Laserfiche WebLink
Y=Bo+gix <br />where: <br />Y =estimated age in years <br />Bo = Y intercept <br />Bi = slope of the line <br />_ x = tree diameter (inches) or shrub height (inches) <br />A test of the hypothesis that Bi = 0 (no relationship exists between <br />size and age) is given by the F ratio. The value of R2 measures the <br />proportion of total variation about the mean Y (age) explained by the <br />regression model. For example, the regression model for age versus <br />height for Gambel oak is: <br />Y = 6.626 + 0.475 x <br />with a R2 value of 0.794 (79.4% of the observed variation in age is <br />accounted for by size). The F ratio is 65.55 resulting in the rejection <br />of the hypothesis that no relationship exists between size and age at <br />the 998 confidence level. Substituting recorded shrub heights for x in <br />the formula provides a means for estimated age (Y); a shrub with a height <br />of 142 inches would have an estimate3 age of 74 years. <br />The range of age for aspen Treasured was 10 (dbh = 1.1 cm) to 116 years <br />(dbh = 29.1 canny. The relationship between age and size was significant <br />at the 99~ confidence level for trees both with a dbh larger than 5.8 can <br />(Table 6) and smaller than 5.8 cm (Table 7). The linear regression <br />model for the larger trees accounted for only 58.30 of the variation in <br />age while the fit was much better for the caller trees (96.1%). <br />Size-age relationship was significant at the 95~ confidence level for <br />serviceberry and at 99o confidence level for the other shrub species <br />measured. The degree of fit of the irodel ranged from 31.1% for <br />~• <br />-37- <br />