Spatial distributions of understory light along the grassland/forest continuum: effects of cover, height, and spatial pattern of tree canopies

Abstract

The understory light environment is a key determinant of vegetation pattern and ecosystem processes, and varies spatially perhaps more than any resource used by plants. Understory light varies along gradients of vegetation structure that range from grassland with no woody canopy cover to forest with nearly complete woody canopy cover. Spatial variability in understory light is largely determined by several characteristics of overstory plants — spatial pattern, height, and cover — which vary concurrently along the grassland/forest continuum. Using a spatially-explicit ray-tracing model, we quantified trends in mean and variance of understory light along the continuum. We modeled understory light over a growing season for two types of plots: (1) generated plots in which cover, spatial pattern, and height of trees were varied systematically, and (2) three actual plots using stand data from piñon-juniper woodland sites for which cover, spatial pattern and height varied concurrently. Mean understory light decreased with increasing canopy cover and was sensitive to changes in height, as expected, but was not sensitive to spatial pattern. Variance in understory light was maximum at an intermediate value of cover that was dependent on both spatial pattern and cover — maximum variance occurred at lower values of cover as height increased and as spatial pattern progressed from regular to random to aggregated. These trends in the overall patterns of understory light were also examined with respect to changes in understory light in canopy and intercanopy locations. Variance in understory light for intercanopy locations was less than that for canopy locations at low canopy cover, but exceeded that for canopy locations as canopy cover increased. The value of canopy cover at which variance in intercanopy locations exceeded that in canopy locations was sensitive to variation in height but not in spatial pattern. The distributions of understory light for the actual plots were generally similar to those for corresponding generated plots, with dissimilarities attributable to differences in cover and height. The general trends highlighted by our simulations are broadly applicable to sites along the grassland/forest continuum.

Introduction

The understory light environment is a key determinant of vegetation pattern and ecosystem processes, and varies spatially perhaps more than any resource used by plants (Bazzaz, 1996). Spatial patterns of understory light are determined by several overstory characteristics, particularly spatial pattern, height, and cover of woody plants. These overstory characteristics vary concurrently along vegetation gradients and may produce complex patterns of understory light across vegetation gradients because the spatial pattern and height of trees can change with increases in woody-plant cover (Padien and Lajtha, 1992). These gradients of vegetation structure can be viewed as a continuum that ranges from grassland with no woody canopy cover to forest with nearly complete woody canopy cover (Belsky and Canham, 1994, Breshears and Barnes, 1999). Many extensive semiarid shrublands and woodlands span substantial portions of the grassland/forest continuum. For example, piñon-juniper woodlands, which are the most extensive ecosystem type in the western USA, can span a large proportion of this continuum (West, 1988, Padien and Lajtha, 1992, Milne et al., 1996). The understory light environment is important in these woodlands because of its effects on microclimate (e.g. solar radiation, soil temperature, leaf temperature, soil evaporation), which is strikingly different in canopy patches (i.e. directly below tree crowns) than in intercanopy patches (Breshears et al., 1997b, Breshears et al., 1998). The center portions of canopy patches receive 40% less near-ground solar radiation than intercanopies through the year, and >50% less during much of the growing season (Breshears et al., 1997b). Consequently, soil temperatures in intercanopy patches can exceed those in canopy patches by >10°C in summer, which in turn results in increased soil evaporation rates in intercanopy patches (Breshears et al., 1998). Further, microclimate affects plant processes such as seedling establishment, germination, facilitation, and growth (Floyd, 1983, Padien and Lajtha, 1992, Martens et al., 1997), as evident in the differences in distributions of understory plants in canopy and intercanopy patches (Arnold, 1964, Lohstroh and Van Auken, 1987, Armentrout and Pieper, 1988, Van Auken and Lohstroh, 1990, Ludwig et al., 1997). Hence, the spatial distribution of understory light—and in particular the difference between canopy and intercanopy patches — can have a large influence on vegetation dynamics for sites along the grassland/forest continuum.

Several studies have quantified heterogeneity of understory light at an individual site along the grassland/forest continuum — in piñon-juniper woodlands (Lin et al., 1992, Breshears et al., 1997b) as well as other systems (see reviews in Scholes and Archer (1997) and McPherson (1997)). However, a systematic analysis is lacking of how understory light patterns vary along the grassland/forest continuum with changing stand characteristics (spatial pattern, height, and cover). Our first objective was to evaluate the relative importance of changes in canopy structural characteristics (cover, spatial pattern, and height) in the understory light environment for sites along the grassland/forest continuum. To address this objective we systematically varied these three structural characteristics independently to generate plots with differing overstories. Along the grassland/forest continuum, however, spatial pattern, height, and cover vary concurrently and may be not be independent (Padien and Lajtha, 1992). Consequently, our second objective was to quantify the spatial pattern in the understory light environment in three actual woodland sites along an elevational gradient for which cover, spatial pattern, and height of overstory trees changed simultaneously. The sites were piñon-juniper woodlands in northern New Mexico, USA. We addressed these objectives using a ray-tracing model to estimate the understory light. Our results quantify how changes in canopy structural characteristics produce complex patterns of understory light. The trends we found are broadly applicable to other sites along the grassland/forest continuum.