1 Fungal Pathogens of Plants in the Homogocene 11 1.4 Inferring Native Ranges of Fungi from Pathogen Release It is tempting to think that fungi with restricted host ranges must be native where their hosts are native.Cmelli,genus of some species.is endemic Asia.with its cent outhem China (Taand Barthol w1984 cove native range in eastern Asia even though other parts of the world may have beer stepping stones.If species of Camellia had never been introduced outside eastern Asia,the inference of sympatry for its host-restricted fungi would be unequivocal Ornamental species of Deutzia provide such an example in that they are also endemic to Asia where seven taxa of rust fungi commonly infect them(Farr et al n.d.).Unlike C.camelliae,records of rust on Deutzia outside the native range absent even though D.scabr was intr oduced to the U.S.as early as 1822(Rehde 1940) Searches of the SMML Fungu s-Host Dis ution Database (Farr et al.n.d. suggest restricte and geographic ranges of fungi too numerous to comprehen- sively list here,but it is instructive to provide examples.Three species of Pucci niastrum occur only on Asian species of Acer,maple,and only in Asia.Rust occurs on the English oak,Ouercus robur,in its native range but not in its introduced range in North America,even though Uredinales is well represented on North American Quercus.Amelanchier alnifolia supports 17 rust taxa in its native range in North America,but none in Eur ope whe it has naturalized (Zerbe and Wirth 2006) Pr and abs in ite nat ralized nge allow fo enative range of a fungus(Table 1.1):presence in both of th host's ranges is problematic only in the absence of historical records of absence of the fungus in one of them It is even more tempting to think that the combination of Fahrenholz's rule and knowledge of the native ranges of plants can be used to further strengthen inferences of native ranges of fungi.Fahrenholz's rule postulates that"parasites and their hosts speciate in synchrony"(Hafner and Nadler 1988) If host switching were not an issue,then native ranges of hosts should also be native ranges of thei parasites.Howe er,host switching is an issue (Jacksor 2004. Host switching is best exemplified by absence in the host's native range an presence in its naturalized range.This requires some explanation,aided by the example of Eucalyptus rust (Grgurinovic et al.2006).Puccinia psidii causes Eucalyptus rust but the first reports of this disease were not from the native range of species of Eucalyptus in Australia.Instead,this rust fungus was first reported on plantations of eucalypts grown in Brazil.Evidently,P.psidii had switched,orjumped,from species of Myrtaceae native to South America to introduced st pecies of Eucalyp s that also belo an also be inf rred fro the early yea of culture.e.g for fo ae spec umeria graminis. Stric coevolution was apparently absent between thi
1.4 Inferring Native Ranges of Fungi from Pathogen Release It is tempting to think that fungi with restricted host ranges must be native where their hosts are native. Camellia, a genus of some 200 species, is endemic in eastern Asia, with its center of diversity in southern China (Ta and Bartholomew 1984). As Ciborinia camelliae is restricted to Camellia, its discovery on C. japonica in Great Britain in 1999 (Jones and Baker 2007) should ultimately be traced back to a native range in eastern Asia even though other parts of the world may have been stepping stones. If species of Camellia had never been introduced outside eastern Asia, the inference of sympatry for its host-restricted fungi would be unequivocal. Ornamental species of Deutzia provide such an example in that they are also endemic to Asia where seven taxa of rust fungi commonly infect them (Farr et al. n.d.). Unlike C. camelliae, records of rust on Deutzia outside the native range are absent even though D. scabra was introduced to the U.S. as early as 1822 (Rehder 1940). Searches of the SMML Fungus–Host Distribution Database (Farr et al. n.d.) suggest restricted host and geographic ranges of fungi too numerous to comprehensively list here, but it is instructive to provide examples. Three species of Pucciniastrum occur only on Asian species of Acer, maple, and only in Asia. Rust occurs on the English oak, Quercus robur, in its native range but not in its introduced range in North America, even though Uredinales is well represented on North American Quercus. Amelanchier alnifolia supports 17 rust taxa in its native range in North America, but none in Europe where it has naturalized (Zerbe and Wirth 2006). Presence in the host’s native range and absence in its naturalized range allow for strong inference of the native range of a fungus (Table 1.1); presence in both of the host’s ranges is problematic only in the absence of historical records of absence of the fungus in one of them. It is even more tempting to think that the combination of Fahrenholz’s rule and knowledge of the native ranges of plants can be used to further strengthen inferences of native ranges of fungi. Fahrenholz’s rule postulates that “parasites and their hosts speciate in synchrony” (Hafner and Nadler 1988). If host switching were not an issue, then native ranges of hosts should also be native ranges of their parasites. However, host switching is an issue (Jackson 2004). Host switching is best exemplified by absence in the host’s native range and presence in its naturalized range. This requires some explanation, aided by the example of Eucalyptus rust (Grgurinovic et al. 2006). Puccinia psidii causes Eucalyptus rust but the first reports of this disease were not from the native range of species of Eucalyptus in Australia. Instead, this rust fungus was first reported on plantations of eucalypts grown in Brazil. Evidently, P. psidii had switched, or jumped, from species of Myrtaceae native to South America to introduced species of Eucalyptus that also belongs to Myrtaceae. Host switching can also be inferred from the early years of agriculture, e.g., for formae speciales of Blumeria graminis. Strict coevolution was apparently absent between this 1 Fungal Pathogens of Plants in the Homogocene 11
12 G.Newcombe and F.M.Dugar an'Dljous (S)ndaosns
Table 1.1 Three categories of first encounters between evolutionarily naive plants and novel pathogens that depend on two factors: (1) which party to the encounter is alien, and (2) whether opportunities for encounters will be prolonged or brief. The category of the encounters in turn affects how susceptible and resistant outcomes of first encounters contribute, or not, to biotic resistance Time Evolutionarily naive parties to first encounter Biotic resistance (BR) of native biotic community versus alien plants or alien pathogens Category Opportunities for encounters (prolonged by pathogen reunion) Naive Plant Novel Pathogen Expected outcome of encounters, if contributing to BR Examples of encounters (discussed in text) 1 Extended, as naturalized, alien plants remain exposed to pathogens of native plants (no) Alien Native Susceptible (S) R (Prunus serotina resistant to Uredinales in Europe). S (individuals of Pinus sylvestris susceptible to Endocronartium harknessii in North America) 2 Extended, when alien pathogens are reunited with alien, naturalized plants (yes) Native Alien Resistant (R) R (6 of 7 taxa of Malus resistant to Podosphaera leucotricha in N. America). S (Malus angustifolia, the seventh taxon, susceptible to P. leucotricha in N. America) 3 Short (no) Native Alien Resistant (R) R (North American pines with Cr genes for resistance to Cronartium ribicola, an Asian fungus). S (individuals lacking these genes, in the same North American pines) 12 G. Newcombe and F.M. Dugan
1 Fungal Pathogens of Plants in the Homogocene 13 powdery mildew and its grass family hosts in western Asia (Wyand and Brown 2003).Likewise,the barley scald pathogen,Rhynchosporium secalis,apparently evolved on other hosts outside the center of diversity for barley (Zaffarano et al. 2006. 1.5 Genetic Criteria for Native Range Host range is sometimes not specific enough to even suggest a particular native range for a fungus.For example.Venturia inaequalis.the apple scab pathogen affects all species of Malus,some of which are native to North America although most are Eurasian.Records of occurrence of V.inaequalis might be misleading that the ppl e first dom sticated in c introdu t it wou uld grow ck1950). It is expl hypothesiz e tha at genetic variation will be greatest in th native range.Using thi criterion,Gladieux et al.showed that V.inaequalis is likely native to the same area in Asia in which apple itself was domesticated (Gladieux et al.2008).Similarly.an Asian origin of the dry rot fungus,Serpula lacrymans,has been inferred from a study of its genetic variation (Kauserud et al.2007).Genetic variation also places the amphibian chytrid fungus,Batrachochytrium dendrobatidis,in a native range in South Africa (Weldon et al.2004)from where it has spread to cause a pander ic Interestingly,the highest rsity for the hu an de is in Afric mo sa itself olved(Gra ere al.2007 Gene diversity s not likely to be by itself an infallibl criterion of native range however.Plants with outcrossing mating systems are frequently as genetically diverse in their naturalized or invaded ranges as they are in their native ranges (Novak and Mack 2005).Ambrosia artemisiifolia.a north American plant.main tains high genetic diversity in its invaded range in France(Genton et al.2005). Studies of Bromus tectorum have shown that even selfing plants may be as diverse in the invaded range,in North America,as in the native,Eurasian range,if the invasion involved multir le introductions (Novak and Mack 2005).A re ent summary of 20 analyzes of netic diversity in invasive plant populations sh ed that ates of genetic diversity vary from"none detected o"high"(Ward et al.2008) Similar caveats may apply to le use of a genetic criterion for determination of native ranges of fungi;genetic diversity of pine-associated Sphaeropsis sapinea is high in South Africa where the fungus must have been introduced from the northern hemisphere(Smith et al.2000).Conversely.North American populations of Entoleuca mammata are genetically more variable than introduced populations in Europe (Kasanen et al.2004).The oak wilt fungus,Ceratocystis fagacearum,is only known to occur in the middle and eastern United States,but its genetic homogeneity has led so chers to hyp othesize an Mexico or Ce ntra rica have been suggested (Juzwik etal.2008) hypothesis to explain lo a new and reproductively solated strain or species (Zambino and Harrington 2005)
powdery mildew and its grass family hosts in western Asia (Wyand and Brown 2003). Likewise, the barley scald pathogen, Rhynchosporium secalis, apparently evolved on other hosts outside the center of diversity for barley (Zaffarano et al. 2006). 1.5 Genetic Criteria for Native Range Host range is sometimes not specific enough to even suggest a particular native range for a fungus. For example, Venturia inaequalis, the apple scab pathogen, affects all species of Malus, some of which are native to North America although most are Eurasian. Records of occurrence of V. inaequalis might be misleading in that the apple, first domesticated in central Asia, was introduced by early explorers everywhere that it would grow (Hedrick 1950). It is common in such cases to hypothesize that genetic variation will be greatest in the native range. Using this criterion, Gladieux et al. showed that V. inaequalis is likely native to the same area in Asia in which apple itself was domesticated (Gladieux et al. 2008). Similarly, an Asian origin of the dry rot fungus, Serpula lacrymans, has been inferred from a study of its genetic variation (Kauserud et al. 2007). Genetic variation also places the amphibian chytrid fungus, Batrachochytrium dendrobatidis, in a native range in South Africa (Weldon et al. 2004) from where it has spread to cause a pandemic. Interestingly, the highest genotypic diversity for the human dermatophyte, Trichophyton rubrum, is in Africa, where Homo sapiens itself evolved (Gra¨ser et al. 2007). Genetic diversity is not likely to be by itself an infallible criterion of native range, however. Plants with outcrossing mating systems are frequently as genetically diverse in their naturalized or invaded ranges as they are in their native ranges (Novak and Mack 2005). Ambrosia artemisiifolia, a North American plant, maintains high genetic diversity in its invaded range in France (Genton et al. 2005). Studies of Bromus tectorum have shown that even selfing plants may be as diverse in the invaded range, in North America, as in the native, Eurasian range, if the invasion involved multiple introductions (Novak and Mack 2005). A recent summary of 20 analyzes of genetic diversity in invasive plant populations showed that estimates of total genetic diversity vary from “none detected” to “high” (Ward et al. 2008). Similar caveats may apply to sole use of a genetic criterion for determination of native ranges of fungi; genetic diversity of pine-associated Sphaeropsis sapinea is high in South Africa where the fungus must have been introduced from the northern hemisphere (Smith et al. 2000). Conversely, North American populations of Entoleuca mammata are genetically more variable than introduced populations in Europe (Kasanen et al. 2004). The oak wilt fungus, Ceratocystis fagacearum, is only known to occur in the middle and eastern United States, but its genetic homogeneity has led some researchers to hypothesize an exotic origin; the oak populations of Mexico or Central America have been suggested (Juzwik et al. 2008). An alternative hypothesis to explain low diversity is “the local genesis of a new and reproductively isolated strain or species” (Zambino and Harrington 2005). 1 Fungal Pathogens of Plants in the Homogocene 13
14 G.Newcombe and F.M.Dugan However.low genetic variation can also characterize ancient and relictual spe cies.For example,the Wollemi pine,Wollemia nobilis,is the only extant member of its genus,surviving only as a single,small population in a canyon in Australia (Peakall et al.2003).No genetic variation whatsoever has been detected in W.nobilis.Genetic drift can reduce genetic variation in small and isolated populations of plants(Ouborg et al.2006).and of organisms generally,but this explanation for genetic homogeneity of fungal species has not been widely considered. 1.6 Inferring Native Ranges of Pathogenic Fungi from Resistance Given historical examples of extreme susceptibility of plants to novel pathogens (e.g.,chestnut blight,and white pine blister rust).and the resistance of related plants elsewhere,it is te ng to think that host resistance c an indicate native pathogen Many ve followed this line evolution anation is that of electio for res stanc e in the umptive native range.and the absence of such selection elsewhere.The host in the presumptive native range is thought to be evolutionarily adapted,whereas the host in the invaded range is said to be evolutionarily naive with respect to the novel pathogen.The Asian species of Castanea and Pinus were certainly more resistant to the chestnut blight and white pine blister rust fungi,respectively,than were North American species of Castaned and Pinus.With these examples as paradigm,researc other pa hoger fungi that we nide oI p y known prior s in do the pathogen has to be present i the native range of a mostly re istant,putative host. Secondly,th nce of the latter has to be adaptive and cannot be complete;the pathogen has to be able to survive and reproduce.so there must either be some susceptible individuals o an otherwise resistant species.or the adapted host could be tolerant (Rov and Kirchner 2000).Even then,it may prove surprisingly difficult to distinguish adap tive resistance,which is associated with ongoing selection,from exapted resistance (Newcombe 1998)Exaptations are chara ers that in a new volutio nary cont ha if the ion els (Gould nd Vrh 982).F 0 poplar nativ to eastern As a,to species of on Ca ada' ancouver Island could only have been construed as adaptive if there had beer evidence that those fungi were native to eastern Asia.Needless to say.that would not have been a parsimonious interpretation (Newcombe 2005). The evolution of plant-pollinator and plant-herbivore interactions may be linked through exaptations (Armbruster 1997).However.the linkages of exaptations fo resistance to pathogens are unknown.One can only speculate that the function of nes in a poplar species in eastem Asia for resist nce to Van ver Island fungi
However, low genetic variation can also characterize ancient and relictual species. For example, the Wollemi pine, Wollemia nobilis, is the only extant member of its genus, surviving only as a single, small population in a canyon in Australia (Peakall et al. 2003). No genetic variation whatsoever has been detected in W. nobilis. Genetic drift can reduce genetic variation in small and isolated populations of plants (Ouborg et al. 2006), and of organisms generally, but this explanation for genetic homogeneity of fungal species has not been widely considered. 1.6 Inferring Native Ranges of Pathogenic Fungi from Resistance Given historical examples of extreme susceptibility of plants to novel pathogens (e.g., chestnut blight, and white pine blister rust), and the resistance of related plants elsewhere, it is tempting to think that host resistance can indicate native range of a pathogen. Many plant pathologists have followed this line of thinking. Their evolutionary explanation is that of selection for resistance in the presumptive native range, and the absence of such selection elsewhere. The host in the presumptive native range is thought to be evolutionarily adapted, whereas the host in the invaded range is said to be evolutionarily naive with respect to the novel pathogen. The Asian species of Castanea and Pinus were certainly more resistant to the chestnut blight and white pine blister rust fungi, respectively, than were North American species of Castanea and Pinus. With these examples as paradigm, researchers have tried to infer the native range or origin of many other pathogenic fungi that were unknown or poorly known prior to an epidemic. The difficulties in doing so are twofold. First, the pathogen has to be present in the native range of a mostly resistant, putative host. Secondly, the resistance of the latter has to be adaptive and cannot be complete; the pathogen has to be able to survive and reproduce, so there must either be some susceptible individuals of an otherwise resistant species, or the adapted host could be tolerant (Roy and Kirchner 2000). Even then, it may prove surprisingly difficult to distinguish adaptive resistance, which is associated with ongoing selection, from exapted resistance (Newcombe 1998). Exaptations are characters that in a new evolutionary context can have selective value even if they resulted from selection for something else (Gould and Vrba 1982). For example, the resistance of Populus maximowiczii, a poplar native to eastern Asia, to species of Venturia and Taphrina on Canada’s Vancouver Island could only have been construed as adaptive if there had been evidence that those fungi were native to eastern Asia. Needless to say, that would not have been a parsimonious interpretation (Newcombe 2005). The evolution of plant-pollinator and plant-herbivore interactions may be linked through exaptations (Armbruster 1997). However, the linkages of exaptations for resistance to pathogens are unknown. One can only speculate that the function of genes in a poplar species in eastern Asia for resistance to Vancouver Island fungi 14 G. Newcombe and F.M. Dugan
1 Fungal Pathogens of Plants in the Homogocene 15 would presumably be related to defense of some kind.Adaptive resistance must also be distinguished from nonhost resistance that is predicted,but not explained. by phylogenetic signal(Gilbert and Webb 2007:Newcombe 2005). Unfortunately,little attention has been paid to these distinctions when inferring e rang by Phytophthora late alis,the ens.In dis ed relati lateralis was des ribed in 1942.It i suspected to be of Eurasian origin because Asiatic species of Chamaecyparis resist it;resistance may have arisen through coevolution with the pathogen(Sinclair et al. 1987)."Eighteen years later however.Sinclair adopted a much more cautious position:"P.lateralis,origin unknown"(Sinclair and Lyon 2005).This caution in inferring native ranges of pathogens from the geographic distribution of resistance is warranted as the examples which follow will hopefully make clear.In the instance just disc ssed,the nature of res sistanc ada versu exapted non ost)of Asian Chamaecyparis to P.la ains unclear Dogwood anthracnose,for example is also thought to be caused by a high profile,alien pathogen,Discula destructiva.Discula destructiva has been regardec as an alien pathogen in North America (Redlin 1991),that has more recently appeared in the U.K.(Jones and Baker 2007).and in western Europe (Holdenrieder and Sieber 2007).Its appearances in the U.K.and Europe have been on Cornus florida that is native to North America.Susceptibility associated with severe damage and mortality in nature is only seen in the North American species c florida and C.nuttallii (Sinclair and Lyon 2005).Inferr ing the nativ D destru ould c clude that it ally,resistance is not a helpful criter on b ecause it characterize at least some species of Cornus s native to each one of the three con nents wher D.destructiva could be native:North America,Europe,and Asia(Holdenrieder and Sieber 2007:Sinclair and Lyon 2005).Although the native range of D.destructivo has been hypothesized on the basis of host resistance to specifically coincide with that of Cornus kousa in eastern Asia (Redlin 1991),the fungus has never been reported on C.kousa in its native range.Fourteen fungi have been recorded on C.kousa in eastern Asia (Farr et al.n.d.).but D.destructiva is not one of them ntrodu ced orna ta susceptibl re to ccinia in Japan a numb of tax and a few oth fungi.but D.destructiva has not been record on it.This evidenc of absence in Japan is not definitive but it does suggest that resistance-based inferences of origin can be quite misleading. Butternut canker provides another example.Sirococcus clavigignenti-iuglanda cearum is thought to be an alien pathogen in North America that might have been introduced on seed of Asian species of Juglans such as J.ailantifolia (Ostry and Moore 2007).But the only record of S.clavigignenti-juglandacearum on J.ailantifolia is from the U.S.(Ostry 1997).and re nd) dsof this fungus from Asia are lacking o Nine fungal taxa have be en recorded so the fur iof the putative,adaptive host are no completely unresearched (Farr et al.n.d.)
would presumably be related to defense of some kind. Adaptive resistance must also be distinguished from nonhost resistance that is predicted, but not explained, by phylogenetic signal (Gilbert and Webb 2007; Newcombe 2005). Unfortunately, little attention has been paid to these distinctions when inferring native ranges of plant pathogens. In discussing root rot of Port Orford cedar caused by Phytophthora lateralis, the assumed relationship between resistance and origin of a pathogen was stated in this way: “P. lateralis was described in 1942. It is suspected to be of Eurasian origin because Asiatic species of Chamaecyparis resist it; resistance may have arisen through coevolution with the pathogen (Sinclair et al. 1987).” Eighteen years later however, Sinclair adopted a much more cautious position: “P. lateralis, origin unknown” (Sinclair and Lyon 2005). This caution in inferring native ranges of pathogens from the geographic distribution of resistance is warranted as the examples which follow will hopefully make clear. In the instance just discussed, the nature of resistance (i.e., adaptive versus exapted/ nonhost) of Asian Chamaecyparis to P. lateralis remains unclear. Dogwood anthracnose, for example, is also thought to be caused by a highprofile, alien pathogen, Discula destructiva. Discula destructiva has been regarded as an alien pathogen in North America (Redlin 1991), that has more recently appeared in the U.K. (Jones and Baker 2007), and in western Europe (Holdenrieder and Sieber 2007). Its appearances in the U.K. and Europe have been on Cornus florida that is native to North America. Susceptibility associated with severe damage and mortality in nature is only seen in the North American species, C. florida and C. nuttallii (Sinclair and Lyon 2005). Inferring the native range of D. destructiva from these observations, one would conclude that its origins were Eurasian. But, actually, resistance is not a helpful criterion because it characterizes at least some species of Cornus native to each one of the three continents where D. destructiva could be native: North America, Europe, and Asia (Holdenrieder and Sieber 2007; Sinclair and Lyon 2005). Although the native range of D. destructiva has been hypothesized on the basis of host resistance to specifically coincide with that of Cornus kousa in eastern Asia (Redlin 1991), the fungus has never been reported on C. kousa in its native range. Fourteen fungi have been recorded on C. kousa in eastern Asia (Farr et al. n.d.), but D. destructiva is not one of them. Interestingly, C. florida, grown as an introduced ornamental in Japan, has proven susceptible there to endemic Pucciniastrum corni, a number of taxa of Erysiphales, and a few other fungi, but D. destructiva has not been recorded on it. This evidence of absence in Japan is not definitive but it does suggest that resistance-based inferences of origin can be quite misleading. Butternut canker provides another example. Sirococcus clavigignenti-juglandacearum is thought to be an alien pathogen in North America that might have been introduced on seed of Asian species of Juglans such as J. ailantifolia (Ostry and Moore 2007). But the only record of S. clavigignenti-juglandacearum on J. ailantifolia is from the U.S. (Ostry 1997), and records of this fungus from Asia are lacking on any host (Farr et al. n.d.). Nine fungal taxa have been recorded on J. ailantifolia in Asia, so the fungi of the putative, adaptive host are not completely unresearched (Farr et al. n.d.). 1 Fungal Pathogens of Plants in the Homogocene 15