4 

 At each site we collected 25 randomly-chosen fruits. We 

 dissected each fruit and recorded the number of viable-appearing 

 seeds in each. 



Population growth rate was calculated by dividing the increase 

 in individuals over the previous year divided by the number of 

 individuals present in the plot the previous year. Negative growth 

 rates reflect a decreasing population size. 



Many of the plants at Vipond Park were still flowering when we 

 read the transects in 1989, thus we read these transects in late 

 June in 1990. 



In 1990 the No. 10 quadrat of the Vipond Park East transect 

 was so crowded that interpretation was impossible, thus this 

 quadrat was dropped from the analyses. 



RESULTS AND DISCUSSION 



A summary of the data collected in 1989 and 1990 is presented 

 in Table 1. At both sites plant density increased substantially, 

 and this increase is reflected in the high positive growth rates. 

 The percentage of plants producing fruit and the mean number of 

 fruits per fruiting plant were also substantially higher at both 

 sites. This change was particularly dramatic at the Lime Gulch 



