|
if./u;, U1 THU 15:15 F.~L ".-'rJ °'S °451 `:LSC/SAES
<br />a • III IIIIIIIIIIIII III •
<br />ss9
<br />Responses of Riparian Cottonwoods
<br />to Alluvial V•Jater Table Declines
<br />MICHAEL L. SCOTT'
<br />PATRICK B. SHAFRO7H
<br />,~ G
<br />3
<br />5~~3~oi
<br />n?n. e~-~
<br />~ 001
<br />GREGOR T. AUBLE
<br />United States Geological Survey,
<br />Midcontinent Ecological Science Center
<br />4572 McMurry Avenue
<br />Fort Collins, Colorado 80525-34170. USA
<br />ABSTRACT !Human demands for surface and shallow allu-
<br />a:al groundwater have conVibuted to the loss, fragmentation,
<br />and simplification of riparian ecosystems. Populus species
<br />typically dominate riparian ecosystems throughout Arid and
<br />semiarid regions of North American and efforts to minimize
<br />loss of riparian Populus requires an integrated understand-
<br />ing of the role of surface and groundwater dynamics in the
<br />establishment of new. and maintenance of existing, stands.
<br />In a controlled, whole-stand field experiment, we quantified
<br />responses of Populus morphology, growth, and mortality to
<br />water stress resulting from sustained water table decline fol-
<br />lowing in-channel sand mining along an ephemeral sandbed
<br />stream in eastern Colorado. USA. We measured live crown
<br />volume, radial stein growth, annual branch increment, and
<br />mortality of 6891ive Populus delfoides subsp. monilifera
<br />Riparian corridors occupy important landscape posi-
<br />tions between upland and aquatic ecosystems and are
<br />uniquely productive, phy5ical]}• dynamic, and biologi-
<br />cally diverse (Brinson and others 1981, Gregory and
<br />others 1991, Neiman and odtets 1993). Depletions of
<br />surface and shallow alluvial groundwater have contritr
<br />used to the loss, fragmentation, or severe ecological
<br />impairment of these systems (Dynesius and Nilsson
<br />1994. Svomberg and others 1996). Species of Pppulur
<br />are the most abundant veer of riparian ecosysums
<br />throughou t aria and semiarid regions of North America.
<br />Populus-dominated stands provide unique structural
<br />habitat (Brinson and others 1981) and are vulnerable to
<br />reductions in surface and groundwater availability.
<br />Declines in Populus forest cover have been observed
<br />where severe drought or land and water management
<br />KEY Wt7ROS: Calorado~. Water stress. Groursdwater, Gravel mining;
<br />PopuNS de/raider: Ripadan; Water table tleclines
<br />'Author ro whom correspondence should he addressed.
<br />stems over four years In conjunction with localized water
<br />sable declines. Measurements began one year prior to min-
<br />ing and included trees in both affected and unaffected ar-
<br />eas. Populus demonstrated a threshold response to water
<br />table declines in medium alluvial sands: sustained declines
<br />?7 m produced leaf desiccation and branch tlieback within
<br />three weeks and signifcant dec!;nes in live crown volume,
<br />stem growth, and 1j8q mortality over athree-year period.
<br />Declines in live crown volume proved to be a significant
<br />leading indicator of mortality in the following year. A Iegistic
<br />regression of Tree survival probability against the prior year's
<br />live crown volume was significant (-21og likelihood = 270.
<br />~ with 1 dl = 232, P < 0.0001) and trees with absolute de-
<br />clines in live crown volume of >30 during one year had sur-
<br />vival probabilities <0.5 in the following year. In contrast,
<br />Hare gradual water table declines of -0.5 m bed no measur-
<br />able effect on mortality, stem growth, or live crown volume
<br />and pmduced significant declines only in annual branch
<br />growth increments. Developing quantitative information on
<br />the timing and extent of morphological responses and mor•
<br />taliry of Populus to the rate, depth, and duration of water
<br />table declines can assist in the design of management pre-
<br />scriptions to minimize impacts of alluvial groundwater deple-
<br />tion on existing riparian Populus forests. .
<br />activities have decreased water availability by reducing
<br />surface flows or depleting alluvial grotmdwater aquifers
<br />(Albertson and Weaver 1945, Groeneveld and Criepen-
<br />vog 1985, Rood and Heinze-Milne 1989. Rood and
<br />others 1995, Svomberg and others 1996). For example,
<br />approximately 1125 knt of once perennial streams in
<br />Kansas are now intermittent, in part a consequence of
<br />groundwater pumping in the High Plains Aquifer (lay-
<br />her 1986, Luckey and others 1988), and flow depletion
<br />along the Arkansas River is associated with loss of
<br />riparian vees (Hromm and White 1992). Similarity,
<br />large areas of riparian forest have been lost to groundwa-
<br />ter pumping and associated flow depletion in [he
<br />southwestern United States (Svomberg 1993). Human
<br />activities that directly or induecdy influence alluvial
<br />groundwater sources include damming and diversion of
<br />rivers and svcams, groundwater pumping, and channel
<br />incision resulting fiom altered flows of Hater and
<br />sediments, bank stabilvation, and insveam gravel mitt
<br />ing (Btavard and others 1997, I{ondolf 1994. 1997, Rood
<br />and others 1995, Strombetg and others 1996, 1997).
<br />Environmental Management Vol. 23, No. 3, pp. 347-358
<br />
|