Rhododendrons are economically important plants in horticulture, and many species are threatened in the wild by habitat degradation. It is therefore doubly important that their nutritional needs should be understood.

A survey of the current status of Rhododendron ponticum in the Peak District and Sheffield area was carried out. It was found to be widespread over much of the area, particularly on free-draining, nutrient-poor, acidic soils, in sheltered, moist situations. It was largely absent from the Carboniferous and Magnesian Limestone Series, from the high altitude Kinder/Bleaklow massif and from areas subject to intensive agricultural or industrial development. Disturbance of habitats through forestry, grazing or recreational activity seems to encourage invasion of suitable areas. The role of mycorrhizal infection in R. ponticum was investigated. Mycorrhizal plants had lower Root/Shoot Ratios, higher Relative Growth Rates and showed increased yield compared to non-mycorrhizal plants. The benefits of infection were strongest on nutrient-poor soils without added nutrients. Infection of roots was visible after around six weeks and the effects of mycorrhizas were increasingly apparent during the following six weeks. The source of fungal inoculum was investigated and considered. The 'interference' phenomenon described by earlier workers was investigated. A mixture of competitive and allelopathic influences upon test seedlings in bioassays was demonstrated. The toxicity . found, was closely related to the presence of R. ponticum roots in the soil. Both living or dead roots produced the effects. These were not removed by nutrient addition. Interference was not dependent on mycorrhizal infection of the R. ponticum roots. To help an understanding of the interference demonstrated, a survey was undertaken of the 'free' phenolic compounds occurring in R. ponticum tissues, associated soil and litter, and in canopy throughfall. These compounds have been implicated in allelopathic interactions involving other members of the Ericaceae. R. ponticum tissues were found to have very high concentrations of 'free' phenolic compounds compared to other plant species examined. Considerable variation in form and amount was found with tissue type and age. Phenolic compounds were detected in associated soil and litter, as well as in canopy throughfall from R. ponticum. It was shown that the interference cannot be fully explained by competition for water and/or nutrients. In some situations a toxic influence perhaps due to aromatic and aliphatic acids released from the roots, has a major effect on the interaction between R. ponticum and associated vegetation. Competition for nutrients and/or water clearly occurs in some field situations, particularly when R. ponticum bushes are encroaching on established vegetation. However, with bare-zones (either in the field or under artificial conditions) competition factors may be almost totally eliminated by the toxicity which inhibits root formation. Since the seedlings have very restricted root development, they are barely able to compete for nutrients or moisture, and the toxic effects dominate the interaction. The natural situation in the field is complicated by the acidification of soils associated with R. ponticum, the physical and chemical effects of its litter, shading and the overall influence of the plant on soils and nutrient cycling. The large quantity of 'free' phenolic compounds in R. ponticum tissues (especially new leaves and new stems) probably have anti-herbivore and/or anti-pathogen functions. This would explain the observed lack of damage to the plant by invertebrate herbivores, diseases or parasites. These compounds would thereby enhance the growth and competitive ability of R. ponticum.

The investigation of hybrid zones has proven to be one of the most promising approaches to advance our understanding of species barriers, and to elucidate evolutionary processes involved in speciation. Due to the improvement of molecular techniques it will soon be possible to investigate the genetic composition of non-model species in much greater detail, and also include species that defy investigation using controlled laboratory conditions. To be able to draw further reaching conclusions about the generality of certain evolutionary factors, it is crucial to investigate a wide spectrum of organisms differing in traits, life histories and relatedness. This study investigates patterns of hybridisation between two pairs of closely related species in the genus Rhododendron. AFLP data for 346 loci, from twelve populations in total comprising 390 individuals, were obtained. Additionally, the abundance of three alkane components in the leaf waxes of 115 individuals was determined. For the species pair R. clementinae and R. roxieanum low levels of recent hybridisation were found, however, the wax composition of R. roxieanum var. cucullatum suggests historical introgression. Two types of hybrid zones were found for R. aganniphum and R. phaeochrysum, one mainly comprising F1 individuals, and the other frequent backcrosses to R. aganniphum. Furthermore, evidence for genomic incompatibilities at several loci for the two species will be presented, and hybrid identity of R. aganniphum var. avorufum and R. phaeochrysum var. agglutinatum is suggested.

The southern Appalachian forests of the United States are undergoing changes due to the death of hemlock trees attacked by the hemlock wooly adelgid. This paper addresses the management impacts of Rhododendron maximum removal in the riparian and upslope areas previously occupied by hemlock. This study measured the consequences macroinvertebrates faced due to riparian Rhododendron removal from 300 m reaches of two low order streams. Two additional low order streams served as reference sites for the experiment. The stream macroinvertebrate communities were assessed using a before-after controlled impact model comparing communities between fall 2014 to those from fall 2015 and from spring 2015 to spring 2016. Macroinvertebrate collections consisted of 288 samples with a total of 61,056 individuals. There was a significant increase in collector-gathers in both removal sites, mostly from increases in Ephemerellidae and Chironomidae. There was also a significant decrease in filter feeding organisms in the removal reaches. Traits analysis also revealed that several traits that are shared by collector gathers also increased, e.g., short life cycles that are related to the increase in Chironomidae. Using Nonmetric Multidimensional Scaling (NMDS) and permutational MANOVA significant annual differences in macroinvertebrates were found in all of the stream reaches during both seasons. However, the trait based NMDS and permutational MANOVA found significant change only in one removal site between fall collections based on traits. These finding are consistent with findings from logging and other riparian removal projects; suggesting that the short-term impacts of selective Rhododendron removal on benthic macroinvertebrates are comparable to that of logging activity. / Master of Science

This thesis is based on literary resources, and original and current catalogues. First part of this thesis pays attention to description of main characteristics of garden and landscape architecture in the period between first and second world war in Czechoslovakia. This part also describes the use of Rhododendron L. genus, its history, its breeding in Europe and Czechoslovakia, and its breeding process in Průhonice. Main objective of this thesis is to complete and specify informations about available Rhododendron L. genus product range in Czechoslovakia in the period between first and second world war, to describe possibilities of use of the genus, and to describe and evaluate recent product range of the genus available on the market. Final part of this thesis is a summary of well known species produced until 1938 according to original resources. This summary is then compared with the recent product range of Rhododendron L. genus available on the market. As a result of this thesis we get the summary of well known Rhododendron L. species used in the chosen period, and comparison of those with the recent product range.

Suppression of canopy tree recruitment beneath rapidly spreading thickets of Rhododendron maximum L. (Ericaceae) in southern Appalachian forests is an issue of major concern because of the potential impacts on forest productivity, hydrology and wildlife habitat. Many studies have investigated the causes of seedling inhibition beneath dense shrub understories, but few have uncovered specific mechanisms leading to seedling decline. In this study, I have examined the influence of the evergreen understory (R. maximum and Kalmia latifolia L.) on tree recruitment processes at multiple stages - seed rain, seed bank, and post-establishment seedling growth and survivorship. Effects of dense shrub cover on seed rain and seed bank density and composition were examined using a paired treatment design in which samples were collected beneath shrub-influenced and open understories. A second experiment investigated the influence of R. maximum and K. latifolia density on the growth and survivorship of Quercus seedlings, resource availability, and the rates / causes of seedling damage. I found that neither seed rain, nor seed bank density or species richness was inhibited by the presence of R. maximum or K. latifolia. Forest seed banks were dominated by sweet birch (Betula lenta L.), and were compositionally disparate from the overstory. Analysis of resource competition between shrubs and seedlings indicated that seedling performance and survivorship was a negative function of R. maximum density. Open-canopy light availability, nitrogen content in the organic horizon (litter and humus), and soil nutrient availability were potential resource-related mechanisms. Further, I found that the rates of insect herbivory on Quercus seedlings were positively correlated with R. maximum density. Kalmia latifolia had little influence on resource availability, seedling performance or herbivory rates, and does not appear to have a suppressive effect on tree seedlings. Overall, this research indicates that resource competition is the primary mechanism by which seedling suppression occurs beneath R. maximum, and that increased herbivory on seedlings may be an additional mechanism that demands further study. / Master of Science

Invasive species introductions, namely the chestnut blight fungus (Cryphonectria parasitica) and hemlock woolly adelgid (Adelges tsugae), have permanently altered the overstory canopy of Appalachian forests by causing the dramatic die-offs of two ecologically significant tree species, American chestnut (Castanea dentata) and eastern hemlock (Tsuga canadensis). These canopy dominants once had significant roles in regulating understory communities. The loss of these trees, along with fire suppression, has driven two common evergreen shrubs, rosebay rhododendron (Rhododendron maximum) and mountain laurel (Kalmia latifolia), to expand and proliferate in areas where they were once restricted. These two common shrubs are recognized agents of change in Appalachian forests because of their abilities to modify soil seed banks, regulate light, and alter the local soil chemistry. This study documented evergreen shrub expansion across the Coweeta Creek basin over the past 36 years analyzing changes in winter greenness using harmonized multi-decadal archives of Landsat imagery. We found the greatest change in winter greenness in relatively dry areas: higher elevations (1275–1300 m), steeper slopes (33°–35°), southward aspects, and far from streams (600–800 m). Historical field data collected in three unmanaged watersheds at Coweeta showed a simultaneous decrease in T. canadensis and increase in R. maximum. We also documented the decline of a xerophytic canopy tree species, pitch pine (Pinus rigida), and an associate understory shrub, K. latifolia. Our analysis of the influence of terrain variables on evergreen shrub expansion allowed us to determine which of the two species was expanding in various locations with reasonable certainty. This study provides spatially explicit data on the expansion of two evergreen shrub species at the Coweeta Hydrologic Laboratory that could be used to pinpoint areas for future management interventions. / Master of Science / Forests in the eastern United States have changed substantially in response to the introduction of highly competitive invasive species. Some overstory tree species have been virtually eliminated from their functional role in regulating understory vegetation in many southern Appalachian ecosystems. Die-offs of these trees have allowed understory evergreen shrubs to expand into areas where they were once restricted. Shrubs that have expanded in response to overstory tree die-offs can alter the ecology of forests for the foreseeable future. Our work leveraged multi-decadal archives of wintertime satellite imagery to document the spread of understory evergreen shrubs in a watershed located in western North Carolina. We investigated the relationship of this spread to local environmental characteristics like elevation, steepness (slope), slope direction (north, south, east, west), and distance-from-stream. The greatest changes in evergreen vegetation were documented on terrain at relatively high elevations, locations farther from streams, on southerly aspects, and on relatively steep terrain. We included historical field data collected during the same time period that showed a simultaneous increase in two understory evergreen shrub species after the die-off of ecologically significant overstory tree species. This information will help forest managers by describing areas where substantial spread has occurred and potentially use this information to inform future management action.

Thickets of Rhododendron maximum (Rm) in the southern Appalachians impose severe limitations on the regeneration of hardwood and coniferous seedlings. Interactions between Rm thickets and ectomycorrhizal colonization were examined to explain seedling inhibition. Experimental blocks were established in and out of Rm thickets in a mature, mixed hardwood/conifer forest in Macon County, North Carolina. Planted seedlings of red oak (Quercus rubra) and hemlock (Tsuga canadensis) were harvested at the end of the first and second growing seasons. Litter manipulation had no effect on total mycorrhizal colonization. Mycorrhizal colonization and ramification index (# mycorrhizae cm-1) were depressed and colonization by Cenococcum geophilum increased in blocks with versus without Rm. After the first year, percent colonization of T. canadensis not in Rm thickets (62 %) was three times higher than in Rm thickets (19%), and the ramification index was increased by more than a factor of four (2.83 versus 0.61). Mycorrhizal colonization levels were correlated with root weight and shoot weight in both hemlock and oak seedlings, but did not explain most of the variation observed. Sporocarps of 69 putatively ectomycorrhizal species were collected on the blocks. Species diversity and overall community structure was similar in and out of Rm thickets. Individual species, i.e. Lactarius speciosus and Russula krombholzii, were significant indicators of forest without Rm thickets. Rhododendron maximum thickets probably affect the process of mycorrhization. The reduced level of mycorrhizal capacity under Rm thickets could be a factor in the increased seedling failure in Rm thickets. / Master of Science

<I>Rhododendron maximum</I> L. (Ericaceae) is an evergreen shrub that grows in dense thickets and currently covers large areas of the understory in the deciduous forests of the southeastern United States. Thickets of R. maximum are inhibitory to recruitment and regeneration of many understory plants including canopy tree seedlings. By effectively lowering the survivorship of woody species trying to establish within thickets, <I>R. maximum</I> could influence stand level regeneration patterns and ultimately the community structure of these deciduous forests. This dissertation outlines research conducted to determine if: 1) below and above ground resources are lower within thickets of <I>R. maximum</I> when compared to forest sites where <I>R. maximum</I> is absent; 2) <I>Quercus rubra</I> and <I>Prunus serotina</I> seedlings growing in thickets have lower mid-day photosynthetic rates; 3) <I>Quercus rubra</I> and <I>Prunus serotina</I> seedlings growing within thickets are low light acclimated when compared to seedlings growing in forest without <I>R. maximum</I>; 4) the presence of <I>R. maximum</I> constrains CO₂ assimilation of <I>Quercus rubra</I> seedlings exposed to light flecks of different durations and intensities; 5) the presence of <I>R. maximum</I> constrains the light fleck responses of <I>Quercus rubra</I> seedlings exposed to eight light flecks in rapid succession; and 6) canopy openness regulates the capacity of <I>Quercus rubra</I> seedlings to assimilate carbon when exposed to eight consecutive light flecks. <I>Rhododendron maximum</I> thickets altered resource availability for seedlings when compared to areas of forest without <I>R. maximum</I>. Diffused photosynthetically active radiation (PAR) averaged less than 5 μmol m⁻² s⁻¹ throughout the growing season in sites with <I>R. maximum</I> in comparison to 10-30 μmol m⁻² s⁻¹ in sites without <I>R. maximum</I>. Soil moisture content, measured using Time Domain Reflectometry was approximately 6% lower in forest sites with <I>R. maximum</I> compared to sites without <I>R. maximum</I> throughout the growing season. Most nutrient concentrations (e.g.,, C, N and most cations) and nitrogen mineralization rates were significantly lower in sites with <I>R. maximum</I>. Temperature and atmospheric relative humidity are slightly lower under thickets of <I>R. maximum</I>. In general, sites with <I>R. maximum</I> are associated with lower resource availability above and below ground in comparison with sites without <I>R. maximum</I>. Attenuation of below canopy PAR by thickets of <I>R. maximum</I> negatively influences the photosynthetic capacity of <I>Quercus rubra</I> and <I>Prunus serotina</I> seedlings as indicated by measurements of mid-day photosynthesis. In 1996, the seasonal mean mid-day photosynthetic rate of first year <I>Q. rubra</I> seedlings growing in <I>R. maximum</I> thickets (1.3 μmol m⁻² s⁻¹) was 62% lower than the seasonal mean mid-day photosynthetic rate (2.1 μmol m⁻² s⁻¹) of seedlings growing in forest sites without <I>R. maximum</I>. For second year seedlings in 1997, seasonal mean mid-day photosynthesis was 183% higher for plants growing outside of thickets (1.7 μmol m⁻² s⁻¹) compared to the mean rate (0.6 μmol m⁻² s⁻¹) for plants located within thicket sites. The mean mid-day PAR available to seedlings located in forest sites without <I>R. maximum</I> during measurements of photosynthesis was 354% higher in 1996 and 257% higher in 1997. First year <I>Prunus serotina</I> seedlings growing in forest without <I>R. maximum</I> also had greater seasonal mean mid-day photosynthesis (0.7 μmol m⁻² s⁻¹) when compared to the mean rate (-0.1 μmol m⁻² s⁻¹) for plants growing within thickets. <I>Prunus serotina</I> seedlings located in the presence of <I>R. maximum</I> received on average 67% less PAR. Photosynthetic acclimation to low light was assessed for <I>Q. rubra</I> and <I>P. serotina</I> seedlings growing under both forest conditions by measuring photosynthetic responses to light <I>in situ</I> using even aged one-year old seedlings. <I>Quercus rubra</I> seedlings growing in forest sites without <I>R. maximum</I> had significantly higher light saturated rates of photosynthesis. For both species, photosynthetic responses to light were otherwise similar irrespective of the presence or absence of <I>R. maximum</I>. The impact of the <I>R. maximum</I> subcanopy on understory PAR and subsequent influence on canopy tree seedling photosynthetic capacity implies that sunflecks are critical for seedling net carbon gain in these forest understory environments. To determine the effect of <I>R. maximum</I> on the photosynthetic response to sunflecks of oak seedlings, light flecks were simulated on 288 randomly chosen, even aged, two-year old seedlings <I>in situ</I>. Half of the seedlings were located within <I>R. maximum</I> thickets. Seedlings were randomly assigned one of four light fleck durations (30, 60, 120, and 300s) and one of three intensities (100, 500, 1000 μmol m⁻² s⁻¹). Half of all seedlings were dark pre-acclimated prior to light fleck simulations by covering with aluminum foil for at least 12 hours, while the remaining seedlings were pre-acclimated under ambient conditions. Analysis of covariance showed that a significant, positive, linear relationship exists between the length of a light fleck and total carbon gain during a light fleck for seedlings in forest sites with and without <I>R. maximum</I> regardless of pre-acclimation status, or light fleck intensity. Furthermore, there was a significant effect of <I>R. maximum</I> on the slope of the relationship such that following ambient pre-acclimation, seedlings located within thickets assimilated significantly less carbon with increasing light fleck length than seedlings located in forest sites without <I>R. maximum</I>. When seedlings were dark pre-acclimated there was no difference in carbon gain with increasing fleck length between seedlings in forest with and without <I>R. maximum</I> except for flecks of 1000 μmol m⁻² s⁻¹. The data lead to the conclusion that under natural conditions the presence of <I>R. maximum</I> likely prohibits <I>Q. rubra</I> seedlings from utilizing sunflecks as effectively as seedlings growing in forest sites where <I>R. maximum</I> is absent. Because sunflecks often occur clustered together during a short period of time during the day, another field study was conducted to further characterize the effect of <I>R. maximum</I> on the photosynthetic response of oak seedlings to eight consecutive light flecks. Within 10 paired sites, (i.e., with and without <I>R. maximum</I>) 3 even aged three-year old <I>Q. rubra</I> seedlings were selected. Over each seedling, a hemispherical canopy photograph was taken and analyzed for percent canopy openness. Each seedling was dark pre-acclimated for 12 hours and then exposed to eight light flecks in rapid succession during which time photosynthesis was logged every two seconds. Each light fleck was 500 μmol m⁻² s⁻¹ in intensity and lasted for 120s. Following each light fleck, leaves were exposed to 10 μmol m⁻² s⁻¹ PAR for 60s before the next light fleck. Mean carbon gain and maximum photosynthesis achieved during each light fleck was significantly lower for seedlings located in the presence of <I>R. maximum</I> for all flecks in an eight-fleck simulation. In addition, seedlings located within thickets generally had significantly lower pre-illumination photosynthesis following the first of eight light flecks. The mean photosynthetic light use efficiency of seedlings located in forest with <I>R. maximum</I> was significantly lower for the first six of eight light flecks in succession. Using regression analysis and analysis of covariance, percent canopy openness was used to explain the variation in carbon gained from all eight light flecks in succession for seedlings under both forest conditions. However, significant relationships failed to exist between under either forest condition and precluded using analysis of covariance. The results from these studies lead to the conclusion that light limitation is a major mechanism responsible for the extirpation of canopy tree seedlings from within thickets of <I>R. maximum</I>. Tree seedlings growing in forest sites with <I>R. maximum</I> receive less solar irradiance, have lower mid-day photosynthesis, fail to acclimate to the lower light conditions within thickets, and utilize sunflecks less effectively as well as less efficiently when compared to plants growing in forest sites without <I>R. maximum</I>. / Ph. D.

Rhododendron maximum L. is a dominant subcanopy species in the southern Appalachian Mountains. R. maximum undergo distinct thermonastic leaf movements (TLM). The purpose of these movements has not yet been determined. Previous studies have suggested TLM are a photoprotective mechanism for the dynamic light environment of the subcanopy in a deciduous forest during winter. The present study aimed to determine the effects of restricting TLM on photoinhibition, net photosynthesis, and other gas exchange parameters, particularly during the early spring. After restricting TLM on certain leaves, we observed the above parameters from autumn 2005 to late spring 2006. Our results indicated that photoinhibition increased (lower Fv/Fm) in treatment leaves over reference leaves throughout the winter. The difference became greater during the early spring, when reference leaves began to return to normal levels of photochemical efficiency and treatment leaves sustained low Fv/Fm. Net photosynthesis was lower for treatment leaves than reference leaves. This became most significant during the early spring, when maximum carbon gain is possible. Finally, gas exchange parameters as measured by light and CO2 response curves did not indicate any significant difference between treatment and reference leaves post canopy closure. Out results suggest that TLM are an important mechanism for photoprotection, allowing leaves of R. maximum to recover quickly during the early spring and maximize their early spring carbon gain. / Master of Science