Nardus and Molinia dominant grasslands (Nardetum and Holinietura respectively) cover large areas of the rough grazings of Scotland. There has been a general view that, in many cases, these two grasslands were derived from heathland as a result of a deflected succession caused mainly by heavy grazing and over-frequent burning. From the point of view of economics, the change from heathland to Nardetum or Molinietum is considered as a 'loss' since the former type supplies greater grazing value, all the year round, to sheep, cattle, deer and grouse. Both Nardus and Molinia can exist in a wide range of habitats, constituting several communities which, though dominated by one of them, are distinct in their composition, structure and dynamics. This study aimed to understand the processes and mechanisms (including management) that cause these patterns and to assess the possibility of re-introducing Calluna either by sowing seeds or, if the seeds are already present in the substratum, the treatment/s that encourage them to germinate and the seedlings to establish. The study also includes a review of the autecology of Nardus and Molinia and observations on their phenology and interactions with other species as affected by various grazing regimes.

The purpose of this study was to test the hypothesis that food availability was responsible for the low population density, larger home range size, high degree of daytime activity and longer activity periods of wood mice living in a sand dune habitat compared to those living in deciduous woodlands in Grampian, North-East Scotland. An experiment was set-up whereby one population of mice was provided with superabundant food in the form of wheat and the parameters mentioned above were measured and compared with an adjacent control population with natural levels of food. Seasonal changes in population density, demography and breeding biology were obtained via live-trapping. The home range size and activity patterns were measured by the use of radio-telemetry. The presence in. or absence from the nest was recorded automatically by a minichart recorder. The supplemented population had a higher density (2-3 times), higher recruitment rate (2-8 times), earlier start to the breeding season, increased juvenile recruitment and lowered weight of sexual maturity compared to the control population. Supplemental food did not prevent the normal population decline over winter, improve survival rates or extend the breeding season into winter. It was concluded that food interacts with behavioural and other environmental factors to. limit population density. The supplemented mice had significantly smaller home range sizes both in the breeding and non-breeding seasons compared to the control mice. The supplemented mice moved significantly shorter distances per night compared to the control mice. The supplemented mice were strictly nocturnal, emerging from the nests after sunset and returning before sunrise, and spent significantly less time active, both in the breeding and non-breeding seasons compared to the control mice. No regular short-term rhythms were observed in the individuals monitored in both populations. Diet analysis showed that the mice were feeding mainly on arthropods throughout the year. The low density and mobility of the food resources would account for the differences in home range size and activity patterns of the control mice mentioned previously.

This study investigated the ecology of a boreo-alpine crane-fly, <I>Tipula montana</I> Curtis (Diptera: Tipulidae), on a Scottish mountain summit. It explored adaptation to the upland environment, placing emphasis on sources of variation in growth, abundance, and low temperature tolerance. The development stages of larvae were reliably identified from measurements of the spiracular disc diameter. The sex of larvae were distinguished by the bimodal frequency distribution of weights in the 4^th instar. Larvae were sampled by hand searching and collecting off snow patches, between 1994 and 1996. The timing and size of development stages indicated that the population possessed a 2-year life cycle, supported by evidence from adult emergence. However, by rearing larvae of known developmental stage at the field site, it was shown experimentally that microhabitat could affect growth, and that it was likely to be related to temperature. In the laboratory, pre-pupal development was shown to be temperature dependent. 2^nd instar larvae were reared on single bryophyte diets, from a range of genera. They were all capable of growth and moulting, although final weights differed by a factor of 2. A pair-wise preference experiment, using 4^th instars, showed the individuals broadly chose bryophyte foods as expected from their quality, although the sedge <I>Carex bigelowii</I> was preferred over all mosses. No intersexual difference was detected. The observational method of assessing preference, compared to faecal pellet analysis, showed a bias towards some species, perhaps because they acted as a refuge. The pellet contents of a small number of larvae from the field were largely unidentifiable. Larval cold tolerance was investigated in the laboratory. The supercooling points of 3 different species (<I>T. montana, T. varipennis & T. confusa</I>), from upland and montane habitats, were very similar in autumn and spring, ranging from means of -4.5 to -6.0^oC.

The microtine rodent <i>Microtus arvalis orcadensis</i> is endemic to 5 of the Orkney Islands. Aspects of the ecology of this animal were studied on Mainland Orkney during the period June 1988-December 1990. Subject areas included habitat-related variations in population density, social organisation and short-term and seasonal variations in activity. The importance of the Orkney Vole in the diets of Hen Harriers, Short-eared Owls and Kestrels was quantified and assessed, both in the context of optimal foraging theory and in relation to the contribution made by female avian predators to nest provisioning. The selection of hunting habitat by avian predators, in relation to spatial variations in vole population density and temporal variations in vole activity was also considered, particularly with respect to Short-eared Owls. Using land-cover data derived from air-photo interpretation, estimates of the total population of Orkney Voles were made and the implications of changing land-use for the conservation of both Vole and avian predator populations considered. Specific types of land management are proposed to ensure that populations of both the Orkney Vole and the avian predators which depend to a varying degree upon them, are effectively conserved.

An increase in fox numbers in Aberdeenshire,dating from the early 1950's, was documented from parish bounty returns. Extending its range, the fox reached the coastal National Nature Reserve, "the Sands of Forvie", in about 1962 and subsequent depredations of nesting birds there initiated this study. Numbers were considered still to be increasing, though data on fox range sizes suggested that this will cease within a decade. Fox scats from Forvie showed the diet consisted mainly of rabbits in all seasons with wildfowl second in importance. In surrounding areas stomach analysis showed similar results though game-birds replaced the wildfowl. Rabbits were counted on Forvie on dawn transects and there was little annual change in numbers. Consistent seasonal changes were not reflected in any difference in fox diet. Rabbits may act as a 'buffer' protecting birds from more serious predation. Nesting female eider duck were the birds most frequently subject to predation, though overall their breeding success was similar to that found before foxes arrived. In one rushy area fox predation was significant, possibly related to denser nesting there. It was estimated that adult foxes and cubs consumed up to 4% of the adult female eiders in any year,which represents almost half of their annual mortality previously. The Forvie eider population, which does recruit from elsewhere, has not decreased or shown a detectable change in sex ratio since foxes arrived. In the past the nesting of terns on Forvie was seriously disrupted by foxes but the continued use of an electric fence has contributed to successful seasons since 1974. Suggestions for improving its efficiency were given, together with methods for monitoring the levels of fox predation, which will indicate if fox numbers should be reduced. The impact of foxes on prey species in other studies was considered and it was concluded/that the Forvie eider population could withstand present predation levels.

Predatory arthropods such as carabid beetles and spiders are generally thought to be favoured by crop management that results in dense vegetation. In grassland, silage production increases vegetarian density compared to grazed grass but involves disturbance from cutting, while intercropping with clover increases the vegetation density of vegetable crops, but causes yield loss through interspecific competition. The first part of this thesis compares the predatory arthropod fauna of silage with grazed grass. There were small differences between the carabid assemblages of the two habitats before the first silage cut, when the contrast between silage and grazed grass was greatest, and a general reduction in species richness following silage cutting. However, the responses of individual species varied greatly. Agonum muelleri and Calathus fuscipes were associated with grazing management, but no species were consistently favoured by silage production. For Loricera pilicornis, this may have been due to cool weather creating similar conditions in silage and grazed grass, while for Nebria brevicollis, which was susceptible to disturbance, the timing of cutting determined its distribution. Spiders showed a more uniform association with structurally complex vegetation and were thus generally found in higher numbers under silage management. The second part of the thesis investigates daily movements of predatory arthropods between dense vegetation, such as clover intercropped vegetables, and adjacent open, weeded crops. The results suggested that nocturnal carabids and spiders of the genus Erigone sheltered in dense vegetation by day and moved into the open at night. Thus intercropping only parts of a vegetable crop would enhance predatory arthropod activity throughout the crop, while reducing interspecific competition. These results showed that the relationship between increased vegetation density and predatory arthropod activity is less important in the habitats studied than other factors, such as disturbance and daily movements in determining the distribution of these animals.

The deep-sea has long been perceived as a food limited environment. The recent discovery of seasonal inputs of organic matter into the deep, in areas of the North Pacific and North Atlantic Oceans, has led to questions about the behaviour of deep-sea demersal animals. This thesis describes an autonomous acoustic tracking system (AUDOS: Aberdeen University Deep Ocean Submersible) developed to address the behaviour question; primarily of the demersal grenadier fish, <i>Coryphaenoides (Nematonurus) armatus</i>, but also other shallower living species. Deep-sea acoustic pinger and Code Activated Transponder (CAT) ingestible fish tags, capable of operating at depths of 600m, are described. Deep-sea pingers operated with an early version of AUDOS, returning directional information on fish which ingested tags. The CAT, which is acoustically interrogated by a scanning sonar mounted on AUDOS, returns precise direction and range of individually tagged fish. AUDOS is an autonomous free-fall vehicle incorporating a microprocessor based logging unit which also controls a camera, current meter and compass, enabling several tagged fish to be individually located within a 500m radius. On-board environmental monitoring allows relationships between fish movements and the localised current to be addressed. Three locations were visited during 5 research cruises, Station M (32<SUP>o</SUP>50'N, 122<SUP>o</SUP>50'W) in the North Pacific Ocean, station MAP (29<SUP>o</SUP>30'N, 21<SUP>o</SUP>16'W) in the North Atlantic Ocean and the Porcupine Seabight area (51<SUP>o</SUP>20'N, 14<SUP>o</SUP>0'W) also in the North Atlantic Ocean. Tracking results indicate that <i>C. (N.) armatus</i> adopt an active foraging strategy at each of these stations. Evidence was found for a faster mean swimming speed, for the grenadier, in October (0.0692 m.s<SUP>-1</SUP>) compared to February (0.0262 m.s<SUP>-1</SUP>) at station M. Fish activity is elevated in October following a seasonal enrichment of the benthos by downward flux of particulate organic matter from the surface. At the Continental slope region of the Porcupine Seabight, the grenadier was found to be moving up the slope rather than directly using the current to forage for food (Station MAP).

This PhD studies the ecology and fishery of the veined squid Loligo forbesi in Scottish waters by making use of different spatial and statistics tools, such as Geographic Information System (GIS), Generalised Additive Models (GAMs) and Geostatistics. The PhD thesis is divided in 6 chapters, arranged in a sequential order driven by the hypothesis and question which have arisen throughout the study. Chapter 1 aims to provide a general introduction of the objectives, state of the art and techniques utilised along all this work. Chapter 2 reports the use of Geographical Information Systems (GISs) to study squid resources in the North East Atlantic. Long-term monthly average Landings Per Unit of Effort (LPUE) series are mapped against SST, SBT, SSS, SBS and Depth to reveal plausible spatio-temporal relationships between oceanographic conditions and squid abundance. Hypotheses about the squid distribution and abundance in relation to environment, habitat selection and spatial migration are suggested. Chapter 3 reports a GAM application to describing intra-annual and inter-annual variation in squid abundance in Scottish waters is also reported. Geostatistical tools were applied in chapter 4 to model spatial patterns in squid abundance in Scottish waters to explore spatial heterogeneity and distribution. Spherical and gaussian model variograms and neighbourhood analysis were made. Results provide enough evidence to indicate that there are differences between the east and the west coast of Scotland. Chapter 5 reports a two-stage model to study distribution and abundance by a Binomial and Gamma sequential approach. GAMs and Regression trees on two stages were utilised using all the available environmental variables. These non-parametric modelling was used to forecast probability of presence and abundance of 1994 by a environmental data set. Finally, chapter 6 comprises the conclusions of the work, particularly from a point of view of the ecology of the squid.

No description available.

No description available.