Summary• Paleobotanists have long used models based on leaf size and shape to reconstruct paleoclimate. However, most models incorporate a single variable or use traits that are not physiologically or functionally linked to climate, limiting their predictive power. Further, they often underestimate paleotemperature relative to other proxies.• Here we quantify leaf-climate correlations from 92 globally distributed, climatically diverse sites, and explore potential confounding factors. Multiple linear regression models for mean annual temperature (MAT) and mean annual precipitation (MAP) are developed and applied to nine well-studied fossil floras.• We find that leaves in cold climates typically have larger, more numerous teeth, and are more highly dissected. Leaf habit (deciduous vs evergreen), local water availability, and phylogenetic history all affect these relationships. Leaves in wet climates are larger and have fewer, smaller teeth. Our multivariate MAT and MAP models offer moderate improvements in precision over univariate approaches (± 4.0 vs 4.8°C for MAT) and strong improvements in accuracy. For example, our provisional MAT estimates for most North American fossil floras are considerably warmer and in better agreement with independent paleoclimate evidence.• Our study demonstrates that the inclusion of additional leaf traits that are functionally linked to climate improves paleoclimate reconstructions. This work also illustrates the need for better understanding of the impact of phylogeny and leaf habit on leaf-climate relationships.

Abstract,•Leaf mass per area (M^) is a central ecological trait that is intercorrelated with leaf life span, photosjnthetic rate, nutrient concentration, and palatability to herbivores. These coordinated variables form a globally convergent leaf economics spectrum, which represents a general continuum running from rapid resource acquisition to maximized resource retention. Leaf economics are little studied in ancient ecosystems because they cannot be directly measured from leaf fossils. Here we use a large extant data set (65 sites; 667 species-site pairs) to develop a new, easily measured scaling relationship between petiole width and leaf mass, normalized for leaf area; this enables M^ estimation for fossil leaves from petiole width and leaf area, two variables that are commonly measurable in leaf compression floras. The calibration data are restricted to woody angiosperms exclusive of monocots, but a preliminary data set (25 species) suggests that broad-leaved gymnosperms exhibit a similar scaling. Application to two well-studied, classic Eocene floras demonstrates that M^ can be quantified in fossil assemblages. First, our results are consistent with predictions from paleobotanical and paleoclimatic studies of these floras. We found exclusively low-M^ species from Republic (Washington, U.S.A., 49 Ma), a humid, warm-temperate flora with a strong deciduous component among the angiosperms, and a wide M^ range in a seasonally dry, warm-temperate flora from the Green River Formation at Bonanza (Utah, U.S.A, 47 Ma), presumed to comprise a mix of short and long leaf life spans. Second, reconstructed M^ in the fossil species is negatively correlated with levels of insect herbivory, whether measured as the proportion of leaves with insect damage, the proportion of leaf area removed by herbivores, or the diversity of insect-damage morphotypes. These correlations are consistent with herbivory observations in extant floras and they reflect fundamental trade-offs in plant-herbivore associations. Our results indicate that several key aspects of plant and plant-animal ecology can now be quantified in the fossil record and demonstrate that herbivory has helped shape the evolution of leaf structure for millions of years.

No abstract