Forest Management Effects on Channel Wood and Wood-Channel Interactions in Caspar Creek, California

First-cycle logging in much of California's redwood region converted streams with some of the highest known wood volumes and piece sizes to efficient log transport channels. A century later, second-growth trees are still much smaller than old growth, and later logging and stream cleaning have further affected potential wood inputs and large woody debris (LWD) volumes in channels. At Caspar Creek, a 50-year paired watershed study creates an opportunity to compare the effects of two second-growth forest management strategies on wood dynamics in these channels, and to examine how the resulting differences in LWD affect channel form and process. Both the North and South Forks of Caspar Creek started the 20th century with almost no in-channel wood and little potential LWD as a result of clearcut logging, burning, and channel clearing. Stands had partially regrown by 1968, when near-channel roadbuilding and selective logging in the 424-ha South Fork watershed again reduced potential channel LWD. Trees that fell into the channel during logging were removed, along with some instream wood. Logging began in the 384-ha North Fork in 1989 using ridgetop roads; buffer strips were left between the mainstem channel and upslope clearcuts. Potential LWD in the buffer strips was reduced by selective cutting, but channel LWD was not immediately affected. LWD mapping, inventories, and tagging, channel cross-sections and photos, and pool mapping and volume measurements show differences in channel wood and LWD-channel interactions between the two watersheds. Windthrow from buffer strips increased the total channel LWD volume in the North Fork in the mid 1990's while reducing potential future LWD. These higher LWD loads increased pool volumes and enabled increased sediment storage, particularly upstream of logjams. In the South Fork, total LWD volumes are lower and a higher proportion of the wood is residual old growth pieces, some of which entered the channel during the 1970's logging. Mean bed elevations have been decreasing, particularly in areas with little wood. Much of the wood in both channels was mobilized by high flows in HY 2005. LWD input, survival, and mobility data, along with channel-adjacent stand information and growth models, allow estimation of future LWD in these channels and of how that wood might affect channel behavior.