Favorable fragmentation: river reservoirs can impede downstream expansion of riparian weeds

Abstract

River valleys represent biologically rich corridors characterized by natural disturbances that create moist and barren sites suitable for colonization by native riparian plants, and also by weeds. Dams and reservoirs interrupt the longitudinal corridors and we hypothesized that this could restrict downstream weed expansion. To consider this "reservoir impediment" hypothesis we assessed the occurrences and abundances of weeds along a 315-km river valley corridor that commenced with an unimpounded reach of the Snake River and extended through Brownlee, Oxbow, and Hells Canyon reservoirs and dams, and downstream along the Snake River. Sampling along 206 belt transects with 3610 quadrats revealed 16 noxious and four invasive weed species. Ten weeds were upland plants, with Canada thistle (Cirsium arvense) restricted to the upstream reaches, where field morning glory (Convolvulus arvensis) was also more common. In contrast, St. John's wort (Hypericum perforatum) was more abundant below the dams, and medusahead wildrye (Taeniatherum caput-medusae) occurred primarily along the reservoirs. All seven riparian species were abundant in the upstream zones but sparse or absent below the dams. This pattern was observed for the facultative riparian species, poison hemlock (Conium maculatum) and perennial pepperweed (Lepidium latifolium), the obligate riparian, yellow nut sedge (Cyperus esculentus), the invasive perennial, reed canary grass (Phalaris arundinacea), and three invasive riparian trees, Russian olive (Elaeagnus angustifolia), false indigo (Amorpha fruticosa), and tamarisk (Tamarix spp.). The hydrophyte purple loosestrife (Lythrum salicaria) was also restricted to the upstream zone. These longitudinal patterns indicate that the reservoirs have impeded the downstream expansion of riparian weeds, and this may especially result from the repetitive draw-down and refilling of Brownlee Reservoir that imposes a lethal combination of drought and flood stress. The dams and reservoirs may also interrupt hydrochory, the downstream flow of seeds and clonal fragments. We thus conclude that with some operational patterns, dams and reservoirs can impede the downstream expansion of riparian weeds.

Sand and sandbar willow: a feedback loop amplifies environmental sensitivity at the riparian interface

“…Also Xowing into Hells Canyon, the Salmon River represents one of the only major non-dammed rivers in the contiguous United States. Our study investigated riparian zones along these two rivers enabling: an upstream (Weiser) versus downstream (Hells Canyon) comparison along the Snake River, a progressive downstream comparison below the HCC, and a comparison of the dammed Snake River versus the free-Xowing Salmon River (Braatne et al 2008;Rood et al 2010a). …”

“…We previously described historic reservoir patterns (Rood et al 2010a) and in the present study we analyzed discharges of the Snake River at Weiser, Idaho (USGS13269000, 1910-present) and below Hells Canyon Dam (Idaho Power and USGS13290450, 1965-present), and the Salmon River at Whitebird, Idaho (USGS13317000, 1910-present). Data were summarized for annual minimum, mean and maximum discharge, and seasonal, daily Xow patterns.…”

“…From 1998 to 2001, we established 235 permanent transects along these river and reservoir reaches (Table 1; Rood et al 2010a). Transects were located with a stratiWed-random scheme with the study corridor divided into 0.2-mile (0.32-km) segments, and one segment in each river mile was randomly selected (permanent mile markers have been positioned along the river).…”

“…For herb quadrats, canopy cover of each species was estimated using an eight class scale (Daubenmire 1959;Rood et al 2010a) and for shrub quadrats, cover of each species was estimated for the shrub layer, from 1-4 m above ground using the same eight classes. Surface substrate was classiWed for each quadrat as: (1) Wne = sands and Wner particles (<2 mm); (2) Wne-coarse = Wnes and gravel (0.2-6 cm); (3) Wne-very coarse = Wnes and cobbles (6-25 cm); (4) coarse = gravel, cobbles, small boulders (>25 cm) without surface Wnes; and (5) boulder/bedrock.…”

“…Following implementation of a number of upstream dams (Parkinson et al 2003) but prior to the HCC, there was major attenuation of peak Xows of the Snake River, and reduced summer Xows ( Fig. 1; Rood et al 2010a). After implementation of the HCC, there was some further change in the seasonal regime but this primarily reXected upstream river regulation since Xows from Hells Canyon Dam were closely aligned with upstream inXows (Fig.…”