Students of Biology often ask questions such as, How can anyone know what the vegetation and climate were like millions of years ago? A related concept that many students find difficult is the role of modeling in science. Paleobotanists estimate temperature and precipitation for the distant geologic past through the functional morphology of fossil plants, and in particular features that correlate with modern climate, such as the shape and size of leaves. We demonstrate the basis for inferring past climates, plus the role of scientific modeling in paleobotany, through a laboratory exercise in which students use the leaf physiognomy of modern vegetation to estimate mean annual temperature and precipitation for their local region. Students divide into working groups, collect samples of leaves from a local forest, segregate their leaves into species, then calculate: 1) the percentage of species with toothed margins, and 2) average leaf size. Students then estimate local climate by fitting their data to graphs that relate the percentage of species with toothed margins to mean annual temperature (MAT), and the natural logarithm of average leaf size to the natural logarithm of mean annual precipitation (MAP). For San Antonio, Texas, students estimate a MAT of 16 C and MAP of 77 cm (average of 25 working groups), close to average values of 20 C and 76 cm derived from meteorological data. The laboratory exercise demonstrates the basis for estimating past climate from trends in living species and provides a venue for discussing issues of study design, sampling error, and biases caused by the misidentification of species. It is designed for the high-school classroom but is adaptable to the university and younger grade levels. Students develop an appreciation for how scientists reconstruct the past and for the breadth of information that can be inferred from fossil plants beyond simple identification.

Key words: climate, laboratory, modeling, plant, teaching