9371 - Laserfiche WebLink

Home Browse Search

1 1 e 1 1 1 1 Table 3.6. Percent of each cover type at each levee removal site in August 1998 (based on three transects per site). Site Barren Native trees or shrubs Saltcedar Whitetop Other herbs Bonanza 59 7 18 4 12 Stirrup 43 6 18 8 25 Horseshoe 83 2 6 1 7 Baeser Bend 62 15 12 4 7 Above Brennan 60 21 4 4 11 J4 65 16 14 1 3 Upper L7 66 11 10 3 26 Lower L7 53 1 6 14 11 Old Charley diked 44 13 8 3 32 Site Characteristics Influencing the Distribution of Giant Whitetop The discriminant function analysis identified three key variables that seem to strongly influence the distribution of giant whitetop. These are relative topographic position, distance along a meander wavelength, and distance to the river channel (in order of decreasing significance). The partial r2 values and p-values are presented in Table 3.7. Table 3.7. Variables affecting the distribution of giant whitetop. Partial r' 12-value Topographic position 0.35 <0.0001 Distance along a meander wavelength 0.21 0.0032 Distance to the river channel 0.10 0.0798 Total r2 = 0.66 By examining the values of our data for these three variables, we can identify the sites that giant whitetop is most likely to invade. Our data showed that, on average, giant whitetop occurs on upland sites that are on the outer edges of meanders at an average distance of 490 meters from the river channel. The model created by the discriminant function analysis was tested using a jackknife technique on the original 53 observations. This method estimated tha our model will correctly predict the presence of giant whitetop 86% of the time, and it will correctly predict its absence 70% of the time. 1 28