The saguaro population in a 700-ha area at the Desert Laboratory on Tumamoc Hill, Tucson, Arizona, was censused in 1908, one year after livestock were excluded. In 1964, four 10-ha plots were established within the original area to determine the effect of slope aspect on plant growth and demography. The plots were recensused in 1970 and 1993. We developed a model for determining saguaro age, using the growth rates of over 3000 plants from 1964 to 1970. The model was verified with 1993 data. Changes in population size and the estimated age structures were then used to infer regeneration trends.

Saguaro populations on all slope aspects nearly doubled since 1908. Yet, during the same period, relative abundances of saguaros remained higher on the south and east aspects than on the north and west aspects. Higher recruitment and survival of young plants rather than mortality of older ones are largely responsible for the differences between aspects.

The estimated age structures show large, multi-decadal fluctuations in saguaro regeneration. Prior to 1908, populations on all slopes experienced an extended period of decline beginning around the late 1860s. The recent surge in recruitment began in the 1920s and peaked in the 1970s. Populations currently are again in decline. Better regeneration generally corresponds with relatively wet conditions, and poorer regeneration with drier conditions. However, extended periods of decline often included episodes of relatively wet conditions (e.g., from the 1860s to the 1920s), indicating that other climatic and biotic factors also determine recruitment success. Establishment may have been suppressed by colder winters, livestock grazing, and rock quarrying.

This long-term study demonstrates that, despite low regeneration and population decline during much of the last two centuries, saguaros have persisted on Tumamoc Hill because of episodic surges in seedling establishment.