| I examined the vegetation dynamics of ponderosa pine ( Pinus ponderosa) forests in northern Arizona. Most field work was conducted on the Hill plots, five grazing exclosures established in 1912 and sampled in 1941 and 2004. Meta-analytic techniques were used to combine the results from individual sites. Overstory vegetation increased in dominance since 1941, but less where subject to continued livestock grazing than where protected from livestock grazing since 1912. A state-and-transition model for overstory vegetation dynamics is proposed, suggesting that historical development of the overstory is the result of interactions between livestock grazing, fire history, climate, and seed production.; Between 1941 and 2004, understory species density and herbaceous plant density both declined by 37%, shrub cover by 69%, herbaceous cover by 59%, graminoid cover by 45%, and forb cover by 82%. Declines were due to the increased overstory rather than grazing effects, as were apparent differences between grazing treatments in 2004. Community-level variables masked interspecific differences in temporal dynamics and in response to grazing treatment and overstory condition. Using Indicator Species Analysis, three times more species were identified as indicators of 1941 as of 2004; few species increased in cover during this interval. More species were indicators of grazing treatments in 1941 and of overstory condition in 2004, suggesting that the dominant structuring force in this ecosystem has changed over time. Some species responded to grazing, some to overstory condition, and some to both forces. Most species were consistent indicators across sites.; I developed a new, proportional method of reconstructing historical tree diameters. This method can be used to model overstory growth and estimate historical tree sizes; such information is essential when accounting for overstory-understory relations in these forests.; This study demonstrated that livestock grazing can have complex long-term effects on overstory vegetation, and provided a quantitative analysis of long-term changes in the understory vegetation of southwestern ponderosa pine forests. Since understory community-level variables are affected primarily by the overstory, the observed declines might be reversed if the overstory is thinned as part of ecological restoration activities, though responses of individual species may vary. |
