RESUMO
In 1939, an experiment was established on the Jornada Experimental Range to evaluate the effects of shrub removal, rabbit exclusion, furrowing, and seeding in creosotebush [Larrea tridentata (DC.) Cov] vegetation. Sixteen plots (21.3×21.3 m) were laid out in four rows of four plots per row with a buffer zone of 7.6 m between plots and rows. A barbed wire fence excluded cattle and poultry wire fencing excluded lagomorphs. Treatments were factorially applied at two levels. Plant cover in the plots was sampled in 1938 (before treatment), 1947, 1956, 1960, 1967 and 1989 with randomly located, line-intercept transects. Data from all sampling dates were analyzed as a split plot in time and main effects for 1989 tested by analysis of variance for a 2×4 factorial experiment. There were significant (P<0.10) year x treatment interactions. Seeding and furrowing treatments were ineffective but lagomorph exclusion and shrub clearing treatments resulted in significant treatment differences for several species. In 1989, basal area of spike dropseed (Sporobolus contractus A.S. Hitchc.) was 30-fold greater on the lagomorph excluded than on the lagomorph unexcluded treatment. Canopy cover of honey mesquite (Prosopis glandulosa Torr. var. glandulosa), tarbush (Flourensia cernua DC.) and mariola (Parthenium incanum H.B.K.) were affected by lagomorph exclusion. None of the responses were viewed as successional in nature. They principally represented individual species sensitivities to either absence of a primary herbivore or removal of aboveground shrub biomass. Though the physical treatments could be regarded as relatively severe disturbances of the system, the impacts on community vegetation dynamics were relatively insignificant.
