This study investigated the zonation of the halophytic shrub Atriplex portulacoides (chenopodiaceae) on three main habitats on the high marsh, shingle ridge and low marsh at Stiffkey saltmarsh, on the north Norfolk coast. The aim was to examine the physical and biological factors controlling the distribution of this species within an apparently wide amplitude of elevation within the tidal frame. The field investigation revealed significant differences in some of the physical characteristics of the sediment of five sites. Soil drainage and soil texture were significantly different between the shingle ridge and the general marsh habitat. Other environmental factors, including duration of tidal submergence proved to have significant effects on the distribution and percentage cover of the plant. Sediment accretion rates were greatly variable between the high and low marsh and may be responsible for low seed germination rates of this species on the low marsh. Laboratory experiments showed that the seed germination was largely inhibited at high salinities. Seed germination can occur under anoxic conditions but only in the presence of light. Seed germination in the laboratory was significantly higher under alternating temperature (20/10 °C) than at continuous lower temperatures (1 or 4 °C). Seed germination under burial with sediment was significantly reduced and this is due to the failure of seeds to germinate in the first place and also due to the difficulty in emergence from burial. Glasshouse experiments showed that long-term waterlogging created hypoxic conditions that were harmful to growth of A. portulacoides at all three stages of the life history tested; mainly changes to morphology occurred. However, the seedlings are the more adversely affected than young established plants or the mature plants. Unlike the seedlings, the young established and mature plants were relatively tolerant to waterlogging, mainly because they produced abundant adventitious roots from their aerial branches. Seed and seedling transplant experiments in the field showed that plants can not establish lower on the marsh than their current lower limit because of physicochemical conditions. The obstacle is mainly at the seed germination, establishment phase as transplanted seedlings had much higher survival rates generally than transplanted seeds. However, the low marsh was generally more conducive to seedlings than the high marsh. Certain biotic interactions were also investigated. Competition of A. portulacoides with Puccinellia maritima tested in a prolonged experiment in the glasshouse with and without waterlogging showed no significant effect on the dry mass of either species. Fruits (seeds) of A. portulacoides were significantly predated by larvae of the moth Coleophora atriplicis.

Soil samples were collected and processed, mainly from woodland, but also from non-woodland soils, particularly from designated Sites of Special Scientific Interest. The survey focussed on the East Anglian region and the north of England and Scotland, where Cavender & Raper's semi-quantitative isolation method was used to assess dictyostelid population densities at the sites surveyed. Qualititative species data only were also obtained from additional soil samples collected from within the United Kingdom. Isolation and identification methods were examined, and a total of 12 described species were recorded. Of these 12 species, 5 were the first reports for the United Kingdom, D. fasciculatum, D. polycarpum, P. candidum, D. implicatum and D. septentrionalis, the latter two being the first isolations of these species in Europe. In addition, one new species of Polysphondylium was isolated from soil collected from Silverdale Glen in the Isle of Man and a putative new Dictyostelium species was isolated from Thetford Forest, Norfolk. The variant, D. aureo-stipes var. helvetium, was also reported for the first time and one D. mucoroides variant was encountered at two sites. Images are presented ofall of these, together with a detailed description of the new species of Polysphondylium, and the characteristics of the putative new Dictyostelium isolate, together with measurements of both of these, which were obtained using image analysis. Allozyme electrophoresis was employed in an attempt to establish the affinities of the new species of Polysphondylium. A relational database was designed and constructed to facilitate identification of isolates and analysis of dictyostelid population densities. The data obtained during this survey were combined with that obtained by Turner (1978) to present a complete species distribution list of dictyostelids, to date, in the United Kingdom. A dictyostelid culture collection has been constructed, containing representatives of most of the strains encountered during this survey, which is available for future taxonomic study.

Management of the southern lowland heaths is usually for conservation, and aims to create structural diversity, prevent succession to scrub and maintain the low nutrient status of the system. The main aim of this study was to investigate the effects of burning, fire temperature and intensity on seed banks, nutrients, vegetative regeneration and seedling demography. In the first experiment (1994), vegetation fuel loads were manipulated in stands representing two different growth phases of vegetation (mature and degenerate) at two sites (Arne in Dorset and Aylesbeare in Devon) to test the hypothesis that frre temperature and intensity depend chiefly on fuel load rather than other factors associated with the pre-burn age of the stand. There were two burning treatments of low and high fuel loads. Fire temperatures were measured with temperature-sensitive paints on tiles ('pyrometers'). Intensity (the heat released per unit area) was highly correlated with fuel load. Fire temperatures were higher in the high fuel load treatments than in the low fuel load treatments at all stands except the Aylesbeare degenerate stand. where temperatures were low in both treatments. In the second experiment (1995). paraffin was added to treatment plots at the mature stand at Aylesbeare in an attempt to produce higher temperatures. but this aim was not achieved. Thermocouples and pyrometers were used to measure temperatures and indicated that these fires were hotter than the fires at the same stand in the first experiment. There was no effect of burning or of temperature on the size of the seed banks in the soil and litter in the frrst experiment. The litter seed bank of Erica tetralix was depleted by a similar quantity in both treatments in the second experiment. There was no effect of burning on the nutrients (P. Ca. K) in the humic soil in either of the experiments. The concentration of nutrients was higher in the ash than in the litter of control plots after burning (except at the Aylesbeare degenerate stand). There were significant but inconsistent effects of temperature on the quantities of nutrients. Vegetative regeneration after burning depended chiefly on the pre-bum age of the stand and possibly on grazing. but fire temperature had no effect. There was rapid resprouting at both the mature stands, but regeneration at the degenerate stands was sparse. Calluna rapidly regained dominance at the Arne mature stand but was relatively less abundant at the Aylesbeare mature stand after burning than it was before. Agrostis curtisii. Ulex gallii and Erica tetralix were co-dominant in the regrowth at this stand. However. there was a decline in the relative abundances of U. gallii and A. curtisii by the third growing season after burning. At the mature stands. seedlings affected by resprouting vegetation were sheltered from microclimatic extremes and experienced higher survivorship than seedlings in areas of bare ground. However. seedling density under resprouts decreased over time as continued growth of resprouts prevented seedling germination and establishment. Seedling populations on bare ground at both types of stand oscillated widely, and there was a rapid turnover. Although bare ground persisted at the degenerate stands, there was no invasion of non-heathland species. A canopy of ericaceous species should eventually develop via gradual accumulation of seedling density and vegetative spread from the few stem bases that were able to resprout. The range of temperatures generated by the fuel loads tested did not have significant effects on regeneration from seed or from stem bases. However. the removal of vegetation by burning had a large impact on vegetation dynamics by enabling seedling emergence and resprouting from stem bases. The balance between regcneration from seed and from stcm bases depended on the pre-bum age of the stand.

The evolution of chemos tat E. aoZi-Avir communities was studied with the aim of clarifying the mechanism of their persistence. Under the physical and nutrient conditions of the experiments these organisms may coexist indefinitely but the p'attern of population dynamics varies considerably from culture to culture. Generally, the evolution of phage-resistant bacteria stabilises the total bacterial numbers and reduces the density of the phage population. The latter persists, however, at very high densities and fluctuates irregularly. Two obvious evolutionary events, ·the evolution of resistance in bacteria and of a delayed-plaque-growth phenotype in phage, were associated with major changes in'population densities. Through extensive phenotypic characterisation of the "selected" strains, however, many more evolutionary events were traced which did not correspond to density fluctuations in any obvious manner. The "selected" bacterial population consisted of several distinct maZ+- (ability to utilise maltose)'phenotypic subpopulations some of which were totally resistant to both the "selected" and the original phages; some were sensitive to the selected phage and to host-range mutants of, but resistant to, the original phage (partially resistant); some appeared resistant but acquired sensitivity in the presence of maltose (semiresistant); some were fully sensitive to all phages. Other characteristics of the "selected" bacteria included mucoidity, larger cell size and longer intrinsic generation time, high spontaneous mutability, inhibition of phage intracellular development in the ,. presence of mitomycin C and low frequency of lysogenisation by temperate . strains, all suggestive of evolution of their SOS-repair, recombination, DNA metabolism and cell division gene-system. The altered antigenic specificity, the increased sensitivity to chloroform and the increased adsorbing ability of the "selected" phage are interpreted as pleiotropic effects of the evolution of its adsorption site. Its ability to infect partially resistant bacteria was, however, due to a second evolutionary event since the partially resistant isolates were found to block the infective process at a stage subsequent to the irreversible adsorption reaction. The evolution of the phage's adsorption site is explained a~ a response to the evolution of "semiresistance". A third independent response of the phage was an extraordinary increase in its'burst size which was expressed specifically on the genetic background of the coexisting selected hosts. An alternative to the existing model of the genetic Control of A-development is proposed, to explain the usual variation in burst ,size and in frequency of lysogenization as well as the evolution of these parameters; according to this model no "decision" is taken as to which developmental pathway will be followed. The delay of the selected phage in forming plaques on all hosts is suggested to be a ii side effect of the evolution of phage genes which control its development. The proposed model of A-development predicts that, 'although not isolated, lysogens and plasmid-carriers do exist in such "selected" communities and may, under certain circumstances, play important evolutionary roles. However, for as long as the extracellular conditions do not inactivate the virulent variants, the latter may persist by evolving their virulence (host range and productivity of infection) in response to any ev&lution of resistance, or immunity, in the coexisting bacteria. The extent and, mainly, the nature of evolution in these cultures, the conclusions of an extensive discussion of the E. aoZi-Avir system~s iii biology, as well as the conclusions of a theoretical and empirical study of "fitness" and its determinants,severely contradict fundamental assumptions and specific (for this sytem) predictions of previous equilibrium and non-equilibrium theories of ecological stability. It is proposed that coexistence of E. coZi and 'Aviro is achieved through continuous, random, indefinite coevolution of resistance and virulence, respectively, and that the apparently erratic deviations of densities from constancy are the ecological - manifestations of the successive extinction and replacement of genotypic sUbpopulations.

Ph. D. thesis submitted to the University of London for the degree of Doctor in Philosophy. Sampling of the surface waters of the Firths of Clyde and Lorne, on the Scottish west coast, established the seasonal presence of detached floating macrophytes from April to October. The neuston of the Firth of Lorne was characteristic of coastal waters and relatively impoverished. In contrast the neuston of the Firth of Clyde tended to be abundant and was often dominated by the calanoid copepod Anomalocera patersoni, characteristic of oceanic neuston. The presence of floating weed provided an additional assemblage of weed associated species to those of the neuston. The association of juvenile C. lumpus with detached floating weed was confirmed. When floating weed was absent (November to March) trawling over rough ground established the year round presence of juvenile C. lumpus in inshore waters. Changes in the size distribution of juvenile C. lumpus caught were negligible compared to the growth of captive fish and the capture of larger juveniles from around moored rafts indicates that the association with floating weed is probably size dependent. The incidental capture of larvae in December in addition to the Summer months indicates a breeding season more protracted than previously reported with the probability that some level of breeding occurs year-round. The diet of juvenile C. lumpus differed fundamentally between the two study areas. The principal foods were harpacticoid copepods in the Firth of Lorne and the calanoid copepod Temora longicornis in the Firth of Clyde. Neither food type was well represented in the neuston samples described but complementary samples obtained with a fine meshed plankton dipnet indicated the numerical dominance of these relatively small components of the macrofauna. Comparison of the composition of the diet with that of the macrofauna showed that lumpsuckers feed as size generalists. The energetic costs of sucker attachment and continuous swimming were compared by respirometry. Although swimming was more costly than attachment, as lumpsucker weight increased the weight specific cost of swimming decreased at a higher rate than that of attachment so that the energetic advantage declined. Behavioural aspects of the lumpsuckers association with floating weed were investigated in the laboratory with plastic structures as weed analogues. The lumpsuckers preferred black structures located at the surface. Structural complexity was also preferred, particularly when concealment was possible. The balance between feeding motivation and predation risk was investigated at two hunger levels in a series of habitat selection experiments. The predicted use of shelter and activity were not fully realised in the results.

This study was part of an ongoing investigation into the long-term effects of nitrogen enrichment and environmental stress on lowland heath ecology and nutrient cycling. The original experimental site, 4 replicate blocks of 2m x 1m plots, was established on pioneer-phase Calluna vulgaris - Deschampsia flexuosa heathland in 1996 (Cawley, 2000). Nitrogen treatments (0, 20, 60 & 120 kg N ha" yr") have been applied on a fortnightly basis, against a background deposition of -20 kg N ha'! yr" (NOx & NHy). During 1997 the plots were split and a six-month drought versus non-drought treatment introduced (full details in Cawley, 2000). The period of environmental stress experienced by the dominant C. vulgaris canopy was extended over the following two years (1998 & 1999) by a natural outbreak of Lochmaea suturalis. Zero-tension lysimeters were installed under the non-droughted half of these plots (2001), facilitating study of increased nitrogen deposition effects upon nitrogen losses through leaching.' A second set of experimental plots were started in 2000 to enable study of germination and establishment of D. flexuosa under conditions of increased nitrogen deposition and canopy gap creation in an otherwise closed C. vulgaris canopy. This study has shown that elevated, long-term deposition of atmospheric nitrogen in combination with periods of environmental stress can contribute to significant changes in the ecology and nutrient cycling of a lowland heath. Elevated nitrogen deposition was found to induce significant increases in: tissue nitrogen content of Calluna and Hypnum sp.; litter nitrogen content; the loss of nitrogen from the system through leaching. Additionally, there were significant reductions in the C/N ratio of both vegetation and soil. This could have major implications for changes in the rates of nitrogen mineralisation and immobilisation in this nitrogen-limited ecosystem. By 2001, few effects of the 1997 drought continued to be apparent in the heathland vegetation. However differences were recorded in the nitrogen content and C/N ratio of the soil components of the droughted and nondroughted plots. Some responses were not as expected, such as the transitory increase in the Deschampsia population. The predicted outcome, based on studies of other European lowland heaths, would have been for a marked transition to a Deschampsia-dominated canopy, especially in those high nitrogen (80N & 140N) treatment plots which had been droughted in 1997. Nitrogen deposition levels in excess of the critical load for dry heathlands (10 - 20 kg N ha" yr"), in combination with Calluna canopy opening induced by environmental stress, have been associated with such transitions in vegetation on Continental heaths. An equally unexpected outcome of this study was the significant recovery of the moss layer after the cessation of drought conditions in 1997. Again, reports from other European heaths have suggested that increased nitrogen deposition would normally be associated with a loss of bryophytes and cryptograms. Elevated nitrogen inputs, at this study site, have been associated with a reduction in bryophyte species diversity. Conversely, there has been a corresponding, significant, increase in the cover of nitrophilic Hypnum species.

1.Myxogastrids were shown to be common on bark of trees in the Bristol area and a variety of species were isolated. 2.Myxogastrids were more common on old bark than young bark. 3.Myxogastrids were isolated from banana peel but not from a variety of dung samples. 4.Myxogastrids were isolated from soil and a medium and method for their enumeration were developed.Soil was chosen as the habitat for further quantitative research. 5.The recor"ding unit for myxogastrids was the Plasmodium Forming Unit (P.F.U.). 6.Myxogastrids were found to be patchily distributed in soil and more abundant in the surface layers than from several centimetres deep. 7.A precise and repeatable sampling system was devised consisting of many small samples pooled to form a single large sample. Samples from nine sites were taken at monthly intervals for one year. These samples were assayed for myxogastrids by a standard proceedure. 8.The biological and edaphic parameters of the series of soil samples were also measured and weather records from the Long Ashton recording station collected. 9.A series of soil samples was taken from Sites of Special Scientific Interest in the Bristol area and assayed as for the temporal samples. IO.A number of soil samples were obtained from various countries and assayed for myxogastrids. 11.Myxogastrids were not abundant in typical woodland soils but they were in many other sites, even including a Barley field. 12.Graphical analysis of some of the biological parameters and of the soil moisture of the temporally sampled sites showed that there was a high negative correlation between soil bacterial numbers and P.F.U.'s in response to changes in soil moisture. 12.Populations of soil ciliates and amoebae showed a response to soil bacteria similar to that shown by the P.F.U.'s but not always synchronously. 13.Frost appeared to lead to increased numbers of myxogastrids in soil. l4.Statistical analysis of the the ecological data showed a high positive correlation between P.F.U.'s and pH and negative with organic matter. Numbers of P.F.U.'s were highly positively correlated with those of other soil protozoa which, with the exception of dictyostelids, all showed similar responses to soil physical properties • IS. The S.S.S.I. data analysis showed similar correlations to that of the temporal data but not as strongly. 16.The foreign samples showed myxogastrids to be widespread and not confined to normally moist soils for some were recovered from deserts. 17.Myxogastrids were amenable to experimental manipulations used to follow their population dynamics, and were shown to possess forms resistant to freezing. The latter appeared to stimulate excystment of dormant P.F.U.'s. 18.Microcosm experiments showed a pronounced rhizosphere effect, but a role for myxogastrids in the enhancement of nutrient cycling and uptake was not conclusively shown. 19.A series of samples from Czechoslovakia showed similar characteristics to samples from the Bristol area. 20.Most species of myxogastrids isolated from soils were of the genus Didymium but with the exception of Didymium squamulosum • I and D. difforme conclusive identification of most 1solates was not.possible. 21. Calculations suggest myxogastrids represent a significant biomass in some soils and that this biomass corresponds to a similarly significant biomass of bacterial prey.

Experiments were made to determine the effect both of shading and of the level of the water-table on the growth of S~hagnum. Growth in mass is reduced by shading but is little affected by water level. Growth in length is greatest with high water levels and reaches a peak when about 80% of the incident radiation is absorbed by shading. Etiolation (straggliness) increases with shade and with higher water levels. There is little interaction between the effects of shade and water level. Different species behave differently; there is some indication that growth is greatest in a species' 'natural' environment although on average, production is the same for all three species investigated. (2) Investigations were also made to ascertain how Sphagnum modifies the supply of light and water to its growing apex. Suitable instruments for measuring light flux in a Sphagnum canopy and moisture profiles in a core are described. The attenuation of light in a Sphagnum profile follows Beer's Law with depth (or more precisely with cumulative dry mass). Attenuation is affected by the previous environment. Shaded conditions and a high water-table both reduce absorbance, an inverse relationship with etiolation. Moisture profiles in Sphagnum cores with various water levels are shown. Profiles of the volume occupied by pores of different sizes are obtained. S. rubellum has a smaller average pore size than S. papillosum and is able to maintain a larger water supply to the apex. A hysteresis effect due to the previous history of the watertable level is investigated. (;) A computer model of a Sphagnum lawn is shown to be capable of simulating the growth of the moss and in particular to reproduce the observed increase in 'surface roughness' with high water levels and shading. In mixed lawns of two species, each is shown to 'out-grow' the other in its own 'natural' environment.

(1) Transect data from the Lydd-on-Sea area at Dungeness establishes the detail of the primary shingle succession. This culminates in what is, probably, a climax community about 2 km inland in the RSPB reserve. (2) A pH gradient was found at the Lydd-on-Sea site, and this was shown to be correlated with the age of the ridges (F-14.78, P < 0.001). concentrations of "total" elements in the humic soils of the shingle ridges, are not correlated with ridge age but are shown to be correlated with concentrations in sea water (r2• 0.526, F-20.04, P < 0.001). (3) Lichens are seen as playing no significant part in the early succession on the ridges at Lydd-on-Sea, but do compete successfully with higher plants in the mid- to late-succession communities. (4) A quantitative and, to a degree qualitative, assessment of mite populations was made on seven lichen species of the late succession communities at Dungeness. No specific mite numbers/lichen species association was found, but a single cyclic, bi-modal response curve was obtained for total mite numbers on the seven lichen species sampled (F - 25.01, P < 0.001). (5) A range of climatic parameters were assessed, both in relation to the ecology of Dungeness, and in particular to the mite populations on lichens. Only temperature was shown to be correlated (negatively) with mite numbers (F-6.3, p < 0.05), whilst rainfall and humidity were not significantly correlated (F- 3.99, P > 0.05, and F - 0.52, P > 0.4 respectively).

published_or_final_version / Botany / Doctoral / Doctor of Philosophy