The study was conducted within nine reach-pair sites located within the McKenzie River Basin in the western Cascade Mountains of Oregon. Seven of the sites were located within the HJ Andrews Experimental Forest (HJA), a 6,400 ha research forest encompassing the entire Lookout Creek drainage basin. Cook Creek and Fritz Creek were located outside of the HJA. Cook Creek is a tributary of Blue River upstream of Blue River reservoir and Fritz Creek is a tributary of Deer Creek, which flows directly into the McKenzie River.
Each pair consisted of two reaches: one within a section of stream with old-growth riparian forest, and another in a nearby section of stream bordered by second-growth riparian forest on at least one stream bank. Harvesting in the previously managed reaches occurred on just one stream bank in three reach pairs (MR404, LO701, and LO703) and on both banks for all other pairs. Sites were selected based on the presence of old-growth and second growth riparian forests close in proximity on the same stream (within 500 m). Having distinctly different forest types along two nearby sections of the same stream reduces inherent stream-to-stream environmental variability (e.g. temperature, gradient, geology, substrate, etc.) that often arises in comparisons between whole-stream systems in basins with managed versus unmanaged, late succession forests. Reaches ranged from 90-200 m and reaches within a reach pair were separated by a 90-325 m buffer section.
Metrics of habitat and productivity relative to fish and salamander biomass in nine stream reach pairs were collected. Considering differences among streams and between reaches within each stream, relationships between both biotic and abiotic covariates and the biomass of coastal cutthroat trout (Onchorhynchus clarkii clarkii), coastal giant salamanders (Dicamptodon tenebrosus) and total vertebrates (fish and salamanders) were evaluated.
Five of the six fish-bearing stream reach pairs originally surveyed by Murphy and Hall (1981) were visited to determine how stream conditions, benthic biofilms, invertebrate predators, and ultimately resident coastal cutthroat trout have responded to nearly four decades of riparian forest regeneration. Using the upstream reference reaches identified by Murphy (1979), which were bordered by old-growth riparian forests, this design is similar to a before-after control-impact study with riparian regeneration as the treatment. The same study reaches were sampled and evaluated how 40 years of riparian regeneration influenced our response variables.
Kaylor, Matthew J.; Warren, Dana R. 2017. Linking riparian shade and the legacies of forest management to fish and vertebrate biomass in forested streams. Ecosphere. 8(6): e01845. doi: 10.1002/ecs2.1845
Kaylor, Matthew J.; Warren, Dana R. 2017. Canopy closure after four decades of postlogging riparian forest regeneration reduces cutthroat trout biomass in headwater streams through bottom-up pathways. Canadian Journal of Fisheries and Aquatic Sciences. {Volume}: 1-12. doi: 10.1139/cjfas-2016-0519
Murphy, M. L.; Hall, J. D. 1981. Varied effects of clear-cut logging on predators and their habitat in small streams of the Cascade mountains, Oregon. Canadian Journal of Fisheries and Aquatic Sciences. 38: 137-145.
Murphy, Michael Louis. 1979. Predator assemblages in old-growth and logged sections of small Cascade streams. Corvallis, OR: Oregon State University. 72 p. M.S. thesis.
Field Sampling-Abiotic Variables: All data were collected in the summer of 2014 during low-flow conditions. Across all nine pairs of sites, we collected a suite of physical habitat variables in each reach including: canopy cover, bankfull width, wetted width, pool area, large wood abundance and volume, temperature, nutrient concentration, and stream gradient. Canopy cover was quantified using a convex spherical densiometer (Forestry Suppliers Model A). Measurements were taken in each cardinal direction at 11 regularly spaced locations in each reach. All densiometer measurements were taken by the same individual to avoid user bias. Bankfull width and wetted width were measured at transects across the same 11 reach locations. The mean wetted width for each reach was multiplied by reach length to obtain total reach wetted area, which was used to standardize fish, salamander, and invertebrate abundance and biomass estimates per square meter of stream. Pools were identified during summer-low flow conditions as slow velocity habitats connected to the main-channel. Pool area was calculated using the length and width of each pool, and percent pool area was calculated as the total pool area divided by total wetted reach area. We quantified all large wood pieces greater than 1 m in length and 10 cm in diameter (Richmond and Fausch 1995, Young et al. 2006, Warren et al. 2009) We only measured the portion of wood pieces located within the bankfull channel for wood volume estimates. Total channel area (mean bankfull width multiplied by reach length) was used to standardize large wood volume among reaches. Temperature loggers (HOBO Pro v2) were deployed for two weeks during mid-summer to evaluate relative temperature among streams and differences between paired reaches. Due to a limited number of sensors, not all sensors were deployed for the same time interval. Sensors were deployed from 7/20/14-8/3/14 in Cook Creek and Fritz Creek and from 8/4/14- 8/24/14 in all other streams. Water samples were collected in September 2014 at all reaches during a two-day period prior to the onset of autumn rain events in this region. Water samples were filtered (25 mm Whatman GF/F filters), frozen, and analyzed for nitrate-N (NO3-N) and phosphate-P (PO4-P) using a Dionex 1500 Ion Chromatograph. Nitrogen is the limiting nutrient for stream autotrophy in the streams evaluated in this study (Gregory 1980, Warren et al. 2017).
Field Sampling-Biotic Variables: Periphyton chlorophyll a (hereafter chl a) accrual was quantified on 10 ceramic tiles (15 x 15 cm) per reach. Tiles were placed in the stream in mid-July and were retrieved after six weeks. Tiles were spaced at regular intervals and positioned within riffle sections of the stream at a depth of 10-25 cm. After six weeks, tiles were scraped using a wire brush and the slurry was filtered through 47 mm glass fiber filters (Whatman GF/F). Filters were placed in 20 ml glass vials and frozen for 24-48 hours.
Benthic invertebrates were sampled in late July (7/15/14-7/29/14). Both reaches within a reach pair were always sampled on the same day. In each reach, six surber samples (363 µm, 0.0625 m2) were collected from riffle habitats at regular intervals. Substrate within the surber sample quadrate was disturbed to a depth of 10 cm for approximately 30 seconds. Samples were stored in 90% alcohol until laboratory processing.
Fish and salamanders were collected using a backpack electroshocker (Smith-Root model LR-20B). Block nets were set at the upper and lower ends of each reach to prevent movement and close the system for the duration of the surveys. Population estimates were conducted using single-pass mark-recapture methods for all reach pairs except Mack Creek. For mark-recapture surveys, fish and salamanders were anesthetized using AQUI-S 20E (AQUI-S, Lower Hutt, New Zealand), weighed (nearest 0.01 g), measured (total length for fish, and snout-vent length for salamanders), and marked. Fish were marked with a small caudal clip and salamanders were marked with a visual elastomer tag (Northwest Marine Technology, Shaw Island, Washington). Fish and salamanders were released and the reach was resurveyed after approximately 24 hours. The number of marked and unmarked individuals was recorded for each species. Abundance was estimated using the Lincoln-Peterson mark-recapture model, modified by Chapman (1951), and biomass was estimated by multiplying abundance estimates by mean weight. Juvenile (0+) and adult (1+) trout were analyzed separately. Trout were distinguished as juvenile or adult based on length frequency histograms and, in general, trout < 65 mm were classified as juveniles.
Multiple pass depletion methods were used to survey fish and salamanders at Mack Creek. Mack Creek is a long-term ecological research (LTER) site where fish and salamanders are sampled annually using depletion estimates. The long-term research project provided the 2014 fish and salamander data used in this study (S.V. Gregory). Multiple pass depletion and mark-recapture methods can produce significantly different population estimates (Rosenberger and Dunham 2005). To standardize population estimates across all reaches, we applied a correction factor that was obtained from simultaneous mark-recapture and depletion estimates conducted in Mack Creek in 2015 (S.V. Gregory) per Thompson and Seber (1994).
In addition to the primary fish and salamander surveys conducted in mid-summer, a second single-pass survey was conducted in late September 2014 to capture juvenile (age 0+) cutthroat trout and assess summertime relative growth rates for this age class. We did not sample Mack Creek in the second juvenile assessment as we did not want to interfere with long-term research efforts occurring annually at this site. In LO701, MCTE, and Fritz Creek, juvenile trout were surveyed on two sampling dates but there were few surveyed fish in at least one of these surveys (n < 5). Therefore, we were only able to evaluate juvenile relative growth rates in 5 of the 9 sites. Relative growth rates were determined by subtracting the mean weight at survey date 1 from the mean weight of survey date 2 and then dividing this number by the number of days between sampling events.
Chlorophyll a; We extracted Chl a from filters with 15 mL of 95% acetone for 2 to 4 h and estimated Chl a based on fluorometric methods with phaeophytin corrections as outlined in US Environmental Protection Agency (EPA) method 445.0 (Arar and Collins 1997). Fluorescence of a subsample of the extraction solution was measured before and after the addition of 0.1 N HCL (0.15ml/5ml solution).
Invertebrates: In the laboratory, the contents of each of the six surber samples from each reach were combined into a single pooled sample. This pooled sample was then subsampled using a plankton splitter until a minimum of 500 individuals were picked from the subsample. We conducted a 60 second visual search of the remaining sample (less the subsample) to collect large bodied predators to more effectively quantify invertebrate predator biomass. Invertebrates were identified to Family or Genus (Merritt et al. 2008) and individually measured using an ocular micrometer mounted on the dissecting microscope. Invertebrate lengths were converted to biomass using established length- weight relationships (Sample et al. 1993, Sabo et al. 2002, Mark Wipfli; unpublished data). We summed the biomass of individuals within a subsample and divided this summed value by the proportion of the total sample that was subsampled. The addition of this value and the biomass of the 60-second sample to identify large bodied individuals (which was not subsampled) was then divided by the total area sampled (0.375 m2) to obtain biomass estimates per square meter (g/m2).
The previously harvested cutblocks were cleared 40-60 years prior to this study. In all cases, timber was removed down to the stream bank with no riparian buffer. Trees were replanted within 5 years post-harvest in seven of the nine cuts in accordance with forest management practices at the time. Stands in MCTW and Mack Creek were regenerated without any post-harvest planting. In 2014, the second growth riparian forests were predominantly Douglas fir (Pseudotsuga menziesii) but red alder (Alnus rubra) was also a common canopy species in areas directly adjacent to streams and provided substantial stream shading. Old-growth forests were comprised of Douglas fir, western hemlock (Tsuga heterophylla) and western red cedar (Thuja plicata). Red alder was present adjacent to streams within old-growth forests as well, but it was not as common as in second growth sections.
Coastal cutthroat trout and coastal giant salamanders (Dicamptodon tenebrosus) were present in all 18 stream reaches and were the dominant vertebrates. Sculpin (Cottus spp.) were present in both reaches of MR404, but were not found in any other reaches. Tailed frogs (Ascaphis truei) were found in low abundance in some of the streams, but were not evaluated in this study.
