THE BALCONES ESCARPMENT Edited by PATRICK L. ABBOTT and C.M. WOODRUFF, Jr. 1986 THE BALCONES ESCARPMENT Geology, Hydrology, Ecology and Social Development in Central Texas Editors PATRICK L. ABBOTI Department of Geological Sciences San Diego State University San Diego, California 92182 C. M. WOODRUFF, JR. Consulting Geologist P.O. Box 13252 Austin, Texas 78711 Published for Geological Society ofAmerica Annual Meeting San Antonio, Texas November, 1986 For copies of this book please write or phone: PATRICK L. ABBOTT Department of Geological Sciences San Diego State University San Diego, California 92182-0337 (619) 265-5591 Copyright© 1986 by Patrick Leon Abbott COPYRIGHT The papers in this volume were prepared for the Geological Society of America annual meeting in San Antonio, Texas, November 9-14, 1986. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior written permission of the copyright owner. Printed by Comet Reproduction Service Santa Fe Springs, California 90670 124p-USBWF: The twilight emergence of Mexican Free-tailed Bats (Tadarida brasiliensis) from the entrance of Bracken Cave. This is one of nature's most spectacular events. Young bats of this species first fly in about 4-5 weeks. Adults attain speeds of more than 40 miles per hour and may cover hundreds of miles in a single night. Up to 9 million of these bats once inhabited Carlsbad Caverns, but the colony has been reduced to a mere 300,000. The largest colony of Mexican Free-tailed Bats ever known has declined by an alarming 99% in only two decades. These. like most other bats, are highly beneficial, essential to a healthy environment, and deserving of our utmost respect and consideration. Photograph courtesy ofMer/in D. Tuttle, Bat Conservation International. FRONT COVER: Artesian wells, City Waterworks, San Antonio, circa 1897 from photograph in Hill and Vaughn (plate 39. 1897, U.S. Geological Survey Annual Report). Drawing by Margaret Campbell. PREFACE The Balcones Escarpnent is a line of low hills that extends through Central Texas. As pointed out by Fenneman in his classic Physiography of Western United States, this escaq:ment marks the break between two grand physiographic divisions of North Airerica: the Great Plains Province on the west and the Coastal Plains on the east. In Central Texas this major physiographic break is denoted by the change fran Hill Country/F.dwards Plateau uplands on the western side of the escaq:ment to the Blackland Prairie on the east (Fig. 1). The Balcones Escarpnent lies along the major line of dislocation of the BalCXllles fault zone, a series of en echelon mainly down-to-the-coast normal faults. Fault N ~ 1 0 50M1les 0 70 Kilometers displacerients have resulted in Lower Cretaceous lilllestones to the west being juxtaposed against Upper cretaceous claystones, chalks, and marls to the east. The Balcones fault zone is a surface expression of a deep-seated crustal disoontiru.ri.ty. The Ouachita orogen extends soutiward fran the Arruck.le/ouachita juncture in Oklahoma beneath Central Texas to the Rio Grande where the orogen is displaced laterally into Trans-Pecos Texas. The Ouachita c::arplex forms a hinge between the stable continental interior and the still-subsiding Gulf Coast basin (Fig. 2). Balcones faulting was probably a result of periodic adjustments across this buried hinge. EXPLANATION E:f=:=:4 Inner (Tertiary) Gulf Coastal Plain ~Blocklond Prairie k<>>lRio Grande Plain l\',/(j Hill Country-Edwards Plateau f~ihd Lampasas Cut Plain -Grand Prairie ~'?l>fJ Eastern Cross Timbers Pn) Western Cross Timbers );:/) Llano Basin Figure 1 Physiographic provinces along the Balcones/OUachita ~end~ Central rrexas.(figure courtesy of Bureau of Economic Geology, The University of Texas at Austin). 'll1e main tectonic events of Balcones faulting are generally thought to have occurred during the Miocene, rut there is considerable evidence that periodic structural adjustments also took place in the Cretaceous. For exanple, mafic alkalic volcanic plugs of Late Cretaceous age occur at the surface and in the subsurface all alcng the Balcones fault zone. The geochemistry and petrology of these igneous rocks suggest that they penetrated the entire crust of the Earth. 1 l I 0 0 7QK1lomelers . ··J . ~ Dramatic changes in the landscape occur across this crustal discontinuity. On the west are plateau uplands and ruggedly dissected limestone hills; soils are thin and stony, and the main agricultural use of the land is for range; the dominant native vegetation assemblage is juniper-live oak savannah; and groumwater is generally of good quality with ample quantities occurring at shallow depths frcm limestone aquifers. On the east side of the escarpnent, terrain consists of rolling prairies and EXPLANATION SURFACE FEATURES ,...A Normol fo~ lt (!1<.:k 1ncl1catrs dONnlhrown s1df' l ••. · UpJrp hm11 of "' t1ory ou11·roo / L1;.id•p f m1 of Cretaceous outcro;> SUBSURFACE FEATURES POST-OUACHITA FEATURES OUACHITA FEATURES // ~v~~~~~·~~~~~1)ll (lefllh on dowrd1p s1d P 1'~J(~~nn~'.~~~~f ~~11~~l~~~c~~l~!:lntarv rocks _] o•bo1100",u" mPtnshulp,elc {oqc unknown) j r t•y1 de, slolri,P-IC (oqe unknown) Figure 2. StrUctural/ tectonic features along the Balcones/Ouachita trend, central Texas (figure courtesy of Bureau of Econanic Geology, The University of Texas at Austin. broad river bottoms; soils are thick fertile clays, which ~prirre arable land; native prairie grasses were dominant before the agricultural activities supplanted them; and groundwater in large aJIK)UI}ts is available only at considerable depth and is oc:rmnnly tepid and brackish. The escarµcent is also a major weather-maker. Maxim.un relief is only a few hundred feet, but it is the first topographic break inland fran the Gulf of Mexico and thus acts as an orographic influence on unstable, water-laden air masses. These combined factors have resulted in the Balcones Eacarµcent being the locus of the largest flood-producing storms in the conterminous United States. The greatest single rainfall event ever recorded in the contiguous U.S. occurred in 1921, when 38 inches of rain fell in 24 hours near Thrall in Williamson County. Physiographic changes have had their impact on human culture as well. As pointed out by Peter T. Flawn in his article "The Everlasting Land" (in A President's Country, Shoal creek Press, 1964) , the Balcones Escarpnent marks the line where the Alrerican West really begins. It marks the boundary between the cotton econany of the Old South and the cattle econany of the Old West. This geocultural break has been the site of many major towns and cities in Texas: Del Rio, San Antonio, Austin, 'l'enple, Waco, and (in its broader geographic context) Dallas/Fort \'brth and Shennan/Denison. The cities are a response to the geologic break in the same way that geologic changes along the Fall Line of the eastern United States created favorable sites for Wustry and oonmerce. People settled along the Balcones Escarprent in order to draw on both ecananies-cotton and cattle. Also, their endeavors were aided by the dependable water supplies provided by the great springs that issue forth along the fault zone. A dramatic expression of the Balcones/0uachita discontinuity is the localization of the F.dwards aquifer. The F.dwards aquifer is composed of karstified groundwater reservoirs that extend along the Balccnes Escarpnent in Central Texas and west beneath the vast F.dwards Plateau. The F.dwards Lirrestone, the main host rock of the aquifer, once extended unbroken across llUlCh of Texas--fran the middle part of the Gulf Coastal Plain to the Panhandle and Trans-Pecos regions. This once­ccntinuous rock unit has been flexed and broken by tectonic events, dissected by erosion, and dissolved by the actions of percolating water. The vagaries of erosion and of structural defonnation have resulted in different hydrodynamic regirres within different geologic/geographic provinces. In areas where the F.dwards Lirrestone crops out, groundwater occurs under water-table conditions. Down dip within the fault zone, a confined zone of groundwater occurs. Farther downdip, water quality changes abruptly across a "bad-water line" that marks the edge of upwelling deep-basin waters. Locally within the Gulf Coast basin, the Edwards Lirrestone is a reservoir rock for hydrocarbons. The Balcones fault zone is where the F.dwards aquifer is JOOSt prolific. There, faults have provided major avenues for directional porosity and perneability, and these conduits have been enlarged by solution so that great voluires of water flCM rapidly from west to east along the escarpnent. In this way, the semi-arid western part of the fault zone provides water for the sub-humid eastern watersheds where the largest springs in Texas occur. It is the purpose of this volume to present nultidisciplinary infonnation on the Balcones Escarpnent. The main focus of these presentations will be on the geology and hydrology along this major break, but there are also papers that treat biological and cultural responses to the geologic features and hydrologic processes in the region. This I1U1ltidisciplinary approach reinforces the message that the manifold resources are interrelated: geology affects landform and soils and plants and animals. Weather patterns affect the geom:irtxtlc and ecological setting as well. H1.111ans have acted in context of what has been established by these natural processes f ran prehistoric tirres to the present. This voluire is aimed at recognizing the controls so that IOOdern humans can live in hallOC>ny with the resources and processes along this borderland. First, we thank all the contributors to this voluire. These men and women have brought their expertise to bear on various facets of this important region. In doing so, we believe that a whole is created that is greater than any individual part. The multidisciplinary approach reinforces the unity of processes-the "big picture"--evident through space and tirre along the Balcones Escarpnent. we especially thank Margaret ~bell for her art work used on the front cover and at various places in the text. We also thank Merlin Tuttle for the use of his photograph of Mexican free-tailed bats emerging fran Bracken cave and Kathy Jessup and Marylou Montross for typing several manuscripts. C. M. \'bodruff, Jr. Patrick L. Abbott Austin, Texas San Diego, california September 1986 Contents FLOODING ALONG THE BALCONES ESCARPMENT, CENTRAL TEXAS ....... . . . .. . ... .. ... ... .. . .. ... . . . . ...........S. Christopher Garan and Victor R. Baker 1 LARGE RAINSTORMS ALONG THE BALCONES ESCARPMENT IN CENTRAL TEXAS .. . .. .. . .. . . . . .. . .. .. . . ... . .. . .. . .. . ... . ... .. ............ . . ... . Raymond M. Slade, Jr. lS PLANT COMMUNITIES OF THE EDWARDS PLATEAU OF TEXAS: An overview emphasizing the Balcones Escarpment zone between San Antonio and Austin with special attention to landscape contrasts and natural diversity .. . .. . .... .David H. Riskind and David D. Diamond 21 EDWARDS HEATH .. . .. ...... . . ... .. .... .. . . .. .. .. .. . .. .. . .. . .. . ... . . .PatrickL. Abbott 33 THE BALCONES FAULT ZONE AS A MAJOR ZOOGEOGRAPHIC FEATURE . .. ................................. . . . . . .. . .. . ... . ........ . .. . .. . . Raymond W. Nick 3S VERTEBRATE PALEONTOLOGY OF THE BALCONES FAULT TREND ... . .... . ................. ..... . . . . .. ... . .... .. ... . ... ..... . .. . Ernest L. Lundelius, Jr. 41 THE BIOTA OF THE EDWARDS AQUIFER AND THE IMPLICATIONS FOR PALEOZOOGEOGRAPHY . . ......... .. ..... . . .. . ... .. ... .... . ..... .. ... Glenn Longley Sl EARLY HUMAN POPULATIONS ALONG THE BALCONES ESCARPMENT . . ......... . . .. ... . .... . ... . .. .. . .. . .. . . ...... .. . ... . Thomas R. Hester SS THE PLEISTOCENT TERRA ROSSA OF CENTRAL TEXAS ... . .. ..... .. . .. .. .. .Keith Young 63 STRUCTURAL STYLE IN AN EN ECHELON FAULT SYSTEM, BALCONES FAULT ZONE, CENTRAL TEXAS: GEOMORPHOLOGIC AND HYDROLOGIC IMPLICATIONS .. ..... ... . . ... ..... . ...... Thomas W. Grimshaw and C.M. Woodruff. Jr. 71 STREAM PIRACY AND EVOLUTION OF THE EDWARDS AQUIFER ALONG THE BALCONES ESCARPMENT, CENTRAL TEXAS .. . . C.M. Woodruff, Jr. and Patrick L. Abbott 77 CA VERN DEVELOPMENT IN THE NEW BRAUNFELS AREA, CENTRAL TEXAS . . .. . . .. . .. ......... . .. . . . . . ... ... ... , ............... . . . ....... . . . . Ernst H. Kastning 91 POST-MIOCENE CARBONATE DIAGENESIS OF THE LOWER CRETACEOUS EDWARDS GROUP IN THE BALCONES FAULT ZONE AREA, SOUTH-CENTRAL TEXAS ... .. . . .. . . . . ...... . . .. . ... .. . . ... . .... . . .Patricia Mench Ellis 101 HYDROCHEMISTRY OF THE COMAL, HUECO, AND SAN MARCOS SPRINGS, EDWARDS AQUIFER, TEXAS . . . . .. .... .. . .. .Albert E. Ogden, Ray A. Quick and Samuel R. Rothermel llS RELATIONS BETWEEN AREAS OF HIGH TRANSMISSIVITY AND LINEAMENTS - THE EDWARDS AQUIFER, BARTON SPRINGS SEGMENT, TRAVIS AND HAYS COUNTIES . .. . . . . . .. . . . .. . .. . . . .. ........ ... ....... . Laura De La Garza and Raymond M. Slade, Jr. 131 GEOTHERMAL RESOURCES OF BEXAR COUNTY, TEXAS ... .. .. ...... .. .. . .... . .. .. . ..... .. ..... . ... .. . . . . Duncan Foley and C.M. Woodruff, Jr. 14S LAND USE AND CULTURAL CHANGE ALONG THE BALCONES ESCARPMENT: 1718-1986 . . ............ . .. ...... . .. . ... ... . . . .... .. . ....... . .. .. . . E. Charles Palmer 1S3 PIPELINE OIL SPILLS AND THE EDWARDS AQUIFERS, CENTRAL TEXAS . . . .. . Peter R. Rose 163 ROADLOG: BALCONES ESCARPMENT ­SAN ANTONIO TO SAN MARCOS, TEXAS .. . .. .. . . ... . . .. . .. . .. . .. . ..... . .. C.M. Woodruff, Jr., Patrick L. Abbott, David H. Riskind 184 COMAL SPRINGS--NEW BRAUNFELS FLOODING ALONG THE BALCONES ESCARPMENT, CENTRAL TEXAS S. Christopher Caran Victor R. Baker Bureau of Economic Geology Department of Geosciences The University of Texas at Austin University of Arizona Austin. TX 78713 Tucson. AZ 85721 A few days before the rains began to fall. a band of Tonkawa Indians that were camped in the river valley just below old Fort Griffin moved their camp to the top of one of the nearby hills. After the flood. on being asked why they moved to the top of the hill . the chief answered that when the snakes crawl towards the hills. the prairie dogs run towards the hills. and the grasshoppers hop towards the hills. it is time for the Indian to go to the hills. (Oral testimony attributed to an unnamed weather observer in Albany. Texas. following a memorable flood on the Clear Fork of the Brazos River in the late 1870's: recounted by Vance. 1934. p. 7.) ABSTRACT High-magnitude floods occur with greater frequency in the Balcones Escarpment area than in any other region of the United States. Rates of precipitation and discharge per unit drainage area approach world maxima. The intensity of rainstorms is compounded by rapid runoff and limited infiltration. producing episodic flooding. Effects of urbanization may be superimposed on meteorologic and physiographic factors. thereby increasing flood hazards in metropolitan areas throughout the region. INTRODUCTION The Balcones Escarpment area. comprising parts of the Edwards Plateau. Hill Country. and northern and westernmost Coastal Plains (fig. 1). is one of the most severely flooded regions of the United States (Leopold and others. 1964. fig . 3-16: Baker. 1975: Beard. 1975. fig . 13; Crippen and Bue. 1977. fig. 12. table 1: Patton and Baker. 1976. p. 945. fig. 5). Floods of record include the catastrophic 1954 inundation of the lower Pecos River valley where peak instantaneous discharge approached 1.000.000 cubic feet per second (cfs). or more than 600 billion gallons per day (International Boundary and Water Commission. 1954). This reach of the Pecos normally is an intermittent stream completely dry for several months each year. But during the 1954 event. its rate of discharge was more than 1 1/2 times mean flow of the world's third longest river. the Mississippi (table 1). What's more. only part of the Pecos drainage basin had received significant rainfall and provided runoff: the contributing area was less than 0.3 percent of the Mississippi's watershed. The 1954 Pecos River flood was a remarkable occurrence. estimated to represent the 2.000-yr recurrence interval flood (0.05 percent yearly-probability flood) in that basin (Kochel and others. 1982. p. 1179). This and other major discharge events are easily and instructively compared by examining the ratio of peak ~scharge (in cfs) to contributing drainage area (in mi ). During the 1954 flood of the Pecos. this ratio was approximately 261 :1. compared to 0.5:1 for mean discharge of the Mississippi (table 1). Although the rate of peak discharge of the Pecos was exceptional. floods yielding comparable discharge:drainage area ratios have been recorded in most drainage basins and in Abbott. Patrick l. and Woodruff. C.M .. Jr., eds.. 1986. The Balcones ~t,Central Teau: Geological Society of America. p. 1-14 subbasins in Central Texas. Intense ramstorms over small watersheds throughout the region have produced numerous examples of discharge in excess of 100.000 cfs. Flooding of this magnitude exacts a heavy toll from area residents who incur the high cost of flood-control structures on trunk streams (fig. 1). but also sustain casualties and damages associated with floods on small. unregulated or under­regulated tributaries. CAUSE OF MAJOR FLOOD EVENTS Baker (1975: 1977) and Patton and Baker (1976) described a number of factors that contribute to flooding in the Balcones Escarpment area. Principal among these are: (1) the intensity of sporadic rainstorms. particularly those associated with incursions of tropical storms and hurricanes: and (2) the high-percentage yield and rapidity of runoff from the steep bedrock slopes that characterize much of the region. (NOTE: Meteorologic conditions in the Central Texas region are discussed in greater detail in another section of this guidebook and in references cited here.) To these factors may be added the many drainage problems inherent in urban areas including large municipalities along the Balcones Escarpment (fig. 1). Although not unique to Central Texas. the role of urbanization in flood enhancement is especially significant when superimposed on adverse characteristics of the natural environment of this region . Meteorologic Factors Easterly Waves The climatic provenance and topography of Central Texas. and its proximity to the Gulf of Mexico. combine to make the incidence of torrential rains in the area extremely high. The region lies within the zone of convergence of polar air masses and easterly waves (Orton. 1966. p. 10­11). Polar air is characterized by cool temperatures. high pressures. and low moisture. Easterly waves. which are westward-moving troughs of low pressure. convey warm. moist air of tropical origin. When a well-developed easterly wave approaches a lobe of high-pressure. such as that associated with a strong polar surge into middle latitudes. pronounced instability and heavy rains may result. N / \ +31°N 99°W + + EXPLANATION + 29°N + 97°W • City N ~ Lake ""'-. River or stream 0 20 40 mi ~ 0 25 50 km ""' Major drainage divide Figure 1. Balcones Escarpment area. Central Texas. Only major streams and those mentioned in text are named. Relief across the Balcones Escarpment varies from 100 to 500 ft. Table 1. Representative flood discharge of Central Texas streams compared with mean discharge of some of the world "s great rivers (A) (B) Draina!e Discharge Ratio 23 River/stream area (x 10 mi} (x 10cfs} B:A Amazon1 2.722.a 4.200b 2:1 2 Nile1.293.a 110b 0.1:1 Mississippi-Missouri3 1.243.7a 620b 0.5:1 Texas (Source: International Boundary and Water Commission. 1954; Crippen and Bue. 1977; Schroeder and others. 1979; Moore and others. 1982) Pecos (U.S. Hwy 90. 1954) 3f 967d 261:1 Little (Cameron. 1921) 7.1c 647d 91:1 North Prong of Medina (Medina. 1978) 0.07c 123d 1.800:1 Medina (Pipe Creek. 1978) 0.5c 281d 600:1 Guadalupe (Comfort. 1978) 0.8c 240d 300:1 Guadalupe (Spring Branch. 1978) 1.3c 158d 122:1 Seco (D'Hanis. 1935) 0.14c 230d 1.500:1 Walnut (FM Hwy 1325. 1981) 0.01c 15d 1.500:1 1 World's largest drainage area and discharge; second longest2 World's longest; fourth largest drainage area; tenth largest discharge 3 World's third longest: fifth largest drainage area and discharge (Source: National Oceanic and Atmospheric Administration. 1971) a Entire basin b Mean discharge at mouth c Contributing portion of drainage area d Flood discharge at point of measurement This combination is comparatively uncommon but has produced extremely heavy rains and associated flooding. The most severe rainstorm ever recorded in the continental United States occurred under these conditions on September 9 and 10. 1921. in Thrall. Williamson County (Jennings. 1950: Bowmar. 1983. p. 69) (fig. 1.2). A total of 36.4 inches of rain fell in 18 hr. which is the world's record for this period. The 24-hr total was 38.2 inches. exceeding in one day the expected precipitation of an entire year (Larkin and Bowmar. 1983. p. 18). At the town of Cameron. Milam County. a few miles northeast of Thrall. peak discharge of the Little Ri~1r was 647.000 cfs from a drainage area of 7.088 mi (Crippen and Bue. 1977. table 1) (figs. 1. 3: table 1) . This storm. which spread over a large area of Central Texas. produced 215 deaths and 19 million dollars in property damage (Bowmar. 1983. table E­3). Orographic Effects The easterly wave that produced the Thrall storm of 1921 was augmented by topographic conditions in the region. Relief across most of the Balcones Escarpment ranges from 100 to 500 ft (fig. 1 caption). Warm. moisture-laden air from the Gulf of Mexico is pushed northward across the gently sloping Coastal Plains by dominant southerly winds. As these winds encounter the escarpment they rise abruptly to higher altitude. If the Gulf air is nearly saturated at lower elevations. rainstorms may occur along the escarpment because of orographic cooling of the air mass. The climate of Central Texas is semiarid; drought years offset wet periods. thereby reducing mean annual precipitation. But cumulative rainfall increases sharply at the escarpment compared to adjacent regions (fig. 4). and rains may be extremely intense for periods of 1000 Note: Regression line and equation pertain to original data set analyzed by Baker (1975) 100 R = 15.3 oOASG .7 6 i ~. = ~ ~· 5 6 c 5 0 a: 10 e 1 Central Texas (see caption) e I e ~ Other sites worldwide 10 60/1 3 6 12 24/1 5 30/1 6 24 min hr day mo Duration Figure 2. Magnitude-duration relationships of the most intense rainstorms in Central Texas and the rest of the world. Sites in Central Texas: (1) Trough D 0 Creek. 1973: (2) Austin. 1981; (3) New Braunfels. 1973; (4) Hanis. 1935: (5) Thrall. 1921; (6) Voss Ranch. 1978; (7) Manatt Ranch. 1978. Adapted from Baker (1975. fig. 2). Additional data sources: Hansen (1979); Massey and others (1982. table 2); Moore and others (1982. figs. 2.2. 2.4); and sources cited by Baker (1975. fig. 2). hours to days over small areas (Carr. 1967. p. 20-21; Bowmar. 1983. p. 56). An astonishing example of this orographic effect is the storm of May 31. 1935. near D·Hanis. Medina County (Jennings. 1950: Morgan. 1966. p. 37. 40) (fig. 2). A total of 22 inches of rain fell in just 2 hr and 45 min . which is the world-record precipitation for that period. At a point a few miles abo~ D0Hanis. Seco Creek has a drainage area of only 142 mi yet briefly discharged at a rate of 230.000 cfs (Crippen and Bue. 1977. table 1) (table 1: fig . 3) . Tropical Disturbances Tropical storms and hurricanes are regular seasonal occurrences over the warm waters of the Caribbean and Gulf of Mexico. Their paths do not often extend far inland but occasional storms penetrate well into the interior of the state and beyond. Some of the Central Texas region· s heaviest rainfalls are products of these events. A recent example is tropical storm Amelia. which produced catastrophic flooding throughout the area in August. 1978. The largest three-day total rainfall ever recorded in the United States occurred on the Manatt ranch. Medina County. where more than 48 inches of rain fell during the period August 1 to 3 (Hansen. 1979) (fig. 2). Near this ranch . on the North Prong of the Medina River. peak di:'fharge was 123.000 cfs from a drainage area of 67.5 mi (Schroeder and others. 1979. p. 6) (table 1; fig. 3). Farther downstream. discharge of the Medina River near Pi~ Creek was 281.000 cfs from a drainage area of 474 mi (Schroeder and others. 1979. p. 111). Medina Lake near San Antonio overflowed its spillway as storage increased by 93.000 acre-feet in 35 hr (Schroeder and others. 1979. p. 6) . Flood stages at 13 stations exceeded previous records and/or projected stages of floods with recurrence intervals greater than 100 yr (Sullivan. 1983. P. 47) . Physiographic Factors and Urbanization Climatic factors control precipitation but once rain reaches the ground it is the character of the land itself that controls runoff. The Balcones Escarpment area has steep sparsely-vegetated slopes. narrow valleys. thin upland soils on limestone bedrock. and. in the Coastal Plains. soils with low infiltration capacity (Baker. 1975. 1976. 1977; Patton and Baker. 1976) (fig. 5). Each of these factors increases runoff and. therefore. discharge per unit drainage area. Development practices in metropolitan areas also tend to increase runoff but may reduce flow through urban stream channels. as well. Urbanization generally increases: (1) impervious cover (that is. the areal extent of roofs. parking lots. and roadways that reduce infiltration): (2) channel rectification (reduces channel storage thereby increasing discharge farther downstream): (3) channel obstruction (causes damming behind bridge abutments. low­water crossings. waterside recreational facilities. etc.): and (4) floodplain development (inhibits high-water throughflow) (Leopold and others. 1964: Costa. 1978: Morisawa and LaFlure. 1979: Rahn. 1984). Espey and others (1966) demonstrated that land-use practices alone can increase Central Texas peak flood discharges by as much as 300 percent. 1 ·~----=* ~2 • t:. 19 .20 18 e 15 16 (/) r '­0 .c. u .'!J "" t:. ........t:. / /4 ~/· 6• "O ~ 0 Cl) CL 10,000 t:// /~2 .3 /~ t:, /,..? Nationwide trend line e 1 Central Texas (see caption) t:. Other U.S. states I (Hoyt and Longebein, 1955) t:. / • / 1000+-----------.--------,..------------.---~----.-----------r 0.1 10 100 1000 10,000 Drainage area (mi2) Figure 3. Discharge-watershed relationships of the most severe floods in Central Texas and other U.S. states. Sites in Central Texas: (1) Trough Creek near New Braunfels. 1972: (2) Bunton Creek at Kyle. 1936: (3) Walnut Creek at Austin. 1981: (4) Little Red Bluff Creek at Carta Valley. 1948: (5) Calaveras Creek near Elmendorf. 1946. Blieders Creek near New Braunfels. 1972. and Spring Creek near Fredricksburg. 1978: (6) Purgatory Creek near San Marcos. 1972: (7) Sink Creek near San Marcos. 1972; (8) North Prong of Medina River near Medina. 1978: (9) Mailtrail Creek at Loma Alta. 1948: (10) Guadalupe River at New Braunfels. 1972: (11) Hondo Creek near Hondo. 1919: (12) Seco Creek near D'Hanis. 1935: (13) West Nueces River near Kickapoo Springs. 1935: (14) Medina River near Pipe Creek. 1978: (15) Guadalupe River at Comfort. 1978: (16) Guadalupe River near Spring Branch. 1978: (17) West Nueces River near Brackettville. 1935; (18) Pedernales River near Johnson City. 1952: (19) Nueces River below Uvalde. 1935; (20) Devils River near Del Rio. 1932; (21) Pecos River at U.S. Highway 90. 1954: (22) Little River at Cameron. 1921. Adapted from Baker (1975. fig. 4) and Crippen and Bue (1977. figs. 2. 12). Additional data sources: Schroeder and others (1979. p. 11): Massey and others (1982. table 1): and source cited by Baker (1975. fig . 4) . Urban flooding is a serious problem in many Central Texas communities (Baker. 1975. 1976). For example. within the small. largely rural Guadalupe River basin. the Federal Emergency Management Agency has designated 17 cities with significant flood hazards (Texas Department of Water Resources. 1984. p. 111-18-6). Annual flood losses throughout the Balcones Escarpment area remain high despite a network of flood-control structures (fig. 1). During the "Memorial Day" flood of May 24 to 25. 1981. the city of Austin sustained 13 deaths and 35.5 million dollars in damages from flooding along small unregulated urban streams (Moore and others. 1982. p. 15). In response. the city constructed several discharge-retention dams and completely revamped its procedures for assessing flood hazards and issuing warnings. But although this system may reduce future casualties and property losses it represents a significant infrastructural investment that few area communities could make. Better planning at an earlier stage of urban development might have prevented foreseeable problems experienced during the 1981 flood and eliminated costly retrodesign. Urbanization merely compounds the natural tendency of Central Texas streams to produce damaging floods with greater frequency than do comparable drainage basins elsewhere. But the causes and effects of flooding in rural and urban settings differ in important ways. Two case studies. one concerning an undeveloped stream reach. the other an area undergoing urban growth. are reviewed in order to assess these differences. CASE STUDIES Rural Flooding: Guadalupe River. 1978 A striking example of flooding in a rural watershed is the August. 1978 event on the upper Guadalupe River. which was associated with the deep inland incursion of tropical storm Amelia. Amelia's climatic history was described in detail by Bowmar (1978. 1979. 1983) and the National Weather Service (1979) . One of the most severe droughts in more than 20 yr was underway just prior to the advent of this storm. A subtropical ridge of high pressure had maintained dry conditions across much of the state throughout the summer. This ridg did not begin to deteriorate until the end of July when tropical storm Amelia formed in the Gulf of Mexico less than 50 mi off the southernmost Texas coast. Amelia was a minimal tropical storm (technically an "extratropical storm" because it originated north of the Tropics) when it made landfall in South Texas. causing little damage along the coast. But as the storm moved northwestward. eventually crossing the Balcones Escarpment near San Antonio. it began producing extremely heavy rains. Amelia followed a path "virtually unique in Texas' weather" (Bowmar. 1979. p. 29). This slow-moving storm drifted over the escarpment and eastern Edwards Plateau. inundating small drainage basins. Rains exceeded 10 inches in 48 to 72 hr across a large area of Central Texas. The heaviest rains were those at the Manatt ranch near Medina. Bandera County. which set. the U. S. 3-day rainfall record of more than 48 inches (Hansen. 1979) (fig. 2). Amelia remained a significant cyclonic system for six days following landfall. producing very intense rains all along its track into North­Central Texas. Flooding associated with tropical storm Amelia was severe. Records of flood discharge in the Medina River basin are summarized above (under "Tropical Disturbances") . For a more complete discussion see Schroeder and others (1979). Sullivan (1983). and Baker (1984) . Remarkable stage heights and discharge peaks were attained on the upper Guadalupe River. as well. At Comfort. Kendall County. water level rose to nearly 41 ft. breaking the previous record established in July. 1869 (Schroeder and others . .;.979. p. 106). Drainage area at this location is 838 mi and peak discharge was 240.000 cfs (table 1: fig. 3). Farther downstream. the U. S. Highway 281 bridge was flooded even though it stands 59 ft above stream bed (Bowmar. 1983. p. 52). Near Spring Branch. Co'2'al County. where the contributing drainage area is 1.315 mi . stage height was greater than 45 ft. However. discharge in this reach had attenuated to 158.000 cfs (Schroeder and others. 1979. p. 107) (table 1: fig. 3). Even this figure is phenomenal: 158.000 cfs is substantially greater than mean discharge of the Nile at its mouth. And the upper Guadalupe has only 1/1000th the Nile's watershed area. And yet. the Amelia flood was only the third largest recorded at the Spring Branch station. The highest stage. observed in 1869. was approximately 53 ft. Figure 5. Block diagram representing geomorphic features that affect flood potential in the Balcones Escarpment area. From Baker (1975. fig. 3). Damage resulting from the Amelia flood was enormous. In Central Texas. 25 persons were killed. 150 were injured. and 50 million dollars in property losses were sustained (Bowmar. 1979. p. 29). Another six persons were killed in North-Central Texas. All flood waters in the upper Guadalupe River watershed were contained by Canyon Lake reservoir in Comal County. Fortunately. lake level was low prior to the storm. Water storage increased by 226.200 acre-feet or approximately 7 4 billion gallons (Schroeder and others. 1979. p. 6). Areas downstream were not subjected to flooding but the lake afforded no protection of sites higher in the basin where rains were heaviest. Geomorphic effects of the flood were pronounced. Devegetation. channel and flood-plain scour. large-scale deposition. modification of channel form . and temporary avulsion of meanders were common. Along both the Guadalupe and Medina Rivers. riparian woodlands including bald cypress trees six feet in diameter were scoured from miles of channel. Sullivan (1983. table 8) estimated 62 to 92 percent reduction of tree-crown cover in some reaches of the Medina. Van Auken and Ford (in preparation) will present a detailed account of effects of the Amelia flood on plant communities along the upper Guadalupe. Baker (1977. p. 1069-1070) discussed the dynamic relationship between riparian vegetation and hydrologic characteristics of channels in flood. This discussion serves to illustrate effects of the Amelia flood on the upper Guadalupe River. Baker's model notes that dense stands of woody plants typically occupy the lower terraces. channel margins. and even the point bars of area streams. As water level begins to rise during a flood . the irregular floor of the low-flow channel is submerged. Boulders and bedrock outcrops that obstruct ba~e flow are completely covered. which reduces resistance or channel roughness. With further increase in depth the stream, now "bankfull". overtops the sinuous low-flow channel. lowest terraces. and vegetated bars. Plants below the level of inundation increase roughness and tend to retard flow. but stream velocity actually increases in response to heightened discharge as the flood crest advances . Within the constricted bedrock channels common to this region. increased discharge is accommodated by rapid increase in stream depth. At this point. the mid-water zone of maximum velocity. the thalweg. shifts laterally inward across the slip-off bars. thereby increasing the effective channel radius and straightening the flood course around meanders. Transition to the next phase of stream flow is governed by a critical threshold that in turn is dependent on the height and density of vegetation. If plants do not choke the flood channel. and if tree canopies remain above water there may be little additional damage. However. if canopies are submerged or flow is greatly restricted trees are uprooted. toppled. or sheared by the force of the water and impact of transported debris. Partial clearing of the channel reduces drag and increases local flow velocities. Rapid flow around remaining obstructions creates macroturbulence. causing intense scouring of gravel bars and low terraces at peak discharge. The coarsest sediment. including boulders and megaboulders. is transported only a short distance. Chute bars and gravel berms form at the downstream ends of bends on which scour was initiated. Valley-bottom scour is selective. partly because the combined resistance of the gravel fill and anchoring vegetation is variable. Following peak stage. as flow subsides. dragged and floated vegetation is deposited in the stream bed where it may inhibit waning discharge. Figure 6. Aerial photographs of a meander bend of the upper Guadalupe River. Comal County. Tops of photos is north. Drainage is from west to east. Maximum east-west diameter of bend is approximately 2.250 ft. (A) U.S. Department of Agriculture. vertical black and white. BQU-UJ-47, October 31. 1969. (B) General Land Office of Texas, vertical black and white. 1-2-114. November 29. 1978. (C) U.S. Department of Agriculture. vertical black and white. 40-48091-180-19. A variation from Baker's model occurred in the the model. the thalweg shifted radially inward across the Guadalupe basin approximately eight miles upstream from slip-off bar. However. this shift completely cut off the the U. S. Highway 281 bridge in western Comal County. neck of the meander. Figure 6B shows scoured chutes. Water depth in this area exceeded 50 ft . As predicted by chute bars. large-scale gravel ripples. and aligned fallen trees at the cutoff. Peak flow bypassed this bend entirely. Consequently. this was one of the few reaches that sustained no serious loss of riparian vegetation such as large bald cypress and other trees. In fact. slack-water deposits began filling this channel segment at the same time contiguous areas were being scoured. A dam of fine sediment and plant debris temporarily blocked the mouth of Honey Creek. a tributary entering this bend from the south. Effects of high-magnitude. low-frequency floods are much greater and more enduring in the bedrock-channel streams of Central Texas than in fine-grained alluvial channels of humid regions. Wolman and Miller (1960) have shown that. in stream systems of the latter type. relatively frequent low-magnitude events are most significant. In contrast. post-flood monitoring of Elm Creek (Comal County) and other streams in the Balcones Escarpment area indicate hydrologic characteristics typically are affected for years and perhaps decades (Baker. 1977). A sequence of aerial photographs (fig. 6) shows the avulsed meander bend of the Guadalupe before. shortly after. and two years after the Amelia flood. Evidence of older (pre-1969) cutoffs at yet higher elevations attest to the episodic nature of these events (fig. 6A). Following the 1978 flood , the river occupied a deep. sharply-defined base-flow channel against the cutbank (fig. 6B). This channel cuts through gravel bars that in 1969 nearly blocked the river at several twists and tributary junctures. Coarse. open-work gravel deposits on slip-off slopes and low terraces show little evidence of reworking or revegetation between 1978 and 1980 (fig. 6C). Low adventitious plants such as grasses and forbs had completely covered the deep cutoff chute by 1980 but no large woody plants had been established. Trees that had fallen or been stranded at this meander bend in the 1978 flood were still in place in 1980. As of summer. 1986. changes in channel geometry and alignment and vegetation patterns that were effected in this reach by the Amelia flood had not been significantly modified. Urban Flooding: Walnut Creek. 1981 Urban flooding generally is more complex than that in rural settings because it often results from failure or inadequacy of engineered drainage systems as well as excessive rains. A recent example of urban flooding in the Balcones Escarpment area is the .. Memorial Day" flood of May 24 to 25. 1981. in Austin. Travis County (fig. 1). Bowmar (1981) presented a detailed review of the meteorologic causes of this flood . Late in the afternoon of May 24. warm moist air from the Gulf of Mexico was moving rapidly northwestward into Central Texas at middle levels of the atmosphere. Near-surface air had been heated throughout the day making the lower third of the atmosphere convectively unstable. Only 10.000 ft of additional vertical movement of surface air was needed to form significant thunderstorms. An upper-level trough of low pressure moved through Central Texas early in the evening and provided the needed lift. Cloud tops reached 40.000 to 45.000 ft and remained in that range for more than 7 hr. Heavy rains began falling at about 9:30 p.m. Within a few hours. 8 to 10­inch rains had covered a large area of the city. The most intense rainfall aod greatest total precipitation in the area were measured at stations in northern and northwestern Austin in the watersheds of Shoal and Walnut Creeks (Massey and others. 1982. fig. 6. table 2: Moore and others. 1982. figs. 2.2-2.4). One site near the headwaters of Walnut Creek recorded almost 6 inches of rain in one hour and 10 inches in 2 1 /2 hours. which are intensities approaching the trend of worldwide precipitation maxima (fig. 2). The effect of so much rainfall in a short period was severe flooding of parts of the city. Conditions were worsened by lighter but substantial rains of the day before which had saturated the ground (Moore and others. 1982. p. 4) . A high percentage of impervious land cover is characteristic of urban areas and reduces further the potential soil infiltration. Under these circumstances. runoff was nearly complete. A remarkable aspect of the 1981 storm was the concentration of moisture in small. relatively stationary cells. Rains produced by these cells were highly localized within a widespread pattern of general though less intense rainfall. Small drainage basins were overwhelmed. producing massive flooding. Massey and others (1982) analyzed flood hydrographs and field observations and reconstructed areas of inundation along parts of Shoal. Little Walnut. and Walnut Creeks. The following discussion pertains to the headwaters of Walnut Creek. which were beyond the area covered by Massey and others. Some of the most intense flooding resulting from this storm occurred in the uppermost reaches of Walnut Creek. The stream skirts well east of the Balcones Escarpment except in the upper part of the basin. There. tributaries drain off a segment of the escarpment which has subdued relief (fig. 7) . These short but steep bedrock slopes enhance runoff onto adjacent Coastal Plain surfaces with low-permeability soils (Werchan and others. 1974) . In addition. the small watersheds of these tributaries are areas of residential and small commercial development with 25 to perhaps 50 percent impervious cover (U.R.S./Forrest and Cotton and others. 1977. table 2-5) . Each of these factors tends to amplify runoff. Only a few years prior to the 1981 flood . upper Walnut Creek basin primarily comprised cultivated fields and rangeland. Until the late 1970's. the area was outside the corporate limits of Austin and other communities and therefore was not governed by construction codes sensitive to flood hazards. Earlier landowners evinced little voluntary concern: for example. initial construction had predated widespread recognition of risks inherent in development on flood plains. Railroads had been constructed along contours on high linear berms that obstruct movement of runoff. Rural roads with low narrow bridges. low-water crossings. and no storm culverts had been only partly replaced by urban streets and drains designed for 25 to 50-yr recurrence floods . Old and new roads and drainage ways were poorly integrated. Few of these problems had been corrected because urbanization was incomplete at the time of flooding. The area was a patchwork of modern urban streets. storm drains. housing. and businesses interspersed with undeveloped tracts. unimproved roads. and small industrial sites adjacent to streams. These conditions exaccerbated meteorologic and topographic factors associated with the flood of May. 1981. Eight to ten inches of rain fell over most of the upper Walnut Creek drainage between 9:30 p.m. and midnight on May 24 (Massey and others. 1982. fig. 6. table 2) . At FM Highway 1325 (Burnet Road). water level reached 19.5 ft. correspondrg to 15.000 cfs discharge from a drainage area of 12.6 mi (Massey and others. 1982. table 1) which approaches the nationwide trend line for high-discharge events (table 1: figs. 3. 7) . Numerous homes and buildings were damaged by rising water along the channel or unchanneled flow on nearby slopes. At Waters Park Road just upstream from Burnet Road. a few commercial buildings on the flood plain were completely destroyed or badly damaged. One small manufacturing plant was submerged by more than 15 ft of very rapidly - N ~ 0 .5mi 0 .5 km Contour interval 50 ft ~ Cultura l and phys109roph1c base from U.S. Geolo91co r survey 7.5·min. topographic mops: Jollyvdle, Texas r:=> / \ ( c I f ( I I ID I })fl [',I \ s·J '1 ~:~~=i: Pflugerville West, Texas Figure 7. Map of part of upper Walnut Creek drainage basin. Austin. Subbasins of reachP.s discussP.d in text are crosshatched. Figure 8. Aerial photographs of streams in the upper Walnut Creek drainage basin. Austin. Photo data: General Land Office of Texas. vertical color infrared. June 4. 1981. (A) Photo number 18703009. Walnut Creek along Waters Park Road . Left side of photo is northeast. Local drainage is from east to west. Trestle is approximately 120 ft long. Much of the debris deposited by the May. 1981 flood had been cleared but a few vehicles and large tanks remain . A long section of the rail bed that was damaged by high water had been repaired with highly reflective grave. seen as bright areas in this photo. (B) Photo number 18700202. Unnamed tributary of Walnut Creek along Dorsett Road. Upper left corner of photo is north. Local drainage is from northwest to southeast. Rails are approximately 90 ft from bridge. Road at lower left is not shown in figure 7. Much of the flood damage had been repaired. Debris had been cleared from the Dorsett Road bridge and washed out portions of the railroad berm had been repaired. Figure 9. Panoramic view of railroad trestle inundated by Walnut Creek. Austin (photographs match imperfectly). Photographs taken May 25. 1981. View is downstream toward west. Top of trestle is approximately 15 ft above water level. Note automobile standing vertically in left photo. moving water. This high-velocity macroturbulent flow transported heavy industrial equipment. large commercial trucks. and passenger cars more than one mile downstream from the plant site (figs. BA. 9). To accomplish this the stream carried some of its load over a 15-ft high railroad trestle partly blocking the channel just downstream. An unnamed tributary of Walnu2 Creek that has a drainage area of approximately 2.5 mi probably was entirely within one of the zones of 10-inch rainfall depicted by Massey and others (1982. fig. 6). Only part of this drainage area contributed to a reach where flood waters damaged a bridge and washed out a railroad berm along Dorsett Road (fig. 7. 8B). Just downstream. a woman was killed when her automobile was submerged at a newly constructed bridge on Duval Road (Massey and others. 1982. p. 22: fig. 3.1 of Massey and others is in error). Twelve additional fatalities occurred along other streams which also destroyed homes and businesses. CATASTROPHIC DAM FAILURE: AUSTIN. 1900 Floods have posed serious hazards throughout the history of Central Texas. In an effort to control flooding and harness the Colorado River for water supplies. recreation . and hydroelectric-power generation. the city of Austin and . later. the Lower Colorado River Authority constructed and maintained a dam in western Austin. The present structure. known as Tom Miller Dam. impounds Lake Austin. An earlier dam at this site was the world's largest masonry structure when it was completed in 1893 (Lower Colorado River Authority. undated). The reservoir formed by this early dam was called Lake McDonald (fig. 10A). Design problems and controversy surrounding the advisablity of the site raised some concern although the dam appeared stable (Taylor. 1930. p. 25). But on April 7. 1900. a major flood in the Colorado watershed caused the dam to fail. draining Lake McDonald. Sections of the dam were displaced downstream yet remained upright (fig. 10B). Other sections were washed away entirely. The dam was reconstructed. only to fail a second time in 1915 (Lower Colorado River Authority. undated). Further construction was delayed. Another flood in 1935 did additional damage (fig. 10C). Finally. in 1938. the existing structure was completed and has operated with few interruptions since that time. CONCLUSIONS The Balcones Escarpment area is one of the most flood-prone regions of the world. Intense rainstorms occur in the area with surprising frequency. Physiographic factors produce rapid runoff which results in phenomenal stream discharge. Urbanization reinforces these natural conditions and increases the probability of casualties and property losses. Numerous flood-control structures throughout the region provide some measure of security but heavy rains are so localized that catastrophic floods may occur almost anywhere else in the drainage basin Small. completely unregulated streams may undergo enormous increases in discharge. posing a considerable threat particularly in urban settings. Within the Balcones Escarpment area. the distribution of major flood-producing storms in time and space is random. Therefore. the only completely effective approach to flood protection is avoidance of geomorphically­defined flood plains and channels. A B (A) Photo number Chai 8484. Dam soon after construction (photo taken about 1895). View is toward east. Note paddlewheel steamboat Ben Hur at left. (B) Photo number Chai 1613. Remnants of dam soon after flood of August 7. 1900 (photo taken about 1900). View is toward northwest. Section in center has been displaced downstream. Note wreak of Ben Hur at right. (C) Photo number Chai 65. Dam during flood of June 15. 1935. View is toward west. ACKNOWLEDGMENTS Personnel of the Texas Natural Resources Information System. particularly Charles Palmer and Lou Falconieri. assisted in compiling aerial photographs used in preparation of this article. The Austin History Center provided historic REFERENCES Baker. V. R.. 1975. Flood hazards along the Balcones Escarpment in Central T exas--alternative approaches to their recognition. mapping. and management: The University of Texas at Austin. Bureau of Economic Geology Geological Circular 75-5. 22 p. __ 1976. Hydrogeomorphic methods for thr regional evaluation of flood hazards: Environmental Geology, v. 1. no. 5. p. 261-281. 1977. Stream-channel response to floods. with --examples from Central Texas: Geological Society of America Bulletin. v. 88. no. 8. p. 1057-1071. 1984. Flood sedimentation in bedrock fluvial systems. --in Koster. E. H.. and Steel. R. J.. eds.. Sedimentology of gravels and conglomerates: Canadian Society of Petroleum Geologists Memoir 10. p. 87-98. Beard. L. R.. 1975. Generalized evaluation of flash-flood potential: The University of Texas at Austin. Center for Research in Water Resources Technical Report CRWR-124. 27 p. Bowmar. G. W.. 1978. An analysis of weather conditions relative to the occurrence of flash flooding in Central Tex as (during the period of July 30 to August 4. 1978): Austin. Texas Department of Water Resources LP-69. 37 p. 1979. 1978--drought in the east. floods out west: --Austin. Texas Department of Water Resources LP-89. 37 p. __ 1981. Appendix--Report on meteorological aspects of the Austin flash flood of May 24-25. 1981. !.!!. Moore. W. L.. and others. The Austin. Texas. flood of May 24-25. 1981: Washington. D. C.. National Academy Press. p. 39-49. 1983. Texas weather: Austin. University of Texas --Press. 265 p. Carr. J. T.. Jr.. 1967. The climate and physiography of Texas: Austin. Texas Water Development Board Report 53. 27 p. Costa. J. E .. 1978. The dilemma of flood control in the United States: Environmental Management. v. 2. no. 4. p. 313-322. photographs of Austin. Kerza Prewitt. Joel Lardon. and Ted Samsel of the Bureau of Economic Geology drafted the final figures . Jim Morgan of the Bureau of Economic Geology rephotographed aerial photos for use as figures. Special thanks .go to each of these individuals and institutions. Crippen. J. R.. 1977. Maximum floodflows in the conterminous United States: Washington, D. C.. U.S. Department of the Interior. Geological Survey Water­supply Paper 1887. 52 p. Espey. W. H. K.. Morgan. C. W.. and Masch, F. D.. 1966. Study of some effects of urbanization on storm runoff from a small watershed: Austin. Texas Water Development Board Report 23. 110 p. Hansen. E. M..1979. Study of the reported 48+ inch rainfall in the storm of August 1-3. 1978. near Medina. Texas: Silver Springs. Maryland. National Oceanic and Atmospheric Administration. Weather Service unpublished report. 12 p. Hoyt. W. G .. and Langbein. W. B.. 1955. Floods: Princeton. New Jersey. Princeton University Press. 469 p. International Boundary and Water Commission. 1954. Flow of the Rio Grande: Washington. D. C.. Water Bulletin 24. 60 p. Jennings. A. H.. 1950. World's greatest observed point rainfalls: Monthly Weather Review, v. 78. no. 1. p. 4­ 5. Kochel. R. C.. Baker. V. R.. and Patton. P. C.. 1982. Paleohydrology of southwestern Texas: Water Resources Research. v. 18. no. 4. p. 1165-1183. Larkin. T. J.. and Bowmar. G. W.. 1983. Climatic atlas of Texas: Austin. Texas Department of Water Resources LP-192. 151 p. Leopold. L. B.. Wolman. M. G.. and Miller. J. P..1964. Fluvial processes in geomorphology: San Francisco. W. H. Freeman and Company. 522 p. Lower Colorado River Authority. undated. Lake Austin and Tom Miller Dam: Austin. brochure. Massey. B. C.. Reeves. W. E.. and Lear. W. A.. 1982. Floods of May 24-25. 1981. in the Austin. Texas. metropolitan area: Washington. D. C.. U.S. Department of the Interior. Geological Survey Hydrologic Investigations Atlas HA-656 (2 sheets) . Moore. W. L.. Cook, Earl. Gooch. R. S.. and Nordin, C. F.. Jr.. 1982. The Austin. Texas. flood of May 24-25. 1981: Washington. D. C.. National Academy Press. 54 p. Morgan. C. W .. 1966. Characteristic meteorology of some large flood-producing storms in Texas--thunderstorms. i.!! Symposium on Consideration of Some Aspects of Storms and Floods in Water Planning: Austin. Texas Water Development Board Report 33. p. 31-44. Morisawa. Marie. and LaFlure. Ernest. 1979. Hydraulic geometry. stream equilibrium. and urbanization. in Rhodes. D. D .. and Williams. G. P.. eds.. ­Adjustments of the fluvial system: Dubuque. Iowa. Kendall/Hunt Publishing Company. p. 333-350. National Oceanic and Atmospheric Administration. 1971. Principal rivers and lakes of the world: Washington. D. C.. U.S. Department of Commerce. 24 p. National Weather Service. 1979. The disastrous Texas flash floods of August 1-4. 1978: Washington. D. C.. National Oceanic and Atmospheric Administration. Natural Disaster Survey Report 79-1. 153 p. Orton. Robert. 1966. Characteristic meteorology of some large flood-producing storms in Texas--easterly waves. in Symposium on Consideration of Some Aspectsof Storms and Floods in Water Planning: Austin. Texas Water Development Board Report 33. p. 1-18. Patton. P. C .. and Baker. V. R .. 1976. Morphometry and floods in small drainage basins subject to diverse hydrogeomorphic controls: Water Resources Research. v. 12. no. 5. p. 941-952. Rahn. P. H .. 1984. Flood-plain management program in Rapid City. South Dakota: Geological Society of America Bulletin. v. 95. no. 7. p. 838-843. Schroeder. E. E .. Massey. B. C .. and Waddell. K. M .. 1979. Floods in Central Texas. August. 1978: Washington. D. C .. U.S. Department of the Interior. Geological Survey Open-file Report 79-682. 121 p. Sullivan. J. E .. 1983. Geomorphic effectiveness of a high­magnitude rare flood in Central Texas: The University of Texas at Austin. Master's thesis. 214 p. Taylor. T. U .. 1930. Silting of reservoirs: Austin. University of Texas Bulletin No. 3025. 170 p. Texas Department of Water Resources. 1984. Water for T exas--technical appendix. v. 2: Austin. pages not numbered consecutively. U.R.S./Forrest and Cotton. Inc.: Espey. Huston. and Associates. Inc.: and City of Austin Engineering Department. 1977. Drainage criteria manual: City of Austin. pages not numbered consecutively. Van Auken. 0. W .. and Ford. A. L.. in preparation. Flood­induced changes in a Central Texas riparian forest. Vance. A. M .. 1934. Excessive rainfall in Texas: Austin. Texas Reclamation Department Bulletin 25. 60 p. Werchan. L. E.. Lowther. A. C.. and Ramsey. R. N .. 1974, Soil survey of Travis County. Texas: Washington. D. C. . U.S. Department of Agriculture. Soil Conservation Service. 123 p. Wolman. M. G .. and Miller. J. C.. 1960. Magnitude and frequency of forces in geomorphic processes: Journal of Geology. v. 68. no. 1. p. 54-74. LARGE RAINSTORMS ALONG THE BALCONES ESCARPMENT IN CENTRAL TEXAS Raymond M. Slade, Jr. U.S. Geological Survey Federal Building, 6th Floor Austin, TX 78701 ABSTRACT Many lives have been lost and much property damage incurred over many years from floods in Central Texas--most of which have been caused by large rainstonns. The meteorological characteristics of Central Texas, along with an orographic influence caused by the Balcones e sea rpme nt, produce co nd i tons which causf> large rainstorms in the area. Many of the highest rainfall intensities in the world have occurred in Central Texas--a 1921 stonn in Thrall, Texas, for example, produced 32 inches of rain in 12 hours, and a 1935 stonn near D'Hanis, Texas produced 22 inches of rain in 2 hours 45 minutes. The accurate determination of the recurrence probability of large design rainfalls in Central Texas is hindered by a lack of documentation of many stonns. llany large stonns have been undocumented because of inadequate areal and temporal coverage of rain gages. Lack of a complete data base contributes to the lack of information concerning the recurrence of large storms. Accurate prediction of design rainfalls is al so hindered by the no nu nifo nn a real and temporal distribution of large stonns. Large ranges occur areally in depths of the largest stonns, and at individual sites, large differences exist between the depths of the greatest stonns, along with temporal clustering of the large storms. Site specific rainfall data are commonly used to predict design rainfalls for use in delineating flood plains and designing urban developments. Because of the nonunifonn occurrence of large stonns in Central Texas, standard statistical methods of predicting design rainfalls can produce inaccurate results. Regional studies of the climatic and physiog raphic conditons in Central Texas, along with analyses of the areal and temporal occurrences of large stonns, could be beneficial in providing methods to better predict large design rainfalls in the area. INTRODUCTION AND BACKGROUND Despite illllle nse publ ic expenditures for flood protect.ion, flood losses remain substantial, costing many lives and averaging several bill ion dollars per year nationally (U.S. Water Resources Council, 1968). A major part of the national flood losses are from "catastrophic" fl oods--fl oods which have a return period of 100 years or more, or cause failure of a flood protection project by exceeding the project design flood (Holmes, 1961). Many catastrophic floods have occurred along the Bal cones fault zone in Central Texas--most as a result of extraordinary rainstonns. Many of these floods are catastrophic because rain­sto nn depths exceed design amounts. The recurrence probability of 1 arge design rainfalls in Central Texas cannot be accurately predicted. The purpose of tnis reoo rt is to surrrnarize the development and occurrence in Abbott, Patrick L. and Woodruff, C.M., Jr., eds., 1986, n.e IWameo l'Marpmnt, c-cnJ Tena: Geological 15 Society of America, p. 15-20 of large rainstonns in Central Texas, to present rain­fall data for some of the larger storms, and to dis­cuss problems in accurate detennination of the recur­rence probability of design rainfal 1s. Many large stonns have occurred in Central Texas. Most of these storms have occurred during the months of May-July or September-October . A detailed discussion of the causes for large storms during these months is presented by Carr {1967). Those two µeriods experience much precipitation because of convective thunderstorm activity, and because migration of cooler air from the north often encounters \\ell established moisture-laden winds from the Gulf of Mexico. Also, upper level areas of atmospheric convergence are then moving over Texas from the l'ESt and east. During the period May-July, the winds have intermittently prevailed from the south long enough to have carried large quantities of water vapor from the Gulf of Mexico far into the interior of Texas. The last of the cold air from the winter season migrates from Canada and the Great Basin, ard springtime low pressure troughs aloft in the westerly winds all contribute to precipitation during this period. By September, the first cold air of the autumn­winter season has begun to clash with the long estab­1 ished moisture-laden prevailing southerly winds. Al so, the severest hurricanes to affect Texas have occurred in September. The remains of many of these hurricanes move inland to Central Texas, carrying much moisture from the Gulf of Mexico. Benson {1964) and Baker (1975) suggest that the physiography along the Balcones fault zone al so con­tributes to conditions which produce large storms. The Balcones e sea rpment, which occurs along the Balcones fault zone, separates the gently sloping and lower altitudes of the Coastal Plains from the dis­sected 1 imesto ne terrains of higher altitudes p rev a­1ent in the Edwards Plateau (figure 1). The escarp­ment 1ies at right angles to the general direction of winds from the Gulf. Moisture laden air is cooled as it rises up the slopes, causing condensation and subsequent precipitation along the escarpment. Close spacing of mean-annual isohyets along the escarpment have been used to illustrate its orographic influence (Carr, 1967). The locations, dates, and amounts of many of the 1arger storms that have been documented in Central Texas are shown in figure 1. Many long-duration storms with large rainfall depths have occurred along the escarpment, ho\\ever, many shorter duration storms of extremely high intensities have al so occurred. Some of the highest reported rainfall intensities of less than 24-hour duration in the world have occurred in Central Texas (figure 2). The storm of September 9­10, 1921, in Thrall, Texas, for example, produced 32 EXPLANATION ::::-=to= RAINSTORM DEPTH CONTOUR--ln Inches. Interval variable MONTH AND YEAR OF RAINSTORM July 1932 BRACICETTIJILLE '\, Loco t ion mop Figure 1.--Locations, dates, and depths for selected large rainstorms in central Texas. inches of rain in 12 hours, and 38.2 inches in 24 hours--the greatest known depths of these durations to occur in the continental United States. The storm of May 31, 1g35 produced 22 inches of rainfall in 2 hours 45 minutes that duration. near D'Hanis--al so a record rate for CHARACTER JSTICS OF STORMS The characteristics of many large storms in Central Texas, however, are unknown due to lack of documentation. Almost all storm documentation is from rain gages, most of which are operated by the National Weather Service. The areal ard temporal coverage of these gages, as ~ll as the tyµe of data being collected are inadequate to properly document many of the large storms. Jn many areas, distances between rain gages are greater than 40 miles--gaging density is as 1 ow as one gage per 1,000 square miles. Al so, many of the gages have only short periods of record-­many less than 10 years. Another gaging problem occurs because most of the gages in Centr~ Texas are ron-recording collectors of rainfall. At those gages, rainfall depths are measured once per day by observ- Modified from Baker (1977) ers, thus only daily rainfall values are avail able. Storm intensities are avail able only for those few gages which record incremental rainfall. Because of these facts, the greatest depths and intensities for many storms are not recorded, and many storms are totally urdocumented. Lack of a complete data base contributes to the lack of information concerning the recurrence of 1 arge rainstorms. Another problem in predicting large storms is caused by 1 arge areal ranges in the depths of the greatest storms.and large differences between the 1argest storm depths at individual sites. The rain­fall records for many rain gages in Central Texas are analyzed to demonstrate these characteristics. Ten gages operated by the National Weather Service with long-term data are chosen for the analyses. The mean-annual precipitation for those sites, which are shown in Figure 1, range from about 26 to about 36 inches. With the exception of the Smithville gage, all the gages were installed before 1900. A common period of 1900-84 is chosen for the analysis. The values of the greatest daily rainfalls for 2000 1000 CHERRAPUNJI, INDIA • 800 0 CILAOS, LA REUNION 600 $ BELOUVE, LA REUNION 400 A ALVIN, TEXAS (JULY 26, 19 7 9) THRALL, TEXAS (SEPT. 9, 1921) •0 200 MEDINA, TEXAS (AUG. 2, 1978) (/) 6. ODEM, TEXAS (OCT. 19, 1984) I u z 80 0 D'HANIS, TEXAS (MAY 31, 1935) 60 0 HOLT, MISSOURI $ z 40 + PLUMB POINT, JAMAICA _j •• 8 GALVESTON, TEXAS _J <{ (JUNE 4, 1871) LL 20 z x FUSSEN, BAVARIA <{ NOTE: Texas ci ties along Balcones a: 10 escarpment represented by boxes. 8 Other Texas c ities, represented by 6 triangles, located along Gulf Coast 4 2 Modified from Baker ( 1975) 10 60 3 6 12 24 5 30 6 24 .,..------___,~~~ MINUTES HOURS DAYS MONTHS DURATION Figure 2.--Magnitude-duration relationships for selected largest rainfalls of the world and of Texas. each of the gages are shown by bar graphs in fiyure 3. The horizontal 1 ines in each bar represent the depths of the highest 3-6 daily values for each of the gages during the period 1900-84. ThesP vcilues are listed at the top of each bar. These data illustrate the range in the highest daily rainfall values between gages in Central Texas. The maximum-daily rainfall for 2 of the gages is less than 7. 5 inches while 3 of the gages have had daily rainfalls greater than 15 inches. Figure 3 also shows the large differences between the 1argest stonns at individual sites. For example, the highest daily value at the Smithville yaye is 16. 05 inches, while the secom through fifth hi<:Jhest values are between 6.60 and 6.01 inches. Incremental rair.­f a 11 s can v a ry a s s i g n i f i cant1 y a s t he d a i l y r a i n-f a 11 s. These variations and inconsistencies in rainfall illustrate the difficult in predicting rainfall magnitude and intensity at specific sites. LargP stonns are also unevenly distrihuted in time throughout sites in Central Texas. Tabl e 1 shows, for five-year periods. the number of months for which the monthly rainfall for each gaye exceeded 10 inches. The irregular frequency at which large stonns occur at each gage is indicated in the table. For example , at the Austin gage, 12 of the 19 months whi ch exceeded 10 inches of rainfall occurred during the first 30 years of the 85-year period. The Austin gage, installed in 1856, represents the first rainfall gage in Central Texas. The data for that gage demon­strate that the large stonns can be irregular or "clustered" in tirne. For example, 11 of the 12 "~ttest" months on record occur before 1930. A rain gage that records incremental rainfall was installed in Austin in about 1928. Rainfall frequency-duration stat i st i cs , based o n v al ue s from t he gage , a re used t.hroughout the area as the basi s for fl ood-plain de­1ineations and dpsigns for urbanization. It is l ikely, however, that these data are not representa­ti ve of the "wet" period occurring before 1930. In Au st in's case , the 130 years of rainfall data indicate that thr> first half of tre period had many fllOre large sto nns a rd g reatpr sto nn depths than the second half of the record. The l argest stonns for the other gages also are tf>mporally clustered--a prohl em which can bias statistical studies of the depths and frequencies of l arge rainsto nns. The most commo n met hod used to predict des i gn rainfalls can be inadequate because of the areal and temporal characte r ist ics of these stom1s. Rainfall frequency-duration statist i cs arf> commonly used by governing officials as the basis for delineating flood plains , and for designirg urban developments. Gener­ally, rainfall statistics for a comriunity are based on one raingage in the area. Stand ard statistical fllethods for rainfall prediction assumes the recorded depths or intensities to be linearly related to fre­quency of occu rrence . This fllethod of predicti o n can­rot account for the large ranges in depths of the l crgest sto nns at the site, or tefllpo ral cl uste ring, bo th of which may bias the statistics. 20 r-1-7-.4-7~~-6-.-9-5~~12___ .-28~~1-3-.9-8~~-9-.0-0~~-8-1-6­ 5_3~-1-6_:_8_6~~9-.-3-8~~-7-.0-8~~.0-5~~~~~~~~--, en w 8.13 6.75 I 0 8.12 6.05 z 6.88 z 18 6.30 0 z 16 <( 0 0 (J) z w 14 w $: 1­ w en z 0 12 I­<( !:::: a._ 0 ~ 10 a._ >­ ...J <( 0 :::E :::) :::E x <( :::E 7.20 6.75 6.35 5.59 9.56 7.97 7.08 13 .03 6.73 7.75 7.82 6.97 10.00 6.45 7.57 6.65 10.00 6.35 6.76 6.44 6.25 8.00 6.48 6.60 6.59 6.54 6 .01 Temple Luling Smithville lt NOTE: Maximum daily rainfall values for each gage listed above bar Braunfels Antonio Marcos llSmithville gage installed in 1918 Data from Douglas Fenn, National Weather Service, oral communication, 1986, and George Bomar, Texas Water Commission, written communication, 1986 Figure 3.--Maximum daily precipitation values, 1900-84 . SUMMARY In summary, areal ard temporal documentation of large stonns is hirdered by lack of appropriate gagirfij. Al so, large ranges occur areally in depths of the 1 argest sto nns. At i rdividual sites, 1 arge differences between the depths of the largest stonns occur, alorfil with temporal clustering of the large storms. These characteristics present problems in planning and managing lard and water resources. Re­gional studies of the magnitudes, frequency, ard locations of large stonns l'.Quld probably be very beneficial in developing methods for better predicting these occurrences. If all relevant climatic, physio­graphic, ard rainfall data ard infonnation were gathered, analyzed, ard interpreted, better planning and managing may reduce the threat to life and pro­perty caused by rai nsto nns. Table 1. Number of months for which monthly precipitation exceeds 10 inches, 1900-1984 Location of precipitation gages Period of Gages v.est of Gages located on Gages east of precipitation Bal cones escarpment Bal cones escarpment Bal cones escarpment New San San Blanco Lampasas Llano Austin Braunfels Antonio Marcos Temple Lul i ng .!._!Smithville 1900-04 3 3 4 2 1905-09 1 1 1910-14 2 1 2 1 1 1915-1 9 3 1 1 1 1920-24 2 1 2 2 1925-29 2 1 2 1 3 1930-34 1 2 1935-39 2 3 4 1 1 1 4 1940-44 3 3 1 2 2 3 1945-49 1 2 2 1950-54 1 1 1955-59 1 1 2 2 1 1960-64 2 2 1 1 2 1965-69 2 1 2 1 1970-74 2 1 1 4 4 3 1975-79 1 1 1 ~ 2 1980-84 2 1 1 1 Total number lR 10 9 19 16 10 21 17 18 15 of months !_! Smithville gage installed in 1918 REFERENCES CITED Baker, V. R., 1975, Flood hazards along the Bal cones Carr, J. T., Jr., 1967, The Climate and physiography Escarpment in central Texas; Alternative of Texas: Texas Water Development Board Report approaches to their recognition, mapping, and 53, 27 p. management: Texas Univ. Austin, Bureau of HoJmes, R. C., 1961, Composition and size of flood Economic Geology Circular 75-5, 22 p. losses, in White, G. F., ed., Papers on flood Raker, V. R., 1977, Stream-channel response to floods, problems: Univ. Chicago, Dept. Geography in Geological Society of America, vol. 88, No. 8, Research Paper 70, p. 7-20. p:--1057-1071. U. s. Water Resources Council, 1968, The Nation's Benson, M. A., 1964, Factors affecting the occurrence water resources, the first national assessment: of floods in the southv.est: U.S. Geological Washington, U. S. Goverrment Printing Office, Survey Water-Supply Paper 1580-D, 72 p. 32 p. BouTELOUA CURTIPENDULA SIDE OATS GRAMA STATE GRASS OF TEXAS PLANT C0"'4UNITIES OF THE EDWARDS PLATEAU OF TEXAS: AN OVERVIEW EMPHASIZING THE BALCONES ESCARPMENT ZONE BETWEEN SAN ANTONIO AND AUSTIN WITH SPECIAL ATTENTION TO LANDSCAPE CONTRASTS AND NATURAL DIVERSITY DAVID H. RISKIND TEXAS PARKS AND WILDLIFE DEPARTMENT 4200 Smith School Road Austin, Texas 78744 INTRODUCTION The Edwards Plateau of west central Texas comprises about 93,240 sq. km. of territory (LBJSchool of Public Affairs 197B). It contains several distinct subregions. It is species rich and its mesic canyons harbor a number of endemic and insular species (Amos and Rowell 1984). Although dominantlylimestone, the southern margin of the Plateau is bounded by the Balcones fault system with limestone, chalk, marl, claystone, and localized outcrops of intrusive igneous features (Lonsdale 1927). Hence, the Edwards Plateau is a large distinct region that supports a diversity of habitats. The followingsections will provide a description of the variation in physiography, geology, climate, soils, and vege­tation that compose the Edwards Plateau; but emphasis will be on the landscape lying between San Antonio and Austin. Physiography and Topography Hill (1892) was the first to recognize the Edwards Plateau as a distinct physiographic pro­vince, but the definition of its extent has varied. Tharp (1939) described the vegetation of Texas and included the Grand Prairie to the north and Hill Country to the south and southeast in his definition of the Plateau, but excluded the Central Mineral Region (= Llano Uplift) and the flatter, central and northwestern portions. Dice (1943)provided a map of biotic provinces of North America based primarily on faunal distributions, and included the Plateau with the Rolling Plains in his C011111anchean Biotic Province. This treatment was later modified by Blair (1950), who separated the Plateau (including the Llano Uplift) as the Balconfan Province. Gould (1975) included the Llano Uplift and Stockton Plateau west of the Pecos River, but not the Lampasas Cut Plain in a widely recog­nized treatment of the vegetational areas of Texas. Godfrey, et al. (1973) also used a similar definition, but excluded the Llano Uplift. The Lyndon B. Johnson School of Public Affairs (1978)published a map of the natural regions of Texas which was essentially similar to one adopted by the United States Fish and Wildlife Service (1979). These treatments excluded the Llano Uplift but included the Lampasas Cut Plain in the Edwards Plateau natural region. The Edwards Plateau, taken in broad context, is a southern extension of the Great Plains of North America (Fenneman 1931, Hunt 1974). To the south and east ft is separated from the lower-lying West Gulf Coastal Plain by the Balcones Fault Zone, where elevations drop sharply to less than 180 m. To the ;,, Abbott. Patrick L. and Woodruff. C.M., Jr.. eds., 1986, Tiie llelooMs Eecmpmeat, Ceatnl Tena: Geological Society of America. p. 21 -32 21 DAVID D. DIAMOND TEXAS NATURAL HERITAGE PROGRAM General Land Office Stephen F. Austin Building 1700 North Congress Avenue Austin, Tex1s 78701 north it grades gradually into the Rolling Plains, while to the northwest it grades into the High Plains (=South Sandy Plains). To the west it is separated from the Stockton Plateau by the Pecos­ Devi ls River divide. The Stockton Plateau is geologically similar to, and has been considered by some as part of the Edwards Plateau (Gould 1975); however, it has more often been lumped with the more desertic Trans-Pecos region (Tharp 1939, LBJ School of Public Affairs 1978). Figure 1 provides a schematic rendering of this physiographic region. The elevation of the Edwards Plateau generally increases from the southern and eastern margins to the northwest . Austin and San Antonio on the south are at 167 m and 213 m, respectively, while Junction near the center of the Plateau is at 521 m and BigLake on the northwest is at 734 m. The southern and southeastern margins of the Edwards Plateau are highly dissected, and could hardly be considered a plateau. This "Hill Country" (= Balcones Canyonlands) consists of steep canyons, narrow divides and high gradient drainages. These short streams originate in the Hill Country and generally flow south or southeast to the Gulf of Mexico. They include, from west to east, the Nueces, Frio, Sabinal, Medina, Guadalupe and Blanco Rivers. The Pedernales flows eastward through the region, joining the Colorado just west of Austin (Fig. 2). There are numerous springs in this region at the edge of the upthrust area of the Balcones Scarp. These springs are important water sources of cities situated along the boundary of the Edwards Plateau (= Great Plains physiographic province) and the West Gulf Coastal Plain physiographic province. The granitic Central Mineral region or Llano Uplift centered in Llano, Mason and Burnet counties is likewise not a plateau; but topographically, it is a basin with respect to the main body of the Plateau to the south and west. Its geologic origins are as an uplift; hence, the name. There are numerous rounded, nearly barren granitic outcropsand the landscape is gently rolling except near drainages such as the Llano and Colorado Rivers and their tributaries or near granite outcrops, where steep slopes an~ some sheer cliffs appear. The Lampasas Cut Plain on the northeast is generally flatter than the Llano region or south­eastern margins of the Plateau previously discussed. It consists of broad valleys and wide stream divides with relatively few steep, high gra­dient canyons. The Lampasas and San Gabriel Rivers are the only two major streams that bisect the area. Figure 1. Landforms of the Edwards Plateau and adjacent areas. r----~-------------­ ,,._ ' ".< \:. ',,. , '• ' ' ... ! "~-­ ' ' ' ' '· 1 Paluxy River 11 Nueces River 2 Bosque River 12 Devil's River 3 Leon River 13 Pecos River 4 Lampasas River 14 Guadalupe River 5 San Gabriel River 15 Blanco River 6 Medina River 16 Pedernales River 7 Hondo Creek 17 Llano River 8 Sabinal River 18 San Saba River 9 Frio River 19 Concho River 10 Leona River 20 Pecan Bayou Figure 2. Approximate delineation of the Edwards Plateau natural region showing major drainages. From the central Edwards Plateau to the north and northwest the topography is generally flat to gently rolling with rounded hills, wide stream divides and few steep slopes. Much of the area could be described as a broad plaf n. Several major streams cut west to east paths across this plain,including, from north to south, the Concho, San Saba and Llano Rivers. These eventually join the Colorado, which flows southerly through the Llano Uplift and eventually to the Gulf of Mexico. The Devils River and its tributaries also bisect thf s plain fn the southwest but flow south to join the Rfo Grande. Geology Most of the Edwards Plateau consists of limestone rock of Cretaceous origin. The less eroded central and western portions are dominated byLower Cretaceous rocks within the Edwards Limestone group, while southward and eastward Edwards Limestone has largely been eroded exposing older Cretaceous material, primarily the Glen Rose forma­tion (Sellards et!}_. 1932). The Lampasas Cut Plain, which represents a generally more mature landscape than the main portion of the Edwards Plateau to the south and west, is composed of strata from both the Glen Rose and FredericksburgDivisions. Patches of limestone, dolomite, chert and marl alternately crop out at the surface across the area. Some Upper Cretaceous material, consist­ing primarily of chalk and marl, crops out along the southern and western margins of the Plateau. The geology of the Central Mineral Region or Llano uplift is strikingly different from that of the remainder of the Edwards Plateau. It is an intrusive outcrop of Precambrian rock which com­prises about 1.5 million ha in the northwestern partof the Plateau. The material overlying this intru­sive granite, where it has not been eroded away(around the perimeter, especially the northern border), consists of early Paleozoic sedimentary rocks including limestone, dolomite, sandstone, siltstone and shale. Mineralogy of the graniticmaterial varies, with hornblende schist, graphiteschist, quartz-feldspar gneiss and quartz­plagioclase-microcl ine rock corrmon. In addition, local Precambrian outcrops are scattered throughout the southern and eastern margins of the Plateau. Soils Variation in substrate and a generally hillylandscape have led to the development of a large nulllber of different sofl types on the Edwards Plateau. Excluding the Llano Uplift, upland sofls of the Plateau have generally developed fn place and occur over limestone or calfche. They are shallow and rocky or gravelly on slopes and deep fn broad valleys and on flats. Most are dark colored and calcareous, although pH f s variable depending on base saturation of the substrate, and the degree of soil profile development (Godfrey et al. 1973).Surface texture also varies from loamy to clayey, depending on substrate and profile development. These upland sofls are generally classified as Mollisols on flats and valleys (deeper soils) or Inceptisols on slopes (shallow sofls). Many have vertic properties due to montmorillonitfc clay mfnerology. These soils shrink and swell on wettingand drying, developing deep cracks fn the dry months. Clayey Vertisols are also present, especially in the east or run-on areas fn the north and northwest. Both Mollfsols and Vertfsols have surface layers that are high in organic matter, but nitrogen, phosphorus, potassium, iron and magnesium may still be limiting factors to plant growth when water is sufficient. lnceptisols may also have fairly htgh organic matter content, although they are not generally as fertile, mature, or deep as Mollisols and Vertisols. Over less alkaline parent materials or where sofl profile development has occurred for long periods over moderately or non­calcareous secondary colluvium or alluvium (for example, on old stream terraces or in former shallow depressions), loamy Alfisols have developed. They are often less fertile than Mollisols or Vertisols, although plant-soil water relations may be good. Soils of the Llano Uplift have generallydeveloped over long periods from granitic materials or, around the margins of the region, from a variety of shale, limestone, dolomite or siltstone. Most have acfd, loamy surface layers and are classfffed as Alfisols. Some deep, well-watered, sandydeposits occur around the base of major granite outcrops and in stream bottoms. These have poorprofile development and are classified as Inceptisols. Climate The climate of the Edwards Plateau becomes increasingly arid to the west and cooler to the north. The eastern and central portion is classi­fied as sub-tropical, subhumid, whfle the western one-fourth is classified as sub-tropical, semi-arid (Larkin and Bomar 1983; Fig. 2). These categoriescorrespond to Thornthwait's (1948) dry sub-tropical and semi-arid moisture regions. The generaldecrease in moisture content of Gulf air as ft flows northwestward across the Plateau is the controlling factor responsible for this difference in moisture regime. Mean annual precfpftatfon decreases from east to west, ranging from about 85 cm/yr on the eastern edge to 35 cm/yr on the western edge of the Plateau (Table 1); (Bomar 1983). There is a concomitant increase fn mean lake surface evaporation rates from east to west. July plus August evaporation rates increase from 46 cm in the east to 57 cm fn the west, while annual rates increase from 160 cm/yr to 206 cm/yr from east to west. The July plus Augustprecfpftation rates also decrease from east to west, ranging from 13 cm to 9 cm (Larkin and Bomar 1983). Hence, there fs a pronounced decrease fn surrmer precfpftatfon and an increase fn surrmer evapotranspfratfon, and thfs effect fs fncreasfngly severe to the west. In addftf on, there are perfodfcdrought years, such as those that occurred in the mid-1950's and in 1980 that cause even more severe moisture stress on plants. The average frost-free period ranges from approximately 260 days in the south (early March through late November) to 230 days in the north. Summer average highs and lows do not vary signifi­cantly across the Plateau and average about 35°C and 22°C respectively. Average January lows decrease northward, ranging from approximately 4°C to 0°c. Hence, there fs little varfatfon in environment related to north-south varf atfon fn temperature. Along with normal surrmer moisture deficiencies and periodic severe drought, high intensity rainfall events caused by tropical cyclonic disturbances are characteristic of the Edwards Plateau. These torrential storms are most common in the Hill Country along the southern and southwestern margins of the region (Baker 1975). Flooding and erosion caused by the storms are major factors in the environment of the Edwards Plateau. Table l. Normal annual and growing season (April­October) precipitation based on 1951-1980 means. for stations along a east to west transect across the central Edward's Plateau. STATION PRECIPITATION (cm)ANNUAL GROWING SEASON Austin 80.0 54.5 Fredricksburg 72.8 54.6 San Antonio. TX 69.96 56.80 Junction 57.2 41.5 Ozona 46.3 37.2 VEGETATION The climate of the Edwards Plateau becomes markedly drier to the west. and the topography becomes less dissected. Soils of the Llano Upliftregion are generally sandy and non-calcareous, in contrast to the calcareous. clayey or loamy soils of most of the remainder of the region. The southern and southwestern margins (=Hill Country; Balcones Canyonlands) are markedly more dissected. and the topography rougher than that of the Lampasas Cut Plain on the northwest. These observations have been made by early (Bray 1906, Johnson 1931. Tharp1939. 1952) as well as later (LBJ School of Public Affairs 1978. USFWS 1979) investigators, who have all separated these regions into separate vegeta­tional or at least physiographic subregions. A recent map of the current vegetation of Texas based on LANDSAT data (McMahan et al. 1984) and a map of potential natural vegetationtiy Kuchler (1964) have noted the differences among these regions. The Balcones Canyonlands or Hill Country region is more mesic and supports more forest or woodland vegeta­tion on slopes and in canyons; the Lampasas Cut Plain ls also mesic but flatter and more open and. therefore. grassier; the central and western Plateau becomes more xeric and more open; and the Llano Uplift region contains a species composition similar to but distinct from the remainder of the Plateau. Hence. the interaction of climate. topography and soils cause major shifts in vegetation patterns evident across the region. These factors. along with past and presentdisturbance regimes. also interact to cause coarse and fine scale variations in vegetation on the Plateau. The demise of free-roaming bison. intro­duction of domestic livestock and exotic herbivores and the drastic change in fire regime since 1700 have led to widespread increase in density of woody species and loss of grasslands across the Plateau (see Smeins 1980). In addition. variations in the timing and density of grazing by domestic livestock together with mechanical and chemical brush control have led to an even more patchy landscape in which the influence of natural variation in soils. slopesand aspect are obscured. The following will provide a general regional characterization of the contemporary and potentiallate seral vegetation of the Plateau; however, the principal focus will be on the San Antonio-Austin segment of the Balcones Canyonlands and adjacent lands east of the fault zone. including the southern extension of the Blackland Prairies. Affinities of the Vegetation Modern flora and fauna of the Edwards Plateau are comparatively well known. Pleistocene fauna. known primarily from caverns and sinkholes is like­wise fairly well known (Lundelius 1967); however. we know almost nothing of the last 22.000 years of vegetational history on the Plateau except through inference from Quaternary pollen records to the east and west (Bryant and Schafer 1977). There are hints of an exciting complexvegetational history which is manifested in the modern occurrence of certain insular woodland communities such as the temperate deciduous Acer-Ti l la-uercus or evergreen Pisatacia-Quercus---or=­Lacey oak Quercus glaucoides). woodlands restricted to mesic canyons; the restricted, insular Pinus remota evergreen pygmy woodlanas; the insu~ faXOdium-Sabal grotto swamps; the tropical ferns in isolated sinkholes, and too. from such excitingstories as the apparently rapid colonization of Ashe juniper (Juniperus ashei) onto the Plateau from a source on the margins of the Mexican Plateau (Adams 1977). Plant communities of the more mesic, dissected portions of the Plateau owe much of their origin to the Sierra Madre Oriental and its outliers. One could also characterize the Balcones Canyonlands of the Plateau as northern facies of the eastern pied­mont of the Sierra Madre Oriental. Mesic habitats in the protected eastern canyons are stronglyinfluenced by floristic contributions from the eastern (Austroriparian) deciduous forests, includ­ing tall grass prairie species. The Plateau on the undissected uplands owes much of its influence to the Great Plains grasslands to the north. On the more xeric western plateau and its canyons. the biotic contribution is from the dry plateaus and massifs of northern Mexico and Trans­Pecos Texas where semidesert grasslands prevail. To the northwest. centered in Reagan. Irion, Schleicher and Crockett Counties, the mesquite-tobosa community seems more akin to the Rolling Plains. as does the mesquite savannah on heavy textured soils of the Llano Basin. Other parts of the Llano Basin. over lightertextured soils. are covered in an open oak-hickorywoodland whose affinities are with the Cross Timbers and oak woodlands to the north and east. Oak wood­lands are also widespread on limestone uplands across interfluvial divides on the eastern margins of the Plateau where Alfisols occur, usually over karstic features or Quaternary terrace deposits. The southern segment of our treatment area, near San Antonio and environs. is influenced by yetanother suite of elements whose origins are the Tamaulipan thorn woodlands/shrub of the Mexican Gulf Coastal Plain. Taxa such as spiny hackberry (Celtis pallida). catclaw acacia (Acacia ~). fern acacia (A. berlandieri). persinrnon (()_i_QspTrostexana) and mesquite (Prosopis glandulosa tend to be more colllllOn on dry, edaphic sites or where dis­ turbance has played a role in landscape develop­ment. Disclimax or disturbed grasslands on heavy soil usually have an abundance of huisache (Acacia smallii). while a sub-tropical component, anaqua (Ehretia anacua). is found occasionally along riparian corridors. Balcones Canyonlands This region of steep slopes and high gradient streams is dominated by evergreen woodlands and deciduous forests. Grasslands are restricted primarily to drainage divides; usually in the context of open woodlands. Although more quantitative data on plant ecology are available for this region than for other subdivisions of the Plateau (Buechner 1944; Solcher 1927; Lynch 1962. 1971; Van Auken et al. 1979. 1980, 1981; Ford and Van Auken 1982; Busiland Van Auken 1984, 1985; Fowler 1985: Van Auken and Bush 1985; Fowler and Dunlap 1986) the composition and structure of the plant conrnunities of this zone are still not well known. Conrnunity composition reflects exposure, edaphic factors and microclimate, and although vegetation changes covered by the factors are qualitatively obvious, only one study (Van Auken et al. 1981) has investigated this topic for the Escarpment, and none have compared communities of similar habitats across moisture and exposure gra­ dients in the zone. An idealized profile of the canyons contains at least three major community types. Streams ides Along perennial watercourses, the streamside component is dominated in our area south of the Colorado by bald cypress (Taxodium disticum), sycamore (Platanus occidentalis) and to a lesser extent black willow (Salax nigra). Buttonbush (Cepalanthus occident~ is often conspicuous in the shrub stratum. Quite often, bald cypress forms monodominant stands. This streamside community is always very narrow, often less than 2 m. Dwarf Palmetto (Sabal minor) occurs occasionally. This conrnunity ~western expression of eastern swamp conrnunities. although it is adapted to periodic flooding of great magnitude, which may be essential for its maintenance (see Gehlbach 1981). Cypress swamps are well expressed at grotto sites like Hamilton's Pool and West Cave Preserve (Travis County) and at Honey Creek and Curry Creek in Comal County. Intermittent drainages support sycamorewoodlands or in the case of very "dry" sites, cedar elm usually predominates. If deep soils accumulate, the streamside component is often indistinguishable from some mesic lower slope or floodplain woodlands within canyonlands. Floodplains Like the streamside conrnunity, floodplains are subject to periodic catastrophic flooding, and are dominated by some combination of oak-elm-hackberry gallery forests. In our area this gallery woodland also may include Arizona walnut (Juglans major). box elder (Acer negundo). chittarnwood (Bumilia lanu inosa), soapberry (Sapindus), Ashe juniper, pecan Carya illinoensis}, eastern cottonwood (Populus deltoides}, live oak, Texas oak, chinkapin oak (Quercus muhlenber ii). ash (Franixus pennsylvanica , American elm (Ulmus americana}, cedar elm, (Q. sinuata), red mulberry (Morus rubra). and rarely basswood (Tilia caroliniana). although there is considerable east to west variation (Buechner 1944, Ford and Van Auken 1982). Species such as pecan, scalybark oak, chinkapin oak, and black walnut are more important in the east or on more mesic bottoms. Live oak, cedar elm, and sugar­berry increase to the west or on more xeric bottom­land sites. Floodplain forests are usually at least two-layered, with deciduous holly (Ilex decidua), roughleaf dogwood (Cornus drummondi~elderberry (Sambucus spp.), Mexican plum (Prunus mexicana}, and hoptree (Ptelea trifoliata) often present. Sugar­berry and cedar elm increase in disturbed flood­plains. The lower Devils River along the southwestern margin of the Edwards Plateau is a mesic outlier with a riparian forest of live oak, pecan and sycamore (Smith and Butterwick 1975a. Gehlbach 1981). Elevated, Quarternary gravel terraces occasionally support post oak (Quercus stellata) woodlands. Early descriptive accounts for the eastern portions can be found in Bray (1906) and Palmer (1920). Riparian vegetation changes in response to an east-west moisture gradient, as well as available riparian water and soil depth. Most eastern decid­uous species such as pecan, chinkapin oak, bur oak (Q. macrocar a), elms (Ulmus spp.), ash (Fraxinus spp. , etc., extend no further ;1es t than on a line t hrough Tom Green, San Saba, Menard, Ki mble and Real count; es. Steep Slopes The steep slopes of the Balcones canyonlands support short-stature woodlands which vary from evergreen juniper and juniper-oak on south and west exposures, to deciduous mi xed-oak hardwood woodlands on north and east exposures (see Table 2a.-d.). Texas oak (Quercus texana) is usually the dominant in the east, but westward to the Nueces River on the southern margins of the Plateau, Lacey oak may dominate. Farther west to the Pecos, vasey oak (Q. vaseyana) is dominant. Some northern exposures are dominated by Texas ash (Fraxinus texensis) or locally big-tooth maple (Acer grandidentatum). In our treatment area, these forests often contain a distinct understory shrub layer, with yaupon (Ilexvomitoria), American beautyberry (Callicarpa ~­americana), hoptree, Mexican buckeye (Ungnadia spe~iosa), red or yellow buckeye (Aesculus pavia}, deciduous holly and rough-leaf dogwood are variously present. A few of these communities have been docu­mented by Van Auken et ~· (1979, 1980). These studies, as well as earlier works (Anderson 1904, Cuyler 1931) noted that substrate has an effect on the vegetation. Texas madrone (Arbutus xalapensis) and pinyon pine (Pinus remota) are Sierra Madrean elements which occur in~community but are restricted to favorable exposures and elevations west of the Colorado River. Slope communities on dry southern and eastern exposures are primarily evergreen and dominated by Mexican juniper, often in nearly pure stands called cedar breaks. Live oaks, Mexican persimmon (Diospyros texana), shin or scalybark oak (Quercus sinuata var. sinuata), evergreen sumac (Rhus BALCONES CANYONLANDS: SLOPE WOODLANDS 2a. Generalized transect@ Ft. Hood, Bell Co., Tx. Lampasas Cut-Plain North and East exposures South and West exposures deep soils shallow soils Trees Trees Quercus muhlenbergii Juniperus ashei Acer grandidentatum Quercus sinuata var. breviloba Ulmus crassifolia Q. fusiformis Quercus sinuata Fraxinus texensis Juniperus ashei Juglans majorCeltis laevigataQuercus texana Shrubs Ilex decidua Forestiera pubescens I. vomi tori a Rhus virens Viburnum rufidulum R. lanceolata Cornus dru11111ondii Sophora secundiflora Rhamnus carolinia Diospyros texana Ptelea trifoliata Berberis trifoliolata Symphoricarpos orbiculatus Zanthoxylem hirsutum Forestiera pubescens Yucca pallida Berchemia scandens Ungnadia speciosaSophora secundiflora Cercis canadensis Rhus aromat1ca BALCONES CANYONLANOS: SLOPE WOODLANDS 2b. Generalized transect at Austin, Travis Co., Tx. North and East exposures South and West exposures deep soils sha11 ow soils Trees Trees Quercus texana Juniperus ashei Q. fusiformis Quercus fusiformis Juniperus ashei Diospyros texana Juglans major Fraxinus texensis Prunus serotina Ulmus crassifolia Q. sinuata Arbutus xalapensis Shrubs Shrubs Quercus sinuata var. breviloba Berberis trifoliolata Cercis canadensis Sophora secundiflora Ungnadia speciosa Rhus virens Eupatorium havanense Eysenhardtia texana I lex vomiter i a Mimosa borealis Forestiera pubescens Rhus aromatica Garrya lindheimeri Bernardia myricaefolia Aesculus pavia Yucca treculeana Ptelea trifoliata Dasylirion texanum Callicarpa americana Nolina texana Viburnum rufidulum Prunus mexicana Table 2 a.-d. Generalized transect of slope woodland co11111Unities on dissected uplands from east to west (mesic to xeric) across the Escarpment (=Balcones Canyonlands). Characteristic species for each co11111unity are ranked according to relative dominance of the most important woody perennials only. BALCONES CANYONLANDS: SLOPE WOODLANDS 2c. Generalized transect @S. Central edge of Escarpment Bexar, Medina, Bandera, Kendall Counties North and East exposures South and West exposures deep soils shallow soils Trees Quercus texana Q. glaucoidesJuniperus ashei Prunus serotina Juglans major Arbutus xalapensis Ulmus crassifolia Fraxinus texensis Celtis sp.Acer grandidentatum Garrya lindheimeri Ungnadia speciosaAesculus pavia var. Diospyros texana Trees Juniperus ashei Quercus fusiformis Diospyros texana Quercus sinuata var. breviloba Sophora secundiflora Acacia spp. Rhus virens Eysenhardtia texana Mimosa borealis Yucca spp.Dasylirion texanum Marus microphylla flavescens BALCONES CANYONLANDS: SLOPE WOODLANDS 2d. Generalized transect @SW edge of Escarpment. Uvalde, KinneyEdwards and Real Counties North and East exposures South and West exposures deep soils shallow soils Quercus texana Q. glaucoidesArbutus xalapensis Prunus serotina Juniperus ashei Juglans major Q. pungens var. vaseyana Q. s1nuata var. brev1loba Pi nus remota Shrubs Forestiera reticulata Garrya 11ndhe1meri Lon1cera alb1flora Ptelea tr1fo11ata Cerc1s canadens1s Cercocarpus montanus Rhus v1rens Aesculus pav1a var. D1ospyros texana Crataegus sp. flavescens Table 2. (Can't.} Juniperus ashei Quercus fusiform1s Q. pungens var vaseyanaBumelia lanuginosaMorus microphylla Shrubs Berberis trifoliolata Condalia hookeri Acacia spp. Yucca spp. Eysenhardtia texana Salvia ballotaefolia Leucana retusa Sophora secundiflora virens), skunkbush sumac (R. aromat1ca), elbow bush ~tiera pubescens), and Texas mountain laurel (Sophora secundiflora) may also be present. Scrub oak (Quercus pungens) is important in the west. These xeric woodlands usually contain no understorywoody layer and are less diverse in woody speciesthan deciduous woodlands on mesic north and west slopes previously discussed. Slopes of the dissected portions of the LampasasCut Plain support communities like those of the Balcones canyonlands in the Escarpment zone between Bexar and Travis County, although Texas oak and Texas ash seem to be more important in the Cut Plain. Quantitative data of analogous communities may be found in Van Auken et al. (1979, 1980, 1981). Scalybark oak is very important on the Cut Plain. Neither madrone, Lacey oak nor pinyon occur on the Cut Plain and the endemic Yucca rupicola of the Plateau is replaced on the Cut Plain by the endemic Yucca pallida. Composition of slope communities west of the Frio River changed dramatically. Woodlands are usually restricted to northern and eastern exposures and canyon bottoms, and taxa with a Mexican affinitybecome more important. Ashe juniper declines markedly in importance and at the Rio Grande in Val Verde County is almost absent. Shrubs such as blue sage (Salva ballotaefolia), sumac (Rhus spp.), lead­tree (Leucaena retusa), cenizo (Leucophylumfrutescens), Spanish dagger (Yucca treculeana),scrub oak, Vasey oak, sotol (DdSYTirion spp.),agarito (Berberis trifoliolata), Acacia spp. and other xeric adapted species are among the domi­nants. The western manifestation of the escarpment vegetation is best described by Bray (1905), Tharp(1944), Webster (1950}, Flyr (1966), Smith and Butterwick (1975a, 1975b) and by Johnston, et al. (unpubished). ~ ~ Special features of the Escarpment zone include assorted karstic features such as sinkholes and grottoes which are well known but little studied (Smith and Butterwick 1975b; Williams 1977b). These features are especially significant because they harbor peripheral and insular biota representative of Mexican/Tropical or eastern temperate deciduous elements. Mesic microenvironments of these features and of steep, protected canyons in general harbor numerous plants whose main distribution lies in the forests of the Gulf Coastal Plain and include yaupon, eastern red cedar (Juni erus virginiana),Indian-cherry (Rhamnus carolineana , Scalybark oak, Carolina supplejack (Berchemia scandens), inland seaoats (Chasmanthium latifolium), spicebush(Lindera benzoin), and dwarf palmetto. Narrow endemics include sycamore-leaf snowbell (Styraxplantanifolia) and Philadelphus ernestii. Some typical "Mexican" species at the eastern distributional extreme include madrone, Mexican tea (Ephedra antisyphlitica), and the fern Anemia mexicana. Communities of the Relatively Undissected Uplands and Broad Valleys Uplands of the Edwards Plateau are not today and historically never were an expansive, open, treeless grassland. However, exclusive of the Llano Upliftregion, a grassland-woodland mosaic currently exists on relatively deep upland soils across extensive portions of the Plateau. Historically, grasslands were probably more extensive than today, having been reduced by encroachment of woody species, due in part to introduction of domestic livestock and elim­ination of fire (see Smeins 1980). Likewise, tall and mid grass have been replaced by short grasses on much of the eastern two-thirds of the Plateau. The Lampasas Cut Plain also historically supportedgrasslands, although a more mature landscape with fewer flat or gently rolling areas suggest that grasslands probably formed a patchy mosaic with woodlands. The Llano Uplift contained some grass­land, although more favorable soil moisture rela­tions in some areas indicate that oak-hickorywoodland along with mesquite or mesquite-oak wood­land predominated. Grassland Communities Grasslands of the Balcones region are generally restricted to relatively flat divides and adjacentmoderate slopes and broad, mature stream valleys. These areas have been heavily grazed by domestic livestock and subjected to various brush control techniques. Hence, they are patchy and dynamic in time. Variation in species composition caused bysoils and aspect is difficult to separate from that due to past disturbance (Dunlap 1983, Fowler and Dunlap 1986). Allred (1956) considered the Plateau region a southern extension of the Mixed Prairie. Thus, well-watered, moderately grazed uplands of the region resemble tall grass communities, but increas­ing aridity to the west causes mid-and short-grass components to become increasingly important. Little bluestem (Schizachyrium scoparium), Texas wintergrass (Sti}a leucotricha), white tridens (Tridens muticus , Texas cupgrass (Eriochloasericea), tall dropseed (Sporobolus asper}, sideoats grama (Bouteloua curti endula), seep muhly(Muhlenbergia reverchonii and common curlymesquite(Hilaria belangeri) are among the dominants of moderately grazed areas (Smeins et al. 1976, Dunlap1983). Heavily grazed grasslandS-and ~ore xeric soils contain a larger proportion of short grasses such as curlymesquite, three-awn, Texas grama (Bouteloua rigidiseta), red grama (~. trifola), hairy grama (B. hirsuta), hairy tridens (Erioneuron pilosum) and Tridens muticus. Cedar sedge (Carex planostachys) is common in these grasslands. Soil depth and texture is highly variable in most areas, and hence the grasslands may be extremely heterogeneous (Smeins et~· 1976). An example of the interaction of grazing and soils is found in Fowler and Dunlap (1986). Uplands of the Hays-Travis County area may support shortgrass communities, while slopes have more tall and mid grasses due to 1) a clayeyer, and ~ence more droughty soil on ridges, and 2) heavier grazing on ridges than adjacent slopes because of the behavior patterns of domestic livestock. Live oak, shin oak and woody species associated with these are compo­nents of the grasslands, forming clumps or mottes. These mottes, along with frequent short but steep scarps dominated by woody species give many areas a park-like physiognomy. On deep, mainly non­calcareous or moderately calcareous soil, Texas oak, post oak (Quercus stellata) and, especially on the east, blackjack oak (Q. marilandica) may be scattered or form woodlands alternating with grasslands in the uplands. Similar oak woodlands also occur along well-drained stream terraces. Essentially all of the grasslands of the region are in some stage of secondary succession, and thus highly dynamic (see Beaty 1973). Although those grasslands may not have been devoid of Ashe juniper,invasion or thickening of this species has been observed to cause "cedar breaks" to form in former grasslands (Buechner 1944; also see Smeins 1980}. Mesquite is also a woody component of these grasslands that has increased in density in many areas, and live oak, shin oak and other woodyspecies such as persi1T1110n, agarita, sumac, etc., may cover more area than in pre-European settlement times. Prickly pear (Opuntia spp.), noseburn (Tragia spp.), rabbit tobacco (Evax spp.) and zexmania (Zexmania hispida} are also conwnon components. Lampasas Cut Plain Plant co11111unities of the Lampasas Cut Plain are hardly distinct from those of the Balcones Canyon­lands, but the general topography is flatter, there are fewer drainages and the character of the region as a whole is that of a grassland or open woodland (sensu Driscoll, et al. 1984; = savanna, Kuchler 1964), rather thailaC:losed woodland or forest. Also, there are more northern elements and a larger extent of post oak-blackjack oak woodlands, especially in the east and where the Cut Plain con­tacts the western Cross-Timbers in the northwest. Southern elements such as Texas madrone, Lacey oak and Mexican pinyon are absent from woodlands while scalybark oak and bur oak are more important. Although usually considered most closely related to Mixed Prairie (Allred 1956, Dodd 1968, Risser et !)_. 1981), grasslands to the north and east of the-­Lampasas Cut Plain are considered extensions of the True Prairie (Oyksterhuis 1946, Diamond and Smeins 1985). Thus, grasslands of the region in good con­dition contain tall, mid and short grasses such as little bluestem, lndiangrass, big bluestem (Andropogon gerardii}, silver bluestem, Texas wintergrass, tall dropseed, sideoats grama and curlymesquite. Mesquite is also a woody component,and Ashe juniper forms "breaks,'' althougn not as extensively as in uplands of the Balcones Canyon­lands. Co11111on short grasses of more xeric soils or in heavily grazed areas are the same as those listed for the Balcones Canyonlands. Mesquite is commonly a problem for ranchers, especially in the west, and as in other regions of the Edwards Plateau, the landscape is patchy due to differential pastgrazing, brush clearing, and other land use prac­tices in general. Woodlands In our treatment area, open mixed-oak woodlands occur on interfluvial divides, frequently over karstic features. Important species include post oak, live oak, cedar elm and Texas oak. Where sands occur, blackjack oak and Texas hickory (Caryatexana) appear locally. These woodlands occur within a matrix of grasslands with affinities for the Blackland Prairie (True Prairie) to the east. Open live oak woodlands occur along broad stream valleys coastward from the steep canyonlands. Widely spaced trees occur in a mixed to tall grass­land context. Many refer to this community type as Oak Savannah (Kuchler 1964). Overgrazing has caused a general trend of replacement of open grasslands by shrubby oak species and tall or mid grasses by short grasses (Buechner 1944). The Llano Uplift, or Central Texas Mineral Region, has been the focus of more workers than has the limestone portion of the Plateau. Detailed modern vegetation and floristic studies of the region include Whitehouse (1931, 1933), McMillen et !)_. (1968), Butterwick (1979), Walters (1980) and Walters and Wyatt (1982). Deep, sandy soils support open oak-hickory woodlands. Cedar elm, Texas hickory, live oak, blackjack oak, and post oak are common. Texas oak and Ashe juniper, nearly ubiquitous on the central and eastern parts of the limestone plateau, are conspicuously absent. Where heavy textured soils occur, primarily over shales, a mesquite woodland predominates. Disturbance on all types of deep soil favors mesquite, persimmon and whitebrush (Aloysia gratissima). Specialized vege­tation of granitic massifs (e.g., Enchanted Rock) is well documented by detailed studies (Whitehouse 1933, Walters and Wyatt 1982). These areas are particularly important as locales to investigate successional processes. Fault Zone East of the Escarpment and Blacklands While most investigators recognize the Balcones Fault Zone south and east of the Plateau, few have described the distinctive vegetation which occurs there. A few investigators who describe this vege­tation include Anderson 1904; Tharp 1926; Blair 1965; Collins et al. 1975; Riskind 1980; Gehlbach 1984; Lynch 1962, 1971. This fault zone of downthrown eroded Cretaceous materials, mostly of chalks, claystones, and marls, forms gently rolling terrain with shallow clayey soils (Mollisols}. Frequently, Quarternary lag gravels cap these hills, more prominently on the south than on the north. The zone is dissected by numerous streams (Fig. 2), which shelter a ripariangallery forest. Bald cypress occurs as a componentbetween the Nueces and Colorado while bur oak and bastard oak occur from the Colorado to the Brazos. Otherwise, the gallery forest can also be characterized as an oak-e1m-hackberry forest. Pecan, ash (Fraxinus enns lvanica, texana; berlandieri (north to south and cottonwood are also important. From the Medina River north and eastward the vegetation of this zone is a tension or transition zone of woodlands, savannah, and prairie. Grass­lands grade into Fort Worth Prairie communities through the Lampasas Cut Plain northward and Black­land Prairie to the east. Both of these grasslands are most closely related to the True Prairie of the North American Mid-continent (Dyksterhuis 1946; Diamond 1983). Woodlands grade into the Cross Timbers to the north and the Post Oak Savannah to the east. From the Medina River southward, but extending as far north as the Colorado River, brushy species of the Tamaulipan thorn scrub become more important, especially on well drained substrates. Species such as Texas persimmon, guajillo (Acacia berlanderi), Spanish dagger (Yucca treculea~ sacahuista (Nolina texana), sotol (Das lirion texana), little-leaf sumac (Rhus micro h lla , spinyhackberry, snakewood (Colubrina texensis , mountain laurel, Mimosa spp., lime prickly ash (Zanthoxylem fagara) and blackbrush acacia become more common. Ashe juniper is common in this zone at least to the Nueces River on the south. On deeper soils mesquite, huisache and hackberry (Celtis reticulata)dominate with mid grass generally characteristic of Tamaulipan savannahs, such as Trichloris, Chloris, Bouteloua, Stipa, Sporoblous, Bothriochloa, Aristida, Hilaria and Erioneuron. Northward from the Colorado, where chalk is exposed, vegetation is typical of the Balcones Escarpment (see Blair 1965, Beaty and Gehlbach 1975)and the dissected uplands of the Lampasas Cut Plain. In the zone between Waco and Dallas eastern red cedar may occur together with Ashe juniper, but it occurs as far south as the Blanco River. To the east of the chalky, relatively steepPlateau margins is the Blackland Prairie. These grasslands once occurred over deep, clayey Vertisols. Few remnants of this once vast ~.ussland remain. The climax dominants include little blue­stem, big bluestem, Indiangrass, tall dropseed, and sideoats grama (Diamond and Smeins 1985). Flat divides on the Plateau proper also supported similar grassland. The few degraded grassy areas over native sod that remain in our region contain Texas grama, Texas wintergrass, buffalograss, curly­mesquite, three-awn, muhly, and a variety of short grasses and weedy forbs. CONCLUSION The environment of the Balcones Fault Zone/Edwards Plateau region is highly variable, and supports forests, woodlands and grasslands. Super­imposed on variation caused by these abiotic factors is variation caused by historical land use pat­terns. Variation in grazing history and various brush control techniques have created a particularlypatchy landscape in the grasslands and woodlands. Thus, the influence of environmental variables is often obscured, and "fence line" contrasts can be viewed throughout. We have referred to the Balcones Canyonlands as the most distinctive biotic region of Texas. It has abundant endemic biota: yet, despite its physio­ graphic distinctness, the area stands out because it harbors an intermixture of biotic elements charac­ teristic of adjacent regions. The mesic microsites and deep sandy soils of the Llano Uplift region contain eastern deciduous forest species, the southern margin harbors Tamaulipan thorn scrub elements, and the xeric portions contain elements from the Mexican Plateau and Chihuahuan Desert. Likewise, grasslands contain elements of the True Prairie, Great Plains (= Mixed) Prairie, Short Grass Prairie and Desert Plains Grassland. This mixing of floras, this aggregate biota which is unlike any other adjacent provincial unit (Blair 1950), is a result of the biogeographic history of the Edwards Plateau, along with its size and high degree of climatic, edaphic and topographic variation. Indeed, the Plateau functions as a wide ecotonal refuge, a filter, or melting pot wedged between the equally rich Austroriparian biota to the east and the Mexican biotas to the south and west. Its rich­ ness and distinctiveness cries out for investiga­ tion, discovery, and most importantly, for responsible, enlightened stewardship. ACKNOWLEDGEMENTS The authors gratefully acknowledge the tireless typing and editorial skills of Bernie Rittenhouse; and to T. B. Samsel, and Tom Diltz for facilitatingreproduction of the landform figure. Recognition should also go to Suzanne Davis and Ann Morse in the Information Processing Center for their patience and skill in handling the revision process and in the production of the final printed copy. REFERENCES CITED Adams, R. P., 1977, Chemosystematics -analysis of populational differentiation and variability of ancestral and recent populations of Juniperus ashei. Annals of Missouri Botanical Garden. 64:184-209. Allred, B. W. and H. C. Mitchell, 1954, Major planttypes of Arkansas, Louisiana, Oklahoma and Texas and their relations to climate and soils. Texas Journal of Sciences 7:7-19. Allred, B. E., 1956, Mixed prairie in Texas. in Grasslands of the Great Plains. Edited by-:J. E. Weaver and T. J. Fitzpatrick. Johsen Publishing Company. Lincoln, Nebraska, pp. 209-254. Amos, B. and C. R. Rowell, 1984, Phytogeographicanalysis of the Edwards Plateau, Texas based on woody taxa and endemic species.. Proceedings of the Symposium-Floristics and Vegetation of the Edwards Plateau. Southwestern Association of Naturalists, Junction, TX. In preparation. Anderson, E., 1904, Plant societies of the Austin quadrangle. M.A. thesis. University of Texas, Austin. Baker, V., 1975, Flood hazards along the Balcones Escarpment in central Texas: alternative approaches to their recognition mapping and management. Bureau of Economic Geology.Geologic circular 75-5. University of Texas, Austin. Beaty, H. E., 1973, Revegetational dynamics of an overgrazed grassland. M.A. thesis, BaylorUniversity, Waco, Texas. ---~and F. R. Gehlbach, 1975, Vegetational map of the central Texas region. Institute of Environmental Studies, Baylor University, Waco, Texas. Bentley, H. L., 1898, A report upon the grasses and forage plants of central Texas. USDA Bulletin No. 10. G.P.O. Washington, D.C. Blair, L. S., 1965, A structural analysis of the cedar-oak woodland on the Austin chalk of Waco, Texas. M.S. thesis, Baylor University, Waco, TX. Blair, W. F., 1950, The biotic provinces of Texas. Texas Journal of Science 2:93-117. Bomar, G. W., 1983, Texas weather. University of Texas Press, Austin, TX. Bray, W. L., 1904, The timber of the Edwards Plateau of Texas; its relation to climate, water supply and soil. USDA Division of Forestry Bulletin 49. 30 pp. --=--' 1905, Vegetation of the sotol country in Texas. Bulletin of the University of Texas, 60. Science Series No. 6, Austin, Texas. ---..,....--• 1906, Distribution and adaptation of the vegetation of Texas. University of Texas Bulletin 82. Austin, TX. Bryant, Y. M., Jr. and H. J. Shaffer, 1977, The late Quaternary paleoenvironment of Texas: A model for the archeologist. Bulletin of the Texas Archeological Society 48:1-25. Buechner, H. K., 1944, The range vegetation of Kerr County, Texas, in relation to livestock and white-tailed deer. American Midland Naturalist 31:697-713 Bush, J. K. and A. 0. Van Auken, 1984, Woody speciescomposition of the upper San Antonio River gallery forest. Texas Journal of Science 36:139-148. Butterwick, M., 1979, A survey of the flora of Enchanted Rock and vicinity, Llano and Guillispie Counties, Texas. in Enchanted Rock: A natural area survey No. 14. LBJ School of Public Affairs, University of Texas, Austin. Collins, o. B., F. E. Smeins and D. H. Riskind, 1975, Plant con111unities of the Blackland Prairie of Texas. in M. K. Wali (ed.) Prairie: a multiple view. University of North Dakota Press, Grand Forks, North Dakota. Correll, D. S. and M. C. Johnston, 1970, Manual of the vascular plants of Texas. Texas Research Foundation, Renner, TX. Cuyler, Robert H., 1931, Vegetation as an indicator of geological formations. Bulletin of the American Association of Petroleum Geologists, 15:67-78. Diamond, D. D., 1983, Composition, diversity, and interspecific relationships of grasslands within the true and upper coastal prairies of North America. Ph.D. dissertation, Texas A&M University, College Station, TX. ____, and F. E. Smeins, 1985, Composition, classification, and species response patterns of remnant tallgrass prairies in Texas. The American Midland Naturalist 113(2):294-308. Dice, L. R., 1943, The biotic provinces of North America. University of Michigan Press, Ann Arbor, Michigan. 78 pp.Driscoll, R. S., E. L. Merkel, D. L. Radloff, D. E. Snyder, and J. S. Higihara, 1984, An ecological land classification framework for the United States. USDA Forest Service. Miscellaneous Publication 1439. Dunlap, D. W., 1983, A quantitative descriptive study of the grassland vegetation and soils of the eastern Edwards Plateau, Texas. M.A. thesis, University of Texas, Austin. 185 pp. Dyksterhuis, E. J., 1946, The vegetation of the Fort Worth prairie. Ecological Monographs 16:2-29. Fenneman, N. M., 1931, Physiography of western United States. McGraw-Hill, New York. Flyr, David, 1966, The contemporary vegetation of the Amistad Reservoir area, in: A preliminary study of the Paleoecology of~he Amistad Reservoir area. D. A. Story, Y. M. Bryant, Jr. assemblers. Report to the National Science Foundation. Ford, A. L. and 0. W. Van Auken, 1982, The distribution of woody species in the Guadalupe River floodplain forest in the Edwards Plateau of Texas. Southwestern Naturalist 27:383-392. Fowler, N. L., 1985, Density dependent population regulation in a Texas grassland. Ecology 67:545-554. and D. W. Dunlap, 1986, Grassland vegetation of the eastern Edwards Plateau. American Midland Naturalist 115:146-155. Gehlbach, F. R., 1981, Mountain islands and desert seas: A natural history of the U.S.-Mexican borderlands. Texas A&M University Press, College Station, TX. -~~-· 1984, Woodlands and forests of the Balcones Scarp zone in Central Texas: struc­ture, succession, human modification and evaluation for landscape planning. Proceedings of the Symposium--Floristics and Vegetation of the Edwards Plateau. Southwestern Association of Naturalists, Junction, TX. In preparation. Godfrey, C. L., G. S. McKee and H. Oakes, 1973, General soils map of Texas. Texas Agricultural Experiment Station, College Station, TX. Gould, F. W., 1975, Texas plants--a checklist and ecological sun111ary. Texas Agricultural Experiment Station, Texas A&M University, College Station, TX. Hill, Robert T., 1892, Notes on the Texas-New Mexican Region. Bulletin of the Geological Society of America 3:85-100. Hunt, C. B., 1974, Natural regions of the United States and Canada. W. H. Freeman and Company, San Francisco. Johnson, E. H., 1931, The natural regions of Texas. University of Texas Bulletin 3113. University of Texas, Austin. Johnston, M. C., D. H. Riskind, M. Butterwick, J. Lamb, and S. Osborn (in press). A botanical survey of the Lower Canyons of the Rio Grande. Chihuahuan Desert Research Institute Special Publ., Alpine, Texas. Kuchler, A. W., 1964, Potential natural vegetation of the conterminous United States. American Geographical Society Special Publication No. 36, New York. Larkin, T. J. and G. W. Bomar, 1983, Climatic atlas of Texas. Texas Department of Water Resources, Austin, TX. Lonsdale, J. T., 1927, Igneous rocks of the Balcones Fault region of Texas. University of Texas Bulletin 2744, Austin, TX. Lundelius, E. L., Jr., 1967, Late Pleistocene and Holocene faunal history of central Texas. in Pleistocene Extinctions: the search for a ~ cause. P.S. Martin and H. E. Wright, Jr. Yale University Press, New Haven. pp. 287-319. Lynch, D., 1962, Study of a grassland mosaic at Austin, Texas. Ecology 43:679-686. ____, 1971, Phenology, community composition and soil moisture in a relict at Austin, Texas. Ecology 52:890-897. Lyndon B. Johnson School of Public Affairs, 1978, Preserving Texas' natural heritage. LBJ School Policy Research Project Report 31. Austin. McMahan, C. A., R. G. Frye and K. L. Brown, 1984, The vegetation types of Texas, including cropland. Texas Parks and Wildlife Department, Austin, TX. McMillan, D., C. E. Jens, W. R. Adler, R. Y. Blystone, W. H. Gillespie, J. R. Irwin, R. E. Janowsky, D. 0. Kille, T. R. McGlathery, R. R. Martin, R. W. Morey, C. R. Mynard, and T. S. Patty, 1968, Factors influencing the narrow restriction of Pilularia americana in Texas. Southwestern Naturalist 13:117-127. Palmer, E. J., 1920, Canyon flora of the Edwards Plateau of Texas. Journal of Arnold Arboretum 1:233-239. Riskind, D. H., 1980, A general introduction to the vegetation and dominant plant communities of the McKinney homestead, Travis County, TX in Archaeological Investigations at the Thomas F. McKinney Homestead. Part I. Bulletin of the Texas Archeological Society. 51:128-135. Risser, P. G., C. E. Girney, H. D. Blocker, S. W. May, M. J. Patton and J. A. Weins, 1981, The true prairie ecosystem. Hutchinson Ross Publishing Company, Strousburg, PA. Sellards, E. H., W. S. Adkins and F. B. Plummer, 1932, The geology of Texas, Vol. 1. Stratigraphy. University of Texas Bulletin 3232, Austin, TX. Smeins, F. E., T. W. Taylor, and L. B. Merrill, 1976, Vegetation of a 25-year exclosure on the Edwards Plateau, Texas. Journal of Range Management 29(1):24-29. ____, 1980, Natural role of fire on the Edwards Plateau. in White, L. D., Prescribed range burning in the Edwards Plateau, Texas Agricultural Experiment Station Proceedings, pp. 4-16. College Station, TX. Smith, J. and M. Butterwick, 1975a, A vegetational survey of the Devils River-Dolan Creek area. in Devils River: A natural area survey, pp. 36­ 57. University of Texas Division of Natural Resources and Environment, University of Texas, Austin. ---=-~· 1975b, A vegetational survey of the Devil's Sinkhole-Hackberry Creek Area. 21-46 1n Devil's Sinkhole Area -Headwaters of the Nueces River. Division of Natural Resources and Environment, University of Texas, Austin. Solcher, E. A., 1927, An analysis of the plant associations of Bexar County, Texas. M.A. thesis, University of Texas, Austin. Tharp, B. C., 1926, Structure of Texas vegetation east of the 98th meridian. University of Texas Bulletin 2606, Austin, TX. ----.,--• 1939, The vegetation of Texas. The Anson Jones Press, Houston, TX. ---=--' 1944, The mesa region of Texas: an ecological study. Proceedings of the Texas Academy of Science 27:81-91. --~~· 1952, Texas range grasses. Universityof Texas Press, Austin. Thornthwaite, C. W., 1948, An approach toward a rational classification of climate. The Geographic Review. 38:55-94. United States Fish and Wildlife Service, 1979, Unique wildlife ecosystems of Texas. Albuquerque, NM. 164 pp. Van Auken, 0. W., A. L. Ford and A. Stein, 1979, A comparison of some woody upland and riparian plant communities of the southern Edwards Plateau. Southwestern Naturalist 24:165-180. --~=--· A. L. Ford, and J. L. Allen, 1981, An ecological comparison of upland deciduous and evergreen forests of Central Texas. American Journal of Botany 68:1249-1256. , A. L. Ford, A. Stein and A. G. Stein, -----.-19""'8""'0,.-,-.Woody vegetation of up 1and pl ant communities in the southern Edwards Plateau. Texas Journal of Science 32:23-35. ____ and J. Bush, 1985, Secondary succession on terraces of the San Antonio River. Bulletin of Torrey Botanical Club 112:158-166. Walters, T. W., 1980, The vascular flora and vegetation of granite outcrops in the Central Mineral Region of Texas. M.S. thesis, Texas A&M University, College Station, TX. and R. Wyatt, 1982, The vascular flora of granite outcrops in the Central Mineral Region of Texas. Bulletin of the Torrey Botanical Club 190:344-364. Webster, G. L., 1950, Observations on the vegetation and summer flora of the Stockton Plateau in northeastern Terrell County, Texas. Texas Journal of Science 2:234-242. Whitehouse, E., 1931, Ecology of Enchanted Rock Vegetation. M.S. thesis. University of Texas, Austin. ____, 1933, Plant succession on central Texas granite. Ecology 13:391-405. Williams, J. E., 1977a, The vegetation of Wild Basin. Wild Basin Institute for Environmental Studies. Wild Basin Publication 1. Wild Basin Wilderness, Inc •• Austin, TX. ____, 1977b, Vegetation of Hamilton's Pool, Travis County, Texas. Unpublished report submitted to Natural Areas Survey, c/o Texas Conservation Foundation, Austin, TX. Qurncus TEXANA TEXAS OAK EDWARDS HEATH Patrick L. Abbott Department of Geological Sciences San Diego State University San Diego, CA 92182 The Edwards Plateau has been dissected by stream erosion to yield a rugged topography referred to locally as the "Hill Country." Its' main phyaiographic component is gently sloping interstream uplands. These relatively level uplands are interrupted by steep slopes and canyon walls of stream courses. The Edwards Plateau is a limestone terrane rife with fissures which carry water to springs which in tum keep the streams supplied. The limestones of the upland divides are slowly carried off in solution by carbonic acid-laced water. The only particulate matter available to form a soil residuum is the minor percent of clay and sand admixed within the limestone. But the steepness of the slopes allows a rapid runoff of rain that commonly results in erosion of elastic material before a mature soil profile can develop. Thus the area is characterized by thin soils mixed with broken rock slabs that rest on hard limestone. The region annually receives from 15 to 30 inches of rain but i ta distribution in time and space is highly irregular. Several years may see far leas than the mean annual rainfall but then the precipitation during one week may exceed the yearly average. Hean annual tempera ture is in the high 60' s but winter readings drop below freezing for s~ort periods and au1D1Der values sometimes exceed 100 • Winds dominantly come from the southeast from the Gulf of Mexico and evaporation rates are considerably in excess of precipitation. The eastern Edwards Plateau is covered by open grassland and scattered scrub timber. The timber of the divides is a dry-climate forest picturesquely described by Bray in a 1904 USDA Forestry Bulletin: "The growth is stunted, the wood dense and hard, the branches rigid, the foliage somber, the leaves small and stiff; the climate is written in every feature." The native vegetation is largely short grasses, bunch grasses, abundant junipers, various oaks, mesquite, cacti, and many shrubs. It is used predominantly as range land and is commonly stocked with combinations of cattle, sheep and goats to make beat economic uae of the variety of plants. The steep limestone slopes and gentler uplands are dominated by a juniper-oak-grass floral aasocia tion. The moat abundant tree is Juniperua ashei. Thia juniper (known as cedar to 11Hill Country" folk) flourishes in the harsh calcareous soils of central Texas and on similar limestone terranea in southern Oklahoma and southeastern Missouri. Oaks common to the area have geographic distributions over large parts of the Atlantic and Gulf coastal plains (Fowella, 1965). These oaks include uercua ainuata (white or shin oak), g. virginiana live oak), and .9.· ahumardii (Texas or Spanish oak). The westernmost extent of each oak in Abbott, Patrick L. and Woodruff. C.M .. Jr., eds .. 1986, n.e e.i.-~I,Cmlnl Te-. Geological Society of America, p. 33-34 33 species on a range map is separated by a dashed line essentially delimiting the Balcones fault trace. West of this line these wide-ranging oak species have had to undergo ecotypic differentiation in order to adjust to the thin, seasonally dry, calcareous soils of the Edwards Plateau. Thus they are further described by the respective varieties: breviloba, fusiformis, and ~· The harsh soils derived from the limestone terrane and the spasmodic rainfall have caused habitat-correlated variation within each species that has created genetically fixed ecotypes. These varieties are found also in southwest Oklahoma and on the east face of the Sierra Hadre Oriental in Hex ico. The distinctive stunted vegetation on the Edwards Plateau forms an extensive tract of wasteland known in other regions as a heath. This broad area of rather level, open, uncultivated land with poor soil and a dominant floral element creates an ambience that affects some people to the essence of their being. In psychological effect the Edwards Plateau and its vegetation create a mood not unlike that of the great heaths of southern England so memorably described by Thomas Hardy in the Return of the Native (1878). The following excerpt has been modified slightly from Hardy's original words. "The face of the heath by its mere complexion added half an hour to evening; it could in like manner retard the dawn, sadden noon, anticipate the frowning of storms scarcely generated, and intensify the opacity of a moonless midnight to a cause of shaking and dread. In fact, precisely at this transitional point of its nightly roll into darkness the great and particular glory of the Edwards waste began, and nobody could be said to understand the heath who had not been there at such a time. It could best be felt when it could not clearly be seen, its complete effect and explanation lying in this and the succeeding hours before the next dawn: then, and only then, did it tell its true tale. The sombre stretch of rounds and hollows seemed to rise and meet the evening gloom in pure sympathy, the heath exhaling darkness as rapidly as the heavens precipitated it. And so the obscurity in the air and the obscurity in the land closed together in a black fraternization towards which each advanced halfway. The place became full of a watchful intentness now; for when other things sank brooding to sleep the heath appeared slowly to awake and listen. Every night its Titanic form seemed to await something; but it had waited thus, unmoved, during so many centuries, through the crises of so many things, that it could only be imagined to await one last crisis--the final overthrow. It was a spot which returned upon the memory of those who loved it with an aspect of peculiar and kindly congruity. Smiling champaigna of flowers and fruit hardly do this, for they are permanently harmonious. Twilight combined with the scenery of Edwards Heath to evolve a thing majestic without severity, impressive without showiness, emphatic in its admonitions, grand in its simplicity. The qualifications which frequently invest the facade of a prison with far more dignity than is found in the facade of a palace double its size lent to this heath a sublimity in which spots renowned for beauty of the accepted kind are utterly wanting. Men have oftener suffered from the mockery of a place too smiling for their reason than from the oppression of surroundings oversadly tinged. Haggard Edwards appealed to a subtler and scarcer instinct, to a more recently learnt emotion, than that which reaponds to the sort of beauty called charming and fair. The moat thorough-going ascetic could feel that he had a natural right to wander on Edwards: he was keeping within the line of legitimate indulgence when he laid himself open to influences such as these. Colours and beauties so far subdued were, at least, the birthright of all. Intensity was more usually reached by way of the solean than by way of the brilliant, and such a sort of intensity was often arrived at during winter darkness, tempests, and mists. Then Edwards was aroused to reciprocity; for the storm was its lover, and the wind its friend. Then it became the home of strange phantoms; and it was found to be the hitherto unrecognized original of those wild regions of obscurity which are vaguely felt to be compassing us about in midnight dreams of flight and disaster, and are never thought of after the dream till revived by scenes like this. It was at present a place perfectly accordant with man's nature--neither ghastly, hateful, nor ugly: neither commonplace, unmeaning, nor tame; but, like man, slighted and enduring; and withal singularly colossal and mysterious in its swarthy monotony. As with some persons who have long lived apart, solitude seemed to look out of i ta countenance. It had a lonely face, suggesting tragical possibilities. The untameable, Iahmaelitish thing that Edwards now was it always had been. Civilization was its enemy; and ever since its beginning its soil had worn the same an tiq ue brown dre as, the natural and invariable garment of the particular formation. In its venerable one coat lay a certain vein of satire on human vanity in clothes. A person on a heath in raiment of modern cut and colours has more or less an anomalous look. We seem to want the oldest and simplest human clothing where the clothing of the earth is so primitive. To recline on a stump between afternoon and night, where the eye could reach nothing of the world outside the summits and shoulders of heathland which filled the whole circumference of its glance, and to know that everything around and underneath had been from prehistoric times as unaltered as the stars overhead, gave ballast to the mind adrift on change, and harassed by irrepressible New. The great inviolate place had an ancient permanence which the sea cannot claim. Who can say of a particular sea that it is old? Distilled by the sun, kneaded by the moon, it is renewed in a year, in a day, or in an hour. The sea changed, the fields changed, the Tivers, the villages and the people changed, yet Edwards remained. Those surfaces were nei theT so steep as to be destructible by weather, nor so flat as to be the vietims of floods and de po Iii ta." THE BALCONES FAULT ZONE AS A MAJOR ZOOGEOGRAPHIC FEATURE Raymond W. Neck Texas Parks and Wildlife Department 4200 Smith School Road Austin, Texas ABSTRACT The Balcones Fault Zone is a major influence on the geographic distribution of animals in central Texas. This influence is most evident in species which are directly influenced by edaphic factors or are closely associated with plants which are similarly influenced by edaphic factors. INTRODUCTION In an attempt to understand the non-random distribution of animal species on the face of the earth, zoogeographers have long attempted to deline­ate discrete geographical areas which possess sig­nificant internal homogeneity of faunal assemblages,especially in relation to adjacent areas with a different set of assemblages. Lines are drawn between such discrete areas after analysis of geographical ranges of individual species. Exact delineation of the boundary between adjacent regions is usually not possible, because zoogeographers realize (if only on a subconscious level) that these lines are an invention of the human mind. The resulting set of biotic regions is simply a model of the real biological world--not an exact picture of that world. Meaningful analysis of the North American fauna from a zoogeographical viewpoint dates to the efforts of Dice (1943) . He published a major con­tribution with cartographic delineations and verbal descriptions of many biotic provinces in North America from the Arctic frontiers to the Tropics Panama. Dice placed portions of biotic provinces within the boundaries of Texas. The treatment by Oice of the biotic provinces of Texas was illuminating but fell short of being satisfactory. Such a lack of demonstration of faanal reality in the l940's was not unexpected. Details of the distribution of animal species in Texas were unknown to most works outside the boun­daries of the state, as Texas was far removed from the intellectual centers of the time. The worker of the early twentieth century with the best knowledge of distribution of animal speciesin Texas was self-trained and had died in early 1933. Yet most areas of southern and western Texas were unknown to John K. Strecker of the Baylor University Museum. Hence, the total picture of the Texas fauna was unavailable to most biologicalworkers in Texas at that time. However, studies by one of the new breed of field biologists of the mid-twentieth century were sufficient to allow a refinement of Oice's biotic provinces with the resultant creation of a new in Abbon. Patrick L. and Woodruff, C.M., Jr., eds., 1986, n.e ....._ ~c-cna Te-. Geological Society of America, p . .lS-40 35 78744 biotic province -one which was totally enclosed bythe capricious geographical boundaries of Texas. W. Frank Blair came to The University of Texas in Austin in 1946. His early training was in mannnals, but Blair later was to concentrate on amphibians and especially the true toads of the genus Bufo. Blair's analysis of the biotic provinces of Texas was published in 1950. This paper marks the beginning of the study of zoogeography in Texas. No significant alterations of Blair's biotic provinceshave been forthcoming. The original analysis (Blair 1950) remains the most often cited reference in the zoogeographic literature of Texas. The creation of the Balconian Biotic Province by Blair (1950) was brought about by the realization that the Balcones Fault Zone was a major physical factor in the dis­tribution of animals in central Texas. Figure 1. Biotic provinces of Texas according to W. Frank Blair (1950). Personal field studies and perusal of pertinentliterature by this author for over a decade have demonstrated variation in the relative distinctive­ness of adjacent biotic provinces. Of all the biotic provinces now accepted by zoogeographers to occur in Texas (except for the Guadalupe Mountains which are included in the Navahonian Biotic Prov­ince), the Balconian Biotic Province is by far the most distinct. Clearly related to the surface expo­sure of an enormous (though highly faulted) block of Lower Cretaceous calcareous sedimentary rocks, the Balconian Province is most distinct along its east­ern and southern edges. Boundary lines drawn on its western edge to delineate it from the Chihuahuan Biotic Province are related to increasingxericity. Similarly, the boundary along the north­western and northern margins are rather vague but are related to changes in geological substrate. The sharper delineation of the Balconian Biotic Province along its eastern (Texan Biotic Province) and southern (Tamaulipan Biotic Province) is a reflection of the sharper break in geological sub­strate. Biotic provinces are artifacts of human desires to order and classify the natural world; in reality, animals and their environments know no such tidy boundaries. This reality is the result of adjustments by animal species in response to vari­ations in substrate, microclimatic, and vegetativeenvironments along and to either side of the Bal­cones Escarpment. The purpose of this paper is to analyze the distributions of animal species along the eastern edge of the Balconian Biotic Province. In this zone of complex geological changes, the environmental factors most important in controlling the distribu­tions of animal species should be easily discern­ible. The zone in the area from Austin southwest­ward to Hew Braunfels and then west into the Texas Hill Country is the area discussed herein because of greater scientific knowledge in this area. Faunal range changeovers appear to be more dramatic in shorter distances here than elsewhere. In recognition of his seminal work in the field of Texas zoogeography and his creation of the Bal­conian Biotic Province, I am pleased to dedicate this paper to the memory of W. Frank Blair (1912­1985). I can hope only that Frank would have enjoyed this analysis and, further, that this analy­sis may encourage future workers to devote addi­tional intellectual investments to this area. METHODS Lists of vertebrate species and selected invertebrate groups which occur in central Texas have been generated. Species (or subspecies) have been designated as limited (east or west) or non­ limited in reference to the effect of the Balcones Fault Zone on geographical range. Analyses of the effect of the Balcones Escarpment upon species distributions for several faunal groups have been published (Buechner 1946; Smith and Buechner 1947). However, the narrative thrust of this paper is aimed more toward specificexamples than a comprehensive numerical analysis of the faunal changes exhibited on either side of the Balcones Escarpment. A summary numerical analysisof the influence of the Balcones Escarpment is pre­sented for various faunal groups, but more effort has been expended toward providing examples of paired species and subspecies. Detailed numerical analyses will be presented in a later paper. NUMERICAL ANALYSIS Analysis of the herpetofauna of central Texas bySmith and Buechner (1947) revealed that the majority of species (77%) of reptiles and amphibians are limited (either eastward or westward) by the Bal­cones Escarpment. The percentage of species thus limited varied among the major orders of the herpe­tofauna from 100% (salamanders) to a mere 67% (chelonians) Other orders had intermediate percent­ages of limited species as follows: snakes, 70%; anurans (frogs and toads), 74%; and lizards, 95%. The dynamic nature of the boundaries of the geographical ranges of birds complicates anyzoogeographical analysis of the effect of the Bal­cones Escarpment upon specific birds. Buechner (1946) reported that geographical ranges of 56% of the bird species occurring in central Texas were limited by the Balcones Escarpment. Of the 12B species of nonmarine mammals listed by Davis (1974) from Texas, a total of 65 species(50.8%) occur along the Balcones Escarpment. Of these 65 forms, 34 species (52.3%) are limited bythe escarpment while 31 species (47.7%) have an overlapping distribution. Of the 34 limited spe­cies, 18 (52.9%) are found only west of the Balcones Escarpment while 12 (35.3%) are found only east of this line. GEOGRAPHICAL RANGE PAIR TYPES Parapatric Species Taxa whose geographical ranges meet with no significant overlap have parapatric ranges. Recog­nition of such pairs is dependent upon the ability to pair taxa as ecological analogues and/or phylo­genetic relatives. Several pairs of birds can be placed in this category. The eastern tufted titmouse (Parus bicolor bicolor) and the black-crested titmouse (Parus bicolor atricristatus) form an east-west pairof subspecies which have a very narrow zone of intergradation along the eastern margin of the Bal­cones Escarpment (Dixon 1978). Also forming an east-west pair are the red-bellied woodpecker(Melaner es carolinus) and the golden-fronted wood­pecker Melanerpes aurifrons). The eastern taxa of these two pairs (and additional unlisted pairs) are typically found in broadleaf, deciduous woodlands while the western taxa is more typical of the more open, xeric woodlands of the Texas Hill Country(both broadleaf and coniferous woodlands). An exceptionally interesting species pair is given by two salamanders which are members of two different families. The wide-mouthed salamander (Ambystoma texanum) of the family Ambystomidae ranges to the east of the Balcones Escarpment. While individuals of this species spend dry periods under logs in protected areas, eggs are deposited in water where larval development occurs. The slimysalamander (Plethodon glutinosus) of the family Plethodontidae is found in mesic canyons, caves, and limestone slopes in the Texas Hill Country. Dryperiods are spent deep in talus slopes, rocky ter­races, or caves; larval development is terrestrial but must occur in very mesic microhabitats. These two species of salamanders with differing phylo­genetic and ecological histories meet along the eastern face of the Balcones Escarpment. Overlap is very narrow and occurs as isolated populations in a series of mesic canyon environments in the Texas Hill Country. Another notable parapatric species pair involves the eastern ranging southern leopard frog (Ranasphenocephala) and the southern and western--raiigingRio Grande leopard frog (Rana berlandieri). The boundary zone between these two forms is very narrow and is a definite example of true parapatry. How­ever, this boundary zone is actually about 30 kilo­meters east of the Balcones Escarpment. The range boundary zone of these two species is apparently related to substrate and moisture relationships which are unrelated to the presence of a major physiographic break to the west. Ground squirrels provide a mammalian example of a parapatric range pair. The Mexican ground squir­rel {Spermophilus mexicanus) is found east of the escarpment in well-drained, generally non-rocky soils especially in open terrace habitats without significant woody vegetation. West of the escarp­ment the rock squirrel {Spermophilus variegatus) is found in talus slopes, canyons, and rocky uplands. Eastern Sympatry -Western Allopatry Other pairs of taxa exhibit overlapping geographical ranges {sympatry) east of the Balcones Escarpment, but only one species occurs to the west­ward of the escarpment (allopatry). In other words, one taxa of each pair ranges on both sides of the escarpment while the other pair occurs only east of the Balcones Escarpment. Several examples of this distribution pair type are exhibited by reptilian species. The eastern box turtle (Terrapene carolina) ranges over most of the eastern United States; a western subspecies, the three-toed box turtle (Terrapene carolina triunguis), ranges westward to within 30 kilometers of the Balcones Escarpment. The three-toed box turtle is typically found in open woodlands dominated by hardwoods (post oak/black hickory). Occurring over much of eastern Texas westward through central Texas, the ornate box turtle (Terrapene ornata ornata) is found in prairies,savannahs, and open woodlands. Another congeneric pair of species with a similar range is provided by the broad-banded copperhead (Agkistrodon contortrix laticinctus) and the western cottonmouth (Agkistrodon piscivorus leucostoma). The western cottonmouth is found in the southeastern United States as far west as the larger streams of the eastern Hill Country of Texas, although this snake is very rare in the Balcones Escarpment area. On the other hand, the broad­banded copperhead ranges from eastern Texas westward through central Texas. In this western portion of the range, the broad-banded copperhead is found in riparian areas as it approaches the aquatic habitat of the cottonmouth. A pair of congeneric aquatic species is provided by the yellow mud turtle (Kinosternon flavescens flavescens) and the Mississippi mud turtle (Kinosternon subrubrum hippocrepis). The yellow mud turtle ranges from southern Nebraska to northern Mexico; in central Texas, it exists both east and west of the Balcones Escarpment. The Mississippi mud turtle ranges from the bottomlands of the Mississippi River westward to the creeks with more dependable flow, such as those occurring along the Balcones Escarpment. The large, vociferous jays provide another species pair example. The blue jay (Cyanocitta cristata) occupies mesic, broad-leaved woodlands from the Atlantic coast westward through central Texas to the well-watered canyons in the Texas Hill Country. The scrub jay (Aphelocoma coerulescens) occupies xeric, evergreen woodlands from the Pacific coast to the western margins of the Balcones Escarp­ment Zone. The narrow-mouthed toads are small anurans which feed on ants and other small insects. The eastern narrow-mouthed toad (Gastrophryne carolinensis) ranges from the southeastern United States westward to about 30 kilometers east of the Balcones Escarp­ment. A related species, the Great Plains narrow­mouthed toad (Gastrophryne olivacea), ranges from western Mexico through central Texas almost to the Texas-Louisiana boundary. While both species are found in terrestrial microhabitats during the dry season, the eastern narrow-mouthed toad is found in areas of more constant water supply while the Great Plains narrow-mouthed toad is found in areas without persistent water supplies. Western Sympatry -Eastern Allopatry As an analogue to species discussed in the previous section, another type of species-range pairs includes species which are sympatric west of the Balcones Escarpment; only one species of each pair exists east of the escarpment. The phylogenetically-varied group of mammals which are classified as "squirrels" provide a spe­cies pair of this type. The wide-ranging fox squir­rel (Sciurus niger) is certainly most common east of the Balcones Escarpment where it is found in bottom­land and upland woods, i.e. woodlands consisting entirely or largely of broad-leaved trees. However, a significant portion of the natural range of this species exists in riparian and canyon woodlands of the Texas Hill Country. The pairing of the fox squirrel, a "true squirrel," with the previously­mentioned rock squirrel, a "ground squirrel," may seem anomalous. However, these two species have substantial overlap in their diet, and, thus, form a pair of ecological analogues. Ecological inter­actions occur in ecotonal areas of mesic canyon systems in the Texas Hill Country where the fox squirrel and rock squirrel are found in mesic and xeric areas, respectively. Two scorpions also occur sympatrically west of the Balcones Escarpment, whereas only one occurs east of that area. The striped scorpion, Centruoides vittatus, ranges over most of Texas where it is found in mesic and xeric woodlands, savannahs, and prairies. A rock scorpion, Vejovis reddelli, is found over much of the Texas Hill Coun­try but is unknown east of the Balcones Escarpment. Similarly, two congeneric woodland snails have partially sympatric ranges. Polygyra texasiana ranges from eastern Texas through central Texas to the western Hill Country, where it is found in wood­lands, savannahs, and those prairies with sufficient downed wood to provide cover for reduction of water loss. The similar-appearing, but smaller Polygyra mooreana, is found in the Texas Hill Country, where it occurs in xeric broadleaf and coniferous woodlands. Endemic Species A number of aquatic species are endemic to thermally and physicochemically stable waters in spring-run streams of the Balcones Escarpment and the Texas Hill Country. Freshwater mussels (family Unionidae) endemic to the Hill Country include BLACK-CAPPED VIREO GOLDEN-CHEEKED WARBLER Quincuncina mitchelli, Lampsilis bracteata, Quadrula petrina, and Quadrula aurea. The freshwater gastro­pod, Elimia comalensis~restricted to spring-run streams of the Balconian Biotic Province. Two bird species are endemic to the central Texas area, at least as far as breeding range is concerned. Both the black-capped vireo (Vireoatricapilla) and golden-cheeked warbler (lieiidroicachrysoparia) breed nowhere but in xeric woodlands of central Texas. Substantial portions of the breeding range occur north of the true Hill Country in the Lampasas Cut Plains. The Lampasas Cut Plains also are located west of the Balcones Escarpment, although the vertical displacement is not visible at the surface. Interestingly, these two endemic breeders are found in very distinct but often adja­cent (even anastomosing) habitats. The black-cappedvireo nests in thickets of scrub deciduous oaks, while the golden-cheeked warbler nests in woodlands of mature Ashe juniper trees with long bark strands. Endemic amphibians include a variety of species which are found in rocky epigean habitats and hypogean (cave) habitats. Species include the cliff frog (Syrrophus marnocki), Texas blind salamander (Typhlomolge rathbuni), San Marcos salamander (Eurycea nana), and several blind salamanders (Eurycea latitans, Ł. troglodytes, Ł. tridentifera). A large number of species of aquatic snails are found in aquifer habitats of both the Edwards and Trinity-Edwards aquifers (Hershler 1986). Other invertebrate and vertebrate species are also restricted to such subterranean habitats (Longley1981). At least one terrestrial gastropod is endemic to the Balcones Fault Zone; Mesodon leatherwoodi is restricted to very mesic box canyons. Range-Limited Single Species A number of species which are limited geographically by the Balcones Fault Zone have no obvious ecological analogue on the opposite side of this zone. Eastern species in this classification include American alligator (Alligator mississippi­ensis), southern flying squirrel (Glaucomys volans),and red-headed woodpecker (Melanerpes erythro­cephalus). Western species in this group i ng include cliff frog, barking frog (HTlactophryne au usti latrans), red-spotted toadBufo punctatus , greentoad (Bufo debilis), and black-tailed rattlesnake (Crotalli"S"molossus). A small group of western species are known to have outlier populations east of the Balcones Escarpment. Both the Texas alligator lizard (Gerrhonotus liocephalus infernalis) and the wood­land snail Polygyra mooreana are known to have popu­lations in an area of caleareous sandstone down­stream on the Colorado River (Oakville Cuesta at La Grange, Fayette Co.). Polygyra mooreana also lives on Goliad Sandstone at Goliad State Park, Goliad Co. (downstream on San Antonio River). FAUNAL HISTORY Prior to the initiation of major faultingactivity in what is presently central Texas, no sharp environmental boundary is believed to have existed. Certainly, environmental conditions were quite different from modern conditions due to the closer proximity of the coast as well as differences in global climate and continental configuration.The historical origin of the Balconian Biotic Province occurred when the environment on either side of the newly-created (and probably yetdeveloping) escarpment became differentiated sufficiently to restrict animal species to only one side of the escarpment . This general statement merely states the obvious and demonstrates ignoranceof the temporal genesis and development of the Balconian Biotic Province. While faulting beganduring the Miocene, the length of the activity period is unknown as is the degree and constancy of the rate vertical displacement. Whether significanteffects on geographical ranges of animal speciescommenced in Late Miocene, Pliocene, or Pleistocene times is unknown. Portions of the aquifer fauna maydate to Late Cretaceous time when an extensive cave system developed in Lower Cretaceous sediments. Fossil Quaternary faunas known from sediments in the Texas Hill Country and Edwards Plateau have revealed that significant biological changes have occurred since the beginning of the Holocene. Unfortunately, correlative faunas from east of the Balcones Escarpment are not available (and may not exist). As recently as 8000 B.P., the eastern chipmunk (Tamias striatus) was living in the western Hill Country (Schultze Cave, Edwards County -Dalquest et al. 1969). This species is not known to live anywhere in Texas today. Historical records of the pine vole (Pitymys pinetorum) are known from Kerr County separated from its main range by at least 450 kilometers. Persistence of this speciesto the present indicates that a portion. of the mesic-adapted fauna survived in the Balconian Biotic Province to the present. HUMAN IMPACT Alteration of the natural environments along the Balcones Escarpment by human activities in the last 150 years have also affected the distribution of numerous animals in this area. Eastern species, such as the bluejay, fox squirrel and the snail Polygyra texasiana, are more common now or are now present west of the Balcones Escarpment after majorhuman utilization of central Texas. Alteration of intact natural communities in the Texas Hill Countryhas caused a general xerification of the environment. This xerification has allowed certain western species, e.g., the house finch (Carpodacusmexicanus), to extend eastward. In this instance, the western-ranging house finch now exhibits a narrow sympatry with the eastern-ranging purple finch (Carpodacus purpureus). SUMMARY A large percentage of the animal species occurring in central Texas are limited to either the eastern or western side of the Balcones Escarp­ment. Going from east to west. general environ­mental conditions change in several ways. Ambient conditions become less humid and somewhat cooler. Precipitation decreases in total amount but short­term rainfall totals increase. Substrate changes abruptly at the Balcones Escarpment, although some substrate changes occur east of the escarpment. Species most likely to have geographical ranges limited by the escarpment are those whose life cycle is totally or partially associated directly with the substrate, e.g .• soil burrows, or hydrologic effects of faulting {the spring fauna). Other species may be associated with plant communities whose distribution is limited directly by the escarpment. REFERENCES Blair, W. F., 1950, The biotic provinces of Texas: Texas Journal of Science, v. 2, no. 1, p. 93­ 117. Buechner, H. K., 1946, Birds of Kerr County, Texas: Transactions of the Kansas Academy of Science. v. 49, no. 3, p. 357-364. Dalquest. W. w.• Roth, E., and Judd, F.• 1969, The marnnal fauna of Schultze Cave, Edwards County, IcHTHYOSAUR Texas: Bulletin of the Florida State Museum. Biological Sciences. v. 13, no. 4, p. 204-276. Davis, W. B., 1974, The mammals of Texas: Texas Parks and Wildlife Department, Bulletin 41, 294 p. Dice, L. R., 1943, The biotic provinces of North America: Ann Arbor, University of MichiganPress, 78 p. Dixon, K. L., 1978, A distributional history of the black-crested titmouse: The American Midland Naturalist. v. 100, no. 1, p. 29-42. Hershler. R.• and Longley, G, 1986, Phreatic hydrobiids {Gastropoda: Prosobranchia) from the Edwards (Balcones Fault Zone) Aquifer region, south-central Texas: Malacologia, v. 27, no. 1, p. 127-172. Longley. G •• 1981, The Edwards Aquifer: earth's most diverse groundwater ecosystem?: Inter­national Journal of Speleology, v. 11, no. 1, p. 123-128. Smith. H. M., and Buechner, H. K •• 1947, The influence of the Balcones Escarpment on the distribution of amphibians and reptiles in Texas: Bulletin of the Chicago Academy of Sciences, v. 8, no. 1, p. 1-16. VERTEBRATE PALEONTOLOGY OF THE BALCONES FAULT TREND Ernest L. Lundelius, Jr. Department of Geological Sciences University of Texas at Austin Austin, Texas ABSTRACT Vertebrate fossils are known from numerous localities of Cretaceous and Quaternary age in the Balcones fault zone. Cretaceous vertebrate re­mains range in age from Neocomian to Maastrichtian and represent the following groups, Chondrichthyes, Osteichthyes, and Reptilia (lchthyosauria, Plesiosauria, Squamata, Crocodilia, Pterosauria, Saurischia and Ornithischia). Trackways of both ornithischian and saurischian dinosaurs are known from the Glen Rose Formation. Quaternary vertebrates are known from cave deposits of the Edwards Plateau and terrace deposits on the Gulf Coastal 'Plain. The mammalian assemblages from these deposits provide data on the Pleistocene and Holocene environments of Cen­tral Texas. INTRODUCTION The Balcones fault trend, as a fault system, extends southward from Waco, Texas through Austin and San Antonio. It then turns westward towards Del Rio and crosses the Rio Grande River into Mexico. This fault system, and its associated fault-line scarp, is the dividing line between the Edwards Plateau and the Gulf Coastal Plain. Over most of this distance the faulting juxtaposes Lower Cretaceous marine limestones, forming the Edwards Plateau, against Upper Cretaceous marine chalks and shales that underlie the western margin of the Gulf Coastal Plain. The area considered here is approximately two counties (50 miles) wide on each side of the fault zone; data from other parts of the Edwards Plateau and the Gulf Coastal Plain have also been used. The vertebrate fossils found in these sedimentary rocks have added significantly to knowledge of the faunal history of this region. Most of these fossils have been recovered from the Cretaceous limestones, shales and cha1ks and from the Pleistocene cave deposits of the Edwards Plateau and alluvial deposits of the streams that traverse the area. A few marine vertebrates are known from the Kincaid Formation of Pa 1 eocene age. CRETACEOUS Vertebrate fossils of Cretaceous age are sparsely distributed throughout the Balcones fault trend. Those on the Edwards Plateau are from Comanchean rocks ranging in age from Neocomian to lower Cenomanian. Cretaceous vertebrates from the Gulf Coastal Plain are from Gulfian rocks ranging in age from upper Cenomanian to Maastrichtian. In general the specimens are much more fragmentary and less common in central Texas than in north and northeast Texas. This is probably the result of much shallower water and lower depositional rates on the San Marcos platform of central Texas (Young, 1986). in Abbott. Patrick L. and Woodruff. C.M .. Jr., eds., 1986, Tbe IW44 kilograms) that became extinct in North America at the end of the Pleistocene. A conservative estimate of the number of species involved is 28 large and 4 small mammalian species. The time of extinction has been investigated by means of radiocarbon dates by Jelinek (1957), Hester (1960) and Martin (1958, 1967, 1984). The lastest comprehensive study (Martin, 1984) involving new Figure 5. Map showing the .modern d~stributions of Mustela erminea (horizontal lines) and Reithrodontomys fulvescens (vertica~ lines) and the location of Schulze Cave (filled circle) where both occur in a Pleistocene fauna. Scale in miles. dates and re-evaluation of many radiocarbon dates indicates that the majority of the species disappeared 10,000-11,000 years ago. The cause or causes of this extinction have been discussed by many authors for decades. Martin (1967, 1973, 1984) has presented the case for human overpredation. Guilday (1967), Slaughter (1967), Axelrod (1967), Lundelius (1967), and Graham and Lundelius (1984) have argued that the climatic change that ~ook place at the end of the Pleistocene was the primary cause. The case for human overpredation rests heavily ~n the close timing between the arrival of humans in North America and the disappearance of the large mammals. It assumes that the early human . populations were numerous and were ~echnolog1cally capable of exterminating these species. The case for climatic change rests largely on the assumption that niches were eliminated by the climatic change. This hypothesis is supported by the disappearance of the disharmonious assemblages at the same time that the large animals became extinct. Holocene Faunas and Climates After the extinction of the large mammals 10,000-11,000 years ago, the fauna of central Texas consisted of the modern fauna plus a few species that are now found north and east of central Texas in areas of cooler and more mesic climates. A few species, such as the armadillo, arrived later in the Holocene. Except for Procyon lotor simus, the species that are found today farthest north disappearerl earliest from central Texas. Sorex cinereus, Microtus pennsylvanicus and Must:ela erminea are last known from Cave Without a Name {10,900 yrs B.P.). They apparently disappeared from central Texas with the extinct fauna. Synaptomys cooperi is known from Miller's Cave associated with a radiocarbon date on bone of 7,300 yr B.P. This date should be regarded as a limiting date. The actual age is probably older as radiocarbon dates on bone tend to be somewhat young (Tamers and Pearson, 1965). Another group of species that disappeared later are found today various distances from the Edwards Plateau. Microtus ochrogaster (prairie vole) which today does not occur south of Oklahoma was present at Miller's Cave in Llano County 3,000 years ago. Blarina carolinensis, now found in east Texas, was at Austin 1,000 years ago (Barton Road site) and possibly as recently as 600 years ago (Mac's Cave). A microtine rodent (either Microtus ochrogaster or Pitymys pinetorum) is also represented in the Mac's Cave deposits. There is some evidence that Blarina carolinensis, Microtus ochrogaster and/or Pitymys pinetorum retreated eastward through time across the Edwards Plateau. Their latest occurrence in the western part of the Edwards Plateau is at Felton Cave in Sutton County 7,800 years ago. None are known from Holocene deposits of either Centipede or Damp caves in Val Verde County. This absence is significant because these caves, which have produced abunda~t small mammal faunas and are located along the canyon of the Rio Grande, would have been expected to be more mesic than the uplands. The deep, mesic canyons along the dissected edge of the Edwards Plateau are refugia today for a number of organisms such as the sugar maple (Acer saccharum), a salamander (Plethodon gTiitinosus) and the pine vole (Pitymys pinetorium) whose primary distributions lie east and/or northeast of central Texas today (Blair, 1958). This is likely to have been the case for the entire Holocene during which species formerly widespread on the Edwards Plateau persisted as isolated populations for varying lengths of time in these locally favorable areas as the regional climate changed. Several species, Geomys bursarius (plains pocket gopher), Scale us a uaticus (eastern mole), and Cryptotis parva least shrew are now absent from most of the Edwards Plateau but occur to the north, south and east. The record of Geomys bursarius is the best of these and the restriction of its range seems to follow no pattern, geographic or chronologic. This, plus the fact that its modern distribution in central Texas is spotty and related to locally favorable soil conditions, suggests that soil erosion on the Edwards Plateau is responsible for its disappearance over much of the area. Remains of the pocket gopher have beeri found in very young deposits in this area which suggests that restriction of range due to soil erosion has been operative very recently and probably is still going on. Mus musculus (house mouse), a European animal broug~to North America by Europeans, has been . found in the black fill unit at Longhorn Cavern in Burnet County indicating a date younger than 1700 A.O. (Semken, 1961). This unit also contains Geomys bursarius which does not live today in Burnet County. Its nearest occurrence is in Llano County approximately 40 miles away. The top twelve inches of a black soil unit in Rattlesnake Cave in Kinney County also contains both Mus musculus and Geomys bursarius. The nearest modern occurrence of G. bursarius is in Dimmit County 75 mi 1 es to the south. In both instances Geomys remains have been recovered from a black sediment similar to the soil found on the Edwards Plateau today. The record of the burrowing mole Scalopus aguaticus is similar but is less complete. The species listed earlier as being absent from or sparsely distributed on the Edwards Plateau are fossorial and require a reasonable depth of soil. Their scarcity in this area today seems to be related to the generally thin soil. The stage was set for the removal of soil from the Edwards Plateau by the entrenchment of the streams during or before the Pleistocene. The process was probably accelerated by the post-Pleistocene _ change to drier conditions. The last stage in the removal of the topsoil from this area seems to have taken place since 1800 as the result of extensive overgrazing which destroyed much of the vegetation (Semken, 1961). There is no evidence in the known Holocene faunas that the climate during the interval 4,000­6,000 years B.P. was any drier or warmer than at present as indicated by the occur~ence_ of ~utra canadensis (otter) in the Wunderlich site in Comal County 5,000 years ago and the fauna from Centipede Cave in Val Verde County 5,000 years ago. This picture may be somewhat biased by the concentration of sites along the eastern and southern edges of the Edwards Plateau where erosion has opened many caves which are the source of much of the fossil material. Most of these caves are adjacent to canyons which maintained more moist environments than the uplands. Vertebrate fossils from more localities away from the canyons will be needed to settle this question. ACKNOWLEDGEMENTS I thank Keith Young, Wann Langston, Jr. and Jeff Pittman for helpful comments, David Stephens for assistance with the illustrations and my wife Judith Lundelius for editorial help. The Geology Foundation of the University of Texas provided financial assistance. REFERENCES Adkins, W. S., 1923. Geology and mineral resources of Mclennan County. University Texas Bulletin, v. 2340, p. 1-202. Axelrod, D. I., 1967. Quaternary extinctions of large mammals. California University Publications Geological Sciences, v. 74, p. 1-42. Bell, B. A., Murry, P. A. and Osten, L. W., 1982. Coniasaurus Owen, 1950 from North America. Journal Paleontology, v. 56, no. 2, p. 520­ 524. Bird, R. T., 1939, Thunder in his footsteps. Natural History, v. 47, p. 254-261. Bird, R. T., 1941, Adinosaur walks into the museum. Natural History, v. 47, p. 74-81. Bird, R. T., 1944, Did Brontosaurus ever walk on land? Natural History, v. 53, p. 61-67. Bird, R. T., 1954, We captured a "live" brontosaur. National Geographic, v. 105, p. 707-722. Bird, R. T., 1985, Bones for Barnum Brown. Adventures of a dinosaur hunter. pp 225. Texas Christian University Press, Ft. Worth. Blair, W. Frank, 1950, The biotic Provinces of Texas. Texas Journal Science, v. 2, p. 93­ 117. Blair, W. F., 1958, Distributional patterns of vertebrates in the southern United States in relation to past and present environments, in Hubbs, C. L. ed., Zoogeography: American Association for the Advancement of Science. Publication 51, p. 433-568. Buchanan, G.D. and Talmadge, R. V. , 1954, The geographical distribution of the armadillo in the United States. Texas Journal Science, v. 6, p. 142-150. Dal quest, W.W., Roth, E. and Judd, F. 1969. The mammal fauna of Schulze Cave, Edwards County, Texas. Bulletin of the Florida State Museum. v. 13, p. 206-276. Farlow, J. 0., 1981, Estimates of dinosaur speeds from a new trackway site in Texas. Nature, v. 294, p. 747-748. Foley, R. L., 1984. Late Pleistocene (Woodford ian) vertebrates from the driftl ess area of southwestern Wisconsin, the Moscow Fissure local fauna. Illinois State Museum Reports of Investigations, no. 39, p. l-50. Gillette, D. D. and Ray, C. E., 1981. Glyptodonts of North America. Smithsonian Contributions to Paleobiology, no . 40, p. 1-255. Gould, C. N., 1929. Comanchean reptiles from Kansas, Oklahoma and Texas. Geological Society of America Bulletin, v. 40, p. 457­ 462. Graham, R. W., 1976A. Pleistocene and Holocene mammals, taphonomy and paleoecology of the Friesenhahn Cave local fauna, Bexar County, Texas. Ph.D. dissertation, University of Texas, Austin. Graham, R. W., 19768. Late Wisconsin mammal faunas and environmental gradients of the Eastern United States. Paleobiology, v., p. 343-350. Graham, R. W., 1984. Paleoenvironmental implications of the Quaternary distribution of the eastern chipmunk (Tamias striatus) in Central Texas. Quaternary Research, 21(1) :111-114. Graham, R. W., 1985. Response of mammalian communities to environmental changes during the late Quaternary, in Diamond, J. and Case, T. J. eds. Community Ecology: Harper and Row, New York, p. 300-313. Graham, R. W. and Lundelius, E. L. Jr., 1984. Coevolutionary disequilibrium and Pleistocene extinctions, in P.S. Martin and R.G. Klein eds. Quaternary Extinctions, A Prehistoric Revolution: University Arizona Press, Tucson, Arizona, p. 223-249. Guilday, J. E., 1967. Differential extinction during late-Pleistocene and Recent times. in Martin, P. S. and Wright, H. E. eds. Pleistocene Extinctions, The Search for a Cause: Yale University Press, New Haven, p. 121-140. Guilday, J.E., Hamilton, H. W., Anderson, E., and Parmalee, P. W., 1978. The Baker Bluff Cave Deposit, Tennessee, and the Late Pleistocene faunal gradient: Bulletin Carnegie Museum in Natural History, v. 11, p. 1-67. Hansen, T. A., 1982. Macrofauna of the Cretaceous/Tertiary boundary interval in east-central Texas. in Maddocks, R. F. ed. Texas Ostracoda. Guidebook of Excursions and Related Papers for the Eighth International Symposium on Ostracoda: Department of Geosciences, University of Houston, p. 231­ 237. Hester, James J., 1960. Late Pleistocene extinction and radiocarbon dating. American Antiquity, v. 26, p. 58-77. Hibbard, Claude W., 1951. Animal life in Michigan during the Ice Age, Michigan Alumnus Quarterly Revue, v. 57, p. 200-208. Hibbard, C. W., 1960. An Interpretation of Pl io­cene and Pleistocene Climates in North America. Annual Report Michigan Academy of Science, Arts and Letters, 62:5-30. Hill, R. T., 1901. Geography and geology of the Black and Grand Prairies, Texas. U.S. Geological Survey Report 21(7):1-666. Jelinek, A. J., 1957. Pleistocene faunas and early man. Michigan Academy of Science, Arts, and Letters Papers. 6:225-237. Jiang, Ming-Jung, 1980. Calcareous nannofossils from the uppermost Cretaceous and the lower­most Tertiary of Central Texas. Master's Thesis, Texas A&M University, p. 1-121. Johnson, E. and Holliday, V. T., 1985. AClovis­age megafaunal processing station at the Lubbock Lake Landmark. Current Research in the Pleistocene. v. 2, p. 17-19. Langston, W., Jr., 1974. Nonmammal ian Comanchean tetrapods. Geoscience and Man, 8:77-102. Lundelius, E. L., Jr., 1967. Late Pleistocene and Holocene Faunal History of Central Texas, in Martin, P. S. and Wright, H. E. Jr. eds . Pleistocene Extinctions, The Search for a Cause: Yale University Press, New Haven, p. 288-319. Martin, P. S., 1958. Pleistocene ecology and biogeography of North America in Hubbs, C. L., ed. Zoogeography: Publication 51, American Association for the Advancement of Science, p. 375-420. Martin, P. S., 1967. Prehistoric overkill. in Martin, P. S. and Wright, H. E., Jr., eds. Pleistocene Extinctions, The Search for a Cause: Yale University Press, New Haven, p. 75-120 . Martin, P. S., 1984. Prehistoric overkill: the global model. in Martin, P. S. and Klein, R. G. eds. Quaternary Extinctions, A Prehistoric Revolution: Univ. Arizona Press, Tucson, Arizona, p. 354-403. McGowan, C., 1972. The systematics of Cretaceous ichthyosaurs with particular reference to the material from North America. University of Wyoming Contributions to Geology, v. 11:-29. McNul ty, C. L., Jr. and Slaughter, B. H., 1962. An ichthyosaurian centrum from the Albian of Texas. Journal of Paleontology, v. 36, p. 346-347. McNulty, C. L., Jr. and Slaughter, B. H., 1972. The Cretaceous selachian genus, Ptychotrygon Jackel 1594. Eclogae Geologicae Helvetiae, v. 65, no. 3, p. 647-655. Ostrom, J. H., 1972. Were some dinosaurs gregarious? Palaeogeography, Palaeoclimatology, Palaeoecology. v. 11, p.287-301. Patton, T. H., 1965. Anew genus of fossil microtine from Texas. Journal of Mammalogy, v. 46, no. 3, p. 466-471. Perkins, B. F., 1974. Paleoecology of a rudist reef complex in the Comanche Cretaceous Glen Rose limestone of central Texas. Geoscience and Man, v. 8, p. 131-173. Perkins, B. F., and Stewart, C. L. 1971. Stop 7. Dinosaur Valley State Park. in Perkins, B. F., Fry, R. W., Hanor, J. S. etal 1971. Trace fossils, a field guide to selected localities in Pennsylvanian, Permian, Cretaceous and Tertiary rocks of Texas and related papers. B. F. Perkins ed. Baton Rouge, Louisiana State Univ. Misc. Publ. n. 71-1, p. 56-59. so Sams, R. H., 1982. Newly discovered dinosaur tracks, Comal County, Texas. South Texas Geological Society Bulletin, v. 23, p. 19-23. Semken, Holmes A., 1974. Micromammal distribution and migration during the Holocene. American Quaternary Association Abstracts. 3rd Biennial Meeting, p. 25. Semken, H. A., Miller, B. B. and Stevens, J.B., 1964. Late Wisconsin woodland musk oxen in association with pollen and invertebrates from Michigan. Journal of Paleontology. v. 38, no. 5, p. 823-835. Shuler, E. w., 1917. Dinosaur tracks in the Glen Rose limestone near Glen Rose, Texas. American Journal of Science v. 44, p. 294­ 298. Slaughter, B. H., 1967. Animal ranges as a clue to Late Pleistocene extinction. in Martin, P S. and Wright, H. E. Jr. eds. Pleistocene Extinctions, The Search for a Cause: Yale University Press, New Haven, p. 155-168. Slaughter, B. H. and Hoover, B. R., 1963. Occurrences of ichthyosaurian remains in the Cretaceous of Texas. Texas Journal of Science, v. 15, no. 3, p. 339-343. Storrs, G. w., 1981. A review of occurrences of the Plesiosauria (Reptilia: Sauropterygia) in Texas with description of new material. M. A. Thesis. University of Texas, Austin, 226 pp. Stricklin, F. L., Jr. and Amsbury, D. L., 1974. Depositional environments on a low-relief carbonate shelf, middle Glen Rose limestone, Central Texas. Geoscience and Man v. 8, ~ c;·:i_i;i; Ag In Tamers, M.A. and Pearson, F. J., Jr., 1965. Validity of radiocarbon dates on bone. Nature, v. 208, no. 5015, p. 1053-1055. Taylor, A. J., 1982. The mamma1 i an fauna from the mid-Irvingtonian Fyllan Cave local fauna, Travis County, Texas. M.A. Thesis, University of Texas, Austin, p. 1-106. Thurmond, J. T., 1969. Lower vertebrates and paleoecology of the Trinity Group (Lower Cretaceous) in north central Texas. Dallas, Southern Methodist University, dissertation. Visher, s. S., 1945. Climatic maps of Geologic interest. Geological Society of America Bulletin, v. 56, p. 713-736. Wright, Thomas and Lundel ius, E., Jr., 1963. Post-Pleistocene raccoons from central Texas and their zoogeographic significance. The Pearce-Sellards Series, Texas Memorial fo'useum, no. 2, p. 5-21. Young, K., 1986. Cretaceous, marine inundations of the San Marcos Platform, Texas. Cretaceous Research, v. 7, P• 117-160. Zangerl, R., 1953. The vertebrate fauna of the Selma formation of Alabama. Part III. The turtles of the family Protostegidae. Fieldiana, Geology Memoir, v. 3, p. 57-132. Part IV. The turtles of the family Toxochelidae. Ibid. p. 137-277. A SNAIL FROM THE EDWARDS AQUIFER. Oo 0.50 mm Morphoiogy ol a lemale Phreatodrob1a nugax (minus the head-foot). as seen from the right (and sl•gh!ly dorsal) sode The k.dney tissue is not shown. Ag = albumen gland, Bu = bursa copula!rix; Cg = capsule g•and . Dg = O•ge st.ve gland . Cgg = digestive gland granule, Edg = posterior end ol diges!:ve gland. In = 1ntest1ne Me = mantle edge: Oo = oocyte. Ov = oviduct: Ova = ovary. THE BIOTA OF THE EDWARDS AQUIFER AND THE IMPLICATIONS FOR PALEOZOOGEOGRAPHY Glenn Longley Edwards Aquifer Research &Data Center Southwest Texas State University San Marcos, Texas 78666 ABSTRACT The Edwards Aquifer is home to a very diverse assemblage of forty, highly adapted, aquatic, subterranean described species. Several additional invertebrate species have been discovered, but have not yet been described. The most unusual of the species are blindcatfish existing more than 600 m beneath the land surface. Possible explanations regarding the existence of this community include marine organisms adapting to the aquifer from a time when paleokarstification occurred and the caves were then inundated in a marine situation similar to that today in Bermuda. Some organisms may have entered the aquifer during the Miocene when extensive faulting occurred along the present Balcones Escarpment. Finally some organisms may have entered the aquifer through spring openings to escape the colder surface temperatures during the ice ages. The paleogeographic implications of the diverse fauna are astounding. INTRODUCTION The unusually diverse subterranean aquatic conrnunity of the Edwards Aquifer poses some interesting questions regarding its origin. There have been a total of 40 species described from the aquifer -Table 1 (Hershler and Longley, 1986). The two most diverse groups in the faunal assemblage are the crustacean, ganrnarid amphipods and the gastropod, hydrobiid snails. In both groups there are species apparently derived from both marine and freshwater ancestors. This diversification would have logically occurred since the Cretaceous period. The major question posed by these species occurrence in the deep confined aquifer is how did they get there. The zoogeographic and ecological implications are considerable. The Edwards (Balcones Fault Zone) Aquifer. The aquifer extends for 282 km from Brackettville to Salado (Fig. 1). Within the aquifer, areas below the Balcones Escarpment are confined (artesian) and those above are typically unconfined (water table). The large size and the amount of cavernous porosity in the confined region of the aquifer makes this aquifer one of the worlds most unique karst aquifers. The deposition of the Edwards formation began almost 100 million years ago. The deposition occurred in a shallow sea with a variety of environments from tidal flats to coral reefs. The area may have been similar to shallow areas in the Bahama islands today. The early limestones were alternately submerged and then exposed allowing early formation of cavernous porosity. It is possible that caverns similar to those in Bermuda that connect Blue Holes to caverns inland may have formed. In time all of Central Texas was covered by a deep Cretaceous sea. Many hundreds of feet of sediments were deposited over this early aquifer. As the North American in Abbott, Patrick L. and Woodruff, C.M., Jr.. eds., 1986, Tiie IWoaDee F.aairtmeat, c-cnl Te-. Geological Society of America, p. 51-5'4 51 continent was slowly uplifted the Cretaceous seas receded. Rivers formed that cut across the overlying sediments again gradually exposing the underlying Edwards with its cavernous porosity. By the Miocene another major event in the history of the Edwards Formation occurred (12 -17 million years ago). This was a period of extensive faulting resulting in the Balcones Fault Zone concurrent with subsidence in the Gulf of Mexico area. The formations dipped toward the present day Gulf coast. The faults allowed new movement of groundwater in the Edwards, in some areas acting as recharge points, and in others as resurgence points (springs). Further solutioning continued with enlargement of the cavernous porosity along the fault zone. The large major geologic event that may have ultimately influenced the biological composition of the aquifer began with the onset of the ice ages some three million years ago. The brief synopsis of the history of the Edwards will relate to when organisms of the various groups represented there may have first colonized the Edwards Aquifer. DISCUSSION The two most diverse groups represented in the aquifer are amphipod crustaceans and prosobranch gastropods. Most of the discussion will center on these two groups. The vertebrate fauna is also very interesting. AMPHIPOOS Information obtained from samples of an artesian well on the Southwest Texas State University campus indicate that both in numbers of genera and species the subterranean amphipod diversity far exceeds any other groundwater community studied in North America. The family Crangonyctidae is restricted to the Holarctic region and is believed to have originated on the old Laurasian landmass prior to the separation of North America and Eurasia in the Jurassic (Holsinger, 1978). The crangonyctids are allied morphologically at the superfamily level with several families living on landmass remnants of Gondwanaland in the Notogaean region. The crangonyctids are therefore believed to be an ancient group that was probably present in North American freshwaters prior to the Cretaceous. Since the Edwards Aquifer is developed in Cretaceous age limestones, the presence of Stygobromus there would imply that members of this genus have invaded and colonized subterranean water in this part of North America since the Cretaceous. It is presumed that the invasions were by ancestral inrnigrants from a part of the continent that remained above the marine waters during late Mesozoic times. Table 1. Described biota of the Edwards Aquifer. Platyhelminthes Kenkiidae Sphalloplana mohri Hyman Mollusca Hydrobi idae Phreatodrobia micra (Pilsbry &Ferriss) Phreatodrobia nugax nugax (Pilsbry &Ferriss) Phreatodrobia nugax inclinata (Hershler & Longley) Phreatodrobia rotunda (Hershler &Longley) Phreatodrobia confca {Hershler &Longley) Phreatodrobfa plana (Hershler &Longley) Phreatodrobia imftata (Hershler &Longley) Phreatodrobfa punctata (Hershler &Longley) Balconorbis uvaldensis {Hershler &Longley) Stygopyrgus bartonensus (Hershler &Longley) Arthropoda Cypridae Cypridopsis vidua· obesa Brady & Robertson Cyclopidae Cyclops cavernarum Ulrich Cyclops learii Ulrich Cyclops varicans rebellus Lilljeborg Entocytheridae Sphaeromicola moria Hart Asell idae Lirceolus smithi {Ulrich) Asellus pilus Steeves Asellus redelli Cirolanidae Cirolanides texensis Benedict Monodellidae Monodella texana Maguire Crangonyctidae Stygobromus flagellatus (Benedict) Stygobromus russelli (Holsinger) Stygobromus pecki (Holsinger) Stygobromus balconis {Hubricht) Stygobromus bifurcatus (Holsinger) Hadziidae Texiweckelia texensis (Holsinger) Texiweckelia insolita Holsinger Texiweckelia samacos Holsinger Allotexiweckelia hirsuta Holsinger Bogidiell idae Parabogidiella americana Holsinger Art es ii dae Artesia subterranea Holsinger Sebidae Seborgia relicta Holsinger Palaemonidae Palaemonetes antrorum Benedict Palaemonetes holthuisi Strenth Dyt i sci dae Hadeoporus texanus Young &Longley Chordata Ambystomidae Typhlomolge rathbuni Stejneger Typhlomolge robusta Longley Ictal uri dae Satan eurystomus Hubbs &Bailey Trogloglanis pattersoni Eigenmann EDWARDS (Balcones Fault Zone) AQUIFER REGION ~---­ '\ r '\ AUSTIN )-­ SUBREGION \ ··· SAN ANTONIO SUBR~ L ...~co -.... + , / ---.:Iii llUCAllC>IMllO O&t• CIC•TEI Cl.AllDCI -~.. Tl:Ul IUTI _...,........ -'1:UI -l - _J Figure 1. Balcones Fault Zone Edwards Aquifer. The family Hadziidae are part of a group that is composed of species that inhabit marine, brackish or freshwater habitats, largely in temperate or tropical regions. Many are found in the old Tethys Sea region (ie., the greater Caribbean and Mediterranean regions in particular). All freshwater species are troglobites or phreatobites. Closely related subterranean hadziid genera also occur in brackish and freshwater habitats in the Mediterranean region, in shallow marine and anchialine habitats at a few spots in other tropical oceans. The genera occurring in the Edwards were probably derived from marine and/or brackish water ancestors at various times from the late Cretaceous to the late Tertiary (Holsinger, 1974). These forms were probably relicted during the recession of marine waters in the late Cretaceous. The family Bogidiellidae is part of a larger complex of the superfamily Bogidielloidea. The majority of species are recorded from the greater Caribbean and Mediterranean regions. Ruffo (1973) and Stock (1977, 1978) have made strong cases for the evolution of freshwater members of Bogidiellidae from marine ancestors, with freshwater invasion occurring at different places over a long period of time. The family Artesiidae is known only from the Edwards Aquifer. Its probable affinity with Bogidiellidae makes it likely that this family had a marine origin, with ancestral forms relicted in freshwater following the Cretaceous embayment of central Texas. The family Sebidae, which is predominately marine, are small, weakly pigmented, mostly eyeless species from benthic habitats. Due to their characteristics, this group would have been good candidates for colonization of interstitial freshwater habitats during marine transgressions. The presence of species in a land-locked, oligohaline-brackish water lake in the Renell Islands of the South Pacific, assumed to have been isolated there since the Late Pliocene (Bousfield, 1970), may indicate the manner in which the ancestors of the Edwards Aquifer form became isolated in, and adapted to, the transitional aquatic environment of south-central Texas during recession of sea water in the late Cretaceous or early Tertiary. GASTROPODS The Hydrobiidae (Rissoacea) are a large family (over 100 genera and 1000 species; Davis, 1979) of dioecious, gill-breathing snails that have radiated into diverse fresh-and brackish-water habitats worldwide. Minute, unpigmented hydrobiids occupy groundwater habitats in numerous areas, with a large fauna occurring in karst regions of Europe, lesser numbers occurring in North America, Mexico, Japan, and New Zealand (Hershler and Longley, 1986). Little is known of the zoogeography of these taxa, in large part due to their small size (often less than 2 mm) and the difficulties in sampling the·ir habitat. The described species of the Edwards Aquifer have been found in 14 artesian wells and four springs. The wells that yielded snails ranged in depth from 59 -582 m. All of these wells are cased and there is no doubt that the snails were expelled from the deep artesian zone. Their habitat probably includes fractures, joints and caverns in the Edwards formation. It is also possible, given their small size, that they even inhabit interstices. OTHER INVERTEBRATES In samples of wells in the San Antonio area, some specimens of a Foraminiferan from the family Lagenidae were found. They appear to be fresh not fossil forms. They were tentatively identified as a species of Robulus. For identification Cushman, 1928 was used. Vandel, 1965 discusses a discovery of A.L. Brodsky of an abundant population of Foraminifera in some wells i n the Kara-Kum desert in the Trans-Caspian Province. The wells were about 20 m deep and were fed by slightly brackish ground water. Further work has not been done on the San Antonio forms, but will be completed in the future . The only known North American Thermosbaenacean, Monodella texana, is found from the Edwards Aquifer. This representative of the family Monodellidae is most closely related to species that occur around the Mediterranean and West Indies (Stock and Longley, 1981). The species found in the Mediterranean region are all freshwater including groundwater forms. The absence of marine Thermosbaenacea in the Mediterranean may be explained by events in the late Miocene hydrographic history of that basin. The sea level dropped considerably and much of the remaining water was temporarily transformed into brine. Conceivably the marine ancestors became extinct in the Mediterranean during the late Miocene. In the West Indies, where no such drastic salinity change took place, marine Thermosbaenacea did survive. It is likely that ancestors of this species also entered the Edwards aquifer during a time when central Texas was a marine area. Troglobitic isopods in the freshwater Asellidae and predominantly marine Cirolanidae show contrasting patterns of distribution and speciation. Asellids are derived from an ancient Holarctic group probably already established in freshwater prior to the breakup of Laurasia (Barr and Holsinger, 1985). Few cirolanid isopods live in fresh water, most of them are troglobites. Species from Bermuda, the Bahamas, and Aruba occur in anchialine waters; the remainder inhabit freshwater habitats in Texas, Virginia, Mexico, and several West Indian islands. Except for the species in the Bahamas, which were probably derived from marine ancestors by direct dispersal, most cirolanids appear to have originated by stranding during regression of marine embayments or through uplifting. The shrimp of the family Palaemonidae were probably derived from ancestral forms that gave rise to the forms found in the aquifer and Texas estuaries today. VERTEBRATES Longley, 1978 discusses the status of two troglobitic salamanders occurring in the groundwaters of the San Marcos area. The species Typhlomolge rathbuni and Typhlomolge robusta are highly adapted neotenic species of the family Plethodontidae. They are considered the amphibians most adapted to cave existence in the world. They reproduce while retaining their gills, a larval characteristic. They have long legs with little musculature, an obvious adaptation to their total aquatic existence. Two species of blind catfish of the family Ictaluridae occur in the groundwaters of the San Antonio area. They are highly adapted species having no airbladders, unlike their surface relatives. One species, the Toothless blindcat, Trogloglanis pattersoni has a suckerlike mouth on the underside of its head unlike any other member of the family. The fish are found in outflow from wells that penetrate the Edwards formation between 402 mat the Artesia well and 582 mat the O.R. Mitchell (nows. Kleburg) well. These are flowing artesian wells having considerable artesian pressure (Longley and Karnei, 1979). Considering the amount of change in these species when compared to other species in their families it seems logical to postulate that they found their way into the aquifer in prepleistocene times perhaps as a means of escaping the colder temperatures on the surface. The temperatures of the groundwater tend to remain constant and would naturally dampen the effects of periods of extreme cold, an advantage to these fish. The same type of retreat into springs and then further down in caves was probable for all of the vertebrates in the aquifer. SUMMARY The unique community of aquatic subterranean forms inhabiting the Edwards Aquifer raise many questions regarding their origins. Additional study is needed to relate marine relict species occurrences in this system. When adequately sampled, other aquifer systems, such as the Floridean Aquifer, will possibly show similar relationships. REFERENCES Barr, T.C. Jr., and J.R. Holsinger, 1985, Speciation in cave faunas: Annual Review Ecological Systems, 16 p. 313-337. Bousfield, E.I., 1970, Terrestrial and Aquatic Amphipod Crustacea from Rennell Island: The natural history of Rennel Isl and British Solomon Islands, 6 p. 155-168. Cushman, J.A., 1928, Foraminifera, their classification and economic use: Cushman Laboratory for Foraminiferal Research 1, Sharon, Massachusetts, 401 p. Davis, G.M., 1979, The origin and evolution of the gastropod family Pomatiopsidae, with emphasis on the MeKong River Triculinae: Monograph of the Academy of Natural Sciences of Philadelphia 20, 120 p. Hershler, R. and G. Longley, 1986, Hadoceras taylori, a new genus and species of phreatic Hydrobiidae (Gastropoda: Rissoacea) from south-central Texas: Proceedings of the Biological Society of Washington, v. 99, no. 1, p. 121-136. , 1986, Phreatic ---.H,_.y-.d_r_ob.-1'"''i'°'d"'s-r(G,..a,..,s'""t:-::r-=-o=-po=-d=a : Prosob ranch i a ) from the Edwards (Balcones Fault Zone) Aquifer Region, South-Central Texas: Malacologia, v. 27, no. 1, p. 127-172. Holsinger, J.R., 1974, Zoogeography of the subterranean Amphipod Crustaceans (Gammaridae, Padzia Group) of the greater Caribbean Region: Virginia Journal of Science, v. 25, no. 2, p. 64. ---..,....,...---' 1978, Systematics of the subterranean Amphipod Genus Stygobromus (Crangonyctidae), part II: Species of the Eastern United States: Smithsonian Contributions to Zoology 266, 144 p. Longley, G., 1978, Status of Typhlomolge (=Eurycea) rathbuni, the Texas Blind Salamander: U.S. Fish and Wildlife Service, Albuquerque, New Mexico, Endangered Species Report 2, 45 p. Longley, G. and H. Karnei Jr., 1979, Status of Trogloglanis pattersoni Eigenmann, the Toothless Blindcat and status of Satan eurystomus Hubbs and Bailey, the Widemouth Blindcat: U.S. Fish and Wildlife Service, Albuquerque, New Mexico, Endangered Species Report 5, 48 p. Ruffo, s., 1973, Studie sui Crostacei Anfipodi, LXXIV: Contribute alla revisione del genere Bogidiella hertzog (Crustacea Amphipoda, Gammaridae): Bolettino del' Institute di Entomologia della Universita di Bologna, v. 31, p. 49-77. Stock, J.H., 1977, The zoogeography of the Crustacean Auborder Ingolfiellidea with descriptions of new West Indian taxa: Studies on the fauna of Curacao and other Caribbean islands, v. 55, no. 178, p. 131-146. , 1978, Bogidiella martini, un nouvel ---amp~h-.-ip-ode souterrain de L' Ile Saint-Martin (Antilles) et la zoogeographie des Bogidiellidae: International Journal Speleology, v. 9, no. 2, p. 103-113. -----,...-• and G. Longley, 1981, The generic status and distribution of Model la texana Maguire, the only known North American Thermosbaenacean: Proceedings of the Biological Society of Washington, v. 94, no. 2. p. 569-578. Vandel, A., 1965, Biospeleology -The Biology of Cavernicolous animals: (Translation) New York, Pergamon Press, 524 p. EARLY HlJ4AN POPULATIONS ALONG THE BALCONES ESCARPr.ENT Thomas R. Hester Center for Archaeological Research The University of Texas at San Antonio San Antonio, Texas 78285 ABSTRACT Evidence of prehistoric human habitation of the Balcones Escarpment region of Texas can be traced to at least ll,000 years ago. The cultural chronology of the Balcones Escarpment is divided into three major prehistoric periods: Paleo-Indian, Archaic, and Late Prehistoric. A brief overview of this cultural sequence is provided. However, the paper focuses on the early occupations of Paleo-Indian tfmes. Some paleontologfcal sftes, such as Friesen­hahn Cave, have yfelded chfpped chert flakes sug­gestive of even earl fer occupation, but the situation fs certainly not resolved at this tfme. other sftes have associated human artifacts and late Pleistocene fauna; at others, fluted pofnts, such as Clovfs and Folsom. are fndicatf ve of great antiquity. The latter part of the Paleo-Indfan period, in the early Holocene, presents some interpretative problems that are discussed here. Continuing archaeological fnvestfgatfons in thfs regfon, particularly large survey and excavation programs, need the partfcipatfon of geomorphologf&ts. Through fnterdfscipl fnary coll aboratfve research, a better Job can be done of locating deeply burfed Archafc and Paleo-Indfan sftes. Geomorphologists can contribute to the fnterpretatfon of site deposft formation and can address the problem of ancfent climatic patterns that may be revealed in gravel deposits at some sftes. INTROOUCTION The focus of this paper is on the archaeology of the Balcones Escarpment area of south-central Texas, and especially on those sftes that can be dated to late Pleistocene-early Holocene times. First a sum­mary fs provided of the prehistoric cultural chron­ology that has been established for thfs area, and then the earliest sftes found fn thfs chronology are reviewed. For purposes of geographic control, most of thfs dfscussfon fs confined to those counties along the southern and southwestern edge of the Edwards Plateau--the area traversed by the Balcones Escarpment (fig. l>. These include the counties of (from west to east) Kfnney, Uvalde, Medfna, Bexar, Comal, Hays, Travfs, and Williamson. It fs fortunate that a considerable amount of archaeological research has been done fn several of the counties, both on the escarpment and be1ow it. AREA CULTURE HISTORY Archaeologists have deffned four broad periods of prehistoric human occupation in this area: Paleo­Indfan, Archafc, Late Prehistoric, and Historic. These span some ll,000 years. Paleo-Indian Texas archaeologists use thfs term to refer to the earl fest human occupation of the state, roughly ;,. Abbott, Patrick L. and Woodruff, C.M., Jr., eds., 1986, n.e JWc.-~t,c-ml Te_, Geological Society of America, p. 55-62 Figure l. 9200-6000 B.C. The fnitial part of the perfod encompasses the late Plefstocene. Both occupation and kfll-sftes, wfth associated human artffacts and Pleistocene fauna, have been identified along the Balcones Escarpment. Whfle the datfng of the onset of the Holocene remafns somewhat ambiguous in the evaluation of many of these sftes, ft fs clear that stylfstfc and technological trafts of the projectfle pofnts of the early phase of the Paleo-Indian continue fnto late Paleo-Indfan times. We assume that population sfze, settlement patterns, and a hfghly mobfle lffeway lfkewise characterize the Paleo-Indfan cultural pattern as late as ca. 6000 B.C. 11 Locatfons of Selected Archaeological Sites Along the Balcones Escarpment. l, Fries­enhahn Cave; 2, Panther Springs site (4lBX228); 3, St. Mary's Ha11 C4lBX229), and Granberg II C4lBX27l>; 4, Orchard site C41BXll; 5, 4lBX52; 6, La Jfta site C41UV29l; 7, Kfncafd rockshelter; 8, Leona Watershed; 9, Monte11 rockshe l ter; 10, 4lBN63; ll, Schulze Cave; 12, Baker Cave; 13, Bonffre rockshelter; 14, 41VV162A; 15, Devf 11s Mouth site; 16, Gamenthaler sfte; 17, Wheatley site; 18, Canyon Reservof r CWunderlfch sfte); 19, Sprfng Lake site; 20, Levi rockshelter; 21, Wflson-Leonard site; 22, John Ischy sfte; 23, Rowe Valleysfte. Whfle Paleo-Indfan sites wfth clear evidence of Pleistocene faunal associatfons are few, the projec­tile pofnts that characterize the early part of this period (9200-8000 B.C.) are wfdespread; these fnclude Clovfs, Folsom. and Plainview pofnts (fig. 2). Simf­larly, the later phase diagnostfcs are quite convnon, even though in sftu, stratified components are fnfrequent. These df agnostfc pofnt types are: Golondrfna, Scottsbluff, Angostura (Ffg. 2), and some hfghly localized styles stfll under analysts (e.g. Barber pofnts; see Turner and Hester, 1985, for further illustrations arrd discussions of all of these types). SS • a Ffgure 2. Paleo-Indf an Pofnt Types. a, Clovfs; b, Folsom; c, Plafnvfew; d, Angostura; e, Golondrfna. Dots fndfcate extent of edge dullfng (related to haftfng technfques). Length of a, 126 mm. Drawf ngs from Turner and Hester 1985. Archafc The term "Archafc" fs used to denote a long tfme span of huntfng and gatherfng cultural patterns that began around 6000 B.C. and cont1 nued untfl 800 A.O. The perfod fs broken up fnto several subperfods, largely on the basfs of changes fn projectfle pofnt styles. along wfth shffts fn settlement patterns, other lfthfc tool forms, use of certafn plant and anfmal resources, and the lfke. Some areas of the Edwards Plateau are better known than others due to fntensfve archaeologfcal research 1n those areas. For example, Prewftt (1981) has deffned a tfghtly control led chronologfcal sequence for the Travfs, Wfllfamson, and Bell Countfes area (see also Sorrow and others, 1967), but ft cannot be applfed 1n all areas of the Edwards Plateau--the "central Texas archaeologfcal area." It fs clear that other sectfons of thfs vast regfon share some sfmflarftfes to Prewftt's sequence, but manffest--as we would expect from human cultures--localfzed dffferences (e.g. Black and McGraw, 1985). A recent overvfew of the Archafc chronology of the regfon has been wrftten by Black (n.d.), It 1 s based fn part on chronologfcal data from the Balcones Escarpment zone, from sftes such as 41BX228 (Panther Spr1 ngs; Black and McGraw, 1985), La J 1ta (Hester, 1971), and the Canyon Reservofr s1tes (Johnson and others, 1962), and buflds on other surrmar1es of local chronology (e.g. Md< f nney, 1981; Story, 1985), e The Early Archaic (6000-3000 B.C.) f s typf ffed by specific dfagnostfc dart pofnt types (Bell, Gower, Early Corner-Notched, etc.; Fig. 3) and tool forms (Guadalupe and Clear Fork fmplementsl. It is sug­gested that population densities were low and groups were organfzed into smal 1, hfghly mobfle bands. Interestfngly, many of the key sites of this era are clustered along the edge of the Balcones Escarpment • Both McKinney (1981) and Story (1985) have speculated that thfs phenomenon mfght be related to a greater ava1labilfty of water resources fn this physfographic area. durfng a hypothesfzed arid climatic epfsode in the Early Archaic t1mes. An 1mportant Early Archafc sfte fs Granberg II (41BX271) along Salado Creek in northern Bexar County (Hester. 1980), Excavat1ons were directed there 1n 1979 by thfs author. In a 3-meter excavation proffle exposed at the s1te. Glen L. Evans (Markey, n.d.) noted that the lower 2.4 meters consisted of alternating gravel strata attributable to flood deposits, pof nt bar format1on, and heavy erosfon. Early Archaic materials are mfxed within these deposfts. Further studies of the Salado Creek gravels may one day give us a better idea of clfmatic fluctuations on the edge of the Balcones Escarpment (see Black and McGraw, 1985). The M1ddle Archaic (3000-1000 B.C.) is clearly a period of population increase, with the native peoples develop1ng special1zed adaptations to the hunting and gathering of abundant regional food resources--especi a11 y acorns and white-tafled deer. The Pedernales dart point type is a diagnostic of the period (Ffg. 3), as are large accumulations of fire­cracked rock known as "burned rock middens." These F1gure 3. Archaic Artifacts. a, Early Corner Notched; b, Bell; c, Gower; d, Pedernales; e, Guadalupe tool. a-c,e, are from the Early Archaic; d, dates to Middle Archaic tfmes. Length of e, 93 mm. Draw1 ngs from Turner and Hester 1985. apparently represent fntensfve utflfzatfon of acorns. wfth the burned rock deposfts fndfcatf ve of certafn kfnds of processfng (e.g. removfng tannfc acf ds from the acorns> and food preparatfon (perhaps stone­bof 1 fng fn baskets of acorn mush. and roastfng platforms of stone for cookfng acorn bread and deer meat>. These sftes are very cOlllllOn throughout the Balcones Escarpment. Among the publfshed burned rock mfdden sftes of the Mfddle Archafc along the escarpment are La Jfta (Hester 1971; see also Hester. 1970> and the Leona Watershed sfte Call fn Uvalde County; Lukowskf, fn press), 41BN63 fn Bandera County (Hester, 1985), Panther Sprfngs fn Bexar County (Black and McGraw. 1985), Wunderlfch fn Comal County (Johnson and others. 1962>. and John Ischy in Wflliamson County (Sorrow. 1968). The Late to Terminal Archaic (1000 B.C.-A.D. 800) represents a contfnuation of the huntf ng and gathering patterns of the Archaic. with some researchers seefng less specfalizatfon. but others notfng evidence of bf son-hunting in certain areas and the presence of cemetery sites . Panther Springs in Bexar County (Black and McGraw. 1985), Wheatley in Blanco County (Greer. 1976) and Rowe Va11 ey in Wi l 1 i amson County. The peop1es of the Toyah Phase may be the ancestors of the historic Tonkawa Indians of central Texas. but this has not been clearly established; work of the sort being done by Elton R. Prewitt at the Rowe Valley site may help to resolve the issue. b Figure 4. Late Prehistoric Artifacts. a. Perdiz arrow point; b. end scraper (cross-section fs shown); c. perforator; d. beveled knffe. Length of d, 113 mm. Drawfngs from Turner and Hester 1985. Historfc Historfc Indfan sites in central Texas are extremely rare. usually represented by the occurrence of glass trade beads. natfve-made gunflints. and metal projectile points. The Tonkawa are presumed to be the historic natives of the region dat1 ng to 8,000-9,000 years ago. This 1s well w1th1n the known range of d1stinct1ve human occupations 1n the region. A mod1f1ed mussel shel 1 fragment also comes from this context, along with bones that were apparently butchered or altered by man. It 1s st111 unclear as to whether these few items represent human use of the cave or whether these are objects washed 1nto the cave from the surrounding area. It 1s my opinion, and one that 1s shared by the excavators, that 11pre-Clov1s11 artifacts are absent from the early deposits at Friesenhahn. Later human cultural materials, possibly referable to late Paleo­Ind1an times (or at least that general time frame> are found in the later sediments. At Schulze Cave 1n Edwards County, Walter Dalquest and others (1969) report three human teeth associated with Pleistocene mammalian fauna which they dated at about 20,000 years B.P. The site was dug by paleontologists, and the finds have not been studied either by physical anthropologists or archaeologists, and thus we must await further data in order to properly evaluate it. Montel 1 Rockshelter, in Uvalde County has been excavated at the St. Mary's Hall site. along Salado Creek in northern Bexar County. Two preliminary reports have been published {Hester, 1978, 1979) and work on a final manuscript 1s underway. The site is located on a colluvial downslope overlooking Salado Creek, a major tributary of the San Antonio River. The site 1s situated atop one of the highest points in the Salado valley, at an elevation of approximately 760 feet above sea level, about 35 meters west of the present stream channel. In this paper, only the materials from area A are discussed. It was within this area that a Plainview occupation was found. The typical stratigraphy is as follows: the upper unit is a brown1sh-gray m1dden with scattered burned rock, hearths, and one extensive accumulat1on of large burned rock. Th1s stratum extends to a depth of 40­50 cm. and contains Late Prehistoric artifacts 1n the upper part, with Late Archa1c {and occasional Middle Archaic) mater1als 1n the lower port1on. At 60-75 cm. there 1s a stratigraph1c un1t composed of brownish so11 and ca11che gravels and w1th1n this stratum Early Archa1c art1facts were found. Below th1s is a stratum referred to as the "gravels," composed of caliche nodules or gravels with interstices of weathered limestone clasts. Geomorphologists Dr. Charles M. Woodruff and Glen L. Evans Cpers. conn., 1977) describe the unit as having been formed by co11 uv i al s l opewash. On top of the gravel unit, Late Paleo-Indian specimens such as Golondrina and Angostura were found, with Golondrina at a lower stratigraphic position. Part of the gravel unit has been badly disturbed by the formation of cal i che "ba 11 s" or conglomerates. The mechanisms which caused the formation of these very disruptive features are poorly known. Woodruff offered two possible explanations: (l) they are a local soil phenomenon caused by underground water flow or percolation; or (2) a local ephemeral stream once coursed through a portion of the site leaving limey depos1ts. The possibility of an erosional or streamlike area 1s supported by our excavations. Fortunately, a large part of area A had been spared the presence of caliche "balls," and this was where we concentrated our excavation efforts. In the gravel units, an occupation tentatively identified as the Plainview period was found about 15-20 cm 1nto the stratum. Cultural materials were extensive and were precisely documented. This occupation w111 be focused on below. The occupation 1s considered 1n sitµ by our geological consultants and is sealed within the gravel unit. Except on the northern margins, where the caliche conglomerates occur, it is undisturbed. The best measurements available at this time ind1cate that the area utilized by the Plainview peoples 1s 8 meters long, north to south and about 6 meters wide (48 m2 or about 157 square feet). Diagnostic projectile points were clustered near the central part of this area CFig. 5). Several hundred pieces of chert ch1pp1ng debris were scattered throughout the occupation area. Other stone tool forms include trimmed or edge­modified flakes, steep bitted un1faces in the form of end scrapers, a large b1facial Clear Fork tool, a heavily worn chopper, thinned bifaces perhaps used as knives, numerous preforms (representing unfinished points> and a number of cores CFig. 5). The set of points, tri11111ed flakes, formal unifac1al and bifacial tools, some scattered animal bones (deer and b1son­sized) and burned hearthstones are indicative of campsite activit1es. The numerous preforms and cores, and the substantial amount of debitage, suggest lithic workshop activities associated with the campsite. That is, there was considerable emphasis on chert-work1ng, over and above that necessary for maintenance purposes. The cal1che gravels 1n which the materials are buried do not, apparently, indicate any part1cular climatic and environmental situation. Both Woodruff and Evans believe the "calichefication" is a normal so11 process in the s1te area; Evans believes that the occupat1on was probably originally buried in clay-loam so1ls of the type that constitute the uppermost so11 hor1zon 1n the valley today. Two Val Verde County sites have also yielded dist1nctive Late Paleo-Indian materials. The Devil's Mouth site (Johnson 1964) is a deeply buried terrace site. In area C of the site, a number of Paleo­Indian projectile points were found, and a radio­carbon date of ca. 8700 B.P. was obtained (Sorrow 1968). This date is related to the Golondrina •d c a f Figure 5. Artifacts from the Pla1nview Occupation at St. Mary's Hall Site. a,b, both sides of Plainview point; c,d, Plainview point basal fragments; e, preform (unfinished Plainview point), f, bifacial Clear Fork tool (cross-section is shown). Length of e,· 107 mm. Drawings of a-d by Margaret Greco; e,f, were drawn by Denn1s Knepper. Complex, a Late Paleo-Indian cultural pattern that was later also recognized at Baker Cave --._ Bell r' i. ' .......__ , lOCOllOn .._~Map ( / ·-..... ./ Burne1? / W1lhomson \, Milam Te•osv> I ~, Bandera J,,.". :~::~:~-) ./ Gonzoles Guadalupe Lavaca I ~ Bexar "'L-. I ;{_ __ Approximate trace of 0 , • -· main Balcones Fault ~ I .~.Location of Areas of ~ ~-......__ I ~Terra Rosso (boundaries Wilson inexact) Atascosa • I ·---....._, 0 25mi Frio Figure 1. Four areas containing major outcrops of Central Texas Terra Rossa. Terra rossa is not continuous since it has been completely removed by erosion in some areas. Silicified fossils and remnants of red clay indicate that terra rossa was once very widespread. Associated with much of the Central Texas Terra Rossa is extensive, Pleistocene silicification of Cretaceous fossils. Some other areas, not within the localities of Figure 1, yielding silicified Cretaceous fossils (Ikins and Clabaugh, 1940; Stanton, 1947; Moore, 1964) may represent Pleistocene silicification, with other evidence for terra rossa either less obvious or unrecorded. On the divide between Kerrville and Fredericksburg along State Highway 16, there is terra rossa. This terra rossa seems to be associated with the collapse of the Kirschberg Gypsum described by Barnes (1946); it may be Pleistocene. Other terras rossas, such as that on the Ellenberger Limestone just a few miles southeast of Brady on the divide between Brady Creek and the San Saba River, are probably Paleozoic. Studies of all of these terras rossas would lead to a greater understanding of the geology of Central Texas. SOIL PROFILE The soil profile for a terra rossa is not easily defined; This apparently accounts for terms such as "BC-horizon" in Duchanfour (1970). Figure 2 represents a profile from the Central Texas Terra Rossa along Loop 1604 north of San Antonio, 1.4 miles east of its intersection with U.S. Highway 287. The profile is developed on Person Formation, the upper formation of the Edwards Group (Rose, 1972), and it is not truncated at the base as are some profiles. A truncated soil is a soil in which the red clay of the B-horizon rests directly on unaltered limestone. Frequently the larger karren are not developed on bare rock, but on the limestone beneath the B-horizon. With such a truncated soil it is generally assumed that there has been some movement of soil into or onto these areas (Duchanfour, 1979), and that the base of the profile is not a natural product of mechanical or chemical weathering • .. with "' at base 0 " a: .. ::. and pseudospar breccia and pseudospar c cave-c lay 0 .. E 0 u. c 0 ., ~ "­ ., E K . Young , 1986 leve l of pavement on old highway Figure 2. Profile of Central Texas Terra Rossa overlying Person Formation on Loop 1604, north San Antonio, approximately 2.2 km. east of the intersection of Loop 1604 with U.S. Highway 287. The A-horizon and the upper part of the B-horizon have been removed by erosion. It would appear from Figure 2 that the A­horizon and at least part of the B-horizon have been removed by erosion. Most of the B-horizon is red, somewhat plastic clay. The lower part of the hori­zon is apparently B2, with limestone and caliche nodules; the latter could, of course, be post-terra rossa. If there is a C-hori zon, it is now repre­sented by a cave-fill-like breccia composed largely of pseudosparite at the base of the profile. This pseudosparite is almost identical to the pseudo­sparite, which is cave-fill associated with rede­posited, red clay, within the underlying Person Formation. According to Duchanfour (1970) such diagenetic alterations are typical of the basal parts of profiles of terras rossas. At other localities the Central Texas Terra Rossa can be seen to be truncated, and on the Kainer Formation there may be well-developed C-horizon. GEOLOGY It is generally stated that a m1n1mum of 1500 mm of rainfall per year is necessary to produce terra rossa. In addition to the minimum rainfall, Mediterranean terras rossas also required yearly dry-wet cycles, but not sufficiently dry to remove moisture from horizons below the upper part of the B-horizon (Duchanfour, 1970). I have seen no hori­zon of plinthite in the Central Texas Terra Rossa to indicate a marked seasonal change in the level of the water table. The above requirements already place certain restrictions on the age of the Central Texas Terra Rossa, because 1500 mm or more of rain­fall is not normal to Central Texas. Surficial Geology The Central Texas Terra Rossa is sufficiently ancient in the Pleistocene that it has been dissected, sometimes removed completely, and dia­genetically altered. In addition to the Central Texas Terra Rossa shown in Figure 1, other areas once covered by it, but not shown, may be indicated by accumulations of red clay in caves and shallow subsurface or by the occurrence of extensive silicification of Cretaceous fossils. The area mapped as high karstic plain by Woodruff and Abbott (1979) in the Cibolo Creek drainage was probably covered with terra rossa at one time, since the caves and collapse zones in the Person Formation contain much red clay and red­stained rock and pseudosparite (Newcomb, 1971). So~e of this area has not been studied closely, and there may still remain outcrops of terra rossa. Between Purgatory Creek and the Guadalupe River along much of Purgatory Road (locality B of Figure 1), the Central Texas Terra Rossa is associated with a pediment cutting across Kainer, Person, Georgetown, Del Rio, Buda, and Dessau formations. Much of this area is the high, karstic plain of Woodruff and Abbott (lg79). The Kainer and Person formations and the Buda and Dessau formations are in fault contact. At some localities along this pedi­ment the terra rossa has been removed. Along the Freeman Ranch-Bear Creek fault zone and along the Bat Cave fault there are rows of dolines (Figures 3 and 4) (Noyes and Young, 1960). These faults have the reverse drag on the down-thrown block (Bills, 1957; Tucker, 1968), which allowed water to flow ~aterally along the faults toward Purgatory Springs instead of down the regional dip. This further resulted in a series of dolines, probably collapse dolines, along the faults. One of these dolines, illustrated in Figure 4, is on the old Wegner Ranch on the northeast side of Purgatory Road. The silicified, fossil trees described as coming from an unconformity between the Edwards and Georgetown Formations by Cronin (1932) are actually Pleistocene. They come from the red clay that ~ccumu~ated in the dolines along Purgatory Road, 1nclud1ng the doline pictured in Figure 4. Such fossil wood should not be confused with Cretaceous wood found around the Devils River Trend. Along Loop 1604, in San Antonio (locality A, Figure 1), the Central Texas Terra Rossa occupies an area that is nearly flat but dissected locally. The streams in this area with rejuvenated meanders (Shaw, 1974) may have originated on this surface. Although there is no visible terra rossa on the Welch Ranch (now southwest Round Rock, Williamson County), terra rossa occurs just southwest of thr area (locality D, Figure 1). On the Welch Ranch, grainstone at the top of the Edwards Limestone con tained several cenotes that had been fenced so that cattle could not fall in and drown. One cenote was full of water to within one meter of the surface of the ground in 1965 when I mapped the area. That lithology was important in the development of the Central Texas Terra Rossa is suggested by its rarity on the Glen Rose Formation and its dominance on rocks on the Edwards Group. Even within the limestones of the Edwards Group, karstification of the Person Formation is much more thorough than karstification of the Kainer Formation, even though the Central Texas Terra Rossa extends uninter­ruptedly across their contacts (usually fault contacts). Terra rossa has not been observed on the formations of the Austin Group, but outcrops of Dessau Formation on the edge of the Edwards Plateau are so smal 1 and uncommon (Young, 1985, 1986) that this may have no meaning. Silica, usually microquartz, is commonly asso­ ciated with the Central Texas Terra Rossa as with other terras rossas (Duchanfour, 1970). Along Purgatory Road at localities where the terra rossa has largely been removed, the surface of the pedi­ ment is commonly strewn with fragments (mostly from 5 to 30 cm in size) of silicified rudists, largely of specimens of the genera Caprinuloidea and Texicaprina. The section of Person Formation (Figure 5) along Loop 433 South, New Braunfels, Comal County, contains about 50 percent pseudosparite and asso­ ciated, red cave-clay and cave-breccia. Other beds show that at least some of the rock was originally cross-bedded, coarse grainstone. Although no terra rossa is on the surface, red cave-fi 11 in Inner Space Caverns (Woodruff et al., }g85) near Georgetown, Williamson County, indicates a source of red clay (terra rossa) not too distant. Subsurficial Geology In this section only the shallow subsurface is considered--sufficiently shallow that there could be evidence of an overlying terra rossa or of a former terra rossa. In 1969 a coring program was carried out for a proposed quarry along Alligator (Geronimo) Creek just above the Balcones scarp, south of Hunter, Hays County. (Samples of the cores are in the collections of the Texas Memorial Museum, University of Texas at Austin.) Cores verified some of the previously mentioned surficial observations, because the Person Formation was much more highly altered, diagenetically, than was the Kainer Formation. These cores also demonstrated three intervals of formation of cave-fill. Of course there were more than three, but they cannot all be demonstrated readily. In these cores it was possi­ble to see (1) an older, unstained cave-fill, which was cut by (2) an iron-stained (red) cave-fill related to the Central Texas Terra Rossa, which in turn was cut by (3) a younger, unstained cave-fill, deposited after the period of formation of the Army, Corps of Engineers (1933). This area isFigure 3. Alignment of sinkholes along faults. part of the high karstic plain of Woodruff andMany of these sinkholes have been plugged with Abbott (1979) and lies between the Blanco Riverred clay from the terra rossa and are full of and the Guadalupe River. water at least during wet weather. This figure is somewhat reduced from that of the U. S. Figure 4. Doline on the old Wegner Ranch. This is one of the sinkholes along the Freeman Ranch­Bear Creek Fault Zone just off Purgatory Road, Comal County, Texas. "'•l•r• .., ~ l•utt conuct w11h 1ec11on I ,, ...,,,..,. p11uCIOU>•1 wllhlr1gmen11d nodut•• ol c;:her! I 0 01'11 areaa ol wacll.111ona - 15phl "'OlllJ' paeudoaoar '---·­ 1 o to 1 5 phi P••UdOJOlf • Uh red Cll)'and raMnll'llaol •oauy p11udo1oar with •1dcl1yandundlg1111d graonatona r1mn1n11 1 0 to 1 .S phi with 1c;a111r1d l1rg11 gr11n1 011udo101r rad clay , c1...1-1111 1 0 IO I .S phi 1 .s 10 2 O phi PllUCIOIOlf 1 0 IO 1 $ Ohl p11udOIOlr •tlh 101111 tld c•17 O .S to 1 0 phi IC 'f•••t. 1111 Figure 5. Section of Person Formation along Loop 433 South, New Braunfels, Comal County, Texas. The rock is approximately 50 percent cave­deposits. Figure 6. Etched and embayed chert nodule from the Fort Terrett Formation, about 20 feet below the collapse zone of the Kirschberg Gypsum, 2.0 to 2.5 km. southeast of Junction, Kimble County, Texas, on Interstate 10. Figure 7. Two views of geologic organs (solution cavities and widened fractures filled with red clay derived from overlying terra rossa) in the Person Formation, north San Antonio, along Loope 1604 just east of its intersection with Interstate 10, Bexar County, Texas. Central Texas Terra Rossa. However, the ages are relative because, at present in areas of terra rossa, cave-fill that is being deposited may be red, whereas in areas of no terra rossa the cave-fill is unstained. During coring the bit would frequently drop the full length of the kelly, and the travelling block would bounce on the rotary table. Most often, however, the caverns had been collapsed, and the recovered core would pass from limestone into red cave-clay, clay-breccia, or pseudosparite. According to Duchanfour (1970) there has been some diagenesis of all terras rossas. The diagene­~is of the Central Texas Terra Rossa has not been studied, but one can assume that it is generally one or more of the types attributed by Ellis (1985) to meteoric water. Certainly the superficial appear­ance of what may be the C-horizon (Figure 2) is similar to the pseudosparite that represents ancient cave-fill (probably derived from the terra rossa) in the underlying Person Formation. That silica (usually microquartz) is a common constituent of terras rossas (Duchanfour, 1970) does not mean that all silicification of the Cretaceous fossils occurred during the formation of the Central Texas Terra Rossa, but the many occurrences of sili­cified fossils in red cave-clay and in the profile of the terra rossa itself suggest high mobility of silica at the time of formation of the terra rossa. In the highway cuts on Interstate Highway 10 in the hills southeast of Junction, Kimble County, there is exposed a collapsed zone where the Kirschberg Gypsum has been removed by solution. Removal of the gypsum would have been most active during the period of higher rainfall represented by formation of the terra rossa. At this particular locality, for depths up to 10 meters below the base of the Kirschberg level, the undersides of the chert nodules have been deeply etched and embayed (Figure 6). This embayment presumably occurred with the abnormal salinities produced by solution of the gypsum by meteoric water. Certainly at this time silica seems to have been mobilized and redeposited by the silicification of fossils. Many fractures, widened by solution (Figure 7), and other geologic organs appear to be associated with, or were just subsequent to, the formation of the Central Texas Terra Rossa. ATTEMPTS TO DATE THE CENTRAL TEXAS TERRA ROSSA Dating the Central Texas Terra Rossa has not been easy. Approaches can be made from the fol­lowing disciplines: (1) geomorphology, (2) paleo­cl imatology, (3) diagenesis, (4) redeposition, (5) paleontology, and (6) paleomagnetism. Geomorphology The relation of outcropping areas of Central Texas Terra Rossa to earlier Pleistocene channels of the Guadalupe and Blanco Rivers (Koenig, 1940; Woodruff, 1977) indicates that this terra rossa formed prior to the capture of these streams. These captures may also correlate chronologically with primary drainage change on the Brazos River as described by Stricklin (1961) and Hibbard and Dahlquest (1967). Hibbard and Taylor (1960) con­sider major changes of drainage in Kansas and adjacent Oklahoma to be at the end of the deposition of the "Yarmouthian" (Crooked Creek Formation of Kansas = Seymour Formation of North Texas). Since "Yarmouthian" and "Illinoisian" do not mean the same to everyone, I should point out that I am using the terms as I read them in Hibbard (1970), Hibbard and Dahlquest (1967), and Hibbard and Taylor (1960). Hibbard and Dahlquest (1967) suggested that the climate of North Texas in the Late "Yarmouthian" was subhumid, mesothermal, frost-free, and maritime. Later, Hibbard (1970) dated most of the drainage changes of the Great Plains, including those of North Texas, as occurring with a change of climate from subhumid to much drier at or near the end of the "Yarmouthian." If the tectonics or climatic changes that altered the courses of the streams and resulted in rejuvenation were as regional as they seem to be, one would suspect that the streams of south-central Texas changed at the same time--that is, during the Late "Yarmouthian" of Hibbard (1970). Paleoclimatology The climatic requirements of greater humidity and particularly of greater rainfall for a terra rossa do not tell us much about age, other than that the Central Texas Terra Rossa is not recent. How­ever, the required greater rainfall (at least more than twice the present rainfall average of 700 mm per year at San Antonio) tells us that all of the Central Texas Terra Rossa was formed at the same time, because a consistent, long-term, high rainfall at different times in different local areas would be impossible. Furthermore, glacial stages would be excluded because of low temperature and probably insufficient rainfall. These data agree with those for the rejuvenation and change in stream courses mentioned above. Diagenesis Duchanfour (1970) considers the amount of diagenesis of terras rossas of the Mediterranean region too great to have been completed in the Recent Interglacial. Redeposition At some localities red clay, reworked from the Central Texas Terra Rossa, has been deposited before modern drainage developed. There is a deposit of reworked red clay on Waller Creek, just below 5lst Street in Austin, Travis County, that was trans­ported across the present area of drainage of Shoal Creek prior to the development of this drainage area. Also redeposition in caves occurred both during and after the formation of the Central Texas Terra Rossa. From all of this evidence, then, the Central Texas Terra Rossa was formed between the Kansan and the Holocene. Glacial stages would be excluded because of climatic restrictions. AGE OF THE CENTRAL TEXAS TERRA ROSSA Recently, both paleontologic and geomagnetic data have greatly reduced the margin of error in dating the Central Texas Terra Rossa. From a small cave filled with red clay derived from the terra rossa in the Murchison Quarry (Fyllan Cave Local Fauna), northwest Austin, Texas, Taylor (1982, 1986) has described a Middle Irvingtonian fauna. This fauna is Late "Yarmouthian" and pre-"Illinoisian." Furthermore, during deposition of the red clay in the cave there was a magnetic-reversal anomaly, since some samples are reversed and some are not. The reversal from Matuyama to Brunhes is known to have occurred in Late Middle Irvingtonian (Taylor, 1982). Also. the Jaramillo event just preceded the Matuyama-Brunhes reversal. If this anomaly repre­sents reversal from Matuyama to Brunhes, then the fauna is about 0.73 m.y. B.P. (Taylor, 1986). The paleo-ecological analysis of Taylor (1982) would indicate that the cave-fill for the Fyllam Cave Local Fauna was deposited after the formation of the Central Texas Terra Rossa, because the fauna seems to represent an environment of less humidity and rainfall than is required for the development of terra rossa. Both the paleomagnetic and paleoeco­logic data of Taylor (1982, 1986) agree with geomorphologic, climatic, and paleoecologic con­clusions of Hibbard and Taylor (1960), Stricklin (1961), Hibbard and Dahlequest (1967), and Hibbard (1970). Thus the age of the Central Texas Terra Rossa would appear to be older than Middle Irvingtonian (0.73 m.y. B.P.) and younger than "Kansan," probably Early and/or Middle "Yarmouthian." SUMMARY The Central Texas Terra Rossa is a widespread paleosol with all the implications ascribed to terras rossas by Duchanfour (1970). It was wide­spread before partial removal. It represents a time of higher humidity and greater rainfall (1500 mm or more per year) than occurs in Central Texas at present (about 700 mm per year). The best dates for the formation of the Central Texas Terra Rossa are Early and/or Middle "Yarmouthian," that is, between 0.73 m.y. and 2.0 m.y. B.P., prior to the Brunhes Normal and the regional climatic change at the end of or within the Late "Yarmouthian: ACKNOWLEDGEMENTS I first became aware of the Central Texas Terra Rossa in the sping of 1949, while visiting outcrops northeast of Wimberly with F. L. Whitney. The Geology Foundation and the Research Institute, both at the University of Texas at Austin, have supported a number of projects in this area, and, therefore, have unknowingly supported this project. Geologic mapping for the Bureau of Economic Geology on other projects also aided this presentation, and Patrick L. Abbott and C. M. Woodruff, Jr., encouraged me. Jeff Horowitz prepared Figure 1, and Rosemary Brant prepared the camera-ready copy. REFERENCES CITED Baldwin, Mark, C. E. Kellogg, and James Thorp, 1938, Soil Classification: p. 979-1001 in Hambridge, G., ed., Soil and Men: United States Department of Agriculture Yearbook, Washington, D.C., United States Government Printing Office, 1232 p. Reprinted as p. 145-168 in Finkl, Charles W., Jr., ed., Soil Classification: Benchmark Papers inS'oil Science/!, Stroudsburg, Penn., Hutchinson Ross Pub1 i shi ng Co., 391 p. Barnes, Virgil E., 1946, Gypsum in the Edwards Lime­stone of Central Texas: Austin, Texas, University of Texas, Publication 4301, p. 35-46. Bills, T. V., 1957, Geology of the Waco Springs Quadrangle, Comal County, Texas: Austin, Texas, University of Texas, unpublished thesis, 110 p. Cronin, Stewart, 1932, Disconformity between Edwards and Georgetown in Hays and Comal Counties: Austin, Texas, University of Texas, unpublished thesis, 32 p. Duchanfour, P., 1970, Precis de P~dologie: Paris, Masson et Cie., 481 p. Ellis, Patricia Mench, 1985, Diagenesis of the Lower Cretaceous Edwards Group in the Balcones Fault Zone area, south-central Texas: Austin, Texas, University of Texas, unpublished thesis, 326 p. Hall, R. D., 1976, Stratigraphy and origin of surfi­cial deposits in sinkholes in south-central Indiana: Geology, v. 4, p. 505-509. Hibbard, Claude W., 1970, Pleistocene mammalian faunas from the Great Plains and Central Lowland provinces of the United States: p. 395-433 in Dort, Wakefield, Jr., and J. Knox Jones, Jr., Pleistocene and Recent environments of the Central Great Plains: Lawrence, Kansas, University of Kansas Press, Department of Geology, Special Publication 3, 433 p. Hibbard, Claude W., and Walter W. Dahlquest, 1g&7, Fossils from the Seymour Formation of Knox and Baylor Counties, Texas, and their bearing on the Late Kansan climate of the region: Ann Arbor, Michigan, University of Michigan, Contributions from the Museum of Paleontology, no. 21, p. 1­ 66. Hibbard, Claude W., and Dwight W. Taylor, 1960, Two Late Pleistocene faunas from southwestern Kansas: Ann Arbor, Michigan, University of Michigan, Contributions from the Museum of Paleontology, no. 16, p. 1-233. !kins, W. S., and S. E. Clabaugh, 1g40, Some fossils from the Edwards Formation of Texas: Ithaca, New York, Bulletins, American Paleontology, v. 26, no. 96, p. 261-282 (1-22). Joffe, Jacob S., 1949, Pedology: New Brunswick, New Jersey, Pedology Publications, 662 p. Koenig, J. B., 1g40, A consideration of the Blanco River terraces north of San Marcos: Austin, Texas, University of Texas, unpublished thesis, 41 p. Moore, C. H., 1964, Stratigraphy of the Fredericksburg Division, south-central Texas: Austin, Texas, University of Texas Bureau of Economic Geology, Report of Investigations 52, 48 p. Newcomb, John H., 1971, Geology of Bat Cave Quadrangle, Comal and Bexar Counties, Texas: Austin, Texas, University of Texas, unpublished thesis, 104 p. Noyes, A. P., Jr., and Keith Young, 1960, Geology of Purgatory Creek area, Hays and Comal Counties, Texas: Texas Journal of Science, v. 12, p. 64­ 104. Quinlan, James F., 1978, Types of karst, with empha­sis on cover beds in their classification and development: Austin, Texas, University of Texas, unpublished thesis, 323 p. Rose, P. R., 1972, Edwards Formation, surface and subsurface, Central Texas: Austin, Texas, University of Texas Bureau of Economic Geology, Report of Investigations 74, 197 p. Ruhe, R. V., 1975, Geohydrology of karst terrain, Lost River Watershed, southern Indiana: Bloomington, Indiana, Indiana University, Water Resources Research Center, Report of Investigations 7, 91 p. Shaw, S. L., 1974, Geology and 1 and-use capability of the Castle Hills Quadrangle, Bexar County, Texas: Austin, Texas, University of Texas, unpublished thesis, 130 p. Stanton, T. W., 1947, Studies of some Comanche pele­cypods and gastropods: Washington, D.C., United States Government Printing Office, United States Geological Survey Professional Paper 211, 256 p. Stricklin, Fred L., Jr., 1961, Degradational stream deposits of the Brazos River, Central Texas: Geological Society of America, Bulletin, v. 72, p. 19-35. Tanasijevic, D., G. Antonovic, z. AleksiC', N. Pavicevic, o. Fil ipoic, and N. Spasjevic, 1966, Pedologic cover in western Serbia: Belgrade, Institute of Soil Science; translated by Bozidar Filipovic, and published in English by the United States Department of Agriculture and the National Science Foundation in 1968, by Nolit Publishing House, Belgrade, 279 p. Taylor, Alisa J., 1982, The mammalian fauna from the mid-Irvingtonian Fyllan Cave Local Fauna, Travis County, Texas: Austin, Texas, University of Texas, unpublished thesis, 105 p. Taylor, Alisa J., (1986, in preparation), The mam­malian fauna from the Mid-Irvington Fyllan Cave Local Fauna, Travis County, Texas. Tucker, Delos R., 1968, Lower Cretaceous geology, northwestern Karnes County, Texas: American Association of Petroleum Geologists, Bulletin, v. 52, p. 820-851. U. S. Army Corps of Engineers, 1933, Hunter, Texas, Quadrangle: Fort Sam Houston, San Antonio, Texas, Eighth Corps Area Engineer, Special Edition, Tactical Map, 2509: 3750/99. Woodruff, C. M., Jr., 1977, Stream piracy near Balcones Fault Zone, Central Texas: Journal of Geology, v. 85, p. 483-490. Woodruff, C. M., Jr., and Patrick L. Abbott, 1979, Drainage-basin evolution and aquifer development in a karstic limestone terrain, south-central Texas, U.S.A.: Earth Surface Processes, v. 4, p. 319-334. Woodruff, C. M., Jr., Fred Snyder, Laura De La Garza, and Raymond M. Slade, Jr., 1985, Edwards Aquifer--northern segment, Travis, Williamson, and Bell Counties: Austin, Texas, Austin Geological Society, Guidebook 8, 104 p. Young, Keith, 1985, The Austin Division of Central Texas: p. 3-51 in Young, Keith, and C. M. Woodruff, Jr., Austin Chalk in its type area-­stratigraphy and structure: Austin, Texas, Austin Geological Society, Guidebook 7, 88 p. Young, Keith, 1986, Cretaceous, marine inundations of the San Marcos Platform, Texas: Cretaceous Research, v. 7, p. 117-140. f ·-;:.·:; ::~·-----.-·· -. : ~ :.-::.: .. ~-~--~:~::·_· ..... --........ -....--; GAY FEATHER STRUCTURAL STYLE IN AN EN ECHELON FAULT SYSTEM, BALCONES FAULT ZONE, CENI'RAL TEXAS: GEOMORPHOLOGIC AND HYDROLOGIC IMPLICATIONS Thomas W. Grin~haw c. M. Woodruff, Jr. Radian Corporation Consulting Geologist P. O. Box 9948 P. 0. :Rox 13252 Austin, Texas 78766 Austin, Texas 78711 ABSTRACT iu twc 7-1 /2 o.inute te>p0graphic qt:adr angles in the rapidly growing San Marcos area (Grimshaw, 1976). Detailed geologic mapping in the Balcones fault zone in the San Marcos area has revealed a structural style The principal features of the case study area, besides that may have had a profound effect on the geomorpho­the city of San Marcos, are Interstate 35, the town of logic and hydrologic evolution of the area. Two uiajor Kyle, the Blanco River, Purgatory Creek, Sink Creek, en echelon step fault zones are present in the area, San Marcos Springs, and the San Marcos River (Figure and a highly faulted ramp structure has formed in the 1). zone between the en echelon fault zones. Differential erosion of rock units in the ramp struc­ ture may have determined the course of a stream which captured the Blanco River from an easterly flow direction into the Onion Creek basin to its current A 150 southeasterly flow direction. Subsequently, the Blanco may have "tapped" the Edwards aquifer by down-cutting or side-cutting action at or r.ear tbe San Marcos Springs location. Thus, both the capture of the Blanco and the current location of San Marcos Springs hay have been indirectly caused by the local structural setting between the two major en echelon fault zones of the Balcones system. Similar major ramp structures are apparent by map inspection in at least three other locations in the Balcones fault zone, one near Austin and two west of San Antonio. A fourth structure may also be present near New Braunfels. INTRODUCTION The Balcones fault zone is a tensional structural system consisting of numerous normal faults, cross faults, grabens, horsts, step faults, en echelon faults, and similar features in central and south Texas. The fault zone extends from Waco south1o1ard to Austin and San Antonio and then westw!lrd to Del Ric. Generally, the rocks exposed at the surface west of the fault zone are Lower Cretaceous stn1tigraphic units consisting of resistant limestones, dolomites, and marls; east of the zone, the rocks exposed are Upper Cretaceous nonresistant chalk and calcareous clay units. The difference in resistance to erosior: has resulted in a fault-line scarp known as the Balcones Escarpment. Soils east of the scarp are deep and well developed, and the predominant historical agricultural land use has been for crorland. West of the scarp, the soils are thin and rocky, and ranchland is the the predon.inant agricultL·ral land llf'E'. The Balcones Escarpment and fault zone are especially well developed in the area around San Marcos, Texas 123 about 35 miles south of Austin. The purpose of this paper is first to describe the structural style in a case study area between two major en echelon step fault systems and then to set forth hypotheses on geomorphologic and groundwater implications of the zone of adjustment between these en echelon fault systems. The discussion of structural style and resulting outcrop patterns is based upon detailed Figure 1. Principal Features Of The San Marcos Case geologic mapping for environmental geologic purposes Study Area. m Ahholl. Patrid l. and Woodruff. ( .M .. Jr., eds .. 1986. The llakone. EKarpment, Ceatnd T~us: Geological 51.x.·icty of America. p. 71-76 71 GEOLOGIC OVERVIEW A generalized geologic map of the San Marcos area is sh~wn ~n Figure 2. The principal rock-stratigraphic units in the area, in descending stratigraphic order, are shown below. All are Cretaceous in age. Approximate Unit Thickness (ft) Taylo~oup (clay) 92S Austin Group (chalk) 170 Eagle Ford Formation (clay) 2S Buda Formation (limestone) so Del Rio Clay so Georgetown Formation (marl) 33 Edwards Group (limestone) 47S Major faults of the Balcones system traverse the study area from northeast to southwest. The net displace­ment in the case study area, as elsewhere in the Balcones fault zone, is downward to the southeast. The faults of major displacement strike about N30 E. The Edwards Limestone crops out over IIIOst of the area west and north of these faults, and the Austin chalk and Taylor clay compose the subsurface east and south of the faults. Outcrops of the thin intervening formations between the Edwards and Austin occur in numerous fault blocks within the fault zone. The regionally important Edwards aquifer is especially significant in the San Marcos area, both as an essen­tial source of copious fresh water and as a recrea­tional resource associated with San Marcos Springs. These springs are a major discharge point of the Edwards aquifer; discharge averages about 161 million gallons per day. The extensive outcrop area of the Edwards limestone in the western and northern portions of the area is an important part of the aquifer recharge zone. The Balcones fault zone enters the northeastern part of the case study area as two major step faults -­Mustang Branch fault and Mountain City fault (Figure 2). These faults are referred to hereafter as the "northeastern step fault zone". The Kyle fault, located farther to the southeast, is another step fault in this succession. Similarly, the Balcones fault zone is represented in the southwest part of the area by three major step faults ~ Comal Springs fault, San Marcos Springs fault, and Bat Cave fault, referred to hereafter as the "southwestern step fault zone". The major northeastern and southwestern step fault zones are thus in en echelon relationship across the study area. The fact that fault traces are not at all influenced by topography indicates that all faults are vertical or nearly so. STRUCTURAL Hll'ERPRETATION The cumulative displacement across the northeastern step fault zone generally decreases to the southwest. Similarly, cumulative displacement across the south­western step fault zone decreases to the northeast. Thus, the total displacement downward to the southeast remains relatively constant across the study area, but is "transferred" from the northeastern to the south­western step fault zone as is typical for en echelon fault zones in a tensional fault system like the Balcones. []] Alluvium [I] Taylor Group OJ Austin Group [I] Georgetown, Del Rio, Buda, and Eagle Ford Formations D:J Edwards Group + Fault Stratigraphic Contact Figure 2. Simplified Geologic Hap Of San Marcos Area, Showing Step Fault Zones and Associated Faulting In The Area Between. The transfer of displacement between the step fault zones is shown in block diagram form in Figure 3. The area of adjustment between the step fault zones forms a ramp-like structure which bends downward to the northeast from the upthrown side of the southwestern step fault zone to the downthrown side of the north­eastern step fault zone. If this ramp were eroded to a level surface parallel to the bottom of the block in the diagram, an outcrop pattern with older rock units exposed in the southwest and successively younger units exposed to the northeast would be displayed. Just such a pattern is exhibited in the case study area, where Edwards limestone exposures in the south­ west part of the area, on the upthrown side of the southwestern fault zone, give way northeastward to Austin chalk and Taylor clay bedrock on the downthrown side of the northeastern step fault zone. The ramp structure is broken up into several irreg­ularly shaped grabens, horsts, and step fault blocks which are, in turn, broken up into numerous small, irregular fault blocks as small as 100 yards (or less) in dimension as shown in Figure 4. The overall structural grain of the faulting in the ramp is consistent with the regional Balcones fault zone, with the larger-displacement faults having a northeastward orientation, but with the smaller cross faults oriented in all directions. The geometry of the larger grabens and horsts, as well as the much smaller individual fault blocks within them, suggests that the ramp was subjected to some torsional stresses, in addition to the dominant Figure 3. Block Diagram Of En Echelon Step Fault Zones Showing Intervening Ramp Structure, tensional stresses, as the southeastern part of the area "dropped away" from the northwestern part and the ramp area adjusted to the transfer of displacement from the northeastern to the southwestern step fault zone. Figure 4. Detail From Geologic Map Of San Marcos Area, Showing Intensity Of Faulting In The Ramp Structure Area (from Grimshaw, 1976). The resulting overall outcrop pattern. with the Edwards giving way to the Austin and Pecan Gap north­eastward along the ramp. and with the complex. irreg­ular faulting between the en echelon step fault zones, is shown clearly in Figure 2. The intervening units between the Edwards Limestone and the Austin Chalk are exposed in the complexly faulted area extending generally southward from the northeastern step fault zone to the southwestern step fault zone. The outcrop pattern on a larger scale and in more detail is a mosiac of irregular fault block outcrops having an appearance not unlike a shattered pane of glass (Figure 2). The complex faulting of the ramp into irregular grabens and horsts at the smaller scale (Figure 3), and the small, irregular fault blocks at the larger scale (Figure 4), represent the adjustment of the ramp area to the tensional and lesser torsional stresses during the Balcones faulting. The intensity of faulting depicted in Figure 4 is most clearly displayed in the band of outcrops of inter­vening stratgraphic units between the Edwards lime­stone and the Austin chalk in the graben and horst area between the en echelon step fault zones. It is likely that this intensity of faulting also exists in the areas of exposure where the Edwards, Austin, and Taylor Groups are exposed. but the lithologic homo­geneity of these units does not allow the individual fault blocks to be mapped in such detail. It is only where the succession of thin, lithologically dis­similar units represented by the Georgetown (marl), Del Rio (clay). Buda (limestone). Eagle Ford (clay with thin siltstone layers) is affected by the intense faulting that sufficient stratigraphic control allows the very small individual fault blocks to be mapped. Along the northeastern section of the southwestern step fault zone there is a major, high-standing fault block, herein named the San Marcos horst, which is undisturbed by the intricate faulting that charac­terizes most of rest of the area northwest of the step faults. The cause of this large. undisturbed, mono­lithic fault block remains problematical. The Bal­cones Escarpment is especially well developed along the eastern margin of this horst; the Old Main buil­ding of Southwest Texas State University is built on this prominent scarp and is a striking local landmark which is easily seen by travelers on nearby Interstate 35. Much of the western half of San Marcos is built on this large fault block. and San Marcos Springs discharges along its northeastern margin. POTENTIAL GECMORPHOLOGIC IMPLICATIONS: A HYPOTHESIS Woodruff (1977), in a discussion of development of drainage patterns near the Balcones Escarpment, has shown that the Blanco River formerly discharged into what is now the Onion Creek drainage basin. The "elbow" in the course of the Blanco River in the northern part of the case study area is near the point at which a headward eroding smaller stream captured the Blanco and diverted its flow from generally eastward to a southeastward direction. The point designated "A" in Figure 1 is the location of a distinctive erosional feature which is interpreted to be a former channel of the pre-capture course of the Blanco. This feature is now located on the drainage divide between the Blanco River and Onion Creek drainage basins. Not long after Balcones faulting occurred (currently believed to be during the Miocene). the stratigraphic units exposed in the vicinity of the en echelon faults and associated ramp structure in the study area were Upper Cretaceous or younger. It may reasonably be expected that as erosion occurred and the land surface lowered in the area, the more resistant Austin chalk was exposed in the southwestern part of the ramp structure while the less resistant Taylor clay was still present in the northeastern part of the struc­ture. Further, because of marked difference in erodibility of these units, it may be expected that a small east-facing escarpment would have formed along the ramp. The hypothesis or question then arises. "Could this small escarpment then have determined the course of the stream which ultimately captured the Blanco River?" If the answer is in the affirmative, then the physiographic development in the area was controlled by the presence of the en echelon faulting and asso­ciated ramp. and the present southeasterly course of the Blanco was ultimately determined by the presence of the ramp. POTENTIAL HYDROLOGIC IMPLICATIONS: ANOTHER HYPOTHESIS Woodruff and Abbott (1979) have hypothesized that the current principal discharge points of the Edwards aquifer, such as San Marcos Springs and Comal Springs, have developed at or near the locations where major streams and rivers traverse the Balcones Escarpment. In effect, these locations represent the lowest elevation points where the Edwards limestone is exposed. There, the actively downwardly eroding rivers have opened drains for the aquifer. After the Blanco River was captured into its current southeastward course, its grade was increased and the currently visible deeply incised canyon where the river crosses the Edwards limestone was formed. This canyon opens to a wider valley containing alluvium at the point where the river crosses a fault and begins to flow on outcrops of younger strata. primarily the Aus~in chalk (Figure 2). The valley widens again, this time to a much greater extent, at the point farther downstream where the river begins to flow on the Taylor Clay. Although the river is still downcutting after it passes out of the Edwards, it is also effectively sidecutting as evidenced by the presence of alluvium in the valley in this stretch. The continued down­cutting is indicated by the presence of bedrock exposures in the river channel. The thickness of the alluvium and the presence of a single sequence of fining upward grain size pattern in the alluvium indicate that the river is removing and redepositing its alluvial deposits as it shifts course in its valley. The question naturally arises: "Was the Blanco River responsible for cutting a drain into the Edwards aquifer at the current location of San Marcos Springs?" The current locations of the river. the springs. and Sink Creek indicate that if the Blanco did open this discharge point of the aquifer by downcutting or sidecutting at the location of the springs. it occurred many years ago when the land surface and the river were at somewhat higher eleva­ tion than at present. If the hypotheses outlined above and in the previous section are correct, then the current location of San Marcos Springs was determined by the presence of the Blanco River, whose location was in turn determined by the northeastern and southwestern en echelon step faults and the associated ramp structure. The role of the San Marcos horst in this sequence of events awaits further study. POSSIBLE SIMILAR STRUCTURAL PATTERNS ELSEWHERE IN THE BALCONES FAULT ZONE Inspection of maps of the Geologic Atlas of Texas published by the University of Texas Bureau of Econo­mic Geology (Barr.es, 1974a, 1974b, 1974c) suggests that similar structural patterns exist elsewhere in the Balcones fault zone. For example, a similar ramp structure is indicated in the Austin area, where outcrops of Edwards limestone in an en echelon step fault setting give way northeastward to Austin chalk outcrops. A band of outcrops of Georgetown, Del Rio, Buda, and Eagle Ford Formations extends southward from the Colorado River to the town of Buda. This band separates Edwards outcrops to the southwest from Austin chalk outcrops to the northeast, just as in the San Marcos area. A smaller scale but similar pattetn may also exist near New Braunfels, where outcrops of Edwards lime­stone gives way northeastward to Austin chalk outcrops in an apparent ramp structure, with a transition zone of Del Rio, Buda, and Eagle Ford Formation outcrops between. Southwest of New Braunfels, the en echelon faulting in the Balcones fault zone reverses, with major displace­ment shifting from southeastward fault zones to northwestward fault zones. Ramp structures in this area are thus also reversed, with bending downward to the southwest rather than to the northeast as in the San Marcos area. One such "reversed" ramp structure, where Edwards outcrops give way southwestward to Austin chalk outcrops, and with intervening exposures of Georgetown through Eagle Ford fault blocks, is apparent immediately northwest of San Antonio. Another larger one appears to be present across most of Media County and far eastetn Uvalde County. SUMMARY Detailed geologic mapping of intensely faulted Creta­ceous strata in the Balcones fault zone in the San Marcos area has revealed a structural style and bedrock geometry which have potential implications for the geomorphologic and hydrologic development of the area. ToucASIA MoNOPLEURA The current location of the Blanco River and San Marcos Springs may have resulted from a sequence of events whose course was determined by this structural style and the relative resistance of the Cretaceous units to erosion. The presence of the two en echelon step fault zones and associated ramp structure may have been the ultimate cause of both the current, captured course of the Blanco River and subsequently the location of San Marcos Springs. Similar en echelon structural pattetns with ramps are indicated in at least three other locations in the Balcones fault zone. REFERENCES CITED Barnes, V. E., 1974a, Austin sheet: The University of Texas at Austin, Bureau Of Economic Geology, Geologic Atlas of Texas, scale 1:250,000. Barnes, V. E., 1974b, San Antonio sheet: The University of Texas at Austin, Bureau Of Economic Geology, Geologic Atlas of Texas, scale 1:250,000. Barnes, V. E., 1974c, Seguin sheet: The University of Texas at Austin, Bureau Of Economic Geology, Geologic Atlas of Texas, scale 1:250,000. Grimshaw T. W., 1976, Environmental Geology of Urban and Urbanizing Areas: A Case Study From the San Marcos Area, Texas: The University of Texas at Austin, Ph.D. dissertation, 244 p. Woodruff, C. M., Jr., 1977, Stream piracy near the Balcones fault zone, central Texas: Journal Of Geology, v. 85, p. 483-490. VERBENA BIPINNATIEIDA PRAIRIE VERBENA STREAM PIRACY AND EVOLUTION OF THE EDWARDS AQUIFER ALONG THE BALCONES ESCARPMENT, CENTRAL TEXAS C.M. Woodruff, Jr. Consulting Geologist P.O. Box 13252 Austin, TX 78711 INTRODUCTION Three river systems dissect the southern margin of the Edwards Plateau in south-central Texas: the Nueces, the San Antonio, and the Guadalupe (Fig. 1). The Edwards Plateau is a karstic upland, and its dissected margin consists of plateau outliers, narrow incised stream courses, and intervening areas of steeply sloping terrain kIXJWn locally as the Central Texas Hill COUntry. The upper reaches of these three drainage basins constitute an ~rtant hydrogeologic entity; they include the catchment watersheds, the recharge areas, and the points of discharge for the central segment of the F.dwards artesian aquifer. The Edwards aquifer is a major cavernous limestone system that extends for over 400 Jan along the BalOCl'leS fault zone fran Val Verde County on the Mexican border to Bell County in north-central Texas (Fig. 1) • The central part of the aquifer is the m:>at prolific water-yielding segment and thus is the main focus of this report; it constitutes the main water Sut4>1Y for a region that includes the city of San Antonio and a population of ioore than one millioo people. Sooe attention also will be given to the 390 km2 Barton Springs segment that lies i.Jllnediately north of the central aquifer segment. Major exchanges on a regional scale occur between surface stream flow and groundwater levels in the central segment of the aquifer (Sayre and Bennett, 1942; Pettit and George, 1956; Arnl:Tw, 1963; Klemt and others, 1975; WOOdruff and Abbott, 1979). In brief, m:>at recharge occurs within the semiarid western part of this aquifer segment, while DDst discharge occurs in the subhumid eastern portion. Interactions between the surface and subsurface water regimes are likely to have occurred during earlier developnental stages of both the aquifer and the surface drainage network. It is our purpose to 800w that drainage-basin evolution and aquifer developnent have ~ated DUltually. That is, within larger structural geologic and climatic controls, physiographic developnent near the Balcones fault zone predetermined both geographic configuration and magnitudes of recharge and discharge in the Edwards aquifer. Moreover, aquifer developnent has influenced the evolution of surface drainage configurations by the diversion of surface flow via recharge in one area while augmenting stream flow via spring discharge elsewhere. We propose that these relations are due in large measure to stream piracy that chiefly occurred within the San Antonio and Guadalupe watersheds. Similar piracy also affected landform and hydrologic developnent farther north within the Bart.al Springs segment of the Edwards aquifer (Woodniff, 1984a). in Abbott, Patrick L. •nd Woodruff, C.M., Jr., eds., 1986, The IWcGnes ~I,C-tnl Teua1 Geological Society of America, p. 77.90 77 Patrick L. Abbott Department of Geological Sciences San Diego State University San Diego, CA 92182 Stream piracy greatly increased the catchment area of the through-flowing Hill Country rivers where they cross resistant limestone strata within the fault zone (Woodruff, 1977; Woodruff and Abbott, 1979). The higher average rainfalls that occur in the pirated basins also enhance the ability of these streams to cut deep canyons. Deeply incised canyons were necessary to provide spring sites for groundwater that otherwise would have been trapped and then equilibrated chemically with host rocks and therefore would have ceased the dissolution of the surrounding limestones (Abbott, 1975). If piracy had not occurred, the dynamic hydrologic situation would not have developed as rapidly and the formation of cavernous porosity would have been retarded. As it happened, a region-wide circulation system developed during two diverse time periods. The first was near the middle of the Cretaceous Period when the Edwards Limestone was deposited, exposed subaerially, and buried. The second was during the Miocene Epoch when Balcones faulting occurred, and the erosion of the fault-rejuvenated streams exhl.lllai the Edwards Limestone. Eventually, extensive cavern systems developed as the main conduits for transmission of groundwater. GIDUXiIC SE'ITING The structural and stratigraphic frameworks of the study area are the basic controlling factors for both surface and subsurface drainage developnent. 1-k>st aquifer developnent occurred within rocks of the Edwards Group that were deposited on the San Marcos platform and in the F.dwards-equivalent limestones of the Devils River trend (Fig. 1). The stratigraphic and lithic characteristics of the (Albian Stage) Edwards Limestone originated with the differing depositional environments, and resultant facies, of late Early Cretaceous time. To swmiarize Rose ( 1972) , the San Marcos platform acted as an area of lesser subsidence during the time of Edwards deposition. That platform was the site of accumulation of about 150 m of shallow marine and tidal flat sediments. At the same time along the Devils River trend roughly a 300 m thickness of grainstone and rudist boundstone was formed. Subsequent uplift along the oortrwest-trending axis of.the San Marcos platform caused irore than 30 m of the uppenrost Edwards Group to be removed by erosion during late Early Cretaceous ti.Die (Fig. 2) • Subaerial erosion of carbonate rock was accarpanied by pore-space enlargerent and cavern developnent resulting fran circulation of shallow meteoric waters. The part of the Edwards Group that makes up the present aquifer in Bexar, Canal and Hays COUnties-(where discharge daninates today)-was on the San Marcos platform and received significant ----, I enhancement of porosity during Cretaceous time. The parts of the aquifer in Medina, Uvalde and Kinney Counties--(where recharge dominates at present)­were _sou~est of _the axis of uplift and apparently received little, if any, solution enlargeirent of porosity during Cretaceous time. During the remainder of the Early Cretaceous and throughout Late Cretaceous time the entire region was covered by shallow marine shelf waters. Deposition of argillaceous and micri tic sediments resulted in the F.dwards Group being covered on the San Marcos platform by a 260 m thickness of low-perneability rocks. This burial sealed off the F.dwards Group and precluded the circulation of ground-water necessary to further increase porosity. At about the end of the Cretaceous, slow upwarping of the nort!Mestern margin of the subsidin;r Gulf of Mexico basin lifted the region of the present-day F.dwards aqui fer above sea level. Continued deformation gave a generally southeastward dip to the sedimentary rock units of central and south Texas. At this time, deep groundwater might have augmented earlier developed porosity except this groundwater system would have been largely static, having no means for egress (Abbott, 1975). This stasis would have resulted in chanical equilibration between host rock and the waters contained therein, thus preventin;r extensive cavern developnent at that time. The dcmi.nant geologic feature in the region is the Balcones fault zone, a system of en echelon mainly down-to-the-ooast, normal faults that ~ about 545 km fran Del Rio on the Mexican border to near waco in north-central Texas. Faultin;r probably occurred primarily during the late Early Miocene (Youn;r, 1972), as evidenced by the abundance of reworked Cretaceous fossils and limestone fragments in the fluvial sandstones (calclithite) of the oakville Formation (Wilson 1956; Ely, 1957). The strike of individual faults within the study region is predaninantly northeast-souttwest, but the overall structural aligment subtly chan;Jes to a ioore east-west trend in the soutl'western part of the region. Faulting within the San Antonio, Guadalupe, and Colorado River basins has juxtaposed the awrox:i.nately 150 m-thick Edwards Limestone against the older Cretaceous Glen Rose Formation which consists largely of alternatin;J beds of limestone, dolanite, and marl. On the downthrown (eastern) side throughout the region the F.dwards Limestone abuts against less resistant chalk, clay, and marl units of youn;rer cretaceous age. Displacement alon;r the main fault-line scarp is as little as 60 m in the westerrm::>st part of the region, whereas a maximJm displacement of about 185 m occurs in the Guadalupe River basin (Klemt and others, 1975). Similarly, total stratigraphic NORTHEASTWEST I ()P(N SHELF I -~ I~,-:_~~~-~~ :1~~---~~~_J~I ~ B LATE CRETACEOUS I QJAOAWPŁ Rr\ (R BAS1'11 L~Ut• ':>,.Id I I C MIOCENE .., iCES 1'11\.fR &ASN Q_ PLE1STOCENE/HOLDCENE ------------' Figure 2. Schematic cross sections of stages in the devel~t of the Edwards Aquifer (m::xilfied fran Rose, 1972, and Abbott, 1975). displacement decreases fran east to west. Total displacement in canal County is as much as 520 m over a maxinun width of 39 Ian (George, 1952). This fault-boond exposure of linestone has resulted in CXllpartmentalization of the aquifer into a narrow belt that includes IOOSt of the recharge and discharge areas within the eastern basins. Farther west, to..Onent streams of the Nueces system have flowed generally southward throughout their developnental history. CLIMATIC SE'ITitli Besides bedrock conditions, another controlling factor affecting water regimes and land.form develoEJllellt is climate. In the western part of the region the climate is semiarid, with mean annual rainfall as lc:M as 48 an in sane areas (Fig. 3). This, coupled with high evaporation rates, means that streamflc:M and erosional potential of western streams are necessarily lc:Mer than that of the subhumid eastern basins. The Guadalupe River basin, for exanple, lies in the center of an oblong 80-c:m isohyet, whereas the 50-c:m isohyet follc:Ms the West Nueces River (Fig. 3). Climatic differences are especially reflected in magnitudes of streamflc:M. For example, Guadalupe River, draining 3,932 kili2 where it crosses the Balcones escarpnent, has a mean flc:M of 10.54 m3/s. This is about three times larger than the canbined discharge of the Nueces and West Nueces Rivers where they flc:M together south of the fault zone; they have a watershed of 5,043 kili2 with a mean discharge of 3.12 m3/s. Hc:Mever, these values reflect water losses caused by infiltration into the aquifer by streams of the Nueces River system above and beyond those <:Ming to rainfall deficiencies or increased rates of evaporation. Cooparing stream gage date upstream and dc:MnStream fran the fault zone, it is seen that where the East and West Forks of the Nueces River converge belc:M the recharge zone their basin areas increase 74 percent while their mean total discharge decreases 59 percent (U.S. Geological Survey, 197 4) • No such recharge loss is included in the Guadalupe River water budget. These differences between the Guadalupe and Nueces flc:M regimes danonstrate the self-ramifying conditions that are evident in many limestone aquifers. Where water is mai.ritained predaninantly in surface flc:M, more stream erosion and thus incision can occur. Where water infiltrates underground, not only is there a lessened amount available to perfonn surface erosion, but, because of soluble bedrock, these recharging waters enlarge their flc:M paths, thus ensuring further underground infiltration. HYDROI.OGIC SE'ITitli The F.dwards aquifer consists of two carpments­-an unconfined (water-table) aquifer in the plateau lands and Hill Country upstream fran the main fault­line scarp of the Balcones fault zone, and a confined (artesian) aquifer within the eastern and southeastexn part of the fault zone. Recharge to the water-table aquifer results fran precipitation occurring throughout nuch of the F.dwards Plateau: this catchment area extems beyond the drainage basins CCXll>OSing this study region (Fig. 4). Groundwater beneath the F.dwards Plateau moves mainly tc:Mard the southeast dc:Mn the regior..'il dip of the aquifer: part of this water discharges through myriad seeps and springs that provide base flc:M for headwater streams in the Nueces, San Antonio, Guadalupe, and COlorado River basins. Surface streams sustained by this spring-derived base flow eventually cross the highly fractured, cavernous limestones in the Balcones fault zone. There, infiltration into the confined aquifer occurs. In addition, about 6 percent of the recharge into the F.dwards occurs by underflc:M fran adjacent rock units such as the Glen Rose Formation. This recharge moves directly into the confined aquifer without having been discharged first as surface flc:M (William B. Klemt, writ. oami., 1977). Major recharge occurs in two types of terrane­stream bottoms underlain by faulted or cavernous limestone, and low-relief uplands underlain by karstic limestones. The more inl>ortant of the two recharge areas is where stream courses cross NUECES RIVER BASIN 55% long-term, regional recharge 9% tota l d ischarge EXPLANATION Edwards L1mestone--plo1eou 1erro1n i recnor<;ie zone for unconfined aquifer f.dwords L1mes1one--d1ssec1ed hill countr y; locus of many small springs i recharge 1n kors11c plains ona sireom bo11oms Łd..,..ords L1mes1one--foull zone i recharge along most stream bonoms i discharge from springs Mo1or d1rt!,l1ons and qcncrohzed magnitude of water -movement In unconf1r.eC1 aquifer In confined aquifer J\quifer (IOOdified from Woodruff and Abbott, 1979). permeable limestone. Water-budget studies in the Barton Springs segment of the aquifer have shown that about 85 percent of incident rainfall is cycled through evapotranspiration, about 9 percent runs off, and the remaining 6 percent recharges the aquifer. Of the recharge fraction, about 85 percent occurs within stream bottans (Slade, 1984; Woodruff, 1984b). Recharge zones along bottanlands are especially apparent because stream discharges decrease through these reaches, dry or nearly dry stream beds ccmnonly are incised into bedrock and there is a conoanitant attenuation of alluvial deposits. About 55 percent of the estimated annual recharge into the central segment of the confined aquifer is supplied by the ocmponent streams of the Nueces basin (Fig. 4). Most natural discharge occurs fran springs along the Balcones escarµnent­notably fran canal Springs and San Marcos Springs. Most well discharge occurs in the San Antonio area, and well puopage is increasing with growing popllation demands. Total discharge fran wells TOI often exceeds total discharge fran springs (Klant and others, 1975). In the eastern part of the 0 JO ml f--r--J 0 10 km ~ajar droinooe divide Effluen1 stream lnfluen1 s1reom "Bod -water line" Major springs I-Leona Springs 2-Son Pedro Springs 3-San Anionic Springs 4-Cornol Springs 5-Hueco Springs 6-Son Morcos Springs 111112 a:rrrr:tI> lJndedlow from unconfined 10 confined aquifer--__,zoo-EEdle;o0~~5°~0:~'~~=~~q-~iler olono stream bo11oms where Edwords L1mes1one 1s con11nuousl y exposed Figure 4. Recharge-discharge relations and potentiometric levels, central segment of the Edwards central aquifer segment the yields fran water wells are greater, discharge fran springs is more volwn:i.nous, and water levels tend to fluctuate more uniformly, cxxrpared to aquifer-discharge characteristics farther west (Table 1). These observations imply that cavernous porosity is best devel~ near the distal end of the groundwater flc:M system in the areas farthest resooved fran the major loci of recharge. The sane relations are seen in the smaller Barton Springs segment(Slade, 1984). Furthenrore a oounty-by-oounty enumeration of vadose caverns conducted by the Texas Speleological Survey documents an increase in the nuni:>er of caves fran west to east along the Balcones fault zone. As of August, 1977, there were 22 surveyed caves in Kinney County, 63 in Uvalde County, 39 in Medina Coonty, 81 in Bexar Coonty, 92 in Canal County, and 86 in Hays County. Thus, caves surveyed in the recharge zone number 123 cxxrpared to 259 in the discharge end of the fault zone (R. Fieseler, writ. cxmn., 1977). Similarly, caverns have been shown to be major conduits for groundwater flc:M in at least part of Table 1. Long-term recharge and discharge from central segment of the Edwards artesian aquifer. (Data from unpublished U. S. Geological Survey source, The Edwards Underground Water District, and the Texas Department of Water Resources) RECHARGE DISCHARGE Springs Wells Mean Annual Mean Annual Extremes (hm3l hm:l Extremes (hm 3) Mean Annual Extremes (hm3) Basin Area hm3 High Low (period of High Low hm3 High Low km 2 11934-1971) (Year) (Year) record) (Year! (Year} 11934-1971) (Year) (Year} NUECES RIVER SYSTEM Nueces/West Nueces River 5,043 124.3 507 .1 10.6 Leona Springs 24.9 54.3 + lnterfluve : Nueces/ Dry 119351 (19341 11963-19751 119731 Frio 91 insig. Frio/Dry Frio Rivers 1,712 112.4 369 .9 5.2 lnterfluve: Frio/Sabinal 142 insig. Sabinal River 640 39.9 275.9 0.7 119581 119551 lnterfluve : Sabinal/ f\l\edina 1.228 75.1 363.6 4.4 119581 119561 Subtotal, Nueces System 8,856 369.2 24.9 38.6 159.5 5.1 11971) 119341 SAN ANTONIO RIVER SYSTEM Medina River 1,738 63.9 128.2 7.8 (19601 119561 lnterfluve : Medina/Cibolo 813 71.0 235.4 2.5 San Pedro/San 16.2 67.2 + 119581 119561 Antonio Springs (1963-19751 (19631 Cibolo Creek 710 82.6• 311 .8 1.5 (19571 119561 Subtotal , San Antonio System 3,261 217.5 16.2 212.8 321 .6 118.0 11971) (1934) GUADALUPE RIVER SYSTEM Dry Comal Creek 337 46.8• 178.8 119571 0.4 119551 fComal Springs 250.8 476 .7 insig . Guadalupe River 3,932 insi g. · · · . . . . . ....... . 11928-19741 (19731 119561 Hueco Springs 32.7 116.9 + (1944-19741 119681 Blanco River 1.311 39.4 94.2 1.4 San Marcos 143.8 143.8 41 .1 Springs 11956-19741 119731 (1956) Subtotal, Guadalupe System 5,580 86.2 427 .3 7.6 20.5 2.5 11971 I (1937) TOTAL 17,697 672.9 468.4 259.0 Total Mean Annual Discharge 727.4 + Numerous Periods of No Flow •Period of Record 1954·1973 the artesian aquifer. Blind catfish have been found in waters discharged fran wells as deep as 610 m in the San Antonio area (Hutl:Js, 1971). A notable cave fa\llla also exists in the waters of San Marcos Springs (Upa and Davis, 1976; Holsinger and Longley, 1980). QUFSI'IONS AND HYPCmiESF.S There are anomalies in this regional hydrologic picture. The drier western areas subsidize (by recharge) the water supplies for the areas with higher perennial rainfall and streamflCM. Evidently, tqx>graphy is one main control of the recharge-discharge couplet; recharge occurs primarily at higher elevations such as occur in the Nueces Basin and discharge occurs mainly at lCM points. Moreover, the total catchment area of the CCIJIX>nent streams in the Nueces basin represents 49 percent of the areas of the three major basins of the region. A nruch larger drainage-catchlrent area plus a base flCM augmented by spring discharge fran the unconfined aquifer on the F.ciwards Plateau, apparently oarpensates for a decrease in precipitation in these westernmost basins. But why has erosion been less in the Nueces watershed? Why was there an initial impetus for transfer of water fran the semiarid west to the subhumid eastern part of the region? Why does the largest single, integrated basin in the region (the Guadalupe River) contribute insignificant amounts of recharge where it crosses the fault zone? Why did the eastern river systems incise irore vigorously into l~topographic levels to create initial discharge sites that controlled aquifer developnent while the Nueces basin was left "high and dry?" The hypotheses posed here are that present surface drainage conditions, aquifer recharge-­discharge relations, and direction of potentianetric gradient can be explained by several geologic determinants and by the activity of processes that have occurred as a result of these determinants. Fracturing and displacement of pre-existing strata set into 100tion the overall drainage evolution of the region. These structural events affected rocks that, in turn, reflected their specific histories of deposition, diagenesis, and weathering. In brief, faulting established the structural grain that: (1) controlled cavern developrent during post-Miocene time, (2) established the tqx>graphic breaks that localized rapid stream incision and (3) provided gross lateral bo\mdaries for the aquifer host rock in the eastern river basins. There are several processes which acted on this structurally prepared ground. Mechanical and chemical erosion by surface streams occurred in response to a change in base level. Erosion, torever, did not occur equally in all areas. The diversion of large volwnes of surface flCM in eastern basins by stream piracy locally enhanced capabilities for surface erosional processes. oawnwasting in the eastern basins also was abetted by higher rainfall rates. Because of increased dcMncUtting, 1C7fl topographic levels were reached that intersected the aquifer and allowed a few loci for discharge of groundwater to become established. Near the intersection of the Balcones escarpnent with the major drainage courses, the surface flC7fl is allographically low points for spring discharge that bec:ane engrained as base levels toward which JOOSt of the artesian aquifer flowed. AIXi 2) in the Guadalupe River basin, incision occurred at such a high rate that JOOSt of the upper aquifer levels were ooupletely breached, and discharge fran the aquifer (instead of recharge into it) bec:ane the major process. Part of the aquifer system draining to San Marcos Springs ~tlydoes extend beneath Guadalupe River, and the Edwards Li.Irestone crops out along a short reach of this deeply-incised river. Yet no long-term recharge is sl:x:iwn to have occurred for Guadalupe River (Table 1). This anomaly may be explained by the relatively small volume of caverns within the part of the aquifer that wrlerlies Guadalupe River. Thus, the pore space beneath Guadalupe River may be essentially full of water under normal climatic cooditions, and only during extreme drought conditions might this cavern system be able to aocept recharge fran Guadalupe River. This thesis is substantiated to sare extent by the lesser fluctuations of discharge fran San Marcos Springs during ti.Ires of drought carpared to the normal~y larger canal Springs (Brune, 1975). Not only is canal Springs probably more adversely affected by increased discharge by well-p.unpage in the San Antonio area, but San Marcos Springs may be recharged to sare extent during very dry years by Guadalupe River--a condition that does not occur during wetter ti.Ires sinply because the cavernous pore space within the small segment of the aquifer that underlies Guadalupe River is normally filled to capacity. Where Guadalupe River crosses the Balcones fault zone, its drainage area increases by 15 percent while its mean annual discharge increases by 34 percent (U.S. Geological Survey records-courtesy of Texas Natural Resources Information System). This increased rate of flow across the fault zone is a result of higher rainfall in the middle part of the basin, input of water fran Hueco Springs, and presumably little or no recharge into the Edwards aquifer. During the peak drought year of 1956, l"lcMever, discharge increased across the fault zone by only 12 percent--a decrease in expected flow of about 84 11m3. This is a 36 percent decline caipared to long-tenn yearly averages. Assuming that the drought's effects on the water budget had equal inpact throughout the river basin, there are only two means for effecting this relative decrease in discharge: diminished flow fran Hueco Springs or infiltration into the Edwards aquifer. Hueco Springs did iooeed experience decreased flow: no spring discharge was recorded during 1956 (Texas Board of Water Engineers, 1959). But in order to attribute all of the decreased Guadalupe River discharge through these reaches to diminished spring discharge would require a decline of 2690 l/s in the long-term rate of flow fran Hueco Springs. There is no exact figure for mean annual discharge fran Hueco Springs because of their erratic flow. However, Brune (1975, p. 38) cited 3710 l/s as the maximum discharge for Hueco Springs. He further listed 33 spring-discharge rates measured over a period of 48 years, the highest of which was 2322 l/s while the lowest was zero. Moreover, all 136 discharge measurements for Hueco Springs cited by the Texas Board of Water Engineers (1959) present values less than the average rate required to account for the CXJll>Uted water loss fran Guadalupe River. It is likely that sare fraction of the diminished flow in Guadalupe River during drought periods is a result of recharge into the Edwards aquifer. IN'l'EX>RATION OF l..ANJ1"Utographic levels. As soon as the Edwards Limestone was breached by the pirate streams, pent-up groundwater was released fran the proto-aquifer, thus beginning the engrail"lnent of the flowpaths of groundwater roving taoiard these few discharge points (see Figs. 2c, 2d). Notwithstanding the presence of primary porosity, Cretaceous solution-enlarged porosity, and Miocene fracture porosity, an effective groundwater flow system could not have developed until the overlying blanket of fine-grained sedimentary rocks was breached, thus exposing the soluble limestone to both recharge and discharge. Continual region-wide groundwater circulation developed as oarponent streams of the Nueoes system exhumed the i;..\\\\\\\~ LOCu$ of loult1nq -Incipient ucorpment .......---.... Miocene stream courses Inferred 1ome 01 modern 1rends -·--.,. lnleHed tor mer 11reom courses d1tltrtnl from modern lre!Kls Modern drolnooe dlvldu ----... _ lnlerrt'd lormtr dralnooe dJvlClu \\\\\\\\\\ Balcones EM:orpmen1 "-.,.. Locus of pruumeCI ropld heodword ero11on Pre·cop1ure llrtom COJr1u Modtrn slrtom courH p. Korst1c plain olom;i Slreom courws MOgraphically higher western areas to the lower, more permeable discharge points to the northeast. This set in motion the continuously circulating, self-ramifying groundwater flow system that converged toward the loci of the few springs. In this way, the initial discharge sites becaire the "drains" for the central segment of the aquifer, drawing an waters througOCiut the several drainage basins that encx:q:iass more than 17,695 km2. In the same way, Barton Springs becaire the drain for the underground watershed carprising the surface catchment areas of Barton, Williamson, Slaughter, Bear, Little Bear, and Onion Creeks. The ancient engrai.nlrent of the aquifer helps explain noteworthy features of the present :Edwards aquifer system such as the paucity of springs and the origin of the "bad-water line." Although the :Edwards aquifer in the fault zone is about 400 km long, there are only about a dozen large springs discharging fran the system (8ayre and Bennett, 1942). Six of the major springs occur in the 280 km-long central segment: Leona Springs in Walde eounty; none in Medina County; san Antonio and san Pedro Springs (four km apart and rising along the same fault) in Bexar County; Comal and Hueco Springs in canal C.ounty; and 8an Marcos Springs in Hays C.ounty (Fig. 4). These springs issue forth at progressively lower elevations to the northeast, and all but Leona Springs occur near major pirate streams. They probably discharge fran enlarged lower-level conduits of the initial flow systems honed to the earliest discharge sites. That is, the loci of discharge probably migrated to progressively lower topographic levels near fault traces as dictated by changing base levels. The controlling base levels were in turn established by vigorously downcutting pirate streams. The general lowering of interfluvial areas and exposure of the F.dwards Limestone over large areas have not caused more springs because the plumbing system was engrained long ago toward the few original discharge sites. The southeastern boundary of the aquifer is a "bad-water line" that separates the potable water of the cavernous, high-yield aquifer fran the high salinity water on the downdip side. Although the bad-water line is roughly parallel to the trend of the Balcones fault zone, its detailed course generallly disregards individual faults and facies boundaries (Abbott, 1975; Woodruff and others, 1982) • It can be understood as the solution­engrained original flow bouOOary of groundwater that moved toward the earliest discharge sites. This hydraulically-controlled bouOOary probably marks the down-dip potentiometric boondary as originally affected by the subtle draws of low elevation springs. Initially,-recharge probably occurred through karstic plains at high tqx>graphic levels within the san Antonio and Guadalupe watersheds (Fig. Sc), and much of the cavern developnent was confined to iOOividual fault blocks in a trend s~allel to the modern system that lies beneath the western edge of the Gulf Coastal Plain. When streams trending normal to the fault scarp effected piracy, the greater voltureS of flow rapidly cut through the high-level karstic plains and stranded sane of the subjacent cavern systems that fonrerly recharged their local areas. Upland karstic plains are the relicts of this fonrer equilibrated landform/process couplet. The post-piracy couplet enhanced while the developrent of caverns at lower topographic levels, as is presently seen along Cibolo creek and Medina River (coopire Figs. 2d and Sc). These low-lying karst plains provide a periodic influx of recharge undersaturated with respect to calcite and dolanite which is so i.rrportant for the continued solutional growth of the :Edwards aquifer. The recharge-discharge geanetry described here does not fit the general concept which holds that dissolution is concentrated near recharge sites where grourx'lwater is least saturated with respect to calcite and dolanite and thus is most aggressively able to form caverns. If this oomoon view of cavern formation held for the :Edwards artesian aquifer, then the cavern systems in Kinney, UValde, and Medina Counties should be more highly developed, and yields fran springs and wells should be greater there. Likewise, cavern developnent in Bexar, Comal, and Hays Counties should be nuch less than it is, because groundwater that has traveled a long distance should be saturated or supersaturated with respect to calcite and dolanite and hence unable to ac:x:oDt>lish any further dissolution. If the groundwater had been mostly saturated during the developnental history of the aquifer, then caverns in the eastern areas would be poorly developed, and yields fran wells and springs would be low. Since the caverns are best devel~ near the distal end of the groundwater system, then clearly the groundwater passing through Bexar, Comal, and Hays Counties has primarily been undersaturated with respect to calcite and dolanite. Today groundwater within the :Edwards aquifer is at least seasonally undersaturated, as shown by negative saturation indices calculated for calcite and dolanite from well and spring water ~les neu canal, 8an Marcos, and Hueco Springs by Pearson and Rettman (1976) and by Abbott (1977a). The mechanism that maintains at least seasonal undersaturation in the ground-water appears to be the mixing-