Table 5— Quantification of the percentage differences expected between 

 populations located along elevatlonal and geographic dines 



Elevatlonal 



Variable 



Difference 

 across 1,000 m 

 elevation 



Difference 

 across 7° 

 latitude 



Interval (m) 

 associated with 

 Isd (.05)' 



7-year height 



(cm) 



16 



28 



463 



Adjusted height 



- 









PREF 640 



(cm) 



14 



10 



544 



PREF 1,500 



(cm) 



6 



5 



2,461 



Lost Valley 



(cm) 



3 



12 



4,735 



Late growth 



(cm) 



7 



6 



224 



Needle cast 



(0/0) 



48 



38 



354 



Mites 



(O/O) 



5 



14 



2,519 



Shoot borer 



(%) 



3 



10 



5,129 



Spring frost injury 



(%) 



2 



2 



2,268 



Snow damage 



(0/0) 



28 



17 



643 



■"Calculated for linear regressions as the ratio [lsd{.05)]lb and for nonlhiear regressions 

 as the solution to the quadratic equation for elevations associated with V + y2[/sd(.05)] 

 at the mean geographic intercept. 



imply that the elevational transfer of seeds should be 

 greatly restricted, but the gentle geographic clines imply 

 that the lateral movement of seeds can be relatively 

 liberal. These conclusions are corroborated by others for 

 the same species in adjacent geographic regions (Rehfeldt 

 1983a, 1985a; Stoneman 1985). 



Regardless, there is little doubt that adaptive variation 

 among populations of P. contorta reflects physiological 

 specialization for relatively small segments of the environ- 

 mental gradient. Specialized populations develop from a 

 balance of environmental stimuli that harmoniously adjust 

 the developmental cycle to the physical and biotic environ- 

 ment. By means of this specialized evolutionary mode, 

 coadaptive traits have been designed by natural selection 

 to overcome ecological problems as diverse as variable 

 frost-free periods, insect infestations, and disease epi- 

 demics. Therefore, the large geographic and broad 

 ecological distributions of this species result from the 

 existence, of innumerable populations, each of which is 

 specialized. Perpetuation of specialized populations is 

 readily facilitated by cycles of even-aged reproduction 

 whereby individuals tend to become established on sites 

 that supported their ancestors. 



As evidenced by substantial genetic variances within 

 populations (Ying and others 1985; Rehfeldt 1985b), 

 specialization has not led to genetic uniformity. Although 

 migration, mutation, and the founder effect undoubtedly 

 contribute to intrapopulation variation, much of this varia- 

 tion likely results from variable selection pressures 

 associated with environmental heterogeneity in time. For 



long-lived stationary organisms, temporal heterogeneity 

 not only is pronounced but also places a limit on the 

 degree to which specialization can develop. Bryant (1976) 

 noted that, if populations are to survive, the process of 

 specialization cannot deplete the genetic variability re- 

 quired for accommodating environmental heterogeneity in 

 time. 



Forest trees of the Rocky Mountains have achieved 

 adaptation to heterogeneous environments according to 

 different modes. Pseudotsuga menziesii, like P. contorta, 

 exhibits specialization, but Larix occidentalis Nutt. and 

 Pinus monticola Dougl. are generalists: adaptive clines are 

 relatively flat, and populations express a high fitness to a 

 broad range of environments (Rehfeldt 1984). Pinus 

 ponderosa Dougl. ex Laws., moreover, displays an inter- 

 mediate mode. Thus, the specialist mode is characterizing 

 species with broad ecological distributions while the 

 generalist mode is characterizing species with relatively 

 narrow distributions. Indeed, P. contorta and Pseudotsuga 

 menziesii have greater botanical distributions and larger 

 ecological amplitudes than any other western conifer; but 

 the ecological distribution of Pinus monticola and Larix 

 occidentalis is relatively small while that of Pinus 

 ponderosa is intermediate (U.S. Department of Agriculture 

 1965). An understanding of the ecological genetics of 

 species such as Thuja plicata Donn ex D. Don and Tsuga 

 heterophylla, both of which have narrow ecological ampli- 

 tudes and small geographic distributions, is necessary for 

 assessing this apparent relationship between abundance 

 and the adaptive mode. 



9 



