Effects of Plant Functional Groups on Vegetation Dynamics and Ecosystem Properties

LOSS OF BIODIVERSITY has likely been the most dramatic change humans have imposed on ecosystems in the past century (Chapin et al., 2000); the global extinction rate is currently between 100 and 1000 times faster than pre-human extinction levels (Pimm et al., 1995). There is growing concern that this species loss will have important effects on ecosystem functioning: that speciespoor ecosystems may perform differently, or less efficiently than the species-rich systems from which they are derived (Zedler et al., 2001). This concern has prompted much research to focus on how biodiversity loss affects ecosystem functioning (e.g., Hector et al., 2000; Pfisterer and Schmid, 2002) and the response of ecosystems to global change (Reich et al., 2001). The resulting studies have created a decade-long debate on the relationship between biodiversity and ecosystem function. An emerging conclusion is that the composition of the community, as well as diversity, plays a major role in controlling ecosystem function (see Hooper and Vitousek, 1998; Scherer-Lorenzen et al., 2003): in fact, the types of species in the community may play an even larger role than the number of species. Despite this realization, virtually no studies have specifically examined the independent effects of species composition on the functioning of ecosystems. Species composition is likely to play an important role in determining ecosystem function because species differ in their traits. The effect of the loss of a species on an ecosystem is the result of both (1) the loss of the direct effects of the organism on ecosystem functioning and (2) the response of other organisms to that loss. These effects and responses occur through numerous mechanisms. For example, species can directly affect soil nutrient and water content through varying root mass. In addition, specific species can alter plant community composition through varying competitive abilities and facilitative effects, which in turn may affect ecosystem function. To date, most experimental biodiversity work has used random assembly experiments, which contain artificially assembled communities of local plants (e.g., Hooper and Vitousek, 1997; Tilman et al., 1997; Hector et al., 1999; Fridley, 2003). Recently, however, removal experiments in natural communities are being promoted as a more realistic way to examine the consequences of biodiversity loss (Diaz et al., 2003). The major difference between random assembly experiments and removal experiments is that the manipulated communities have gone through different assembly processes: removal experiments are based on naturally assembled communities and therefore may include important natural processes that might be underestimated by random assembly experiments. My PhD research uses a removal experiment to examine the roles of different plant functional groups (groups of plants that have similar roles in a community, e.g., grasses, legumes) both in influencing plant community dynamics (responses of other functional groups to the loss of a particular group) and in determining ecosystem function (properties and processes of an ecosystem affected by the biota). Specifically, my questions are: