This paper describes the 'recruitment box,' an integrative model that defines the stream stage patterns that enable successful establishment of riparian cottonwood seedlings. In western North America, cottonwood seed dispersal generally occurs after annual peak river flows. The receding stream exposes moist sites upon which seeds land after transport by wind and water, Germination is rapid, and initial seedling establishment is often prolific. However, the vast majority of seedlings die, primarily due to drought stress, as root growth is insufficient to maintain contact with the receding zone of moisture. Cottonwood roots grow about 0.5 to 1 cm per day or 60 to 100 cm in the first year. Along the 'losing' streams in semi-arid regions, the riparian water table is an almost horizontal extension from the stream stage. A capillary fringe exists above the water table and is often 30 to 40 cm in elevation, but can range from about 5 to 130 em depending on substrate texture. The combination of root growth and capillary fringe define the successful recruitment band, which is usually from about 0.6 to 2 m in elevation above the late summer stream stage. Within this range, higher elevation establishment occurs (i) for the Aigeiros cottonwoods, Populus dettoides, and P. fremontii, which grow more rapidly than Tacamahaca species and occur in warmer areas with longer growing seasons; (ii) along larger rivers that are characterized by more gradual stage fluctuations; and (iii) along streams with finer substrate. The rate of stream stage decline is also critical for seedling survival and should not exceed 2.5 cm per day. The recruitment box model is consistent with dendrochronological interpretations that moderate flood events are naturally required for cottonwood recruitment. Flood events with recurrences of about 1 in 5 to 1 in 10 years often satisfy the model and provide stream stage patterns with a gradual decline through the recruitment box. The model will facilitate analyses of the reproductive ecology of riparian cottonwoods and also permit the prescription of stream stage patterns for cottonwood seedling recruitment along danuned rivers.

No abstract

River damming has dramatic environmental impacts and while changes due to reservoir flooding are immediate, downstream impacts are more spatially extensive. Downstream environments are influenced by the pattern of flow regulation, which also provides an opportunity for mitigation. We discuss impacts downstream from dams and recent case studies where collaborative efforts with dam operators have led to the recovery of more natural flow regimes. These restoration programs, in Nevada, USA, and Alberta, Canada, focused on the recovery of flow patterns during high flow years, because these are critical for riparian vegetation and sufficient water is available for both economic commitments and environmental needs. The restoration flows were developed using the “Recruitment Box Model”, which recommends high spring flows and then gradual flow decline for seedling survival. These flow regimes enabled extensive recruitment of cottonwoods and willows along previously impoverished reaches, and resulted in improvements to river and floodplain environments. Such restoration successes demonstrate how instream flow management can act as a broadly applicable tool for the restoration of floodplain forests.

Completed in 1951, the St. Mary Dam enables water storage and diversion for irrigation; river flows downstream are consequently dramatically reduced during summer months. To assess historical changes in the abundance of riparian cottonwoods (Populus balsamifera, Populus angustifolia, and a few Populus deltoides), airphoto analyses were conducted for 40-km river reaches upstream and downstream from the dam and along adjacent dammed and undammed rivers. Cottonwoods along the lower St. Mary River are confined by steep-walled canyons to narrow bands and consequently analyses of the lineal river distance associated with cottonwoods were conducted. These revealed a 68% decline from 1951 to 1985. The decline was progressive, with 28.9, 27.6, 15.1, and 7.6% of the reach associated with cottonwoods in 1951, 1961, 1981, and 1985, respectively. Ground surveys from 1985 to 1994 indicated further decline after 1985 and an absence of cottonwood seedlings and saplings. Cottonwood stands upstream from the St. Mary Dam and along adjacent rivers are more extensive and analyses of the areal extent of stands were consequently appropriate. These indicated minor change along the upper St. Mary (−0.5%), the upper (+1.9%) and lower Waterton (+3.5%), and the upper Belly (−9.1%) rivers, and an increase in forest abundance along the lower Belly River (+52.2%), between 1951 and 1985. Thus, the decline of cottonwoods along the lower St. Mary River was not symptomatic of a general pattern of decline in the region. Analyses of historical stream flows indicated that the cottonwood mortality was drought induced as a result of insufficient flows during the hot, dry summer periods and abrupt flow reductions following the high-flow period in the late spring. The riparian water table was determined to be closely coordinated with river stage, as changes in river elevation were followed by quantitatively similar changes in water table depth. Along the St. Mary River, reduced sedimentation downstream from the dam was not considered to be responsible for the cottonwood decline. The historically sparse cottonwood abundance along the lower St. Mary River may have reflected environmental conditions that were naturally only marginally suitable, and those groves may have been particularly vulnerable to the impacts of river flow regulation. Key words: Populus, cottonwoods, instream flows, mortality, riparian vegetation.

No abstract