Characterization of Ecosystems Exposure | Landscape Ecology Research

Development of Landscape Indicators for the Chihuahuan Desert

Co-Principal Investigators: Maliha S. Nash, Ph.D. & Walter G. Whitford, Emeritus

For other projects and field works at the Chihuahuan Desert contact: Kris Havstad, Ph.D.

Introduction: This project is a collaboration with the USDA, ARS, and the Jornada Experimental Range, Las Cruces, New Mexico, in an effort to define and quantify indicators of rangeland condition that can be used in assessment and monitoring of rangelands in the western U.S. The research encompasses spatial scales from patch to landscape and time scales ranging from seasons (to assess inter-annual variation) to decades and centuries. Studies examining indicators at the patch scale have been conducted as part of a long-term experiment designed to examine interacting and multiple environmental stressors on ecosystem properties and processes.

Jornada on the Map: http://Jornada-www.nmsu.edu

Experimental Design:
Studies at the Chihuahuan Desert were carried out by applying different methods of sampling. First, the multiple stressors site has been in place since 1994. The treatments were a combination of two factors: (1) removal of the mesquite shrubs (Prosopis glandulosa Torr) and (2) intensive seasonal grazing (winter or summer). These treatments were applied in a randomized complete block design. A complete description of the site can be found in Nash et al. (1998, 2000; see publications). Vegetation, soil and animal population measurements were made from 1994 through 1999 to characterize the exposure variables and to evaluate system responses to these variables for assessment of useful indicators. Measurements completed by 1999 included cover and composition of the vegetation, fetch lengths of unvegetated patches, fine-scale topographic variability in soil surfaces on grazed and ungrazed plots, soil aggregate stability and compaction, arthropod and mammal communities, and soil microarthropod communities. Various papers have been published on ants, mammals and fine-scale topographic variability and soil disturbance by animals (Nash et al., 2000 and others; see publications). There has been high mortality of grasses on the grazed plots and current studies are examining the role of herbivory by small mammals in keeping the plots in a degraded state.

Second, systematic sampling over grids for many variables was carried on in the multiple stressor sites and other sites. Geostatistical techniques were used for mapping to follow the changes of biota behavior with space and time (Nash et al., 1998, 1999, 2001; see publications). If the effort is to determine whether a biological variable is an indicator of an environmental stressor, spatial and temporal variability of that variable must be considered. Mapping abundance of particular ant species, for example, Solenopsis and Conomyrma groups that characterize their response to disturbance caused by grazing with time indicated that the overall abundance decreased over time and did not revert to their original values. Changes in Solenopsis and Conomyrma abundance were more pronounced following grazing and were in a specific pattern as a function of disturbance. Because ants construct nests deep in the soil, surface disturbances resulting from human and/or animal activities may be difficult to assess if measurements are made only once. Ants do respond to environmental changes which can be explained by their spatial distribution and by how they relate with other variables such as soil properties, vegetation and nest locations.

Third, characterization of biological and nonbiological variables were measured along transects at equally spaced intervals. We summarized results from two studies as an example of impact of domestic livestock on soil properties (e.g., microtopography) and perennial vegetation. This impact was greatest close to water points and generally decreased exponentially with distance from water. In one study (Nash et al., 1999; see publications), we used multivariate analysis and semi-variograms to locate boundaries, and to determine the number and width of different annual-plant-zones (referred to as biotic zones) on long-term livestock disturbance gradients in the northern Chihuahuan Desert, NM. Tansy mustard, Descurainia pinnata, was abundant in severely disturbed areas and also in areas which are known to have high soil nitrogen content. Amaranthus palmeri was abundant in half of the transects in the zones nearest the water points. The number of boundaries and sizes of zones varied with distance from water points, with seasons, and with duration of grazing. The first biotic zone (most severely impacted by cattle) ranged from 75 to 795 m in radius for winter-spring annuals, and from 165 to 1065 m in radius for the summer annuals. Variability in the number and size of biotic zones along grazing gradients was spatially correlated with the frequency and intensity of disturbance, with landscape position, and with patchiness of soil features. There were fewer but larger zones of summer annuals than the winter-spring annuals. Boundary analysis of livestock disturbance gradients provided a method with replication for assessing the impact of long-term livestock grazing on annual plant communities. Livestock create nutrient-rich patches near water points by mixing dung with soil by hoof action. These nutrient-rich patches support species of annuals that are rare or absent in areas where soils are subjected to low-intensity disturbance.

Another study of topography used systematic sampling. In testing the significant impacts of livestock in the creation of piospheres centered on water points, the loss of soil microtopography across a "landscape" has been influenced by many years of livestock grazing (Nash et al., 2003; see publications). The size, height, and spatial distribution of micromounds and surrounding depressions were measured by a modified erosion bridge at three distances (50, 450, and 1050 m) from water points in desert grassland pastures in the Jornada Basin, NM. Microtopography of plots 1050 m from water points was significantly different from that of plots nearest to water points. We found that the dependence of a strong correlation between microtopography and the cover of long-lived perennial grasses could be used to assess the trend of organic matter content with the concordance of microtopography. Loss of microtopography from the impact of livestock in piospheres exacerbates erosion processes and contributes to desertification.

Remote Sensing: Remotely sensed and ground-based data are being used to develop indicators of ecosystem health and exposure to landscape stress. The project is developing analytical techniques to integrate information about stress across scales that will lead to regional risk assessment tools. Previous research at the landscape scale has addressed the question of "how to identify the threshold for degradation that may lead to an alternate stable but degraded state." Current research is focused on regional analysis using remotely sensed imagery and ground studies to examine relationships among degraded states, geomorphology – lithology, and man-made features that support the livestock industry in the western U.S.

Publications
Desertification: Much of the literature on desertification in the desert Southwest has focused on drought and overgrazing by livestock management. However, there may be important but unknown or undocumented ecological consequences of activities associated with the livestock industry. The present day rangelands are crisscrossed by a network of roads used by ranchers to manage livestock herds. In general, ranch roads were not designed to have minimal impacts on the landscape and were simply developed to get from point A to point B with the least amount of effort and in the shortest time.
There are a number of variables that potentially influence the impact of desert rangeland roads on landscape processes and ecosystem function. Many ranch roads are simply cut into the soil to remove vegetation and to provide a suitable track for ranch vehicles. These roadbeds are generally below the grade of the surrounding landscape. Such roadbeds disrupt rill flow and channel-flow water during intense rains and may become ephemeral stream channels that fill with sediment. Elevated roadbeds may function as dams and disrupt overland flow to downslope areas. The effect of ranch roads on landscape processes and ecosystem function will also vary with soil textural characteristics and geomorphology of the landscape (e.g., slope, aspect, density and size of channels). We are initiating studies of the impacts of roads in desert rangelands using a GIS modeling approach combined with ecological studies of hydrology and vegetation composition and production at critical points on roads identified by the GIS model.