Grantee Research Project Results
2009 Progress Report: Hydrologic Thresholds for Biodiversity in Semiarid Riparian Ecosystems: Importance of Climate Change and Variability
EPA Grant Number: R833025Title: Hydrologic Thresholds for Biodiversity in Semiarid Riparian Ecosystems: Importance of Climate Change and Variability
Investigators: Dixon, Mark A. , Hogan, James F. , Lite, S. Joy , Stromberg, Julie , Meixner, Thomas
Current Investigators: Meixner, Thomas , Dixon, Mark A. , Hogan, James F. , Lite, S. Joy , Stromberg, Julie , Baird, Kate
Institution: University of Arizona , Arizona State University
Current Institution: University of Arizona
EPA Project Officer: Packard, Benjamin H
Project Period: March 1, 2008 through February 28, 2011
Project Period Covered by this Report: July 1, 2008 through June 30,2009
Project Amount: $841,881
RFA: Nonlinear Responses to Global Change in Linked Aquatic and Terrestrial Ecosystems and Effects of Multiple Factors on Terrestrial Ecosystems: A Joint Research Solicitation- EPA, DOE (2005) RFA Text | Recipients Lists
Research Category: Climate Change , Aquatic Ecosystems
Objective:
Problem Statement: Riparian ecosystems of the arid and semiarid Southwest are linear corridors of high productivity and diversity. These ecosystems are sensitive to even small changes in the riparian water balance, with sharp changes in vegetation as streams become intermittent and as groundwater declines below survivorship thresholds. As a result, riparian vegetation has declined on many rivers due to water abstraction or has been altered due to the hydrologic impacts of climate variability. Despite much disciplinary work on individual rivers, a regionally comprehensive and integrated understanding of how aquatic-terrestrial ecotones respond to hydrologic change, including those imposed by climate change, awaits development.
Objectives: We will determine region-wide sensitivity of riparian vegetation to climate change. Our hypotheses:
i) Decadal scale climate change and variability alter riparian aquifer recharge through mechanisms that depend on the magnitude, frequency and seasonality of flooding, and exert the greatest change in reaches that receive minimal groundwater inflow from the regional aquifer.
ii) Riparian vegetation structure responds non-linearly as riparian aquifers are dewatered and as key hydrologic thresholds for survivorship of plant species are exceeded.
iii) Decadal scale climate variability and change alters riparian ecosystem water budgets that in turn changes vegetation structure and function and the ecosystem services provided to society.
Progress Summary:
Hypothesis 1 - Decadal scale climate change and variability alter riparian aquifer recharge through mechanisms that depend on the magnitude, frequency and seasonality of flooding, and exert the greatest change in reaches that receive minimal groundwater inflow from the regional aquifer.
Extending previous research results from the San Pedro to the Bill Williams and Hassayampa river systems has begun The aridity of the Bill Williams River system has allowed us to study flood recharge processes in relative isolation of basin scale groundwater processes. Determining the source of baseflow in the lower Bill Williams/NWR, and the residence time of this water in the Planet Valley aquifer, will provide insight into the dependence of streamflow on earlier recharge-inducing floods and the timescale over which large floods influence river system hydrology and ultimately vegetation ecology. Isotopic (d18O, d2H) analysis indicates that groundwater in Planet Valley (the main baseflow source to the lower Bill Williams) ranges from recent recharge (i.e. similar to current outflows from Alamo Dam at the upstream end) to old (>10 years at a minimum or possible predating dam closure in 1969) floodwater recharge at the downstream end. Measurements following two experimental floods on the river have not caused an appreciable change in flow in the river. This result is mirrored in chemical composition data that shows that to date there has been no appreciable change in river composition. The 2009 experimental flood was larger than the ones in 2007 and 2008 but samples have not yet been analyzed from this most recent flood.
On the Hassayampa river we have installed two new monitoring wells along the floodplain that, along with one preexisting well, have been equipped with water level/ conductivity/ temperature sensors. Groundwater samples are also being collected from these wells (approximately every two months) as well as more frequent (monthly to weekly depending on river flow condition) and numerous river water samples. To date there have been no floods large enough to induce a physical response or alter the underlying chemical composition of the groundwater system.
Additionally statistical classification of streamflows in the San Pedro, Bill Williams, Hassayampa and other Arizona River systems has been pursued. Initial analysis indicates the approach of Sanz and del Jalon (2005) will be successful at understanding the variability among these rivers across the three defined seasons of Monsoon (July-September), Winter (October-March) and pre-monsoon (April-June). As expected the rivers are widely divergent in their flow characteristics with the San Pedro having very consistent and pronounced high flows during the monsoon season while the Bill Williams is dominated by winter floods with the Hassayampa having few floods in both seasons.
The two years of the project saw the completion of a simple groundwater surface water interaction model for the San Pedro River system. This modeling furthered coupled understanding of stream-aquifer interactions in flood and baseflow conditions. The model consisted of a simple daily time-step stream-aquifer exchange model that was parameterized using observed groundwater levels, and daily river discharge data. The model competently simulated overall riparian system hydrologic behavior; more significantly the model matched chemical/isotopic of river and groundwater data at a number of locations. This study showed that a simple, spatially-discretized lumped parameter model can be used to simulate the impacts of short time-scale floods on seasonal and inter-annual time-scale hydrologic conditions (Simpson, 2007). Further efforts to link this simple surface water model to the groundwater model of the basin will focus on the use of response functions to link the two models. Simple scenario results with this model demonstrate the importance of flood flows to sustaining streamflow for longer periods of time on the more ephemeral reaches of the river.
Our efforts to develop a seasonal groundwater model of the upper San Pedro basin have currently focused on better estimating seasonal and inter-annual variability of 1) groundwater recharge from mountains that bound the groundwater basin (mountain system recharge) and 2) the groundwater – surface water exchange processes. For the controls on mountain system recharge the overall objective is to quantify seasonal mountain system recharge in semi-arid catchments by analyzing different components of water balance. First, a water balance model was developed for the Garden Canyon catchment using the Soil and Water Assessment Tool (SWAT2005) and forty six years of meteorological data. The result showed annual variability of MSR and threshold response of MSR with respect to seasonal precipitation and soil moisture in the catchment. We have also pursued using dynamically downscaled forecasts of precipitation and temperature to understand how recharge might change under a warmer climate. This modelling has only recently been completed and analyses are not yet complete.
To examine the adequacy of SWAT in modeling hydrologic response in mountainous catchments, a dataset from the Marshall Gulch catchment was used to construct a SWAT model. Availability of stream flow and soil moisture data enabled us to evaluate model performance in soil water partitioning. Although calibration has enhanced model performance for stream flow prediction (NSE = 0.77), the model is still behaves poorly in estimating soil water content (NSE = 0.04). Poor behavior in estimating soil water content is related to soil water storage in second layer of the model, and over-simplified structure of SWAT in soil water partitioning. The next step in our modeling exercise is to apply a physically based model of coupled soil moisture –hillslope Boussinesq equation developed by Troch and others (2003). We are also pursuing using the method of Kirchner (2009) to understand how changes in streamflow can be used to understand change sin recharge at the mountain block scale
We are also focused on developing the data sets needed to utilize the RIP-ET package in a MODFLOW groundwater model. This work includes georeferencing and classification of historical air photos of the San Pedro basin to have solid information on vegetations current location and its changes in the recent past. Additional data on groundwater depth ET relationships is also being gathered.
Hypothesis 2- Riparian vegetation structure responds non-linearly as riparian aquifers are dewatered and as key hydrologic thresholds for survivorship of plant species are exceeded.
2A. Threshold #1: Flow permanence and decline of wetland herbaceous plants.
To anticipate effects of changes in stream hydrology on riparian vegetation, we are using multiple approaches including modeling efforts, controlled experiments, and extrapolation of future temporal dynamics from existing temporal and spatial patterns. By assessing the response of vegetation to the spatial hydrologic gradients that exist along dryland rivers, one can project potential vegetation changes resulting from temporal changes in stream hydrology. Relationships between stream flow permanence and low-flow channel vegetation have been previously described for one spatially intermittent river in a semiarid portion of Southwestern USA (the San Pedro) (Stromberg, et al., 2005). To determine how regionally robust these relationships are, and to determine how patterns vary between rivers with somewhat different climatic signals, we are collecting data at three other spatially intermittent rivers (Hassayampa River, Cienega Creek, Santa Cruz River). The Hassayampa River and Cienega Creek, like the San Pedro, are spatially intermittent owing to underlying variations in geology. Short stretches with perennial stream flow grade into those with ephemeral flow. The Santa Cruz is a river that has been dewatered by excessive ground water pumping, and perennial flow is maintained today at in some sections by discharge of effluent from wastewater treatment plants.
Preliminary analysis has been initiated to relate cover and richness of plant groups to various measures of stream flow permanence and to contrast vegetation-hydrology relationships among study rivers. Contrasts reveal a fairly high degree of similarity between rivers in response of key plant functional groups for data collected in the pre-monsoon dry season. This multi-river analysis supports the idea of a threshold response to change for a key functional group- herbaceous wetland perennials. This group of plants declines sharply in abundance as stream flow becomes non-perennial. This group plays a key functional role in riparian ecosystems by serving to stabilize channels and provide cover for aquatic animals.
For all rivers, strong relationships are evident between low-flow channel vegetation traits and stream flow permanence for the pre-monsoon sampling time. However, given the differences in strength of the summer monsoon between rivers (i.e., several of the rivers occur within a portion of the Southwest that receives large amounts of late summer (monsoon) precipitation), the post-monsoon patterns differ sharply. During late summer at the San Pedro, the strong summer monsoon flood pulse equalizes water availability among sites classified as ephemeral, intermittent and perennial, and thus equalizes many traits of the herbaceous riparian plant community. Such a pattern was not evident at the Hassayampa.
2B. Threshold #2: Groundwater depth and decline of shallow-rooted and deep-rooted obligate riparian herbaceous plants.
Our original intent was to address the response of riparian plants to changes in water table depth primarily through field observations but we have decided to focus more heavily on an experimentally controlled approach. Two greenhouse experiments are being implemented at Arizona State University, one to determine rooting depth and root growth rate of a suite of riparian species and another to investigate interactions of water table depth, surface water duration, and nutrient availability on plant community response.
One experiment, to be initiated in 2009, will determine rooting depth and root growth rate of a suite of herbaceous and woody riparian plants. The other experiment was initiated in August of 2008. This experiment investigates interactions between water table depth, surface watering duration, and nutrient content in a fully balanced design. Rather than selecting particular plant species to investigate, soil and litter samples were collected from riparian forest understory at the Hassayampa River collected to capture the soil seed bank plant community.
2C. Threshold #3: Groundwater depth and decline of shallow-rooted and deep-rooted pioneer trees
Data for this portion of the study are being collected at five study rivers (Cienega Creek, and Hassayampa, Verde, Bill Williams and Santa Maria rivers). The intent is to quantify relationships between water table traits (depth to water table, degree of seasonal fluctuation) and the abundance of various riparian tree and shrub species, and to determine how regionally robust such relationships are. Relationships between water table conditions and key riparian tree species have been previously described for the San Pedro River (Lite and Stromberg 2005). Cienega Creek and Verde River, although not initially originally identified as study rivers for this proposal, have been included because of the availability of hydrology data and/or study site access being provided by local managers.
Hypothesis 3: Decadal scale climate variability and change can alter riparian ecosystem water budgets that in turn change vegetation structure and function as well as the services provided to society by these ecosystems.
We have made less progress on hypothesis 3 of this project as was expected since this hypothesis is the integration of the first 2 hypotheses. Still the work we have completed in hypothesis 1 and 2 has laid a solid foundation for coupling the various products we foresee needing to complete the plan outlined in the proposal.
3A. Climate-Vegetation Change Scenarios
The streamflow classification work from hypothesis 1 has proceeded and we have classified our three flow season Monsoon (July-September), winter (October-March) and Pre-Monsoon (April-June). We have pursued preliminary scenario analysis with years classified as wet, dry and normal and seeing how the systems respond to these changes.
3B. Floods, groundwater and recruitment of hydromesic pioneer trees and shrubs.
We have made substantial progress on the data (hydrologic scenarios, vegetation data) needed for parameterizing bio-hydrologic models to simulate vegetation responses to hydrologic change. Work on this component will be continued during 2010 by Dr. Mark Dixon, on a subcontract between ASU and the University of South Dakota. Dr. Dixon will build off of his previous work on riparian vegetation dynamics models, including models representing ‘recruitment box’ relationships between flow pattern and woody vegetation establishment on the San Pedro River, to develop models that will couple with the hydrologic scenarios developed under other components of this project (e.g., 3A). We will use size and age structure data for Populus fremontii and Salix gooddingii, and other vegetation data collected at perennial and intermittent flow sites along the San Pedro and other rivers to validate recruitment functions for these species and overall projections of vegetation change, and then model these under varying hydrologic scenarios.
3C. Floods and riparian vegetation dynamics: data collection along spatial gradients
Taking advantage of the longitudinal variability in stream hydrology along the San Pedro River, we examined how low-flow and high-flow stream attributes interact to shape riparian vegetation at the population level and community level, as a basis for inferring future response of eh vegetation to stream flow changes resulting from climate change (Stromberg, Lite and Dixon in review) Specifically, we asked which biotic variables were influenced to the greater extent by low-flow conditions, which by flood intensity, which by both, and which by some interaction thereof. Given that identification of positive and negative feedbacks among ecosystem components is important for predicting effects of climate change, we also speculate in this in-review paper as to how the particular changes in vegetation ensuing from hydrologic change might reciprocally affect hydrologic and geomorphologic processes.
3C. Floods and riparian vegetation dynamics: aerial photography analysis
A time series of aerial photographs are being analyzed to accomplish two goals: one, to assess long-term temporal patterns of change in response to extreme river flooding, and two, to contrast response between stream reaches with different stream flow conditions (i.e., perennial, intermittent and ephemeral). Archived aerial photographs from 1955, 1978 and 2003 were obtained to provide continuous coverage of the Upper San Pedro riparian corridor from the United States/Mexico border to the Benson Narrows. The 1955 photos were digital, scanned, black and white images obtained from the USDA Aerial Photography Field Office; this was the oldest photo series with high-resolution imagery. The 1978 photographs were digitized, scanned, color images obtained from the U.S. Geological Survey EROS Center. The 2003 photos provided a recent end-point, and were color infrared Digital Orthoimagery Quarter Quadrangles (DOQQ). The scanned aerial photographs were georeferenced in ArcMap 9.2 using spatially referenced DOQQs as target data. Seven cover types (among them Populus-Salix forest, woody shrubland, and riparian grassland/forbland) were identified using ground-truthed field data, and these were discriminated on the imagery based on shape, texture, and color. To map cover types, polygons were drawn around homogeneous vegetation patches while viewing the imagery at a scale of 1:3,000. A minimum polygon size of 10,000 square meters was used. Within each polygon, the percent cover of each cover type was visually estimated using cover classes. The cover classes were 0%, 1-5%, 6-20%, 21-40%, 41-60%, 61-80%, and 81-100%. The area of each cover type was calculated by summing, across polygons, the product of the cover-class midpoint for the polygon and the relative area of each polygon. Preliminary results show that significant areas of bare ground from 1955 have been converted to vegetated cover particularly Populis-salix forest which dovetails with the generally observed increased encroachment of woody species that has been observed along the San Pedro.
3C. Floods and riparian dynamics: field monitoring through time
Desert rivers have a high degree of intra- and inter-annual variability in streamflow conditions. Monitoring changes in riparian vegetation through time can provide insights into changes that may arise from climatic shifts. Herbaceous vegetation is particularly responsive to changes in stream flow conditions, given the short live span of many species in this group.
To investigate changes in riparian plant abundance during years with different stream flow conditions, and to contrast response between sites with ephemeral vs. perennial stream flow, we established monitoring sites at two rivers (Hassayampa River and Cienega Creek). Riparian plant cover and richness was sampled during multiple years and seasons. These studies have implications for predicting future climate change and for restoration and conservation of riparian ecosystems. Results from the Hassayampa River reveal the important role of large winter flood pulses in allowing for development of ephemeral wetlands at typically-dry river sites and in creating a diversity-productivity pulse at all sites, including those with perennial flow (Stromberg, et al., 2009a). Results from Cienega Creek similarly reveal the role of wet years in increasing plant species diversity, but also reveal that the response to flood pulses varies with the long-term flow conditions at the site: the relative diversity increase following floods can be greatest at ephemeral-flow sites because the chronically dry conditions create large expanses of bare ground available for colonization by opportunistic species. (Stromberg, et al., 2009b).
3D. Develop integrated vegetation change model to evaluate climate scenarios.
In the work described in hypotheses 1-3 we are developing the building blocks for this task. We have developed the first step in our surface water routing model and in developing the interaction of surface water with groundwater in the San Pedro system. We are on a path to developing the necessary seasonal groundwater model and need to pursue the student who will be able to build and execute such a model for the San Pedro. Work has also proceeded on the historical aerial photos and the mortality and recruitment model that will be needed to combine into the overarching integrated vegetation change model. The work on this integration cannot be completed until the various pieces are integrated.
3E. Regional Sensitivity of Rivers to Climate Change
While we have not explicitly worked on this aspect of the project our ongoing investigations in the Bill Williams, San Pedro and Hassayampa do lead us to be able to make some conclusions about the regional sensitivity of these systems to climate change and variability. The Bill Williams and Hassayampa are both heavily dependent on large flows that then sustain the baseflow in the system over longer time scales. The San Pedro is much less sensitive to short term variations in storm flows but longer term variations do seem to have an impact on baseflow and likely effects that cascade into the vegetation community.
3F. How will ecosystem services be impacted under climate change?
Again this task is a final integrative task of the project but again initial results from the impact of hydrologic alterations on vegetation change lead us to some preliminary conclusions. With less frequent floods vegetation might change to be more sparse and upland in character with more annuals over perennials in the herbaceous community. Such a shift would decrease the riparian systems ability to resist stream erosive power during the remaining high flows leading to an increase of sediment export downstream. As we make progress on other tasks we expect to gain insight into how the coupled hydrologic and biological system will respond to alterations ion climate and how this will affect services that society gains from these ecosystems.
Awards
-
Outstanding student presentation at fall AGU 2007
- Hoori Ajami a graduate student on this project has successfully won two internal Graduate Fellowships and a travel award to the ESRI GIS conference.
Journal Articles on this Report : 4 Displayed | Download in RIS Format
Other project views: | All 51 publications | 16 publications in selected types | All 15 journal articles |
---|
Type | Citation | ||
---|---|---|---|
|
Stromberg JC, Hazelton AF, White MS, White JM, Fischer RA. Ephemeral wetlands along a spatially intermittent river:temporal patterns of vegetation development. Wetlands 2009;29(1):330-342. |
R833025 (2008) R833025 (2009) R833025 (Final) |
Exit Exit |
|
Stromberg JC, Hazelton AF, White MS. Plant species richness in ephemeral and perennial reaches of a dryland river. Biodiversity and Conservation 2009;18(3):663-677. |
R833025 (2008) R833025 (2009) R833025 (Final) |
Exit |
|
Stromberg JC, Tluczek MGF, Hazelton AF, Ajami H. A century of riparian forest expansion following extreme disturbance:spatio-temporal change in Populus/Salix/Tamarix forests along the Upper San Pedro River, Arizona, USA. Forest Ecology and Management 2010;259(6):1181-1189. |
R833025 (2009) R833025 (Final) |
Exit Exit Exit |
|
Treese S, Meixner T, Hogan JF. Clogging of an effluent dominated semiarid river:a conceptual model of stream-aquifer interactions. Journal of the American Water Resources Association 2009;45(4):1047-1062. |
R833025 (2009) R833025 (Final) |
Exit |
Supplemental Keywords:
Riparian ecosystems, groundwater hydrology, isotope hydrology, riparian vegetation, invasive species, water quality., RFA, Scientific Discipline, Air, Ecosystem Protection/Environmental Exposure & Risk, Hydrology, climate change, Air Pollution Effects, Monitoring/Modeling, Environmental Monitoring, Regional/Scaling, Ecological Risk Assessment, Atmosphere, coastal ecosystem, biodiversity, environmental measurement, ecosystem assessment, global change, greenhouse gases, anthropogenic, climate models, water quality, environmental stress, coastal ecosystems, habitat diversity, ecological models, climate model, Global Climate Change, land use, regional anthropogenic stresses, atmospheric chemistry, stressor response model, climate variability, habitat preservationProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.