Genetic resistance to Phytophthora lateralis in Port-Orford-cedar ( Chamaecyparis lawsoniana ) – Basic building blocks for a resistance program

of mortality in the and important conifer species Chamaecyparis lawsoniana (Port-Orford-cedar). The lateralis stabilize in


Societal Impact Statement
Non-native pathogens and pests cause high mortality to tree species globally and may imperil the future viability of associated forest ecosystems. Phytophthora lateralis, an oomycete, causes Port-Orford-cedar root disease and is a major cause of mortality in the ecologically and economically important conifer species Chamaecyparis lawsoniana (Port-Orford-cedar). The P. lateralis resistance program shows promise to help stabilize C. lawsoniana in its native range of northwestern California and southwestern Oregon, USA, and serves as a leading example of disease resistance breeding in forest trees Summary • A non-native, invasive pathogen, Phytophthora lateralis, has caused extensive mortality within the native range, northern California and southern Oregon USA, of Chamaecyparis lawsoniana (Port-Orford-cedar), as well as in horticultural and amenity plantings in the USA and Europe. Restoration of affected sites is contingent upon development of populations with genetic resistance. Naturally occurring genetic resistance has been identified in C. lawsoniana, and an active selective breeding program seeks to characterize and increase resistance levels.
• Two seedling root dip inoculation trials, assessed for mortality for nearly three years each, are used to examine the types and levels of genetic resistance in C.
lawsoniana. Most seedlings utilized in these trials are progeny of crosses from parent trees that exhibited apparent resistance to the disease in earlier trials.
• Seedling trials suggest that both qualitative major gene and quantitative disease resistance occurs in C. lawsoniana. Both types of resistance to P. lateralis appear to be present at levels high enough to be immediately useful for restoration and reforestation. The data suggest that the qualitative resistance is conditioned by a This article has been contributed to by US Government employees and their work is in the public domain in the USA.

| INTRODUC TI ON
Non-native pests and pathogens have negative impacts on forest health globally, especially where there is a high degree of susceptibility resulting from a lack of co-evolution alongside the disease or pest (Aukema et al., 2010;Britton & Liebhold, 2013;Fei, Morin, Oswalt, & Liebhold, 2019;Jung et al., 2018;Santini et al., 2013;Showalter et al., 2018). Despite the lack of co-evolution with these pests and pathogens at least some forest species will have genetic resistance (Sniezko & Koch, 2017;Sniezko, Smith, Liu, & Hamelin, 2014). In general non-coevolved resistance arises when a trait that evolved for a different function provides some resistance to a novel pest or pathogen; known as exaptation (Gould & Vrba, 1982). Knowing the type(s), frequency, and geographic distribution of resistance existing across the range of a tree species will provide valuable information useful to applied resistance programs interested in restoration and reforestation.
Phenotypic patterns of resistance are usually divided into two categories, qualitative and quantitative (Poland, Balint-Kurti, Wisser, Pratt, & Nelson, 2009). This is a simplistic division and we must acknowledge that there is a resistance continuum, and in many cases little can be inferred about the underlying inheritance. In the simplest case, qualitative resistance can be conditioned by a single major gene (e.g. Kinloch & Dupper, 2002), and at the other extreme, quantitative disease resistance (QDR), which can have a complex pattern of inheritance, is often the result of several to many genes interacting (King, David, Noshad, & Smith, 2010;Poland et al., 2009;St Clair, 2010). Durable resistance will be a necessity in longlived forest trees, and QDR is generally thought to provide a higher likelihood of being durable than qualitative resistance (Cowger & Brown, 2019;Poland et al., 2009;Sniezko, Johnson, & Savin, 2019).
However, there are only a few examples documenting QDR in forest tree species, and studies are needed to guide applied forestry and conservation programs (see Sniezko et al., 2019 for an example). Because there is growing recognition that in many forest trees other management activities will have limited impact on slowing the progress of a disease or pest epidemic, there may be a need for more resources to be directed toward development of resistant populations useful for restoration or reforestation (Showalter et al., 2018). We also note that plants can marshal defenses to pathogens that fall into two categories: resistance and tolerance. Tolerance, in contrast with resistance is not thought to impact the pathogen directly, but instead influences traits that can limit the health and fitness costs of being infected (Martins et al., 2019). Though both tolerance and resistance aid in improved survival of infected individuals it can be difficult to untangle the difference using phenotyping approaches alone.
The species has been extensively impacted by the non-native pathogen Phytophthora lateralis, which is responsible for the Port-Orfordcedar root disease (Betlejewski et al., 2003(Betlejewski et al., , 2004Hansen, Goheen, Jules, & Ullian, 2000). The pathogen, which was first discovered in a North American nursery in 1923, has subsequently led to high mortality of C. lawsoniana in portions of its native range ( Figure 1) (Hansen et al., 2000;Zobel et al., 1985). P. lateralis, an oomycete, is generally a soil and water-borne root pathogen and can kill trees of all sizes, with riparian areas most heavily affected (Hansen et al., 2000;Zobel et al., 1985). There are also examples of aerial infection, most notably in plantings in Europe (Robin et al., 2011). C. lawsoniana has also been a valuable tree for horticultural use in North America and Europe (often known as Lawson's cypress), but the impact of the pathogen has greatly curtailed its use (Zobel et al., 1985). Recent findings of P. lateralis on C. obtusa in Taiwan suggest Asia as the origin of this pathogen (Brasier, Vettraino, Chang, & Vannini, 2010;Webber et al., 2012). Although its main impact has been on C. lawsoniana, mortality of Taxus brevifolia (Pacific yew) has also been documented (DeNitto & Kliejunas, 1991;Hansen et al., 2000;Murray & Hansen, 1997), and a few additional conifers, including Juniperus communis and Microbiota decussatta also appear to be very susceptible (Ebba Peterson, personal communication).
In many of the affected areas, it is difficult to eliminate the pathogen directly, and land managers are reluctant to plant C. lawsoniana without access to seedlings with identified genetic resistance to the disease. The first testing of C. lawsoniana, using rooted cuttings from several hundred surviving trees in areas of high disease incidence, concluded that there was little or no resistance to P. lateralis or that refinements in testing techniques would be needed to uncover resistance (Hansen, Hamm, & Roth, 1989;Zobel et al., 1985). single major gene (designated here as Pla), but nothing is known about the number of genes involved in quantitative disease resistance.
• Seedling progeny from resistant parent trees in containerized seed orchards are now being used for restoration and reforestation. Resistant seedlings or clones could also be used to re-establish C. lawsoniana in urban forests.

K E Y W O R D S
major gene resistance, non-native pathogen, Phytophthora lateralis, Port-Orford-cedar root disease, qualitative resistance, quantitative disease resistance A subsequent trial suggested that there was variation in susceptibility among parent trees and their progeny, and that the resistance was due to a slowing in the rate of advance of P. lateralis and was not immunity (Hansen et al., 1989). This information and concerns about population viability and conservation status of C. lawsoniana raised awareness and helped provide the support needed to undertake a resistance program (Betlejewski et al., 2004;Farjon, 2013;Potter, Escanferla, Jetton, Man, & Crane, 2019). Large operational resistance screening trials began in the 1990s in which more detailed results, including some evidence for both QDR and MGR disease resistance emerged (Sniezko, 2004(Sniezko, , 2006. These trials were part of a comprehensive breeding program that was established at the USDA Forest Service's Dorena Genetic Resource Center (DGRC), Cottage Grove, Oregon, to aid in restoration and reforestation of C. lawsoniana (Sniezko et al., 2012b).
To improve success in restoration and reforestation, it is important to determine the type(s) of genetic resistance (MGR, QDR, or both) that exist so that breeding efforts can maximize the chance to deploy resistance that is durable. In this paper, we report on the use of two seedling inoculation trials from DGRC to examine the phenotypic range of P. lateralis resistance available in C. lawsoniana.
Using the data from these trials, which incorporate a more extended assessment period than used previously and many parent trees and their progenies, we seek to clarify whether progeny phenotypes display a qualitative or quantitative resistance response. The two screening trials, established nine years apart and with different isolates of the pathogen, provide the most comprehensive examination of resistance in the species to date.
We show here that there is evidence of both QDR and MGR resistance to the disease. These data will provide managers and breeders information that can be used to assign parent trees to seed orchards or breeding populations, make forward selections from within seedling families, and provide managers with estimates of usable resistance from the initial field selections (Sniezko et al., 2019).

F I G U R E 1
Chamaecyparis lawsoniana (Port-Orford-cedar) killed by Phytophthora lateralis (a) in South Slough National Estuarine Research Reserve (SSNERR), southern Oregon; (b) mortality along Munger Creek, southern Oregon; (c) dying tree with characteristic P. lateralis stain above root collar in SSNERR; (d) mortality in Redwood National Park, northern California; and (e) seedling with lesion at the root collar typical of phenotype in root dip and raised bed tests

| MATERIAL S AND ME THODS
The two greenhouse inoculation trials used here were initiated in 2005 and 2014 and are part of an annual series of trials for an applied resistance breeding program that is evaluating progenies of over 1,000 parents selected in an earlier stage of screening. The parents used in these trials included some of the earliest field selections from the resistance program as well as more recent field selections. The early selections had been used for more than 10 years in numerous trials (of one-year or less in duration) and were the progenitors of most of the resistant and susceptible controls in these two trials. The susceptible controls used are typical of the level of resistance that might be found in most parent trees in wild C. lawsoniana populations. The other parents in these trials were a subset of the >14,000 trees that had been evaluated to a lesser extent (generally in one or two earlier trials). Prior to the 2005 trial, none of the parents had been evaluated using a 3-year testing period.
The trials include seedling families from full-sib and self-pollinations as well as open-pollinated families from wild stands and containerized orchards. Self-pollinated families were used because of the ease of making the cross (although they had somewhat lower seed yields). The self-pollinated families showed little or no evidence of inbreeding depression in traits that could cause these families to show higher and faster mortality (R. Sniezko, unpublished). A self-pollinated family from a susceptible control parent was included in each trial. Prior to sending the seedlings to Oregon State University (OSU) for inoculation, seedling tops were cut in Blocks 1, 2, and 3 for rooting at DGRC for potential later inclusion in a breeding orchard. In addition, for eight families some of the tops were cut in the remaining blocks for a different study. The seedlings were treated with a foliar application of Chlorothalonil and Benomyl approximately two months prior to inoculation to discourage foliar disease.

| Experimental design
Seedlings were inoculated at OSU in early 2005 using the root dip inoculation method (Hansen et al., 2012). Inoculum consisted of a mixture of zoospores from isolates 366 and 368 of P. lateralis collected previously from diseased trees; cultures for block 1 were grown in pea broth while those for blocks 2-7 were grown in carrot broth. Before inoculation, any roots growing out of the bottom of the containers were clipped. The bottom 1 cm of each tube was immersed in the P. lateralis zoospore suspension for 48 hr. Inoculation of the seedlings was spread out over four weeks, beginning on  forward selections (five of them crossed with the susceptible CON1 parent). The trial had two susceptible and two resistant controls (Dataset S1). Most of the parents used in full-sib crosses or selfs had been rated resistant in either previous stem dip trials or both a stem dip and root dip trials that used rooted cuttings (R. Sniezko unpublished data).
Seed were sown in 164 ml Ray Leach supercell cone-tainers during early Spring 2013, and the seedlings were cultured in an unheated greenhouse at DGRC. Seedlings were treated with beneficial nematodes to reduce potential root weevil damage. Seedlings were placed in a RCB design, with six blocks of up to seven seedlings per family, arranged in row plots (19-42 total seedlings per family, an average of 37.2 seedlings per family). In early December 2013 before transport to OSU for inoculation, seedling tops were cut to reduce the risk of foliar disease and to produce rooted cuttings for potential inclusion in seed orchards. The trials were terminated when plant health was no longer tenable in the pot system being used.

| Survival and Days-to-mortality
Survival was calculated by assessment date, and survival curves were generated for each family. DTM was calculated for each dead seedling as the number of days between its inoculation date and the date it was rated as dead. For families with no survival, the mean DTM included all seedlings in a family, but for families with survival >0%, it included only those dead (censored population). DTM provides an indicator of the presumed rate of progression of the P. lateralis pathogen in each seedling following inoculation. We used the survival curves to help characterize the variation in resistance both within and among families and interpreted the variation in survival and DTM as a reflection of the broad range in resistance of parent trees and their progeny in C. lawsoniana.

| Classifying resistance type
The Mendelian segregation ratios 3:1, 1:1, and 1:3 for alive:dead seedlings were tested for each family using exact binomial tests in R version 3.5.0 (R Core Team, 2018). Parent trees whose families did not significantly differ from the 3:1 and 1:1 ratios potentially possess MGR, controlled by a single dominant gene, while those with 1:3 ratios may possess a single recessive gene for resistance. For those parents represented in more than one cross or family, examination of all families in-

| RE SULTS
Both seedling root dip inoculation trials were successful, with the susceptible controls reaching 0% survival and averaging less than 130 and 100 DTM in trials one and two respectively ( Table 2, Dataset S1). In trial one, there was no statistical difference in survival or DTM in the three blocks where tops had been cut, compared to the three blocks where seedling tops were intact. In both trials, the range of survival among the many families spanned nearly a continuous distribution from 0 to 100 (Figures 2a, 3a), and family variation in DTM occurred across the range of survival in both trials ( Figure 4). The five families in common to the two trials showed very similar performance, despite the difference in isolates used and the 9-year difference in initiation of the trials (Table 1). There were significant differences among families for survival (p < .05) in both trials during the four time intervals tested and for DTM (Table 3, Dataset S1), with the family means varying widely (Figures 2, 3, 4 for both survival and DTM). In some families, seedling survival plateaued early, but for others survival continued to decrease over the duration of the two trials (Figures 2, 3). By the end of both trials, only 8 and 29 families had 0% survival in trial one and tworespectively.

| Major gene resistance
In trial one, three of the CF1 families showed no significant difference from a 3:1 Mendelian segregation ratios of live:dead seedlings. Where the 4four open-pollinated (OP) families in trial one were from field selections from southern Oregon forests, and the six OP families in trial two were orchard open-pollinated. The full-sib (CC) families include crosses among parents that showed differing levels of resistance and susceptibility in prior resistance testing. The S 1 and S 2 are first-and second-generation self-pollinated families, respectively, from parents that were either among the top 10% selections in previous stem dip trials or showed some resistance in prior root dip trials. The resistant controls included families that have major gene resistance (MGR), quantitative disease resistance (QDR) or both, and all but one were orchard open-pollinated seedlots; the remaining resistant control was one controlled cross (CC) in trial two. The susceptible controls showed high mortality and low DTM in previous resistance screening trials, and include two self-pollinated lots, one controlled cross, and one bulked collection from fast-dying susceptible parents in the Dorena Genetic Resource Center clone bank.
for MGR (Table 4 further illustrates the survival of parents 117490 and CF1 in several crosses, relative to crosses involving the susceptible control). We designate this locus Pla.
Parents CF1 and 117490 are the most studied MGR candidates, but in trial one, 44 of the 125 families (including 10 with parent CF1) showed no significant difference from alive:dead segregation ratios, either 3:1 (12 families) or 1:1 (33 families, including one family which also showed no difference for 3:1) (Dataset S1, Figure 5), suggesting that there is not enough evidence to preclude the possibility that one parent (1:1 ratio) (Figure 5b) or both parents (3:1) (Figure 5a) might be heterozygous for MGR. Multiple crosses with these parents are needed to help determine the MGR status of the parents. There were 15 parents involved in the crosses of families that failed to show differences from 3:1 segregation, and some of those parents were involved in other crosses. In the simplest case, we would expect a parent heterozygous for Pla to show no difference from a 3:1 or 1:1 alive:dead ratio in crosses with other heterozygotes or susceptible parents respectively.
Examination of 11 parents that were represented in more than one family showed that four parents had moderate to high survival in all crosses (including CF1, Figure 2d), providing further support that those four parents were heterozygous for MGR, while the seven other parents had at least one cross with 2.0 to 28.6% survival (Dataset S1), indicating that these parents were not MGR. The S 1 family of parent Forty-four of the families failed to show significant differences from a 1:3 segregation ratio and nine of these also failed to show significant differences from a 1:1 ratio (Dataset S1, Figure 5c). A Mendelian segregation ratio of 1:3 may signify that both the pollen parent and the seed parent are heterozygous for a recessive gene; additional crosses will be needed to determine if a recessive gene is responsible in any of these cases.

| Quantitative disease resistance
Compared to the susceptible controls, most of the other families in both trials displayed higher survival and/or longer DTM ( Figures   2-6, Tables 4, 5, S1). Although as noted above, resistance in some

F I G U R E 3 Variation in Chamaecyparis lawsoniana survival over time (days post-inoculation) following root dip inoculation with
Phytophthora lateralis in Trial 2 among (a) 142 families and one bulked seedlot (two families with 100% survival not shown); (b) four parents represented by both a self-pollinated S1 family (solid lines) and a full-sib cross involving susceptible CON1 parent (dotted lines), and compared to susceptible control (COS-31276 S1 family). Two of the four parents show putative major gene resistance (MGR), one parent shows only quantitative disease resistance (QDR), and the fourth parent has little resistance; and (c) the 10 families with the highest days to mortality (DTM 517.9 to 968.0), including several with possible major gene resistance ( families appears to be due to a single major gene, Pla, that leads to high survival, but for most other families showing resistance MGR does not appear to be present. The range of DTM found within and between these non-MGR families (Figures 2-6, Tables 4, 5, Table S1) suggests a more complex inheritance. This finding also suggests that resistance is under quantitative genetic control and presumably influenced by several to many genes, although a single gene could also be responsible for such a phenotypic pattern (Poland et al., 2009). It is also notable that in many cases, mortality does not reach 100% in these QDR families. Several examples discussed below illustrate the level and quantitative pattern of resistance.
The progeny of some parents were particularly notable for the extended length of time in which the mortality occurred relative to the susceptible control ( Figure 2f, Tables 5, Table S1).  (Table 5). Parent 70016 appears to be a strong candidate for having MGR, but also QDR; while parent 70072 displays only QDR, (note from Table 4 that it shows only 3.6% survival when crossed with a susceptible control CON1 parent that has little evidence of QDR).
The eight crosses involving parent SIS-41924, a parent previously characterized in the earlier trials as relatively fast-dying and highly susceptible (Sniezko, unpublished), showed a range of resistance much higher than the two susceptible controls (Figure 6a

| D ISCUSS I ON
The phenotypic distributions of the data from the seedling families suggest that both MGR and QDR are present in C. lawsoniana and further illustrate that by extending the screening trial duration to 3 years it is possible to more clearly delineate several types of genetic resistance. Here, we designate a parent (or family) QDR if it differs in survival and/or DTM from the susceptible control and if there is no evidence to support that it is MGR. Many families in these two trials show QDR and careful study will be needed to dissect it further and see if the inheritance, and basis for the resistance, varies by family, and whether factors such as modifier genes or recessive genes are also involved in at least some cases. We have shown that the range of QDR varies widely between parent trees, and at least for some parents a level of QDR is present in all crosses involving that parent. The frequency of parents with MGR appears to be rarer than those with QDR. Many of the putative QDR families had surviving seedlings at the end of the greenhouse trials, and those individuals would be good candidates for forward selection for breeding to increase the level of resistance and for inclusion in seed orchards. As with many other species, we acknowledge that the underlying inheritance of QDR is still unknown (Poland et al., 2009). Nevertheless, the phenotypes presented here provide the basis for future studies, including using genomic resources, to examine inheritance in greater detail. The individuals in this study, and the many others being tested in this breeding program, also provide information on many parent trees to select for inclusion in seed orchards or breeding to increase resistance, while maintaining a high level of genetic diversity in populations used in restoration or reforestation.
We suggest that the large differences in DTM within and among families indicate variation in the ability to slow the progress of the P. lateralis pathogen in the 20.6 cm length from the bottom of the seedling container to the root collar, since only the bottom 1 cm of root tissue was exposed to inoculum. Logistical constraints, foremost the general vigor of the seedlings in the small tubes, precluded continuing the greenhouse trials for longer periods; yet, since the  Relatively little is known about the underlying basis of the chemical and structural defenses against pathogens in forest trees (Kovalchuk et al., 2013). In another study of C. lawsoniana, some seedlings or rooted cuttings from resistant parents died, but the dead seedlings or rooted cuttings from some MGR parents often showed sunken, necrotic bark lesions at the root collar, often marked by resin flow, and many of those trees also died later relative to those of susceptible parents (Oh, Hansen, & Sniezko, 2006).
In stem-wound inoculations of the materials from the same parents, stems of resistant parents had shorter lesions, often marked by resin, while in foliage inoculations, expanding patches of necrotic scale leaves were apparent for the susceptible parents and one resistant parent but reduced on foliage of a different resistant parent, which also exuded resin (Oh et al., 2006). A histopathology investigation indicated that P. lateralis hyphae in roots of rooted cuttings or seedlings from MGR parents grew slower in cortical cells and did not seem to penetrate the vascular tissues. Additionally, infection in the roots of resistant parents was marked by thickening of the cortical cell walls, collapsed cells, wall thickening, and apposition of electron-dense materials and crystals in cell walls were some of the structural changes common in resistant stems (Oh & Hansen, 2007). At least some differences between the two MGR parents or their progeny were noted, including some limited xylem colonization and indications of differences in zoospore attraction to roots (Oh & Hansen, 2007). This may suggest that different genes are responsible for resistance between them, but further investigations on the nature of resistance are warranted in MGR families as well as the QDR families noted in the trials reported here.
The presence of a single major gene, designated Pla, is just the simplest of several hypotheses to explain the segregation pattern of the families denoted MGR here, we also acknowledge that there could be more than one locus involved. Most MGR parents identified thus far appear to be heterozygotes for the Pla gene; however, 117490, a field selection made in 1989 on the Rogue River-Siskiyou National Forest in southern Oregon, appears to be homozygous for Pla, showing little or no mortality in seedling families in both trials.
The low percentage of mortality in some families involving parent 117490 may be due to factors such as incomplete penetrance, modifier genes or microsite influences on stability of the resistance that have been suggested in other species where some deviations from expected ratios have occurred (Blaker & MacDonald, 1981; F I G U R E 5 Survival over time for Chamaecyparis lawsoniana seedling families showing failure to differ significantly (p = .05) from within family segregation ratios of (a) 3:1 (12 families), (b) 1:1 (33 families), and (c) 1:3 (44 families) for alive:dead in Phytophthora lateralis inoculation Trial 1. The 3:1 and 1:1 ratios provide evidence of major gene resistance (MGR), but data from other crosses with some parents suggest some of these families have quantitative disease resistance (QDR) and not MGR. One possible interpretation of the 1:3 ratio is the presence of a single recessive gene but see text for further discussion. Note: One family shows both 3:1 and 1:1 ratio; nine families did not differ significantly for both 1:1 and 1:3 ratios. Color and line types are used to specify type of genetic resistance: MGR (solid sea green line), possible MGR (MGRp) (long dashed gold line), QDR (solid gray line), and low level QDR (QDR-) (dashed orange line) Junghans et al., 2003;Kinloch, Sniezko, Barnes, & Greathouse, 1999). In trial two, the S 2 family from the 71560 parent also showed 100% survival, and thus parent 71560 is likely homozygous for Pla.
This parent was a forward selection from the S 1 family of parent SIS-42770 that had shown 92.5% survival in a previous trial (R. Sniezko, unpublished (Kovalchuk et al., 2013;Poland et al., 2009;St Clair, 2010).
In addition to MGR and QDR, some evidence suggests that there may also be resistance controlled by a recessive gene, notably, the failure to reject the hypothesis of a 1:3 segregation ratio. Recessive genes for resistance have been documented in several crop species (St Clair, 2010), and it has also been documented in at least one tree species, although the number of genes involved was not resolved (Newcombe & Ostry, 2001). Another more likely explanation is that the efficacy of QDR in some families could lead to survival levels that result in significant 1:3 segregation like that expected for a single recessive gene.  (3) although we have shown that both QDR and MGR appear to be present in C. lawsoniana, and potentially occur in combination, we still do not know how durable the resistance is over the scale of decades in the field. It is important to note that in this study we are not attempting to be definitive on patterns of inheritance (as described in Poland et al. (2009); that is still unknown in many systems with QDR. However, we are illustrating that of those C. lawsoniana parent trees identified with resistance we can expect a wide range of survival and time to mortality of progeny of parents with QDR as well as a low frequency of parents with MGR and relatively high survival. Future studies must untangle the many questions relating to the number of genes and their interactions leading to QDR.

| CON CLUS ION
The results we have reported here are encouraging. Although most of the trees in forests are highly susceptible, we show here that a large concerted applied program involving over 14,000 initial field selections has located hundreds of parent trees with resistance that can be incorporated into seed orchards and future breeding efforts. The large number of confirmed resistant selections will help maintain the genetic diversity needed in future forests. Ongoing breeding will use this information for forward selections to increase resistance in individuals that will be used in seed orchards to produce seed for reforestation and restoration. The resistance should also be valuable to re-establish horticultural use of C. lawsoniana.
Field trials are ongoing to assess the stability, durability, and usability of resistance in C. lawsoniana and to further develop the capacity to restore a species impacted by a non-native invasive pathogen.
Reforestation and restoration efforts with resistant seed has been underway for over a decade and the knowledge that there are several types of resistance provides us with cautious optimism that resistance will be durable.
The C. lawsoniana program to develop resistance to P. lateralis offers a successful and concrete example for those contemplating resistance programs in other species. Thousands of field selections were made initially, breeding zones established, several types of screening methodologies developed, tested, and utilized, field trials and seed orchards established (Hansen et al., 2012;Sniezko et al., 2012a;Sniezko et al., 2012b). The development of the program involved the participation of interdisciplinary teams of tree breeders, pathologists, foresters, and organizations, as well as the sustained support from several government agencies. The concerted effort, aided by some biological aspects of this species, has made this ongoing resistance program one of the fastest to develop resistance populations of trees. The work reported here on the types and magnitude of genetic resistance will be incorporated into the breeding program to help ensure that genetically diverse populations of C.
lawsoniana with resistance to P. lateralis are available for restoration and reforestation. Recent testing of some resistant C. lawsoniana with other lineages and sources of P. lateralis (Robin et al., 2014), and in the field (Sniezko et al., 2012a) provide some support for future durability of resistance, but as with all resistance programs, ongoing monitoring will be needed. and Genetic Resource programs and the BLM, without which these trials and the successful development of resistant seed currently being used in restoration and reforestation could not have been accomplished. We also thank the many members of the public and land managers interested in C. lawsoniana, whose interest in C. lawsoniana helped and continues to help spur this program. The USFS and BLM silviculturists and FHP staff have been instrumental in making tree selections, establishing field trials and planting resistant seedlings. We thank the reviewers for their suggestions in the earlier version of this paper.