Effects of rainbow trout (Oncorhynchus mykiss) cage culture on Western Cape irrigation reservoirs.

[摘要] ENGLISH ABSTRACT: Water storage in reservoirs forms an integral part of the agricultural landscape in the WesternCape Province, South Africa. A few large reservoirs serve primarily as a drinking and industrialwater supply, while on private farms, small reservoirs provide irrigation water for the dry summerperiod. Protection of water quality to secure irrigation and drinking water quality, and theextension of water use efficiency are priority issues in the Western Cape. In the current study,the suitability of rainbow trout (Oncorhynchus mykiss) cage farming as a non-abstractive wateruse was investigated. The current study concentrated on the identification, and where possiblequantification of aquaculture impacts, the identification of successful sites and a description ofrequirements in which net-cage aquaculture has none or a very low negative impact on waterquality (e.g. <15 % change from previous water quality conditions for phosphorusconcentrations).In order to study the effects of 5 t trout cage production units in reservoirs <15 ha in area, thegeneral ecology of the irrigation reservoirs was assessed. Sixteen reservoirs withoutaquaculture production (reference reservoirs) were compared to 26 reservoirs with aquacultureproduction (production sites with varying production histories). Catchment characteristics werealso monitored. Water from different depths (0 m, 2 m, 6 m and near bottom) was tested forphysical and chemical qualities as well as nutrient concentrations. Phytoplankton andzooplankton biomass and species composition was monitored. In addition to the generalphytoplankton findings, cyanophytes were evaluated for their contribution to algal taint problemsthat emerged at a number of production sites. Sediments were tested for total phosphoruscontent and phosphorus release capacity. Indicators and minimum conditions to avoid the mostcommon production problems were formulated. In order to determine long-term productionsuccess, which prevents trophic level changes of reservoirs, a mass balance approach (nutrientbudget) was employed to indicate the limits for nutrients that can be added. The phosphorusbalance indicated long term trends for reservoirs with and without aquaculture. The detailedmass balance approach was compared to a 'ready to use carrying capacity model thatestimated the maximum fish load each reservoir could support.The anthropological input of phosphorus into the reservoirs causes a decreasing water quality inthe studied reservoirs and this development was also reported for lowland rivers. Twentypercent of the studied reservoirs are in a condition that could be an immediate threat to fish orwater bird health (e.g. free ammonia concentrations and pH). Harmful algal blooms were notobserved.Aquaculture production evoked changes in water chemistry and ecology in most of the studiedreservoirs. Adverse effects of aquaculture sites versus non-aquaculture sites were: increasedphytoplankton biomass and species shifts towards sizes >80 μm. The increased phytoplanktonabundance influenced pH maxima to values >9 at mid-day. The high pH fluctuations weregreatly influenced by the naturally low alkalinity and hence low calcium buffering capacity ofWestern Cape waters. The deoxygenation of the hypolimnion during stagnation (summer)occurred faster in reservoirs of certain character, greatly dependent on elevation and surfacearea, with consequent acidification of the hypolimnion, as well as ammonia and totalphosphorus (TP) accumulation. In this context, a diversity of each reservoir with and withoutaquaculture production, with a similar ratio of undisturbed reservoirs to reservoirs with influenceof e.g. agriculture, were compared to each other. When grouping the respective differencesfrom the average reference reservoir (i.e. no trout production), a low impact on water qualitywas manifested at four sites (15 %) with <15 % increase of bottom TP and ammonia, while eightsites (31 %) showed medium effects (59 % mean increase), and a high impact was found at 54% of the sites (312 % mean increase). In reservoirs without aquaculture, the extent of incomingphosphorus (which could represent an influence by runoff from agricultural land) was very high.However, in small reservoirs (<5 ha), these values were exceeded by the incoming phosphorusfrom aquaculture practices. In the case of small reservoirs where the carrying capacity wasclearly surpassed, effects caused by aquaculture were severe and the assimilation of waste bythe system was not possible (in extreme cases aquaculture waste delivered 60 to 90 % of allincoming phosphorus – two to nine times the phosphorus brought in by rivers and runoff).With regards to sediment, only indirect conclusions could be drawn. Aquaculture productionincreased hypolimnetic anoxia and the latter was shown to increase potential phosphorusrelease from sediments. This implies that not only will aquaculture increase the phosphorusconcentration of surface waters directly, but it will also increase internal loading. Thesedimentation rate was increased with cage aquaculture which affects a hypothesized area ofapproximately 0.2 to 1.0 ha depending on reservoir hydrology. The composition of the sedimentincreased organic components which can impact on sediment processes. It can be postulatedthat increased sedimentation of aquaculture waste and extended anoxic conditions impact onmacrozoobenthos.Hydrological and nutrient mass exchange of the reservoirs indicated that no annual increase ofphosphorus was achieved with low nutrient input (good inflowing water quality) or good waterexchange (>5 times per year), and sometimes with extraction of hypolimnetic water during thestagnation period (summer). A model developed by Beveridge (1984) showed similar results tothe mass balance approach and can therefore serve as a more ready model to determinesuitable stocking rates.The small (man-made) reservoirs in the Western Cape are in a eutrophication process which farexceeds the speed of natural eutrophication (trophic states indicating highly eutrophic orhypertrophic conditions after approximately 10 to 20 years following construction of thereservoirs) and this process is triggered by agricultural practice (indicated by literature – not asubject of this study). However, it is concluded from the data of this study that trout-cageaquaculture duplicated the total phosphorus already present (independent of continuation of theprojects, the phosphorus introduced was trapped in the closed systems the reservoirsrepresent) in only 1 to 2 years of production - which means a significant acceleration of theeutrophication process already in place. There are positive exceptions where trout-cageproduction is possible without negative effects.Careful site selection is the most important step in successful and sustainable trout production.No impact of aquaculture was recorded at four reservoirs (15 % of the investigated reservoirs)which shared the characteristics of good water exchange (>3 times per year) and a minimumsurface area of 5 ha. Additionally, criteria that reduced the risk of algal taint included a minimumwater depth of 6 to 7 m in a reservoir at its lowest water point (to avoid intermediate mixingduring the stagnation period) as well as cold hypolimnetic conditions (<17 °C) to minimizecyanophyte cyst remobilization.Further improvement of food conversion ratio (feeding management) and feed quality are thenext (after site selection) two most important components that determine if a reservoir can beutilised for cage production without any long-term changes. There is potential in advances infeed quality, feed management and waste collection systems. These measures (e.g. the cagesize could be decreased to efficate feeding management) can increase the number ofsustainable sites and achieve multiplication of water use without water quality deterioration.

[发布机构] Stellenbosch University