Table 4. --Results of multiple regression analyses for relating population performance 

 to physiognomic criteria of the seed source 



Variable 



Regression coefficient 



(b) 



(0.2) 



Elevation (X,) -0.026 466 ra (1500 ft) 



Latitude {X2) 8.791 1.4° (100 miles) 



Longitude (X3) 5.134 2.3° (120 miles) 



bo = 268.84 



r2 = 0.51 



s = 10.99 

 yx 



Statistical analyses have detected differences among populations in 16-year height and 

 have related differentiation to geographic and ecologic conditions of the seed source. The 

 applicability of these results to practices of seed zoning depends on the degree by which tree 

 height reflects adaptations to the natural environment. While there is little doubt that seed 

 zones should be based on numerous adaptational characteristics, no data other than height are 

 available for ponderosa pine in southern Idaho. But, adaptations in ponderosa pine may be 

 manifest much like adaptations in Douglas-fir of north Idaho: the growth, phenology, and cold 

 hardiness of populations are so highly intercorrelated that variation among populations in a 

 single character reflects numerous characters (Rehfeldt 1979) . Consequently, it is assumed 

 that observed patterns of variation in 16-year height of ponderosa pine are adaptive. 



Table 4 reveals the distance in elevation, latitude, and longitude that is associated 

 with mean differences among populations equal to a value of isd(o.2)- At minimum distances of 

 1,533 ft (466 m) elevation, 100 miles (1,4 degrees) latitude, or 120 miles (1.9 degrees) 

 longitude, differences among populations are detectable with a probability of 80 percent. Or 

 as a corollary, seed for afforestation may be transferred from the collection area: +_ 750 ft 

 (230 m) elevation, + 50 miles (0.7 degrees) latitude and -t^ 60 miles (1.2 degrees) longitude. 



Yet, it should be recognized that the recommended seed zones are based solely on the 

 probability of detecting differences in height among populations. Not only do these zones 

 fail to account for traits other than height, they also fail to account for administrative 

 factors such as patterns of land ownership or location of future planting programs. However, 

 these guidelines can be altered readily provided that an increased risk of maladaptation is 

 recognized. A difference in height of about 10 percent can be expected from populations 

 separated by 1 000 m (table 4) ; populations separated by 1 degree of latitude or longitude can 

 be expected to differ by 3 percent and 2 percent, respectively. 



Genetic gains from tree improvement are readily attained by collecting seeds for refores- 

 tation from natural stands of proven genetic value. For instance, if the present region is 

 divided into four zones separated at 1 400 m elevation and 44.3 degrees latitude, a gain in 

 height of about 4 percent could be expected within zones when seed collections are made from 

 the following populations: 



Zone Superior populations^ 



Low north 27,38,53,21A 

 High north 14**, 28 



Low south 42,47,21B 

 High south 14**, 19 



*Coded in table 3. 



**Suitable for both northern and southern zones. 



5 



