Research Article Notes: Insect-Based Fish Feed in Decoupled Aquaponic Systems

Abstract

• Background:

  • Using insect meal-based fish feed is a promising alternative to fish meal to enhance circularity in aquaculture and aquaponics.

  • There's limited research on its use in aquaponics.

  • No reports exist on the effects of fish waste water from Black Soldier Fly (BSF) meal-based diets on lettuce growth performance.

• Objective:

  • Compare the effect of reusing fish waste water from tilapia culture (as a nutrient solution base) fed with:

    • Fish meal-based diet (FM).

    • BSF meal-based diet.

  • Assess impact on:

    • Resource use.

    • Lettuce growth in decoupled aquaponic systems.

• Control:

  • Conventional hydroponics nutrient solution (HP).

  • Inorganic fertilisers were added to all nutrient solutions to reach comparable target concentrations.

• Experiment Setup:

  • Conducted in a controlled climate chamber.

  • Nine separate hydroponics units, three per treatment.

• Measurements:

  • Lettuce:

    • Fresh and dry weight.

    • Number of leaves.

    • Relative leaf chlorophyll concentration.

  • Water consumption.

  • Usage of inorganic fertilisers.

  • Micro- and macronutrients in the nutrient solutions were monitored over time.

• Results:

  • Similar lettuce yield across all treatments.

  • No significant effects on fresh and dry weight, number of leaves, and relative chlorophyll values.

  • Water use per plant was similar between treatments.

  • Total inorganic fertiliser required was 32% lower in FM and BSF compared to HP.

  • Higher sodium concentrations were found in the FM nutrient solutions compared to BSF and HP.

• Conclusion:

  • BSF-based diet is a promising alternative to FM-based diet in aquaponics without negative effects on lettuce growth.

  • BSF-based diet might be beneficial in intensive, professional aquaponics applications due to the lower sodium concentration in the nutrient solution.

Introduction

• Aquaponics Definition: Food production system reusing nutrient-rich water from aquaculture for plant nutrition and irrigation $(1-5)$.

• System Combination: Combines aquatic animal production (fish in recirculating aquaculture systems - RAS) with soilless plant production (hydroponics) $(4, 6)$.

• Resource Sharing: Aquaponics relies on sharing resources between RAS and hydroponics units; the extent varies based on system operation and unit connections $(1)$.

• Decoupled Aquaponic Systems:

  • Units are partially connected with a unidirectional flow of water and nutrients from RAS to hydroponics $(1, 7)$.

  • Benefits:

    • Allows optimal management and control of environmental conditions for all organisms.

    • Achieves high yield and efficient resource use $(8-10)$.

• Nutrient Input & Fish Feed:

  • Plant evapotranspiration drives water flow $(11)$.

  • Nutrients primarily enter via fish feed $(12, 13)$.

  • Supplementing fish waste water with missing nutrients is common $(8, 14, 15)$.

  • Fish feed composition affects nutrient availability $(13, 16)$ and environmental impacts $(17, 18)$.

• Fish Meal Issues

  • Fish feed relies heavily on fish meals as a key protein ingredient in intensive systems like RAS and aquaponics $(12, 19)$, impacting sustainability.

  • Considered a "gold standard ingredient" due to palatability, digestibility, amino acid profile, and low anti-nutritional factors $(20)$.

  • Problems:

    • Overexploitation of wild forage fish stocks.

    • Increasing prices $(19, 21, 22)$.

    • Undesirable sodium accumulation in the water, which is not beneficial for most freshwater crops $(12)$.

• Alternative Protein Ingredients:

  • Studies have tested animal by-products, plant-based, and insect-based meals to substitute fish meals $(22-24)$.

  • These can partially provide the protein required by Nile tilapia (Oreochromis niloticus) $(22-24)$.

  • However, non-insect-based, fishmeal-free diets don't always significantly decrease environmental impact compared to biotic resources use $(26)$.

• Insect Meals (Black Soldier Fly - BSF):

  • Presented as a sustainable alternative ingredient for the future of fish feed $(27)$.

  • BSF (Hermetia illucens) is promising due to its high content of protein, lipids, minerals, and similar essential amino acid patterns to fish meal $(28-30)$.

  • BSF meal is easily digested by tilapia, improving feed use efficiency and nutrient availability in the water $(16, 31-34)$.

  • BSF can be fed a vast range of substrates (e.g., food waste, municipal sewage sludge, and fish sludge), making its production more eco-friendly $(35-38)$.

  • Using fish sludge discharged from the RAS unit to produce BSF would boost aquaponics circularity $(24)$ and system sustainability.

• Limited Research on BSF in Aquaponics:

  • Some studies investigate BSF frass (mineral-rich by-product) as a source of supplementary nutrients $(39, 40)$.

  • Others use BSF meal in fish feed, evaluating fish growth and nutrient release in RAS, discussing results from a decoupled aquaponics perspective $(16, 24, 41)$.

  • No scientific research has been reported on reusing BSF meal-based fish waste water in the hydroponics unit of a decoupled aquaponic system.

• Study Aim: Compare the effect of using fish waste water from Nile tilapia culture fed with fish meal-based diet (FM) and BSF meal-based diet on lettuce (Lactuca sativa) growth and resource use (water and inorganic fertiliser) in decoupled aquaponic systems.

Material and methods

• Lettuce Choice: Lettuce was selected because it's one of the primary plants scientifically and commercially produced in aquaponics $(4)$.

• Experimental design and setup: The experiment was conducted at the Leibniz Institute of Freshwater Ecology and Inland Fisheries (IGB, Berlin, Germany) for 35 days in a climate-controlled growth chamber.

  • Target room temperature of $18^(circ)C$ and relative humidity of 60% controlled by an integrated cooling system.

  • The chamber hosted nine experimental tents (Royal Room C120S, 120x60x180 cm), each equipped with a hydroponics AeroFlo10 unit (Growstream1, GHE) and a light-emitting diode (LED) lighting system.

  • The hydroponics unit consisted of two gutters (110 x 50 x 50 cm) with five plant-sites each, connected to a 45 L nutrient solution reservoir by a pump (compactON 1000, Eheim GmbH, Germany).

  • The nutrient solution was constantly oxygenated using compressed air and air stone diffusers.

  • The lighting system composed of one LED lamp (SANlight Q4WL S2.1 Gen2, 165W) positioned 90 cm to the top of the hydroponic unit and programmed to supply $91 (μmol \cdot m^{-2} \cdot s^{-1})$ photosynthetic photon flux density (PPFD) at plant level for 12h each day.

  • Experiments used digital timers to automatically switched the lights on in every tent at 6 am and off at 6 pm (LOGILINK ET0007 timer, digital, max 1800 W).

  • An automated fan (PK125 EC-TC Prima Klima) was introduced in every tent to mitigate possible lighting heating, with a target air temperature of $22^(circ)C$ and humidity of 60%, using minimum speed of 10% and maximum speed of 20%.

  • Air temperature and relative humidity in the tents were monitored by data loggers (HOBO UX100-011).

• Treatments: Lettuce growth responses were evaluated using three nutrient solution sources, assigned by a completely randomised experiment:

  • Conventional hydroponics solution (HP, control).

  • Fish waste water solutions from tilapia culture fed with FM-based diet.

  • Fish waste water solutions from tilapia culture fed with BSF meal-based diet.

• All treatments were tested in triplicate using the nine experimental units.

• The fish water was taken from six experimental recirculating aquaculture systems (RAS) at IGB. The RAS were stocked with juvenile Nile tilapia (Oreochromis niloticus) that were reared for 56 days on two experimental diets differing by the type of protein ingredient: an entirely FM-based diet and an entirely BSF meal-based diet.

• Lettuce Material and Growth Conditions:

  • Lettuce (Lactuca sativa L., Aquino RZ cv., Rijk Zwaan; De Lier, The Netherlands) seeds were sown in stone-wool cubes (4 cm, Rockwool1, Grodan,The Netherlands) and germinated during 14 days under controlled environmental conditions in a climate chamber (temperature: 18˚C, PPFD: 250 μmol m-2 s-1, and humidity: 90% for 2 days and 60% for the remaining 12 days).

  • Then, ten plants were allocated in every hydroponics unit at a plant density of 18 plants m-2 and produced for 35 days until harvest.

  • The plants were irrigated with a continuous flow (1.5 L min-1, over 24h) and had a light interval of 12h of light followed by 12h of darkness.

• The solutions were changed twice during the experiment, on the 14th and 27th day. On both occasions, the "old" nutrient solution was pumped out and replaced with a "fresh" nutrient solution specific to the treatment being tested.

Nutrient solutions preparation

• Two days prior to each reservoir filling procedure, samples of tap water and fish waste water from both aquaponics treatments (FM and BSF) were analysed for their nutrient concentrations using inductively coupled plasma-optical emission spectrometry (ICP-OES) and continuous flow analysis (CFA).

• The results of these analyses were used to calculate the amount of inorganic fertiliser to be added in the form of stock solutions so that all solutions had a similar nutrient profile, targeting the Howard-Resh recipe for hydroponics lettuce nutrient solution (165 NO3-N, 15 NH4-N, 50 P, 210 K, 45 Mg, 190 Ca, 65 S, 4 Fe, 0.1 Zn, 0.5 B, 0.5 Mn, 0.1 Cu, 0.05 Mo, 0 Na, Si and Cl–all values in mg L-1).

• For the control treatment (HP), the nutrient solution was prepared with 50% tap water and 50% distilled water and added nutrients.

• The FM treatment nutrient solution was prepared with waste water from the RAS rearing tilapia fed with FM-based diet plus the addition inorganic fertilisers. For the BSF treatment, nutrient solution was prepared with waste water from the RAS rearing tilapia fed with a BSF meal-based diet plus inorganic fertiliser.

• 135 L of fish waste water used in the hydroponics units of each aquaponics treatment were collected from three RAS-corresponding treatments (45 L per RAS replicate) three times during the experiment: prior to day 0, day 14, and finally, day 27.

• The amount of mineral nutrients added to the water sources as fertiliser mix was calculated using Hydrobuddy v1.91 program, formulating A+B concentrated stock solutions composed of multiple nutrients. For that, we fed the Hydrobuddy program with information on the target nutrient profile in the final solution, the nutrient profile of the water sources (Table 2), and the composition of the fertilisers available at IGB. For stock solution A, the following fertilisers were used: Potassium Nitrate–KNO3, Ammonium nitrate–NH4NO3, Iron EDDHA–FeEDDHA, Yara Calcium Nitrate–Yara_Ca(NO3)2, Magnesium Nitrate Solution–Mg(NO3)2.6H20; for stock solution B: Potassium Monobasic Phosphate–KH2PO4, Boric Acid–H3BO3, Zinc Sulphate (dihydrate)–ZnSO4.2H20, Manganese Sulfate (monohydrate)–MnSO4.H20, Copper Sulfate (pentahydrate)–CuSO45H20.

• 135 L of solution per treatment was considered in every reservoir filling procedure, and the concentration factor and volume of each A or B stock solution were set to 100 and 1.35 L, respectively. The water parameters for the final nutrient solution were set at EC 1.6 mS cm-1 and pH 6.0.

  • Separate fertilisers were weighted and added to the respective six beakers (A and B stock solutions vs three treatments).

  • The beakers were filled with 1 L distilled water and stirred until all the salts were dissolved.

  • Each beaker was then filled with distilled water to a final volume of 1.35 L, representing a 100-fold concentration (i.e. 1% of the nutrient solution volume).

  • Afterwards, the stock solutions A and B were added to the HP, FM and BSF tanks and adjustments were made to reach a pH value between 6.1 and 6.3, using nitric acid or sodium hydroxide to decrease or increase the pH, respectively.

• Once all the nutrient solutions were ready, samples were collected in duplicate for analysing nutrient content by ICP-OES and CFA methods.

• Then, nutrient solutions were distributed from the 135 L tanks to the 45 L replicate reservoirs.

Plant measurements

• Relative leaf chlorophyll concentration (soil plant analysis development—SPAD value) was measured in every lettuce plant one week before the harvesting (day 28) using a Chlorophyll Meter SPAD-502Plus device (Konica Minolta, Japan).

  • The SPAD value was taken in five leaves per plant.

  • The device calculated automatically the mean SPAD value of the five leaves resulting in the mean SPAD value of the lettuce plant.

• At the end of the experiment, all lettuce heads were harvested by cutting the organ directly above the stone-wool cube.

  • Immediately after harvesting, lettuce heads were weighted individually using precision balance (Kern EWJ, Reichelt elektronik GmbH, Sande, Germany) to determine the fresh weight.

  • Then, four lettuce heads per replicate were randomly selected to count the number of leaves and prepare subsamples for further analyses.

  • The subsamples were prepared by sampling one-quarter of each selected plant, weighted, placed into plastic bags, and stored at $-80^(circ)C$.

  • Every subsample was weighted again using analytical balance (Sartorius CP225D, Sartorius AG, Goettingen, Germany) to determine the dry weight and to calculate it for the whole lettuce head (multiplying sample value by 4).

Water measurements and resource use

• The abiotic parameters of the nutrient solutions, pH, temperature, electrical conductivity (EC), and dissolved oxygen (DO) concentration, were measured in all replicates on all weekdays during the experimental period.

• The pH, EC and temperature of the nutrient solutions were measured with the Hach Lange HQ40d probe. The DO concentration was measured using a dissolved oxygen meter, OxyGuard Handy Polaris.

• Water consumption was measured three times at every water exchange moment and at the end of the experiment.

• These data were used to estimate the volume of water needed per plant per day $V<em>w$ by summing the differences of the initial $V</em>0$ and final $V<em>f$ volume of the nutrient solution reservoirs of every replicate in the three measuring points (i = on day 14, 27 and 35): $V</em>w = \sum(V<em>{0,i} − V</em>{f,i})$.

• For nutrient analysis, all water samples were collected in duplicate, filtered with a 0.2 μm cellulose acetate membrane filter (GE Healthcare, United Kingdom), preserved by adding 150 μL of 2M of hydrochloric acid (HCl) to 12 mL filtered sample and stored at 4^[circ}C for subsequent analysis.

• The samples were analysed right after the experiment ended.

• The concentrations of phosphorus (P), potassium (K), calcium (Ca), magnesium (Mg), sulphur (S), iron (Fe), boron (B), manganese (Mn), copper (Cu), zinc (Zn), sodium (Na), silicon (Si) and aluminium (Al) were determined by the Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES) using the Thermo Scientific iCAP 7400 ICP-OES (Thermo Fisher Scientific Inc., USA).

• The continuous flow analysis (CFA) was performed using the FSR Seal High Resolution AA3 chemical analyser (Seal Analytical, Germany) to determine nitrate (NO3-N) and ammonium (NH4-N) concentrations.

• In addition to the samples of tap water and fish waste water, water samples were collected for characterising the fresh nutrient solutions on days 0, 15 and 26, and every week after the water exchange procedure in all replicates on days 6, 21, and 33.

• The amount of inorganic fertiliser (IF, in grams) needed to prepare a suitable nutrient solution for lettuce production was estimated by averaging the differences in the initial and final amount of all nutrients per treatment at all measuring points (i.e. on days 14, 27, and 35), using the following equation: $IF<em>j = \sum [(n</em>{0,i} \cdot V<em>{0,i})–(n</em>{f,i} \cdot V<em>{f,i})] / N</em>{mp}$;

  • Where:

    • $IF_j$ = amount of inorganic fertiliser in the j treatment.

    • $n_{0,i}$ = concentration of each nutrient (n) in the initial water source.

    • $n_{f,i}$ = final concentration of each nutrient (n) in the fresh solution (after adding the inorganic fertiliser).

    • $N<em>{mp}$ = number of measuring points, i.e. reservoir filling procedures ($N</em>{mp} = 3$).

• Using the results obtained from the equation, the proportion between the total amount of inorganic fertiliser needed to produce a control hydroponic nutrient solution, compared to FM and BSF, was calculated.

• The average percentage difference between the aquaponics and control HP treatments for every macronutrient (NO3-N, NH4-N, K, P, Ca, Mg, and S) in the form of inorganic fertiliser was calculated by: $Dif_IF<em>n ($

  • Where:

    • $Dif_IF_n$ is the percentage difference of the amount of inorganic fertiliser for the specific macronutrient n.

    • $IF_{n,j}$ is the amount (g) of the inorganic fertiliser for the specific macronutrient n in the j treatment.

    • $IF_{n,control}$ is the amount (g) of the inorganic fertiliser for the specific macronutrient n in the control HP treatment.

Data processing and statistical analysis

• Data were pre-processed using Microsoft Excel and analysed in MATLAB (ver. 2020b, "Statistics and Machine Learning" toolbox, TheMathWorks Inc., Portola, CA, USA).

• Shapiro-Wilk’s and Levene’s tests were applied to all data to check the normality and homogeneity of variances, respectively.

• Once the assumptions were met, a one-way ANOVA was used to determine whether the treatment affected the measured response variables: lettuce growth parameters, SPAD values, and water consumption.

• The data on the amount of inorganic fertiliser was not Gauss-distributed, thus it was subjected to nonparametric Kruskal-Wallis and Dunn’s tests to observe whether the medians varied significantly between the different treatments.

• All data were analysed at 5% significance level.

• Descriptive statistics were used to present the results for abiotic conditions and nutrient concentrations in nutrient solutions.

• In the tables, values represent means ± standard deviations.

Results

• The use of waste water from tilapia culture fed with a FM-based diet and BSF meal-based diet resulted in similar lettuce yields compared to conventional hydroponic nutrient solution (HP, control treatment).

• The lettuce fresh and dry weight, the number of leaves and SPAD values were statistically not significant different among all treatments ($\alpha$ 0.05).

• The same was found for water use per plant (p > 0.05).

• The amount of fertiliser needed, however, differed between treatments, being approximately 32% lower in FM and BSF compared to HP.

• The abiotic growth conditions in the tents were kept at the desired values in all tents, with mean air temperature and relative humidity of 22.1 \pm 1.0^[circ}C and $62.7 \pm 6.0$, respectively.

• The water quality condition, with values of electrical conductivity (EC), temperature, pH, and dissolved oxygen, in general, comparable between treatments

• For the nutrient concentration in the nutrient solutions over the experiment, no clear differences between the treatments were observed.

• The $Na$ concentrations showed remarkable differences, as the values were lower in HP, followed by BSF and higher in FM, with mean values of $25.8 \pm 4.2 mg \cdot L^{-1}$ $Na$ in the HP treatment, $46.0 \pm 2.9 mg \cdot L^{-1}$ in the BSF treatment, and $59.4 \pm 7.9 mg \cdot L^{-1}$ in the FM treatment.

Discussion

• Main Finding: Waste water from tilapia culture in RAS fed with either fish meal-based diet (FM) or Black Soldier Fly meal-based diet (BSF) can be used as a base for preparing a nutrient solution for lettuce cultivation, as an alternative to conventional hydroponic nutrient solutions (HP).

• Lettuce Growth: FM and BSF treatments showed no statistical difference in lettuce fresh and dry weight, number of leaves, and SPAD values, supporting the idea that alternative feed ingredients like BSF meal don't have obvious effects on aquaponics lettuce production.

• Comparison to Hydroponics: No plant growth differences were found when comparing aquaponics and control HP treatments.

  • Some studies suggest aquaponics outperforms hydroponics due to plant growth promoters from RAS (humic organic matter, beneficial microorganisms) $(14, 42)$, but these findings align with recent studies showing comparable plant growth in decoupled aquaponic and conventional hydroponic systems $(10, 15, 43, 44)$.

• Abiotic Conditions: All abiotic condition parameters were kept within limits for lettuce production $(45, 46)$ without affecting plant growth or results.

• Resource Optimization: Achieving similar yields to standalone hydroponic systems is desirable in aquaponics, as the primary goal is to optimize resource use $(47)$.

  • Vegetable production in decoupled aquaponics can reduce global warming potential and turn the water footprint of vegetable produce from positive to negative $(48)$.

• Resource Input: Overall resource input was lower in aquaponics compared to hydroponics without fish feed composition (FM or BSF) having an effect.
  Plant Water Use:* Similar in all treatments, suggesting the nutrient solution source didn't affect plant water uptake

• Aquaponics is more efficient in water use than hydroponics since it mostly relies on fish waste water, where hydroponics relies on fresh water.

• Inorganic Fertiliser Use: Both aquaponics nutrient solutions saw a 32% reduction in inorganic fertiliser addition, meaning at least 32% of required nutrients were derived from fish water.

  • Saved inorganic fertiliser amount can be increased depending on system type, fish stocking density, amount of feed, and daily water exchange rate $(14, 49-51)$.

• Inorganic fertiliser production contributes to greenhouse gas emissions by releasing nitrous oxide during manufacturing

• Sodium Concentration BSF-fish waste water contains considerably less sodium compared to FM

  • Safe $Na$ concentration in hydroponic nutrient solution is around $50 mg L^{-1}$ $(13)$.

  • $Na$ concentrations can easily reach high levels in aquaponics depending on the recirculation method used in the hydroponic unit.

  • It is known that salt stress reduces the growth, photosynthesis, and stomatal conductance of lettuce $(55, 56)$.

Conclusion

• The use of insect-based fish feed in decoupled aquaponic systems is a resource-efficient alternative to produce lettuce without a reduction in yield.

• Using waste water from tilapia culture in RAS fed with fish meal-based diet (FM) and Black Soldier Fly meal-based diet (BSF) in decoupled aquaponic systems resulted in similar lettuce yields compared to conventional hydroponic nutrient solution (HP), as well as no adverse effects on lettuce growth were observed.

• Moreover, FM and BSF resulted in approximately 32% lower inorganic fertiliser uses compared to HP.

• BSF-based diet is a promising alternative to FM-based diet in aquaponics, particularly for commercial, intensive applications.

• Our study supports the potential of using insect-based fish feed as an effective circular way to reduce the reliance on fish meal-based diets in aquaculture/aquaponic systems and the use of inorganic fertiliser for hydroponics systems.