GAS-EXCHANGE OF 20-YEAR-OLD BLACK SPRUCE FAMILIES DISPLAYING A GENOTYPE X ENVIRONMENT INTERACTION IN GROWTH-RATE

Gas exchange and xylem water potential were measured on 20-year-old black spruce (Picea mariana (Mill.) B.S.P.) trees from four ful-sib families from a 7 X 7 diallel. These families constitute a 2 parent X 2 parent factorial breeding structure. One female parent (59) produced families that have displayed high productivity on three plantation sites, whereas the other female parent (63) produced families that had high growth rates on two of the sites but not on the third, least productive, site. In 1991 (a hot, dry year) measurements were obtained from trees at two sites; in 1992 (a cool, wet year) measurements were obtained from only one site. Gas exchange was measured on detached branches cut from the upper third of the crown; measurements were initiated within 2 min from cutting. In 1991, daytime xylem water potential, predawn xylem water potential, net photosynthesis (P-n), and needle conductance (g(wv)) were all lower in trees growing on the least productive site (2) than in trees growing on the most productive site (3), indicating that trees on site 2 were under greater water stress than those on site 3. Progeny of female 59 displayed higher P-n than progeny of female 63 over both years of contrasting rainfall and on both the wetter and drier sites during the first year. However, progeny of the two female parents did not differ with respect to g(wv), in either year nor on either site during the first year. The genetic variation in P-n/g(wv), observed over sites and years, is positively related to growth rate only on the dry site. Thus, the genotype X environment interaction in growth appears to be due to relatively stable genetic differences in P-n. We hypothesize that water stress on the dry site reduces P-n below a threshold value at which point genetic variation in P-n impacts on productivity and results in genetic variation in growth rate. Furthermore, we hypothesize that on the wet site carbon gain per unit leaf area is not as limiting to growth rate; thus, the observed genetic differences in P-n do not influence genetic rankings in growth rate.