Interest in the subject of fish welfare is continuing to grow, with increasing public awareness and new legislation in the UK. Water quality has long been recognised as being of prime importance for welfare: water provides the fish with oxygen and removes and dilutes potentially toxic waste metabolites. This thesis investigates the interactions between water quality and the welfare of farmed rainbow trout (Oncorhynchus mykiss Walbaum). A literature review was undertaken to identify current recommended water quality limits for the health and welfare of farmed rainbow trout. Contradictions in the literature regarding suggested ‘safe’ water quality limits were also identified, as were deficiencies in some of the methods used to arrive at conclusions for recommended limits. The literature relating to the effects of poor water quality on welfare were also reviewed. The review ends with a discussion about water quality monitoring in the context of on-farm welfare assessment and how the information might be used in such a scheme. A telephone survey of UK rainbow trout farmers was undertaken to ascertain the level of water quality monitoring currently conducted. Participants in this study accounted for over 80% of 2005 UK rainbow trout production. It was established that 54% of farmers monitored dissolved oxygen to some extent and 69% monitored temperature, the most commonly measured water quality parameters and among the most important for health, welfare and growth. Subsequent visits were made to a sample of the participants in the telephone survey to obtain more detailed information of the farming operations, such as frequency of water quality monitoring, retention of production data and slaughter methods. Monitoring water quality will be an integral part of any on-farm welfare assessment scheme, and while measuring some water quality parameters requires specialist equipment, farmers should be able to monitor the essential parameters, dissolved oxygen and temperature. Any on-farm welfare assessment scheme for rainbow trout should incorparate fish-based measures in addition to resource-based parameters in order to provide as complete an overview of trout welfare as possible. An epidemiological study was undertaken to investigate the current status of welfare on UK rainbow trout farms and to identify risk factors for welfare. Forty-four trout farms from throughout the British Isles were visited between July 2005 and April 2007, sampling a total of 3700 fish from 189 different systems. Farms were visited twice, once in winter and once in summer, to account for any seasonal differences in fish physiology and environmental conditions. Data were collected on a range of fish parameters, together with background information on the batch from which the fish originated. Particular emphasis was placed on water quality due to the potential effects this can have on welfare. The water in each system sampled was monitored for 24 hours, with measurements of dissolved oxygen, temperature, pH, specific conductivity and ammonia taken every 15 minutes. A welfare score was developed for each fish using a multifactorial method, combining data on the condition of the fins, the condition of the gills, the stress hormone cortisol, the splenosomatic index and the mortality levels for the population of fish in the system. Using this welfare score and the individual components of the score as response variables, multi-level models were developed using the water quality, system and husbandry data collected. The primary risk factor that was associated with deteriorating welfare was disease. The purpose for which the fish was being farmed was also important, as fish farmed for the table market had on average worse welfare than those farmed for restocking fisheries. Seasonal effects, linked to higher water temperatures in summer, were associated with poorer welfare scores. Aside from seasonal effects, there is not much evidence that poor water quality is a major problem for the welfare of farmed rainbow trout in the UK. While deteriorating water quality certainly has the potential to affect the welfare of farmed rainbow trout, water quality measurements were within recommended ranges for the majority of farms visited. The results of this epidemiological study suggest that factors other than water quality may have a greater impact on trout welfare, such as exposure to diseases and production differences between farming for the table and restocking markets.

M.Sc. / The Kruger National Park is internationally one of the best known conservation areas. The perennial rivers draining eastward towards and through the park are of great importance in maintaining a healthy ecosystem. The Olifants River catchment is the largest of all the rivers flowing through the Kruger National Park. Although this river has ceased flow for three days in 1968, problems experienced in the Olifants River are of a qualitative rather than a quantitative nature. Several factors contribute to the deteriorating water quality of the Olifants River. Urbanization, agricultural, industrial and mining activities in the Phalaborwa area pose a threat to the lower parts of the Olifants River. Pollution has on several occasions lead to mortalities of populations of fish in these parts...

Data were collected from natural and introduced fishes present in Ferndale Lake, a small (120 ha) sport fishing reservoir in Camp County, east Texas. Levels of pH measured in the lake during the study period ranged from 3.5 to 5.3. Monthly field surveys and experimental manipulations were designed to evaluate quantitatively the signs of stress at various biological levels. Lethal limits to low pH were quantified for largemouth bass (Micropterus salmoides) and bluegill (Lepomis macrochirus) to be pH 3.8 and 4.0,respectively. Mean blood pH (+ 1 SD) of 59 bluegill was 7.41 (j 0.16), with no significant difference (P-0.05) among groups from Ferndale Lake and Moss Lake (Cooke Co., Texas) under experimental conditions, even when severe stress was externally apparent. In a dual-trough horizontal pH gradient, bluegill behavioral avoidance was observed at pH levels below 7.0. Individual testing of 40 bluegill in pH gradient of 5.2 to 7.6 resulted in median occupation of pH 7.1,with an interquartile range of pH 6.9 to 7.3. Decreased community structure and population "well being" compared to early studies cannot be attributed entirely to recent acidic condition. Separating potential stress due to lake conditions from that due to heavy biotic predation by sport fishing in a small reservoir is difficult.

M.Sc. (Zoology) / Concern has been shown by Ezemvelo KZN Wildlife for the state of the Lower Pongolo River and Floodplain due to the ever increasing urban and industrial development in the area. This increase in development poses a threat to the fish populations within the system and their ability to support the surrounding communities. Continued spraying of pesticides e.g. DDT within the area are adding to stresses placed on the fish populations through decreasing water quality. The effectiveness of environmental water flows suggested by White et al. (1984), have never been assessed by conservation authorities. This means that the authorities have been unable to fulfil their legislative obligations to establish management and conservation plans (Hughes et al., 2001). Work on the biological responses of fish to anthropogenic stressors such as organic pollutants and metals; have been conducted by a group of experts in the field of ecological health. This section of the project focusses on two fish species, Hydrocynus vittatus and Synodontis zambezensis, sampled within the study area and used for biomarkers and tissue chemical level determination. Standard techniques were applied for Organochlorine Pesticide (OCP) analysis according to Yohannes et al. (2013) and these results have been compared to biomarker responses which have also been completed following standard protocols. Results found within both fish species, show the relationship between accumulation of toxicants and biomarker responses. The highest levels of ΣOCPs where found within H. vittatus at (Mean ± SE) 1101.61 ± 610.97 ng/g lipid weight, the highest constituent being technical grade Dichlorodiphenyltrichloroethane (DDT). High levels of ΣDDT were also found in S. zambezensis, but were lower than levels in H. vittatus and these levels correspond with positioning of each species within the food web. H. vittatus occupies a much higher level in the food web as they are top predators and S. zambezensis falls lower down on the food web due to their reliance on detritus and only small invertebrate species for food. Biomarker responses are clearly affected by bioaccumulation levels and this is especially evident within H. vittatus as Achetylcholinesterase (AChE) is greatly inhibited by high levels of ΣDDT. Changes in the level of biomarker responses are not as apparent in S. zambezensis and this is believed to be food web linked, due to differences in bioconcentration. Data collected during this study will contribute to baseline data on S. zambezensis and add to already accumulated data on H. vittatus. This new data will also help with the improvement of or alteration to already present monitoring programmes in the study area.

Grand Coulee Dam created Franklin D. Roosevelt Lake as part of the Columbia Basin Project. Located in northeastern Washington State, the Project provides economically important hydropower (19 billion kilowatt hours per year), irrigation (225,000 ha), flood control, and sport fishing ($5 to 20 million annually). A good system understanding aids in balancing these beneficial uses for the 230 km long reservoir. The reservoir's atypical 45-day mean residence time is much shorter than a typical lake, and much longer than for a riverine dam. The spring freshet requires drawdowns of 15 to 20 m for flood control—the driving characteristic of reservoir operations. A physically based two-dimensional hydrodynamic and water quality model, CE-QUAL-W2 Version 3.5 (Cole and Wells, 2006), is coupled with a fish bioenergetics model based on the Stockwell and Johnson model (1997, 1999) to examine the effects of hydrodynamics on the reservoir algae-zooplankton-kokanee food web. This model was applied and calibrated to Lake Roosevelt with model improvements of multiple zooplankton compartments and zooplankton omnivory. Calibration parameters included temperature, dissolved oxygen, nutrients, algae, and zooplankton. The fish bioenergetics model is applied over the entire reservoir model space to generate a spatial and temporal fish growth potential distribution. The fish model refinements include sub-daily time-steps and an optimized vertical foraging strategy. The linked model suggests that kokanee fish growth potential is seasonally limited by both warm water and prey densities. While the lake ecology is significantly affected by the reservoir operations in general, the pelagic fish growth potential did not appear sensitive to minor changes in reservoir operations. However, the model suggests that the advantageous foraging locations shift seasonally and that optimal foraging strategies are dependent on fish size.

The fish community structure of two contrasting estuaries, one with a well developed mangrove forest (Mngazana) and the other without mangroves (Mngazi) was compared. Both the Mngazi and Mngazana estuary fish communities were dominated by marine species, reflecting the importance of these systems as nursery areas for marine fishes. The Mngazi Estuary contained 18% more estuarine fishes in terms of catch per unit effort (CPUE) than the Mngazana Estuary. The reduced tidal influence due to the narrow mouth opening is a possible reason for the heightened CPUE of estuarine species in the Mngazi estuary. The recorded higher diversity of fish species in the Mngazana Estuary when compared with the Mngazi Estuary was attributed to the greater influence of the marine environment due to the wide permanently open mouth, as well as the presence of a variety of habitats in this system. In both estuaries tropical and temperate species were captured, confirming the transitional nature of their biogeographic location which is situated close to the boundary between the subtropical and warm temperate regions of the Southern African coastline. Contrary to previous studies, which recorded seasonal changes in the proportions of tropical and temperate species, the proportions of tropical species remained unchanged at approximately 70% during the January and June sampling occasions. Global warming as a possible reason for the increased dominance of tropical species is discussed. Although several studies in southern Africa have investigated estuarine food web structure, none have compared mangrove and non-mangrove estuaries. In this study, the primary sources of carbon utilised by the fish fauna in the Mngazi and Mngazana estuaries was investigated. The carbon isotopic values of fishes in both estuaries displayed a continuum rather than a tight clustering around particular energy sources. Most detritus feeders of the family Mugilidae (mullets) from both estuaries were relatively more enriched than other fish taxa. The isotopic values of the mullet species suggest a diet derived from relatively enriched carbon sources such as benthic microalgae, the eelgrass Zostera capensis and associated epiphytes. Based on the isotopic values, piscivorous fishes from both estuaries could not be linked to specific prey fish taxa, but clearly the mullet species were not their main food source. The invertebrate feeders that were found in both estuaries showed greater isotopic variations in the Mngazana Estuary than in the Mngazi Estuary, probably reflecting the higher diversity of habitats (carbon sources) and invertebrate prey species in the Mngazana system. Generally the isotopic signatures of fishes from the Mngazi Estuary were more enriched than those from the Mngazana Estuary, thus indicating the possible effect of δ¹³C depleted mangrove derived carbon in the latter system.