Elsevier

Aquaculture

Volume 311, Issues 1–4, 3 February 2011, Pages 175-186
Aquaculture

Estimation of waste outputs by a rainbow trout cage farm using a nutritional approach and monitoring of lake water quality

https://doi.org/10.1016/j.aquaculture.2010.12.001Get rights and content

Abstract

Nutrients released by cage fish farms to the aquatic environment are an issue of concern since these can result in deleterious environmental changes. In the present study, a mass balance model was used to calculate nutrient loadings from an experimental rainbow trout cage farm located in a freshwater, oligotrophic lake. Detection of these nutrient loadings using water quality monitoring was then investigated.

The loading of total solids of faecal and feed origin (TS), solid phosphorus (SP) and nitrogen (SN), and dissolved P (DP) and N (DN) wastes from the farmed fish were estimated over two production cycles. Waste outputs were estimated using the Fish-PrFEQ feed requirement and waste output model using measured inputs including diet composition, nutrient digestibility, intake and retention by the fish, and water temperature. Nutrient loading predictions were compared with measured nutrient concentrations of lake water.

In 2003, TS, SN and SP and DN and DP waste outputs were 236.0, 12.8, 5.3, 41.3, and 3.4 kg tonne 1 of fish produced, respectively. In 2004, the TS, SN and SP and DN and DP waste outputs were 220.0, 12.2, 5.3, 38.0, and 3.4 kg tonne 1 of fish produced, respectively. Over 60% of the P waste output from the cage was predicted to be solid for both years while over 65% of the total N waste from the cage farm was predicted to be excreted as ammonia. Concentrations of ammonia and of dissolved and particulate phosphorus were not reflective of waste loading of cage origin, suggesting efficient removal through uptake by biota and/or in the case of ammonia by nitrification.

Fish-PrFEQ model is a valuable cage management tool that allows realistic estimation of waste outputs for cage farms and can be used to examine effect of management and feeding practices on waste outputs. However, the model is of limited use as a lake management tool, as it does not consider effects and fate of wastes released by fish. Similarly, reliance on periodic water quality monitoring at stations near cage farms may not be protective of the environment, as our results demonstrate that rapid diffusion, uptake, transformation and removal of nutrients resulted in water quality measures that were relatively insensitive to cage loading. Combining the Fish-PrFEQ model with a consideration of assimilative capacity of the system in addition to the monitoring of chemical and biological variables of the lake is recommended for environmental impact assessment of cage culture operations.

Introduction

The environmental impact of waste from the fish culture industry, notably from cage culture operations, is increasingly a matter of close scrutiny by the public and various levels of government in Canada and elsewhere around the world. The main concern is the release of solid and dissolved phosphorus (P) and nitrogen (N) wastes since these wastes may lead to environmental degradation. Environmental compliance of cage culture operations is generally assessed through measurements of nutrient concentrations in receiving waters to assure that certain target water quality criteria are not exceeded (Ontario Ministry of the Environment, 1993, Boyd et al., 2001). For example, in Ontario, cage farms are required to take depth-integrated (over depth of cages) water samples for measurement of total P, 30 m from each side of the cages and at two remote reference stations eleven times during the ice-free period (Boyd et al., 2001). There is reason to believe, however, that these measures might not be particularly sensitive to cage outputs.

Net-cage farm sites very often are located in a dynamic, multi-user environment where it is difficult to differentiate farm-loaded nutrients from natural and other anthropogenic sources, and where the effects from separate point sources such as cage farms will not be easily detected (Reid et al., 2006). The rate of horizontal mixing in epilimnetic waters is very rapid (Quay et al., 1979), and dissolved phosphorus, the primary nutrient of interest, is taken up rapidly (within minutes) by bacteria and phytoplankton (Levine et al., 1986) and is removed to sediments through the processes of consumption, senescence and settling (Schindler et al., 1971). Cage rainbow trout production results in the release of approximately 7.3 kg of P to the environment for every tonne of fish produced (Bureau et al., 2003). However, the majority of the waste P (> 60%) is particulate and has a settling rate of ≥ 6 cm/s (Reid et al., 2009), meaning that within a few minutes, most of the faecal material that is generated above the epilimnion will pass to the sediments. Only a portion of the particulate (faecal) P transported to the hypolimnion will be bioavailable (Hua and Bureau, 2006), and once in the hypolimnion, slow diffusion across the thermocline density gradient will largely prevent the movement of P back up into the epilimnion. Thus, P becomes largely unavailable for the production of nuisance algae or detection by epilimnetic water monitoring programs until thermal mixing occurs. As thermal mixing occurs only two times a year in a dimictic system such as Lake Huron, there can be a time lag of up to 6 months between the deposit of P into the water and its availability for sampling over the depth of the cages. The horizontal mixing rates of epilimnetic waters, the dynamics of P loading from fish farms, and the behavior of aquaculture-origin P cycling in the dimictic lake environment all combine to make detection of waste release by monitoring water quality challenging.

Several nutritional-based models have been proposed to estimate the waste outputs from aquaculture operations (Cho and Bureau, 1998, Stigebrandt, 1999, Davies, 2000, Bureau et al., 2003). The models of Stigebrandt, 1999, Davies, 2000 are currently used as environmental management tools for the salmon cage farming industry in Scotland and Norway, respectively. The application of a modified version of the Fish-PrFEQ model as an environment management tool has been adopted by the land-based industry in France (Papatryphon et al., 2005) and has been proposed for the cage aquaculture industry in Canada (e.g., Bureau et al., 2003, Reid, 2004).

The Fish-PrFEQ model is based on feed quality, quantity of feed used, biomass of fish produced and composition of that biomass. Because of its nutritional and bioenergetic nature, provided that accurate measures of the feed (nutrients) ingested and retained by the fish are known, the same degree of accuracy would be expected of waste estimations. Furthermore, the bioenergetic nature of the Fish-PrFEQ model allows incorporation of the effects of water temperature on feed intake, growth, metabolic and excretion rates. This is particularly important when modeling excretions by fish growing under less controlled conditions, such as cages in lakes, where fish may be exposed to seasonally and spatially (i.e., over depth) varying water temperatures.

The present study is part of multidisciplinary, whole-lake experiment to investigate the effects of cage aquaculture on the freshwater environment. This component of the project is concerned with both the quantification of waste loading from the farm and the investigation of its effects. The first objective was the quantification of the farm nutrient loading. For this purpose, the Fish-PrFEQ model was calibrated with actual (measured) feed composition, nutrient digestibility, quantity of feed used, water temperature, growth rate and biomass of fish produced at a trout cage and the fish farm waste loading estimated. The second objective of this study was to examine the utility of water chemistry measures in detecting nutrient loading of cage origin. The relationship between predicted waste outputs and measured water chemistry parameters was investigated to test the suitability of these parameters as tools for environmental assessment of fish cage farms.

Section snippets

Location of experimental area and cage settings

Two production trials were carried out between May and October 2003 and 2004 in Lake 375, an oligotrophic double basin (north and south basins) lake located at the Experimental Lakes Area (ELA), northwestern Ontario, Canada (49°44′43.61″N, 93°47′15.56″W). This lake was the object of frequent monitoring over two decades and long term pre-cage (1982–2002) mean TP and TN values (mean ± SD, n = 314) of 6.3 ± 4.5 μg L 1 and 245.2 ± 92.0 μg L 1, respectively were available. The lake has a surface area of 23.2 ha

Model inputs — feed, fish growth and water temperature

Measured nutrient contents (Table 1) and nutrient digestibility (Table 2) of the commercial diet used during the 2003 and 2004 production trials were used as model inputs to calculate nutrient intake and digestible nutrient intake, respectively. The nitrogen content of the diet (%CP/6.25, as is basis) was 7.1% and the phosphorus content was 1.1% (as is basis) resulting in a dietary N:P ratio of 6.4.

Measured initial body weight, weight gain and growth rate (expressed as TGC) (Table 3) were used

Farm waste loading estimations

The present study is part of an ongoing, multidisciplinary, whole-lake experiment to examine the ecological effects of freshwater cage aquaculture. This component of the experiment was concerned with testing our ability to quantify farm nutrient loading and our ability to detect this loading by monitoring of water quality. Some observations from this project of the effects of waste loading have been reported elsewhere (e.g., Bristow et al., 2008, Findlay et al., 2009, Kullman et al., 2009,

Acknowledgements

This project was supported by the Aquaculture Collaborative Research and Development Program (ACRDP) of the Department of Fisheries and Oceans, Canada and by the Northern Ontario Aquaculture Association, Ontario, Canada. Martin Mills, Inc. provided the fish feed for this project. We thank Dr. Michael Turner, Kelly Hille and Adam McFee for constructive comments on this manuscript. We thank Dr. Patrick Buat, Adam McFee, Kelsey Fetterly, Rebekah Rooney, Adrianne Lewis, Erin Forster and Jesse Hamel

References (70)

  • L. Häkanson et al.

    A new approach to calculate the phosphorous load to lakes from fish farm emissions

    Aquacult. Eng.

    (1998)
  • K. Hua et al.

    Modelling digestible phosphorus content of salmonid fish feeds

    Aquaculture

    (2006)
  • M.S. Islam

    Nitrogen and phosphorus budget in coastal and marine cage aquaculture and impacts of effluent loading on ecosystem: review and analysis towards model development

    Mar. Pollut. Bull.

    (2005)
  • T. Johansson et al.

    Empirical mass balance models calibrated for freshwater fish farm emissions

    Aquaculture

    (2002)
  • T. Johansson et al.

    Direct flows of phosphorous and suspended matter from a fish farm to wild fish in Lake Southern Bullaren

    Aquacult. Eng.

    (1998)
  • M.E. Mcdonald et al.

    Fish simulation culture model (FIS-C): a bioenergetics based model for aquacultural wasteload application

    Aquacult. Eng.

    (1996)
  • K.D. Shearer

    Factors affecting the proximate composition of cultured fishes with emphasis on salmonids

    Aquaculture

    (1994)
  • A. Stigebrandt et al.

    Regulating the local environmental impact of intensive marine fish farming: III. A model for estimation of the holding capacity in the Modelling–Ongrowing fish farm–Monitoring system

    Aquaculture

    (2004)
  • AOAC

    Official Methods of Analysis of AOAC International

  • P.A. Azevedo et al.

    Growth and feed utilization of large size rainbow trout (Oncorhynchus mykiss) and Atlantic salmon (Salmo salar) reared in freshwater: diet and species effects, and responses over time

    Aquac. Nutr.

    (2004)
  • S.M. Bartell et al.

    Individual parameter perturbation and error analysis of fish bioenergetics models

    Can. J. Fish. Aquat. Sci.

    (1986)
  • E.G. Bligh et al.

    A rapid method of total lipid extraction and purification

    Can. J. Biochem. Physiol.

    (1959)
  • S. Blomqvist et al.

    A review on sediment traps in aquatic environments

    Arch. Hydrobiol.

    (1981)
  • D. Boyd et al.

    Recommendations for the operational water quality monitoring at cage culture aquaculture operations

  • A.E. Brafield et al.

    Animal Energetics

    (1982)
  • Bristow, C.E., 2006. Experimental Freshwater Cage Aquaculture: Short term effects on Carbon, Nitrogen and Phosphorus...
  • C.E. Bristow et al.

    Cage aquaculture and the evolution of eutrophication in an experimental lake

    Can. J. Fish. Aquat. Sci.

    (2008)
  • D.P. Bureau et al.

    Chemical composition and preliminary theoretical estimates of waste outputs of rainbow trout reared on commercial cage culture operations in Ontario

    N. Am. J. Aquacult.

    (2003)
  • S.R. Chipps et al.

    Bioenergetics modeling in the 21st century: reviewing new insights and revisiting old constraints

    Trans. Am. Fish. Soc.

    (2008)
  • C.Y. Cho et al.

    Quantitation of fish culture wastes by biological (nutritional) and chemical (limnological) methods; the development of high nutrient dense (HND) diets

  • M.J. Costello et al.

    Seasonal changes in food conversion ratio as an indicator of fish feeding management

    Bull. Aquacul. Assoc. Can.

    (1996)
  • I.M. Davies

    Waste production by farmed Atlantic salmon (Salmo salar) in Scotland

  • P.J. Dillon et al.

    The phosphorus–chlorophyll relationship in lakes

    Limnol. Oceanogr.

    (1974)
  • D.L. Findlay et al.

    Aquaculture impacts on the microbial communities in a small boreal forest lake

    Can. J. Fish. Aquat. Sci.

    (2009)
  • J.A. Green et al.

    Effects of dietary phosphorus and lipid levels on utilization and excretion of phosphorus and nitrogen by rainbow trout (Oncorhynchus mykiss). 2. Production-scale study

    Aquacult. Nutr.

    (2002)
  • Cited by (42)

    • Bioenergetic and water quality modeling for eutrophication assessment of El Pañe Reservoir, Peru

      2021, Ecohydrology and Hydrobiology
      Citation Excerpt :

      An increase in aquaculture activities is expected according to the Peruvian National Aquaculture Plan (Produce, 2010), which may impact the water quality of lakes and reservoirs as a result of the increased nutrient loads that could lead to different states of eutrophication, for example, as has occurred in El Pañe Reservoir, Peru (Mamani Larico and Zúñiga Medina, 2019). Although several bioenergetic models have been proposed to estimate the fish feeding rate and nutrient waste load (Azevedo et al., 2011; Bureau et al., 2011; Cho and Bureau, 1998; Papatryphon et al., 2005; Stigebrandt, 1999), few studies have linked a bioenergetic model to a water quality model for management purposes (Canale and Chapra, 2016; Dalyander and Cerco, 2010; McKillip, 2008). There has been evidence of algal blooms reported by the local population in El Pañe Reservoir, which has caused the death of animals (Autodema, 2012).

    • Influence of Nile tilapia (Oreochromis niloticus) fish farming in net cages on the nutrient and particulate matter sedimentation rates in Umari reservoir, Brazilian semi-arid

      2020, Aquaculture Reports
      Citation Excerpt :

      The impacts caused on aquatic environments from the fish farming activity using net cages can lead to loss of biodiversity, through introduction of exotic species, and changes in the trophic web. This activity generates residues that become food sources for the local biota (Brandão et al., 2012; Ortega et al., 2015) and increased concentration of nutrients that can alter the natural conditions of aquatic ecosystems (Azevedo et al., 2011; Zhou et al., 2011; Venturoti et al., 2015). The wastes from fish farming are released directly into the environment as excreta and unconsumed ration, increasing the levels of organic residues in the natural environment (Gunkel et al., 2015).

    • Survival of purebred and hybrid Serrasalmidae under low water temperature conditions

      2018, Aquaculture
      Citation Excerpt :

      The ideal temperature range for tropical fish growth is 25 °C to 32 °C and, in the warmer seasons of the year, food consumption increases (Dayton et al., 1995; Dobson and Holmes, 1984). However, temperatures exceeding the optimal level for the species hinder digestion processes, inhibiting the absorption of nutrients, reducing growth rates and, in extreme cases, may be lethal (Azevedo et al., 2011; Jobling and Baardvik, 1994). In contrast, low temperatures reduce metabolic activity, decrease the immunity of the animals and favor the emergence of diseases that are triggered mainly by bacteria and fungi (Sardella et al., 2007).

    View all citing articles on Scopus
    View full text