Since the time of Linnaeus, plant relationships have been determined largely by looking at fruit and flower characteristics. This is a useful approach, but it doesn't always work. Botanists studying modern tropical plants sometimes have to wait years for a particular plant to flower. For paleobotanists, the problem can be even worse. Think of a tree in a temperate forest. Every spring, it sprouts a new set of leaves and flowers. Some of the flowers flutter away in the breeze. Of the flowers that become fruits, many get eaten, and all of them eventually fall off the tree. In the autumn, the tree sheds all its leaves. In other words, the tree spits out parts all year long but rarely do those parts come off together. Of all the parts plants shed, leaves are by far the most common, but they are seldom attached to the reproductive organs that botanists have traditionally used to classify plants.

• When studying ancient landscapes, morphotyping allows a paleobotanist to classify leaves to determine extinction events and the emergence of new species without having to track down the nomenclature for each fossil. New species found during the study can be formally named through the Linnaean system later.
• Every species is defined by an example specimen, known as the type specimen. In the Linnaean system, once the type specimen is selected, it can never be changed, even though much better specimens may be unearthed later. (Excellent specimens found later can be used to supplement the type specimen.) Because the morphotype approach is informal, holomorphotype specimens can be replaced with better examples as often as necessary.

Toothed leaf (From Manual of Leaf Architecture)

Venation patterns

To morphotype leaves, you start by sorting them on the basis of toothed versus entire (smooth) margins, primary and secondary vein patterns, and the presence and types of lobes. With some exceptions, these characters are usually stable within morphotypes. Leaf size and shape are the least reliable characters in identifying leaves. After the leaves are sorted into these broad categories, you can further divide them by looking at higher order venation patterns and tooth type.

In 1993, Museum volunteer Beth Ellis initiated a six-year project based on Kirk's morphotyping method and leaf architecture terminology established by Leo Hickey of Yale University. Beth's goal was to standardize the terminology used by botanists and paleobotanists to describe leaves. The result is a 65-page Manual of Leaf Architecture, available in PDF format via http://www.peabody.yale.edu/collections/pb/MLA/ (a new browser window will open). Accompanied by a leaf architecture database entry form, this manual walks researchers through the steps of morphotyping leaves based on stable, diagnostic characteristics. Others who worked with Beth and Kirk on this manual were Amanda Ash and Scott Wing of the Smithsonian Institution, Leo Hickey of Yale University, and Peter Wilf of the University of Michigan. Collectively, these individuals are known as the Leaf Architecture Working Group.

Kirk, aided by his interns and Leaf Whackers, has also completed a digital leaf database at the Museum. To ease comparison on venation patterns, fossil leaf specimens are now photographed and converted to high-resolution digital format. These images are burned onto CD-ROM for permanent storage, and can be compared to cleared leaf images from the Smithsonian Institution. (Cleared leaves are modern leaves that have been chemically treated to make their venation patterns more apparent. Many of the images in the Manual of Leaf Architecture are from the Smithsonian's cleared leaf collection.)

A major part of Kirk's research focuses on Late Cretaceous and Early Paleocene leaves from North Dakota, from the Hell Creek and Fort Union Formations. The North Dakota vicinity where Kirk does much of his fieldwork preserves one of the world's best terrestrial records of the K-T boundary, the time when the last dinosaurs went extinct. In 1999 alone, Kirk and his Leaf Whackers identified and catalogued nearly 7,000 fossil leaf specimens, collected from eighty-four different localities between 1991 and 1999.