Peninsular Malaysia has a diverse flora and fauna, much of which is yet to be documented. The freshwater fishes are one important group that have received little attention. Accordingly, the overarching goal of my study is to investigate the pattern of species richness and analyse the community composition and assemblage structure of fishes in the small streams in Peninsular Malaysia. Small stream habitats appeared to be particularly important repositories of fish biodiversity in this region thus obtaining a reliable census of species occurring in such habitats is critical for conservation and management of biodiversity. Although samplings were far from completed, these habitats support a great variety of species with more than 100 species were recorded from fifty streams sampled in this study. A few are extremely rare with restricted distribution and can thus be considered important in biodiversity conservation of the Peninsular Malaysian ichthyofauna. Human-influenced modification of lowland, headwater stream habitats in Peninsular Malaysia is common and often exemplified by the creation of pools in stretches of rapids and riffles. However, it was not possible to separate pristine and disturbed sites which contained almost identical for species diversity. These findings suggest that local habitat modification does not necessarily cause a decrease in freshwater fish diversity, with only minor negative consequences for other community variables recorded in this study, and therefore raise interesting issues regarding conservation. That said it remains premature to conclude that small stream fishes are insensitive to disturbance and thus their potential utility as bioindicators of disturbance-influenced community changes remain to be confirmed. The maintenance practises being applied to small streams modified for recreational usage were not imposing detectable negative consequences, at least across the sites sampled in this study. The rich diversity of tropical stream environments is the result of both within-habitat (alpha) diversity and between-habitat (beta) diversity. The results showed that there was substantial beta diversity particularly amongst sites that are geographically separated from one another. On the contrary, the lowest beta diversity values were portrayed by contiguous sites. Many fishes exhibited discontinuous patterns of distribution and were considered to be rare while only a handful were widely distributed and abundant. Ordination based on the relative resemblance of fish communities to one another support the existence of two distinct ichthyogeographic divisions in Peninsular Malaysia. It was possible to assign the species recorded to all seven of Rabinowitz's categories of rarity, with at least 10 restricted to a single stream and locally scarce, although not all of these could be described as hyper-endemic. It is recommended that a sizeable augmentation of the existing protected areas is needed to safeguard Malaysia's exceptionally diverse stream-dwelling fauna of which fishes are simply the most well-known inhabitants. Conservation managers should therefore place particular emphasis on small streams since localities in close proximity to one another can exhibit surprisingly high beta diversity, meaning that partial or small-scale habitat protection may prove insufficient.

Fishes are a major component of modern aquatic ecosystems where they regulate some key biological processes such as the proliferation of algae. Their diet is an important and highly variable aspect of their ecology and role in the environment. Yet fossil fishes are often interpreted as ecologically restricted to a single function. Fossil fishes from the order Pycnodontiformes are considered as specialised shell-crushers that drove the evolution of shelled animals in the escalation event known as the Mesozoic Marine Revolution. But this hypothesis is still untested for want of a reliable tool. The manner in which vertebrates wear their teeth provides a direct link between a resilient, easy to fossilise structure and their feeding ecology. It has recently become possible to quantify and compare the roughness of this wear at the microscopic scale. The potential of dental microtexture analysis as a tool to compare and make inferences of dietary preferences in fishes is further investigated here. Here we show that the technique can track subtle differences in diet between morphologically similar animals and correctly assign them to a relevant trophic niche in a taxon, size and environment-independent way. Dental microtexture analysis is applied to teeth from pycnodontiform fishes to test their role in Mesozoic foodwebs. The hypothesis that they were specialised shell-crushers is refuted. A broader range of realised trophic ecologies is highlighted in pycnodonts. They had more generalist or herbivorous habits and potentially were among the first major consumers of macroalgae, a key ecological role filled by extant analogues which allows light to reach zooxanthellate corals on modern reefs. The results of early investigations comparing dental microtexture between different histologies in extant and early gnathostomes are also presented, and show that trophic diversity and shell-crushing are 360-million years old phenomena that need to be considered in scenario of macro-evolution.

Introductions of non-native species can result in the release of their parasites. Although the majority of parasites are lost during the introduction process, those that do get released can spill over to native species and potentially result in pathological, physiological and ecological impacts. Whilst it is increasingly recognised that native parasites can play important ecological roles, the ecological consequences of non-native parasites remain unclear. Consequently, through study of three host-parasite models, this research investigated the ecological consequences of non-native parasites in UK freshwater fish communities through assessment of their effects on hosts (individuals to populations), and on food web structure. The three non-native parasite: host systems were Ergasilus briani and roach Rutilus rutilus and common bream Abramis brama, Bothriocephalus acheilognathi and common carp Cyprinus carpio, and Anguillicoides crassus and the European eel Anguilla anguilla. These parasites were chosen as they reflect a range of life cycle complexity in parasites. The pathology of each parasite was identified using histology, with E. briani having substantial effects on host gill structure, B. acheilognathi impacted the intestinal structure of their hosts, and A. crassus substantially altered the structure and functioning of the host swimbladder. Whilst infections of E. briani and A. crassus had minimal effects on the body size, growth and condition of their hosts, chronic infections of B. acheilognathi did impact the growth and condition of C. carpio when measured over a 12 month period. Differences in the trophic ecology of the infected and uninfected components of the host populations were identified using stable isotope analysis and associated metrics, and revealed considerable differences in the trophic niche breadth of the infected and uninfected fish. In the component infected with E. briani, their trophic niche was constricted, indicating diet specialisation and a shift to feeding on less motile food items. For C. carpio infected with B. acheilognathi, their niche shifted away that of uninfected fish as they fed on higher proportions of planktonic prey resources. Whilst differences in the trophic ecology of infected and uninfected A. anguilla were apparent, this related to differences in their functional morphology that enabled the infected eels to prey upon greater proportions of fish paratenic hosts that resulted in their higher rates of infection. The wider ecological consequences of the introduced parasite were then investigated using topological and weighted food webs. The topological webs revealed that lifecycle and host specificity were important factors in how each parasite impacted the food web metrics, but in all cases the combined effects of including native parasites in food web structure exceeded that of adding the non-native parasite. However, weighting these food webs by using the dietary data outlined above revealed that these infections were predicted to have greater consequences than predicted topologically, and enabled scenarios of differing parasite prevalence and environmental change to be tested on food web metrics. These revealed that under increasing nutrient enrichment, infected individuals generally benefit via having access to greater food resources, a counter-intuitive resulting from increased algal biomass. Thus, this research revealed that introductions of non-native parasites have pathological and ecological consequences for their host populations that have measurable effects at the food web level. These outputs have important implications for the management of non-native parasites and their free-living hosts, and should be incorporated into risk-management and policy frameworks.

Introduced species are a major threat to native biodiversity and, whilst there are socio-economic benefits often gained from introductions, there are also inherent risks to receiving ecosystems, particularly from those species that develop invasions. This is problematic, as once in the wild, their management is inherently difficult. The most effective method to avoid an invasion is preventing the initial introduction; if this is unsuccessful, then methods should aim to minimise the dispersal opportunities of the species into the wider environment. This research used the Asian fish of the Cyprinidae family, topmouth gudgeon Pseudorasbora parva, as the model species, a highly invasive fish in mainland Europe. England and Wales was the model area, where 26 P. parva populations have been recorded but the species is not yet invasive. The research identified how aspects of P. parva ecology can underpin the application of methods to prevent their wider dispersal into the environment to prevent their invasion. It covered work on: (i) the expression of their life history traits; (ii) preventing their introduction into inland waters via fish movement audits; (iii) identifying the risk of their dispersal into river catchments from aquaculture sites; and (iv) assessing how their populations can be controlled to minimise dispersal opportunities. Assessments of their somatic growth rates in relation to temperature and latitude revealed that whilst temperature was a strong regulator of growth in controlled conditions, in the wild, population density was important, with this having a stronger influence on growth rates than latitude (temperature). Individuals in recently founded populations and at low densities had very fast growth rates, with this advantageous in facilitating their establishment, with these growth rates slowing as population abundance increased. Assessments of their reproductive traits revealed that, contrary to European and Asian populations, they did not show an extended reproductive period involving fractional spawning, but instead had a discrete reproductive season that lasted no longer than three weeks in the early summer period. Approximately 8000 fish movements (i.e. removals from fisheries, stocking into fisheries) are legally consented by the Environment Agency in England each year. On each of these stockings, there is a concomitant risk of also accidentally introducing a ‘hitch-hiking’ species that has not been detected in the batch of fish. To test whether a ‘fish movement audit’ scheme could reduce this risk, an experiment was completed on batches of native fish containing known proportions of P. parva and using auditors of varying non-native fish knowledge and pre-defined searching effort levels. Outputs revealed that although the scheme could be successful at preventing 95 % of accidental introductions, this required highly experienced auditors using high searching efforts. The presence of P. parva on an aquaculture site in Southern England enabled completion of field studies on assessing their drifting rates from the farm and into an adjacent stream. Drift rates were negligible during normal summer operations. However, when aquaculture ponds on the site were drained down for husbandry purposes, the elevated flow of water displaced high numbers of mature P. parva, resulting in high dispersal rates. Whilst no P. parva were recorded in the three hours prior to ponds being drained down, over 200 were captured in drift nets in the 21 hours after draining commenced. Thus, whilst the overall risk of their dispersal was low, this was elevated considerably during specific on-site activities, and would require pro-active management to prevent accidental dispersal during these activities. A long-term field experiment, allied with a field study, revealed that in small fishing and aquaculture ponds, P. parva population densities can be maintained at very low abundances over long time periods, but required a combination of biocontrol (via manipulating of numbers of native piscivorous fish) and targeted fish removals (by trapping in spring and autumn). Fish removals alone were unable to reduce population sizes due to the compensatory responses of the remaining fish. These responses were not, however, detected when native piscivorous fish were present, probably through their predation of nest-guarding male P. parva that then inhibited large numbers of in young-of-the-year being produced. In summary, the research revealed utility of non-native fish management methods being applied to populations following their testing in more controlled conditions. It revealed that introductions into fisheries of fast growing, small-bodied, non-native fishes can be prevented by auditing, and where this is unsuccessful then their dispersal into the wider environment can be minimised, if not eliminated, by a variety of method that should reduce their invasion risk.

The present study investigated the fish populations of the lower Forth Estuary, east Scotland. Cooling water extraction by the 2400 MW Longannet Power Station (LPS) inevitably removes a certain quantity of fish from the estuary, all of which experience mortality. The present study employed a sampling regime of greater intensity than previous studies to investigate the extent of mortalities from January 1999 - December 2000. Collections of fish impinged on intake screens were made eight times monthly, at LW or HW of spring or neap tides during the day or by night. Marine species dominated the assemblage of fish collected, with sprat, herring, and whiting contributing > 80% of total abundance. Sprat was twice as abundant as herring in 1999, while the proportions were very similar in 2000. Total abundance of all species collected in 1999 was estimated at 1. 09 x 107, while the value of 3.29 x 107 in 2000 was three times larger. These figures were the largest recorded among British estuarine and marine power stations, but were precisely the correct order based on an exponential relationship between total impingement and water abstraction rate established from data from other locations. Validation of the estimated total biomass of fish removed was given by comparison with the known total mass of all materials disposed to landfill. Statistical analysis of impingement data showed that tidal range and season were the most important environmental variables influencing the rate of removal of fish from the estuary. That light was not significant for most species is attributed to high levels of turbidity and the resulting low visibility by day and night. Demersal and benthic fish abundances collected from 1982 - 2000 in 30 annual trawls at three sites in the mid-lower Forth Estuary were analysed. Species tended to be present in greatest abundance at the most seaward of the sites. Patterns of seasonal abundance reflected those observed in the impingement study at LPS, and catches tended to be greatest at L W. Total species richness showed no significant trend over time, whilst total annual abundance of fish captured in trawls showed a significant negative trend. This was largely due to significant declines in the two most abundant species, namely whiting and eel pout, attributable in the latter case to increasing temperatures. Changes in the ichthyofaunal composition were largely driven by whiting, eel pout, cod and plaice. Eight of ten common species showed no significant trend in abundance over the length of the time series, suggesting them to perhaps be at equilibrium densities. Quantities of commercially fished species above minimum landing size limits that were removed by LPS were very low, and restricted to herring and occasional whiting. The quantity of juveniles that could have recruited into the fished populations was expressed as equivalent adults. The values were larger than any previously reported in the UK, primarily due to the quantities of juvenile fish impinged being greater than at any other British power station, and the importance of the Forth as a nursery area for marine species. It was concluded that LPS is the dominant UK power station in terms of magnitude of impingement losses. It may be prudent to consider a precautionary approach to mitigate losses, and to this end options for reduction of the magnitude of impingement are discussed.