The objective of this study was to clarify the relationships between drift rates, population density, production rates, key environmental factors, and movements of adults within two populations of stream insects (a caddisfly--Oligophlebodes sigma and a mayfly--Baetis bicaudatus). Samples of benthic invertebrates (424 total) were collected every 2 8 days at four stations on Temple Fork of the Logan River, Utah, from October, 1967 to September, 1969. Samples of drift invertebrates (181 total) were collected every 14 days at three stations on Temple Fork during the same period. During summer months (June-September) a day and a night drift sample (681 total) were collected every other day. Drift rates of O. sigma larvae were greatest (5,987 gm/year for O. sigma) when density in the benthos (256 mg/0.1 m2) and production (430 mg/0.1 m2 /year) were greatest. Drift rates were not related positively to density in the benthos over an entire year, but drift rates were correlated positively and significantly (r = .78 and .55 for day drift of O. sigma and B. biaaudatus, respectively) with density during the months of June-September for both O. sigma larvae and B. bicaudatus nymphs. Flow, distance below the spring source of Temple Fork, and densities of competing aquatic insects were other factors of significance in the multiple regression analyses of factors affecting drift rates of the two insects. The 17 independent variables in the multiple regression analyses accounted for 65 percent and 55 percent of the variability in day drift rates of o. sigma and B. bicaudatus, respectively. The adults of O. sigma (but not those of B. bicaudatus) undertook a definite upstream migration estimated at 2-3 km. This flight of adults resulted in a concentration of eggs being laid in the upper reaches of the stream. The advantage of the upstream flight may be that it stores reproductive products in areas where they are relatively safe from the effects of anchor-ice in the winter and floods in late winter and early spring.

Chlorine has a wide variety of applications in water treatment. Because of its disinfectant efficacy, it is used world wide for the treatment of potable water, sewage, swimming pools and for the control of nuisance organisms in cooling towers. A problem arises when such chlorinated water enters the natural environment, as chlorine's greatest advantage, i.e. its germicidal capacity, becomes its greatest disadvantage. In particular, the discharge of heated, chlorinated water from cooling towers and chlorinated, treated sewage into rivers have severe consequences for the riverine flora and fauna. This study focused on the effects of chlorinated, treated sewage effluent on the community structure of benthic macroinvertebrates in two rivers in KwaZulu-Natal viz. the Umsunduze River in the Pietermaritzburg area, and the Umbilo River in the Durban area. The study was conducted in three phases. The first two phases comprised a toxicological investigation of the effects of chlorine on a selected riverine macroinvertebrate, and the third phase comprised an ecotoxicological investigation of the effects of chlorinated treated sewage on benthic macroinvertebrate community structure. The first phase of the study involved the development of an artificial stream system which would be suitable for determining the response of a selected macroinvertebrate species to chlorine. Chlorine is both reactive and volatile, so this necessitated the development of a specialised flow-through system with apparatus which would allow continuous dosing of a sodium hypochlorite solution. The system was set up at the Process Evaluation Facility at Wiggins Waterworks, Durban, where raw water from lnanda Dam was used. The second phase involved the use of this artificial stream system to conduct acute 96 h toxicity tests. Baetid mayfly nymphs (Baetis harrisoni Barnard) were selected as the test organisms after a preliminary investigation found them to be suitable for survival under laboratory conditions. For comparative purposes, tests were run first on B. harrisoni from a relatively uncontaminated stream in a residential area of Westville, then on specimens from the severely impacted Umbilo River. The LC₅₀ of chlorine for organisms from both sources was found to be in the region of 0.004 mg/l (free chlorine). This value was well below the general effluent standard of 0.1 mg/l in effect at the time. The recommended acute environmental guideline is 0.001 mg/l. The third phase of the study involved field validation of the toxicity test results. It was hypothesised that since the LC₅₀ for free chlorine was 0.004 mg/l, B. harrisoni would not be found downstream from a point source of chlorinated effluent where the concentration of free chlorine ranged from 0.06 to 0.2 mg/l, and that the macroinvertebrate community structure would also be altered. In order to test these hypotheses, benthic macro invertebrate community structure was investigated at several sites up- and downstream from the outlets of the Darvill Wastewater Works in the Umsunduze River and the Umbilo Sewage Purification Works in the Umbilo River. In addition, in order to differentiate between the effects of chlorinated and unchlorinated treated sewage, a section of the Umbilo River (upstream from the chlorinated discharge) was exposed to unchlorinated, treated sewage. In this way, a limited "before and after" sewage and an "upstream and downstream" from sewage investigation could be carried out. Organisms were collected from riffles (and from pools in the Umbilo River) and the samples were then sorted and organisms were identified to species level, where possible, otherwise to genus or family. Changes in community composition were shown graphically as pie charts of relative proportions of organisms found at each site, graphs of the average number of taxa at each site; and graphs of the average number of individuals at each site; Data from the Umbilo River were also analysed using TWINSPAN (Two-way indicator species analysis). In both the Umsunduze and the Umbilo rivers, the deleterious effects of the chlorinated effluent were clearly evident. At Umsunduze Site 3 and Umbilo Site 5 (both immediately downstream from the chlorinated effluent) both the number of taxa and number of individuals were substantially reduced, sometimes to zero. Where organisms were found at the next sites downstream (Sites 4 and 6 respectively), the samples were dominated by Chironomus. In contrast, the unchlorinated effluent in the Umbilo River caused very little difference in community structure. As predicted, B. harrisoni was not found in downstream samples in which chlorine was present, yet appeared to be relatively unaffected by the unchlorinated effluent, suggesting that chlorine, rather than the effluent was responsible for its absence at downstream sites. In conclusion, it would appear that while treated sewage effluent certainly causes changes in macroinvertebrate community structure, chlorination of this effluent leads to large scale destruction of the riverine community. This in turn delays the recovery process of the river, rendering a longer stretch unfit for use. The consequences of this delayed recovery are that the failure to meet the water quality requirements of the natural environment results in those of the other water users (agriculture, industry, domestic and recreation) not being met. This reduces the natural capacity of the riverine community to process organic waste and recover from the discharge of sewage effluent. Chlorination increases the distance of impaired water quality and environmental integrity which result from organically enriched treated sewage effluent. The results of the study indicated that the draft water quality guidelines for aquatic ecosystems, derived from inadequate data, and calculated with a safety factor, were the correct order of magnitude. The approach followed in the study will be useful in the development and refinement of water quality guidelines for aquatic ecosystems.