Wagner et al. (1984). Beginning in 1990 we also recorded the 

 length and degree of pinnation of the sterile frond segment. 



In 1991 plants were too dense to map in portions of plots 3 

 and 4 in the upper line. These areas have been eliminated from 

 the analysis. 



Data Analysis 



We found that some plants would go undetected for one or 

 more years but reappear in subsequent years. The presence of 

 underground plants can be inferred by comparing transect maps 

 from the full sequence of years. We recorded these plants as 

 "U." The proportion of underground plants ranged from 3-20% in 

 1991-92 among the three species. In 1990-91 among the three 

 species, 93% of the plants stayed underground for only one year. 

 Thus, we have chosen to eliminate the first and last years of the 

 study from demographic analysis. Montgomery (1990) reported that 

 plants of Botrvchium dissectum would often be "missing" for one 

 or two years before appearing again. 



Mortality is the proportion of the plants present in year t 

 that died between years t and t+1. Recruitment is the proportion 

 of plants present in year t that were new recruits. Statistical 

 significance of mortality and recruitment proportions among 

 species were compared with an overall chi-square goodness of fit 

 test. If a 2 X 3 test showed a significant result, we used 2X2 

 tests to determine which pairs of sites were different. 

 Probability values were not. corrected for multiple tests. We 

 employed an a-level of 0.05. 



RESULTS 



No plants "changed species" during the four years of the 

 study. The few plants that we previously reported as changing 

 from one species to another were based on misidentif ications. 



Sample population sizes in 1990-92 were relatively stable; 

 all three species had slightly more plants in 1992 compared to 

 1990 (Fig. 1) . Survivorship of the 1989 sample population was 

 highest for Botrychium hesperium . lowest for B^ paradoxum and 

 intermediate for B^ watertonense (Fig. 2) . Mortality was higher 

 in B^ paradoxum compared to B_j. hesperium for all three years 

 measured, and this difference was statistically significant in 

 two of those years (chi-square test, P<0.05; Fig. 3). Mortality 

 in B^ watertonense was more variable than the other two species, 

 as high as B^ paradoxum in one year and lower than B^ hesperium 

 another (Fig. 3) . Recruitment among the three species showed a 

 pattern very similar to mortality. Recruitment was higher in B. 

 paradoxum . and this difference was statistically different in two 

 of the three years (Fig. 4). Recruitment proportion for B_^ 



