The likelihood for evolution of resistance by the host depends strongly on the amount of genetic variation on which selection

Disease epidemics caused by exotic forest pathogens can have devastating effects on native ecosystems ( Burdon et al., 2006 ; Desprez-Loustau et al., 2007 ). The consequences of severe mortality may be far-reaching for ecosystem health if the host species is an important biological component of the habitat and if environmental hazards of fi re and erosion are signifi cant risks. This is the case of tanoak [ Lithocarpus densifl orus (Hook. & Arn.) Rehder, Fagaceae], recently attributed to a new genus [ Notholithocarpus densifl orus (Hook. & Arn.) Manos, Cannon & S. Oh] that occupies several forest types in coastal California, southern Oregon, and the Sierra Nevada, Klamath, and Cascade mountains ( Fig. 1 ). In interior northern California, it commonly occurs in shrub form [ Lithocarpus densifl orus subsp. echinoides (R. Br. Campst.) Abrams]. Tanoak is the most susceptible host to Phytophthora ramorum (Werres deCock and Man in ’ t Veld), an exotic pathogen that causes sudden oak death (SOD) ( Davidson et al., 2003 ; Rizzo and Garbelotto, 2003 ; Rizzo et al., 2005 ). Although tanoak has not coevolved with P . ramorum and is therefore unlikely to have specifi c resistance genes (Deprez-Loustau et al., 2007; Suarez and Tsutsui, 2008 ), some degree of resistance, or tolerance may be present as is the case for other exotic pathogen interactions, such as white pine blister rust ( Cronartium ribicola ) ( Kinloch et al., 1970 ) and jarrah dieback ( Phytophthora cinnamomi ) ( Stukely and Crane, 1994 ). Dodd et al. (2005 , 2008 ) have reported genetic-based variability in response of coast live oak ( Quercus agrifolia N é e) to inoculation with P. ramorum , and preliminary studies indicate withinand between-population variation in response to inoculation in tanoak (K. Hayden, A. Nettel, R. S. Dodd, and M. Garbelotto, University of California, Berkeley, unpublished data). The likelihood for evolution of resistance by the host depends strongly on the amount of genetic variation on which selection can act ( Crow, 2002 ), and the level of genetic structuring among host populations should play an important role in the evolutionary trajectory of the host response ( Parker and Gilbert, 2004 ). Therefore, an understanding of genetic structure of susceptible host species is crucial in implementing resistance screening. Where populations are at risk for extinction, as is the case for tanoak, discovery of divergent population lineages will also provide a rationale for conservation priorities. Recent advances in DNA analysis permit studies about past demographic changes in natural populations. Phylogeographic patterns observed in maternal lineages, such as chloroplast DNA in angiosperms, help to infer paleogeographic population processes such as prePleistocene vicariance and late Quaternary population fl uctuations caused by past climate change. Different Pleistocene refugial patterns have been detected, from mass extinctions of high latitude populations and restriction to southern refugia (Petit et al., 1993; Brunsfeld et al., 2001 ; Hewitt, 2004 ) to in situ refugia for alpine species ( Stehlik et al., 2002 ; Sch ö nswetter et al., 2005), Mediterranean pine ( Afzal-Rafi i and Dodd, 2007 ) and sky island taxa in the southwestern United States (DeChaine and Martin, 2005). Surprisingly little is known about plant restrictions during the Quaternary in the California Floristic Province or of subsequent effects of Holocene climate changes on plant distributions and fl uctuations in population size. Comparative analysis of animal and plant taxa suggests 1 Manuscript received 7 October 2008; revision accepted 14 July 2009. The authors thank W. Mayer for assistance in the laboratory work and anonymous reviewers for advice on the manuscript. This study was funded by the United States Department of Agriculture, Forest Service. 2 Author for correspondence (e-mail: dodd@nature.berkeley.edu)