Thresholds, memory, and seasonality: understanding pulse dynamics in arid/semi-arid ecosystems.

Abstract

Changes in resource availability can alter the functioning of ecosystems, especially with regard to both population dynamics and the cycling of organic matter and nutrients. At perhaps one end of the spectrum, arid and semi-arid ecosystems represent an extreme in which essential resource availability (e.g., water) is discontinuously available and the availability of these resources impact the ecosystem as discreet pulse events interspersed among long periods of limited resource availability. The objective of the special section that follows is to contribute to our understanding of how pulsed events shape population dynamics, species interactions, and ecosystem processes in arid and semi-arid ecosystems. There has been a renewed interest in recent years in the relationships between precipitation and ecosystem processes in arid and semi-arid lands. For many years, studies focused on the effects of mean precipitation, averaged at annual, seasonal, or monthly time scales. These studies suggested broad-scale relationships between precipitation amounts and seasonal distribution and the productivity of ecosystems (Le Houerou 1984; Le Houerou et al. 1988), or the plant functional type composition of water-limited systems (e.g., Neilson 1995; Smith et al. 1997). Though some researchers asked questions about the effects of single rainfall events early on (e.g., Went and Westergaard 1949; Beatley 1974; Sala and Lauenroth 1982), this perspective received wider attention only in the last decade, in part through the development of stable isotope tracer methods in plant ecology (e.g., Caldwell et al. 1985; Ehleringer et al. 1991; Lin et al. 1996). More recently, studies also have begun to focus on intra-seasonal precipitation patterns, asking whether variability in rainfall event size, frequency, and timing alone affect the biological processes in water-limited ecosystems (e.g., Goldberg and Novoplansky 1997; Knapp et al. 2002; Schwinning et al. 2003). With a view to addressing these questions, an international workshop was held at the University of Arizona in Tucson, USA (http://ag.arizona.edu/research/schwinn/ workshop.html). This special issue contains both peerreviewed synthesis papers that were generated through the collaboration of workshop participants, as well as original research contributions related to pulse dynamics in arid and semi-arid ecosystems. A generation ago, two major paradigms for understanding rainfall effects in water-limited ecosystems were published: (1) the pulse-reserve paradigm of Noy-Meir (1974) and co-workers, and (2) Walter’s (1971) two-layer soil water-partitioning model. Our understanding has advanced significantly since then, moving from the largely conceptual paradigms to an invariably more complex body of data and theory. Reynolds et al. (2004) critically review both paradigms, concluding that the accurate description of soil moisture dynamics is pivotal for addressing the dynamics of primary production and soil water partitioning. While Walter (1971) envisioned water partitioning in vertical space to govern major patterns of plant functional type distribution, Ehleringer et al. (1991) and Lin et al. (1996) have provided strong evidence of this spatial partitioning. Evidence is mounting that partitioning over time may be at least as important (Reynolds et al. 2004; Schwinning et al. 2004a, b). However, the consequences of the hydraulic redistribution of water for species S. Schwinning (*) School of Natural Resources, University of Arizona, Biological Sciences East 325, Tucson, AZ, 85721 e-mail: schwinn@Ag.arizona.edu