H11K-01
Multiple Stable States in Hydrological Models: An Eco-hydrological Example
Many physical based models of surface and groundwater hydrology are constructed without the possibility of
multiple stable states for the same parameter set because they do not include positive feedbacks. For such a
conceptualization, at the cessation of a transient hydrological disturbance of any magnitude the model will
return to the same stable state, and thus show an infinite resilience.
To highlight and challenge this assumption a numerical distributed eco-hydrological model (coupled hillslope
Boussinesq-vertically lumped vadose zone) was developed in which qualitatively different steady state water
table elevations exist for the same parameter set. The multiple steady states are shown to emerge from a
positive feedback arising from a reduction in leaf area index (LAI) and, thus, transpiration, as a saline water
table approaches the surface. Limit cycle continuation (LCC) was undertaken to quantify the state space
location of the threshold (repellor) between the steady states (attractors) and quantify the resilience.
Time series simulations and LCC were undertaken for two lower boundary conditions: a fixed hydraulic
gradient; and a general head, which simulates the interaction with a river. For both conditions three ranges in
saturated lateral conductivity (ksat) existed in which i) only the shallow; ii) both the shallow and deep;
and iii) only the deep water table attractor exist. For the general head boundary, these three parameter
regions also produced multiple attractors of runoff and baseflow such that, for ksat of 0.5 m day-1,
a change from the shallow to deep attractor resulted in a reduction to baseflow and total flow of 93% and
71% respectively. Such an attractor shift occurs if there is a sufficient low-rainfall disturbance. Upon
cessation of the disturbance, the system has a high probability of remaining in the deep water table basin of
attraction and thus stream and baseflow remain low. While the model is bio-physically simple it is sufficiently
realistic to challenge the implicit assumption of hydrological systems having a single attractor and to
investigate the implications for water resource management.
H11K-02 INVITED
Interacting controls on eco-hydrologic responses to warming in mountain ecosystems
The response of ecosystems to a warmer climate depends upon both the direct impact of temperature on ecophysiological function and indirect effects related to a range of physical and biological processes. In mountainous regions, reduced snow accumulation and earlier melt of seasonal snowpacks are well-know hydrologic responses to climate warming. In the Western US, hydrologic consequences of this change have been shown to include reduced summer streamflow and increased drought. For forests, earlier soil moisture recharge is likely to increase summer drought and reduce late summer plant productivity and evapotranspiration. At the same time, increased air temperature alters plant respiration and growth. Most models of climate change impacts focus either on hydrologic behavior or ecosystem structure or function. In this study we address the interactions between them. We use a coupled model of eco-hydrologic processes to estimate changes in evapotranspiration, vegetation growth and streamflow under a range of temperature warming scenarios for both a snow-dominated site and a semi-arid coastal site in California. Model results suggest that for snow-dominated elevations, the shift in the timing of recharge is the dominant effect, and leads to declines in both productivity and vegetation water use. For the semi-arid coastal site, changes in vegetation productivity are strongly controlled by potential changes in water-use efficiency with higher atmospheric CO2. Results demonstrate that climate-driven changes in vegetation water use may be substantial and can strongly influence streamflow responses. Further, regional and relatively local scale variations in vegetation responses shown in this work emphasize the need for mode-based analysis at finer scales than are commonly used in regional climate modeling. Finally we note that while model results must be interpreted as a best guess, results from this analysis provide a framework that can be used to develop strategic point-based measurement campaigns and a context for scaling them.
H11K-03 INVITED
Resilience to Changing Snow Depth in a Shrubland Ecosystem.
Snowfall is the dominant hydrologic input for high elevations and latitudes of the arid- and semi-arid western United States. Sierra Nevada snowpack provides numerous important services for California, but is vulnerable to anthropogenic forcing of the coupled ocean-atmosphere system. GCM and RCM scenarios envision reduced snowpack and earlier melt under a warmer climate, but how will these changes affect soil and plant water relations and ecosystem processes? And, how resilient will this ecosystem be to short- and long-term forcing of snow depth and melt timing? To address these questions, our experiments utilize large- scale, long-term roadside snow fences to manipulate snow depth and melt timing in eastern California, USA. Interannual snow depth averages 1344 mm with a CV of 48% (April 1, 1928-2008). Snow fences altered snow melt timing by up to 18 days in high-snowfall years, and affected short-term soil moisture pulses less in low- than medium- or high-snowfall years. Sublimation in this arid location accounted for about 2 mol m- 2 of water loss from the snowpack in 2005. Plant water potential increased after the ENSO winter of 2005 and stayed relatively constant for the following three years, even after the low snowfall of winter 2007. Over the long-term, changes in snow depth and melt timing have impacted cover or biomass of Achnatherum thurberianum, Elymus elemoides, and Purshia tridentata. Growth of adult conifers (Pinus jeffreyi and Pi. contorta) was not equally sensitive to snow depth. Thus, complex interactions between snow depth, soil water inputs, physiological processes, and population patterns help drive the resilience of this ecosystem to changes in snow depth and melt timing.
H11K-04
Dynamics and Resilience of Desert Ecosystems under Changing Climate
Ecological models have been used to probe the causes of spatial complexity and predict specific responses of desert ecosystems. However, many models have been limited in their focus: models of dynamics have been developed with no consideration of the inherent patchiness or patterns in the vegetation, or else models have been developed to generate patterns with no consideration of the dynamics. Models that attempt to address both pattern and dynamics have been qualitative and descriptive. Furthermore, if, as is commonly believed, both dynamics and patterns/patchiness are a function of resource (specifically water) limitation, then there has been little integration of this relationship into such models. Consequently, these models have limited utility for understanding resilience of desert ecosystems. Here we present an integrated approach to the observed patchiness and dynamics in desert vegetation that is based on a sound process-based understanding and is formulated to accommodate previous conceptual models within an overarching framework. This framework is implemented as a mathematical model. Our contribution represents an advance over previous work in that we propose a general model framework for the analysis of ecosystem change in deserts that explicitly considers spatial interactions among multiple vegetation types and multiple resources, and predict specific responses to a variety of endogenous and exogenous disturbances. We present an application of this model to investigate conditions the conditions that would result in observed desert vegetation patterns in south-western desert systems of North America. In particular, we investigate the encroachment of shrubs (Larrea tridentata) into formerly pure stands of grass (Bouteloua eriopoda). We present the results of simulations that rest on rainfall data that was reconstructed for the Jornada Basin Long-Term Ecological Research site in southern New Mexico, USA based on 300-year tree- ring records. The results show that populations of native grasslands were stable under all historical rainfall conditions and should have remained stable over the last century. The observed encroachment of shrubs can only be explained by the combination of periodic droughts with a second disturbance, which in this work is simulated as varying grazing intensities. We also show that the spatial pattern of the invading species is a function of the behaviour of the previously dominant species in redistributing resource and propagules during times of resource stress.
H11K-05 INVITED
Impacts of climate variability on functional diversity and species distribution in dryland ecosystems
The role rainfall heterogeneity is an important but poorly understood aspect of ecosystem function in dryland communities. While general relationships between annual rainfall and tree cover have been revealed through meta-analysis and inter-site comparisons, the impact of rainfall variability on both functional and species-level biodiversity has received less attention. However, it is likely that shifts in the seasonality, and distribution of rainfall (i.e. changes in daily storm frequency and/or storm intensity) could substantially alter waterbalance of dryland ecosystems. In particular, the distribution of plant water use and plant water stress (proxies for growth and survival, respectively) are highly dependent on the temporal signatures of rainfall events within and between growing seasons. Therefore, ecosystem resilience is likely to be dependent on the functional - and species-level - responses of ecosystems to shifts in rainfall timing and intensity. This presentation will explore the potential and observed impacts of both seasonal and inter-annual variability in rainfall distribution in two classic dryland ecosystems. The first is the Kalahari region of southern Africa, where seasonal variability in rainfall patterns leads to functional organization of trees and grasses across a large climate gradient that optimize water use and minimize water stress. The second example is drawn from an east African savanna in Kenya, where observed shifts in rainfall seasonality are used to infer changes in the spatial distributions in the dominant woody species. Finally, I will address the critical need for new observational frameworks capable of directly measuring plant and ecosystem-scale responses to rainfall heterogeneity and propose a sampling design capable of resolving the functional response of vegetation to rainfall heterogeneity in water-limited landscapes.
H11K-06
Global change enhances vegetation vulnerability to drought: Warmer drought kills pinyon pines faster.
The resilience or vulnerability of dominant, overstory vegetation to climate change is of great interest in understanding potential feedbacks from vegetation losses to ecosystem function and the global carbon budget. Many dryland regions of the world face the potential loss of overstory plants from a combination of warmer temperatures and increases in the frequency and severity of drought events. In the Southwest US, recent, unprecedented, widespread drought-induced die-off of pinyon pine (Pinus edulis) could be a harbinger of global change impacts in other dryland ecosystems. Research revealed that this drought was slightly warmer than past drought events, for which tree mortality was not widespread. We tested the hypothesis that warmer drought kills trees faster in an experimental setting. Using the environmental controls of the Biosphere 2 facility in Oracle, AZ, we placed transplanted pinyon pines in two areas of the glasshouse, one with temperatures close to ambient for the species, the other elevated by 4.3° C over ambient. We simulated a severe drought on half of the trees in each area by restricting all precipitation or irrigation until they died. Drought trees in the elevated temperature treatment died nearly 30 percent faster than those in the ambient treatment, a disproportionate response of drought mortality to temperature. Given expected future increases in temperatures and variability of precipitation for pinyon-dominated ecosystems, our results imply a high degree of vulnerability of these ecosystems to global change that will impact their function and the regional carbon budget. Understanding the mechanisms of how trees die from drought will improve our understanding of the links between processes of vegetation dynamics, hydrological function, and feedbacks to the climate system.
H11K-07 INVITED
Ecosystem Structure and Survival
Ecosystem survival through chronic and/or episodic water stress requires not only physiological adaptations by the vegetation to the water stress but also an accessible store of water removed from environmental evaporative demand. The existence of such a store at depth is corroborated by recent observations from many biomes, including the seasonally-dry Amazon, which show enhanced productivity during the dry season thanks to hydraulic redistribution of soil water resources by deep-rooted trees. Additionally, we present new hydrologic observations from a small, steep, temperate forested watershed along the Northern Californian coast. The differences between upslope and downslope hydrology and ecosystem dynamics suggest that ecosystem resilience may be dependent on diverse functional and structural attributes of plants, which are species dependent. These plant attributes may, in addition, reflect plant adaptation to, and plant modification of, the local hydrologic setting that feeds back on other biogeochemical cycles and even on climate.
H11K-08
Regional watershed discharge patterns in Southeast Alaska: implications of climate change
The Tongass National Forest in Southeast Alaska encompasses 69,000 km2 of forest, glacier and rock. The Tongass is a coastal temperate rainforest characterized by moderate cool summer temperatures and high but variable (0.8-8 m y-1) precipitation. Annual discharge from the Tongass is about 207 km3, comparable to the Yukon or Columbia Rivers. Comparing modeled precipitation versus USGS discharge records from gages throughout the Tongass reveals three distinct annual hydrograph types: I) a fall/winter discharge maximum with daily discharge that closely matches precipitation; II) a bimodal spring snow melt/autumnal rain driven system and; III) a winter low/summer high discharge glacier/snowfield melt system. Because of differences in basin morphology and relationship to average winter snow line, watersheds vary in their vulnerability to potential climate change effects. Climate models in this highly variable, rugged terrain are imprecise, but predicted increases in temperature, and consequent predicted increases in the mean elevation of the snow line put much of southeast Alaska below the snow line. We hypothesize that hydrologic responses to future increases in snow line elevation will be driven by watershed morphology. In general, Type I watersheds have maximum elevations below 1,500 m, Type II systems have maximum elevations between 1500 and 3200 m and Type III watersheds have maximum elevations >3,200 m and typically >3% glacier coverage. Twenty percent of total regional discharge is from Type I watersheds, with Type II and III watersheds providing 46 and 34 percent of the total annual discharge respectively. As the snow line increases in elevation and storage of precipitation in snow and ice declines, watersheds will transition from Type III to Type II to Type I hydrographs. The resulting redistribution of discharge seasonally will alter fundamental channel-forming processes such as flow magnitude, return frequency and depth of scour. These changes in channel morphology, seasonal instream flow and water temperature may have dramatic impacts on salmon spawning and rearing habitats with presently unknown impacts on reproductive success and productivity. At present, the three watershed types also vary in the seasonal pattern of carbon, nitrogen and phosphorus flux through freshwater habitats and into coastal estuaries. Changes in biogeochemistry wrought by climate change, for example the cessation of summer phosphorus inputs from elimination of glacier scouring, will have large impacts on river habitat and estuarine productivity. Such alterations in freshwater salmon habitat, combined with predicted changes in ocean conditions affecting the marine portion of salmonid life history must be considered when predicting the impact of climate change on one of the last remaining healthy salmon fisheries