Fish and Plant Production in a Recirculating Aquaponic System:

a New Approach to Sustainable Agriculture in Canada

N.A. Savidov and E. Hutchings J.E. Rakocy
Crop Diversification Centre South Agriculture Experiment Station
Alberta Agriculture Food University of the Virgin Islands
and Rural Development Kingshill, St. Croix
Brooks Virgin Islands
Canada USA

Keywords: aquaponics, greenhouse production, vegetable production, aquaculture

Abstract
An aquaponic system was built, based on the University of Virgin Islands
design, as a prototype for commercialization in Alberta, Canada in 2002. It consists of
four fish rearing tanks (5 m-3 each) and four raft hydroponic troughs (29 m-2 each). To
test commercial feasibility of aquaponics under the climatic conditions of Alberta the
food fish tilapia was selected in combination with several conventional greenhouse
plants (cucumber, tomato, etc.), herbs, medicinal plants, and nutraceutical plants. A
protocol was developed for producing aquaponic crops in Alberta. More than 60
different crops and varieties were tested in a greenhouse. Based on this preliminary
evaluation, 24 crops (five greenhouse vegetables and 19 herbs) were grown in trials to
determine production levels. Yields of tomatoes and mini-cucumbers reached 20.7 kg
plant-1 year-1 and 33.4 kg plant-1 year-1 and exceeded average values of commercial
greenhouses in Alberta that employ conventional hydroponic technology. During the
2-year study, the yield of Genovese basil increased from 13 kg m-2 year-1 to 42 kg m-2
year-1 as production and harvesting methods were refined. This study demonstrated
the technical feasibility of the aquaponic technology in Alberta. Evaluation of the
economic feasibility is under way.

INTRODUCTION
Aquaponics is considered as a potentially important industry for Alberta with two
profit centres: fish and plants, locally grown high value products, which can be produced
and marketed year round. A grant of $100,000 in 2002 by provincial government of
Alberta gave a start to a first research aquaponics facility in Prairies.
The funds were used to construct an aquaponics module based on the model
developed at University of Virgin Islands. The evaluation of this model under Alberta
conditions in greenhouse system with supplemental heating for both the fish and the crop
was the major objective of the project in Brooks.
Preliminary results in 2002/2003 year showed that fish waste could be an adequate
source of nutrients for intensive crop production (Savidov, 2004). The yields under the
conditions of standard greenhouse technology and plant density were about 40 kg of
tomatoes m-2 year-1, 100 cucumbers m-2 year-1. These yields far exceeded the average
yields of greenhouse vegetables produced by organic soil-based technologies in Canada.
However, the limited time did not allow evaluation of the system when operating in full
mode production, as not all tanks were occupied by fish. Further research was required to
measure the yield of the fish and crops. Therefore, the project continued in 2003/2004 for
evaluation crop yield, fish yield, and the market potential for high value niche products in
Alberta. Objectives of the project included evaluation of the stand-alone, warm water fish
model at Crop Diversification Centre South under Alberta conditions; achieving a
sustainable balance between fish and plant parts of aquaponics system and optimizing
plant crop yields for greenhouse vegetables including tomato, mini cucumbers, basil and
other crops in a computer controlled environment and marketing study.

Proc. Int. Conf. & Exhibition on Soilless Culture

Ed. K.K. Chow 209
Acta Hort. 742, ISHS 2007
MATERIALS AND METHODS
Recirculation System in Brooks
The stand-alone aquaponics system at the Crop Diversification Centre South,
Brooks was based on the J. Rakocy model (Rakocy et al., 2004) adapted to greenhouse
conditions (Fig. 1).
The facility consisted of three greenhouses (each 7.6 m wide x 15.5 m long) in a
straight line separated by storage areas (each 2.9 m wide x 7.6 m long). One greenhouse
contained the aquaculture equipment and the other two contained the plant trays. The
aquaculture area held four fish tanks, two clarifiers, five settling/degassing tanks, one
central sump tank and a base mixing barrel for a total system capacity of 71750 L. Fish
were raised in four fiberglass culture tanks (2.4 m dia x 1.2 m deep, 5600 L capacity)
arranged in two series of two tanks each. Fish tank effluent moved through two conical
clarifier tanks (each 4500 L) that removed most of the solids through a series of baffles.
Accumulated solids were drained from the clarifiers daily and stored in a holding tank for
later application to field crops. Water moved from the clarifiers into two rectangular
settling tanks (each 750 L) then into a joint degassing tank. These small tanks removed
the rest of the solids and CO2 from the fish effluent by filtration through plastic netting.
The net filter provided extended surface area for residing ammonifying and nitrifying
bacteria to mineralize organic waste. Water from the degassing tank flowed into four
plastic-lined, concrete plant troughs (each 0.9 m wide x 30.5 m long x 0.45 m deep, 9000
L capacity) arranged in two series of two. The outflow from the plant troughs moved into
a small sump tank (1000 L) where a submersible pump continuously circulated the water
back to all fish tanks via a mixing barrel. Fresh water was plumbed into the sump tank
area, through a heat exchanger and boiler system capable of keeping the water
consistently warm (~ 24.8ºC for tilapia). When the float valve in the sump tank lowers
with the water level it triggered the addition of more fresh water. Thus, the replacement
rate adjusted automatically. Water circulated through the system at 400 L min-1. Each fish
tank received a flow of 100 L min-1 and each plant tray received a flow of 200 L min–1.
This gave a turnover time of once per hour for the fish tanks and once every 45 minutes
for the plant trays. The greenhouse and recirculation system is under full-computerized
control (Argus Control System Ltd). The computer collects some data on a daily basis
using specific probes. Environmental parameters in the greenhouse such as temperature
and humidity are maintained at stable levels by the computer using heaters/coolers and
humidifiers. Irradiation in the greenhouse is also being monitored. The recirculation
system is aerated using air blowers and diffusers and had a liquid oxygen backup. Water
temperature, oxygen levels, electric conductivity (EC), and pH are monitored
continuously by the computer control system.

Fish Culture
Fish growth trials at the Brooks aquaponics facility were carried out at 24.8°C in a
24-week growth cycle with staggered production. Each tank initially received 600 finger-
lings of tilapia of 100 g mean wet weight. The fish fingerlings were supplied every 6 weeks.
The expected food conversion ratio (FCR) was 1.3 at 90% feeding efficiency. Fish were fed
3.2 mm pellets up to a mean size of 300 g, and 4.8 mm pellets beyond 300 g. Food was
provided through automatic feeders linked to the computer control system. Feeding rates
started at 2.5% day-1 for 100 g fish and was gradually reduced to 1.25% day-1 for fish of
400 g. With this regime, fish were expected to reach a market size (700 g) in 24 weeks. At
the end of the trials, all fish were returned to the Alberta Fish Farmers Association.

Plant Culture
The plants grown were selected according to their commercial importance and
their conductivity factor (CF, 100 μS = 1 CF) which indicates their tolerance to different
concentrations of minerals and their ability to extract minerals including nitrogen. Three
groups were tested:

210
• Group 1, high CF (20-45, tomato and egg plant)
• Group 2, medium CF (10-20, lettuce, basil, chives, spinach, parsley and cucumber)
• Group 3, low range of CF (2-10, water cress)
Plants were also selected based on their ability to grow fast and resist disease.
Plant seedlings were grown in rockwool and transferred to holes in styrofoam sheets
floating in the plant troughs. The plants were grown in the greenhouse at an air
temperature of 22-25°C, an irradiation level ≥300 µmol photons m-2 sec-1 photo-
synthetically available radiation (PAR), and a 16:8 day:night photoperiod provided by
natural and artificial lights. Water pH was maintained near 6.5-7.0 by the addition of
either Ca(OH)2 or KHCO3 (alternate on weekly basis) to increase pH, or H3PO4 to reduce
pH. This pH was considered optimal to maximize mineral uptake and plant growth and it
was not harmful to the tilapia. As with the fish, plant growth trials were staggered so that
the total plant production was roughly balanced with fish production. Each crop was
rotated to avoid spikes of high mineral concentration from excessive fish waste input.
Seedlings of cucumbers and tomatoes were transferred to the facility every three weeks
and basil every two weeks to ensure uniform consumption of the minerals during
operation. The crops were routinely monitored for pests and diseases. Biological crop
protection was carried out as required through integrated pest management (IPM).
Predatory insects and hyperparasites were used for chemical-free protection. The plant
growth trial protocol was standardized among the stand-alone and add-on facilities.
RESULTS AND DISCUSSIONS

Plant Production
The facility in Brooks started in December after 600 fingerlings were placed in the
tank #1. The first plants were transplanted in the middle of January in 2003. From April 1,
2003, till March 31, 2004, a number of crop species were tested in stand-alone aquaponics
system. The total number of crops was sixty. Three major crops tested in the Brooks
facility included tomato, cucumber and basil. Each crop occupied one trough. The forth
trough was used to grow a variety of crops. The first two months of plant production was
characterized by slow plant growth and number of deficiencies due to lack of nutrients
mostly iron and magnesium (Fig. 2).
The absence of nutrients in the beginning visibly affected biomass accumulation.
That was anticipated due to lack of nutrients in the beginning of aquaponics production
cycle. In the beginning, the symptoms included nitrogen, phosphorous, potassium, iron
and other microelement deficiencies. However, there was a significant difference in
severity of the observed symptoms among different species. Lettuce proved to be the most
affected and portulaca showed minimum deficiency symptoms. Aquatic plants such as
water hyacinth, frog bit, crystal wort and azolla were not affected by the nutrient
deficiency at all. Aquatic plants are known for their ability to extract nutrients from very
diluted solutions. The situation with conventional greenhouse crops dramatically
improved when aquaponics solution accumulated enough nutrients for crop production
after 3-4 months of operation. Mini cucumber variety Alamir reached peak of production
by April 2003 (1.29 kg plant-1 week-1, Fig. 3).
The high production period continued until September, then yield started
decreasing. The problem was identified as a root rot caused by Pythium aphanidermatum.
It has been suggested that high occurrence of the disease was due to accumulation of
waste material in the biofilter, which served as a breeding ground for P. aphanidermatum.
The problem was contained by weekly cleaning filter tanks (Fig. 1). One of the possible
reasons of increased susceptibility of cucumbers to P. aphanidermatum was also shorter
light period in the fall. The combined improved sanitation procedures and increased
length of day light in the end of February 2004 led to higher yields, which were restored
to the levels in summer 2003. The projected level of mini cucumbers production in stand-
alone aquaponics facility per year during P. aphanidermatum-free period in 2003/2004
even surpassed the average level in the industry using conventional hydroponics. Cv.

211
Alamir was compared with three other varieties of mini cucumbers including Harmony,
Kian and Melita. There were no significant differences between four varieties except cv.
Harmony showing slightly slower growth in August and September (Fig. 4).
Heirloom tomato varieties performed well in aquaponics exceeding the average
reported yield level by 10-15% (Fig. 5).
However, the low yield of Roma tomatoes in the winter 2004 indicated the need
for higher light conditions for this cultivar (Fig. 6). The adequate level of irradiation in
May increased the yield more than four times.
The basil crop was definitely one of the most successful crops in aquaponics (Fig.
7). The yield was steadily growing from 13 kg m-2 year-1 to 42 kg m-2 year-1 increasing
more than 3 times. There were not any problems observed with pests and diseases for the
reported fiscal year. We suggest that this growth in production reflected gradual
accumulation of favorable factors for this crop in the aquaponics system.
Among other vegetables, Japanese eggplant and bitter melon, were tested. Both
crops produced good yields (Fig. 8) especially for bitter melon (37.2 kg plant-1 year-1). Up
to 60 different crops were tested for the reporting period. Most of them were culinary
herbs (Fig. 9). Culantro was the slowest growing crop producing only 4 kg m-2 year-1. The
highest yield was produced by Swiss Chard and water spinach (51.5 kg m-2 year-1 and
58.3 kg m-2 year-1 respectively).

Fish Production in Aquaponics System

The fish production was monitored through fish sampling every second week. The
results showed that the biomass increased steadily in all fish tanks (Fig. 10). However, the
mortality was high in the beginning of the reported fiscal year reaching 30%. The number
of fish in tank #1 dropped from 600 to 420 in July 2003 (Fig. 11).
It was suggested that the low quality of fingerlings was the major cause of the
problem. The situation was considerably improved after changing a supplier. A significant
number of fish (up to 25%) was lost in September due to a power outage. The incident
suggested that a backup power generator should be a compulsory requirement for running
aquaponics operations. Despite of the lower than expected production in 2003/2004, no
major problem in aquaculture was associated with the water quality.

Water Quality, pH and the Nutrient Balance in the Fish Effluent

The nitrogen cycle is a central factor of bioproductivity in natural and artificial
ecosystems. Ammonia is the main component in the excrements of freshwater teleosts.
Ammonia is oxidized in two-step reaction by nitrifying bacteria with production of
nitrate. The nitrification is a crucial process in aquaculture as it reduces level of
ammonium, which is a major cause of toxicity for farmed fish. The efficiency of
nitrification is higher in alkaline solution, pH 7.5-8.0, which is the reason for relatively
high pH in most aquaculture facilities. We used pH 6.5-7.0, which was a compromise
between requirements for active nitrification and plant nutritional requirements. The N-
NO3- level gradually increased from 0 to 10 mM, which provided sufficient source of
nitrogen for plants (Fig. 12). The iron and microelements were added to the solution to
reverse the deficiency symptoms.
Since the concept of aquaponics implies use of fish feed as a major source of
nutrient for the plant production, the nutrient balance in the fish feed is crucial for the
plant production. The requirements for potassium are different for plants and for fish.
Fishmeal, the major component of the fish feeding formulations is not always rich in
potassium. The measured level of potassium in the fish effluent was 10 fold less than that
of calcium and 5 fold less than sodium in the beginning of the experiment. The
recommended Ca:K ratio for hydroponic production of most crops is between 2:1 and 1:1.
Ca and Na interfere with K uptake. The increased level of these elements can cause severe
K starvation. Thus, the preliminary observations in aquaponics system revealed an
intrinsic nutrient imbalance in the system based on fish feeding formulations as the only
source of the plant nutrients. The existing aquaponics systems use either calcium or

212
potassium hydroxide supplements in order to control pH. In such systems, however,
potassium level is not controlled by the plant demands, but rather by pH. We proposed to
add potassium supplement, due to its deficiency in fish feed. The balance between the
plant nutrients in the fish effluent was controlled by the addition of supplements limited to
iron and potassium to provide the best nutrient regime for the maximum plant production.
On the other hand, the fish effluent provided the major portion of the nutrients. After six
months of operation the macronutrient balance in aquaponics facility closely mimicked a
standard commercial mixture with minimum supplements with minerals (Fig. 12).
However, we found persisting problem with iron deficiency (Fig. 13). Concentration of
Zn and B considerably exceeded recommended norm for hydroponic production (Fig. 13).
Cucumber leaf analysis conducted in summer 2004 showed no deficiencies in both
macro- and microelements and relatively high level of B, Zn, and Mn in established
aquaponics system (Fig. 14 and 15).

CONCLUSIONS
1. The study demonstrated that Rakocy aquaponics model developed for outdoor
conditions can be successfully adapted for greenhouse operations in Canada.
2. The same yields of vegetables as in hydroponics are achievable using aquaponics
technology.
3. The rate of fish biomass production in aquaponics is comparable with conventional
aquaculture operations.
4. Aquaponics system has an intrinsic capacity of self-regulation and balancing nutrients
in the solution.
5. The nutrient balance necessary for optimal crop production can be reached within six
months of operation or earlier.
6. Biological control is an essential tool for success of aquaponics operations.
7. Staggered crop and fish production and maintenance schedules prevented spikes in
nutrient concentrations.
8. Economic analysis is under way. However, a preliminary estimation indicates that
aquaponic operations are economically feasible when growing high value crops like
basil.

ACKNOWLEDGEMENTS
The authors are thankful to a number of funding agencies including AAFRD- New
Initiatives Fund, Department of Fisheries and Oceans Canada, and Diversified Livestock
Fund of Alberta and Alberta Aquaculture Association of financial support of the project;
Drs. R. Howard and M. Bandara for useful discussions; and to Patricia Cote and Don
Bailey and all other technical personal involved in aquaponics project at Crop
Diversification Centre South for technical assistance.

Literature Cited
Rakocy, J.E., Shultz, R.C., Bailey, D.S. and Thoman, E.S. 2004. Aquaponic production of
tilapia and basil: comparing a batch and staggered cropping system. Acta Hort.
648:63-69.
Savidov, N. 2004. Annual Report of Greenhouse Crops Program for 2003/2004. CDC
South, Brooks, Canada.

213
Figurese

Filter
Fish Clarifiers
tanks
Plant growing troughs

Fig. 1. Brooks aquaponics facility (BAF).

Fig. 2. Iron deficiency in Faba vulgaris plants grown aquaponically.

214
 Reduced yield from Pythium infection during winter 2003/2004
1.4

1.2
Yield, kg plant-1 week-1

0.8

0.6

0.4

0.2

0
3-Mar 5-Apr 1-May 2-Jun 2-Jul 4-Aug 22-Sep 27-Oct 30-Nov 15-Dec 5-Feb 9-Jan 23-Jan 2-Feb 30-Mar 7-Apr 16-Apr 5-May 26-May 15-Jun 21-Jun

Crop date

Fig. 3. Mini cucumber production, cv. Alamir, 2003/2004.

30

Harmony
25 Kian
Alamir
Yield, kg plant-1 year-1

Melita
20

15

10

0
August September November

Month

Fig. 4. Yield of different mini cucumber cultivars grown aquaponically in 2003.

215
 0.

Yield, kg plant-1 week-1 0.

0.

0.

0.

0.

June July August September

Month
Fig. 5. Production of Heirloom tomatoes in aquaponics in 2003.

0.5

0.4
Yield, kg plant-1 week-1

0.3

0.2

0.1

January February March April May

Month
Fig. 6. Production of Roma tomatoes in aquaponics in 2003.

216
 50

40
Yield, kg plant-1 week-1

30

20

10

0
Crop date
Fig. 7. Annual production of basil, cv. Genovese, in aquaponics.

2
Yield, kg plant-1 year-1

0
Heirloom Roma Cucumbers Eggplant Bitter
Tomato Tomato Mini melon

Fig. 8. Annual production of vegetables in aquaponics.

217
218

60

50
Yield, kg m-2 year-1

40

30

20

10

0
Amaranth Basil Lemon Chives Cilantro Culantro Fenugreek Parsley Spinach Water cress
Basil Genovese Basil Osmin Choi Coriander Dill Lettuce Portulaca Swiss Chard Water spinach

Fig. 9. Annual production of herbs in aquaponics. The fish production was monitored through fish sampling every second week.
The results showed that the biomass increased steadily in all fish tanks (Fig. 10).

218
 60 60
Number of fish
Weight
Number of fish tank-1 50 50

Yield, kg tank-1
40 40

30 30

20 20

10 10

0 0
Jul-03 Dec-03 Mar-04
Dates of harvest
Fig. 10. Fish production in 2003/2004.

0.6
Tank1
Tank2
0.5 Tank3
Average fish biomass, kg1

0.4

0.3

0.2

0.1

0
27-May 12-Jun 24-Jun
Date
Fig. 11. Fish production rate in May-June, 2003.

219
 500
aquaponics
Nutrient concentration, ppm

hydroponics
400

300

200

100

0
Ammonium Nitrate Phosphorous Potassium Calcium Magnesium Sodium

Fig. 12. Concentration of macroelements and sodium in aquaponic solution as compared

with standard hydroponic mixture.

3.5
Nutrient concentration, ppm

3 aquaponics
hydroponics
2.5

1.5

0.5

0
Iron Manganese Boron Zinc Copper

Fig. 13. Concentration of six major microelements in aquaponic solution as compared

with standard hydroponic mixture.

220
 Deficiency Normal range

Young leaves Old leaves

Mg

Ca

0 2 4 6 8 1 12
Nutrient level, % Dry Matter

Fig. 14. Diagnostics of macroelements in leaf tissue of cucumber plants grown in aqua-
ponics.

Deficiency Normal range

Young leaves Old leaves

Cu

Zn

Mn

Fe

0 5 10 15 20 25 30 35
Nutrient level, ppm

Fig. 15. Diagnostics of five major microelements in leaf tissue of cucumber plants.

221
222