The optimal formulation of a readily compostable horticultural growing substrate for vertical farming was determined using Design of Experiments

doi:10.21203/rs.3.rs-4351439/v1

A novel, optimized, polysaccharide and biochar-based, compostable hydrogel horticultural growing substrate for use in hydroponics and vertical farming was created based upon empirical methods and statistical design of experiments (DOE). A 15-run D-optimal mixture DOE was completed that increased the 2-week plant growing ability of a five-component hydrogel nearly ten-fold from 4.3695 g to 41.2623 g per 100 plants. The data were analyzed using a standard least squares method with an effect screening emphasis, and a model was created that maximized the signal to noise ratio. There was a good correlation between the measured and predicted values of the model, with an r-squared value of 0.90. The predictions of efficacy and compostability were confirmed with subsequent experiments that showed the hydrogel was composted in less than 12 weeks and that the plant growth predicted by the model differed from the experimental growth by 0.65%. The resulting optimized formulation had a high fertilizer content for a growth medium. We therefore suggest that an empirical approach to formulation research can produce superior outcomes with a statistically designed study.

Physical sciences/Materials science/Soft materials/Gels and hydrogels

Biological sciences/Plant sciences/Plant biotechnology/Field trials

Earth and environmental sciences/Environmental sciences/Environmental impact

Plant based foods can be grown outdoors on naturally irrigated land, outdoors on artificially irrigated land, indoors using artificial irrigation and natural light (for instance, in a greenhouse or hoop farm), or indoors using both artificial irrigation and light (in a vertical farm)¹. All these approaches have their advantages and disadvantages. However, vertical farming is becoming increasingly economically viable as Earth’s population and average temperatures continue to increase.² Vertical farming uses far less water, provides more optimal conditions for plants to grow, and uses less land than any other farming technique.³ Therefore, there is increasing interest in creating vertical farms near population centers.⁴ However, while much attention has been paid to optimizing the growing conditions of vertical farms by controlling the atmosphere,⁵ photosynthetic active radiation (PAR),⁶ and hydroponic solution attributes,⁷ far less attention has been paid to optimizing the horticultural growing substrates.⁸ The state of the art in horticultural growing substrates is dominated by rockwool,⁹ peat,¹⁰ and coconut coir.¹¹ Rockwool is made of mined materials that are neither biodegradable nor compostable.¹² Peat is harvested from the environment, and while it is compostable, it is not readily renewable (as peat can take 1000 years to regenerate naturally).¹³ Coconut coir is also compostable but can transfer pests, bacteria, and fungi to vertical farms.¹⁴ The ideal substrate would contain renewable ingredients, be readily compostable, pest-free, and grow plants well.¹⁵ We report here the development of seaweed-based,¹⁶ compostable, pest-free substrate that grows plants well.

We endeavored to find a plant-based gelling agent that was beneficial for crops, compostable,¹⁷ easily renewable,¹⁸ and with enough structural integrity to work as a horticultural growing substrate for 1-2 months.¹⁹ Various candidates were screened, including agar,²⁰ gellan gum,²¹ chitosan²² and carrageenan.²³ K-carrageenan emerged as the best candidate because it had strong gelling characteristics²⁴ and was a known plant growth promoter.²⁵ Furthermore, as many fertilizers are cationic,²⁶ and K-carrageenan gel strength was increased in the presence of cations²⁷ there was a good overlap in function.

The other ingredients studied were fertilizers, a buffer, and a biochar purification agent. Murashige & Skoog plant growth medium was selected as the fertilizer because it supported the growth of a wide variety of crops.²⁸ Calcium citrate was chosen as the buffer because it was known to be compatible with Murashige and Skoog fertilizers.²⁹ Finally, activated charcoal derived from biochar wood was designated as the purification agent because it was also a known plant growth promoter.³⁰

Optimization of a five-component hydrogel by changing one variable at a time (OVAT) would have been at best inefficient and at worst would have failed to find the optimal formulation.³¹ Conversely, DOE has emerged as a technique that maximizes the value of research and produces more robust formulations.³² We have found that by varying the concentrations of ingredients with wide ranges we could find optimized formulations with unusual properties. Of course, as we were conducting mixture DOE, all the components were restricted to adding up to 100%.

As the goal of this research was to produce a growing medium useful for today’s vertical farms, the DOE was tailored to maximize the growth of crops common to them (lettuces, kale, bok choy, and spinach).³³ The state of the art in vertical farming varies widely from company to company and there are many watering systems that are employed. Some of the most common hydroponics systems are ebb and flow, nutrient film technique (NFT), vertical drip tower, and deep-water culture.³⁴ Therefore, we only studied the initial 14-day propagation stage of plant growth to make our results more universal. We know that plants that are larger at 14-days are more likely to grow faster in the production phase because they have more leaf surface area.³⁵

Likewise, the conditions chosen for the plant propagation study were chosen to be consistent with best practices: humidity domes for germination,³⁶ a commonly used fertigation solution,³⁷ and LED lighting with appropriate cycling.³⁸

Therefore, a five-component formulation of established plant growth promoters was created for species commonly grown in vertical farms using hydroponics best practices. We were able to optimize the formulation using only 15 experiments because of the efficiency of DOE. The purpose of this paper is to serve as a model for the rapid research and development of plant growing media.

Statistical Design of Experiments (DOE). The formulation was optimized to maximize the plant growth of leafy greens in 15 experiments using 1500 samples (15 half-trays of 200 x 1.75” grow plugs). Statistical Design of Experiments (DOE) using JMP software (licensed from SAS) was utilized to assess the effects of changing multiple formulation components simultaneously and to ensure product quality and process capability. A mixture design was selected which constrained the ingredients to add up to 100%. JMP recommended a 15-run study and a D-optimal design (D efficiency = 0.003277). A D-optimal design was appropriate for this application because the desire was to execute an inexpensive study that would serve as a template for the design of several future products. Furthermore, a D-optimal design was suitable because this was a mixture design, and the parameters of study were tightly constrained. The 15-runs were compared to a previously developed hydrogel formulation control and a coconut coir control.

Raw materials. All raw materials were purchased as food grade. Purified water was generated in-house by treating municipal water by reverse osmosis just as one would do at a vertical farm. Refined K-carrageenan was purchased from W Hydrocolloids. Acidic biochar, Murashige & Skoog basal medium, and citric acid were purchased from Sigma-Aldrich.

Plant substrate formulation. The horticultural substrates were made in 2L batches by adding the proportions listed in Table 1 by mass. RO water was added to a 3L beaker on a hotplate and overhead stirring was initiated at 400 RPM. The solid ingredients were mixed by hand in a separate beaker and then added portion wise to avoid agglomeration. Heating was initiated on the hotplate to warm the mixture to 85ºC. When the mixture was substantially homogeneous, stirring was slowed to 100 rpm. The mixture was maintained at 85ºC +/- 3ºC for 30 minutes, at which time the heat was turned off and the mixture allowed to cool to less than 60ºC. The mixture was poured into a 200 x 1.75” tray that had the bottom holes plugged with a silicone mat. The plugs were allowed to cool to room temperature and then they were dibbled with a pen in preparation for seeding.

EC and pH measurement. The EC and pH of prepared gels were measured by placing an EC/pH probe (Extech EC-500) inside a gel plug and waiting for 30 seconds or until the measured value stabilized.

Plant growth studies. The performance of the formulations was evaluated for the growth and health of five species (N=1500): fairly butter lettuce and F1 crystal lettuce (purchased from Paramount Seeds Company), and bok choy, spinach, and kale (purchased from Johnny’s Selected Seeds). 200 x 1.75” trays and the appropriate humidity domes were purchased from T.O. Plastics. During germination the plugs were placed under a humidity dome and spritzed with fertigation solution daily with an EC of 0.3 mS/cm. Lighting in the vertical nursery carts was set to 250 PAR and with a 20/4-hour day/night cycle. Climate in the nursery carts ranged from 65-82°F with a relative humidity of 50-90%. After 7 days, the humidity domes were removed. At 14 days the plants were cut at crown level, grouped by species, and immediately weighed for wet mass.

Compostability study: Compostability was measured per ISO 20200 (lab scale disintegration). The composting vessels were 5 liters in size (14” X 8” X 4.5”) and were made from non-degradable plastic. The vessels’ lids had 6 latches and a gasket to ensure a tight seal. Per the method, for aeration a 5 mm diameter hole was drilled on either end of the container, approximately 6.5 cm from the bottom. The containers were incubated in an incubator that kept the temperature at 58 +/- 2 ºC throughout the 12 weeks of composting.

As inoculum, a synthetic compost recipe was made using rabbit feed, mature compost, corn starch, saccharose, corn seed oil, urea, saw dust, and wood chips (30%, 10%, 10%, 5%, 4%, 1%, 40% by mass, respectively). This feedstock was inoculated using a 3-month-old mature compost from the Monterey District composting facility. The compost was sieved through a 9.5 mm sieve and then mixed. The amount of urea was adjusted to achieve a C:N ratio of 28:1 to 32:1 and the appropriate volume of water was added to adjust the moisture content to 55% by weight.

Into each vessel 1 kg of the inoculum described above was added, then 500g of the test sample was added as whole plugs. This was mixed well to spread the compost evenly throughout the container. The mixture was composted for 12 weeks, and the temperature was controlled at 58 +/- 2 ºC. The compost was controlled and monitored using the ISO 20200 procedure listed in Table 1. Observations were periodically recorded. At the end of 12 weeks the material was emptied from the composting vessels and screened through a 2 mm sieve. The test was considered passed if no more than 10% of the original dry weight of the product was retained on the sieve. The compost made from these disintegration studies was mixed with control compost for the quality of compost test (ecotoxicity).

Table 1: The ISO 20200 list of reactor control and measurement used for the compostability study.

Time (days)	Operation
0	Initial mass of reactor was recorded
1,2,3,4,7,9,11,14	Reactor was weighed and water was added to restore the initial mass. The reactor was mixed.
8,10,16,18,21,23,25,28	Reactor was weighed and water was added to restore the initial mass. The reactor was not mixed.
30-60	Reactor was weighed and water was added to restore the mass to 80% of the initial mass. The reactor was mixed.
60-84	Reactor was weighed and water was added to restore the mass to 70% of the initial mass. The reactor was mixed.

Quality of compost test (ecotoxicity): The quality of compost produced from the plant growing substrate was compared to control compost following the OECD Guideline 208 with the modifications found in Annex E of EN 13432³⁹. In brief, pots with clear plastic covers were added the compost mixtures to be studied (0% control, 25% and 50% by weight compost mixed with soil). Two species of plants were used for the study, barley, and cucumber. 100 cucumber seeds were planted in each dilution. 50 barley seeds were planted in each dilution. Triplicates of each dilution were made. Percent germination was determined using percent germination compared to the control as 100 percent. The dry weight of the plants was used to determine biomass.

The formulation components effects on plant growth. After the final masses of the plants were measured in the plant growth studies, JMP was used to fit a model using standard least squares analysis with an emphasis on effect screening. The model effects were each of individual formulation components crossed with the mixture. The least squares fit found that all the individual formulation components had a significant effect on plant growth with >95% confidence. The actual versus predicted plot for plant mass had an R² = 0.90. The order of ingredient importance was water>biochar>carrageenan>calcium citrate>Murashige & Skoog with log Worth values of 6.2, 5.4, 4.9, 2.1, and 1.3, respectively (log Worth is -log of P value). The JMP prediction profiler used the model to maximize predicted plant mass. The model’s predicted optimized formulation is included in Table 2 and Figure 1 along with the control formulations and the 15 DOE runs.

Table 2: Comparison of the compositions and performance of the 15-run DOE formulations to the gel control coir control and optimized formulation (predicted) and average of the confirmation run.

Run	Water	K-carrageenan	Biochar	Murashige & Skoog	Calcium Citrate	14-day plant mass (g)
1	0.981	0.01	0.001	0.004	0.004	1.6486
2	0.98	0.017	0.001	0.002	0	25.3899
3	0.98	0.01	0.004	0.002	0.004	13.9062
4	0.982	0.01	0.004	0.004	0	35.9626
5	0.981	0.01	0.001	0.004	0.004	1.5914
6	0.987	0.01	0.001	0.002	0	22.4868
7	0.98	0.011	0.001	0.004	0.004	1.5302
8	0.98	0.01	0.004	0.002	0.004	10.5598
9	0.98	0.012	0.004	0.004	0	30.9997
10	0.98	0.017	0.001	0.002	0	26.0163
11	0.987	0.01	0.001	0.002	0	35.0747
12	0.987	0.01	0.001	0.002	0	35.9909
13	0.98	0.01	0.004	0.002	0.004	19.5847
14	0.982	0.01	0.004	0.004	0	43.8156
15	0.98	0.017	0.001	0.002	0	18.3488
Gel control	0.97925	0.015	0.003	.00275	0	4.3695
Coir control	N/A	N/A	N/A	N/A	N/A	28.2394
Optimized (predicted)	0.984	0.01	0.004	0.002	0	40.9936
Confirmation of optimized	0.984	0.01	0.004	0.002	0	41.2623

Confirmation run of the model’s predicted optimized formulation. One 100 kg confirmation batch of the optimized formulation recommended by the DOE model was completed and a portion of it was tested for plant growth. The EC and pH of the batch were measured to be 4.26 and 6.99, respectively. Three side-by-side trials of 20 plants of each species (tray size 200 x 1.75”) were subjected to the same plant growth testing as completed in the DOE. The average plant mass per tray was found to be 41.2623 g (SD=1.4016 g) whereas the statistical model predicted 40.9936 g of plant mass per tray (a percent error of 0.65 %).

Compostability study. The plant growing substrate was found to be 100% disintegrated after the test compost was passed through a 2 mm sieve per ISO 20200.

Quality of compost study (ecotoxicity). The plant growing substrate was found to have no significant effect on the germination rate nor the biomass versus control for barley and cucumber (see Table 3). In the barley with 25% compost from the disintegration test there was a slight decrease in the rate of percent germination versus control and percent biomass versus control (97.9% and 92.0%, respectively), however this effect was not dose-dependent because when 50% compost was used with the barley the percent germination and percent biomass were like control soil.

Table 3: The effect of test compost on germination and biomass in the growth of barley and cucumber.

Species	% Compost from disintegration test	% Germination rate of control	% Biomass of control
Barley	25	97.9 +/- 1.2	92.0 +/- 2.0
Barley	50	100.7 +/- 3.3	100.2 +/- 10.1
Cucumber	25	99.7 +/- 0.6	99.2 +/- 1.5
Cucumber	50	98.9 +/- 6.9	98.9 +/- 6.9

The purpose of this study was to determine whether a mixture DOE could be used to find the optimal formulation of a five-component, compostable, plant-based plant growing substrate for use in hydroponics systems and vertical farms. A 15-run D-optimal study was created using the JMP Custom Designer software platform. A D-optimal study was recommended by JMP, and this was appropriate because the mixture design was necessarily constrained by the fact that all the formulation components must have added up to 100% and was further constrained by the ranges of ingredient percentages that were workable for this application: that too high of polymer percentage would create a mixture that was not able to be stirred nor poured, that too high a fertilizer concentration would create a mixture that was too salty for plants, that too high a citric acid concentration would lower the pH too low for plant growth, and that too much biochar would create a substrate that had weak structural characteristics. There was many other possible DOE design optimalities that could have been chosen including A- and G- optimal, however, for a highly constrained mixture design D-optimal tends to give superior results with the minimal number of resources allocated.

The D-optimal mixture design recommended by JMP minimized the variance of the constrained design space by distributing the prescribed mixture fractions of the 5 ingredients throughout the design space. Of course, the variance is always lowest near the center of the design space, and one could decrease the variance near the extremes of the design space by adding tested ingredient fractions outside the design space via design augmentation. We should comment that the fractions recommended by the model are at the extremes of the design space for 4 of the 5 mixture components (K-carrageenan, biochar, Murashige & Skoog, and calcium citrate) and therefore further refining of this model is desirable before finalization of this formulation. However, it should be noted that there was a strong agreement between the predicted plant growth of the optimized formulation by the model and the average plant growth of the test batch, with percent error <1%. This was impressive considering the tests could not be run simultaneously and because plant growth over a 14-day period was necessarily variable. Moreover, Run 14 (Table 2), was like the optimized formulation and demonstrated a similar level of plant growth in 2 weeks. Had the confirmation batch produced a substantially different plant growth to the model prediction then design augmentation would have been even more appropriate.

The DOE test batches were compared with a gel control, which was the original gel substrate formulation that this study was attempting to improve. One can see from Table 2 that the plant mass at 2 weeks was increased nearly 10-fold from 4.3695 g to 41.2623 g per 100 plants. This demonstrated how sub-optimal the original gel substrate formulation was. It also showed that with only 15 test batches substantial refining of a formulation was possible using a DOE mixture design.

The confirmation batch had a relatively high EC of 4.26, which was about twice the EC of a typical hydroponics solution. This result supports the theory that effective mixture DOE models rely upon purposely exiting the first principles regime and moving instead toward an empirical approach. In other words, the purpose of a study is to maximize, minimize, or match the target value of a chosen experiment response without regard to the reasoning behind the result. An empirical approach to DOE was therefore more likely to be successful.

The biodegradability testing results were consistent with what we predicted as for the disintegration test the only portion of the substrate that needed to break down was the crosslinked K-carrageenan, a polysaccharide derived from seaweed. Therefore, it is reasonable that a compost mixture containing microorganisms would be able to disintegrate this hydrogel.

The quality of compost study (exotoxicity) result was consistent with our predictions, because K-carrageenan is used as a plant growth stimulator, Murashige & Skoog is an established fertilizer and biochar is known to be compatible with a wide variety of plants.

A 5-component mixture was rapidly optimized using a 15-experiment mixture DOE made with JMP software. The substrate that was created by this study was made from biocompatible and compostable ingredients, and therefore the mixture was also biocompatible and compostable as no chemical reaction occurred during the process except for the gelling of the K-carrageenan polysaccharide. The optimized substrate was confirmed to grow plants at nearly 10 times the rate of the previous gel substrate formulation using the same ingredients and was confirmed by compostability studies to disintegrate readily and to not be toxic to plants afterwards. Because the optimized formulation component fractions were at the edges of the design space for 4 of the 5 ingredients, augmentation of the DOE is recommended before finalizing this formulation.

Author Contribution

T.J.C. conceived and designed the experiments, conducted the JMP statistical analysis, and wrote the main manuscript text.J.M.R. conducted the plant studies wrote the associated text.I.N.S. made the plant substrate formulations, collected formulation data, and wrote the plant formulation text.M.G. conducted the biodegradability study and wrote the associated text.All authors reviewed the manuscript.

Data Availability

The datasets generated during the current study are available from the corresponding author on reasonable request.

Ampim, P. A. Y., Obeng, E. & Olvera-Gonzalez, E. Indoor Vegetable Production: An Alternative Approach to Increasing Cultivation. Plants 11, 2843 (2022).
O'Sullivan, C. A. et al. Vertical farms bear fruit. Nat Biotechnol. 38(2), 160-162 (2020).
Dutta, M. et al. Evaluation of Growth Responses of Lettuce and Energy Efficiency of the Substrate and Smart Hydroponics Cropping System. Sensors. 23(4), 1875 (2023).
Yuan, G. N. et al. A review on urban agriculture: technology, socio-economy, and policy. Heliyon 8(11), e11583 (2022).
Fleisher, D. H., Cavazzoni, J., Giacomelli, G. A. & Ting, K. C. Adaptation of SUBSTOR for controlled-environment potato production with elevated carbon dioxide. Trans ASAE 46(2), 531-538 (2003).
Zhen, S. & Bugbee, B. Far-red photons have equivalent efficiency to traditional photosynthetic photons: Implications for redefining photosynthetically active radiation. Plant Cell Environ. 43(5), 1259-1272 (2020).
Mourantian, A., Aslanidou, M., Mente, E., Katsoulas, N. & Levizou, E. Basil functional and growth responses when cultivated via different aquaponic and hydroponics systems. PeerJ. 11, e15664 (2023).
Blok, C., Jackson, B. E., Guo, X., de Visser, P. H. B. & Marcelis, L. F. M. Maximum Plant Uptakes for Water, Nutrients, and Oxygen Are Not Always Met by Irrigation Rate and Distribution in Water-based Cultivation Systems. Front Plant Sci. 8, 562 (2017).
Dannehl, D., Suhl, J., Ulrichs, C. & Schmidt, U. Evaluation of substitutes for rock wool as growing substrate for hydroponic tomato production. J. Appl. Bot. Food Qual 88, 68-77 (2015).
Allaire-Leung, S. E., Caron, J. & Parent, L. E. Changes in physical properties of peat substrates during plant growth. Canadian Journal of Soil Science 79(1): 137-139 (1999).
Noguera, P., Abad, M., Noguera, V., Puchades, R. & Maquieira, A. Coconut Coir Waste, A New and Viable Ecologically-Friendly Peat Substitute. Acta Hortic. 517, 279-286 (2000).
Reddy, N. & Crohn, D. M. Effect of composted greenwaste and rockwool on plant growth of okra, tomato, and chili peppers. Compost Science & Utilization, 26(4), 217-224 (2018).
Van Seters, T. E. & Price, J. S. The impact of peat harvesting and natural regeneration on the water balance of an abandoned cutover bog, Quebec. Hydrological processes 15(2), 233-248 (2001).
Kumara, A. D. N. T., Chandrashekharaiah, M., Kandakoor, S. B. & Chakravarthy, A. K. Status and management of three major insect pests of coconut in the tropics and subtropics. New horizons in insect science: Towards sustainable pest management, 359-381 (2015).
Verdonck, O., De Vleeschauwer, D. & De Boodt, M. The Influence of The Substrate to Plant Growth. Acta Hortic. 126, 251-258 (1982).
Newlove, A. A., Mahunu, G. K., Akbar, A. & Shakeel, A. Extraction, structural properties, and applications of carrageenan gum in Natural Gums (ed. Shakeel, A. & Akbar, A.) 647-668 (Elsevier, 2023).
Salma, S., Nurida, N. L. & Dariah, A. Bio-decomposer of seaweed composting In IOP Conference Series: Earth and Environmental Science, 637(1), 012080 (IOP Publishing, 2021).
Jenkins, E. H. & Street J.P. Manure from the sea. No. 193. (Connecticut Agricultural Experiment Station, 1917).
Berton, S. B. R. et al. Properties of a commercial κ-carrageenan food ingredient and its durable superabsorbent hydrogels. Carbohydrate research 487, 107883 (2020).
Barrett-Lennard, E. G. & Dracup, M. A porous agar medium for improving the growth of plants under sterile conditions. Plant Soil 108, 294–298 (1988).
Giavasis, I., Harvey, L. M. & McNeil, B. Gellan Gum. Critical Reviews in Biotechnology 20(3), 177-211 (2000).
Coelho, N. & Romano, A. Impact of chitosan on plant tissue culture: recent applications. Plant Cell Tiss. Organ Cult. 148, 1–13 (2022).
Carmo, L. P., Moura, C. W. N. & Lima-Brito, A. Gelation of Culture Medium with K-Carrageenan Improves and Reduces the Cost of in vitro Propagation of Comanthera mucugensis (Giul.) LR Parra & Giul. Brazilian Archives of Biology and Technology 66, e23230191 (2023).
Rey, D. K. & Labuza, T. P. Characterization of the effect of solutes on the water‐binding and gel strength properties of carrageenan. Journal of Food Science 46(3), 786-789 (1981).
Pacheco, D. et al. Seaweeds’ carbohydrate polymers as plant growth promoters. Carbohydrate Polymer Technologies and Applications 2, 100097 (2021).
Arnon, D. I. Microelements in culture-solution experiments with higher plants. American Journal of Botany, 322-325 (1938).
Zabik, M. E. & Aldrich, P. J. Gel strength of kappa‐carrageenan as affected by cations. Journal of Food Science 33(4), 371-377 (1968).
Murashige, T. & Skoog, F. A revised medium for rapid growth and bioassays with tobacco tissue cultures. Physiol. Plant 15, 473-497 (1962).
Said, E., Gaffar, A. & Murashige, T. Continuous cultures of tomato and citron roots in vitro. In vitro 15(8), 593-602 (1979).
Thomas, T. D. The role of activated charcoal in plant tissue culture. Biotechnology advances 26(6), 618-631 (2008).
Garg, R. K. & Singhvi, I. Optimization techniques: an overview for formulation development. Asian J. Pharm. Res. 5(3), 217-221 (2015).
Fukuda, I. M., Pinto, C. F. F., Moreira, C. D. S., Saviano, A. M. & Lourenço, F. R. Design of experiments (DoE) applied to pharmaceutical and analytical quality by design (QbD). Brazilian journal of pharmaceutical sciences 54 (2018).
Beacham, A. M., Vickers, L. H., & Monaghan, J. M. Vertical farming: a summary of approaches to growing skywards. The Journal of Horticultural Science and Biotechnology 94(3), 277-283 (2019).
Van Gerrewey, T., Boon, N. & Geelen, D. Vertical farming: The only way is up? Agronomy 12(1), 2 (2021).
Erickson, R. O. Modeling of plant growth. Annual Review of Plant Physiology 27(1), 407-434 (1976).
Murashige, T. Plant propagation through tissue cultures. Annual review of plant physiology 25(1), 135-166 (1974).
Patil, S. T., Kadam, U. S., Mane, M. S., Mahale, D. M. & Dekale J. S. Hydroponic nutrient solution: A review. Journal of Pharmacognosy and Phytochemistry 9(5), 2095-2099 (2020).
Wong, C. E., Teo, Z. W. N., Shen, L. & Yu, H. Seeing the lights for leafy greens in indoor vertical farming. Trends in Food Science & Technology 106, 48-63 (2020).
OECD, Test No. 208: Terrestrial Plant Test: Seedling Emergence and Seedling Growth Test in OECD Guidelines for the Testing of Chemicals, Section 2 (OECD Publishing, Paris, 2006)

No competing interests reported.

The optimal formulation of a readily compostable horticultural growing substrate for vertical farming was determined using Design of Experiments

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Results

Discussion

Conclusion

Declarations

Author Contribution

Data Availability

References

Additional Declarations

Status:

Version 1