392
JAMES E. DEACON AND W. L MINCKLEY
(Gautier, 1935) which certainly are inconsistent with present climatic
regimes, and fossil assemblages of vertebrates along these watercourses attest to big river habitats across the presently arid lands (Joleaud, 1935).
Similar situations of ancient, now dry valleys, or remnants of terraces produced by streams larger than those now present, occur in many other
desert regions (Leopold and Miller, 1954; Wright, 1956; Alimen, 1965).
In some instances, remaining waters of such extinct systems support
remnant fish populations, reconfirming drainage connections long since disrupted and obscured (Pellegrin, 1914, 1931, 1934; Braestrup, 1940;
Miller, 1946a, 1949a; Hubbs and Miller, 1948a).
Man's activities over the past few decades in North America—and far
longer in some other deserts—have accelerated desiccation of streams, especially larger ones. These have almost all been beheaded by impoundment
and their waters spread over lowlands to irrigate domesticated crops.
At the time of settlement, large rivers such as the Colorado and Gila
of the Sonoran Desert had complex marshes, lakes, and sloughs along their
lower courses, as did their major tributaries (Sykes, 1937; Leopold, 1953;
Miller, 1961a). Desert creeks also were far more permanent in their lower
parts. Smaller streams in southern Arizona flowed through broad, marshy
floodplains in multiple channels, or moved by seepage (Hastings, 1959a,b;
Miller, 1961a; Hastings and Turner, 1965; Minckley, 1969a). Dark, organic soils form parts of the walls of now dry washes. These soils are remnants of marsh, or ciénega deposits (Martin and Mehringer, 1965).
Riparian galleries of cottonwood (Populus), and other, broad-leafed, deciduous trees, so conspicuous at present along some desert waterways
(Lowe, 1967), may well have been restricted to places of relatively good
drainage in the recent past. Areas of entrenched channel doubtless were
present (Hastings, 1959a,b), bordered by mesquite (Prosopis) and forming
bosques on terraces. Under such conditions stream bottoms are typically
of mud and organic debris, except in places of channel degradation or the
presence of stony dikes. Currents are slow, impeded by debris and the
tortuous nature of channels, and flooding effects are minimized (Miller, 1961a; Minckley, 1969a). The present aspect of many streams in
southwestern United States, with broad, sandy, braided channels, may
therefore be a disclimatic phenomenon.
A trend toward erosion of channels originated in the 1880's with a cycle
of arroyo cutting that may be continuing today (Miller, 1961a; Hastings
and Turner, 1965). Similar erosive events were recorded, at about the
same time, over a large geographic area, and the relative uniformity
"makes it difficult to link settlement. . . with arroyo cutting, the uniform
o n s e t . . . on the other hand, points toward operation of a broad regional
factor like climate" (Hastings and Turner, 1965). However, in southern
JAMES E. DEACON AND W. L MINCKLEY
(Gautier, 1935) which certainly are inconsistent with present climatic
regimes, and fossil assemblages of vertebrates along these watercourses attest to big river habitats across the presently arid lands (Joleaud, 1935).
Similar situations of ancient, now dry valleys, or remnants of terraces produced by streams larger than those now present, occur in many other
desert regions (Leopold and Miller, 1954; Wright, 1956; Alimen, 1965).
In some instances, remaining waters of such extinct systems support
remnant fish populations, reconfirming drainage connections long since disrupted and obscured (Pellegrin, 1914, 1931, 1934; Braestrup, 1940;
Miller, 1946a, 1949a; Hubbs and Miller, 1948a).
Man's activities over the past few decades in North America—and far
longer in some other deserts—have accelerated desiccation of streams, especially larger ones. These have almost all been beheaded by impoundment
and their waters spread over lowlands to irrigate domesticated crops.
At the time of settlement, large rivers such as the Colorado and Gila
of the Sonoran Desert had complex marshes, lakes, and sloughs along their
lower courses, as did their major tributaries (Sykes, 1937; Leopold, 1953;
Miller, 1961a). Desert creeks also were far more permanent in their lower
parts. Smaller streams in southern Arizona flowed through broad, marshy
floodplains in multiple channels, or moved by seepage (Hastings, 1959a,b;
Miller, 1961a; Hastings and Turner, 1965; Minckley, 1969a). Dark, organic soils form parts of the walls of now dry washes. These soils are remnants of marsh, or ciénega deposits (Martin and Mehringer, 1965).
Riparian galleries of cottonwood (Populus), and other, broad-leafed, deciduous trees, so conspicuous at present along some desert waterways
(Lowe, 1967), may well have been restricted to places of relatively good
drainage in the recent past. Areas of entrenched channel doubtless were
present (Hastings, 1959a,b), bordered by mesquite (Prosopis) and forming
bosques on terraces. Under such conditions stream bottoms are typically
of mud and organic debris, except in places of channel degradation or the
presence of stony dikes. Currents are slow, impeded by debris and the
tortuous nature of channels, and flooding effects are minimized (Miller, 1961a; Minckley, 1969a). The present aspect of many streams in
southwestern United States, with broad, sandy, braided channels, may
therefore be a disclimatic phenomenon.
A trend toward erosion of channels originated in the 1880's with a cycle
of arroyo cutting that may be continuing today (Miller, 1961a; Hastings
and Turner, 1965). Similar erosive events were recorded, at about the
same time, over a large geographic area, and the relative uniformity
"makes it difficult to link settlement. . . with arroyo cutting, the uniform
o n s e t . . . on the other hand, points toward operation of a broad regional
factor like climate" (Hastings and Turner, 1965). However, in southern