Hydraulic and Geomorphic Effects of Large Woody Debris Additions in the Narraguagus River Watershed, Coastal Maine
Maine coastal rivers host the last remaining runs of endangered anadromous Atlantic salmon in the United States, whose populations have decline from ~500,000 returning adults in the 1880s to only ~1000 in 2000. Restoration projects have focused on these coastal river systems to bring natural populations back to the area, and recent efforts involve adding large woody debris (LWD) to small tributaries to improve salmon rearing habitat. Large woody debris actively changes the hydraulics and geomorphology of small streams by acting as a barrier to flow and creating decreased velocity zones, scour pools, and sediment storage and sorting. I study the effects of LWD additions in early August 2008 on hydraulics and substrate in Baker Brook, a west-flowing tributary of the Narraguagus River. Hydraulically, I focus on the treatment reach nearest the confluence with the Narraguagus River (Baker1), and I also study changes in substrate in Baker1 and the upstream treatment location (Baker3). Both study locations are divided into two reaches, treatment (Baker1-T and Baker3-T) and control (Baker1-C and Baker3-C). In Baker1, the treatment and control reaches are further divided into four 50 m sub-reaches based on channel gradient (~1% in Baker1-C-Flat and Baker1-T-Flat; >2% in Baker1-C-Steep and Baker1-T-Steep). In Baker3, we use two 50 m sub-reaches of similar gradient (ranges from ~1% to 2%) to determine substrate changes. Significant post-LWD addition changes are determined by comparison with the control sub-reaches. Changes in the treatment sub-reaches must be larger than those in the control sub-reaches to be deemed significant. I seek to answer three research questions: (1) how much does mean velocity through the study sub-reaches change as a result of additions; (2) how much does hydraulic roughness change; and (3) does sediment storage and spatial sorting result from the LWD additions? I measured reach-average velocities (Ureach) in Baker1 using the salt dilution method in May, July and August 2008 and May 2009. I use rating curves to compare the post-treatment to the pre-treatment Ureach-stage relationship. A temporary decrease in Ureach occurred in October 2008 in Baker1-T-Flat, whereas the other sub-reaches experienced no change in Ureach. A localized change in cross-sectionally averaged velocity (U) measured with a flow meter, is also evident at Baker1-T-Flat, but this is because an added tree lies directly in the downstream cross-section where measurements are recorded. I assessed channel roughness changes by comparing roughness rating curves created using the Manning roughness parameter, n (back-calculated from velocity measurements) for each sub-reach. Because of the short-term decrease in Ureach, roughness increased in Baker1-T-Flat in October 2008 as well. No change in roughness is evident in the other sub-reaches because post-treatment values of n plot on the same decreasing trend with respect to stage as pre-treatment values. I quantified pre- and post-treatment sub-reach substrate median grain size (D50) with intensive clast counts in July 2008 and May 2009. In Baker1, analysis of pre-treatment substrate size show that the flat sub-reaches have a finer substrate size (34-38 mm) than the steep sub-reaches (88-134 mm). Baker3 pre-treatment grain size is similar to that of the flat Baker1 sub-reaches, with a median grain size of 38 mm in Baker3-T and 28 mm in Baker3-C. Two of the three treatment sub-reaches exhibited significant fining (D50 decreased by 37-54%) between the surveys, and the third changed less than measurement uncertainty. One of the three control sub-reaches coarsened significantly (D50 increased by 29%), one fined significantly (-42%), and one coarsened less than measurement uncertainty. In summary, I find that LWD additions in Baker Brook had little effect on reach-scale hydraulics during the flows we observed, but did influence bed-grain size during the 10-month study interval, underscoring the importance of floods on channel change.