Oral Session Presentations

Coastal Stream Ecosystem Metabolism: A Comparison of the Conventional Open Water Technique and Eddy Correlation

Koopmans, Dirk* and Peter Berg

Inland waters transform or store at least half of terrestrial net ecosystem production , making metabolism in these waters significant in the global carbon budget. However, the conventional technique for measuring benthic metabolism in these systems relies on a prediction of gas exchange across the air-water interface, increasing uncertainty. To reduce uncertainty we have combined the conventional technique with a new technique that does not suffer from this limitation, eddy correlation, which measures benthic oxygen flux under in situ hydrodynamic conditions and at high temporal resolution. This talk will present parallel seasonal measurements made with the two techniques in a stream at the Virginia Coast Reserve LTER. The conventional technique provided multi-day measurements during which the eddy correlation technique was used to measure fluxes over shorter intervals, allowing direct comparisons between the two. The combination of techniques is complementary, with eddy correlation ground-truthing the longer term measurements made with the conventional technique.

Keywords: stream, river, metabolism, open water, eddy correlation, VCR-LTER

Conditional heteroskedasticity warns of impending regime shift in a whole-ecosystem experiment

David A. Seekell*

Abstract: Regime shifts in stochastic ecosystem models are often preceded by early warning signals such as increased variance and increased autocorrelation in time series. There is considerable theoretical support for early warning signals, but there is a critical lack of field observations to test the efficacy of early warning signals at spatial and temporal scales relevant for ecosystem management. Conditional heteroskedasticity is persistent periods of high and low variance that may be a powerful leading indicator of regime shift. We evaluated conditional heteroskedasticity as an early warning indicator by applying moving-window conditional heteroskedasticity tests to time series derived from a whole-lake experiment designed to create a regime shift. There was significant conditional heteroskedasticity at least a year prior to the regime shift in the manipulated lake but there was no significant conditional heteroskedasticity in an adjacent reference lake. Conditional heteroskedasticity was an effective leading indicator of regime shift for the ecosystem manipulation.

Keywords: regime shift; early warning; ecosystem experiment; conditional heteroskedasticity


Jonathan A. Walter*, Ottar N. Bjørnstad1, and Kyle J. Haynes

[1] Pennsylvania State University Department of Entomology

An important question in biotic invasions is why invaders spread more rapidly through certain environments than others.  A number of challenges make this question difficult to answer; monitoring spread over time and space is resource-intensive and may be unfeasible, and conventional methods of calculating spread rate from available data are not suitable for estimating spread at local scales where variability in invasion rate may occur.  This presentation introduces local polynomial fitting as a method for deriving local estimates of invasion speed and direction.  Using the gypsy moth as a model system, we demonstrate the utility of this method for detecting geographic variability in invasion rate and use geographically weighted regression (GWR) to explore relationships between environmental variables and invasion rate.  Measures of model fit indicate that the combination of human population density and forest fragmentation (the cohesion index) provide the best predictions of gypsy moth invasion rate.

Impacts of gypsy moth defoliation and Bacillus thuringiensis on Coleopteran diversity in Shenandoah National Park, Virginia

Hilary Wayland*, Rea Manderino, Kyle J. Haynes

The species richness, abundance, and diversity of forest Coleoptera was assessed at 15 sites in Shenandoah National Park. Sites belonged to one of three defoliation histories: defoliated, defoliated and treated with Bacillus Thuringiensis var. kurstaki (Btk), and undefoliated and untreated (control). Beetles were captured during two trapping sessions using blacklight traps and were sorted to family and morphospecies. Individuals were assigned into feeding guilds based on the dominant feeding behavior of their family. Understory plant richness, abundance, and diversity were measured along transects at each site. Beetle species richness and abundance were negatively impacted by defoliation. These effects were evident even two years since the last defoliation. Control sites had the greatest number of individuals and species, as well as the highest plant richness. Abundance and richness of beetles at Btk-treated sites fell midway between defoliated and undefoliated, suggesting that Btk was somewhat effective at reducing the impacts of defoliation on non-target insects. A significant relationship between understory plant and beetle species richness suggested that low plant richness may be partially responsible for low beetle species richness at defoliated sites. The abundances of beetles in dominant feeding guilds (Herbivore, Predator, Fungivore, and Xylophage) did not significantly vary among defoliation histories, except xylophages which were lowest at Btk-treated locations. The effect of Btk on xylophage abundance may be explained by indirect interactions between forest Lepidoptera and Coleopteran xylophages. Overall, gypsy moth defoliation had a strongly adverse effect on forest Coleoptera and should be considered a serious threat to forest biodiversity. 

Keywords: beetle, Lymantria dispar, conservation


David M. Hondula*and Robert E. Davis 

High temperature and humidity conditions each summer lead to elevated mortality rates across the United States and other mid-latitude climates. Our research focuses on the spatial and temporal aspects of the health response within large metropolitan areas to identify places and populations that are at greater risk when the apparent temperature is dangerously high. We integrate 25 years of daily mortality records from seven U.S. cities with a suite of sociodemographic and environmental variables to determine the specific factors that contribute to the geographic distribution of heat-related deaths.  

We identify several postal codes within each of the study cities associated with mortality rates significantly above baseline levels when the summertime temperatures exceed a critical threshold. Heat-sensitive locales are generally associated with low income, high building density, and a higher percentage of elderly residents. Regression tree predictions are significantly correlated with observed mortality rates. Model performance did not improve with the addition of an exposure variable derived from remote satellite imagery of the urban heat island. 
Current heat alert systems provide warning of the onset of dangerous conditions for large geographical areas. Our results suggest that specific locations within cities are at greater risk when such conditions are forecast. Local emergency management agencies could adapt more localized intervention and mitigation measures to reduce the health burden of extreme summertime weather.

KeywordsBiometeorology, heat waves, extreme temperature, climatology

Mortality Displacement - A Characterization of Lagged Mortality Response for Heat Events in Boston and Philadelphia

Michael Saha* 

Heat waves cause significant number of deaths each year. It is difficult to assess the sum mortality effect of a single heat event due to the complexity of lag effects, namely mortality displacement. The goal of this study is to develop a metric that characterizes this phenomenon, and see how it varies with time of year and the strength of the event. The mortality displacement ratio (MDR) is the proportion of deaths above average during a heat event that are accounted for in the following negative mortality anomaly. The MDR was found to be 21% for Philadelphia and 32% for Boston. In Philadelphia there was a strong negative trend in mortality displacement as events got stronger, confirming the results of previous studies. Further, the MDR was found to be significantly higher in the later season than the early season for both cities. This difference was mainly in the lagged displacement period, with the mortality spikes generally similar in magnitude for different events. These findings indicate that there are more “new” deaths earlier in the season and with stronger events. Though heat warning systems already consider the severity of higher temperatures, they should also consider early season events to maximize life-time preserved by minimizing “new” deaths.

Keywords: Heat waves, Human Mortality, Seasonality, Apparent temperature

Looking for Saharan Dust in the Forests of the Yucatan: A Multi-scale Approach

Rishiraj Das*, Deborah Lawrence, Amato Evan, Julie Pett-Ridge [1] and Natalie Mahowald [2]

 [1] Crop and Soil Sciences, Oregon State University;  [2] Department of Earth and Atmospheric Sciences, Cornell University

 Atmospheric phosphorus (P) inputs are rarely considered in models of terrestrial P cycling, but may be critical in balancing losses of P from ecosystems over the long term, especially in tropical dry forests. Previous work suggest that long-distance dust transport may be an important source of P to forests in the Amazon basin, Caribbean and Hawaiian islands over the long term, but relatively few studies combine field measurements, remote sensing and transport modeling to evaluate the importance of dust as a source of atmospheric P inputs. Here we quantify the contribution of long-distance dust transport to atmospheric P inputs to tropical dry forests in the Yucatan peninsula by comparing measurements of total P in field collections of atmospheric bulk deposition with MODIS AOD remote sensing and MATCH atmospheric transport modeling simulations of dust transport and deposition. Long-distance dust transport is relatively smaller than local sources of atmospheric P inputs, and contributes 55±15 g P/ha/yr, which is approximately 20% of total atmospheric P inputs measured in bulk deposition. Local biomass burning and biogenic inputs are larger sources of atmospheric P. There is distinct seasonal variation in atmospheric P inputs, with biomass burning and biogenic inputs during March-May, dust transport during June-August, and other local inputs between September and November. Interannual variation is influenced by variation in biomass burning and a 2-3 fold variation in dust inputs. Even though they are relatively small, long-distance dust inputs are large enough to offset leaching losses of P and are important in maintaining fertility and productivity over the long term. 


Gerald V. Frost*, Howard E. Epstein, Donald A. Walker [1], and Georgiy Matyshak [2]

[1] University of Alaska, Institute of Arctic Biology, Fairbanks, AK, USA, [2] Lomonosov Moscow State University, Department of Soil Science, Moscow, Russia

We integrated field data and remote-sensing to identify key environmental drivers of tall shrub expansion in a tundra ecotone near Kharp, northwest Siberia. Comparison of high-resolution Corona and QuickBird satellite imagery shows that alder (Alnus viridis ssp. fruticosa) cover increased by ~10% since 1968. In the field, we found that most shrub expansion occurred in patterned-ground and virtually all recently-established shrubs occur on small, mineral-dominated microsites (“circles”) that are annually disturbed by differential frost-heave.

In order to test if shrubland development is facilitated by circles, we established a series of transects in shrublands in varying stages of development. We mapped the location of alders and circles and measured surface organic depth and mineral horizon thickness systematically along the transects, and at alders. In recent colonization zones, young alders occur almost exclusively on circles that lack vegetation and surface organic matter. Spatial statistics indicate that most alders occur in evenly-spaced clumps that mirror the spacing of circles; this distribution pattern persists in mature shrublands where circles are large and widely-spaced. However, dense shrub thickets develop where circles are close-packed.

We conclude that circles strongly facilitate alder recruitment at Kharp, likely because soils are relatively warm and lack competing vegetation. Physical processes in patterned-ground promote continuous, rather than episodic, disturbance that maintains a mosaic of seedbeds that promote development of fast-growing, minerotrophic shrubs during periods of favorable climate. Modern high-resolution remote-sensing data indicates that alder proliferation in patterned-ground has promoted persistent changes in land-cover across a broad swath of the northwest Siberian Low Arctic in recent decades. These changes have broad implications for a range of biophysical system components, including surface energy balance, ground temperatures, disturbance regime, and permafrost stability.

Keywords: shrub expansion, patterned-ground, facilitation, Low Arctic. 

The impact of mountaintop carbon dioxide mixing ratios on inversion model estimates of carbon exchange over North America

Temple R. Lee*, Stephan F. J. De Wekker, Bjorn Brooks [1]

[1] Department of Plant Biology; University of Illinois; Urbana, IL

 Large uncertainties exist in the North American carbon budget.  Constraining surface fluxes requires inverse models that use CO2 mixing ratio measurements from a network of tall towers and, more recently, mountaintops.  Because the network of monitoring stations is sparse, there are large sensitivities to measurements from any one station.  In this study, we investigate the sensitivity of an inverse model, CarbonTracker, to measurements from Pinnacles, a new mountaintop CO2 monitoring station in the eastern US.  We perform CarbonTracker simulations for two one-month periods (December, 2008, and July, 2009) with and without the assimilation of CO2 mixing ratio measurements from Pinnacles and estimate the concentration footprint of Pinnacles using a trajectory model.  We find that the impact of Pinnacles on continental CO2 measurements is 4.78 Tg CO2 month-1 in December, with the largest impacts in the Appalachian Mountains and Southeast US.  The impact of Pinnacles on continental fluxes is larger in July and shifts northward to capturing CO2 uptake in the Great Plains and Upper Midwest.  These results show the significance of one CO2 monitoring site to regional- to continental-wide estimates of CO2 flux and show how the impact varies in seasonal extremes.

 Keywords:  Carbon dioxide, carbon budget, Blue Ridge Mountains

Characterization (δ13C & TOC) of marine sediments from restored seagrass (Zostera marina) meadows in coastal lagoons of Virginia.

Noah E. Egge* & Stephen A. Macko

Restoration of the Chesapeake Bay and its neighboring coastal areas has been a major focus of scientific research in the last ten years. The decline of the area is often linked with the explosive population growth and subsequent increases in agriculture and urbanization of the region. In the coastal bays of Virginia, large efforts have been made to restore seagrass (Zostera marina) meadows, which were devastated by a combination of events in the 1930s. As a result of the restoration efforts begun in 2001, seagrass has come to dominate portions of South Bay and Hog Island Bay.

Sedimentary characterization (δ13C and total organic carbon) is used to assess sources of organic matter in marine environments. Push cores were used to collect samples from sites that have different records of seagrass cover. Sediments dominated by seagrass contain organic matter enriched in 13C relative to autochthonous algal sources. Within the sediment record there are two regions that show heightened 13C content—one near the surface and another at lower depths. Organic matter enriched in 13C near the surface is from current seagrass cover. Similar characteristics of sediment deeper in the core imply that the area was previously dominated comparable vegetation. Results may be used as a proxy to infer the spatial extent of the historic (prior to 1930) presence of seagrass and guide future restoration projects.

Keywords: seagrass, carbon isotopes, organic matter, Delmarva coastal bays

Simulating tundra plant community dynamics in response to climate change and grazing pressures at regional and pan-arctic scales

 Qin Yu *, Howard E Epstein , Donald A Walker [1], Gerald V Frost, and Bruce C Forbes [2]

[1] Institute of Arctic Biology, University of Alaska, Fairbanks, [2]Arctic Center, University of Lapland

Understanding the responses of the arctic tundra biome to a changing climate requires knowledge of the complex interactions among the climate, soils and biological system. This study investigates the individual and interaction effects of climate change and reindeer grazing across a variety of climate zones and soil texture types on tundra vegetation community dynamics at the regional and pan-arctic scales using an arctic vegetation model.  The model incorporates the reindeer diet, where grazing is a function of both foliar nitrogen concentration and reindeer forage preference. At the regional scale, we found that grazing is important, in addition to the latitudinal climate gradient, in controlling tundra plant community composition, explaining about 13% of the total variance in model simulations for all arctic tundra subzones. The decrease in biomass of lichen, deciduous shrub and graminoid plant functional types caused by grazing is potentially dampened by climate warming. Moss biomass had a nonlinear response to increased grazing intensity, and such responses were stronger when warming was present. Our results suggest that evergreen shrubs may benefit from increased grazing intensity due to their low palatability, yet a growth rate sensitivity analysis suggests that changes in nutrient uptake rates may result in different shrub responses to grazing pressure. Heavy grazing caused plant communities to shift from shrub tundra toward moss, graminoid-dominated tundra in subzones C and D when evergreen shrub growth rates were decreased in the model. At the pan-arctic scale, simulation results of warming only and grazing only scenarios were compared. Greater magnitude of warming caused greater biomass increase in the Arctic, particularly in southern subzones where shrubs can likely respond to warming more quickly than other species. Biomass decrease caused by grazing was correlated with reindeer population density (e.g. Russian Arctic, Norwegian Arctic).

Our results suggest that warming, grazing, and soils can all have substantial effects on tundra vegetation. However, heavy grazing and trampling are becoming issues with increased reindeer herds in the Russian Arctic, and the response of vegetation to these disturbances remains unknown. Additionally, the interaction of warming and grazing can have greater impacts on some species than others.  Deciduous shrubs for example, can respond to warming quickly, but also contribute largely to the reindeer diet during the growing season.

Keywords: tundra plant communities, plant functional types, climate change, grazing pressure, interaction effects


Long, M.H.*, D. Koopmans, P. Berg, S. Rysgaard [1,2], R.N. Glud [1,3,4] and D.H. Søgaard [1,4]

[1] Geenland Climate Research Centre, Greenland Institute of Natural Resources, Nuuk, Greenland, [2] Centre for Earth Observation Science, CHR Faculty of Environment Earth and Resources, University of Manitoba, Winnipeg, Canada, [3] Scottish Association of Marine Sciences, Oban, Scotland, UK, [4] Southern Danish University and NordCee, Odense M, Denmark

This study uses the eddy correlation technique to examine fluxes across the ice-water interface.  Temperature eddy correlation systems were used to determine rates of ice melting and freezing, and O2 eddy correlation systems were used to examine O2 exchange rates as driven by biological and physical processes.  The research was conducted below  0.7 m thick sea ice in mid-March 2010 in a southwest Greenland fjord and revealed low average rates of ice melt amounting to a maximum of 0.80 ± 0.09 mm d-1 (SE, n = 31). The corresponding calculated O2 flux associated with release of O2 depleted melt water was less than 13% of the average daily O2 respiration rate.  Ice melt and insufficient vertical turbulent mixing due to low current velocities caused periodic stratification immediately below the ice.  This prevented the determination of fluxes during certain time periods, amounting to 66% of total deployment time. The identification of these conditions was evaluated by examining the velocity and the linearity and stability of the cumulative flux. The examination of unstratified conditions through velocity and O2 spectra and their cospectra revealed characteristic fingerprints of well-developed turbulence.  From the observed O2 fluxes, a photosynthesis/irradiance curve was established by least-square fitting. This relation showed that light limitation of net photosynthesis began at 4.2 μmol photons m-2 s-1, and that the algal communities were well-adapted to low-light conditions as they were light saturated for 75% of the day during this early spring period.  However, the sea ice associated microbial and algal community was net heterotrophic with a daily gross primary production of 0.69 ± 0.02 mmol O2 m-2 d-1 (SE, n = 4) and a respiration rate of -2.13  mmol O2 m-2 d-1 (no SE, see text for details) leading to a net primary production of -1.45 ± 0.02 mmol O2 m-2 d-1 (SE, n = 4).  Modeling the observed fluxes allowed for the calculation of fluxes during time periods when no O2 fluxes were extracted.  This application of the eddy correlation technique produced high temporal resolution O2 fluxes and ice melt rates that were measured without disturbing the environmental conditions while integrating over a large area of approximately 50 m2 which encompassed the highly variable activity and spatial distributions of sea ice algal communities.