GENETIC-VARIATION AND PRODUCTIVITY OF POPULUS-TRICHOCARPA AND ITS HYBRIDS - X - TRAIT CORRELATIONS IN YOUNG BLACK COTTONWOOD FROM 4 RIVER VALLEYS IN WASHINGTON

A common-garden study of Populus trichocarpa Torr. & Gray was establis hed in spring 1986 with 128 clones collected from sites along two mesi c (Hoh and Nisqually) and two xeric (Dungeness and Yakima) river valle ys in Washington. Two replicate plantation, one in Puyallup and the ot her in Wenatchee, Wash., were established with this material. Over 2 y ears data were taken on stem growth, leaf/crown characters, spring/aut umn phenology, and the incidence of Melampsora occidentalis leaf rust. Combining clones from all four sources, correlation/regression analys es were used to examine clonal stability of traits between lest sites and trait relationships with stem growth; broad-sense heritabilities ( H-2) and genetic correlations revealed the genetic strength of these t raits. At Puyallup, many leaf/crown traits predicted stem growth moder ately to very well (r(2)>0.50), e.g., total leaf area (TLA) to diamete r gave an r(2) of 0.91 and current-terminal leaf size, of 0.79. Some r egressions were quadratic, suggesting a threshold level in a trait (e. g., leaf size) beyond which stem growth levels off. Upper-crown TLA wa s more closely related to height than TLA of the lower sylleptics, but the reverse was true for diameter. A decline in r(2) values from uppe r to lower crown positions was sharper for correlations of TLAs with h eight than with diameter. Thus, leaf area al location seems to differe ntially affect stem growth. When autumn leaf fall (LF) and rust incide nce (R) were regressed with growth, r(2) values ranged from 0.58 to 0. 71, but those of spring flush (SF)were only 0.10 to 0.12. Early LF and high R, both negatively affecting growth, had a strong geographic com ponent as it occurred mainly on lower-elevation Yakima clones. At Wena tchee, field conditions were harsher and microsites more variable, so trait/growth relationships were weaker. Genetic correlations with grow th revealed similar trends as phenotypic analyses. Unlike leaf/crown t raits, clonal scores of LF, SE and R were fairly stable across the two test sites (r(2): 0.58-0.80). These traits also showed strong genetic control (H-2: 0.96-0.98). The trait/growth relationships as well as t rait stability within clones have implications for selecting clonal st ock in poplar culture and conservation.