Global circulation models consistently forecast an increase in the frequency of extreme events such as severe storms and droughts. These changes will alter species interactions and ecosystem functions shaped by precipitation, such as productivity. Ecosystem management will need to anticipate, and where possible, mitigate the effects of increased climate variability in order to maintain ecosystem services and biodiversity. This is a pressing issue for California rangelands, which host a high percentage of California’s endemic plants and support a large ranching industry that depends on reliable forage production. My dissertation uses observational and experimental approaches to understand the implications of increased precipitation variability for the stability of forage production (i.e. aboveground productivity) and the persistence of rare species in California rangelands.
Chapter 1 explores how species interactions affect the stability of aboveground productivity and whether these patterns change along precipitation gradients. I compiled and analyzed nine long-term datasets of plant species composition and aboveground productivity from grassland sites across the United States. I found that productivity in mesic grasslands was stabilized by species richness, whereas productivity in climatically variable grasslands was stabilized by species asynchrony over time. The latter pattern was exemplified by California rangelands, which experienced the most variable precipitation as well as exhibited the most species asynchrony.
Chapters 2 and 3 experimentally test the relationship between precipitation variability and species asynchrony in California rangelands and its implications for the stability of cover and aboveground productivity over time. In Chapter 2, I used rainout shelters and irrigation to experimentally create dry and wet conditions, which I replicated across areas with both low and moderate grazing histories. In moderately grazed areas, my rainfall treatments generated a classic pattern of “grass years” in wet conditions and “forb years” in dry. This pattern helped to stabilize cover across rainfall treatments and is a likely reason for the relationship between precipitation variability and species asynchrony that I observed in Chapter 1. In low grazed areas, however, my treatments essentially generated “grass years” in wet conditions and “no-grass years” in dry; forb cover was both low and unresponsive to rainfall in these areas. This suggests that moderate grazing may be an important management tool to maintain the functional responsiveness of California rangelands to precipitation variability.
Chapter 3 tests whether competitive and functional differences between grasses and forbs affect the degree to which asynchrony stabilizes total biomass production. Within wet and dry plots I manipulated species interactions to create monocultures of Avena barbata (the most abundant grass), Erodium botrys (the most abundant forb) and a mixture of Avena and Erodium. I found that Avena exerted a stronger competitive effect on Erodium under wet conditions relative to dry, which should help stabilize community productivity. However, this effect was overwhelmed by highly unequal production capacity between the two species; Erodium productivity was much lower than Avena and, consequently, tradeoffs between the species did not increase the stability of the mixture relative to either monoculture.
Chapter 4 further investigates tradeoffs between grass and forb years, but in the context of species population dynamics in a ecosystem of conservation concern. Serpentine grassland patches in California host a unique, predominately native flora that is threatened by non-native grass invasion. I focused on a serpentine site that over the past 32 years has exhibited high fluctuations in native forb abundances, and has experienced a series of invasions and subsequent recessions by a non-native annual grass, Bromus hordeaceus. Effective native species conservation and invasive species management require an understanding of what drives such variation in species abundances. I applied a population model to the six most-abundant species at the site – four native annual forbs, a native annual grass and Bromus – to test factors affecting their population size and stability. I found that species could have large population sizes (measured as mean abundance over time) for different reasons – three species had high intrinsic growth rates, whereas the other three, including Bromus and the native grass, had minimal self-limitation. Population stability was highly affected by these differences: species with both low intrinsic growth rates and minimal self-limitation had less stable populations and were more sensitive to rainfall. These findings suggest a framework to describe population stability and to identify which species are likely to be sensitive to environmental change.