Plants often modify microsite conditions important for seedling establishment. Palms may reduce light levels and produce deep leaf litter; both may suppress seedlings, affecting the local abundance, distribution, and species composition of tree seedlings. The abundance and composition of palms were examined along a tropical forest chronosequence at the Barro Colorado Nature Monument, Panama. Palms were sampled along transects in secondary (20, 40, 70, and 100 y in fallow) and old-growth forests. Palm abundance and basal area significantly increased with forest age. Palm distribution was driven by the most abundant palm, Oenocarpus mapora, which accounted for 60% of the total stems. I examined the effects of O. mapora, on seedling distributions on Barro Colorado Island, Panama. I asked: a) what are the effects of O. mapora on light availability and leaf litter accumulations; b) how does O. mapora affect abundances of seedlings; c) does seed size of established seedlings differ between palm and non-palm sites? Seedling composition, light availability, leaf litter, and palm presence were measured in 1-m² contiguous plots along transects in one late second growth and two old growth forest stands. I conducted a seedling transplant experiment to determine the effect of O. mapora and its leaf litter on growth and mortality rates of Gustavia superba seedlings. Leaf litter depth was higher and light availability was lower in plots where palms were present. There was a lower probability of seedling presence where large palms were present. The probability of seedling presence increased with light availability. Seed size was greater for seedlings in palm plots relative to non-palm plots. Seedling survival was significantly lower for palm and litter treatments relative to controls. I conclude that O. mapora creates poor establishment conditions for seedlings which may affect the local abundance, distribution, and species composition of the seedling pool.

An increasingly urgent task in the field of conservation biology is to identify changes in abiotic and biotic interactions that result when large areas of forest are converted to small fragments surrounded by anthropogenic landscapes. My research, conducted in lowland dipterocarp rain forest at the 2,500-ha Pasoh Forest Reserve in Peninsular Malaysia, investigated a novel but strongly negative edge effect - namely, a tremendous increase in the density of wild pigs (Sus scrofa) and the resulting deleterious impacts on the understory plant community. The absence of feline predators due to the small size of the reserve and the presence of a year-round food supply in the agricultural areas surrounding the reserve are believed to be the main factors contributing to increased pig density. Line transects were conducted to determine pig density within Pasoh. Density in 1996 and 1998 was estimated to be 47 and 27 pigs/km², respectively, or 10 100 times historical levels. Fences were constructed to exclude pigs from control plots to quantify the impact of soil rooting and seed predation on plants in the understory. After two years, plots inside exclosures had three times more recruits, greater species richness, and 53% more height growth among plants 1 7 m tall than did adjacent plots to which pigs had access. Surveys were conducted in 1995, 1996, and 1998 to determine the number of reproductive nests constructed by pigs. Pigs constructed an estimated 6.0 nests/ha/year during this time, with an average nest composed of 145 snapped saplings and 117 uprooted saplings. Nest building accounted for 28.9% of all mortality for trees 1 2 cm diameter at breast height. Uprooted stems died, but snapped stems produced a leafless stump that could resprout. Observations of >1,800 stumps for 36 months revealed large differences in resprouting among species, families, and groups of plants with similar life history characteristics. Overall, the results of the different studies suggest that if elevated pig densities continue there could be a shift away from the currently dominant Dipterocarpaceae and Euphorbiaceae.

Phylogenetic relationships and classification of Chrysothamnus, Ericameria, Xylothamia and related genera were investigated. The internal transcribed spacer and 3' external transcribed spacers (ITS and ETS) of the nuclear ribosomal (nr) DNA were analyzed separately and combined employing different optimality criteria. These analyses indicated that the previous classifications and hypotheses of relationships were not monophyletic. Chrysothamnus, Ericameria, Xylothamia, and related genera were placed in separate lineages irrespective of data set and optimality criteria. Chrysothamnus species, as traditionally delimited, were resolved in four, not necessarily closely related lineages affiliated with the Solidagininae. Previous sectional classification of Chrysothamnus based primarily on morphology was not supported by the present molecular data. Ericameria was placed in a clade separate from both Chrysothamnus and Xylothamia. Associated with, but basal to, the Ericameria lineage was a clade composed of Pentachaeta, Rigiopappus, and Tracyina. Prior infrageneric classification of Ericameria was in part consistent with the results of this investigation. Species were placed in three, rather than four, lineages within the genus. The three annual genera and Ericameria represent a lineage separate from the Solidagininae and Hinterhuberinae. Species of Xylothamia were not monophyletic but were placed in at least five separate lineages. Four species were aligned with Gundlachia, while the others were strongly supported in a separate clade. Within that clade, however, the other species were usually in distinct, but unresolved lineages. Xylothamia and its relatives were resolved in a clade distinct from other Solidagininae and merits recognition of their distinctiveness. Both Stenotus and Tonestus were polyphyletic. Type species of both genera were associated with other clades, and the relationship of most of the other species remains unclear. These results suggest a reclassification of these taxa into novel, distinct genera. In general, the results of this study were incongruent with relationships inferred from morphology.

Aerenchyma - a specialized tissue with abundant, interconnected gas spaces - facilitates oxygen transport from shoots to roots in many emergent wetland plants. The importance of oxygen transport to root elongation and aerenchyma development in rice (Oryza sativa L.) seedlings was demonstrated in this investigation. Germinated seeds of Lemont (a commercial cultivar) and red rice (an agricultural weed) were positioned at 4.5, 2.5, or 1.0 cm below the surface of aerated or hypoxic water to control the duration of leaf submergence and the availability of oxygen. Primary root and leaf lengths were measured daily. Aerenchyma was compared among emergent and completely submerged seedlings that grew in aerated and hypoxic water. Root elongation rates were slower in hypoxic water than in aerated water. Root elongation rates increased after leaves emerged from 1.0 and 2.5 cm of overlying hypoxic water but not after leaves emerged from 4.5 cm of overlying hypoxic water. The benefits of oxygen transport to root elongation may therefore depend upon submergence depth. Measurements of transverse shoot sections taken between 2 and 3 mm above the seed revealed that 1) red rice coleoptiles grown in hypoxic water had 3 to 12 times more gas space than aerated coleoptiles and 2) the amount of gas space in primary leaves was two to five times greater in emergent seedlings that grew in hypoxic water than in aerated seedlings and completely submerged seedlings that grew in hypoxic water. Aerenchyma gas space formation occurred closer to the root tip and over a shorter distance in roots of emergent seedlings that grew in hypoxic water than in aerated roots. Roots of emergent seedlings that grew in hypoxic water also had a larger maximum amount of gas space than aerated roots. The increase in gas space between tissue ages of 10 and 45 h was similar among aerated roots and roots of emergent seedlings that grew in hypoxic water. This study is the first to report that hypoxia greatly reduced the amount of gas space and the rate of gas space formation in roots when oxygen transport was prevented by complete submergence in hypoxic water.

Exotic pest plants often grow to greater stature, become more abundant, and display increased shade tolerance in their introduced ranges than in their native ranges. These differences have been hypothesized to result from genetic shifts in biomass allocation, growth, or photosynthesis between genotypes in native and introduced ranges or from plastic, phenotypic responses to different environmental conditions, such as lower herbivore or fungal pest loads in areas of introduction. I used the tropical shrub Clidemia hirta (Melastomataceae) as a model exotic pest plant to test these two non-mutually exclusive hypotheses of invasion. Clidemia hirta invades forest understory and is more abundant in much of its introduced range in parts of Oceania, Asia, and Africa than in its native range in Central and South America, where it does not occur in forest understory. Contrary to predictions, I found less genetic variation, as detected with allozymes, within and among native, Costa Rican populations than introduced, Hawaiian populations of C. hirta. Hawaiian and Costa Rican populations also were markedly dissimilar genetically (Nei's I = 0.64), but there were few ecologically important differences in biomass allocation, growth, or photosynthetic parameters between Costa Rican and Hawaiian genotypes grown under high or low light in a common garden experiment. The absence of C. hirta from forest understory in its native range likely results, at least in part, from the strong pressures of insect herbivores and pathogens (natural enemies). A natural enemy exclusion study conducted in the field showed that insect herbivore and fungal pathogen damage was substantially greater on Costa Rican than Hawaiian plants and that these natural enemies caused substantial mortality of C. hirta planted into forest understory in Costa Rica but not Hawaii. These results coupled with demographic data collected over three years in two Hawaiian populations suggest that biological control could cause a decline in C. hirta population growth rates in Hawaiian forests. For now the expanded habitat distribution and vigor of C. hirta in its introduced range seems to result from an ecological response to enemy release rather than a genetic shift in resource acquisition, allocation, or growth.

Unresolved phylogenetic problems in pyrenomycetes and population structure of an asexual plant pathogen (Discula destructiva) were investigated. Analyses based on DNA sequences of nuclear encoded small and large subunit RNA genes (SSU and LSU nrDNA) and deduced amino acid sequences of a subunit of RNA polymerase II gene (RPB2) indicated that five genera previously considered in Ceratostomataceae of Sordariales were related to Hypocreales. Melanospora (including the type) and a genus of Ceratostomataceae formed a basal clade, but monophyly of the five genera and of the genus Melanospora was rejected. Discula destructiva and four other Discula species were derived from within Diaporthales according to the SSU and LSU nrDNA and RPB2 phylogenies. There were three clades in the LSU nrDNA phylogeny. The five Discula species were in one of the clades but they were not monophyletic. This delimitation is congruent with anamorph and pigmentation distributions in the diaporthalean taxa but is not congruent with family concepts based on other phenotypic characters. Taxa of Magnaporthaceae were excluded from the Diaporthales, but its phylogenetic position remains unresolved. Preliminary studies of the evolutionary mechanisms of D. destructiva involved investigation of its population structure. Two distinct groups of D. destructiva isolates, one from the western U. S. and the other from the eastern U. S., were identified with amplified fragment length polymorphism (AFLP) markers and sequences of several genes. Discula destructiva is thought to have been introduced to North America in late 1970s. The remarkably low genetic diversity compared to other asexual fungi indicated that D. destructiva is still under intense selection pressure and that episodic selection may still be in effect. However, the transition to a less virulent, heterogeneous population might have begun in the New York City area, a possible epidemic center in the east, which had relatively higher genetic diversity than the samples from other areas.

Variation in fire intensity may affect the structure and composition of frequently burned plant communities. I hypothesize that small-scale variation in fuel loads affects fire intensity within single fires in frequently burned savannas. Furthermore, I expect that local differences in fire intensity produce predictable patterns to the demography of groundcover shrubs. I tested these hypotheses by manipulating fuel loads in a longleaf pine savanna. Fuel load treatments consisted of fine fuel removal, pine needle addition, wood addition and unaltered control, mimicking naturally occurring fuel load variation. Treatments were repeated in 1m2 subplots over two sets of biennial burns and shrub responses were followed for 5 years. Addition of fuels increased fire intensity (maximum fire temperature, fuel consumption, heat output), while removal of fuels decreased fire intensity relative to controls. For all species pooled, addition of fuels, particularly wood, increased damage to shrubs and decreased resprouting relative to control. Removal of fuels did not affect shrub resprouting. Similarly, fuel addition increased the probability of species loss (genet mortality), while fuel removal had no effect. Fuel addition decreased resprouting of native rhizomatous resprouting species relatively more than that of native root-crown resprouting species or non-native species. However, density of rhizome sprouting shrubs tended to increase over the course of the study. The demography of Rhus copallinum, a common rhizomatous species, was affected by fuel load treatments. Matrix projection models indicated that population growth was positive in the control treatment, nearly stable in the pine needle addition treatment, and negative in the fuel removal and wood addition treatments. In all treatments, stasis of large ramets was the largest contributor to population growth. Effects of fuel loads and fire intensity on population dynamics of R. copallinum may result from an interaction between damage to rhizomes and post-fire environmental conditions. My overall results indicate that small-scale variation in fuel loads predictably affects local fire intensity in pine savannas. By altering fire intensity, fuel loads alter patterns to the local demography of shrubs. Small-scale variation in fire intensity may contribute significantly to the observed patterns of shrub abundance in frequently burned pine savannas.

Arabidopsis thaliana is a model organism used for genetic, molecular, and biochemical analyses of higher plant processes. Leaf hairs (trichomes) on the surface of Arabidopsis leaves are macroscopic single cells that develop by elaborate cell expansion and differentiation of the epidermis. Arabidopsis trichomes provide an excellent model to study plant cell expansion and differentiation and the cell cycle. Upon maturation, an Arabidopsis trichome becomes a branched, stellate structure with a highly birefringent cell wall. Arabidopsis mutants have been isolated that have trichomes that look glassy and have reduced birefringence. CONSTITUTIVE PATHOGEN RESISTANCE5 (CPR5) is a previously isolated Arabidopsis gene observed to have trichomes with reduced branching. Work presented here shows cpr5 trichomes to have reduced birefringenceand leaves to have less paracrystalline cellulose. The leaves, stems, and roots of cpr5 plants have defects in cell expansion. The cells of cpr5 leaves prematurely stop expanding, resulting in leaves that are much smaller than wild-type leaves. Stem and root cells of cpr5 plants have reduced polar expansion, resulting in longitudinally shorter cells. The roots of cpr5 plants grow at a slower rate than wild-type roots. In trichome development, CPR5 acts downstream of genes involved in trichome initiation. Synthetic genetic interactions between CPR5, TRIPTYCHON, and NOECK (NOK) suggest these three genes might interact to control trichome cell expansion. An allele of NOK was isolated in a screen of cpr5-2 phenotypic modifiers. Trichomes on nok plants are extrabranched, glassy, and have reduced birefringence. NOK maps to the top of chromosome three. Additionally, a novel 4-branched trichome mutant was identified in the modifier screen and would be classified as an asymetric branch mutant. This mutation maps to the top of chromosome one. In a separate screen, a novel glassy trichome mutant called DEFLATED TRICHOMES (DFT) was isolated because the trichomes had collapsed cell walls. Trichomes on dfl plants also have reduced birefringence. DFT maps to the top of chromosome five. Further characterization of NOK and DFT is integral to understanding the degree to which the glassy trichome phenotype is indicative of the nature of the protein.

Chlamydomonas reinhardtii is a photosynthetic alga that has the ability to concentrate CO2 around Rubisco to achieve enhanced rates of photosynthesis in a low CO2 environment. This dissertation presents results obtained from the generation and analysis of four high CO2 requiring mutants of C. reinhardtii. The use of reverse genetics is a very powerful tool to dissect out the individual components of metabolic pathways. Two reverse genetics methods were utilized in this study: a random insertional mutagenesis method to discover genes that are required for growth in a low CO2 environment, and a directed mutagenesis approach, RNA interference, to determine the role of two low CO2 inducible genes in the carbon concentrating mechanism. The first high CO2 requiring mutant was determined to be defective at the Rubisco activase locus. The second mutant, cia6, had an insertion in a SET domain containing protein that may be involved in the regulation of the carbon concentrating mechanism. The third mutant, slc23, had an insertion in a gene that encodes for multiple splice variants that encode for at least four distinct WD40 repeat proteins that vary in their number of WD40 repeats. A targeted mutagenesis approach was utilized to silence the expression of the two low CO2 inducible, nearly identical genes, Ccp1 and Ccp2. RNA interference was successfully used to reduce the expression of Ccp1 and Ccp2 mRNAs and proteins to undetectable levels. Results suggest that the Ccp1 and Ccp2 proteins are required for growth in a low CO2 environment, but that these two proteins are not required for efficient photosynthesis at low levels of CO2.

Carbonic anhydrase (CA) is a zinc containing metalloenzyme that catalyzes the reversible interconversion of CO2 and HCO3-. There are three evolutionarily unrelated CA families designated alpha, beta and gamma CA. Vertebrates have members of the alpha CA family, while higher plants, algae and cyanobacteria have members belonging to all three CA families. In the green alga, Chlamydomonas reinhardtii, five CAs have previously been identified including three alpha CAs and two beta CAs. This dissertation describes the identification and characterization of new CA genes from C. reinhardtii. Four new CA or CA like genes have been discovered including two beta CAs and two gamma CAs. Three CAs were investigated further including the alpha CA Cah3, one of the new beta CAs, Cah6; and a new gamma CA designated Gclp1 for gamma CA like protein. Cah3 is an alpha CA located in the thylakoid. Past studies with two Cah3 mutants, ca-1 and cia3 have shown that Cah3 plays an important role in the CO2 concentrating mechanism. In this work, the mature Cah3 protein was overexpressed as a fusion protein in E. coli and found to have significant CA activity. This is the first report of detection of CA activity in the Cah3 protein and its partial biochemical characterization. A novel beta CA (Cah6) and a putative gamma CA (Gclp1) gene were identified in C. reinhardtii. Gclp1 is one of two putative gamma CAs found in C. reinhardtii. Both the Cah6 and Gclp1 open reading frames (ORFs) were cloned in the overexpression vector pMal-c2x and expressed as recombinant fusion proteins. The purified Cah6 had significant in vitro CA activity but Gclp1 did not. Gclp1 was designated as a gamma CA like protein because it lacked detectable CA activity. Cah6 has a leader sequence consistent with a chloroplast localization. Although Cah6 is constitutively expressed under low and high CO2 conditions, it is slightly upregulated under low CO2 conditions. Immunolocalization studies confirmed that the Cah6 is localized to the chloroplast stroma particularly in the starch sheath around the pyrenoid. A possible role of Cah6 in the CO2 concentrating mechanism and photosynthesis is discussed.