456 BOHUN B. KINLOCH, JR. 



This type of approach seems well adapted to studying rust diseases 

 of pines, especially the autoecious peridermia. Isogenic (or nearly so) 

 aecial clones could be obtained in quantity from single spore inoculations 

 (or perhaps even single gall isolates), tested on different host clones to 

 determine reaction spectra, and then on selfed and Fj progenies to 

 determine modes of inheritance, in the manner just described. 



With heteroecious species of Cvonavtivm , the technical problems are 

 more difficult, especially in securing genetically homogeneous inoculum. 

 Since the basidiospore stage that infects pine is a product of meiosis, 

 each spore is potentially a different genotype. Relative uniformity could 

 be obtained by establishing clones derived from single aeciospore or 

 urediospore inoculations on alternate hosts. Complete homozygosity in 

 different isolates would be assured in both haplophase and dicaryophase 

 if the fungus in question were homothallic, as Hirt (1964) has suggested 

 for C. vibicola. 



In the context of this discussion, I would now like to return to 

 experience of my own with fusiform rust, described earlier in this paper 

 and in Kinloch and Stonecypher (1969) . This was a population study of 

 open- and control led-pollinated offspring derived from a large number of 

 randomly selected parent trees and planted on different sites. Early 

 plantings sustained negligible rust infection for several years. But 

 after an epidemic in 1964, infection was widespread in all plantations, 

 though the intensity varied greatly among sites and families. I have 

 already mentioned the high heritabilities of resistance obtained. While 

 characterizing a population by quantitative parameters is valid and 

 helpful for utilitarian purposes, it is not likely to lead to an under- 

 standing of the biological mechanisms involved. Another approach to the 

 same raw data is possible, which may provide a deeper insight to the 

 genetic structure of the host-pathogen population and point to alterna- 

 tive routes of investigation. 



When individual trees on a given site, irrespective of family 

 relationship, were grouped into successive categories according to the 

 number of separate infections (galls) they had, their frequency distribu- 

 tion appeared as in Fig. 1, A and B. Similar data were obtained in 

 another sample of trees from bulk seed lots of unknown parentage (Fig. 1C) 

 The latter were in a different experiment at a distant site, but were 

 approximately the same age and infected in the same year as the former. 

 The close resemblance of these sample distributions suggests that they 

 characterize the response of a young population following exposure to 

 relatively heavy infection. Individual families, on the other hand, had 

 distributions peculiar to each. Fig. 2 illustrates the frequency 

 distributions of individual trees within the same full-sib families 

 representative of high, intermediate, and low susceptibility on each of 

 three sites that differed greatly in the amount of over-all infection. 



It is clear from the overall distribution of the sample populations 

 (Fig. 1), as well as the differences in the distributions among full-sib 

 families (Fig. 2) , that a wealth of genetic resistance exists in the 

 population. It is also apparent that different degrees of resistance 

 exist, and that most individual trees have some--even in relatively 

 susceptible families. In these tests (and commonly in natural stands), 

 individual trees with more than 10 and up to 50 galls were often 

 adjacent to those with few or none. Most families included individuals 

 with more than four times the number of galls as the mean for that family 

 (though individual extremes in resistant families always had fewer galls 



