Drivers of Plant Community Dynamics in Sagebrush Steppe Ecosystems: Cattle Grazing, Heat and Water Stress


Michael Reisner, Postdoctoral Researcher, Oregon State University,
Paul Doescher
Dave Pyke

Study Dates:

Jan 2008-November 2010

SageSTEP Study Plots:

Hart Mountain – Rock Creek and Gray Butte Sites, and other non-SageSTEP plots in Oregon

About the Study:

Sagebrush ecosystems across the Great Basin are threatened by cheatgrass invasion. In order to find how land managers might most effectively control the spread of this aggressive non-native, Michael Reisner of Oregon State University, working in collaboration with the SageSTEP team, dedicated his PhD work to understanding the drivers of change in Wyoming big sagebrush communities. Reisner’s work is unique among the SageSTEP projects for its focus on understanding the role of cattle herbivory in sagebrush ecosystems. Cattle herbivory, unlike other factors such as heat stress and water stress, can be actively managed and controlled, making Reisner’s work particularly valuable to land managers. The following three objectives structured this study:

  1. Describing the role of sagebrush as a driver of herbaceous species abundance
  2. Describing the role of sagebrush as a driver of community stability
  3. Evaluating the relative importance of the causal network of factors and processes driving invasibility

1. Sagebrush as a driver of herbaceous species abundance

Artemisia tridentata, commonly known as sagebrush, is a foundational shrub across many of the semi-arid Great Basin communities, dominating the landscape and defining the structure of many of these communities. Seeking to better understand how sagebrush might promote or limit the presence of native and non-native herbaceous species, Reisner’s study tested the stress gradient hypothesis (SGH) in the semi-arid Artemisia tridentata spp. wyomingensis (Wyoming big sagebrush) communities of the northern Great Basin.

The stress gradient hypothesis (SGH) predicts that facilitation and competition vary inversely along stress gradients with facilitation more frequent and stronger when stress is high and competition more frequent and stronger when stress is low. Previous studies have emphasized that the SGH is context-dependent, with relationships varying at the species level and between resource based and non-resource based stresses. In the case of Great Basin communities this is important because native bunchgrasses might interact differently with sagebrush than the non-native cheatgrass, and these interactions might differ under non-resource based (cattle herbivory and heat) stress as compared with resource based (water) stress.

The data used for this study were gathered from 75 sites in Wyoming big sagebrush communities chosen to capture one resource based stress gradient (water) and two non-resource based stress gradients (heat and herbivory). This figure shows examples of the different types of sites chosen in order of decreasing resilience and resistance.

Reisner analyzed the spatial patterns of association between the Wyoming big sagebrush and eight focal herbaceous species, six native and two non-native, in order to determine under what conditions the sagebrush competes with or facilitates different focal species. Greater focal species cover beneath sagebrush canopies as compared to the cover in adjacent interspaces was interpreted as facilitation. Greater cover in the interspace as compared to beneath the canopy was interpreted as competition. The two stress gradients observed were:  cattle herbivory stress overlapped with heat stress and cattle herbivory stress overlapped with water stress.

At the community level Reisner’s findings support the general applicability of the SGH across overlapping resource and non-resource based stress gradients. At the species level, comparing interactions between Wyoming big sagebrush and each focal species, Reisner’s findings supported the prediction that the strongest facilitation should occur with competitive species at the limits of their stress tolerance while the strongest competition should occur with stress tolerant species located at their ecological optimum.

The differences observed between native and non-native focal species was striking. The two non-native species (B. tectorum and L. perfoliatum) are at their ecological optima under high stress levels and correspondingly exhibited the strongest competition at the highest stress levels. The native focal species, in contrast, showed greater facilitation at the highest stress levels.

This means that in Wyoming big sagebrush communities under high levels of heat, water and cattle herbivory stress, the removal of sagebrush will simultaneously end its competition with non-natives and its facilitation of natives, favoring the spread of non-natives. Maintaining a minimum level of sagebrush cover may therefore be important in communities experiencing high cumulative stress levels when controlling the spread of non-native species like cheatgrass is a priority.

2. Sagebrush as a driver of community stability

Next Reisner sought to determine whether species level observations would translate into meaningful effects on community compositional and functional stability by testing the following two hypotheses:

  1. Wyoming big sagebrush facilitation of native bunchgrasses would increase functional stability,
    (i.e. decrease invasibility) by maintaining greater bunchgrass composition in under-shrub communities as compared to interspace communities, thereby limiting non-native cover in these communities.
  2. Wyoming big sagebrush facilitation of native bunchgrasses would both increase and decrease
    community compositional and functional stability. Facilitation would increase stability at intermediate stress levels, but decrease stability at high stress levels if obligate facilitation resulted in many bunchgrass species persisting only beneath sagebrush canopies.

Sagebrush facilitation of native bunchgrasses was found to decrease the invasibility of communities underneath the shrub canopy; however, this unfortunately did not translate into lower invasibility at the community level due to the limited spatial scale of the sagebrush facilitation. Consistent with the second hypothesis, facilitation did increase community stability at intermediate stress levels, but not at high stress levels. At high stress levels native bunchgrasses were largely absent from interspaces, aggregating beneath sagebrush and leaving increased resource availability in the interspaces, resulting in decreased stability.

These findings indicate that sagebrush communities experiencing high levels of heat, water and cattle herbivory stress are highly vulnerable to cheatgrass invasion. In contrast, communities experiencing intermediate cumulative stress levels are stabilized by sagebrush facilitation of native bunchgrasses which increase stability. In the first case, if stress levels are not reduced transformation into cheatgrass dominated annual grasslands could ensue. In the second case, management actions or natural disturbances that reduce sagebrush cover may decrease community stability unless cumulative stress levels are also reduced.

3. Causal factors and processes driving invasibility

As part of his research looking into the causes of invasibility, Reisner worked with a team of ecologists to develop a detailed conceptual model of the function of sagebrush ecosystems. Controlling for some factors, such as amount and timing of precipitation, in order to focus on a few factors of particular interest, such as grazing disturbance, heat stress, and water stress, Reisner used structural equation modeling to find the relative importance of individual factors. The conceptual model can be seen below.

The model shows the connections between factors predicted to influence invasibility while the pictures on the right show sagebrush ecosystems as invasibility increases.

In examining the causes of invasibility, plant community structure measured as the distance between perennial vegetation was found to be the most significant causal factor. Cattle grazing had no direct impact on invasibility, but rather indirectly increased invasibility by negatively impacting the community structure. This can be seen in the schematic below. 

Final inferential model of the susceptibility of sagebrush ecosystems to cheatgrass invasion. The relative importance or strength of a given causal effect is indicated by the thickness of the arrow and the standardized path coefficient. Because distance from water is inversely related to cattle grazing intensity (i.e. cattle grazing intensity increases with decreasing distance to water), positive path coefficients and correlations indicate an inverse relationship between cattle grazing intensity and the variable (i.e. increasing cattle grazing intensity decreases bunchgrass abundance). R2 depict the proportion of variation of each endogenous (response) variable explained by the model.

The results of this study will be useful to land managers seeking to control the spread of cheatgrass in grazed areas. Management strategies that promote the spread of native bunchgrasses will be most effective. Reisner also found that cheatgrass spread more readily and native bunchgrasses were less common in areas experiencing high cumulative stress levels. This means that grazing will have a more detrimental impact on the resilience of areas already experiencing some form of environmental stress, such as south facing slopes that receive high heat stress and areas with coarse grained soils that drain quickly, limiting water availability.

As global climate change is likely to increase heat and water stress in the Great Basin region, reducing cumulative cattle grazing intensities by altering utilization rates and/or seasons of use and other management strategies (such as changing the location of watering sources) may be the only effective means to maintain structurally and functionally sound sagebrush ecosystems. Cumulative cattle grazing levels must be reduced to levels that prevent the most susceptible communities within a grazing management unit from crossing these thresholds. Otherwise, the resilience of more vulnerable communities is likely to be compromised and they are likely to be invaded by cheatgrass. Once invaded, these communities will increase the risk of fires and may serve as foci for subsequent invasions of surrounding communities.

If the management goal is to restore ecosystem resilience, Reisner’s findings suggest that efforts should focus on restoring biotic resistance and preemption of resources provided by the native bunchgrasses within the interspaces between sagebrush. Managers should focus on maintaining high overall bunchgrass abundance/dominance and community structure characterized by spatially dispersed bunchgrasses in interspaces and small basal gaps between such individuals to capture large amounts of otherwise available resources in space. Additionally, it is important to maintain a diverse assemblage of bunchgrass species with different spatial and temporal patterns of resource use to capture available resources at different soil depths and times. Lastly, Reisner also found that biotic soil crusts limit safe sites for cheatgrass establishment in gaps between perennial native vegetation, making their presence useful in preventing cheatgrass spread.


Reisner, M. 2010. Drivers of community dynamics in sagebrush steppe ecosystems: cattle grazing, heat, and water stress. Ph.D. Dissertation,Oregon State University, Corvallis, OR. Available here.