<br />
<br />D (70 ({7 cr
<br />
<br />)uvial adjustments to the spread of tamarisk
<br />h the Colorado Plateau region
<br />
<br />,Ie
<br />
<br />rn
<br />se
<br />
<br />,;~';i
<br />
<br />WILLIAM L. GRAF Department of Geography, University of Iowa, Iowa City, Iowa 52242
<br />I \' ~ i
<br />
<br />ld
<br />)1-
<br />le-
<br />'n-
<br />Ise
<br />
<br />
<br />ABSTRACT
<br />
<br />Tamarisk, a shrub or low tree that was
<br />'" artificially introduced into the American
<br />!lSouthwest in the late 1800s, has spread
<br />';:'r.r,oughout the Colorado Plateau region by
<br />!OfX\lPying islands, sand bars, and beaches
<br />.Iong streams. Historical photographs
<br />~!$bow that tamarisk spread from northern
<br />~Fona to the upper reaches of the Col-
<br />~'orado and Green Rivers at a rate of about
<br />"20 kmIyr, Detailed studies on the Green
<br />River in Canyonlands National Park, Utah,
<br />Show that the plant has trapped and
<br />'stabilized sediment, causing an average re-
<br />duction in channel width of 27%. Photo-
<br />'grammetric analysis of historical ground
<br />photography, including photos from John
<br />Wesley Powell's 1871 expedition, and re-
<br />tent aerial photographs supplemented by
<br />field surveys provided quantitative data.
<br />I Expanded islands and channel-side bars
<br />: exhibit allometric relationships as they
<br />Idtange, apparently maintaining a balance
<br />I between turbulence and friction. Overbank
<br />lOooding is common on the tamarisk-
<br />'~tabilized features.
<br />
<br />;sa
<br />In-
<br />
<br />se,
<br />76,
<br />iak
<br />
<br />~es
<br />
<br />ac- n
<br />
<br />f::
<br />
<br />
<br />ler- 1
<br />
<br />cli-!I
<br />uic
<br />;0. ,
<br />of H
<br />)pe (1
<br />les: :1
<br />>1m ij
<br />H
<br />
<br />
<br />i1NTRODucnON
<br />hi
<br />~ H
<br />;1 A review of historical ground photog-
<br />',,
<br />:1SPhy taken by explorers, surveyors, and
<br />;'!iprofessional photographers indicates that
<br />'"within the past century the most dramatic
<br />:jlchange in the fluvial landscape of the Col-
<br />. do Plateau has been in the riparian vege-
<br />rion, The changes shown in.Figure 1 are
<br />ical of most of the major water courses
<br />the Colorado Plateau, where shifting
<br />nd bars, beaches, and islands that were
<br />ce sparsely vegetated are now covered by
<br />e thickets of tamarisk, In most cases,
<br />'thickets consist of a few fully developed
<br />risk trees 10 to 12 m (30 to 36 ft) taU,
<br />rounded by closely spaced plants 5 to
<br />(15 to 18 ft) tall, that exclude other
<br />.~. By the late 1960s, tamarisk proba-
<br />
<br />bly covered 500,000 ha (1.3 million acres)
<br />in the American Southwest (Robinson,
<br />1965),
<br />Because tamarisk colonizes moist sand
<br />surfaces, grows rapidly, and stabilizes sed-
<br />iment, it is an effective geomorphic agent,
<br />The early photography, beginning with
<br />John Wesley Powell's second expedition, in
<br />1871, shows substantial growth of stable
<br />surfaces near channels as a result of the
<br />tamarisk invasion (Shoemaker and Ste-
<br />phens, 1969). Baars and Molenaar (1971),
<br />Lohman (1974), and Shoemaker and Ste-
<br />vens (1975) offered examples of compara-
<br />tive photographs, Islands have become
<br />longer and wider, channel-side bars have
<br />widened, and some alluvial-fan surfaces
<br />have expanded, accompanied by concomit-
<br />ant reductions in channel widths. The
<br />largest changes in near-channel landforms
<br />have taken place in canyons with relatively
<br />gentle gradients and floors that are at least
<br />two channel widths wide. Changes are
<br />much smaller in canyons with steep gra-
<br />dients, rapids, and floors that are almost
<br />completely occupied by the river channel.
<br />The net result of channel restrictions in
<br />the canyons is increased frequency of over-
<br />bank flooding and increased roughness of
<br />near-channel surfaces that serve as channel
<br />perimeter during high-water periods (Had-
<br />ley, 1961). An appreciation of fluvial re-
<br />sponses to the growth of tamarisk in areas
<br />already infested will permit accurate pre-
<br />diction of changes in areas yet to be in-
<br />vaded. Additionally, the responses of fluvial
<br />systems to biotic disruptions may shed
<br />some light on the general nature of such sys-
<br />tems and may be useful in the analysis of
<br />other types of disruption in the canyons of
<br />the Colorado Plateau, 1ncreasing recre-
<br />ational use of the canyons necessitates a
<br />clear understanding of environmental sta-
<br />bility for effective land management,
<br />In this paper, I discuss the introduction
<br />and spread of tamarisk in the Colorado
<br />River system and the adjustments of the
<br />
<br />gica! Society of America Bulletin, v, 89, p. 1491-1501,13 figs.," tables, October 1978, Doc, no, 81005,
<br />p
<br />~'
<br />
<br />1491
<br />
<br />fluvial geomorphic systems of the Colorado
<br />Plateau to the growth of tamarisk.
<br />
<br />Tamarisk
<br />
<br />The accepted common names of tam-
<br />arisk, salt cedar and sa!tcedar, are usually
<br />applied to many of the 90 species of the
<br />plant, but in the southwestern United States
<br />and for purposes of this paper, the term
<br />"tamarisk" applies to the species Tamarix
<br />chinensis (Lour.) (Horton and Campbell,
<br />1974).
<br />When it was introduced into the south-
<br />western United States in the mid-1800s,
<br />tamarisk was cultivated as an ornamental
<br />shrub in California and later as an erosion-
<br />control plant in New Mexico (Robinson,
<br />1958), but after escaping cultivation in the
<br />late 1800s the plant occupied moist sand
<br />surfaces with little competition from native
<br />plants (Horton, 1962, 1964), The cause of
<br />the spread of the plant throughout the
<br />drainage systems of the region has been a
<br />matter of controversy, Most previous
<br />workers have relied on written accounts
<br />and field collections (for example, Christen-
<br />sen, 1962), which are useful but sometimes
<br />less accurate than photographic evidence.
<br />Bowser (1960) reported that the plant
<br />was growing under natural conditions
<br />along the Virgin River in southwestern
<br />Utah in the 1890s, while Christensen
<br />(1962) stated that the plant was not grow-
<br />ing in the Green or Colorado systems before
<br />1925, Christensen (1962) suggested that the
<br />most rapid spread of the plant was from
<br />1935 to 1955, but Harris (1966) and
<br />Clover and Jotter (1944) contended that the
<br />plant was common in most major drainage
<br />systems of the Southwest after 1920.
<br />Photographic evidence resolves some of
<br />these issues,
<br />The geomorphic effects of plant growth
<br />near channels is significant in humid re-
<br />gions, as shown by Zimmerman and others
<br />(1967) and Thornes (1970), and in arid re-
<br />
|