Recovery of habitat heterogeneity in restored and remnant riparian forests along the Sacramento River, California, United States

Abstract Introduction Ecosystems are naturally heterogeneous, though humans often homogenize them through land use practices. Increasing physical heterogeneity in degraded ecosystems can enhance biodiversity and habitat resilience but is rarely a focus in restoration. Objectives We compared topographic and vegetation structural heterogeneity in 3251 ha of restored and remnant riparian forests along 100 km of the Sacramento River, California, which was degraded by intensive agriculture and flood control and has undergone large‐scale restoration since the late 1980s. Methods We used Light Detection and Ranging (LiDAR) data to derive three measures of topographic and vegetation structural heterogeneity that influence plant and wildlife diversity—topographic ruggedness index (TRI), overstory rugosity, and foliage height diversity (FHD)—at 2‐m (fine) and 10‐m (coarse) resolution. Results Fine‐scale TRI was 76% lower in restored than remnant forests and slowly recovered with forest age. After 11 years of age, remnant forests had 13% greater FHD and 23% greater overstory rugosity than restored forests, whereas by 20 years of age both forests had similar levels of vegetation structural heterogeneity. Restored forests with higher TRI had greater vegetation structural heterogeneity. Conclusions Restored forests increased slowly in fine (microtopography) but not coarse‐scale topographic heterogeneity with age, potentially from scale‐dependent sediment erosion and deposition patterns in this dam‐controlled river. Thus, future forest restoration efforts should consider recreating more topographic heterogeneity at the outset of restoration, which would likely result in greater vegetation heterogeneity. Our study shows the potential of LiDAR data to evaluate recovery of forest structure over large and inaccessible areas.