<p> South Florida's loggerhead (<i>Caretta caretta</i>), green (<i>Chelonia mydas</i>) and leatherback (<i>Dermochelys coriacea </i>) sea turtles hatchling have environmentally determined sex. The <i> in situ</i> nest mean hatchling sex ratios (SR) were highly femalebiased: loggerhead <i>F</i>=0.89) and green turtle <i>F</i>=0.81; leatherback's SR was nearly balanced (0.55<i>F</i>). Nest temperatures and SRs differed between leatherbacks and loggerhead and green turtles. The latter two did not differ. The loggerhead response parameters were estimated within biological limitations by both 50-65% of incubation and mean middle 1/3 temperature. The maximum middle 1/3 temperature was the best-fit predictor for green turtles. No best-fit sex ratio-temperature response could be identified for leatherbacks. Clutches incubating under natural conditions can vary greatly in SR; TRT differences may account for differences among species' sex ratios.</p>

I used sapling population data collected since 1980 to investigate how canopy disturbance interacts with chronic understory disturbance to determine regeneration patterns. I reconstructed flooding history using a combination of river gauging, a local water-level recorder, and contour maps. 1979 was an extreme flood year for this site, as was 1989. Reduction in flooding frequency since dam construction in 1965 was significant for all elevations, while reduction in flooding duration was significant only for the upper half of the site. Using constrained ordination, I showed that sapling occurrence varies primarily along a flooding/soil moisture gradient, and secondarily along a canopy-openness gradient. This confirms that both flooding and light influence local variation in species composition. Small sapling density increased by more than five times during the decade, while large sapling density increased over 70%. I suggest that this increase is related to the decline in frequency and duration of flooding, and specifically, to the pattern of flooding from 1979-1989, characterized by severe flooding in 1979 and 1989. Sapling survivorship also responded to temporal variation in flooding over the decade. A damage survey revealed a size component to flooding damage, with small individuals making up a disproportionate share of those in the highest damage classes. In addition, there is a relationship between river flow and the elevation of saplings that died, suggesting that the effects of elevation are not simply related to flooding, but perhaps to soil moisture as well. Analyses of sapling growth suggest that there is a trade-off between tolerating flooding in wet years and being subject to moisture stress in dry years. I presented evidence that canopy gaps play an important role in determining sapling growth. Gaps appear to be crucial to the continued success of large saplings, and the fastest-growing individuals are found in them. This may also confer a measure of flood-tolerance related to sapling size.

This study shows that fire had a strong short-term effect on the small-sized individuals in the understory of two dry upland vegetation types, sandhill and upland pine. Large trees were less affected than saplings and small trees. The changes in these types were mainly structural rather than compositional. The influence of fire diminished toward the wet end of the vegetation gradient. Post-fire recruitment via resprouting and germination from seeds was rapid. The interaction of fire and vegetation type was apparent in the comparison of the two dry upland types with slope types. Simulation of fuel and fire provided additional, although not very strong, support for the existence of differences in fire behavior corresponding to an environmental gradient from dry upland forests to mesic and wet slope forests. Ordination using pre- and post-fire data suggests that fire had limited effects on current vegetation patterns and there was no indication that any present pre-fire type would be converted to another type. Furthermore, there were no clear convergent or divergent patterns in overall vegetation change; hence it was difficult to predict long-term changes. An eight-year record of vegetation change at a tornado damaged savanna site in Hickory Creek Unit demonstrates that the tornado had resulted in the succession of the vegetation toward a midslope forest, but two prescribed fires partially reversed such a trend. Phytolith analysis revealed that upland pine and some upperslope forests had abundant grass phytoliths. This indicates that these types had a grass component in the past under an open canopy which might have disappeared as a result of fire suppression, logging, and encroachment of shrubs and hardwoods.

A bottomland hardwood forest stand located in east Texas experienced a hurricane-related flood event in 1989. After this event, Carpinus caroliniana Walt. had a large pulse in mortality, decreasing dramatically in abundance (47%) over a 5-year period. The decline was greater at low elevations suggesting that at least part of the mortality was related to flooding. The flood was the longest summer flood on record. The duration of this growing season flood may have been partly a consequence of upstream dam management. Linear regression showed that in areas where Carpinus death was greater, several sapling species had increases in basal area growth. The invasive species Sapium sebiferum (L.) Roxb. had the greatest increase. This suggests that the Carpinus death and concomitant increase in light availability resulted in accelerated invasion of this forest by Sapium.

A mixed pine-oak forest in the Hickory Creek unit of the Big Thicket National Preserve was hit by a tornado in December of 1983. Subsequently the forest was subjected to three prescribed fires in an effort to restore this area to its former savanna state. The tornado has allowed for both pines and oaks to regenerate. Both groups have regained much of the basal area lost in the tornado. However, the biggest compositional change has occurred at the understory level. There has been a huge increase in the upland shrub Ilex vomitoria (yaupon). Vegetation composition and density seem to be involved in a positive feed-back loop with fire. Plots dense with shrubs experienced low flame heights, which has allowed for increased shrub growth. This process seems to be leading the tornado plots down a spiral towards unflammability. The current fire regime of periodic winter fires has only been able to slow this process in the most open plots. A more intense use of fire, such as repeated summer burns appears to be necessary to reclaim this area as a fire-dependent savanna.

Despite a 20 year reintroduction effort into northeastern North Carolina, little is known about the natural history and ecology of red wolves (Canis rufus). In 2005, I studied home ranges, habitat use and pup attendance of red wolves during the pup rearing season. Data indicated that red wolves have home range sizes intermediate between those of coyotes (Canis latrans) and gray wolves (Canis lupus). Similar to other canids in the eastern United States, red wolves preferred to use extensive agricultural fields during the summer months rather than adjacent wooded areas. Red wolves abandoned dens early to move pups into adjacent agricultural fields. Consistent with pup rearing studies done on gray wolves, red wolf pups were rarely left alone indicating that red wolves share duties of pup rearing and that males play a significant role in the rearing of red wolf pups.

For insectivorous bats living in temperate areas such as Nova Scotia, underground sites are a critical resource for over-wintering, as well as for swarming in the late summer and early fall, which is when mating occurs. The objectives of this study were to 1) identify additional abandoned mines and caves in Nova Scotia that are important swarming sites, and 2) quantitatively characterize factors which best differentiate between caves and mines that are used for swarming/hibernating, and those that are not. Acoustic and/or trapping surveys of 17 abandoned mines and nine caves in Nova Scotia were conducted in 2010. Five site characteristics were analysed to explain differences between used and unused sites. Surveys indicated that twelve of the 26 sites are used by bats during the swarming period. Results of a logistic regression analysis of nine a priori selected models indicated that chamber length was the best predictor of swarming.

The reproductive success of American kestrels (Falco sparverius) infected with Trichinella pseudospiralis was significantly reduced when compared with uninfected controls. Egg laying was delayed, and infected females produced fewer eggs than uninfected controls. Infected females were more aggressive and copulated less frequently. Egg breakage among infected birds was greater than among controls. Infected females tended to lay eggs outside the nest box. Infected females did not incubate as frequently as the controls, thus accounting for the higher prevalence of embryo mortality among these birds. The above effects translated into a mean production of 2.1 hatchlings per control pair and 0.6 per infected pair of kestrel. Both in breeding and non-breeding kestrel, T. pseudospiralis infections gave rise to a general decline in mobility. Behavioural effects are primarily attributable to the presence of muscle larvae. The parasite showed no predilection for specific muscle groups.

No description available.

This dissertation presents a new perspective on the history of conservation in North America. In contrast to historiography that locates conservation-oriented approaches to the North American environment as the product of late-nineteenth-century concerns within Canada and the United States, this study links wildlife conservation in Quebec directly to longstanding European land tenure, estate management, and associational strategies. Through a range of materials including state documents, associational records and personal and family papers, I show how advocates of fish and game protection in the province drew heavily on Old World customs and traditions, particularly those of British landowners, who displayed in their varied social, economic, and political commitments an ongoing engagement with improvement. These 'patrician sensibilities,' I argue, formed the basis of the regulatory system that developed in Quebec during the period 1850-1914, first on the remote salmon rivers of the north shore and Gaspe peninsula, and by the First World War on the bulk of the province's best and most easily accessible hunting and fishing territories. In addition to the regulatory strategies that developed during this period, the dissertation deals with forms and limits of resistance on the part of aboriginal and non-aboriginal subsistence, commercial, and sport hunters and fishers. The dissertation's major contribution lies in its demonstration of the longstanding patterns that underpinned the development of conservation strategies in North America. Class and gender are central to the project, and it also has important implications for our understanding of civil society and state formation.