The rust fungus, Puccinia psidii, is a devastating pathogen of introduced eucalypts (Eucalyptus spp.) in Brazil where it was first observed in 1912. This pathogen is hypothesized to be endemic to South and Central America and to have first infected eucalypts via a host jump from native guava (Psidium guajava). Ten microsatellite markers were used to genotype 148 P. psidii samples from eucalypts and guava plus five additional myrtaceous hosts across a wide geographic range of south-eastern Brazil and Uruguay. Principal coordinates analysis, a Bayesian clustering analysis and a minimum-spanning network revealed two major genetic clusters among the sampled isolates, one associated with guava and another associated with eucalypts and three additional hosts. Multilocus genotypes infecting guava differed by multiple mutational steps at eight loci compared with those infecting eucalypts. Approximate Bayesian computation revealed that evolutionary scenarios involving a coalescence event between guava- and eucalypt-associated pathogen populations within the past 1000 years are highly unlikely. None of the analyses supported the hypothesis that eucalypt-infecting P. psidii in Brazil originated via host jump from guava following the introduction of eucalypts to Brazil approximately 185 years ago. The existence of host-associated biotypes of P. psidii in Brazil indicates that this diversity must be considered when assessing the invasive threat posed by this pathogen to myrtaceous hosts worldwide.

Fusarium species can cause severe root disease and damping-off in conifer nurseries. Fusarium inoculum is commonly found in most container and bareroot nurseries on healthy and diseased seedlings, in nursery soils, and on conifer seeds. Isolates of Fusarium spp. can differ in virulence; however, virulence and colony morphology are not correlated. Forty-one isolates of Fusarium spp., morphologically indistinguishable from F. oxysporum, were collected from nursery samples (soils, healthy seedlings, and diseased seedlings). These isolates were characterized by amplified fragment length polymorphism (AFLP) and DNA sequencing of nuclear rDNA (internal transcribed spacer including 5.8S rDNA), mitochon-drial rDNA (small subunit [mtSSU]), and nuclear translation elongation factor 1-alpha. Each isolate had a unique AFLP phenotype. Out of 121 loci, 111 (92%) were polymorphic; 30 alleles were unique to only highly virulent isolates and 33 alleles were unique to only isolates nonpathogenic on conifers. Maximum parsimony and Bayesian analyses of DNA sequences from all three regions and the combined data set showed that all highly virulent isolates clearly separated into a common clade that contained F. commune, which was recently distinguished from its sister taxon, F. oxysporum. Interestingly, all but one of the nonpathogenic isolates grouped into a common clade and were genetically similar to F. oxysporum. The AFLP cladograms had similar topologies when compared with the DNA-based phylograms. Although all tested isolates were morphologically indistinguishable from F. oxysporum based on currently available monographs, some morphological traits can be plastic and unreliable for identification of Fusarium spp. We consider the highly virulent isolates to be F. commune based on strong genetic evidence. To our knowledge, this is the first reported evidence that shows F. commune is a cause of Fusarium disease (root rot and dampingoff) on Douglas-fir seedlings. Furthermore, several AFLP genetic markers and mtSSU sequences offer potential for development of molecular markers that could be used to detect and distinguish isolates of F. oxysporum nonpathogenic to conifers and highly virulent isolates of F. commune in forest nurseries.

9410 or call (800) 795-3272 (voice) or (202) 720-6382 (TDD). USDA is an equal opportunity provider and employer. AbstractKliejunas, John T.; Geils, Brian W.; Glaeser, Jessie Micales; Goheen, Ellen

White pine blister rust disease, caused by the introduced pathogen Cronartium ribicola, has severely disrupted five-needled pine ecosystems in North America. A 100-year effort to manage this disease was predicated in part on the premise that the pathogen utilizes only species of Ribes (Grossulariaceae) as alternate hosts on this continent. The current study presents the first conclusive demonstration that some species in the family Orobanchaceae (Pedicularis racemosa and Castilleja miniata) are functioning as alternate hosts in a natural ecosystem of North America. This finding has implications for improving our understanding of epidemiology, pathogen adaptation and host-pathogen interactions within white pine blister rust.

Armillaria possesses several intriguing characteristics that have inspired wide interest in understanding phylogenetic relationships within and among species of this genus. Nuclear ribosomal DNA sequence-based analyses of Armillaria provide only limited information for phylogenetic studies among widely divergent taxa. More recent studies have shown that translation elongation factor 1-α (tef1) sequences are highly informative for phylogenetic analysis of Armillaria species within diverse global regions. This study used Neighbor-net and coalescence-based Bayesian analyses to examine phylogenetic relationships of newly determined and existing tef1 sequences derived from diverse Armillaria species from across the Northern Hemisphere, with Southern Hemisphere Armillaria species included for reference. Based on the Bayesian analysis of tef1 sequences, Armillaria species from the Northern Hemisphere are generally contained within the following four superclades, which are named according to the specific epithet of the most frequently cited species within the superclade: (i) Socialis/Tabescens (exannulate) superclade including Eurasian A. ectypa, North American A. socialis (A. tabescens), and Eurasian A. socialis (A. tabescens) clades; (ii) Mellea superclade including undescribed annulate North American Armillaria sp. (Mexico) and four separate clades of A. mellea (Europe and Iran, eastern Asia, and two groups from North America); (iii) Gallica superclade including Armillaria Nag E (Japan), multiple clades of A. gallica (Asia and Europe), A. calvescens (eastern North America), A. cepistipes (North America), A. altimontana (western USA), A. nabsnona (North America and Japan), and at least two A. gallica clades (North America); and (iv) Solidipes/Ostoyae superclade including two A. solidipes/ostoyae clades (North America), A. gemina (eastern USA), A. solidipes/ostoyae (Eurasia), A. cepistipes (Europe and Japan), A. sinapina (North America and Japan), and A. borealis (Eurasia) clade 2. Of note is that A. borealis (Eurasia) clade 1 appears basal to the Solidipes/Ostoyae and Gallica superclades. The Neighbor-net analysis showed similar phylogenetic relationships. This study further demonstrates the utility of tef1 for global phylogenetic studies of Armillaria species and provides critical insights into multiple taxonomic issues that warrant further study.