Figure 4. Since scale insect abundance was the most important predictor in our multiple logistic regression model, we use simple logistic regression to illustrate the relationship between scale insect abundance and tree condition.

Points represent recorded measures of M. tenebricosa abundance on the x-axis but are not associated with y-axis cumulative probabilities. Colored sections are associated with the cumulative probabilities on the y-axis as they change across measured M. tenebricosa abundance on the x-axis.