Predicting growth response of shrubs to clear-cutting and site preparation in coastal Oregon forests

Cover-projection models were developed based on algebraic difference formulations of an exponential-power function to describe shrub recovery and development patterns after clear-cutting and site preparation. We tested the effect of six treatments on shrub growth patterns by incorporating indicator variables into the rate and shape parameters of the models for salal (Gaultheria shallon Pursh), thimbleberry (Rubus parviflorus Nutt.), salmonberry (Rubus spectabilis Pursh), and all shrubs. For salal, the shape parameter included an adjustment for burning treatments that delayed maximum cover by several years as compared with unburned treatments. The rate parameter in the thimbleberry model was adjusted for burning treatments; maximum cover occurred about 2 years earlier in burned than in unburned treatments. Both rate and shape parameters in the salmonberry model were adjusted for burning treatments; delayed establishment but increased growth rate and less salmonbeny cover are characteristic of burned treatments as compared with the unburned treatments. The rate and shape parameters in the model for the shrub group included adjustments for burning treatments. Overstory removal fostered shrub development, whereas site preparation treatments slowed and curtailed it. The final cover-projection models accounted for 68-92% of the total variation in cover, with the adjustments for burning accounting for 1.5-3.3% of the variation. The predicted growth patterns are consistent with trends in site occupancy and published autecological characteristics.