Sonoran Desert annual plants: Empirical tests of models of coexistence and persistence in a temporally variable environment.

Abstract

The desert offers windows of opportunity to annual plants, but they must deal with temporal variation in environmental conditions. This dissertation explores the idea that temporal variation plays a role in species coexistence, enhancing the diversity of desert annuals. Theory suggests that for temporal variability to promote coexistence among annuals: (1) species must differ in their years of highest fitness, (2) species must have long-lived seed banks, (3) the success of an abundant species must be limited in its otherwise good years by competition. Chapter 1 reviews mechanisms of coexistence applicable to terrestrial plants. Chapter 2 reviews the population biology of Sonoran Desert annuals. Chapter 3 investigates whether temporal variability might allow three different species (Pectocarya recurvata, Plantago patagonica, Schismus barbatus) on a creosote flat to out-perform each other in different years. In a two year experiment, I simulated additional year-types by manipulating factors that vary across years (water and seedling density), and incorporated shrub-covered and open microhabitats. I mapped seedlings, observed rates of herbivory and reproductive success. To compare species, I calculated the average value that seeds have for population growth from 10 years of data. I found shifts in competitive hierarchies for two species pairs, depending on year-type factors. Furthermore: (1) herbivory may contribute to shifts in competitive hierarchies and (2) habitat partitioning was not evident. Chapter 4 quantifies dormancy and germination fractions in the field for a guild of winter annuals. Dormant seeds were removed from soil samples collected after germination, but prior to new seed set. Seedling densities and reproductive success were followed in nearby plots. The species with the largest dormant seed bank had higher temporal variation in reproductive success (over the last 10 years) and tended to have smaller seeds, consistent with the theory that seed dormancy and large seeds are partially substitutable bet-hedging strategies. Plants germinated more in years of higher reproductive success, suggesting that germination could be "predictive". In addition, species responded differently to years. I discuss how these experiments demonstrate that this system possesses the traits required for temporal variation to promote coexistence.