Dynamics of wood recruitment in streams of the northeastern US

Wood is an important component of forested stream ecosystems, and stream restoration efforts often incorporate large wood. In most cases, however, stream restoration projects are implemented without information regarding the amount of wood that historically occurred or the natural rates of wood recruitment. This study uses a space-for-time analysis to quantify large wood loading to 28 streams in the northeastern US with a range of in-stream and riparian forest characteristics. We document the current volume and frequency of occurrence of large wood in streams with riparian forests varying in their stage of stand development as well as stream size and gradient. Linear models relating stream wood characteristics to stream geomorphic and forest characteristics were compared using Akaike's Information Criterion (AIC) model selection. The AIC analysis indicated that the volume and frequency of large wood and wood accumulations (wood jams) in streams was most closely associated with the age of the dominant canopy trees in the riparian forest (best models: log10(large wood volume (m3 100 m−1)) = (0.0036 × stand age) − 0.2281, p < 0.001, r2 = 0.80; and large wood frequency (number per 100 m) = (0.1326 × stand age) + 7.3952, p < 001, r2 = 0.63). Bankfull width was an important factor accounting for wood volume per unit area (m3 ha−1) but not the volume of wood per length of stream (100 m−1). The empirical models developed in this study were unsuccessful in predicting wood loading in other regions, most likely due to difference in forest characteristics and the legacy of forest disturbance. However, these models may be applicable in other streams in the northeastern US or in streams with comparable riparian forests, underlying geology, and disturbance regimes—factors that could alter long-term wood loading dynamics. Our results highlight the importance of understanding region-specific processes when planning stream restoration and stream management projects.