Sudden Oak Death (SOD), caused by the oomycete pathogen, Phytophthora ramorum, is an introduced forest disease causing large-scale tree mortality in the oak woodlands of California. Pathogen dynamics are bound to be affected by site ecology, forest structure and composition and by interactions among sympatric pathogens. In the four chapters included in this dissertation I address various aspects of the both ecology and epidemiology of the Sudden Oak Death disease, including the competition between three sympatric species of Phytophthora, the population structure of the P.ramorum on its three main hosts, California bay laurel (Umbellularia californica), tanoak (Notholithocarpus densiflorus) and coast live oak (Quercus agrifolia), and at sites with different levels of disease incidence and the population structure of isolates found in two different substrates, soil and bay laurel leaves, in various weather conditions.

Using molecular tools coupled with an intensive multi-year field sampling approach, all within a coastal California watershed with a relatively long and uniform history of disease presence, I determined the following:

When in competition with other Phytophthora species, specifically P. nemorosa, P. ramorum prevalence increases to levels higher than those of the competing species when abundant rainfall triggers its sporulation. Despite P. nemorsa having a broader geographic range, it exhibits a narrower ecological amplitude and, in any given region, occupies fewer sites than P. ramorum. Results additionally suggest that, perhaps due to priority effects, P. nemorosa can persist at levels comparable to those of P. ramorum in ecologically suitable plots when climate favors P. ramorum dormancy.

Shifts in weather, primarily levels of rainfall are accurate predictors in the likelihood of both bay laurel and oak infection as well as transitions between dormant and active infection of the pathogen itself. Additionally site specific factors such as aspect, bay laurel density and bay laurel basal area drive differences in levels of both disease incidence and prevalence on both bay laurel and oak hosts. For oaks specifically, infection rates are greater among larger trees, yet levels of mortality are greater among smaller trees. Neighborhood effects such as proximity to infected bay laurel foliage, and surrounding bay laurel density and basal area are important factors in predicting oak infection.

Ten microsatellites were used for genetic analyses on cultures from successful isolations to determine differences in population structure between different substrates, namely populations isolated from soil versus aerial populations isolated from bay laurel leaves in various weather conditions. Migration of genotypes among sites was low and spatially limited during dry periods, but intensity and range of migration of genotypes significantly increased for leaf populations during wet periods. Only leaf genotypes persisted significantly between years, and genotypes present in different substrates were distributed differently in soil and leaves. It was concluded that epidemics start rapidly at the onset of favorable climatic conditions through highly transmissible leaf genotypes, and that soil populations are transient and may be less epidemiologically relevant than previously thought.

Population structure of the pathogen in each of the three main hosts was examined to provide evidence of contagion pathways among hosts, as well as differences in population structure in wet vs. dry years and at sites with various levels of disease incidence and prevalence. The relationships among P. ramorum populations in bay laurels, oaks and tanoaks were analyzed and structure was found to exist among hosts. It was determined that bay laurel is the source population for both tanoak and oak infection, and that tanoak contributes minimally to oak infection. In spite of their common source of inoculum, oaks and tanoaks were found to be sinks that select for different pathogen genotypes, due to the variance in selection pressure in each host type. Additionally, different sites supported a dominance of different genotypes, more genotypes overall and more persistent genotypes, when compared to other sites, and these ‘hotspots’ are likely to play a more significant epidemiological and evolutionary role for the pathogen.

Together these results help to advance the state of knowledge surrounding the ecology, epidemiology and population genetics of the Sudden Oak Death pathogen in California, add to the growing body of research on invasive plant pathogens, support theories of invasion biology and most importantly can be applied to regulatory and land management practices in attempt to mitigate the spread of this disease.

Chair Date

Since its emergence in the late 1990s, sudden oak death has killed mature oak trees and tanoaks in 14 California counties. Treatment options are now available to safeguard these trees from infection by Phytophthora ramorum, the aggressive and exotic pathogen responsible for sudden oak death. We provide an update on current knowledge regarding this emergent disease in California, and present results from three controlled experiments of two chemical treatments to manage the disease in oaks and tanoaks. Phosphonate treatments, legally registered in California to control sudden oak death, were effective in slowing both infection and growth rates for at least 18 months. Conversely, an alternative method consisting of an azomite soil amendment and bark lime wash was always ineffective, and did not reduce either growth or infection rates.

I am curious if any researcher has linked the inadvertent introduction of Phytophthora to restoration areas to the current practice of using dead plant tissue (compost) as part of the growing media.

Composted municipal green waste is a potential vehicle for the transmission of Phytophtora ramorum, the pathogen responsible for the disease known as sudden oak death. To assess the survival rate of the pathogen in compost, we introduced zoospores — a type of infectious propagule — into six composts of varying provenance and maturity. The compost samples represented three production facilities, two production techniques (turned windrow and forced air static pile) and two levels of maturity (fresh, defined as aged for less than 1 week; and mature, aged for more than 4 weeks). Positive re-isolations — indicating survival of the pathogen — were obtained from all composts. The re-isolation rate from the compost from one of the three production facilities was greater than that obtained from an inert substrate (filter paper) inoculated with the pathogen (P < 0.01), while re-isolation rates from the other two sources were statistically indistinguishable from those obtained from the inert substrate (P < 0.01). There was no significant difference in re-isolation rate between composts produced by the turned windrow method and composts produced by the forced air static pile technique. Re-isolation rates were greater from mature composts than from fresh composts (P < 0.01). The results show that P. ramorum may be present and infectious if introduced into finished compost, and that variations in compost characteristics appear to influence survival rates.

This dissertation presents a body of work collected from the very beginning of the outbreak of the introduced forest pathogen Phytophthora ramorum. The first data were collected the same year that P. ramorum was identified as the causal pathogen of the disease in California. The research, aimed at gathering the scientific knowledge necessary for an understanding and informed response to a new problem, placed an emphasis on bridging evolutionary ecological theory and management practice.

First, my colleagues and I outline two quantitative polymerase chain reaction (qPCR) diagnostic assays for P. ramorum and their use as part of a widespread survey screening California plants for the pathogen. In Chapter 3, my colleagues and I report the first surveys of variation in susceptibility to P. ramorum on detached leaves of field-collected tanoak trees. We found low but non-zero heritability of susceptibility, and that the structure of phenotypic variation in this measure of susceptibility echoed that of neutral genetic markers. Finally, I describe the establishment of a greenhouse common garden population of tanoak seedlings concurrently with a field test garden in an infested natural area. Using these study populations, we calculated the heritability of susceptibility in detached leaves with much greater precision, and expanded the assay techniques and types of host resistance we measured. In so doing we discovered highly variable responses of seedling families to intact-seedling tip inoculations, the majority of which variation was due to genetic factors. With the qPCR detection techniques described in Chapter 2, we confirmed seedlings' infection within the first 7 months of the field test garden's establishment, with disease incidence rising over the course of the study. We found that both P. ramorum infection status and family origin have significant effects on tanoak survival in natural disease conditions.

Together with the laboratory assays, our findings show that tanoak does possess some forms of genetic resistance to P. ramorum. The gardens we established provide the infrastructure necessary to determine if this resistance affects survival under natural conditions, to further investigate its inheritance and action, and to test the effect of varying environment and host genotype on its expression.

“Sudden oak death” refers to a complex set of symptoms that has already culminated in the death of tens of thousands of California oak trees. Now confirmed in seven coastal counties, SOD attacks California tanoak, coast live oak and California black oak. Although several fungal species and the western oak bark beetle and ambrosia beetles have been associated with the syndrome, we now have solid evidence that a newly discovered Phytophthora species is the primary causal agent. This Phytophthora species was recently isolated in rhododendron as well; it may be the same species that was isolated, but not described, on rhododendrons in Germany in 1993. The discovery of SOD on rhododendron has serious implications. The disease may well be present at the ecosystem scale, and its appearance on an ornamental plant suggests the possibility of wider dissemination. A team of UC scientists has developed an integrated approach to managing this disease, including practices to enhance tree health. Early disease detection and targeted chemical treatment may also hold some promise for disease management. In addition, we have developed a molecular probe that will enable rapid identification of SOD from any infected part of the plant. Ultimately, the fate of the oak species affected by SOD will be determined by the levels of disease resistance present in natural populations of these trees.