MAR 1 3 2013 l . , r» g. '1 l . , .3 3" ABSTRACT LATE- AND POSTGLACIAL VEGETATION CHANGE IN SOUTHWESTERN NEW YORK STATE BY Norton George Miller Pollen stratigraphy in sediments from four small lake basins was determined and used as evidence for vege- tation change on the Allegheny Upland of southwestern New York State. The sites studied are within 35 mi of the un- glaciated Salamanca reentrant and are on an important mi- gration.route for species spreading northward following the Late Wisconsin glaciations. Forests of the hemlock-northern hardwoods type occur in southwestern New York at the present time. Point-quarter sampling of upland stands shows Acer saccharum, gags grandifolia and Tsuga canadensig to be the leading species in order of decreasing Importance Values. An analysis of bearing—trees recorded" in the original lot survey notes for areas around three of the sites studied palynologically re- vealed the pre-colonial forests to be dominated by the same leading species, except _E;a_g_u_s was first in importance and d- R values were calculated using the Acer secon \ l Nbrton,George~Miller pre-colonial data and a recent survey of existing timber re- sources in the region. The basins studied included the Genesee Valley Peat werks in central Allegany County--on Olean drift (pre—Cary), .Allenberg bog in east-central Cattaraugus County-—near the Kant terminal moraine (pre-Cary) and Houghton and Protection bogs in southeastern Erie County--on Valley Heads drift (= Port Huron). The profiles obtained were divided into A, B and C zones following the Deevey classification. In ad- dition, a T zone characterized by high nonarboreal pollen The T zone (NAP) percentages occurred at Allenberg bog. pollen assemblages compare well with the modern pollen rain at Fort Churchill, Manitoba. The A zones differ according to the age of the drift on which the basins are situated. Mbst unique was the Genesee Valley site where spruce (EE- 25%) occurs with abundant NAP (40 to 45%) . Spruce values decreased upward. The significance of the assemblages is obscure, but taken at the presence of an open vegetation type, perhaps face val ue , At Allenberg bog, Similar to park-tundra, is indicated. fluctuations in Fraxinus nigrg and Quercus percentages Sug-‘- ge'st correlation With climatic modifications associated with glacier [advance and retreat. :rom this site indicate that the fluctuations oc- However, absolute pollen fre- quency data f Curred as a response to increasing deposition rates for pine h Norton George Miller and spruce pollen. WOod at the bottom of zone A at Houghton bog has been dated at 11,880_1 730 years B.P. (I-3290). Upper A zone spectra, except for the presence of pollen from temperate deciduous trees, are similar to surface spectra occurring today in the boreal woodland of central Quebec. The spruce woodland disappeared around the Valley Heads sites about 10,500 years ago and was replaced by B zone forests dominated by Pinus strobus. At several sites lower pine-birch and upper pine-oak subzones can be dis- tinguished. At Protection bog where the pine peak has been dated at 9030 _-+_- 150 years B.P. was recovered from sediments deposited about 10,500 years (I-3551), a g. strobus cone ago 0 Zone C-l records the development of hemlock-northern hardwoods forests. With the exception of gradually in— creasing Fagus values, the profiles demonstrate stability in the regional vegetation during the interval between about 8000 and 4400 years B.P. An abrupt decline in hemlock per— centages marks the end of the C-1 which was dated at 4390 i 100 years B.P. (I-3550) at Protection bog. Increased relative numbers of Acer saccharum, Betula, Cgrxa W and w occur in zone C-2. {Tsuga percent- .9 ages remain lOW- Absolute pollen frequency determinations he Cel/C-Z hemlock decline but show only slight in- affirm 1: he numbers of broadleaf tree pollen types being cI‘eases in t NOrton George Miller émposited. This fact and the tendency for hemlock to ex- tfibit high drought mortality indicate that a series of mmere droughts occurring over a relatively short time span may explain the relative frequency fluctuations in C zone sediments from western New York. Zone C-3 which began 1270 i 95 years ago (I—3549) at Protection bog was divided into the following subzones: C-3a across whichg‘gggg regains its position of prominence in the smofiles and C-3b in which abruptly increasing percentages CE'NAP, including Ambrosia, Plantago and.§gmgx, record European settlement and attendant forest clearance. LATE-.AND POSTGLACIAL VEGETATION CHANGE IN SOUTHWESTERN NEW YORK STATE BY Norton George Miller A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY -Department of Botany and Plant Pathology 1969 '- w" -~‘-— - ACKNOWLEDGMENTS The Guidance Committee for this thesis was comprised of Dr. Aureal T. Cross (Chairman), Dr. John H. Beaman, Dr. John E. Cantlon, Dr. Gerald W. Prescott, Dr. Jane E. Smith and Dr. Stephen N. Stephenson. I appreciate the substantial help given to me by all members of the committee but extend Special thanks to Dr. Cross for assistance with various Phases of my doctoral program and to Dr. Beaman for generous- 1y allowing me to use the facilities of the Beal-Darlington Herbarium. I am grateful to The Buffalo Audubon Society, Inc. and The Nature Sanctuary Society of Western New York, Inc. f°r Permission to sample at Allenberg bog and Houghton bog, respectively. Russell B. Fales of the Erie County Bureau of Forestry kindly allowed me to work at Protection bog and Paul Butt-011 Of Belmont, New York gave me access to the Genesee Valley Peat Works. Robert A. Stickney and Arthur Flick, f”esters with the New York State Conservation Department, provided much appreciated help in locating forest stands suite'ble for study. I am especially thankful to my father, George C. Miller: for his aid in collecting the sediment samples. Brandt J. .Miller, Heather S. Miller, John M. Swan and ii Gary G. Thompson were helpful assistants during various as- pects of the field work. My wife, Heather, has provided much encouragement and has aided in the preparation of the manuscript. Additionally, I extend my sincere appreciation to the numerous other individuals who. offered many courtesies while I was pursuing the research reported herein. .My postgraduate study was supported by an NDEA Title IV Fellowship held from 1963 to 1966 and by a Michigan State University Graduate Council Fellowship held during the 1966 to 1967 school year. Field expenses and funds for radio- carbon dating were provided by a New York State Museum and Science Service Graduate Student Honorarium. To the people and institutions that made the granting of these awards POSSible, I offer my thanks. iii -~,--a L'J.‘ .‘. ‘ i 5'. . ... - \ ho. . H... I \ h. i ... . 5 5-. ‘ w, u \o. . ‘1. I ~y‘tv‘ \ n TABLE OF CONTENTS Page MHGmWLEDGMENTS - . . . . . . . . . . . . . . . . . . ii LISTOFTABLES.................... vii LISTOFFIGURES................... ix LISTOFPLATES.................... x LIST OF APPENDICES O o o o o o o o o e o e e o o o o o Xi IJST OF POLLEN DIAGRAMS . . . . . . . . . . . . . . . xii IHHRODUCTION . . . . . . . . . . . . . . . . . . . . . 1 THE REGION . . . . . . . . . . . . . . . . . . . . . . 3 Physiography . . . . . . . . . . . . . . . . . . . 7 Bedrock Geology . . . . . . . . . . . . . . . . . 12 Glacial Geology . . . . . . . . . . . . . . . . . 14 Soils . .. . . . .. . . . .. . . . .. . . . . 26 Climate . . . . . . . . . . . . . . 32 Flora .. . . . . .. . . . .. . . . . . 40 Vegetation . . . . . . . . . . . . . . . . 50 General Statement . . . . . . . . . . . 50 Forests of the Erie- Ontario Lowland . . . . . 60 Deciduous Forest Region of Southern Ontario . . . . . . . . . . . . . . . . 61 Deciduous Forests of Northwestern New York . . . . . . . . . . . . . . . 63 Forests of the Allegheny Upland . . . . . . . 67 Stands Sampled by the Quarter Method . . . . . 76 R Values . . . . . . . . . . . . . . . . . . 88 U. S. Forest Service Survey Statistics . . 89 Original Lot Survey Statistics . . . . . . 93 Discussion . . . . . . . . . . . . . . . . 102 Summary . . . . . . . . . . . . . . .'. . . . 109 iv Page THE POLLEN DIAGRAMS . . . . . . . . . . . . . . . . . 112 Methods . . . . . . . . . . . . . . . . . . . . . 112 Field Techniques . . . . . . . . . . . . . . . 112 Laboratory Techniques . . . . . . . . . 115 Sites Associated with the Valley Heads Moraine . . 122 Protection 809 . . . . . . . . . . . . . . . . 122 Sediment Stratigraphy . . . . . . . . . . 127 Houghton Bog . . . . . . . . . . . . . . . . 128 Sediment Stratigraphy . . . . . . . . . . 131 Pollen Stratigraphy . . . . . . . . . . . . . 132 Zone A . . . . . . . . . . . . . . . . . . 133 Zone B . . . . . . . . . . . . . . . . 140 141 143 Zone C-l . . . . . . . . . . . . . Zone C-2 . . . . . . . . . . . . . . . Zone C-3 . . . . . . . . . . . . . . . 144 Lockport Site '. . . . . . . . . . . . . . . . . . 146 Sites Associated with Pre-Valley Heads Moraines . 155 Allenberg Bog . . . . . . . . . . . . . . . . 155 Sediment Stratigraphy . . . . . . . . . . 163 Pollen Stratigraphy . . . . . . . . . . . 164 Zone T . . . . . . . . . . . . . . . 164 .zone A . . . . . . . . . . Zone B . . . . . . .. . . . . . . . . . 168 Zone C-l . . . . . . . .. . . . . . . . 169 Zone C-2 . . . . . . . . . . . . 171 Zone C-3 . . . . . . . . . 173 Pollen Size-Frequency Measurements . . . . 175 Betula . . . . . . . . . . . . . . . . 177 Picea . . . . . . . . . . . . . . . . 181 Pinus . . . . . . . . . . . . . . 184 Genesee Valley Peat Werks . . . . . . . . . . 188 Sediment Stratigraphy . . . . . . . . . . 190 Pollen Stratigraphy . . . . . . . . . . . 191 Zone A . . . . . . . . . . . . . . . . 192 ,Zone B . . . . . . . . . . . . . . . . 193 :Zone C-l . . . . . . . . . . . . . . . 196 INTERPRETATION.................... 198 Zonecr......................198 ZoneA......................208 Genesee Valley Peat Werks . . . . . . . . . . 209 mAllenberg Bog . . . . . . . . . . . . . . . . 216 Houghton and Protection Bogs . . . . . . . . . 229 l‘leB......................241 ZoneC-l....................252 .Zonec-z....................261 Zonec-a....................274 3;! SUMMARY AND CONCLUSIONS LITERATURE CITED . . . APPENDICES . . . . . . POLLEN DIAGRAMS - . vi Page 280 290 307 315 Table 10. LIST OF TABLES Selected climatic data from weather stations in western New York State by physiographic region 0 O I O O O O O I I O O O O O O O 0 Selected climatic data from weather stations in western New York State by physiographic region 0 O O 0 O I O O O 0 O I O O O O 0 Potential evapotranSpiration and growing degree months at selected weather stations in western New York State . . . . . . . Total area, non-forest land area and forest land area of western New York State by counties . . . . . . . . . . . . . . . . . Percent of total commercial forest land by forest type in eight western counties of New York State 0 I O O O 0 O O I O O O I 0 Forest stand data: general information Forest stand data: trees and saplings, Canadaway Creek Game Management Area, Chautauqua County, New York . . . . . . . Forest stand data: herbs and seedlings, Canadaway Creek Game Management Area, Chautauqua County, New York. . . . . . . . Forest stand data: trees and saplings, Forestry Department Plantation #11, Erie County, New York . . . . . . . . . . . . . Forest stand data: herbs and seedlings, Forestry Department Plantation #11, Erie County, New York . . . . . . . . . . vii Page 35 36 38 56 59 79 80 81 82 83 Table 11. 12. 13. 14. 15. 16. 17. 18. 19. Page Forest stand data: trees and saplings, Zoar Valley PrOperty #12, Erie County, New York . 84 Forest stand data: herbs and seedlings, Zoar Valley Property #12, Erie County, New York . . . . . . . . . . . . . . . . . . 85 Percent total volume of trees on commercial forest land in eight western counties in New York State . . . . . . . . . . . . . . . 92 Pre-settlement survey data: Relative fre- quency, relative density, relative dominance, importance values and importance percentages of bearing-trees used in the original lot survey of the areas around Allenberg, Houghton and Protection bogs . . 100 Pre-settlement survey data: Frequency of mention of tree species along lot survey lines for areas around Allenberg, Houghton and Protection bogs . . . . . . . . . . . . 101 R values calculated using various estimates of vegetation composition . . . . . . . . . 103 Data for check on mounting techniques used in the determination of absolute pollen frequency 0 O O O O O O O O O O O O O 0 O O 120 Lockport site bryophyte fossils by habitat type C O O O O O O O O O O O O I O O O O O O 15 3 Pollen spectrum from a silty-clay lacustrine deposit of Lake Iroquois age, Lockport site, Niagara County, New York . . . . . . . 156 viii LIST OF FIGURES Figure Page 1. western New York State . . . . . . . . . . . . 8 2. Drift borders and strandlines in western New York State . . . . . . . . . . . . . . . l7 3. Vegetation maps . . . . . . . . . . . . . . . 54 4. Size-frequency graphs of Betula pollen from Allenberg bog . . . . . . . . . . . . . . . 178 5- Size—frequency graphs of Picea pollen from Allenberg bog using wingtip to wingtip measurements . . . . . . . . . . . . . . . 182 6- Size-frequency graphs of Pinus Diploxylon pollen from Allenberg bog using measure- ments of external body diameter . . . . . . 187 7- Houghton bog-Section B: number of ter- restrial pollen and spores per m1. of wet sediment . . . . . . . . . . . . . . . . 318 3- Allenberg bog-Section C: number of ter- restrial pollen and spores per m1. of wet sediment . . . . . . . . . . . . . . . . 323 ix Plate Figure A. Figure Figure Figure B. B. LIST OF PLATES Page Protection bog, view looking southeast, August 23, 1967 Houghton bog, view looking east, August 26, 1967 . . . . . . . . . . 125 Allenberg bog, view looking 'northwest, August 28, 1967 Genesee Valley Peat WOrks, view looking northwest, September 2, 1967 o o o o o o o o o o o I o o o 160 LIST OF APPENDICES Appendix Page A. Pollen spectra above and below gyttja samples used for C-l4 age determination. at Protection bog O O O O O O O O O O O O 0 307 B. Percentages of minor pollen and Spore types not shown on pollen diagram: Protection bog O O I O O I Q 0 O O 0 O O O O I O O O O 308 C. Percentages of minor pollen and Spore types not shown on pollen diagram: Houghton bog-section A o o o o o o o o o o o o o o o 309 D. Percentages of minor pollen and spore types not shown on pollen diagram: Houghton bog-section B . . . . . . . . . . . . . . . 310 E. Percentages of minor pollen and spore types not shown on pollen diagram: Allenberg bog-section A . . . . . . . . . . . . . . 311 F. Percentages of minor pollen and Spore types not shown on pollen diagram: Allenberg bog- SGCtiOD B o o o o o o o o o o e o o o o 312 G. Percentages of minor pollen and Spore types not shown on pollen diagram: Allenberg bog-SGCtion C o o o o o o 0 ' o o o o o o o 313 H. Percentages of minor pollen and Spore types not shown on pollen diagram: Genesee Valley Peat werks . . . . . . . . . . . . . 314 xi LIST OF POLLEN DIAGRAMS Diagram 1. Protection bog: Relative pollen frequency . 2. Houghton bog-Section A: Relative pollen frequency . . . . . . . . . . . . . . . . 3. Houghton bog-Section B: Relative pollen frequency . . . . . . . . . . . . . . . . 4- Houghton bog-Section B: Absolute pollen frequency . . . . . . . . . . . . . . . . S. Allenberg bog—Section A: Relative pollen frequency . . . . . . . . . . . . . . . . 6- Allenberg bog—Section B: Relative pollen frequency . . . . . . . . . . . . . . . . 7- Allenberg bog-Section C: Relative pollen frequency 0 O O C O O O O O O I O O C O O 8. Allenberg bog-Section C: Absolute pollen frequency . . . . . . . . . . . . . . . 9- Genesee Valley Peat Werks: Relative pollen frequency . C C C C O O C O C 0 O O C . xii Page 315 316 317 319 320 321 322 324 325 to. a. .. I. ‘ ""‘ I . f, . H u u a i. \u .1«_ /' V v INTRODUCTION This thesis treats a problem that is primarily historical in nature-—vegetation change through time follow- ing deglaciation. The region involved is at the west end of New York State where surfaces, according to available data, have been ice-free for at least 12,500 years or longer, southward toward the Pennsylvania border. Temperate broad— leaf deciduous or deciduous-coniferous forests now character- ize western New York. The history of their development, as Well as information on other vegetation types that no longer exist in the area, are major objectives in this investigation. The principal technique I have used is pollen analy- sis whereby the vertical succession of pollen and spores is determined in sediments that have been accumulating in small lake basins for thousands of years. Changes in the relative and absolute frequency of various pollen types provide a b38is for inferences concerning the history of past vege- tation. These are supplemented, when possible, by additional data from plant macrofossils. Conclusions on both floristic and vegetational history and character can be reached through Pollen analysis, although the latter are often tentative and difficult to relate to the frequency and abundance of species “OW encountered on the landscape. Since major vegetation classes presumably develop in response to regional climates, pollen analysis further provides a way to determine former climates. Pollen analysis is ecological in approach and empha- sis. An understanding of the requirements of individual species and the relationships of species to associated organisms is necessary for the interpretation of pollen pro- files. Indeed, an assumption basic to paleoecologic recon- structions is that fossil representatives of an extant species had the same ecological characteristics as now occur in the living representatives. Since the unglaciated Allegheny Plateau and Appala- chian Mountains doubtless served as a refuge for certain plant:Species during Pleistocene glaciations, southern New York is a particularly critical area to conduct such re- search because it is one of several important migration routes for species that participated in the revegetation of eastern Nerth America. THE REGION For purposes of this study southwestern New York State includes Cattaraugus, Chautauqua, Allegany and the southern portions of Erie and Wyoming Counties. It is diffi- cult and misleading to discuss this geographically restricted area, however, without reference to the surrounding region, particularly to the three northern counties of western New Yerk, Genesee, Niagara and Orleans. Thus data from eastern New Yerk, New England, Pennsylvania and southern Ontario will frequently be used, but because the whole of western New York forms a coherent unit, historically, vegetationally and otherwise, this area will receive the most emphasis. The political geography of the region is shown in Figure l. The eight—county region includes nearly 6550 Square miles. It lies between 42° 00' and 43° 25' N Lat and 770 45' and 790 45' W Long, or, in more general terms, extends south- ward from Lake Ontario nearly 100 mi to the Pennsylvania border and eastward from Lake Erie and the Niagara River 100 to 65 mi, depending on the latitude, to the Genesee River. The counties along the Pennsylvania border are by far the largest in area, but they are the least populated. At the present time, the greatest concentration of population occurs around Buffalo and extends northward into the southwestern corner of Niagara County. Smaller population centers occur at Dunkirk, Jamestown, Salamanca, Olean, Lockport and Batavia. As political units the counties date from the early part of the 19th century. Prior to this time the region was visited by few Europeans, although in 1679, slightly more than 50 years after the formation of the Plymouth Colony in Massachusetts, an expedition under Robert Cavélier de LaSalle established a short-lived outpost at the mouth of the Niagara River (Williams, 1947). It wasn't until 1720, however, that the region had its first permanent resident and not until the late 18th century and early 19th century that more than a handful of settlers were present. In 1810, for example, there were only about 16,000 inhabitants in the eight county region, but a decade later the population totaled about 75,000 and hardly any district lacked the beginnings of settlement (Meinig, 1966). The principal impetus to colonization was provided by the Helland Land Company, a group of Dutch businessmen who were able to purchase from Robert Morris, through their American agents, the title to nearly all the land which com- prises the eight western counties of the state. The eastern portions of what are now Orleans, Genesee, wyoming and Allegany Counties were sold by Morris to various individuals in tracts of large acreage. Joseph Ellicott, one of the principal agents of the Holland Land Company, directed the surveying of the land owned by the Hollanders and did much to stimulate the development of the region while working from .his headquarters at Batavia in Genesee County. Ellicott planned the location of roads, established villages and di— rected the pattern of settlement, basing his decisions principally on economic considerations which favored his foreign employers. Many of his original developments have persisted in somewhat modified form to the present day (Evans, 1924). Before the arrival of settlers of European descent, western New York was occupied by a succession of Indian tribes. The historically important Iroquois controlled much of the state prior to the Revolutionary War. According to one current theory, they themselves did not actually inhabit New Yerk until about 1300 when a general northward movement of Iroquoian tribes occurred, supplanting the resident Algonkian Indians who had themselves migrated to the state from the west about three centuries earlier. The five Iroquois tribes which lived in central and eastern New York joined together in 1570 to form a confederation. One of these tribes, the "Keepers of the Western Door," as the Senecas were called, originally occupied the region between Cayuga Lake and the Genesee River, but later in the 1650's, they extended their influence to Lake Erie and the Niagara River by conquering the Erie and Neutral tribes which pre- viously controlled these areas. The Iroquois generally lived in stockaded villages containing about 250 people. Their homes were often sur- rounded by small, partially cleared fields where corn, beans and squash were cultivated. The Indians frequently chose rich agricultural lands on alluvial soils for their villages and these sites were also highly favored by the later white settlers. It has been estimated that about 20,000 Iroquois lived throughout New York State during the 18th century (Rayback, 1966). During the Revolutionary War the Iroquois confederation which had allied itself with the British was crushed by the Americans. The Clinton-Sullivan expedition destroyed nearly every one of their large villages and ultimately paved the way for the easy extinction of the Indian title to much of the land in central and western New York. A complete summary of the archeology of Iroquoian and pre-Iroquoian Indians in New York has recently been provided by Ritchie (1965). Today, outside the rapidly growing urban centers of the region, farming constitutes the largest percentage of land use. The most heavily cultivated areas occur in Niagara, Orleans, Genesee and northeastern Erie Counties and in a narrow strip in Chautauqua and southern Erie Counties im- mediately adjacent to Lake Erie (Thompson, 1966). Vegetables and fruits are the principal crops throughout this region. Of these counties, Niagara and Orleans contain the highest percentages of nonforest land—-83% and 80% respectively (see Table 4; p. 56). Southward, but north of the Allegheny River and southeastern Allegany County, dairy farming and the supportive growing of feed crops for the cattle accounts for most of the land use. Here and in the largely non- agricultural lands of southern Cattaraugus and Allegany Counties forest covers about 60% of the area, but much of this is of secondary character. PHYS IOGRAPHY One of the principal physiographic regions of western New York State extends south from Lake Ontario to east- central Erie and southern Genesee Counties and southwestward along the south shore of Lake Erie in a strip of land about 2 miles wide at the Pennsylvania border and about 15 miles wide in central Erie County (Fig. 1). This region is general- ly called the Erie-Ontario Lowland (Muller, 1965) and is mostly underlain by easily-eroded shales, although two prominent east-west trending limestone escarpments occur in the area adjacent to Lake Ontario. These features subdivide the lowland into three more or less flat plains in part covered by lacustrine sediments deposited from the ancestral stages of Lakes Erie and Ontario. The Ontario plain is located directly south of Lake Ontario and terminates about 10 mi to the south at the .1. 3.11) A 1:.- ..0 3:06.; $ >0. 5.5:! ‘OuO—eecl I > ( r 1 Id It" 0.25.20 324 a Z (4“? v) melatxoo... MRI... 0 , a Z a .. 3.30.2! . _ _2 .l..<. . . . so.!“(o n :32}... 4 ,u . 531......” .IJJ 2 n o 2.0::- 50‘30’ o .00. :5an a :01553 #:3052135 III-I .3153; teacqul MEGS—kw ¥¢O> >>NZ ZKMPWM; .r MKDO.& Niagara escarpment which, at places, rises 200 feet above the plain. Between the Niagara escarpment and the Onondaga escarpment, 10 mi to the south, lies the Huron plain. The soft shales of the Salina Group between the two limestone ridges were easily eroded to form a trough which held a shallow lake during late- and early postglacial time.1 The Onondaga escarpment is less prominent, rising only 100 feet or less above the surrounding lowland. It is also the northern edge of the third subregion of the Erie-Ontario Low- land, the Erie plain, which ends at the northern edge of the Allegheny Plateau and includes the previously mentioned narrow strip of land bordering the southern shore of Lake Erie. The elevation of these plains increases southward from 246 feet above sea level at the surface of Lake Ontario to 572 feet at the surface of Lake Erie to about 1000 feet at the north edge of the upland plateau. The Allegheny Plateau, or Appalachian Upland as it is sometimes called, is the second principal physiographic region of western New York. The northern edge of this unit is marked by the Portage escarpment which locally is more or less coincident with the northern erosional edge of strata belonging to the Upper Devonian Canadaway, Java and west Falls Groups. The Allegheny Plateau is characterized by 1The term 'lateglacial' is used to include the time up to the disappearance of spruce forests in western New Ybrk. 'Postglacial' encompasses the time from this point to the present. q... - - a o . ‘v; I' n.‘. I i h I I z;- _ ':Q l I ‘- a . \ K ‘ uq’ ‘- ‘._ .u ‘ ~ .- v '(V 10 flat-topped hills alternating with broad valleys. This tOpography is largely the result of pre-Pleistocene stream erosion, but glaciation has rounded the hills and widened many of the valleys. Elevation and relief increase south- ward. The highest point in western New York, 2548 feet, is found near Pikesville in southern Allegany County. The Allegheny Plateau is sometimes subdivided into glaciated and unglaciated sections. Southeastern Chautauqua, southern Cattaraugus and southwestern Allegany Counties, an area approximately bounded by the Allegheny River, were ap- parently never completely invaded by ice. Muller (1963) points out that the unglaciated region is characterized by less smoothly eroded ridges, more continuous crest lines and deeply incised V-shaped valleys. He further notes that, as far as 15 miles north of the limit of glaciation, summit re- duction by glacial scour was as little as 50 to 100 feet, al- though farther to the north greater lowering occurred. Throughout most of southern New York less than 200 feet of bedrock was removed from the plateau t0ps (Muller, 1964a). The glaciated Allegheny Plateau is covered with drift of varying thickness. Deposits are generally thinnest at the approaches to the summits of hills where the bedrock is often only several feet below the surface. The drift is usually thicker in the valleys where it may reach a depth of several hundred feet. Many of the valleys are notable also because they were the sites of short-lived proglacial lakes 11 and now contain lacustrine sediments as evidence of their former status. rGlacial Lake Zoar, for example, covered a large area in northern Cattaraugus County. Other proglacial lakes existed in the northwest—southeast trending valleys of southern Erie County, in the upper Genesee River valley and elsewhere. The drainage of western New York is generally north— ward into the St. Lawrence River (see Fig. l). Streams in only the southern portions of Chautauqua and Cattaraugus Counties and the southwestern corner of Allegany County empty into the Allegheny River which is connected to the Mississippi by way of the Ohio River. The drainage divide separating the St. Lawrence and Mississippi watersheds ex- tends northeastward in Chautauqua County approximately following the crest of the Lake Escarpment moraine, which is several miles inland from Lake Erie. From there it may be followed eastward, across northern Cattaraugus County with a dip southward toward Little Valley and finally southeasterly across eastern Cattaraugus and western Allegany Counties. Part of eastern Allegany County is drained eastward into the Susquehanna River. The only large inland lake in the region is Lake Chautauqua which occupies the axis of a through valley in south central Chautauqua County. It is approximately 15 miles long, but only 2 miles wide. A number of smaller lakes exist and many of these originated as pits formed where ice 12 blocks melted. Eastward, but outside western New York proper, lie the Finger Lakes. Several large artificial lakes occur in lowland areas and along the beds of major rivers and streams. The Genesee River, which starts at the Pennsylvania border and follows a general northward course to Lake Ontario, is one of the major waterways of western New York. Of importance also is the Allegheny River which follows a shallow, inverted U-shaped course in southern Cattaraugus County. Cattaraugus Creek, Buffalo Creek and Tonawanda Creek and their tributaries, which flow approximately west— ward to Lake Erie or the Niagara River, drain a large part of central and northern western New York. Several other streams course generally northward across the Onondaga es- carpment, drain a portion of the Huron plain, and empty into Lake Ontario. ' BEDROCK GEOLOGY western New York State is entirely underlain by Paleozoic sedimentary rocks (Fisher-25.31., 1961), which are exposed at the surface in only limited areas. The beds dip gently to the south so that rocks of greater age are en- countered successively northward. More or less east-west trending belts of shale, siltstone and sandstone are present throughout the region, but the most important exposures of limestone and dolomite are found in the lowland north of the l3 Allegheny Plateau. Complete summaries of the palenotology and bedrock geology have recently been published by Buehler and Tesmer (1963) and Tesmer (1963), for Erie and Chautauqua Counties respectively. The red Queenston shale and siltstone of Late Ordovician age outcrops in the area immediately south of Lake Ontario and is the oldest rock exposed in the region. The predominantly calcareous rocks which comprise the Ni- agaran escarpment were deposited during Middle Silurian time and are included in the Lockport Group, although other Huddle Silurian rocks belonging to the Clinton Group are present toward the foot of the scarp. The Lockport dolomite, a member of the Lockport Group, is the erosion-resistant cap rock of present-day Niagara Falls. The Onondaga escarpment to the south is composed of limestone which accumulated during the early part of Middle Devonian time. The rocks comprising the escarpments were deposited during different periods of time when western New York was covered by warm shallow seas. The 10 miles separating the two zones of calcareous rock are mostly occupied by weak shales belonging to the Upper Silurian Salina Group. Later in Middle Devonian time, shales of the Hamilton Group succeed the Onondaga limestone, a result of a massive influx of sediments eroded from an upland to the east. Upper Devonian rocks, mostly shales and siltstones, but occasional- ly sandstones, and rarely, conglomerates and thin limestones, . v; .. . I\ or. t. . ”K “ :7 I. on; I . - .ru 14 are found southward from central Erie and southern Genesee Counties. These belong, in order of decreasing age, to the following Groups: Genesee, Sonyea, Java, West Falls, Canada— way, Conneaut and Conewango. The sediments comprising these rocks gradually become coarser upward, indicating a westward movement of shoreline during the latter part of Upper De— vonian time, and are part of the Catskill Delta which ex- tends from what is now the Catskill MDuntains region west to (mio. Near the Pennsylvania border, largely within the un- glaciated area of western New York, are found shales, sand- stones and conglomerates which belong to the Lower Missis- sippian Pocono Group and to the Lower Pennsylvanian Potts— \dlle Group. These deposits have been greatly dissected by erosian and may be considered outliers of more continuous Strata of the same age to the south. Muller (1963) suggested that the resistant conglomerates in this area may have terminated southward glacial movement. GLACIAL GEOLOGY In contrast to the age relationships of the bedrock units, surficial deposits resulting from Pleistocene glaci- ations are oldest in the south and youngest in the north. Dr'ift from only the Illinoian and Wisconsin glaciations has been identified in western New York, although study of the clevelopment of the present Allegheny River suggests the 15 pmesence of pre-Illinoian glacial activity (Muller, 1963, 1965). Deposits of Illinoian drift have weak geomorphic ex- pmession. They consist mainly of terrace remnants in the Allegheny River valley 50 to 100 feet above the present river level: Wisconsin terraces are found from 10 to 20 feet above the water surface (MacClintock and Apfel, 1944). Post- Illinoian erosion and weathering have left only isolated patches of Illinoian drift in this region. Muller (1960) Showed that some of the gravel terrace remnants recognized by MacClintock and Apfel bordered a lobe of an ice sheet that projected southeastward across the Allegheny valley CF19» 2% .A separate pre-Wisconsin glaciation is indicated by the occurrence beyond this border of terrace remnants pro- duced by an additional period of ice-margin deposition. This tWofold expression of Illinoian deposits has long been recog- nized in adjacent northwestern Pennsylvania. But it has only recently been determined that what had been called "Inner Illinoian" drift is actually referable to late mid-Wisconsin time (White and Totten, 1965). This may also be true of Similar deposits in southwestern New York State. In south- eaStern Chautauqua County patches of attenuated drift, Emesent beyond the Wisconsin terminal moraine, are thought t0 be of Illinoian age (Muller, 1963). The earliest recognized deposit of Wisconsin age is the Clean drift which in western New York State is found at Q. 16 the surface on the north and east sides of the unglaciated :nrtion of the Allegheny Plateau (Fig. 2). This roughly tri- angular area is often referred to as the Salamanca reentrant because it is located at the junction of deposits produced by ice moving southeastward out of the Lake Erie basin and similarly derived deposits from the north and east from the Lake Ontario basin. Olean drift as mapped by MacClintock and Apfel (1944: see algg Denny and Lyford, 1963) forms the terminal moraine on the north and east sides of the reentrant and follows a greatly convoluted course extending slightly South of east from about 3 miles northeast of Napoli to near Humphrey where its position extends abruptly southward to- ward the village of Allegheny. From this point it angles mbutheastward toward the corner of Cattaraugus County. It is a very subdued terminal moraine but the limit of glaci- ation.can be accurately determined by the presence of erratic Cdbbles (Muller, 1965). The exact age of the Clean moraine is not known. MacClintock and Apfel (1944) considered it IoWan or Tazewell in age; other authors have concurred with this disposition. However, Denny and Lyford (1963) point out that it may be a pre—Farmdale—post-Sangamon substage. There is, at any rate, general agreement that it is pre-Cary in age . The section exposed near the village of Otto in northwestern Cattaraugus County displays evidence of a long hiStory of successive glaciations which are now considered l7 8:: i=3: 3:: 3.3... 320.0 .2920 Gig g :‘\.\.!/. DZ‘IF. . . 20.044 . .......I.. a. 0.0340... . 102(1km .0:S.oc.oum >UNUJFLLI5> O— n O o... .toaauOL 3.2.3 3.; >0:u> 0:0:00 no. cote-.9; no. .3295: can otencoz< :2... Oumoo_ocx_ou to. to; u:— tos... to on .o .3130.- : cots-.0: .3:- 3:93; :eaou eta «p.10: >u=o>sooNrion eo_.o:::ou .n.».z of; EOE-06.3.1 2:00 1.0.0 xolovucou 1.: _au_oo_0uo .0. $3: 3:39) .0 e23»: —flfl'0 0 N p 4 on? 5.3.22.2: 3 Earl! ulfi:‘flflbomvl Soo- isid YE >530 lot: :11 ....§.c2iaga£- 3:23:11 «2:30! 5:33.} .0 2:..3' w._.<.rm zmo> >>mz ZIP—hm; z. mmz_..oz06 666 0>onm poem :6 cowum>66m .om umAEoumom ou 6 an! m N .uflmnswunmm mooummo :6 monouMHOQEou «Amwmdv mucouoz 806m muwn 6 66- 66 6.66 6.66 6.66 6.66 6666666666 1.6666. 66666o> 66- 66 6.66 6.66 6.66 6.66 66666666660 1.6666. ouuo 66- 666 6.66 6.66 6.66 6.66 6666666660 1.6666. 636666666 66- 66 6.66 6.66 6.66 6.66 66666666660 1.66666 66666666 66- 66 6.66 6.66 6.66 6.66 66666666660 A.6666c 6666>666x66um 66- 666 6.66 6.66 6.66 6.66 6666666: 1.66666 66>6666 66- 66 6.66 6.66 6.66 6.66 6666663 ..6666. 666664 66- 666 6.66 6.66 6.66 6.66 66666666 ..6666c 66666666 66- 666 6.66 6.66 6.66 6.66 66666664 1.6666c 6666666 66- 666 6.66 6.66 6.66 6.66 66666666666 1.6666c 6666 oumum acosomadd 66- 666 6.66 6.66 6.66 6.66 66666664 1.66666 66666< oCMHQD aconmoddfl 66- 66 6.66 6.66 6.66 6.66 6666666660 1.66666 66666666: 66- 666 6.66 6.66 6.66 6.66 6666 1.6666c 66663 66666 oumwooeuoucH 66- 666 6.66 6.66 6.66 6.66 6666660 1.666. 66666666 6- 66 6.66 6.66 6.66 6.66 6666666 1.666c 66366666 66- 666 6.66 6.66 6.66 6.66 6666662 1.666c 66666666 66- 66 6.66 6.66 6.66 6.66 6666666 1.666c 666 66 66- 66 6.66 6.66 6.66 6.66 6666666660 1.666. 66666666 66- 666 6.66 6.66 6.66 6.66 6666666 1.666. 6666 66- 66 6.66 6.66 6.66 6.66 6666 1.666. 6666666 66- 666 6.66 6.66 6.66 6.66 6666662 A.666c 6666666< m uc66306 owumucono6um .mEoB .QEOB .QEoH Noemmom .m509 .m509 .QEoB huaooo :O6um00Q 660306 amonmwm mc63ouo c602 M666 one: humscmb c662 Hasscd :66: HZOHUNm UHMA‘flGOHmMmm Mm mfiflam KKON 3mz zmwamm3 2H mZOHHGBm mama06 mom w>onm noon :6 co6um>06m .66 666666666 66 6 66: .606026 66 666w3ocm pom COHHmuwmwooun “Amemdc «mound! 506w mumo m N H 666 6.66 66.66 66.66 66 .6666 66 66: 6666666660 6.66666 666666> 666 -- 66.66 66.66 66 .606 66 66: 66666666660 6.66666 o66o 666 6.66 66.66 66.66 6 .606 66 66: 6666666660 6.6666. 636666666 666 6.666 66.66 66.66 66 .6666 66 66x 66666666660 6.66666 66666666 666 6.66 66.66 66.66 66 .6666 66 66: 66666666660 6.66666 6666>66666666 666 6.66 66.66 66.66 66 .6666 6 6666 66666666 6.66666 66>6666 666 6.66 66.66 66.66 66 .6666 66 66: 6666663 6.6666. 666064 666 6.66 66.66 66.66 66 .6666 66 66x 66666666 6.6666. 66666666 666 6.66 66.66 66.66 66 .6666 66 66: 66666666 6.6666. 66>6666 666 6.66 66.66 66.66 66 .6666 66 66: 66666666660 6.66666 6666 oumum acmmodad 666 6.66 66.66 66.66 66 .6666 66 66: 66666666 6.66666 666666 McMHQD 560660664 666 6.66 66.66 66.66 66 .606 6 66: 6666666660 6.6666. 666666663 666 6.66 66.66 66.66 6 .606 66 66: 666m 6.66666 6666: 66666 oumwooeuoucH 666 6.66 66.66 66.66 6 .606 66 66: 6666666 6.6666 06666666 666 6.66 66.66 66.66 6 .606 6 66: 6666666 6.666. 66366666 666 6.66 66.66 66.66 66 .6666 6 66: 6666666 6.666. 66666066 666 6.666 66.66 66.66 6 .606 6 66: 6666660 6.6666 666 66 666 6.66 66.66 66.66 66 .606 6 66: 6666666660 6.666. 66666666 666 6.66 66.66 66.66 6 .606 6 66: 6666660 6.6666 6666 666 6.66 66.66 66.66 66 .606 66 .666 6666 6.666. 6666666 666 6.66 66.66 66.66 66 .606 6 66: 6666662 66.6666 66666666 0:66:06 Owumusoloaum .o6006m Hamwsocm .mwooum .owuoum 6606b umoum 2663 when Honest cannon 666::4 666m umuwm mcwumm 6666 666664 :60: :60: new 060 can! :66: no open .oal no open .o>¢ aucsou sawumooq “6"” “6H" HZOHUQK UHEH‘flUOHmMmm Mm wHHHHHHHHHHHHHLWHHHHHHHHHHHHHmHHHHHHHHHHHHHHHHHHHHHHHHH ma3 mnmoud «.maa o.~m msmsmumuumo xumm mumum mammmaad GGMHQD mamsmmaad m.m.v._u h..VN MHMUMHZ COumszQH H.¢ma H.¢m mflum oamwmsm ©Cma3oq OAHMHGOImwum manna: ,cowumuflmmcmuuomm>m a G50 cod no wmummn mcw3ouw Hmwucmuom u U .u on HBdBm_KMOM.3mZ zmflamm3.ZH mZOHaflem mmma¢fi3 QWBUMAmm H 9‘ @3920: NNMUNQ QZH30MU 92¢ ZOHHfiMHmmZm AdHBZflBOm m m4m<9 ‘r I. .I ‘u .l \- I1 l.‘ ‘I. H‘. .Io. 39 Niagara County and nearly all of Wyoming County, in contrast, are among the driest parts of the whole state. Precipitation is fairly evenly distributed throughout the year, but summer months characteristically receive more than the others. During the period from May 1 to September 30, the limits of the growing season (see Table 2), less rain falls in lowland areas than on the upland. Although summer is the season of greatest precipitation, it is also the time of greatest moisture need, so small moisture defi- cits occasionally occur. These are most critical in the heavily cultivated lowland region and may sometimes lead to crOp failure. Few major droughts, however, have affected the region. A serious one occurred in 1899 when the total precipitation for the three summer months was less than 3 inches at many localities bordering Lake Ontario (Mordoff, 1949). There has been at least one noteworthy period of drought every 20 years. The prevailing winds are westerly across the region but often shift to the north in winter and toward the south during the summer. At Buffalo the winter months are characterized by west-southwest winds, while during the rest 0f the year, except the fall months when the wind is from the south, wind direction is from the southwest. Winter is the time of maximum cloud cover. At Buffalo only 1 to 3 days, during the months of December,~January and March are listed as clear days with less than 0.2 percent cloud cover. During IA-O n ‘U" ‘.-O 0U“. y.- - In. 'I On ( ‘- O,- - h.“ t I II! {It It i 40 most of the spring, summer and fall nearly 10 days of every month are clear. The whole of New York State lies within the eastern cloudy belt which ranks well below central and western states in the amount of sunshine received (ibid.). FLORA Floristic research in western New York state began in.earnest during the mid—1800's, Before this time, however, Niagara Falls had attracted naturalists and plant collectors to the region, many of whom published botanical observations made at the Falls gggg Dow, 1921) or while travelling across the Erie-Ontario Lowland. Unfortunately, most of this infor- mation seems to have been gathered casually or, in some cases, by untrained people, making it of questionable scien- tific value. Three of the more notable visitors were Peter Kalm, who viewed the cataract in 1750 (Kalm, 1751), Frangois André Michaux, who travelled throughout the eastern Great Lakes area in 1806 or 1807 (Zenkert, 1934), and Thomas Nuttall, who undertook a pedestrian trip from Philadelphia to Canandaigua and west to Niagara Falls in 1809 (Graustein, 1967). Zenkert (1934) has traced the history of botanical exploration in western New York from early times through the 1930's. members of the Buffalo Society of Natural Sciences ‘Were particularly active collectors during the last half of the 19th and the early part of the 20th centuries. In recent 41 years field work has been largely carried out under the auspices of the New York State Museum at Albany. Much of this more recent study remains to be published. A total of 1587 species are listed by Zerkert (1934) as growing in the Niagara Frontier region which encompasses the area within aASO mile radius of Buffalo. Of these, 1187 are native species; the remaining 400 are introduced. The total number of species probably has been enlarged somewhat since 1934. In comparison, House (1924) enumerates nearly 2950 native and introduced species in his flora of New York State, and for the area covered by Gray's Manual (Fernald, 1950), 4425 native and 1098 introduced species are listed. Three more or less distinct floristic regions occur at the west end of the state. The most northern is bounded on the south by the Wisconsin terminal moraine which approxi— mately separates the glaciated and nonglaciated parts of western New York (see Fig. 2). This boundary marks the southern limit of a number of boreal Species. In large part these are bog plants which presumably are not found south of the drift limit because of the absence of kettle holes and other suitable habitats generally associated with glaciated terrain. Zenkert (1934) has noted 55 species which exhibit this distribution pattern in western New York, although a few occur at disjunct stations southward beyond the New York- Pennsylvania border. Some of the species he lists are: I a It... ‘Pt- 42 VAndromeda glaucophylla,l Brasenia schreberi, Eriophorum spissum, Galigm.l§bradorigum, Habenagia dilatata, Kalmia pglifolia,.§a£;x_l§£icina, Ledum groenlandicum, Myrica gale, .2igga mariana, Pinguicula vulgaris, Populus balsamifera, Primula mistassinica, Rhynghospora alba, Saxifraga aizoides, Vaccinium oxycoccus_and Viola renifglia. An interesting plant community containing relict boreal species occurs within the unglaciated Salamanca re- entrant in the Red House valley at the west side of Allegany State Park (House and Alexander, 1927). When this site was visited during the summer of 1966, only a quarter acre patch of what was once an extensive Pinus strobus stand was still in existence. The principal trees present at this time were .Abigg balsamea, Age; rubrum, Betula alleghaniensis, Fraxinus .gig£§,_gigu§ strobus,_gggg§ canadensig and glans americana. Gaultheria hispidula, a boreal Species once found there, could not be located, although Vaccinium myrtilloides, an- other northern plant, was still present. Borings made with a Davis sampler revealed that beneath 10 to 13 cm of humified peat lay a bed of heavy blue clay, the thickness of which was not determined. .This inorganic sediment was apparently de— posited at some time during the Pleistocene from a lake :nnded in the valley of Red House Brook, a small tributary 1Nomenclature follows Fernald (1950) with the ex- cmption of the binomials used for yellow birch and leatherleaf which are Betula.alleghaniensi§ Britton and Cassandra calyculata (L.) D. Don, respectively. 43 of the Allegheny River. The community seems to be a remnant of a swamp forest developed on soil of large water holding capacity (Taylor, 1928). Superimposed on the glaciated-nonglaciated distri- bution pattern is another type which shows no relationship to the presence or absence of glaciation. The floristic regions delimited by the new pattern correspond to the Erie- Ontario Lowland and the Allegheny Upland and are best illustrated by species which occur in the former area but not in the latter. The boundary between the two regions is usually depicted as part of the well-known tension zone that crosses Minnesota, Wisconsin, Michigan, southern Ontario and western and central New York to the eastern end of Lake Ontario. In New York State it is coincident with the Portage escarpment (Fig. 1). Typical upland species occur on both sides of the glacial boundary. In central and western New York this tension zone has not been extensively studied. Plant distribution maps similar to those which support its existence in Wisconsin and elsewhere have not been published for New York State. However, in western New York it is clear that the upland is characterized by species which are absent or are of greatly restricted occurrence in the lowlands bordering Lakes Erie and Ontario. Conversely, Asimina_triloba, Celtis occidentalis and Sassafras albidum grow in the lowland or on the flank of 44 the upland, and either do not occur in the upland or are rare there. Soper (1962) has discussed a large number of species which are found in southern Ontario below a line extending from Grand Bend on Lake Huron southeastward past London, then northeastward in the vicinity of Aylmer to Toronto. These species occur nowhere else in Canada and because of their obvious southern affinities they are considered as part of the Carolinian flora of Canada, using the terminology of Merriam (1898). The ranges of only a few of them extend into the coniferous-deciduous forest region of northern Ontario. Of the plants listed by Soper, those which occur in western New York State are restricted to stations in the lowland, with the exception of a few found along the Allegheny River or in the region intermediate between the lowland and the higher parts of the upland. .Zenkert (1934) recognized the distinction between the lowland and upland flora and enumerated 96 species which he considered to represent an austral element best developed in the region adjacent to Lakes Erie and Ontario. This distribution pattern is the basis for Bray's recognition (1915) of two zones in western New York separated approxi- mately by the Portage escarpment and characterized by differ- ent tree species. His Zone B, in which chestnut, oaks, ruckories and tulip-poplar are typical, occurs in the Erie- (kmario Lowland, while his Zone C, characterized by sugar 45 Inaple, yellow birch, hemlock and white pine, corresponds in area to the Allegheny Upland. It is noteworthy that the northern elements of the flora are best developed on the Allegheny Plateau in southern ‘western New York, while the southern elements are best de— 'veloped to the north in the Erie-Ontario Lowland, a situation ‘which is directly opposite that found in Michigan and Wiscon- sin. Climate may be the major factor controlling this distribution, although soil and other edaphic factors may also have a role. The more rigorous climate of the upland ‘would tend to eliminate species adapted to higher mean xvinter temperatures and a longer growing season. The flora of western New York can be divided into a number of phytogeographic elements, each of which is made up of species that shareaa similar type of geographical range 1 today. These species are typical of a certain natural area, that is, the entire region of distribution of a taxonomic unit attained through natural dispersal mechanisms, whether they now grow within that area or not (Cain, 1944). They often have the same center of dispersal, but may or may not share a common center of origin. The identification of ele- ments in a regional flora is based upon their being typical cm certain well-defined phytogeographical areas elsewhere. Elements may be categorized as either extraneous or intraneous. The former contains species at or near the lindxs of their range which may, therefore, exhibit 46 disjunctions of various types, while the latter includes ‘widespread species whose occurrence in a particular region is well within the plants' total range (Braun, 1937; Cain, 1944). Intraneous species, which may comprise as much as 60 percent of a flora (Parker, 1936; Thompson, 1939), tell us little about the geographical affinities of that flora,' but extraneous ones are considerably more helpful in this re- «gard. Most of Curtis' discussion (1959) of the extraneous (elements in the flora of Wisconsin can be applied to New “York State, so much of the following account is drawn from this source. The Alleghenian element contains a group of species «of Arcto-Tertiary origin which center in the southern Appa- lachians and extend northward into southern Canada. Such ‘well-known and important forest trees as Acer saccharum, iBetula alleghaniensis, Fraxinus americana, Ostrya virginiana, Pinus strobus, Quercus alba, Tilia americana and Tsuga canadensis are members of this element. Also of Tertiary origin is the Ozarkian element which contains more drought tolerant species developed in isolation from the southern «Appalachians on the Ozark upland of Missouri and Arkansas. gggr saccharum var._gigrum, Carya spp., Quercus macrocarpa, __Q- muehlenbeszii and g. velutina are components of this ele— ment. Steyermark (l939).has compiled a list of plants common to both the Appalachians and the Ozarks which presumably acted together as a single center of origin and dispersal I 3.? 47 for certain species. Magnolia acuminata is one he mentions that is found in western New York. Mombers of the Boreal element are not rare in western New'York but in most cases they occupy positions in bog com- xnunities generally associated with senescent kettle hole .1akes. Abies balsamea, Larix laricina and_gigg§ mariana oc- cur here as members of this element. These trees are charac- teristic of the boreal forest which ranges across central Canada from eastern Alaska to the Atlantic seaboard, south to the upper Great Lakes with a discontinuous extension down the Appalachian mountains at higher elevations. Also oc- curring in western New York are such typical boreal shrubs and herbs as Andromeda glaucophylla, Cassandra calyculata, Ledum groenlandicum,*§inpaea borealis var. americana and W M among others- Another group of species characteristic of the region farther to the north but often found southward at mountain stations of high elevation belongs to the Arctic-alpine ele- ment. As expected it is very poorly represented in western New York but has better expression in the high peak areas of the Adirondack and Catskill Mountains to the east. Pinguigula vulgaris and Saxifraga aizoides which grow together on a wet vertical gorge wall near a falls of the Genesee River in southeastern Wyoming County, their only station at the west end of the state (Zenkert, 1934L are members of this element. 48 Species typical of western North America but which are also found eastward are members of what can be broadly called a Western element. Actually this category includes several types of distinct distribution patterns, two of *which particularly pertain to western New York. The Prairie (element is made up of species whose ranges center on the existing prairies. Certain.members of this element such as .Andropogon gerardi, A” scoparius and Sorghastrum nutans have a wide distribution through the Erie-Ontario Lowland in New 'York.State where they apparently grew mostly in oak Openings Ibefore and for a while after the settlement of the region. !They are now generally found in abandoned fields, in hedge- rows and in thin second growth oak stands and are often associated with prairie forbs. Shanks (1966) felt that the oak openings in this area were essentially edaphic prairies, remnants of more extensive grasslands which occurred in this region when the Prairie Peninsula extended farther east. The shallow dry soils and the occasional water deficits characteristic of the Erie-Ontario Lowland would seem to favor the persistence of prairie species and exclude more mesophytic competitors. The Cordilleran or western Mountain element, another type of Western element, was early recognized in eastern North America by Fernald (1925). More recently Iltis (1965, 1966) has redirected attention to it, pointing out that " . . . the ranges of many of our commonest as well as rarest 49 species in the northeastern United States . . . fall into the standard pattern of eastern North America -- western Nerth America vicarious species pairs with the post-glacially produced modern ranges overlapping in glaciated northeastern NOrth America" (1965, p. 149). AS examples, Iltis (ibig.) cites a substantial list of paired species including the following in which both members are found in western New York: .Actaea.£ub£§ (western-w) “.§° pachypgda (eastern-e),_gigg§ latifolia (w) -—.9. arundinacea (e), Cypripedium parviflorum (w) -- g. pubescens (e), Gentiana procera (w) -- g. crinita (e), Junipgrus horizontalis (w) --.g. virginiana (e), Pogulus 'tremuloides (w) -—_g. grandidentata (e), Salix serrissima (w) --.§. lucida (e) and yiglg adunca (w) -- y. conspersa (e). Certain additional Species claimed by Iltis (ibig.) and others to be members of this western element are found at the ‘west.end of New York as well. These include_ggrex flava, Oryzogsis asperifolia, Potentilla arguta,_g. fruticosa, Pterospgra andromeda,.§glix candida and Valeriana uliginosa. Another group of species belonging to the Atlantic Coastal Plain element has a limited distribution along the beaches of Lakes Erie and Ontario and westward around the upper Great Lakes (Peattie, 1922). These species apparently attained their current range sometime during late- or post- glacial time, perhaps by migrating along the St. Lawrence or Mohawk River valleys and thence along the shores of the ancestral Great Lakes. Cakile edentula, Euphorbia / 50 polygonifolia, Lathyrus maritimus and Xyris caroliniana are a few of the coastal plain species which occur in western New York. The Exotic element containing non-native species *which have entered the region through the activities of man remains to be discussed. Of particular interest palynologi— cally are certain Species of Plantago, especia11y_g. lanceo- _l§£§ and E. major, two EurOpean species widely naturalized in North America. The appearance of Plantago pollen in postglacial sediments which can be mostly attributed to these species clearly marks the arrival and Spread of EurOpeanS in America. However, there are many other examples of the Exotic element in the flora and, as has been pointed out earlier in this section, 25 percent of the total flora of the Niagara Frontier region is comprised of introduced species. VEGETATION General Statement Authors of the earliest histories which have been published about western New York are uniform in stating that the region was completely covered by forest except for dis- continuous openings in the oak forests of the Erie-Ontario Lowland and other small partially cleared areas associated with Indian villages. (A general account of the whole region published a few years after settlement began contains an 51 interesting discussion of the kinds of timber present and the agricultural potential of the land on which grew certain forest types (Munro, 1804). . . . in the better or most even parts of the country . . . the most common sorts of timber . . [are] . . . sugar maple, beech, lyn (here called bass- wood), oak, ash, and elm; and the hilly parts are mostly timbered with oak. Where the sugar maple and basswood are most common, the land is generally es- teemed best for grass, and probably grain, and is ex- perienced to be durable; the lands which produce mostly beech timber, are considered as generally clayey, wet, and cold. A considerable portion of the better part of the country is timbered with oak, and lands on which it is of a large growth are by many esteemed the most durable, although at first not productive of as good crops as maple lands, and harder in tillage. The sorts of trees and shrubs which are most scarce are hemlock fir, cucumber tree, white poplar, white and black birch, turmeric tree, Spruce pine, locust tree, prickly ash, spice wood, hazel nut, willow, and alder (p. 1173-1174). The observations of Rev. Mr. E. J. Hill, an ac— complished botanist, although made four to six decades later, provide a more complete description of the forest cover of western New YOrk. Rev. Hill was born at Le Roy in Genesee County in 1833 and spent much of the early part of his life in this region at a time during which undisturbed tracts of forest were still fairly abundant. Hill (1895) reports that: The most abundant trees of the upland woods are the Beech and Hard Maple. On light soils, and where there is a considerable mixture of sand or gravel with the clay loam, the Oaks predominate, inter- spersed with Hickory, and sometimes with the Chest- nut. In colder and higher tracts or along the banks of streams, the Hemlock, is frequent or even abundant. The Basswood is common in the richer uplands, among O“ IV C: ‘- lfll 52 Beeches and Maples. Here also the White Ash is most often seen. . . . Where the Beech and Maple abound the White Oak is occasionally mixed with them, but is mostly con- fined to the low land, where it is much more common than the Swamp White Oak. The Red Oak is much more commonly-seen with the Beech and Maple. In flinty and gravelly soils the most common Oaks are the White, Red and Black Oaks. Here also occurs the Chestnut Oak; it is usually less abundant than the other kinds and may also be found in the wet lands (p. 382). Turning briefly to historical records which pertain to either of the two physiographic regions, an informative account of the original timber covering of Orleans County indicates in a general way the nature of the forests through- out the Erie-Ontario Lowland during the period of EurOpean settlement (Arad, 1871). In its natural state Orleans County was thickly covered with trees. On the dry, hard land, the pre- vailing varieties of timber were beech, maple, white, red and black oak, white wood or tulip tree, bass- wood, elm, hickory, and hemlock. Swamps and low wet lands were covered with black ash, tamarack, white and yellow cedar, and soft maple; large sycamore or cotton ball trees were common on low lands and some pine grew along Oak Orchard Creek, and in the swamps in Barre; and a few chestnut trees grew along the Ridge1 in Ridgeway, and in other places north of the .Ridge (p. 29). In comparison, C. G. Locke's description of the forests of Cattaraugus County, which pertains to much of the western Allegheny Upland in New York State, emphasizes the prevalence of hemlock-and pine in this region at the time of settlement (in Adams, 1893). lLake Iroquois strandline. 53 This table-land was originally covered with a heavy growth of deciduous trees intermixed with hem- lock and some pine, and this same description of the original forest would apply to the entire northern portion of the county, excepting that pine was generally found along the low-lands. The southern part of the county was covered with forest of the choicest pine and hemlock, with a mixture of de- ciduous trees. Here we find the home of the white and red oak and chestnut, which apparently did not cross the dividing ridge, as very little of this timber is found in the northern part of the county (p. 50). These passages clearly indicate that the forests of the lowland and upland were of a different type‘with beech maple, oaks and other deciduous species predominating in the former, while a mixed forest of conifer and deciduous trees occurred in the latter. Most botanists who studied the vege- tation of this region in more recent years have also made this distinction. For example, Kuchler (1964) has recognized three main types of forest in western New York in his treat- ment of the potential natural vegetation of the contiguous United States (see Fig. 3B): (1) Beech-maple forest dominated by_§gg; saccharum and_§ggg§ qrandifolia, (2) Northern hard- woods forest dominated by_Acer saccharum, Betula alleghanien- sis, Fagugqrandifolig and Eggqg canadenSiS and (3) Ap- palachian oak forest in which Quercus alba and Q. EEQEQ are dominant but generally occur with many other subdominant. species. The boundary between (1) and (2) roughly corresponds to the Portage escarpment with the Nerthern hardwoods forest area in the upland and the Beech-maple forest area in the lowland, although inclusions of one type are mapped in the 54 _ cab cab AODOSCSED E9; not-goon .n AtQOCLo.:o=¥ .3:( .( < n.<_2 >9 pero>oo .33.! «09.01 caesguoog «neon $0032.... user—econ UOOLOK .ICLOD m \\\\\\\\~\ one 9 o o v u v v .0.0.0.0.0.0.0 >>wZ ZZMPMM; n—O 20....(Fm0m) J .m WEDGE 55 other and-gigglggrga. The Appalachian oak forest is re- stricted to the Allegheny River valley and to several small areas in the upland north of the Salamanca reentrant and adjacent to the Genesee River. It occurs more widely in the region of the Susquehanna River drainage immediately to the east of the area being treated in this study. Kfichler has drawn heavily on the map of major forest types in Armstrong and Bjorkbom's study (1956) of the timber resources of New York State. Although the boundaries of the units being mapped are essentially the same in both publi- cations, the units themselves differ somewhat. This is a re- sult of two different approaches used in the preparation of the maps. In one case the potential natural vegetation, or "the vegetation that would exist today if man were removed from the scene and if the resulting succession were tele- scoped into a Single moment" (Kachler, 1964, p. 2), is mapped, while in the other, the actual or "real" vegetation de- termined by a survey of existing forests carried out during the period 1949-1952 is represented. Armstrong and Bjorkbom's work does tell us what the general pattern and composition of existing forest vegetation is and for this reason a brief discussion of their units and those of Kfichler follows. The area of currently existing forests in western New YOrk is given in Table 4. The Nerthern hardwoods forest mentioned earlier is dfletributionally equivalent to the Maple-beech-birch forest 56 .mnqu Home: .oow>uom umouom .m.D .GOaumum ucofiflnomxm .ocma umoHOM ©o>uomou pan o>auosoonm pom m>auos©oumlcoc mooSHOCH d .oooa mOEHB m .mouom mo mocmmsonu SH N .ma:m>a>mccom umouom.cuoommo£uuoz :.hmmallxuow 3oz .muwcs pom mmaucsoo an moflumfiumum >m>nsm umouom wumcwaflaoum= .Eoum momma e.mH k.me m.o N.H m.mk m.~o~ e mmm mammauo m.m~ «.mm e.o m.H m.o> H.s- e.o~m mmmmcme o.ka o.mm o.o H.o o.mm o.mm~ a.aem mummmaz 4.0m m.eaa k.a e.e m.ke m.mmm s.~mm usages: m.om k.mo~ v.0 m.~ e.me e.mee k.eke were m.oe m.moe .e.o m.~ s.mm e.mm~ k.Oke scmmmaae k.em o.see ~.k m.ae ~.mm H.6mm 4.4mm msmsmnmuumo a.om e.eem m.o m.a e.me o.mem ~.ame msvamusmau ezmummm menace ezmommm mammoe memes azmommm menace mmeezsoo quommzzoo eanommzzoonoz nzeq umeoe Ir, «use name summon . «mac nzea ammmom.zoz n (I illlllll N in - i i iLnlllllliulg: HmmHBZDOU Hm fiBflBm XMON.3HZ zmmamflz ho dfldd 9244 BmmmOh Dad 4mm¢ 924A BmHMOhJZOZ «flfimd A4909 w wAmH>mccom .mnumn Home: .moa>uom umouom .m.D .SOaumum ucofiauomxm ummuom cumummonuuoz :.homaluxuofi 3oz .muacs 0cm mowucsoo an moaumaumum >o>usm uwouom humcwawaoum: .EOHH mama a e.m e.ae ~.He «.4 e.m e.a mammauo m.e m.me m.oe m.m m.e s.a momocme o.e ~.ae e.me s.e m.e m.a mummmaz ~.m e.ae e.mm s.m m.m m.~ maesosz k.m e.ae a.oe m.m ~.m e.~ mane m.k m.em m.ka o.m s.m H.e memmmaam o.m m.em m.oa a.m m.e m.e mememumuumo m.s H.em k.ea m.m m.e ~.e msvamuemao 1mmwmm umommmmmmmez. mmmmmsmwm mxeo mnoozuamom memes ommmwMHmz sazooo Ahmad .maflam MmOM_3fiZ ho mmHBZDOU 2mm8mfl3 Bmefi 2H NQMB BmflMOh Mm QZQQ Banach AflHUMHZSOU 4(908 ho Bzmummm m NAm4B 60 latter forest region extends westward from maritime Canada and northern New England across southern Ontario and Quebec to western Minnesota and includes seven subdivisions, each of which is characterized by forests of somewhat different composition which occur in distinct parts of the total region. The Mixed mesophytic and the Oak-chestnut forest regions, as they are mapped by Miss Braun, closely approach western New York. The former extends from the Allegheny and .Cumberland plateaus northward along the Allegheny River to southern Cattaraugus County, while the latter occurs across the east flank of the Appalachians to central Pennsylvania and northward to southern New England, with an extension up the Hudson River Valley. Forests.g§.5h§_§£i§-Ontario Lowland The forests of the Beech-maple region in northwestern New York are imperfectly known. Although as its name im- plies,.§§gg§ grandifolia and Acer saccharum are the pre— dominant forest trees throughout the whole forest region, many other species are present, and in the Erie-Ontario Low- land, oaks and hickories are particularly abundant. This suggests that the beech-maple area in western New York may not be solely an eastward extension of the deciduous forest Of the midwest but that it may also have an affinity to the (Dakrchestnut region of the eastern United States. This :I ha' i -.| Cu.‘ ‘A '- 61 relationship has been emphasized by Bray (1915). It is also clearly depicted by Shantz and Zon (1924) who, in their treatment of the natural vegetation of the United States, map what they call Chestnut-chestnut oak-yellow poplar forest throughout the lowland areas adjacent to Lakes Erie and Ontario, in a wide band on either Side of the Hudson River up to about Glens Falls and in most of the larger river valleys in the southern part of the Allegheny Upland. They represent these areas as northern extensions of oak forests of the same type which occur in broad areas on both sides of the Appalachians. Deciduous Forest Region of Southern Ontario AS the deciduous forests of southern OntariO'and northwesterna New York are placed in the same forest region, the communities which comprise both display many Similarities. .For this reason, Rowe's description (1959) of the deciduous forest region in canada, which he designates the Niagara Section, generally applies to northwestern New York State and provides a capsule summary of the forests in these areas: The forest communities are dominated by broad- leaved trees. The characteristic association . consists primarily of beech (Fagus grandifolia) and sugar maple (Acer saccharum), together with bass— wood (Tilia americana), red maple (Acer rubrum), red oak (Q_ercus rubra), white oak (Q.a1ba) and bur oak (Q. macrocarpa). Also within this area is found the main distributiOn in Canada of black walnut (Juglans nigra), sycamore (Platanus occidentalis), swamp white oak (Quercus bicolor) and shagbark hickory (Carya ovata) with the more widely distributed 62 butternut (Juglans cinerea), bitternut hickory (Carya cordiformis), rock elm (Ulmus thomasii), Silver maple (Acer saccharinum) and blue beech (Carpinus carolini- _g§§ var. virginiana). Other species with a sporadic occurrence as scattered individuals or groups, either on specialized sites or within the characteristic forest types of the Section, are the following: tulip tree (Liriodendron tulipifera), black cherry (Prunus serotina), mockernut and pignut hickories Car a tomentosa,._. glabra a), chinquapin oak (Quercus muehlenbergii), chestnut oak (Q. prinus), pin oak (Q. palustris s), scarlet and black oaks (9. coccinea, _Q. velutina), black gum (Nyssa sylvatica), blue ash (Fraxinus guadranqulata), cucumber tree (Magnolia acuminata) [and] papaw (Asimina_triloba). . . . There is . . . a poor representation of needle-leaved .Species, though eastern hemlock (Tsuqa canadensis) is sometimes scattered through upland forests, [and] white pine (Pinus strobus) occurs locally in small stands on coarse-textured soils . . . (p. 43-44). .Maycock (1963) has made a detailed study of the de- ciduous forests of the Niagara Section of Ontario. By sampling a large number of woodlots and relating his findings to the water-retaining capacity of the soils upon which a particular stand was located” he has demonstrated a gradual shift in the importance of the tree species which comprise the forests of this area along the moisture gradient. He as- serts that, rather than being found in distinct natural groupings or associations, the Species behave individualisti- cally and display independent relationships to the moisture gradient. For nearly all of the tree Species encountered in his study, Maycock has calculated an "index of regional im- portance" which serves well to summarize the overall im- portance of»a particular species in this region. An index 63 is computed by adding together the average importance values for a species in each of the five sections of the moisture gradient and then dividing this figure by the grand sum of the importance values for all species in all of the five parts of the gradient. The sum of the indices will thus total 100. Of the 56 species treated by Maycock, the follow- ing have the greatest influence in the deciduous forests of Ontario: Acer saccharum (13.4), Ulmus americana (11.3), .ggggg grandifolia (10.1), Fraxinus_americana (6.0),.5235 rubrum (5.7), guercus alba (5.0), _g. _r_ub_r_§ (4.4), 9. velutina (4.4), Age; gaccharinum (3.2) and_Tilia.§merican§ (2.8). To- gether these comprise 66.3 percent of the total regional influence. Deciduous Forests of Northwestern New York The only available detailed study of the vegetation of the deciduous forest region in western New Yerk deals specificallvaith Monroe County (Shanks, 1966; see Fig. 1). -However, its findings in general apply to other parts of the Erie-Ontario Lowland. Although recently published, it was written.before 1943, summarizing field work carried out from 1938-1940, and thus it reflects outlooks and methods differ- ing from those of Maycock's study. Analysis of notes made by the first land surveyors, in conjunction with study of ex- isting woodlots and forest remnants, permitted Shanks to prepare a map of the original vegetation of Monroe County. 64 Measurements by planimetry of the areas covered by the recognized vegetation types showed that the Beech-sugar maple type accounted for 61 percent of the original vege- tation cover. In order of decreasing areas the remaining types were Hemlock-hardwoods (12 percent), Upland oaks and Oak—hickory (11 percent), swamp forest (6 percent), Oak- chestnut-pine (4 percent), Mixed mesophytic (2 percent) and bog forest (2 percent). The remaining 2 percent was occupied by either marsh land or oak Openings. Today in contrast only 16 percent of the total area is forested (NOrtheastern.Forest Experiment Station, 1967). In Monroe County, the Oak-chestnut-pine type occurred: on the driest sites underlain mostly by sandy deltaic sedi- ments deposited in glacial lakes. An exact equivalent probably did not occur westward through Genesee, Orleans and Niagara Counties, for, although the three dominant oaks, Quercus.glb§,_g._£gb£§ and_Q. velutina, have a wide distri- bution in this area, as does one of the dominant pines, _gigg§ strobus, g. rigida is at the present time native no farther west than the vicinity of Rochester. There are no historical records of its occurrence west of this area. Castaneg dentata, originally a member of this forest type has been removed as a forest canopy dominant by the chestnut blight. Several other types of oak forest grew at Slightly nmue mesophytic sites. The Upland oak type, in which Quercus 65 .glb§,_g..£ub£§ and-Q. velutina were the usual dominants, occurred on the tops and Sides of drumlins and kames and on dry, flat-lying, gravelly soils of high porosity. Deficient soil moisture during the growing season probably character- ized such sites. Carya_9yat§ was frequently a codominant in such stands and at certain locations additional Species of _Q§£y§ attained dominant rank along with the oaks, resulting in the Oak-hickory type. Transitional Oak-sugar maple as- sociations occurred at favorable locations between lowland Beech-sugar maple and Upland oak types. Other transitional types called Mixed mesophytic forests occupied positions be- tween Oak-chestnut-pine and Beech-sugar maple types, between Oak—chestnut—pine and Hemlock-northern hardwoods types and in some cases in conjunction with Upland oak forests. Such transitional forests which are generally characterized by a large number of tree species occurring in about equal abundance rather than just a few dominant species were of limited occurrence in Monroe County. Originally Beech-sugar maple forest occupied more than half of Monroe County where it occurred on a wide variety of soil types. At the better Sites in the region such forest tends to maintain itself and some data are avail- able which suggest that it succeeds adjacent less mesophytic forest types. Typical beech-sugar maple stands exhibit abundant regeneration of the dominant species. Sugar maple seedlings often form a continuous undergrowth and beech is M '1 Ir: 9... 66 abundantly represented by rootsprouts. Tilia americana is often codominant in Beech-sugar maple forests and other com- mon associates include Ulmus americana, Fraxinus americana, Ostrya virginiana, Acer rubrum, A. nigrum,_guercus rubra, _Qg£y§ ovata, Prunus serotina and Liriodendron tulipifera. Areas of Hemlock-northern hardwoods forest were originally found in the northeastern and northwestern corners of the county, mostly on the Lake Iroquois plain, and in sheltered ravines in the vicinity of the Genesee River. Stands of this forest were dominated by_Tsuga cana- densis, Fegus qrandifolia and Acer saccharum which in many places occurred with Betula alleqhaniensis, Tilia americana, ,5235 rubrum, Fraxinus amerigan§,_guergg§ rubra, Ostrya virginiana, Prunus serotina and Ulmus americana. Hemlock— northern hardwoods forests have a dense canopy and light in- tensity on the forest floor is typically very low. Shrubs and herbs often listed as characteristic of the Hemlock-white pine-northern hardwood Forest region were present in some. remnant stands but were absent from others. Those with the greatest frequency include: Acer pensylvaniCum, ,Aster acuminatus, Dryopterig spinglgsa var. intermedia, Lonicera canadensis, Lycopgdium lucidulum, Maiagthemum canadense, Sambucus pubens and-Taxus_gan§d§n§i§. Large swamp forests were widely distributed in Monroe County and probably occupied about 6 percent of the total area. Deficient soil aeration is the most important factor 67 preventing invasion of the Swamp habitats by more mesophytic species. In order of decreasing abundance the following species occur in various combinations as dominants in differ— ent phases of the swamp forest: .glmus americana,_Acer rubrum, .5. saccharinum, Tilia americana, Fraxinus americana, Quercus bicolor, Fraxinus pennsylvgnica and Fraxinus nigra. The Dutch elm disease has made serious inroads on Ulmus populations. The former role of_g. americana as a dominant or subdominant has ceased to exist over wide areas. Forests'2£.§hg.glleqheny,Upland Considering now the forest vegetation of the upland south of the Portage escarpment, Braun (1950), as already noted, has mapped this area as part of the Hemlock-white pine-northern hardwoods Forest region. More specifically, She assigns nearly all of southern New York and northern Pennsylvania to the Allegheny Section of the Northern Appa- lachian Division. This and the Great Lakes-St. Lawrence Division are the two major parts of the forest region as it is treated by Miss Braun. The Northern Appalachian Division, including two other sections in addition to the one already mentioned, occupies much of the northeastern United States and extends into maritime Canada south of the Gulf of St. Lawrence. It differs from the Great Lakes-St. Lawrence Di- vision, which stretches westward from the St. Lawrence River valley to western Minnesota, in the presence of_gicea_£ubens 68 at higher elevations throughout the northeastern mountain region, in the absence of Pinus banksiana and the rarity of _g. reSinosa, in.the admixture of_Liriodendron tulipifera and ubgnolia acuminatg and other Species characteristic of the central deciduous forest and in the presence of Aster acuminatus, Tiarella cordifolia and Viburnum alnifglium and some shrubs and herbs which are rare or absent in the western divieion. The Hemlock-white pine-northern hardwoods region in- cludes the Birch-beech-maple-hemlock (northeastern hardwoods) forest of Shantz and Zon (1924): the Beech-birch-maple forest type as it is recognized in Pennsylvania (Illick and Frontz, 1928); the Lake forest of weaver and Clements (1938); the Maple-beech-birch forest type of Armstrong and Bjorkbom (1956) as it is applied by them to New York State; the Great Lakes-St. Lawrence forest region of Canada (Rowe, 1959); the Northern hardwoods region as identified in south-central New York State and north-central Pennsylvania (Goodlett and . Lyford, 1963): the Northern hardwoods, the Northern hardwoods-fir, the Great Lakes pine, the Great Lakes and Northeastern spruce-fir and the Conifer bog forests of Kfichler (1964); and the Beech-birch-maple and White pine-hemlock- hardwood forest regions as applied throughout the north- eastern United States by Lull (1968). For further equivalents and a review of the literature pertaining to the recognition \u-J *— 69 of the Hemlock-white-pine northern hardwoods Forest region, the reader is referred to Nichols (1934). .The original forest cover of upland southwestern New York has been greatly modified by lumbering and by clearing for agricultural purposes, The upland counties have at the present time the largest areas of commercial and noncommercial forest land of any at the west end of the state because of the poorer agricultural potential of this area (gee Table 4). Here, as elsewhere in western New York, an important stimulus for forest clearance during the early period of settlement was the demand for the ashes which remained after burning the cut trees. Crude field ashes were worth 4 to 9 cents a bushel and, if the settler wished to refine these somewhat, 600 bushels could be leached and boiled down into a ton of pot or pearl ash (also called black salts) worth 125 to 150 dollars (Munro, 1804; Young, 1875). Lye manufactured in this manner was used to make soap. The nature and composition of the original forest is indicated by several virgin tracts which have been preserved in northwestern Pennsylvania. These include Cook Forest (Morey, 1936), the East Tionesta Creek Natural Area (Hough, 1936a) and Hearts Content (Lutz, 1930b) which are all within the Allegheny National Forest. These and additional studies in northwestern Pennsylvania have been summarized by Hough and Forbes (1943). 70 Judging from early land survey records for a 175,000 acre tract in northwestern Pennsylvania (Lutz, 1930a), the forest existing today along the East Tionesta Creek is fairly typical of that which originally covered dissected areas of the Allegheny Plateau, particularly Nefacing Slopes. .EEEES canadensis and-gagus grandifolia in both abundance and fre- quency values are the dominant canopy trees on the plateau tops and on the middle and lower SlOpes (Hough, 1936a). Betula alleqhaniensig is third in order on middle and lower slopes, but on the plateau top Age; saccharum holds this rank. In order of decreasing totals of abundance and fre- quency values associated species are: Betula_alleghaniensis, ’gggr rubrum, Prunus serotina, Fraxinus americana,_Li£ig— dendron tulipifera, Magnolia acuminata and Tilia americana. Viburnum alnifolium is the most abundant Shrub in this forest and common herbaceous plants include pryopteris Spinulosa, Lycopgdium lucidulum, Maignthemum canadense,_Mitchell§ repgns, Oxalis acetosella and Tiarella cordifolia. Within forests of this type there is a tendency toward segregation of hemlock-beech, beech-hemlock—sugar maple and beech-sugar maple communities which differ from one another in the relative abundance of the dominants. Hough (1936a) and.Morey (1936) have suggested that there is an alternation in the occupation of a given spot by hardwoods and hemlock- hardwoods.. Uprooting or death of the hemlocks permits the hardwoods in the understory to become established as the 71 canopy trees, while removal of the hardwoods, either catas— trophically or by aging, releases the hemlocks in the under- growth so that they eventually grow to occupy the canopy again. In Cattaraugus County forest of the Beech-sugar maple type originally occupied the better drained soils near ridge tOps and was apparently more extensive in the glaci- ated portion of the plateau (Gordon, 1940). However, most of it in this region today is of secondary origin, having developed after the removal of hemlock and white pine for lumber. .gaqgg grandifolia and Acer_§accharum comprise 97 percent of the canopy in an undisturbed beech-sugar maple stand on a northeast slope in the Big Basin at Allegany State Park (Braun, 1950, Table 82). White pine has an interesting position in the virgin forests of northwestern Pennsylvania. It is absent from the East Tionesta tract, but at Hearts Content, 30 miles to the west, it is abundant both in the hemlock and in the hemlbck— beech communities. An age analysis Showed that the pine started there as an even-aged stand at about 1680 (Lutz, 1930b). Similar data gathered at other localities on the Allegheny Plateau suggest that the presence of white pine stands can nearly always be correlated with fire, windfall or some other event that opens a portion of the forest for seeding (Hough and Forbes, 1943). If no openings are made, the white pine apparently matures, dies and is replaced by tmmlocks or hardwoods but not by other white pines. 72 Forest communities essentially the same as those in the Allegheny National Forest have been preserved at a few places in southwestern New York State, both outside and in— side the glacial boundary. Gordon's map (1940) of the vege- tation of Cattaraugus County at the time of settlement, which was prepared by analyzing the original lot survey data in conjunction with an examination of existing stands, Shows that the prevailing forest type in this area belonged to the Hemlock-white pine-northern hardwoods forest. Hemlock- northern hardwoods communities with little or no white pine comprised the typical stand. Quantitative data are un- fortunately sparse but the virgin tract of forest in Stoddard Hollow in the Big Basin at Allegany State Park has "a composition almost exactly Similar to the Hemlock-Beech association at Heart's Content" (Gordon, 1937, p. 39). The principal trees on lower Slopes in the Big Basin in order of decreasing abundance are-ggqg§_g£gndifglia, Tsuga canadensis, ,Acer sacchgrgm and A. rubrum (Braun, 1950, Table 82). To- gether they total 87 percent of the canopy. The leading dominants in the Hemlock-beech association at Hearts Content are-ggggg canadensis,.§aqg§ grandifolia,_Age£ rubrum,_ging§ strobus and.ggstangg dentata in order of decreasing totals of the relative frequency and density for each of the Species, recalculated in part from the data provided by Lutz (1930b). A number of other forest types are also recognized in Cattaraugus County (Gordon, 1940). AS in Monroe County 73 the drier sites, which in the upland occur on ridges and ex- posed south and southwest slopes, were occupied by forest of the oak—chestnut type. _Qgercus alba, 9°.EEEEE: Q. prinus and .9. velutina are now typically the dominant trees at these sites, but prior to about 1934, before being eliminated by the chestnut blight, Castanea dentata was codominant with them. Associated species which sometimes reach dominant status include Pinus strobus, Acer rubrum, Carya glabra, Betula alleqhaniensis, Populus tremuloides and occasionally others. Similar communities also occur on dry S—facing Slopes throughout northwestern Pennsylvania (Hough, 1936a). In Cattaraugus County secondary forests rich in oak species commonly result after fires and excessive logging. To the east just beyond Allegany County in the region straddling the New York-Pennsylvania border, Goodlett and Lyford (1963) have mapped the current extent of oak forest using Quercus .albg as an indicator species. Species with frequencies greater than 50 percent in oak forests studied by these workers include Quercus rubra,_g. alba, Acer rubrgm,.gigu§ strobus,Quercus velutina and_Q. prinus. Such forests oc- cupy a greater area in this region than they do in western New York and Pennsylvania. Communities transitional between the Oak-chestnut type of dry slopes and ridges and the Beech—sugar maple and Hemlock-beech types of the lower, more mesophytic sites are considered to be somewhat-attenuated examples of the Mixed 74 mesophytic forests which occurs across much of the southern part of the unglaciated Allegheny-Plateau. Similar com- munities have been recognized at places in the Erie-Ontario Lowland associated with Upland oak forest (Shanks, 1966). .Mixed mesophytic forests in Cattaraugus County occupy moist well-drained and well-aerated sites which are favorable to the growth of a wide variety of tree species. Such forests are dominated by Quercus-Egb£§,.§§qg§ grandifolia,.Ager rubrum, Betula alleqhaniensis, Fraxinus americana, Prunus serotina,,and, formerly,_gastanea dentata. Magnolia acuminata, Quercus alba, Liriodendron tulipifera, Pinus strObus,.Tili§ americana, Qagya cordiformis,_gg§; saccharum, Ostryg virginiana and Acer pensylvanicum also occur but less frequently. Similar communities occur today in other areas of the upland. For example, in a stand near Lily Dale in Chautauqua County (Braun, 1950, Table 85), the following trees comprised the canopy (in percent): Tsuga_ganadensis (20.9), ggqgg grandifolia (16.9),_Prunus serotina (12.4), Acer rubrum (11.3),_Acer saccharum (8.5), Pinus strobus (8.5), Magnolia_acuminata (5.7), Quercus rubra (5.6) Fraxinus americana (4.5), Betula alleghaniensis (2.8), Tilia americana (1.7) and Carya ovata (1.1). Acer_§accharum accounted for 37 percent of the second layer trees suggesting greater dominance by this species at the site in the future. 75 Two additional forest types restricted to bottomlands have been recognized in Cattaraugus County (Gordon, 1940). The White pineeAmerican elm forest occupied flood plains of the major rivers and streams in the county, especially those which were filled with impervious lacustrine sediments. The great value of_gigu§ strobus as a timber tree led to the early destruction of these forests, but their former distri- bution has been well-documented (ibig.). The largest area of this forest, occupying many hundreds of acres, occurred along the axis of the preglacial Allegheny River whose course ran toward Lake Erie from the northwest side of the Salamanca reentrant along what is today Connewango Creek (egg Muller, 1963 for further details). These forests also contained Fraxinus.nig£§, Quercus bigolor Acer_£ubrum, Betula alleqhaniensis,-Egggg canadensis and occasionally_§big§ balsamea and.Lg£ix laricina. Of less widespread occurrence are the Bottomland hardwood forests which are found on recently deposited alluvium, especially along the Allegheny River and Cattarau- gus Creek and their tributaries. In composition these forests are variable, but they are by no means as rich in ”numbers of species as are the bottomland forests of Ohio and southeranichigan. Populus_de1toides, Sali§_nigra, Acer negundo and A. rubrum are frequent along disturbed stream courses, while_Platanus_9ccidentalis and Juglans cinerea are typically found on the more stabilized flood plains. 76 Stands Sampled by_£hg_guarter Method A good understanding is needed of the distribution and composition of existing forests in the upland around the sites where cores for palynological analysis were collected. During the summers of 1966 and 1967 I studied about 35 wood- lots scattered across southwestern New York State from western Chautauqua County to central Steuben County. Notes were taken on nearly all of the stands, but three of them were studied more carefully by the point quarter method (Cottam and Curtis, 1956). Briefly, this technique involves taking measurements at a number of sampling points Spread throughout a woodlot. I used a grid of 48 points (eight points by six points) in which the points along a traverse line were 20 m apart. The initial point was chosen randomly. Within a meter-square quadrat centered over each of the 48 points, the presence of seedlings1 over and under 30 cm in height and the presence of herbs were noted and tallied by species. The area around a point was then divided into four quarters using the transect line as a bisect and an— other line which passed through the point at right angles to the bisect. Bamboo wands were used to temporarily mark these lines. For each of the quarters the distance between 1The size classes used follow Curtis (1959): trees-- greater than 4 inches (ca. 10 cm) diameter at breast height (d.b.h.), saplings--between 1 inch (2.5 cm) and 4 inches d.b.h. and seedlings--less than 1 inch d.b.h. 77 the point and the nearest tree, its Species and diameter at breast height were determined and recorded. Each of the four trees generally occurred well beyond the original 1 m quadrat. The number and species of saplings which occurred within an area 1 m on each side of a line between the points were also recorded. (The three stands selected for sampling were chosen because they met the following criteria: (1) greater than 15 acres in Size permitting the influence of surrounding fields and secondary forests to be reduced, (2) absence of disturbance in the form of fire, grazing or excessive cutting (none or very little during the past 40 years) and (3) oc- currence on upland soil types. After the information was col- lected, it was divided into four equal parts and a Chi-square test of homogeneity was applied to determine if the number of major tree species within any segment deviated Signifi- cantly from the number expected on the basis of uniform distribution (Curtis and McIntosh, 1951). In no case did the Chi-square values exceed the expected values at the 5 ~percent level indicating that the stands were homogeneous according to this test. The location of the stands in relation to the sites selected for palynological study is indicated in Figure 2. Other pertinent data concerning the sites are summarized in Table 5. Relative frequency (percent frequency), density (percent occurrence) and dominance, (percent basal area), and 78 the sum of these three figures, the Importance Value, in addition to the absolute density (no.treeS/acre) and domi- nance (basal area/acre) were calculated for each of the tree and sapling species tallied during the sampling (Curtis, 1956). The relative frequency of seedling and herb species was also computed. These figures are given in Tables 6 through 11. The stands sampled, although too few in number to permit a complete assessment of the variability in the existing forests of southwestern New York, nevertheless do provide quantitative data on the composition of several up- land communities. Acer saccharum,.§§gus grandifolia and 2333; canadensis in order of decreasing importance values are the leading dominants in all three stands. The simi- larly high importance values for sugar maple, beech and hem- lock saplings suggest continued dominance by these Species at the sit‘es,,although hemlock seedlings were infrequent with- in the sampled areas. Quantitative data for comparison are unfortunately not available, but the three stands clearly seem to belong to the Hemlock-white pine-northern hardwoods forest (Nichols, 1935). This unit is recognized by Gordon (1937, 1940) to be the climatic climax of the entire upland in southwesternNew York State where it is expressed by as- sOCiations in which Tsuga canadensis occurs by itself or mixed with 3319?. wdifolia, Acer gccharum and Betula illeghaniensis. Although present in all three of the stands 79 .AmmmH .Hosomv GOHumHUOmmm MHmSHo>IcH Hmzlcumm on» on mchcoHon omosu ou HMHHEHm auo> mHHOm mchocoo noHn3 coHumHUOmme mHHmlouow can ecu mo mHHOm an oouo>oo mono or» m3ocm mucsoo msdsmusmno How one coHuMHOOmmm HHom ucooou one .ocmum ecu sundown >MH momma HHom o3u was no SOHSB ocHEnouoo on oHnHmmOQEH mH 9H SOmmoH anu Mom .monm one CH ucoEGUSWHHm Emouum pew moHHoH uconMMHo umczoeom 302m meme ucoomu woos mm ocmum ecu mo muHcHOH> on» cH c3muomHE mHucmnmmmm mH poocHum mm3 one HHOm chu noHr3 so mean UHcmmumomou one H oQOHm mcHomMJzz oHusom anamumoaou OHSHMHOE xmoS .msHHHou AmemH .umHHszc moses IHoom ocHHuwSUMH nmcHhHIHMHm aHHmo: .mcHHHou >Hucom lemma HHo3 ..Hm om MOHmmBV EmoH oomH omsmse ho uHHm Auoumoozv comm «at muumnoum HHm> neon mwz lemma ..Hm no Hoasmec HHm3 EmOH aHHo>mHm oomH mesh Auoumoo3v comm HH#.c0HumocmHm mussoo oHuu AmHmH ..Hm um mHHooa um£3 SMmHuuozv EmoH uHHw Imeom Ho .HHe3 Anoumoozv room no 1 .._ EMOH uHHm mHmSHo> oomH HonEoumom xooHO amzmomcmu mammnmomoa mmmHO ommchun mama HHom oonsmm open . osmz venom I ll l‘ LII ZOHB¢Z&OEZH Q4MWZNU «49¢Q QZflBm BmfiMOh m mflmde .ucosuummon coHum>uomcoo oumum xuow 3oz 80 H we Tm m.~ See and $44 a amalfleme MHHouchouHm mssuoo 6.6 TN m.~ see sad e34 3533 MGMHSHmHH> mNHumo m.~m me Wm .85 e8; maze w regime HmcoHdmcmoHHm «Haven 9mm 93 93 slim sfim seam Nocaanmm memoHHoEm mscmeuh «.3 To mé see some 2&4 mmmua mundwfidom mwdoflmmz 9: To mi See sad SN; mamas mcHuonom mscsum «.8 TS new .86 seam see 353mm 935 To To... slam use 5.: 38a. mflmGOUMGMo MMSQB m. m2 a . 3 T? as . 3 SN. 3 x6. 3 mecfimmm m.~mm.m 83 Tom x113 made soda moose MHHOMHoeon human week 0.3... To: 86.? x92 mode 3533 e. gm .3 e. on N. mam Sade so. we mm. mm moans ESHMSOUMm HOUC 9MQ¢\MDZ¢ZHZO9 NMQ¢\HBHmZm9 HDA<> NDZ¢ZHSO9 MBHmZN9 MDZHDOHMN mmflHU mNHm HBDHOmmd flBDHOmmfl NDZ¢BMOQZH 9>HHHH¢HMM H>HH¢AHM 92¢ mflHUNmm K90H.392 «M92900 flbcaflanflmu H.aumzua usa¢> anaesnxoo seaweed spamscasu needs name mangoeme uaoaomea nozuaua¢qu¢ usuaaqus one ennoenm KBOfi Ill .HHIDOO 9H9” .HH* BOMB‘BZ‘AG SEMIHM¢AH9 #3999809 walndm¢w 92¢ wflflxfi H a fldn RC.N momwcH Nleoncs Re.o morocH NH uo>o mm.o mHHowHocsuou .> memoHHoEm MHHHB mwmmmmmmmm.mmmmm. mm.o mHmcoomcoo MHOH> x&.o nonocH NH Ho>o 8v.o mOpHonuHumouom EssoHumNHOm RC.N monocH NH Hones Km.o maHowHouoo MHHoumHB MHHoprcme msmmm xe.o ssomuemm.esHH>noooom Xm.o .mm monHm xo.N mHooEuoucH .Hm> x&.o muoOHHo>Hp XOHcm mmOHscHQm mHumuQO>H9 xw.o enough mHHmnouHE $v.o mHHomoqum mHmeo RO.N MHHNSNHG MHHmuco9 Km.o OEHOMHHHoAMHw .um> xe.o HcouamHo oNHnoosmo ESDNEMHmEoo SSHommooNq Re.N monucH NH Hopes mamoHuoem nsEHD xe.o measnmae magmas: Sm.o .am mamaummsa RV.N Escmwuuonou ESHcmuoo No.0 enactmcmo EsEozucmHmz x&.o MEMHchuH> mammemm x&.N ESHchmHHu msommHH< Xv.o .nm EsHOMNOHm Rm.o mHHOwHEAOUHm mscuoo .I $N.m ouHcmoocH .> *v.o onHHusom ouHqumm Rm.o muoOHSmHHcmsU oomOHHU $6.4 .Em maoa> R€.o ocHuonOHHmz mHuommamm xw.o wooHouuoHHmnu a“SHHKEQOHemU Rm.o monocH NH nope: Xv.o .dm muucmoHo xm.o .nm xoumo MCMOHHoEm nacHXNHh Xv.o wOLUCH NH Moos: xw.o mooawnooa oomu0¢ x&.o nonucH NH Ho>o mHENOMHpHOO m>umo I1, RH.OH monocH NH popes RN.H mcoomonsa EsumcouwHom Edundu Homfl x&.o mocwmeSMHQ xmumu SN.H sacmacamua> eyanu>uuom s~.aa nuance «a noose *e.o .mm ESHLONuuom meuouom sacrum XO.H monocH NH Hopes Re.o mooHoHuouoaHmnu .m. mcoHckuH> owuumo mm.m monueH NH Ho>o Rb.mH nonocH NH Hoods Xv.o mcHEmmuxHHmu ESHH>£u< xo.N meOHwHocmuu ESHHHHAB Esumnoomu H004 wuzmbommm wmdqo MNHw Mozmoommm mmdao mNHm Muszommh mmtqu MNHm m>HsHaflqu 92¢ mmHommm N>HBHQ ..um09 :oHuo>Homcoo .xuow 302 no oumum cosmom oUUmU * COflUflmflSwnv‘ a 04 TN 04 and x: and «933m MSNHSHHoumo mschHMU 2mm «.4 93 .886 warm x8e machine mquchuH> mxuuwo we... we ed xfio same .84 moose mHn:OHcosmOHHm mHsuom m.mmo m.H N.n XO.N m6.H *6.m moons mfimOfiHQEm mdcwxmuh meme etc. To xma SEN $04 $39 MEMOHHOEM mHHHB ENS 82 6.3 roam «.82 x83 «maiden Tame 0.3 Tam xfim $73 8.3 moose mHucooocoo omens ~63 5.: Emma 32.: $4.3 x82. «333m salsa 5.3 9:. sfimm $83 8.8 moons mHHOMHUSmum acumm s.ama e.mm a.os xe.k~ x~.- xm.o~ masseuse 453.3 92 ~63 $0.3 $83 «8.3 339 asumnousn uoo< secede—2280a mmuebFHmzma 88> mozezazoo €5sz genomes $48 fine menace? mesaomma "mo—2285 Seeded E935 E935 92 munoumm KMOM 392 «M92900 HHME H.311. awesome Ens, "EON wgaqmem 82 mamas 3.55 82.3 Summon HH 999(9 85 .pcmq mmD OHQHHHOZ .mumeOh can nosed mo :OHmH>H9 ..u&w9 SOHum>HochU .xuofi 302 no ouwum .0.m.0 .uumo # coHuHmHon< H «3.0 .3 $304 024 333.3» H93 as??? $0.0 morocH NH Hones XH.H mommxcomm moaned $6.0 mumooum NOHOHD Esunsu Hood $¢.H omsoomcmo EsEonucmHmz .0210 .nm .H eke .mm 33> Xv.H mum0HsmHHoosu monouHO $0.0 Esuuwuo ameHHua Xh.0 monocH NH Hopes Al osmoHHmEm useHD XH.N mHmsoomcmo .> fie.0 MHHOMHOHOO mHHoumHB I Xe.0 monoca NH Hopes Xm.N wowcm>Hchom .> 80.0 mHmcoomcmo mavenEmm MSNOHuoEm MHHHB Xm.N mHHowHocsuou oHoH> .086 .am 252 End 29:83 nuanuosmo Xm.N mQOHHusom MUHummom sate .nm 32.238 .0330 333:: 3032.02 *0.N mooHouuoHHmsu ESHHNEmoHsmO x&.0 ESHSQHUSH EsHoommumm $n.0 OCHHOQOHHOS mHuomaHQm xm.m mHOOEhoucH .Hm> $0.0 .Qm mcwHuomEH *n.0 MHHOMchouHm mscuou mmoHschm mHnoqu>H9 $v.0 Eamomsn 95Hu0umosm *n.0 EscMHchHH> acho>Huom mm.v ouHcmoocH MHOH> fiv.0 mCHmHm ammonHo $>.0 monocH NH uo>o $0.0 monocH NH Ho>o $0.0 morocH NH hopes 8m.N monocH NH Homes .510 .nm .m meauouwm museum 3339.30 memes $v.0 mmchmocmHm xonmo *H.H awcmHuuonou awacmuoo XH.N monucH NH Ho>o XN.0 mogocH NH Hops: x&.0 mcHEmwuxHHmm ESHuwrud XH.H .mm ESHHMU acmoHHoEm wsstmum Xv.0 msuoHMHMOOMH .m. $H.H mQOHHuocsm MHuoomumccoo RN.0H EMHHNSQHHO mEoomHMfl. Xv.0 msumoHum>Ho mound $6.0 monucH NH uo>o Xv.NH monueH NH uo>o *h.0 morocH NH some: X¢.MH monocH NH Hope: *&.0 Esumomm.E:ocmHnfl. mHEHomHouoo maumo Esumnoumm Hood Mozmbcmmh mm<90 MNHm 302990929 mm490 mNHm NU299092M mmdflo mNHm m>HB¢Hmm 924 mmHUmmm 9>HHHH¢HNM 92¢ mMHommm 290% 392 .MBZbOU HHMW H.NH# whmmmomm MNHA<> 9(0N m02HHQmmm 92¢ 09292 NH 94948 u¢B 1 are over-represented and those \Mith values about 1 are proportionately represented. I have calculated R values for nearly all of the tree pollen types which occurred in the sediments I studied fron1 southwestern New York State. At the present time, the Pollen Of different species of Betula, £93219: Quercus and 31% apparently cannot be identified except, in some cases, by tiJne consuming size-frequency measurements. Therefore, 90 only generic level R values can be obtained for these taxa. However, because other pollen types can be identified to the species level, R values for these apply to units which are Of greater usefulness in ecological interpretation. The other portion of the ratio, the abundance of given components of the vegetation, has been estimated in three ways, but all of these involve certain limitations including the absence of data corresponding to a number of pollen types recognized in the spectra and the accuracy of the sampling technique used to provide the information on vegetation composition. U.S. Forest Service Survey Statistics The NOrtheastern Forest Experiment Station (1967) has published the most recent source Of quantitative data on the present vegetation of southwestern New York State. This information was collected during a resurvey of the forest resources of the state that was undertaken to Obtain an esti- Inate of the total timber volume, total periodic tree growth .and other statistics of use principally to foresters and economic planners and to update the initial survey of the state completed in 1952 (Armstrong and Bjorkbom, 1956) . The sampling design used in both surveys is fully described by Bicflcford.gt_§l. (1953). The initial survey involved the use Of aerial photographs on which a large number of points were randomly selected for photo-interpretation. Each point was assisgned to a class based on the volume of timber occurring 91 on a one acre plot surrounding it, and subsequently a certain number of the plots from each of the classes was measured on the ground. The resurvey used the same sampling technique, but in addition some plots from the initial survey were re— measured and new ground plots were established. The most meaningful figures in the resurvey data for use in calculation of R values are those Of total volume by species on commercial forest land. To facilitate this compu- tation, raw data listing commercial timber species five inches in diameter and upwards in millions of cubic feet have been recalculated in percentages. The percent total volume figures for species recognized in the survey are listed by county in Table 13. The figure which pertains to a given species within the county where the bog was located was used in the R value calculation. The surface samples analyzed were collected at the exact sites where the sediment sampling was done. In each case the sample taken was from a fairly dense but actively «growing sphagnum polster and comprised the upper l to 2 cm from an area of about 10 sq cm. Two subsamples from each were macerated in the laboratory and their residues were uliaimately combined and counted together. Inasmuch aS about 50 percent of the total pollen in the surface spectra was conixributed by herbaceous plants, reflecting the large area °f EKanorest land in southwestern New York State (see Table 4), the counts were recalculated using the sum Of the arboreal €92 .chm>H>mccom .Houmoum Ho Hou656Ho CH monocH m mOHUOQm oouu HmHouoEEoo .600036906 0cm 60003066: Horuo mOQSHUcH m .Noumn Home: .moH>uom #66606 .m.D .coHumum ucoEHHomxm umouom cuoummonuuoz :.600Hunxuow 3oz .muHcs 6:6 weaucsoo an muHumHumum >o>usn u6660w NumcHEHHOHmz SH HH OHnt .Eouw 6669 N H 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6662 6.6 6.6 6.« m.« s.« 6.6 6.6 6.6 .666 no .66 mmHmmmM. 6.« 6.6 6.« «.6 6.« 6.6 6.6 6.6 mmmmmmmm..m. 6.6 6.6 «.6 6.6 6.6 6.6 H.« 6.« menouum 63:66 6.6 6.6 6.6 6.6 6.6 6.« 6.« H.« .666 mummm. 6.6 6.6 6.6 6.6 6.6 6.66 6.66 6.66 .666 mammmma. 6.66 6.66 «.66 6.66 6.66 6.66 6.«6 «.66 saunas 6666 6.6 6.66 6.6 6.6 6.6 6.« 6.6 6.« .666 no .66 6:266 N.6 0.m 6.6 m.m 0.m m.m o.m ¢.m wwwuouom magnum 0.6 0.6 6.0 m.m 0.0 0.m 0.6 m.m McNOHHoEm MHHHB 6.6 6.6 «.6 6.6 6.6 6.6 6.6 6.6 666666656 6666x666 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 66666 .m. 6.6 6.6 6.6 6.6 6.6 6.6 «.6 6.6 66666666666446.666666 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6666666666 mummm. 6.6 6.6 6.6 6.6 6.6 6.6 6.6 6.6 mHHmMHmmmmm.mmmmm. 6.66 6.66 6.66 6.66 6.66 6.6« 6.6« 6.6« 266666666 6666 6666666 6666. 6266666 6666266 6666662 6262663 6666 62666666 66666669666 6666666666 M92900 mafiem 220% 322 2H mmHBZDOO 2298mm3 BflUHN 2H N 92 ACBOB 8220229 mH mflm¢9 93 pollen as the percentage base. This is necessary because the forest composition percentages are based on total forest land, not on land of all classes. R values calculated from these two sets Of data are listed in Table 16. Original Lot Survey Data The original land survey of western New York was privately sponsored, but was similar in organization to the rectangular pattern used in the General Land Office Survey of public lands west of the Appalachian Mountains. This system of surveying was authorized by Congress in 1785, but prior to this time metes and bounds, the establishment Of property boundaries according to natural features, was the rule (Bourdo, 1956). In the rectangular survey of western New York a grid of north-south range and east-west township lines bounding townships generally Six miles square was sur- veyed west of a line extending north from central Allegany County through eastern Wyoming, Genesee and Orleans Counties, the area encompassed by the Holland Purchase (Evans, 1924; fPurner, 1850). The number of townships per range varied :Erom sixteen in the eastern ranges to three in far western CHlautauqua County between Lake Erie and the Pennsylvania border. These townships have only partly retained their identity as political units in contrast to those in the re- giiln.surveyed by the rectangular method to the west. 94 The laying out of township and range lines began under the direction of Joseph Ellicott in the Spring of 1798, Shortly after the Holland Land Company acquired clear title to western New York, and continued until 1800. The internal survey of townships into lots was largely completed by 1810 but a few of the townships in southern Cattaraugus County were not divided until 1819. The first 24 townships surveyed were subdivided into sixteen sections 120 chains (1.5 mi) square, and each section was then partitioned into twelve 60 chain (0.75 mi) by 20 chain (0.25 mi) lots containing 120 acres each. The remaining townships were divided into lots of 360 acres, 60 chains square. These in turn were Often subdivided into three parts of 120 acres each depending on the requirements of the prOSpective buyer. It was the custom in the Ellicott survey to mark each lot corner with a post. Records were kept of the direction, diameter and Species of from one to four bearing-trees lo- cated in different quarters surrounding the post. Since .1and sale was of paramount concern to Ellicott, he directed 1118 surveyors to take careful notes on the topography, soils, timber, windfalls, Springs and other natural features along the survey lines. When running a lot line the surveyors made a list of the predominant timber encountered and, if the forest composition varied along a lot line, several lists were: recorded. It is perhaps logical to assume that the spatties listed were arranged in order of decreasing abundance, 95 as they were in the General Land Office Surveys, but no evi— dence exists that this was the case. The richness of the data included in the Ellicott . survey is illustrated by the notes1 pertaining to lot 60, T.9 R.5 (southeastern Erie County). The boundary line of this lot begins at a beech post marking the northeast corner and passes westward across an upland of the first quality timbered with beech (Fagus grandifolia), hemlock (Eggga canadensis), bass (Tilia americana) and elm (Ulmus Sp.), to an upland of the second quality, then across a deep gully and abruptly back again to an upland of the first quality with hemlock, beech and sugar maple (Acer sacchgrum),timber, and finally to a beech post at the northwest lot corner. Southward from this point the line passes through land con- sidered excellent for meadow and timbered with sugar maple, cherry (gggnus serotina), bass and elm, to a sugar maple post at the southwest corner of the lot. A short distance to the east of the post the line crosses an intervale of the first 6 woaumHumum >w>usm umquu ozo “wumnHounm macHXMNm mOOSHoaHo “2666:0066 6606 £UH3 oomEsHu “mHmcwaem: OHHm .m nuH3 ponESH NLEOH 6Hsuwm mo mommucwuuon >o>usm umouom 16660 :OHumuomo> uSoEoHuuou-oun :H anemone wane A m--- m--- will mnun mutu m-l- 0N.0 m-I- m-II manuaoo wocmuumo mu-I mn-I m-I- 00.H mull m-I- mull 60.0 OI-I mouseHo mcmHmsh mu..- HH.N m-I- m-II mun: mu..- MIII MN.0 Olli mNHu60\m=:HmHmU m-- m-- 66.6 m--- m-- 66.6 m-- m-- 66.6 .666 6666666 66.6 m--- 6-- m-- m-- 6-- 66.6 m-- 6-- 6666666-6 66.6 666.6 66.6 666.6 m-- 66.6 6«.6 66.6 ««.6 wumnHoo-w wSCMxmuh 66.6 6-- 66.6 «6.6 6--- 66.6 66.6 6-- 66.6 666666626 66669 66.6 66.6 66.6 66.6 66.6 6«.H 66.6 m-- 66.« .666 65566 m--- m-- 6«.6 6-- m-- 6«.6 m-- 6-- 66.6 .666 «News 6«.c o... 66.6 o--- u-- 66.6 66.6 m--- ««.6 55666666666 I sunsu .1 66.6 66.6 «6.6 6«.o «6.6 66.6 6«.6 66.6 66.6 566666666 6666 m-- m--- 66.6 m-- m--- 66.6 66.«6 m-- 66.6 .666 6566660 «6.6 «6.6 66.6 66.6 66.6 66.6 66.6 ««.6 66.6 66606666666 6:666 66.« 66.« 666.6 6«.6 66.6 66«.66 66.6 66.6 6«6.6 .666 665666 6«.6 66.« 66.6 66.6 66.« 66.« «6.6 66.6« 66.6 6666666666 66:66 66.66 m-- 66.6 66.66 66.6 66.6 66.66 m-- 66.6 .666 6:666 :oHuawE 606 X :oHusoE 6mm X :oHucoE 606 R mxmu :oHHOA wo .vouh ovumuuomEH mo .wmum oocmuuomEH mo .Uoum ovumuuomEH 66660: 66600: :uooo: UEOEOHuuwm-oum 6:6EOHuuwm-wum ”a: OEOHHUOD log 606H6> & mom coHuomuoum mom cobrmsom mom muonsoHH¢ wH 99918 20HBHm09200 20HH¢BHGH> ho mfiaflxHBmw mDOHm¢> U2Hwb AWHZADQHCU 99994> 2 104 The best example of the latter group is_ginu§ Spp. whose values range from about 1 to 5 using the modern data and from 7 to 57 using the pre-settlement data. These figures, particularly the ones derived from the pre-settlement data, emphasize pine pollen's typical over-representation. One difficulty in obtaining truly accurate R values for pine from surface sample pollen percentages, however, is that both native and exotic species now grow and contribute to the pollen rain in western New York, and thus the pollen dis- persal capacity of pines of both types influences surface counts. Since only native species are:represented in pre- settlement spectra, the R values obtained from such data can— not be applied indiscriminately throughout postglacial time.- Another problem is that locally growing reforested pine stands may contribute more pollen to surface spectra than is typical of the region from which the vegetation composition data were derived. The R value of 5.0 at Protection bog probably reflects the influence of several nearby mature, planted ginus resinosa and g. strobus stands. That local over-representation is indeed operative at this site is sub- stantiated by the sharp decrease in_gigu§ Diploxylon and_g. Haploxylon percentages in the upper 5 cm of sediments. Pollen below this depth accumulated before the plantations existed. .R values for_glmg§ Spp. are also variable but to a lesser degree. They range from 0.5 to about 5.0, but most 105 of them are over one, implying over-representation. Simi- larly, over-representation is indicated for Quercus Spp. The R values for this pollen type calculated from the modern data suggest that the oaks are somewhat less than twice over- represented, while the only pre-settlement value implies that they are nearly thirteen times over-represented. Little emphasis should be given to this one pre-settlement value, however, because of the rather small area used to obtain the percentage of oaks in the vegetation. Since Quercus pollen seems to be transported easily for long distances, a much larger area should be sampled to gain an accurate estimation of the abundance of oak trees contributing to the pollen rain. In northern Vermont where Quercus rubra is the principal pollen producing oak, Davis and Goodlett (1960) have found oak pollen to be the most over-represented of all pollen types in their spectra. The R values for_g§gg§, with one exception, are fair- ly constant and imply that Tgugg pollen is somewhat over- represented. My figures compare fairly well with those calculated for hemlock in northern Vermont (ibig.) but differ from the correction factor for the species used in Wisconsin by Curtis (1959) who has multiplied numbers of hemlock pollen in fossil spectra by three to obtain pro- portional representation. Since R values seem to differ from one major geographic region to another (gee Comanor, r967), the use of one correction factor over a wide area may lead to incorrect results. 106 The R values for Betula spp., using both modern and pre-settlement data, indicate that it is greatly over— represented, a finding that is substantiated by the studies in Vermont and Wisconsin cited above. Although I cannot be certain which species of birch produced a given pollen grain, it is likely that only two species, Betula lenta and_§. alleghaniensis, have contributed most of the birch pollen to the spectra because the other species reported from western New Ybrk, g. papyrifera and g. pumila, occur either at the periphery or well beyond the source area of the bogs. According to my data only gagug grandifolia.and Juglans cinerea are about proportionately represented. To these may be added Carpinus/Ostrya if the average of the two R values in the table is taken to be meaningful. The pollen of Carpinus caroliniana and Ostrya Virginiana are morpho- logically very similar, so no attempt was made to tally them separately. Since the autecology of the two species differs and the abundance of Carpinus in the vegetation is unknown, undue significance cannot be given to the R values for this pollen type. Davis and Goodlett (1960) found the basal area percentage of Ostryg virginiana in the forest north of Brownington Pond, northern Vermont, to be nearly equal to the percentage of its pollen in the surficial pond sediments. These authors also have found-gaggg grandifolia to be_i pro- portionately represented in this region, although the per- centage of beech pollen in the surface sediment was slightly 107 greater than the percentage of beech in the surrounding forest. Acer rubrum/saccharinum, 5° saccharum, Carya spp., Castanea dentata, Fraxinus 4-colpate (incl. §._american§ and 'g. penpsylvaniga), g. 3-colpate QE’.££9£§): PoEulus spp. and .gilig amerigang are under-represented. Also in this cate- gory, although not listed in Table 16, is Prunu§_§erotina, a Species which is apparently entirely insect pollinated be- cause its pollen was not found in any of the spectra analyzed even though black cherry did occur in the surrounding forest. The low values for_gili§ and éggr species may likewise re- flect the influence of insect pollen vectors. Sangster and Dale (1964) have emphasized an addition— al reason for under—representation. They have demonstrated that pollen of different species vary in their resistance to degradation, implying that species of low resistance will be under-represented regardless of the amount of pollen they produce or the effectiveness of its dispersal. working mostly with species native to eastern Nbrth America, these authors have experimentally shown Agg£.§gcch§rinum and 5. saccharum pollen to be less well preserved in peat than the Pollen of Betula,_§raxinus,_gigu§,.Quercus,_glmus species, and that Pogulus_tremuloigg§ pollen is the most severely de- graded of all the types they investigated. Similar results pertaining to Pogulus are implied in data from New York State. At the four stations in the 108 southwestern corner of the state where airborne pollen data have been collected (Ogden and Lewis, 1960), Populus pollen is well represented, exceeding_gigg§ in numbers of grains per sq cm of slide surface, and occasionally equalling such other heavy producers as Quercus and_g§mg§. Similar counts do not occur in samples analyzed from bog surfaces. This may reflect.both the ease and the speed with which Populus pollen is decomposed, as Sangster and Dale in an earlier study (1961) showed that 80 percent of fresh p0plar pollen samples planted in the surface of a peat bog was degraded within 32 days.. The most accurate R values from the standpoint of the method used to determine vegetation composition are those calculated from the modern data. Since the three sites where the current pollen rain was measured are close enough to be considered in about the same source area as far as the regional pollen rain is concerned, the most useful way to summarize the R values is to compute the average for each of the taxa. These averages arranged in decreasing magnitude and placed in the three classes of representation are: (1) over -- Betula spp.,_ginu§ spp., Quercus Spp.,.g§ug§ canadensis; (2) proportional -- qugg grandifolia; and (3) under -- Populus spp., Acer saccharum, Carya spp., Fraxinus americana and/or pennsylvanica, Acer rubrumflgagcharinum and M t—n__ Tilia americana. \ 109 Few differences occur between this list and another in which decreasing average values calculated from both modern and pre-settlement figures are arranged. Use of these averages is less justifiable, however, because different techniques were used to obtain the basic data. Nevertheless the order of the taxa remains the same as above except Betula spp. and Pinus spp. switch positions. In class 2 Carpinus[gstrya and Juglans cinerea are added before ggqug and in class 3 Ergxinus_nigra and Castanea dentata are placed between Pogulus spp. and Acer saccharum and between Acer rubrum/saccharinum and Tilia americana respectively. Summary New York State occupies a position intermediate be- tween the northern coniferous boreal forest and the southern deciduous forest region. Two more or less well—defined floristic provinces occur at the west end of the state. Species confined to the warmer and drier plain south of Lake Ontario are mostly of southern affinity. Those of the cooler and wetter upland are characteristically more northern in distribution and reach their maximum abundance in the plateau but many also extend to the lowland. This difference in abundance is apparent in the plant communities which Characterize the two provinces. At the best sites the typi- Cal upland community is a mixture of hemlock and northern hardwoods including beech, sugar maple and yellow birch. Although these species form similar communities in the 110 lowland, they are much less frequent than deciduous tree communities which contain mainly beech and sugar maple at mesic sites and oak species at drier sites. In the plateau region of southwestern New York each of three typical stands sampled on upland soil types was dominated by sugar maple, beech and hemlock in that order. No precise line can be drawn separating the deciduous forest and the coniferous- deciduous forest in western New York emphasizing the broad ecotonal nature of this region. The pre—settlement forests, judging from studies which have used the first land survey notes, had this same basic pattern. However, due to the settlement of the region beginning about 1800, the forests have been greatly modified. The three counties along the shore of Lake Ontario are today the least forested of any area in the Northeast (Lull, 1968), and in general less than 50 percent of the land in the eight western counties is forested. Since about 1880, however, an ever increasing amount of the area originally cleared for farms has reverted to forest. Armstrong and Bjorkbom (1956) and Lull (1968) show the current forest vegetation to be com- posed of beech-birch-maple and white pine—hemlock—hardwoods types in the upland throughout the Allegheny Plateau but at Places also extending northward to Lake Ontario. These authors also recognize oak-yellow poplar and elm-ash—maple fiyPes in the lowland adjacent to Lakes Erie and Ontario. 111 Superimposed on this general pattern are other forest communities whose presence reflects local favorable edaphic situations. Bog and swamp forests, mostly occupying small areas, and riverine communities occur throughout the region, but of more restricted occurrence are the oak rich forest types that occupy well-drained sites associated with sandy deposits or areas of moderate relief in the lowland and dry south and southeast hill slopes in the upland. Other deciduous tree communities called Mixed mesophytic Forests have been recognized at sites between dry hilltOps and more moist lower slopes. “‘3 s ‘1 ‘ I 0“ ”I .I THE POLLEN DIAGRAMS METHODS Most of the techniques which I have used to prepare the pollen diagrams are detailed in Faegri and Iversen (1964), but a brief review of them is presented here. ‘gield Techniques The four basins chosen for palynological study (age Fig. 2), all appear to be ice block depressions. Protection and Houghton bogs are associated with the Valley Heads moraine and were chosen because their proximity permitted a check for parallel trends in the pollen profiles. Such trends, if found duplicated at many sites throughout a region, have greater significance in reconstructing the pattern of vegetation change than smaller, short-term changes present at only one site. The other two deposits are located south of the Valley Heads moraine on drift deposited at an earlier time. Allenberg bog is near the terminal position of the lfient moraine, while the Genesee Valley Peat WOrks is on a Still older surface, apparently of Clean age. Thus, two of the deposits began to accumulate at approximately the same tine; while the other two began at earlier times. 112 113 At each site a series of compass-oriented transects was run, along which depths and sediment lithology were de- termined using a Davis sampler. As three of the four sites investigated lack a central bog lake, traverses were easily made across the semi-firm bog surface. At the fourth site, however, a small lake 50 m in diameter is present, so a series of soundings was made around the mat at the edge of the open water. The goal in all cases was to find the deepest spot in the basin at which samples for pollen analysis could be collected. At two of the sites the upper sediments were too watery to be sampled with the instruments used, and a supplementary sample series was collected in the firmer sedi— ments to one side of the main sampling point. A set of three standard samplers was available and these were employed depending on the nature of the sediments encountered. At most sites a Hiller sampler (Borros, Solna, Sweden) with a 50 cm long chamber was used. This instrument worked best in coarse, fibrous peat of variable compactness, in finer peat deposited from water and in the stiffer lake muds or gyttjas. Lake sediment cores were also collected with a Livingstone sampler of the style described by Cushing and wright (1965). An adapter which allowed the sturdy Livingstone rods to be coupled to a Davis head (Eberbach, Ann Arbor, Michigan) was found to be particularly effective in penetrating heavier clayey sediments, mainly because the 114 piston.of the Davis sampler I used had a stronger catch than the piston of the Livingstone sampler. Two of the samples for radiocarbon analysis collected at Protection bog were obtained with the Livingstone sampler, equipped with a 2-inch barrel, from a location 50 cm to one side of the point where samples were taken for pollen analysis. Subsamples 10 cm in length were removed from the cores and submitted for age determination. To check the lo- cation of these age-date samples, in reference to the pollen diagram, the sediments immediately above and below were also collected and their pollen content determined. These spectra matched well those expected on the basis of depth alone (9; Diagram 1 and Appendix A). The uppermost date at this site was based on peat collected with a specially made piston sampler four inches in diameter built after the principle of the Livingstone sampler. The bottom 10 cm of wet peat from the core was sub- mitted for age determination. The Houghton bog date was obtained from wood which lodged in the Davis instrument during an exploratory probe of the area adjacent to the site where samples for pollen analysis were taken. An insufficient amount of wood was col- lected to permit both C-l4 analysis and microsc0pic study, so the species was not identified. Extreme care was taken not to contaminate the samples collected for pollen analysis. When using the Hiller sampler, 115 successively deeper samples 50 cm long were collected alter- nately from two holes 50 cm apart. The outer several mm of sediment uncovered by turning the outer sleeve to open the sediment chamber were cut away and discarded. The sediments thus exposed were removed with a micro-spatula from points midway between lines five cm apart stamped on the sampler head. Individual samples were placed in pint polyethylene bags labelled with the appropriate data. The cores col— lected with the Livingstone sampler were extruded in the field, wrapped in aluminum foil, labelled and placed in ply- wood core boxes. Cores taken with the Davis sampler were also wrapped in foil and labelled in the field. Subdivision of the cores was done in the laboratory. All samplers were washed with clean water after each use. The sediment samples were refrigerated until macerated in the laboratory. _Lgbor§tory_gechniques The standard methods for separating and concentrating pollen from sediments by removing inorganic and unwanted organic material were used in the preparation of the samples. A 1 cc subsample, measured with a graduated cylinder cut down to hold 1.: 0.05 ml of water, was macerated in each case. The schedule employed during the maceration procedure depended on the type of sediment being analyzed. 116 Two main types of organic sediments were encountered. The pollen in peat was concentrated by following the suc- cessive steps detailed below: (1) place sample in 40 ml tapered centrifuge tube, cover with 10 percent KOH and heat for 5 min in boiling water bath; (2) Wash sample through #60 mesh seive (250 mu) with enough distilled water to stop the KOH reaction; use seive residue for identification of plants making up the peat; (3) centrifuge and decant; (4) glacial HAc wash, acetolysis solution (Erdtman, 1960) in boiling water bath for 3 min and glacial HAc wash. Since gyttja, the other principal organic sediment found, generally does not deflocculate in KOH, a partial breakdown of this sediment was accomplished using cold 10 percent HCl. The sediment was gently teased apart with a glass rod. After centrifuging and then pouring off the HCl, the sample was acetolyzed and, if deflocculation had not completely taken place, the sample was heated for 5 min in 10 percent KOH in a boiling water bath. The KOH was removed from the residue by centrifugation and distilled water washes. Inorganic sediments containing calcium carbonate were treated first with 10 percent HCl, then acetolyzed and, if necessary, finally treated with 10 percent KOH. If the samples contained silts and clays, these were generally treated with cold, 72 percent HF for 24 hours culminated by an additional 30 min in a boiling water bath. After centri- fuging and decanting the HF, 10 percent HCl was added to the residue and the mixture heated in a boiling water bath for 117 3 min to remove colloidal silicon dioxide and silico- fluorides. The acids were then thoroughly washed from the residue with distilled water. Acetolysis, and in some cases a 10 percent KOH treatment, completed the maceration. If heavy minerals such as pyrite were present in the sample, heavy liquid separation using zinc chloride (sp. gr. 1.93) immediately followed the HF treatment. After maceration, all residues were treated in the same manner. They were thoroughly washed in distilled water, put through successive washes in 96 and 100 percent ethyl alcohol, stained with 2-4 drops of safranin—O in 100 percent ethyl alcohol, given a final 100 percent ethyl alcohol wash and then pipetted into labelled 3 dr vials. The staining was best if the residues were soaked for 24 hours in dis- tilled water previous to alcohol dehydration. Residues were stored in vials at room temperature until mounted for counting. The technique used in mounting residues for micro— scopic study was devised to enable the calculation of the number of pollen grains per unit volume of sediment and is an adaptation of the method described by Davis (1965a, 1966). The residue was washed into a 12 ml graduated centrifuge tube with tertiary butyl alcohol (TBA) and brought to a known.volume. After thoroughly mixing the residue and the TBA by vigorous pumping with a large bulb pipette, a certain volume of the mixture was removed from the tube with the 118 pipette and from 3 to 15 drops were released onto a small amount of silicone oil (2000 cs) placed at the center of a slide on a slide warming table (Ea. 75°C). The heat from the warming table rapidly evaporated the TBA leaving a mixture of silicone oil and residue. To produce an even distribution of pollen grains under the cover slip, a clean dissecting needle was used to further blend the pollen and the silicone oil. The silicone oil-residue mixture clinging to the needle was wiped off on the underside of the cover slip which was promptly placed downward over the preparation. After making three slides from each of the residues, TBA was removed from what re- mained of the residue by a 100 percent ethyl alcohol wash. After centrifuging, the residue—ethyl alcohol mixture was pipetted back into the appropriate vials. Because of the noxious character of vaporized TBA, it is necessary to carry out these procedures in a fume hood. The volume of TBA- residue mixture in the graduated centrifuge tube and the number of drops of this used to make each slide were recorded. Since the same pipette was used to mount every resi- due, the volume of one drop delivered by this pipette was presumably constant for all preparations. It was found, after 20 trials, that the pipette delivered 48.95 drops per ml. When counting, the number of traverses completed across a cover slip was recorded. Since in most cases the basic 119 sum was reached before the entire area was examined, the following equation was used to determine the total number of pollen and spores under the cover slip: number of traverses / total number of possible traverses = sum of terrestrial pollen and spores / x, where x is the number of grains per slide or per y drops of residue-TBA mixture. Multiplying x by a factor, 2 ml of residue-TBA mixture in the centrifuge tube - 48.95/’y drops delivered to the slide, gives the number of grains per cc of wet sediment. Although many potential sources of error are present in this technique, it is considered to give reasonably re- liable results for the amount of time expended. To test the accuracy of the mounting technique, I determined the number ‘ ' I of arboreal pollen grains for an equal number of traverses in each of two or three slides prepared from the same resi— due. These data, listed in Table 17, show that the method delivers similar numbers of pollen grains to individual slides in a series. One of the most serious causes of error during count- ing occurs when the residue is concentrated under some sector of the cover slip. If, for example, the counts were made only in a zone of high concentration, fewer traverses would be necessary to reach a given sum than if the residue were evenly dispersed under the whole cover slip area, and thus, the calculated estimate of the number of grains per unit volume would be greater than was actually true. Visual 120 TABLE 11 DATA FOR CHECK ON MOUNTING TECHNIQUE USED IN THE DETERMINATION OF ABSOLUTE POLLEN FREQUENCY m SAMPLE NUMBER NUMBER AP NUMBER OF TRAVERSESl Rb 7020-1 262 22 Pb 7020-2 236 22 Pb 7156—1 116 44 Pb 7156-2 117 44 Pb 7156-3 122 44 Pb 7157-1 264 44 Pb 7157-2 261 44 Pb 7180-1 266 22 Pb 7180—2 272 22 Rb 7906-1 159 44 Pb 7906-2 156 44 Pb 7906-3 160 44 Pb 7907-1 '188 44 Pb 7907-2 185 44 Pb 7907-3 207 44 1-Using a magnification of 250 diameters 44 traverses are generally possible with little or no overlap across a 22 sq cover slip. 121 inspection and adjustment in the location of the traverses was used to overcome this potential source of error, and in many cases parts of several slides were also counted to ob- tain a representative sample of the residue. Counts were uniformly made using equispaced traverses controlled by a mechanical stage. The distance between the traverses was initially chosen with reference to the density of the pollen under the cover slip. The lower the density, the closer together were the traverses. Counting was done routinely at a magnification of 250 diameters using a Leitz Ortholux microscope. Grains difficult to identify were ex— amined using a 95 x oil immersion lens. As an aid to the identification of the pollen encountered in the preparations, a large collection of reference slides was assembled, and the standard pollen identification manuals were also fre- quently used (Erdtman, 1943, 1957, 1965, 1966; Faegri and Iversen, 1964: WOdehouse, 1935). At least 500 pollen grains of trees were counted in nearly all samples. The percentage base used to calculate the relative pollen frequencies at a given level was the sum of arboreal pollen (AP) and nonarboreal pollen (NAP) at that level, as this figure best represents the regional upland pollen rain (wright and Patten, 1963). The percentage base varied from spectrum to spectrum with the greatest differ— ences occurring in the uppermost post-settlement and in the basal inorganic sediments where NPA is abundant. Relative 122 frequencies of other pollen taxa-including mostly bog plants, aquatics and pteridophytes-—everything to the right of the sums column in the pollen diagrams——were calculated using the sum of the first percentage base and the total number of pollen and spores of the miscellaneous taxa at a given level as a new percentage base. Pollen grains and spores which were well-preserved but which could not be identified due to inadequacies in the pollen reference collection were classified as unfamiliar. Grains in the unknown category were in part corroded or broken, in part obscured by debris and in other ways rendered undeterminable. SITES ASSOCIATED WITH THE VALLEY HEADS MORRAINE Protection Bog This basin is a large, fairly shallow ice-block kettle now nearly completely filled with sediment. Peat com- prises the upper 3 to 4 m and the bog plant communities which currently occupy the surface continue to add to the deposit. The kettle occurs at an elevation of 1410 feet in an area of morainic topography of Valley Heads age somewhat north of the head of the Chaffee outwash plain. To the east, north and west the surrounding hills rise 300 to 400 feet. The bog is located in the Town of HOlland near the southeast corner of Erie County, 0.2 mi west of the Erie-wyoming County line, 1.8 mi northeast of Protection at 420 37' 10" N Lat and 123 780 28' W Long. It is shown as a wooded marsh in the north- west sector of the Arcade 75' quadrangle. The bog and much of the surrounding land is currently owned by the County of Erie whose Bureau of Forestry administers the area as its Plantation #5. The bog surface is uniformly covered with vegetation (age Plate 1, Fig. A) and there is no standing water except for a temporary lagg along the north edge of the mat. As measured by planimetry on an aerial photograph, the original lake occupied an area of about 22 acres. Of this, 15 acres are now nonforested and covered mainly by ericaceous shrubs with occasional clumps of tamaracks. The basin has an ir- regular outline with the long axis trending east—west. There is a prominent bay extending northward and smaller bays, which are now covered with deciduous swamp forest, occur at the east and west ends of the basin. The steepest slopes above the basin are found on the north and south sides. An intermittent shallow stream occurs at the east end, but the basin has no permanent outlet or inlet. The stream, which discharges into a tributary of Buffalo Creek, apparently functions only at the peak of the spring runoff as it is nearly dry during the summer. .In spite of the advanced stage of basin infilling, most of the major vegetation zones characteristic of bog suc- cession are still evident. On the north, east and west sides a prominent and almost impenetrable, high shrub zone occurs. 124 Plate 1 Figure A. Protection bog, view looking southeast, August 23, 1967. Figure B. Houghton bog, view looking east, August 26, 1967. 125 126 Vaccinium corymbosum, Pyrus melanocarpa and Nemopanthus mucronata are the major species. Rhododendron nudiflorum is also present but is less frequent. The western one-third of the open mat is occupied by a low shrub heath in which Cassandra calyculata is dominant. Andromeda qlaucophylla, Kalmia polifolia and Vaccinium myrtilloides are other shrubs which occur in this zone. Eriophorum_yirqinicum and Sarra- cenia purpgrga are typical herbs. Sphagnum_gapillaceum,.§. magellanicum and_§. recurvum form a continuous carpet be- neath the shrubs and at places Polytrichgm juniperinum var. gracilius and Sphagnum fuscum have built up hummocks. _§§£l§ laricina and_gigg§ strobus seedlings are present throughout the low shrub zone. An island of Larix laricina trees 4 to 6 inches in diameter occurs near the middle of the mat and extends eastward and southward. Large_ginus strobus trees are present on humified peat along the west and northwest sides near the hinge line. Picea mariana does not occur at this bOg. Much of the upland around the basin was formerly under cultivation, although conifer stands planted during re- forestation projects and secondary forests developing on abandoned fields currently occupy much of the area. The woodlot to the east and southeast of the bog is the least modified of any nearby forest remnant. On muck in the lower areas Acer rubrum, Betula alleghaniensis, Carpinus_garoliniana, Prunus serotina and Tsuga_ganadensis are the principle trees. 127 Typical herbs in this area include Clintonia borealis, Coptis groenlandica, Oxalis acetosella, Medeola virginiana and Trillium undulatum. Upslope on better drained soil, Acer saccharum and Fagus grandifolia are abundant and they occur with Fraxinus americana, Tilia americana and Tsuga_ganadensi§. Plantations of Pinus resinosa,_g. strobus and Larix_decidua interspersed with untilled fields occur south, east and north of the bog. To the west and northwest, contiguous with the bog, is a narrow zone of open swamp forest in which_Aggr rubrum and Ulmus americana are the dominant trees. Populus qrandidentatg and Crataegus sp. are common at disturbed sites and abandoned fields around the bog. Sediment Stratigraphy The bog was sampled on August 24, 1967 near the center of the basin. The Hiller sampler was used from the surface downward to 6 m but, because of the compactness of the sediments, it was necessary to substitute the Davis head attached to the Livingstone extension rods to sample beyond this depth. The stratigraphy at the sampling point is (Diagram 1): 0.00-0.10 m : peat, with sphagnum leaves; humified, dark brown; 0.10-0.90 m : peat, undifferentiated but with sphagnum leaves; reddish brown; 0.90-3.25 m : peat, undifferentiated but with sedge leaf fragments and other plant debris; coarse near the top but gradually be- coming finer downward; reddish; 128 3.25-6.30 m gyttja; soft, somewhat gelatinous, brown at top, gradually becoming stiffer and rubbery downward, with Najas seeds from 4.75 to 5.50 m; silt and clay ad- mixture beginning at 5.60 m; mostly brown or reddish brown; silty-clay, with some medium sand at bottom; bluish-gray. The sediments could not be penetrated further with equipment used. 6.30-6.33 m Houghton Bog Located 12 mi southwest of Protection bog, Houghton bog occupies an ice-block depression in a pitted outwash plain which extends southward from the Valley Heads moraine past the village of Springville to Cattaraugus Creek. Houghton bog is one of the larger of the many basins which dot this plain. Most of the depressions have filled with sediment, but several open lakes are also present. Of these, Dead Man's Lake is the largest. It occurs in a deep kettle hole just northwest of East Concord and is surrounded by an extensive bog mat. Most of the smaller basins lack bog plant communities. The outwash fan forms a minor divide separating drainage in a general northward direction to Eighteen Mile Creek and the west Branch of Cazenovia Creek from drainage southward to Cattaraugus Creek. Houghton bog is situated in the Town of Concord, 2.3 mi north of Springville between US 219 and Sharp Street at 42° 32' 30" N Lat and 780 40' 13" W Long and is shown as an area of wooded marsh on the Springville 78' quadrangle. The basin occurs in a forest remnant about 45 acres in size 129 which is completely surrounded by tilled and fallow fields. Little of the original forest remains within a 3 mi radius of the boq. The basin measured on an aerial photograph, covers about 18 acres. The shoreline occurs between the 1400 and 1410 foot contour lines. Except for about 5 acres occupied by an open mat (gee Plate 1, Fig. B) the depression is now entirely forested. The bog is owned by the Nature Sanctuary Society of Western New York, Inc. which attempts to maintain it in an unmodified condition. The long axis of the basin has a north-south orien- tation. Two shallow bays extend beyond the main depression to the north and south (Brosius, 1953). These must have rapidly filled with sediment for the upper layer is strongly humified and now supports a swamp forest. The 5 acre bog mat extends entirely across the surface in a northwest- southeast direction. The mat is thinnest near the southeast end where the weight of a man is sufficient to cause water to seep upward. Elsewhere the mat is firm. No permanent outlet or inlet is present, although a low area which ap- parently is the route for excess spring runoff extends south- ward to join a tributary of Spring Brook. This temporary drainage channel is dry in the summer and its bed is mostly filled with water-saturated muck. The slopes immediately above the basin are gentle. The surface of the outwash plain is rolling and uneven and in general no more than 20-30 130 feet above the mat surface. The surrounding hills rise about 300 feet above the surface of the plain. The bog is surrounded by a strip of forest of vari- able width. On the slope above the mat the main tree species are Acer saccharum, Betula alleqhaniensis, Fagus grandifolia, Prunus serotina, Tsuga canadensis and Ulmus americana. Adjacent to the north and northeast bog margin occurs a some- what larger woodlot. This stand was heavily logged in the past and is now characterized by trees of small diameter, mostly-Age; saccharum and Fagus grandifolia. A few Prunus serotina trees are also present, and Tsuga canadensis seed- lings were noted. On the wetter, organic-rich soil between the hinge line and the open mat, a swamp forest of Acer rubrum, A. saccharinum, Betula alleqhaniensis, Pinus strobus, Prunus serotina, Tsuga canadensis and Ulmus americana occurs. At places in the swamp forest, Taxus canadensis forms a dense cover. Pinus strobus is particularly abundant along the south and west margins of the bog mat. A narrow high shrub zone of_Nemopanthus mucronata, Pyrus melanocarpa and.yiburnum cassinoides occurs between the swamp forest and the open mat. Cassandra calyculata is the dominant shrub throughout most of the open mat, but Kalmia pglifolia and Vaccinium myrtilloides and y. oxycoccus are also present. Carex canescens, Q. pauciflora,_g._trisperma, Rhynchospora alba, Sarracenia purpurea and various bog orchids are restricted to certain parts of the mat. The main Sphagnum 131 species include_§. capillaceum var. tenellum, §. fuscum, g. magellanicum and_§. teres; §. cuspidatum occurs in the wetter areas near the southeast end. Pinus strobus seedlings are abundant on the grounded mat, but Larix laricina is re- stricted to the southeast and northeast corners of the mat. Picea mariana does not occur at Houghton bog. Sediment Stratigraphy Because of the water pocket beneath the mat, two series of samples were collected at this site. The first, designated as section B, was taken on September 1, 1966 some- what east of the center of the open mat using a Hiller sampler between 4.00 and 8.00 and a Livingstone sampler with a 2 inch tube beyond 8.00 m. The stratigraphy at this point is: Section B (Diagrams 3 and 4) 0.00-0.75 m peat. fibrous, with sphagnum leaves, with water; no samples taken; finely comminuted plant debris in water, few seeds present near bottom; no samples taken; gyttja, soft & gelatinous at top with small amount of plant debris, gradually becoming stiffer and more rubbery down- ward; brownish throughout; marl, with dark brown and reddish laminae; mollusca and charophyte oogonia abundant, strongly calcareous; whitish gray; clayey silt, dense with gravel near bottom, carbonized wood fragments present sporadically; mostly dark gray but some- what brownish; sharp contact with marl above. Further penetration impossible. 0.75-4.00 m 4.00-9.43 m 9.43-9.95 m 132 Section A was collected on October 17, 1966 with the Hiller sampler near a small stand of tamarack located one- half the distance toward the center of the basin 185 feet s 100 E of the sampling point for section B. The stratigraphy of section A is: Section A (Diagram 2) 0.00-0.10 m 0.10-0.60 m peat, humified; dark brown; peat, undifferentiated but with sphagnum leaves; light reddish brown; 0.60-0.85 m : peat, undifferentiated but with sedge leaf fragments; somewhat coarse grading downward into finer texture; gray to dark gray; gyttja, soft, with some sedge leaf debris near top, becoming stiffer downward; gray. 0.85-4.00 m Pollen Stratigraphy In the following discussion the pollen diagrams have been divided into zones using the letter designations that Deevey (1939) first applied to his New England pollen pro- files. These have found wide usage in the northeast. Zone C, the uppermost, is characterized by an assemblage of hard- woods and hemlock and can generally be subdivided into three main parts. Below this in order are zone B which is dominated by pine pollen and zone A in which abundant spruce pollen is found. The T zone (Leopold, 1956b) records an interval be- neath the A zone in which NAP percentages are high. I am aware of the recent trend of describing pollen assemblage zones from bog and lake sediments in accord with the Code of Stratigraphic Nomenclature (egg Cushing, 1967) rather than extending the use of letter zone designations 133 developed in one region to distant geographical areas. But I feel, as does Livingstone (1968), that zones are "to be regarded as temporary divisions of convenience, to be used as reference points in discussions of the underlying trends in the pollen curves . . . [and that they] . . . should not be enshrined under the protection of a code involving strict rules of description and priority" (p. 95). There are chronological reasons, and perhaps climatic ones as well, for extending Deevey's zones to western New YOrk. It should be understood, however, that floristically and vegetationally the zones are not the same in New England and western New York or, for that matter, nearly any- where else they are used. This is clearly brought out by Deevey (1957) in a summary table in which he compares pollen sequences from northern Maine, southern Connecticut and Michigan by subdividing them into A, B and C zones. Because of the close similarity between the pollen diagrams for Protection bog (see Diagram 1) and Houghton bog (gee Diagrams 2 and 3), they will be discussed together. Minor pollen types not included in the diagrams are listed in Appendices B, C and D. Zone_A. In the lowermost spectra at both bogs 40 to 50 percent of the sum is comprised of Picea pollen. It is about five times more abundant than pollen of any other indi- vidual type. Although both records may be truncated at the bottom, there is no indication of a T zone beneath the A 134 zone at either site. At Protection bog, however, the abrupt decline in the spruce curve, which below the maximum at 6.265 m is coordinated with a 25 percent NAP high, may, in part, record the transition from herb to spruce dominated vegetation. The number of terrestrial pollen and spores rises rapidly from about 18,000 to 140,000 grains/ml of wet sediment between 6.325 and 6.195 m. Assuming a constant rate of sedimentation across this interval, the change parallels that reported for Rogers Lake, Connecticut during the T to A zone transition (Davis, 1967b). At Houghton bog a similar change occurs, but it is not as readily interpreted (gee Diagram 4). The deepest sediment sampled at this location was a dark gray silty clay, apparently barren of pollen. Upward, passing abruptly into marl, the number of grains per unit volume was at first very low, 2000 to 3000 grains/ml (see Fig. 7), but rose to 60,000 grains/ml in the first gyttja sample immediately above. Be- cause of the change in sediment stratigraphy, it is unlikely that the rate of sedimentation was constant from the clay upward through the marl to the base of the gyttja. It seems possible instead that, in relation to the accumulation rate of the gyttja, the marl was deposited rapidly, resulting in a lower number of pollen and spores per unit volume. Unfortunately close interval radiocarbon dating is the only method at present which can be used to determine accurately the sedimentation rate, and the necessary age 135 determinations are not available for Houghton bog. At .Rogers Lake (Davis and Deevey, 1964; Davis, 1967b), Seth's Pond, Massachusetts and Silver Lake, Ohio (Ogden, 1966), however, fairly uniform rates have been demonstrated. These involve about a two- to threefold increase from lateglacial time through nearly all of the postglacial except the most recent. At Protection bog, subdivision of the A zone into a Picea-Abies subzone is suggested by the prominent peak in the_§big§ curve near the top of the zone. The greatest per- centage of Abigg pollen at Houghton bog also occurs in the upper part of the A zone. _L§£ix, although not encountered in A zone sediments at Houghton bog, accounts for about 5 per- cent of the sum in the middle portion of the A zone at Pro- tection bog, decreases upward and finally drops out of the counts in zone B. .EEEEE pollen regularly comprises 10 to 15 percent of the total in the lower levels of both bogs, but upward, its percentage gradually increases until the maximum is reached in zone B. Three different categories of_ginu§ pollen were counted. The basic separation was between grains which could be identified as belonging to the softwood pines, subg. Haploxylon, which in east-central Nerth America includes only_ginus strobus, and the hardwood pines, subg. Diploxylon, which in this region includes g. banksiana and_g. resinosa. The most readily observed differences between the pollen of 136 these two subgenera is that the germinal furrow, located on the distal face between the bladders, is verrucose in Haploxylon pines, whereas in Diploxylon pines it is smooth (Ueno, 1958). The third category,_gigg§ undifferentiated, contains grains that could not be oriented to permit obser- vation of the furrow, those grains in which the exine be- tween the bladders was missing and reassembled grains, the number of which was determined by keeping track of the larger fragments and then dividing the sum of these by an appropri— ate figure to reduce the sum to the number of whole grains. Diagram 1 shows that diploxylon pine pollen was the major type identified in zone A. This also is illustrated in Dia- gram 3, but less prominently. The sum of the three cate- gories is graphed as_gigu§ total. ygmgg pollen is found in low percentages near the bottom of zone A but gradually increases to about 7 percent near the top of the zone at Protection bog. From 1 to 2 percent_glmg§ occurs at an equivalent stratigraphic position at Heughton bog. About 5 percent of Carpinus and/or Ostrya pollen is present in all A zone spectra. At both bogs there is a small but definite peak near the A to B zone transition. ,Low percentages of Corylus pollen occur in all A zone spectra at both bogs. Betula pollen is regularly present in A zone sediments, although in fairly low percentages. At Protection bog there is a gradual increase in percentage up- ward and near the beginning of the B zone a maximum is reached 137 which persists throughout the lower part of this zone. A similar but sharper peak is present at Houghton bog. Populus pollen accounts for 2 to 3 percent of the total at Protection bog in zone A. Low percentages of 3—colpate Fraxinus pollen occur in the A zone of Protection bog. Pollen of this type, which at most sites was tabulated separately from Fraxinus pollen with 4 and 5 colpi is, judging from reference slide exami- nations, produced mainly by-g..nlg£§. Fraxinus americana and_§. pennsylvaniga, on the other hand, typically have quadricolpate pollen, although a few tricolpate grains are occasionally found in reference slide preparations of these species, as are some quadricolpate grains in reference slides of_§. nigra. Fraxinus pollen was not differentiated in this way at Houghton bog, but it is reasonable to assume that 4- colpate grains are as poorly represented in the A zone at this site as they are at Protection bog. The pollen of many taxa characteristic of the Hemlock-white pine-northern hardwoods and Beech-sugar maple forest regions appear in zone A. At Protection bog, 2 to 3 percent of.g§gg§ pollen occurs in the upper part of this zone, and at Houghton bog, 1 percent or less is found in the same stratigraphic position. At one or both sites, sporadic grains of Castanea, Fagus, Fraxinus 4-colpate, Juglans Eggerea, Liquidambar and Platanus were also encountered. At Protection bog about 1 percent of Acer saccharum pollen is 138 present in A zone spectra and trace percentages of undiffer- entiated Age; grains also occur at Houghton bog. One percent .Qggya pollen and about 10 percent of Quercus pollen occurs throughout this zone at both bogs. The presence of such pollen is somewhat difficult to explain ecologically, since at the time A zone sediments were deposited the region was presumably cool and moist and dominated by spruce forests. One possibility which has found favor with a number of workers is that such pollen is of secondary origin having been eroded from the surrounding till, into which it had become incorporated from older de- posits. It then subsequently entered the basin carried with the mineral sediments that characterize the lower portions of these deposits. As small quantities of deciduous tree pollen have been found in surface samples in regions far re- moved from the place of origin of such pollen, a certain percentage may have been blown to the sites from distant sources as well. With the exception of the top 20 cm of the sediment column, the A zone contains the highest percentages of non- arboreal pollen anywhere in the diagrams. At both bogs the pollen produced by unknown members of the Cyperaceae and Gramineae amounts to 5 to 8 percent of the total. Associated with them is the pollen of Ambrosia, Artemisia,_§umex, Thalictrum, periporate grains belonging to species in either the Chenopodiaceae or Amaranthaceae (Cheno-Am), and other 139 herbaceous taxa listed in Appendix 2 and 4. Significant percentages of pollen belonging to unknown members of the Asteroideae (Compositae) were regularly present in the A zone. In all diagrams these are graphed under the heading high—spine Compositae. From 1 to 5 percent of_512g§ and _§glix pollen occurred in this zone at both bogs. Low per- centages of Myrica pollen were found in the A zone of Houghton bog upward to the base of zone B. An age determination of 11,800.: 730 B.P. (I-3290) on wood near the bottom of the marl at Houghton bog affords a minimum date for the beginning of the A zone at this site. This is a good correlation with a comparable date of 12,000.: 300 B.P. (wa507; Rubin and Alexander, 1960) on wood from a marly silt deposited in a depression near Cheery Tavern Crossroads on the Chaffee outwash plain 10 mi to the east. Both provide minimum dates for the formation of the Valley Heads moraine in this region. Mollusks found in the marl at the Cheery Tavern site indicate the sediment was de- posited near the margin of a heavily vegetated pond. Pollen analysis and the fossil snails suggest a climate somewhat cooler than the present at the time of deposition (Daily, 1961). Remains of a mastodon were also uncovered at this site. The Mollusca which occurred in the marl at Houghton bog were not identified, but charophyte oospores removed from the residue after HCl treatment and part of the original 140 core were sent to Fay Kenoyer Daily for study. The col- lection contained only Chara sejuncta A. Br. (letter, April 11, 1968), a species which often grows in ponds with mud bottoms (Daily, 1961). WOOd (1965) treats it as a variety of Chara zeylandica and reports that it occurs from Massachusetts to the Great Lakes southward to the West Indies, Brazil and Uruguay. Its New York State distribution is listed in WOod and Muenscher (1956). _§9£§ g. Pine pollen is by far the predominant type in zone B at both bogs, although substantial percentages of Quercus pollen are also present. The boundary separating A and B zones was drawn where Quercus pollen begins to increase. This also is at about the middle of the_gig§§ decline marking the demise of spruce forests in the area. Pine pollen ac- counts for about 50 percent of the total in this zone, oak pollen for an additional 20 percent. The identification of high percentages of_§ing§ subg. Haploxylon suggests that Pinus strobus was the dominant pine surrounding the site when the B sediments were deposited. At Protection bog a Pinus strobus cone was found in stiff gyttja at a depth of 5.75 m near the beginning of zone B in the 2 inch diameter core collected for radiocarbon assay. This establishes the presence of white pine immediately adjacent to the basin during early B zone time. Small but significant percentages of Ulmus and Carpinus-Ostrya occur in the B zone and a slight increase is 141 shown in the amount of birch pollen present across the A to B zone transition. About 2 percent of the total pollen in the B zone is contributed by nonarboreal species. This level continues upward throughout both diagrams until the pre- to post-settlement boundary is reached. Diagram 4 in which the number of grains per ml of wet sediment is plotted shows that the greatest numbers of pollen grains in the B zone were contributed by_gigu§ and Quercus. .gigug reaches a maximum of 176,000 grains/ml at 9.25 m. At Protection bog the B zone_§ing§ peak has been dated at 9030 i 150 B.P. (1-3551). _§2§§_§:l. The boundary between zones B and C-1 was drawn at the middle of the_3§gg§ increase. Pine pollen at this point still accounts for about 30 percent of the total but it subsequently decreases to 7 percent in the lower third of the C-1 zone and to about 3 percent at the end. _§big§ disappears in the beginning of zone C-l. The percentage of Quercus remains high and amounts to nearly 20 percent of the total at Protection bog. Its decline is gradual at this site and, with a slight lag, parallels that of Pinus, al- though at the end of the C-1 zone it is still 10 percent of the total. A similar pattern in the curve for this species occurs at Houghton bog. The lower third of C-1 is dominated by the pollen of_T§gg§, Quercus and.ginu§. Upward, the latter decrease in abundance and are replaced in part by 142 _§ggg§ which at the end of zone C-l forms 35 percent of the total at Protection bog and about 10 percent at Houghton bog. At the beginning of the C-1 zone at Houghton bog are peaks in the curves for Fraxinus and Juglans. Fraxinus is strongly represented through much of the zone, and at Pro- tection bog the highest C zone percentages of Fraxinus 4- colpate occur in the C-1. Betula increases somewhat over its percentage in lower levels at both bogs and reaches a peak in the lower half of the zone. Acer, Ulmus, Carpinus- Ostrya and Carya are represented in all C-l spectra in amounts ranging from 5 to 10 percent. Both Acer rubrum (incl. 5. saccharinum) and_A. saccharum were present at Pro- tection bog in this zone. Smaller percentages of Tilia and Corylus occur throughout. At Protection bog there is a Tilia maximum of modest size at the beginning of zone C-l. Castanea first appears in the lower half of the C-1 zone at Houghton bog and, although sporadic grains occur in the same zone and at lower levels at Protection bog, Castanea was not regularly encountered in the counts until just above zone C-l at this site. The highest percentages in the Platanus curve occur near the middle of this zone at Houghton bog and a similar but smaller peak occurs at about the same stratigraphic position at Protection bog, although in both cases fairly high percentages persist into the lower part of zone C-2. 143 As shown in Diagram 4, the largest number of pollen grains of any one type per ml in the C-1 zone at Houghton bog was contributed by Tsug . Tsuga replaces Pinus and Quercus as the major contributor to the pollen rain upward from zone B to zone C-l. -§22§.§:2. The middle of the prominent-Eggqa decline was chosen to mark the C-l/C-2 boundary and at Protection bog this point has been dated at 4390 i 110 B.P. (I-3550). Associated with the decreasing_g§ug§ percentages at both bogs are increases in Acer, Betula, Carya,_§agus and Quercus curves. At Protection bog where species identifications of .Aggr pollen were made,_A. saccharum percentages are larger and increase more than those of A. rubrum (incl. 5' _§gcch§rinum). The number of_T§uga grains per ml decreases from 30,000 to 5000 across the C-l/C-2 transition. Small in- creases in the number of grains per volume for Fagus, Acer, Quercus and Betula are evident. At Protection bog there is a small decrease in the percentage of_§aqg§ pollen at about the middle of zone C-2. This probably does not reflect a decrease of Fagus in the surrounding forest, as the decrease is mainly compensated for by an increase in Cyperaceae pollen which is most likely of local origin. Other than the two spectra at Protection bog in which Cyperaceae pollen accounts for about 7 percent of the total, NAP percentages average less than 3 percent of the sum. 144 Zone C-3. It is difficult to place the C—2/C-3 boundary, but at both sites it was drawn after the decline in Quercus which was taken to mark the end of zone C—2. _T§gq§ percentages increase across this interval. At Pro- tection bog these changes are dated at 1270 i 95 B.P. (I-3549). It cannot be conclusively determined whether sedi- ments of the same age are present in Diagram 3 for Houghton bog because of the obvious absence of the upper spectra. Diagram 2 was prepared to overcome this deficiency. The exact relationship between these two perhaps can be de- termined only by radiocarbon dating, but an examination of the pollen curves in relation to the complete diagram for Protection bog indicates that zones C-3, C-2 and part of C-1 are present in Diagram 2. In the absence of dates, however, the boundaries have been drawn to indicate their questionable positions. The best markers in Diagram 2 are the low in the Eggqg curve and the corresponding highs in the_gg£ya,_ggqu§ and Quercus curves. The percentages are about the same magnitude at the edge and at the center of the basin. Zone C—3 has been divided into two subzones. In sub- zone C-3a, the lowest, Tsuga increases to 25 percent and .Egggg correspondingly decreases at both sites. At Protection bogqgggrgus, Betula, Carya and Acer saccharum, which decrease slightly at the C-2/C-3 transition, increase somewhat toward 145 the end of the C-3a. These changes are not evident at Houghton bog. The C-3a/C-3b boundary records the influx of settlers to the area and the associated forest clearance. The change is quite abrupt and about 50 percent of the total pollen in sub— zone C-3b above the bonndary is contributed by nonarboreal Species, mainly those associated with agriculture. Since the percentage base includes both AP and NAP, decreases in tree taxa percentages are directly related to the large numbers of NAP. At both sites_Ambrosia accounts for about 25 percent of the total. Other important herbaceous taxa include Gramineae (incl. Ceralia),_§gm§x and Plantaqo. The latter is perhaps the best zone marker, since it appears abruptly at the pre- to post-settlement boundary. Occasional grains which are found in the C zone below this level are best interpreted as contaminants, although species of_glantago were presumably native at this time to the region surrounding the bogs. Cheno-Am pollen also occurs in the" C-3b and the small increases in Artemisia and high spine Compositae may be attributed to introduced weedy species. Zea occurred in this zone at both bogs and_§agopyrum was found in the surface spectrum at Houghton bog. Clay and silt sized mineral parti- cles, presumably blown into the basin, were abundant in the C—3b subzone attboth bogs. 146 PoEulus and_§ig§§ pollen reappear in upper C—3 spectra. Increases in the percentages of Acer, Betula, Larix and_§igg§ occur at one or both sites between 2.5 and 7.5 cm levels and the surface. In spite of the absence of mature trees in nearby forests, a few Castanea grains were found in the surface samples at both bogs. LOCKPORT SITE Several years ago a rich deposit of plant debris was discovered on property belonging to Neil Malloy between Ewings Road and Eighteen Mile Creek, 4.5 mi north and a few degrees west of the center of Lockport in Niagara County at 430 14' 6" N Lat and 780 42' 30" W Long (see Fig. 2). Al- though somewhat north of the region on which this study concentrates, I have included the results of my research on the site because of the added information it provides on the lateglacial vegetation of western New York. The plant material lay beneath about 3 m of sorted sand and gravel which at the present time is being actively quarried by Malloy. Numerous animal fossils have been found at the site including a well-preserved mammoth tooth and abundant gastropods. Richard L. McCarthy of Lockport has obtained an age determination of 12,000 i 100 B.P. (I-838, Buckley £5 31., 1968) on a sample of spruce wood from the organic bed. 147 I visited the locality with McCarthy on October 16, 1966 at which time the water table was low enough to permit access to the organic bed. The overlying gravels had been completely removed and the organic material which was still in place lay beneath several cm of inorganic sediment de- posited largely out of a shallow pond which covers the bottom of the excavation every spring. At the sampling point the organic deposit was about 15 cm thick. Using a shovel, over 5 liters was readily collected and the contents of this sample will be discussed in following paragraphs. At the time the water table was just at the top of the deposit, and any hole being dug rapidly filled with water. The principal constituent of the deposit is wood. Large fragments of branches, stems and roots are present intermingled with abundant smaller woody debris 1 cm or less in length. .The larger pieces have rounded corners and other evidence of having been water transported. Spruce needles, mosses, seeds and other plant remains also occur, and very little inorganic sediment is intermixed. Judging from ex- ploratory digging elsewhere, the organic bed is of limited distribution. At 50 m N 200 W of the sampling point a residual "island" of sediment was left in place when the surrounding sands and gravels were stripped away. According to McCarthy, sediments of similar lithology occurred above the organic bed. The section was measured as follows: 148 0- 15 cm : soil, A zone; 15- 53 cm : sandy loam, with cobbles, reddish; 53-100 cm : sandy loam, yellowish; 100-104 cm : pebble lens; 104-112 cm : sandy loam, yellowish; 112-135 cm : sand, fine to medium, with dark stain; 135-174 cm : sand, medium to coarse, carbonate cementation into stringers trending north- south; 174-198 cm : gravel, with silt lenses, grading from 1t brown to reddish brown at base; 198—262 cm : sand, fine to medium with pebbles and cobbles, yellow brown to olive; 262-277 cm : sand, coarse, dark gray with yellow brown patches; 277-282 cm : silty clay, pink; 282-295 cm : sand, coarse, dark gray with yellow brown patches; 295-300 cm : silty clay with coarse sand, pink; 300-306 cm : sand, fine, dark gray; 306 cm : water table; 306-311 cm : silty clay; 311- cm : sand, fine. No organic layer was found at this point. On April 11, 1968 I again visited the site and at this time a backhoe was being used to dig a small pit 350 m north of the organic bed. The overlaying sand and gravels had also been removed from this new site. The excavation had begun in medium sand and had proceeded downward through fine sand with silt and clay to a depth of 2.5 m. A sample of the silty-clay sand, 1.8 m below the surface, was collected for pollen analysis. It seems<31ear that the plant material was deposited after having been abraded in moving water, but how this re- lates to the Late Wisconsin geological history of the area is not precisely known. Several possibilities exist. The age of the Lockport organic deposit corresponds closely to 149 dated wood samples from Lake Iroquois sediments found at 340 ft A.T. near Lewiston in western Niagara County (eee Karrow .EE;21°: 1961). The Lockport deposit is located 1.75 mi lake— ward from the Iroquois strand, more or less equi—distant be— tween two strand extensions which course southward toward the city. The Iroquois beach is about 400 ft A.T. in this area and the deposit occurs 50 ft below this. In view of these facts, it is likely that the organic material was deposited under comparable conditions at both sites. The ice front at this time was some distance north of the sites and drainage was to the east through the Rome outlet to the Hudson River. Lake Tonawanda which was trapped between the Onondaga and Niagara escarpments in northwestern New YOrk during the lowering of early Lake Algonquin was draining through five spillways. The two westernmost out- lets located at the present cities of Lockport and Lewiston carried the most water and functioned the longest. Separate deltas were built up north of these spillways and, as shown on the map of Kindle and Taylor (1913), the Lockport organic deposit occurs at the edge of a stony delta deposit which grades lakeward into lacustrine sand. A subaerial delta with shifting stream courses easily accounts for the burial of organic debris carried down the spillway or derived from the near shore vegetation. It is likely that this would have occurred in the early history of Lake Iroquois at a time when the lake was north of the main 150 strand, perhaps building the weakly expressed Newfane beaches of Kindle and Taylor (leie.). Subsequent ice retreat and isostatic uplift raised the lake to the level of the main beach, inundating and planing most of the delta and the New- fane beaches. It is also possible, however, that the organic debris became buried along the shoreline during these adjust- ments in lake level and the gravel and sand above represents off-shore beach deposits. In either case there is no evi- dence at present which suggests that the two events were greatly separated in time. The spillways gradually became extinct westward with further uplift and the draining of Lake Tonawanda. As ice retreated north of the St. Lawrence River valley, permitting drainage along its axis, Lake Iroquois came to an end. It was followed by a low stage presumably contemporaneous with the Champlain Sea episode, although marine waters did not invade the Ontario basin because of the low sea level at this time. Isostatic rebound subsequently tilted the basin roughly toward the west causing the lake level to rise initi- ating Lake Ontario. The sample of the organic bed was washed with water through a series of three seives with 1190, 500 and 177 mu screens arranged in decreasing size downward. The residues retained by each were then examined under a dissecting micro- scope and potentially identifiable fossils were removed and placed in vials. Thirty gms of the wet silty clay were 151 macerated with treatments in 10 percent HCl, 72 percent HF, ZnCl2 (heavy liquid separation) and acetolysis. About 240 grains were present per gram of wet sediment. The plant macrofossils in the deposit are of several types. Seeds are infrequent, but of the many species repre— sented I have been able to identify only a few. These are Abies balsamea, a single trigonous Carex achene, Eleocharis spp., Hippuris vulgaris, Menyanthes trifoliata, Potentilla anserina and Potomogeton spp. Branch tips with needles of _g;eee sp. or spp., cones of Picea mariana, spruce seed wings, leaves or floral bracts of Myriophyllum, a bracket fungus, a microscopic vesciculate fungus similar to that described by Rosendahl (1943) as Rhizophegites_butleri, megaspores of Selaqinella selaginoides, charophyte oospores identified by Fay Kenoyer Daily (letter, April 11, 1968) as Chara sejuncta and Tolypella qlomerata and thirty different moss taxa have also been identified. The mosses are the most useful macrofossils for making inferences about the type of plant communities present in the area 12,000 years ago. The moss assemblage is a mixed one indicating that a number of different communities contributed fossils to the deposit. This implies that the plants did not grow in place where they were found, but rather became intermingled prior to their burial, a fact which is inde- pendently confirmed by the abraded nature of the large pieces of wood in the deposit. '44 v 152 The identified moss species are listed in Table 18, and as none of them is extinct, knowledge of their current autecological requirements enables inferences to be made about the habitats present near Lockport 12,000 years ago. These are listed in an idealized transect beginning at the lake shore and proceeding inland, although, instead of the zonation which this implies, a habitat mosaic may have been present. A fairly good analogue for this vegetation exists today at places in the Straits of Mackinac region of Michigan where rich fens (ders, 1961) occur between sandy beaches and.g;eee glauca stands. Many of the moss species which are found in or near the fans in this area are considered boreal forest or subarctic disjuncts. The Lockport moss assemblage is comprised mostly of calciphiles, and presumably the near- by Niagara escarpment and local deposits of calcareous till provided a ready source of carbonates. A dry, sandy beach, perhaps near the lakeshore or some distance inland, but in any case above the zone of per— petual wave disturbance, is the first habitat indicated. Be- hind the beach, between dunes and elsewhere, rich fens oc- curred in shallow ponds enriched with carbonates. On drier, gentle lepes adjacent to the fen a somewhat shaded habitat characterized by shrubs and perhaps occasional spruce trees occurred, and this likely graded further inland into a forest of spruce and balsam fir, perhaps with occasional tamaracks. Exposed areas of the Niagara escarpment_ee. 5 mi southward 4.4 It}? 1153 menu CH pmuoomxo kHQmQOHQ O. .lamma .ameaaaoumxv ment buoaxooq so ceaou aaueouuaoe .umeanms “when u m luaumwuouumumco Hon u I “powwow: was» no Uaumwuouumumnu n + +-+I m + + I+-+ I + + ll+0§ m-+ + +-+ +-++-+ + + (\0 + + +-+I I I 1.:Hoalmmmmmmmmmm.Eocmuofin I I «.3omm ammmmmmmmmfl.mmmmmmmxmmm. I I .meanom w Auoum aocooomesoum Ho {300$ a a I I .Bpom mwocamwws semen: I I .o.m.m A.3oomv mmmmmmmmm.mmmmammmmxm. I I .3pom mopwoapcsfimo ecopammwh I I .o.m.m 1.3omm. mmmmmeaoce no .o.m.m 5.3pemv Edoomaaaomm.Esasowunwo I I .meaaom ..3bomv mcmwuonwunum Md#o:muuwn I I tomcma .b A.pflumv mmwa>nm0m>unu E=HH>QEMU I I :030 A.3©omv ewmwmmufl>uooou EUHH>£QOHQU>HQONMM. .umomsnum ..ncooamzv mmmmmmmm..m. .Hmm Aaaoeue .m e .beeaqc sauaeaaeum mammammmmme. I «mxmooq ..3pomv mcouwc mwcmamummmmm. I oxwmoq A.ucob .Ih.mdoomowusu Edwpwmwowmn I s.umcum3 A.euumm xo .Apcaqv momoowcuo> .Q I .umeumz 1.3eemv mammmmwm..m. «.umcHMS A.3cmmv mummmmm.mmWMflmmmmmmmm. I acauooee e euomum .sauo mmmmmmmm..wo mmmmmm. I ¢.ncob .U A.3©Umv mmumaaoum Esflaammmm. I «.nuozom a Howe: ..ouuomw A.3Comv Ezwuomkuu0pooma Eo>um «.Hmmm3num A.3Uomv ouunsamm EBMGEOUMHsd .nuonom 6 Home: ..:uuomw A.3©omv maflmuou maounoa .uaaau A.m .3eomv mmmeaaoea maoouuoe mason A.Hmomznumv oasmuaxcam Eonuauuwn «.0wwm 5.3pomv msousmusn cooommuoo e.aoaaoam 1.3bem. mwauoamm.aaaoeauoamm. +.+4.+.+4.+.+ I |+ +| I +-++-+ wanna no monouzo xuom ummHom cmcmnm when: .ompm com doom nomom moooumoamo modem H. com no powwow axes mmMB BflBHmdm Mm mflHmmom waflmmowum NBHm BEGQfiUOA ma manta 154 provided habitats for those species which today occur in rocky situations. Certain species are found in more than one habitat, but others are of more restricted occurrence. Twenty-six species are listed in Table 18 and additional species of Bryum and Campylium and species of Brachythecium and Sphagnum were also found. According to Ketchledge (1957), 46 percent of the species occur in the Lockport area today, although some are not of widespread occurrence and are considered rare in the area. Careful collecting would probably increase the per- centage somewhat. Nearly all of the species are character- istic of the boreal forest and the region northward. For example, 88 percent of them are listed by Steere (1947) as found in the Canadian eastern arctic which encompasses the area from the Ungava Peninsula to the north end of Ellesmere Island. The two species of greatest phytogeographical in- terest are Aulacomnium acuminatum and_§. turgidum. Steere (ihifij) considers the former to be an important component of the high arctic vegetation. It is a circumpolar species widespread in the North American arctic and extends into the subarctic in the western half of the continent. The most southern station, a major range disjunction, occurs in the Thunder Bay District of Ontario near the north shore of Lake Superior between Nipigon and Port Arthur (Williams, 1968). Aulacomnium turgidum is also a circumpolar species which 155 occurs throughout arctic subarctic North America. Although very rare in northeastern United States, it is found south- ward through the White Mountains to the Adirondacks. Macrofossils of various animals were also found in the organic bed. Beetle exoskeleton fragments were common. Ostracod valves, statoblasts of fresh water bryozoa and several snail shells were present. I have not attempted to identify these fossils, but the beetle remains have the greatest potential for contributing paleoecological data. The pollen Spectrum from the site is given in Table 19. Of the total pollen 41 percent is comprised of non- arboreal species, principally members of the Cyperaceae. eggeee accounts for 38 percent of the AP:.£123§ for 13.6 per- cent and.;e£;§ about 1 percent. About 5 percent of the total was contributed by_Qeercus, Betula and Fraxinus 3-colpate. Single grains of Acer, Carya, Juqlans cineree and Ulmus were also found. MicrOSpores of Selaginella selaginoieee were present and complement the presence of the megaspores of this species found in the organic bed. A number of blackened pre-Pleistocene spores were also found. SITES ASSOCIATED WITH PRE-VALLEY HEADS MORAINES Allenberg Beg Situated in Cattaraugus County in the Town of Napoli, a few miles north of the Wisconsin drift limit, this bog occupies a deep, northeast-southwest trending basin about 10 156 TABLE 19 POLLEN SPECTRUM FROM A SILTY-CLAY LACUSTRINE DEPOSIT OF LAKE IROQUOIS AGE,LOCKPORT SITE, NIAGARA COUNTY, NEW YORK — - — —_ _;i — 'No. Grains Percent TREES Picea 129 38. Larix Pinus undifferentiated _£. Diploxylon Acer Jeglage cinera Carya* Quercus Ulmus Betula Fraxinus 3-colpate SHRUBS Alnus salix HERBS Cyperaceae Gramineae Ambrosia Artemisia High spine Compositae PERCENTAGE BASE (SUM AP + NAP) 339 PERCENT AP 59.0 PERCENT NAP 41.0 MISC. Polypodiaceae Selaginella_ee1aginoides Sphagnum Fungus spores Pediastrum Dinoflagellate cysts pro-Pleistocene spores Ostracoda Broken Abietineae Unknown PERCENTAGE BASE (339 + SUM MISC.) 478 .p has H wtd bknhdmrdhdth HP‘CDNCDCDOPH NLHUJbWHUJwLDFJNFJ H [—1 L11 0‘00 U) l—‘O (DKO wkmm Ol-‘I-‘OUJ \ONCDO‘kD I—' \l wmwflOWDI—‘nfim Ikl-‘Ol-‘NOU‘IOOH mflmU'Il-‘O‘U'INWW N *Badly corroded. 157 acres in extent located at the outer edge of an area mapped as Kent (Binghamton) moraine by MacClintock“ and Apfel (1944). It is approximately 30 mi southwest of Houghton and Pro- tection bogs and is shown on the southeastern quarter of the New Albion quadrangle at 420 15' 4% N Lat and 780 52' 57" w Long as a small lake with a marsh on the southwest side, 2.7 mi south of New Albion and 1.2 mi north of the Pigeon Valley cemetary. The lake, which occurs near the northeast end of the basin, is about 40 m in diameter. Less than one-half mile north of Allenberg bog is Waterman swamp, a roughly triangular tract of swamp and bog forest about 300 acres in extent. The swamp probably began as a lake ponded between drift deposits to the south and north. Allenberg bog does not seem to have been connected originally to the lake, and at present is separated from it by a hill and other intervening high ground. Since both oc- cur at 1620 ft A.T., however, a possible connection between them may have been present around the south and east edge of the upland. Black pond, a small bog lake, is located at the west end of the swamp. Both occur in a valley above which hills to the east and west rise 250 to 300 ft. In the vicinity of Allenberg bog the valley floor is slightly higher in elevation than the area to the south permitting drainage in this direction through Cold Spring creek. Waterman swamp is the headwaters of Little Valley creek, which, as an outlet, functions 158 mostly in the spring carrying waters charged with humic acids to the northeast away from the swamp. The two streams eventually empty into the Allegheny River. The swamp, Allen- berg bog and some of the surrounding land are currently owned by the Buffalo Audubon Society which maintains the area as a wildlife sanctuary. The vegetation of the region has been described by Gordon (1940) and Schick and Eaton (1963). Most of Waterman swamp is characterized by the latter as an elm-ash— rhododendron swamp. On the several knolls which rise above the level of the swamp Pinus strobus is particularly abundant, and Betula alleqhaniensis and Tsuga canadensie are common as— sociates. Rhododendgep maximum and_yiburnum alnifolium are typical understory shrubs in this area. Below the 1720 ft contour epies balsamea, Fraxinus nigra, Larix laricina, Picea mariana,_gyrus americana and Ulmus americana are frequent. At certain places dense thickets of Nemopanthus mucroneee and Vaccinium Sp. occur. Large Abies, Larix and_g;eee trees, 18, 21 and 16 inches d.b.h. respectively, have been found at the southeastern corner of the swamp. Black pond is sur- rounded on all sides by an invading Cassandra_ealyculata heath, but an extensive sedge mat is absent. Small Picea, Larix and Pinus strobus trees are scattered across the heath. At Allenberg bog the zonation of plant communities around the lake is fairly distinct (see Plate 2, Fig. A). The photographs in Gordon (1940) taken in the mid—1930's 159 Plate 2 Figure A. Allenberg bog, view looking northwest, August 28, 1967. Figure B. Genesee Valley Peat werks, View looking north— west, September 2, 1967. 160 161 show a narrow low shrub zone separating the lake from the bog forest. Beavers, sometime after these pictures were taken, dammed the outlet and raised the water level high enough to kill most of the trees and many other plants then inhabiting the mat. They cut a shallow channel through the peat to the lake and constructed houses near the north- and southeast ends of the basin. The beavers were last seen in 1951 and the disappearance of the dam has since allowed the water level to return to normal. The pollen stratigraphy seems not to have been disturbed by their activity. Nuphar microphyllum and Nymphaea odorata have been reported from the edge of the open water, although currently only the former is present. A narrow quaking mat of gegeg limosa and Sphagnum spp. is located along the south and south- west margin of the lake, but northward becomes grounded. Here_Qessandre calyceLeEe and Decodon verticilletus are in- vading the open water directly. A discontinuous low shrub zone interspersed with dead trees occurs across the south- west two-thirds of the bog. .Andromedaglaucophylla, Carex spp., Cassandra, Decodon, Eriophorum virqinicum, Ledum qroenlendicum, Rubus hispidus and Vaccinium geegocarpeg are the main species present. _§y£;e.eerolipiana, an apparent coastal plain disjunct, has been found near the south end of the bog mat. Around the periphery of the basin, particularly along the north edge, Larix laricina and Picea mariana occur. 162 Nearer the upland, they are found with Acer rubrum, Betula alleghaniensie, Fraxinus nigra, Pinus strobus and Tsuga canadensge. The understory shrubs in this area are Viburnum cassinoides and Pyrus melanocarpa. Rhododendron maximum is present along the west edge above the hinge line. The upland vegetation has been described as a_Teege _eanadensis:§egus grandifolia forest with an admixture of _Acer_eacche£um (Gordon, 1940). Other upland trees include Juglans cineree, Ulmus rubra, Fraxinus americana, Prunus serotina, Acer pensylvanicum,_§. rubrum and Betula alleghan'- ensis. The typical forest herbs are Dryopteris spinulosa var. gptermedia, Lycopodium luciduleg, Mitchella repens, Medeola virginiana, Oxalis montana, Trillium undulatum, Viola incognita and y. rotundifolia. Oak forest does not occur in the nearby upland. Cultivated fields surround the entire area and ap- proach within 0.25 mi on the west and northeast sides of Allenberg bog. However, about one-half of the area in a 3 mi radius of the bog is forested. A narrow strip of cut— over forest occurs on the east and west sides and a similar but more extensive forested area occurs immediately to the south. Much of Waterman swamp has been heavily logged. Secondary forests on abandoned farmland are abundant in the area, but mature conifer plantations are rare. Southeast of Allenberg bog the New York State Conservation Department has 163 flooded about 30 acres for use as a waterfowl breeding preserve. Sediment Stratigraphy Sediments from Allenberg bog were collected in three series. Section A was taken on October 17, 1966 with a Hiller sampler southwest of the bog lake from solid peat peripheral to the sedge mat. The stratigraphy at this point is: Section A (Diagram 5) 0.00-0.15 m : peat, sphagnum leaves abundant, humi- fied; dark brown; 0.15-3.00 m : peat, undifferentiated, fibrous at top grading into medium to fine dissected peat downward, sphagnum leaves abundant above 1.5 m, Drepanocladus fluitans from 1.7 to 3.0 m; reddish brown throughout. Section B was collected on October 15, 1966, again with the Hiller sampler, at a point 60 m N 390 E of section A. Sampling was discontinued at 12.5 m because insufficient extension rods were available to reach beyond this depth. At this point the stratigraphy is: Section B (Diagram 6) 0.00-0.70 m : peat, fibrous, not compacted, watery; no samples taken; 0.70-4.50 m : water, some fine plant debris; no samples taken; peat, undifferentiated, finely dis— sected; brown; 4.50-6.25 m 6.25-7.20 m : peat, undifferentiated, finely dis- sected but with Drepanocladus fluitans; brown; 7.20-7.93 m : peat, with abundant sedge leaf fragments, gyttja percentage increases downward; brown; 164 7.93-11.90m gyttja, soft gelatinous at top, be- coming stiffer downward; dark brown; 11.90-12.50m gyttja with silt and clay; dark brown. Section C was taken through a Cassandra heath on April 12, 1968, 1.5 m west of the site where section B was collected. At this time more extension rods were available and the Hiller sampler was used to a depth of 14.5 m. The Davis head coupled to the Livingstone rods enabled further sampling to 15.17 m. The stratigraphy of this section is: Section C (Diagram 7 and 8) 11.50-12.30 m : gyttja, soft: dark brown; 12.30-14.90 m : gyttja, with increasing amounts of clays and fine sand, some plant debris present; dark brown above, be- coming light brown to gray to light gray at bottom; clay, stiff and dense, with dark brown stains, small specks of vivianite present; bluish gray; not sampled further because of the difficulty of withdrawing the sampler from the sediments. 14.90-15.17 m 15.17. m Pollen Stratigraphy Zone'g. The lowest sediments sampled at Allenberg bog, including the basal clay and a portion of the clay- gyttja above, contain a pollen assemblage rich in NAP see Diagram 7 and Appendix G). At 14.87 m, just above the base of the T/A zone boundary as it was placed in the diagram, 28 percent of the sum was contributed by nonarboreal species. In the next lower spectrum NAP increases to over 51 percent, and at 15.165 m it reaches a maximum of 55 percent. 165 The largest NAP contributor to the zone is the Cyperaceae which accounts for over 20 percent of the total. Also present is about 10 percent Gramineae pollen. From 5 to 7 percent of Artemisie pollen occurs, and Ambrosia, .Thalictrum and high—spine Compositae pollen are found regu- larly but in lower percentages in all T zone Spectra. Of the less common pollen types listed in Diagram 7 and Appendix G, pollen belonging to the Caryophyllaceae, Chenopodiaceae, Cichoriodeae and Labiatae and to Plantego, Ranunculus and Bflfléfi appear most regularly. MicrOSpores of Selaginelle selaginoides were found at 14.985 and 15.085 m. Pollen from the shrubs-Ainus, Myrica and_§e;;§ aggregate 15 percent of the total. The-most frequent AP type in the T zone, Picea, ac- counts for nearly 20 percent of the total. About 10 percent of_§;eee pollen is present, and in general half of this is of the Diploxylon type. Very low percentages 0f.§lflfl§ Haploxylon pollen also occur. Quercus pollen is uniformly present in amounts which range from 5 to 8 percent. _A high in the Quercus curve occurs near the top of the T zone and carries over to the lower A zone spectra where a peak oc- curs. Increasing percentages of Fraxinus 3-colpate pollen are found from the lowest spectrum upward across the T/A zone transition where a maximum of 9 percent occurs. .AQEEE: .eetule, Carpiee_-Ostrya and_g;mee are weakly represented and a few grains of Acer saccharum, Carya, Corylus, Fraxinus 166 4—colpate,.geglans cinerea, Larix and Populus occur in some or all of the Spectra. Diagram 8, which shows the number of grains per ml of wet sediment, was prepared for the same spectra graphed in Diagram 7. The number of grains in the three lowest samples is relatively small and ranges from 26,000 to 40,000 per ml (eee Fig. 8). In the sediments above, the absolute number rises gradually to about 200,000 at 14.425 m near the bottom of the A zone, and it fluctuates near this figure to the end of the zone at which point the number of grains again increases until the maximum of about 380,000 grains is reached at 12.425 m. Assuming that the rate of sedimen- tation was constant, Diagram 8 Shows that in relation to the A zone relatively few pollen grains of any type were de- posited during the accumulation of T zone sediments. The change upward into the A zone is marked not only by an in- crease in the percentage of Spruce pollen, but also by a six- fold increase in the numbers of spruce grains being deposited. Contrary to the implication of the relative percentage dia- gram, larger numbers of Cyperaceae pollen occur above the T zone than within it. .§2£§.§° Spruce pollen dominates Slightly over 2 m of sediments at Allenberg bog. It accounts for about 40 percent of the total in nearly all spectra except those near the bottom of the zone, where at one level over 60 percent was found. This peak is associated with lows in the Pinus 167 total and_gipee Diploxylon curves and high (but not the highest) percentages of Quercus and Fraxinus 3-colpate. The absolute pollen frequency diagram, however, does not Show these fluctuations, although lower numbers of_g;eee grains occur below the level of the_g;eee peak than above it. Except near the T/A transition,_g;eee accounts for about 20 to 25 percent of the total in all A zone spectra. Nearly half can be assigned to the Diploxylon type. Haplox- ylon grains also occur and, in the upper two thirds of the zone, they account for about 5 percent of the total, but nearer the bottom, lower percentages are found. With the exception of somewhat higher percentages in the lower part of the zone, Quercus averages about 7 percent of the total in all A zone spectra. Highs in Carpinus- Ostrya, Fraxinus 3-colpate and Polypodiaceae curves are also present near the bottom of the zone. Two peaks in the gpiee curve occur near the beginning and end of the zone at 13.075 and 14.425 m. From 2 to 5 percent 0f.HlEE§: Betula, Corylus and Populus pollen are present throughout. .LQEEE: although poorly represented in the lowest A zone spectra, increases to about 5 percent just below the middle of the zone and re- mains near this level upward to zone B. NAP percentages above the T/A zone transition are about one-third of what they were in the T zone. .A122§: Cyperaceae and Gramineae have the highest percentages. ‘Salix is more weakly represented in this zone than below and 168 finally drops out upward in zone B (eee Diagram 7) or near the bottom of zone Crl (eee Diagram 6). High-spine Com- positae, Ambrosia and Artemisia are present in nearly all spectra and the curve for the latter has three highs at vari- ous points throughout the zone. The relative and absolute pollen frequency diagrams agree closely across the A zone. .ZQB§.§° The A zone to the B zone change is Shown in Diagrams 6 and 7. In spite of the close proximity of the sample series, the B zone begins 30 cm lower in Diagram 7. However, the percentages in both match well. Just the top of the A zone occurs in Diagram 6. The A/B zone transition is marked by several im- portant changes in pollen percentages. In a Span of 25 cm _§;eee decreases from 35 to 5 percent, and it finally drOps out near the end of zone B. The boundary between the two zones was drawn at the middle of the Picea decline which also corresponds to about the middle of the.g;eee increase. This transition is characterized by peaks in the Betula, Fraxinus 3-colpate, Larix and Populus curves. Quercus percentages steadily increase in the lower half of the B zone and reach a high of 25 percent near the end of this zone. Associated with this is an increase in Carpinus-Ostrye percentages which remain high but decrease somewhat in the lower part of zone C-l over a peak reached near the end of zone B. 169 NAP percentages decrease at the A/B zone transition and are low throughout the B zone. Small but steady per— centages of_§£eee, Cyperaceae, Ambrosia, Artemisia, high— spine Compositae and Thalictrum occur. In no B zone spectrum do NAP percentages rise above 2 to 3 percent of the total. Both the absolute and relative frequency curves are Similar across the B zone. The high numbers of total pine pollen which occur in a broader interval than is evident in the corresponding relative frequency curve help to define the zone. The two taxa which contributed the greatest numbers of grains in the B zone of Allenberg bog, as was the case‘in the two Valley Heads bogs discussed previously, are_§;eee and Quercus. The total number of_g;eee grains at 12.425 m, the peak of the pine curve, was 145,000. Quercus at the same level was represented by 91,000 grains. The_§etula, Carpinus—Ostrya, Frexipus 3-colpate and Pepulus highs shown in the relative frequency diagram also appear when the data are plotted on an absolute basis. If the sedimentation rate was constant across the A/B zone tran- sition, these peaks occurred at a time of high pollen de- livery to the basin which presumably reflects a greater abundance of the plants producing these pollen types in the region neighboring the basin. Zone C-l. A COmplete C—l zone is shown in Diagram 6, but in Diagram 7 only the lower third of it is present. The 170 B/C-l transition is marked by rapidly increasing percentages of-Teege pollen coordinated with decreasing percentages of .Elflflé grains. However, the_g;eee decline is not as abrupt as the increase in.Teege and there is a small interval across which the percentages of both are high. Quercus remains strongly represented across the transition and high per- centages persist through the lower third of the zone. Near 11 m, Quercus decreases from 20 to 10 percent, while Eegee and Betula percentages increase. Quercus continues to de- cline upward through the C-1 until just below the 9 m level where only 7 percent occurs, its lowest postglacial per- centage at this site. .gegee is weakly represented in B and lower C-l spectra, but it begins to increase above 11.675 m, after_geege becomes stabilized at near 30 percent of the sum. Near the middle of the zone,.§egee accounts for 13 percent of the total, but in the upper one—third of the zone, it comprises 27 percent. The curves for_Acer saccharum, Betula and Juglans cinerea Show highs near the middle of the zone. Although Fraxinus 4—colpate and-glige first appear in the B zone, the former has a maximum in the lower third of the C-1, while in the same interval the latter has two highs, one near the beginning and one near the end of the zone. From 3 to 5 percent Fraxinus 3-colpate pollen occurs regu- larly in the lower two—thirds of the C-1. In the rest of the zone only about 1 percent is present. A parallel change also occurs in the Carpinus—Ostrya curve. Ulmus is uniformly 171 present in all C-l spectra and accounts for about 7 percent of the sum. The C zone maximum for Platanus is reached at the end of the C—1. .gegye pollen, which is found in low per- centages at the beginning of the C-1, increases slightly in percentage upward in the zone. Except for sporadic grains in lower levels, Castanea occurs regularly from the upper one-third of the zone to the topmost spectrum. Total NAP percentages vary from 1.2 to 3.6 in zone C-l. Cyperaceae and high Spine Compositae are most con- sistently present. Less frequently counted pollen types are graphed in Diagrams 6 and 7 or listed in Appendices F and G. As in zone B, the relative and absolute frequency curves parallel one another (ef. Diagrams 7 and 8). In that portion of the C-1 studied the largest numbers of grains be- longed to.Teege and Quercus. .EEEEE is also an important com- ponent of lower C-l Spectra. Zone_§-2. The boundary between this zone and C-1 is easily placed at about the midpoint of the_Teege decline, but the upper boundary of the C-2 is more difficult to locate. The gradually increasing_1eege percentages which characterize the C-2/C-3 transition in the Valley Heads bogs are not evi- dent in Diagram 6, although they do occur in Diagram 5. .Teege percentages increase somewhat in the upper 0.5 m of Diagram 6, and for this reason the zone boundary was placed just below this level. A case might also be made for locating it beneath the 8 m level at which point Quercus and Betula 172 percentages have decreased somewhat over their previous highs. However, this creates an unusually thin C-2 zone, especially when taking into account the amount of sediment above this level and a comparison between Diagram 6 and the complete Protection bog profile. A similar comparison sug- gests that a part of the C—2 zone may be represented in Diagram 5. At the C-l/C-2 transition_Teege pollen drops from 38 to 7 percent in 50 cm. A small decrease in ggmee percentages is also apparent across this interval. These reductions are mainly compensated for by increases in percentages of Age; _eeeeharum, Betgie and Quercus and to a lesser degree by .Qegye,_§egee, Fraxinus 3- and 4-colpate,_g;eee undifferenti- ated and gieee Haploxylon. The Eegee curve shows a gradual increase from 22 percent near the end of the C-1 to 30 per- cent just below the C—2/C—3 boundary at a depth of 6 m. Higher percentages of Acer saccharum, Betula and uercus are maintained throughout the C-2 than occur in the upper C—l spectra. Castanea is weakly represented across the C-l/C-2 transition but reaches its maximum of 4 percent at 7.175 m, well into zone C-2. Two Eeege highs, each about 15 percent, occur in the middle of the C-2. These represent an increase over the 7 percent_geege low present near the beginning of the C—2 and the 9 percent low which occurs near the end of the zone. After a sporadic presence in much of the C—1, Larix regularly 173 appears from the beginning of the C-1 to the uppermost spec- trum in Diagram 6. Similarly, Picea pollen, after an absence from all C-l Spectra except the lowest, occurs in low per- centages from near the beginning of the C-2 upward. Total NAP percentages remain small throughout the C-2. One or two percent of Alnus, Ambrosia, Artemisia, Cyperaceae and Gramineae pollen is most regularly present. The peak in the Cyperaceae curve just above the 7 m level may be associated with intrabasinal succession, since sedge peat, which is evidence of the presence of a sedge mat at the surface, occurs just below it. The associated Ambrosia high is more difficult to explain, although contamination during sampling may be the cause. _g92§_92§. The complete C-3 zone is shown in Diagram 5; percentages of minor pollen types are listed in Appendix B. At Allenberg bog the increasing_Teege percentages of this zone are associated with decreasing Eegee values. As in the Valley Heads bogs, the upper NAP rich sediments are placed in subzone C-3b. The remainder of the zone, where NAP per— centages are minimal, belongs to the C-3a. Except for minor fluctuations, the percentages of most pollen types remain more or less constant across the C-3a. The percentage of Betula pollen increases below the C-3a/C-3b boundary and remains higher in the IC-Bb than in the C-3a. Together_§;eee and_§e£;§ account for about 4 per- cent of the sum in most C-3 spectra. NAP percentages are 174 low in the C-3a subzone and average about 3 percent of the total. Alnus,_Ambrosia, Cyperaceae and high spine Compositae are most regularly present. The higher percentages of Cyperaceae in the lower half of the C-3a than in the upper seem to be related to intrabasinal succession. Upward from the lowest spectrum in Diagram 5, the aquatics_§rasenia, Nuphar and Nymphaea abrupt- ly drop out of the counts. Above the level of their dis- appearance, Cyperaceae percentages increase and above this, increases occur in the curves for Ericaceae and Osmundaceae. These changes match those expected during succession from open water to an ericaceous shrub association of the type which occurs at the surface today. Pollen typical of the terminal stage, a bog forest, does not replace the Ericaceae upward, but the occurrence of Larix and_g;eee pollen through- out zone C-3 is evidence for its presence somewhere on the bog mat. A few spruce needles recovered from the peat in the lower C-3b spectra imply the presence of spruce trees near the sampling point at the time this part of the zone accumulated. These were no doubt produced by black spruce which is found on the bog mat today. The C-3a/C-3b transition is abrupt and clearly marked by a decrease in AP and an increase in NAP. The largest re- ductions in tree pollen percentages occur in Fagus,_g;eee Haploxylon, Quercus and Teeg_. Acer sacchegeg percentages drop somewhat but generally remain high, as do those for 175 Betula. Only Acer rubrum/saccharinum and Populus show a marked increase in C-3b spectra. At 2.5 cm beneath the surface total NAP reaches 57 percent of the sum, the highest in the C-3b subzone at Allenberg bog. Pollen from herbs which today grow mostly in disturbed habitats is abundant. AS in the Valley Heads bogs, Ambrosia has higher percentages than any other NAP type. Pollen from Cheno-Ams, Cichorioideae, Gramineae_p;pe,.g;ee— .Eege and-gemeg, probably produced by weedy Species, also oc- curs. Fagopyrum, Gramineae_p;p; (incl. Ceralia) and.§ee pollen, representing cultivated plants, was present but in much lower percentages than the weeds. Highs in the Cyperaceae, Ericaceae, high—spine Compositae, Nemopanthus and Polypodiaceae curves probably reflect conditions on the bog mat favorable to the growth of local species. Pollen Size-Frequency Measurements In certain cases size is a useful species character in pollen which on other morphological grounds can be identi- fied only to genus (Cain, 1940; Cain and Cain, 1948; LeOpold, 1956a). For this reason measurements were made on as many well-preserved Betula,-gleee and Pinus Diploxylon grains as possible while counting the Allenberg bog sections. Sili- cone oil is a particularly effective mounting medium in such studies because gentle tapping of the cover slip rotates grains permitting access to the length chosen to be measured. 176 The data collected have been plotted in size—frequency graphs (Figs. 4, 5 and 6). Some difficulties in using size—frequency data have been reviewed by Whitehead (1964). The maceration procedure employed to isolate pollen from sediments, the nature of these sediments and the medium in which the pollen is mounted for microscopic study apparently affect grain size. When these sources of variability are coupled to the fact that in only a few cases has an analysis been made of the geographi- cal variation in pollen Size of a given species, not to mention the absence of an evaluation of differences within populations or within a single individual, fossil size-fre- quency data must be interpreted cautiously. Certain of these factors have been studied but less variability was found than anticipated. For example, Faegri and Deuse (1960) exposed Betula tortuosa pollen to different lengths of treatment in boiling 10 percent KOH and did not find a significant size change with longer treatment, al- though they did observe an increase in size when acetolysis followed exposure to.KOH. These authors have also shown negligible changes over a period of 5.5 years in pollen preparations mounted in water, glycerol and glycerine jelly. Clausen (1960), who studied freshly gathered pollen from different parts of Betula catkins located on different parts of two birch species, was unable totdemonstrate any significant size variations within a Single species. Similar studies 177 need to be extended to all species identified on size charac- teristics alone, but, if the same maceration procedure and mounting medium is used for the fossil samples and the modern preparations employed to identify peak frequencies within a fossil spectrum, variability caused by these factors can be minimized. Betula. The most extensive study of Size variation in pollen of the North American species of Betula has been published by Leopold (1956a). Birch pollen is triporate and the dimension usually measured extends from the tip of a pore across the grain to the edge of the exine in the inter- poral area on the opposite side. This, the maximum diameter of the grain, can be measured at three places. The grain diameterzpore depth ratio, used recently by Birks (1968) to identify Betula_eeee pollen, may also be useful for working with temperate North American species. Measurements collected from individual spectra were lumped by zones or major portions thereof to increaSe sample size and to produce graphs that were characteristic of the main subdivisions of the diagram (eee Fig. 4). As only one mode occurs in nearly all parts of the C zone, one or possib- ly two birch Species with pollen grains of similar size seem to have been dominant during C zone time. Today only Betula _;ee£e and e. alleghaniensis occur in the region (Zenkert, 1934) and, although the former is not listed by Schick and Eaton (1963) as growing within the Allenberg bog-Waterman PIICINY OI SAI'II 178 Io- ";mt '“.m°~& “I”... //—— r3345?" Kiaet/ \ ”1 /"———1\ c-3- l... m: Aw/ II. / :o- _//\ canuommm _ "t / \\ ,3 / ~% / INS‘ ...: b‘w/N ”N 1.5:: '9 /\ ..... .I N/ .. . / / 179 swamp area itself, it is reasonable to conclude that these species contributed most of the birch pollen to C-3b spectra. The gross volume data collected by the U. S. Forest Service Show both species to be about equally abundant in Cattaraugus County (Northeastern Forest Experiment Station, 1967). Be- cause nearly identical modes occur throughout zone 0:.2- .lEEEE and g. alleghaniensis are likely to have been the only species present. The highest size-frequencies occur over a l to 2 mu interval with the mode at 26 mu in the C-3b, C-3a and the upper and lower halves of the C-2, but in the C-1 the mode shifts to 27 mu. Leopold (1956a) has reported the mode for Betula_leepe pollen to be 28 mu in acetolyzed samples and to be near 24 mu in those treated only with KOH. In.§. alleghaniensis the mode for KOH treated samples is 28 mu but it is 45 mu for acetolyzed ones. My samples were exposed to both KOH and acetolysis, so it is expected that their modal classes would be near the larger figures. The fact that they are smaller may be related to the Shrinkage phenomenon re- ported to occur during fossilization in peat by Buell (1946) and others. The shift in the position of the mode in zone C-l to slightly larger grains may indicate the presence of Betula papyrifera, a species which, although now rare in western New York, may have been more abundant in the past. The modes shift to smaller size classes in zone B and below. The presence of at least some grains greater 180 than 25 mu suggests that Betula lenta, e. alleghaniensis and perhaps e. pepyrifera grew at an unspecifiable distance from the basin during the period of time represented by these zones. However, it is difficult to identify the main birch species contributing pollen to zone B sediments because none of the taxa studied by Leopold (1956a) has modes at 25 mu. Perhaps some aspect of the depositional environment caused pollen from the three tree species to shrink more in zone B than in those zones above—or possibly another species was present. If so, it may have been the shrub Betula pumila which has been found only once in western New York within the area of glacial lake Tonawanda, but likely had a wider distribution in the past. Leopold (1956a) considers the pollen of this species to be smaller than that of the tree birches mentioned above, but the mode at 30 mu for the aceto- lyzed sample she reports does not match the mode in the fossil material. Betula pumila pollen treated only with KOH is smaller, however, and six samples prepared this way have a mean size of 24 mu. The modes at 22 mu in the A and T zones can be identi- fied with more certainty. Leopold (1956a) has found Betula glandulosa, an arctic-alpine species found as far south as the Adirondack. mountains in eastern North America (Fernald, 1950), to have small pollen with modes at 20, 22 and 23 mu in the three acetolyzed modern samples she studied. These 181 comfortably overlap the mode for fossil grains in both zones at Allenberg bog. .Picea. The technique of using the size-frequency graph at the surface to interpret those beneath can also be used in reference to Picea (see Fig. 5). There are three spruce species that could have been members of the late- and postglacial vegetation of western New York,_gicea glauca, g. mariana and g. rubens. The last named is now found in mountain forests stretching from the southern Appalachians to New Brunswick and Nova Scotia (Fowells, 1965) but, in spite of some evidence that it may have occurred as far west as Michigan during A zone time in this area (Cain, 1948), the distributional history of the taxon is largely unknown. A number of recent workers have consistently sepa- rated the pollen of Picea glauca and_g. mariana. Most use only size characteristics, but there seem also to be morpho- logical differences between these species (ieig.). Size- frequency measurements indicate that the smaller grains generally belong to g. mariana and larger ones to g. glauca (Cain, 1948; Davis and Goodlett, 1960; Heusser, 1960). west (1960) has used 100 mu based on wingtip to wingtip measure- ments as the point of separation between them in his work in eastern Wisconsin. Unpublished measurements of this di- mension made by James H. Anderson of Michigan State Uni- versity (personal comm.) on three collections of_g. glauca treated with 10 percent KOH and acetolyzed have the following 182 If. '1' mm FIGURE 5. Size Frequency 6 183 means: 116 mu (Arnold Arboretum, Massachusetts), 104 mu (Cheboygan County, Michigan) and 99 mu (Neultin Lake, N.W.T.). Similarly treated samples of_g. mariana have smaller means: 85 mu (Ingham County, Michigan) and 79 mu (Thunder Bay District, Ontario). Since these data indicate that a total length greater or lesser than 100 mu is a reasonable point of division between the pollen of these species, this figure was used in the present study. Maximum internal diameter (excluding the wings) measurements of five collections of Picea rubens pollen show that the average of the means is about 3 mu greater in this species than an average of the same dimension in_g. glauca (Davis and Goodlett, 1960). Wingtip to wingtip measurements are unfortunately not available for g. rubens. The size—frequency graphs are readily interpreted with this information. We can be fairly certain that Picea mariana was the only species present in zones C-3a + b and C-2 because of the probable absence of habitats for the two other spruces near Allenberg bog during the past 4000 years. Only two grains larger than 100 mu were found in these zones, and it is likely that these originated from introduced species because they occurred in post-settlement spectra. As mature Picea eplee trees are common near farm dwellings and in cemetaries in the area, this species is a possible source. The mean grain size in the C-3a + b is 81 mu; in the C-2 it is 82 mu. Spruce pollen is practically absent in 184 zone C-l at Allenberg bog. It also occurs in low percentages in the C-1 at the other Sites included in this study. In zone B the mean grain size is 91 mu, but 87 per- cent of the grains are still less than 100 mu in length. Progressively lower in the bog, the mean size is larger and a maximum of 101 mu is present at 14.750 m near the bottom of zone A. At this level 53 percent of the grains are greater than 100 mu. At 15.175 m in zone T the mean size decreases slightly to 98 mu but 43 percent of the sample is greater than 100 mu. In the lower spectra the maximum wingtip to wingtip length found was 120 mu. Whether such grains belong to.g;eee rubens cannot be determined, but occasional grains of.g. glauca attain this size. .2192fi- Early work on size-frequency distributions in_g;eee pollen suggested the feasibility of species identifi- cation on this basis. Cain (1940), for example, found three modes in a size—frequency curve of fossil pine pollen ex- tracted from the Spartanburg buried soils on the Piedmont of western South Carolina. He related the smallest mode to .g. banksiana and the larger ones to_g. glabra and_g. rigida or g. palustri . In a study of pine pollen from sediments in a southeastern Michigan lake, Cain and Cain (1948) found bimodal and trimodal distributions which they considered evi- dence forthe occurrence of g. banksiana, g. resinosa and _P. str obus . 185 The presence of a verrucose furrow (Ueno, 1958) al- 1ows pollen from the only member of the subg. Haploxylon in eastern NOrth America, Pinus strobee, to be tabulated sepa- rately from the smooth furrowed pollen of the subg. Diploxy- -lee. Size measurements of Diploxylon pine pollen can thus be directed at identifying the occurrence of species ex- clusive of.g. strobus. Whitehead (1960) has studied pollen of all_g;eee species occurring in eastern North America and has published the most complete Size-frequency data available. Our macer- ation techniques were Similar, so his data Should be com- parable to mine. Whitehead found that, of the three Diploxy- lon pines expected in western New York pollen profiles on the basis of modern distribution patterns,.g. banksiane, g. resinosa and_g. rigida, the last named has the largest grains with a mean size of 44.95 mu based on 9 collections. .§l£E§ benksiane and_g. resinosa pollen grains are smaller. The mean site of g. banksiana based on 24 different samples was 37.01 mu, while the latter, based on half as many, was 40.11 mu. Whitehead concludes, "it is . . . doubtful if one could separate grains 0f.§1£2§ banksiana and-g. resinosa in size- frequency analysis even though the means differ by 3.10 mu . . . _[and] . . . a size-frequency curve for fossil grains to which both species contributed would be perfectly unimodal" (p. 772). His measurements are of internal body diameter. 186 My own measurements of fossil Diploxylon pine pollen from Allenberg bog are graphed in Fig. 6. Since I measured the external body diameter, 2 mu should be subtracted from my data to obtain figures which are equivalent to those of Whitehead (gelg.). In zone B the modal class at 42.5 mu minus the 2 mu correction factor gives a figure which com- pares well with the mean size of Pinus resinosa pollen as de- termined by Whitehead. In the upper two-thirds of zone A the mode for the Allenberg data shifts to 37 mu (39 mu) sug- gesting dominance by_g. banksiana. A fairly prominent shoulder at 40.5 mu (42.5 mu) corresponding to the mode in the B zone also occurs in these graphs. In the lower third of zone A the mode again shifts to the larger Size class. NOne of my size-frequency graphs is strictly bimodal, but the correlation in size implies the presence of both species in zones A and B. In zone T, two modes occur, one at 37 mu (39 mu) and another at 42 mu (44 mu). The lesser corresponds to the two upper A zone modes and presumably reflects the presence of .ELBEE banksiana, while the greater has no precise counterpart elsewhere in Fig. 6. Since the T zone was apparently a time of low pollen delivery to the basin by the regional vegetation, the probability of finding far-travelled pollen types in T zone spectra is greater than elsewhere in the profile be- cause they would not have been masked by nearby pollen pro- duCers. Perhaps the higher frequency of larger grains is 187 FIGURE 6. Size Frequtnchroph: of Pinus Diploxylon'l’ollen From Allenberg log " /‘/P\ Using Measurement: of External Body Diameter _ I Zone \-—_—-\ "'32 _\ A lone: Upper l/3 — 0' 1., PERCENT OF SAMPLE \\ A loneIMlee Ill '- A lone: lover I]! n- 96 ' \//\ f 45 MlCIONS 188 related to input by one of several possible pine Species, such as g. rigida, which occurred at some distance south of the glacial boundary. However, no clear evidence of the presence of_g. rigida is shown in the pre-C zone samples graphed in Fig. 6. .eenesee Valley Peee_W6rks This site occurs in an area mapped by Connally (1964) as part of the pre-Kent Olean moraine. Similarly, MacClintock and Apfel (1944) and Muller (1960) place the Kent terminal moraine north of the locality. The peat, which has accumu- lated in what seems to have been originally a shallow 10 acre lake in a pitted valley train, is being actively mined by the owner, Paul Button of Belmont, New York. The site is located in Allegany County in the Town of Amity, 2.6 mi northeast of Belmont on the north Side of NYS 244 about 0.3 mi west of Baker Valley Rd at 420 15' 10" N Lat and 780 59' 37" W Long. The surface of the peat deposit lies near the 1620 ft contour line. The peat works is shown as an area de- void of vegetation, but with three small ponds that mark the periphery of the peat deposit, on the northwest quarter of the wellsville North quadrangle. The basin occurs on a flat terrace about 100 ft above Phillips Creek which flows to the southwest and empties 3 mi downstream into the Genesee River. Local relief is about 350 ft and many of the surrounding hills reach 2000 ft in elevation. 189 The peat deposit was wooded when Button purchased it in 1951 and at this time no outlet or inlet existed. He removed the trees and built a dike at the west end of the basin to flood it for use as a trout pond. During the en- suing years the stumps and root mat became freed from the underlying peat and Button decided to drain the pond and be- gin to excavate the peat. The surface was bulldozed clear, a channel was cut through the drift at the west end to facili- tate drainage and the peat along the south rim of the basin was removed and dragged up onto the land for drying (eee Plate 2, Fig. B). Mining has continued along this edge. The peat after being screened is sold in nearby cities for use in gardening. According to Button the bog surface was covered with a forest of Acer rubrum, Betula alleqhaniensis, Pinus strobus and_Tsuga canadensis. At the basin edge Prunus_eerotina and Ulmus Sp. occurred. Apparently Picea mariana and Larix laricina were absent. In a somewhat open area near the east end, cranberries (either Vaccinium macrocarpon or y. oxycoc- cus) grew and Arisaema sp. and Cypripedium acaule and_§. ‘ealceolus were mentioned as noteworthy for their abundance on the forested mat. At the present time the site is entirely surrounded by fields or secondary forests. To the north and contiguous with the basin margin is a highly disturbed forest remnant, long cut over and now dominated by trees of small diameter. 190 The following species were noted in the summer of 1967: Acer rubrum, Crataegus sp., Pinus strobus, Popelus grandi- dentata, Prunus pennsylvanica, g. serotina,_Qgercus alba and Tsuga canadensis. Cultivated land occurs on the east and south sides and a plantation of small conifers in an abandoned field is located immediately to the west of the basin. In general a large percentage of the surrounding hill tOps are forested but valley floors and lower slopes are usually under cultivation. Oak-rich forests are more abundant in this area than to the west in Cattaraugus and southern Erie Counties. Sediment Stratigraphy The deposit was sampled on September 12, 1966 near the west end of the basin at a point where Button said that the maximum depth occurred. The Hiller borer was used from the surface to 3.25 m. The Livingstone sampler equipped with a 1 in diameter barrel was employed beyond this depth. The stratigraphy at the sampling point is: 0.00-0.50 m : peat, undifferentiated, humified; dark brown, no samples taken because of disturbance; peat, undifferentiated, mostly coarse near top, finer at bottom, well com— pacted, Meesia trifaria layers at 2.63 and 2.83 m: dark brown and humified at top, reddish brown commencing at 0.75 m: peat, mostly sedge leaf debris, with Meesia triferia layer at 3.18 m; light brown; gyttja: brown; gyttja, with Silt and clay; 0.50-3.00 m 3000-3e25 m 3.25-3.35 m 3.35-4.16 m 191 4.16-4.80 m silty-clay, compact, with shale frag- ments and quartz sand and granules throughout but most abundant near bottom, layer of gastropod shells at 4.60 m, small vivianite nodules from 4.70 m, mostly greenish gray: further sampling not possible because of the compactness of the sediment. 4.80- m Pollen Stratigraphy The pollen zonation at the Genesee Valley Peat Works is basically similar to that found in the three bog deposits discussed previously. However, several important differences are apparent: the tripartite C zone does not occur, pine pollen is predominant over an exceptionally broad interval and very high NAP percentages are associated with rather low percentages of spruce pollen in the sediments beneath zone B. These divergences from the basic pattern have made the placement of zone boundaries difficult and the two shown on Diagram 9 have been placed with question. The precise age of the Clean drift on which the present Site is located is unknown, but it generally is con- sidered to be older than the Kent glaciation, dated at 23,250 B.P. near Cleveland, Ohio (White, 1968), which in turn predates the deposition of the Valley Heads moraine. The temporal equivalence of the A and B zones at the Genesee Valley site and similar zones elsewhere in southwestern New York cannot be exact because of the differences in the times of basin origin. Radiocarbon dates for all the boundaries 192 are needed to determine the precise relationship between the Genesee Valley profile and the others. .§2£§.§° Spruce pollen reaches a maximum of 29 per- cent in one level near the bottom of the zone, but in general its percentages are half what they are at the other Sites in the study area. Spruce gradually decreases in abundance up- ward and is replaced mainly bY.£l£2§- About 5 percent_g;eee Diploxylon occurs;_g;eee Haploxylon pollen was only sparsely represented. Also present in A zone spectra is 2 percent .EEAQE pollen. Pollen from broadleaf deciduous trees consistently occurs in the A zone. Fraxinus and Quercus each account for about 8 percent of the sum, and lesser, but nonetheless sub- stantial, percentages of Betula and Carpinus-Ostrye also were found. In addition Acer, Carya, Corylus, Fegus, Juglans, _glmee and geege pollen occur, but in very low percentages. Typically, these taxa amount to 15-20 percent of the total. The most unusual aspect of the A zone is the high percentages of NAP. In similar stratigraphic positions at the other Sites, total NAP reached only 10 percent, but here it uniformly accounts for over 30 percent of the sum and reaches 48 percent in the bottom spectrum. Near the A/B transition between 3.45 and 3.19 m, NAP decreases from 33 to 5 percent. Pollen from a variety of herbs was identified throughout zone A. Cyperaceae, Gramineae, high spine Com- positae and Salix have by far the highest percentages and 193 total grass and sedge pollen varies from 25 to 30 percent. Lesser amounts of Alnus, Ambrosia, Artemisia, Cheno-Ams, Myrica, Plantago, Rosaceae, Thalictrum and Umbelliferae oc— cur and, of the minor types listed in Diagram 9 and Appendix H, the most significant ecologically are grains similar to Emgtrum sp.,QEyee spp.. and Saxifrage spp. which were found in spectra below 4.5 m. The number of pollen and spores per ml of sediment does not help to further define the A zone. Below 3.19 m the grain number varies from 42,000 per m1 at 4.8 m to 85,000 at 3.45 m, with most spectra having about 50,000 grains. The increase in the number of pollen and spores upward is not great enough to overcome the decrease in spruce percentages, so similar curves are obtained whether the data are plotted on an absolute basis or not. Zone B. The A/B zone boundary was located at the middle of the increase in the_g;eee total curve. Greater Pinus percentages are compensated for by a decrease in_g;eee and nonarboreal pollen. Both_§;eee Haploxylon and_g;eee Diploxylon grains are present and, while the former continues to occur in abundance upward, the latter falls to less than 1 percent near the middle of the zone and stays at this level until it drops completely out of the counts in the C-1. In some spectra, total Pinus accounts for 70 percent of the sum. .epiee percentages reach a peak just below the middle of zone B. 194 High percentages of Quercus pollen which occur with pine in zone B at the other sites are not present until near the end of the zone. In the bottom half Quercus was found in slightly lower percentages than were present throughout zone A, but above the middle of zone B Quercus increases from 7 to about 15 percent. The B/C-l boundary was placed at the middle of the interval where total_gieee percentages decrease and_geege percentages reciprocally increase. The abrupt rise in the .geege curve marks the beginning of zone C-l at this Site as well as at the others in southwestern New York. A slight in- crease in percentages of_g;mee and Carpinus-Ostrye is ap- parent near the end of zone B, but percentages of other AP types remain fairly constant across the zone. NAP percentages in zone B are about one-fourth of what they were in lower spectra. Cyperaceae and eeggg pollen are the two most abundant types, but it is likely that these were produced by the local rather than the re- gional vegetation. Intrabasinal succession is well-defined by peak pollen frequencies of bog and lake indicator species en- countered at various levels in the sediments, which by them- selves are evidence for such change. Pollen from Sagittaria and Sparganium, two shallow water, near-strand aquatics, oc- cur in the lowest spectrum and upward for over 1.5 m. They imply that the water was probably too deep during most of 195 this interval to permit other aquatics to grow near the sampling point. Somewhat higher in the sediments,_§otamogeton makes its first appearance and later Brasenia occurs in abundance. These are rooted, open water aquatics which generally grow in shallow ponds. Both have their peak frequencies higher in the section, near the level where Sagittaria and Sparganium drop out of the counts. Potentilla palustris and Potomogeton first appear together, but the peak percentage of the former is slightly above that of the latter, a situation perhaps caused by the occurrence of the Potentille at the leading edge of an ad- vancing bog mat. High percentages of Cyperaceae pollen oc- cur in these levels and in those immediately above, at which point both Potentilla and Potamggeton have nearly dropped out of the counts. The Cyperaceae high occurs precisely between the last occurrence of Potentilla palustris and the first occurrence of Ericaceae pollen. In absolute numbers of pollen per m1 of sediment, sedge pollen is more than twice as abundant in this part of the profile than in the A zone. Within this interval occur three separate horizons of the moss Meesia trifaria, a predominantly boreal forest species with disjunct stations throughout the lake states. It grows in bogs and swampy woods and is often found in somewhat calcareous habitats. 196 Upward, Cyperaceae percentages decrease and per- centages of Ericaceae and Polypodiaceae rise. Ericaceae reach a peak near the top of the profile and, slightly above this, maximum Osmundaceae percentages occur. These changes record the presence of several distinct plant communities which probably occurred in response large- ly to the degree of basin infilling. In a developmental se- quence open water, sedge mat, ericaceous shrub heath and lastly bog forest are the main ones indicated. The bog forest is the least clearly defined, but evidence of its presence is afforded by the high percentages of Osmundaceae spores similar to those produced by Osmunda cinnamomea and _Q. regalis, species which are characteristic of this habitat. Zone C-l. The entire C zone is truncated and no subdivisions,as they are defined in the previous profiles, can be discerned. The_Teege and-gegee curves suggest that only the C—1 is present. The two-step-Teege curve is un- paralleled elsewhere in the study area. High_g;eee per- centages persist well above the B/C-l boundary and_g;eee Haploxylon pollen is the predominant type present. Between 0.5 and l m_g;pee drops below 10 percent of the total. Pollen from deciduous tree species show a gradual increase in the upper C-l spectra. .Aeeg, Betula,.gepye, Fagus, Fraxinus, Juglans, Tilia and Ulmus have higher per- centages at the end of the zone than at the beginning. The temporary decrease in the Betula, Faggs and Quercus curves 197 at 1.41 m is associated with a reciprocal increase in.£l£2§- This probably reflects a short term change during which -g;eee became more abundant locally perhaps due to disturbance. Total NAP percentages are low and account for less than 2 percent of the sum in all C-l spectra except at 0.67 m where Nemopanthus, Rhus, Salix and Viburnum pollen, probably from species growingcnr the bog mat, contributed over 12 percent to the sum. INTERPRETATION ZONE T Among the sites included in this study, pre-A zone spectra with total NAP amounting to 50 percent or more of the sum occur only at Allenberg bog, although the lowest spectra in the two Valley Heads bogs contain enough non- arboreal pollen to suggest that they are transitional be- tween the NAP-rich T zone and the comparatively NAP-poor A zone. The large nonarboreal pollen content in what I have called the A zone at the Genesee Valley Peat Works site is difficult to interpret and will be discussed under the head- ing, Zone A. T zones, or Herb Pollen Zones as they have come to be called more recently, have been found in various parts of glaciated eastern North America (Sirkin, 1967), Michigan (Anderson, 1954) and Minnesota (Cushing, 1967). High NAP frequencies in basal lake sediments were overlooked by North American pollen analysts until Deevey (1951) demonstrated abundant sedge and grass pollen in clays from several lakes in northern Maine. A band of marly sedi— ment within the clay contained lower NAP percentages and greater relative numbers of spruce and birch pollen than the clay above or below. The similarity in both pollen and 198 199 sediment stratigraphy between these changes and those associ- ated with the European Younger ngee--Aller¢d--old r Egyee sequence indicates that the marl was deposited during the warmer Two Creeks interval. High NAP percentages above the marl were evidence of tundra vegetation occurring in re- sponse to the probable colder climate that prevailed during the subsequent Valders glaciation; those below presumably recorded a similar type of vegetation developed during some phase of the Port Huron (Mankato) glaciation. In southern Connecticut lake sediments determined by radiocarbon analysis to predate Two Creeks time, Leopold (1956b; Leopold and Scott, 1958) found similar fluctuations in pollen percentages. The youngest of these so-called "Pro-Durham" oscillations (T-3) is characterized by abundant herb and birch pollen. It succeeds spectra with low herb and relatively high Spruce percentages (T-2), which in turn overlies sediments containing high herb pollen percentages (T-l). These fluctuations were correlated with the European Older ngee--B¢lling--Oldest.ngee sequence (Deevey, 1958) and were interpreted (Leopold, 1956b) as evidence of birch park-tundra developed during a pause or readvancement of the Port Huron ice (T-3), preceded by spruce park-tundra repre- senting a period of climatic amelioration (T-2) which fol- lowed another interval of tundra vegetation (T-l). Zone A spectra from southern New England have been internally divided into subzones, the "Durham" oscillations, 200 interpreted as responses to the Two Creeks--Valders climatic changes. These will be discussed later. No "Pre-Durham" oscillations are present in the T zone at Allenberg bog, but the pollen assemblage compares fairly well with the T-3 from New England. Davis (1967a) has contrasted the basal herb zone pollen assemblages from vari- ous sites in New England, Wisconsin and Minnesota. As— semblages from Allenberg bog have much in common with com- parable spectra from both of these regions, but in general, New England Sites have less spruce and more sedge pollen than occurs in the Allenberg bog T zone. At the Minnesota and Wisconsin locations more spruce but about the same per- centage of sedge pollen occurs. Similar herb-rich Spectra have been reported from only two sites in upstate New York: Crusoe Lake near Syra- cuse in north-central New York (Cox and Lewis, 1965) and .Kernochan bog southwest of the Catskill Mountains near the Pennsylvania border (Stingelin, 1965). Erratic fluctuations of tree and herb pollen percentages in the basal gravelly clay at the former Site and the unknown age of the sediments make interpretation difficult, but the authors tentatively suggest correlation with Deevey's Maine zonation. From 15 to 45 percent NAP occurs in a 2.5 m section of silty clay differentiated as zone T at Kernochan bog. The entire inter— val is interpreted as evidence of a period during which the vegetation near this site was taiga-like. It is of interest 201 that Quercus pollen is almost completely absent from the bottom 1.5 m of the section. In recent years pollen diagrams have been most ef- fectively interpreted by relating the current pollen rain from regions of known vegetation to fossil pollen assemblages. If fossil and modern pollen assemblages are Similar, a high probabilily exists that the vegetation producing them was also Similar and, as an extension of this, it can be further reasoned that the climate controlling the vegetation was alike at both points in time. The pollen content of upper- most lake sediments, moss polsters and other superficial pollen traps has been determined at many localities across northern Nerth America, but the pollen rain of much of this region still remains imperfectly known. Nevertheless, Davis' (1967a) use of the analogue technique has produced several informative correlations. NOne of the surface counts she illustrates are ex- actly like the Allenberg bog assemblage, although some spectra from the tundra and boreal forest regions of Laborador are Similar. Of the sites in arctic and boreal Canada studied by Ritchie and Lichti—Federovich (1967), the pollen rain at Fort Churchill, Manitoba in the forest-tundra ecotone is per- haps the closest to the Allenberg bog T zone assemblage. Major pollen types found at the surface at Fort Churchill include: .21222"13 percent,_g;eee--25 percent, Betula--ll percent, Gramineae--6 percent and Cyperaceae--24 percent. 202 Lesser amounts of Larix, Alnus, Salix, Myrica and pollen from a number of herbaceous species were also found. At. this particular site only 0.1 percent Artemisia pollen occurs, but at other sampling localities in the same vegetation type up to 17 percent was counted. Unfortunately, the modern pollen rain in the tundra and tundra-forest ecotone, as it is currently known, is too Similar to permit differentiation of these vegetation types in the fossil record (Davis, 1967a). Hopefully, as more data become available, better correlations should be possible. The occurrence of an open, park-tundra vegetation surrounding Allenberg bog during the deposition of zone T might be questioned in light of the nearly equal represen- tation of AP and NAP. However, the presence of 20 percent .gieee and 10 percent_gieee pollen indicates that species be— longing to these genera grew within pollen dispersal dis- tance but not necessarily close to the basin. Ritchie and Lichti-Federovich (1967) report finding 4 to 35 percent spruce pollen in various parts of the subarctic where spruce trees occur in widely scattered stands surrounded by heaths and dwarf-birch tundra. In the high-arctic, still farther away from the region of provenance, less than 3 percent was encountered. Similarly, Hafsten (1961) working in the south- western U.S., found about 10 percent_gieee pollen 150 mi from its nearest source, and Ritchie and Lichti—Federovich (1967) report over 10 percent pine pollen at the surface in 203 the three high-arctic sites they investigated. These find- ings are evidence that a considerable quantity of pine and spruce pollen is normally blown into areas where the parent plants either do not occur or are of patchy distribution. The presence of other tree pollen types in zone T can be similarly explained. Pollen contributed to a given part of a profile is derived from both nearby and distant sources, but because the absolute pollen frequency data indicate a low delivery rate of pollen and spores to the basin during the T zone interval, the area contributing to the pollen rain may have been much larger at this time than during any subsequent period. Fewer than 1000 grains/ cm2 / year were deposited in zone T at Rogers Lake, Connecticut in sediments older than 12,000 B.P. (Davis, 1967b). The absence of radiocarbon dates at Allenberg bog makes similar calculations impossible, but the number of grains/ml of sediment is nearly equal in pre-A zone levels at both sites. With a relatively low number of grains reaching the basin every year, the few grains that are occasionally shed in the air by local entomophilous species and those derived from distant sources would have an increased chance of being expressed in the counts. This iS borne out by data collected by Ritchie and Lichti—Federovich (1967). They found about 1400 grains/ cmz/ year being deposited at Fort Churchill in the region which provides a close modern analogue to the 204 Allenberg bog T zone fossil pollen assemblage. Pollen from both entomophilous taxa and species which grow many hundreds of miles away are represented in their counts. With this phenomenon in mind, species represented in the T zone at Allenberg bog can be discussed. Among the arboreal pollen types identified in zone T, those of temperate thermophilic species are incongruous in a pollen assemblage which otherwise indicates the vege- tation surrounding the site was similar to that in the boreal forest-tundra ecotone. The pollen types in question include .ggexinge nigra (from 3 to 5 percent), Quercus (from 3 to 8 percent) and a few grains of Carpinus—Ostrya, Carye, Corylus, Fraxinus americana and/or_§. pennsylvanica,_ggglans cinerea and.g;gee. Such pollen frequently occurs out of place in lateglacial T and A zone spectra throughout glaciated eastern Nbrth America. Its presence has been explained by claiming redeposition from older sediments (Andersen, 1954) and long- distance wind transport (Deevey, 1951). A third alternative, that species producing these types actually grew nearby, seems less likely at least during the deposition of T or Herb Pollen Zones. I know of no source near Allenberg bog for rebedded grains. Furthermore, pollen in the T zone at this site is extremely well-preserved and shows little evidence of abrasion. A strong case can be made, however, for wind transport of exotic grains using recently published data of 205 Ritchie and Lichti—Federovich (1967). They measured the. pollen rain at Fort Churchill with a Hirst spore trap for six days in early May, a time when local plants had not begun to flower. During each 24 hour period at least one Quercus grain was trapped and a maximum of four grains were collected on one of the Six days. Other types were recovered less fre- quently, but often were more abundant. For example, 17 .gigee grains, 15 Corylus grains and 5 Fraxinus grains (pollen with 3 and 4 colpi were not separated) were trapped in one 24 hour period. All of the exotic pollen types occurring in the Allenberg bog T zone except_gerpinus-Ostrya were found capable of being carried to Fort Churchill from sources 500 to 1000 mi away. Calculated as a percent of the sum of all other pollen and spore types trapped by the Hirst sampler, however, such pollen amounts to only 0.6 percent of the total. Since by comparison higher percentages of Quercus and Fraxinus occur at Allenberg bog, pollen producing trees of these taxa perhaps grew much closer to the site than did species of .gegye, Corylus, Jugleee and_g;me§. The nearness of Allenberg bog to the unglaciated Salamanca reentrant and other ice-free areas further south suggests that black ash and certain oak species may have occupied favorable habitats several tens of miles beyond the terminal moraine. Of the ash species in eastern Nerth America, Fraxinus nigra currently has the most northern distribution, so it seems likely that it would be 206 able to survive closest to the ice front. Similarly, Quercus macrocarpa and_Q._£ep£e, which are found northward to the edge of the boreal forest, are the most probable members of the .1ateglacia1 T zone vegetation near Allenberg bOg. Davis (1958) has pointed out that T zone herb pollen assemblages are mixtures of taxa which have both northern and southern affinities. The Allenberg bog samples yielded pollen from Artemisia, Caryophyllaceae, Rosaceae, Ranunculus, _§e;;§, Cichorioideae and Asteroideae (= high-Spine Compositae .2121), all of which have species in both arctic and temperate regions. Other pollen types identified in T zone sediments are from taxa which are mainly temperate in distribution. At Allenberg bog these include Ambrosia, Chenopodiaceae- Amaranthaceae, Labiatae and umbelliferae. Taxa found ex- clusively in arctic and subarctic regions have not been identified in the basal layers at Allenberg bog. However, the presence of Betula glandulosa is indicated by the size- frequency measurements of birch pollen. This is an arctic- alpine species which today is widespread in the North American subarctic with an extension southward in eastern North America to the Adirondack MOuntainS (Fernald, 1950). Micro- spores of Seleginelie seleginoides were also found in zone T sediments at this site. Currently, this species grows at exposed, calcareous sites in the boreal forest and subarctic across Nerth America as far south as the upper Great Lakes 207 and the mountains of southern New England (Hultén, 1958). The occurrence of this Species, and pollen from other taxa with high light requirements,give support to the interpre— tation of an open vegetation. In the absence of radiocarbon dates, it cannot be determined if the semi-treeless landscape which occurred adjacent to Allenberg bog was a successional stage transi- tional to a spruce forest or whether it was a climatically controlled group of communities equivalent to several of the many expressions of tundra vegetation. In southern New England where tundra persisted several millenia prior to 12,000 B.P. (Davis, 1967b), the treeless interval probably represented a period during which the climate was too severe to allow the deve10pment of a spruce forest near the site. Trees may also have been absent because they had not migrated to the region. In view of the length of time involved, how- ever, climatic control is the most acceptable hypothesis, even though both factors may have interacted. The Allenberg bog T zone pollen assemblage seems clearly to imply the occurrence of park-tundra vegetation at some time prior to the development of A zone spruce forests in the region. Size-frequency measurements indicate that jfieee glauca was the main spruce Species present, although _3. mariana undoubtedly also occurred. Evidence suggests that spruce trees were either sparsely scattered across the landscape or occurred some tens of miles away. The presence 208 of Pinus banksiana and/or_g. resinosa is also indicated, but these species may have grown much further south of the site, perhaps even a distance of several hundred miles. Plant com- munities rich in sedges, grasses and heliophytic herbs must have dominated much of the surrounding region. Few truly tundra species were apparently present, although a dwarf arctic birch, Betule glandulosa, and Selaginella selaginoides, both of which are northern species, are represented in the deposit. The vegetation may also have been prairie-like. The occurrence of Ambrosia and Cheno-Am pollen supports this conclusion but the maxima in Artemisia, Cyperaceae and Gramineae curves, which also would support such an inference if prairie species could be identified, can be taken as evi- dence of either prairie or park-tundra vegetation. Since both ragweed and Chemo-Am pollen could have been wind carried to the basin from a distant source, however, the park-tundra interpretation remains the strongest. ZONE A One of the most consistent stratigraphic features in basal lake and bog sediments across glaciated eastern North America is a zone in which spruce pollen accounts for 30 to 70 percent of the sum. The occurrence in New York State of a spruce zone, or A zone as Deevey (1939) named it, was early established by MCCulloch (1939) from a bog near 209 Syracuse. Later workers have confirmed its presence and have since found similar zones in sedimentary basins through— out central and eastern New York (Cox, 1959; Durkee, 1960). An A zone was also found in each of the four southwestern New York bogs I studied, but Since this zone varies from site to site, it will be discussed separately in reference to each of the localities. Genesee Valley Peat Works Low spruce percentages and the high values of total NAP set the Genesee Valley Peat Works A zone apart from the others. At this locality most A zone spectra contain less than 25 percent spruce pollen and the maximum value of 29 percent was reached in only one sample. By comparison, from 40 to 65 percent spruce was counted in equivalent strati- graphic positions at the three other sites. Associated with spruce in the Peat Works sediments are unusually high per- centages of Cyperaceae, Gramineae, high-spine Compositae (= Asterioideae.B;E;), Salix and other NAP types. Replotting A zone spectra on an absolute basis does not greatly change the form of the curves. Although the overall pollen strati- graphy of the entire diagram is basically similar to the other profiles from the region, the unusual character of the spruce zone makes its meaning difficult to determine. In the para- graphs that follow two alternative explanations are proposed and evaluated in light of the evidence at hand. 210 Taken at face value, the nearly equal representation of AP and NAP throughout zone A at the Genesee Valley site, but particularly near the bottom where_g;eee percentages are low, implies that the regional vegetation was open and perhaps similar to that which occurred in Cattaraugus County during the deposition of zone T at Allenberg bog. Widely spaced spruce stands interspersed with herb communities is one possible interpretation of the data based on the assumption that A zone pollen assemblages from the Peat Works fully represented the regional vegetation. The slightly larger spruce percentages in the lower part of the zone indicate that spruce trees were more abundant around the site early in the depositional history of the basin. Pine pollen in- creases at the expense of Spruce higher in zone A suggesting that pines became more frequent in the surrounding vege- tation with the passage of time. _g;eee banksiana and/or g. resinosa seem to have been the only species present because Haploxylon grains are only sparsely represented. Since much of the pine pollen at the lowest levels could have been blown in from a distance, the actual abundance of pine in the vegetation around the site must remain conjectural for this part of the profile. An open spruce-pine woodland interpretation is hard to accept, however, because there is no evidence in the pro- file that a more closed forest was present later in the history of the region adjacent to the basin. Such a change 211 would be expected to have occurred on the basis of pollen studies elsewhere. A climate too cold to permit the develop- ment of a spruce forest may have been the controlling factor in the lower part of zone A where pollen tentatively identi- fied as.2£yee spp., ggpetrum niqrum and Saxifraga spp. was found. The proximity of an ice front may have induced con- ditions favorable to tundra communities and might at-the same time have kept_§;eee from becoming more dominant in the landscape. High frequencies of sedge, grass and forb pollen would be expected in vegetation of this type. A warming trend and other factors that aided the colonization of the region bY.2££E§ banksiana and/or_§. resinosa may have oc- curred in the upper 1.25 m of the A zone. The NAP contri- bution remained more or less constant during this interval implying the persistence of an open vegetation. Since the Peat works is situated on the Tazewell or pre-Tazewell Olean drift (Muller, 1965), the lower part of the profile, if complete, may antedate most other pollen records in eastern North America. The other sites I investi— gated in southwestern New York had their inception during sabsequent glacial advances and therefore provide no parallel data. However, in eastern North America several additional deposits on Clean drift have been studied. Highland Lake southwest of the Catskill mountains (Cox, 1959) is truncated basally, as is the Cranberry bog profile from eastern Pennsylvania (fide Stingelin, 1965; Gehris, 1965), but the 212 long sedimentary record at Kernochan bog, located near High- land Lake about 150 mi east of central Allegany County, con- tains a sequence that is similar to the Genesee Valley pro— file. The shape of the spruce silhouettes at both Sites is almost identical, but at Kernochan bog, below the point where spruce reached its A zone maximum of 35 percent, there is a long interval of lower-spruce percentages and high_g;eee and NAP values Which are absent from the Genesee Valley site. (Above this level the pine and spruce curves at both sites are essentially identical, but at Kernochan bog NAP values are about half those at the Genesee Valley site. The parallel nature of the spruce and pine curves in the Kernochan bog and the Genesee Valley profiles indicates that the sequence may have regional significance. This raises the interesting possibility that, for an unknown period following the Tazewell (?) glaciation and perhaps contempor- aneous with subsequent ice advances, a more or less open pine-spruce woodland existed in the Allegheny Upland of southern New York. Pollen analyses of sediments below a radiocarbon date of 13,630 i 230 years B.P. (Y-479: Martin, 1958b) in unglaciated southeastern Pennsylvania have pro- duced spectra somewhat similar to A zone samples at the Genesee Valley Peat werks, but without C—l4 dates at this site the contemporaneity of the deposits cannot be proved. The main differences at the Pennsylvania locality include larger total NAP percentages (50 to 75 percent) and a weaker 213 expression of_g;eee (e_. 5 percent); Pinus comprises from 15 to 25 percent of the total. In light of the uncertain age of the various drift sheets east of the Salamanca reentrant and of the Genesee Valley and Kernochan bog pollen profiles and in the absence of radiocarbon dated profiles between these two localities, the open spruce-pine woodland hypothe- sis is presented as one of several explanations of existing data. The vegetation inferences in the preceding discussion are based in part on the assumption that the fossil Spectra represented the regional pollen rain. An alternate interpre- tation is possible, however, if the NAP was derived mainly from a source near the basin. In this case the regional vegetation contributing to A zone sediments was probably not an open woodland. ,As a correction for local over— representation, a portion of the data were recalculated using the total number of arboreal pollen at a given level for the percentage base rather than the sum of AP and NAP. The new curves still Show the same trends because the total NAP contribution was more or less constant throughout the inter- val. However, what is different is the percentage of the main arboreal components in the counts. In such a calcu- lation spruce attains a maximum of nearly 50 percent and the overall transition from the spruce dominated basal sediments to pine dominated sediments above is not unlike the A/B zone transition at the other sites in western New York with the 214 exception that it is more gradual. The new data indicate that the tree cover was more dense and was perhaps similar in overall aspect to the boreal forest of northern Ontario and central Quebec. If this were so, the previous hypothesis would have to be modified to accomodate the new vegetation interpretation.‘ Pending C—l4 dating, therefore, the possible existence of a denser pine-spruce forest on the glaciated Allegheny Upland of southern New York during pre-Cary times must also be considered. There are several lines of evidence pointing toward local over-representation. The slopes above and leading to the center of the basin are gentle and the depression itself is large and relatively shallow. Therefore, an abundance of habitats for marsh plants could have existed around the margin of the basin during its early history. The presence of Sagittaria and_§parqanium pollen throughout zone A is confirmation that marshy shallows existed in or near the basin. Both genera are not represented in the counts higher in the deposit, probably because littoral habitats were eliminated by the development of a bog mat. Other parts of the valley floor on which the depression is located may have supported marsh communities but bedrock highs are abundant in the area and well—drained upland sites must also have been a regular feature of the landscape. An incomplete knowledge of the current pollen rain in many sections of northern North America again hampers the 215 search for a modern analogue of the A zone vegetation near the Peat werks. If pollen spectra from this zone represent the regional pollen rain, they cannot be matched with any of the surface samples reviewed by Davis (1967a), because modern pollen assemblages in which Carpinus—Ostrva, Fraxinus, Quercus and-gimee pollen are minor but significant components of spectra otherwise dominated by Picea, Pinus and NAP are unknown. The anomalous occurrence of pollen from these thermophilic taxa is not without precedent. I have already discussed the possibility of their having been derived from a southern source by wind transport. The spectra may also be contaminated by redeposited pollen. I have not demonstrated that rebedding has occurred, since a source of older pollen is unknown in the area, but the presence of_gegee grains in the silty clay at the bottom of the deposit and their absence from the more organic sedi- ment immediately above, suggests that pollen eroded from the surrounding drift may have been carried into the basin along with the inorganic sediments. Andersen (1954) has asserted that the optimum time for redeposition is during an ice ad- vance when frost activity would be continuously exposing potential pollen-bearing deposits. This is also the time of maximum inorganic sedimentation because the basin would be relatively free of living organisms. Thus the occurrence ofpgegee grains together with pollen of_eree, Empetrgm and Saxifraga might be explained by claiming a climate favorable 216 to solifluctiOn. If there were some way to determine ob- jectively what part of the spectra were composed of rede- posited and extra-regional pollen, subtractions could be made, and the modified fossil spectra might compare more favorably with one or more of the surface pollen assemblages now known. Allenberg Bog At Allenberg bog the A zone overlies an interval which I interpret as a record of a more or less treeless landscape with herb communities covering much of the region surrounding the basin. Above this, a rapid increase in spruce percentages occurs and high spruce values are main- tained for about 2 m of sediment. Fluctuations in the rela- tive numbers of subdominant AP types, however, allow the interval to be divided into several subzones that are reminiscent of those reported from southern New England by Leopold (1956b; Leopold and Scott, 1958) and Davis (1958). The correspondence of the Subzones in the two areas is not exact and, to my mind, the percentage changes in the Allen- berg bog profile are not evidence of a modification in forest composition or density induced by advance and with- drawal of an ice sheet as the New England workers have claimed for their area. Davis (1967b; Davis and Deevey, 1964) has demonstrated that at Rogers Lake, Connecticut, the fluctuations in relative pollen frequency which have been 217 taken as evidence of vegetation change, can be interpreted differently when the data are replotted on the basis of numbers of grains deposited per unit area per year. Four divisions of the A zone are recognized at sites studied by Davis (1958) in central Massachusetts. The A zone sequence in her Tom Swamp profile is the closest to what I have found at Allenberg bog. The Tom Swamp_g;eee curve shows an increase in the lowest subzone, the A-l, while_g;pee values decrease slightly over the same interval. These changes are also present across a comparable strati- graphic interval in the Allenberg bog profile. However, the high values for Betula and Populus pollen (e_. 10 percent each) which characterize the A-l at the former site are ab- sent from the latter and without other stratigraphic markers, an A-l cannot be as readily defined at Allenberg bog. At Tom Swamp Betula and-gepeiee drop to.5 percent or less of the total in the next highest subzone,=the Ar2, and a pronounced maximum in the relative numbers of_g;eee occurs. Somewhat lower Quercus percentages are found in this subzone than are present in the A-l or in the A-3 above. The spruce peak at Allenberg bog occurs between 14.70 and 14.92 m and, in contrast to the New England locality, is found with maxi- mum percentages of Quercus and Fraxinus BESEE- The Allen- berg Quercus curve shows no important fluctuations below the _g;eee peak, but the maximum A zone percentage of_§raxinus nigra pollen occurs just beneath it. 218 The New England A—3 sszone has somewhat lower gleee percentages and increased values for_g;eee, Quercus and other AP types. At Allenberg bog spectra between 14.22 and 14.70 m can be assigned to subzone A-3. Higher relative per— centages of CarpinustOstrya occur in this interval than else- where in zone A, although the maximum for this pollen type is reached just above the_g;eee peak. Spruce percentages in- crease in the A-4 which at Allenberg bog encompasses sedi- ment between 12.80 and 14-22 m or about the upper two-thirds of zone A. _g;eee values are lower at the middle of the A-4 than they are in the A-3, but the curve of the sum of the three pine categories increases gradually across the upper part of the A-4. In further contrast to the New England diagrams Carpinus-Ostrye, Fraxinus_gigra and Quercus per- centages do not decline in the A-4. The interpretation of the New England A zone sequence was based on the Late Wisconsin glacial chronology as it was known in the mid-1950's (eee Beetham and Niering, 1961; Davis, 1965b). As the climate improved following the re- treat of the Port Huron ice, subzones A-1 and A-2 were de- posited in which Spruce percentages increase and attain a maximumfi The lower relative numbers of_g;eee and the in- creased percentages of_§;eee and deciduous tree pollen types implied further warming during the A-3. Several radiocarbon age determinations permitted correlation of the A-3 with the Two Creeks Interstade, then dated at about 11,000 years B.P. 219 (eee Leopold, 1956b). The return to higher spruce percentages in the Ar4 and drops in oak and pine values were taken as a record of a spruce dominated forest developed during the colder climate which presumably prevailed during the ensuing Valders Stade. The more recent dating of the Two Creeks forest at 11,850 i 140 years B.P. (Broecker and Farrand, 1963) makes its correlation with the New England A—3 doubtful. The similarity between the New England and Allenberg bog profiles may be fortuitous. The lower part of the A zone at Allenberg bog is perhaps not temporally equivalent to spectra with comparable pollen assemblages in the New England diagrams where the zone is broadly contemporaneous with the Valders and Two Creeks events. Accepting 23,250 years B.P. as the age of the Kent drift (White, 1968) upon which Allenberg bog is located, not only the Valders--Two Creeks climatic changes may have influenced the vegetation surrounding the site but also those associated with the pre- ceding Port Huron and Cary glaciations. Records of these events in the pollen profile may be in part preserved be- neath the A zone lower in the incompletely sampled clay de- posit, but the A zone itself seems not to show any well- defined changes that might be related to them. At Houghton bog, 25 mi northeast of the Allenberg site, wood from near the bottom of zone A dated at 11,880 i 730 years B.P. (I—3290) is of Two Creeks age. The sample occurs with_g;eee dominated pollen spectra which are similar to those in the upper part 220 of zone A at Allenberg bog. Between the dated level and the end of the spruce zone at Houghton bog, which encompasses the time of the Valders readvance, changes in pollen percentages similar to those just reviewed from the southern New England profiles do not occur. Therefore, the vegetation in south- western New York State does not seem to have been affected by climatic changes accompanying the Valders readvance. The maximum southward extension of Valders ice was more than 100 mi north of southwestern New York. Valders ice apparently never reached the Lake Ontario basin (Karrow e; _;., 1961) which was then occupied by Glacial Lake Iroquois. Replotting section C of the Allenberg bog profile on an absolute basis permits an alternate interpretation to be made (9;. Diagrams 7 and 8). Variations in numbers of differ- ent pollen types per unit volume through time is meaningful only if the sedimentation rate was more or less constant across the interval being considered. At Rogers Lake, Connecticut, the only site in eastern NOrth America where data on the pre-Valders sedimentation rate has been obtained, Davis and Deevey (1964) reported a constant rate of 0.036 cm per year (later corrected to 0.037; Davis, 1967b) between 14,000 and 10,000 years B.P. In certain parts of a profile deposition rates for each pollen type are potentially the most useful data for assessing vegetation change. If the accumulation rate of the sediment was constant, however, variations in the absolute numbers of pollen types with depth 221 will Show the same trends. It would have been preferable to determine directly the sedimentation rate at Allenberg bog with close interval C—l4 dating but in the absence of the necessary age determinations, the following discussion is based on the assumption that the time taken to accumulate a unit volume of sediment was the same in all divisions of zone A. At Rogers Lake the spruce dominated A zone begins at 12,000 years B.P. and ends about 9,500 years B.P. Quercus pollen is present during the entire interval but 10,500 years ago it fell from about 15 to 5 percent. A Similar but less pronounced change also occurs in the Caprinus—Ostrya and Fraxinus curves at this site. These changes, clearly ex- pressed in the relative frequency diagram, are not maintained when the data are converted to the numbers of pollen and spores accumulating per unit area per year. The pollen in- put from Quercus and other temperate deciduous trees re- mained relatively constant during the entire interval and the maximum and minimum of oak pollen at 11,000 and 10,000 years, respectively, that occur in the percentage diagram no longer exist. Davis (1967b) concludes that percentage change in Quercus is a reflection of increasing deposition rates for coniferous tree pollen 10,000 years ago and is not due to a climatic oscillation correlated with the Aller¢d—- Younger Dryas sequence. 222 The same reasoning can be used to explain fluctu- ations in the percentages of Quercus and_§raxinus nigra near the bottom of zone A at Allenberg bog. In the interval across which the percentages for these taxa are high, the number of grains per m1 of sediment increases (eee Fig. 8). Between 14.540 and 14.425 m the pollen and spore total stabilizes near 200,000 grains per ml where it remains until the end of zone A. The increase is due mainly to a greater number of spruce and pine pollen being deposited in the basin and this indicates an increase in the number of spruce and pine trees in the region. By 14.425 m the maximum attainable density of the spruce-pine forest may have been reached and, signifi- cantly, above this level no important changes are evident in Quercus and.§£exinus niqra percentages. The higher relative pollen frequencies just above the T/A zone boundary may re— flect the openness of the developing spruce forest when fewer numbers of_g;eee pollen relative to Quercus pollen were being deposited. If the input of Quercus remained constant, as was the case according to Diagram 8, an increase in the abso- lute numbers of spruce and pine pollen being deposited would depress the percentage values for Quercus. Allenberg bog upper A zone spectra, with the ex- ception of somewhat higher percentages of Carpinus-Ostrya and Fraxinus, agree fairly well with southern New England spectra from equivalent stratigraphic positions. Davis (1967a) considers the New England fossil pollen assemblages 223 to be similar to surface samples deposited today at about 530 N Lat in northern Quebec in the Nichicun Lake area west of Schefferville. This region has been characterized as an open, park-like woodland in which closed black spruce forests with an admixture of larch are present at wet lowland locali- ties, while black and white spruce with balsam fir occur in open stands interspersed with lichen communities on the better drained upland sites (Terasmae and Mett, 1965). Davis (1967a) suggests that the lower part of the New England A zone represents tundra-forest transitional vegetation which developed into a boreal woodland later in the zone. She further suggests that such a change is evidence of gradual climatic warming. The data from Allenberg bog would seem to fit this interpretation, but it is to be treated as tentative until confirmatory information is obtained from other sites in southwestern New York State. The gradual increase in the number of Quercus grains in the sediment from zone T to the beginning of zone A may - represent northward migration of oaks to positions nearer the basin. The occurrence of from 5 to 7 percent oak pollen throughout the upper part of zone A suggests that oaks were present somewhere within 100 mi or less of the bog. Quercus percentages of this magnitude are not known to be deposited in the boreal forest or in the more open subarctic woodland to the north, which on other evidence are the closest ana- logues of the fossil vegetation. In eastern Nerth America 224 at about 460 N Lat, near the northern distributional limit of the genus, similar percentages occur. But since this is located in the mixed coniferous-deciduous forest of mid- Ontario, significant percentages of Acer, Ulmus and other temperate AP types are present also. Similar percentages of these types are absent from the A zone of Allenberg bog. Less than 1 percent Quercus pollen, calculated using the sum AP as the percentage base, was found by King and Kapp (1963) at the southern edge of the boreal forest north of Georgian Bay. A relatively high representation of temperate tree pollen in existing vegetation dominated by spruce and larch has been found by Janssen (1967) at Myrtle Lake on the Lake Agassiz plain of north-central Minnesota. By comparing an estimate of the regional forest composition derived from the General Land Office Survey notes with pollen deposited on the surface at a number of points along transects at the lake, he showed.gieee and_§e£;§ to have high importance values in the surrounding vegetation, but to be relatively poorly represented in the surface pollen spectra. On the other hand, Fraxinus, Quercus and.g;mee pollen were distinctly over-represented in reference to the regional vegetation. If spruce and larch had a similarly low 'delivery capacity' during lateglacial time, they would be under-represented in pollen profiles, while higher percentages for certain extra- regional deciduous trees with greater 'delivery capacity' 225 would be expected in Spite of the probability that they com- posed only a minor part of the regional vegetation. It would seem, therefore, that the problem is defining the area contributing pollen to a given basin. During the deposition of the lateglacial Spruce zone at Allenberg bog, judging from studies on the current pollen rain, it may have in- cluded the area within a 100 mi radius of the site. Using 100 mu as the dividing point between the smaller pollen of Picea marieee and the larger_g. glauca grains, size-frequency measurements of A zone spruce pollen at Allenberg bog confirm the presence of both species (Fig. 5). .gieee rubens may or may not have been present also. At 14.750 m near the bottom of the zone, the mean size of measured spruce grains was 101 mu; wingtip~ to- wingtip measurements were greater than 100 mu in 53 percent of the sample. Higher in the profile the mean size decreases until near the end of the zone, at 12,925 m, it was 89.2 mu and only 17.6 percent of the measured sample was over 100 mu. This implies a gradual loss of_g. glauca upward and perhaps replacement by_g. mariana. Most of the pine pollen in the Allenberg A zone is the Diploxylon type. The configuration of the modes in the, size-frequency curves for this pollen type (Fig. 6) indicates!‘ that both Pinus banksiana and.g. resinosa contributed to zone A sediments. In view of the similarity in pollen size of these species (Whitehead, 1964), however, it will be 226 necessary to find macrofossils to establish conclusively the presence of either one. The occurrence of about 20 percent of pine pollen throughout the zone definitely establishes that one or both of the Diploxylon pines grew near the basin. This is in contrast to the situation in the western Great Lakes region where significant amounts of pine pollen are not found in the profile until near the end or following the spruce zone (Wright, 1964; 1968b). Judging from present-day habitat preferences, both 3. banksiana and_g. resinosa grew on dry sandy soils, although the former was probably re- stricted to the driest sites. The low relative frequency of Haploxylon pine pollen further implies that_g. strobus was not a part of the regional vegetation, because the relatively small amount of Haploxylon pine pollen present in the counts could have been blown to the basin from afar. The broad size class spread and the occurrence of several modes in the A zone size-frequency curve for Betula indicates that more than one species was present near Allen- berg bog (Fig. 4). The smallest grains, 20 mu or less in size, may have been derived from the arctic-alpine dwarf birch, §° glandulosa. This species was apparently also present in the underlying T zone. The modal classes center- ing near 22 and 24 mu, however, have no exact modern counter- parts among the taxa investigated by Leopold (1956a), which include 9 of the 11 Species native to northeastern Nerth America. Although it could be postulated that extinct 227 species contributed to the modes, it is more likely that the maceration technique or some aspect of the fossilization pro- cess modified the grain size of extant taxa. In the upper part of zone C, for example, where the only contributors to the regional pollen rain were §._£§EE§ and_§. alleqhaniensis, the modal class in the size-frequency curve is smaller than that reported for pollen from herbarium specimens of either species (eele.). On the basis of modern distribution patterns and pollen size-frequency characteristics, specu- lation leads to two additional birches that may have been members of the lateglacial flora near Allenberg bog. One of these,.§. pepulifolia, is a species with small pollen (mode 27 mu in three acetolyzed preparations; epie.). Davis (1958) suggests that it may have occurred in the New England A zone vegetation where it likely occupied disturbed sites. Betula papyrifera is the other expected Species because it now grows mainly in the boreal forest. However, its relatively large pollen (mode 33 mu in one acetolyzed preparation; Leopold, 1956a) does not correspond to any measurements of fossil grains at Allenberg bog. Reviewing briefly the nature of the vegetation in the A zone of Allenberg bog, as it has been interpreted here, the lower third of the zone seems to record the development of a more or less open boreal woodland, similar to that which today occurs in the subarctic of northern Quebec. This appears to have persisted throughout most of the zone 228 because few meaningful changes occur in the A zone above 14.5 mg. The woodland was preceded by a transitional vegetation type in which spruce and pine greatly increased in abundance. These changes may have been in response to a warming trend in the climate, as the A zone overlies an interval of tundra- 1ike vegetation apparently dominated by herbaceous communi- ties with spruce probably infrequent in the entire region. Without information on the duration of the tundra, however, a simple successional change may be represented instead. The density of_g;eee and Pinus in various parts of zone A must remain conjectural until additional surface samples prove that the pollen rain in an open woodland is different from that in a more closed forest. If the landscape con- tributing to the regional pollen rain was incompletely covered by stands of Picea glauca and g. mariana, the latter being more abundant at wetter sites, various non-tree communities dominated by sedges, grasses, Artemisia, other Compositae and additional herbs occupied the openings. An alternate interpretation would have most of the herbs at the basin margin. Alnus and Myrica were certainly present at pond and lake edges and in other wet habitats. fifilifi was also part of the vegetation, but it is not known whether dwarf or shrub species are represented. Pinus banksiana and/or_g. resinosa grew at dry sandy sites in the vicinity, and both apies balsamea and Larix laricina were members of the region- al vegetation, although it is not possible to tell in what 229 proportion they occurred in the forest because their pollen is usually under-represented. Carpinus caroliniana and/or Ostrya virgiana, Fraxinus nigra and Quercus spp. probably occurred within 100 mi of the basin, but perhaps closer. Houghton and Protection Bogs The two sites that remain to be discussed are associ- ated with the Valley Heads moraine. Both have relatively thin A zones in comparison to the long interval of spruce domination at Allenberg bog. In neither of them is there any clear indication of a zone with high NAP percentages. The lowest spectra in each do contain from 15 to 24 percent herb and shrub pollen but this is in association with high values (40 to 50 percent) for_g;eee. The absence of a T zone at both localities may be a sampling deficiency, al- though the samplers were pushed as deeply as possible. Pollen is present in the basal clay at Protection bog, but is absent from similar sediments at Houghton bog. As has been mentioned already, the A zone pollen as— semblages from the Valley Heads bogs compare favorably with the upper A zone spectra from Allenberg bog. The vegetation that presumably produced the assemblages has just been re- viewed and little new information can be added here. In common with the other sites in southwestern New York, Cerpinus—Ostrya, Fraxinus and Quercus pollen are apparent in the A zone of both profiles. Lesser amounts of Carya, 230 Corylus and_glgee pollen also occur. Low relative numbers of_geege pollen first appear in the A zone of the two bogs and also near the beginning of this zone at the Allenberg site. It is probable that hemlock was an extra-regional species at this time because the few grains present could have been wind carried to the site from a distant source. All of these are minor pollen types, however, and the zone is clearly dominated by Spruce and pine. Pinus banksiana and/or_g. resinosa were present. Except for the pollen of the temperate tree Species, the assemblages match the modern pollen rain accumulating today in the open, boreal woodland of subarctic northern Quebec and at certain points to the south within the boreal forest itself. A maximum in the_§e;ee curve occurs near the end of zone A in all four profiles, but is best developed at Pro- tection bog. At this locality, and perhaps at the others as well, this increase may represent an actual increase in the number of balsam fir near the basins. The rapidly declining spruce percentages associated with the fir maximum imply an abrupt and perhaps catastrophic change in the vegetation. If balsam fir was growing suppressed in a Spruce-dominated woodland at this time, deterioration of the spruce overstory might have released the fir seedlings and saplings growing in the understory. The period during which fir thrived must have been relatively short because its pollen drops out of the counts soon after the maximum is reached. At the present 231 time Abies balsamea persists under a dense forest cover but nearly full sunlight is needed for best development (Fowells, 1965). This behavior is in agreement with its known quick response to release. High fir percentages near the end of the spruce zone occur over a wide area in the Nertheast, although the peak is sometimes just within the zone and other times at its and (Cox, 1959; Deevey, 1943; and others). Deevey (1943) has suggested that the high fir values, which often occur with a spruce maximum at the New England sites he studied, may represent a vegetation response to the last glacial advance. As such it would correlate with subzone A-4 discussed pre- viously. Since the Protection bog fir peak occurs in sedi- ments accumulated about 10,500 years B.P. (by extrapolating from the two higher dates at this site assuming a constant sedimentation rate), considerably after the last or Valders readvance, it seems best to view the peak as a successional event. The radiocarbon dated Valley Heads profiles enable time stratigraphic correlations to be made between these sites and others in eastern North America. In New Yerk State itself, few dated pollen spectra comparable in age to the lateglacial A zone assemblage at Houghton bog, the lower part of which has been dated at 11,880 3'730 (I-3290), have geen puinshed. 'I am aware of only one, the King Ferry site in the Finger Lakes region of central New York (Cox, 1959; 232 Brown_;e Deevey.ee(el., 1959). There, spruce wood 11,410 i 410 years old (Y-460; Deevey 2E.él°: 1959) in close associ- ation with a mastodon skeleton was embedded in sediments dominated by spruce and pine pollen. Spruce accounted for over 80 percent of total AP; NAP, unfortunately, was not tallied. The microfossil flora was taken to record the presence of a boreal coniferous forest in central New Yerk at this time (Brown_ie Deevey _E e;., 1959). Spruce wood dated at 12,100_: 400 years B.P. (I—838; Buckley_ee_e;., 1968) from the Lockport site near the Glacial Lake Iroquois strand in central Niagara County is approxi- mately the same age as the lower part of the Houghton bog A zone. I was able to extract pollen from a silty clay lake deposit associated with the organic bed from which the wood was taken and found the assemblage to agree, in a general way, with Houghton bog spectra of the same age. The main difference is the relative frequency of Cyperaceae pollen: 34 percent was found at the former site, while only 4 to 8 percent occurred at the latter. If all of the sedge pollen is considered to have been produced by the upland vegetation, it is likely that a considerably less dense spruce woodland occurred near Lockport than existed 50 mi to the south near Houghton bog. If, on the other hand, habitats near the strand were especially favorable to the aquatic lowland members of the family, local over-representation could ex- plain the difference. The abundant seeds of Eleocharis spp., 233 members of the Cyperaceae with pollen typical of the family (Sears, 1930), in the Lockport organic bed substantiates the case for a near-site origin of much of the 'sedge' pollen. Eleocharis species probably grew along the Lake Iroquois strand in beach pools and along the stream that passed through the Lockport spillway. Since pollen recovered from the lake sediments was carried there by both wind and moving water, near-site aquatic and semi-aquatic species shedding pollen into the water would have an excellent chance of being well represented in the counts. The largest amount of wood in the Lockport deposit confirms the suspected dominance by forest communities. Cones of black spruce were positively identified and perhaps one other species of spruce is represented also. In addi- tion spruce needles, seeds and twigs are exceptionally abundant in the deposit. A single seed of_ge;ee balsamea and pollen of_§e£;§ laricina indicate that these trees occurred also. The rich moss assemblage that was found in the organic deposit permits recognition of several other communities. Rich fens must have been relatively common because both fen and fen edge mosses are frequent. Drier habitats, perhaps on beach ridges, were present, and species which may have grown on or among the calcareous rocks of the nearby Niagara escarpment also occur. Only one species which today typi- cally grows in shaded spruce forests was identified. Others which sometimes grow in this habitat were also found, but 234 are less useful indicators because they occur at open sites as well. The absence of a dominant forest element in the moss flora probably means that the landscape along this part of the Iroquois strand was occupied by a patchwork of dry and wet site herb and moss communities and that spruce oc- curred some distance behind the beach. Most of the spruce macrofossils were probably carried to the site from inland by drainage through the spillway. Nearly all of the fossil mosses are characteristic boreal forest species. .Most range northward to the arctic tundra but many also occur in the Great Lakes states. For example, 92 percent of the assemblage presently grows in the Straits of Mackinac region of Michigan where the present zonation of the strand vegetation is similar to that which apparently existed in northwestern New York State 12,000 years ago. An important difference, however, is that a temperate coniferous-deciduous forest occurs today in the up- land of northern Michigan but was absent from western New York at this time. The mosses of greatest phytogeographic interest are Aulacomnium aeeminetgm and_§. turgidum whose present Nerth American ranges center in the arctic and sub- arctic. The southernmost station for the former is along the north shore of Lake Superior, an area well-known for its relict arctic—alpine plants. Aulacomnieg turgggum has a greater number of occurrences along the southern edge of its range but it also is found widely in the subarctic and 235 arctic. In the East disjunct stations are known from the high peak region in the Adirondack mountains of New York and from the White Mountains of New Hampshire. .The presence of these taxa indicates that arctic- alpine vascular plants also may have grown near the Lake Iroquois strand at Lockport 12,000 years ago and raise the possibility that limited areas of tundra may have occurred in the region. It is possible, of course, that at this time both mosses were relicts from a period of 'tundra' vege- tation which may have existed in the area 500 to 1000 years prior to their burial. The presence of the two taxa along the strand could be explained, therefore, by postulating microhabitats favorable to their survival. Since fossils of other tundra plants have not been found in the deposit and forest species are clearly present and abundant, this seems to be the better hypothesis. Beyond New York State, but within glaciated eastern North America, spruce-rich forests were widely distributed 12,000 years ago. Their presence in southern New England (eee Davis, 1965) has already been mentioned. To the west in southern Ontario an age determination of 11,950 i 350 years B.P. dates the beginning of organic sedimentation and the upper part of the A zone at Crieff Kettle bog near Hamilton (Karrow, 1963). Spruce pollen accounts for about 80 percent of total AP in the A zone (Terasmae_;g Karrow, 1963) and the proportion of white to black spruce pollen is 236 approximately 6 to 1. Abies, Betula, Pinus banksiana and Quercus are the other main tree pollen types present. From 15 to 45 percent NAP occurs in the zone (calculation based on sum AP) and Ambrosia, Artemisia, other Compositae, Cyperaceae and Gramineae are the principal types identified. .QEXEE pollen occurs near the bottom of the zone. “The correspondence of the Houghton bog date and the newly determined age of the Two Creeks forest bed (11,850 i 100 years B.P.; Broecker and Farrand, 1963) has been noted. west's (1961) reanalysis of the type Two Creeks locality along the shore of Lake Michigan at the base of the Door peninsula, Wisconsin, produced spectra dominated by up to 90 percent spruce pollen (calculation based on sum AP + NAP) in all levels except the bottom. At this level Shepherdia canadensis, a heliophytic shrub that may have been one of the first colonizers of surfaces freed for plant occupation in the area, attained over 95 percent of the total. One out of every 6 spruce grains was identified as_gicee magiana. Spruce forest was also present in southeastern Minnesota at this time (the Picea-Larix Assemblage Zone of Cushing, 1967). It seems clear that the spruce dominated vegetation 12,000 years ago was not uniform in composition between New England and Minnesota. The most obvious difference is the presence of high values for gleee pollen in western New York and New England at this time and their absence from sites in Michigan, Wisconsin and Minnesota. Apparently pines were a 237 part of the lateglacial A zone vegetation in the East but did not occur in the contemporaneous vegetation of the Mid- west. The available data (Wright, 1964; 1968b) indicate that the Appalachian region served as a full- and late- glacial survivium for the three main pine species,_g;eee banksiana,_g. resinosa and g. strobus, which participated in the revegetation of the glaciated Northeast. The relative numbers of temperate deciduous tree pollen types also vary from site to site within the region. An accurate assessment of the variability in terms of climate depends in part on a detailed knowledge of the pollen rain in existing boreal forest and woodland, the forest-tundra, transition and the tundra itself. This is not available at the present time. It also must be kept in mind that modern analogues for cer- tain lateglacial pollen assemblages may never be found be- cause the vegetation which produced them may have been a mixture of species brought together by differential migration rates and may thus represent chance combinations of species which coexisted for varying periods of time following the withdrawal and disappearance of the ice. inouth of the glacial boundary the vegetation in several parts of Pennsylvania 12,000 years ago was apparently much different from that found at this time near Houghton bog in western New York. At Bear Meadows in central Pennsylvania (Kevar, 1964; Stingelin, 1965) pollen analysis of sediments below a radiocarbon date of 10,320 i 290 years B.P. 238 (Westerfeld, 1961) produced spectra dominated by pine pollen (60 to 70 percent) and associated with relatively low values for spruce (10 to 15 percent). NAP generally totaled 10 percent of the sum. Similar spectra have been obtained by Paul S. Martin from sediments below a C-14 date of 11,300 i 1000 years B.P. (Y-727; Guilday e3 el., 1964) at the New Paris Sinkhole No. 4, in south-central Pennsylvania, 65 mi from Bear Meadows and 100 mi from the glacial boundary cross- ing northwestern Pennsylvania. The 3 m of cave filling be- neath the dated horizon is dominated by_g;eee pollen which accounts for about 60 percent of the AP plus NAP sum. From 6 to 15 percent_g;eee pollen occurs in the same interval and the rest of the sum, from 20 to 30 percent, is comprised of Compositae, Cichorioideae, Cyperaceae and Gramineae pollen. Both categories of Compositae pollen likely were produced by near- and on—site herbs. Some grass and sedge pollen may have had a similar origin. The vegetation producing this assemblage may have resembled an open boreal woodland with spruce and jack (?) pine separated by open ground (1219-)- Above the dated level Pinus remains dominant but_g;eee drops to less than 5 percent of the sum and Betulaceae, Quercus and other temperate arboreal pollen taxa become strongly represented. This apparently records the movement of temperate forest elements into the area. Sediments below the date contain bones of a large number of vertebrates whose modern ranges center southeast 239 and west of Hudson Bay in boreal Canada. Of particular interest are the remains of at least three Labrador collared lemmings, a species that today occurs mainly within the tundra of northern Quebec. Also found were the bones of the thirteen-lined ground squirrel and the sharp-tailed grouse, two prairie species whose occurrence substantiates the con- tention (Schmidt, 1938; see also Benninghoff, 1964) that an eastward extension of prairie elements occurred in late- rather than postglacial time. If the vegetation in central and southern Pennsylvania indeed was an open, boreal woodland like that existing beyond the north edge of the boreal forest today, an interpretation which is in part substantiated by the fossil vertebrates, the presence of a more closed spruce-pine forest to the north on glaciated terrain is difficult to understand because this zonation is the reverse of the current arrangement of these vegetation types in North America. Existing data are too sparse to establish the presence of the boreal forest in the region south of Pennsylvania and a taiga-tundra in a wide band between the ice margin and the forest during the full- glacial 18,000 years ago as has been claimed by Martin (1958a). If such a zonation existed, however, the low rela- tive numbers of spruce pollen in Pennsylvania about 11,500 years ago might have been produced by the stragglers of the spruce migration that participated in this phase of the re- vegetation of the glaciated region to the north. The apparent 240 abundance of pine trees in Pennsylvania at this time as shown by the work of Martin (Guilday_ee_e;., 1964 and Kevar: 1964), indicates that certain species of this genus may have dominated the landscape northward toward New York State. A floristic boundary separating spruce- and pine-rich forests must have existed somewhere between the two areas. If pines indeed were dominant behind the spruce forest during A zone time, this would help to explain the rapid development of the B or pine zone following the disappearance of spruce from the region. In western New York_g;nus strobus was the principal B zone pine. It would be interesting to determine whether it was present in central Pennsylvania during the lateglacial. The presence or absence of a verrucose furrow, a characteristic which allows white pine pollen to be sepa- rated from pollen produced by the other pines in eastern North America makes such differentiation possible. The end of the A zone at both Houghton and Protection bogs can be dated by extrapolation. At the former locality, assuming that the pine peak occurred at the same time as it did at nearby Protection bog and that the sedimentation rate was constant, the mid-point of the_g;eee decline is about 9500 years B.P. Since the basal marl at Houghton bog may have accumulated at a more rapid rate than the silty gyttja at Protection bog, this date may be somewhat too young. The same type of calculation applied to data from Protection bog yields an age of 10,500 years B.P. for the same point in the 241 spruce decline. Both age determinations are in accord with those listed by Ogden (1967) who has concluded that the ap- proximate syndhroneity in the extinction of the spruce forest across the mid-latitudes of eastern North America points toward a sudden climatic change at this time. ZONE B As the spruce dominated A zone vegetation near the two Valley Heads sites disappeared about 10,500 to 9500 years ago, the pollen record indicates that pines became increasingly abundant in the region. At Houghton, Pro- tection and Allenberg bogs the transition was abrupt. In contrast, spruce percentages at the Genesee Valley Peat works gradually decline, although_g;eee values increase rapidly. This is achieved mainly at the expense of various nonarboreal pollen types. High_g;eee values are maintained upward in the section well into spectra which seem to be equivalent to those in zone C-l at the three other sites. Although_g;eee drops to about 7 percent of the sum in the Genesee Valley profile above a depth of l m, that portion of the diagram in which percentages of both_§;eee and_Teege are high may be strongly influenced by on-site pine trees. _Pinus strobus pollen was the main type identified from this interval and the occurrence of white pine cones at various levels in the peat implies that white pine was growing locally. For this reason the end of zone B was drawn near 2.25 m at the 242 mid-point of the ngqa increase, even though above this level giggg percentages are still high. The A to B zone transition has not been dated at either Allenberg bog or the Genesee Valley Peat werks but, considering the proximity of all four sites, the disappearance of spruce may have been synchronous across the entire region. The interval over which maximum pine percentages oc- cur in Protection bog sediments has been dated at 9030 i 150 years B.P. (I—3551). This age determination compares well with a date of 9310.: 150 years B.P. from an equivalent stratigraphic position at Crystal Lake in northwestern Pennsylvania (Walker and Hartman, 1960) where the entire postglacial pollen sequence parallels my profiles from western New York. The dated sample at Crystal Lake was taken from the level at which maximum_gigu§ values occur, al- though at this depth_§ig§§ still amounts to 10 percent of the sum. In southern New England maximum relative and absolute members of pine pollen have been found in sediments about 9000 years old (Davis, 1967b; see also Davis, 1965b). The ecological meaning of zone B has been discussed at length by Dansereau (1953) who presents a number of hy- potheses to explain the widespread occurrence of maximum pine values following the disappearance of the A zone spruce forests. This publication appeared at a time when specific identifications of_§igg§ pollen were less certain than they are now. A part of the difficulty in interpreting zone B 243 lies in the well-known over-representation of pine pollen in sediments. With this in mind, Davis (1963, 1965b) has ap- plied correction factors derived from a comparison of surface pollen accumulation and vegetation composition to a profile from northern Vermont. Her data indicate that maximum B zone pine percentages are an artifact caused by the low pollen productivity of the rest of the B zone vegetation. Pine trees were thought to have been rare in the region sur— rounding the basin in spite of the high relative pine pollen frequencies. This interpretation was later revised, however, when absolute pollen frequency data from Rogers Lake in southern Connecticut became available (Davis, 1967b). The deposition rate of oak, pine and other arboreal pollen types was found to actually increase in zone B, and at certain levels the rate for pine was 18 times higher than later in postglacial time, implying that pines were truly abundant in the region during B zone time. My absolute pollen frequency determinations from Allenberg and Houghton bogs corroborate these findings, providing the sedimentation rate was uniform across zone B at these sites. One or both of the Diploxylon pines, Pinus banksiana or_§. resinosa, appear to have been members of the regional vegetation that contributed pollen to zone A sediments in western New York. Fairly high values for the Diploxylon species persist through the lower part of zone B at all sites, but by the end of the zone B time, only 1 to 3 percent 244 occurs. Similar values are found in early post-settlement spectra before extensive plantings of Diploxylon pines were made in the area. At this time presumably only_g. resinosa, which is today restricted to stations along the Genesee River (Zenkert, 1934), was contributing Diploxylon pine pollen to the sediments. Utilizing data given by Whitehead (1964), a shift of the mode to a larger size class, as seen in the size-frequency graphs of_gigg§ Diploxylon pollen from Allenberg bog (Fig. 6), suggests that g. resinosa was the principal B zone Diploxylon pine, whereas both P. banksiana and_g. resinosa may have been members of the zone A vege- tation. However, the closeness in pollen size of these two species, as Whitehead has emphasized, makes positive identifi- cation impossible. Diploxylon pines were infrequent in the regional vegetation at the west end of New York State after about 9000 years B.P., while they appear to have been more abundant before this date throughout the state (see Cox, 1959). Prior to 10,500 years B.P., red or jack pine or both were absent from Minnesota, but about this time they arrived at the southeast corner of the state, having migrated, probably around the northern end of the Great Lakes, from their survivium in the Appalachian Highlands of eastern North America (Wright, 1968b; Yeatman, 1967). Jack pine does not seem to have persisted in other refugia south or west of the Great Lakes. 245 The relatively few Pinus strobus grains which occur in the lower half of zone A indicate that white pine was not initially near any of the basins. However, as the deposition of zone A progressed,_g. strobus percentages increased, indi- cating that white pine became more abundant regionally. This is most clearly seen in the Allenberg bog section C diagram. It is certain, however, that white pine was present near Pro— tection bog during the A to B zone transition because a white pine cone was recovered from silty-clay gyttja at a depth of 5.75 m. Based on an extrapolation from the two higher radiocarbon dates at this site, assuming a constant sedimentation rate, the cone was deposited approximately 10,000 years ago. Total pine pollen at 5.75 m accounted for 35 percent of the sum. Haploxylon and Diploxylon types each amounted to 6 percent but the remainder could not be identi- fied to subgenus. Pine was associated with 18 percent spruce pollen at this level. A maximum of 25 percent_§. strobus pollen is present higher in zone B. At the other sites in southwestern New York, white pine also appears to have been one of the principal species which replaced spruce. _ging§ strobus arrived in eastern Minnesota 7000 years ago from the east (Wright, 1968b). Its further migration toward the west was limited by the eastward expansion of the prairie and oak savanna which began 8000 years ago in the upper Midwest. The expansion ceased 4000 years ago and white pine began again to migrate westward reaching the northwest part of 246 Minnesota about 2700 years B.P. This is presently the western edge of its distribution in North America. White pine in western New York 10,500 years ago supports wright's contention that the species survived full-glacial conditions in eastern North America. Studies in the Allegheny National Forest of north- western Pennsylvania (Hough and Forbes, 1943), indicate that Pinus strobus may have played a successional role in the change from spruce to pine forests. In this area today oc- cur even—aged pine stands whose origin, in many cases, has been traced to an event that opened a part of the forest to seeding from nearby mature individuals. Understory white pines are absent because its seedlings do not survive in the shade. It is easy to visualize white pine seeding into open- ings created in the deteriorating spruce forest 10,500 years ago. _we know that mature, Seed-producing white pines were established at this time near Protection bog and probably elsewhere in the region. They seem to have coexisted temporarily with spruce whose actual abundance in the total vegetation at this time is not precisely known. Size- frequency measurements indicate that Picea mariana was the main spruce near Allenberg bog at the end of zone A. Because this is principally a lowland, wet-site species, the upland _gigg§ glauca forests, which according to the pollen size data were present earlier in the A zone, may have been re- placed by other communities. Considering the narrow interval 247 across which spruce drops from high to low values, spruce forests must have rapidly disappeared, freeing more and more surfaces for occupation by pine and other B zone species. Wright (1964) has suggested that spruce regeneration at this time was limited by summer temperatures which exceeded the tolerance of the species. The total duration of zone B in western New York seems to have been between 1500 and 2000 years, extrapo- lating from the Protection bog age determinations. This is equivalent to about four white pine lifetimes, if we accept 450 years as the normal life span of the species (see Fowells, 1965). The occupation of a given site by successive generations of white pine may mean that other species that would normally replace it in the region today,_gflq..g§Eq§ canadensis, had not yet migrated to the area. Since hemlock pollen does not occur in large numbers until some time after the B zone peak at 9030 i 150 years B.P., this hypothesis seems to be supported by my data. There was at least a four millenium lag in the migration of hemlock northward from somewhere in the unglaciated Appalachians following ice with- drawal from the Valley Heads moraine 13,000 years ago. Rather than assuming a climatic control for the zone B vege- tation, differential migration rates of species back onto glaciated terrain may explain the basic pattern of post- glacial pollen succession in this region. 248 In some of my profiles zone B can be divided into a lower pine—birch subzone and an upper pine-oak subzone. A birch peak in the lower part of the zone is best deve10ped at Allenberg bog where it is associated with a high in .Qggpinus-Qstgyg, Fraxinus.gig£§ and Populus curves. These features are less apparent at the other sites, although at Houghton bog high percentages of Betula and Carpinus-Ostrya occur in the equivalent stratigraphic interval. Whether these changes had regional significance is doubtful, however, because they are less clearly defined at nearby Protection bog where no peak is discernable in the Betula curve. .9lfl2§ is well represented at most sites suggesting that elms were an important part of the regional vegetation. In fact the magnitude of elm percentages in zone B is only slightly less than that present in later postglacial time. The presence of high birch values is not unique to western New York. Similar findings from southern New England have been reported by Davis (1958) and by Whitehead and Bentley (1963). These authors refer to the interval as sub- zone B-l. Since the peak occurs across the A/B zone boundary, birches may have been locally important members of the vege- tation that existed during the transition from spruce to pine domination. In western New YOrk Carpinus caroliniana and/or Ostrya virginiana, Fraxinus_gig£§ and Populus spp. appear to have been present during this interval as well. Davis (1967b) mentions that the New England B—l pollen assemblages 249 compare well with modern surface samples from northern Minne- sota and from central Canada near Lake Timagami. These lo— calities are in the mixed coniferous-deciduous forest for- mation about 60 mi south of the boreal forest and indicate that the climate of southern New England during the B-1 was cooler and drier than it is at present. Slightly lower Betula percentages and higher Quercus and glmgg values characterize the western New York State sites, but otherwise B-l assemblages from this region agree with those from New England. At Allenberg bog where size measurements are avail- able for B zone birch pollen, the configuration of the size- frequency curve suggests the presence of two species (Fig. 6). However, these cannot be identified with the size data currently available from herbarium specimens of eastern North American birch species. Possibly B. populifolia,_§. pumila or both were present, as these species have pollen inter— mediate in size between the small grains of the shrub birch, _§. glandulosa, and the larger pollen of the tree species,_§. _lgg§a and-B. alleghaniensi . If B. populifolia and.§. pumila produced the mode at 25 mu, then the larger modal class near 27 mu may indicate the presence of one of the tree birches during the deposition of zone B. In zone C-l the mode also occurs at 27 mu, but higher in the section it shifts to 26 mu. Although I can be certain that only B. lggtg and_§. alleghaniensis gave rise to the mode at 26 mu in the upper C 250 zone spectra, the meaning of the modes at 25 and 27 mu in zone B is obscure and I must conclude that, although other species seem to be represented, their identity is unknown- The upper portion of zone B in western New York is dominated by_g;gg§ strobus and Quercus pollen. In some of the profiles Betula, Carpinus-Ostrya, Fraxinus and Populus values are lower than they were near the bottom of zone B. .Aggr saccharum was likely established in the region by the end of zone B. Arrival of oaks and expansion of the area occu— pied by them at locations near the basins is indicated by rapidly increasing relative numbers of oak pollen, although before this time some oaks were probably growing within 50 to 100 miles of the basins. The species involved are un— known, and either macrofossil evidence or improved pollen identification techniques are needed for specific determir nations. However,_Quercu§ rubra is a good possibility be- cause of its current 'northern' distribution and pioneer status, but others could have been present also. The entire B zone would seem to record the develop- ment of a white pine-oak forest. However, it is likely that the vegetation was quite complex at this time. Because of its broad ecological tolerances, white pine probably oc- curred in lowland valleys with elm and black ash and in the upland with cake and/or sugar maple. The former community may have been similar to the White pine-American elm swamp forest that originally occupied the axes of some of the major 251 valleys in Cattaraugus County (Gordon, 1940). Forest types containing white pine and oak species are also known from western New York at the present time. White pine and red, black and white oak originally occupied dry sites in about 2 percent of Menroe County (Shanks, 1966), and similar stands undoubtedly occurred elsewhere in the Erie-Ontario Lowland. Castanea dentata and, at certain places, Pinus rigida were additional important members of this community. The pollen rain of this forest type has not been determined and, con- sidering the present distribution of vegetation in the low- land, it seems unlikely that a sample which was not influenced by pollen output from the surrounding mesophytic forests could be obtained for comparative purposes. In any case, neither Castanea nor Pinus rigida appears to have been a member of the B zone vegetation according to pollen data currently available. Forests containing white pine and oak species are also known from well—drained sites, usually S- facing slopes, in the Allegheny Upland. The pine-oak subzone pollen assemblages seem to have no exact modern analogue, but Davis (1967a) has pointed out that they are closest to the modern pollen rain in southern Ontario near the boundary between the deciduous and coniferous-deciduous forest (gee King and Kapp, 1963, sample 4). However, an important difference is the higher pine and oak percentages found in upper B zone spectra from south- western New York State. Although the suggestion that an 252 analogue of the pine-oak subzone is not in existence today seems premature, it is possible that a unique assemblage of species brought together by differential migration rates was present in western New York 9000 years ago. ZONE C-l Post-zone B sediments contain a record of the de- velopment and persistence of forests which contain the same species that now dominate existing forest types in western New York. Hemlock is an important tree in this region today and its pollen record is especially interesting and signifi— cant. The C-l is set apart from the zones above it by high relative numbers of hemlock pollen and gradually increasing beech values. These features are retained when the relative frequency data from Houghton and Allenberg bogs are re- plotted on an absolute basis. Tsuga percentages increase markedly at the B/C-l boundary and total Pinus values, with _g. strobus pollen predominating, reciprocally decline. Across a 30 cm interval at Protection bog Tsuga increases from 2 to 25 percent. Assuming that it took 26 years to de- posit 1 cm of gyttja in the basin (the sedimentation rate between the two radiocarbon dates immediately higher in the section), about 800 years was needed to accumulate this thickness of sediment. The abrupt nature of the increase and the weak ex- pression of Tsuga in zone B suggests that the interval may 253 record the initial invasion and expansion of hemlock in the region. The low relative numbers of hemlock pollen found in zones A and B at most of the sites probably represent wind- blown grains. Unfortunately, detailed information on the dispersal of hemlock pollen is not available, but up to 7 percent has been found in surface sediments near Lansing, Michigan (Parmelee, 1947) at locations about 75 mi south of the limit of continuous hemlock distribution in the state as mapped by E. L. Little, Jr. (in Fowells, 1965). According to the R values which I have calculated using several estimates of forest composition, hemlock pollen is somewhat over- represented in both surface and pre-settlement spectra. This may also have been true during earlier postglacial time sug- gesting that hemlock trees were actually somewhat less abundant in the total vegetation than the pollen record implies. The ultimate cause of the replacement of white pine by hemlock is speculative. Arrival of hemlock in the region during its migration northward onto glaciated terrain has already been mentioned as one possibility, but whether hem- lock was migrating at its fullest potential during the time preceding its arrival in western New York or whether its movement was held in check by climate or soil development is not known. The latter would seem not to have been too criti- cal because hemlock seeds are able to germinate on a wide variety of substrates: moist, well-decomposed litter, rotted 254 wood, mineral soil and moss mats on soil and rocks (Hough, 1960). Once hemlock was present, however, white pine re- placement can be viewed as a successional event. Although in existing forests both white pine and hemlock are often periodic in occurrence, studies in the Allegheny National Forest (Hough and Forbes, 1943) have shown that when both are found together, white pine will drop out of the associ- ation as time passes because its seedlings do not become es- tablished under a dense canOpy, while those of hemlock can. At the present time hemlock occurs from the southern Appalachians northward across the glacial boundary to northern Maine, New Brunswick and Nova Scotia and westward to eastern Kentucky, central Ohio and through southern Ontario and the northern part of the southern and the entire northern peninsula of Michigan to northeastern Wisconsin (Little_ig Fowells, 1965). The pollen record for_3§uq§ is not identical across this entire region, however, and al- though the main difference is the absence of two hemlock maxima from certain areas, a feature which will be discussed more fully under the heading zone C-2, another variable is the magnitude of hemlock representation in sediments de— posited immediately following the high B zone pine per- centages. In glaciated eastern North America from north- western Pennsylvania to northern Maine (gee Cox, 1959; Davis, 1965b; Deevey, 1951; Krauss and Kent, 1944; Potzger and Otto, 1943; Terasmae_i§ Karrow, 1963; Walker and Hartman, 1960) 255 maximum hemlock values appear early in the profiles. At several sites in the southern part of this region where radiocarbon dates are available, the appearance of hemlock can be estimated at between 9300 and 8500 years B.P. (Davis, 1967b; Walker and Hartman, 1960). Although hemlock first appears near Halifax, Neva Scotia about this time, maximum values were not reached until about 7100 years ago (Living- stone, 1968). In general, at sites from eastern Nerth America which fall within the present limits of the Hemlock-white pine- northern hardwood forest (Nichols, 1935), Tsuga accounts for 25 to 35 percent of the sum directly above zone B. However, in southern New England, south of the forest boundary but still within the total range of hemlock, maximum C-l hemlock values reach only 10 percent (Davis, 1967b). To the west in the Hemlock-white pine-northern hardwood Forest region of Michigan and Wisconsin,_2§gqa pollen is also weakly repre— sented in the equivalent stratigraphic interval. Un- fortunately few C-l4 dated pollen profiles are available from either state but hemlock would seem to have reached the Douglas Lake region of Michigan (Wilson and Potzger, 1943) early in the period of oak-hardwood domination which may be temporally equivalent to zone C-l in the East. About 5 per- cent or less occurred until some point later in postglacial time when an increase to 20 percent took place. In central Michigan hemlock is consistently a part of the pollen record 256 above a C-14 date of 7982 i 250 years B.P., but it accounts for only 5 percent or less of the sum (Gilliam_gt.§l., 1967). west's diagram (1961) from Seidal Lake in eastern Wisconsin shows similarly that_T§gg§ appeared fairly early during the period of oak domination that followed the_giflg§ maximum, but hemlock never exceeded about 3 percent of the total until much higher in the section. Increasing-Tsuga percentages in later postglacial sediments from this lake (=C-3 in western New York?), parallels the same trend at sites in northern Michigan. This change can also be observed in profiles from many other sites in the region (Messenger, 1967; Potzger, 1946). Hemlock appears to have entered Michigan from the east, north of Lake Erie, and not from the south across the Prairie Peninsula which apparently acted as an effective barrier to migration of hemlock, beech and perhaps other species from the central Appalachians (Benninghoff, 1964). For example, in the diagram from Silver Lake in western Ohio (Ogden, 1966), low relative numbers of hemlock pollen (<5 percent) first occur 9800 years ago, but at several points higher in the section it completely drops out of the counts. Whenever hemlock pollen is found, it comprises only 2 to 3 percent of the sum, indicating that.2§uq§ was never very abundant in western Ohio. In this region during postglacial time hemlock likely occurred intermittently in small, iso- lated stands, perhaps on N—facing slopes or in other suitable 257 edaphic situations. At the present time Silver Lake is about 50 mi west of the limit of continuous hemlock distri- bution in Ohio. Although it seems that hemlock pollen and probably hemlock trees first appeared at about the same time in western New York, Michigan and northern Wisconsin, the period of hemlock dominance characteristic of my western New York profiles is absent from sites to the west. In western New York where the C-1 occurs between about 8500 and 4300 years B.P., the vegetation appears to have been remarkably stable. The only significant changes occur in the Fagus and Quercus curves. The former shows a long-term increase, while the latter undergoes a corresponding decline. However, during the same period, the prairie and oak savanna expanded east- ward in Minnesota (Wright, 1968a). It has not yet been es- tablished whether Wisconsin and Michigan were affected by the drier and warmer climate that probably induced this vege— tation change, but if they were this might explain the meager representation of hemlock pollen in sediments accumulated during this interval at sites in the northern part of these states. Hemlock is known to grow best in a humid, cool cli- mate and to be sensitive to drought which, if excessive, will result in death of the trees. As a corollary to this hypothesis, it follows that western New York State, where hemlock pollen is abundantly represented between 8500 and 4300 years ago, was moister and cooler than the Midwest. In 258 short, it would seem that when Minnesota and perhaps surround- ing areas were undergoing a 'xerothermic' interval, western New York was not. Too few pollen profiles are available from the Erie- Ontario Lowland in central and western New York to determine conclusively whether the vegetation during zone C—l time was the same on both sides of the tension zone which new exists in the area, or whether beech-maple and oak forests dominated the lowland vegetation as they did immediately preceding settlement of the region. In pre-settlement forests hemlock was apparently much less abundant than in the upland, and since the development of this difference should be apparent in the pollen record a weaker representation of hemlock °pollen in the lowland than the upland during the C-1 might indicate that the tension zone was established fairly early in postglacial time. Little difference is apparent in the C-l_g§uqa values between available profiles from lowland and upland sites, however. For example, in the Bullhead Pond profile (Cox, 1959), from a small lake 20 mi south of Lake Ontario in central New Yerk,_T§gq§ accounts for about 20 per- cent of the sum. Slightly higher percentages of hemlock pol- len occurred at some of my upland sites, but the difference hardly seems significant. At Cicero Swamp (ibi§.) and Penn- ville Hidden Lake (Durkee, 1960), about 40 mi further east but near the edge of the lowland deciduous forest region, 259 .ngqa reaches about 40 percent of the sum in zone C—l. In general, the pollen diagrams from both the upland and the lowland are enough alike to indicate that only minor differ- ences occurred across the entire region, but further data are needed to more adequately treat this problem. The forest vegetation which developed during zone C—l in southwestern New York was very similar to that exist— ing in the region just prior to colonial settlement. Upper C-l spectra closely match pollen assemblages which accumu— lated in the region between 1000 and 200 years ago. This means that the regional vegetation, and very likely the cli- mate, during both periods was the same. .gsgqa and Fagus did not arrive in western New York at the same time, and communi- ties containing hemlock must have been well-develOped when beech entered the region and began to expand. The long-term increase in.§§gg§ values, which took place mainly at the ex- pense of Quercus, can be interpreted as a trend toward in- creased mesophytism in the total vegetation. The prominence of beech in postglacial sediments from western New York is scarcely surprising in view of the important position this species holds in the Allegheny National Forest where it ranks highest of all forest species in establishment capacity, survival and competition. Beech is also less dependent on special seedbeds, soil moisture or light than hemlock (Hough and Forbes, 1943). During zone B time, oak forest types may have occurred at a variety of sites, although at the present 260 time they are found mostly on S- and SWefacing slopes in southern Cattaraugus County and in limited areas to the north. The pollen record indicates that with the passage of time these forests shrank in size and were in part replaced by more mesophytic associations containing hemlock, beech, sugar maple and other northern hardwoods. Overall, the C-1 vegetation was probably a mixture of forest types as complex as now occurs in the region. Pollen from most of the important tree species which at present exist in the area are represented in various magni- tudes in the zone; those that are not, such as Prunus serotina and Magnolia acuminata, are mainly insect pollinated and therefore are rarely found as fossils. Both Tilia and Fraxinus americana and/or_§. pennsylvanica (4 colpate Fraxinus grains) first appear at the beginning of the C—1 and are as well represented in this zone as higher in the profiles. Likewise, pollen from Platanus occidentalis was first encountered at about this time in the two Valley Heads bogs; however, to the south at Allenberg bog small percentages occur throughout zone B. High Platanus values are prominent in zone C-l at Houghton bog where the outwash plain sur- rounding the bog may have been an especially favorable habitat for this species. Juglans cinerea first occurs in low rela- tive numbers in the upper part of zone A and a few grains were encountered in zone B sediments, but at all sites the postglacial maximum is reached at some point within the C-1. 261 .glmgs and Betula continued to hold prominent positions in the vegetation of the region. At Allenberg and Protection bogs birch values are somewhat higher near the middle of the zone than at either beginning or end. Size-frequency measurements of birch pollen at the former location indicate that the tree birches,_§._lgg£§ and_§. alleghaniensis were the main Species present. Low relative numbers of Castanea dentata pollen first appear in zone C-l at Allenberg and Houghton bogs, although it regularly occurs from near the beginning of the C-2 upward at Protection bog. Local habitat differences around the basins probably explain the disparity- ZONE C-2 In southwestern New York zone C-2 is characterized by low hemlock percentages and increased values for broad- leaf deciduous tree taxa. It is an interval between two successive hemlock maxima. The zone is represented in the Houghton, Protection and Allenberg bog profiles but is ab— sent from the Genesee Valley Peat Works diagram because the uppermost sediments at this site were not sampled due to dis- turbance by bulldozing. In my profiles the lower zone boundary can be readily located at the midpoint of the abrupt hemlock decline which, at Protection bog, has been dated at 4390 i 110 years B.P. (I-3550). However, the placement of the upper boundary is arbitrary because of the absence of any clear stratigraphic markers. I have chosen a point where 262 percentages of deciduous tree taxa are reduced over their C-2 maxima and where hemlock just begins to exceed 10 to 15 percent of the sum. Accepting this placement of the boundary, the culmination of the C-2 at Protection bog occurred 1270 i 95 years ago (I-3549). This is from 500 to 800 years younger than other age determinations of the C-2/C-3 transition from eastern North America (gee Davis, 1965b). Using the rate of sediment accumulation between the two highest C-l4 dates at Protection bog (0.069 cm/year), hemlock percentages decreased from 23 to 8 percent in about 350 years. This is merely an estimation, however, because the sedimentation rate may have been less in the upper part of the gyttja than between the dated levels in the gyttja and peat. Furthermore, the 25 cm over which the reduction in hemlock percentages took place reflects the sampling interval I used in this part of the profile. Since the same change could have taken place in less than 25 cm, the time interval may actually have been shorter. The pollen stratigraphy across the transition should be determined in detail in future studies. In the relative frequency diagrams the-Tsuga re- duction is compensated for by increases in a number of other arboreal pollen types, principally-Fagus, Acer_§accharum, Betula, Quercus and Carya. Lesser increases also occur in Pinus strobus, Fraxinus americana and/or.F._pennsylvgnica, .g. nigra and at one site, Carpinus-Ostry . In no case are 263 any of the increases as prominent as the Tsuga decline. Size— frequency measurements from the upper and lower halves of zone C-2 at Allenberg bog indicate that_§etula lenta and/or _§._lg£g§ were the main birches contributing to the pollen rain, but possibly a third species was present during the deposition of the upper part of the zone. These changes indicate modifications in the regional vegetation which simultaneously favored the expansion of dry site oak and hickory forests and mesic communities contain- ing beech, sugar maple and birch. Traditionally zone C-2 in eastern North America has been interpreted as a xerothermic interval, a period of warm and dry climate during which oak and hickory forests expanded at the expense of more meso- phytic associations (gee Deevey, 1949). Although a decrease in the representation of hemlock, a strongly mesophytic species, and the corresponding increase in oak and hickory in western New York State is the expected patttern, if we accept the xerothermic interpretation, the coordinated in- creases in.Ag§£ saccharum,_Betul§ lenta and/or_§. lute; and .Egggs grandifolia, all of which are mesophytes, are contra- dictory. A drier, more continental climate during the C-2 would be better documented if an analogue for the vegetation could be found where such a climate prevails at the present time. The Beech-maple forest region of central Ohio and Indiana is a loqical place to look for surface pollen assemblages 264 similar to C-2 spectra from southwestern New York, but un- fortunately no systematic study of the recent or subrecent pollen rain in Ohio and Indiana has been made. Pollen pro— files from this area provide some comparative data, however. The topmost spectra in the diagrams from north-central and northeastern Ohio presented by Sears (1942), which in most cases probably represent the subrecent pollen rain, do not match any of my C—2 assemblages, nor do spectra from just beneath the post-settlement Ambrosia peak at Silver Lake in western Ohio. In both areas Quercus and Carya values are larger and Fagus representation is much weaker than at any of my southwestern New York sites. However, the lack of correspondence is not conclusive proof of the absence of a relationship because the pollen rain has been determined at so few sites in this part of the Midwest. Until such infor— mation is forthcoming, it seems best to reserve judgment on the current existence of a probable analoguleor the C-2 vegetation. It has been postulated that a wedge of prairie vege— tation extended through central Indiana and Ohio during the putative mid-postglacial xerothermic interval (gee Benningr hoff, 1964). This hypothesis was used by Shanks (1966) to explain the origin of prairie remnants in oak openings in the Erie-Ontario Lowland of western New York. There is, how- ever, no C-2 increase in grass pollen at any of the sites I have studied in this area. All C-2 spectra are dominated by 265 arboreal pollen types and in only a few cases does nonarboreal pollen account for more than 3 percent of the sum. In these instances the NAP is clearly of local derivation. Of some interest in this regard, however, is the presence of Ephedra pollen in zone C-2 at Protection bog. If it could be es— tablished that Ephedra species were growing in the region during this time, the argument for an interval of xeric, continental climate would be improved. However, Maher's re- cent review (1964) has shown that Ephedra pollen, which has been found widely in the Great Lakes region, is not limited to any one constant stratigraphic interval, but rather it oc- curs sporadically in both late- and postglacial deposits. This fact and the presence of Ephedra pollen in a surface sample near Lake Simcoe, north of Lake Ontario (King and Kapp, 1963) and elsewhere implies that an extra-regional origin, through long-distance wind transport from the southwestern United States, is the most likely explanation for the presence of Ephedra pollen in western New York. The oc- currence of Liquidambag pollen in both late- and postglacial sediments in this region can be explained in the same way. At‘the present time the northern limit of sweet gum is southern Ohio and central west Virginia about 300 mi south and southwest of the sites included in this study. In eastern North America pollen profiles with two hemlock maxima separated by a single interval of low hemlock percentages are known from a broad area including northwestern 266 Pennsylvania (Walker and Hartman, 1960), New York State (Cox, 1959; McCulloch, 1939; Dunham, 1965; Durkee, 1960), northern Vermont (Davis, 1965b), southern Vermont (Whitehead and Bentley, 1963), southern New England (Davis, 1967b; Deevey, 1939, 1943) and Maine (Deevey, 1951). In Canada a hemlock decline is present in what appears to be a top-truncated pro- file obtained from a site west of Hamilton, Ontario (Terasmae _ig Karrow, 1963), a.T§gq§ minimum occurs in profiles from the Gatineau Valley region of Quebec, 30 to 60 mi north of Ottawa (Potzger and Courtemanche, 1956) and a mid-postglacial decline in_T§gqa percentages occurs farther east in Nova Scotia (Livingstone, 1968). South of the glacial boundary at Bear Meadows bog in central Pennsylvania (Kbvar, 1964) a mid—postglacial_T§gg§:minimum also occurs. It remains to be established, however, whether the hemlock decline was strict- ly synchronous over the entire region just described. Radio— carbon dates are available from only scattered localities but they show a surprising degree of accordance. The Pro- tection bog date of 4390 i 110 years B.P. agrees favorably with the-Tsuga minimum which started at Rogers Lake, Connecti- cut about 4100 years ago (Davis, l967b) and with the abrupt .gggqg decline dated at 4540.: 140 years B.P. at Crystal Lake near Halifax, Nova Scotia (Livingstone, 1968). Can the hemlock decline be explained in any other way than by postulating a xerothermic interval? Viewing the tripartite C zone as a unit, the decline seems to occur at a 267 time when soil development had progressed to a point where, by late C-l time, the soil supported the same forest types that exist in the area today. Furthermore, the pollen evi— dence indicates that no new taxa entered the region follow- ing deposition of the lower third of the C-1, although data are not available on the immigration of such Species as Liriodendron tulipifera, Magnolia acuminata, Prunus serotina and a few others that, while not important species regionally, are nonetheless significant members of certain forest com- munities. Whatever was responsible for the modifications in the mid—postglacial vegetation of southwestern New Yerk would seem to have effected a fairly stable situation, as least in respect to entry of new species into the area and probably soil development as well. In any event hemlock appears to be the key to the interpretation of zone C-2. The relative frequency diagrams clearly depict a reduction in_T§gqa percentages and an en- largement of values for temperate deciduous tree pollen types. It is difficult, however, to determine which was cause and which effect because of the nature of expressing data in per- centages, 122' when the relative numbers of one category in- crease, a concomitant reduction in one or several others must occur. Therefore, the hemlock decline could represent an actual reduction in the number of.T§gqa grains being de- posited per year or an increase in the deposition rates of Fagus, Acer saccharum, Quercus and other pollen types, while 268 EEeqe pollen deposition remained constant. In the former case, increases in relative numbers of deciduous tree taxa would be artifacts of the percentage system providing their deposition rates remained constant; in the latter the con— verse would pertain. Insofar as the absolute pollen frequency or the number of grains/unit volume of sediment represents actual deposition rates of the various pollen types, the absolute frequency diagram from Houghton bog shows that a significant reduction in the numbers of Teeqe pollen/m1 took place across the C-l/C-2 boundary and that low absolute numbers persist throughout zone C-2. Of interest also is the fact that _E§qg§, Ace; saccharum, Quercus and Betula, which showed the greatest relative frequency increases during the change from zone C-l to zone C-2, are not any more strongly represented in the C-2 than in the upper part of the C-1. Although these features are meaningful only if the rate of sediment accumulation was constant across the interval (and un- fortunately I was not able to obtain radiocarbon dates which would enable its determination in Houghton bog), at sites in eastern North America where sedimentation rates have been determined (Davis, 1967b; Ogden, 1967) it was more orlless constant between late C-1 and late C—2 time. This may not be universally true in all small lake basins but, in the ab- sence of differences in sediment lithology, it seems 269 reasonable to extend the assumption of a uniform sedimen- tation rate to Houghton boq. The absolute pollen frequency data indicate that the C-2 modifications in the relative frequency diagrams were produced mainly by a decrease in the absolute numbers of hem- lock, a species whose silvical characteristics are fairly well known (Fowells, 1965; Hough, 1960). Hemlock mortality can be caused by a variety of environmental factors, but drought is most important because of hemlock's shallow root system. Severe damage to hemlock stands over a broad area following the droughts of the early 1930's is well docu- mented in the literature. For example, Secrestuee_e1. (1941) estimate that 50 million board feet of hemlock died in the 230,000 acre Menominee Indian Reservation in Wisconsin during the three years between 1931 and 1933. These authorities demonstrated that under drought conditions root tips are rapidly killed and gradually the larger roots become weakened leaving the affected trees open to fungus and insect attack. To the east, hemlock mortality during the same drought period has been recorded at the Allegheny National Forest (Hough, 1936b) and near New Haven, Connecticut where in a 0.1 acre sample plot dead hemlock saplings and trees comprised 75 per- cent of the total (Stickel, 1933). In the latter region seedlings were killed outright and mortality in all size classes probably was enhanced by shallow soils develOped di- rectly over bedrock. 270 Data from a more detailed study on the effect of the 1930 drought on different forest types near State College in central Pennsylvania collected by McIntyre and Schnur (1936) supplies additional pertinent information. These authors ex- amined 23 plots each 66 by 100 feet spread among chestnut oak, hemlock, scarlet oak-black oak and white pine—chestnut oak forest types. Eighty-four percent of the total basal area of hemlock was lost from the entire sample. By comparison, black oak lost 52 percent, chestnut oak 28 percent, red oak 26 percent and sugar maple 11 percent. Before the drought the four hemlock type plots contained abundant_Teeqe canaden- 'eie, 60, 67, 78 and 60 percent expressed as a percentage of total basal area, while after the drought the relative dominance of hemlock was reduced to 2, 6, 66 and 24 percent in the same plots. The hemlock type changed from one domi- nated by this species to one of mixed composition, mainly chestnut and red oak with much smaller amounts of hemlock. In a general way this change parallels that observed in zone C-2 sediments from western New York with the exception of the importance of beech, sugar maple and yellow and/or sweet birch in the C-2 sediments from this region. I suggest, therefore, that the hemlock decline can be viewed as a response to several severe drought years. The recorded devastation of hemlock during the 1930 drought from Wisconsin to southern Connecticut and the accordance of the dates of the hemlock decline between western New York, 271 New England and Nova Scotia might furnish the basis for postulating that widespread droughts might also have oc— curred about 4400 years ago. This new hypothesis, which can be partly tested by obtaining additional radiocarbon dates, to further check the synchroneity of the mid-postglacial hem- lock decline, differs from the xerothermic interpretation in the nature and duration of the warm-dry climatic optimum. I feel that the modifications in the pollen record can be as well explained by postulating a series of severe droughts, perhaps distributed over several centuries or even over much less time, as by postulating an interval of xeric, continental climate lasting several millenia. The return of_geeqe to a position of prominence in the pollen diagrams by zone C-3 time may represent the order— ly and gradual succession of hemlock into communities to which it formerly belonged. Competition between hemlock and other mesophytes, particularly beech, would accompany this change and would affect the speed of hemlock reestablishment. IEESQ was never completely eliminated from the region because its pollen was continuously being deposited, even though in one instance, it reached as low as 4 percent of the sum. _Tegge likely survived in especially favorable edaphic situ- ations, perhaps in the deeper gullies that were cooler and moister than the surrounding upland. If the decline in hemlock alone produced the C-2 modifications in the pollen diagrams from southwestern New 272 York, biotic factors rather than climate must also be con- sidered as possible causative agents. Certain insects in- cluding two species of hemlock lOOpers and the hemlock borer are known to cause local mortality, as are foraging deer, porcupines and rabbits (Fowells, 1965). Man may also have played a role. The Protection bog date for the C-l/C-2 boundary corresponds to a period during which central and western New Yerk were occupied by Indians of the Lamoka culture (Ritchie, 1965). They subsisted by hunting, fishing and gathering; agriculture came somewhat later, about 1000 B.C. Little is known about the hunting techniques of these Indians, but apparently their main weapon was a javelin pro- pelled by a throwing board which was used to secure large game, mostly the white-tailed deer. They probably used dogs during the hunt and it is not inconceivable that fire was used to drive game. Hemlock is known to be vulnerable to fire damage and, although old trees may survive light sur- face fires because of their thick bark, the roots are easily damaged by a burn that extends deeper than the loose surface litter. I attempted to measure the influence and periodicity of past fires in southwestern New York by recording the number of charcoal fragments over 30 mu in size while count— ing to the basic pollen sum, but charcoal frequency does not seem to have been any greater in zone C-2 than in C-l. How— ever, my counting technique needs refinement before fire 273 damage can be ruled out completely. For one thing, charcoal is brittle and larger pieces probably fragment during macer- ation, suggesting that a smaller minimum size should have been established before counting. Of more serious conse- quence is the sampling interval which in this part of the profile was 25 cm, definitely too great to regularly document an annual event such as a fire. While I cannot rule out di- rect or indirect biotic interaction as the cauSe of the hem- lock decline with the existing data, it seems unlikely that a biological agent would have led to a reduction in hemlock percentages over the broad area in which they occur. Local- ly they may have been important but certainly not across many hundreds of miles. One last topic remains to be discussed before leaving zone C-2. In certain regions two Eeqee maxima occur in sedi- ments which are approximately contemporaneous with the C-2 in western New York. For example at Silver Lake in western Ohio, Ogden (1966) considers the xerothermic interval to be represented by a minimum in beech pollen covering the inter- val between 3600 and 1300 years B.P. This is about the same time period during which a Eeqee minimum has been found in southern New England deposits (Davis, 1967b;.eee_e;ee Deevey, 1943). The only diagram from New York State in which a simi— lar change occurs is Cox's undated Consaulus bog profile (1959) from eastern New York near Albany. In this profile the Fagus and Tsuga minima are not coordinated, as they are 274 in southern New England, but rather the former occurs slight- ly above the latter. Although.§que is characterized by erratic fluctuations in certain other of Cox's diagrams, a strong C—2.§eqee representation occurs in all. The signifi- cance of the bimodal.§egee curve in deposits to the east and west of southwestern New York is not known. Since the Eeqee and_geeqe declines are not entirely synchronous, it seems likely that different factors were responsible in each case. The relationship between the two should be pursued in future research. ZONE C-3 Two important changes in the pollen record character- ize this zone, the most recent in origin. These are the re- turn of.geeqe (subzone C-3a) and the occurrence of high per- centages of NAP above the pre-/post-settlement boundary (sub- zone C-3b). Following the C—2 Teeqe minimum, hemlock values steadily increase higher in the profiles until near the end of the C-3a, the lower of the two subzones, where they are similar in magnitude to those of the C-1. Percentages of de- ciduous tree taxa are reduced over their C-2 maxima, but they still remain strongly represented upward to the C-3a/C-3b boundary. Across this interval at the three sites where zone C—3 sediments were sampled,_geeqe increases mainly at the expense of Fagus, indicating an increased role for the former in the regional vegetation. This may have been 275 enhanced by a trend toward a moister and a somewhat cooler climate which many feel has prevailed during the past sever- al millenia (Sears, 1932) and may represent a continuation of succession begun during zone C—2 time. This climatic trend seems affirmed at sites in northern New York (Durkee, 1960) and Canada (Potzger and Courtemanche, 1956) by 3.21222 increase in sediments that appear equivalent to zone C-3 sediments in western New York where spruce, although very sparsely represented in the upper 25 cm in Houghton and Protection bogs, shows a dis- tinct increase upward in zone C-3 at Allenberg bog. Spruce pollen encountered during the counts at the two former sites is mostly restricted to post-settlement spectra and, there— fore, probably originated mainly from planted trees. At Allenberg bog, however, spruce occurs regularly throughout zones C—2 and C-3 but apparently was absent near the sampling point during the deposition of zone C-l. The small size of the spruce pollen (generally < 92 mu) in both the C—2 and C-3 indicates the presence of only Picea mariana which likely grew on the bog mat. Two grains larger than 100 mu found in (zone C-3b at Allenberg probably were contributed by intro- duced cultivated species. The Allenberg bog spruce increase needs further documentation in western New York because, rather than indicating a climatic trend, changes in the hy— drology of the peat deposit, induced by either physiographic modifications or biotic factors {229° beavers), may explain 276 what at the present time appears to be only a localized increase. In New England, Castanea pollen shows a decided in- crease in the C-2 (Davis, 1967b; Deevey, 1939) but this is not true in southwestern New York. Although regularly present upward from either zones C-l or,C—2 in the deposits I have studied, maximum Castanea values are reached near the end of the C-2 at Houghton bOg (1.8 percent) and near the middle of this zone at Allenberg bog (4.2 percent). Less than 1 per- cent occurs at equivalent positions in the Protection bog profile. Castanea was recorded in the original lot survey data only around Allenberg bog, and according to Gordon (1940), the pre-settlement distribution of chestnut mainly included the southern part of Cattaraugus County where it grew with oak on dry upper plateau lepes and tops and in Mixed mesophytic forests. Since Allenberg bog is near the area of maximum chestnut occurrence, while the two Valley Heads sites are about 25 mi to the north, my profiles, taken at face value, indicate that chestnut was never very abundant north of central Cattaraugus County in the Allegheny Upland of western New Yerk. Increasing_geeqe percentages in sediments that ap- pear to be stratigraphically equivalent to the C-3a in western New York occur across an area that approximately coine cides with the Hemlock-white pine-northern hardwood Forest region. A clearly defined hemlock increase is not apparent 277 in profiles from Nova Scotia, however. As was the case in zone C-l, the maximum values attained by_geeqe vary from dis— trict to district. For example, hemlock does not exceed 10 percent in the C—3 just south of the Forest region at Rogers Lake, Connecticut, while in western New York it reaches over 25 percent. At Rogers Lake, the highest C-3 hemlock per- centages occur between about 1500 B.P. and the Ambrosia peak which marks the advent of EurOpean settlement. To the west but still within the Forest region, the hemlock increase is more pronounced and parallels my findings in western New York. The C-l4 dated Maple River Township bog diagram prepared by Hushen and Benninghoff (unpublished ms) from Emmet County near the northern tip of the Lower Peninsula of Michigan shows that hemlock was weakly repre- sented (10 percent or less of the sum) between 4000 and about 3200 years B.P. at which time the increase began. After several erratic fluctuations upward in the profile, hemlock accounts for 50 percent of the sum in two spectra just be- neath the pre-/post-settlement boundary. This change was ac- complished largely at the expense of_g;eee. The weak character of hemlock representation during the C-1 and C-2 in profiles from Michigan and Wisconsin has already been mentioned, but nevertheless a few_geeqe stands probably ex- isted at favorable sites in the region. The geeqe increase can be viewed as an expansion of these colonies or, alter— nately, immigration from some source area may be represented. 278 In all, the pollen record for hemlock is worthy of continued study. As more C-14 dated profiles become available from southern Ontario, which appears to have been the principal westward migration route for hemlock, a more critical analy- sis of its postglacial history will be possible. The settlement of western New York, which began about 1800, and the attendant forest clearance is sharply marked by increasing NAP percentages and by the presence of wind- blown silt and clay in the Allenberg, Houghton and Protection bog profiles. Arboreal pollen drops to 50 percent or less of the sum over a very narrow interval indicating the catastro- phic effect of EurOpean settlement on the natural vegetation. Although there is no clear evidence of Indian agriculture in any of my diagrams, low but perhaps significant percentages of Amerosia and_§emex pollen which expand upward from zone C-3a below the pre-/post-settlement boundary, may have originated from weeds occupying cleared areas where corn, squash and beans were being grown by the Indians. It might be claimed that these pollen types were intrusions from more recent sediments, but Plantago pollen which would be expected to show the same behavior does not occur below the boundary except for rare single grain occurrences. Rayback's map (1966) of known Indian settlements shows a concentration of villages just west of Allenberg bog in eastern Chautauqua and western Cattaraugus Counties. Other villages are known from near the head of Cattaraugus Creek fairly close to both 279 Houghton and Protection bogs. I do not know how many of these sites were inhabited by agricultural Indians or ex- actly at what times they were occupied, but the high incidence of Indian habitation in certain parts of southwestern New Yerk indicates that any associated agricultural activities could be recorded in zone C-3 sediments. Pollen from cultivated species was found from the be- ginning of the C-3b to the surface. Agriculture indicators, including Zéé and pollen from other cereals (counted as Gramineae), occur at Protection, Houghton and Allenberg bogs, while Fagopyrum, the buckwheat, was found only at the two last named sites. Ambrosia pollen is the dominant NAP type in the zone and reflects the high incidence of disturbed nonforest habitats where the common.A. artemisiifolia and_§. trifida and the less frequent adventive, A. psilostachye, continue to flourish today. Maximum Plantago, Rumex, Cheno- Am, and Cichorioideae values are further evidence of the abundance of surfaces occupied by weedy species. Peak high- Spine Compositae percentages may be in part due to increased frequencies of weedy Species, but they could also reflect a local change in bog surface conditions favoring an increase in on-site taxa, such as certain BidenS species. SUMMARY AND CONCLUSIONS All of western New York, except the Salamanca reentrant, a semicircular area on the south approximately bounded by the present course of the Allegheny River, was ap- parently ice-covered sometime during the Wisconsin glaciation. The various drift sheets in the region lack definitive dates, but the following correlations have been used in a recent review of the Pleistocene geology of New York State (Muller, 1965): Lake Escarpment-Valley Heads Moraine: assigned to the Port Huron (Mankato) Substage,_ee. 13,000 years B.P.; Kent (Binghamton) Moraine: assigned to Cary Sub- stage, minimum date 14,000 years B.P.; Olean Moraine: pre-Cary (may be Tazewell or earlier). Other moraines associated with these are considered Short recessional still-stands or minor readvances of the ice margin. Recently obtained C-l4 age determi- nations, which indicate the Kent ice overrode the area around Cleveland, Ohio about 23,250 years ago (White, 1968), will necessitate some revision in the above chronology. Recession to a point north of the Niagara escarpment in northwestern New York State was complete 280 281 by 12,500 years B.P. and ice apparently never again re- advanced into western New York. Pollen succession was studied in sediments from four basins located on drift sheets of different age: Houghton and Protection bogs associated with the Valley Heads moraine, Allenberg bog on Kent drift and the Genesee Valley Peat Werks on Clean drift. Houghton and Protection bogs are 10 mi apart and Allenberg bog is 25 to 35 mi southwest of these. The Peat Works is 35 mi Southeast of the former two Sites and 50 mi east of Allenberg bog. The Portage escarpment separates western New York into two physiographic divisions: the Allegheny Upland in the south and the Erie-Ontario Lowland in the north. At the time of arrival of EurOpean settlers, the entire region was forested except for limited areas of prairie- like openings in the lowland. Forests of the Hemlock- white pine-northern hardwood Formation covered the up- land, while beech—sugar maple and oak-hickory communi- ties belonging to the Deciduous forest Formation oc- curred in the lowland. The ecotone between the two was not Sharp and large inclusions of hemlock-hardwoods have been identified in the lowland. Upland oak forests are mainly limited to dry plateau tOpS and S-facing lepeS near the Pennsylvania border. 282 An analysis of the bearing-trees recorded in the original lot survey notes for areas around Houghton, Protection and Allenberg bogs Shows Fagus grandifolia, followed by.Aee£ saccharum, to have the largest Im- portance Values. Tsuge canadensis is third in im- portance in two of the three areas. When frequency of mention data were computed from the same survey notes, EESEE continues to head the list. Second and third in frequency are Acer saccharum and Tsuqe canadensis around Allenberg bog,_Acer saccharum and_$ilia americana about Houghton bog and_gegqe canadensis and.Aee£ saccharum around Protection bog. Point-quarter sampling of three existing forest stands shows dominance by the same three leading Species but with a change in the order of de- creasing Importance Values. At all three sites Acer saccharum heads the list followed by Fagus grandifolia and Tsuga canadensie. The relative frequency of different pollen types in sur— face and pre-settlement Spectra divided by a percentage estimate of the importance of Species in a vegetation sample contributing a given pollen type provides a measure of the degree of representation of these pollen types in relation to the vegetation surrounding a depositional basin. These ratios or R values were calculated in several ways using surface pollen spectra compared with composition data collected by the U.S. 283 Forest Service in existing forests and pre-settlement Spectra compared with importance percentages and fre- quency of mention values derived from the original lot survey data. The computations indicate that in Recent and Sub-Recent sediment samples from western New York, pollen from Betula spp., Pinus spp., Quercus spp. and Tsuga canedensis are over-represented, Carpinus caroliniana and/or Ostrya virginieee, Fagus grandifolia and Juglans cinerea are proportionately represented and that_§ee£ rubrum and/or_§._eaccharinum, A. saccharem, Castanea dentata, Carya spp., Fraxinus_emericana and/or .g. pennsylvanice,_§._nigra, POpulee spp. and Tilia americana are under-represented. VA clearly defined T zone characterized by 50 percent or more nonarboreal pollen underlying a zone of spruce pollen domination was found in basal inorganic sediments only at Allenberg bog. T zone pollen assemblages con- tain 20 percent Picea, 10 percent_gieee, 3 to 8 percent Quercus, 3 to 5 percent Fraxinus_eiqge, 15 to 25 percent Cyperaceae, about 10 percent Gramineae and numerous other NAP types and closely match the pollen rain today in the boreal forest-tundra ecotone at Fort Churchill, Manitoba where discontinuous Spruce stands occur inter- spersed with herbaceous communities in a park-tundra. This implies that the climate in southwestern New York during the deposition of zone T was probably similar to 284 that in this part of the subarctic today. No positive tundra indicator pollen types were found, although microspores of a subarctic species, Selaginella selagin- oides, occur in several Spectra. Abundant herb pollen was present in basal sediments at the Genesee Valley Peat works, but a zone in which Spruce dominates is not present higher in the profile making the meaning of the basal herb-spruce-pine as- semblage at this site somewhat obscure. If local over- representation and redeposition were not operative, then an Open vegetation perhaps similar to the park— thndra of T zone time at Allenberg bog existed around the Genesee Valley site. However, there is some evi- dence to suggest that the pollen rain was heavily in- fluenced by near- and on-site herbs and therefore that the regional vegetation was a denser spruce-pine forest. Since the peat works are on the oldest drift Sheet in western New York, the basal sediments may antedate comparable deposits elsewhere in eastern North America. If subsequent C-l4 dating bears out their antiquity and if the pollen assemblage is taken at face value, a park- tundra may have covered the Alleghany Upland in southern New York during the "classical" Wisconsin glaciations. More data are needed from additional Sites in the region to further document this hypothesis. 285 Zone A at Allenberg bog is a long interval of domination by spruce and pine pollen. Changes in Fraxinus nigra, “Quercus, Pinus andcgieee percentages permit subdivision of the zone following the sequence recognized in certain profiles from southern New England where such changes have been interpreted as vegetation modifications in re— Sponse to the Two Creeks--Valders climatic changes. However, absolute pollen frequency data from Allenberg bog indicate that an increase in the absolute numbers of spruce and pine pollen being deposited per unit volume of sediment--evidence of an increased abundance of Spruce—pine forests on the 1andscape--was responsible for changes in the Quercus and.§£exinus niqre curves at this site. The absolute numbers of these pollen types remained more or less constant across the interval in which relative pollen frequency changes took place. Zone A pollen assemblages from Allenberg, Houghton and Protection bogs contain both Picea glauca and g. mariana and are similar to existing surface pollen accumulations in the open boreal woodland of central Quebec. In con- trast to the situation in Michigan, Wisconsin and Minnesota, Pinus banksiana and/or.g. resinosa grew in southwestern New YOrk during zone A time. Abies balsamea and Larix laricina were members of the A zone forests and deciduous trees, whose pollen consistently occurs in the zone, may have occupied favorable Sites within 10. 11. 286 some tens of miles of the basins. This is particularly true of Quercus spp. and-ggexinee nigra and perhaps Carpinus-Ostrve also. The presence of Acer, Carya, Juglans, Tsuga and_g;mee pollen probably reflects wind transport from distant sources. The bottom of zone A at Houghton bog has been dated at 11,880 i 730 years B.P. Mosses from an organic bed deposited 12,100 i 400 years ago along the southern edge of Lake Iroquois near Lock- port, N.Y. and a pollen Spectrum from associated lacustrine sediments indicate the existence of a mosaic of plant communities in northwestern New York at this time. Species characteristic of dry dune sand, rich fens and better drained fen edges probably occupied the area between the lake edge and a spruce-fir-larch forest occurring some distance inland. Exposed rocky habitats may have existed also. The occurrence of two typical arctic and subarctic mosses, Aulacomnium acuminatum and_A. turgidum, indicates the possible presence of limited patches of tundra vegetation. The spruce-pine woodland disappeared from 9500 to 10,500 years ago near the Valley Heads sites and was succeeded by zone B forests in which Pinee strobus held a dominant position. .éélfié balsamea flourished briefly during the transition. At some sites lower pine—birch and upper pine-oak subzones can be distinguished. A 12. 13. 287 Pinus strobus cone was recovered from sediments about 10,500 years old at Protection bog and clearly es- tablishes the presence of this Species in southwestern New York during the deterioration of the spruce-pine woodland. The B zone pine peak was dated at 9030 i 150 years B.P. at Protection bog. Zone C-l is characterized by high percentages of Tsuga canadensis and increasing values for Fagus grandifolia. Other Species which grow on the Alleghany Upland at the present time are also represented in this zone. The similarity of the pollen assemblages near the end of zone C-1 and those found immediately beneath the pre-/ post—settlement boundary indicate that the zone records the regional development of forests of the hemlock— northern hardwoods type. Forest composition likely was as complex as now occurs in the upland. No major changes took place in the vegetation of southwestern New York during the duration of the C-1, although the_§egee in- crease may indicate a trend toward increased mesophytism. An abrupt hemlock decline at Allenberg, Houghton and Protection bogs which has been dated at 4390 i 110 years B.P. at the last named Site, marks the beginning of zone C-2. The relative frequency of Acer saccharum, Betula, _Qe;ye,.gqu§, Fraxinus,_g;eee strobus and Quercus pollen all Show small increases in this zone. However, abso- lute pollen frequency data imply that these changes were 14. 15. 288 induced by a decrease in the total number of hemlock grains being deposited per unit volume. Rather than a long interval of xeric, continental climate, the C-2 in southwestern New York seems to be a result of differ- ences in hemlock abundance alone. A series of severe droughts, which are known to cause heavy hemlock mortality in existing stands, occurring over several years or tens of years is postulated as the cause of the hemlock decline. Biotic factors, including man, may or may not have played a secondary role. Hemlock was never completely eliminated from southwestern New York during the C-2. Zone C-3a began 1270 i 95 years ago at Protection bog and records the return of hemlock to a position of prominence in the regional vegetation. This change may have been influenced by a climatic trend toward greater moisture during the past several millenia but the hem- lock return following the low, early in C-2 time, may represent successional recovery. There is some evi- dence that Indian agriculture is recorded in the upper half of this subzone. European settlement and forest clearance occurred during the deposition of the topmost pollen assemblages belong- ing to subzone C-3b. Pollen from agricultural indi- cators, including Fagopyrum, gee and other cereals, was found in this subzone and the high frequencies of 289 Ambrosia, Cheno-Am, Plantaqo and Rumex pollen, species which grow in disturbed habitats, are characteristic. Zone C-3b sediments contain large quantities of Silt and clay blown in from bare areas around the basins. LITERATURE CITED Adams, W. (Ed.). 1893. Historical gazetteer and biographi- cal memorial of Cattaraugus County, N.Y. Syracuse, Lyman, Horton and Co. 1164 p. Andersen, 8. Th. 1954. A late-glaCial pollen diagram from southern Michigan, U.S.A. Danmarks Geol. Unders¢g. Ser. II, nr. 80: 140-155. .Arad, T. 1871. Pioneer history of Orleans County, New York. ’ Albion, H.A. Bruner. 464 p. . Armstrong, G. R. and J. C. Bjorkbom. 1956. The timber re- sources of New York. u.s. Dept. Agriculture, Forest Service, Northeastern Forest EXperiment Station, unnumbered publication, 37 p. Beetham, Nellie and W. A. Niering. 1961. A pollen diagram from southeastern Connecticut. Amer. Jour. Sci. 259: 69-75 0 Benninghoff, W. S. 1964. The Prairie Peninsula as a filter barrier to postglacial plant migration. Indiana Acad. Sci. Proc. (1963) 72: 116-124. and R. O. Kapp. 1962. Suggested notations to indi— cate identification status of fossil pollen (abs.). Pollen et Spores 4: 332. ——.-—-’-“I Bickford, C. A., C. E. Mayer and K. D. Ware. 1963. An ef- ficient Sampling design for forest.inventory: The northeastern forest resurvey. JOur. Forestry 61: 826-833. Birks, H. J. B. 1968. The identification of Betula nana pollen. New Phytol. 67: 309-314. Bourdo, E. A., Jr. 1956. A review of the general land office survey and of its use in quantitative studies of former forests. Ecology 37: 754-768. 290 1.0: 291 Braun, E. Lucy. 1937. Some relationships of the flora of the Cumberland plateau and Cumberland mountains in Kentucky. Rhodora 39: 193-208. ‘ 1950. Deciduous forests of eastern NOrth America. Philadelphia, Blakiston. 596 p. Bray, W. L. 1915. The development of the vegetation of New York State. New York State College of Forestry, Syracuse Univ., Tech. Publ. 3, 186 p. Broecker, W. S. and W. R. Farrand. 1963. Radiocarbon age of the Two Creeks forest bed, Wisconsin. Geol. Soc. Amer. Bull. 74: 795-802. Brosius, N. 1953. Study of present conditions in an ice age kettle hole. Sanctuary News, Sept. 1953 suppl., pp. 3-38. (mimeo., published by the Nature Sanctuary Society of Western New York, Inc., Buffalo, N.Y.). Buckley, J. D., M. A. Trautman and E. H. Willis. 1968. Iso- topes' radiocarbon measurements VI. Amer. Jour. Sci., Radiocarbon Suppl. 10: 246-294. Buehler, E. J. and I. H. Tesmer. 1963. Geology of Erie County, New York. Buffalo Soc. Natural Sciences Bull. 21(3), 118 p. Buell, M. F. 1946. Size-frequency study of fossil pine pollen compared with herbarium preserved pollen. Amer. Jour. Bot. 33: 510—516. Cain, S. A. 1940. The identification of species in fossil pollen of Pinus by Size-frequency determinations. Ibid. 27: 301-308. 1944. Foundations of plant geography. New York, Harper and Brothers. 556 p. 1948. Palynological studies at Sodon Lake: I. Size-frequency study of fossil Spruce pollen. Science 108: 115-117. .and Louise G. Cain. 1948. Palynological studies at Sodon Lake. II. Size-frequency studies of pine pollen, fossil and modern. Amer. Jour. Bot. 35: 583-591. Calkin, P. E. 1966. (Lake Pleistocene history of north- western New York, pp. 58-68. _I_r_13ueh1er, E. J., Ed. Geology of western New York, Guide Book. New York State Geological Assoc. 38th Ann. Meeting. SUNY at Buffalo, Dept. of Geological Sciences. 292 Carter, D. B. 1966. Climate, pp. 54-78. .lg Thompson,,l. H.. Ed. Geography of New York State. Syracuse, Syracuse Univ. Press. 543 p. Clausen, K. E. 1960. A survey of variation in pollen Size within individual plants and catkins of three taxa of Betula. Pollen et Spores 2: 299-304. Cline, M G. 1955. Soils and soil associations of New York. New York State College of Agriculture, Cornell Univ. , Extension Bull. 930, 64 p. Comanor, F. L. 1967. Forest vegetation and the pollen Spectra: an examination of the usefulness of the R value (abs). Diss. Abstracts 27: 1804B. Connally, G. G. 1964. The Almond moraine of the western Finger Lakes region, New York. Ph.D. Thesis, Michi- gan State University, 102 p. (abstract in Diss. Abstracts 25: 6532-6533, 1965). Cottam, G. and J. T. Curtis. 1956. The use of distance measures in phytosociological sampling. Ecology 37: 451-460. Cox, D. D. 1959. Some postglacial forests in central and eastern New York State as determined by the method of pollen analysis. New York State Museum and Sci. Service Bull. 377, 52 p. and D. M. Lewis. 1965. Pollen studies in the Crusoe Lake area of prehistoric Indian occupation. New York State Museum and Sci. Service Bull. 397, 29 p. Curtis, J. T. 1956. Plant ecology workbook, laboratory, field and reference manual. Rev. Ed. Minneapolis, Burgess. 86 p. . 1959. The vegetation of Wisconsin. Madison, The Univ. of Wisconsin Press. 657 p. and R. P. McIntosh. 1951. An upland forest con- tinuum in the prairie-forest border region of Wis- consin. Ecology 32: 476-496. Cushing, E..J. 1967. Late-Wisconsin pollen stratigraphy and the glacial sequence in Minnesota, pp. 59-88. _;g Cushing, E. J. and H. E. Wright, Jr., Eds. Quaternary Paleocology. New Haven, Yale.Univ. Press. 433 p. 293 and H. E. Wright, Jr. 1965. Hand-operated piston corers for lake sediments. Ecology 46: 380-384. Daily, Fay K. 1961. Glacial and poSt-glacial charophytes from New York and Indiana. Butler Univ. Bot. Stud. 14: 39-72. Dansereau, P. 1953. The postglacial pine period. Royal Soc. Canada Trans. 47, Series 3, Sect. 5: 23-38. Davis, Margaret B. 1958. Three pollen diagrams from central Massachusetts. Amer. Jour. Sci. 256: 540-570. . 1963. On the theory of pollen analysis. Amer. Jour. Sci. 261: 890-912. . 1965a. A method for determination of absolute pollen frequency, pp. 674-686. “£3 Kummel, B. and D. Raup, Eds. Handbook of paleonotological techniques. San Francisco, w. H. Freeman and Co. 852 p. . 1965b. Phytogeography and palynology of north— eastern United States, pp. 377—401. .Ig Wright, H. E,, Jr. and D. G. Frey, Eds. The Quaternary of the United States. Princeton, Princeton Univ. Press. 922 p. . 1966. Determination of absolute pollen frequency. Ecology 47: 310-311. 1967a. Late-glacial climate in northern United States: a comparison of New England and the Great Lakes region, pp. 11-43. _;g Cushing, E. J. and H. E. Wright, Jr., Eds. Quaternary paleoecology. New Haven, Yale Univ. Press. 433 p. 1967b. Pollen accumulation rates at Rogers Lake, Connecticut, during late- and postglacial time. Rev. Palaeobotany and Palynology 2: 219-230. and E. S. Deevey, Jr. 1964. Pollen accumulation rates: Estimates from late-glacial sediment of Rogers Lake. Science 145: 1293-1295. and J. C. Goodlett. 1960. Comparison of the present vegetation with pollen-spectra in surface samples from Browington Pond, Vermont. Ecology 41: 346-357. Deevey, 294 E. 8., Jr. 1939. Studies on Connecticut lake sedi- ments. I. A postglacial climatic chronology for southern New England. Amer. Jour. Sci. 237: 691-724. 1943. Additional pollen analyses from southern New England. Ibid. 241: 717-752. 1949. Biogeography of the Pleistocene. Geol. Soc. Amer. Bull. 60: 1315-1416. 1951. Late-glacial and postglacial pollen dia- grams from Maine. Amer. Jour. Sci. 249: 177-207. . 1957. Typical pollen-stratigraphic sequences in northeastern North America, p. 351. .;e Flint, R. F. Glacial and Pleistocene Geology. New York, John Wiley and Sons, Inc. 553 p. 1958. Radiocarbon-dated pollen sequences in eastern North America. Geobot. Inst. Rfibel Verdffent. 34: 30-37. , L. J. Gralenski and Vaind Hoffren. 1959. Yale natural radiocarbon measurements IV. Amer. Jour. Sci., Radiocarbon Suppl. 1: 144-172. Denny, C. S. and w. H. Lyford. 1963. Surficial geology and Dow, C. Droste, Dunham, Durkee, Erdtman, soils of the Elmira-Williamsport region, New York and Pennsylvania. U.S. Dept. Interior, Geological- Survey Prof. Paper 379, 60 p. M. 1921. Anthology and bibliography of Niagara Falls. 2 vols. Albany, State of New York. 1423 p. J. B., M. Rubin and G. W. White. 1959. Age of marginal Wisconsin drift at Corry, northwestern Pennsylvania. Science 130: 1760. V. L. 1965. A general ecological study of Moss Lake Nature Sanctuary. Unpublished pollen diagram, Science Teaching Library, Syracuse University. Syracuse, New York. L. H. 1960. Pollen profiles from five bog lakes in New York State. Ph.D. Thesis, Syracuse University, 60 p. (abstract in Diss. Abstracts 21: 2446-2447, 1961). G. 1943. An introduction to pollen analysis. New York, Ronald Press. 239 p. (Chronica Botanica Vol. 12). 295 1957. Pollen and Spore morphology / plant tax- onomy. Gymnospermae, Pteriodophyta, Bryophyta (illustrations). Stockholm, Almqvist and Wiksell. 151 p. 1960. The acetolysis method. A revised de- scription. Svensk Bot. Tidskr. 54: 561-564. . 1965. Pollen and spore morphology / plant tax- onomy. Gymnospermae, Bryophyta (text). Stockholm, Almqvist and Wiksell. 191 p. . 1966. Pollen morphology and plant taxonomy. Angiosperms. Corrected reprint. New York, Hafner. 553 p. ‘ Evans, P. D. 1924. The Holland Land Company. Buffalo Historical Society Publ. 28, 469 p. Faegri, K. and P. Deuse. 1960. Size variations in pollen grains with different treatments. Pollen et Spores 2: 293-298. and J. Iversen. 1964. Textbook of pollen analysis. Copenhagen, Munksgaard. 237 p. Fernald, M. L. 1925. Persistence of plants in-unglaciated areas of boreal America. American Acad. Arts and Sci. Mem. 15: 239-342. 1950. Gray's manual of botany, 8th Edition. New York, American Book Company. 1632 p. Feuer, R., W. L. Garmon and M. G. Cline. 1955. Chautauqua County soils. New York State College of Agriculture, Cornell Univ., Soil Assoc. Leaflet 3, 6 p. Fisher, D. W., I. W. Isachsen, L. V. Rickard, J. G. Broughton and T. W. Offield. 1961. Geologic map of New York. New York State Museum and Sci. Service. Fowells, H. A. (Ed.). 1965. Silvics of forest trees of the United States. U. S. Dept. of Agriculture, Forest Service, Agriculture Handb. 271, 762 p. Frederick, R. H., E. C. Johnson and H. A. MacDonald. 1959. Spring and fall freezing temperatures in New York State. New York State College of Agriculture, 'Cornell Univ., Misc. Bull. 33, 16 p. 296 Gehris, C. w. 1965. Pollen analysis of the Cranberry Bog Preserve, Tannersville, Monroe County, Pennsylvania (abs.). Diss. Abstracts 25: 4372. Gilliam, Jeanne A., R. O. Kapp and R. D. Bogue. 1967. A post-Wisconsin pollen sequence from Vestaburg bog, Montcalm County, Michigan. Michigan Acad. Sci., Arts and Letters Papers (1966) 52: 3-17. Goldthwait, R. P., A. Dreimanis, Jane L. Forsyth, P. F. Karrow and G. w. White. 1965. Pleistocene deposits of the Erie lobe, pp. 85-97. .;3 Wright, H. E., Jr. and D. G. Frey, Eds. The Quaternary of the United States. Princeton, Princeton Univ. Press. 922 p. Goodlett, J. C. and w. H. Lyford. 1963. Forest regions and great soil groups, pp. 54-55. .le Denny, C. S. and W. H. Lyford, Surficial geology and soils of the Elmira-Williamsport region, New York and Pennsylvania. U. S. Dept. Interior, Geological Survey Prof. Paper 379. 60 p. Gordon, R. B. 1937. The botanical survey of the Allegany State Park. New York State Museum Handb. 17: 23—88. . 1940. The primeval forest types of southwestern New York. New York State Museum Bull. 321, 102 p. Graustein, Jeannette E. 1967. Thomas Nuttall, naturalist, explorations in America, 1808-1841. Cambridge, Harvard Univ. Press. 481 p. Guilday, J. E., P. S. Martin and A. D. McCrady. 1964. New Paris No. 4: A Pleistocene cave deposit in Bedford County, Pennsylvania. Nat. Speleological Soc. Bull. 26: 121-194. Hafsten, U. 1961. Pleistocene development of vegetation and climate in the southern High Plains as evidenced by pollen analysis, pp. 59—91. _Ig Wendorf, F. (Ed.). Paleoecology of the Llano Estacado. Fort Burgwin Research Center Publ. 1, 144 p. Heusser, C. J. 1960. Late-Pleistocene environments of North Pacific North America. Amer. Geogr. Society Special Publ. 35, 308 p. Hill, E. J. 1895. Notes on western New York woodlands. I, II. Garden and Forest 8: 342-343; 382-383. 297 Hough, A. F. 1936a. A climax community on East Tionesta Creek in northwestern Pennsylvania. Ecology 17: 9-28 0 1936b. The dying of hemlock and other species in the Allegheny National Forest. U. S. Dept. Agri- culture, Forest Service, Allegheny (Nertheastern) Forest Experiment Station Tech. Note 7, 2 p. 1960. Silvical characteristics of eastern hem- lock (Tsuge canadensis). U. S. Dept. Agriculture, Forest Service, Northeastern Forest Experiment Station, Station Paper 132, 23 p. .and R. D. Forbes. 1943. The ecology and Silvics of forests in the high plateaus of Pennsylvania. Ecol. Monogr. 13: 299-320. Hough, J. L. 1958. Geology of the Great Lakes. Urbana, Univ. of Illinois Press. 313 p. 1963. The prehistoric Great Lakes of North Ameri- ca. Amer. Scientist 51: 84-109. House, H. D. 1924. Annotated list of the ferns and flower- ing plants of New York State. New York State Museum Bull. 254, 759 p. and w. P. Alexander. 1927. Flora of the Allegany State Park region. New York State Museum Handb. 2, 225 p. Hultén, E. 1958. The amphi-Atlantic plants and their phyto- geographical connections. Kungl. Svenska Vetenskapsakad. Handl., Fjgrde Serien, 7, Nr. 1, 340 p. Illick, J. S. and L. Frontz. 1928. The beech-birch-maple forest type in Pennsylvania. Penn. Dept. Forests and Waters Bull. 46, 40 p. Iltis, H. H. 1965. The genus Gentianopsis (Gentianaceae): transfers and phytogeographic comments. Sida 2: 129-153. . 1966. The western element in the eastern North American flora and its phytogeographic implications (abs.). Amer. Jour. Bot. 53: 634. 298 Janssen, C. R. 1967. A comparison between the recent re- gional pollen rain and the sub—recent vegetation in four major vegetation types in Minnesota (U.S.A.). Rev. Palaeobotany and Palynology 2: 331-342. Johnson, E. C. 1960. Climate of the states--New York. U.S. Dept of Commerce, Climatography of the United States No. 60-30, 20 p. Kalm, P. 1751. Cataracts at Niagara, pp. 79-94. _;g Bartram, J. Travels from Pensilvania to Onondago, Oswego and the Lake Ontario in Canada. London, Whiston and White (facsimile edition, March of America Facsimile Series, No. 41, University Microfilms, Ann Arbor, Michigan, 1966). Karrow, P. F. 1963. Pleistocene geology of the Hamilton- Galt area. Ontario Dept. of Mines, Geological Rept. 16, 68 p. , J. R. Clark and J. Terasmae. 1961. The age of Lake Iroquois and Lake Ontario. JOur. Geology 69: 659-667. Ketchledge, E. H. 1957. Checklist of the mosses of New York State. New York State Museum Bull. 363, 55 p. Kindle, E. M. and F. B. Taylor. 1913. Niagara, New York folio. _;§ Geological atlas of the United States. U. S. Dept. Interior,Geological Survey Folio 190, 26 p. King, J. E. and R. O. Kapp. 1963. Modern pollen rain studies in eastern Ontario. Can. Jour. Bot. 41: 243-252. Kovar, A. J. 1964. Pollen analysis of the Bear Meadows bog of central Pennsylvania. Penn. Acad. Sci. Proc. 1964: 16-24. Krauss, R. W. and G. N. Kent. 1944. Analyses and corre- lation of four New Hampshire bogs. Ohio Jour. Sci. 44: 11-17. Kfichler, A. w. 1964. Potential natural vegetation of the conterminous United States. Amer. Geogr. Society -Specia1 Publ. 36, 116 p. Leopold, Estella B. 1956a. Pollen size-frequency in New England Species of the genus Betula. Grana Palyno— logica (n.S.) 1: 140—147. 299 . 1956b. Two late-glacial deposits in southern Connecticut. National Acad. Sci. Proc. (U.S.) 52: 863-867. and R. A. Scott. 1958. Pollen and spores and their use in geology. Smithsonian Inst. Report 1957: 303-323. Leverett, F. 1902. Glacial formations and drainage features of the Erie and Ohio basins. U.S. Dept. Interior, Geological Survey Monogr. 41, 802 p. Livingstone, D. A. 1968. Some interstadial and postglacial pollen diagrams from eastern Canada. Ecol. Monogr. 38: 87-125. Lull, H. W. 1968. A forest atlas of the northeast. U. S. Dept. Agriculture, Forest Service, Nertheastern Forest Experiment Station, unnumbered publication. 46 p. Lutz, H. J. 1930a. Original forest composition in north- western Pennsylvania as indicated by early land survey notes. Jour. Forestry 28: 1098-1103. . 1930b. The vegetation of Heart's Content, a virgin forest in northwestern Pennsylvania. Ecology 11: 1-29 0 MacClintock, P. and E. T. Apfel. 1944. Correlation of the drifts of the Salamanca re-entrant, New York. Geol. Soc. Amer. Bull. 55: 1143-1164. and J. Terasmae. 1960. Glacial history of Covey Hill. Jour. Geol. 68: 232-241. Maher, L. J., Jr. 1964. Ephedra pollen in sediments of the Great Lakes region. Ecology 45: 391-395. Martin, P. S. 1958a. Pleistocene ecology and biogeography of North America, pp. 375-420. _;g Hubbs, C. L., Ed. Zoogeography. Amer. Assoc. Advancement Sci. Publ. 51, 509 p. . 1958b. Taiga-tundra and the full-glacial period in Chester County, Pennsylvania. ‘Amer. Jour. Sci. 256: 470-502. (Maycock, P. F. 1963. The phytosociology of the deciduous forests of extreme southern Ontario. Can. Jour. 300 MCAndrews, J. H. 1966. Postglacial history of prairie, savanna, and forest in northwestern Minnesota. Torr. Bot. Club Mem. 22(2), 72 p. McCulloch, W. F. 1939. A postglacial forest in central New York. Ecology 20: 264-271. McIntyre, A. C. and G. L. Schnur. 1936. Effects of drought on oak forests. The Penn. State College, Experiment Station Bull. 325, 43 p. Meining, D. w. 1966. Geography of expansion, 1785-1855, pp. 140-171. .22 Thompson, J. H., Ed. Geography of New York State. Syracuse, Syracuse Univ. Press. 543 p. Messenger, A. S. 1966. Climate, time and organisms in re- lation to podzol development in Michigan sands. Ph.D. Thesis, Michigan State University. 241 p. Merriam, C. H. 1898. Life zones and crop zones. U. S. Dept. Agriculture, Div. of Biological Survey Bull. 10, 79 p. Mordoff, R. A. 1949. The Climate of New York State.. New York State College of Agriculture, Cornell Univ., Extension Bull. 764, 72 p. , Morey, H. F. 1936. A comparison of two virgin forests in northwestern Pennsylvania. Ecology 17: 43-55. Morrison, T. M., C. C. Engle and G. L. Fuller. 1919. #8011 survey of Chautauqua County, New York. U. S. Dept. Agriculture, Field Operations of the Bureau of Soils (16th Report), 1914, pp. 271-326. Muller, E. H. 1960. Glacial geology of Cattaraugus County, New York. Friends of Pleistocene Geology, Eastern Section, 23rd Reunion, Guidebook. Syracuse Uni- versity, Department of Geology. 33 p. . 1963. Geology of Chautauqua County, New York. Part II. Pleistocene geolOgy. New York State Museum and .Sci. Service Bull. 392, 60 p‘.- . 1964a. Quaternary geology in New York State. The Empire State Geogram 2: 12-16. . 1964b. Quaternary section at Otto, New York. Amer. Jour. Sci. 262: 461-478. 301 . 1965. Quaternary geology of New York, pp. 99- 112. _;e Wright, H. E., Jr. and D. G. Frey, Eds. The Quaternary of the United States. Princeton, Princeton Univ. Press, 922 p. Munro, R. 1804. A description of the Genesee Country in the State of New York. Reprinted in O'Callaghan, E. B., Ed., The documentary history of the State of New York, Vol. 2, pp. 1168-1188, 1849. Albany, Weed, Parsons and Co. Nichols, G. E. 1935. The hemlock-white pine-northern hard- wood region of eastern North America. Ecology 16: 403-422. Northeastern Forest Experiment Station. 1967. Preliminary forest survey statistics, New York--l967. U. S. Dept. Agriculture, Forest Service, Northeastern Forest Experiment Station, unnumbered and unpaged publication. Ogden, E. C. and D. M. Lewis. 1960. Airborne pollen and fungus Spores of New York State. New York State Museum and Sci. Service Bull. 378, 104 pp. Ogden, J. G., III. 1966. Forest history of Ohio. I. Radiocarbon dates and pollen stratigraphy of Silver Lake, Logan County. Ohio Jour. Sci. 66: 387-400. . 1967. Radiocarbon determinations of sedimentation rates from hard and soft-water lakes in northeastern North America, PP- 175-183. .;g Cushing, E. J. and H. E. Wright, Jr., Eds. Quaternary paleoecology. New Haven, Yale Univ. Press. 433 p. Parker, Dorothy. 1936. Affinities of the flora of Indiana: Part I. Amer. Midl. Nat. 17: 700-724. Parmelee, G. W. 1947. Postglacial forest succession in the Lansing area of Michigan: A study of pollen Spectra. M.S. Thesis, Michigan State University. 74 p. Peattie, D. C. 1922. The Atlantic coastal plain element in the flora of the Great Lakes. Rhodora 24: 57-70, 80-88. Potzger, J. E. 1946. Phytosociology of the primeval forest in central-northern Wisconsin and upper Michigan, . and a brief post-glacial history of the Lake Forest Formation. Ecol. Monogr. 16: 211-250. 302 and A. Courtemanche. 1956. Pollen study in the Gatineau Valley, Quebec. Butler Univ. Bot. Stud. 13: 12-23. and J. H. Otto. 1943. Post-glacial forest suc- cession in northern New Jersey as Shown by pollen records from five bogs. Amer. Jour. Bot. 30: 83-87. Rayback, R. J. 1966. The Indian, pp. 113-120. _;g Thompson, J. H.,Ed. Geography of New York State. Syracuse, Syracuse Univ. Press. 543 p. Ritchie, J. C. and Sigrid Lichti-Federovich. 1967. Pollen dispersal phenomena in arctic-subarctic Canada. Rev. Palaeobotany and Palynology 3: 255-266. Ritchie, W. A. 1965. The archeology of New York State. Garden City, The Natural History Press. 355 p. Rosendahl, C. O. 1943. Some fossil fungi from Minnesota. Torr. Bot. Club Bull. 70: 126-138. Rowe, J. S. 1959. Forest regions of Canada. Canada Dept. Northern Affairs and National Resources, Forestry Branch Bull. 123, 71 p. Rubin, M. and Corrinne Alexander. 1960. U.S. Gedlogical Survey radiocarbon dates V. Amer. Jour. Sci., Radiocarbon Suppl. 2: 129-185. Sangster, A. G. and H. M. Dale. 1961. A preliminary study of differential pollen grain preservation. Can. Jour. Bot. 39: 35-43. . 1964. Pollen grain preservation of underrepre- sented species in fossil Spectra. Ibid. 42: 437-449. Schick, Sister Mary Salesia and S. w. Eaton. 1963. (Liver- worts, mosses and vascular plants of Waterman swamp and Allenberg bog. St. Bonaventure Univ. Sci. Studies 21: 5-51. Schmidt, K. P. 1938. Herpetological evidence for the post- glacial eastward extension of the steppe in Nerth America. Ecology 19: 396—407. Sears, P. B. 1930. Common fossil pollen of the Erie basin. Bot. Gaz. 89: 95-106. . 1932. Postglacial climate in eastern Nerth America. Ecology 13: 1-6. 303 1942. Forest sequences in the north central states. Bot. Gaz. 103: 751-761. Secrest, H. C., H. J. MacAloney and R. C. Lorenz. 1941. Causes of decadence of hemlock at Menominee Indian Reservation, Wisconsin. Jour. Forestry 39: 3-12. Shanks, R. E. 1966. An ecological survey of the vegetation of Monroe County, New York. Rochester Acad. Sci. Proc. 11: 108-252. Shantz, H. L. and R. Zon. 1924. Natural vegetation, 29 p. _lg Atlas of American agriculture, Physical basis, 1936. Washington, U. S. Government Printing Office. Sirken, L. A. 1967. Correlation of late glacial pollen stratigraphy and environments in the northeastern U.S.A. Rev. Palaeobotany and Palynology 2: 205-218. Soil Survey Division. 1938. Soils of the United States, pp. 1019-1161. .19 Soils and Men, Yearbook of Agri- culture, 1938. Washington, U. S. Government Print- ing Office. 1232 p. Soper, J. H. 1962. Some genera of restricted range in the Carolinian flora of Canada. Royal Can. Inst. Trans. 34: 3-56. ders, H. 1961. Forest and peat land at Hawley Lake, northern Ontario. National Museum Canada Bull. 171: 1‘3].- Steere, W. C. 1947. (Musci, pp. 370-490. _;e Polunin, N., Ed. Botany of the Canadian eastern Arctic. Ibid. 97, 573 p. Steyermark, J. A. 1939. Some features of the flora of the Ozark region in Missouri. Rhodora 36: 214-233. Stickel, P. w. 1933. Drought injury in hemlock-hardwood stands in Connecticut. Jour. Forestry 31: 573-577. Stingelin, R. W. 1965. VLate-glacial and post-glacial vege- tational history in the north central Appalachian region. Ph.D. Thesis, The Pennsylvania State Uni- versity, 191 p. (abstract in Diss. Abstracts 26: 6650, 1966). Suess, H. E. 1954. U. S. Geological Survey radiocarbon dates 1. Science 120: 467-473. 304 Taylor, A. E., F. B. Howe, C. S. Pearson and w. J. Moran. 1929. Soil survey of Erie County, New York. U. S. Dept. Agriculture, Bureau of Chemistry and Soils, Series 1929, No. 14, 52 p. Taylor N. 1928. The vegetation of the Allegany State Park. New York State Museum Handb. 5, 126 p. Terasmae, J. 1958. Non-glacial deposits in the St. Lawrence lowlands, Quebec, pp. 13-28. .;E Terasmae, J. Contri- butions to Canadian palynology. Geol. Survey Canada Bull. 46, 35 p. . 1959. Notes on the Champlain Sea episode in the St. Lawrence lowland, Quebec. Science 130: 334-336. and R. J. Mott. 1965. Modern pollen deposition in the Nichicun Lake area, Quebec. Can. Jour. Bot. 43: 393-404. Tesmer, I. H. 1963. Geology of Chautauqua County, New York, Part I. Stratigraphy and paleontology (Upper Devonian). New York State Museum and Sci. Service Bull. 391, 65 p. Thompson, Isabel. 1939. Geographical affinities of the flora of Ohio. Amer. Midl. Nat. 21: 730-751. Thompson, J. H. 1966. The primary sector, pp. 201-231. .12 Thompson, J. H., Ed. The Geography of New York State. Syracuse, Syracuse Univ. Press. 543 p. Turner, 0. 1850. Pioneer history of the Holland purchase of western New York. Buffalo, Jewett, Thomas and Co. 670 p. Ueno, Jitsuro. 1958. Some palynological observations of Pinaceae. Osaka City Univ., Jour. Inst. Polytechnics, Series D, 9: 163-187. Wall, R. E. 1968. A sub-bottom reflection survey in the central basin of Lake Erie. Geol. Soc. Amer. Bull. 79: 91-106. Walker, P. C. and R. T. Hartman. 1960. The forest sequence of the Hartstown bog area in western Pennsylvania. Ecology 41: 461-474. Weaver, J. E. and F. E. Clements. 1938. Plant ecology. 2nd Ed. New York, McGraw-Hill. 601 p. 305 West, R. G. 1961. Late- and postglacial vegetational history in Wisconsin, particularly changes associ- ated with the Valders readvance. Amer. Jour. Sci. 259: 766-783. Westerfeld, w. F. 1961. An annotated list of vascular plants of Centre and Huntington Counties, Pennsyl- vania. Castanea 26: 1-80. White, G. w.' 1968. Age and correlation of Pleistocene de- posits at Garfield Heights (Cleveland), Ohio. Geol. Soc. Amer. Bull. 79: 749-752. and S. M. Totten. 1965. Wisconsinan age of the Titusville till (formerly called "Inner Illinoian"), northwestern Pennsylvania. Science 148: 234-235. Whitehead, D. R. 1964. Fossil pine pollen and full-glacial vegetation in southeastern North Carolina. Ecology 45: 767-777. and D. R. Bentley. 1963. A post-glacial pollen diagram from southwestern Vermont. Pollen et Spores 5: 115-127. Williams, Claire. 1968. New and additional moss records for Ontario. Bryologist 71: 282-284. Williams, E. T. 1947. Niagara County, pp. 1-448. .22 Horton, J. T. e; e;., History of northwestern New York, Vol. 2. New York, Lewis Historical Publishing Co., Inc. 599 p. Wilson, I. T. and J. E. Potzger. 1943. Pollen study of sediments from Douglas Lake, Cheboygan County and Middle Fish Lake, Montmorency County, Michigan. Indiana Acad. Sci. Proc. (1942) 52: 87-92. wodehouse, R. P. 1935. Pollen grains, their structure, identification and significance in science and medi- cine. New York, McGraw-Hill. 574 p. Wood, R. D. 1965. Monograph of the Characeae. .ig WOOd, R. ID. and K. Imahori, A revision of the Characeae, Vol. 1. weinheim, Cramer. 904 p. and W. C. Muenscher. 1956. The Characeae of the State of New York. New York State College of Agri- culture, Cornell Unin, Agriculture Experlment Station Mem. 338, 77 p. 306 Wright, H. E., Jr. 1964. ASpects of the early postglacial forest succession in the Great Lakes region. Ecology 45: 439-448. . 1968a. History of the Prairie Peninsula, pp. 78— 88. .lg Bergstrom, R. E., Ed. The Quaternary of Illinois. Univ. of Illinois, College of Agriculture, Special Publ. 14, 179 p. 1968b. The roles of pine and Spruce in the forest history of Minnesota and adjacent areas. Ecology 49: 937-955. and H. L. Patten. 1963. The pollen sum. Pollen et Spores 5: 445-450. Yeatman, C. W. 1967. Biogeography of jack pine. Can. Jour. Bot. 45: 2201-2211. Young, A. w. 1875. History of Chautauqua County, New York. Buffalo, Matthews and Warren. 672 p. Zenkert, C. A. 1934. The flora of the Niagara Frontier re- gion. Buffalo Soc. Natural Sciences Bull. 16, 328 p. APPENDICES A. Pollen Spectra Above and Below Gyttja Samples Used C-l4 Age Determinations at Protection Bog B-H. Percentages of Minor Pollen and Spore Types not Shown on Pollen Diagrams NOTE The notation "comp." after certain pollen taxa indicates the degree of certainty in the identification of difficult pollen grains (Benninghoff and Kapp, 1962). It is used when a grain is provisionally assigned to a taxon and to convey that un- certainty exists about the conclusiveness of identification. If no notation follows a taxon, the identification is con- sidered positive. POLLEN SPECTRA ABOVE AND BELOW GYTTJA SAMPLES USED FOR C-l4 AGE DETERMINATION AT PROTECTION 306 307 APPENDIX A Depths (m) Taxa 3.56 Arboreal Pollen (AP)1 Picea Abies Larix Pinus undifferentiated 'g. Haploxylon '2. Diploxylon Juniperus Tsuga lfifllEE Acer saccharum Tilia Fraxinus 4-colpate Juglans cinerea Carya Quercus Ulmus Betule Frax1nus 3-colpate Acer rubrum/saccherinum Carpinus—Ostrye Corylus Platanus % AP NOnarboreal Pollen (NAP)l .5122E Salix Viburnum Rosaceae Cyperaceae Gramineae Ambrosia Artemisia Xanthium High-spine Compositae Cheno-Am. .3222222125 Thalictrum % NAP Misc. pollen (Mp)2 Ericaceae Aweher Sarracenia Polypodiaceae Osmundaceae broken Abietineae unfamiliar unknown H 0‘ 0) U1 00003-500000an I'-‘ OI wo—amcsr—at-wom m 0 £0 97.3 00000 wNuwm 0.5 0.2 0.5 2.7 DON mu: MN mumbme-Iwmcsq h‘lcac3Hanhqh-or-c3q \D m dub N MN HO‘O‘NHOOOI—‘ON:U)NOOOO . O O O O O O O C O O NMOWOHO‘WUQHINO‘OwO‘Q 000‘ w £~¢9 mdomm 92¢ Zmddom ZOZHZ m0 mmmCsZmummm n XHDZWQQ‘ 3()9 H.o m.o I I I H I ' I O I N00 III III III III III III III III III III III III III III III III III III III N00 N00 III III III III III N00 III III III III III III III III III III III III III III III III III N00 III III III III III III III II III III III III Noo III III III III III III III III Noo III III III N00 III III III III III III III Noo III III III III III III III III III III III III III III III III III III III III III N00 III III III III III H00 III III III III III III ”.0 III III -- -- «.0 «.0 «.0 «.0 -- «.0 -n -- -- -n- -n I I l I I I N O I I I I I I I I I m 0 I I I V0 0 I I I I I I _m«m.m mNh.m mN¢.m mNN.m mnm.m mNh.N mh¢.N mNN.N mum.H mNh.H va.H mNN.H mNm.o mNh.o mhm.o mNN.o mbH.o mNH.o who.o mmo.o ooo.o aeaaessotbotqdo erzetnatzqn mnrue 12 s 9:19:31 as aeze;tttaqmn mn130119qm aeaoegqnu sntnaunueu aesouymnbaq 89191qe1 31; oo aeaaettfiqdofizeo aeeaeaou s 'dmoa anumnqu anqqunt q as (u) aqadaa exam "iddudHQ zmddom 20 4 ZOHBUNm I won ZOHEUDOE 230% 82 man; Edam a flag 82H! m0 mflagflummm U anmemfl 3IL() N.o m.o III III III III III III III III III III III III III ”no III III III III III 0.0 -n n-n -- -- -- -n «.0 -n- -n -- -- n- -- -n- -- -- -- -n III III III III III III III III N00 III III III III N00 III III III III III III III III III III III III III III III III III III III III III III Noo III n- -- -- -- -- -n- -- 0.0 «.0 n- -- -- -- -n -- -- -n «.0 -- n- -- -- n- -- -- «.0 0.0 «.0 «.0 n- -- -n- -- -n- -- -- -- -n- III NCO III III III III III III III III III III III III III III III III III III III III III III Nco III III III III III III III III III III III III III n- nnn -- «.0 -- -- -- «.0 -- -- -- -- -n -- n- -n- n- «.0 «.0 III III III N00 III III III III III III III III III III N00 III III III III III III III N30 III Noo III III III III III III III III III III III III III III III III III III III III III III III III III N00 III III III III III III III III III 0 III III III III III III III III III III III III III ”.0 III -- -- «.0 . -- -- -- -- -- «.0 -n- n- n-n -- -- -- n- -- - Hum Hum Hum mwm Hum Hun Hum mnm Hun um“ ”H” ”H” H”- -HH Hum Hun Hum nun ”- -H- H-H ”mm 0.0 -- -- -- «.0 -- «.0 «.0 -- -- -n «.0 n- «.0 n- -n Hm” “um um“ Hum mwm Hun Hun Hun ”um ”um nun nnn -HH -nu -HH -Hn nnn nnn Hun EIJEIDDIIQQ 9 eruaoelzes 911233! as eoeu; JaSOJd seqque new eeeq N Jeq nu etuaseza mntqauex aezaJIIIaqmn WUJQDTIEQL xamnu aeeaetqna ~dmoo eueyquas aeeaeIIquofizeo aeaoesoa -dmoo snumequ anque omen 929020113 239 -dwoa sua,: xeqmeptn 11 9X9 (m) sqndaa m ZOHBUmm I 000 20830902 HZ9 mmomm 02¢ ZNAAOQ MOZHZ ho mMQQBZNUMNm U ZOHBUmm I 00m Ommmzmqflfl O NHQZNQA‘ 31L4: 0.0 v.0 N.o v.0 v.0 m.o H.o H.o H.o H.o H.o m.o 0.0 III III III III III III III III IIIIII III III III III v adK: ITIN93U“ aeaoessotbotqdo alods 12 Brzetnorlqn B erzazqanaqas etuaoexles eaeu: 135015 wntt q 0:1 saqque new aeaoeatnzn e ezgrxes ~dmoo sntnaunuea 'dmoa mnxuomatod eyqoemts q 'dmoa acnexqeq . -owoo euetquas -dmoa mnzn m3 -dmoa se 10 8"! BUBpEUED BDUJOD PIIOJIPUDIOI ntnue meg mntqoqqnaazv snonqmes anque omen anqqueteq 93 0d N snIn a T1 Jeqmeptn a: In suet np exam (m) sundaa WXMOS Edam wmdd<> mmmmzmu uzJ anooaOcaEmO III mmoofuonfol ceamoEoSQ s T ‘\‘J 1 Ll \ ‘ LIT ‘‘‘‘‘ Eniozugk + omnocfa 7 on:monEoU ofam zm_I , ouoEoIozoflv u_m_EoT_< , [I ‘ . , , u_mo.nF_< r IIL HON ,T oquEuco 7+ I T 1 I‘ + ;r III} annoucon>0, ,’ xzum ,.| VI m::_< L * AT IIIK I ’ l ‘T‘ wisdom I+++ W + mJZCOU + extanOLmzcacuu In _ Earns. too< L I o~mn_oo.n mafxutu 33.00 mar—:3 accounuo , + l f LII nautoae uxcao , cocofo magma... I fl . 7 l I m.ma.ou.v mafxutl _ . I + ++ 2.; i , Eatnguompu. tood‘ m3m5l Muzak :o_>xo_a_om:c_n. r . + + Ia , 0 co_>xo_nmz wafl I r T , s U l n _ , i , P ,, ‘ \ ,f r ‘ \ no.m2:ota:_u:3 mafn. r I . ‘ x \ , ‘ ‘ IT wk . men r , ww_n< : l cmoiu I i P l, . 53. T O P 2 3 ON 4 5 O B 6 7 B 9 3 2 5 30 0% ,1 9 4H 91 +_ #. 4. \\ \\\\\\\\ Ailinn: A A.llil,atrlll A A @ Silty-clay gyttja I‘Tr" .1 NGM 1968 1 50 Percent DSilty clay geynja ‘Undifferentiated Peat 316 DIAGRAM 2v HOUGHTON BOG-SECTION AI RELATIVE POLLEN FREQUENCY b mocoN _3 CI IIIII‘ :30”..ch I tm___rcnwc3 + mmoczfnxx coxocm omoofloonooxn , ommouU::EmO W onoofuonfol Econ—rm I anazz couamoEooon Eceoatnmm wzcacnnoEoZ DflOUGU_Lm omuucfll E3t>aomum onam .mE<.o:o:0 onZwonEoOoanzBI I onoEOCOLBU TI + m_mo._nE< neN + ouwaaLO , + + t + i amouncoaxo I I v 525 3:3 FI %\ \\\\\\\§ Mn__enmmiflIlIl VILII + m>§m0.m3c_ntno_r + + + i «Tao, m3c_xmtu_, 5:; , I , I l . mammm mmzmk to;xo_a_u maria II + J ’ 7 ,, co_>xo_QMz m3:_n_ L I :30. wsciu/I no.m_~cmc0:_ucz w:c_n_ ‘ 334 : mourn rl+ i m . TI'IIzzt.‘ $5530 0 - 0 5 NGM 1968 ,II Percent I L -Gyttja {-IUndifferentiated Peal 317 DIAGRAM 3. HOUGHTON BOG-SECTION Bl RELATIVE POLLEN FREQUENCY outwoarcoo 050m LEI 7 7 , II uoEichotgca m::.& 7 I mocoN 7 c30cxc3 LflzEEEJ 7 7 onocCEn< coxoLm , osouflooaoutj , oamonuczEmO NGM 1968 LIJ sums .mEdTOCozU a_m_EuI< n_moLnE< ouoEEuLO amountonxo mszvcflmznoo 52532 12.5 m:c_< EILIIII II §N\\\\\ mzcnufl mgaaon. 943.00 a> . C : .LLuU Eaaom mar—:3 mocnumao _ 17l7L77I 77777777i7717 I 7 m e 3.9.0307 7 c , r e u>LaO , , 7 P magmas. I I +7 77 7 WI 7 7 I maExSL7 7 II n___.r7 r I, 7 ,7 7 .004 ,A L ,7 707 mamml 77 7 7 7 7 7 , .7 7 amamk 1 co_>xo_n_u mafn. :o_>xo_nmc wJEn. I I LL; 7L II I LI I::IC'3Y Pinus IoIaI a 1“ H a 3:2 I M L 7 I 7 7 m mega 7 . L + t A r + ,. ............. L ............. k .................. , ......... . ............. p 4 . 7, 7.. W IIIIIIIIIIIII J IIIIIIIIIIIII 7. ............. 7 ............. a ............. a ..... AEvfiaoo 4 5 6 7 8 9 318 Figure 7. Houghton 809- Section B: Number of Terrestrial Pollen and Spores Per Ml. of Wet Sediment ‘ I C-3 —? 5 .. C’2 N e - g e O A O E _ 2 V a. 5 a 3 7 - 2*. ° 3 C-1 8 I- 9 b $0 8 A 10L 1O 2O 30 4O 50 Number of Terrestrial Pollen end Sporee (x104) 319 DIAGRAM 4‘ HOUGHTON BOG-SECTION Bi ABSOLUTE POLLEN FREQUENCY wecoN, :30cxc3 +Ln_:rcue:37 emoc23n4 cexotm wooxznouteol, ante; .um__2 Ewfitofxx, u_mo._nrc< mDDLLm _nuOI—. maznufn. wisdom. w3_>LoU me.mO.w::_QLn0 Ezaem 22.53 nocm~an maoLmJO M>LmU mcEmjfi w:c_xmLL 32F Loo< mamnu, amamk co_>xo_o_o wast c0_>xo_nmr_ m::_n_ cognizmteetpca mafl ma_n< muo.n., AEZEmo /Pir1ustotal 50 Thousands of grains per ml. 0 77747: NGM 1968 320 SECTION A: RELATIVE POLLEN FREQUENCY DIAGRAM 5. ALLENBERG BOG Lu___rw_cndxfl _ onec:e_n<.coxo._m m— . I + 7 Escmaznw oneusuczEmO enoufooaooj + + + + .7 Ln. . mmconwcm 7 meazaExz mmzazu)c02 mazacunchZ ouoouqu s M u s omgcfu EaanomuL xméam .mEdTOCOLU outmanoU ofnm LEI osoEOCOLBU 335.324 n_moLnE< noN omeEEaLmv unconcoacao x__uw mac—4 macuufl magnoa na_>LoO M>L«m0.m:c_QLaU EJEanoonflEatnsc L004 oauu_oo-n m3,:_xu‘_u 5:500 m3E_D cocn«un 2.0.030 nxcuo decaf". m:u_m:1 o«nn_0u.v m:c_xu..L SIP EILEIOUGm LOU< usual swank co_>xo_u_v wafn co_>xo_uaz mask :30“ m:c_l Unificcetowtucz «:50. xCuJ m0_n< mega. “Evfdoo 7 50 Percent -Undifferentiated peat NGM 1968 321 DIAGRAM 6. ALLENBERG BOG-SECTION BI RELATIVE POLLEN FREQUENCY :30c1CD Lm___Emch wnwczflnxx :mxocm ._ 0 mmwofloioaog: _ C Ecwwmtm nonLnE>Z mazacmaoEmZ unounoiw E320__m:.r .wE<.0csz entmonEoO wEnm LEI , 22:57.4 EwoLnE< wmeEnLO mmwumeQ>U x__mw w::_< 7 I \\\\\\\\\\\\\\\\\\\\ \M 7 macnena miaaou I 7 I c +7 . +7 + + w3_>LoU 7 + 50 Percent n mfuwonocatmo . .:3:_ J .32. L004 I 7 7 I 7 , 7 mangooIm m:c_xmLL 7 msuvm mar—:3 NOCNumflo 7 7 m>LnU mmeEu w:n_m:1 wungooIv waEanm «7:... EJLMLUUNW .004 maan nmamk co_>xo_n:o marzn. 'I‘Pinus Total wDEQ £3.73 mJEl meI. $54 mega TS £an |7t r7477 A 747‘ 0 j Undifferentiated fine peat ‘7 I @Gyma @ CIay—gyttia jSame, with Drepanocladus C NGM 1968 ‘ISedge peat 1 DIAGRAM 7. ALLENBERG BOG-SECTION C: RELATIVE POLLEN FREQUENCY 3V BM”. 3 IGV NIMOUC WhNOVI-Idfi EVIDV IOOdODAW QVIOVONI'IWBO BVBOVIOOdA‘IOd WLMWOLOd VINIBVUG UVHan VHdAL QOHLN N "£9513; any I DI ‘I VH1. 09 VLNVTd xam ’SWV‘ 3H EVLISOde INIdG HDIH iVBOIOIUOHDID VISIWBLUV VISOUCWV EVENIWVHS BVQOVNEdAD VOIHAW XI'IVS SnN‘IY SfiNVLVWd smnaoa Sf'nAUOO VAULBO- SandUVD wnusnu 833V BLVdWOD‘C BONIXVHJ V‘IOLES 3 “N1” Bnouanb V3 VBUINIO BNVTMI‘ ilVd'IOO-' enmxvun VIWIL wnavuoavs 833V snow VBOSI NO‘IAXOWdIO Sand NO‘IAxdeI-I Sand Oi .LVIINBUIiiIW BflNld XIUVT SEIOV V3 OId “ADI-H.630 euue . r- I Dy 0 v V I I -71 _ 77I 7‘ A I _-_7—__—7_ 4 ‘ __.—.__ _ I0 » I . 7 A—7 A - - ,- , .__..__._—_ A 7 A vv 9 7 _ .,-. ‘ , 7 ¢ : AIR, TITA I 7 7i“ .1 Ir. it 7IIiYI7II II 7.7L: ‘ + 7 ‘ I #777 _I 7AA ‘“ I I I “— _a————+ +| I I I ‘OPINUS TOTAL 7|7 rv-—r 4 i74.i.7r.i7ii7il.i.i.i7i.i7l7.li7n.iti.l.iti.i.1.1.1777ii1.7t|77itl.i.iti.i.i.1.l.1.i7171 AL‘I.I. SCI-y Percent 50 17.77.7777 0 H‘ECley-gyme [jayw- NGM 1968 Depth (m) 323 Figure8. Allenberg Bog-Sectionc= Numberof Terrestrial Pollen and Spores Per Ml. of Wet Sediment 11 - O 0 C-1 0 12 - . . — O ' e ' _ N ' 3 13 . ' . e e 0 e ‘ 0 ° on e :I e A g 14 - ' 3' e a O O O O O . — 15 - '. 1' e 10 2O 30 4O 50 Number of Terreetriel Pollen end Spores (x1 04) DIAGRAM 8. ALLENBERG BOG-SECTION C: ABSOLUTE POLLEN FREQUENCY saNozI NMONMNFI BVI‘IIWVdNn EVENILBIEV NBMOBS 'IVIOL VLAHdOOIUSLd SBHQH 'DSIW VISIWSLHV BVJJSOdWOD BNIdS HQIH VISOHEWV EVENIWVBQ EVBOVHSdAO VQIHAW XI‘IVS an'lV snMaoo snwneoa SDNVLV'Id VAHLSO-SnNIdHVD WOHBHH 830V BLVdWOD-C snmxvas VWnLBQ an‘ln SHOUBOO VAMVO VEHENIO smnenr 31Vd103-7 SONIXVUA ‘ VIWII ‘ WHHVHDOVS H! 3V 3 OOVd V90 SJ. NO‘I A XO‘IdIO SnNId NO'IAXO1dVI-I Sand OEIVIINBUEAJIONH SrINld XIHV'I SBISV VBOId (\u) Hie-130 324 A 7‘ A77 ‘ I *— I + M __ _. A I 77 ‘ 4 A77 ‘ - ——_,7 A7 _ 7A - I 777*.— I» 7I,_ ' 7L 7 V ‘ —._ r I ++ ‘+ I A I I ‘7 7441 77‘ +++ ++ 12‘: I [:ICley .IIIIILLU-L.A.;7A.I.A77L‘4 . “.1. “WMWAAILM Iti.Itttl.i.iLi.I ”II .I.-. NGM 1968 50 - w Thoueende of grelne per ml. 0 LA Ink Cley- gyttja A' 1' Ifi‘ L7} Gym. 1 RELATIVE POLLEN FREQUENCY .INIUII VALLEY FIAT WORK. DIAGRAM I. J'IIIIIII‘I‘n v «emu. Ipeleel‘l ewe-e:- uooelewe‘e‘ elueefl‘ Illuflflfld Onealbnj I e eeee' UHIQOII‘i n seeing 'IIIV -eueu3 elueeluoa eulde «III-t II IIOHOUOIO euweuy eaewv .IIUIUII‘ eIAD new!" III'. enupy elm-um enilneo "l..°"fl|l|¢l.3 lune. enwln enamel. in”. :eev ent-1 elne; ‘l! "we ueM-elluu Illa" eeuuue‘euleun Inn" IOIQV (IIINAI‘OO 325 i .. I e z - C I C r O I. O 3 D I O O E I I. 8:. 2 O O O a C = h I D 3 g? :l‘ h: éi “wanna!"