Lesion formation and host response to infection by Armillaria ostoyae in the roots of western larch and Douglas‐fir

The process of lesion formation and host response to natural infection by Armillaria ostoyae were studied in the roots of western larch (Larix occidentalis) and Douglas‐fir (Pseudotsuga menziesii ssp. glauca) trees in the three age classes, 6–8, 18–19 and 85–95 years. The characteristics of lesions on infected roots were recorded and bark samples were dissected from infection points and lesion margins in the field and stored in liquid nitrogen for macroscopic study in the laboratory. Infection in the roots of 6‐ to 8‐year‐old trees advanced freely, overcoming any host resistance, quickly girdling the root collar and killing the trees. In 18‐ and 19‐year‐old trees, however, 43% of infections on western larch and 27% of the infections on Douglas‐fir roots were confined to lesions bounded by necrophylactic periderms with multiple bands of phellem. Host response was similar in 85‐ to 95‐year‐old trees, but the percentage of confined lesions was higher than in younger trees. The results suggest that larch shows resistance to A. ostoyae at a younger age and with greater frequency than Douglas‐fir.     

Fungal lesion length and expansion rate for the root pathogen Armillaria ostoyae in Douglas‐fir affecting root colonization and damage

Plant lesions affect disease impact, progression and host resistance. Root lesions caused by Armillaria ostoyae (Romagn.) Herink were inspected on 24‐ to 34‐year‐old planted Douglas‐fir (Pseudotsuga menziesii (Mirb.) Franco) pulled from the soil in five locations. Four hundred seventy‐nine lesions were measured for length (mm), expansion rate (mm/year) and lesion type. Lesion types were patch lesions on a root side, girdled lesions covering the root circumference and root collar lesions on the stem below ground. Non‐linear mixed modelling revealed that fungal lesion length at girdled and collar lesions increased but the expansion rate slowed as a power function of time, indicating increasing secondary inoculum and host resistance. Average fungal lesion expansion rate in girdled root and collar lesions was 127 mm the first year which then declined non‐linearly to 26 mm/year after 10 years. Fungal lesion length in patch lesions showed little change after the first year of infection. For girdled and collar lesions, fungal lesion length increased with tree size at the time of infection. Sites with a longer frost‐free period had shorter girdled root and stem lesions, probably because of more active host defence. Results indicate that mortality would be rare from one infection event. Furthermore, patch lesions pose a lower risk of mortality than other lesions because of a lower fungal lesion expansion rate and secondary inoculum. Tree size, lesion position, lesion type and fungal lesion expansion rate need to be considered when assessing the risk of lesions to individual tree survival. Techniques to minimize the disease impact are discussed.     

Port‐Orford‐cedar resistant to Phytophthora lateralis

Phytophthora lateralis, an exotic root pathogen, is the primary cause of Port‐Orford‐cedar (Chamaecyparis lawsoniana) mortality throughout its native range in southwest Oregon and northwest California. Most trees in the field are very susceptible, but genetic resistance to this pathogen has been demonstrated. Since the late 1980s, the USDA Forest Service and the Bureau of Land Management in cooperation with Oregon State University have conducted an intensive programme to identify and test resistant trees from the field, and propagate them in a seed orchard with the goal of providing resistant seedlings for regeneration. Susceptible families showed only 0–10% survival using a variety of inoculation techniques. Rooted cuttings of resistant parents are seldom killed, and seedling families of these parents exhibit 25–100% survival, depending on family and inoculation technique. Symptom development on resistant trees, including sunken lesions, and resinosis as well as reduced colonization and re‐isolation success, was consistent with a hypersensitive reaction. In a long‐term field test, five resistant families had 20–80% survival after 16 years, while three susceptible families had 0–8% survival in the same interval.     

Epidemiology of Armillaria root disease in Douglas‐fir plantations in the cedar‐hemlock zone of the southern interior of British Columbia

Results are presented from several studies on the epidemiology of Armillaria ostoyae in Douglas‐fir plantations in the interior cedar‐hemlock (ICH) biogeoclimatic zone of British Columbia. Two plantations were monitored for mortality by A. ostoyae and other agents for 35 years after establishment. In these and other plantations ranging in age from 7 to 32 years, one or more of the following factors were determined: source of inoculum, mode of spread and characteristics of lesions on roots of excavated trees; symptom expression in relation to tree age and damage to the root system and years from initial infection to death on trees killed by the fungus. Mortality from A. ostoyae began in both plantations about 5 years after planting, reaching 30% in one and 11% in the other after 35 years. The spatial pattern of mortality was similar to that reported from New Zealand, France and South Africa; however, the temporal pattern differed, beginning later and, instead of declining, continuing at a nearly constant rate to the present. To age 10, nearly all infections were initiated by rhizomorphs; as plantations aged, the proportion of infections occurring at root–root contacts increased. In seven plantations, in moist and wet subzones of the ICH, from 23 to 52% of Douglas‐firs had root lesions, with the higher incidences occurring on moist sites. The occurrence of aboveground symptoms, reduced leader growth and basal resinosis, was related to the percentage of root length colonized by A. ostoyae on trees with more than 30% of root length killed. Average time from infection to death increased from 1 to 2 years at age 6 to 22 years at age 33. The outlook for timber yield from Douglas‐fir plantations in the cedar‐hemlock zone is discussed. Management alternatives for reducing damage from A. ostoyae when regenerating sites are reviewed.     

Signs and symptoms associated with Heterobasidion annosum and Armillaria ostoyae infection in dead and dying mountain pine (Pinus mugo ssp. uncinata)

Symptoms and signs associated with root rot caused by Heterobasidion annosum or Armillaria ostoyae in mountain pines (Pinus mugo ssp. uncinata) were investigated in the Swiss Alps. A sample of dying or recently dead mountain pine trees (≥12 cm d.b.h.) and saplings (<1.3 m height) was assessed for root pathogen infection by taking root samples followed by isolations in the laboratory. From a subsample, an additional core was taken from the butt of each tree and evaluated in the same fashion. A total of 157 dying or recently dead mountain pine trees and 184 saplings with roots infected by either of the two pathogens or which lacked infection were analyzed using logistic regression models. The main objectives were to determine the most prominent symptoms induced by the fungi (resinosis), signs of the fungi (mycelia, fruiting bodies and rhizomorphs), and tree characteristics (d.b.h./height and evidence of wounds) that would allow an easy and reliable determination of H. annosum and/or A. ostoyae infection of mountain pines in the field. Heterobasidion annosum caused both root and butt rot on mountain pine, whereas A. ostoyae was mostly restricted to the root systems of the trees sampled. The most discriminating sign for the presence of A. ostoyae infection was the presence of characteristic mycelial fans, and for H. annosum root rot the presence of H. annosum mycelia (sheets of paper‐thin mycelium and mycelial pustules). In addition, resinosis was a powerful predictor for A. ostoyae in trees. Symptoms and signs indicating A. ostoyae or H. annosum infections were more reliable for saplings than for mature trees. Armillaria rhizomorphs were not useful in detecting A. ostoyae infection and, if present, were often formed by saprophytic Armillaria species. Heterobasidion annosum fruiting bodies were rarely observed and poorly reflected the widespread occurrence of this pathogen in the mountain pine forests.     

The relationship between Swiss needle cast symptom severity and level of Phaeocryptopus gaeumannii colonization in coastal Douglas‐fir (Pseudotsuga menziesii var. menziesii)*

This study examined 108 15‐year‐old Douglas‐fir (Pseudotsuga menziesii var. menziesii) trees to investigate whether trees exhibiting less severe Swiss needle cast (SNC) symptoms were more resistant (had less fungal colonization) or more tolerant (maintained healthy foliage under similar infection levels). Trees were sampled from six open pollinated families that were categorized into three disease severity groups (two families for each group; mild, moderate and severe disease symptoms). The amount of retained foliage and level of discoloration were visually assessed on trees in the field. Fungal colonization (as determined by proportion of stomata occluded with pseudothecia and by amount of Phaeocryptopus gaeumannii DNA in sampled needles) was measured on 1‐ and 2‐year‐old needles in the laboratory. Trees in the different disease severity groups were similar with respect to amount of fungus in their needles, yet the trees in the mild symptom group retained higher proportions of needles and maintained greener foliage. The relationship between amount of P. gaeumannii in needles and SNC symptom severity was statistically significant (p < 0.05) for amount of fungal DNA in 1‐year‐old needles and average needle retention (NR) over the last four growing seasons. Average NR decreased with increased amount of pathogen DNA in the mild disease symptom families. This relationship was reversed in the severe disease symptom group and there was no relationship in the moderate disease symptom group. Because the amount of P. gaeumannii DNA in foliage did not differ significantly among the groups, differences in symptom severity were attributed to tolerance, not resistance. Visual scoring of individual trees for average NR over the past four growing seasons could be used to effectively assess for SNC tolerance in Douglas‐fir.     

Tree vigour and the susceptibility of Douglas fir to Armillaria root disease

In this study the effects of thinning, fertilization and pruning on the vigour of Douglas fir (Pseudotsuga menziesii) and its susceptibility to Armillaria root disease were explored. Tree vigour was defined as the relative capacity for tree growth, expressed as the above-ground biomass increment per unit of photosynthetic tissue, or growth efficiency (GE). It has been hypothesized that trees with higher GE can better resist pathogen attack, and that GE can be used as a predictor of tree susceptibility to disease. In a previous study, four Douglas fir plantations were thinned, fertilized and pruned in all combinations, and the effects of these treatments on tree vigour were measured after 10 years. Root disease was not a factor in the initial study design, and mortality was ignored until 8 years after the treatments were applied. The results of an earlier study were utilized and the correlation between Armillaria root disease incidence and the effects of earlier stand treatments on tree growth was tested. Armillaria ostoyae was the primary mortality agent in the study area. Disease incidence on infested subplots ranged from 2 to 20%. Armillaria ostoyae incidence was the highest at medium tree density (6.1%), slightly lower on the low density (5.6%), and lowest on the unthinned plots (3.8%). There were no significant correlations between disease incidence and previous tree growth. The vigour of trees that became symptomatic or died by 1993 was not significantly different in 1983–85 from the vigour of trees that remained asymptomatic. On these sites, in areas of infestation, A. ostoyae is killing the largest, fastest growing trees, as well as less vigorous trees. Armillaria continues to cause mortality, regardless of the growth efficiency or growth rate of the host.     

Temporal evolution of collar lesions associated with ash dieback and the occurrence of Armillaria in Belgian forests

Ash dieback, caused by the fungus Hymenoscyphus fraxineus, has been observed in Europe for several years. In Belgium, the disease was first reported in 2010. Besides crown defoliation and dieback, collar lesions have sometimes been reported. To evaluate the prevalence and the progression of collar lesions and crown defoliation in ash dieback‐affected stands of various ages, a survey was conducted in 2013 and 2014 on 268 ash trees (Fraxinus excelsior) originating from 17 Walloon forest stands. The results showed that the proportion of trees with collar lesions greatly increased between June 2013 and September 2014 and that there appeared to be no significant link between a tree's diameter‐at‐breast height (DBH) and collar lesion occurrence. The mean percentage of defoliation increased in each forest stand across time, with observations conducted in September 2013 and 2014 showing a positive correlation with the mean percentage of trees with collar lesions. Molecular tests were carried out on 103 additional trees originating from 12 of the 17 stands to evaluate the occurrence of H. fraxineus and Armillaria spp. at the collar level. Most of the trees (98%) were infected by H. fraxineus. In contrast, only 41% of the samples were infected with Armillaria spp., most commonly A. gallica and A. cepistipes. This study discusses the role of Armillaria spp. and the rapid increase in the number of trees with collar lesions within the context of the evolution of ash dieback in Europe.     

Tolerance of Pinus patula full-sib families to Fusarium circinatum in a greenhouse study

The pitch canker fungus, Fusarium circinatum, has caused large-scale mortality of young Pinus patula Schiede and Deppe ex Schltdl. and Cham. seedlings in nurseries in South Africa since 1990. Diseased seedlings have been inadvertently carried to the field, which in turn have died and has reduced stocking below an acceptable level. Tree breeders have suggested that the only long-term solution to limit infection by this pathogen is to identify and deploy tolerant P. patula families. A commonly used technique to identify tolerant clones is to artificially inoculate open-pollinated progeny from orchard clones with F. circinatum under greenhouse conditions. In these trials, large variation in tolerance to the pathogen among seedlings within open-pollinated families was observed and this could be influenced by the pollen parent. Therefore, identifying individual full-sib families, where both parents are known, should improve the identification of tolerant families, which can then be repeated. In this study, cuttings from control-pollinated P. patula seedling hedges were inoculated with F. circinatum in a greenhouse. The results showed large family variation where some of the full-sib families were similar in tolerance to P. elliottii Engelm. var. elliottii seedlings. Therefore, it is recommended that breeders focus on identifying specific family combinations that are more tolerant to F. circinatum.     

Comparison of the virulence of Armillaria cepistipes and Armillaria ostoyae on four Norway spruce provenances

The basidiomycetes Armillaria cepistipes and Armillaria ostoyae frequently occur in the same forest stand. In this study, we determined the virulence of 20 isolates of A. cepistipes and 16 isolates of A. ostoyae on four different provenances of 2‐year‐old Norway spruce (Picea abies). Within 30 months after inoculation, 1.1 and 19.1% of the seedlings inoculated with A. cepistipes and A. ostoyae, respectively, had died or were dying. The incidence of dead and dying seedlings varied between 3 and 49% among the A. ostoyae isolates. The virulence of an isolate was positively correlated to its ability to produce rhizomorphs. One Norway spruce provenance showed significantly lower susceptibility to A. ostoyae than the other three. Rhizomorphs of both Armillaria species were attached to the root surface. The attached rhizomorphs of A. ostoyae, however, were associated with significantly more lesions. The virulence of the isolates was not correlated with their wood‐degrading capability for either of the Armillaria species.     

Root and aerial infections of Chamaecyparis lawsoniana by Phytophthora lateralis: a new threat for European countries

Phytophthora lateralis has been isolated from root and collar lesions in Port‐Orford Cedar (POC) trees (Chamaecyparis lawsoniana) in north‐western France (Brittany). These trees, planted in hedgerows, displayed symptoms similar to the typical symptoms of POC root disease. Until now, the disease has been found outside of the nurseries only in western North America. Aerial symptoms, not associated with root or collar infections, were also observed, and P. lateralis was isolated from branch lesions. Similar symptoms were previously observed only in one POC root disease site, located in the Pacific coast of Oregon where climatic conditions are similar to those occurring in Brittany. The reported aetiology as well as the morphological characteristics (deciduous sporangia) of P. lateralis suggests that this species could be air‐dispersed, as described for P. ramorum, a closely related species. This outbreak of P. lateralis in Brittany in farming landscapes associated with the aerial spread of this pathogen represents a new threat for European countries.     

Relationship of site and stand characteristics to Armillaria root disease incidence on ponderosa pine in the Black Hills,South Dakota

Ponderosa pine (Pinus ponderosa) in the Black Hills National Forest, SD, USA, was surveyed for Armillaria root disease (ARD). The root pathogen Armillaria ostoyae occurred on ponderosa pine seedlings, saplings, pole‐size trees and large‐diameter trees. The mean incidence of aboveground disease symptoms by stem count was low (0.2%), but in certain areas, the incidence was higher, affecting the regeneration success and tree longevity. Symptomatic ponderosa pine were in areas characterized by having higher elevation, greater annual precipitation, more seedlings, bigger large‐diameter trees and greater odds of past harvesting activity than in areas without root disease. Stump density was positively spatially correlated with root disease incidence. No particular soil type was related to disease occurrence; though, in areas with symptomatic trees, soil available water holding capacity (AWC) was greater and soil permeability was less where root disease was present. Spatial analysis confirmed the relationships found in linear correlations, with soil AWC and stump density positively and soil permeability negatively correlated with per cent infected stems ha^?1 and basal area infected.     

Traumatic resin canals as markers of infection events in Douglas‐fir roots infected with Armillaria root disease

In response to an infection, traumatic resin canals (TRCs) are formed in the roots of many conifers, which may be used to determine the timing and sequence of infection events essential for epidemiological studies of root diseases. Juvenile Douglas‐fir (Pseudotsuga menziesii) tree roots at coastal and interior sites in British Columbia were wounded at various times of the year or were inoculated with an isolate of Armillaria ostoyae, and root sections were taken to determine the timing and extent of TRC formation. Naturally infected Douglas‐fir were also examined to determine the extent of the TRCs in infected and uninfected roots on infected trees and in the lower stem. Wounds made in March and October had poor or no TRC formation while the summer wounds responded strongly and were associated with resin soaking. Roots wounded in October did not respond until the following year in all trees except one. Trees produced TRCs and resin soaked tracheids at all times in response to the fungal inoculations. The most striking difference between wounding or fungal inoculation was the multiple bands of TRCs produced in response to the fungus. TRCs at natural A. ostoyae infections were found 92% of time in roots at the stem junction and 74% of the time in the stem at soil line. TRCs were produced in uninfected roots on infected trees but disappeared with increasing distance from the initiating lesion. TRCs can be used to time yearly and seasonal root infections when they can be traced from an identified lesion.     

Effects of wounding and fungal infection with Armillaria ostoyae in three conifer species. II. Host response to the pathogen

Structural responses in the bark and wood were described following penetration by Armillaria ostoyae in the roots of 20‐ to 30‐year‐old Douglas‐fir, western hemlock and western redcedar trees. Tissue necrosis presumably caused by fungal exudates was commonly observed at inoculum contact. In Douglas‐fir and western hemlock, A. ostoyae interfered with the initiation of active defence mechanisms involving the development of a lignified zone of impervious tissue (IT), necrophylactic periderm (NP) formation and compartmentalization of infected woody tissue. Breaching of IT and NP barriers was frequent, particularly around the clusters of sclereid cells in western hemlock. In western redcedar, the IT zone was inconspicuous. Induced rhytidome formation occurred in western redcedar either simultaneously with or after completion of NP development. The formation of this tissue facilitated en masse sloughing of infected tissue from the surface of roots. In western redcedar, traumatic phloem resin ducts formed in tangential bands surrounding the margin of expanded lesions. Effective compartmentalization in western redcedar was achieved by a barrier zone comprised of a higher‐than‐average number of axial parenchyma that accumulated polyphenolic deposits. A combination of host‐mediated defence mechanisms in western redcedar resulted in a significantly higher frequency of effective resistance reactions than in western hemlock or Douglas‐fir.     

Penetration of Picea sitchensis root bark by Armillaria mellea,Armillaria ostoyae and Heterobasidion annosum

Penetration of root bark tissues of Picea sitchensis by Armillaria ostoyae, Armillaria mellea and Heterobasidion annosum was examined in the absence of wounds, in superficial wounds (rhytidome tissues removed to expose the secondary phloem) and in wounds to the depth of the vascular cambium (deep wounding). Both species of Armillaria penetrated bark without prior wounding, but neither species formed rhizomorphs in this treatment. Armillaria ostoyae penetrated to 39 cell layers in depth by 48 days after inoculation of unwounded bark, whereas A. mellea penetrated 25 cell layers in the same time. Armillaria mellea penetrated superficial wounds significantly more rapidly than did A. ostoyae. Both species produced rhizomorphs within wounded host tissues. Inoculation of deep wounds with Armillaria resulted in a greater depth of bark necrosis with A. mellea than with A. ostoyae. In the absence of wounding, H. annosum failed to penetrate root bark tissues, but in both superficial and deep wounds hyphae penetrated beyond the ligno–suberized boundary zone (LSZ) by 12 days after inoculation. Where no inoculations were made, superficial or deep wounding led within 25 days to the restoration of a structurally continuous LSZ, and by day 48 the wound periderm (WP) was fully differentiated. In inoculated wounds, however, formation of the LSZ and WP was delayed or inhibited in most trees, particularly following inoculation with A. ostoyae or A. mellea. Suberization in the LSZ and WP remained diffuse and discontinuous 48 days after inoculation. Moreover, the presence of WP did not prevent further penetration of the tissues by the pathogens. Variations between trees in the depth of pathogen penetration were noted, possibly indicating differing susceptibilities of individual host genotypes. The possible host factors involved in resistance to penetration of root bark tissues by Armillaria and Heterobasidion are discussed.     

Screening resistance of Dalbergia sissoo clones against Ganoderma lucidum root rot disease in field conditions

Ganoderma lucidum is a serious root rot pathogen of trees and is known to cause large scale mortality in Dalbergia sissoo. Clones prepared from selected plus trees of D. sissoo have been established in a clonal seed orchard and a seed production area in northern India where 5‐year‐old plants of various clones were attacked by G. lucidum to varying degrees. Regular assessment was carried out for further 5 years to screen disease progress in the field and to identify clones of greater and lesser resistance. Data for height, clear bole and girth were collected to select trees with better growth form and disease resistance. Five highly resistant and five susceptible clones were categorized on the basis of total score for growth form characters and disease development.     

Variation in resistance of loblolly pine (Pinus taeda L.) families against Leptographium and Grosmannia root fungi

Pine decline poses a serious threat to forest sustainability in the south‐eastern United States. Complex interactions between biotic and abiotic factors are involved in the decline and include root‐feeding bark beetles and their associated fungal genera, Leptographium and Grosmannia. A study was conducted to determine the relative tolerance of loblolly pine (Pinus taeda L.) families when challenged with either Leptographium or Grosmannia species. In the first study, bare root seedlings from 23 loblolly pine families were screened with L. procerum, L. terebrantis, G. huntii and G. alacris using an artificial inoculation method. In a second study, containerized seedlings from 27 loblolly pine families were screened with G. huntii and L. terebrantis. Measured seedling responses to the inoculations included lesion length, lesion width and occlusion of vascular tissues, measured 8 weeks after inoculations. The most common host response was dark brown lesions and resinous occluded stem tissue. Seedling families had a wide range of host responses to the different Leptographium and Grosmannia species and showed that it could be possible to select the families that may be tolerant to Leptographium and Grosmannia fungal species based on these results.     

Parasitic association of the mycetophagous wood nematode,Bursaphelenchus fraudulentus with the honey fungus Armillaria ostoyae

Honey fungi, Armillaria spp. are common and economically important pathogens of a wide range of tree species grown both in the forest and orchard cultures worldwide. Our field research in 73 forest stands across Poland has shown a common association of the mycetophagous wood nematode, Bursaphelenchus fraudulentus with Armillaria‐infected trees. The data reported here provide the first insight into the parasitic interaction of a nematode and Armillaria ostoyae. In laboratory conditions, B. fraudulentus reproduced on A. ostoyae, caused substantial damage to its mycelium and, finally, killed the whole colony within a short time. In ageing cultures, the nematode produced resting (dauer) juveniles. After artificial inoculation to Armillaria‐infected trees, the nematode population persisted under the bark, continued feeding and mass reproduction on the mycelium, and dispersed over the mycelial fans. The ability of B. fraudulentus to develop and reproduce on the surface and inside the fungal rhizomorphs suggests that these structures could facilitate the nematode dispersion in the environment. Since B. fraudulentus is not pathogenic to the host tree, its parasitic association with A. ostoyae may contribute to natural regulation of this fungus in the environment. The observed characteristics of this nematode suggest that it could potentially be used as a biocontrol agent of honey fungi in forest and orchard trees.     

Pathogenicity and virulence of Armillaria sinapina and host response to infection in Douglas‐fir,western hemlock and western redcedar in the southern Interior of British Columbia

The frequency of infection, lesion characteristics and anatomical changes in phloem and cambial tissues caused by Armillaria sinapina were studied on inoculated trees of Douglas‐fir, western hemlock and western redcedar and compared with results of A. ostoyae inoculations previously reported on the same host species. Similar percentages of inoculations resulting in infection of roots on the three hosts indicate that A. sinapina and A. ostoyae are equally pathogenic. Armillaria ostoyae was more virulent than A. sinapina as demonstrated by fungal exudates from A. ostoyae inoculum blocks, which appeared to cause lesions on roots; the higher frequency at which lignified impervious tissue (IT) and necrophylactic periderm (NP) developed in bark and following cambial invasion, compartmentalization; the large proportion of roots that showed no visible host response; the large zones of IT formed under continuous stimulation by A. ostoyae advancing in inner bark; and the high frequency of breaching of NP barriers. Spread of A. sinapina mycelium in host species appeared slower than that of A. ostoyae, particularly in Douglas‐fir and western hemlock. In western redcedar, A. ostoyae induced stronger host responses than those following invasion by A. sinapina, which included further expansion of the induced rhytidome response, traumatic phloem resin duct formation and higher numbers of polyphenolic parenchyma comprising its barrier zone. Where damage by A. sinapina ensued, it was always associated with high inoculum potential. The ecology of virulent and less virulent species of Armillaria in natural forests is discussed.     

Variography and kriging allow screening Pinus pinaster resistant to Armillaria ostoyae in field conditions

Armillaria ostoyae is an important disease of Pinus pinasterin north-west Spain, which kills trees following a heterogeneousspatial structure. In a progeny trial of P. pinaster seedlings,spatial heterogeneity and autocorrelation of neighbour mortalitycaused by A. ostoyae impeded proper analysis of the diseaseincidence. We used variography and kriging methods to describethe spatial distribution of the infection probability and thegenetic variation of the resistance to A. ostoyae among families.The spatial structure of disease incidence was modelled, andthe probability of survival was corrected by kriging at eachtree location. Cumulative mortality 3 years after planting was65.1 per cent. Significant differences among P. pinaster familiesin terms of mortality to A. ostoyae were found, with low individual(h = 0.08) and moderate family (h = 0.35) heritability estimates. According to a theoretical semivariogram, the patch size ofthe disease incidence was 63 m wide. This is the first timevariography and kriging are used to select P. pinaster resistantto Armillaria root rot. It is concluded that geostatistics providesforest pathologists with a powerful tool for screening resistanttrees in field conditions.