PLANT MACROFOSSIL METHODS AND STUDIES/Validation of Pollen Studies 2413 Validation of Pollen Studies pollen grains. This imparts one of the three unique features of plant macrofossil data compared to pol- S T Jackson, University of Wyoming, WY, USA len: high taxonomic resolution. Macrofossils allow R K Booth, Lehigh University, PA, USA us to surmount the taxonomic smoothing frequently ª 2007 Elsevier B.V. All rights reserved. imposed by pollen data. The second unique feature of plant macrofossils is their local nature. Plant macrofossils generally do not Plant macrofossils and pollen play complementary travel far when airborne; the primary mode of trans- roles in paleoecology. Each has its own advantages port beyond 101–102 m is by surface waters or ani- and shortcomings. Used separately, they provide mals. For depositional basins with no inflowing valuable information about past vegetation. Used streams, most plant macrofossils derive from within together, the two types of data can corroborate or, 101–102 m of the shore. For depositional environ- occasionally, challenge each other, and they can pro- ments with a fluvial component, macrofossils must vide more detailed and refined understanding of past have been derived from plants growing somewhere vegetation at multiple spatial scales. In this article, within the catchment above the point of deposition. we briefly compare and contrast the flora- and vege- In contrast, pollen grains can be transported by wind tation-sensing properties of pollen and macrofossil for distances up to 104 km, although more typical data and then present a series of case studies in travel distances are on the order of 101–105 m. which the two data sets have been used together Pollen assemblages in a lake or wetland basin repre- profitably at spatial scales ranging from local to con- sent a distance-weighted integration of vegetation tinental. We focus on macrofossils from ‘wet’ sedi- surrounding the basin, with significant representa- ments (lakes, peatlands, buried soils, fluvial deposits, tion of distant vegetation (i.e., >103 m). In contrast, etc.) and do not explicitly treat macrofossils from plant macrofossils provide a more-local sampling of rodent middens and archeological sites. vegetation, undoubtedly distance-weighted, but with little or no representation beyond ca. 102 m (Table 1). Vegetation-Sensing Properties of Pollen Many plant taxa are poorly represented in pollen and Plant Macrofossil Data assemblages due to low productivity, poor dispersal, Pollen grains comprise a distinct and uniform stage in poor preservation, or adaptations for zoophilous the life history of seed plants—the gametophyte gen- transport. Many of these taxa—Pseudotsuga, Larix, eration. Pollen grains are small, with a restricted Abies, Dryas, Populus, Oxyria, Polygonum,and range of sizes (5–150 mm), and most are packaged Nymphaeaceae—are well represented in plant in sporopollenin, a refractory organic compound. macrofossil assemblages from lakes and other Plant macrofossils from seed plants represent the sporophyte generation. However, because seed plant sporophytes are multicellular, differentiated Table 1 Systematic studies of modern representation of terres- entities, plant macrofossils can take any of a variety trial plant macrofossils in lake sediments of forms: seeds, fruits, anthers, microsporangia, Birks HH (1973) Modern macrofossil assemblages in lake megaspores, ovulate cones, leaves, needles, petioles, sediments in Minnesota. In: Birks HJB and West RG (eds.) bud scales, buds, twigs, branches, etc. The size range Quaternary Plant Ecology, pp. 173–189. Oxford, UK: Blackwell. is potentially vast, from seeds <1mm3 to buried Drake H and Burrows CJ (1980) The influx of potential 3 macrofossils into Lady Lake, north Westland, New Zealand. trees >20 m . Materials are also variable, although New Zealand Journal of Botany 18:257–274. macrofossils preserved in sediments tend to be com- Dunwiddie PW (1987) Macrofossil and pollen representation of posed of the more refractory compounds (particu- coniferous trees in modern sediments from Washington. larly cellulose and lignins). Ecology 67:58–68. Pollen grains are typically morphologically invar- Hill RS and Gibson N (1986) Distribution of potential macrofossils I: Lake Dobson, Tasmania. Journal of Ecology 74:373–384. iant among species within most anemophilous genera Jackson ST (1989) Postglacial vegetational changes along an and among many genera within certain families. In elevational gradient in the Adirondack Mountains (New York): contrast, interspecific morphological variation of A study of plant macrofossils. New York State Museum and sporophyte plant organs is typically large since Science Service Bulletin 465. Wainman N and Mathewes RW (1990) Distribution of plant plant species are generally described and distin- macroremains in surface sediments of Marion Lake, guished based on sporophyte morphology. southwestern British Columbia. Canadian Journal of Botany Accordingly, many plant macrofossils can be identi- 68:364–373. fied to species, in contrast to the corresponding 2414 PLANT MACROFOSSIL METHODS AND STUDIES/Validation of Pollen Studies depositional environments. Thus, plant macrofossils of high decomposition and low sediment accumula- can provide information about taxa for which the tion rates. Deeper waters with good preservation and pollen record may be unreliable—the third unique high sediment accumulation rates will also have few feature of macrofossil data. This complementarity macrofossils because of poor offshore dispersal. does not hold for all taxa, however; some plants are Highest macrofossil concentrations will occur in poorly represented in both pollen and macrofossil intermediate depths, deep enough to preserve macro- assemblages. For example, Liriodendron, an impor- fossils and close enough to shore to ensure high tant tree species of eastern North America, is ento- macrofossil influx. Since most paleoecologists obtain mophilous and hence almost completely absent from cores from the deepest portions of lakes, it should pollen records, regardless of local or regional abun- come as no surprise that macrofossils are often dance. Fruits, bud scales, and leaves of Liriodendron absent or rare in sediment cores. are large and poorly dispersed, and hence show up Significant sorting of macrofossils occurs with dis- only rarely in macrofossil assemblages. Some taxa tance from shore due to differential dispersal (and, in well represented in pollen assemblages (e.g., some cases, flotation) properties of various plant Quercus, Carya,andCorylus) are poorly represented organs. Large, coarse materials (large, gravity- or as macrofossils in many depositional settings. animal-dispersed seeds and fruits, ovulate cones, large leaves, branches, etc.) are rarely dispersed more than a few tens of meters from shore, where Macrofossil and Pollen Representation they are likely to sink and then decompose in shallow benthos. Smaller organs (small seeds and fruits, con- Within and Among Depositional ifer needles, and bud scales) are dispersed farther Environments from shore and hence are better represented in off- Representation of plant macrofossils is far more sub- shore sediments. In deep waters more than 100 m ject to influences of site type and local depositional from shore, only small fragments of small-diameter environment than pollen. Pollen grains may be sub- conifer needles, papery bud scales, and small winged jected to some secondary sorting by fluvial transport fruits (e.g., Betula papyrifera and B. pubescens) are and sediment focusing. However, pollen assemblages typically present. The relative contributions to off- from diverse depositional environments (lakes, wet- shore deposition of airborne dispersal, flotation, and lands, forest-floor moss polsters, soil samples, allu- skating across winter ice are inadequately known. vium, caves, and rodent middens) are generally Selection of coring sites for macrofossils thus similar within a region due to widespread airborne represents a trade-off between maximizing macrofos- transport and mixing prior to deposition. In contrast, sil preservation (deep water) and deposition (near macrofossils are not subject to the same degree of shore). For large lakes, it is advisable to run a transect mixing due to large size and poor dispersal. of short cores or dredge samples to determine the Furthermore, macrofossils are more subject to sort- ideal site for macrofossil deposition. Alternatively, ing by airborne and fluvial dispersal processes than careful selection of lakes can yield good macrofossil pollen (Spicer, 1989; Greenwood, 1991). results. In particular, small or moderate-sized lakes with deep waters 10–50 m offshore and steep sides can have very high macrofossil concentrations, yield- Lakes ing excellent records of upland terrestrial vegetation. Lake sediments are generally the most widely used Lake sediments are typically sampled using piston- sources of both pollen and plant macrofossil data. corers with diameters ranging from 5 to 10 cm. Thus, Within-lake variation in pollen assemblages is signif- sample volume is limited, and some plant organs of icant, but the variation is generally small relative to moderate size or intermediate dispersal properties variation among lakes. In contrast, plant macrofossil (e.g., Acer samaras, most deciduous leaves, and assemblages within a single lake vary substantially, coarse conifer needles) are not reliably sampled: depending on water depth, distance from shore, and