|
<br />50
<br />
<br />F. ZAMORA-ARROYO ET AL
<br />
<br />their riparian corridors (Dynesius & Nilsson, 1994; Nilsson et al., 1997). In south-
<br />western United States, flow-regulated riparian zones have been damaged by loss of the
<br />natural pulse-flood regime, that formerly washed excess salts from riverbanks and
<br />germinated native trees (Briggs, 1996; Busch & Smith, 1995; Poff et aI" 1997;
<br />Stromberg, 1998a). On the lower Colorado River, the largest and most altered river in
<br />the south-western United States and Northern Mexico, an exotic, salt tolerant shrub,
<br />Tamarix ramosissima (salt cedar), in association with native halophytes, has almost
<br />completely replaced the mesophytic native forest that historic\\lly dominated the riparian
<br />corridor from Grand Canyon to the delta on the Gulf of California (Ohmart et aI" 1988;
<br />Busch & Smith, 1995). Loss of Populus fremontii (Fremont's cottonwood) and Salix
<br />gooddingii(Goodding's willow) trees and associated epiphytes and understory plants,
<br />has lead to a collapse of supporting habitat for numerous species of plants, birds,
<br />mammals and reptiles, so that today 45 species in the lower Colorado River ecoregion
<br />are listed as sensitive, threatened or endangered (U.S. Bureau of Reclamation, 1996).
<br />Deterioration of native habitat on regulated rivers cail be progressive and irreversible
<br />(Nilsson et al., 1997). It is not known if restoring elements of a natural flow regime
<br />would by itself permit native species to repopulate south-western riparian zones (Briggs,
<br />1996). We report on the effects of pulse floods on vegetation in the delta region of
<br />the Colorado River, below the last diversion of water at Morelos Dam in Mexico. These
<br />discharges began with the filling of the last large reservoir on the river, Lake Powell
<br />behind Glen Canyon Dam, in 1981 (Glenn et aI., 1996, 1999).
<br />The entire discharge of the Colorado River is considered to be over-apportioned for
<br />human use with no water remaining for ecosystem maintenance (Morrison et aI., 1996).
<br />However, the flows of arid-zone rivers are inherently variable. The river's dams are
<br />operated to keep the large reservoirs full, to accommodate electric power generation,
<br />recreation, and storage for downstream water use. Hence, when there is excess
<br />runoff in the watershed, large volumes of so-called 'waste spills' are released to the
<br />Colorado River delta and upper Gulf of California. Since 1981, the major releases have
<br />been associated with ENSO cycles in 1983, 1993 and 1997, which brought above-
<br />normal winter or spring precipitation into the lake system (Glantz et aI., 1996; Li
<br />& Kafatos, 2000)(Fig. 1).
<br />We analysed the effects of river discharge on abundance of vegetation from 1992
<br />to 1999 and characterized the species composition of existing vegetation using remote
<br />sensing and ground methods. We found that pulse floods have reestablished cohorts of
<br />native trees in a 100 km stretch of the riparian zone of the delta, and that the extent of
<br />vegetation cover in this stretch is responsive to flood flows.
<br />
<br />Materials and methods
<br />
<br />Study area andjlow data
<br />
<br />The general study area is the zone of natural (non-agricultural) vegetation in the riparian
<br />corridor from Morelos Dam to the mouth of the Colorado River in the Gulf of California
<br />(Fig. 2). The main vegetation types and hydrological features of this zone have been
<br />described and entered into a Geographic Information System database (Valdes-Casillas,
<br />1998), The riparian corridor is confined within 6 m tall, earthen levees that keep flood
<br />waters out of adjacent agricultural fields. The corridor is < 2 km wide in the northern
<br />stretch as it passes through the agricultural district, then widens to 30 km in the southern
<br />stretch as it approaches the ocean. The river in the delta is a series of braided channels
<br />interspersed with straight sections which have been dug to facilitate water movement,
<br />Our primary interest for the present report is the 100 km (14,316 ha) stretch of river
<br />from Morelos Dam to the junction of the Colorado River with the Hardy River which
<br />contains a mixture of native trees and shrub vegetation. Below the juncture the river
<br />
<br />I
<br />t
<br />i
<br />r
<br />i
<br />~
<br />I
<br />~
<br />~
<br />~,
<br />~
<br />p
<br />I
<br />~
<br />~
<br />I~
<br />~
<br />'if
<br />*
<br />~
<br />tt
<br />f;:'
<br />~:
<br />%
<br />~:
<br />~
<br />~
<br />(i
<br />t
<br />f'
<br />~
<br />~
<br />~
<br />~
<br />~
<br />~
<br />~~
<br />~
<br />
<br />REGENERATION OF TREES IN RESPONSE TO FLOOD RELEASES
<br />2??oo
<br />
<br />51
<br />
<br />
<br />
<br />o
<br />
<br />
<br />1970
<br />
<br />1990
<br />
<br />2000
<br />
<br />1980
<br />Year .
<br />
<br />Figure 1. Annual water releases from the United S . ,
<br />the Southerly International Boundary Th's ,tates .to ~exlco In the Colorado River below
<br />water and flows to the Gulf of Califo~ia ~:;~g~n~ staCtJon IS bel~w the last ?ive;-;ion point for
<br />g e olorado River delta npanan zone.
<br />
<br />1960
<br />
<br />carries saline agricultural drain water to th .
<br />native trees. . e sea and IS apparently too saline to support
<br />
<br />We used a combination of low-level aerial surve
<br />wens to document the vegetation es and ys, ground .~sects and monitoring
<br />locations are shown in Fig 2) WetythP hI yddrology of this nver stretch (sampling
<br />. .. en corre ate past eta' '
<br />events, using satellite images of the delta and fl dV ge no':1 cover With past flow
<br />Boundary and Water Commission US De t ow ata prOVided by International
<br />measured at the Southerly Internati~n~1 130 ~. of State, EI Paso, Texas, USA. Flows
<br />for water, were assumed to flow to the sea u~ ary' 3.~ km b~low ~e last diversion point
<br />(AI Goeff, IBWC, pers. comm.). WI a resl ence time In the delta of 3-5 days
<br />
<br />~
<br />.~
<br />i
<br />f.
<br />I,
<br />~
<br />t
<br />k
<br />"
<br />~
<br />~
<br />&
<br />~
<br />f
<br />1\.
<br />I
<br />I
<br />
<br />Aerial SUI'VeyS
<br />
<br />We conducted a medium-altitude (lOOOm) 'al
<br />documented by oblique videography to ga' aen su~ey on 27 February 1997,
<br />. . ' In an overview of geomo hId
<br />vegetation of the nparian ecosystem and to rp 0 ogy an
<br />during a release event of known m . obse~e patterns of water distribution
<br />(150 m) aerial photographic survey o~~~~' This was foll~wed by a low-level
<br />releases, using a multiband (red blue and NIR) ad~ ,19199, follOWing 3 years of water
<br />2001). ' Iglta camera (DyCam) (Nagler et aI.,
<br />
<br />DyCam images taken within the nati (
<br />% of bare soil, trees, shrubs and groun;:o~:~ ~~n~ n = 63) were used to. determine the
<br />approximately 67 x 100 m ground area Each h e present s~dY. Eac~ Image covered
<br />viewing program and overlain with a iliin-li~d O~Oograo ph,was l~dPorted Into a computer
<br />, -pOint gn . Land cover class was
<br />
|