Climate Forcing of Wetland Connectivity in the Great Plains:
An Exploratory Study Using Graph Theory
Embedded Files
Who we are
PIs –
Chris Wright (South Dakota State University)
Nancy McIntyre (Texas Tech University)
Co-PIs –
Geoff Henebry (South Dakota State University)
Katharine Hayhoe (Texas Tech University)
Frank Schwartz (Ohio State University)
Associated personnel –
Ganming Liu (Ohio State University)
Sharmistha Swain (Texas Tech University)
NSF-Macrosystems Biology grants
1065773 and 1065845
What we do
Land conversion and climate change are acknowledged threats to habitat connectivity and thence to the long-term viability of many wildlife species. Of all the habitats being affected by land conversion and climate change, wetlands are simultaneously among the most important and yet most impacted. For example, the three main wetland complexes of the Great Plains--the prairie potholes of the northern Great Plains, the Rainwater Basin in south-central Nebraska, and the playas of the southern Great Plains--are crucial breeding, overwintering, and migratory stopover habitats for aquatic and amphibious wildlife, but losses of these wetlands (to drainage, drought, etc.) compromise the ability of wetland-dependent species to disperse between patches of suitable habitat. We are examining these wetland networks to establish links among climatic drivers and functional habitat connectivity, using a graph-theoretic approach as a relatively rapid assessment of the metapopulation potential of wetland landscapes and how that capacity varies as a function of climatic forcing.
We are examining responses by birds and amphibians to changes in wetland network topology as a function of land cover change and climate change within the prairie potholes, Rainwater Basin, and playas of the Great Plains (Fig. 1). Our research is using advanced sensing and information technologies to test whether species occurrence in Great Plains wetlands is a function of habitat availability and functional connectivity. If so, then weather-driven variation in habitat connectivity, as quantified using graph-theoretic metrics, should be positively associated with variation in species presence and abundance.
Fig. 1. Clockwise from top left: prairie pothole in South Dakota, playa in Texas, Northern Leopard Frog (Rana pipiens), and waterfowl (mostly geese). Photos by N.E. McIntyre.
Results to date
Climate projections suggest that precipitation patterns will be disrupted across the Great Plains, with resulting effects on the availability of wetlands. We project increased risk of drought for portions of all three of our focal regions, particularly during the breeding season, although there are differences with season, time period, and emissions scenario (Fig. 2).
Climate model projections predict increased risk for drought, particularly during summer:
Fig. 2. Anomalies in average Standardized Precipitation Index (SPI) for January and June, as simulated by 28 climate models relative to the historical baseline period 1961-1990 under two future emissions scenarios (RCP 4.5 and 8.5) over two time periods (2041-2070 and 2071-2100). Anomalies are calculated as the difference between future and historical SPI, scaled from -100% to 100%. The “warm” colors indicate futures drier than the historical period whereas the “cool” colors indicate a wetter future. The prairie pothole region is outlined in red, the Rainwater Basin area in green, and the playa lakes region in blue.
With respect to climate change, the primary question is whether wetland landscapes in the Great Plains of the future will resemble network configurations found in drought years. In these cases, even though some wetlands may persist under drier conditions, land conversion has resulted in most wetlands being embedded within a relatively hostile matrix, effectively isolating them within components that are likely too small to support viable metapopulations (Fig. 3).
Some differences in connectivity are seen with increasing wetland density (more redundancy, fewer stepping-stones):
Fig. 3. These false-color composites of mid-infrared, near-infrared, and visible spectral bands (Landsat-5 Thematic Mapper imagery from a portion of path 31/row 27) show prairie potholes in central North Dakota during dry (1989) and wet (1997) times. Surface water appears black, actively transpiring vegetation is green, and senescent vegetation and bare soil appear pink to purple. A graphical representation of the prairie pothole network is superimposed, with linkages as yellow lines between potholes >20 ha in area within 0.5 km. Wetlands that may act as stepping-stones are circled in white; those that are hubs linked to >8 other wetlands within 0.5 km are circled in red. A difference in network topology between drought and deluge periods is evident.
We have found that size distributions of prairie potholes are power-law distributed throughout a drought to deluge cycle, meaning that between years but at similar calendar dates, the slopes of these distributions were remarkably similar. In other words, the number of potholes exhibited a consistent pattern of size abundances in spite of variability in landscape wetness and in the interplay of surface water flow, groundwater flow, and evaporation. This power-law scaling relationship indicates that wetland habitat networks under projected climate scenarios will exhibit constrained connectivity; moreover, dispersal capacity will be as important as wet/dry conditions to an organism’s ability to traverse the wetland network (Fig. 4).
Network topology appears to differ as much with scale as with drought:
Fig. 4. Centroids of playas within a 185 x 185 km region of the Texas panhandle (Landsat path 30/row 36), from a relatively dry date (12 August 2010, 2519 wetlands) (panels a and b) compared to a relatively wet date (14 January 1986, 3223 wetlands) (panel c). In panels a and c, playas within 5 km of each other are grouped into clusters; in panel b, playas within 10 km of each other are grouped. Members of the same cluster are denoted by the same color, with linkages shown as gray lines. Whereas multiple clusters are evident during the relatively dry date at a lower dispersal capability (a), a single, spanning cluster of wetlands indicates complete network traversability for an organism with a greater dispersal capacity (b) or for one with a lower dispersal range but during greater wetland availability (c). In other words, during drought, an organism must effectively double its dispersal capacity if it is to traverse the network.
Coalescence of the playa network usually occurred at >10 km dispersal distance and depended on wetland density, indicating that critical thresholds in connectivity arose from synergistic effects of dispersal ability (spatial scale) and habitat availability (wet vs. dry periods). Organisms with dispersal capabilities limited to <10 km routinely experienced effective isolation during our study.
We quantified fluctuations in the status of playas in supporting connectivity through the wetland network. We ranked playas as stepping-stones, cutpoints, and hubs at different levels of habitat availability (regionally wet, dry, and average periods of precipitation) and for dispersal distances ranging from 0.5 to 34 km, representing a range of species’ vagilities, to provide baseline dynamics within an area likely to experience disrupted connectivity due to anthropogenic activities. An individual playa’s status as a stepping-stone, cutpoint, or hub was highly variable over time (only a single playa was a top 20 stepping-stone, cutpoint, or hub in >50% of all of the dates examined). Connectivity is thus a dynamic emergent landscape property, so management to maintain connectivity for wildlife within ephemeral habitats like playas will need to move beyond a patch-based focus to a network focus by including connectivity as a dynamic landscape property (Fig. 5).
There are very few playas that are consistently important in maintaining connectivity through the network, posing a significant challenge for conservation:
Fig. 5. Map of centroids of wet playas (blue circles) from Landsat scene 30/36 during the single wettest date over the period we examined (9 June 2010), superimposed with the top 20 betweenness centrality stepping-stones (red dots), top 20 hubs (orange dots), and all cutpoints (black dots) for all 37 dates examined, as calculated for the coalescence distance at each date. The gold star indicates the single playa that was a top 20 stepping-stone for a majority of dates and a top 20 hub for the top quartile of dates. Axes are UTMs (zone 14N).
Hydroperiod (a key ecological parameter) of playas of the southern Great Plains was affected by the land-use type surrounding a wetland. Land use affected the likelihood and duration of inundation, with those playas that never held water over a 4-year period (2008-2011) being surrounded by significantly more cropland and less pasture/grassland than those playas that were wet at least once. In contrast, playas in a urbanized setting had prolonged hydroperiods (Fig. 6).
Anthropogenic land use can shorten or lengthen hydroperiod:
Fig. 6. Proportion of land use within 100 m of playas that were always wet (blue), always dry (red), and those that were wet at least once (gray) during 2008-2011, Landsat scene 30/36.
Using monthly precipitation, potential evapotranspiration, and average temperature as inputs, a hydrological model of the prairie pothole complex was constructed that simulates surface water dynamics; this model allowed us to re-construct the spatio-temporal dynamics of wetland networks and relate these to the occurrence of wetland birds. We have found that bird abundance is closely associated with this climate-driven fluctuation in precipitation (Fig. 7) and wetland availability (Fig. 8).
Synchrony in focal bird species’ time series, and in response to precipitation:
Fig. 7. Time series of precipitation and abundance of some focal bird species within the playa region of the southern Great Plains, 1983-2009; data are from the Breeding Bird Survey and Christmas Bird Count, with each species represented mainly during only one of those two time periods (SaCr, UpSa, NoPi, NoSh = winter; SnPl, WiPh, AmAv = breeding season). SaCr = Sandhill Crane (Grus canadensis), SnPl = Snowy Plover (Charadrius nivosus), WiPh = Wilson’s Phalarope (Phalaropus tricolor), UpSa = Upland Sandpiper (Bartramia longicauda), NoPi = Northern Pintail (Anas acuta), AmAv = American Avocet (Recurvirostra americana), NoSh = Northern Shoveler (Anas clypeata). Abundances have been normalized as proportion of the maximum value by species to display species that differ in abundance by orders of magnitude on the same graph. Normalized precipitation data are from Amarillo, Texas, as a representative of precipitation patterns within the playa region overall. Sandhill Crane abundance in playas is related to accumulated precipitation over the previous three years, and marginally to autumn precipitation within a given year; Northern Pintail abundance is also related to autumn precipitation. Similar relationships to precipitation patterns (most lagged 1-3 years) are also seen in the Rainwater Basin and prairie pothole regions for various species.
Bird abundance is positively associated with wetland density:
Fig. 8. Tracking relationship between one of our focal species (Northern Pintail, Anas acuta) and prairie pothole availability from 1967-2005. Mean wetland density (blue diamonds) was calculated from simulated pond numbers (from the pothole complex hydrologic model developed by Liu and Schwartz 2011) for the month of May within 13 250-km2 sampling blocks in North and South Dakota. Northern Pintail occurrence was quantified for the following month (June) for each year with respect to the number of stops along each Breeding Bird Survey (BBS) route within the study area where at least one bird was observed (yellow circles).
Products
Planning tools for conservation –
N.E. McIntyre invited to be part of the Texas Playa Decision Support System, Playa Lakes Joint Venture, summer 2013-present (click here for more information)
Presentations –
*undergraduate student; #graduate student; +postdoctoral researcher
Bucher*, J., M.R. Al Mamun#, C.K. Wright+, and G.M. Henebry. 2012. Poster - Delineating open wetlands in Landsat 4-7 imagery using rule-based segmentation. Eastern South Dakota Water Conference, Brookings, SD, 30 October 2012.
Bucher*, J., M.R. Al Mamun#, C.K. Wright+, and G.M. Henebry. 2012. Talk - Delineating open wetlands in Landsat 4-7 imagery using rule-based segmentation. Rocky Mountains-Great Plains Regional AAG meeting, Park City, UT, 11-13 October 2012.
Collins#, S.D., L.J. Heintzman#, S.M. Starr#, and N.E. McIntyre. 2013. Poster: The hydrological dynamics of temporary wetlands in the southern Great Plains as a function of surrounding land use. Annual meeting of the U.S. Chapter of the International Association of Landscape Ecology, Austin, TX, 14-18 April 2013.
Hayhoe, K. 2013. Talk - Climate Change: Understanding the Science and the Uncertainties. Holtry Distinguished Lecture at South Dakota State University, Brookings, SD, 9 April 2013.
Parikh*, N.N., L.J. Ruiz*, L.J. Heintzman#, S.D. Collins#, S.M. Starr#, and N.E. McIntyre. 2013. Poster: Dynamic connectivity of temporary wetlands. Annual meeting of the U.S. Chapter of the International Association of Landscape Ecology, Austin, TX, 14-18 April 2013.
Parikh*, N.N., L.J. Ruiz*, L.J. Heintzman#, S.D. Collins#, S.M. Starr#, and N.E. McIntyre. 2013. Poster: Dynamic connectivity of temporary wetlands. Undergraduate Research Conference, Lubbock, TX, 22-25 April 2013.
Parikh*, N.N., L.J. Ruiz*, L.J. Heintzman#, S.D. Collins#, S.M. Starr#, and N.E. McIntyre. 2013. Poster: Using graph theory to quantify dynamic connectivity of temporary wetlands in the southern Great Plains. Annual meeting of the Ecological Society of America, Minneapolis, MN, 4-9 August 2013.
Wright+, C., N. McIntyre, G. Liu+, F. Schwartz, G. Henebry, K. Hayhoe. 2012. Poster - Collaborative study: Climatic forcing of wetland landscape connectivity in the Great Plains. First MacroSystems Biology PI meeting, Boulder, CO, 11-14 March 2012.
Publications –
*undergraduate student; #graduate student; +postdoctoral researcher
Collins#, S.D., L.J. Heintzman#, S.M. Starr#, C.K. Wright+, G.M. Henebry, and N.E. McIntyre. 2014. Hydrological dynamics of temporary wetlands in the southern Great Plains as a function of surrounding land use. Journal of Arid Environments 109:6-14. DOI: 10.1016/j.jaridenv.2014.05.006.
Liu+, G., and F.W. Schwartz. 2011. An integrated observational and model-based analysis of the hydrologic response of prairie pothole systems to variability in climate. Water Resources Research 47 (2):W02504. DOI: 10.1029/2010WR009084.
Liu+, G., and F.W. Schwartz. 2012. Climate-driven variability in lake and wetland distribution across the Prairie Pothole Region: From modern observations to long-term reconstructions with space-for-time substitution. Water Resources Research 48 (8): W08526. DOI: 10.1029/2011WR011539.
Liu+, G., and F.W. Schwartz. 2014. On modeling the paleohydrologic response of closed-basin lakes to fluctuations in climate: Methods, applications, and implications. Water Resources Research 50 (4):2975-2992. DOI: 10.1002/2013WR014107.
Liu+, G., C.K. Wright, N.E. McIntyre, F. Schwartz, and G. Henebry. In prep. Climate-driven variability of wetlands and waterbirds in the Prairie Pothole Region.
McIntyre, N.E., C.K. Wright+, S. Swain+, K. Hayhoe, G. Liu, F.W. Schwartz, and G.M. Henebry. 2014. Climate forcing of wetland landscape connectivity in the Great Plains. Frontiers in Ecology and the Environment 12:59-64. DOI: 10.1890/120369.
Rüegg, J., C. Gries, B. Bond-Lamberty, N.E. McIntyre, P.A. Soranno, B.S. Felzer, K.L. Vanderbilt, G.J. Bowen, and K.C. Weathers. 2014. Information management underpins open science and enhances research in macrosystems ecology. Frontiers in Ecology and the Environment 12:24-30. DOI: 10.1890/120375.
Ruiz*, L., N. Parikh*, L.J. Heintzman#, S.D. Collins#, S.M. Starr#, C.K. Wright+, G.M. Henebry, N. van Gestel+, and N.E. McIntyre. 2014. Dynamic connectivity of temporary wetlands in the southern Great Plains. Landscape Ecology 29:507-516. DOI: 10.1007/s10980-013-9980-z.
Swain+, S., and K. Hayhoe. 2014. CMIP5 projected changes in spring and summer drought and wet conditions over North America. Climate Dynamics. DOI: 10.1007/s00382-014-2255-9.
Websites –
https://sites.google.com/view/macrosystems-biology1/home
For "part 2" of our project, click here: https://sites.google.com/view/macrosystems-biology2/home
Links
NSF news about our project (scroll down to next to last entry on page): http://www.nsf.gov/news/news_summ.jsp?cntn_id=121279
NSF press release about our research in a special journal issue: http://www.nsf.gov/news/news_summ.jsp?cntn_id=130218
Texas Tech University press release: http://today.ttu.edu/2014/02/researchers-predict-future-birdwetland-scenarios-under-climate-change/
Disclaimer: Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Last updated: 16 June 2025
Page updated
Google Sites
Report abuse