Lecanosticta acicola: A growing threat to expanding global pine forests and plantations

Summary Lecanosticta acicola causes brown spot needle blight (BSNB) of Pinus species. The pathogen occurs mostly in the Northern Hemisphere but has also been reported in Central America and Colombia. BSNB can lead to stunted growth and tree mortality, and has resulted in severe damage to pine plantations in the past. There have been increasingly frequent new reports of this pathogen in Europe and in North America during the course of the past 10 years. This is despite the fact that quarantine practices and eradication protocols are in place to prevent its spread. Taxonomy Kingdom Fungi; Phylum Ascomycota; Subphylum Pezizomycotina; Class Dothideomycetes; Subclass Dothideomycetidae; Order Capniodales; Family Mycosphaerellaceae; Genus Lecanosticta. Host range and distribution Lecanosticta spp. occur on various Pinus species and are found in North America, Central America, South America (Colombia), Europe as well as Asia. Disease symptoms Small yellow irregular spots appear on the infected pine needles that become brown over time. They can be surrounded by a yellow halo. These characteristic brown spots develop to form narrow brown bands that result in needle death from the tips down to the point of infection. Needles are prematurely shed, leaving bare branches with tufts of new needles at the branch tips. Infection is usually most severe in the lower parts of the trees and progresses upwards into the canopies. Useful websites The EPPO global database providing information on L. acicola (https://gd.eppo.int/taxon/SCIRAC) Reference genome of L. acicola available on GenBank (https://www.ncbi.nlm.nih.gov/genome/?term=Lecanosticta+acicola) JGI Gold Genome database information sheet of L. acicola sequenced genome (https://gold.jgi.doe.gov/organism?xml:id=Go0047147)


SUMMARY
Lecanosticta acicola causes brown spot needle blight (BSNB) of Pinus species. The pathogen occurs mostly in the Northern Hemisphere but has also been reported in Central America and Colombia. BSNB can lead to stunted growth and tree mortal ity, and has resulted in severe damage to pine plantations in the past. There have been increasingly frequent new reports of this pathogen in Europe and in North America during the course of the past 10 years. This is despite the fact that quarantine prac tices and eradication protocols are in place to prevent its spread. Taxonomy: Kingdom Fungi; Phylum Ascomycota; Subphylum Pezizomy cotina; Class Dothideomycetes; Subclass Dothideomy cetidae; Order Capniodales; Family Mycosphaerellaceae; Genus Lecanosticta. Host range and distribution: Lecanosticta spp. occur on various Pinus species and are found in North America, Central America, South America (Colombia), Europe as well as Asia. Disease symptoms: Small yellow irregular spots appear on the infected pine needles that become brown over time. They can be surrounded by a yellow halo. These characteristic brown spots develop to form narrow brown bands that result in needle death from the tips down to the point of infection. Needles are prema turely shed, leaving bare branches with tufts of new needles at the branch tips. Infection is usually most severe in the lower parts of the trees and progresses upwards into the canopies. Useful websites: The EPPO global database providing informa tion on L. acicola (https ://gd.eppo.int/taxon/ SCIRAC) Reference genome of L. acicola available on GenBank (https :// www.ncbi.nlm.nih.gov/genom e/?term=Lecan ostic ta+acicola) JGI Gold Genome database information sheet of L. acicola sequenced genome (https ://gold.jgi.doe.gov/organ ism?xml:id=Go004 7147) Keywords: brown spot needle blight, Lecanosticta acicola, Lecanosticta species, Mycosphaerella dearnessii, pine pathogen, Pinus spp.

I N T RO D U C T I O N
Lecanosticta acicola is an ascomycete fungus that causes a dis ease of Pinus spp. known as brown spot needle blight (BSNB). The pathogen was first described by de Thümen (1878) and it owes its notoriety to a disease problem that arose in the south eastern USA on Pinus palustris, better known as long leaf pine in that area (Siggers, 1932). This tree species, which is highly susceptible to infection, is peculiar in having a socalled 'grass' stage during the first five years of its growth. This mass of young needles provides a favourable environment for infection to occur.
The BSNB pathogen completes its life cycle ( Fig. 1) on pine nee dles that are shed prematurely. This leads to reduced or stunted growth that can result in significant yield losses (Wakeley, 1970) or tree death. In some cases, pine plantations have been suffi ciently damaged that they have needed to be cleared (Huang et al., 1995;Lévy, 1996;Markovskaja et al., 2011).
Lecanosticta acicola has been recorded on 53 different Pinus species and hybrids in native and nonnative pine stands in the USA, Canada, several European countries and Asia as well as in Central America and Colombia (Table 1). Due to the severity of the disease, the pathogen has been afforded an A1 quarantine status in Africa, Argentina, Chile, Uruguay, Bahrain, Kazakhstan, Ukraine and Russia, and A2 quarantine status in Europe (https :// gd.eppo.int/taxon/ SCIRA C/categ oriza tion). However, reports of new outbreaks of the disease in various European countries have increased significantly since 2008 (Adamson et al., 2015(Adamson et al., , 2018Anonymous, 2012;Cleary et al., 2019;Hintsteiner et al., 2012;Jankovský et al., 2009a;Markovskaja et al., 2011;Mullett et al., 2018;Ortíz de Urbina et al., 2017).
Quarantine measures rely on accurately identifying the presence of pathogens on symptomatic tissues. This is com plicated in the case of L. acicola where the symptoms of BSNB closely resemble those of Dothistroma needle blight (DNB). DNB is caused by two species: Dothistroma septosporum and D. pini . Due to their similar symptoms, field diagnoses of the causal agent based on symptoms and/or on morphology alone have commonly been incorrect (Shishkina and Tsanava, 1967;Siggers, 1944;Thyr and Shaw, 1964). Consequently, past reports of L. acicola based only on morphological descriptions and symptoms must be treated with caution and verified using molecular identification tech niques (van der Nest et al., 2019).
Lecanosticta acicola has been wellknown in the southeastern USA since the early 1900s, but is rapidly spreading in northern parts of the USA, Canada and in some parts of Europe (Broders et al., 2015). Its complete host range is not known but appears to be expanding (Mullett et al., 2018). A recent taxonomic reevalu ation of isolates previously identified as L. acicola, applying phy logenetic analyses based on DNA sequences, has led to various isolates being recognized as distinct species (Quaedvlieg et al., 2012;van der Nest et al., 2019). This and a number of recent publications (Adamson et al., 2018;Cleary et al., 2019;Mullett et al., 2018;Ondrušková et al., 2018;Ortíz de Urbina et al., 2017;Sadiković et al., 2019;Schneider et al., 2019;Wyka et al., 2017) justifies the need for a review of current knowledge regarding BSNB and the Lecanosticta species that cause this disease. This is the first review of the topic to be presented in 75 years subse quent to that of Siggers (1944).

L EC A N OST I C TA S PEC I E S
The genus Lecanosticta, which includes nine species with the type species being L. acicola (previously known as Mycosphaerella dearnessii, Table 2), is characterized by stromata and septate, pigmented conidia. The genus was erected by Petrak in 1922 (Sydow andPetrak, 1922). The taxonomic history and no menclature of Lecanosticta acicola has been succinctly presented previously (Evans, 1984;Siggers, 1944) and is summarized and updated in Table 2.
Lecanosticta acicola is the oldest known species in the genus and owes its notoriety to the disease of long leaf pine, which it was first associated with, in the southeastern USA (Chapman, 1926;Hedgcock, 1929). Although the pathogen was identified in           M o l ec u l a r P l a n t Pat h o lo gy (2019) 20(10), 1327-1364 Siggers (1944) USA, Alabama 1948-1967 P. palustris Symptoms were observed annually on seedlings and the proportion of seedlings affected were recorded.
In a 4-year study, 78% or more seedlings were infected yearly with L. acicola. M o l ec u l a r P l a n t Pat h o lo gy (2019) 20(10), 1327-1364  Siggers (1944) the identification was based on characteristics that do not fit Lecanosticta and therefore this record should be verified.

Boyer (1972)
Hedgcock (1929) USA, Iowa 1967-1971 P. sylvestris Symptoms were observed in the field and the proportion of needles affected were noted. In some cases, microscopic examinations of conidia were used for identification. Skilling and Nicholls (1974) USA, Kansas 1929 P. nigra var. austriaca Hedgcock reported on collections of the pathogen at the office of Forest Pathology at Washington, D.C. and the Mycological collections of the US Department of Agriculture. According to Siggers (1944) the identification was based on characteristics that do not fit Lecanosticta and therefore this record should be verified.
Hedgcock (1929) USA, Kansas 1951 P. nigra, P. ponderosa Reports in the field and mycological identification. Rogerson (1953) USA, Kansas 1967-1971 P. sylvestris Symptoms were observed in the field and the proportion of needles affected were noted. In some cases, microscopic examinations of conidia were used for identification. Skilling and Nicholls (1974) USA, Kentucky 1929 P. nigra var. austriaca Hedgcock reported on collections of the pathogen at the office of Forest Pathology at Washington, D.C. and the Mycological collections of the US Department of Agriculture. According to Siggers (1944) the identification was based on characteristics that do not fit Lecanosticta and therefore this record should be verified. Hedgcock (1929) USA, Kentucky 1967-1971 P. sylvestris Symptoms were observed in the field and the proportion of needles affected were noted. In some cases, microscopic examinations of conidia were used for identification. Skilling and Nicholls (1974)   Hedgcock (1929) USA, Louisiana 1929Louisiana -1930Louisiana , 1960 P. palustris Symptoms were observed and the proportion of seedlings affected were recorded at 4-5 years of age and again at 30 years.
Most of the trees were affected.
Wakeley ( Siggers (1944) the identification was based on characteristics that do not fit Lecanosticta and therefore this record should be verified.
Hedgcock (1929) USA, Missouri 1947-1949 P. ponderosa Symptoms were observed in the field and morphological identifications were made. Both the sexual state and asexual state were observed.
All trees were affected. Excessive needle defoliation and in some cases tree mortality was observed.
Luttrell (1949) USA, Missouri 1967-1971 P. sylvestris Symptoms were observed in the field and the proportion of needles affected were noted. In some cases, microscopic examinations of conidia were used for identification. Skilling and Nicholls (1974) USA, New England 2016 P. strobus Severe needle browning was observed and L. acicola was identified as part of a complex of species causing premature defoliation. This is possibly WPND although it was not defined as such.
Brazee (2016) USA, New Hampshire 2011 P. strobus Isolates were collected and morphologically identified in a survey. These isolates were later identified with TEF 1 sequencing and mating type 1 was detected.
Honduras, Yoro 1981 P. caribaea, P. oocarpa These isolates were previously identified as L. acicola based on morphological characteristics (Evans, 1984). Multigene phylogenetic analysis revealed it to be L. guatemalensis. * van der Nest et al.    Central America based on morphological characteristics (Evans, 1984), it is now recognized as a Northern Hemisphere pathogen for which phylogenetic analyses of the translation elongation factor 1α gene ( (Tables 1 and 2) (Evans, 1984;Marmolejo, 2000;Quaedvlieg et al., 2012;van der Nest et al., 2019). Evans (1984) recognized    The ex-type of L. guatemalensis (IMI281598) was initially described as L. acicola based on morphological characteristics (Evans, 1984   considerable morphological variation amongst his collections of L. acicola. In that study, he described a second species, L. gloeospora from Pinus pseudostrobus in Mexico, and the fun gus remains known only from Mexico on this host (Evans, 1984;Marmolejo, 2000). The novelty of this species was recently vali dated using DNA sequence data (van der Nest et al., 2019). Lecanosticta longispora was first described based on mor phological features from P. culminicola in Nuevo León, Mexico (Marmolejo, 2000). This species was characterized in a phyloge netic study by Quaedvlieg et al. (2012), and was distinguished from L. acicola based on differences in the TEF 1 and βtubulin 2 (BT 2) gene sequences. That study was the first to delineate spe cies of Lecanosticta based on phylogenetic inference (Quaedvlieg et al., 2012). These authors included several samples from Central America that had previously been identified as L. acicola, as well as the collection used by Marmolejo (2000) to typify L. longispora. In their phylogenetic analyses (Quaedvlieg et al., 2012), L. acicola was not identified from Central America but two new species, L. brevispora and L. guatemalensis, were described (Tables 1 and 2). Evans (1984) observed that ecotypes or morphotypes exist amongst isolates of L. acicola in Central America, depending on the altitude and hosts from which the isolations were made. He therefore hypothesized that Central America could be the centre of origin of Lecanosticta. This was later supported by analysis of TEF 1 sequence data that revealed high genetic diversity in this geographical region (Janoušek et al., 2016). An extensive collec tion of isolates from Central America was recently studied using a phylogenetic approach (van der Nest et al., 2019). Interestingly, L. acicola was not identified amongst isolates from Guatemala, Nicaragua or Honduras. Furthermore, the isolates considered to be L. acicola by Evans (1984) were sequenced and identified as L. gua temalensis and a new species, L. variabilis (van der Nest et al., 2019, Table 1). Lecanosticta brevispora was identified in Guatemala and Honduras on Pinus oocarpa and P. pseudostrobus (Table 1), ex panding the host range and distribution for that species. Likewise, L. guatemalensis was also identified in Guatemala, Honduras and Nicaragua on P. caribaea, P. oocarpa and P. tecunumanii (Table 1) With only one exception, which is probably a taxonomic incon gruity, Lecanosticta species are all associated with Pinus species. Petrak (1954) described Phragmogloeum gaubae on Callistemon sieberi in Australia (Petrak, 1954). von Arx (1983) attempted to re duce various species with overlapping characteristics to fewer gen era and found that Phragmogloeum had the same morphological   Barr, 1996), Lecanosticta gaubae was transferred to that new genus (Crous, 1999). The genus Eruptio was further evaluated and it was found that L. acicola and D. septosporum were not conge neric (Crous, 2009). Consequently, Lecanosticta was selected as the correct name for Eruptio acicola following the one fungus one name convention Hawksworth et al., 2011). Because Eruptio gaubae is morphologically similar to Lecanosticta, phylo genetic analyses are required to resolve this taxonomic confusion.
Lecanosticta acicola is the only species in the genus known to be a significant pathogen. This is particularly important because it is spreading rapidly in Europe and the northeastern parts of North America. Therefore, all data collected over time regarding Lecanosticta pertain to the organism that was assigned the name L. acicola, and the remainder of the review will focus on this spe cies. However, it is relevant to recognize that other species of Lecanosticta cause symptoms similar to those of L. acicola and that they have the potential to emerge as pine pathogens if they were accidentally moved to new environments. They would then be recognized as members of a complex of BSNB pathogens.

S Y M P TO M S O F B RO W N S P OT N E E D L E B L I G H T
Symptoms of infection can vary depending on the host species af fected. Typically, a small and yellow, sometimes light greygreen or reddish brown, irregular circular spot, with defined margins, ap pears at the point of infection (Hedgcock, 1929) (Fig. 2C-E). These spots soon become brown as the infections mature and they are often surrounded by a yellow halo (Skilling and Nicholls, 1974). In severe cases, infections can occur on several parts of a needle, leading to more rapid necrosis (Fig. 2E). The characteristic brown spots are the first conspicuous symptoms on the pine needles and this has led to the common name 'brown spot needle blight' proposed by Siggers (1932). These brown spots can also appear resinsoaked depending on the host species (Skilling and Nicholls, 1974). In some cases, as has been reported in P. strobus, symp toms may only be displayed as chlorosis of the needles without banding (Broders et al., 2015). Infected needles die from the apex to the base (Fig. 2B) and they are eventually shed from the trees (Hedgcock, 1929;Skilling and Nicholls, 1974). Usually only the second and thirdyear needles are affected, leaving healthy new growth at the tips of the branches. The new growth tips are then infected in the subsequent season by inoculum on older needles (Skilling and Nicholls, 1974). Generally, infection is more severe in the lower parts of the canopy and then progresses upwards in the trees (Sinclair and Lyon, 2005;Skilling and Nicholls, 1974).
An asymptomatic phase in which L. acicola establishes within needles can last several days (Setliff and Patton, 1974) to 3 months (Skilling and Nicholls, 1974). This is dependent on the strain of the pathogen (Kais, 1972) and length of the wet season. This delay in symptom development could lead to the accidental movement of infected plants to new areas.
The symptoms of BSNB (Fig. 2) can easily be confused with those of DNB, which is caused by Dothistroma septosporum and D. pini (Barnes et al., 2004. On some host species, symp toms of DNB are similar to those of BSNB (Fig. 2B) but rather than the characteristic brown discoloration and spots, a distinct red band forms around the point of infection in the case of DNB (Pehl and Cech, 2008). However, in some cases the characteristic red banding pattern associated with DNB is not formed or alter natively the red bands are sufficiently dark to give a false impres sion of brown spots. This can easily lead to incorrect pathogen diagnoses Petrak, 1961).

L I F E C YC L E
Lecanosticta acicola can occur in either its asexual or sexual state (Fig. 1) (Siggers, 1939). The pathogen overwinters in ac ervuli (asexual) (Fig. 1Aa) or ascostromata (sexual) (Fig. 1Ba) in the dead tissue of either dead or living pine needles. It can also overwinter as vegetative mycelium in the infected needles that remain attached to the host (Siggers, 1944). Conidia are released in gelatinous masses (Fig. 1Ab) or ascospores are released from asci in ascostromata (Fig. 1Bb) on the needles when the light, temperature and humidity are favourable (Kais, 1975;Tainter and Baker, 1996).
Conidia begin to germinate on the needle surfaces by devel oping one to four germ tubes, depending on the number of cells in the conidia (Setliff and Patton, 1974). It is uncertain whether the germ tubes are attracted to the stomata, or whether they grow randomly over the needle surface (Patton and Spear, 1978;Setliff and Patton, 1974). Light plays an indirect, but essential role in the infection process as it stimulates the opening of sto mata, allowing the germ tube to penetrate the needle (Fig. 1c) (Kais, 1975). Infections can also occur through wounds (Kais, 1978). Once a germ tube enters the stomatal antechamber, it increases in diameter and becomes thickwalled and melanized (Patton and Spear, 1978). Appresoria, such as those found in Dothistroma (Gadgil, 1967), have never been seen (Patton and Spear, 1978).
Once the mesophyll tissue has been invaded by L. acicola mycelium, conidiomata begin to form. These begin to integrate with the needle tissue and increase in size until they are visible to the naked eye (Wolf and Barbour, 1941). The conidiophores produce conidia towards the leaf exterior (Evans, 1984), which exerts pressure on the needle epidermis. This causes the epider mis to rupture, leaving a flap that partly covers the conidiomata (Wolf and Barbour, 1941). The conidia are released from the co nidiomata during wet weather and the disease cycle is repeated.
In the case of the sexual state, asci are formed within the ascostromata on necrotic distal parts of living needles or on dead needles (Henry, 1954;Jewell, 1983). Ascospores are released from asci and dispersed through wind and rain. Asci and asco spores develop more rarely than conidia and have been reported only from Nicaragua, Honduras, Colombia and the southern parts of the USA (Table 1) (Evans, 1984;Henry, 1954;Kais, 1971;Luttrell, 1949;Siggers, 1944). The reports from Nicaragua and Honduras probably represent species other than L. acicola.

TOX I N PRO D U C T I O N
Many plant pathogenic fungi have adapted to produce toxic sec ondary metabolites in their plant hosts and these could influence colonization and sporulation, as has been seen in D. septosporum (Kabir et al., 2015). Lecanosticta acicola is known to produce the toxic compounds LAI and LAII, which are heatresistant and nonhost specific phytotoxins (Yang et al., 2002(Yang et al., , 2005. The two Fig. 3 Maximum likelihood (ML) tree representing the nine known species of Lecanosticta as well as the three lineages of L. acicola generated from the translation elongation 1-α region. ML bootstrap support (>70%) are indicated first, followed by maximum parsimony (MP) bootstrap support values (ML/MP, * = insignificant value). Phaeophleospora gregaria was used as the outgroup taxa. All represented type species are indicated in bold.
compounds interact with the host independently and do not pro mote or inhibit the interaction of one another (Yang et al., 2002). Different Pinus species have different reactions to LAI and LAII. When rooted cuttings of P. thunbergii were exposed to the toxin, they showed little sensitivity to it. In contrast, when P. elliottii and P. taeda, both highly susceptible to BSNB infection, were exposed to the toxin, the results showed high sensitivity to LAI (Ye and Qi, 1999). It seems likely that these toxins are involved in the destruction of mesophyll tissue of the pine needles at the point of infection (Jewell, 1983).

B I O LOG Y A N D D I SS E M I N AT I O N
Conidia and ascospores are released throughout the year at temperatures ranging from -5.5 to 28 °C (Kais, 1971;Siggers, 1944;Wyka et al., 2018). However, warm and wet weather is particularly conducive for the development of BSNB, irrespec tive of whether infection takes place by sexual or asexual spores. The conidia do not germinate below 5 °C, although most survive this temperature and commence germination once the tempera ture increases (Siggers, 1944). At the other extreme, tolerance to high temperature was found to vary depending on the strain of Lecanosticta involved. It was shown that conidia of isolates from the northern parts of the USA could not germinate at 32 °C, whereas cultures isolated from the southern parts of the USA, as well as China, had a germination success of 80% at the same temperature (Huang et al., 1995). This physiological distinction is reflected in population genetic studies which define two lineages of the pathogen in the USA (Janoušek et al., 2016). The success of the pathogen may therefore be a result of isolates in each line age adapting to local temperature conditions. The maximum temperature for the germination of L. acicola conidia is 35 °C (Siggers, 1944). It was also found that high hu midity pre and postinfection is required for high levels of in fection (Kais, 1975). The optimal temperature for infection to occur is 30 °C during the day and 21 °C at night, and Kais (1975) showed that these temperatures gave positive results in inocu lation trials.
Conidia are dispersed predominantly by rain splash to ad jacent trees, and they contribute significantly to rapid disease buildup in pine stands (Tainter and Baker, 1996). High levels of conidial dispersal were recorded during the rainy season in the USA, especially between late spring and summer, as well as when there were rain spells after a long period of dryness (Kais, 1971). In other reports, conidial production and dispersal were recorded throughout the year (Siggers, 1944). Dispersal was not influenced by the temperature range but conidial release was connected to rainfall patterns. In Wisconsin, two peaks of co nidial release were recorded, with the first peak in early summer when young pine needles are present and the second in late sum mer (Skilling and Nicholls, 1974), which was similar to that found in the northeastern USA (Wyka et al. 2018). In Japan, it was found that conidia were produced by the pathogen from early spring to autumn with peak dispersal in midsummer. However, for a second year of infection, the dispersal was most abundant from late summer to midautumn the following year (Suto, 2002). A study in Fujian province (China) showed that the greatest num ber of conidia were detected between early spring and midsum mer and again in late summer to late autumn in Pinus elliottii plantations (Li et al., 1987). It consequently appears that conidial dispersal varies depending on the rainfall season in any particu lar geographical region.
Spore traps in several studies failed to capture ascospores (Kais, 1971;Siggers, 1939;Wyka et al., 2018). It was found, however, that conidia could be dispersed to a distance of up to 60 m (Wyka et al. 2018). A recent investigation of the dispersal of Dothistroma, where the mechanisms of conidial and ascospore dispersal are similar to those in L. acicola, showed that conidia could be naturally disseminated over more than 1 km . The assumed distance of dispersal in L. acicola may, consequently, be similar.
The ascospores of L. acicola are forcibly expelled into the air (Wolf and Barbour, 1941) and dispersed by wind currents (Kais, 1971) or rain splash driven by wind (Siggers, 1939). Ascospores can also be released during periods of fog, rain and dew (Tainter and Baker, 1996). Ascospores were recorded in the USA mainly during periods when temperatures were above 15 °C and are found in late summer to autumn. Small numbers of ascospores were detected when temperatures were below 10 °C (Kais 1971).
The main component that facilitates spread of conidia and ascospores is moisture, but other factors may also aid in their dispersal. Insect dissemination was suggested as a mechanism of conidial spread when two Lepidopteran wing scales were found to have conidia attached to them (Skilling and Nicholls, 1974). Given the biology of L. acicola, it seems unlikely that insects are involved in its dissemination. It has also been suggested that an imals grazing in forests might aid in dissemination of the conidia when spores stick to their coats or hooves (Skilling and Nicholls, 1974;Tainter and Baker, 1996). Again, this mode of dissemina tion seems unlikely to be particularly important.
Anthropogenic movement of infected plant material has con tributed to the dissemination of many tree pathogens (Wingfield et al., 2015). This has been clearly demonstrated for Dothistroma septosporum (Barnes et al., 2014), which has a biology very sim ilar to that of L. acicola. A study that used microsatellite markers has demonstrated that two separate lineages of L. acicola have most likely been introduced into Europe from North America (Janoušek et al., 2016). Long distance dispersal of L. acicola is, therefore, likely to be the result of anthropogenic movement of infected plant material. This would not include seed transmis sion as L. acicola conidia cannot survive on a pine seed's surface longer than 30 to 34 days and it is thus not considered seed borne (Jianren and Chuandao, 1988).

D I S E A S E M A N AG E M E N T
Several measures have been suggested to prevent BSNB during plantation establishment. The most effective is to plant disease free seedlings of superior quality (Cordell et al., 1990;Skilling and Nicholls, 1974). It is also advisable to avoid establishing new plantations alongside old, infected pines that could potentially serve as reservoirs of inoculum (Tainter and Baker, 1996). For natural pine stands, the application of thinning treatments was investigated as a silvicultural practice against pine needle dis eases . This practice, conducted on native stands of P. strobus in the USA, showed promise in reducing the fungal load of L. acicola, resulting in reduced severity of the disease over time in stands already infected with the pathogen . This practice is recommended as a preven tative measure in stands that are at risk of infection by L. acicola and other pine needle pathogens .
Pruning of infected pines can contribute to the spread of BSNB if it is conducted during rainy or wet periods. This is be cause conidia are exuded during these conditions and can at tach to the pruning shears, providing a means of spread from infected to healthy trees (Skilling and Nicholls, 1974). Cutting blades should be cleaned during pruning and clipped needles and shoots should be removed (Kais, 1978). In the case of infec tion on Pinus palustris, which begins growth as a grass stage, stimulation of growth during the first 3 years of growth reduces the levels of infection (Tainter and Baker, 1996). Because this treatment is economical, effective and environmentally safe, it is widely used in the southeastern USA (Cordell et al., 1990), where BSNB occurs on P. palustris.
Breeding for resistance to L. acicola has been successfully used to reduce the impact of the disease on P. palustris in Alabama. The source population of these trees found in south western Alabama was used in breeding programmes (Snyder and Derr, 1972) where seed was made available to the public (Phelps et al., 1978). Since 1982, resistant phenotypes of P. elliottii have also been selected for in plantations affected by BSNB in the Fujian province in China. Over time, and using artificial inocula tions, resistant clones were selected and resistant seed orchards were established (Ye and Wu, 2011).
Fungicide treatment can protect pine seedlings from infection by L. acicola. For example, when P. palustris was sprayed with fungicide, the seedlings displayed increased diameter growth in a single growing season, compared to untreated plants (Siggers, 1932). Seedlings, seed orchard trees and Christmas tree plan tations have been protected by Bordeaux mixture of copper sulphate and lime, which inhibits conidial germination, by a benomyl root treatment or by ferbam (Fermate ® ). Chlorothalonil, a broadspectrum organochlorine pesticide (products include Bravo ® , Daconil ® and Maneb ® ), has also been applied to provide efficient control against BSNB. Chlorothalonil is also very effec tive against Lophodermium needle cast, which could be advan tageous when both pathogens are present (Cordell et al., 1990;Kais et al., 1986;Skilling and Nicholls, 1974). Practical details and recommendations concerning fungicide treatment can be found in Skilling and Nicholls (1974). However, the use of chemicals is not considered a desirable solution for disease control due to negative environmental factors and many of these treatments are no longer available.
Controlled burning in pine forests can eliminate competing vegetation and reduce the impact of needle pathogens, espe cially in P. palustris where a grass stage is relevant (Barnett, 1999;Chapman, 1932). This pine species is completely adapted to survive fires as it concentrates all its energy into root develop ment during the first 5 years of growth (Chapman, 1932). Siggers (1934) showed that a single controlled fire can significantly de crease BSNB in P. palustris until the next season and that during the initial growth stage, before seedlings begin to increase in height, a winter burn every 3 years is the most beneficial for disease control. The efficacy of controlled burns differs depend ing on the Pinus spp. involved and on the ability to tolerate fire damage.
In countries and regions where L. acicola is a quarantine or ganism, it is suggested that complete eradication of diseased trees or pine stands should be performed once the disease is detected (Pehl and Cech, 2008). This is achieved by felling and burning of infected trees and litter found under infected trees (Sosnowski et al., 2009). In Lithuania, for instance, after posi tive identification of the pathogen in the Curonian Spit in 2009 and 2010, effective eradication measures were implemented (Markovskaja et al., 2011). Due to this rapid action, the disease has remained under control in that country and is under con stant monitoring by the state plant service of the Ministry of Agriculture of Lithuania (https ://gd.eppo.int/taxon/ SCIRA C/distr ibuti on/LT). Eradication efforts are, however, not always effec tive, and the best preventative method is to limit the movement of plant material across borders and between regions. As new knowledge is emerging regarding different genetic entities of the pathogen, including strains of different mating types (Sadiković et al., 2019), the importance of avoiding new introductions is becoming increasingly obvious.

H OST R A N G E , H OST SU SC E P T I B I L I T Y A N D G EOG R A PH I C D I ST R I B U T I O N
In an effort to consolidate 140 years of literature with regards to the geographical distribution and host range of L. acicola, a de tailed list of these data has been compiled (Table 1). This shows that the pathogen has been reported in 31 countries and on 53 pine species and pine hybrids. The majority of the host records on native and nonnative trees are from the Americas, followed by Europe. The pathogen has not been found in Africa, Australia or New Zealand and in South America it is known only in Colombia. Of the 69 reports of the pathogen (Table 1), 31 were made in the last decade (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019). This suggests that incidences of the pathogen are most likely increasing.
In North America, the first report of L. acicola was in 1876 on native Pinus echinata as Cryptosporium aciculum (de Thümen, 1878). Since then, the pathogen has been reported in the USA on several susceptible species, including nonnative P. caribaea and P. pinea, and native P. elliottii, P. echinata, P. glabra, P. pon derosa, P. rigida, P. taeda and P. virginiana (Hedgcock, 1929;Siggers, 1944;Sinclair and Lyon, 2005;Webster, 1930) as well as on regionally planted exotic species such as P. attenuata, P. coulteri, P. muricata and P. sabiniana (Siggers, 1944). Pinus pa lustris seedlings are the most severely affected, largely due to the grass stage associated with early growth and where BSNB can cause complete defoliation (Siggers, 1934). Here it can result in mortality reaching 50% and higher in the southeastern USA (Cordell et al., 1990). New reports of L. acicola causing damage on P. strobus have emerged since 2005 in the northeastern USA and Canada and these have been attributed to changes in pre cipitation and climate in the regions (Broders et al., 2015;Wyka et al., 2017Wyka et al., , 2018. Lecanosticta acicola is also recognized as a component of a complex of pathogens that cause white pine needle damage (WPND) in this region (Broders et al., 2015). Additionally, the pathogen has been reported on P. banksiana and P. contorta var. latifolia in Canada (Laut et al., 1966).
Lecanosticta acicola has been reported from 17 European countries (for a complete list of records see Table 1). The patho gen was first recorded in northern Spain in 1942 (Martínez, 1942), where it still occurs on P. radiata (Ortíz de Urbina et al., 2017). In southwest Europe, L. acicola has caused severe defolia tion of P. radiata × P. attenuata, leading to the felling of 100 ha in the 1990s (Lévy, 1996). Lecanosticta acicola is spreading through the valleys in the Alps in Switzerland (Holdenrieder and Sieber, 1995), Austria (Cech, 1997;Hintsteiner et al., 2012), Italy (La Porta and Capretti, 2000) and Slovenia (Jurc and Jurc, 2010;Sadiković et al., 2019), which can be attributed to high humidity in deep valleys or the proximity of lakes. In Europe, L. acicola often infects P. mugo, a susceptible species on which it has re cently caused severe outbreaks in Austria (https ://gd.eppo.int/ repor ting/artic le5139). It also infects other pine species such as P. sylvestris and P. nigra. The pathogen has been recorded in sev eral peat bog sites in southern Bavaria (Germany) and southern Bohemia (Czech Republic). These locations are naturally humid throughout the year and the susceptible pine species P. mugo and/or P. uncinata subsp. uliginosa can be heavily infected, lead ing to considerable mortality. Similarly, L. acicola was recorded in the Baltic states (Drenkhan and Hanso, 2009) and, most recently, also in Sweden (Cleary et al., 2019). These records usually come from stands close to the sea or, very frequently, from botanical gardens or urban areas.
Other pine species such as Pinus × rhaetica and P. ponder osa have also been affected by L. acicola (Adamson et al., 2015(Adamson et al., , 2018. Lecanosticta acicola has been present in Croatia on P. halepensis for more than 40 years (Milatović, 1976;Sadiković et al., 2019). Interestingly, the pathogen was identified only at a single site in Ireland despite largescale screening throughout the British Isles (Mullett et al., 2018). From all these records, it is reasonable to conclude that L. acicola is spreading in Europe in native and nonnative pine species, in plantations and natural forests, and associated with different climatic conditions.
In Asia, BSNB has been reported in China in plantations of nonnative P. thunbergii, P. elliottii and P. taeda where the trees were severely damaged by the pathogen (Huang et al., 1995), and on P. caribaea, P. palustris, P. clausa and P. echinata that were reported to be susceptible to infection (Li et al., 1986). It was suggested that native pines such as P. taiwanensis, P. fenzeliana and P. massoniana were highly resistant to infec tion (Huang et al., 1995;Li et al., 1986). BSNB has been re ported on native P. thunbergii in Japan (Suto and Ougi, 1998) as well as on native P. thunbergii in South Korea but the disease was not severe (Seo et al., 2012).
Although some species of Pinus seem to not be suscepti ble to infection by L. acicola, the pathogen has the potential to overcome host resistance in a favourable environment and expand its host range, as is suggested for D. septosporum and D. pini (Drenkhan et al., 2016). For example, L. acicola is rarely reported on native P. sylvestris in Europe. Considering the importance of P. sylvestris in Europe, it will be important to monitor the presence of the pathogen on this host. Only single incidences of L. acicola have been reported on P. syl vestris in Austria (Cech and Krehan, 2008), Slovenia (Jurc and Jurc, 2010) and most recently in Estonia (Adamson et al., 2018) and Ireland (Mullett et al., 2018). In contrast, L. acicola is an important pathogen of P. sylvestris grown as part of the Christmas tree industry since the 1960s in the USA (Skilling and Nicholls, 1974). This implies that under favourable conditions this host could be infected by the pathogen. Investigations on the impact of DNB on P. sylvestris revealed that there is high intraspecific variability of P. sylvestris in Europe and that susceptibility of the host to the pathogen varies between in dividuals (Perry et al., 2016a,b) and this could also influence the potential importance of L. acicola. Unusually high humidity associated with climate change could increase pathogen pres sure on P. sylvestris (Perry et al., 2016a) and the single inci dences in Europe should carefully be monitored. Caution must also be taken when planting susceptible exotic hosts alongside native forests, as this could influence the vulnerability of na tive forests (Piotrowska et al., 2018).
Of the 69 reports of L. acicola, only 22 used DNA sequence comparisons for species verification. This is of concern as there might be an over or underestimation of hosts affected by BSNB globally. In Central America, for example, L. acicola was reported based on identifications using morphological characters. Because the pathogen has not yet been confirmed as occurring in this region using DNA sequences (Quaedvlieg et al., 2012;van der Nest et al., 2019), those reports could be erroneous and may represent different species which could possibly cause new outbreaks if not contained in their native environment.

M O L EC U L A R D I AG N OST I C S A N D F U T U R E PROS PEC T S Molecular markers used for species identification
Three molecular methods are currently being used to accu rately identify L. acicola. These include sequencing of vari ous gene regions, an ITSRFLP method and a conventional PCR that uses speciesspecific primers. The most common of these approaches is comparison of DNA sequences for the ITS gene region (Adamson et al., 2015(Adamson et al., , 2018Cleary et al., 2019;Markovskaja et al., 2011;Mullett et al., 2018). However, the TEF 1 (Fig. 3) and BT 2 gene regions have been recommended to distinguish between species of the Mycosphaerellaceae (Quaedvlieg et al., 2012). In order to accurately distinguish between different species of Lecanosticta, van der Nest et al. (2019) used a multigene phylogenetic approach using sequences for the ITS, TEF 1, BT 1, MS204 and RPB 2 gene regions. The outcome was the discovery of four new species, with the ITS and TEF 1 proving to be the gene regions showing the best amplification success across all species. Pehl et al. (2004) developed an ITSRFLP method to distinguish between L. acicola, D. septosporum and ten other plant pathogens. However, whether this method remains valid after the recogni tion of various new species (van der Nest et al., 2019) will need to be established.
Another rapid method allowing for the identification of L. ac icola, D. septosporum and D. pini is a conventional PCR that uses speciesspecific primers (Ioos et al., 2010). These were developed to partially amplify the TEF 1 gene for L. acicola and D. pini, and partially amplify the BT 2 gene region in D. septosporum (Ioos et al., 2010). Importantly, this method can be used to identify the pathogens directly from infected needles (Adamson et al., 2015;Ortíz de Urbina et al., 2017;Schneider et al., 2019) and is now widely used for preliminary identification of L. acicola (Adamson et al., 2018;Sadiković et al., 2019). A multiplex qPCR was also recently developed to detect L. acicola as well as Dothistroma species from needles simultaneously using probelabelled prim ers developed by Ioos et al. (2010) and Schneider et al. (2019), which could become more widely used once that technology is more easily available.

Population genetic studies
Knowledge regarding the population structure and diversity of pathogens such as L. acicola allow for an understanding of mi gration patterns as well various aspects of their invasion biology. Eleven polymorphic microsatellite markers and mating type prim ers have been developed for this purpose (Janoušek et al., 2014). The first population genetic study using these markers revealed that two lineages of L. acicola were introduced into Europe, pos sibly on two separate occasions (Janoušek et al., 2016). These results are similar to an earlier study where RAPD analysis of L. acicola, collected in the northern and southern parts of the USA and China, showed that the Chinese population originated from the southern USA and that the collection from the north ern USA was unique (Huang et al., 1995). A second population genetic study compared populations from Croatia and Slovenia and revealed four distinct populations with possible introduc tions from other sources within the two countries (Sadiković et al., 2019). Currently available knowledge suggests a Northern American centre of origin for this pathogen (Huang et al., 1995;Janoušek et al., 2016;van der Nest et al., 2019) but further sam pling and analyses are required to support this hypothesis. In the population genetic study of Janoušek et al. (2016), the micros atellite markers amplified poorly for the L. acicola isolates from Mexico and Central America. A later study (van der Nest et al., 2019) showed that these isolates were L. variabilis, a new and recently described species.
The study by Janoušek et al. (2014) showed that L. acicola is heterothallic and that two individuals, one with a MAT111 idiomorph and the other with a MAT12 idiomorph, are needed for sexual reproduction to occur. Consequently, to understand whether sexual recombination might occur in a region, it is im portant to have a knowledge of the mating type idiomorph dis tribution. Mating type primers that amplify the MAT111 and MAT12 idiomorphs and that tested positive for Dothistroma species as well as L. acicola, L. guatemalensis and L. gloeospora have been developed (Janoušek et al., 2014). It is, however, not yet known whether these markers will amplify these gene re gions for the other, newly described Lecanosticta species. Janoušek et al. (2016) considered the global L. acicola pop ulation and showed that the ratio of mating type idiomorphs in Mississippi, Austria, France and Germany reflected sexual recom bination in these regions/countries. In contrast, only asexual re production occurs in the Czech Republic and northern parts of America. Using the mating type markers of Janoušek et al. (2014), the distribution of MAT1 and MAT2 isolates was detected in stud ies with isolates from Croatia (Sadiković et al., 2019), Estonia (Adamson et al., 2015(Adamson et al., , 2018, Ireland, Portugal, Russia (Mullett et al., 2018) as well as Spain (Ortíz de Urbina et al., 2017). In Spain, both mating types were detected whereas only single mat ing types were detected in all other areas studied. However, in Estonia it was suggested that a second introduction of the patho gen occurred since only MAT1 was initially present but that later both mating types were detected in the same region (Adamson et al., 2015). In populations with equal ratios of mating types or with both mating types present, sexual reproduction could occur, possibly giving rise to more virulent strains. This emphasizes a need to exercise caution and thus to prevent introduction of new strains into regions where the pathogen is already present.

Future prospects in the age of genomics
Canada's Michael Smith Genome Sciences Centre has recently re leased a full genome for a L. acicola isolate from France (https :// www.ncbi.nlm.nih.gov/assem bly/GCA_00050 4345.2#/def). This genome has not yet been annotated but provides a valuable re source for future studies. Many other genomes of Dothidiomycetes, which have been sequenced and annotated, are available for com parative purposes Ohm et al., 2012). Annotation of putative genes of the L. acicola genome, utilizing knowledge of these other genomes, will provide insights into questions regarding many aspects of the biology of L. acicola. Opportunities also now arise to sequence the genomes of other Lecanosticta spp. and to compare these in order to better understand their relative impor tance. It will also be possible to follow the Dothistroma example where a transcriptomic study considered which genes are expressed during various stages in the infection of P. radiata (Bradshaw et al., 2016) and genome sequencing of global representatives of D. septosporum revealed that gene copy numbers could play a role in dothistromin production by the pathogen (Bradshaw et al., 2019).

CO N C LU S I O N S
Lecanosticta acicola has been known in the southern USA for many decades. Consequently, its life cycle, mode of infection, host susceptibility and strategies to prevent infection, particu larly on P. palustris, have been extensively studied in that region. Yet there is evidence to show that the pathogen, which now has an extensive host range, is spreading rapidly northwards. The reasons for this host range and geographical expansion re quire further study. Contemporary knowledge has also shown that there have been two introductions of L. acicola into Europe. Consequently, BSNB is becoming a disease of great concern in Europe, where it is increasingly being discovered on both non native and native Pinus spp. There are many relevant hypotheses to explain the growing importance of BSNB and these include the effects of climate change, emergence of more aggressive strains of the pathogen and anthropogenic processes leading to new introductions. There is clearly a need for increased attention to and studies of L. acicola, particularly in Europe.
Recent studies have shown that there are eight species of Lecanosticta in addition to L. acicola. All of these other species appear to have a Mesoamerican origin. Much of the literature pertaining to L. acicola needs to be reconsidered given the fact that a single name has been widely used to refer to what we now know represents numerous cryptic species. Lecanosticta acicola identified based on DNA sequence comparisons has not been found in Central America, suggesting a North American centre of origin. Of the 69 reports of L. acicola, only 25 from 12 countries have been confirmed using DNA sequencebased tools (Table 1). Many reports of the pathogen could thus be erroneous and there is an urgent need to resolve this important question.
All the available knowledge regarding BSNB relates to stud ies on L. acicola and these are predominantly from the USA. Nothing is known regarding the relative importance of the remaining eight species of Lecanostica. At least some of these are most likely also important pathogens and their relative threat to global forests and forestry needs to be assessed. A concerted effort must be made to prevent their accidental introduction into new regions of the world and as part of this process DNA sequencebased techniques need to be routinely applied to allow for meaningful identification.
The development of new tools to study Lecanosticta spp. and BSNB provides many exciting opportunities to enhance our knowledge of this important group of pathogens. The population structure and diversity of L. acicola can now be easily studied in the USA as well as where new invasions occur in Europe, and at levels that were previously not possible. For example, application of the available microsatellite markers will enable a more com prehensive understanding of the pathogen as well as determina tion of its centre of origin.
Genome sequencing is rapidly becoming cheaper and more readily available, and an isolate of L. acicola is already available in the public domain for study. We envisage that all the species of Lecanosticta will be sequenced in the relatively near future and many isolates of some species will likely also be studied at this level. These studies, and others relating to the 'omics' level, will surely have a substantial impact on our understanding of a group of pathogens that is growing in importance and relevance. Overall, BSNB (including all species of Lecanosticta) has the po tential to become a pine needle disease of global importance if proper preventative measures for the spread of the causal patho gens are not implemented.
Cooperative Program for financial support. Ariska van der Nest was supported by a Scarce Skills Doctoral Scholarship (no. 89086) provided by the National Research Foundation of South Africa. The authors have no conflicts of interest to declare.