The pea aphid (Acyrthosiphon pisum) was one of the first aphid species for which different biotypes were described. Subsequently, the differences between biotypes were found to be consistent in time and space and several of them were given subspecies status. The differences between the subspecies is mainly ecological, their use of certain plants (the so-called marker hosts). There are hardly any differences in the morphology of the subspecies with the exception of that from Restharrow (Ononis spec. ). The performance and survival of aphids on several host plants were used to determine the degree of separation between the pea aphid subspecies and their marker hosts. To confirm the genetic basis of the host plant relations of the subspecies they were crossed. Few of the crosses showed hybrid dysfunction. The performance and survival of the hybrid clones confirmed that host plant relationships were genetically determined. There was also indication of a trade off. However, there was no indication that "Hopkin's host selection principle" played a big role in the utilisation of non-preferred host plants, with possible exception of clover. The different taxa differ significantly in body sizes. Clones from crop plants were generally bigger than those from wild plants. The genetic component of the size difference accounted for nearly 50 percent of the variances in size in wild clones. By comparing the performance of reciprocal crosses between subspecies on the marker hosts of the parents, no evidence was found that the specialised symbionts are specialised for particular marker hosts. This indicates that the aphid's genotype is the main determinant of host plant usage in the pea aphids. Furthermore, these aphids prefer their respective marker hosts. The connection between preferencea ndp erformancew as partly broken by hybridising the subspecies. The only subspecies that produces winged males and therefore has the ability to colonise other host plants, and thus the opportunity to mate with females of other subspecies, preferred sexual females of its own subspecies. The separation of the subspecies is further enhanced by the behaviour of egg laying females, which preferred to oviposit on their marker hosts. Hatching time of the eggs was also associated with the ecology of their marker host plants and probably the life history of the aphid, i.e. the subspecies that host alternates hatched first. The ecological separation between the subspecies was not confirmed by a molecular analysis, which even failed to separate the morphologically distinct subspecies from Ononis from the others taxa. The pea aphid complex is a good example of sympatric taxa, which is isolated from one another by their preference for particular marker hosts. That is, host plant is the main pre-zygotic separation mechanism, which is likely to lead the development of post-zygotic separation mechanism and eventually to fully independent species.

Abstract Aspects of the ecology and life history of the little known theraphosid Nesiergus insulanus are determined on Frégate Island in the Seychelles archipelago. The general behaviour of this species is revealed to be comparable to that of other theraphosids, with field and captive observations establishing that they are a generalist and opportunistic species. The defensive behaviour of the species is determined to be based primarily on the avoidance of conflict situations, with overt aggressive behaviour seldom displayed. Although the opportunistic existence that this species has developed is probably obligatory due to their sedentary lifestyle and limited availability of resources, this behaviour nonetheless allows the species to occupy diverse habitats and exploit niches unavailable to more specialized species.

This thesis considers the application of metapopulation theory to a field system; two species of Notonecta (Hemiptera: Heteroptera, Notonectidae), a freshwater invertebrate predator, inhabiting a series of small man-made ponds known as dewponds, in the Peak District, Derbyshire. Surveys of pond occupancy and habitat characteristics examined the habitat preferences and spatial population dynamics of the two species, and associations between Notonecta and other taxa. Interspecific competition and predation between nymphs were investigated in the laboratory and in field mesocosms to determine their potential influence on field distributions. The two species have contrasting habitat preferences, and breed in a subset of all ponds in the area. Choice of oviposition substrate appears to be an important mechanism of habitat selection. The landscape scale population dynamics of Notonecta resemble those predicted by metapopulation models, but regional persistence is determined by the availability and distribution of suitable habitat across the landscape, rather than by a dynamic balance of stochastic colonisation and extinction. Where the species co-occur, competition is likely and the outcome is influenced by the amount of submerged vegetation present, which affects the feeding efficiency of the two species. Associations between Notonecta and other taxa largely appear to result from covariance in response to habitat factors; the distributions of Notonecta and potential prey do not seem to be strongly linked. Most metapopulation models assume that habitat is static, but in common with many other field systems, metapopulation dynamics of Notonecta appear to be driven by dynamic changes in habitat of individual patches. This suggests that in order to be of greater practical value, future developments in metapopulation theory must incorporate effects of habitat dynamics on regional persistence and distribution.

The increasing need to study insect flight behaviour in the natural environment has led to the development of a relatively inex- pensive remote sensing instrument (IRADIT) . Flying insects dfferentay iumjmted, under a light cmations, in a intense ba of pulsed near infra-red radiation and detected using a gated image intensifier - CIV system at a rate of 5OHz. This thesis details the methods employed for extracting insect oo-ordinates from the' CI video signals, 'using low~oost electronic circuits ad micro computer (Apple II+) , and for the immediate analysis and display of two-dimensional insect flight trajectories, whilst rejecting environmental and system noise. Flight trajectories are often dominated by the immediate local wind. A new low-cost sensitive three dimensional digital anemometer system hs be developed for use with the IRADIT system, providing the data required to allow the vector wind to b immediately subtracted and interesting aspects of flight behaviour to b revealed. The IRADIT - Anemometer system ws tested i the field, on insects above wheat. _ Small insects with membzßanous wing areas as small as 5mm2 were readily detected and tracked over distances of the order of l at ranges up t 15m against a mid-day sky background. Subse- quent computer analysis of stored insect track and wind data oonfirmed that the effects of immediate local wind cn b smdied or canoelled out. Examples of insect flight behaviour are described. The relatively low cost of the IRADIT system and the practical features in its design result in a powerful new tool for use by the field entomologist.

Consideration is given to the use of.vane (DVA) and propeller anemometers in the sensing of atmospheric turbulence. Careful experimental measurements of the DVA dynamic response contributes to this consideration. Quantification of the airflow within and above a Sitka spruce canopy then follows. Statistics indicate the non-Gaussianity of the air turbulence Such forms are interpreted through the intermittency of turbulence and investigated by a conditional sampling scheme. Nindspeed spectra have ragged forms, showing them to be vastly modified by the combined effects of the canopy-imposed scales, together with the canopy's elemental vortex shedding frequencies and resonances. Studies of tree motions show the possibility of_a simple connection between the resonances of the separate canopy elements, which could provide the tree with an atmospheric shock-absorbing structure, supportive of its observed response to turbulence The ensuing proposal of a hypothetical qualitative mechanical design principle based on aeroelastic similaritynay bebeneficial to the tree's survivability in conditions of strong atmospheric turbulence.

Trophic cascades, together with other indirect interactions are important aspects in shaping the composition and abundance of species in the food web. Theoretically, movement of energy between systems, and coupling between habitats by mobile predators have been suggested as being important for food web stability and evenness. Individual diet specialisations have been shown to be widespread in many animal taxa. Although not widely studied, some studies have indicated that mobile predators that display individual specialisations, may have a reduced ability to couple habitats. In this thesis, by using field studies and an experimental study, my aim was to assess the individual specialisation displayed by Eurasian perch (Perca fluviatilis) and its influence on the ability of the perch to couple habitats. In the experiment, we also investigated the effect of habitat coupling, or the lack of habitat coupling, on the dynamics and stability of the resources in the ecosystem. We show that habitat diversity and resource availability influenced perch individual specialisation and morphological variation. We found that the perch total niche width decreased with decreasing habitat switching ability. We demonstrate asymmetrical habitat coupling ability in perch across pelagic and littoral habitats, providing evidence that not all individuals within a species respond in the same way when it comes to spatial coupling and thereby providing stability within a food web. Our results expand on previous work and suggest that habitat coupling ability can influence individual specialisations and niche width. Furthermore, we show the importance of individual specialisations in relation to habitat coupling. Finally, we provide evidence for the theory that a food web dominated by a food specialist should exhibit more variable resource dynamics than a food web dominated by a generalist predator by showing a greater indirect effect of predation on the phytoplankton levels when no habitat coupling occurs. While many models and theoretical concepts have proposed a stabilising effect of cross movement of energy and mobile predators, little empirical evidence exists that confirms this mechanism. In conclusion, my thesis gives some support for the theoretical predictions that habitats coupled by a generalist predator should be more stable.

Field experiments, with 125 watt Mercury-Vapour light-traps, are described in this Thesis. Catches of moths in water traps arranged in a circular mosaic pattern around the light-trap were distributed bivariate normal. Comparisons between the bivariate mean moth vector and the mean wind vector suggested that the wind vector determined the distribution of moths around the light-trap. In light winds, less than 1 metre/second, moths were randomly distributed around the trap. In stronger winds the direction of moth displacement from the trap was associated with the direction component of the mean wind velocity. A second experiment with light and water traps was used to estimate the effective light-trap radius. Two light-traps, surrounded by water traps, were separated by up to 25 metres on different evenings. When one light-trap was downwind of another the catch was significantly greater in the downwind trap. However, in crosswind conditions the two catches were similar. Bivariate methods were used to estimate that the effective trap radius was between 10 and 25 metres. A further experiment, using multivariate correlation and regression analysis, found that air temperature, windspeed and moonlight affected light-trap catches. An algorithm is described to enable the calculation of moonlight illuminance. Increased windspeed and increased moonlight illuminance tended to decrease the light-trap catch, however, increased air temperature was associated with an increase in light-trap catch. These effects varied depending on family and species. A final experiment used a remote sensing technique to estimate light-trap efficiency. A 10 cubic metre volume around the light-trap was observed using a video camera. Moth tracks recorded were classified into New Arrivals, Passers By and Local Flights. The number of these tracks was compared to the lighttrap catch so that efficiency estimates could be obtained. Results suggested that at ground level, 125 watt Mercury-Vapour light-traps are only 20 % efficient within this 10 cubic metre volume.

The broad objectives of the work were to develop standard methods for the routine biological surveillance of river water quality, using the non-planktonic algae. Studies on sampling methodology indicated that natural substrata should be sampled directly wherever possible, but for routine purposes, only a semi-quantitative approach was found to be feasible. Artificial substrata were considered to be useful for sample collection in deeper waters, and of three different types tested, Polythene strips were selected for further investigation essentially on grounds of practicality. These were tested in the deeper reaches of a wide range of river types and water qualities: 28 pool sites in 14 different rivers were studied over a period of 9 months. At each site, the assemblages developing on 3 strips following a 4, or less commonly, an 8 week immersion period were analysed quantitatively. Where possible, the natural substrata were also sampled semi-quantitatively at each site. and at a nearby riffle. The results of this survey were very fragmentary: many strips failed to yield useful data, and the results were often difficult to interpret. and of limited value for water quality surveillance purposes. In one river. the Churnet, the natural substrata at 14 riffle sites were sampled semi-quantitatively on 14 occasions at intervals of 4 weeks. In this survey, the results were more readily interpreted in relation to water quality, and no special data processing was found to be necessary or helpful. Further studies carried out on the filamentous green alga Cladophora showed that this alga may have some value as a bioaccumulation indicator for metals, and as a bioassay organism for the assessment of the algal growth promoting potential of natural river waters.

Policy-makers increasingly need to use scientific data that are imprecise. This problem is particularly apparent for marine mammal management issues, where practical research constraints leave scientists and managers with the problem of drawing inference from sparse data. Effective use of such data therefore places great demands on our methods of data analysis and statistical inference. In this thesis I introduce novel Bayesian methods for the analysis of data on marine mammal abundance and trends. Bayesian methods are applied to a suite of case studies to inform current management issues of importance both in the UK and overseas. These include estimating the probability of density dependence in the growth of a killer whale (Orcinus orca) population inhabiting the inshore waters of Washington State; estimating the size of a widespread population of bottlenose dolphins (Tursiops truncatus) in the Bahamas; and assessing the population status and abundance trends of bottlenose dolphins within a newly designated Special Area of Conservation in the Moray Firth, NE Scotland. Each of these case studies uses model-based analysis of individual photo- identification data to make inference about unknown population parameters of interest. Specifically, Bayesian inference, based on "posterior" probability distributions and statements, is used to facilitate scientific reporting in the face of uncertainty about these unknowns. Additional issues addressed are the selection of alternative statistical models for inference based on posterior model probabilities; incorporating model selection uncertainty into inference through the estimation of model-averaged parameter estimates; and the use of random effects prior distributions to model the relatedness between unknown parameters and increase estimate precision. The application of these methods is accomplished through the use of Markov chain Monte Carlo sampling methods, which are implemented using the WinBUGS software.

Regeneration of colonising species following disturbance of primary tropical rain forest in Malaysia was studied in artificial gaps ranging in size up to 30% canopy openness. Effects of gap size on soil temperature and water status were investigated. Spatial variation in density and composition of the soil seed bank beneath undisturbed forest was assessed by germination from soil samples, and seed rain was sampled using seed traps. Seed germination, seedling survival and seedling growth were monitored for two years in gaps and beneath intact canopy. Effects on colonisation of gap size, disturbance to microsite and competition from advance regeneration were studied. Maximum soil temperatures increased with gap size but minimum temperature was unaffected. Soil water content showed no relationship with gap size, but might be higher in gaps than forest after dry periods. Primary forest soil seed bank density was very high (mean 1904 seeds m-2). Climbing shrubs/lianes contributed approximately 75% of seedlings and 50% of taxa. Trees accounted for about 25% composition of taxa, but only 15% of seedlings. Soil seed bank composition differed little among primary forest sites. Similarities in composition between one site and secondary forest possibly reflected recent disturbance. Current seed rain intensity was low, very variable, and not consistently different between gaps and closed forest. Germination in the two years after gap creation was < 5% of the soil seed bank, and appeared greatest in medium-sized gaps. Most germination and accession of new taxa occurred in year one, but both continued at low levels during year two. Ordination analysis suggested that some colonising species regenerated only in large gaps. Both exposure and scarification of the soil increased germination, but had no consistent effects on seedling survival or growth. Competition from advance regeneration was potentially important, increasing mortality and reducing maximum growth rate among colonising seedlings.