3.2 | Log vane analysis
The likelihood ratio test for the integrity CLM (Table 5) indicated that slope affected rankings for log vanes (p = 0.056), while preexisting morphology did not (p = 0.571). The difference between threshold coefficients occurred between intact and all higher (damaged through failed) rankings (p = 0.001). Flatter slopes increased the likelihood of higher integrity rankings for log vanes (Figure 3; Table 5). Although the angle of departure was not included as an explanatory variable in the model due to correlation with log-vane slope, the negative slope for that relationship indicated that flatter slopes were associated with greater angle of departures. Neither slope (p = 0.196) nor morphology (p = 0.774) affected rankings for erosion, and no differences between threshold coefficients were observed (p > 0.5). Results from the CLM for log-vane deposition indicated that morphology had a significant effect on rankings (p = 0.013) while slope did not (p = 0.142). Log-vanes that were placed in riffles were more likely to receive higher deposition rankings when compared to pools (Figure 3), and those difference occurred between moderate and severe rankings (p = 0.001).