Contiguous macroscopic charcoal analyses were performed on a 9.03 m long lake sediment core from Roe Lake on Pender Island in the Gulf Islands National Park Reserve of British Columbia, Canada to reconstruct the island’s fire history over the last 10,000 years. Charcoal particles >150μm were counted to quantify charcoal concentrations, charcoal accumulation rates and mean fire return intervals. Results show that the early Holocene was characterized by high charcoal accumulation rates and frequent low-severity fire with a mean fire return interval of 100 ± 29 years. Forests at the time were dominated by Pseudotsuga menziesii with an open canopy and fern taxa, particularly Pteridium aquilinum, being common in the understorey. This open vegetation, coupled with warm and dry summer climate, likely created conditions conducive to this fire regime. Charcoal accumulation rates decreased in the middle to late Holocene, and fire frequency decreased, resulting in a mean fire return interval of 167 ± 43 years. Climate cooled and moistened along with a decrease in seasonality during this time and the canopy closed, establishing closed-canopy Pseudotsuga menziesii forests. Climate appears to be the primary factor controlling fire regimes near Roe Lake for most of the Holocene. At times, shifts in the fire regime cannot be explained by changes in climate. Fire frequency increased between 7000-5000 cal yr BP, coincident with a peak in Quercus garryana pollen, despite cooling and moistening climate. Fire likely maintained patches of Q. garryana savanna during this time. Fire again became more common contrary to trends in climate after ~2500 cal yr BP. This late Holocene increase in fire is also seen elsewhere in the Pacific Northwest and may be a reflection of increased climate variability due to more frequent El Niño events or an increase in human-lit fires. Indigenous populations on southern Vancouver Island commonly used fire as a resource management tool and it is likely that people on Pender Island did as well. As fire management practices shift from fire suppression to more sustainable practices, this study offers the Gulf Islands National Park Reserve important baseline information on the area’s natural fire regime to help guide future conservation efforts. / Graduate / 2023-04-07

The early Pliocene age Gray Fossil Site (GFS) is a fossil site that preserves diverse fauna and flora and represents one of only a few sites of its age in eastern North America. The assemblage of fauna found at GFS is unique compared to what has been found at other contemporaneous fossil sites across the continent. While studies of the fauna and flora have helped us to understand the ecosystem at GFS, detailed study of the diverse mammal community will help us better understand this unique site. Studying mammalian community structures at fossil sites can provide an abundance of information about the past such as, ecological inferences and climate conditions. To understand the paleoenvironment at GFS, I have gathered data from twenty modern communities and five well-known contemporaneous sites. Mammal species from each of these 26 sites have been categorized by body size, locomotor mode, and cheek tooth crown height to characterize the niches occupied by each species. Proportions of species within categories have been compared and analyzed across modern and fossil sites using discriminant function analysis (DFA). Results of the DFA demonstrate that modern communities with different climates and habitats can be distinguished based on the body size, locomotion, and cheek tooth crown heights of mammals living there. Moreover, functions can be used to infer conditions of past sites. Species diversity within categories was also compared across all sites. Results show that GFS has more brachydont and scansorial species present, and fewer cursorial species than occur at other contemporaneous sites. The results of the discriminant functional analysis indicate that the GFS community structure is most similar to modern temperate forest regions of the Himalayan Mountains in China and modern Washington Co., TN. Proportions of brachydont, gliding, and scansorial species appear to be driving variables behind these results.

A dissertation presented to the Faculty of Science, University of the Witwatersrand, in partial fulfillment of the requirements for the degree of Master of Science. August, 2017 / The Karoo Supergroup represents a highly complete sedimentary succession that was deposited in several basins throughout southern Africa during the late Palaeozoic and Mesozoic (Carboniferous-Jurassic). While research in the Lebombo-Tshipise Basin of southern Africa has largely focused on lithological description of Karoo sediments or structural features of the basin, little effort has been made to describe the palaeoenvironments recorded in the sediments, or the basin fill’s response to major tectonic or climatic events. To address palaeoenvironmental reconstruction, lithostratigraphic analysis resulted in defining 11 facies associations from the Tshidize, Madzaringwe, Mikembeni, Fripp, Solitude, Elliot and Clarens Formations in the Pafuri sub-basin of the Lebombo-Tshipise Basin. Twenty boreholes drilled in Kruger National Park by the Council for Geoscience in 1979 record largely stable and consistent deposition of the Tshidize, Madzaringwe and Mikembeni Formations during the Permian. The Mikembeni Formation thins dramatically southwards, whereas the overlying Triassic successions tend to wedge out to the east. Late Triassic and Jurassic sediments directly overlie Precambrian basement in the southernmost boreholes. Significant thickness differences between the western and eastern boreholes indicate a large fault, which likely represents a rift shoulder. The palaeoenvironments in this basin are similar to those of the Main Karoo Basin, but quantitative analyses suggest a more humid environment in the Late Triassic Elliot Formation. Sauropodomorph fossils validate assignment of formerly mapped Solitude Formation as actually being the Elliot Formation. Palaeosols in the Elliot are consistent with either Oxisols or Argillisols. Wet desert conditions, evidenced by burrows produced by invertebrate communities, and tectonic activity, suggested by seismites, persist into the Early Jurassic Clarens Formation. / MT2018

Pre-Pleistocene freshwater fish assemblages in the eastern United States are exceedingly rare, limited to only a handful of localities. The Gray Fossil Site, an early Pliocene sinkhole fill in northeast Tennessee, has yielded an abundance of vertebrate remains, including fish. Comparison with extant and fossil fish taxa reveals a depauperate ichthyofauna consisting of only two centrarchid genera: Micropterus cf. M. salmoides (black bass) and Lepomis sp. (sunfish). This material includes hundreds of specimens and represents the oldest centrarchid material from the eastern United States; it has potential to inform our understanding of centrarchid evolution and diversification. Fishes of the genus Micropterus are opportunistic predators that feed increasingly on fish and even terrestrial vertebrates as they increase in size. Lepomis, smaller Micropterus, and terrestrial vertebrates such as frogs and salamanders likely formed a major food source for Gray Fossil Site Micropterus. Extant fishes in the genus Lepomis are generally predators of aquatic insect larvae, small crustaceans, zooplankton, and other aquatic invertebrates and prey occasionally on small fish. Lepomis from the Gray Fossil Site were likely non-specialized predators of aquatic invertebrates – they are “short-jawed,” indicating low levels of piscivory and lack the robust pharyngeal jaws associated with extensive feeding on mollusks. Extant species of Micropterus and Lepomis occupy a wide variety of freshwater habitats but typically only spawn in shallow, open water with soft substrate overlying gravels. This suggests these conditions were present in the sinkhole pond. The absence of other fish species that are ubiquitous in streams and rivers in the southeastern United States today suggests that the sinkhole pond was not connected to surrounding waterways by surface hydrology. In addition to large numbers of isolated bones, several articulated and partially articulated fish specimens have also been recovered and show minimal evidence of postmortem disarticulation or scavenging. This suggests a cold and/or anoxic hypolimnion within the sinkhole pond as well as an absence of scavengers such as crayfish, which have not been documented from the site. Osteological thin sectioning of fish atlantes suggests slow growth rates in Gray Fossil Site fishes and small overall size for their age, possibly due to high competition or limited surface area of the pond they inhabited.

No description available.

No description available.

The fossiliferous Yorktown Formation (Pliocene) of Virginia was used as a natural laboratory for testing predictions of ecological and evolutionary theories. Specifically, coordinated stasis and ecological locking models have testable elements that can be analyzed using data from the Yorktown Formation. The ecological locking model requires that species within an ecosystem have strong interactions in order to stabilize morphologies of multiple lineages over millions of years. Species intereactions that are strong enough to do this should also be strong enough to be a major ordering force on the composition of paleocommunities. Single and replicate samples were taken from 30 cm stratigraphic intervals within the Rushmere and Morgart's Beach Members at several localities. A total of 142 samples were collected from 5 localites, which yielded 29,000 specimens belonging to 140 species of bivalves, gastropods, and other taxonomic groups. Principle components analysis, ANOVA, MANOVA, and other analyses were used to test the occurence and recurrence of local paleocommunities, paleocommunities, and paleocommunity types. Three paleocommunity types which occured under specific paleoenvironmental conditions were defined: rubbly bottom, transitional, and muddy bottom. Within a single locality samples from the same paleocommunity type yielded very similar faunal compositions, based on the relative abundance of the contained species. However, samples from the same paleocommunity type but different localities displayed low similarity values. This is consistant with local paleoenvironmental control of paleocommunity composition being more important than strong species interactions. The pattern predicted by the model of ecologic locking is absent from these Yorktown paleocommunities. A guild analysis was performed on the data to test whether the same types of organisms recurred in a predictable fashion under similar paleoenvironmental conditions. While the guild structure of the rubbly bottom paleocommunity type did recur at several localities, the guild structure of the other paleocommunity types varied greatly from place to place. / Ph. D.

Identification of pollen in fecal pellets is a potential technique for describing an animals diet and in turn the vegetation of an area. Mule deer and Bighorn Sheep pellets representing both summer and winter browsing and a variety of habitats were examined using relative percentages and the absolute pollen frequencies. In addition, fossil pellets from Stanton's Cave, Grand Canyon, Arizona, were examined and compared with modern pellets. Absolute pollen frequencies of individual pellets showed higher values and greater variation for summer pellets than for winter pellets. Relative pollen percentages for a specific vegetation type showed more variation in fecal pellets than in soil surface samples. Unless specifically being eaten, arboreal pollen types such as Pinus are less abundant in fecal pellets than in soil samples. Unlike soil surface samples, arboreal pollen types in fecal pellets are frequently limited to the immediate source area.

While the archaeology of the Northern Frontier of Mesoamerica is poorly understood, Pedro Armillas' hypothesis that climatically induced environmental change was the limiting factor for cultural change has become the ruling theory. In order to test this hypothesis original lacustrine pollen profiles were compared with a detailed inspection of the known archaeological record and the previously published paleoecological record. The archaeological evidence suggests that there was a dense human occupation in the northern reaches of Mesoamerica between about AD 600-900. The first indications of human settlements are related to the Chupicuaro culture that reached its apogee about 2000 years ago located along the Rio Lerma. A rustic variant of the Chupicuaro culture spread north and is associated with scattered hamlets. About AD 600 the sedentary population expanded considerably in conjunction with the development of regional centers and the Coyotlatelco red-on-buff ceramic tradition. This expansion can be seen all along the Northern Frontier of Mesoamerica from Alta Vista south to Tula. Between AD 900-1000 there was a dramatic change in settlement patterns and by about AD 1000 most of the northern reaches of Mesoamerica were once again under the control of semi-nomadic hunters and gatherers. A suite of four cores was collected in a transect that crosses the Northern Frontier of Mesoamerica. The goal of selecting sites that minimized human impact was not altogether successful since these cores identify the impact of agriculture. Within the limits of the dating and material available, the pollen profiles from these cores suggest an environmental change between AD 1000 and 1500. From the data available it is not clear if this change, or changes, was the result of changes in human settlement patterns or climatic change. As such Armillas' hypothesis remains unproven.

Fire-scarred cross-sections from coast redwood (Sequoia sempervirens) at two sites were dendrochronologically dated and used to develop a fire history. Redwood offers a challenge for dendrochronological study due to ring wedging and complacency. Crossdating was successful in 12 of 24 trees. The fire history was developed by comparison of fire scars and fire-associated ring characteristics (resin ducts, double latewood, growth releases, and ring separations) recorded in ring series. Using only dates of fire scars from the first fire in 1714 to the last in 1962, the mean fire interval (MFI) was 9.9 years. MFI for the best represented presettlement segment 1714-1881 was 8.0 years. Using all fire-associated ring features, MFI 1714-1962 was 7.0 years and 1714-1881, 6.0 years. Use of all fire-associated ring characteristics is argued to be a more accurate representation of past fire frequency. MFIs determined are less than others reported for coast redwood and suggest fire frequency in redwood may have been underestimated in past studies.