-TWO sites, one relatively undisturbed and one heavily used for training activities, were sampled in the upland Juniperus ashei-Quercus fusiformis community. Effects of intensive tank training over a period of years are: (1) reduction of total plant cover by approximately 60%; (2) disturbance and consequent erosion of about 60%of the soil surface; (3) reduction in density and cover of woody vegetation without any major change in species composition; and (4) a major shift in herbaceous plant species from perennials to small annuals. Continued use of the same site for training will probably not result in much further deterioration. No work has been published regarding the enviromental effects of heavy off-road (i.e., tracked) vehicles such as tanks and armored personnel carriers. The purpose of this study was to determine the effects of tracked vehicles on the vegetation of the rock cedar-live oak woodland of central Texas. STUDY AREA.-Two sites were selected for quantitative sampling; selection was based on close similarity of soils, topography, and tree species composition, with the only significant difference being that one be heavily used for tracked vehicle training and the other not be used for such activities. The study was conducted in May 1979. The two study areas are located within the boundaries of Fort Hood Military Reservation west of Temple in central Texas (31?05'N, 97?45'W). Fort Hood is in the southern portion of the Cross Timbers and Prairies Vegetational Area (Gould, 1975). The flora of Fort Hood is influenced by the Edwards Plateau Vegetational Area about 80 km to the southwest and the Blackland Prairie, which is immediately to the east. The relatively undisturbed site, which is referred to as the control area, is located at 258468 on the 1977 Ft. Hood Training Map. The site is about 0.5 km north of North Nolan Road and about 2 km east of East Range Road. The vegetation consists of typical upland forest dominated by rock cedar (Juniperus ashei) and live oak (Quercusfusiformis). Topography is rolling, with a combination of level ridge tops and both eastand west-facing slopes. The soil is light-colored gravely clay. A few cedars have been cut for posts and a few old vehicle trails pass throught the area, but this is true of all the upland forest observed at Fort Hood. The tracked vehicle training site, which is referred to as the Tank Training Area (TTA), is located at 075560 on the Ft. Hood training map. The site is adjacent to the east side of Antelope Road about 1 km south of Table Rock Creek. The vegetation consists of motts of trees in a network of heavily used tracked vehicle trails. Plant species composition conforms to the typical rock cedar-live oak type, at least with respect to the woody plants. Topograpy and soils are very similar to the control area except that much of the topsoil has eroded, leaving the very light-colored gravely clay subsoil exposed. Both the control and the TTA are moderately grazed, and have been for many years. METHODS.-Vascular plants were divided into three categories for phytosociological sampling. The categories are: (1) trees-plants with woody stems greater than or equal to 5-cm diameter at a height of 150 cm; (2) small woody vegetation-plants with woody stems less than 5-cm diameter at a height of 150 cm (this category includes saplings, tree seedlings, shrubs, woody vines, and cacti); and (3) herbaceous plants-vascular plants without perennial aboveground parts. Trees were sampled for density, frequency, and basal area with an augmented variable-radius method (Rice and Penfound, 1955). At 30 points (placed at random distances along a compass line) in each stand, basal area was estimated for each tree species with a 10-BAF prism (Bruce, 1955) and all trees were counted by species in an 18.3-m diameter circular quadrat for density and frequency estimates. At an additional eight randomly selected points in each stand, diameters were measured on all trees within an 18.3-m diameter circle for size-class distribution data. Total area sampled for trees was I ha in each This content downloaded from 157.55.39.120 on Sun, 04 Dec 2016 04:54:25 UTC All use subject to http://about.jstor.org/terms The Southwestern Naturalist TABLE 1.-Tree data summary. Importance percentage is the mean of relative frequency, relative density, and relative basal area. Importance percentage