produced agricultural products.

Initial data suggest that organic crop VNFs are 20-50% less than conventional crop VNFs, which means that organic crop production creates less nitrogen pollution per unit nitrogen produced than conventional crop production. Since there are larger differences between organic and conventional livestock production, the animal protein VNFs reflect a more complicated picture for which the VNF calculation must be adjusted.

In order to minimize the negative impacts of reactive nitrogen, consumers must make lifestyle choices that minimize their nitrogen footprint. With a deeper knowledge of the N losses from organic production relative to conventional, consumers will be more equipped to determine the potential ‘sustainability’ of purchasing organic products. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²University of New Hampshire ³McGill University

A nitrogen footprint links an entity with the reactive nitrogen lost to the environment as a result of their activities. This reactive nitrogen results in a cascade of environmental and human health problems. An institution’s nitrogen footprint is an important component to consider when determining its environmental impact.

The nitrogen footprint of the University of Virginia was calculated for 2010 and determined to be 492 metric tons N. Utilities (48%) and food production (38%) were the most significant sectors of UVA’s nitrogen footprint. In September 2013, the Board of Visitors passed an initiative to reduce UVA’s nitrogen footprint by 25% relative to 2009 levels by the year 2025. Many scenarios have already been suggested to achieve this goal.

This study examines additional scenarios that may be considered to reduce the university’s nitrogen footprint. Scenarios presented for consideration include 1) the purchase of organic food; 2) the use of nitrogen offsets; 3) consideration of alternative energy sources; 4) consideration of student-generated suggestions.

Significant findings included:

 The purchase of 50% organic vegetables, grains and starchy roots would result in a 1.4% decrease to the UVA nitrogen footprint.
 Nitrogen offsets also have potential but the implementation of such programs will depend upon their cost. Possible nitrogen offsets include reforestation and wetland mitigation projects.
 The nitrogen reduction potential of alternative energy sources should be determined along with their cost.
 Finally as suggested by students, working to increase awareness of the problem of reactive nitrogen in regards to food consumption would be beneficial to increase student support of nitrogen reduction strategies.

This semester I will continue this research. Specifically:
 I will explore the impact of the purchase of 50% of organic animal products (e.g., milk) on UVA’s nitrogen footprint.
 Conduct cost-benefit analysis of alternative energy sources
 Assess impact and cost of nitrogen offsets
Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²University of New Hampshire

Growing awareness of sustainability issues is prompting many communities to assess and manage their environmental impact. The amount of reactive nitrogen released as a result of an entity’s activities is an important indicator of environmental impact because it contributes to cascade of detrimental effects in different ecosystems, including smog, soil acidification, eutrophication, loss of biodiversity, and global climate change. The goal of the Nitrogen Footprint Tool project is to educate communities on the environmental impact of their choices. The calculation of a nitrogen footprint gives an indication of how much reactive nitrogen different activities contribute. Using the nitrogen footprint tool will help communities reduce losses of nitrogen to the environment as a result of their activities and use of resources.

With support from EPA’s Sustainable and Healthy Communities program, versions of the nitrogen footprint tool are being developed at various scales for a number of communities – these include ~10 additional universities, several secondary schools, a watershed area, and an urban area. The current focus on institutions means expanding the tool used for UVA’s nitrogen footprint so that other colleges and universities can use it. The combined goals of research, education, and sustainability at universities gives them the potential to be leaders in making positive management decisions and spreading awareness of environmental problems.

In addition to the tool for institutions, a version is being developed for secondary schools, where it will double as a hands-on environmental education project. Current work also includes a calculation for individuals in the Chesapeake Bay watershed, where nitrogen pollution in waterways is both a visible ecological problem and a pressing policy issue. Future work will focus on developing tools for urban areas, using the city of Baltimore as a model. These calculation scales can impact choices from the individual level to institutional management and policy levels. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²University of New Hampshire

The environmental impact from the consumption and upstream production of foods is poorlycommunicated to and seldom realized by consumers. As consumers become increasingly aware in food production and sustainability issues, it is necessary to provide a means by which consumers can compare environmental impacts across and within food product groups. There are a multitude of metrics by which to judge the environmental impact of food products. In this study, we demonstrate how the carbon, nitrogen and water footprints can be utilized in tandem to educate consumers about the environmental impact of food products. We propose different footprint calculation methods and label designs that can display these three footprints on food products at varying levels of detail. Three methods of footprint calculations are presented, two quantitative (footprint weight and % daily value) and one qualitative (sustainability criteria). Footprint calculations can then be displayed on one of four label designs (Star Label, Stoplight Label, Nutrition Label Add-On, and a Detailed Comparison Label) that vary in design and the amount of detail provided. Labels implementation may be carried out by government organizations, retail and local grocers, and farmers. Author affiliations: ¹University of New Hampshire, Durham, NH ²Department of Environmental Sciences, University of Virginia

Global trade can allow countries to overcome regional shocks to food supply, but reliance on international food trade exposes a country to higher risks from external perturbations. Countries which are nutritionally- or economically-dependent on international trade of a commodity may therefore be adversely affected by such perturbations. We studied how the global seafood trade network responds to environmental and policy perturbations from two angles: first, by studying how negative local impacts, such as a natural or environmental disaster or the collapse of a regional fishery, propagate through the trade network, and second, by evaluating how trade flows are redistributed under shock scenarios. Shock propagation and distribution among regions are modelled on a network of historical bilateral seafood trade data. Vulnerability to shocks in the network is then assessed by comparing changes in national fish supplies to indices of each country’s economic and nutritional fish dependency. Shocks were found to be absorbed within the network quickly and across all parameters the regions with higher net imports, particularly West Africa, tended to be most exposed within the network. Comparing exposure to sensitivity revealed West Africa and Central Africa to be relatively vulnerable to shocks within the network, with their vulnerability increasing when a GDP effect is included. Author affiliations: ¹Department of Environmental Sciences, University of Virginia ²Department of Mathematics and Mathematical Statistics, Umeå University, Umeå, Sweden ³International Institute of Applied Systems Analysis, Vienna, Austria

As our global population grows and the demand for food increases, there is an increasingly urgent need to conserve land being used for crop growth and cattle grazing. Invasive plant species continue to diminish crop yields and grazing grounds across the US, as much as $100-$137 billion and 700,000 hectares annually (Miao et al. 2006, Wilfong et al. 2009), raising food production costs. Remote sensing can be used to help distinguish these invasive species and manage their spread. Research on making remote sensing methods more accurate and efficient can improve conservation efforts and thus ameliorate the degradation of land being used for food production. Remote has been successfully used to locate plant species and track their spread using time-series imagery. My research focuses on studying six invasive plant species in northern Virginia including two thistles that are said to be two of the most notorious weeds in the continental US (Allen et al. 2006, Tiley et al. 2010). I am working on being able to locate them and assess their effects on the ecosystem using remote sensing at the ground level and at the satellite level. At the ground level, I have collected hyperspectral data from 350 nm to 1025 nm using the FieldSpec 3 field spectrometer. Using principle components analysis to select wavelengths and discriminant analysis with those wavelengths, I have been able to distinguish species using those spectral data; however, discriminability of the two thistle species continues to be a challenge, perhaps due to phylogenetic similarity. Author affiliations: Department of Environmental Sciences, University of Virginia

Seagrasses are considered foundational species because the meadows they form create three-dimensional habitat for an abundance of biota. Several layers of seagrass connectivity contribute to the persistence and resilience of wild fisheries populations. The connectedness between seagrass patches and other habitat types such as coral reefs and mangroves may impact fish species diversity due to the different feeding requirements during various life history stages. Population connectivity, as defined by demographic processes such as immigration of individuals, controls meadow growth and expansion and thereby the extent of available habitat. Genetic connectivity is influenced by rare, long-distance dispersal events and may reflect historical processes more so than contemporary hydrological patterns. This type of connectivity is particularly important because it largely determines genetic diversity, which may have cascading effects on seagrass species diversity and faunal diversity. In my Master’s thesis, I am specifically interested in testing whether genetic diversity and species diversity are correlated in wild populations of the tropical seagrass, S. filiforme in South Florida. Preliminary microsatellite analyses have revealed moderate genotypic diversity ranging from 0.18 to 0.50, and inbreeding coefficients ranging from -0.25 ± 0.04 to -0.04 ± 0.04 in ten populations of S. filiforme sampled in the Florida Keys, indicating there is greater heterozygosity than would be expected under Hardy-Weinberg conditions. Alone, these data suggest S. filiforme populations in South Florida may be experiencing outbreeding, or selection may favor heterozygotes. Further sampling is required to compare genetic diversity of S. filiforme in mixed species stands and mono-specific stands. Both natural and anthropogenic disturbances can eliminate entire seagrass patches, thus disrupting all types of connectivity. One of the many challenges in marine reserve design lies in determining how to effectively maintain habitat, population and genetic connectivity for the maximal number of species. Author affiliations: Department of Environmental Sciences, University of Virginia

Restoration Agriculture represents the establishment of perennial farming systems that mimic the structure and function of natural ecosystems to produce our staple food crops and reestablish ecological integrity. Mark Shepard introduced the term in his book, Restoration Agriculture: Real-World Permaculture for Farmers.¹ Mark’s work at New Forest farm in Wisconsin², where he grows chestnuts, hazelnuts, apples, mushrooms, annual produce, livestock, and other products, has inspired academics and regenerative farming advocates across the country, including Versaland³, Bio-Logical Capital⁴, and the Woody Perennial Polyculture Research Site at the University of Illinois.⁵

Restoration Agriculture Development⁶ is a coalition of ecological farmland design and investment experts. RAD’s mission is to design and install restoration agriculture farming systems by utilizing investor capital, tapping the knowledge of ecological design experts, and reducing barriers for beginning farmers. I am an intern for the RAD team⁷, which includes Mark Shepard, WPP research site lead Kevin Wolz, and holistic management expert Owen Hablutzel. My research and work include tree crop market research, outreach at conferences, and further training in ecological farm design. I met the RAD team as a student in an intensive restoration agriculture design course in Wisconsin this past June (2014).

The purpose of my “ignite” session is to introduce the concept of restoration agriculture. This includes restoration agriculture theory, environmental and economic benefits, producing staple crops in regenerative systems at scale, and my experiences on family land in Pennsylvania. I will also talk about how my work with RAD relates to my studies in land use, food systems, and environmental planning.

Farming in nature’s image can help feed the world and combat climate change. Restoration agriculture has the potential to redefine agriculture, while building economic, social, and environmental resilience in rural landscapes around the globe Author affiliations: Department of Urban and Environmental Planning, University of Virginia.

¹www.amazon.com/Restoration-Agriculture-Mark-Shepard/dp/1601730357
²www.newforestfarm.net/our-story.html
³www.versaland.com
⁴www.biologicalcapital.com
⁵wppresearch.org
⁶radpioneers.com
⁷radpioneers.com/team/

View Presentation
Changing atmospheric composition due to human activities, primarily carbon dioxide (CO2) emissions from fossil fuel burning, is already impacting ocean circulation, biogeochemistry, and ecology, and model projections indicate that observed trends will continue or even accelerate over this century. Elevated atmospheric CO2 alters Earth’s radiative balance, leading to global-scale warming, alterations in the water cycle, and climate change. The ocean stores the majority of resulting anomalous heat, which in turn drives other physical, chemical, and biological impacts. Sea surface warming and increased ocean vertical stratification are projected to reduce global-integrated primary production and export flux as well as to lower subsurface dissolved oxygen concentrations. Upper trophic levels will be affected both directly by warming and indirectly from changes in productivity and expanding low oxygen zones. The ocean also absorbs roughly one-quarter of present-day anthropogenic CO2 emissions. The resulting changes in seawater chemistry, termed ocean acidification, include declining pH and saturation state for calcium carbon minerals that may have widespread impacts on many marine organisms. Climate warming will likely slow ocean CO2 uptake but is not expected to significantly reduce upper ocean acidification. Improving the accuracy of future model projections requires better observational constraints on current rates of ocean change and a better understanding of the mechanisms controlling key physical and biogeochemical processes.