Phytophthora ramorum
(Sudden Oak Death (SOD))

Phytophthora ramorum is considered an invasive species due to its ability to spread, persist, and reproduce in new environments. Its rapid life-cycle, propensity to reproduce asexually and splash dispersal via windblown rain, plus its ability to survive through harsh climatic conditions, are elements favouring this species’ potential invasiveness. Spread potential in forests has been elucidated by several studies in California and Oregon employing population genetics approaches. Results have consistently shown that scale of spread of naturalized endemic pathogen populations in natural ecosystems is limited to a few hundred metres and, occasionally, during extremely wet years, spread may reach a few (3-5) km (Mascheretti et al., 2008; Mascheretti et al., 2009; Eyre et al., 2013). Spread events at scales larger than those reported above appear to be associated either with the movement of infected plant parts, normally from large wild infestations, or with the introduction of infected plants, normally from infested ornamental nursery plant stock (Croucher et al., 2013). Spread scales from the hundreds of metres to the few kilometres apply to pathways that involve only foliar hosts, in particular California bay laurels and tanoaks, and are clearly positively correlated with rainfall (Eyre et al., 2013). However, spread from foliar hosts such as California bay laurels, tanoaks and ornamental rhododendrons to stem hosts such as oaks and tanoaks occur at the much lower scale of 10 to 20 metres and are strongly associated with the occurrence of episodic and above average rainy years (Cobb et al., 2012; Garbelotto et al., 2017). Given the limited spatial scale of dispersal of P. ramorum, its spread is strongly driven by structure and composition of individual forest stands and is projected to increase as the density of infectious foliar hosts increases (Cobb et al., 2010; Meentemeyer et al., 2015). Monocultures of Japanese larch in the UK, stands with high proportion of tanoaks in Oregon and California, and oak woodlands with an abundance of California bay laurels have all been the hardest hit systems. Presence of contiguous forests (Condeso and Meentemeyer, 2007), genetics of host populations (Dodd et al., 2005; Hayden et al., 2011), microclimate (Anacker et al., 2008; DiLeo et al., 2014) and climate (Meentemeyer et al., 2004; Venette and Cohen, 2006; Ireland et al., 2013; Meentemeyer et al., 2015) are all know to drive the spread in ecosystems invaded by P. ramorum. In spite of the theoretical tolerance of the pathogen to both high and low temperatures, models validated by extensive field sampling in regions infested by NA1 populations indicate high maximum temperatures strongly limit the spread of the pathogen (Meentemeyer et al., 2015) and may even cause significant reversion from positive to negative infection status in foliar hosts (Lione et al., 2017). High temperatures have also been shown to make water populations of the pathogen not viable (Eyre et al., 2015). In the lab, exposure of Petri dishes to 55°C for 1 hour, to 45°C for 4 hours or to 40°C for 24 hours has blocked pathogen growth (Swain et al., 2006), but survival of the pathogen has been reported up to 1 week at 55°C for infected California bay laurel leaves (Harnik et al., 2004), and in other trials the pathogen has been shown to survive in pant tissue between 40°C  (2 days) and -20°C (4 days) (Tooley et al., 2008). Finally, the pathogen’s broad host range on popular, nursery grown, ornamental plants, and the non-lethal, nondescript nature of the disease on most of the foliar hosts allows for long-term dispersal.

Disease epidemiology in nursery settings has not been extensively studied, however it has been reported that disease cycles are strongly influenced by seasonality and year-to-year climatic variations (Tjosvold et al., 2008). Propagules have been shown to survive in potting mix for over a year but infested mix only caused infection if leaves were in contact with it (Tjosvold et al., 2009), this is in contrast with shorter survival in natural ecosystems (Fichtner et al., 2007; Eyre et al., 2013). Infection does not seem to be strongly associated with irrigation water, but sprinkler irrigation did favour disease development (Tjosvold et al., 2008). Serrano et al. (2020) have provided a first in depth analysis of an outbreak in a mock nursery setting and have identified leaf to leaf, followed by leaf to soil as the two main pathogen dispersal mechanisms in the absence of flooding. In the same study, dispersal from individual infected potted rhododendrons was always less than 2 metres in a single growing season. Anecdotal evidence suggests outbreaks in nursery settings may be associated with flooding of growing beds (see Serrano et al., 2020). Thus, both in natural ecosystems and in nurseries, foliar infections appear to be driving disease outbreaks and the microevolution (i.e. generation of novel genotypes) of the pathogen (Eyre et al., 2013; Serrano et al., 2020). Soil and water are infested but may often be dead ends, epidemiologically speaking, although water populations have been shown to foster microevolution of the pathogen which could lead to the emergence of adaptive traits (Eyre et al., 2015).

Genetic analysis shows P. ramorum infection initially reported only in Europe and parts of North America is also present in northern Vietnam, at the border with China (Jung et al., 2020); three clonal lineages were originally identified: EU1 (Europe and North America, nurseries and forests), NA1 (North America, nurseries and forests) and NA2 (North American nurseries) (Ivors et al., 2006). Information to date indicates that divergence of these lineages occurred ages ago, P. ramorum originated from isolated populations and has migrated at least four times to North America and Europe (Grünwald et al., 2011). In 2012, Van Poucke et al. (2012) identified a fourth lineage, named EU2, in the UK, and additional lineages may be present in Vietnam (Jung et al., 2020). Whole genome resequencing has identified genomic differences among the lineages (Dale et al., 2019), but lineages also differ phenotypically from one another both in vitro and in planta, although estimates of their relative invasiveness may often be biased by the choice of hosts tested (Elliott et al., 2011; Eyre et al., 2014; Franceschini et al., 2014; O’Hanlon et al., 2017). One mating type only is found in each of the four widely studied lineages: EU1 and EU2 are A1 and NA1 and NA2 are A2 (Grunwald et al., 2019). Lineages with different mating types are co-mingled in some North American nurseries (Grünwald et al., 2012) and in a few forest stands of southwestern Oregon (Grünwald et al., 2016) raising the issue of potential sexual reproduction between lineages carrying different mating types. In vitro tests have indicated mating is possible (Boutet et al., 2010), however ‘hybrid’ progeny appears to be unstable possibly due to aberrant aneuploid genome size (Vercauteren et al., 2011). In spite of the clonal nature of populations within lineages, significant intralineage phenotypic variation has been observed; such variation is mostly due to the presence of non-wild types (nwt) that are often less virulent and are caused by genomic alterations, including locus copy number variation and chromosomal rearrangements (Kasuga et al., 2012; Kasuga et al., 2016). The emergence of nwt is strongly associated with host species (stem hosts) or habitat (plants grown in nursery settings) (Elliott et al., 2018) and may be a trait identifying pathogen populations from ‘dead end’ hosts.

In Europe, P. ramorum has been reported in commercial nurseries in over 20 countries, and is under regulatory control (Sansford et al., 2009). Outside of nurseries, in the UK infection appeared to be at first limited to ornamental plants in gardens and scattered woodland trees and shrubs until 2009 (Grünwald et al., 2012), when extensive mortality and infection erupted in the UK, on Japanese larch (Larix kaempferi) in timber plantations (Brasier and Webber, 2010). Limited outbreaks have been reported in France (Schenck et al., 2018) and the Netherlands (De Gruyter et al., 2006).

P. ramorum was first identified in 2001 in the USA in the California coastal forests around Marin County (Kliejunas, 2010; COMTF, 2017b; COMTF, 2017c). When P. ramorum was subsequently found infecting woody ornamentals in a California nursery surrounded by P. ramorum-infected Umbellularia californica trees, federal regulations were imposed on the US nursery industry to prevent the spread of the pathogen. In 2004, two large wholesale nurseries in California and Oregon mistakenly sent P. ramorum-infested host plant material to numerous states across the USA. As a result, multiple surveys conducted in all affected states in 2004 detected P. ramorum in nurseries in 41 states (R Bulluck, National Science Director, USDA-APHIS-PPQ, personal correspondence, 2018). These detections were considered episodic and the infested plants were destroyed. P. ramorum has been detected on infected plants outside of nursery or managed landscape perimeters in only two states, California and Oregon. P. ramorum remains under official control in the USA (NAPPO, 2016) and surveys continue on an annual basis in forest habitats, regulated nurseries and nearby areas (USDA-APHIS, 2016).

In California and Oregon, P. ramorum has killed over 50 million trees in coastal regions (Meentemeyer et al., 2011; Frankel and Palmieri, 2014; Cunniffe et al., 2016, Suddenakdeath.org). Tanoak (Notholithocarpus densiflorus) is the most susceptible species in California and Oregon (Frankel and Palmieri, 2014; Swiecki and Bernhardt, 2017). Sudden oak death was first reported in California on the coast near San Francisco (Rizzo et al., 2002b) and it spread along the coast into northern counties. In California, P. ramorum is quarantined in 15 counties, Humboldt, Mendocino, Napa, Lake, Sonoma, Solano, Contra Costa, Alameda, Marin, San Mateo, Santa Clara, Santa Cruz, San Francisco, Monterey and Trinity (USDA-APHIS-PPQ, 2016). The infestation is patchy and limited to areas within 80 km of the Pacific Ocean (Frankel and Palmieri, 2014).

In 2001, sudden oak disease was discovered in Oregon, in the south-western part of the state, Curry County (Goheen et al., 2002). Mitigation efforts in Curry Co. were implemented soon after detection, continued for many years and then discontinued in 2015 when it was apparent that although measures slowed down the spread of disease, they did not eliminate it (Frankel and Palmieri, 2014; Goheen et al., 2017). After treatments were used in the cooperative programme, the pathogen could no longer be detected in 39% of the treated forest plots (Goheen et al., 2013). P. ramorum is quarantined in a forested area of Curry Co., Oregon (NAPPO, 2016).

In Washington State, P. ramorum has been found infecting individual plants along streams passing through plant nurseries and significant outbreaks on planted ornamentals have been identified in a preserve (Strenge et al., 2017).

P. ramorum is a quarantined pest in California and southern Oregon, USA (other than nurseries, the rest of Oregon is not quarantined due to wildland eradication and containment efforts). All host plants are regulated in nurseries shipping out of state in Washington, Oregon and California.

The USA, the EU, Canada, New Zealand, Australia, the Czech Republic, Mexico, Taiwan, South Korea and other countries have identified P. ramorum as a quarantine pest (Frankel, 2008). The European Union Pest Risk Analysis for P. ramorum (Sansford et al., 2009) lists 68 countries that mention P. ramorum in their regulations (Kliejunas, 2010). The European Plant Protection Organization (EPPO) includes P. ramorum in the A2 list of quarantined organisms (www.eppo.int/ACTIVITIES/plant_quarantine/A2_list).

Natural habitats that have become endemic for the pathogen include a few types of coastal California forests. The first is the mixed evergreen forest characterized by coast live oak (Quercus agrifolia), California bay laurel (Umbellularia californica) and Pacific madrone (Arbutus menziesii). The second is the tanoak-redwood forest, characterized by redwood (Sequoia sempervirens) dominance, with a significant tanoak (Notholithocarpus densiflorus), California bay laurel and Douglas fir (Pseudotsuga menziesii) component (Rizzo et al., 2002a; Rizzo et al., 2002b; Garbelotto et al., 2003). In 2009, the pathogen was found infecting tanoak in a bishop pine (Pinus muricata) forest along the Mendocino Coast (Frankel and Hansen, 2011). In Oregon impacted forests are dominated by tanoak and Douglas fir with red alder (Alnus rubra) and Oregon myrtlewood (California bay laurel) (Umbellularia californica) (Hansen et al., 2008).

In Europe, and particulary in the UK, the pathogen is present in woodlands characterized by the presence of hosts such as native beech (Fagus sylvatica), Quercus cerris and Q. falcata, horse chestnut (Aesculus hippocastanum) and naturalized Rhododendron ponticum, all hosts on which infection of the stem and branches can be observed. In these environments, it may also infect the leaves of holm oak (Quercus ilex), sweet chestnuts (Castanea sativa) and ash (Fraxinus excelsior) (Slawson et al., 2006).

It is also established and an aggressive lethal pathogen of Japanese larch in plantations of the UK (Webber et al., 2010), Ireland (McCracken, 2013) and northern France (Schenck et al., 2018).

In northern Vietnam, P. ramorum has been found in subalpine and montane Rhododendron scrub and forests, and in montane broadleaved forests.

Quercus rubra, Q. palustris, Pittosporum undulatum and many other species are regarded as potential hosts: for these species, inoculation experiments have been completed, confirming susceptibility, but no natural infection has been recorded to date (2003). A database of species tested for susceptibility is available at the Risk Analysis for Phytophthora ramorum website (http://rapra.csl.gov.uk/). More information on host range is given in the following references: Werres et al. (2001); Davidson et al. (2002a); Hansen and Sutton (2002); Linderman et al. (2002); Maloney et al. (2002); Parke et al. (2002); Rizzo et al. (2002a); Rizzo et al. (2002b); Tooley and Englander (2002); Garbelotto et al. (2003); Hüberli et al. (2003) and  Kliejunas (2010). A host list is maintained by the USDA Animal and Plant Health Inspection Service (http://www.aphis.usda.gov/plant_health/plant_pest_info/pram/downloads/pdf_files/usdaprlist.pdf. To date (2012) there are over 120 species listed. The California Oak Mortality Task Force (www.suddenoakdeath.org) also maintains a host list with photos of symptoms. In 2017 the abundant rainfall associated with el nino, combined with the extensive distribution of the disease in the San Francisco Bay Area, resulted in the infection of eight Arctostaphylos species commonly known as Manzanita spp., of which at least six are considered rare or endangered (Rooney-Latham et al., 2017; Garbelotto et al., 2020).

In 2009, P. ramorum was confirmed as the cause of extensive dieback and mortality in mature and juvenile Japanese larch (Larix kaempferi) at a number of sites in south-west England. In 2010, P. ramorum was isolated from larch plantations displaying similar symptoms in south Wales. Overall, 2400 ha or ca. 0.6 million mature larch were affected (Webber et al., 2010) in addition to a large area of juvenile larch. This is the first widespread and lethal damage caused by P. ramorum to a commercially important conifer species anywhere in the world. In 2013, it was reported in Japanese larch plantations of Ireland (McCracken, 2013) and in 2018 in France (Schenck et al., 2018). Secondary infection of Fagus sylvatica, Nothofagus obliqua, Castanea sativa, Betula pendula, Rhododendron ponticum, Tsuga heterophylla and Pseudotsuga menziesii was found adjacent to some affected larch sites in south-west England.

Yaupon (Ilex vomitoria), sweetbay magnolia (Magnolia virginiana) and baldcypress (Taxodium distichum) were reported as hosts of P. ramorum when several plant species native to the Gulf Coast and south-eastern US forests were tested for reaction to P. ramorum (Preuett et al., 2013).

Our understanding of P. ramorum is still limited and subject to be modified in the future. For reviews on current knowledge, refer to Rizzo et al. (2005), Sansford et al. (2009), Kliejunas (2010), Grünwald et al. (2011); Garbelotto and Hayden (2012); Grünwald et al. (2019).

The coastal distribution of the disease caused by P. ramorum (Rizzo et al., 2005) in California, USA, suggests that the pathogen is favoured by moist and moderate climates. Moisture in the infested region is provided both by precipitation and fog, and temperature fluctuations are relatively small when compared with the interior of California. Within the coastal region, eastern and southern slopes are significantly drier. A comparison of inoculum viability between these drier areas and the more mesic areas within the same region has indicated a much more pronounced seasonal pattern in drier climates (Davidson et al., 2002c). These observations, combined, indicate that maximum disease progression is to be expected in mesic areas with moderate climate. Meentemeyer et al. (2015) and Lione et al. (2017) have both identified high temperature in spring-summer-autumn as a limiting factor in the distribution and/or persistence of the pathogen in the interior of California. In the latter study, a yearly mean temperature of 20-25°C and a maximum temperature of 35-30°C were extremely limiting for P. ramorum.

Experimental evidence has indicated that infection of bay (Umbellularia californica) leaves is optimal when a film of free water remains on the leaf surface for at least 9-12 hours and temperatures are approximately 18°C (Garbelotto et al., 2003). These conditions are frequently met in the fog-drenched coastal region of California, where over 200 days of fog per year have been recorded. However, Garbelotto et al. (2017) have shown in a 9-year-long study that sporangia production is exclusively limited to the rainy season and is basically nil in other periods, including those when the marine layer brings in significant moisture to coastal forest. In the same study, Garbelotto et al. (2017) were also able to determine that oak infection is extremely episodic and as of 2020 may only have occurred a few times since the arrival or the pathogen in California forests (2000-2001, 2006-2007, 2010-2011, 2016-2017). The reason for the episodic nature of oak infection relies in the requirements necessary for infection namely: inoculum density needs to reach 5 x 10^4 zoospores per mL, temperature needs to be around 20°C multiple hours per day for a week before infection starts to be successful, and rainfall must have surpassed the 250 ml threshold in the 3-6 weeks prior to infection. Interestingly, the overwhelming majority of oak infections were also modelled to occur when oaks are within 5 m from more than one bay laurel (Garbelotto et al., 2017).

In the absence of free water, plant infection is significantly reduced. It has been presumed that zoospores are the main source of infection, and micrographs have shown zoospores producing infection pegs that enter the host through bark lenticels or leaf stomata. The most important feature of this forest Phytophthora is that all plant parts affected by the disease are aerial. Is this an aerial Phytophthora? There is no doubt that an aerial phase has to be invoked to explain the epidemiology of Sudden Oak Death (SOD), but current knowledge on dispersal, that is when sporangia are produced and dispersal scale strongly indicate P. ramorum to be an aerial splash dispersed species, such as P. capsici (Ristaino et al., 2000). Sporangia are extremely deciduous and are produced in abundance on leaves and sometimes twigs of hosts such as bay laurel (U. californica) and tanoak (Notholithocarpus densiflorus). Infections can vary greatly from extremely abundant on almost every leaf in the lower and mid crown of bay trees in infested areas to just a few leaves, and sporangia can be recovered from rainwater. The extent of the aerial spread of P. ramorum is becoming clearer as extensive population genetics studies in colonized forests have shown that leaf pathogen populations rather than soil, water or even populations from stem hosts are epidemiologically relevant and drive the genotypic composition and the microevolution of this pathogen (Hüberli and Garbelotto, 2012; Eyre et al., 2013; Eyre et al., 2015). It is unclear whether in nature, soil and water populations may be epidemiologically significant, possibly because selection pressure leading to adaptation to live in soil and water may not be beneficial to the aerial lifestyle of the pathogen.

In California the disease is almost always correlated to the presence of two species, bay laurel and tanoaks (although a few exceptions are known). In areas where tanoaks and California bay laurel coexist, infection of the latter tree species precedes and overwhelmingly surpasses infection of tanoaks, suggesting a key role is being played by this species in the epidemiology of the disease. In a study where position of tanoaks and bay laurels was studied at the landscape level, it was shown that tanoaks and bays are ecologically overlapping in terms of physiological requirements, but are not epidemiologically identical when studying SOD, as only bay laurels are strongly associated with oak infection (Garbelotto et al., 2017).

After the first few rains, P. ramorum can be detected in the environment (soil, rain traps) (Davidson et al., 2002c). Leaf infection will occur in a few hours, and infected leaves may persist on the branch for more than a year, providing a persistent source of inoculum on some trees. When relative humidity is high, and temperatures are warm, sporangia and chlamydospores will be produced on the infected leaves and will be splashed by rain onto other leaves and into the soil, and may become airborne in rain driven winds. In hosts like tanoaks, foliar infections are often associated with twig infection: the outcome is twig and branch dieback. Oaks and the main stems of tanoaks are probably infected in a final stage of the disease. Sporangia are the most effective transmission propagule, but the motile zoospores that they contain are the most effective infection propagules. Girdling cankers will proceed at varying rates, probably mostly affected by the genetic makeup of the individual infected tree (Dodd et al., 2005). Variations in susceptibility to the pathogen have been noticed in both bay laurel (Hüberli et al., 2012) and oaks (Dodd et al., 2005). Susceptible individuals will potentially be girdled in just a few weeks. Girdled trees will look apparently healthy for several more months, until all resources have been depleted and the trees undergo a rapid decline. Disease tolerance has been identified in tanoaks (Hayden et al., 2011; Hayden et al., 2013) and oaks (Conrad et al., 2017).

Soil, water and leaves may be infected/infested and infectious but only infected plants have been shown to be highly contagious (Serrano et al., 2020): it is possible to experimentally infect leaves by placing them over infested soil, and it is possible to infect wood by placing it under infected leaves. Conversely, it is extremely difficult to infect leaves by placing them near or on infected wood. These results indicate that leaves (including twigs) and not wood play a crucial role in the epidemiology of the disease. Because oak leaves are not generally infected by P. ramorum, oaks may not effectively spread the disease. In support of this hypothesis, surveys have indicated that oaks are much more likely to become infected by P. ramorum if they grow in proximity to bay laurel trees (Kelly and Meentemeyer, 2002; Swiecki and Bernhardt, 2002). While the epidemiological role of soil in infested areas is probably minor (Eyre et al., 2013), the epidemiological role of water remains an open question, and there is anecdotal evidence that occasionally floods may lead to infection. Serrano et al. (2020) and Tjosvold et al. (2008) have indicated that water used for irrigation is not a good infection pathway, however they have not tested flooding. On the other hand, Eyre et al. (2015) have shown that genotypic composition in water is different to that of leaf populations, so it remains to be seen whether genotypes that thrive in water may be able to start aerial outbreaks.

Sporangia can survive for several weeks even if dried, whereas the survival period for chlamydospores is still unknown. Chlamydospores are found in abundance in soil, streams and embedded in leaves. Chlamydospores embedded in bay leaves are quite resilient and will survive for a week with a constant temperature of 55°C (Harnik et al., 2004) and for multiple days at -20°C (Tooley et al., 2008). Both of these propagule types are potential means of long-range spread of the disease.

Both Phytophthora nemorosa and Phytophthora pseudosyringae have been found to be competitors of P. ramorum. Their interaction has been studied in the laboratory and in the field: in dry weather the competitors are more fit than P. ramorum, and vice versa in rainy years (Kozanitas et al., 2017).

P. ramorum will grow on most media used for other Phytophthora species such as cornmeal agar, PDA (not optimal), V8 and carrot agar (Werres et al., 2001). It grows relatively well between 15 and 22°C, but slows down significantly when temperatures go over 25°C. The most widely used selective medium to isolate P. ramorum from infected plants is corn meal based PARP (Rizzo et al., 2002a). While colonies will generally emerge from plated plant material within 6 days, at times up to 3 weeks has been necessary to obtain cultures. Cultures can be obtained from the following substrates: canker margins from oaks (Quercus spp.) and tanoaks (Nothorlithocarpus densiflorus), rhododendron leaves and stems, bay (Umbellularia californica). It is harder to isolate the pathogen from Rhamnus, Pseudotsuga, Arbutus or Heteromeles. From hosts such as Arctostaphylos, Acer, Aesculus and Lonicera, isolations can be extremely difficult. In California, a state characterized by an extremely marked Mediterranean climate, isolation success of P. ramorum follows a seasonal pattern, a trait not uncommon among Phytophthora spp. Isolation success tends to be best during late winter and spring, while it progressively declines as the dry California summer progresses. Almost every host has some specific requirements to maximise isolation success. The length of time that has passed between sample collection and processing has been shown to strongly affect the sensitivity of culture-based diagnostic methods (Vettraino et al., 2010).

For some foliar hosts, such as Umbellularia and Rhododendron, incubation in a moist chamber for 48 to 64 hours results in abundant sporulation. Sporangia and chlamydospores can then be analysed.

It is possible to bait P. ramorum from soil, streams, infected leaves, and to a limited extent from wood (Werres et al., 2001; Davidson et al., 2002c). The most common baits used for P. ramorum are pears, and 'Cunningham White' rhododendron leaves. Baiting is normally done for 5-7 days, and appears to be facilitated by temperatures between 12 and 15°C, at least for some substrates such as wood. Baits are collected at the end of the baiting period, left to dry for a couple of days, and then plated on selective PARP medium and incubated at 18°C. From mid-summer on, baiting from soil becomes unsuccessful. It is unclear whether this corresponds to a temporary dormancy period of the pathogen, or whether viability of the inoculum may be permanently lost. In infested areas it is possible to bait P. ramorum directly from the air by using healthy susceptible foliar hosts such as rhododendrons. Baiting from streams is strongly affected by air temperatures and season: in San Mateo County, Eyre et al. (2015) could not bait the pathogen out of almost 20 baiting stations starting in July and ending in November. Water baiting from plant production facilities may also be influenced by temperature, oxygen levels and by the presence of other microorganisms: Serrano et al. (2020) report that baiting from tanks of recycled irrigation water was only successful when baiting from the top layer of the water tanks and was never successful from the bottom layers. Success of baiting from top layers was also negatively affected by warmer temperatures and by the presence in the water of Phytophthora chlamydospora.

There is a wide selection of DNA PCR-based protocols to identify the pathogen both from cultures and directly from infected plant material (Garbelotto et al., 2003). The first government sanctioned protocol was that of Hayden et al. (2004). A nested PCR approach is needed to obtain a signal from most foliar hosts with the exception of bay and rhododendron leaves. Because of the difficulty in isolating the pathogen from most foliar hosts, it is recommended to back up diagnosis based on isolations, with PCR-based data. Vettraino et al. (2010) concluded that a combination of either culturing and molecular diagnosis, or of two molecular assays was the most successful approach to identifying P. ramorum. The best comparative analysis of various PCR methods by ring trials is that reported by Martin et al. (2009).

Tomlinson et al. (2005) developed a sensitive and specific single-round real-time PCR (TaqMan) assay using a portable real-time PCR platform. This has the advantage of being able to diagnose P. ramorum entirely in the field, independent of laboratory facilities.

Diagnostic methods are reviewed in Kliejunas (2010). For the USA, diagnostic and sampling protocols for official regulatory confirmations are posted at the USDA Animal and Plant Health Inspection Service (APHIS) P. ramorum website http://www.aphis.usda.gov/plant_health/plant_pest_info/pram/protocols.shtml#diag. A duplexPCR detection method, based on the internal transcribed spacer (ITS) regions of the ribosomal DNA, was developed by Schlenzig (2011) to enable sensitive detection of P. ramorum and to distinguish it from the similar P. kernoviae. Chimento et al. (2012) developed a real-time RT-PCR which is able to clearly distinguish between dead and viable P. ramorum pathogens.

P. ramorum is difficult to diagnose from dead trees, even if death has occurred recently. Looking around the dead tree for trees with obvious symptoms (see Symptoms section) that are still green is recommended. Bleeding can be an easily visible symptom, but it can be due to a myriad of causes including mechanical wounding, cracking, insect attack and wood decay. Using a blade, cut the outer part of the bark and check for discoloration under it. A P. ramorum lesion may or may not have a clear zone line, but it will almost never have a regular shape. Borders are in general angular, and shape is often irregular. It is essential to uncover the border of the canker in order to diagnose a tree as infected by P. ramorum. Cultures should also come from the margin of the lesion (see Diagnosis).

Bleeding is not always associated with cankers, and is normally absent in small tanoak (Notholithocarpus densiflorus) branches. In these cases, cankers will be visible as slight depressions in the bark, often appearing slightly water-soaked. A dark-brown or black lesion, without zone lines, will be present under the bark.

For all other hosts, inspection for leaf and branch symptoms is the first step. In rhododendron, look for both leaf blight and branch cankers. The two are often connected, so it will be possible to follow a lesion from the leaf into the bearing stem, and vice-versa. Lesions on rhododendrons often have diffuse margins, or conversely, dark lines in a concentric pattern highlighting growth patterns of the pathogen. Visible branch dieback is normally seen as a second stage of infection.

In the case of Umbellularia foliage, there is a good correspondence between the side of the leaf carried downwards and location of the P. ramorum lesion. This is due to the fact that swimming zoospores (requiring free water) are the main infection propagule responsible for foliar infections. If leaves are carried sideways, lesions will develop on the lower blade; if they are carried with their tip downwards, lesions will develop on the leaf tips, and so forth.

In California, USA, the most valuable diagnostic approach is to focus not on one tree but on an entire area. The diagnosis is strengthened by identifying a few different hosts, each displaying their characteristic symptoms. The use of the database SODmap both on the web (www.sodmap.org) and through the Application SODmap Mobile for smartphones and tables (www.sodmapmobile.org) allows to determine whether the pathogen is present in any given locality (Garbelotto et al., 2014). A new database called Calinvasives and part of the Calflora database will be accessible in 2020 to geolocate all proven infestations by SOD and other emergent pathogens in California (calflora.org).

Symptoms on plants are fairly generic and can be easily confused with those caused by other Phytophthora species, or with other fungal diseases. Culturing or molecular diagnostics need to be used to confirm pathogen presence.

P. ramorum cankers are in general above ground and do not develop in the roots: this feature can be helpful, but at times other Phytophthora species also cause aerial cankers. Lesions on Umbellularia resemble symptoms caused by bay anthracnose and by other Phytophthora species, in particular P. nemorosa and P. pseudosyringae, which are very common on the West Coast of the USA (Davidson et al., 2002b). The lack of fungal reproductive structures can assist in differentiating P. ramorum symptoms from other symptoms caused by fungi. Also, while fungi are facilitated by high relative humidity, infection by Phytophthora spp. requires a film of water, so correspondence between lesion location on a leaf and where water would accumulate can be a helpful diagnostic feature to differentiate fungal from Phytophthora infections. On Rhododendron spp., symptoms are identical to those caused by species such as Phytophthora cactorum, P. citricola, P. citrophthora and P. nicotianae for twig symptoms, and P. syringae for foliar symptoms (Werres et al., 2001).

On some hosts, foliar infection by P. ramorum leads to early leaf abscission. This is particularly noticeable in Camellia species, but other plant species, at times characterized by smaller leaves, and even California Bay laurels growing in warmers sites may develop the same symptoms. Some foliar hosts both in North America and in Europe are known to bear viable infections that are not visible as necrotic lesions, blotches or spots, and can only be detected by isolation, PCR or baiting (Denman et al., 2009).

Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.

Best management practices for P. ramorum are available at: http://www.suddenoakdeath.org/diagnosis-and-management/best-management-practices/ and at www.matteolab.org

-Removal of California bay laurels (Garbelotto et al., 2017), tanoaks (Cobb et al., 2017) and Japanese larch (O’Hanlon et al., 2018) have all been shown to be effective.

-Employment of disease tolerant plants (Hayden et al., 2013; Conrad and Bonello, 2016; Cobb et al., 2019) holds significant promise and some results have been produced on higher survival of tanoak recruitment using disease tolerance.

-Planting non-susceptible species and other strategies are being developed to prevent or manage P. ramorum in wildlands (Lee et al., 2010; Swain and Alexander, 2010; Cobb et al. 2019).

-It has been shown that fresh wounds will be optimal infection courts; pruning of large branches or of stems should occur in the autumn, months in California before the rain-driven spread of P. ramorum occurs, or 4 months before the infectious period in other parts of the world.

-In restoration projects, avoid bay laurel (Umbellularia californica) if possible, especially in areas where oaks may be growing. Eradication has been attempted in southern Oregon, USA, via the burn-and-slash technique (Goheen et al., 2002b).

- Kiln drying: 55°C for 30 minutes was insufficient to kill the pathogen. At least 1 hour is required but only for substrates without abundant chlamydospores such as wood. If substrate supports chlamydospores 2 weeks at 55°C were necessary to kill the pathogen. However, vacuum plus exposure at 55°C, together killed the pathogen after 22 hours (Harnik et al., 2004).

- Composting following EPA guidelines for California is completely successful in eliminating the pathogen (Swain et al., 2006), however mature compost can be re-infected by the pathogen if exposed to large amounts of inoculum (Swain and Garbelotto, 2015).

- P. ramorum is susceptible to label-dosages of copper sulfates and copper hydroxides (Garbelotto et al., 2009). In different formulations it is moderately susceptible to mancozeb. The pathogen is sensitive to phosphites or phosphonates. Phosphite injections are effective in oaks for 2 years (Garbelotto and Schmidt, 2009) and also temporarily in tanoaks (Notholithocarpus densiflorus). Phosphite foliar sprays are not effective on oaks and tanoaks. The pathogen is extremely sensitive to metalaxyl, but drenches and foliar sprays are ineffective in oaks (Garbelotto et al., 2007; Garbelotto et al., 2009).

- Water and moisture management are extremely important, especially when temperatures are between 15 and 20°C. Infection on bay (Umbellularia californica) leaves requires 9-12 hours of leaf wetness. Oak infection requires 250 mm of rain in for 3-6 weeks and at least 1 week when max temperature daily reaches 20°C (Garbelotto et al., 2017).

- Natural contagion from oaks is estimated to be low: susceptible oaks should not be planted near foliar hosts like bays and rhododendrons.

- Early infection can be detected on foliar hosts: new infection on bay leaves are good signs of inoculum level.

- Sites, soil and streams can be monitored by baiting with rhododendron leaves (Davidson et al., 2002c).

- Removal of California bays 10 and 20 m around oaks has been proven highly effective in reducing possible infection events and remains the main prescription to preventatively protect oaks from becoming infected (Garbelotto et al., 2017).

- Sanitation of small tools such as axes, hand saws or chain saws is best done by removing all organic debris, followed, if desired,by treatment with lysol, 70% ethanol, or a 10% dilution of commercial bleach.

11/05/2020 Updated by:

Matteo Garbelotto, Department of Environmental Science, Policy & Management, University of California, Berkeley, California, USA.

09/03/12 Updated by:

Susan J. Frankel, Sudden Oak Death Research, USDA Forest Service, Pacific Southwest Research Station, Berkeley, California, USA