DISCUSSION 

 Geographic Distribution 



Habitat type identification to characterize plant com- 

 munities is becoming a standard tool of forest manage- 

 ment in the Inland West. The techniques of vegetation 

 analysis are well documented (Cooper and others in press; 

 Pfister and others 1977; Steele and others 1981). In this 

 study, we applied these techniques to the analysis of 

 vegetation on plots. In a few cases, where vegeta- 



tion indicated a habitat undefined for that region, we 

 applied habitat type definitions from adjacent forests, pro- 

 viding that other evidence also supported a correspon- 

 dence of habitat types. 



The dichotomy observed in this study between presence 

 and absence of Armillaria agrees in general with findings 

 in Australia (Kile 1980, 1981) where Austrahan Armillaria 

 spp. also did not occur on very dry sites. In this study, the 

 warmest and driest habitat types of the PSME series, as 

 well as coldest and driest habitat types of the ABLA 

 series and the "frost pocket" sites exemplified by the 

 ABLA/VACA and PSME/VACA habitat types, did not 

 seem suitable for Armillaria. The absence of Armillaria 

 from the two cold-wet sites (ABLA/CACA and ABLA/ 

 STAM) indicates that both excessive cold and excessive 

 moisture might limit growth and development of 

 Armillaria. 



The presence-absence dichotomy of Armillaria spp. 

 occurrence prevailed when we examined moist auxiliary 

 plots adjacent to dry or cold random plots (see fig. 1 for 

 location of auxiliary plots). Also, the occurrence of Armil- 

 laria on all of the plots randomly located on the Coeur 

 d'Alene, Clearwater, and Nez Perce National Forests, an 

 area of approximately 5 million acres, reinforces these 

 ideas. The five instances of discontinuous distribution (one 

 plot in each of ABLA/MEFE, ABLA/CLUN, ABGR/LIBO, 

 ABGR/CLUN, and TSHE/CLUN), where continuous 

 distribution was expected, could be attributed to misiden- 

 tified habitat type, a transitional state between suitable 

 and unsuitable habitat types, or stand history (possibly 

 wildfire) that prevented establishment of the fungus or 

 recently eradicated it. The ABLA/MEFE plot was prob- 

 ably in transition to the ABLA/XETE habitat type, and 

 the ABGR/LIBO could be transitional to ABGR/CARU. In- 

 fluential factors of stand history are a particularly attrac- 

 tive explanation of the absence of the fungus from the 

 ABLA/CLUN and ABGR/CLUN plots because the two fir 

 series plots were located about 1,600 m apart in the same 

 drainage of western Montana. We have no ready explana- 

 tion for the absence of Armillaria from one TSHE/CLUN 

 plot. 



The evidence from this study concerning Armillaria 

 spp. distribution can be summarized in the following hy- 

 pothesis: Any 0.04-ha plot of forest land in the TSHE and 

 THPL series located in the Northern Rocky Mountains 

 will yield Armillaria. All the ABGR series in this region, 

 except the SPBE and CARU habitat types, will also yield 

 Armillaria. Greater discontinuity exists within the ABLA 

 and PSME series. Only two PSME habitat types support 

 Armillaria; these are VAGL and PHMA. In fact, the 

 limiting conditions may be at the phase level within these 



habitat types. Four ABLA series habitat types support 

 Armillaria. These are ALSI {Alnus sinuata), VAGL {Vac- 

 cinium globulare), MEFE {Menziesia ferruginea), and 

 CLUN (Clintonia uniflora); however, they may show 

 discontinuities at the phase level. We further hypothesize 

 that Armillaria is absent from both the ponderosa and 

 limber pine {Pinus flexilis) series. Armillaria spp. will 

 behave in the mountain hemlock (Tsuga mertensiana) 

 series as they do in the ABLA series. We have no infor- 

 mation about the Engelmann spruce series. 



Epiphytic Rhizomorphs 



Our frequency of encountering epiphytic rhizomorphs is 

 similar to that of Kile (1980, 1981), who encountered 

 epiphytic rhizomorphs on 85 percent of the dead eucalypts 

 (stumps, fire killed, and so forth) inspected. We found 

 rhizomorphs on 86 percent of the dead conifers inspected 

 on habitat types that supported the fungus. He found 20 

 percent of 202 healthy eucalypts supporting epiphytic 

 rhizomorphs. We found 43 percent of the healthy conifers 

 and 80 percent of the healthy hardwoods (or dead portions 

 of healthy clumps) supporting epiphytic rhizomorphs. 

 These values are based on a large sample and therefore 

 provide good estimates of the probability of encountering 

 epiphytic rhizomorphs on these materials in the Inland 

 West in habitat types suitable for Armillaria. 



CONCLUSIONS 



Results presented in this paper indicate that risk to Ar- 

 millaria spp. in the Northern Rocky Mountains can be 

 predicted by habitat type. The cold-dry and hot-dry sites 

 seemingly are outside the ecological range of all forms of 

 Armillaria native to the Northern Rocky Mountains. 

 These hypothesized patterns of occurrence must be 

 validated before being put to general use for predicting 

 risk of Armillaria root rot. This technique of risk predic- 

 tion has potential application throughout the Western 

 United States, both for Armillaria and for other endemic 

 diseases of forest trees. This approach could be particular- 

 ly effective for predicting the range and presence of root 

 pathogens of woody plants. The occurrence and function 

 of these fungi are likely tied to long-term soil and climatic 

 conditions, similar to the occurrence of indicator plants. 



Perfecting a risk-prediction system will entail additional 

 research. We must do more fieldwork to learn the geo- 

 graphic and host ranges of Armillaria spp. clones. This 

 will enhance our knowledge of ecological genetics of the 

 Armillaria-conifer interaction. We must also taxonomical- 

 ly classify Armillaria clones. Effective progeny testing 

 awaits development of techniques to measure host vigor 

 and susceptibility, along with fungus pathogenicity. 

 Regardless of the ultimate findings of such research, the 

 findings reported here clearly indicate that forest man- 

 agers will need to know habitat type of stands in the 

 Northern Rockies if they are to account for potential 

 Armillaria damage in their management plans. 



5 



