Historical Changes in Streamflows, Channel Morphology, and Riparian Vegetation of the Rio Grande Downstream of Brownsville, Texas
The Rio GrandefRio Bravo drains an area of more than and northern New Mexico; the vast Chihuahuan Desert of 440,300 square kilometers of Mexico and southwestern southern New Mexico, northern Mexico, and southwestern United States (Bartlett. 1984). The Rio Grande flows for Texas; ami the subtropical lower valley of southern Texas 3,000 kilometers from its headwaters in the San Juan Moun (fig. I ). tains of southern Colorado to the Gulf of Mexico down stream of Brownsville, Texas. The "Rio," as it is often Mean annual flow in the Rio Grande ranges from 6.0 mil call ed, drains the southern Rocky Mountains of Colorado lion cubic meters in th e snowmelt-dominated segments of
EXPLANATION
llJ3' f'\ FW MEXfC() ... Stream fl ow-golging station 1 Ri o Crande al Fori Quitman 2 Rio Grande 31John son Ranch 3 Rio Crande n('i1 r Brownsville
MEXICO
" COLORADO I If, \ , \/lJ,L
50 100, KilOMETERS
LOCATION MAP Figure 1. Rio GrandefRio Bravo Basin and location of sel ected streamflow-gaging stations and segment downstream of Brownsville, Texas. u.s. Department of the Interior USGS Fact Sheet 074-00 U.S. Geological Survey July 2000 2,000 1,800 0 1,600 Rio Grande at Fort Quitman Z 1,400 0 1,200 Uw 1,()()() (f) 800 a: 600 w 4()() a. 200 (f) 0 a: ·200 W t-w 2,()()() ::; 1,800 Rio Grande at Johnson Ranch near Castolon 1.600 U 1,400 iii 1,200 => 1,()()() U BOO ;::: 600 I W 400 (!J 200 I I J ,I, '- /I • .,d, A I 11 a: 0 J_Jl.L ,I.IL ~ •. ~ l J. , J ~." I}(I, « -200 I U 2,000 (f) 1,800 Rio Grande near Brownsville is 1,600 Z 1,400 « 1,200 W 1,000 ::; BOO >- 600 ' III, Completion of Falcon International Darn/Reservoir -'« 400 1 , 0 200 II I ' , II I (\ 0 l I' 1,- _!J~ A-.. ..c ·200 0 0 0 0 0 0 ...... 0> 0> 0> '"0> 0> 0>"' 0> "' 0> "' 0> "'0> '" "' "' '" '" " " "' "'~ '" '"~ '" '" '" '" '" Figure 2. Hydrographs showing daily mean streamflow for three gaging stations on the Rio Grande, the river in northern New Mexico to 22 million cubic meters plain of the river. and numerous drainage ditches return this near its mouth at the Gulf of Mexico, The Rio Grande is water, often more sal ine, to the main channel of the river. sometimes thought of as a river that has been disconnected The International Boundary and Water Commission in the middle. Downstream of EI Paso, between Fort (mWC) has undertaken channelmodificalion of the Rio Quitman and Presidio, Texas, the river can be totally dry, Grande over much of its length between Elephant Butte Just upstream of Presidio, the RIO Conchos, which drains Reservoir and Fort Quitman. Between EI Paso and Fort 68,400 square kilometers of the Sierra Madre Occidental Quitman, the river has been straightened from 249 to 142 of Mexico. gives new life to Ihe river. As much as 75 perce11l kilometers in lenglh. of the flow of the Rio Grande downstream of Presidio comes from the Rio Conchos, About 10 million people live in the Rio Grande Basin; the fastest growing areas are the EI Paso and Ciudad Juarez The Rio Grande downstream of Albuquerque. New metropolitan area and the McAllen to Brownsville corridor Mexico, is a regulated river. Four major reservoirs impound in the lower Rio Grande Valley of South Texas (Texas Natu the Rio Grande; Elephant BUlle and Caballo Reservoirs in ral Resource Conservation Commission, 1996). In 1956, southern New Mexico and Amistad and Falcon International Mexico initiated the Border Induslrialization Program, often Reservoirs along the Texas border with Mexico. These res referred to as the maquiladora program. This program, ervoirs, which were construe ,ed for water storage and flood which provides incentives for fo reign companies to con control, have provided a continuous source of water for irri struct factories in Mexico, has fostered the establishment of gation, allowing year-round farming in the Rio Grande Val over 900 maquiladoras along the inlernalional border ley of Mexico and the United States. Irrigation water for between Texas, New Mexico, and Mexico. agriculture accounts for 80 percent of surface-water use in the U.S. part of the river basin (Texas Natural Resource Population growth and agricultural and industrial devel Conservation Commission, 1996). The 1906 Allocation opment along the main stem of the Rio Grande, particularly Treaty ensures that the United States will deliver 74 cubic the segment of the river that forms the international bound hectometers of irrigation water annuall y to Mexico by way ary between Texas and Mexico, has caused changes in of International Dam between EI Paso and Ciudad Juarez waler-quality conditions and in historical flows of the river. (Meuller, 1975). The J11lernational Dam is the major flow Changes to a river's flow patterns, and more specifically, diversion structure on the Rio Grande. However, many changes in the timing and magnitude of a river's peak or smaller diversion dams carry water to crops in the flood- higher flows have important consequences for channel
2 1,000 1.000 I
0 1936-50 (pre-Falcon Reservoir) 1956-95 (post-Falcon Reservoir) Z 0 900 900 ' U w VJ a: w a. 800 800 VJ a: UJ f- UJ 700 I 700 , "u iii ::J U 600 ~ 600 ~ ui a: '" 500 I 500 I U'" VJ 0 Z 400 f 400 '"UJ
".J>- <' 300 0 300 '
200 200 '
100 100
0 0 0 0.1 0.2 0.3 0.4 0 .5 0.6 0.7 0.8 0.9 1.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
PROPORTION OF DISCHARGE MEASUREMENTS Figure 3. Distribution of daily mean discharge at the Rio Grande near Brownsvi lle before and after construction of Falcon International Reservoir.
features and riparian zone vegetation (Leopold. 1994). kil ometers downstream of the International Bridge between These changes in channel and ripari an zone features will in Brownsville and Matamoros, Mexico. An equivalent period tum influence the makeup and status of riparian and aquatic of streamflow record, 1936-98, was selected to characterize biological communities. This report presems the results of streamflow at each station. an assessment of changes in general streamflow pallems at selected locati ons on th e Rio Grande between EI Paso and Daily mean streamflow at th e FOlt Quitman station did Brownsville and of changes in channel and riparian features not exceed 200 cubic meters per second during 1936-98 downstream of Brownsville that are thought 10 have been caused by changes in streamflows. (fi g. 2). This is partially a result of upstream reservoirs in New Mexico regu lating flows. In contrast, daily mean Historical Changes in Streamflows in the Rio streamflow is as great as 1,800 cubic meters per second at Grande th e Johnson Ranch station. which is downstream of the con fluence with the Rio Conch os (fi g. I). Flows at the Johnson Ranch station are more seasonal than flows at th e Fort Quit Three locations (fi g. I) on the Rio Grande between EI Paso and Brownsvi ll e were selected to investi gate changes man station. with increased flows in latc summcr corre in historical streamflows. The streamflow-gaging stations sponding to the onset of monsoonal rains in the Chihuahuan operated by the I B WC at each of these locations have been Desert. These seasonal flows refl ect the flow of the Rio monitoring dail y mean flows since the early 1920s or 1930s. Conchos, which is less regulated th an the Rio Grande above These stati ons are Rio Grande at Fort Quitman, 2.4 river EI Paso. The streamflow record for the station just down kilometers downstream of old Fort Quitman; Rio Grande at stream from Brownsville (fi g. 2) shows that the variability in Johnson Ranch near CaslOlon, 2.2 rive r kilometers upstream flows decreased substantially after the completion of Falcon of the old Johnson Ranch headquarters in Big Bend International Reservoir in 1958. In particular, the proportion National Park; and Rio Grande near Brownsville, 10.9 river of flows greater than 100 cubic mete rs per second decreased
3 1950 1995
Figure 4. 1950 and 1995 aerial photographs of part of segment of Ihe Rio Grande downstream of Brownsville, Texas. from a prop0l1ion of 0.3 or 30 percent during 1936-50 to a (scale I :20,000) taken in 1950 by the Soil Conservation Ser proportion of 0.1 or 10 percent during 1958- 95 (fig. 3). vice (now the Natural Resources Conservation Service) and U.S. Geological Survey digital orthophoto quadrangles The segment of the Rio Grande from downstream of (DOQs) derived from 1995 National Aerial Photography Falcon Reservoir to th e mouth of the Rio Grande at the Program aerial photographs (scale I :40,000) were used to Gulf of Mexico is bordered by tracts of undeveloped ripar compare channel and riparian features (fig. 4). The 1950 ian and floodplain lands that are part of the U.S. Fish and aerial photographs were rectified to match the scale of the Wildlife Service (USFW) complex of wildlife refu ges and 1995 DOQs for comparison. The difference between daily the National Audubon Society Sabal Palm Preserve. These mean discharge when the 1950 and 1995 aerial photographs lands contain some of the last stands of native riverside veg were taken was only 0.5 cubic meter per second: the effect etati on such as the sabal palm. However, these native stands of this difference on channel features in the photographs, at are being displaced in many areas by more aggressive exotic this scale, is thought to be negligible. species such as African bermuda and salt cedar (Larry Ditto, U.S. Fish and Wildlife Service. oral commun .. 1999). Various channel and vegelalive-cover measurements Changes in streamflows in the Ri o Grande, specifically the were made to assess changes along the Rio Grande during lack of higher out-of-bank flows. can be an important cause 1950-95 (table I). The vegetation score for point bars was of the encroachment of these exotic species. based on assigning values from 1.0 to 5.0 according to the area of the bar covered by vegetation. For example, a vege Historical Changes in Channel and Riparian tation score of 1.0 indicates vegetation coverage of 0 to 25 Features Downstream of Brownsville, Texas percent of the bar, and a score of 5.0 indicates coverage of 75 to 100 percent. These measures were determined using A segment of the Ri o Grande downstream of Brownsville geographic information system (GIS) software. (fig. I) was selected to characteri ze hi storical changes in channel and riparian features that might have been caused The largest changes in channel and liparian features in by changes in stream flows. The segment extends about 38 the segment are in liparian vegetation, point-bar area and kilometers fro m just upstream of the IBWC gaging station vegetation, and wetted channel width (fig. 5. table I). near Brownsville to just downstream of the Palmito Hill Between 1950 and 1995, the frequency of breaks in riparian Civil Wa r historical site (inset, fi g. I). Aerial photographs vegetation increased by an average of 35 percent; however,
4 the length of these breaks increased by an average of more to the 1995 aerial photograph for a part of the segment stud than 3UO percent. The average width of riparian vegetation ied (fig. 4). Point bars in th e 1950 photograph are primarily decreased by about 57 percent on the U.S. bank and about unvegetated sand. The same point bars in the 1995 photo 47 percent on the Mexican bank. An increase in th e number graph have little or no sand visible and are largely covered and extent of breaks in riparian vegetation might have been with vegetation. The channel in the 1995 photograph is visi caused by land-use practices including the clearing of vege bly narrower and is more simplified Ulan the channel in the tation to increase land available for cultivation. However, a 1950 photograph. decrease in the intensity and frequency of hi gher flows might have caused a reduction in recharge to the alluvial Many native species of riparian vegetation require sea aquifer, thus lowering the alluvial water table, and similarly sonal flooding to genninate, and the periodic flooding of the reduCing the density and extent of riparian vegetation. riparian zone affects soil chemistry through the import and removal of organic matter, in the reple nishment of mineral The average point-bar area increased by II I percent. and resources, and in maintaining a large flux of energy and the average score of vegetation coverage of the same point nutrients in th e riparian zone (Mitsch and Gosselink, 1993). bars increased by 230 percent. The average wetted channel Riparian zone vegetation provides important nesting and width decreased by 32 percent. A decrease in the frequency feeding habitat for resident and mi gratory birds and often of higher river flows (such as out-of-bank flows) and the provides the only substantial cover for other wildlife, partic resulting increase in rooted vegetation are probably the ularly in more arid regions. This report provides evidence of cause of channel narrowing and the stabilization of point a decreasing trend in higher flow events (which might have bars. caused the narrowing of the channel) and increased vegeta tive coverage leading to more stable channel bars and a less These changes in channel and riparian features in this continuous riparian buffer zone paralleling the river in this segment of the Rio Grande downstream of Brownsville are segment of the Rio Grande downstream of Brownsville, further illustrated by comparing the 1950 aerial photograph Texas.
Table 1. Channel and riparian features for th e segmen t of the Ri o Grande downstream of Brownsville, Texas 2 1m, meters, m • sq u<.tre meters I
Feature 1950 1995 Percent change
Linear segment length ( 111 ) 19,193 19,18 1 -0.063
Curvi lin ear segment length (m) 59,323 60.847 2.57
Sinuosity of segment 3.09 3. 17 2.59
Left bank IU .S. llength (m) 57,770 57,922 .263
Right bank IMex ico llength (m) 57.625 58,615 1.72
Length of hreaks in riparian vegetation, left bank I U.S .I (m) 6,536 26.70t 309
Leng th of breaks in riparian vegetation, right bank IM exicol (m) 4,340 17,834 3 t I
Frequency of breaks in riparian vegetation. left bank I U.S.] 43 56 30.2
Frequency of breaks in riparian vegetation, ri ght bank IMexico) 62 87 40.3
Average width of riparian vegetation, left bank [U.S.I (m) 32.4 13.9 - 57. 1
Average width of riparian vegetation, right bank IM exicol (m) 22.5 12. 0 -46.7
Average wetted channel width (m) 55.9 38.0 - 32.0
Ave rage area of point bars (m2) 10,153 2 1.44 1 I II
Average score of vegetation coverage of point hars 1.0 3.3 230
5 36.000 75 3.5 _ Area of point bar _ Wetted channel width _ Point-bar vegetation score 70 _ Frequency of vegetalion breaks I- 30.000 3.0 Z W ::; w
0 0 35 1950 1995 1950 1995
Figure 5. Channel and riparian zone features for segment of the Rio Grande downstream of Brownsville, Texas.
References -J. Bruce Moring and Rita Setser
Banlett, R.A., 1984, Rolling rivers-An encyclopedia of Any use vf trade, product, or finn names is lor descriptil1e America's rivers: New York, McGraw-Hili, 398 p. purposes ollly alld does 1101 imply endorsement by the U.S. Government. Leopold. L.B ., 1994, A view of the river: Cambridge, Mass., Harvard University Press. 298 p. For more information, please contact: Meuller, J.E., 1975, Restless river: EI Paso. Texas Western Districi Chief Press, 153 p. U.S. Geological Survey 8027 Exchange Dr. Mitsch, W,J ., and Gosselink, J.G., 1993. Wetlands (2<1 ed.): Auslin. TX 78754-4733 New York, Van Nostrom Reinhold, 722 p. E-mail: dc_tx @usgs.gov Texas Natural Resource Conservation Commission, 1996, 1996 Regional assessment of water quality in the Rio Phone: (512) 927- 3500 Grande Basin: Border Environmental Assessment FAX: (5 12) 927- 3590 Team, SFR- 041 , 74 p. World Wide Web: http://tx .usgs.gov/
6
-(;, u.s. GOVERNMENT PRINTING omCE: 2000 - 513-089 / 52011 Region No.8