Data Package Metadata   View Summary

Study area

The study area, a grassland in southeastern Arizona, is dominated by perennial C4 grasses. Stipa neomexicana is the only C3 perennial grass in these grasslands, while approximately 40 C4 grass species are present. The experimental site was located near Sonoita (Santa Cruz County), Arizona (elevation 1450 m). The climate is semi-arid (401 mm annual precipitation), with a summer and a winter rainy season and a dry period in late spring. The experimental site, fenced rangeland, had never been plowed and was rarely grazed since the area was first settled, according to the owner. During the experiment, cattle were grazed on the site for a few days; jackrabbits, cotton-tails, small rodents, badgers, and coyotes also occur. The site included a broad, roughly north-south ridge that sloped gently to fields on the east, and on the west to a wash that drained into the Babacomari River several hundred meters to the north.

Experimental design

Three topographic positions, from the dry ridge crest to the lower slopes, were chosen to represent the dry-to-wet environmental gradient. At each of these positions, an area was chosen at random and divided into 10 blocks, each of which was subdivided into 16 approximately equal subunits. Treatments were randomly assigned to subunits. Experimental treatments consisted of removing all grass within a 0.5 m radius around a "focal" Stipa or Aristida plant, thus creating denuded circles around these center plants. These individuals were selected at random within the central area of a subunit. Control plants had no neighboring grasses removed, although the ground around control plants was disturbed to a degree comparable to that in removal treatments. Removal or control treatments were applied to 20 replicates for each species within the 10 blocks at each of the three positions along the environmental gradient.

Growth and reproduction of mature plants

Basal area was chosen as the best measure of plant size and response to treatments. Basal area was estimated from measurement of the longest axis through the base of a bunchgrass and the axis perpendicular to that, with area calculated from the formula for an ellipse. This estimate of basal area was closely correlated with aboveground dry mass (including crown) and with the number of tillers, although the relationship varied with season. A more accurate measurement was taken by forming insulated copper wire around the base of a plant so that the exact shape of the base was outlined, gently lifting the wire, and tracing the outline onto drawing paper. The cut-out tracing was weighed on an analytical balance, and the area determined from a regression of mass on known area. Correlation between the two methods for measuring basal area was highly significant (r2 = 0.86, P less than .001, n = 160).

Measurements of the initial sizes of mature plants were completed in January 1980. Subsequent measurements were made in January, May, August, and October 1980, and in January, May, and August 1981. Final measurements were made in January 1983. The total number of flowers produced by plants of each species for each treatment and position was sampled and counted when plants were within several days of anthesis.