Krieger, Jonathan , Guralnick, Robert , Johnson, Kirk , Smith, Dena .
Predicting climate using empirically determined continuous measures of leaf shape..
Botanists have developed numerous methods for quantifying correlations between leaf form and climate. Current methodology scores mean leaf area and percentage of species with toothed leaves in an assemblage, two measures that contain little leaf-shape information. Incorporating additional information about leaf form has the potential to both increase our understanding of how climate influences leaf shape, and improve our ability to make predictions about past climates. Our goals in this study were to: 1) identify aspects of leaf shape of non-lobed leaves that can be considered independent characters through empirical measurement of patterns of shape variation; 2) determine which of these characters correlate with broad climate classifications (tropical versus temperate); and 3) assess our ability to predict climate type using leaf shape information from small samples of fossil leaves. Using eigenshape analysis, we identified ten independent aspects of shape variation in a sample of 396 leaves from extant assemblages. Six were correlated with climate: base shape, petiole length, length-to-width ratio, location of widest portion of the leaf, tip shape, and width near the leaf base. There was evidence that length-to-width ratio and width near the leaf base both reflect allometry. Four characters were not correlated with climate: overall asymmetry (wider on left vs. right), extent of widest portion of leaf (variation from elliptic to ovoid), base asymmetry, and tip asymmetry. Using the shape characters generated in the eigenshape analysis, we classified small samples from four fossil assemblages as either tropical or temperate. Castle Rock, a putative tropical assemblage, was classified as tropical in our analysis. All three putative temperate assemblages, Republic, Green River, and Florissant, were classified as temperate. The results underscore the utility of eigenshape analysis as a predictive tool for climate estimation derived from modern and fossil leaf shapes.
1 - University of Colorado, University Museum & Department of Ecology and Evolutionary Biology, 265 UCB, Boulder, Colorado, 80309-0265, USA
2 - University of Colorado, University Museum and Department of Geological Sciences, Boulder, Colorado, 80309-0399
3 - Denver Museum of Natural History and Earth Science, Denver, Colorado, 80205
4 - University of Colorado, Department of Ecology and Evolutionary Biology, UCB 334, Boulder, Colorado, 80309-0334, USA
leaf area analysis
leaf margin analysis.
Presentation Type: Paper
Location: Maybird (Cliff Lodge)
Date: Wednesday, August 4th, 2004
Time: 8:00 AM