I headed to eastern Washington in early spring to measure balsamroot shoots as they began to grow. The northernmost site was a single population of the widespread arrowleaf balsamroot. The site farthest to the southwest consisted of a single population of deltoid balsamroot, along with an outgroup species, northern mule's ears. I had three sites for Carey's balsamroot, a species associated with the Missoula floods that scoured eastern Washington some 12,000 years ago. One of these was near Ginkgo State Park, with its ancient, petrified stumps. I also had a population of the small, carrot-like Hooker's balsamroot here. Another site at Sun Lakes included both Carey's and Hooker's balsamroot. A last population of Carey's balsamroot was located at Palouse Falls. All sites were many miles apart from each other, making pollen transfer among them unlikely. At beautiful Palouse Falls was a plaque recognizing J. Harlen Bretz, the geologist who first suggested that many features of the region were due to massive, catastrophic floods and not gradual erosion.
I met Tom Henn, Head Ranger at Palouse Falls and Lyon's Ferry. We had a fine time in the land of sagebrush, fell in love in spite of the long-distance relationship, and eventually got married. I was sad to leave eastern Washington after my first year of data collection, but had to head to Cedar Point Biological Station in the Nebraska Sand Hills for a month of work before heading back to Lincoln. John taught field ichthyology at the station, and I got to know botanist Dave Sutherland, who would also be on my committee. Tom stayed in touch and visited me in Nebraska. I headed back to eastern Washington the next spring to gather a second year of growth measurements from the same plants for comparison.
Principal components analysis was the method I used to track patterns of development within species, and cladistics provided the phylogenetic framework for my study of these two levels of the evolutionary hierarchy - closely related species and the development of individuals. S.J. Gould offered a nice summary of principal components, and also a warning about relevant variables. Each variable I measured on the plants contributed to an overall correlation matrix summarized by the first two principal components. I used these first components as axes and plotted the values for each variable against them, making star-like graphics that showed changing organization through developmental time. I wanted to see how the changes during growth might be reflected in relationships among species. The patterns weren't clear, but the method was a good way to explore levels of the evolutionary hierarchy.