444 Copyright © 1998, Elsevier Science Ltd. All rights reserved. 0169-5347/98/$19.00 PII: S0169-5347(98)01462-1 TREE vol. 13, no. 11 November 1998 F ungi have provided some of the best model systems in genetics 1 , including the first eukaryotic genome to be completely sequenced 2 . Al- though the genetic capabilities of fungi are well known from lab- oratory experiments, actual pat- terns of genetic transmission in natural populations have been less well understood. In this re- view, we consider recent ap- proaches to two key questions that cannot be answered by direct observation. What are the fun- damental units of fungal popu- lations, and how important are asexual and sexual reproduction in determining their genetic struc- ture? As fungal population gen- etics has been amply reviewed 3–6 , we focus on recent studies identi- fying variation in DNA, especially nucleotide sequence data and genealogical analysis. Fungi do not fit the classical models Most fungi do not fit the classical models of population genetics. They have novel methods of colonization, disper- sal, perennation and their generations overlap. Sexual reproduction can be irregular or absent altogether in some recently derived lineages 7 . Also, most fungi grow inde- terminately within opaque substrates making direct meas- ure of fungal biomass or census of individuals difficult or impossible. Under these conditions, there is no single definition of the fungal individual that serves all purposes. For example, the basidiomycete Armillaria gallica, can legitimately be considered to be one of the largest 8 or one of the smallest organisms. This depends on whether the unit counted is the entire genet or the ramet, which may consist of as little as a single totipotent cell. In fungi, even the fundamental distinction between growth and repro- duction is not always clear. For example, several discrete mushrooms can be produced by one continuous mycelium that is hidden from view and is difficult, if not impossible, to trace physically within its substrate. In this case, which is the appropriate unit to count, the mushroom or the mycelium? Like fungal growth, methods of fungal reproduction are extremely varied. Fungi produce a multitude of different kinds of propagule associated with meiotic and mitotic nuclear division 9 (Fig. 1). After their production, spores and other propagules can lie dormant in their substrates for long periods, and then germinate and grow rapidly under favorable conditions. This means that not all fungi in a locality or a habitat are active players at all times. Dormant fun- gal propagules, like seedbanks, ef- fectively shelter genotypes from selection for various periods. Be- cause of these complexities, fun- gal populations are sometimes difficult to define. By necessity, fungal populations are usually de- fined as a group of conspecific individuals in the same locality at the same time. Genetic exchange is not limited to the sexual cycle In addition to their varied methods of growth and reproduc- tion, fungi have novel mechanisms of genetic exchange and recombi- nation, which are not accommo- dated by existing population gen- etic theory – genetic exchange and recombination are not necess- arily limited to the sexual cycle. Unlike plants, animals and bac- teria, the vegetative cells of many fungi fuse readily with one another – hyphal fusion is a constitutive feature of nearly all ascomycetes and basidiomycetes (32 267 and 13 857 species are recognized in each group, respec- tively 10 ). Vegetative hyphal fusion can be followed by a process of genetic exchange known as parasexuality 1 . In parasexuality, genotypically different nuclei coexist within the same hyphal compartments in a heterokaryon formed after hyphal fusion. Nuclei of different genotypes occasion- ally fuse with one another, and new genotypes are then produced by two independent processes: mitotic crossing over and haploidization with segregation of whole chromo- somes. Both of these processes of recombination occur much less frequently in parasexual crosses than in sexual ones with meiosis. Although parasexuality has been well documented in laboratory experiments, its occurrence in natural popu- lations has not been proven conclusively. This would be difficult for two reasons. First, parasexuality is limited by the widespread occurrence of the somatic incompatibil- ity (or vegetative incompatibility) response. This response is a self–nonself recognition system 11,12 that usually in- volves cell death and therefore restricts heterokaryosis among nuclei carrying different alleles at the loci determin- ing the specificity of somatic recognition. Second, para- sexuality has genetic consequences that are very similar to those of sexual recombination; both result in the reshuf- fling of entire chromosomes and in the recombination of alleles on the same chromosome through crossing over. However, parasexuality could be a potent evolutionary REVIEWS Genotyping, gene genealogies and genomics bring fungal population genetics above ground James B. Anderson and Linda M. Kohn As ubiquitous decomposers, symbionts and parasites, fungi build populations not easily accommodated by population genetic theory. Identifying and delineating individuals and populations is often difficult, and recombination can occur in complex and variable ways. Genotyping and gene genealogies provide the framework for identifying and delineating individuals and for detecting recombination in natural populations. Expanding genomic databases now make fungi ideal subjects for tracking mutation and expression in genes of adaptive importance in experimental populations. James Anderson and Linda Kohn are at the Dept of Botany, Erindale College, University of Toronto, Mississauga, Ontario, Canada L5L 1C6 (janderso@credit.erin.utoronto.ca).