in demography has long been recognized (Turesson 1922, Clausen et 

 al. 1940, Stebbins 1950), and different ecotypes can occur within 

 short distances of each other (Solbrig and Simpson 1977) . Under 

 selection pressure from the environment, different populations 

 can evolve divergent genotypes highly adapted to their 

 surroundings. On the other hand an "all-purpose genotype" adapts 

 to different environments by plastic responses in morphology, 

 phenology and demography (Bradshaw 1965, Schlichting 1986) . 

 These two means of coping with environmental variability are not 

 mutually exclusive, and either or both could be the source of 

 variable demographic strategies in Arabis f ecunda . 



Regardless of the underlying mechanism causing demographic 

 differentiation in Arabis f ecunda , it is interesting to speculate 

 on the environmental factors that drive the differentiation. The 

 Birch Creek and Charleys Gulch populations occur at 1430-1530 m 

 in elevation and are less than 10 km apart in the Sapphire Range. 

 They are found in similar soils derived from the same geologic 

 formation. Lime Gulch and Vipond Park are at 1890 and 2195 

 respectively and are ca. 30 km apart in the Pioneer Range. The 

 two sites in the Pioneer Range are 130 km south of the two 

 Sapphire Range sites. If abiotic factors such as weather or 

 soils were determining life history strategies, we would expect 

 there to be more similarity between the more geographically 

 proximal sites, but this is not the case. The fact that the 

 Birch Creek population is demographically more similar to Vipond 

 Park than Charleys Gulch and that Charleys Gulch is more similar 

 to Lime Gulch than Birch Creek indicates that abiotic factors, 

 meteorological or edaphic, are not responsible for the 

 demographic differentiation. Arabis f ecunda habitats at Birch 

 Creek and Vipond park have, on average, 15% more bare soil than 

 those at Charleys Gulch and Lime Gulch, while the total cover of 

 vascular plants is similar. Thus, A_^ f ecunda populations that 

 experience greater interspecific £:rowding are also the ones with 

 higher and more variable demographic parameters. 



Crowding has been shown to increase mortality of younger or 

 smaller plants (Harper 1977, Powell 1990). Competition for water 

 or nutrients could increase variability in mortality rate by 

 amplifying the effects of adverse weather conditions. For 

 example, a crowded population would be expected to suffer heavy 

 mortality in a dry year, while a population that is not competing 

 for water may not show such a large decrease in survivorship. In 

 wet years, mortality at the two sites would be more similar. 

 Recruitment of A^ fecunda is enhanced by reduced interspecific 

 competition (Lesica and Shelly 1992b) and should be lower in less 

 crowded habitats. However, if differences between the two 

 demographic strategies are under genetic control, increased 

 fecundity and recruitment may have evolved to compensate for 

 increased mortality in crowded habitats. Experimental work is 

 needed to determine the basis for the two demographic strategies 

 found in A^ fecunda populations. Nonetheless, our results suggest 



14 



