POLLEN ANALYSIS AND CHRONOLOGY OF A CENTRAL TEXAS PEAT BOG APPROVED; — To wife JoAnn my for her help and encouragement. POLLEN ANALYSIS AND CHRONOLOGY OF A CENTRAL TEXAS PEAT BOG by TOM SHORT PATTY, B.S. THESIS Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of MASTER OF ARTS THE UNIVERSITY OF TEXAS AT AUSTIN June, 1968 ACKNOWLEDGMENTS I am grateful to Dr. Donald A. Larson for direction, suggestions, criticisms, and general help during the course of this study. I also thank Dr. H. C, Bold for his kind assistance during the preparation of this thesis. Special thanks are due to fellow graduate student Vaughn M. Bryant, Jr., for his generous assistance both in the field as well as in the laboratory. Appreciation is given to Dr. E. Mott Davis and Mr. Sam Valastro of The University of Texas Radiocarbon Dating for assistance in collection and dating of the Laboratory samples. I wish to thank Mr. Fred Alex, landowner, for access to the bog site and information about the recent history of Lee Sam the bog. lam grateful to other landowners, Soefje, Soefje, and Mr. Rutledge, all of Ottine, Texas, for access to property and the fine hospitality during the many trips to and the Hershop adjacent bogs. The initial work on this project was conducted during 1966-67 when the writer was supported as a member of the Academic Year Institute National Science Foundation—supported at The University of Texas at Austin. T.S.P April 3, 1968 IV ABSTRACT Results from two cores (5.4 and 5.0 m) taken from Texas the Hershop Bog, a Central peat deposit analyzed for fossil pollen and radiocarbon data, indicate that a sharp decline in the arboreal components of the vegetation oc­ curred at the end of the late-glacial pluvial period. Fossil pollen records of other bog sites and a tri-county (50 mile) modern pollen transect are compared with Hershop Bog pollen profiles established from dual correlatable cores. The pollen record suggest the occurrence of arid conditions be­ tween 10,000 and 7,000 years B.P. followed by a 4,000 year of mesic conditions. This mesic period was span apparent interrupted about 3,000 years ago by a brief slight-arid period and a return to mesic conditions occurring about 1,000 B.P. years V TABLE OF CONTENTS Page INTRODUCTION 1 LOCATION, TOPOGRAPHY, GEOLOGIC SETTING AND VEGETATION 6 MATERIALS AND METHODS Peat Samples 13 Radiocarbon Samples 15 Modern Pollen Surface Transect 15 Extraction and Preparation 16 Pollen Reference Collection 18 Pollen Counts 19 Pollen Diagrams 20 Observations 21 RESULTS Radiocarbon Dates 22 Pollen Analysis 24 Modern Pollen Rain 34 DISCUSSION 38 SUMMARY 48 APPENDIX 50 REFERENCES CITED 54 VI LIST OF FIGURES Figure Page 2 1. Texas Pollen Sites 2, Texas Peat Localities 5 3. Area Map of Ottine, Gonzales County, Texas 7 .. 14 4. Hershop Bog Coring Diagram 5. Map of Modern Pollen Transect 17 6. Hershop Bog Radiocarbon Dates with Corresponding Depths 25 7. Hershop Bog Pollen Profile 28 8. Percentage of Arboreal in Relation to Non- Arboreal Pollen from Core I and Core II 30 9. Modern Pollen Rain 35 VII LIST OF TABLES Table Page 1. Radiocarbon Dates from Hershop Bog 2 3 2. Pollen and Types of the Hershop Bog Surface Transect viii INTRODUCTION The Pleistocene and post-Pleistocene vegetational is understood. history of Texas as yet incompletely Only a of studies quarter a century ago peat bog began yielding the initial knowledge of late Quaternary vegetation in the Central Texas area. These analytical studies of fossil pollen used to reconstruct past environments in Texas were first con­ ducted by Potzger and Tharp (1943, 1947, and 1954) from three peat bog deposits. Soon to follow was a single probe into a forth peat deposit (Graham and Heimsch, 1960). This probe yielded not only a pollen profile but a radiocarbon date. In the western half of the state, playa lakes on the Llano Estacado have provided pollen analytic information which reflects vegetational and climatic changes from the Wisconsin to glacial period the recent (Hafsten, 1961; Oldfield and Schoenwetter, 1964). A Blancan-age lake deposit has provided an early Pleistocene pollen record in the Texas Panhandle (Harbour, in press), while archeological sites the Rio along Grande in the Amistad Reservoir area have recently yielded post-glacial records (Johnson, 1963; Bryant, 1966a, 1966b; McAndrews and Larson, 1966; Hevly, 1956). These sites are still under examination. Locations of the pollen analytical sites mentioned above are indicated in Figure 1. 1 Fig. 1. Texas Pollen Sites. will The present study (1) present a peat-bog pollen stratigraphy and depositional chronology as substantiated by dual cores and extensive radiocarbon analysis; (2) evaluate and compare the interpretations of the late Quaternary cli­ matic and vegetational sequences in Texas presented in the previous studies of peat-bog deposits; and (3) seek correla­ tions among the post-glacial pollen records as reflected in the playa-lake and archeological sites with the Central Texas bog sites. Peat bogs, along with numerous lake, stream- cave, terrace deposits, and the Gulf Coastal Beaumont and Lissie- age sands and gravels represent the Pleistocene and f>ost- Pleistocene series in Texas. Peat deposits consist of accu­ mulated plant fragments in various stages of decay and con­ tain vast sums of pollen preserved under the reduced and acid environment. By careful extraction and identification of the pollen at various levels within a bog, a reconstruction of the past anemophilous vegetational history of an area can be inferred to a reliable degree, and this yields indications of past climatic conditions. Unfortunately, the humus content and water-absorbing qualities of peat make it a useful item in the horticultural business. With the coming of World War 11, the supply of commercial peat from northern Europe was cut off. Not only had the production of peat in the United States been small, to as compared the European sources, but the freight rates by rail from the peat-bog areas of Minnesota, Wisconsin, and New York were high enough to demand the development of peat deposits in Texas to supply the local market (Plummer, 1941), thus destroying a number of important sites. A unit of the WPA State-Wide Mineralogical Survey of Texas, sponsored by The University of Texas Bureau of Eco­ nomic set out in 1940 to discover and to evaluate Geology, economic deposits of peat in Central and East Texas. Several bogs were located and numerous marshes containing peat deposits were also examined (Fig. 2). Deposits along Yegua Creek in Lee County yielded commercial quantities of peat in 1940 (Plummer, 1941). Several peat deposits in Polk and San Jacinto Counties were examined and probed (Shafer, 1941). Many marshes of associated with all of the containing deposits peat are major river systems in East Texas (Fisher, 1959). One of the deposits in Gonzales County, the Hershop Bog (named after the late George Hershop, owner of the estate on which the bog was found in 1940), is the site for the present study. In a recent Bureau of Economic Geology Report of Investigations (Maxwell, 1962) this bog carried the name "Fred Alex Bog" named after the present landowner, Fred Alex, nephew of the late George Hershop. Fig. 2. Texas Peat Localities. LOCATION. TOPOGRAPHY. GEOLOGIC SETTING. AND VEGETATION The Hershop Bog is located on the Fred Alex farm two miles southwest of the village of Ottine, Gonzales County, Texas, as illustrated on the area in Figure 3. Map dis- map tance is ca. 60 miles E.N.E. of San Antonio, 50 miles S.S.E. of Austin, and 10 miles N.W. of Gonzales, The coordinates of the bog are lat 29°35'N.; long 97035'30"W. Its elevation is 340 feet above sea level. The elevation of the San Marcos River two present miles northeast of the bog is 290 feet. The flood plain of the river, used for cultivation and grazing, has a minimum elevation of 320 feet and lies within a mile distance of the Hershop Bog site, while upland areas, one half mile or so to the west of the site, over 400 feet in elevation. are The meander belt of the San Marcos River, trending southeast in the area, intersects the Carrizo (Eocene) forma­ tion, a ferruginous sandstone which strikes NE-SW through Gonzales County. The Carrizo is a good aquifer and supplies the area with artesian and well water. The formation makes contact with flood-plain deposits at an elevation of ca. 320 feet and forms numerous seeps in the area. Near-by bogs, the Rutledge, and both the North and South Soefje Bogs, owe their origin to seepage from the Carrizo sands. 6 Fig. 3. Area Map of Ottine, Gonzales County, Texas. The Palmetto State Park, about two miles northeast of the Hershop Bog site and dissected by the San Marcos River, lies entirely on the flood plain and contains many swamps and boggy areas. Most of the bogs in Gonzales County exist under condi­ tions which duplicate the areas of other bog sites in Texas (Fig. 2). These include rainfall of 35 or more inches per and hillside meander-belt year, gently rolling topography, or locations. As a matter of convenience, peat deposits in Texas this writer into two main flood- are placed by classes; (a) plain-bounded and (b) upland-area. Most of the Texas bogs fall into the first class. They are formed as a result of the slow filling of oxbow lakes or abandoned meanders. Also included in the first class would be the bogs formed by seep­ age at the base of hillsides in contact with, and normally formed on, the flood plain. Other peat deposits result from fed the filling of upland depressions and sinks, usually are from perched water tables, and fall into the second class. Carrizo Water slowly percolating through the permeable sands ultimately reaches impermeable layers. It then moves laterally and, in regions of rolling topography and stream dissection, can reappear at the surface as springs or seeps. The depositional environment (topography, water availability and climate) in which plant fragments accumulate with incom­ plete decay determines the depth to which the peat can accu- This mulate. writer, using steel rod probes, has found peat accumulations in Central Texas ranging from 0.25 meters to meters just over 5 in depth. The is interpretated by this writer to be Hershop Bog It is situated well beyond an upland-area bog. the upper limits of the flood plain (about 20 feet) and its origin is the result of an upland area depression dissecting a perched water table and being filled by plant fragments. Large blocks and boulders of Carrizo, the aquifer supplying the water, can be observed outcropping along the sandy hillsides above the bog. is situ' The Hershop Bog, as can be seen in Figure 3, ated about a mile south of the South Soefje Bog and is topog raphically above it by about 20 feet. This bog is unsilted and to have been free of erosional breaks appears during active deposition. However, the present peat area exposed is much smaller than the original active bog area because sand has washed down from the hillside covering the recently edges of the deposit. With a steel probe, a total bog area this of almost 500 feet in diameter has been examined by writer. Early study of the bog (Chelf, 1941) included analyses of moisture, ash, and pH. A reported pH range from within the range obtained by this writer. 3.6 to 4.7 was One sample from cores taken for this study was observed to have a pH of 2.8. Chelf (1941) also reported that the Her- shop Bog ranged in thickness to about 20 feet, but on a pro­ file of the bog, the vertical scale was such that the thick­ ness would be 50 feet. This error was picked-up and reported again some years later (Maxwell, 1962). Extensive probing demonstrated Bog has a maximum depth of 5.4 that the Hershop meters, or about 17.5 feet. In order to make significant evaluations of a fossil pollen profile the distribution of local extant species should be considered. Studies of Texas plants, fragmentary and basically unsystematic, date back to field observations in the early 1800's. Plant lists of Palmetto State Park, one and a half miles northeast of the Hershop site, were compiled in the early part of this century by Bogusch (1928) and Parks (1935). The vegetation of the general area has been designated by Tharp (1939) and Blair (1950) as an oak-hickory association. Gould (1952) placed the northwest quarter of Gonzales County into savannah but mentions that others to a post-oak prefer class the area as part of the true prairie association of the formation. This latter view is based on the fact grassland is tall native that the understory vegetation typically grasses. Along the hillsides adjacent to the Hershop Bog site, the dominant trees are Quercus stellata Wang, (post oak), Q. marilandica Muench. (black-jack oak), Q. viginiana Mill, (live oak), Q. macrocarpa Michz. (bur oak), Ulmus americana L. (American elm), and Prosopis glandulosa Torr. (mesquite). Juniperus virginiana L. (cedar) can be seen in the vicinity but is a very minor component. Next to the bog and toward the flood plain are Carya illinoensis (Wang.) K. Koch, (pecan), Salix nigra Marsh, (black willow), Populus deltoides Bartr. ex Marsh, (cottonwood), Vaccinium arboreum Marsh, (huckleberry), Closer and Celtis pallida Torr. (hackberry). to the flood plain Fraxinus americana L. (ash), Myrica cerifera (L.) Small. and vomitoria Ait. (wax myrtle) Ilex decidua Walt, 1. (yaupon) , can be found. not now Plant species growing in the vicinity of the Hershop Bog, as well as their present distribution, should be noted in relation to the fossil-pollen findings. Pines (i.e., £. taeda), in recent years, have been planted along Highway 90 ca. 10 miles north of the study Isolated native stands area. of pines, common in the eastern part of the state, occur 30 miles or so farther north in Caldwell and Bastrop Counties. The species represented in the Bastrop State Park area is P_. taeda L. (loblolly). It has been reported by Critchfield and Little (1965) that a minute isolated stand of P_. palustris Mill, (longleaf pine) exists in far eastern Gonzales County some 25 miles from the bog site. Pinus echinata Mill (short leaf pine) grows in East Texas in association with the loblolly and longleaf pine about 125 miles northeast of the study area. Liquidambar styraciflua L. (sweet gum) grows in East Texas r- some 150 miles away. Alnus rugosa (Dußoi) Spreng. (alder) is also found in East Texas along the Trinity River, and Betula nigra L. (river birch) is native to East Texas, According to the landowner, a long-time resident, the Hershop Bog was active with cattails and other bog plants covering the surface up until the late 1940'5. Where water drains from the bog, ferns, mosses, and cattails (Typha sp.) are still present. At present the bog is "dead" and, with no protective covering of plants, is undergoing surface erosion and silting. The bog is slightly domed in the middle and "quakes" when jumped upon. MATERIALS AND METHODS Peat Samples After to determine its basal probing the Hershop Bog to two conformation and identify its greatest depth, complete peat cores were taken for pollen analytic purposes. These cores were obtained with a Hiller borer. Actually, Core I was collected on August 8, 1967 while Core 11, for strati­ graphic correlation, was obtained on February 1, 1968. Peat samples from each retrieved core were removed in 10-cm seg­ ments and placed in labeled glass jars and sealed. A record was made of color, compactness, odor, degree of decomposition, and sand content of each sample. Permineralization and total replacement of plant stems by pyrite were observed in the lowest bog samples. The deposit was cored to a depth of 5.5 meters before resistance was met. When the bottom cylinder of peat was brought up and opened, the lowermost 10-centimeters was composed of clean, fine-grain, well-sorted Carrizo sand. Fifty-four samples of 10-cm increments were taken for a total recovered core of 5.4 m and designated as Core I. The second core was taken 10 m west of Core I and had a total depth of 5.0 m (See coring diagram Figure 4). 13 Fig. 4. Hershop Bog Coring Diagram. Diagramatic Radiocarbon Samples Peat from the Hershop Bog to be analyzed for Carbon­ -14 was obtained and submitted to The University of Texas Radiocarbon Dating Laboratory in November, 1967. The samples were obtained by means of a Hiller borer in 20-cm increments. Three arbitrary levels were used: 0.3-0.5 m (upper); 2.3-2.5 m (middle); 4.9-5.1 m (lower). Five probe areas designated as A, B, C, D, and E were selected. The first four mark the corners of a square with diagonals 10 m long and probe area "E" in the center as illustrated in the plan view in Figure 4. The five probe areas were centered over the deepest part of the bog "Elocated 50 from Core I used with probe M cm in pollen analysis. A total of 45 penetrations were made in the process of acquiring peat samples for radiocarbon processing. This means that at each probe position three samples were obtained from each of the depth levels. The three samples from each depth level at a probe point were placed together in a plastic bag and sealed. This composite sample was then used in dat­ ing. Washing of the Hiller borer with distilled water after each sample protected against contamination, and peat was re­ moved by a carefully cleaned spatula also insuring against contamination. Modern Pollen Surface Transect Eight surface-soil samples, taken from a 50-mile tri­ county transect, were collected and analyzed during the course of this study. Six samples were taken along a four-mile transect in a general NE-SW direction in the vicinity of the Hershop Bog. Of the other two, one was collected ca. 100 m from a small trend of native pines in northern Caldwell County, and one was taken within Bastrop State Park in Bastrop County. The collecting site in Caldwell County was an oak-mesquite site within woodland, whereas, the Bastrop was an oak-pine woodland. The transect within Gonzales County spans the en­ tire flood plain of the San Marcos River and adjacent upland areas. The sampling sites are designated as MPI through MPB as illustrated on the area as Figure 5. The collections map were made during February, 1968, by the writer, and consisted of several handfulls of surface dirt sampled over an area of 2 ca. 100 m The samples were placed in labeled plastic bags . and sealed. To avoid any gross overrepresentation of a species which might in one occur any spot, the samples in the bags were shaken and The sites were thoroughly mixed. sampling spaced to provide information about the modern pollen rain over a diverse topographic and vegetational range. Extraction and Preparation The University of Texas Brackenridge Field Laboratory houses the facilities used in the processing of the Hershop Bog samples. The fully air-conditioned, positive-pressure Palynology Research room has equipment for all types of pollen extraction techniques. Fig. 5. Map of Modern Pollen Transect. A small amount (2-3 cc portion) of peat from each of the 10-cm increments was processed by a modification of the acetolysis treatment described by Faegri and Iverson (1964). A detailed description of the seven basic steps used in ex- from is in the tracting fossil pollen the peat samples given Appendix. Additional preparation of the samples consisted of mounting the 2.0% Safranin-stained palynomorphic material. A drop of the concentrate suspended in silicon oil (2000 cs) was placed on a labeled microscope slide, covered with a coverslip, and sealed with fingernail polish (Anderson, 1960) All unused portions of Core I and 11, duplicate-processed concentrate, and prepared slides are catalogued and stored as reference collection. These part of the fossil pollen samples are accessible to worker desiring refinement or any verification with future work. any Pollen Reference Collection Reference material used in the identification of the Hershop Bog pollen types consisted of the Amistad Pollen Collection (Bryant, 1966c) and collection of pollen a ex­ tracted by this writer from specimens on herbarium sheets in The University of Texas Herbarium. The selected pollen types for the reference collection are from species known to exist in the Hershop Bog area today and from plants such as Betula and Liquidambar which do not now grow in the vicinity of the bog but appear in the fossil pollen record. The Amistad Pollen Reference Collection has a pre­ pared pollen key (McAndrews, 1965) for 250 pollen types. this Vaughn Bryant, writer, and others have tripled the size of the reference collection which now includes plant types from all over Texas and adjacent areas. This reference col­ lection, including the Amistad collection, is currently housed in the Palynology Research room of The University of Texas Brackenridge Field Laboratory. Identification of pollen types in this report are based upon comparative studies with the mentioned reference material. Pollen Counts counts made with an oil-immersion All pollen were objective at 970X, The 54 samples in Core I were examined for all pollen and spores (excluding fungal spores) using a standard 200-grain count (Barkley, 1934), this provided a primary count, A count which excluded all trilete secondary grains was obtained from Core I. Grain counts for the 5-M correlation core (Core II) were obtained using a fixed sum of 150 grains and excluded all spores and trilete grains. As each grain was identified, it was scored on a tabulation sheet and recorded on a mechanical hand counter. Whole grains and fragments (such as pine bladders) of known pollen types were included in the standard counts. Although the fungal spores were excluded they were not ignored. Highly distorted, crushed, and deteriorated grains which could not be identified were also excluded from the counts. Pollen counts using fixed sums of 200 were obtained from each of the surface soil samples. Pollen Diagrams The primary pollen profile, a standard 200-grain count, included all recognizable pollen types and trilete summarizes grains. The secondary pollen profile the data from a second count totalling 200 grains but excluding trilete types. (Fig. 7). Data for the secondary pollen profile was obtained by using the of the non-trilete grain sum types tabulated from duplicate slides for corresponding levels. For example, from processed peat at a depth of 310 cm (3.1 m), a total of 48 trilete grains (24%) was tabulated in the primary count. From a duplicate slide a second count was made of 48 non-trilete grains and added to the 152 non-trilete grains from the primary count, thus giving a total of 200 grains. Unless otherwise stated the discussion in this re­ port refers to the results of the secondary count. A pollen profile illustrating the arboreal pollen in relation to non-arboreal pollen was constructed by totalling all of the tree-pollen percentages and plotting them against the remaining elements of the pollen counts (Fig. 8). The non-arboreal component of the diagram includes the unknown pollen types which averaged about two percent. Superimposed on the profile of Core I is an arboreal-non-arboreal profile from Core 11. This provides graphic illustration of the relationships between pollen percentages in the two cores. Observations Observations for this report were made with an A. 0. Spencer phase-contrast binocular microscope with objective lenses of 10X, 43X, and 97X, The oculars were 10X. All counts pollen were obtained under oil at 970 X followed by- scanning under 100 X and 430 X for unusual pollen types not observed in the standard sums. RESULTS Radiocarbon Dates Radiocarbon dating of the Hershop Bog has yielded a maximum date of 10,920+160 years B.P. from peat approximately 0.5 m above the sands at the deepest known point (5.4 m) , The set of dates as released by Dr, E, Mott Davis, complete listed in director of the Radiocarbon Dating Laboratory are Table 1. When examining these dates, the reader should under­ stand that an average obtained from the the date represents analysis of a bag of peat collected over a 20-cm stratigraphic For level of Radiocarbon Coare "A" range. example, the upper an age of years B.P, is obtained from a unit of peat ranging between 0.3 m and 0.5 m depth (Fig. 4). In the upper and middle levels of the Radiocarbon Cores, the depth of the 20-cm increments was consistent, but in the lower level, because of variation in bottom depth and sand content of the cores, all increments were not taken from the same level. The exact depths are indicated under the dates listed in Table 1. As shown in the table, an approximate date is also indicated for each of the three levels. The date represents an average and is not intended for the any specific depth but applies to depth range from which the peat was sampled. Henceforth, the 22 BOG HERSHOP FROM TABLE DATES RADIOCARBON Average E D C Core B A Level 2,022 6,006 2,170l90 6,0501100 1,520+80 6,0001130 2,120l90 5,8501120 1,960ll00 5,980ll00 5m) .5m) 2,340180* (0.3-0. 6,1501130 (2.3-2 Upper Middle 10,574 1m) 10,4501160 (4.9-5. Ira) 10,5601160 4.9-5. 0m) 10,4901160 (4.8-5. 6m) 10,450+160 (4.4-4. 0m) 10,920+160 (4.8-5. Lower • P.) (B. present before are ;s date *A11 average dates and the depth ranges are included in the pollen diagram for general purposes but definitive dates must be read from Table 1. Figure 6 illustrates the relationship of the Hershop Bog radiocarbon dates to the corresponding depths. Straight lines connect the points of average depth and average age for each sample taken. Dates used in this report for depths not from curve and are sampled are extrapolated the three-point intended to be only approximations. Pollen Analysis The Hershop Bog cores and the modern pollen transect yielded 33 pollen types representing 25 families (Table 2). Four of these types (Acer, Alnus, Betula, and Liquidambar) Acacia were found only in the fossil record, whereas, was found only in the modern pollen rain. Prosopis, a dominant of tree in the immediate area the bog, was encountered only in the upper 10-cm increment. Prosopis composed three per­ cent of the modern pollen rain collected near the bog (MPI). The results of fossil pollen processed from Core I are repre­ sented by Fig. 7 for the primary and secondary count. The primary fossil pollen profile represents per­ centages of pollen and trilete moss and fern spores based on the standard 200-grain count. A general survey of fungal spores was made not to make some evaluation as to their significance but to see how they rank in number compared to Fig. 6. Hershop Bog Radiocarbon Dates with Corresponding Depths. TABLE 2 POLLEN TYPES OF THE HERSHOP BOG AND SURFACE TRANSECT ACERACEAE Acer AQUIFOLIACEAE Ilex CACTACEAE - CHENOPODIACEAEAMARANTHACEAE COMPOSITAE High-spine types Ambrosiaceae Ligulifloreae Artemisia CORYLACEAE (BETULACEAE) Alnus Betula CYPERACEAE ERICACEAE FAGACEAE Quercus GRAMINEAE HAMAMELIDACEAE Liguidambar JUGLANDACEAE Carya Juglans TABLE 2 (Continued) LIGUMINOSAE Prosopis Acacia LYCOPODIACEAE Lycopodium MYRICACEAE Myrica ONAGRACEAE PALMACEAE Sabal PINACEAE Juniperus Pinus POLYGONACEAE SALICACEAE Salix SELAGINELLACEAE Selaqinella TYPHACEAE Typha anqustifolia T. latifolia ULMACEAE Celtis Ulmus UMBELLIFERAE PROFILE POLLEN Figure BOG HERSHOP the pollen. In the lower meter of the Hershop Core I samples, counts were made of 200 grains which included fungal spores. to The percentages of fungal spores ranged from 11% 30% with an average greater than 20%. Pollen from almost any level of the bog, particularly the grass, often appeared to be in various stages of deterioration perhaps due to some sort of fungal activity. This writer observed several surface-soil samples collected for the transect in which 90% of the pollen showed signs of fungal attact; pollen was absent altogether in certain samples. Graham (1962) discusses the possibility of the of this use fungal spores in palynology but problem is outside the scope of this counts obtained The for the paper. profiles exclude all fungal spores. If the gross trends in the pollen assemblages repre­ sented on the pollen diagrams are examined, it can be seen that the non-arboreal pollen comprises the bulk of the profile for the base and upper segment of the bog. Dividing the except pollen types into two main groups (the arboreal and non-arboreal) and then plotting the percentages, the resulting profile nicely summarizes the trends (Fig. 8). The relationship between arboreal and non-arboreal types can be compared for any depth. For example, it can be shown that in the lower 0.5 m (5.4-4.9 m) of the Hershop sediments, arboreal pollen contributed 40% or better of the total preserved pollen. Birch Pollen Zone. The lower peak in the tree pollen is reached at of 5.0 with maximum a depth m a of 44% (as Fig. 8. Percentage of Arboreal Pollen in Relation to Non-Arboreal Pollen of Core I and Core II from Hershop Bog. which make illustrated in Fig. 7). The major tree types up this bulk are Quercus, Betula, Ulmus, and Pinus, the latter first three add and adding only 3%, whereas the 17%, 14%, 4%, respectively. This stratigraphic zone is designated as the Birch Pollen Zone in accordance to Cushing's (1964) Code of Stratigraphic Nomenclature. The age of the lower tree peak is ca. 11,000 years B. P. as indicated by the results of the radiocarbon analysis. An approximate date of 12,000 years is assumed for the beginning of the bog deposition. The radio­ carbon analysis provides a minimum age for the end of the Birch Pollen Zone of ca. 10,000 years B, P. This lower zone has Salix and possibly Liguidambar present. Salix is not restricted to this zone, whereas, a type identified as Liguidambar was observed only in the 5.2-m sample. Juglans ranged throughout this lower zone in percentages of 2% or less. Betula was re­ covered from level between the bottom and the 4.7-m level. every Both Betula and Juglans dropped from the fossil record above this level, except that the latter was again recovered from the upper 10-cm segment at about 2%. Carya and Celtis also have this basic pattern, present in the lower levels, absent for the most part in the middle portion of the strata, and reoccurring in the upper 10-cm segment. The Ulmus peak is in the lower but observed at other levels in the portion also frequently section. The non-arboreal components of the Birch Pollen Zone can be characterized by a a low percentage of Gramineae, near absence of Umbelliferae, and, as seen in the primary pollen at profile, the trilete (moss and fern) types are their mini- is in this mum. Ilex only represented lower zone in the fossil identified in record along with Onagraceae. Ilex was the modern pollen rain from samples taken along the flood plain boundary (MPS). Cyperaceae reflects a relatively high (averages somewhat above 20%) in this lower pollen percentage zone. Maximum Grass Pollen Zone. With the sharp decline in arboreal increase in pollen types, there is a corresponding non-arboreal types. The grass curve begins increasing at 4.9 m mark at passing the 40% at 4.3 m. The grass pollen peaks a maximum 53% at the 3.8 m level. This zone of sharp rise in grass pollen, above levels in which Betula pollen is found and up to the 2.8 m level, is designated as the Maximum Grass Pollen Zone. During the time represented by this portion of the bog, arboreal pollen was at its lowest Re- percentage. covered pine pollen ranged at only 1 or 2%, Quercus from 5 to 10% and Alnus, Salix, and Fraxinus each contributed 1% or less to the pollen counts, Cyperaceae pollen was also reduced in percentage, whereas, Umbelliferae pollen was beginning its rise. At the end of the Maximum Grass Pollen Zone, the grass pollen fell to 20%. The approximate age of 7,000 years B. P. is assumed for this time, A sudden rise in Typha, Polygonaceae and Cyperaceae with an ever-increasing rise in Umbelliferae marks the end of this period. Umbelliferae Pollen Zone. This zone is characterized rise by the constant increasing in umbelliferous pollen types Another feature is a gentle rise in oak (4% to about 18%). The the decline from Cyperaceae and, generally, grass pollen bottom to top in this zone. The Compositae range about 30% below. compared to an average of 20-25% in the zone Typha is in this to limit of this present zone up 4%. The upper zone corresponds to the sudden decline in the Umbelliferae recovery (from 25% to less than 10%) and the increase in oak and grass pollen. Oak-Mixed Grass Pollen Zone. At the 0.8 m level, there is a pronounced peak in the oak curve. The same level is also marked by an increase in Gramineae and Cyperaceae. For the next 30 cm the latter, as well as Umbelliferae, de­ creases and the oak curve becomes stabilized. The upper por­ tion of this zone is characterized by noticeable decline in a Compositae and Gramineae with an ever increasing Cyperaceae and oak curve. Juglans, Alnus, Carya, and Celtis contribute to this zone especially in the upper 10-cm portion. Prosopis also is observed in this segment. in small Although percentage (2%), Ulmus contributed to the arboreal pollen for this upper pollen zone. An assumed age of just less than 1,000 years B. P. is placed on the reoccurrence of the riparian elements observed in the core segment (an unknown amount of peat upper has been eroded from the present bog surface). Modern Pollen Rain The transect of surface-soil three samples across counties provides information about changes in local modern pollen rains with respect to topography and plant distribu­ tion. The transect covers a distance of almost 50 miles map and NE-SW illustrated in Figure 5. trends roughly as Eight One samples were collected and processed. in Bastrop State Park (MPB) one in mid-eastern Caldwell County Bastrop County, (MP7) and the remaining six in Gonzales County. The elevation ranged from ca. 460 feet in Bastrop County ot ca. 218 feet on the flood plain of the San Marcos River in Gonzales County. A diagram showing the modern-pollen rain obtained It from the surface samples is represented as Figure 9. is possible to separate several geographic and topographic zones from each other by using the modern-pollen rain. High pine recovery is in the pine forest in Bastrop County, but long- distance transport results in pine pollen being found also in surface samples from Caldwell and Gonzales Counties. The percentage of pine in the surface soils within the five-mile radius of the bog site matches the percentage recovered in levels the upper of the peat cores (ca. 2%) Samples taken . from the oak-mesquite upland area (MP6) north of Ottine shows Quercus and Prosopis as dominant tree-pollen types with Acacia and associated Ulmus as types. In the same location, Compositae and Gramineae dominate the herb-pollen segment of the pollen sum. Rain Pollen Modern 9. Fig. From the upland collecting site to the flood plain from 425 boundary (MPS), the elevation drops ca. feet to ca. 340 feet. There is a marked change in the vegetation from the two sites. Seepage zones are common along the flood plain boundary and the vegetation is dense. The modern-pollen rain reflects this change in the percentage of upland-area types and exchanges with several others. Ilex and Malvaceae show up well certain riparian types such as Celtis, Salix, and as as Fraxinus. Ulmus and Carya take over as dominant tree types, along with Quercus, on the rich soil flood plain. Prosopis pollen falls off quickly from the upland area to the flood plain. The pollen rain offers a challenge for the expert in oak The area oaks are Q, stellata and Q. pollen. upland marilandica while the flood plain border oak is Q. virginiana. ratios would make Analyzing the surface pollen for percentage a good problem. Riparian tree pollen, i.e., Salix, Fraxinus, Juglans, and Ulmus, makes the major components of the modern up pollen rain. Juglans and Fraxinus are limited only to the flood plain, according to the pollen rain. The non-arboreal pollen types characteristic have some profile features. The cheno-ams range across the transect area for the Palmetto site (MP3). except swamp collecting Cyperaceae dominate the swamp (MP3) and march (MP4) sites on the flood plain. The grass profile indicates that the high peaks in the pollen rain are in the upland areas with per­ centages above 30%. The grass curve declines across the swamps to a recoverable increase near the Hershop Bog (MPI). the Compositae pollen in the pollen rain, almost parallels grass curve except that the percentages at the peaks in the profile are much higher, especially at the Hershop site (ca. 49%). DISCUSSION Based upon excellent radiocarbon dating and the unsilted condition of the peat, it to be a safe appears assumption that in the Hershop Bog during the past 12,000 a pollen record has accumulated without disruption years the intrusion of stream Thus, or deposited palynomorphs. the fossil record is representative of the bog, local up­ land, and regional pollen rains. This pollen record begins prior to the termination of the Wisconsin glacial period and covers the recent. However, there has been an inexactly known amount of erosion of the most recently deposited peat during the past decade or so. Strong evidence for post-glacial climatic changes is is not in Central Texas present for the pollen record uniform, and a rapid decline in arboreal pollen, especially birch, approximately 10,000 years B.P. appears to mark the end of a pluvial period in Central Texas. The impressive increase in grass pollen, which followed the drop in arboreal elements, suggest a deterioration of upland and stream-side forest caused by a reduction in regional mois­ ture availability. The occurrence of an arid period be­ tween 10,000 and 7,000 years B.P. is also indicated by the reduction in Cyperaceae pollen and the many riparian types 38 i.e., Juglans, Celtis, and Carya At about record of 7,000 years 8.P., the pollen the start of Hershop Bog suggest a gradual change to more mesic conditions as indicated by an increase in Polygonaceae and Umbelliferae pollen and a sudden rise in Cyperaceae and Typha pollen. Oak pollen percentages increase gradually. The apparent mesic conditions were interrupted at approximately 3,000 years B.P. by a brief slight-arid period which lasted some 2,000 years and was characterized by a sudden increase in grass pollen, a sharp decline in Cyperaceae and Umbelliferae pollen, coupled with stabilizing of the rise in oak pollen. By approximately 1,000 years B.P. the postu­ lated slight-arid period terminated and a return to mesic conditions occurred. This last change as seen in the pollen profile was sufficient to bring a return of riparian types i.e., Juglans, Ulmus, Celtis, and Carya; along with aquatic elements i.e,, Cyperaceae and Typha. The pollen record also suggest an invasion of Prosopis and Juniperus occurring dur­ ing the last 1,000 years. In order to critically judge the significance of the various components of the pollen profile, a modern pollen rain transect was made. This has proven to be extremely valuable. By comparing vegetational components in the fossil record to the modern pollen rain, certain general assumptions about the paleovegetation of the area can now be tentatively made. However, it must be recognized that changes in the physical nature of a bog may very well result in changes in the vegetation of the bog itself, thus influencing the pollen profile without significant regional environmental With reference to the latter statement, a question changes. to climate of time arises as the general during the period from about 7,000 to about 3,000 8.P.? The years B.P. years period before and immediately following this 4,000 year Was span of time had prominent grass peaks. the general this mesic condi­ region during period being subject to more tions with increased rainfall, or could changes in the physi­ cal nature of the bog, such as darning of its drainage outlet cause ponding on the bog's surface? Surface ponding could have resulted in certain bog plants supplying increased per- rain without in centages in the pollen any necessary change the regional climate, and increased percentages of bog pollen types would automatically reduce the representation of regional pollen. is in The modern pollen rain transect quite helpful the interpretation of pine pollen percentages in the profile. As shown in the Hershop Pollen Profile pine averages close to 4% in the lower meter of the bog deposit. Although sparce in the middle levelspine persist throughout the sequence but never averages more than 2 or 3% above the 4.0-n level. The modern-pollen rain along the transect demonstrates that current long-distance transport results in 2-3% of pine pollen in the surface-soil samples near the bog. These data suggest that for the past 12,000 years stands of pines have been no closer to the Hershop area than they are at time, and that the "isolated" pine stands in the present Central Texas may have been stable for this period as well. The communities around marshes, vegetational seeps, and bogs and surface samples in the vicinity of the Hershop site provide insights as to what the local paleo-micro­ environment might have been. The dominant understory shrubs growing on and around the South Soefje Bog, one mile north of the Hershop Bog, are Ilex and Myrica. Very active seeps provide abundant water for the "alive" and "quaking" portions of that bog today. Ilex was observed in the modern-pollen rain in the surface collecting area surround­ ing seeps and springs. The fossil-pollen record indicates that Ilex grew in or around the Hershop Bog between approxi­ mately 10 and 12,000 years B.P. The disappearance of Ilex pollen from the profile may have resulted from meander migra­ tion of the San Marcos River rather than the regional changes which had more influence on upland plants. Additional support for a mesic environment during the deposition of the lower meter of the Hershop deposit has been obtained. Analyzed samples from a 0.5-m core taken from the - bottom 5.7 of a discovered (5.2 n) recently peat deposit designated by this writer as East Hershop Bog has revealed a pollen sequence which parallels the Hershop profile. All of observed the arboreal components in the Hershop samples, except Licruidambar, were encountered in the East Hershop samples. The birch pollen peak occurred at the 5.7-m level and gradually decreased upwards. Most of the other pollen types were observed at all levels but the birch pollen was absent above the 5.0-m point. Radiocarbon dating of the East Bog was not obtained and stratigraphic correla- Hershop tions will not be attempted. and Betula pollen from both bogs was carefully suc­ cessfully compared to Betula nigra pollen processed from herbarium specimens. Collection data on the herbarium sheets identified birch in East Texas as close as 250 miles to the Hershop site. The average annual precipitation in the East Texas counties where Betula nigra was collected ranges from 40 to 50 inches per year, compared to the 30-35 inches in Gonzales County, Since the climatic per year differences between the current habitat of Betula nigra and does not include real variation the bog area any temperature the author has considered it prudent to refrain from inter­ preting the time from 12,000 B.P. to 10,000 B.P. as cool- moist. In fact, none of the pollen types encountered in in the bog require any real reduction in temperature to explain their occurrence. As discussed, the Hershop profile is not uniform and definitely reflects changes in the vegetation since late- times. pluvial Near-by Soefje Bog has been reported (Graham 43 and Heinsch, 1960) to reflect a fairly stable vegetational Their the past 8,000 years. interpretation sequence during has been based upon an unchanging non-arboreal vs. arboreal ratio. Before attempting any correlation of the Hershop and Soefje the natures of the two bogs and local plant communities should be taken into account. Thus, differences in the two profiles may be attributed to depositional events and to differences in local plant communities controlled by topography and water availability. The "flood-plain-bounded" Soefje Bog may have stream deposited pollen types laid down during floods. Also, the present Soefje Bog has a thick understory of Ilex vemetoria and Myrica cerifera with species of Quercus, Fraxinus, Ulmus, and Salix composing the surround ing tree types. The local plant community around Soefje Bog may always have been different than that of the Hershop bog. Around the Hershop site and distant hillsides are Prosopis and Quercus; both species have lower water requirements. Carya and Salix do follow the watershed leading from the Hershop Bog. Graham's 4.7-m core from the bottom of the Soefje site was either too shallow or the sediments too young to encounter the arboreal peak seen in the bottom of the Her- shop deposit. The greater age of the Hershop deposit can account for perhaps 4,000 years which is sufficient time for some major climatic changes in the Central Texas area not reflected in the Soefje deposit. The Soefje profile indi­ cated that grass pollen fluctuated only slightly, whereas, the grass curve in the Hershop profile shows major changes Due its in pollen percentages. to flood-plain location, perhaps the Soefje site had abundant supplies of water even when the general area was subjected to less mesic conditions which are reflected in the Hershop profile in the sequence designated as the Maximum Grass Pollen Zone. Some 70 miles to the north, in Lee the County, Patschke Bog (Potzger and Tharp, 1943, and 1947) was cored deeper than the Hershop Bog by 1.5 m. The former probably, but not necessarily, contains a pollen sequence older than the latter. The Patschke Bog has since been destroyed which adds to the following problems in correlation; (a) the Patschke Bog was sampled at one-foot intervals compared to 10-cm intervals sent in the Hershop; (b) the samples were and to Indiana for processing analysis which brings up the question of Abies and other pollen types being contaminants; and (c) from the 22-ft level a total of 10 slides yielded only 50 grains, which seems to be a highly unreliable count. a four-fold climatic However, sequence was proposed by Potzger and Tharp. This sequence is as follows; cool (boreal conifers), to warm-dry (Quercus and grasses), to warm-moist (Alnus and Castanea, and then to warm-dry (Quercus, Carya, and grasses), becoming still drier towards the topmost level of the bog. This climatic sequence is in doubt, and East Texas are in order bogs being sought to study vegeta­ in this area of Texas tional changes The Gause Bog (Potzger and Tharp, 1954) about 35 miles northeast of the Patschke Bog, in eastern Milam County, was cored to 14 ft and also sampled at 1-ft intervals. It was reported that this bog had Abies and Picea in the lower 1-ft sample indicating a boreal forest. Four years later Graham obtained unused portions of the lower 5 ft of the Gause Bog samples and processed them to verify the reported of Abies and Picea. He of presence reports the presence Picea but could not identify any Abies in the material. The re-examined Gause samples also contained no Castanea as previously reported and Graham found differences in the amounts of oak, and Alnus. grass The Franklin Bog (Potzger and Tharp, 1954) in Robert­ son County, about 25 miles northeast of the Gause Bog, was cored to a depth of 10 ft. Abies was not reported but Picea was. The origin of Picea encountered in the fossil record is currently unknown, but could have been bog spruce from either eastern or western forests. Furthermore, the author is reticent to of Picea completely accept the reports pollen as recent studies have not confirmed its presence in peat this supposedly containing pollen type. A mesic and perhaps cooler environment is suggested for the lower strata in several rockshelter sites from the Amistad Reservoir Area of West Texas (Bryant, 1965a; McAndrews and Larson, 1955). The assumed age for these strata is than 10,000 B.P, in pine and the greater years High peaks of strata. possible presence spruce were observed in these From about 10,000 to 5,000 years B.P. the climate in the Amistad region became drier yet the vegetation as reflected by the pollen record could be characterized as remaining mesophytic. This period was followed by a general change to aridity with minor fluctuations. It is apparent that both West and Central Texas pollen profiles indicate a post- glacial dry period of considerable length. The pollen profiles from playa lakes the Llano on Estacado have been summarized by Oldfield and Schoenwetter Lubbock (1954). The Subpluvial between 11,000 and 10,000 years B.P. has been reconstructed as a pine woodland which stretched across the Texas High Plains. The period corre­ to Flint's (1963) Valders Readvance in Central North sponds America. This period was followed by the San Jon Subpluvial and Yellow House Interval which is matched in the Amistad area and to some degree the Hershop area. The late-Quaternary climate in southern Arizona as outlined by Martin (1963) begins with cool-humid conditions during the late-pluvial period to about 11,000 years B.P. And, from about 11,000 to 8,000 years B.P. a period roughly corresponding to Antev's (1955) Anathermal period, the climate was warm-arid. This proposed period corresponds closely to the assumed arid Maximum Grass Pollen Zone in the Hershop Martin sequence. argues that Antev's Altithermal (from 7,500 47 to 4,000 years 8.P.) was not dry but was characterized by intensive summer rainfall. According to Martin, the pollen and records fail to biogeographic support any biologically important drought occurring in the Southwest between 7,500 and 4,000 years B.P. While it is as yet impossible to correlate pollen profiles from Central Texas to those from Arizona, it can be stated that the climate in Central Texas during the Altithermal was not excessively mesic and may have been only less arid than the preceding period (Maximum Grass Pollen Zone). Deevey and Flint (1957) coined the term Hypsithermal to include the warmer part of the post-glacial period between 9,500 to 2,500 years B.P. As discussed earlier, in Central Texas the role that temperature played in the development of plant communities may have been minimal in comparison to available moisture. SUMMARY Fossil pollen from cores (5.4 and 5.0 m) from the located in Gonzales was stud- Hershop Bog, County, Texas, ied to obtain information about the late-Quaternary vege­ tational history and climate of Central Texas. Samples were processed by a technique suggested by Faegri and Iver- and son (1964) counts using fixed sums of 200 grains were made at 10-cm intervals. A tri-county (50 mile) modern pollen transect was employed to obtain a modern pollen rain which was compared to the fossil pollen record. Ex­ tensive radiocarbon data was collected from a series of probes allowing a detailed chronology of vegetational trends to be given. Four pollen zones have been established to delineate the vegetational changes. From bottom to top the four zones are: Birch Pollen Zone; Maximum Grass Pollen Zone; Umbelliferae Pollen Zone; and the Oak-Mixed Grass Pollen Zone. The vegetational sequence indicates that a mesic existed from period the beginning of the bog deposi­ tion to about 10,000 B.P, A trend toward arid condi­ years tions followed indicated by an increase in as grass pollen and decrease in tree pollen. This period existed until about 7,000 years B.P. possibly changing again to slightly more mesic as reflected by the drop in grass and increase in Umbelliferae pollen and aquatic herbs. This period was 48 interrupted about 3,000 years B.P. by a slight-arid period which was terminated by a return to mesic some .1,000 years B.P. The pollen profile of this period is characterized by a high oak peak and the presence of riparian tree types. Problems in evaluating previous pollen analytic studies are such that no real attempt to make correlations among Hershop, Soefje, Patschke, Cause, and Franklin bogs has been made. There is a need for renewed palynological efforts in East Texas. Considering the distance, diversity in present climate, biomes, and topography it is difficult to make meaningful correlations with pollen sites across the Edwards Plateau in West Texas at the present time. APPENDIX Extraction Techniques for Peat Samples the the Among outstanding advantages in having peat as vehicle for fossil pollen are: (1) due to its very origin, peat has enormous amounts of preserved pollen; (2) the fact that most peat deposits are formed in a depositional basin and materials washed in or blown in usuaully remain to be pre­ served; (3) due to the chemical changes of the organic debris in aquatic environments, peat bogs are highly acidic and act like because of the abundant a preservative for pollen; (4) supply of pollen per unit of peat, only a very small quantity is needed for processing. Seven basic steps were followed in extracting pollen form the Hershop Bog samples. Step 1 Physical separation of coarse elastics and the plant debris. Place a small amount (2-3 cc) of the peat sample in 250 ml beaker. Add 15-20 ml of either or 10% HCI to break up the material. Use stiring rod or swirl the suspen­ sion in order to separate the heavy clastic sediments from the organic material. Let the sample sit 10-15 seconds, then decant liquid and suspended material. Discard clastic sedi­ 50 if Screen ments and repeat necessary. sample using 200 mesh screen to separate plant fragments, jet water onto screen to remove pollen from plant debris. Centrifuge and discard liquid fraction. Transfer sediment to 100 ml plastic centri­ fuge tube. Step 2 Hydrofluoric Acid Treatment Add 10-15 ml of concentrated HF to sediment in plastic centrifuge tube and stir with plastic stirring rod (wear rubber gloves and work under fume hood). Place on hot water bath for 10-15 minutes and stir frequently. Remove from water bath and centrifuge. Discard liquid fraction. Step 3 Hydrochloric Acid Treatment Add 15-20 ml of concentrated HCI and stir. Then add enough 10% HCI to fill tube 3/4 full and stir well. Centi­ fuge and discard liquid fraction. Check for colloids and repeat if necessary. Wash sample several (3-5) times with water. Step 4 Potassium Hydroxide Treatment Add 20-30 ml of 10% KDH to sample and place on water bath for 10 minutes. Remove and decant sample, centrifuge, liquid fraction. Wash sample with distilled water 2-3 times, followed by centrifuging and decanting each time. Flush sample onto a 149 micron mesh screen covering a 250 ml beaker. Screen sample using a jet of water and discard plant debris from screen. Pour sample back into a 100 ml tube, centrifuge, once and decant liquid fraction. Acidify sample by washing with 10% HCI and transfer sample to a 12 ml centrifuge tube. Centrifuge and decant liquid fraction. Step 5 Zinc Chloride, Density Separation Add 10-11 ml of Zinc Chloride solution (having specific and stir well with rod. gravity ca. 1.65) to sample Centrifuge at 2500 rpm for 3 minutes. Decant liquid fraction into 100 ml beaker. Check residue for pollen before discarding. Dilute centri­ liquid fraction sample with double amount of 10% HCI, fuge well and discard liquid fraction. Wash sample with Glacial Acetic Acid, centrifuge and decant liquid fraction. Step 6 Acetolysis Treatment Add 10 ml of Acetolysis mixture (9 parts acetic anhydride and 1 part concentrated sulfuric acid added slowly together) to the sample in a 12 ml centrifuge tube. Stir well and place on water bath for 10-15 minutes. Centrifuge the sample and decant liquid fraction. Fill tube with Glacial and Acetic acid, stir, centrifuge, decant liquid fraction. Wash sample 3-4 times with distilled water. Step 7 Dehydration, Staining, and Preservation Add 10 ml of 95% alcohol, stir, centrifuge, and decant liquid. Repeat 2-3 more times. Then use 100% alcohol for and complete dehydration of sample, centrifuge, decant liquid. Add 3-3 ml of Benzene to sample and transfer sample to a 5 ml shell vial. Centrifuge, and decant Benzene. To the concentrate, add enough silicone oil (2,000 cs) to suspend material to de­ sired amount. Allow sample to stand on warming plate to evaporate Benzene. Cork the vial to prevent contamination. REFERENCES CITED Anderson, Svend Th. 1960 Silicone Oil as a Mounting Medium for Pollen Grains. Geological Survey of Denmark, Iv Series. Vol. 4, No. 1, 1-24. Antevs, Ernst. 1955 Geologic-Climatic Dating in the West. American Antiquity 20: 317-355. Barkley, Fred A. 1934 The Statistical Theory of Pollen Analysis. Ecology, Vol, 15: 283-290. Blair, W. F. 1950 The Biotic Provinces of Texas. Texas Journal of Science, Vol. 2, No. 1, 93-117. Bogusch, E. R. 1928 Composition and Seasonal Aspects of the Gonzales County Marsh Associes. Unpublished M.S. Thesis. The University of Texas at Austin. Bryant, Vaughn M., Jr. 1966 a Pollen Analysis of the Devil's Mouth Site. In: A Preliminary Study of the Paleoecology of Reservoir the Amistad Area (assembled by Dee Ann Story and Vaughn M, Bryant, Jr.), multilith 54 sent to the National Science Foundation: report 129-164. 1956 b Pollen Analysis of the Devil's Rockshelter. In: A Preliminary Study of the Paleoecology of the Amistad Reservoir Area (assembled by Dee Ann Story M, Jr.), multilith report sent and Vaughn Bryant, to the National Science Foundation: 185-188. 1955 c The Amistad Pollen Reference Collection. In: A of the Preliminary Study of the Paleoecology Amistad Reservoir Area (assembled by Dee Ann Story M. and Vaughn Bryant, Jr.), multilith report sent to the National Science Foundation: 91-110. Chelf, Carl. 1941 Peat Bogs in Gonzales County, Mineral Resources Survey No. 34, The University of Texas Bureau of Economic Geology: 12 pp. Critchfield, W. B. and E. L. Little, Jr. 1965 Geographic Distribution of the Pines of the World. U.S. Department of Agriculture Forest Service Miscellaneous Publication No. 991, 97 pp. Cushing, E. J. 1964 Application of the Code of Stratigraphic Nomincla ture to Pollen Stratigraphy. Minisota Geological Survey (M.S.). Deevey, E. S. and R. F. Flint 1957 Postglacial Hypsithermal Interval. Science 125: 182-184. Faegri, K. and J. Iverson 1964 Textbook of Pollen Analysis. Hafner Publishing New York. 2nd. Ed., 237 Company, pp. Fisher, W. L. 1965 Rock and Mineral Resources of East Texas. The University of Texas Bureau of Economic Geology Report of Investigations No. 54, 439 pp. Flint, R. F, 1963 Studies of Pleistocene Wisconsin Stages in Central North America. Science 159: 402, Gould, F. W. 1962 Texas Plants-A Checklist and Ecological Summary, The Agricultural and Mechanical College of Texas Publication. Texas Agricultural Experimental Station, R. D, Lewis, Director. 112 pp. Graham, Allan. 1962 The Role of Fungal Spores in Palynology. Journal of Paleontology, Vol. 36, No. 1, 60-69. Graham, Allan and C. Heimsch 1960 Pollen Studies of Some Texas Peat Deposits. Ecology, Vol. 24, No. 4, 751-764. Hafsten, U. 1961 Pleistocene Development of Vegetation and Climate in the Southern High Plains as Evidenced by Pollen Analysis. In; Paleoecology of the Llano Estacado (assembled by Fred Wendorf) Museum of New Mexico Press Publication No. 1, Research Fort Burgwin Center. 59-91. Hevly, Richard H. 1955 A Preliminary Pollen Analysis of Bonfire Shelter. In: A Preliminary Study of the Paleoecology of the Amistad Reservoir Area (assembled by Dee Ann multilith report Story and Vaughn M, Bryant, Jr.) sent to The National Science Foundation: 165-178. Harbour, Jerry 1963 Pollen Analysis of a Blancan-age Lake Near Channing, Texas. IN: Environmental Reconstruction of A Blancan Lake Near Channing, Texas. Edited by: D. W. Kirkland and R. Y. Anderson (Unpublished manuscript). Johnson, Le Roy, Jr. 1951 Pollen Analysis of Two Archeological Sites at Amistad Reservoir, Texas. Texas Journal of Science, Vol. 15, No. 2, 225-230. Martin, P. S. 1963 The Last 10,000 Years, a Fossil Pollen Record of the American Southwest. Tucson, University of Arizona Press, 87 pp. Maxwell, Ross A. 1962 Mineral Resources of South Texas. The University of Texas Bureau of Economic of In- Geology Report vestigations No. 43, 140 pp. McAndrews, John H. and Donald A. Larson 1956 Cave. In: A Pollen Analysis of Eagle Preliminary Study of the Paleoecology of the Amistad Reservoir Ann M. Area (assembled by Dee Story and Vaughn Jr.), multilith report sent to the National Bryant, Science Foundation: 179-184. McAndrews, John H. 1955 A Preliminary Key to Some Texas Pollen Types. In: A Preliminary Study of the Paleoecology of the Amistad Reservoir Area (assembled by Dee Ann Story and Vaughn M. Bryant, Jr.), multilith report sent to the National Science Foundation: 111-127. Oldfield, Frank and James Schoenwetter 1964 Environments Man on Late Quaternary and Early the Southern High Plains. American Antiquity, Vol. 38, No. 151, 226-229. Parks, H. B. 1935 Plant Life of Ottine. Ins First Scientific Field Meet, Palmetto State Park, Ottine, Texas. Plummer, F. B. 1941 Peat Deposits in Texas. Mineral Resource Circular No. 15. The University of Texas Bureau of Economic Geology. 10 pp. Potzger, J. E. and B. C. Tharp 1943 Pollen Record of Canadian Spruce and Fir From a Texas Bog. Science, Vol. 98, 584-585. 1947 Pollen Profile From a Texas Bog. Ecology, Vol. 28, 274-280. 1954 Pollen Study of Two Bogs in Texas. Ecology, Vol. 35, 452-466, Shafer, G. H. 1941 Peat Deposits in Polk and San Jacinto Counties, Texas. Mineral Resource Survey Circular No. 38 The University of Texas Bureau of Economic Geology. 5 pp. Tharp, B. C. 1939 The of Texas. Texas of Vegetation Academy Science Publication of Natural History. The vita has been removed from the digitized version of this document.