Elevational clines are much steeper than the geo- 

 graphic. A comparison of these clines can be made by 

 noting that in figure 4, each geographic band, situated 

 between any two contours, is standardized for elevation. 

 Each band is equivalent to the amount of genetic 

 differentiation that occurs across 302 feet elevation for 

 3-year height, 315 feet for duration of elongation, 935 

 feet for leaf length, and 1.377 feet for fall freezing 

 injury, respectively. Thus, in a region approximately 200 

 miles from north to south, the geographic cHnes (fig. 4) 

 in those variables for which population differentiation 

 was pronounced (table 5) are equivalent to the genetic 

 differentiation that occurs within about 1,000 feet eleva- 

 tion at a single locality. 



Patterns of adaptive variation between populations 

 have direct implications in forest management. To limit 

 maladaptation in planted trees, seed transfer guidelines 

 must reflect adaptive variation. One estimate of an 

 appropriate limit to seed transfer involves the smallest 

 geographic or elevational interval across which differenti- 

 ation can be detected (Rehfeldt 1979). This interval is 

 estimated by the ratio /sd(0.2)/b, where b is the regres- 

 sion coefficient and lsd{0.2) is derived from the analysis 

 of variance as the least significant difference between 

 population means at the 80 percent level of probability. 

 (A rather low level of probability is used to avoid accept- 

 ing no differences when differences actually exist.) 



Elevational intervals associated with lsd(0.2) include, 

 for example. 604 feet for 3-year height, 738 feet for rate 

 of elongation. 1,788 feet for late growth, and 2,752 feet 

 for injury from fall freezing. These intervals suggest 

 that the maximum elevational limits for biologically 

 sound seed zones should not be much greater than 600 

 feet. This means that the transfer of seed from a single 

 source should be limited to ±300 feet. Nevertheless, con- 



siderable genetic differentiation occurs across seed zones 

 that occupy only 600 feet. These differences, for 

 instance, amount to 9 percent in 3-year height and 5 per- 

 cent in susceptibility to injury from the cold. Therefore, 

 transfers of seeds beyond these Umits can result in con- 

 siderable productive losses in artificial reforestation. 



Geographic patterns of variation (fig. 4) were scaled to 

 a value of V2 lsd{0.2), and therefore lateral transfers of 

 seed should be Limited to about ± 1 contour. This means 

 that two geographic zones are sufficient for the upper 

 Snake River Basin. 



The appropriate size of seed zones should be practical 

 operationally and economically. Thus, Umits to seed 

 transfer must accommodate biology, operational feasibil- 

 ity, and economics. When, however, administration 

 demands an alteration in the guidelines proposed above, 

 it should be remembered that each geographic band in 

 figure 4 is equivalent genetically to a relatively small 

 elevational interval within a band. Consequently, the 

 geographic limits should be compromised. From the bio- 

 logical viewpoint, seed zones for the upper Snake River 

 Basin should be considered as elevational zones with 

 geographic stratification. 



Regardless, populations of lodgepole pine from the 

 drainages of the upper Snake River display patterns of 

 genetic variation that typify the adaptation of the spe- 

 cies to heterogeneous environments in other geographic 

 locahties. Populations appear to be physiologically 

 attuned to specific environments, cUnes are steep, and 

 genetic differences occur across relatively small environ- 

 mental intervals. Differentiation between populations is 

 easy to detect experimentally and is closeh' related to 

 geography and elevation of the seed origin. Therefore, 

 seed transfer in artificial reforestation should be greatly 

 limited. 



8 



