Pace

Leading Indicators of Regime Shifts Regime shifts of an ecosystem are unusually large and rapid changes that lead to a fundamentally different condition. These changes may be irreversible or reversed only at great cost. Based on theory and some observations, regime shifts may be preceded by observable increases in variability of system components. Statistical indicators derived from time series data such as variance, autocorrelation, return time, and heteroscadascity may be useful in detecting regime shifts prior to their occurrence. We are testing this idea with a whole lake experiment where we are creating a regime shift through changes in the fish community. We are assessing variabilty of primary production, ecosystem respiration, phytoplantkon biomass, zooplankton biomass, and small fish abundance using high frequency measurements. Regime shifts are a frontier in ecological research because these changes often cause a loss of ecosystem services. Expanding human pressure on ecosystems makes regimes shift more likely but not more predictable. Our hope is to provide means to identify and where possible avoid regime shifts. Terrestrial Subsidies of Aquatic Food Webs Organic matter supporting consumers ultimately comes from two sources: primary production within an ecosystem (autochthonous sources) or primary production imported from other ecosystems (allochthonous sources). Allochthonous subsidies can decouple consumers from the constraints of internal primary production, thereby altering predator-prey interactions, nutrient cycling, and entire food webs. In lakes it is well known that allochthonous inputs are important to both the carbon balance and whole-system metabolism. The importance of allochthonous inputs as subsidies is less well documented. In prior research, significant progress was made by adding inorganic carbon-13 to augment the isotopic contrast between aquatic and terrestrial sources in order to trace the support of consumers by these materials. To broaden the approach we are using stable isotopes of carbon, nitrogen, and hydrogen to test a general model of how external inputs and internal primary production across a variety of lakes affect variation in the support of consumers by allochthonous materials. Long term response of an ecosystem to an invasive species Human activities have moved thousand of species out of their native ranges, allowing them to establish populations in new parts of the world. These alien species affect native biodiversity, biogeochemical processes, and cause enormous economic damage. Long-term effects of alien species invasions are poorly studied. Yet, common ecological and evolutionary mechanisms suggest the chronic, long-term effects of an invader may be very different from acute, short-term effects. We are part of collaborative study of the effects of the zebra mussel (Dreissena polymorpha) invasion on the Hudson River ecosystem. We have documented extensive, transforming impacts of this species on the Hudson ecosystem since it invaded in 1990-1991. Recent work indicates the mussel population is experiencing high mortality that is altering size structure through the elimination of the largest, oldest individuals. We are evaluating the ecosystem consequences of ongoing changes in zebra mussels providing one of the few continuous studies of the chronic impacts of an alien species. Watershed-Scale Analysis of Nutrient Loading to Adirondack Lake EcosystemsLakes of the northern forest regions of the U.S. vary widely in nutrient concentrations and water quality because of natural variation in the composition and configuration of vegetation in the watershed, and because of variation in inputs due to human activities. Air pollution is the dominant human source of nitrogen loading, and lakeshore development is a primary source of phosphorus loading. We are using a novel, spatially-explicit modeling approach to quantify the watershed controls on nutrient concentrations across a set of more than 300 lakes within New York's Adirondack Park. The research involves sampling the chemistry of each of the lakes, and then using datasets on the topography and vegetation of each watershed to estimate the rates of nutrient input from different vegetation types. The research is examining how watershed land cover influences inputs and how these inputs interact with lake processes to determine nutrient concentrations.