Our objective was to evaluate the effects of burrowing activities by banner-tail kangaroo rats (Dipodomys spectabilis Merriam) on plant community structure and species dominance for two patch types at an ecotone between shortgrass steppe and desert grasslands in New Mexico, USA. Ten mounds produced by kangaroo rats were selected in patches dominated by Bouteloua gracilis (the dominant in shortgrass steppe communities) and ten mounds were selected in patches dominated by Bouteloua eriopoda (the dominant in Chihuahuan desert grasslands). Plant cover and density by species were sampled from three locations associated with each mound: the mound proper, the edge of the mound in the transition area, and the off-mound vegetation. Similar cover of B. eriopoda for the edges of mounds in both patch types indicates the ability of this species to respond to animal disturbances regardless of the amount of cover in the surrounding undisturbed vegetation. By contrast, cover of B. gracilis was low for all mounds and mound edges in patches dominated by this species. Much higher cover of B. eriopoda on mound edges compared to the undisturbed vegetation in B. gracilis- dominated patches indicates that kangaroo rats have important positive effects on this species. Lower cover of perennial grasses and higher cover of forbs, shrubs, and succulents on the edges of mounds in B. eriopoda - dominated patches compared to patches dominated by B. gracilis indicate the importance of surrounding vegetation to plant responses on disturbed areas. Our results show that kangaroo rats have important effects on both species dominance and composition for different patch types, and may provide a mechanism for small-scale dominance patterns at an ecotonal boundary; thus providing further support for their role as keystone species in desert grasslands.
Kangaroo rats have dramatic affects on vegetation structure in a number of patch types or landscape units including those dominated by various species of grasses and shrubs. As a result of these large effects on the vegetation relative to their low abundance, bannertail kangaroo rats have been identified as keystone species. At the ecotone between Chihuahuan desert grasslands and shortgrass steppe communities in central New Mexico, kangaroo rats affect patches dominated by Bouteloua eriopoda (Torr.) Torr. (black grama), the dominant grass in Chihuahuan desert grasslands, or Bouteloua gracilis (H. B. K.) Lag ex Steud. (blue grama), the dominant grass in shortgrass steppe communities. These two species differ in their life history traits, especially in response to disturbances that may affect their ability to dominate in the presence of small disturbances, such as mounds produced by banner-tail kangaroo rats. B. eriopoda is a short-lived grass that may respond rapidly to disturbance through the production of long stolons. By contrast, B. gracilis is a long-lived, slow-growing bunchgrass with limited ability to respond to disturbance either vegetatively or through seedling establishment. Differences in species or lifeform composition of patches dominated by one or the other Bouteloua species may result in patch-specific patterns in vegetation associated with mounds.Study Begin Date: 1-Jun-1998 Study End Date: 1-Aug-1998
Research MethodsStudy Methods: Mound selection: Banner-tail kangaroo rat mounds were selected from within patches dominated by either B. eriopoda or B. gracilis where dominance was based on average cover > 750f total plant cover. We identified a total of eight patches, four dominated by each Bouteloua species, from within a 400m x 1000 m area; each patch was ca 200 m x 200m in size. From within each patch, we identified five active kangaroo rat mounds where activity was determined by unobstructed entrances to burrows and/or fresh fecal material near a mound. Only active mounds were used in this study in order to minimize variation in vegetation and soil properties due to recovery processes that begin after a mound is abandoned. Although mound age could not be determined, mounds are sufficiently long-lived (> 30y) that differences in age are unlikely to affect vegetation dynamics. We then randomly selected a total of ten mounds to be sampled in each patch type. The area of each mound was estimated by measuring its outer length in two cardinal directions (north-south, east-west), and assuming an elliptical shape. Because average mound size was similar in B. eriopoda (11.3 m2) and B. gracilis (12.0 m2), mound size was not needed as a covariate in our analyses.Vegetation sampling: Average mound radius was used to locate the center of each mound. Vegetation was sampled in July along four transects radiating within 5o of the four cardinal directions from the center of each mound, and extending to a distance of 25 meters away in vegetation considered typical of the patch type between mounds. Variation in the direction was necessary to avoid nearby mounds. Data were collected using a 2.0-m2 quadrat placed at nine locations along each transect (in m from the mound center): 0.0, 1.5, 3.0, 5.0, 7.5, 10.0, 15.0, 20.0, and 25.0. Distance between quadrats was short near the mound to allow intense sampling of this area and to increase sample size; most of the length of each transect was contained in the typical patch vegetation, thus longer distances between quadrats were used. Each quadrat was recorded as occurring at one of three locations: on the mound ("mound"), at the edge of the mound in the transition area ("edge"), or off the mound in typical vegetation between mounds ("off-mound"). Locations were defined based upon the frequency and intensity of burrowing activities of kangaroo rats as well as the amount of bare ground. Mound locations were characterized by frequent, unobstructed entrances to burrows, predominantly bare ground, and elevated topography. Edges were defined as the transitional area surrounding mounds that contained few burrows with an intermediate amount of bare ground. Off-mound vegetation was characterized by no burrowing activity and low amounts of bare ground.For each quadrat, canopy cover only (to the nearest 1%) was estimated for bare ground and litter combined, and by species for perennial grasses where vegetative spread by tillers or stolons makes identification of individuals difficult, thus density estimates are imprecise. Data were collected by species for B. eriopoda, B. gracilis, and other frequently occurring grasses including Hilaria jamesii, Sporobolus flexuosus, and Aristida purpurea; all other perennial grasses were combined into one cover estimate for each quadrat. Canopy cover (%) and density (no./2 m2) were estimated for species where individual plants are easily determined, including shrubs (Gutierrezia sarothrae, Ephedra viridis) and succulents (Yucca glauca, Opuntia spp.). Cover and density of annual and perennial forbs and sub-shrubs including Psilostrophe tagetina, Glandularia wrightii, Hymenopappus filifolius, Sphaeralcea spp., Machaeranthera pinnitifida, Solanum elaeagnifolium, Plantago patagonica, Salsola kali, Astragalus spp., Chaetopappa ericoides, Cryptantha crassisepala, and Kraschninnikovia lanata were determined as a group for each quadrat due to low frequencies of occurrence. Because sampling was conducted in the middle of the growing season (July), cover of all species are likely under estimates of peak growth for that year.Statistical analyses: The data were analyzed using analysis of variance to test for the significance of patch type (B. eriopoda- or B. gracilis - dominated) and sampling location (mound, edge, off-mound) on cover or density of the vegetation. The design included patch type as the between subjects factor and location as the within subject factor. Dependent variables were analyzed separately, and included cover of five species or species-groups: (1) B. eriopoda; (2) B. gracilis; (3) other grasses combined; (4) forbs, shrubs and succulents combined; and (5) the total. Species-groups were used for grasses other than the two Bouteloua species, and for forbs, shrubs and succulents due to small sample sizes for individual species. Density of the forbs, shrubs, and succulents group was also analyzed. Least significant difference (LSD) means comparison tests were used to identify significantly different means at the 0.05 level. Two separate means comparisons tests were conducted for each response variable. Effects of patch type were determined by comparing cover or density of each species or group within the three locations. A similar analysis was conducted for determining effects of location on cover or density within each patch type.
July 25, 2003 -- File created by Kristin Vanderbilt from Excel metadata and data files submitted by Deb Peters. -- KLV
Additional Information on the personnel associated with the Data Collection / Data Processing
Other field crew members: Jon Erz & Teresa Seamster
See all Sevilleta Publications