Geologic Approaches to the Determination of Long-Term Coastal Recession Rates, T-~ .1 Matagorda Peninsula, Texas U

Geologic Approaches to the Introduction Shoreline erosion and recession re- Determination of Long-Term present a serious environmental problem in the Texas Coastal Zone. Understand- ing of the processes responsible for Coastal Recession Rates, shoreline recession depends not only on analysis of the lateral variation in reces- Matagorda Peninsula, Texas sion rate but also on an analysis of the temporal variation of rate for any spe- cific area. Recession rates commonly are determined through comparison of a chrononological sequence of charts, B. H. Wilkinson maps, and/or aerial photography. Deter- Department of Geology and Mineralogy, The University of Michigan Ann mination of the temporal variation of Arbor, Michigan 48109, U.S.A. rate with such data are limited by the short time period over which measure- ments are taken. In many cases, accurate coastal charts are available for only a century or so, and aerial photographs J. H. McGowen are available for only a few decades. As a result, the use of historical data for Bureau of Economic Geology, The University of Texas analysis of variation of recession rates in Austin, Texas 78712 U.S.A. most areas is limited. McGowen and Brewton (1975) examined historic shoreline changes in the Matagorda Bay area, Texas. They indi- cated that between 1856 and 1957 erosion was a natural process along the Gulf ABSTRACT / The degree to which human modifications in the coastal zone shoreline and that erosional shoreline have increased or decreased coastal erosion rates is difficult to determine trends were established at least 118 owing to the short time period for which shoreline-position data are available. years ago before arty significant modifi- This limitation is circumvented in areas where long-term recession rates can cation of the coastal environment by be determined from geologic data. Three such areas from Matagorda Bay man. Given their data on historic reces- have been examined to determine the temporal variation in recession rates sion rates, one may then question wheth- over the past several thousand years. Preliminary results indicate that er the rates of erosion determined reflect recession rates over the past century may be 30% to 40% greater than those natural conditions or if they have been of prehistoric time_ Although additional data are needed, it is suggested that significantly accelerated or decelerated accelerated rates result from human modification of the coastal zone, and by the activities of man. Comparison of that in the future increased recession rates can be anticipated. long-term recession rates (encompassing several millenia) with short-term recession rates (several decades) along three sections of Matagorda Peninsula pro- vides a temporal context in which the short-term data can be evaluated.

Geological Setting

Matagorda Peninsula is a narrow barrier [peninsula] which separates Mata- gorda Bay from the Gulf of Mexico (Fig.

Environmental geology Vol. 1, pp. 359-365 1977 by Springer-Ver[ag New York Lnc, 359 360 + H. Wilkinson and J. H. McGowen

./ / 2 2 3 '

1). This segment of the coast has been and dissected during the Wisconsin low- temporal variation in erosion and reces- migrating landward for several thousand stand of the Gulf of Mexico. sion rates. years. Shoreline recession is a result of Most of Matagorda Peninsula is in an daily removal of sediment along the erosional state and has been so for many Central Matagorda Peninsula beach by waves and transport of sedi- decades. This is due to several processes Recession rates spanning a 900-year ment bayward into Matagorda Bay dur- described by McGowen and Brewton period can be estimated for Central ing hurricanes. Numerous hurricane (1975) which include: (1) a deficit of Matagorda Peninsula. In this area the surge channels, which cross the peninsu- sand-size material made available to lit- subaerial vegetated barrier is separated la normal to the Gulf shoreline, attest to toral currents from erosion of shelf and from the subaquaeous muds of Matagor- the frequency of this catastrophic pro- shoreface sediments and from rivers to da Bay by a broad, fiat, sand platform cess. Matagorda Peninsula is a sand the northeast, and (2) human activities which extends from the lagoonward body that overlies older bay muds which such as construction of dams across margin to water depths of 1 meter. This accumulated in Matagorda Bay. These major fluvial systems, rice farming adja- platform terminates abruptly as water muds have been overridden progressive- cent to major streams which requires' depths increase to nearly 3 m; bay muds ly by the peninsula during its landward large volumes of water, construction of accumulate at these depths (Fig. 2). retreat. Bay muds in turn overlie an ir- jetties, and river diversion. Analysis of Shoreline retreat in this alea is accom- regular surface developed on Pleisto- the relative importance of these factors plished primarily during the passage of cene sediments which were weathered is possible through the determination of hurricanes when the beach and shore- Geologic Approaches to the Determination of Long-Term Coastal Recession Rates, t-~ .1 Matagorda Peninsula, Texas U /

Figure 1, Matagorda Bay Area. Arrows rates can be made for a part of Matagor- gorda Peninsula accreted over bay muds indicate areas discussed in the text. (a) da Peninsula utilizing a radiocarbon age can be estimated. These data indicate Central Matagorda Peninsula area; (b) determination of 900--- 30 years B.P. that average recession rates for Mata- Southwestern Matagorda Peninsula area; from Crassostrea virginica (the edible gorda Peninsula in this area have been (c) Pass Cavallo area. oyster) recovered from bay mud just approximately 0.9 m/year for the past below the sands which make up the bar- 900 years. rier. The age of the oysters closely ap- In deriving this estimate, two as- proximates the time when sediment de- sumptions have been made: (1) the rate rived from the shoreface and beach of of sediment accretion on the bayward face are eroded, and sediment is trans- the peninsula was transported across the edge of the sand platform closely reflects ported into the adjacent bay. During barrier to accumulate in the area (Fig. 3). the rates of Gulf shoreline erosion, and these brief but high energy events, the The horizontal distance between the (2) the oyster reef grew on the bay bot- entire peninsula migrates bayward. The hole from which this material was re- tom fairly close to the edge of the bay- bayward margin of the sand platform covered and the present bay margin and margin sand platform, and the time of represents the landward limit of sedi- platform edge are known. Since the burial by bayward accretion is closely ment transport during these events. width of this platform is fairly constant reflected by the age of these shells. Nei- An estimate of long-term recession throughout this area, the rate that Mata- ther assumption can be conclusively

2. Dip section across central Figure MATAGORDA PENINSULA Matagorda Peninsula showing the ~Qy-morgLns~3nd plQtform distribution of barrier sands and bay Gu!f of Mexico Matogord~ Boy muds. Shell of Crassostrea collected from a reef penetrated at a depth of 6m. o Dashed isochronous lines indicate 7 bayward accretion of sand and shell resulting from storm surges.

~" -.--~- :~ --~L - - -- -::-_~---..-_':

0 I/2 ] Km i J i

Figure 3. Distribution of the subaerial MAT,4GORDA BAY barrier, bay-margin sand platform, and 0 I 2 Km deeper mud-bottom bay along Central Matagorda Peninsula. Long-term ..... --- A ..~, --...... ~ ~:2mlm recession rate calculated by dividing the horizontal distance between the sample hole and the sand platform margin by 900 ' years. Short-term recession rate from McGowen and Brewton (1975). ;iii ii{!i? iii iiii;i! iiiii iii) i iiii;:iii:

1856-1956 recession rate - 1.6 m/yr.

GULF OF MEXICO 62 B.H. Wilkinson and J. H. McGowen

Figure 4. Development of an erosional unconformity on eastern Matagorda Island. (a) Closing of the tidal pass (Pass Cavallo) by deposition on the flood delta; (b) Tidal pass closed, erosion smoothes the Gulf shoreline by removal of part of __ ,,. ~,J~ __ _~,, __ .... ~_ -~ -- eastern Matagorda Island; (c) Pass Cavallo as it appears today, following reopening of the tidal pass, erosion of the !iiii!iii!ii!!iiiiiiii!i!!iiiiiiii!iiiiiii!ii!i!iiiiiiiiiiiiiiii!iiii!!i - flood delta by waves in Matagorda Bay, accretion of eastern Matagorda Island, !i!!:! !iiiiilJ!i!iiiiiilili!iii!i!iiiiii!i!iiiilililili! and bayward retreat of Matatorda Peninsula. Dashed line represents the position of the Gulf shoreline 1500 years B.P. Modified from McGowen and Brewton (1975). defended, but several lines of reasoning suggest that they are warranted. In order to estimate long-term recession rates, net bayward accretion must be the same as net erosion on the beach and shoreface. Matagorda Peninsula exhibits little variation in the width of the bay margin sand platform, or in the width of the subaerial barrier. This re~ quires that the long-term variations in both shoreline erosion and bayward accretion are insignificant laterally. If long- term coastal erosion was variable along the Gulf shoreline, the beach would not be straight, nor would the width of the C subaeriai barrier be constant. Similarly, if rates of bayward accretion were laterally variable, the landward margin of the bay margin platform would not be straight, nor would the width of this platform exhibit a laterally uniform width. As a result, long-term rates of Gulf erosion and bayward accretion must be laterally similar. If accretion rates and erosion rates for this segment are dissimilar, this must be reflected in a long-term gorda Peninsula suggest that this is prob- Figure 5. Estimated recession rate for change in the combined widths of the ably not the case. Southwestern Matagorda Peninsula. subaerial and subaquaeous peninsula. The possibility that the Crassostrea Long-term rate calculated by dividing the Although this possibility cannot be di- reef grew some distance from the edge of horizontal distance between the Gulfward rectly evaluated, the uniformity of width the bay margin platform and that it is projection of the unconformity along along the peninsula's entire length, and considerably older than the overlying Matagorda Peninsula, and the present the close agreement between calculated sands also cannot be directly evaluated. position of the Gulf shoreline of accretion rates behind central Matagor- Shell material was collected from a l-m Matagorda Peninsula by 1,500 years. da Peninsula and shoreline recession zone beneath the base of the barrier Short-term recession rate from McGowen rates estimated for southwestern Mata- sands. If this reef were significantly old- and Brewton (1975). Geologic Approaches to the Determination of Long-Term Coastal Recession Rates, 363 Matagorda Peninsula, Texas

er than the sands which buried it, one Matagorda Peninsula retreated 156 m Matagorda Island began to accrete, se- would anticipate the presence of some between 1856 and 1956. This rate, 1.6 quentially developing its present form. [thickness of] bay mud overlying the ree- m/year, is 40% greater than that estimat- This period of erosion and reprograda- fal mass. Data on rates of argillaceous ed from long-term geological data. tion is clearly revealed as a line of un- sedimentation in this area are few; how- conformity between the ridges and ever, Shepard and Moore (1960) report- Southwestern matagorda swales on the island's eastern end. ed rates of 1.07 m/century for mud depo- peninsula The period of erosion which sition in this part of Matagorda Bay. If a Recession rates for a period of 1500 smoothed the Gulf shoreline records the maximum thickness of 1 m of mud sepa- years can be estimated for southwestern position of Matagorda Peninsula relative rates this reef from the overlying sands, Matagorda Peninsula where Pass Caval- to Matagorda Island at the time the un- calculated bayward accretion rates Io, a large active tidal pass, separates conformity developed. A projection of would be too rapid by only 11% or 0.1 Matagorda Peninsula and Matagorda the line of unconformity on Matagorda m/year. We believe that the uncorrected Island (Fig. 1). Matagorda Island is a Island to the northeast, approximates recession rate of 0.9 m/year is more ac- wide sand-rich barrier which, unlike the position occupied by Matagorda curate. Matagorda Peninsula, has accreted Peninsula at that time. A series of ra- Short-term recession and erosion throughout most of its history diocarbon dates indicate an age of 1500 rates for this segment of Matagorda Pen- (Wilkinson, 1975). An accretionary his- years B.P. for the unconformity of Mata- insula are highly variable. Shepard tory is reflected in the pronounced ridge- gorda Island (Wilkinson 1973). Compar- (1973) reported that as much as 244 m of and-swale topography paralleling the ison of the present shoreline of south- shoreline recession occured in this area modem shoreline along most of the is- western Matagorda Peninsula with the during the passage of hurricane Carla in land. Accretion of Matagorda Island was projected shoreline position 1500 years 1961. McGowen and Brewton (1975) interrupted at one time by erosion which ago (Fig. 5) yields a recession rate of 1.3 reported that this same area had accret- truncated ridges and swales along the m/year for this time interval. ed as much as 154 m between 1961 and island's eastern end. The causes of this Because of the construction of jetties 1971 (this is 90 m short of the pre-Carla erosion are not well understood. Mc- along the Matagorda ship channel in shoreline position). Clearly, severe Gowen and Brewton (1975) suggested 1965, present erosion rates along this storms cause high variation in short-term that erosion resulted from the closing of shoreline segment are highly variable. accretion and recession rates. Pass Cavallo and smoothing of the Gulf Data from the Galveston Office, U.S. McGowen and Brewton also show shore by wave action (Fig. 4). Pass Cav- Army Corps of Engineers indicates that that in this area, the Gulf shoreline of alio was later reopened, and eastern between 1964 and 1971 1.46 x 106m a of

_,z.

~_,,,L~,.~-~/ ]856-1956 recession rQte-l.9 m/yr.\ ,~.:.i~

:,.

~~ /~-~ 1,500 yr. recession .../:i ...... ( rate-l.5 m/yr. "~.~!:i taUlr snorenne I,aUU yrs tJ.H. ,,

ii:ili!iiiii!i::iii:! o . I I I 364 B.H. WilkinsonandJ. H. McGowen

Figure 6. Strike cross section across MA PA Southwestern Matagorda Peninsula, Pass Cavallo, and eastern Matagorda Island showing the distribution of Holocene Facies. PA refers to the position of the paleoaxis of Pass Cavallo 3000 years ago. MA designates the modern axis of the )!!: pass.

~--~ Bay mud L~.I~ Barriersand sp,t-re,o,od she,,ysooo TidGI delta muddysend Pleistocenemud

Figure 7, Estimated erosion rate for Western Pass Cavallo. Long-term rate OI ]I 2I 3I 4i 5I Km calculated by dividing the horizontal distance between the paleoaxis and the modern axis by 3000 years. Short-term erosion rate from McGowen and Brewton (1975).

_i,~: -~~") /~--PrePr sent ~ ~~ the northeastern end of Matagorda Is- land and simultaneous deposition of sediment on the southwestern end of Mata- j~ ~ ~5//6_ 1956 er.osionrate, 5.5 m/yr. PClS5oxis- 3,O00 yrs. B.P ~ ' gorda Peninsula. Facies relationships of Matagorda Peninsula indicate that the axis of Pass Cavallo originally was 11.4 km northeast of its present position. This is the northeastern limit of shelly spit- associated sands, indicating that Pass Cavallo did not extend northeast of this point. Location of the paleoaxis of Pass Cavallo 11.4 km northeast of its present position coincides with a deep valley cut sand was removed from the beach and channel. This rate is 34% greater than into Pleistocene sediments during Wis- shoreface southwest of the south jetty of that estimated from long-term geologic consin glacial advances (Fig. 6). Morton the ship channel. McGowen and Brew- data. and Donaldson (1973) also found that ton (1975) reported historical recession major tidal inlets are stabilized over old- rates from a point midway between the Pass Cavallo er Pleistocene valleys on the southern present ship channel and the southwest- Pass Cavallo has been migrating to Atlantic coast. ern end of Matagorda Peninsula of 1.9 the southwest since still-stand of the The time that Matagorda Island and m/year for the time period 1856- 1956, Gulf of Mexico, 3000 years ago. Migra- Matagorda Peninsula originated is not which predates construction of the ship tion is accomplished by the erosion of accurately known. Several sources Geologic Approaches to the Determination of Long-Term Coastal Recession Rates, 365 Matagorda Peninsula, Texas

(Frazier 1974; Nelson and Bray 1970; ties and seawalls, (2) subsidence due to their relationship to the Quaternary strati- Curray 1960) suggest that these barriers groundwater withdrawal, and (3) dam- graphic framework in the northwestern ming the major sand-contributing rivers. portion of the Gulf Basin: Univ. Texas, are no older than 3000 years. If an age of Austin, Bur. Econ. Geol. Circ. 74-!, 3000 years can be assumed to be reason- It is difficult at this point to ascertain the 28p. ably accurate, migration of Pass Cavallo future impact of man-made modifica- McGowen, J. H., and J. L. Brewton, 1975, 11.4 km in 3000 years yields an average tions on the coastal zone. Clearly, reces- Historical changes and related coastal annual erosion rate for northeastern sion has been a natural process in these processes, Gulf and mainland shorelines, Matagorda Island of 3.8 m/year (Fig. 7). three areas for a long time. Recession Matagorda Bay area, Texas: Univ. Texas, Austin, Bur. Econ. Geol., Spec. Pub., Short-term erosion rates for this will undoubtedly continue in these areas, 72p. coastal segment are fairly uniform along and may accelerate as man's activities Morton, R. A., and A. C. Donaldson, 1973, the western side of Pass Cavallo. Mc- continue to interrupt natural sediment Sediment distribution and evolution of tid- Gowen and Brewton (1975) reported dispersal systems. Much additional data al deltas along a tide-dominated shoreline, erosion rates directly southwest of Ma- on long-term rates and processes are Washapreague, Virginia: Sed. Geology, v. tagorda Peninsula of 5.5 m/year between needed before the impact of these modi- 10, p. 285- 299. 1856 and 1956. This rate is 33% greater fications can be fully evaluated. How- Nelson, H. F., and E. E. Bray, 1970, Stratig- than that determined from long-term ever, the long-time data available for this raphy and history of the Holocene sediments in the Sabine, High Island area, geologic data. area suggests that short-term recession Gulf of Mexico, in Morgan, J. P., and R. H. rates may represent conditions of acce- Shaver, eds., Deltaic sedimentation, mod- Conclusions lerated erosion, and indicate that this ern and ancient: Soc. Econ. Paleontolo- trend toward more rapid recession may gists and Mineralogists Spec. Pub. No. 15. p. 48- 77. Data available for three areas along continue. Matagorda Peninsula allow one to calcu- Shepard, F. P., 1973, Submarine geology: New York, Harper and Row, 517 p. late short-term shoreline erosion, and to Shepard, F. P., and D. G. Moore, 1960, Bays estimate long-term erosion rates. The ACKNOWLEDGMENTS of central Texas Coast, in Shepard, F. P., differences between long-term (geologi- F. B. Pitleger, and T. H. van Andel, eds., cal) and short-term (historical) rates for The paper was read and criticized by Recent sediments, northwest Gulf of each area is remarkably consistant. In Erwin Seibel, Assistant Director, Sea Mexico: Tulsa, Okla. Am. Assoc. Petro- each of the three areas, short-term rates Grant Program, The University of Mich- leum Geologists, p. 117- 152. appear to be 30% to 40% greater than igan, and C. G. Groat, Acting Direc- Wilkinson, B. H., 1973, Matagorda Island: those for long-term intervals. Reasons The evolution of a Gulf coast barrier com- tor, Bureau of Economic Geology, The plex: Univ. Texas, Austin, unpub. Ph.D. for accelerated erosion rates are not University of Texas at Austin. Support dissertation, 178 p. precisely known. Increased erosion in for portions of this work were provided -- 1975, Matagorda Island Texas: The the past century may be due to a variety by a Faculty Research Grant from the evolution of a Gulf coast barrier com- of natural processes including those re- Rackham School of Graduates Studies, plex: Geol. Soc. America Bull., v. 86, lated (1) to climatic change which re- The University of Michigan. We thank 959-967. suits in decreased sediment supply the Galveston Office of the U.S. Army from fluvial systems updrift, (2) to Corps of Engineers, who provided sev- increased water depths along Mata- eral logs of borings made in southwest- gorda Peninsula by removal of shoreface ern Matagorda Peninsula. Publication is sediment, (3) to increased rates of dune authorized by the Director, Bureau of development which removes sand from Economic Geology, The University of the littoral areas, or (4) to decreased sed- Texas at Austin. iment supply by fluvial sources as streams decrease their gradients by valley down-cutting. With data available at REFERENCES present, it is not possible fully to evalu- ate the significance of each process. Curray, J. R., 1960, Sediments and history of Whereas possible reasons for the appar- Holocene transgression, continental shelf, northwest Gulf of Mexico, in Shepard, F. ent increases are numerous, more rapid P., F. B. Pitleger, and T. H. van Andel, recession during the past century sug- eds., Recent sediments, northwest Gulf of gests that changes may be man-induced. Mexico: Tulsa, Okla., Am. Assoc. Petro- These artificial modifications include leum Geologists, p. 221 - 266. such activities as (1) construction of jet- Frazier, D. E., 1974, Depositional-episodes: