Many plant traits are sensitive to climate (Woodward, 1987; Breckle, 2002) and paleobotanists commonly use plant–climate relationships to reconstruct ancient climates (Chaloner & Creber, 1990; Parrish, 1998). Most plant–climate studies focus on interspecific patterns, for example correlating leaf area to mean annual precipitation (MAP) across distinct biomes (Givnish, 1984; Wilf et al., 1998). However, intraspecific patterns also provide useful information. Most critically, plant–climate relationships within species are affected by both ecotypic variation of plant traits (as in interspecific patterns) and the plasticity of plant traits. In this study, we examined the strength of correlation between leaf size and shape (physiognomy) and climate for two North American species with broad climatic ranges, Acer rubrum (red maple, Sapindaceae) and Quercus kelloggii (California black oak, Fagaceae). We sampled Q. kelloggii across a large elevation gradient (146–2362 m) but restricted sampling of A. rubrum to lowland areas (< 250 m). Our results provide new information about the sensitivity of leaf traits to climate within species and demonstrate the potential for incorporating intraspecific physiognomic data from fossil plants in paleoclimatic reconstructions.
Leaf teeth and climate
It has long been noted that the percentage of woody dicot species in a flora that are toothed inversely correlates with mean annual temperature (MAT) (Bailey & Sinnott, 1916; Wolfe, 1979, 1993; Wilf, 1997; Jacobs, 1999, 2002; Gregory-Wodzicki, 2000; Kowalski, 2002; Greenwood et al., 2004; Greenwood, 2005a; Royer et al., 2005; Traiser et al., 2005). More recently, Huff et al. (2003) and Royer et al. (2005) reported strong correlations among 17 sites (mostly from eastern North America) between MAT and site-level means of a suite of physiognomic variables, including number of teeth, tooth area and perimeter/area relationships. Compared with warmer sites, colder sites contained species whose leaves generally had more teeth, a larger tooth area and a higher perimeter-to-area ratio (Royer et al., 2005).
The biological basis for these correlations may be related to the observed increases in rates of photosynthesis and transpiration within teeth early in the growing season (Baker-Brosh & Peet, 1997; Royer & Wilf, 2006). This increase in sap flow presumably enhances the delivery of solutes to young emerging leaves and to recently dormant leaves, which may confer an advantage to plants in progressively colder environments with shorter growing seasons (Royer & Wilf, 2006). Thus, leaves with many large teeth may be adaptive in cold climates. This proposed mechanism also provides an explanation for why at a given MAT toothed species are proportionately more abundant in physiologically wet environments (Bailey & Sinnott, 1916; MacGinitie, 1953; Wolfe, 1993; Burnham et al., 2001; Kowalski & Dilcher, 2003; Greenwood, 2005b) because in these environments the impact of water costs associated with leaf teeth are less severe. Hydathodes in teeth may also serve to release excess root pressure via guttation, thereby preventing the flooding of intercellular airspaces; this process could be beneficial to plants in cold climates where freeze–thaw embolisms are more prevalent (Feild et al., 2005).
Intraspecific patterns between leaf physiognomy and climate
Building on the interspecific work of Bailey & Sinnott (1916) and others described above, a currently unresolved question is how tooth size, shape, and number respond to climate within species. Royer et al. (2005) reported significant correlations between MAT and physiognomic variables related to tooth size, tooth number, and perimeter/area within four species in the eastern USA (A. rubrum, Prunus serotina, Ostrya virginiana, Carpinus caroliniana). However, because the level of sampling within this data set was low (n ≤ 12 sites for all species), these patterns could only be considered preliminary.
Investigating the role of intraspecific variation in leaf–climate relationships is important for at least three reasons. First, such work bears directly on the broader issue of how plasticity and genotype influence the sensitivity of leaf form to climate. Many studies have sought to tease apart how plasticity and genotype affect the relationships between plant traits and climate, and considerable progress has been made in the areas of physiology, growth, stomatal patterning and leaf size (Gurevitch, 1988; Williams & Black, 1993; Morecraft & Woodward, 1996; Beerling & Kelly, 1997; Cordell et al., 1998; Oleksyn et al., 1998; Imbert & Houle, 2000; Hovenden, 2001; Flann et al., 2002; Hovenden & Vander Schoor, 2004). By contrast, very little is known about the impact of plasticity vs genetic determination on the relationships between leaf shape and climate. To rigorously test the role of leaf plasticity in leaf–climate relationships, transplant studies are required (Hovenden & Vander Schoor, 2004); however, measuring the physiognomic variability across the native ranges of individual species (which reflects both plasticity and differences in genotype) is an important first step.
Second, the paleobotanical community has applied leaf–climate relationships to fossil leaf floras for nearly a century to quantitatively reconstruct climate (Bailey & Sinnott, 1915; Dilcher, 1973; Wolfe, 1978, 1993; Wolfe & Upchurch, 1987; Greenwood & Wing, 1995; Utescher et al., 2000; Jacobs, 2002; Wilf et al., 2003; Wing et al., 2005; Miller et al., 2006). Most of the methods used in these studies are heavily dependent on tooth characters. Implicit in these paleobotanical studies, particularly those at high temporal resolution, is that plant traits respond rapidly to climate change in a predictable fashion and that a given climate will always select for the same range in leaf physiognomy (Christophel & Gordon, 2004). Therefore, if intraspecific responses of tooth morphology to climate were shown to be broadly similar to the interspecific patterns, this would further emphasize the value and reliability of paleoclimatic reconstructions based on leaf physiognomy.
Third, the nature of intraspecific patterns may confer preference to one leaf-paleoclimate method over others. For example, the method that reconstructs paleotemperature from the percentage of toothed species in a flora (‘leaf-margin analysis’) is dependent on a single, binary character (presence vs absence of teeth) that is fixed in most species. By contrast, the method presented by Huff et al. (2003) and Royer et al. (2005) (‘digital leaf physiognomy’) is based on multiple continuous variables (for a method based on multiple categorical variables see Wolfe, 1993). As a result, digital leaf physiognomy more fully captures the spectrum of physiognomic variation: for example, a species with a large variability across its geographic range in tooth variables (e.g. tooth count and tooth area) would be scored identically with leaf-margin analysis, but could be differentiated with digital leaf physiognomy. Therefore, a potential advantage of digital leaf physiognomy is that intraspecific patterns can contribute to the site-level means of the variables, for example if a toothed fossil species had a greater number of teeth in the cold end of its natural range.
Here, in an effort to more firmly ascertain the influence of climate on leaf physiognomy within species, we report results from a large data set that includes two North American woody plants (A. rubrum and Q. kelloggii) whose native ranges span large MAT gradients and are not closely related to each other. The results of our study provide a test for the importance of intraspecific patterns in leaf–climate relationships and for the usefulness of leaf–climate methods that are based on continuous physiognomic variables (e.g. digital leaf physiognomy) vs binary variables (e.g. leaf-margin analysis).