Data Package Metadata   View Summary

We studied three cold-water streams – Walton, Shane, and State creeks – in the Jumbo River drainage of the East Branch Ontonagon River. These streams are fed by groundwater at baseflow and runoff during spring snowmelt and summer/fall rain events. All three streams occupy adjacent northwest-flowing drainages, and have similar elevation (409-429 m), drainage area (24.0-45.5 km2), channel width (low-flow channel, 2.1-2.4 m), water depth (13-17 cm), gradient (0.8-1.2%), and mean discharge (28-47 L sec-1) (Table 1). The active channels (i.e., bankfull width) of these streams have well-defined banks, but floodplains are gently sloped and variably covered by riparian vegetation. All three watersheds including their riparian zones were logged at similar times in the past (~1880-1900), leaving the current forest around all three streams as second-growth mixed hardwoods and young conifers. The streams contained variable but low amounts of wood (of all sizes, but on average 2.7 m L x 0.2 m D; as long but thinner than added wood) from past natural inputs or from the regenerating forests. Existing LW was similar in length and diameter to that in other eastern United States streams but less abundant than in streams in unlogged forests. All streams contain native brook trout populations (Salvelinus fontinalis), a species of concern in Michigan due to habitat loss and competition from introduced salmonids. In addition, a guiding principle of the Ottawa National Forest is to manage streams for native organisms and to implement ecosystem restoration where feasible (https://www.fs.usda.gov/ottawa)

For each study stream, a 100-m treatment reach was designated at a random starting point. A 30-40 m buffer reach separated the treatment reach from an upstream 100-m control reach. For the purposes of this study, we only consider the treatment reaches with the added logs, but several previous studies have compared ecological responses in the treated and control reaches. Commercially harvested and debranched logs (bigtooth aspen of mean size, 2.5 m L x 0.5 m D; volume, ~0.5 m3) were sledded over snow to staging sites in winter 2004 on each stream to prevent damage to riparian habitat at each site. On 13 May, 2004, 25 logs were hand-placed, with the help of a local prison work crew, at random locations throughout each treatment reach at haphazard orientations, angles of inclination, and proportions in the channel to mimic natural input. All logs were placed at least partially within the active channel, with some portion of the log generally placed on the stream bank. Unlike many wood additions, no log was cabled in place or otherwise anchored to the bank. We tagged a cut end of each log (generally the bank end) with a unique metal identifier for tracking purposes. The total volume of wood added to each stream was ~12.5 m3.

Measurements were immediately taken of location, orientation, angle of inclination, and proportion of the log in the active stream channel. Location in the reach was measured as the meter mark (0 – 100 m, to the nearest 0.5 m) for the centroid of the log. Orientation was measured by taking a compass bearing from the tag end of the log, along with a bearing for the linear axis of the channel at that location. As magnitude of deviation from the downstream direction and its change over time were of primary importance, orientation relative to downstream direction was recorded regardless of the direction of deviation. Angle of inclination (i.e., slope) was recorded for the main axis of the log using a clinometer, again from the tag end. Proportion of the log in the stream channel was estimated as the percentage of a log’s length in the active channel and calibrated using a tape measure.