Dryland Research Plan


  1. Quantify the presence of ordered spatial distribution of semi-arid southwestern woody plant species.
  2. Measure the spatial correlation of that patterning with the ecohydrological drivers associated with a specific mechanism of emergent growth pattern, which is tied to non-linear mortality in response to precipitation change.

This project investigates the previously un-examined phenomenon of emergent patterning of vegetation in American drylands. Initial inspection of aerial imagery suggests that there is structured distribution of vegetation in water-limited regions of the southwestern U.S., not previously characterized in the region. Similar phenomenon are described in other semi-arid regions of the world, and have been linked to threshold mortality under reduced rainfall, a probable outcome of global climate change in the American southwest.

Theoretical Justification and Research Significance


The phenomenon of emergent spatial patterning of water-limited vegetation has been described in several dryland regions of the world. Examples include “tiger striping” of bush in the African Sahel [White1970], “mazing” of woody vegetation in the Congo Basin [Couteron and Kokou1997] and banding of chenopod shrubland in Australia [Dunkerley and Brown1995]. In these systems, seedlings in the ‘organic island’ of an established plant have access to water retained in the litter layer, greater soil organic content, altered soil structure, and a more temperate micro-climate. In addition to spatially differentiating seedling establishment, these patches of organic island also alter the down-slope flow of rain water from rare storms, spreading it laterally. As these plant-scale mechanisms of spatially differential growth and water retention aggregate over time and across landscapes, structured patterns of plant presence emerge [Aguiar and Sala1999]. These structures are understood to be more efficient at retaining rain water than would random spatial plant distribution, chanelling it across more absorptive surface rather than allowing direct run-off from the system [Tongway and Ludwig2001]. Emergent pattern is therefore theorized to allow plants to exist at lower levels of precipitation than otherwise possible. Descriptive understanding of these mechanisms is well established. New theories based on dynamic simulations indicate that such patterning may have significant implications for the functional response of dry landscapes to changes in rainfall. Complex emergent structures commonly demonstrate threshold responses to changing drivers, wherein small changes are amplified through the aggregate response of the individual-level mechanisms. Simulations suggest that if a region of patterned vegetation crosses a threshold minimum of precipitation, sudden die-off may occur. Further, re-vegetation may not be possible even if rainfall amount subsequently surpasses that threshold, as no one first plant can be a rain-gathering pattern unto itself [Rietkerk et al.2002].

In America.

Semi-arid drylands cover much of the southwestern United States. The climate sensitivity of these systems is well documented. Regional scale mortality of dominant species has been observed on the order of 1000s of km2 in response to sub-decadal changes in soil moisture [Breshears et al.2005]. Water distribution and retention are predominant drivers of drylands ecology generally, and plant-scale understanding of how water and nutrients are retained and utilized is an active and productive research area. Linking plant-scale mechanisms to landscape-scale outcomes becomes a critical next step in understanding and predicting dynamic regional response to climatic change.

Initial assessment of aerial imagery suggests that structured plant growth patterns do exist at many sites across the American drylands. These patterns are generally consistent with the banding associated with the water-limited emergent structure described above, and the plant-scale mechanism associated with this type of emergent structure–improved success of seedlings near established plants–is well demonstrated in American ecosystems. This is particularly the case in the pinon-juniper woodlands which dominate large areas of the southwest [Breshears2008]. However, water-limited patch-driven emergence has not been proposed to explain banded pattern in American ecosystems; in fact there is no literature regarding banded vegetation in U.S. arid systems (excepting structuring by underlying geology, e.g. Ives1946Woodruff1993). Examples of spatially ordered growth at non-U.S. sites have been visually striking at considerable distances, sometimes functioning at the scale of tens of miles, and/or being very consistent. The patterns identified during initial assessment of U.S. sites are typically less pronounced. It is plausible that a degree of spatial patterning is present in a significant number of semi-arid areas in this country, but remains unidentified.

Self-ordered vegetation is expected to exhibit non-linear, threshold response to change in available water (see World-wide, above). This can mask impending regional mortality, by enhancing a landscape’s efficiency at retaining scarce rainfall up to but not beyond an unpredicted threshold [Rietkerk et al.2004]. In cases where available rainfall varies outside previously observed limits, it would therefore be crucial to have an understanding of self-patterning dynamics to better predict possible thresholds and calibrate management response [Noble et al.2001]. Current climate change models suggest an increase in the frequency and intensity of droughts. There is therefore a strong motivation to test if U.S. ecosystems are indeed emergently self-organized, and to closely characterize any mechanisms of organization.

This project explores the ecohydrological link between plant and landscape scales by testing the novel hypothesis that American drylands are spatially structured in part by emergent, water limited patch-interpatch dynamics. This possibility represents both a potentially critical gap in understanding how these ecosystems will respond to climatic change, and a potential synergy of available plant-scale research; ecological pattern theory; and remotely sensed data of real ecosystems.

Analysis Steps

The spatial distribution of woody plants at several sites in water-limited regions of the American southwest will be mapped using semi-automated extraction from aerial photography. If present, patterns of distribution will be statistically characterized using geostatistical pattern classification methods. The water content of the vegetation will be estimated using aerial hyperspectral imagery. Data regarding the geohydrological template of each study site will be also be gathered, in particular modeled rainfall accumulation. Statistical correlations will be measured among the type and degree of spatial patterning of the vegetation, the geohydrological drivers, and the estimated foliar water content maps. Existing spatially-explicit dynamic simulation models will be implemented, parameterized by the physical drivers of specific sites, and tested to see if they can recreate the statistical character of the patterns observed.


    D.D. Breshears. Structure and function of woodland mosaics: Consequences of patch-scale heterogeneity and connectivity along the grassland-forest continuum. In OW Van Auken, editor, Western North American Juniperus communities A dynamic vegetation type, volume 196 of Ecological Studies, page 58. Springer Verlag, 2008.

    D.D. Breshears, N.S. Cobb, P.M. Rich, K.P. Price, C.D. Allen, R.G. Balice, W.H. Romme, J.H. Kastens, M.L. Floyd, J. Belnap, et al. Regional vegetation die-off in response to global-change-type drought. Proceedings of the National Academy of Sciences, 102(42):15144–15148, 2005.

    P. Couteron and K. Kokou. Woody vegetation spatial patterns in a semi-arid savanna of burkina faso, west africa. Plant Ecology, 132(2):211–227, 1997.

    DL Dunkerley and KJ Brown. Runoff and runon areas in a patterned chenopod shrubland, arid western new south wales, australia: characteristics and origin. Journal of Arid Environments, 30(1):41–55, 1995.

    R.L. Ives. Desert ripples. American Journal of Science, 244(7):492, 1946.

    SK Jenson and JO Domingue. Extracting topographic structure from digital elevation data for geographic information system analysis. Photogrammetric Engineering and Remote Sensing, 54(11):1593–1600, November 1988.

    JC Noble, R. Peltier, P. Montagne, and EHL Mahamane. Toward improved management of arid and semiarid banded landscapes. In David J Tongway, Christian Valentin, and Josiane Seghieri, editors, Banded vegetation patterning in arid and semiarid environments Ecological processes and consequence for management, number 149 in Ecological Studies, pages 210–227. Springer Verlag, 2001.

    D.R. Patton. A diversity index for quantifying habitat edge. Wildlife Society Bulletin, 3(4):171–173, 1975.

    M Rietkerk, MC Boerlijst, F van Langevelde, R HilleRisLambers, J van de Koppel, L Kumar, HHT Prins, and AM de Roos. Self-organization of vegetation in arid ecosystems. AMERICAN NATURALIST, 160(4): 524–530, OCT 2002. ISSN 0003-0147.

    M Rietkerk, SC Dekker, PC de Ruiter, and J van de Koppel. Self-organized patchiness and catastrophic shifts in ecosystems. SCIENCE, 305(5692):1926–1929, SEP 24 2004. ISSN 0036-8075.

    H.C. Stimson, D.D. Breshears, S.L. Ustin, and S.C. Kefauver. Spectral sensing of foliar water conditions in two co-occurring conifer species: Pinus edulis and juniperus monosperma. Remote Sensing of Environment, 96(1): 108–118, 2005.

    DJ Tongway and JA Ludwig. Theories on the origins, maintenance, dynamics, and functioning of banded landscapes. In David J Tongway, Christian Valentin, and Josiane Seghieri, editors, Banded vegetation patterning in arid and semiarid environments: ecological processes and consequence for management, number 149 in Ecological Studies, pages 20–31. Springer Verlag, 2001.

    OW Van Auken and DC McKinley. Structure and composition of juniperus communities and factors that control them. In OW Van Auken, editor, Western North American Juniperus communities A dynamic vegetation type, volume 196 of Ecological Studies, page 19. Springer Verlag, 2008.

    L. P. White. Brousse tigree patterns in southern niger. The Journal of Ecology, 58(2):549–553, 1970. ISSN 00220477. URL http://www.jstor.org/stable/2258290.

    CM Woodruff, Jr. Geology and physiography of glen rose limestone terrains, barton creek watershed, central texas. In CM Woodruff Jr, WM Marsh, and LP Wilding, editors, Soil, landforms, hydrologic processes,and land-use issues: Glen Rose limestone terrains, Barton Creek watershed, Travis County, Texas, pages 1–21. Society of Independent Professional Earth Scientists, Central Texas Chapter, 1993.

Back to Drylands Emergent Plant Pattern


[…] the curious, I’ve posted a more detailed version of my thesis research plan. Which is currently being […]

[…] updated my research plan, again. It’s becoming a hobby. Now with improved narrative, and references to attest to […]

leave a comment