Non-native species invasions threaten the structure, function and biodiversity of ecosystems worldwide, and those of non-native fishes pose amongst the greatest threats to inland waters of the U.K. This PhD investigated the establishment, dispersal and ecological implications of introducing the two non-native fish species, sunbleak Leucaspius delineatus (Heckel, 1843) and topmouth gudgeon Pseudorasbora parva (Temminck and Schlegel, 1842) to inland waters of England. The introduction and initial dispersal of both species can be attributed to the commercial fish trade. Species-specific variability of life history, growth and morphological traits was examined in sunbleak (12 sites) and topmouth gudgeon (3 sites) to assess their role in establishment success. The drift dynamics, i.e. timing and intensity (propagule pressure), of sunbleak and topmouth gudgeon were assessed for source populations to determine dispersal potential. Potential risks for native species posed by these two alien cyprinids were assessed with respect to the parasite fauna and overlaps in resource use. For sunbleak, these were also examined in terms of social integration of this species into a native fish assemblage. Biological resistance to topmouth gudgeon invasion was evaluated by stomach flushing and gut content analysis of native piscivorous fishes. Inter-population variability in life histories and morphological characters were observed in both sunbleak and topmouth gudgeon. Populations of both species matured at small body sizes and between the ages 1 and 2. The fish were of good body condition and exhibited high reproductive investment. In both species, dispersal from source waters followed a diel pattern, with higher rates at night than during the day (e.g. maximum drift densities during May of 2004 and 2005: 9-10 sunbleak per 1000 m -3 at about 23:00 hrs; 40-52 topmouth gudgeon per 1000 m -3 at about 05:00 hrs). Downstream of one source population, microhabitat use of topmouth gudgeon was found to overlap with native species (brown trout Salmo trutta L., European chub Leuciscus cephalus (L.), bullhead Cottus gobio L., stoneloach Barbatula barbatula (L.); both brown trout and chub were observed to prey on topmouth gudgeon. However, predation intensity may be density-dependent and of insufficient level to impede topmouth gudgeon establishment, which was facilitated in the receiving stream by the consistent propagule pressure from on-line source populations. Sunbleak diet and microhabitat use also overlapped with native species (roach Rutilus rutilus (L.) and common bream Abramis brama (L.)) as young larvae, but this decreased with age. Social network analysis of sunbleak-native species interactions revealed that sunbleak creates significantly stronger social bonds with the native species than do natives amongst themselves. No macro-parasites were found in topmouth gudgeon, but two ‘Category II’ non-native parasites Neoergasilus japonicus (Harada, 1930) and Ergasilus briani (Markewitsch, 1932) were found in some populations of sunbleak. The potential for sunbleak to spread beyond their current distribution in England and the species’ social integration behaviour may facilitate the dispersal of these parasites, which may spread faster among communities invaded by sunbleak than in those where this non-native species is absent. The results of this PhD study are discussed within their wider context and their relevance to non-native species risk analysis and management.

Diese Dissertation ist ein Beitrag zum Forschungsfeld der Stressökologie, im Spe-ziellen der Verhaltensökotoxikologie. Das spontane lokomotorische Verhalten der Fischarten Danio rerio und Leucaspius delineatus wurde unter sublethaler Expo-sition mit dem Cyanobakterientoxin Microcystin-LR (MC-LR) und dem Xenobio-tikums 2.4.4`-Trichlorobiphenyl (PCB 28) quantifiziert. Die Schwimmgeschwin-digkeit und Anzahl der Wendungen wurden kontinuierlich mit einem automati-schen Video-Monitoringsystem unter Laborbedingungen aufgezeichnet. In Hin-blick auf zyklische Aspekte wurden die Verhaltensanalysen mit chronobiologi-schen Methoden kombiniert. Hiermit wurde gezeigt, dass MC-LR und PCB 28 zu signifikanten Effekten in Verhalten und Aktivitätsrhythmik beider Fischarten führten. Höhere Konzentrati-onen beider Untersuchungssubstanzen verursachten eine deutliche Aktivitätsredu-zierung bei Danio rerio und Leucaspius delineatus. Einige der festgestellten Do-sis-Wirkungsbeziehungen entsprechen der Hormesistheorie, z. B. war bei geringe-ren MC-LR Konzentrationen ein Aktivitätsanstieg und bei höheren ein Aktivitäts-abfall beider Fischarten zu verzeichnen. Die Exposition mit MC-LR und PCB 28 verringerte bei beiden Testfischarten die Synchronisation der Aktivität mit dem Zeitgeber Licht. Dies führte bei beiden Fischarten zu einer Phasenverschiebung. Bei Leucaspius delineatus war unter dem Einfluss von MC-LR eine Phasenumkehr zu verzeichnen, die Fische wechselten von Tag- zu Nachtaktivität. Die Cosinor Analyse zeigte Dosis abhängige Veränderungen der circadianen Rhythmen der Schwimmaktivität (z.B. MESOR, Akrophase) unter Einfluss von MC-LR und PCB 28 an. Die Power Spektral Analyse indizierte für beide Fischar-ten unter Einwirkung von MC-LR and PCB 28 eine reduzierte Dominanz des cir-cadianen Rhythmuspeaks. Da die registrierten Unterschiede in der Reaktion beider Fischarten auf MC-LR und PCB 28 eher gering waren, sind Ergebnisse der Art Danio rerio, die häufig in Toxizitätstests verwendet wird, mit denen der einheimischen Art Leucaspius deli-neatus vergleichbar. Die Ergebnisse belegen, dass Verhaltensuntersuchungen in Kombination mit chronobiologischen Auswertemethoden eine sensitive und zuverlässige Abschät-zung des Gefährdungspotentials von Substanzen sowohl auf dem Gebiet der Öko-toxikologie als auch für Biomonitoring ermöglichen. / This thesis contributes to the field of stress ecology specifically behavioural ecotoxicology. The spontaneous locomotor behavior of two fish species Danio re-rio and Leucaspius delineatus was recorded and quantified continuously under sublethal exposure to the cyanotoxin microcystin-LR (MC-LR) and the xenobioti-cum 2.4.4`-trichlorobiphenyl (PCB 28). By using an automated video-monitoring system, the swimming velocity and the number of turns were assessed under labo-ratory conditions. For analysing cyclic aspects basic behavioural analyses were combined with chronobiological procedures as cosinor analysis and power spec-tral analysis. Both MC-LR and PCB 28 acted as stressors and caused significant changes in the behaviour and circadian activity rhythms of Danio rerio as well as Leucaspius de-lineatus. So for both species elevated concentrations of the stressors led to a re-duction of their activity. Some dose-responses correspond to the hormesis theory, e.g., there was an increase of daytime activity at lower MC-LR concentrations and a decrease at elevated concentrations of MC-LR for both species. A degree of desynchronisation of activity to the zeitgeber light, which led to a phase shift was caused by the chemicals in both fish species. In Leucaspius de-lineatus this shift was so drastic that this species reversed their significant diurnal activity and became nocturnal under the influence of MC-LR. The Cosinor analysis revealed MC-LR- and PCB 28-induced, dose-dependent al-terations of the circadian rhythms of activity (e.g., MESOR, acrophase). The power spectral analysis showed that the dominance of the circadian rhythmic peak (of 24 h) was reduced under MC-LR and PCB 28 for both species. Since the observed differences in the reactions of both species to MC-LR and PCB 28 were rather small, the results of the species Danio rerio which is widely used for environmental risk assessment tests, are comparable to those of the native European species Leucaspius delineatus. The findings of this study proved that the basic behavioural analyses combined with chronobiological procedures could be valuable tools for the study of stressful or even harmful environmental factors in the field of ecotoxicology as well as for biomonitoring.