The study consists of 24 experimental plots, assigned to various ant, rodent, and plant manipulations. Although a subset of the plots has maintained a consistent manipulation since 1977, other plots have been reassigned over the course of the study.
Experimental manipulations have been altered for some plots over the course of the study. Changes in treatment assignment occurred in 1985 and 1987. Changes in treatment were initiated when the manipulation did not cause significant changes in the rodent community (see Brown 1998 for details). Because responses were either non-existent or minor, plots that were plant manipulations or seed additions from 1977-1987 and were assigned as controls in 1987 are often treated as “long-term controls” for research focusing on the long-term dynamics at the site.
Rodent manipulations are effected through the use of gates in the fencing of each plot. Rodent removals contain no gates and any rodents captured on those plots are promptly removed. All other plots contain 16 gates (4 per plot side); gates consist of holes cut through the hardware cloth of the fencing. Gate size is manipulated to exclude subsets of the rodent community. Dimensions for gates on kangaroo rat removal plots are 1.9 cm x 1.9 cm, D. spectabilis removals were 2.6 cm x 3.0 cm, and control plots are 3.7 cm x 5.7 cm. Species caught on plots from which they are supposed to be excluded are removed from the site and the access point to the plot is located and eliminated. Plots affected by these treatments are listed in Table 1, columns 2-4.
Since 1988 there have been no direct manipulations of the plant community. Before 1988, annuals were “removed” by applying an herbicide (brand: Roundup), but this removal was not considered successful and was discontinued (Brown 1998). Plots affected by these treatments are listed in Table 1, column 3.
Since 1985 there have been no seed additions to any plot. Before 1985, seed additions were conducted by applying 96 kg milo (Sorghum vulgare) and/or millet (Panicum miliaceum) seeds to designated plots (Davidson et al 1985). Plots affected by these treatments are listed in Table 1, column 2.
Since 1977, ant manipulations have been conducted by applying a commercial poison (Mirex [Allied Chemical Corporation] through 1980 and AMDRO [American Cyanamide Company] afterwards) to designated plots (Davidson et al 1985). Removal of all ants was conducted by widespread application of the baited poison across the designated plot. Selected removal of Pogonomyrmex rugosus was conducted by application of the baited poison at conspicuous mounds of that species (Brown 1998). Plots affected by these treatments are listed in Table 1, columns 2-4.