Deep-sea demersal fish are likely to be highly important in structuring deep-sea ecosystems, but a paucity of data means that relatively little is known about the spatial and temporal processes that influence their distributions or how these may change at different scales of observation. As human activities continue to expand into deeper waters, the importance of understanding these processes is becoming increasingly urgent. The oil and gas industry are expanding into deeper waters as coastal oil reserves diminish, but potential long-term effects on the benthos are unknown. Time-lapse photographic data collected from within an active oil field on the Angolan continental slope (tropical eastern Atlantic) detected no significant differences in the community composition of fish compared to a reference site. Significant seasonal changes in the total abundance of demersal fish were detected however, with >20-fold increases in abundance recorded from the oil-field observatory. These results suggest that the deep-sea environment in this region is likely to be highly dynamic over seasonal and decadal timescales, and some possible drivers are discussed. Understanding the spatial distributions of fish is important in understanding their potential ecological roles within an ecosystem and how they may benefit from spatially-explicit management measures. In the bathyal NE Atlantic, the demersal fish communities associated with cold-water corals (CWC) appeared to be influenced by processes operating at multiple spatial scales. At the broadest scales (100s km), depth was a significant predictor of community composition, while habitat type was significant at the finest scales (m). These results highlight the need to account for the effects of scale in observational research and may explain why no consensus has thus far been reached regarding the role of CWC habitats for deep-water fish in the NE Atlantic and provide a possible framework for approaching future deep-water community studies. In abyssal depths, the importance of habitat heterogeneity in structuring fish communities has never been previously studied. Investigations of the spatial distributions of demersal abyssal fish around a small (c. 250 m high) abyssal hill showed that the distribution pattern of the total fish fauna and the two dominant taxa were not significantly different from random. Random distributions are unusual in nature and these results suggest that the environment may be essentially homogeneous to abyssal fish at spatial scales between 100 m – 10 km and suggests no influence of the abyssal hill on the fish fauna. The results highlight the great potential value of autonomous vehicles in extending abyssal research over broader spatial scales. Experimental studies investigating the effects of large-scale ecological processes on deep-water ecosystems are often unfeasible. Mathematical models can provide an alternative methodology, but have not been widely applied to the deep sea. In Chapter 5, a simple mathematical model was developed to explore the effects of large, transient resource pulses on the population dynamics of the abyssal holothurian Amperima rosea. Large, apparently transient increases in the biomass of this species have been observed at the Porcupine Abyssal Plain during two “Amperima events”, but the mechanisms leading to the increases remain unclear. The simulations showed that Amperima biomasses within the observed range of the smaller “Amperima event” could be simulated in some circumstances following the addition of a single, large resource pulse (10000 – 20000 g POC ha-1) to the model, which led to a short-term, transient increase in Amperima biomass before a return to previous levels. None of the simulations produced unstable population dynamics. While the scarcity of empirical data from the PAP means that these results must be treated cautiously, they highlight the potential for temporal changes in food supply to rapidly alter the dynamics of abyssal populations.

No description available.

The life history of a population of Lutraria lutraria in a depth of 7m at Hunterston, Ayrshire is discussed. Much of the present population Is thought to have settled in 1967. The functional morphology of Lutraria is described and related to its life as a large, deep-burrowing bivalve. Lutraria spawned in late spring and continued to do so through the summer in 1979 and 1980. Animals became spent in August and September. Unsuccessful attempts were made to induce spawning in the laboratory. Artificial fertilization was successful but development did not proceed beyond the ciliated gastrula stage. Larvae of Lutraria were not identified in plankton samples and young stages were not encountered in sieved sediment samples. The biochemical cycle of the total animal and five component parts (gonad and visceral mass, digestive gland, adductor muscle, siphon and 'other' tissue) is investigated. A marked increase in weight, reflected in an increase in weight of the component parts, was recorded in Autumn 1979. This is thought to be related to an exceptional increase in the phytoplankton at this time. Although a relationship between the biochemical cycle and reproductive cycle remains uncertain, definite seasonal changes were recorded in the respiration rate of Lutraria. At 10°C, the maximum rate of a standard 20g animal was 0.1283m1s 02/g. dry wt./hr. in May 1980 and the minimum rate was 0.O59mls 02/g. dry wt./hr. in October 1980. The effect of temperature on respiration rate was also investigated. Significant differences were recorded for five experimental temperatures (10°C, 15°C, 20°C, 25°C and 30 °C) in August and October but only between two temperatures (10 C and 30 C) in April. There was a decrease in respiration rate at 30 C in August and October, but an increase in April. Respiration rate is affected by a reduction in oxygen tension. A variety of responses were recorded with a small degree of regulation shown. Individuals of Lutraria were able to survive 48 hours under anaerobic conditions. In fully oxygenated conditions heart rate ranged from 4-15 beats per minute with an average of 8 beats per minute. Heart beat was markedly affected by changes in temperature and oxygen tension, increasing to a maximum 22 beats per minute at 25 C, and decreasing to a minimum 2 beats per minute in anaerobic conditions. Heart rate is reduced (12 beats per minute to 5 beats per minute) on exposure to air. Lutraria exhibits an intermittent pattern of pumping activity. Under normal conditions 35% of the time is spent pumping and this Increases as oxygen is reduced (3.00mls 02/litre) to 65% of the time spent pumping. 15. Under normal conditions the respiratory flow varies between 0.382 litres per hour and 1.023 litres per hxir. Adult Lutraria maintain their ability to burrow, albeit slowly.

Biological indicator species can reveal consequences of changes in physical processes within the environment, through effects on their physiology, behaviour and population dynamics. Long-lived species tend to be positioned at the top of the food chain where they can act as indicators of environmental change occurring at lower trophic levels. During poor conditions, these long-lived top predators have been selected to prioritise their own survival above the current breeding attempt, in order to maximise lifetime reproductive success. Endocrine mechanisms involving corticosterone, the ‘stress hormone’, and possibly prolactin, the ‘parental hormone’, are involved in mediating the abandonment of breeding in response to environmental perturbations. This thesis aimed to assess what the breeding success of a top marine predator indicates about changes in the marine ecosystem and what mechanisms control changes in breeding success, using the black-legged kittiwake Rissa tridactyla as the model species. I combined population-level analyses of long-term datasets (1997–2010) of diet composition, adult body mass, breeding success and foraging behaviour from the Isle of May, National Nature Reserve, Firth of Forth, south-east Scotland (56° 11‘ N, 02° 33’ W) with an individual-level field experiment to simulate chronic stress. Kittiwakes breeding in the north-western North Sea depend primarily on adult (1+ group) lesser sandeels Ammodytes marinus at the start of the breeding season and subsequently switch to depend primarily on young of the year (0 group) sandeels. Analysis of the long-term data showed that the timing of the kittiwake breeding season has become later in recent years, whilst the timing of the switch from 1+ group to 0 group sandeels in the kittiwake diet has become earlier, which may suggest mismatches in the timing of prey availability and predator demand. Increasing proportions of clupeids (mainly sprat Sprattus sprattus) were seen in the diet and further years of study may reveal whether clupeids could be a beneficial alternative prey type for kittiwakes. Foraging trip duration was unrelated to diet composition, suggesting that the main prey types of kittiwakes do not differ in their distance from the colony. Whilst foraging trip duration during incubation was related to changes in adult body mass and hatching success, diet composition was unrelated. There was a weak effect of diet composition during chick-rearing on fledging success, mediated via changes in adult body mass. However, this effect was masked by a stronger, independent, negative effect of foraging trip duration during chick-rearing. To simulate chronic stress in kittiwakes, individuals were implanted with corticosterone, using Alzet® osmotic pumps, for a week at the end of incubation. The methodology applied to kittiwakes was based on a preliminary experiment carried out in Japanese quail Coturnix coturnix japonica. The body mass and prolactin concentrations of kittiwakes were unchanged after this treatment. Corticosterone concentrations had returned to pre-treatment values by the end of the treatment week, which may have been due to down-regulation or suppression of the stress response as a result of the treatment. Corticosterone-implanted males showed lower nest attendance than sham-implanted males but the opposite was true for females. Breeding success at the end of the season was lower in corticosterone-implanted birds, suggesting a prolonged effect of chronic stress. In order to investigate the effects of disturbance to a group or colony of birds prior to the capture of an individual, a preliminary experiment was also carried out to test the stress responsiveness of a captive bird, the Japanese quail. No increase in corticosterone concentrations was seen after a capture-restraint protocol and with increasing time since the group of birds was first disturbed. A suppressed stress response in this bird may be explained by long-term captivity or domestication. These results show that the breeding success of a top marine predator can indicate changes in the timing of prey availability and prey location, mediated through changes in adult body mass. I also found that changes occurring during the chick-rearing period contributed most to the outcome of the breeding season. Chronically elevated concentrations of corticosterone are important in the control of breeding success, whereas prolactin may only play a role close to the timing of breeding failure or after failure has occurred. This thesis demonstrates the need for continued long-term monitoring of wild populations and refining of experimental methodology to better understand the impacts of environmental change on top predators.

It is widely recognised that both nutrient-driven processes acting from the “bottom-up” and predator-driven processes acting from the “top-down” are important drivers of population change. However, studies that examine how these joint forces influence the population dynamics of oceanic species are lacking. In the Antarctic and Sub-Antarctic ecosystem, human-mediated changes have driven biological change at both ends of the food chain; rapid regional warming at the bottom and heavy exploitation of apex predator populations at the top. Consequently, many populations of marine predators have rapidly changed in size over the last 50-years. Unravelling the effects of bottom-up and top-down forcing on these open ocean ecosystems, has thus been highlighted as an immediate priority for polar scientists. The overall aims of this study were to use demographic, environmental and diet data to unravel the processes that contributed to a population of macaroni penguins at South Georgia declining rapidly between 1985 and 2012. I use mark–recapture modelling to examine the survival rates of macaroni penguins. Over 10 years, birds were marked with subcutaneous electronic transponder tags and re-encountered using an automated gateway system fitted at the entrance to the colony. These findings were combined with a 28-year time series of population counts and productivity measurements in an age-structured state-space population model to disentangle the processes underlying the observed population decline. Finally, I combined stable isotope analysis and tracking data to investigate the individual strategies macaroni penguins might employ to mitigate the effects of density-dependence during the breeding season. Macaroni penguins at South Georgia declined at 6% per year between 1985 and 2000, stabilising thereafter. This study indicates that the population declined in response to recruitment rates being lower than adult mortality. This trend was potentially accelerated by three large mortality events that were possibly associated with top-down predation pressure from giant petrels. Survival rates were low and variable during the fledging year, increasing to much higher levels from age 1 onwards. Year-to-year variability in demographic rates was induced by a combination of individual quality, top-down predation pressure and bottom-up environmental forces. The relative importance of these covariates on survival rates was age-specific, whereby predation pressure had a considerably greater effect during the fledgling year compared with birds older than 1-year. The population trajectory stabilised after 2000 in response to an increase in survival, as well as density-dependent feedbacks upon productivity. In order to minimise the effects of density-dependence during the breeding season and optimise daily energy expenditure, macaroni penguins appeared to make distinct dietary choices that remained highly consistent from year-to-year. Individually specialised foraging strategies occurred in response to seasonal variations in foraging range, conspecific density and prey availability at the foraging sites. The future stability of this population will depend on the carrying capacity of the environment supporting productivity rates at their present level, and the population size and breeding success of giant petrels not increasing so that adult survival rates can remain stable. More broadly, this study highlights the importance of considering multiple causal effects across different life-stages when examining the demography of seabirds, and demonstrates the additional insights that can be gained by using models with increased precision.

1. Climate change is predicted to negatively impact ecosystems globally. Freshwaters are particularly susceptible, due to thermal and hydrological sensitivity and poikilothermic fauna. Such change may ultimately imperil ecosystems service provision. Adaptation schemes that limit ecological and economic impacts are therefore widely advocated. The major management action recommended for rivers is restoration of “buffer strips” of riparian tree cover, which can lower stream water temperatures, and should therefore protect thermally sensitive species, including economically important salmonid fish (Family: Salmonidae). 2. Despite thermal consequences being well-quantified, community-wide biotic responses to restoration remain unexplored. Such ecological consequences may ultimately govern the efficacy of this measure in protecting focal species, but there is limited evidence available from which to predict restoration outcomes and inform management efforts. 3. Using 24 upland streams, this thesis assessed ecological consequences of differential catchment tree cover on stream communities. Using surveys and experimentation, the project assessed the role catchment afforestation in mediating the importance of terrestrial resource subsidies to stream food webs, structuring macroinvertebrate communities, and supporting salmonid fish. 4. Several lines of evidence indicated that wider catchment tree cover controls the functional composition of stream communities, by mediating availability of terrestrial resources. However, buffer strips did not approximate the effects of wider broadleaf forest, and supported communities that were functionally indistinguishable from streams draining unafforested moorland. Unexpectedly, catchment broadleaf cover did not ultimately affect salmonid populations, contrary to expectations. 5. In combination, these results contribute to the understanding of land use effects on stream communities, and provide clear evidence needed to guide future management efforts. These findings suggest two distinct adaptation options: restoration of buffer strips should be able to confer benefits to fish populations, by offsetting thermal impacts without jeopardising food availability. However, re-instatement of wider broadleaf forest could potentially re-establish woodland stream communities.

Due to mounting evidence that the early environment experienced by a juvenile can affect the morphology and physiology of the adult, there is currently great interest in how environmental variability may shape the behavioural phenotype, and whether such shaping has adaptive benefits. It is clear for example that the developmental environment will have immediate effects on an animal in terms of its survival and performance. Individuals with access to little food or exposed to high levels of predation will have lower survival, and resource-poor surroundings may mean that a young individual is unable to forage successfully, or disperse as normal. However, there is now increasing evidence to suggest that early environmental conditions are also important in determining the success of the adult, meaning that experiences during early development can have significant long-term effects. In this thesis, I consider the effects of diet quality and exposure to stress in postnatal life on behavioural traits in adulthood, using the zebra finch (Taeniopygia guttata) as a model species. There is already evidence to suggest that the early environment can shape behaviour, and consequently, many aspects of behaviour may have important developmental origins that are not a direct result of the genotype. Two populations of birds were used throughout this collection of studies. The first was raised under different diet regimes, in which both diet quality and consistency were manipulated during the first weeks of life. The second did not undergo any diet manipulation, but were given oral doses of the stress hormone corticosterone or a peanut oil control in the nestling phase. Both populations then underwent a variety of behavioural tests to determine the effects of their early life experiences. In examining the behaviour of zebra finches, I found that these birds displayed distinct and repeatable behavioural traits, and discuss the idea that they may even exhibit behavioural syndromes, since I was able to show strong correlations between various responses. I studied the effect of both exposure to varying diet quality and consistency in early life on various behavioural traits in adulthood, and found evidence to suggest that certain behavioural responses are linked to the consistency of the nutritional environment experienced by the chick. When tested in adulthood, birds that had experienced a consistent early environment (regardless of food quality) showed a strong trend towards being bolder than birds raised on a variable diet, and this result was replicated in a subset of the original population over a year later, suggesting real longitudinal effects. Such an outcome may suggest that certain behavioural traits are developed in response to environmental sampling, so as to maximise fitness in the anticipated environment. Similarly, I investigated the behaviour of birds exposed to corticosterone in early life, but found no evidence to suggest that their behavioural traits had been directly shaped by exposure to this hormone. Interestingly however, relative growth rate was linked to individual boldness, a finding which could again possibly be explained through adaptive environmental shaping, in which growth rate is used as a proxy for environmental richness. Using subsets of the populations of birds that were examined for personality traits, I also considered the effects of the early environment on adult performance in two simple memory tests; specifically studying the ability of birds to search for and remember the location of a food item using environmental cues. In both the diet and hormone manipulated groups, I was able to show that those individuals that had experienced a sub-optimal early environment were compromised in their performance on these simple learning and memory tasks. Again, these were long term effects, since the tests were carried out many months after the birds became adult, suggesting they could not be compensated for, and consequently, since spatial and visual memory are important throughout life, could not be said to be adaptive under any circumstances. Finally, I also looked for effects of the diet and hormone treatments on the ability of the birds to become dominant. In tests using the birds raised on the different diet regimes, results again suggested that diet consistency was important, as birds raised on a consistent diet were generally more aggressive than those raised on a variable diet, and also showed a trend towards winning more of their individual interactions. Consequently, in this study, nutritional stability made individuals generally more successful in conspecific interactions and thus more likely to hold a higher dominance position, though this is complicated by the fact that dominance was not linked to priority of access to other desirable resources such as potential mates. Using the same experimental design on the birds raised under the hormone treatment, there were trends to suggest that exposure to corticosterone could have negative effects on dominance-related behaviour under certain environmental conditions, but sample sizes were too small to allow firm conclusions. Overall, this study provides strong evidence for the idea that behaviour can be modified by the environmental conditions an individual experiences in early life. Since behavioural traits will impinge significantly on individual fitness, such effects are of general interest. Since there are also studies that suggest that environmental shaping can be adaptive, I discuss the costs and benefits involved in the different behavioural phenotypes that arise as a result of environmental variation.

Nutrient enrichment of tropical waters constitutes an increasing threat to the health and biodiversity of coral reefs. In order to manage these ecosystems effectively, the onset of nutrient pollution has to be closely monitored. This thesis examined the possibility of using some physiological responses of hermatypic corals as an early-warning bio-assay, to detect nutrient enrichment before reef deterioration has taken place. To this aim, the physiology of the common branching coral Porites porites and the massive coral Montastrea annularis was studied both in the laboratory and on the reef under different nutrient conditions. By measuring the organic and inorganic productivity of corals and by constructing carbon budgets, it was hoped to relate differences in the fixation, allocation and utilisation of carbon to differences in nutrient regimes. Nubbins of Porites porites and explants of Montastrea annularis were chosen as the experimental units. Nubbins were obtained by cutting coral tips (approx. 20 mm), grounding their cut surface flat, and gluing them onto a perspex tile with cyanoacrylate glue. To obtain explants, a coral head was cored under a drill press fitted with a hole saw. Cores were then cut to fit, and sealed into polyethylene cups with underwater epoxy putty. A new culturing system was developed to grow corals successfully in the laboratory under completely controlled and repeatable conditions. This system (the 'photostat') consisted of glass aquaria (30x21x18 cm) placed in a constant temperature water-bath under metal halide lamps. The aquaria were fitted with specially designed air lines and coral trays to maintain a strong water motion around the corals, independent of the rate of water-flow. A peristaltic pump ensured a daily water turn-over. A new improved carbon budget methodology was developed by comparing the well established methods of Davies (1984) and Muscatine et al (1984) on Porites porites. These methodologies differed in the measurement of zooxanthellae respiration rate (Rz) and zoozanthellae growth rate (). Rz,DAVIES was found to be twice as small as Rz,MUSCATINE (RZ, MUSCATINE (RZ, DAVIES = 18.1 gC cm-2d-1 vs. Rz,MUSCATINE = 33.1 gC cm-2d-1), but this accounted for a difference of only 3% when Rz was expressed as a percentage of the total daily carbon input. By comparison, a 25-fold difference between methods occurred in the component of carbon required for the daily growth of the zooxanthellae. Davies' method measured the net rate of zooxanthellae growth (NET) from the increase in surface area, assuming a constant zooxanthellae population density. In this case NET was only 1.65 gC cm-2d-1.

Energy is the fundamental currency of life that drives organismal growth and development. Energy requirements vary greatly between species but also within species due to differences in physiology, behaviour and life history. The consequence of this variation is of great interest to ecologists, as it is potentially a trait upon which natural selection can act. One of the main components of an organism’s energy budget is its baseline level of metabolism, hereafter termed its standard metabolic rate (SMR). It has been shown in several species of salmonid fish that a high standard metabolic rate correlates with dominance, aggression and boldness. This competitive advantage has been shown to result in higher growth over conspecifics in simple lab environments, but the ecological consequences are less clear. This thesis examined the performance of contrasting metabolic strategies across a range of environmental conditions to ascertain the ecological consequences of SMR variation. Experiments also investigated the relationships between SMR, food intake and absorption efficiency to help relate energetic strategies to performance. The effects of environment on the outcome of different energetic strategies were profound. Higher population densities increased intraspecific competition for preferable feeding territories, but fish with a higher SMR tended to be the best competitors and so were most likely to get a preferred territory (Chapter 2). However, for a given quality of feeding territory, whether relatively good or poor, lower SMR individuals grew best due to their lower energy requirements. The benefit to high SMR fish of being able to secure better territories was diminished under less predictable feeding conditions, and disappeared under a structurally complex habitat, resulting in these fish having no performance advantage over fish with a lower SMR (Chapter 3). These high SMR individuals performed poorly in the presence of low densities of a heterospecific competitor, being subject to a disproportionate proportion of the aggression from a more dominant species (brown trout, Chapter 4). At higher densities of trout, intraspecific interactions appeared much more important for both species, resulting in the salmon with the highest SMR exhibiting the fastest growth. These three chapters demonstrate that environmental conditions, both abiotic and biotic, have great consequences for the success of different energetic strategies. The consequences of metabolic strategy on physiology proved just as interesting. High SMR individuals expended more energy when digesting a given size of meal but reduced the duration of this specific dynamic action (SDA, the rise in metabolism associated with processing and digesting a meal) response (Chapter 5). This suggested that their digestion was more rapid than that of low SMR fish, but this did not lead to a higher rate of food consumption (Chapter 5) nor did they sacrifice absorption efficiency (Chapter 6). This thesis demonstrates that the performance of fish with alternative energetic strategies is dependent on the prevailing environmental conditions, which helps explain the persistence of variation in SMR within populations.

Morphometric analysis of Loch Rannoch charr indicated that in addition to the pelagic morph previously described, the benthic morph previously described consists of two distinct morphs. Thus one form had a relatively small head, short gill rakers and consumed only benthic invertebrates, whilst the other had a relatively large head and consumed benthic invertebrates until a size of 16cm at sizes above which it consumed other fish. These two forms were termed the benthivorous and piscivorous morphs respectively. Morphometric and dietary analysis agreed with the previous description of the third form as a pelagic planktivore which had fine head and jaw features, long gill rakers and a relatively streamlined morphology and differed from the other two morphs in that it was highly coloured as opposed to being cryptically coloured. This form was termed the planktivorous morph. Each morph was reproductively isolated by each spawning in a discrete location. Thus, the benthivorous morph spawned in the mouth of the largest afferent river, the River Gaur which enters the west basin of the loch, the piscivorous morph spawned in sublittoral regions of the east basin of the loch and the planktivorous morph spawned in shallow littoral regions throughout the loch but mainly in the east basin. Outwith the spawning period benthivorous and piscivorous charr were restricted to the west and east basins respectively, indicating that each is adapted to the biotic conditions within each basin and that the basins have effectively isolated these two morphs. Morphometric analysis of the charr of Loch Ericht, a similarly sized lake within the same catchement as Loch Rannoch, demonstrated the existence of two distinct morphs. One morph exhibited cryptic colouration and had a large head and was found to consume benthic invertebrates until a size of 16cm at sizes above which it consumed other fish. The other morph was highly coloured, had a relatively small head and streamlined body and consumed predominately zooplankton.