Plant species richness in riparian wetlands - a test of biodiversity theory

Ecology, Jan, 1998 by Michael M. Pollock, Robert J. Naiman, Thomas A. Hanley

INTRODUCTION

Disturbance, productivity, and spatial heterogeneity are major factors regulating species richness in plant communities (Groombridge 1992, Ricklefs and Schluter 1993, Huston 1994). Along disturbance gradients, species diversity has been found to be highest at intermediate levels (Connell 1978, Fox 1979, Huston 1979, Sousa 1979, Ward and Stanford 1983). However, others have found diversity increases (Tilman 1983, Kneidel 1984, Bailey 1988, Grassle 1989, Kaczor and Hartnett 1990, Phillips et al. 1994), or decreases (Gaedeke and Sommer 1986, Robinson and Minshall 1986, Bailey 1988, Brown and Brussock 1991, Englund 1991, Wilson and Tilman 1991) along disturbance gradients. The relationship between productivity and species diversity has been described as unimodal (Grime 1973, Al-Mufti et al. 1977, Tilman 1982, Wheeler and Giller 1982, Moore and Keddy 1989), but diversity has also been correlated with decreasing (Rosenzweig 1971, Huston 1980b) or increasing (Brown and Gibson 1983, Currie and Paquin 1987, Currie 1991) productivity. Spatial heterogeneity in the physical environment (e.g., substrate, nutrients, soil moisture and structure) has been positively and linearly correlated with diversity at a number of spatial scales (Harman 1972, Cody 1975, Schlosser 1982, Tonn and Magnuson 1982, Crozier and Boerner 1984, Chambers and Prepas 1990, Kaczor and Hartnett 1990, Pringle 1990, Scarsbrook and Townsend 1993).

Most empirical studies and several theories and models (Horn and MacArthur 1972, Grenney et al. 1973, Levin and Paine 1974, Abrams 1988) examined the singular role disturbance, productivity, or spatial heterogeneity had on species diversity, while ignoring possible interactions among these factors. An exception is Huston's (1979, 1994) dynamic-equilibrium model (DEM), which predicts diversity patterns based on the interaction of productivity and disturbance [ILLUSTRATION FOR FIGURE 1 OMITTED]. The DEM also predicts that temporal asynchrony of disturbances across an otherwise homogenous landscape creates spatial heterogeneity (Loucks 1970, Huston 1994). This increases diversity by creating patches of different seral stages containing different suites of species. Related to this idea is that landscape heterogeneity may result in spatial variation of disturbance frequencies. If disturbance frequencies vary over space, a landscape could also contain patches of different seral stages.

In this study a model was developed, based on empirical data, to explain how plant diversity is influenced by a disturbance regime (flooding) and by spatial heterogeneity (microtopography), which controls small-scale spatial variation of the flooding regime. We also used our data to test predictions of the DEM at two spatial scales.

Huston's (1979, 1994) DEM explains how disturbance and productivity influence species richness. The model predicts that when the opposing forces of disturbance and productivity are in dynamic equilibrium, diversity will be high. Fig. 1 illustrates that the diversity maximum should be in the area of the disturbance-productivity continuum where productivity and disturbance frequencies are relatively low, but not extremely low.

Previous studies demonstrated that the DEM has predictive power under controlled experimental conditions (Huston 1980a, Rashit and Bazin 1987). However, we know of no published studies testing the model effectively at the scale of communities, under naturally occurring conditions. If a model has relevance to ecological systems, its predictions should be apparent in natural systems. There are few systems where communities vary widely in natural levels of disturbance, productivity, and diversity. However, low-gradient riparian corridors often contain diverse patchworks of highly dynamic, species-rich communities that range widely in productivity and disturbance frequency (see reviews in Malanson 1993, Naiman et al. 1993), making them ideal systems for field-testing the DEM. Unfortunately, many riparian corridors have been invaded by exotic species, and their disturbance regimes altered by human activities. It is not clear whether the predictions of the DEM are applicable in systems undergoing such fundamental changes in diversity and disturbance patterns. To avoid this potentially confounding problem, study sites were selected in pristine watersheds.

The purpose of this study was twofold: (1) to determine if flood frequency and productivity can predict species richness between (primarily riparian) wetlands (i.e., testing the DEM) and to see how these relationships change with differing scale of observation (1000 [m.sup.2] and 1 [m.sup.2]), and (2) to determine if flood frequency and the spatial variation of flood frequencies can explain patterns of plant species richness at the community (1000-[m.sup.2]) scale.

METHODS

Site location

The 16 wetlands sampled for this study are on Chichagof Island in southeast Alaska, [approximately]115 km southwest of Juneau in the temperate coastal rain-forest ecoregion (Naiman et al. 1992). The climate is maritime, with moderate temperatures throughout the year. The nearest long-term climatological records are from the town of Tenakee Springs (Sidle and Swanston 1982), [approximately]5 km north of the mouth of the Kadashan River [ILLUSTRATION FOR FIGURE 2 OMITTED]. Mean annual temperature is 6.3 [degrees] C, with monthly averages ranging from -1.6 [degrees] C in January to 13.5 [degrees] C in August. Mean annual rainfall is 167 cm. The highest mean monthly precipitation occurs in October (30.2 cm), September (20.2 cm) and November (17.5 cm). Thirteen sites were in the Kadashan River basin (latitude 57 [degrees] 39[minutes]46[seconds], longitude 135 [degrees] 11[minutes]06[seconds]), a 145-[km.sup.2] watershed with a broad, glacially carved valley. One site was in East Crab Creek basin, and two sites were along the shore of Saltery Lake [ILLUSTRATION FOR FIGURE 2 OMITTED].