PHENETIC AFFINITIES IN PINUS ATTENUATA LEMM. X P. RADIATA D. DON. HYBRIDS

Phenetic analysis is the use of numerous measurements of large numbers of morphological and physiological traits to arrive at a conclusion regarding taxonomic relationships and degrees of resemblance between living things. In my Master's thesis, I introduced the Affinity Index (AI), a technique for determining the relative phenetic affinities of hybrids for the parent species by means of phenetic analysis. The hypothesis of this study was that, based upon a finite array of characters, there is a difference between the phenetic affinities of hybrids, as measured by the observed AI, and the AI expected on the basis of actual percent ancestries. The basis for his hypothesis was the presumed influence of non-additive genetic effects and environmental effects upon the AI. The objective of this research was to quantify the deviation of observed AI from expected AI.; Ten specimens of Monterey pine (Pinus radiata D. Don.), ten specimens of knobcone pine (P. attenuata Lemm.), and 60 hybrids from controlled crosses were sampled in Eddy Arboretum, on the grounds of the Institute of Forest Genetics in Placerville, California, and at nearby Camino Arboretum. Data were taken on five needle, one growth, and nine cone characters. The Character State Distance (CSD), a measure of the reliability of a character for distinguishing between two species, was calculated for each character. Based upon the CSD values, an AI was calculated for each hybrid.; The difference between observed AI's and the expected AI's was too great to permit the AI to be used as an estimator of percent ancestry of hybrids, at least for he species studied. A power-of-the-test analysis indicated that a much greater number and diversity of characters, and a larger sample size within trees, would reduce this discrepancy sufficiently to allow the AI to be used for that purpose. This very large sample size is probably necessitated by non-additive genetic effects and environmental effects. If only additive genetic effects were involved, a small sample size would probably be sufficient.; The AI appears to have greater utility as a descriptive device. It was much more sensitive than, for example, Anderson's Hybrid Index to consistent phenetic tendencies on the part of certain types of crosses (F(,1) hybrids, back-crosses to P. radiata, etc.). With few exceptions, the hybrids morphologically resemble P. attenuata more than would be expected on the basis of percent ancestry. The results of a regression analysis raised the question of whether the tendency of the hybrids to resemble P. attenuata might be partially due to maternal effects.; A number of half-sib and full-sib families were included in the hybrids sampled. An analysis of variance indicated that percent ancestry has a much greater effect upon the deviation of observed AI from the expected AI than specific parentage. Inbreeding does not distort the AI. If anything, the AI is more accurate as an estimator of percent ancestry for those individuals.