Kenneth L. Cole

Late Quaternary Zonation of Vegetation in the Eastern Grand Canyon

Fossil assemblages from 53 packrat middens indicate which plant species were dominant during the last 24,000 years in the eastern Grand Canyon. Past vegetational patterns show associations that cannot be attributed to simple elevational displacement of the modern zones. A model emphasizing a latitudinal shift of climatic values is proposed.

Late Pleistocene vegetation of Kings Canyon, Sierra Nevada, California

Seven packrat midden samples make possible a comparison between the modern and late Pleistocene vegetation in Kings Canyon on the western side of the southern Sierra Nevada. One modern sample contains macrofossils and pollen derived from the present-day oak-chaparral vegetation. Macrofossils from the six late Pleistocene samples record a mixed coniferous forest dominated by the xerophytic conifers Juniperus occidentalis, Pinus cf. ponderosa, and P. monophylla. The pollen spectra of these Pleistocene middens are dominated by Pinus sp., Taxodiaceae-Cupressaceae-Taxaceae (TCT), and Artemisia sp. Mesophytic conifers are represented by low macrofossil concentrations. Sequoiadendron giganteum is represented by a few pollen grains in the full glacial. Edaphic control and snow dispersal are the most likely causes of these mixed assemblages. The dominant macrofossils record a more xeric plant community than those that now occur on similar substrates at higher elevations or latitudes in the Sierra Nevada. These assemblages suggest that late Wisconsin climates were cold with mean annual precipitation not necessarily greater than modern values. This conclusion supports a model of low summer ablation allowing for the persistence of the glaciers at higher elevations during the late Wisconsin. The records in these middens also suggest that S. giganteum grew at lower elevations along the western side of the range and that P. monophylla was more widely distributed in cismontane California during the Pleistocene.

Reconstruction of past desert vegetation along the Colorado River using packrat middens

Abstract The paleoecological reconstruction of Pleistocene deserts along the Colorado River of western North America is attempted using data from fossil packrat middens. The Colorado River drainage is set into a physiographic context along a gradient from the hyperarid Colorado Desert to the moist high elevations on the Colorado Plateau. The Pleistocene and modern distributions of individual plant species along this corridor are compared emphasizing records from the Picacho Peak area of the Colorado Desert and the eastern Grand Canyon. In general, plant species are now distributed 700–900 m higher in elevation and 400–700 km further up-river than they were during the late Wisconsin, however, some plant species have not conformed to the general pattern emphasizing the individualistic nature of species distributions in time.

The lower Colorado River Valley: A Pleistocene desert

Abstract A chronological sequence of plant macrofossil assemblages from twenty-five pack rat middens provides a record of desert scrub vegetation for most of the last 13,380 yr B.P. from a hyperarid portion of the lower Colorado River Valley. At the end of the late Wisconsin, and probably during much of the Quaternary, the Picacho Peak area, Imperial County, California, supported a typical Mohave Desert association of Larrea divaricata (creosote bush), Coleogyne ramosissima (blackbrush), Yucca brevifolia (Joshua tree), and Y. whipplei (Whipple yucca). Recent arrivals of Sonoran Desert plants such as Olneya tesota (ironwood) and Fouquieria splendens (ocotillo) suggest that the area supported relatively modern Sonoran desert scrub species for relatively short periods during interglaciations.

Late Holocene vegetation changes in Greenwater Valley, Mojave Desert, California

Abstract Small-scale late Holocene vegetation changes were determined from a series of 13 modern and fossil packrat middens collected from a site in the Greenwater Valley, northern Mojave Desert, California. Although the site is above the modern lower limit of Coleogyne ramosissima (black-brush), macrofossils of this shrub are only present in samples younger than 270 yr B.P. In order to measure changes more subtle than presence vs absence, macrofossil concentrations were quantified, and principal components and factor analyses were used to distinguish midden plant assemblages. Both the presence/absence data and the statistical analyses suggest a downward shift of 50 to 100 m for Coleogyne (blackbrush) communities between 1435 and 1795 A.D.

Geographical and climatic limits of needle types of one- and two-needled pinyon pines.

Aim The geographical extent and climatic tolerances of one- and two-needled pinyon pines (Pinus subsect. Cembroides) are the focus of questions in taxonomy, palaeoclimatology and modelling of future distributions. The identification of these pines, traditionally classified by one- versus two-needled fascicles, is complicated by populations with both one- and two-needled fascicles on the same tree, and the description of two more recently described one-needled varieties: the fallax-type and californiarum-type. Because previous studies have suggested correlations between needle anatomy and climate, including anatomical plasticity reflecting annual precipitation, we approached this study at the level of the anatomy of individual pine needles rather than species. Location Western North America. Methods We synthesized available and new data from field and herbarium collections of needles to compile maps of their current distributions across western North America. Annual frequencies of needle types were compared with local precipitation histories for some stands. Historical North American climates were modelled on a c. 1-km grid using monthly temperature and precipitation values. A geospatial model (ClimLim), which analyses the effect of climate-modulated physiological and ecosystem processes, was used to rank the importance of seasonal climate variables in limiting the distributions of anatomical needle types. Results The pinyon needles were classified into four distinct types based upon the number of needles per fascicle, needle thickness and the number of stomatal rows and resin canals. The individual needles fit well into four categories of needle types, whereas some trees exhibit a mixture of two needle types. Trees from central Arizona containing a mixture of Pinus edulis and fallax-type needles increased their percentage of fallax-type needles following dry years. All four needle types occupy broader geographical regions with distinctive precipitation regimes. Pinus monophylla and californiarum-type needles occur in regions with high winter precipitation. Pinus edulis and fallax-type needles are found in regions with high monsoon precipitation. Areas supporting californiarum-type and fallax-type needle distributions are additionally characterized by a more extreme May–June drought. Main conclusions These pinyon needle types seem to reflect the amount and seasonality of precipitation. The single needle fascicle characterizing the fallax type may be an adaptation to early summer or periodic drought, while the single needle of Pinus monophylla may be an adaptation to summer–autumn drought. Although the needles fit into four distinct categories, the parent trees are sometimes less easily classified, especially near their ancestral Pleistocene ranges in the Mojave and northern Sonoran deserts. The abundance of trees with both one- and two-needled fascicles in the zones between P. monophylla, P. edulis and fallax-type populations suggest that needle fascicle number is an unreliable characteristic for species classification. Disregarding needle fascicle number, the fallax-type needles are nearly identical to P. edulis, supporting Little’s (1968) initial classification of these trees as P. edulis var. fallax, while the californiarum-type needles have a distinctive morphology supporting Bailey’s (1987) classification of this tree as Pinus californiarum.

Past and current trends of change in a dune prairie/oak savanna reconstructed through a multiple-scale history

. The history of a rapidly changing mosaic of prairie and oak savanna in northern Indiana was reconstructed using several methods emphasizing different time scales ranging from annual to millennial. Vegetation change was monitored for 8 yr using plots and for 30 yr using aerial photographs. A 20th century fire history was reconstructed from the stand structure of multiple-stemmed trees and fire scars. General Land Office Survey data were used to reconstruct the forest of A.D. 1834. Fossil pollen and charcoal records were used to reconstruct the last 4000 yr of vegetation and fire history. Since its deposition along the shore of Lake Michigan about 4000 yr ago, the area has followed a classical primary dune successional sequence, gradually changing from pine forest to prairie/oak savanna between A.D. 264 and 1007. This successional trend, predicted in the models of Henry Cowles, occurred even though the climate cooled and prairies elsewhere in the region retreated. Severe fires in the 19th century reduced most tree species but led to a temporary increase in Populus tremuloides. During the last few decades, the prairie has been invaded by oaks and other woody species, primarily because of fire suppression since A.D. 1972. The rapid and complex changes now occurring are a response to the compounded effects of plant succession, intense burning and logging in the 19th century, recent fire suppression, and possibly increased airborne deposition of nitrates. The compilation of several historical research techniques emphasizing different time scales allows this study of the interactions between multiple disturbance variables.

Carbon isotopes from fossil packrat pellets and elevational movements of Utah agave plants reveal the Younger Dryas cold period in Grand Canyon, Arizona

Carbon isotopes in rodent fecal pellets were measured on packrat (Neotoma spp.) middens from the Grand Canyon, Arizona. The pellet samples reflect the abundance of cold-intolerant C4 and Crassulacean acid metabolism (CAM) plant species relative to the predominant C3 vegetation in the packrat diet. The temporal sequence of isotopic results suggests a temperature decline followed by a sharp increase corresponding to the Bolling/ Allerod‐Younger Dryas‐early Holocene sequence. This pattern was then tested using the past distribution of Utah agave (Agave utahensis). Spatial analyses of the range of this temperature-sensitive CAM species demonstrate that its upper elevational limit is controlled by winter minimum temperature. Applying this paleotemperature proxy to the past elevational limits of Utah agave suggests that minimum winter temperatures were ;8 8C below modern values during the Last Glacial Maximum, 4.5‐6.5 8C below modern during the Bolling/Allerod, and 7.5‐8.7 8C below modern during the early Younger Dryas. As the Younger Dryas terminated, temperatures warmed ;4 8C between ca. 11.8 ka and 11.5 ka. These extreme fluctuations in winter minimum temperature have not been generally accepted for terrestrial paleoecological records from the arid southwestern United States, likely because of large statistical uncertainties of older radiocarbon results and reliance on proxies for summer temperatures, which were less affected.

Use of packrat middens to determine rates of cliff retreat in the eastern Grand Canyon, Arizona

Packrat midden data can be used to calculate rates of cliff retreat by relating midden age to the distance between cliff face and midden. Regression analysis using 14 radiocarbon-dated packrat deposits from the Mississippian Redwall Limestone in the eastern Grand Canyon suggests that the Redwall has been retreating at an average rate of 0.45 m/103 14C yr. This rate of cliff retreat, which is comparable to other cliff-retreat rates reported from arid environments, implies that the Colorado River cut through the Redwall Limestone in the vicinity of Horseshoe Mesa about 3.7 m.y. B.P.

In defense of inertia

Markgrafs reassessment of vegetational inertia (Markgraf 1986) touches upon several points of importance, although much of the original article (Cole 1985) seems to have been misinterpreted. First of all, 10,000 yr B.P. was never chosen as the time of southwestern paleoclimatic change. Through a mathematical analysis of the data, I concluded that the major periods of paleovegetational change were gradational (stepwise?) and had occurred between 12,000 and 8000 yr B.P. or 10,000 and 6000 yr B.P., depending upon whether change was considered as species flux (zonal invasions plus zonal extinctions) or was measured as change relative to the modern communities. These conclusions are not inconsistent with Markgrafs reassessment. It is unfortunate that Markgraf interpreted my article as implying that plant species diversity (richness?) was totally a function of climatic regime. Quite to the contrary, the article interprets the low species-richness values between 12,000 and 9000 yr B.P. as the product of vegetational inertia, a lag time between a major climatic change and the eventual biotic equilibrium. My interpretation of changes in species diversity attributable to historic factors is not without precedent. Both temperate forests (MacArthur 1972, p. 174) and deserts (Van Devender 1986, p. 297) vary in species diversity because of past history. Although the species-richness curves were similar for total species and perennial species, only the results for perennial species were published. Ephemeral species were deleted in the published analysis because their fossil seeds are difficult to identify to species. Furthermore, if we can be certain that each packrat midden contains an entire years flora (which we cannot), then some ephemerals might still have been absent (or rare) in that particular year. Thus, I find it far more reliable to stick with perennials for this type of analysis. A Late Wisconsin temperature rise followed later by a summer precipitation increase is one of many possible climatic mechanisms that could explain the changes seen in the fossil record. In fact, this particular scenario seems likely enough that it was considered in a separate section of my article (p. 300). However, this model and the inertial model are not mutually exclusive, and no study has yet adequately discriminated between their effects. I agree with Markgraf that fossil vegetation is by far the best terrestrial source of paleoclimatic information. But this does not mean that it can be interpreted with a total disregard for ecological mechanisms. In times of rapid change such as the Pleistocene-Holocene transition, paleoclimatic reconstruction is complicated by many ecological factors. The concept of vegetational inertia invoked in my article to explain the fossil record from the Grand Canyon is not new or unique. It