Patterns of woody plant proliferation in grasslands and savannas influence rates of erosion, spread of disturbance, and nutrient pools.  Spatial pattern is the outcome of plant dispersal, recruitment, competition/facilitation, and disturbance. We quantified effects of livestock grazing, a widely cited driver of shrub encroachment, on the spatial patterns of velvet mesquite (Prosopis velutina) distribution and recruitment in southeastern Arizona.  Field surveys of plant canopy size and location spanning 74 years (1932, 1948, and 2006) were conducted on areas grazed by livestock since the late 1800s and on areas protected from livestock since 1932.  Point pattern analysis and Moran’s I were used to quantify changes in distribution and spatial autocorrelation over time. 

Results/Conclusions 

Mesquite density and cover were similar on grazed (173 ha-1, 12%) and protected areas (203 ha-1, 8%) in 1932 when livestock exclusion was initiated; yet protection from livestock resulted in greater increases in density by 1948 on protected (620 ha-1) versus grazed (325 ha-1) areas.  By 2006, both density and cover were significantly higher on protected (960 ha-1, 22%) versus grazed (433 ha-1, 15%) areas. Univariate analyses with Ripley’s K revealed mesquite plants were randomly distributed on both areas in 1932.  In 1948, following 16 years of protection from livestock, plants on the protected area were clustered at all distances (1 to 20 m), while mesquite plants on the grazed area were clustered at short distances (1 to 4 m) only.  By 2006, mesquite plants on the protected area remained clustered at all distances, with a marked peak in clustering at short distances (1 to 4 m).  In contrast, mesquite plants on the grazed area were randomly distributed.  Bivariate analyses indicated that juvenile mesquite plants (<1 m height) were not significantly associated with adults in any year on the grazed area.  However, juveniles were significantly clustered around adult plants from 2 to 5 m on the protected area in 2006 only.  Moran’s I revealed that spatial autocorrelation in mesquite density did not deviate from random on the grazed area, whereas associations on the protected area became increasingly positive over time.  Grazing influenced recruitment-driven changes in population structure from 1932 to 2006 and protection from livestock promoted the formation of high density shrub clusters generating high spatial heterogeneity.  Collectively, results are counter to the common generalization that livestock grazing promotes woody plant encroachment.  In this system in the absence of fire, grazing retarded rather than promoted woody encroachment over 74 years.