PS 34-225 - Long-term successional trajectory of restored forests along the Sacramento River, California

Abstract: Convergence with reference systems is a target for restoration, but there is doubt whether this will occur when restoration does not recover natural processes, especially in disturbance-tolerant systems like floodplains. The California Central Valley historically had floodplains kilometers wide populated with riparian forests. Expansive water supply projects and intensive agriculture greatly degraded the natural landscape so that by the 1980s, only ~5% of historical riparian forest area remained with a greatly attenuated flood regime. In 1987, the California State Senate passed legislation to reverse riparian habitat loss by restoring riparian processes and habitat. This effort has resulted in more than 3,000 hectares of restored forest over the past 30 years but has primarily focused on horticultural planting with limited focus on natural riverine processes. Our research evaluated how much restored riparian forest along the Sacramento River has converged with reference forests in terms of vegetation composition and structure and if this mostly came from the recovery of restored forests or the degradation of reference forests. In addition, we evaluated the effect of distance to the river and remnant habitat on recovery trajectories. In 2021 and 2022, we surveyed vegetation composition in 37 restored and 41 reference forest plots in areas previously surveyed over a decade prior. We measured understory composition and cover within ten 1-m understory quadrats and shrub cover in three to five 10-m transects both of which were nested inside rectangular forest plots (600 m²) within which we measured all tree diameters ( >2.5 cm) at breast height. Native species cover decreased in reference forest understories compared to the prior survey. We found that riparian and reference forests are on a trajectory to converge due to degradation, driven mainly by the invasion and dominance of the shrub layer by the exotic, shade-tolerant shrub, Rubus armeniacus. Native forb cover was lower in restored than reference forests, despite having up to 32 years to naturally recruit. Distance to river was negatively correlated with water-dispersed native plant cover, and distance to reference forests was negatively correlated with native wind-dispersed species. R. armeniacus cover was negatively correlated to native understory and shrub cover and positively correlated with overstory cover. Our results suggest that the common forest restoration approach of quickly establishing a native tree canopy to shade out light-demanding invasive species, although effective on shorter time scales, will not work over longer time scales when shade-tolerant species are abundant in the landscape.