topographic position suggest that Vipond Park is the harsher and less productive of the two 

 sites. It is probably also more variable in terms of climatic fluctuations both within and 

 among years. These environmental differences may help explain the differences in A. 

 fecunda life history at the two sites. Arabis fecunda has the demographic characteristics of a 

 short-lived perennial at Lime Gulch, while at Vipond Park, it behaves more like a biennial, 

 with shorter average lifespans, higher fecundity, and a tendency for plants to be monocarpic. 

 In 1990 the Vipond Park population had a very high proportion of plants that "bolted." 

 These plants produce a very large and fecund inflorescence from the terminal bud followed 

 by death of the whole rosette and often the entire plant. At Vipond Park, death of plants that 

 bolted in 1990 was in large part responsible for the high death rate in 1990-91 (see Appendix 

 A). 



These differences, if they are consistent over a period of years, have implications for 

 species management and reserve design. Although differences in A^ fecunda life history 

 between the two sites may be due, in part, to phenotypic plasticity, it is likely that there are 

 genetic differences as well. These differences should be preserved in any reserve design 

 system. The larger fluctuations in population growth at Vipond Park suggest that this 

 population is more prone to local extinction than the more stable Lime Gulch population 

 (Diamond 1984, Pimm et al. 1988), especially during periods of extreme environmental 

 conditions. Thus, if average climate stays the same, long-term persistence of A^ fecunda is 

 more likely at the Lime Gulch site. However, if the climate becomes warmer (Peters 1988, 

 Henderson-Sellers 1990), sites at higher elevations may be more likely to persist. 



