Ge. Rehfeldt, GENETIC-VARIATION, CLIMATE MODELS AND THE ECOLOGICAL GENETICS OF LARIX OCCIDENTALIS, Forest ecology and management, 78(1-3), 1995, pp. 21-37
Provenance tests of 143 populations of Larix accidentalis revealed gen
etic differentiation for eight variables describing growth, phenology,
tolerance to spring frosts, effects of Meria laricis needle cast, and
survival. Geographic variables accounted for as much as 34% of the va
riance among Rocky Mountain populations. Patterns of genetic variation
were dominated by the effects of latitude and elevation, with populat
ions from the north and from high elevations having the lowest growth
potential, the least tolerance to the needle cast, and the lowest surv
ival. However, the slope of the geographic dines was relatively flat.
Populations in the same geographic area, for instance, need to be sepa
rated by about 500 m in elevation before genetic differentiation can b
e expected. Regression models using geographic predictors were develop
ed to describe variation in ten climatic variables from 192 weather st
ations that best represented the Rocky Mountain distribution of L. occ
identalis. Values of R(2) ranged from 0.43 to 0.92 and were higher for
temperature than for precipitation variables. Using climatic effects
predicted for each provenance to describe genetic variation produced v
alues of R(2) between 0.15 and 0.27 which accounted for nearly as much
(68-100%) of the genetic variance as had geographic variables. The an
alyses suggested that genetic differentiation was controlled primarily
by the relative mildness of the climate. Secondary effects of precipi
tation were implicated for variables measuring the impact of Meria nee
dle cast and survival.