Halimeda (Chlorophyta, Bryopsidales) (Lamouroux 1812) is a genus of calcareous green algae that has become abundant on coral reefs of the tropics and subtropics. Though they serve biologically and geologically important roles in these ecosystems, their presence may be an indicator of declining environmental conditions. This study focused on a population of Halimeda in the shallow coastal waters of Tavernier, Florida. Three species were encountered: Halimeda discoidea, Halimeda opuntia, and Halimeda incrassata; Halimeda discoidea was the most abundant. The study was conducted in two parts. Part one involved the observation and quantification of the life history characteristics (e.g. fertility rate, growth rate, mortality rate) of the collected specimens in an artificial environment. Population numbers increased throughout the five week study as new individuals were added through vegetative reproduction. Sexual reproduction was not observed. In Part two the life history characteristics of H. discoidea were used in the construction of a size-classified transition matrix model. Sensitivity and elasticity analyses suggested that the population is most sensitive to changes in the survival rate of reproductive juveniles, and least sensitive to the survival of post-reproductive adults. A standard Chi-square test was used to examine the accuracy of the population distribution predicted by the model against the observed distribution during each week of the study. The test results confirmed the validity of the model; however further analysis of the matrix revealed that without the input of sexual recruits the population will go extinct. A random sexual component was then added to the matrix. Environmental stochasticity can have a direct affect on sexual recruitment in Halimeda. Recruitment interval and recruitment success influence the ability of the population to maintain itself. These factors were manipulated in the matrix. With a 25% recruitment interval of 50 recruits the simulated population was able to avoid extinction over a 25 year period. These results implicate that Halimeda populations require both sexual and asexual recruitment. Additionally, the patchy distribution of Halimeda on and near coral reefs of south Florida indicates that the population may actually be a metapopulation. Recommendations for further studies include an investigation of additional sub-populations within the metapopulation. A mortality analysis would also be useful for gaining a better understanding of the population dynamics and sediment contributions of Halimeda in south Florida.

The conservation biology field seeks to preserve biodiversity and the processes shaping that variation. Conservation biology is intimately tied to evolutionary research, in order to identify evolutionary distinct lineages that may be in danger of disappearing. Interestingly, patterns and processes of lineage divergence and persistence change with respect to spatial and temporal scale. I seek to evaluate biodiversity, the factors that have shaped this heterogeneity, and how this variability persists. To accomplish this I used a phylogeographic approach as well as niche and population modeling on the Peromyscus maniculatus species group found widely distributed in North America. My emphasis was on the southeastern U.S. species P. polionotus and its distinct beach forms. At a continental scale, I found that environmental niches are likely involved in generating and/or maintaining genetic lineages within the P. maniculatus species group. These findings add to a growing number of studies that have identified lineages occupying different environmental spaces. At a regional scale, I supported the hypothesis that barrier islands on the Atlantic coast of Florida were colonized by an ancestral form of P. polionotus by a single colonization, from the central Florida area. Subsequently, at least two distinct lineages diverged (P. p. phasma and P. p. niveiventris). I also found evidence that suggests that the extinct form of beach mouse (P. p. decoloratus) is part of the P. p. phasma lineage. At the population level, I evaluated changes in genetic diversity in historical samples compared to those that experienced recent human encroachment on natural habitat I used tissue preserved in natural history collections to compare with live-trapped specimens, and found that P. p. niveiventris has maintained historical genetic diversity levels. I suggest that the continuation of historical levels of genetic diversity is due to the presence of a single large area iv of continuous habitat in the central portion of the species’ current distribution. Finally, I evaluated the importance of scrub and beach habitat to the population dynamics of beach mice. Beach mice have traditionally have been associated with beach dunes rather than with the scrub habitat found more inland on barrier islands. Using almost three years of capture-recapture data from Cape Canaveral Air Force Station (CCAFS), I created a stochastic matrix model to assess the relative contribution of populations from the two different habitats to a variety of demographic measures. Both field data and model results provided evidence that the population dynamics of beach mice may rely much more on scrub habitat than formerly documented. Overall, my research emphasized a hierarchical approach to evaluate biodiversity and the processes shaping differentiation at different spatial and temporal scales. The methods and findings give insight into speciation at different scales, and can be applied to a wide range of taxa for questions related to evolutionary and conservation biology.

Au cours des dernières décennies, la popularité des plantes médicinales s’est accrue auprès des civilisations occidentales de sorte que la quantité de plantes récoltées, la plupart provenant de populations sauvages, a grandement augmenté. Dans ce contexte, l’objectif principal de mon mémoire est d’évaluer l’impact de la récolte de deux plantes médicinales (Sarracenia purpurea et Rhododendron groenlandicum) utilisées par la Nation Crie du Nord du Québec. Pour y parvenir, des parcelles expérimentales, simulant différentes intensités de récolte (S. purpurea) et différentes méthodes de récolte (R. groenlandicum), ont été mises en place, puis des suivis annuels de la reprise ont été réalisés. Les résultats obtenus suggèrent que les techniques de récolte chez R. groenlandicum devraient exclure les nouvelles pousses, leur exploitation causant une forte mortalité. Par ailleurs, chez S. purpurea, la récolte de 20 % des individus semble peu dommageable, mais critique lorsque plus de 50 % des plants sont récoltés. Un modèle démographique pour S. purpurea a aussi été construit à partir des observations de terrain. Ce modèle a permis de réaliser des projections temporelles en variant les taux de récoltes ainsi que les intervalles entre les récoltes. Les résultats indiquent qu’une récolte de 20 % des individus est acceptable une fois tous les 20 ans. Pour une récolte plus régulière, 5 % tous les trois ans serait soutenable. Mon projet permettra d’assurer une exploitation soutenable de deux plantes médicinales ayant un grand potentiel pour le traitement du diabète de type II. / Over the past few decades, the popularity and use of medicinal plants have increased in Western civilisation. Consequently, harvesting plants from wild populations has increased. Using this perspective, the principal goal of my thesis is to assess the impact of harvest on two medicinal plants (Sarracenia purpurea and Rhododendron groenlandicum) used by the Cree Nation of Northern Québec. To achieve my goal, the simulation of four harvesting intensities (S. purpurea) and of two harvesting methods (R. groenlandicum) were setup in experimental plots and an annual follow up were conducted on establishment and growth. Results show that harvesting methods used on R. groenlandicum should exclude new shoots, since their exploitation results in high mortality rates. For S. purpurea, harvesting of individuals does not seem harmful under 20% of harvest, but is critical when 50% or more are collected. Furthermore, a matrix population model of S. purpurea demography was built from field observations. This model was used to simulate population projection in time under different levels of harvest when varying the number of years between harvests. These results demonstrate that collecting 20% of individuals should be done only once every 20 years to remain sustainable. A more efficient and regular harvest of 5% every three years would be sustainable. My project allowed defining acceptable levels of harvest in order to insure sustainable exploitation of two Cree medicinal plants bearing interesting potential for treating type II diabetes related symptoms.

Au cours des dernières décennies, la popularité des plantes médicinales s’est accrue auprès des civilisations occidentales de sorte que la quantité de plantes récoltées, la plupart provenant de populations sauvages, a grandement augmenté. Dans ce contexte, l’objectif principal de mon mémoire est d’évaluer l’impact de la récolte de deux plantes médicinales (Sarracenia purpurea et Rhododendron groenlandicum) utilisées par la Nation Crie du Nord du Québec. Pour y parvenir, des parcelles expérimentales, simulant différentes intensités de récolte (S. purpurea) et différentes méthodes de récolte (R. groenlandicum), ont été mises en place, puis des suivis annuels de la reprise ont été réalisés. Les résultats obtenus suggèrent que les techniques de récolte chez R. groenlandicum devraient exclure les nouvelles pousses, leur exploitation causant une forte mortalité. Par ailleurs, chez S. purpurea, la récolte de 20 % des individus semble peu dommageable, mais critique lorsque plus de 50 % des plants sont récoltés. Un modèle démographique pour S. purpurea a aussi été construit à partir des observations de terrain. Ce modèle a permis de réaliser des projections temporelles en variant les taux de récoltes ainsi que les intervalles entre les récoltes. Les résultats indiquent qu’une récolte de 20 % des individus est acceptable une fois tous les 20 ans. Pour une récolte plus régulière, 5 % tous les trois ans serait soutenable. Mon projet permettra d’assurer une exploitation soutenable de deux plantes médicinales ayant un grand potentiel pour le traitement du diabète de type II. / Over the past few decades, the popularity and use of medicinal plants have increased in Western civilisation. Consequently, harvesting plants from wild populations has increased. Using this perspective, the principal goal of my thesis is to assess the impact of harvest on two medicinal plants (Sarracenia purpurea and Rhododendron groenlandicum) used by the Cree Nation of Northern Québec. To achieve my goal, the simulation of four harvesting intensities (S. purpurea) and of two harvesting methods (R. groenlandicum) were setup in experimental plots and an annual follow up were conducted on establishment and growth. Results show that harvesting methods used on R. groenlandicum should exclude new shoots, since their exploitation results in high mortality rates. For S. purpurea, harvesting of individuals does not seem harmful under 20% of harvest, but is critical when 50% or more are collected. Furthermore, a matrix population model of S. purpurea demography was built from field observations. This model was used to simulate population projection in time under different levels of harvest when varying the number of years between harvests. These results demonstrate that collecting 20% of individuals should be done only once every 20 years to remain sustainable. A more efficient and regular harvest of 5% every three years would be sustainable. My project allowed defining acceptable levels of harvest in order to insure sustainable exploitation of two Cree medicinal plants bearing interesting potential for treating type II diabetes related symptoms.