Overview
I have diverse research interests in the theory and practice of plant systematics. My taxonomic interests include higher level analysis of seed plant and angiosperm relationships, and relationships of Hamamamelid and Rosid ordinal and family relationships. I work at the generic and species level within Fagaceae, and in particular in Quercus. I anticipate completion of the taxonomic treatment for Quercus for Flora Mesoamaricana this summer (2003). I am also very much involved in paleobotanical studies of Tertiary Fagaceae and a wide assortment of Late Cretaceous angiosperm floral remains in collaboration with Dr. W.L. Crepet. We are attempting to place a diverse Cretaceous flora into phylogenetic context and anticipate a rather extensive body of research in this area over the next few years. I also recently have developed a collaboration with Dr. David Dilcher (University of Florida) on early Cretaceous angiosperms of China, resulting in a publication in Science in 2002 describing what appears to be one of the first flowering plants. My theoretical interests center on methods of phylogenetic analysis, computer programs for analysis and display of phylogenetic trees, molecular and morphological information, supertree construction, molecular sequence alignment. I am also working on theory of diagnostic analysis, molecular diagnostics, and development of interactive morphological keys. My recent work has focused on developing algorithms for rapid parsimony analysis (although these can also be applied to maximum likelihood). I am currently working on new methods for faster DNA sequence alignment that should solve some existing problems that are limited by computation time. I have written several software packages, including CLADOS, DADA, Winclada, and Encino, that are used to analyze character distributions and augment other phylogenetic software such as NONA and Hennig86. I am coauthor of a new software package, TNT, with Pablo Goloboff and James S. Farris, that provides tools for rapid phylogenetic analysis of very large data matrices. In the area of biodiversity, I have developed a simple method for measuring phylogenetic diversity and thus maximizing sample diversity based on phylogenetic information. This method will be expanded and directly implemented in upcoming software that will provide priority lists for sampling while undertaking biological prospecting. I am also actively involved in both theoretical and practical approaches to bioprospecting. Finally, I have developed an extensive website system (www.plantsystematics.org, www.diversityoflife.org) that serves images of plants and animals, provides diagnostic interactive keys, and links all of these with specimen data, phylogenetic trees and classifications. The diagnostic programs I have developed run on desktops, handhelds, and through the web interface and are based on new algorithms for maximizing information retrieval.
Research Focus
My research activities are rather broad, and be categorized as follows:
1. Theory and methodology of phylogenetic analysis.My background in this area includes several papers focusing on various theoretical issues in phylogenetic analysis, and software implementations of "user-friendly" programs for data visualization and analysis.Perhaps my most important contribution in this area was development of a rapid tree-search algorithm (the "parsimony ratchet") that uses a directed random walk by iteratively reweighting variables to increase efficiency of tree search.The parsimony ratchet was the first of the 'rapid" methods that are now available for tree search, and improved tree search efficiency and effectiveness by orders of magnitude over existing strategies as implemented in PAUP.I first implemented the algorithm through the program NONA (written by Pablo Goloboff) and it has since been implemented in at least two versions that utilize PAUP* (e.g., Sikes and Lewis, 2001), as well as in TNT (Goloboff, Farris and Nixon, 2001, 2003). The ratchet has been successfully employed to analyze some of the largest data sets in plant systematics, e.g., the 500 taxon chloroplast data set (Nixon, 1999) and 567 taxon 3-gene data set (Soltis et al., 2000) and is now commonly used in calculation of supertrees.
2. Phylogenetic relationships and diversity of higher seed plants and angiosperm groups.This is intimately tied with the paleobotanical research in collaboration with Crepet and Gandolfo here at Cornell (and others outside of Cornell).I have been involved in several large analyses of seed plant/angiosperm relationships, beginning with a morphological analysis (Nixon et al., 1994) and the first combined molecular/morphological analysis (Albert et al., 1994) and continuing with reanalysis of the 500-taxon rbcL data set (Nixon, 1999; Davis, Nixon, and Little 2004), the original analysis of the "3-gene" angiosperm data set (Soltis et al., 2000)and the combined molecular/morphological analysis of Archaefructus (Sun et al. 2000).I plan to continue these large-scale analyses, combining new approaches for the analysis of large data sets (e.g., the parsimony ratchet and Goloboff tree search methods) with new morphological data sets over large numbers of taxa.
3. Fossil history of angiosperms. I have a long standing research program in fossil angiosperms, beginning with a review of Tertiary fossil oaks in my Ph.D. thesis, continuing with papers on the earliest fossil evidence of Fagaceae in the Tertiary, and continuing with studies of Cretaceous fossil flowers in collaboration with Crepet and Gandolfo (see publication list).My role in these collaborative studies has been to apply my knowledge of angiosperm diversity and identification (see teaching section) in preliminary fossil identifications, followed by intensive cladistic analysis of morphological or combined morphological/molecular datasets to accurately and reproducibly identify fossils.Our studies of Cretaceous fossil flowers remain the most extensive, and I believe the most completely analyzed, of any angiosperm fossil flower sites, and provide the best view of the Late Cretaceous diversification of angiosperms into modern taxonomic groups.These papers will remain important for an indefinite period of time as documentation of this diversity, and are useful in providing calibration points for molecular dating studies. I also was invited to collaborate on an analysis of what may be the oldest angiosperm fossil flower, Archaefructus.My role in this paper was to develop the morphological matrix for extant seed plants, combine this with a selected three-gene data set, and analyze with our tools for large data set analysis.
4. Systematics and taxonomy of Quercus (the oaks). and Fagaceae (the oak family).I am the de facto expert on North American Quercus (I authored the treatment of Fagaceae including the beeches and chestnuts) and Quercus for Flora North America, which includes the U.S. and Canada).I am currently in the final phases of completing the treatment of Quercus for Flora Mesoamericana, covering the region From Chiapas, Mexico to Panama (I am recognizing 40 species in this area).Recently, a species of oak from Guerrero, Mexico was named in my honor (Quercus nixoniana Valencia & ) although it is likely that I will eventually place it in synonymy under another species name.I also have a fungus named after me by a former student.
5. Systematics of Platanaceae (the Sycamores).The sycamores are an extremely important genus that was poorly understood prior to publication of a comprehensive monograph of the North American species (Nixon & Poole, 2003).I began work (not part of my Ph.D. thesis) on this group while a graduate student in Texas, based on my extensive fieldwork in Mexico, and was able to complete this work fairly recently (Nixon & Poole, 2003). In the context of this study we named two new and undescribed species from Mexico.I consider this to be one of my most important taxonomic contributions, given the previous confusion and misunderstanding surrounding how many species of Platanus are present in Mexico, inadequate description of their morphology, and the importance of this group in the early fossil history of flowering plants.
6. Theory and methods of interactive diagnosis (simply put, interactive keying software and algorithms).I have developed (as yet unpublished) algorithms for expediting specimen identification through interactive keys that allow selection of most efficient paths based on the level of expertise and background knowledge of the user (keys can be weighted for either experts or students, for example). An example of such a key can be seen at my website www.plantsystematics.org.
7. Software development for distributed parallel phylogenetic analysis.In addition to developing software for PCs that aids in phylogenetic analysis (Winclada), and being a coauthor on the program TNT (Goloboff, Farris and Nixon, 2001) I am currently developing web-based interfaces that allow distributed analyses using multiple software packages across heterogeneous arrays of computers (Linux, Windows and MacIntosh) based on a simple Perl client/Adobe server/MySQL server model.An important aspect of this system is a relational database model for storage of analytical methods along with databased results, which alleviates many of the issues in managing large numbers of analyses across a single or many permutations of a data set. This system also allows users to contribute computation resources based on the usage of their machine, capabilities (e.g., speed and ram) and what software they have licensed to perform multiple, simultaneous analyses.The system has been implemented as part of my teaching plan, and I am currently also training two graduate students in programming database and web interfaces as part of development of the package.
8. Bioinformatics: Database structure and implementation for biodiversity and classification data.Ancillary to development of a large database and web delivery system for phylogenetic, taxonomic, and biodiversity data (see http://www.plantsystematics.org), I have developed database structures for storage of complex tree-like relationships inherent in taxonomic names.This system provides easy ways to contrast different classification systems and retrieve name usage across different classifications.Currently, the system is focused on delivery of images, interactive maps, and interactive keys for all families and genera of angiosperms, gymnosperms, and ferns (i.e., the vascular plants). I will continue to expand this project to include more types of data.
