Table 4 — Simple correlations among population means. Underlined coefficients failed to reach statistical significance at the 5 percent 

 level. 



Variable 



Code > 



H3-X 



AH-M 



AH-X 



LL 



RED 



DD 



EL 



IN 



CES 



DUR 



RATE 



Growth and development 



























f^-vpar mp^ir hpinht 



H3-M 



0.74 



0.56 



0.31 



0.30 



0.84 



0.02 



0.76 



-0.14 



0.68 



0.68 



0.74 



r^-vpar xprir hpioht 



H3-X 





.28 



.69 



.35 



.25 



-.18 



.64 



- .14 



.48 



.50 



.64 



Adjusted mesic height 



AH-M 







.19 



-.02 



.57 



m 



.31 



-.02 



.30 



.29 



.29 



Adjusted xeric height 



AH-X 









.22 



-.11 



-.29 



.24 



-.01 



JZ 



/t6 



.24 



Leaf length 



LL 











.15 



-.18 



.21 



-.18 



.08 



.12 



.23 





RFD 

 riCIL/ 













18 





- Ofi 









Drought mortality 



UU 















— .V / 



.U f 



.UO 



.U 1 



— .UO 



Periodicity of shoot elongation 



























Elongation 



EL 

















-.37 



.82 



.87 



.98 



Initiation 



IN 



















-.06 



-.29 



-.37 



Cessation 



CES 





















.97 



.72 



Duration 



DUR 























.78 



Rate RATE 



be limited by drought or frost or both. The species in- 

 habits environments that differ by over 60 days in frost- 

 free period and by 60 cm in precipitation (fig. 5). That 

 populations have become differentially adapted to en- 

 vironmental differences of this magnitude is an expected 

 result of natural selection. 



Adaptive differentiation can be expressed by significant 

 interactions involving populations and environments. How- 

 ever, the interactions that were evident for 3-year height 

 and adjusted height arose for different reasons. On the 

 one hand, the interaction for 3-year height represents a 

 scale effect rather than an adaptive response. As evi- 

 denced by strong simple correlations (table 4), populations 



FROST-FREE PERIOD (days) 



retained similar rankings in each environment, but the ab- 

 solute difference in height of any two populations was 

 greater under mesic than under xeric conditions. 



On the other hand, the interaction for adjusted height 

 reflects differential adaptation. Under mesic culture, 

 populations of greatest growth potential would still have 

 been tallest at age 3 even if all trees had been the same 

 height at age 2. Thus, the superiority of these populations 

 has increased between ages 2 and 3. But under xeric con- 

 ditions, populations from dry eastern areas grew the most 

 from a common height at age 2. This suggests that if 

 xeric culture were to be maintained, these eastern popula- 

 tions would eventually become tallest on this particular site. 



PRECIPITATION (cm) 



Figure 5— Regional environmental patterns of variation (enlarged from U.S. Department 

 of Commerce 1968). 



7 



