Field data were collected in two distinct areas, the Hubbard Brook Experimental Forest in New Hampshire, USA and the Adirondack State Park in New York State, USA. Sixteen headwater stream reaches (defined as 1st or 2nd orders) were evaluated in the HBEF. Fourteen reaches were 300m in length while the remaining two were 100m and 200m each. The reaches were largely in mature forests with remnant legacy trees remaining from past land-use. The exception to this is W4 which was progressively strip cut in the early 1970s. The one stream reach sampled in this area is thus younger than the other reaches sampled, with canopy trees ranging from 48-52 years old.
Study streams in the Adirondack region were spread across three areas: the Five Ponds Wilderness Area in the north-central Adirondacks, the High Peaks Wilderness Area in the eastern Adirondacks and a privately owned preserve in the western Adirondacks. All streams in the Adirondacks were determined to occur in old-growth forests with dominant tree ages >150 years and no evidence of historic clearing. Thirteen 1st and 2nd order stream reaches in old-growth forest stands were selected for this study. These reaches ranged from 93m to 300m in length based on geomorphological or riparian constraints (e.g., waterfalls, beaver flats, or significant changes in riparian vegetation).
Data Collection
Field crews inventoried LW using two methods: a total wood census and a line-intercept technique. The first method, the total wood census, was only conducted at the HBEF. For the total wood census, crews collected diameter and length dimensions on all LW within the stream bankfull width. To be considered LW, a log must be ≥10cm in diameter and ≥ 1 m in length. Diameters were measured to the nearest 0.05m using a wading staff, while length was measured to the nearest 0.5m.
Crews also collected LW data using a line-intercept method in both the Adirondacks and at the HBEF. This method followed the protocol in Warren et al. 2008 “A comparison of line-intercept and census techniques for assessing large wood volume in streams”. Line-intercept sampling was the only data collection method utilized when the Adirondack data were collected and by following the same method at the HBEF it allows for more accurate comparisons between the two sites. This was deemed an efficient estimation of wood volume in streams by Warren et al. (2008), despite the possibility of non-random wood orientations. In the line intercept method, a field tape is laid down the middle of the stream, parallel with the flow of water, and the diameter at the point of intercept for any piece of LW ≥10 cm in diameter and ≥1 m long that crosses this transect is recorded. At the HBEF, species and decay class (1-5) (based on Sollins et al. 1987 - "Patterns of log decay in old-growth Douglas-fir forests") were collected for each piece of LW that crossed the transect.
To understand how riparian forest structure relates to carbon storage at the HBEF, field crews collected data on the forest surrounding each of the 16 reaches. This consisted of 2 variable radius prism plots on each side of the stream (4 total) conducted with a 2.3 (metric) basal area factor prism. Diameter at breast height (DBH) and species were recorded for all trees within the plot. Field crews also collected bankfull width and slope measurements every 50m.
