Light occlusion at forest edges: an analysis of tree architectural characteristics

Abstract

We examined the phenomenon of tree canopy closure by lateral branch expansion at the edges of temperate forest fragments bordering fields, in southwestern Ontario, Canada. We evaluated the relationship between tree crown characteristics and the degree of light reduction (a measure of canopy closure) among six tree species of differing shade tolerance (Fagus grandifolia Ehrh., Acer saccharum Marsh., Quercus rubra L., Prunus serotina Ehrh., Juglans nigra L. and Populus grandidentata Michx.). Tree species differed in their capacity to occlude light: the most effective were the most shade tolerant species, F. grandifolia and A. saccharum, while the most shade intolerant trees, J. nigra and P. grandidentata, had the highest percentage of light transmission. The most shade tolerant species presented most of their branches and foliage towards the field and resulted in the most voluminous crowns with the largest values in branch density, branch clumping, canopy depth and canopy width. The best model to predict light occlusion was obtained when we considered the effect of each species as a factor, and foliage biomass facing the field as covariates; this explained 80% of the total variation in light transmission. The ability of shade tolerant trees to enlarge their crowns in response to an increase in light availability allows them to occlude light better. To maximize the protection of microclimates in forest fragments, shade tolerant trees should be encouraged at forest edge communities by favoring recruitment or by selective thinning of the least shade tolerant species.

Introduction

The great majority of forest ecosystems in the world have been fragmented by anthropogenic activities converting continuous forests into landscapes of remnant forest patches, embedded in agricultural matrices. The switch from a continuous cover to small fragments brings with it many changes in physical, chemical and abiotic flows in the altered landscape (e.g. Yates et al., 1994, UNEP, 1995, Malcolm, 1998) as well as a suite of new habitat conditions that may be rare or non-existent in continuous forest communities.

The most obvious effects of forest fragmentation are a decreased forest area with increased isolation of the biota and the appearance of permanent, rather than ephemeral, edge habitats. An edge can be considered a transitional ecosystem between the forest and the surrounding open areas, whose creation may induce a changed forest microclimate (Kapos, 1989, Williams-Linera, 1990, MacDougall and Kellman, 1992, Malcom, 1994, Didham and Lawton, 1999), higher tree mortality (Lovejoy et al., 1986, Kapos et al., 1997), greater tree density (Palik and Murphy, 1990, Williams-Linera, 1990), increased plant growth and foreign invasions (Brothers and Spingarn, 1992, Murcia, 1995) and lower recruitment of herbs (Benitez-Malvido, 1998, Jules, 1998). Avian diversity may also be reduced at the edges by an increased nest predation (Gibbs, 1991, Burkey, 1993, Soderstrom, 1999).

Microclimatic variables, such as air temperature, vapor pressure deficit, and light intensity can vary greatly along an edge-to-interior gradient (Murcia, 1995, Kapos et al., 1997, Didham and Lawton, 1999), due to the increased amount of solar radiation and wind that penetrates into the forest interior (Ranney et al., 1981, Matlack, 1993). Recently created forest fragments possess an open edge where trees lack any significant lateral crown spread and a deep penetration of atmospheric conditions exists (Matlack, 1993). However, trees appear to respond to increased light levels at these boundaries and eventually, edges with trees exhibiting extensive lateral crowns develop (Fig. 1). At long-established edges, these well-developed lateral canopies appear to promote effective closure and may insulate the forest interiors from edge effects (Brothers and Spingarn, 1992, MacDougall and Kellman, 1992). As a result, atmospheric edge effects may become considerably attenuated with time.

Tree species may vary in their capacity to expand their branches laterally. Sakai (1987) and Canham et al. (1994) suggested that those species that have larger crowns, in terms of radius, may have better lateral branching growth in the northern temperate zone. Such trees include species of Acer, Carya and Fagus which are classed as shade tolerant. In contrast, less shade tolerant species appear to be less prone to developing and sustaining lateral canopies.

Although many mathematical models that simulate light interception and crown characteristics have been developed (e.g. Kurth, 1994, Niklas, 1994, Hilbert and Messier, 1996, Pearcy and Yang, 1996, Takenaka et al., 1998), the adaptive and functional significances of lateral branching at permanent forest edges have received far less systematic discussion (for exceptions, see Canham et al. (1994) and Millet et al. (1998)).

The aim of the present study was to examine tree canopy closure at the edges of forest fragments and to evaluate whether there are significant differences between species that could affect their capacity to insulate the fragment from atmospheric edge effects. We ask how species that evolved in forests without permanent edges respond to this new habitat. We present two hypotheses: (1) Species of greater shade tolerance will close an edge more effectively than those of less shade tolerance. (2) Tree species will vary in their capacity to close an edge as a consequence of differences in branching patterns, foliage density and foliage arrangement. To test these hypotheses, the degree of light reduction was used as a measure of canopy closure, and we analyze the relationship between tree architectural characteristics and light occlusion among tree species of differing shade tolerance.