Optimisation of seaweed-based biostimulant application to improve bell pepper yield and quality using Topsis-Shannon entropy

Abstract

The effectiveness of seaweed-based biostimulants is strongly influenced by application concentration and phenology-based timing. This study aimed to optimise these factors to improve the yield and fruit quality of bell pepper (Capsicum annuum L.) using the Technique for Order Preference by Similarity to Ideal Solution combined with Shannon entropy (TOPSIS–Entropy). An Ecklonia maxima-based biostimulant was applied at five concentrations (0–0.8%) and at different phenological stages: one day before transplanting (ODBT), fourteen days after transplanting (14DAT), fruit set, and fourteen days before harvest (14DBH), applied individually or in combination, with no use of fertilisers before and during the experiment. Analysis of variance indicated significant effects of concentration and application timing on yield and fruit quality parameters. The application of 0.4% extract at 14DAT + fruit set + 14DBH achieved the highest overall performance, with the greatest TOPSIS–Entropy closeness coefficient (0.77) compared with the control (0.46). This treatment increased yield from 2.01 to 4.43 kg m⁻² and improved fruit quality, resulting in higher TSS/TA ratio (1.85 vs. 1.26), and increased total phenolic content (238.9 vs. 173 mg 100 g⁻¹ fresh weight). The optimised application strategy consistently enhanced both productivity and internal fruit quality, indicating its potential as an efficient management practice in bell pepper production. Future research should focus on elucidating physiological mechanisms and validating performance under diverse environmental and field-scale conditions.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-56037-z.

Introduction

Bell pepper is a high-value vegetable crop of the Solanaceae family¹, cultivated worldwide for its economic and nutritional importance². It is an essential fruit in the human diet due to its high content of vitamins (A, C, and E) and bioactive compounds such as carotenoids, flavonoids, and phenolics³, which contribute to antioxidant activity and support human health^4,5. Owing to its nutritional benefits, consumer demand for bell peppers has increased production⁶. According to FAOSTAT⁷, global production of Capsicum spp reached about 36.97 million tons, produced on roughly 2.02 million hectares. However, the yield and qualities of bell pepper are affected by environmental conditions such as extremely high or low temperatures⁸ and water stress⁹, which can reduce fruit set¹⁰ and induce physiological disorders, including blossom end rot and sunscald¹¹. Furthermore, pest infestations (aphids, thrips, and whiteflies) and diseases such as bacterial spot and Phytophthora blight significantly constrain bell pepper production and fruit marketability¹².

To overcome these challenges, synthetic agrochemicals have been widely used to enhance yield and quality¹³. However, their overuse has raised concerns regarding environmental sustainability and human health risks¹⁴. The use of agrochemicals has contributed to water pollution¹⁵, soil degradation^16,17, cancer and neurological disorders in humans^18,19, and loss of biodiversity²⁰. This has driven the need for eco-friendly, sustainable cultural strategies that can improve crop productivity and quality, with minimal adverse effects on the ecosystem and human health²¹. Studies have shown that the use of biostimulants plays an essential role in sustainable agriculture^22–24.

Biostimulants are substances other than fertilisers and pesticides that can improve crop growth, yield, and quality²⁵. The biostimulants are grouped into seaweed extracts, microorganisms, humic and fulvic acids, protein hydrolysates, silicon, and polymers^26–28. Among these, seaweed-based products are widely recognised for their beneficial effects on plant growth, yield, and stress tolerance^29,30. They are also known to enhance nutrient uptake and stimulate the production of bioactive compounds, contributing to plant health and the quality of the produce³¹. Among seaweed-based biostimulants, the Ecklonia maxima-based biostimulant is popular among growers globally^32–34. Ecklonia maxima, a brown Kelp, is mainly found and harvested in South Africa for the commercial biostimulant Kelpak^®. Unlike other seaweed-based biostimulants, Ecklonia maxima is extracted using a non-chemical method called cell burst³⁵, suggesting that the extract composition is minimally altered, including their efficacy^22,30,36. Ecklonia maxima extract contains phytohormones, including brassinosteroids, auxins, gibberellins, cytokinins, abscisic acid, and polyamines, as active ingredients^35,37. These compounds improve crop growth and productivity by modulating hormonal signaling and metabolic pathways^38–40. For instance, auxins accumulate in root tips where they activate signalling pathways that stimulate meristematic cell division, cell elongation, and lateral roots initiation, thereby increasing root surface area and nutrient uptake^41–43, which enhances vegetative growth and yield potential³¹. Cytokinins promote cell division and leaf expansion by regulating cell cycle genes⁴⁴, while delaying senescence by preserving chlorophyll pigments⁴⁵, thereby increasing photosynthetic capacity and assimilate production for fruit development^46,47. ABA-like substances regulate stomatal behaviour and nutrient remobilisation⁴⁸, supporting stress tolerance, flowering, fruit set, and maturation⁴⁹. Therefore, the seaweed extract affects crop growth and yield through the collective coordinated hormonal and metabolic effects of the phytohormones it contains.

According to the manufacturer’s recommendation, the Ecklonia maxima extract can be applied at concentrations ranging from 0.2 to 1%, as a root dip before transplanting, as a foliar spray at the 4–5 leaf stage, and once or twice at 14–21 days after transplanting, depending on the crop⁵⁰. Previous studies have demonstrated that the response to Ecklonia maxima extract varies with concentration and application timing. For instance, in blueberries, applications of 0.2 and 0.4% Ecklonia maxima extract at full bloom, early green fruit, and fruit colouring stages increased yield; however, 0.4% significantly outperformed 0.2%⁵¹. Moreover, Wszelaczyńska et al.⁵², observed that a single foliar application of 0.2% Ecklonia maxima extract at the 4-leaf stage significantly improved the nutritional content of carrots compared to repeated applications of 0.3%. Applying E. maxima extract (0.4%) 2 h before transplanting, followed by 3 foliar applications at 21-day intervals after transplanting, significantly enhanced bell pepper (F1 hybrid) yield grown inside a greenhouse⁵³. However, application of 1% Ecklonia maxima extract 3 h before transplanting, followed by 0.5% applied 3 times at 15-day intervals after transplanting, did not affect the yield of Capsicum annuum (F1 hybrid) grown inside a plastic tunnel compared with the untreated⁵⁴). Although these studies differ in plant species and environmental conditions, they collectively suggest that the effectiveness of seaweed-based biostimulants strongly depends on application concentration and timing¹¹⁰. Despite this, current application strategies are mainly based on trial-and-error practices and/or on the selection of treatments using single-response mean comparisons. These approaches do not simultaneously consider multiple yield and quality parameters. To date, there is limited information on objective optimisation of both concentration and application timing of seaweed E. maxima extract using a systematic, data-driven approach. This represents a significant gap that limits the reliability and consistency of biostimulant use in practice.

This need is especially evident in bell pepper, where both yield components and fruit quality traits must be optimised simultaneously to meet agronomic performance and consumer nutritional expectations. The yield performance of bell pepper is largely determined by fruit number per plant, average fruit weight, and fruit size⁵⁵. Whereas fruit quality is commonly evaluated using biochemical and physicochemical indices, including total phenolic content, total soluble solids, titratable acidity, and chlorophyll content⁵⁶. Total phenolic content represents a major group of antioxidant compounds that contribute to human health benefits and antioxidant capacity of pepper fruits⁵⁷. Total soluble solids (TSS) estimate soluble sugar concentration and are widely used as an indicator of fruit sweetness and consumer acceptability. Moreover, the titratable acidity of bell pepper quantifies organic acid concentration and, together with TSS, determines the sugar-acid balance, which influences fruit flavour⁵⁸. These quality parameters are widely used in Capsicum research because they respond sensitively to crop management and environmental factors, providing reliable indicators of nutritional and market quality. However, improvements in yield traits may come at the expense of biochemical quality attributes due to physiological trade-offs in resource allocation within the plant⁵⁹.

As such, identifying the best application strategy is a multi-criteria decision problem rather than a single-variable optimisation. Multi-criteria decision-making (MCDM) methods are well-suited to this complexity⁶⁰ and have been successfully used to evaluate trait importance and rank treatments⁶¹. Among these, the Shannon entropy method assigns weights based on data variability⁶², while TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) ranks treatments based on their proximity to an ideal solution⁶³. However, TOPSIS alone requires subjective weighting, which can lead to bias results⁶⁴. Combining Shannon entropy with TOPSIS overcomes this issue and yields a more accurate, objective ranking⁶². This TOPSIS-Shannon entropy model has already been validated in horticultural research, such as optimising edible coatings to improve strawberry shelf-life⁶⁵. However, it has not yet been applied to optimise biostimulant concentration and application timing in bell pepper production systems, even though these factors are critical to treatment efficacy. We hypothesise that different combinations of Ecklonia maxima extract concentration and application timing will significantly influence bell pepper yield and fruit quality, and that the TOPSIS-Shannon entropy model can be used to objectively identify the best treatment. To address this, the present study aimed to identify the optimal concentration and application timing of seaweed-based biostimulant for bell pepper (Capsicum annuum L.) by (i) measuring defined yield and quality traits across the 25 treatment combinations, (ii) deriving objective weights for those traits using Shannon entropy and, (iii) ranking treatments with TOPSIS to select the single treatment that maximises yield and fruit quality.

Materials and methods

Experimental site description

The field experiment was conducted from November 2024 to January 2025 at the University of Limpopo Research Unit (23˚ 55’ 30.08” S; 29˚ 28’ 17.45” E), South Africa. Four-week-old bell pepper seedlings (cv Star Jupiter) were bought from a commercial nursery (23˚ 48’ 40.66” S; 30˚ 10’ 28.17” E), Tzaneen, Limpopo Province, South Africa. The bell pepper fruit were harvested 60 days after transplanting, thereafter, transported to Plant Production Laboratory (23˚ 53ʹ36.1ʺS 29o 44ʹ28.9ʺ E) at the University of Limpopo for quality assessment.

Treatments, experimental layout, and agronomic details

The experiment was laid out in a randomised complete block design with four replicates. Blocking was used to account for field variability, particularly slope, which can influence water distribution and potentially affect the results. The treatment factors included Ecklonia maxima-based biostimulant Kelpak^® at five concentrations (0, 0.2, 0.4, 0.6 and 0.8%) and application timing based on bell pepper phenology: [1] one day before transplanting (ODBT), [2] 14 days after transplanting (14DAT), [3] ODBT+14DAT+ fruit set (FS), [4] 14DAT + FS + 14 days before harvest (14DBH) and [5] ODBT + FS + 14DBH with four replicates. Each replicate consisted of 5 bell pepper plants, for a total of 20 per treatment. The plants were 40 cm apart within rows and 50 cm between rows, resulting in a planting density of 5 plants m^− 2. The soil was sandy loam (19.2% clay, 6.4% silt, and 74.4% sand), with a pH (H₂O) of 6.7, pH (KCl) of 4.6, ammonium of 0.38 mg kg^− 1, nitrate of 0.49 mg kg^− 1, organic carbon of 0.25%, and electrical conductivity of 38.2 µs/cm. The weather conditions at the experiment site are presented in Table 1.

Table 1.

Weather conditions at the experimental site.

Month	Average daily temperature (°С)	Precipitation (mm)	Humidity (%)
November	21	56.5	57
December	22	121.0	65
January	22	134.0	66

The plots (2 m × 1.6 m) were irrigated with tap water using a 500 mL plastic beaker at 1-day intervals. Weeds in the plots were controlled manually using hand hoes. No fertilisers were applied both prior to and during the experiment, including during soil preparation. Efekto ^® cutworm bait (active ingredient: sodium fluorosilicate) was applied according to the manufacturer’s recommendations to control cutworm after they were noticed during scouting. Furthermore, rust, early blight, and bacterial symptoms were identified^66,67, which was followed by the application of Mycoguard 720^® (active ingredient: chlorothalonil) and Tenazole^® (active ingredient: tebuconazole) according to the manufacturer’s recommendations. The respective Ecklonia maxima extract concentrations, mixed with surfactant Tween 20 (0.5%)⁶⁸, were sprayed on the leaves until runoff⁶⁹. Control crops were sprayed with distilled water containing 0.5% Tween 20. The Tween 20 was used to increase the wettability of the treatments⁷⁰. Approximately 40 days after flowering, bell pepper fruit were harvested once at the green mature stage at the termination of the experiment. There were no staggered or successive pickings; consequently, all yield and fruit quality assessments were based solely on this single, final harvest. Yield was measured from all four replicates per treatment (five plants per replicate), using the full experimental unit for yield determination. For quality analysis, three replicates were used per treatment, with three fruits sampled per replicate, resulting in a total of nine fruits analysed per treatment.

Data collection

Determination of yield

Yield was determined as the average fruit weight per treatment at harvest using a bench scale (Bench Scale LBX 6, Adam Equipment Co. Ltd., Kingston, UK) and expressed as kg m^− 2.

Determination of fruit colour and pigments

Fruit colour was measured using a chromameter (CR-410 Chroma meter, Konica Minolta Inc, Osaka, Japan) on two opposite points of the fruit. The chromameter was calibrated with a standard white tile before use. The data obtained from the chromameter were represented as a* value, which indicates the degree of redness when positive (+ a*) and greenness when negative (-a*), b* value indicating the degree of yellowness, L* indicating lightness, chroma (c^*) representing colour saturation, and hue angle (h°) representing the colour difference⁷¹. Pigments were determined by homogenising 1 g of fresh pepper fruit with 10 mL of 80% v/v acetone, then keeping the mixture on ice for 10 min. Thereafter, centrifuged at 5000 rpm for 10 min (± 25 ˚С). The absorbance for chlorophyll a, chlorophyll b, and carotenoids was read at 663 nm, 646 nm, and 470 nm, respectively, using a spectrophotometer (Jenway, Staffordshire, United Kingdom) (Ali et al.⁷²,. The chlorophyll content and carotenoids were calculated using the following equations, expressed as mg g^{− 1} FW:

Determination of TSS, TA, and TSS/TA ratio

TSS was determined using a hand-held refractometer (Atago, Tokyo, Japan) and expressed in ^oBrix⁷³. Titratable acidity of fruit was determined by titration with 0.1 M sodium hydroxide (NaOH) and expressed as % citric acid⁷⁴. The TSS/TA ratio was determined by dividing the TSS (^oBrix) by the TA (%) value for each sample as described by Tigist et al.⁷⁵.

Determination of total phenolic content

The total phenolic content was determined using the Folin-Ciocalteu reagent, following the methodology outlined by Sari et al.⁷⁶. Briefly, 1 g of a fresh fruit sample was homogenised with 10 mL of 80% methanol using a mortar and pestle. Thereafter, the solution was vortexed, and the supernatant was collected and used for analysis. The 18.5 µL aliquot of the extract was pipetted into a microplate, followed by 36.4 µL of Folin-Ciocalteu reagent. Thereafter, 145.4 µL of 7% sodium carbonate was added to the mixture. The final solution was incubated in the dark at room temperature for 2 h. Thereafter, the absorbance was read on a microplate reader at 750 nm. A standard curve for gallic acid was prepared at concentrations of 0, 0.25, 0.5, 1, and 2 µg/mL (R² = 0.997). The total phenolic content was quantified using the equation:

The total phenolic content was expressed as milligrams of gallic acid equivalents per gram of milled sample (mgGAE g^− 1 FW).

Data analysis

The data were tested for normality and homoscedasticity using the Shapiro-Wilk and Levene’s test, respectively. Where the assumptions of normality and homogeneity were met, data were subjected to analysis of variance (ANOVA) to assess the effects of factor levels and their interactions on the dependent variables. Kelpak^® concentration, application timing, and their interaction were used as sources of variation. Therefore, a factorial ANOVA model was employed for the analysis. Where a significant difference was found, treatments were subjected to post hoc analysis using Tukey’s Honest Significant Difference (HSD) test, due to its robustness in reducing the risk of Type I error⁷⁷. All analyses were computed using Genstat software (version 18, VSN International, United Kingdom). The significance level was set at p < 0.05. Additionally, Microsoft Excel version 16.69 was used to run the TOPSIS-Shannon entropy analytical model to optimise Ecklonia maxima extract concentration and application timing according to the steps enumerated by Tiwari et al.⁷⁸, as described below.

1. The development of a decision-making matrix consisting of treatments (m) and parameters (n).

2. The normalisation of decision-making matrix.

   

   Where:

   Xij: normalised value.

   m: total number of evaluated treatments.

   n: total number of parameters.

3. To determine the weights of the parameters and construct the normalised weight matrix.

   

   Where:

   Vij: weighted normalised value.

   Wij: weight of the parameter.

   rij: normalised value for the evaluated objects.

   The weights of each parameter (yield, a*, b*, c*, L*, TSS, TA, TSS/TA ratio, total chlorophyll, total carotenoids and total phenolics) was determined using Shannon entropy method⁷⁹ using the following steps:

   • 3.1.
     Designing the decision-making matrix.
   • 3.2.
     Designing the normalised matrix.
     
   • 3.3.
     Calculating the entropy for each parameter:
     
     Ej: entropy value parameter.
     K: constant factor.
     M: total number of treatments.
     Pij: normalised probability value.
   • 3.4.
     Calculating the distance for each parameter from the entropy.
     
   • 3.5.
     Calculating the weights for each parameter.
     

4. Calculating the positive and negative solutions.

   Yield, TSS, TSS/TA, total chlorophyll, total carotenoids, total phenolic content, b*, C* and L* value were considered positive parameters (where higher values are preferable), while TA and a* value were considered negative parameters (where lower values are preferred).

   

   

   Where:

   = maximum value for a beneficial parameter and minimum value for a non-beneficial parameter.

   = minimum value for beneficial parameter and maximum value for non-beneficial parameter.

5. Determining the distance of the normalised weighted matrix from ideal positive and negative points.

   

   

   Where:

   Si⁺ is the distance from the positive ideal.

   Si⁻ is the distance from the negative solution.

6. Calculating the distance from the ideal point.

7. 

8. Ranking the treatments.

Treatments were ranked by scores, where the treatments with the highest Ci+ value identified the performing treatment. The TOPSIS-Shannon entropy model was further cross-validated using Multi-Attributive Border Approximation Area Comparison (MABAC) according to the steps enumerated by Wang et al.⁸⁰, (supplementary). Furthermore, a sensitivity analysis was conducted to assess the robustness of the model through ± 10–20% weight perturbation (Deterministic sensitivity)⁸¹ and equal weight scenario (Benchmark Test)⁸².

Results and discussion

Effect of Ecklonia maxima extract concentration and application timing on the yield of bell pepper

The efficacy of the Ecklonia maxima extract on yield was not significantly influenced by the interaction of concentration and application time (p > 0.05) (Table 2). Although treated crops showed generally higher yields than the control, not all increases were statistically significant. Yield ranged from 2.01 kg m^{− 2} in the control to 6.40 kg m^{− 2} in the E. maxima-treated crops. This yield increase is agronomically significant, as it demonstrates that applying E. maxima can improve productivity in commercial bell pepper production. This improvement could be attributed to the synergistic effects of the phytohormones present in the seaweed extract, which have been shown to enhance nutrient uptake, photosynthesis, and fruit development through hormonal regulation⁸³. To the best of our knowledge, this is the first study to determine whether Ecklonia maxima extract’s efficacy on crop yield is influenced by application timing and concentration in bell pepper production. Previous studies have focused primarily on whether the efficacy of Ecklonia maxima extract is affected by concentration, without considering crop physiology during application. Hence, Mbandlwa et al.⁵⁴ applied 1% Ecklonia maxima extract before transplanting and repeated the treatment at 0.5%, 15, 30, and 45 days after transplanting. Although application timing and concentration were not considered fixed factors, the authors found that biostimulant-treated and control crops did not differ significantly in yield. Biostimulants are usually effective at low concentrations^84–86. Therefore, a higher concentration of Ecklonia maximaextract applied frequently to the crop might have an inhibitory effect or no effect, as found by Mbandlwa et al.⁵⁴,.

Table 2.

Effect of Ecklonia maxima extract application timing and concentration on the yield of bell pepper fruit harvested 2 months after transplanting.

Application time	Concentration	Yield (kg m− 2)
Control	Control	2.01 ± 0.32b
ODBT	0.2	3.92 ± 0.45ab
ODBT	0.4	4.70 ± 1.29ab
ODBT	0.6	3.37 ± 0.84ab
ODBT	0.8	2.92 ± 0.49ab
Control	Control	2.01 ± 0.32b
14DAT	0.2	2.97 ± 1.42ab
14DAT	0.4	2.88 ± 0.59ab
14DAT	0.6	2.35 ± 0.71ab
14DAT	0.8	1.78 ± 0.17b
Control	Control	2.01 ± 0.32b
ODBT+14DAT + FS	0.2	3.06 ± 0.64ab
ODBT+14DAT + FS	0.4	6.40 ± 1.16a
ODBT+14DAT + FS	0.6	2.91 ± 0.59ab
ODBT+14DAT + FS	0.8	3.26 ± 0.68ab
Control	Control	2.01 ± 0.32b
14DAT + FS+14DBH	0.2	2.78 ± 1.19ab
14DAT + FS+14DBH	0.4	4.43 ± 0.34ab
14DAT + FS+14DBH	0.6	5.13 ± 0.75ab
14DAT + FS+14DBH	0.8	4.60 ± 0.53ab
Control	Control	2.01 ± 0.32b
ODBT + FS+14DBH	0.2	4.49 ± 0.54ab
ODBT + FS+14DBH	0.4	4.88 ± 1.16ab
ODBT + FS+14DBH	0.6	4.42 ± 1.28ab
ODBT + FS+14DBH	0.8	2.57 ± 0.69ab
Significant level
A		***
B		*
A*B		ns

A-concentration; B-application time; ODBT-one day before transplanting; 14DAT-14 days after transplanting; FS-fruit set; 14DBH-14 days before harvest. NS, *,*** are non-significant or highly significant at p < 0.05, P < 0.01, and p < 0.001, respectively. Mean (n = 100) values with the same letters are not significantly different according to Tukey’s HSD test at p < 0.05.

Effect of Ecklonia maxima extract concentration and application timing on colour and pigments of bell pepper

Ecklonia maxima extract application timing, concentration, and their interaction did not significantly (p > 0.05) affect bell pepper fruit colour (a*, b*, and L*) (Table 3). This suggests that the extract does not affect bell pepper fruit colour. Similar findings were observed in tomato fruit: Ascophyllum nodosum extract (0.3%) applied 3 times at bi-weekly intervals, starting 21 days and 7 days after transplanting, respectively, had no significant effect on colour^85,87. Moreover, Ascophyllum nodosum extract (0.2–0.6%) applied 3 times at 21-day intervals from the 4–5 leaf stage to the beginning of flowering had no significant effect on the external colour of bell pepper⁸⁸. Rouphael⁸⁹, also found that Ecklonia maxima extract (0.3%) applied at a 2-week interval, commencing 10 days after transplanting and continuing for a total of 5 applications throughout the growing period, did not affect the external colour of zucchini squash.

Table 3.

Effect of Ecklonia maxima extract concentration and application timing on colour and pigments of bell pepper at harvest.

Treatments		Parameters
Application time	Concentration	a*	b*	L*	Total chlorophyll (mg g− 1 FW)	Total carotenoids (mg g− 1 FW)
Control	Control	−5.25 ± 0.81a	5.60 ± 0.91a	51.20 ± 0.23a	6.41 ± 0.05h−k	3.02 ± 0.02abc
ODBT	0.2	−4.61 ± 1.14a	3.87 ± 2.06a	50.28 ± 1.23a	8.87 ± 0.07bc	2.81 ± 0.03a−d
ODBT	0.4	−5.62 ± 1.02a	3.84 ± 0.23a	50.56 ± 0.22a	8.57 ± 0.05cd	2.94 ± 0.02a−d
ODBT	0.6	−4.79 ± 0.55a	4.55 ± 0.90a	50.70 ± 0.60a	5.94 ± 0.69jkl	3.18 ± 0.25ab
ODBT	0.8	−3.93 ± 0.41a	2.21 ± 0.50a	50.38 ± 0.32a	3.87 ± 0.07m	1.37 ± 0.08f
Control	Control	−5.25 ± 0.81a	5.60 ± 0.91a	51.20 ± 0.23a	6.41 ± 0.05h−k	3.02 ± 0.02abc
14DAT	0.2	−5.02 ± 0.35a	4.31 ± 0.51a	50.71 ± 0.55a	5.69 ± 0.04kl	2.01 ± 0.02c−f
14DAT	0.4	−5.34 ± 0.94a	4.99 ± 1.52a	51.83 ± 0.97a	5.46 ± 0.05l	1.62 ± 0.04ef
14DAT	0.6	−4.71 ± 0.54a	3.10 ± 1.37a	50.47 ± 1.01a	6.75 ± 0.03g−j	2.49 ± 0.13b−e
14DAT	0.8	−4.33 ± 0.29a	3.34 ± 0.51a	50.35 ± 0.67a	7.64 ± 0.18ef	3.16 ± 0.07ab
Control	Control	−5.25 ± 0.81a	5.60 ± 0.91a	51.20 ± 0.23a	6.41 ± 0.05h−k	3.02 ± 0.02abc
ODBT+14DAT + FS	0.2	−4.60 ± 0.33a	3.45 ± 0.67a	50.07 ± 0.17a	4.58 ± 0.09m	2.30 ± 0.02b−f
ODBT+14DAT + FS	0.4	−4.88 ± 0.49a	4.32 ± 0.64a	51.44 ± 0.37a	8.08 ± 0.06cde	3.18 ± 0.03ab
ODBT+14DAT + FS	0.6	−5.83 ± 0.74a	6.58 ± 2.07a	52.14 ± 1.17a	6.28 ± 0.05i−l	2.53 ± 0.03a−e
ODBT+14DAT + FS	0.8	−4.59 ± 0.87a	3.81 ± 1.60a	50.74 ± 0.86a	6.90 ± 0.18f−i	2.66 ± 0.03a−e
Control	Control	−5.25 ± 0.81a	5.60 ± 0.91a	51.20 ± 0.23a	6.41 ± 0.05h−k	3.02 ± 0.02abc
14DAT + FS+14DBH	0.2	−5.82 ± 0.89a	6.29 ± 1.56a	51.91 ± 1.21a	5.90 ± 0.02jkl	1.90 ± 0.02def
14DAT + FS+14DBH	0.4	−5.80 ± 1.25a	6.31 ± 1.37a	50.93 ± 1.09a	6.59 ± 0.3g−j	2.10 ± 0.02d−f
14DAT + FS+14DBH	0.6	−4.73 ± 0.64a	4.35 ± 1.33a	51.25 ± 0.33a	10.16 ± 0.21a	3.56 ± 0.04a
14DAT + FS+14DBH	0.8	−5.15 ± 0.57a	4.33 ± 1.29a	51.72 ± 0.70a	7.43 ± 0.03efg	2.41 ± 0.03b−f
Control	Control	−5.25 ± 0.81a	5.60 ± 0.91a	51.20 ± 0.23a	6.41 ± 0.05h−k	3.02 ± 0.02abc
ODBT + FS+14DBH	0.2	−5.24 ± 0.71a	5.73 ± 1.54a	51.80 ± 0.87a	7.22 ± 0.03e−h	2.85 ± 0.02a−d
ODBT + FS+14DBH	0.4	−4.68 ± 0.07a	4.01 ± 0.17a	50.71 ± 0.27a	9.46 ± 0.11ab	2.85 ± 0.09a−d
ODBT + FS+14DBH	0.6	−6.05 ± 0.64a	6.40 ± 1.27a	52.55 ± 0.84a	7.06 ± 0.03f−i	3.24 ± 0.06ab
ODBT + FS+14DBH	0.8	−4.84 ± 0.78a	4.13 ± 1.33a	51.34 ± 0.58a	7.74 ± 0.05def	2.44 ± 0.10b−e
Significant level
A		ns	ns	ns	***	***
B		ns	ns	ns	***	**
A*B		ns	ns	ns	***	***

A-application timing; B-concentration; ODBT-one day before transplanting; 14DAT-14 days after transplanting; FS-fruit set; 14DBH-14 days before harvest. ns, **,*** are not significant or highly significant at p < 0.01 and p < 0.001, respectively. Mean (n = 75) values with different letters are significantly different according to the Tukey’s HSD test at p < 0.05.

We further investigated the effects of Ecklonia maxima extract concentration and application timing on the pigments chlorophyll and carotenoids in bell pepper. Our findings revealed that the interaction of the extract’s concentration and application timing significantly (p < 0.001) affected fruit pigments (Table 3). Applying 0.6% at 14DAT + FS+14DBH resulted in the highest total chlorophyll content (10.16 mg g^{− 1} FW) compared to the control (6.41 mg g^{− 1} FW). This suggests that higher chlorophyll retention was not simply due to concentration alone, but rather due to the combined effect of a higher concentration applied at a physiological responsive stage, where nutrient uptake and photosynthetic regulation are more active. These results not only confirm that a seaweed-based biostimulant maintains chlorophyll content in bell pepper, as shown in cabbage and spinach⁹⁰ (Kulkarni et al., 2019), but also that the interaction of concentration and application timing influence this effect. Similarly, total carotenoid content was significantly influenced by the interaction between the extract’s concentration and application timing (Table 3). However, the differences between treatments were not statistically significant. Although the treatment (0.6% at 14DAT + FS+14DBH) showed a higher total carotenoid content (3.56 mg g-1 FW) than the control (3.02 mg g^{− 1} FW), the two were not statistically different. These results indicate that the E. maxima treatment did not significantly increase carotenoid content compared to control.

The bell pepper fruit in this study were harvested at the green stage, when chlorophyll dominates and carotenoids that drive the colour change to red, yellow, or orange hues accumulate in small amounts⁹¹. Therefore, the insignificant effect of the Ecklonia maxima extract concentration and application timing interaction on visual colour parameters (a*,b* and L*), is not contradictory to its significant impact on total chlorophyll, rather, it reflects that the treatment retained the bell pepper green colour, which is commercially desirable for peppers marketed at the green stage⁹². Moreover, the increase in total carotenoids, although not significant and sufficient to alter external colour, suggests that the E. maxima extract has the potential to improve the nutritional quality of bell pepper⁹³.

Effect Ecklonia maxima extract concentration and application timing on the sweetness of bell pepper

In bell pepper, sweetness is generally associated with high total soluble solids (TSS) content, moderate acidity, and an optimal TSS/TA ratio that enhances eating quality. In the present study, the interaction of Ecklonia maxima extract concentration and application timing significantly (p < 0.001) influenced TSS, TA, and TSS/TA ratio of bell pepper fruit (Table 4). The treatment 0.2% applied at 14DAT + FS+14DBH, had significantly higher TSS (3.79 ^oBrix) compared to the control (3.11 ^oBrix). In contrast, the treatment 0.8% applied at ODBT + FS+14DBH resulted in a significantly higher TA (4.77%) than all other treatments, including the control (2.48%). Furthermore, the TSS/TA ratio was significantly higher (1.85) in fruits produced by the treatment 0.4% applied at 14DAT + FS+14DBH, compared to all treatments. These results indicate that the interaction of applying a low concentration of Ecklonia maxima extract at these phenological stages, 14DAT + FS+14DBH, enhances bell pepper sweetness and improves the overall sugar-acid balance. Similar responses have been observed in zucchini squash, where Ecklonia maxima extract (0.3%) applied at 2-week intervals, starting 10 days after transplanting, for a total of 5 applications, improved sweetness⁸⁹. Moreover, Ascophyllum nodosum extract (0.25%) applied 4 times at a monthly interval starting from the day of transplanting significantly improved the sweetness (TSS/TA) of the strawberry⁹⁴. Preharvest application of Ascophyllum nodosum extract (0.5%) at 14 and 60 days after transplanting significantly increased tomato sweetness⁹⁵. The present study demonstrated that fruit sweetness and sugar-acid balance in bell pepper are influenced by the interaction between Ecklonia maxima extract concentration and application timing. This interaction showed that the beneficial effect of the extract depended on matching an appropriate concentration with the correct growth stage, rather than increasing concentration alone. Therefore, to improve the sweetness of bell pepper fruit, Ecklonia maxima extract can be applied at a low concentration at critical growth stages such as 14DAT + FS+14DBH.

Table 4.

Effect of Ecklonia maxima extract concentration and application timing on sweetness parameters (TSS, TA, and TSS/TA ratio) of bell pepper.

Treatments		TSS	Parameters
Application time	Concentration		TA	TSS/TA
Control	Control	3.11 ± 0.07bcd	2.48 ± 0.04ij	1.26 ± 01bcd
ODBT	0.2	2.01 ± 0.05hi	1.72 ± 0.14kl	1.29 ± 0.03bcd
ODBT	0.4	2.55 ± 0.03efg	2.08 ± 0.07jk	1.26 ± 0.04bcd
ODBT	0.6	2.32 ± 0.09f−i	4.32 ± 0.16ab	0.51 ± 0.06ij
ODBT	0.8	2.69 ± 0.04def	2.53 ± 0.03hij	1.19 ± 0.04cde
Control	Control	3.11 ± 0.07bcd	2.48 ± 0.04ij	1.26 ± 0.01bcd
14DAT	0.2	3.16 ± 0.13bcd	1.97 ± 0.06jk	1.60 ± 0.07ab
14DAT	0.4	2.14 ± 0.11ghi	1.66 ± 0.10kl	1.42 ± 0.02bc
14DAT	0.6	2.79 ± 0.08def	4.06 ± 0.15bc	0.70 ± 0.01g−j
14DAT	0.8	3.16 ± 0.13bcd	2.86 ± 0.06f−i	1.16 ± 0.01c−f
Control	Control	3.11 ± 0.07bcd	2.48 ± 0.04ij	1.26 ± 0.01bcd
ODBT+14DAT + FS	0.2	3.53 ± 0.07ab	2.75 ± 0.07ghi	1.26 ± 0.03bcd
ODBT+14DAT + FS	0.4	2.95 ± 0.11cde	3.13 ± 0.05e−h	1.03 ± 0.04c−h
ODBT+14DAT + FS	0.6	3.16 ± 0.13bcd	2.75 ± 0.02ghi	1.18 ± 0.03cde
ODBT+14DAT + FS	0.8	2.43 ± 0.05fgh	3.04 ± 0.05e−i	0.76 ± 0.02f−i
Control	Control	3.11 ± 0.07bcd	2.48 ± 0.04ij	1.26 ± 0.01bcd
14DAT + FS+14DBH	0.2	3.79 ± 0.06a	3.93 ± 0.09bcd	1.04 ± 0.03c−g
14DAT + FS+14DBH	0.4	2.69 ± 0.04def	1.25 ± 0.24l	1.85 ± 0.04a
14DAT + FS+14DBH	0.6	2.95 ± 0.11cde	3.30 ± 0.02de	0.99 ± 0.03d−h
14DAT + FS+14DBH	0.8	2.01 ± 0.07hi	3.42 ± 0.04de	0.63 ± 0.01hji
Control	Control	3.11 ± 0.07bcd	2.45 ± 0.04ij	1.26 ± 0.01bcd
ODBT + FS+14DBH	0.2	2.95 ± 0.11cde	3.13 ± 0.05e−h	1.05 ± 0.02e−g
ODBT + FS+14DBH	0.4	3.31 ± 0.08abc	3.64 ± 0.03cde	0.82 ± 0.03e−i
ODBT + FS+14DBH	0.6	1.92 ± 0.15i	3.96 ± 0.08bcd	0.32 ± 0.09j
ODBT + FS+14DBH	0.8	2.32 ± 0.09f−i	4.77 ± 0.14a	0.47 ± 0.06ij
Significant level
A		***	***	***
B		***	***	***
A*B		***	***	***

A-application timing; B-concentration; ODBT-one day before transplanting; 14DAT-14 days after transplanting; FS-fruit set; 14DBH-14 days before harvest. *** is highly significant at p < 0.001. Mean (n = 75) values with different letters are significantly different according to the Tukey’s HSD test at p < 0.05.

Effect of Ecklonia maxima extract concentration and application timing on total phenolic content of bell pepper fruit

The total phenolic content of bell pepper fruit was significantly (p < 0.001) influenced by the interaction of Ecklonia maxima extract concentration and application timing (Fig. 1). Applying 0.4% Ecklonia maxima extract at 14DAT + FS+14DBH resulted in fruit with significantly higher total phenolic content (238.9 mg 100 g^{− 1} FW) than all other treatments, including the control (173.6 mg 100 g^{− 1} FW). These results confirm that Ecklonia maxima extract enhances the antioxidant potential of the fruit, as shown previously in bell pepper fruit⁵⁴. Our results further indicate that concentration and application timing influence the efficacy of Ecklonia maxima extract on total phenolics. The phenolic content in bell pepper fruit decreases as the fruit matures^96,97. Therefore, our results suggest that to maintain the phenolic content, the extract should be applied repeatedly during the growing period. The mechanism behind the observed treatment effect on maintaining fruit phenolic content in this study is unclear. However, we assume that the extract stimulates key enzymes in the phenylpropanoid pathway, thereby enhancing phenolic biosynthesis⁹⁸. In particular, the application of 0.4% at fruit set may have promoted the biosynthesis of phenolic compounds by stimulating metabolic processes associated with fruit development⁹⁹; Elansary et al.^90,100. Furthermore, repeating the application 14 days before harvest may have modulated endogenous fruit hormones¹⁰¹, which, in turn, could have enhanced phenolic accumulation in fruit¹⁰². These findings show that the combined effect of concentration and application timing influences the extract’s efficacy in regulating phenolic accumulation, thereby enhancing bell pepper’s antioxidant potential.

Fig. 1.

Effect of Ecklonia maxima-based biostimulant concentration (B) and application timing (A) on total phenolic content of fruit. The vertical bars indicate the standard errors of means (n = 75). Different letters are significantly different at p < 0.05 according to Tukey’s HSD test at p < 0.05.

The optimisation of treatments using TOPSIS-Shannon entropy and its cross validation using Multi-Attributive Border Approximation Area Comparison (MABAC)

The Shannon entropy model was used to determine the weights of the parameters, and it revealed that the TSS/TA ratio carried the most significant weight (0.18) (Table 5), highlighting its significant contribution to bell pepper fruit quality. The TSS/TA of fruit is generally regarded as an essential parameter for assessing fruit maturity and sensory quality¹⁰³. Studies have demonstrated a significant correlation between fruit sweetness and quality^56,104, supporting the model used in our study, which indicates that TSS/TA is a significant determinant of bell pepper quality. In contrast, the a* value had the least weight (0.02), highlighting its least significant contribution to quality. The green colour, denoted by a negative a* value, is not considered a reliable quality or harvesting index in green-matured bell peppers because colour changes may occur post-harvest^105,106. These entropy-derived weights were then used in the TOPSIS model for the final treatment ranking, thereby eliminating subjective weights that could lead to biased results¹⁰⁷. To evaluate the robustness of the entropy-derived weighting structure, a sensitivity analysis was performed. The parameter weights were varied by ± 10% and ± 20%, followed by re-normalisation and recalculation of the TOPSIS rankings. Across all perturbation scenarios, the top ranked treatment remained unchanged, indicating that the ranking was stable around the original entropy-based weights. However, when a uniform weight was applied, the ranking order changed and a different treatment emerged as optimal. These differences demonstrated that the TOPSIS ranking outcomes are sensitive to the weighting scheme, emphasising the critical role of objective weight determination¹⁰⁸.

Table 5.

Parameters and calculated weights for each response by the Shannon entropy model.

Criterion	Weight
TSS/TA	0.18
Yield	0.13
Titratable acidity (TA)	0.13
Fruit yellowness (b*)	0.08
Total phenolics	0.08
Total carotenoids	0.07
Total chlorophyll	0.06
Total Soluble Solids (TSS)	0.06
Chroma (C*)	0.04
Lightness (L*)	0.04
Fruit greenness (a*)	0.02

Using the TOPSIS model, the performance of the Ecklonia maxima extract concentration and application timing combination treatments, expressed as closeness coefficients (Ci), was evaluated based on their distances from the positive ideal (Si+) and negative ideal (Si-) (Table 6). The positive ideal represented high yield and fruit quality, while the negative ideal corresponded to minimised yield and fruit quality. The treatments were then ranked from the highest closeness coefficient (ranked 1) to the lowest (ranked last¹⁰⁹. Based on the final ranking by TOPSIS, the optimum treatment was identified as 0.4% applied at 14DAT + FS+14DBH, which achieved the highest closeness coefficient (Ci = 0.72) (Table 6). This indicates that the treatment (0.4%, 14DAT + FS+14DBH) most closely approached the positive ideal solution (high yield and fruit quality), outperforming the other treatments. Our results indicate that the concentration and application timing of Ecklonia maxima extract can be optimised to improve its efficacy using the TOPSIS-Shannon entropy model. The model revealed that the Ecklonia maxima extract concentration for bell pepper should be 0.4%, applied at 14 days after transplanting, repeated at fruit set, and at 14 days before harvest (14DAT + FS+14DBH). To strengthen the robustness of these results, cross-validation was conducted using the MABAC model. The model ranks treatments similarly to the TOPSIS method. The MABAC model similarly ranked the 0.4% (14DAT + FS+14DBH) treatment as the best-performing combination, assigning it the highest Si value (0.22) (Table 7). In contrast, the control was ranked 13/21 in TOPSIS and 14/21 in MABAC, with Ci values of 0.46 and − 0.04, respectively, showing that it was not the worst-ranked alternative in either model. The consistency between the TOPSIS-Shannon entropy and MABAC models confirms the stability and reliability of the ranking outcomes. These results support this study’s hypothesis, as significant differences among treatments were observed, and the TOPSIS-Shannon entropy framework successfully identified 0.4% applied at 14DAT + FS+14DBH as the optimal treatment. These findings provide the biostimulant industry with scientifically validated guidance on how to strategically combine concentration and application timing to enhance the efficacy of Ecklonia maxima extract, moving beyond generalised recommendations. By offering a data-driven optimisation framework, this study not only advances the application of seaweed-based biostimulants but also establishes a model applicable to other crops and biostimulants.

Table 6.

Final rankings on Ecklonia maxima-based biostimulant concentrations and application timing obtained by the TOPSIS method for improving bell pepper yield and quality.

Rank	Alternatives	Si+	Si−	Ci
1	14DAT + FS+14DBH(0.4)	0.03	0.07	0.72
2	ODBT(0.4)	0.03	0.05	0.64
3	ODBT(0.2)	0.03	0.05	0.62
4	14DAT(0.2)	0.04	0.06	0.61
5	ODBT+14DAT + FS(0.4)	0.04	0.05	0.59
6	14DAT(0.4)	0.04	0.05	0.56
7	14DAT + FS+14DBH(0.6)	0.04	0.05	0.55
8	ODBT(0.8)	0.04	0.05	0.53
9	ODBT+14DAT + FS(0.2)	0.05	0.04	0.49
10	ODBT + FS+14DBH(0.2)	0.04	0.04	0.48
11	ODBT + FS+14DBH(0.4)	0.05	0.04	0.48
12	ODBT+14DAT + FS(0.6)	0.05	0.04	0.46
13	CONTROL	0.05	0.04	0.46
14	14DAT(0.8)	0.06	0.04	0.40
15	14DAT + FS+14DBH(0.8)	0.05	0.03	0.38
16	14DAT + FS+14DBH(0.2)	0.05	0.03	0.37
17	ODBT+14DAT + FS(0.8)	0.05	0.03	0.37
18	ODBT + FS+14DBH(0.6)	0.07	0.03	0.32
19	14DAT(0.6)	0.06	0.03	0.30
20	ODBT + FS+14DBH(0.8)	0.07	0.02	0.26
21	ODBT(0.6)	0.07	0.02	0.25

Si⁺- distance from the positive ideal; Si⁻- distance from the negative ideal; Ci⁺- distance from the ideal point.

Table 7.

Final rankings on Ecklonia maxima-based biostimulant concentrations and application timing obtained by the MABAC method for improving bell pepper yield and quality.

Rank	Alternatives	Si
1	14DAT + FS+14DBH(0.4)	0.22
2	ODBT(0.4)	0.17
3	ODBT+14DAT + FS(0.4)	0.17
4	14DAT + FS+14DBH(0.6)	0.16
5	ODBT(0.2)	0.14
6	ODBT + FS+14DBH(0.4)	0.13
7	14DAT(0.2)	0.09
8	ODBT(0.8)	0.05
9	14DAT(0.4)	0.04
10	ODBT + FS+14DBH(0.2)	0.03
11	ODBT+14DAT + FS(0.6)	0.00
12	14DAT + FS+14DBH(0.8)	−0.02
13	ODBT+14DAT + FS(0.2)	−0.02
14	CONTROL	−0.04
15	14DAT + FS+14DBH(0.2)	−0.05
16	0DBT+14DAT + FS(0.8)	−0.05
17	0DBT + FS+14DBH(0.6)	−0.06
18	1DAT(0.6)	−0.08
19	ODBT + FS+14DBH(0.8)	−0.11
20	14DAT(0.8)	−0.12
21	ODBT(0.6)	−0.16

This study was conducted during a single growing season at a single experimental site, limiting the generalisability of the findings across different locations, soil types, and climatic conditions. Seasonal variability and environmental factors may influence bell pepper’s response to biostimulant concentration and application timing. Although ranking stability was demonstrated through entropy-weight sensitivity analysis and cross-validation with the MABAC model, external validation across multiple seasons and locations is required to establish broader applicability. Furthermore, no economic analysis was carried out. Although the 0.4% Ecklonia maxima treatment applied at 14DAT + FS+14DBH achieved the highest TOPSIS and MABAC score, its cost-effectiveness and practical feasibility for commercial growers were not evaluated. Agronomic optimisation should therefore be complemented by economic analysis to determine practical adoption potential.

Conclusion

Application of Ecklonia maxima extract improved bell pepper yield compared with the untreated control, regardless of concentration or application timing, increasing production from 2.01 to 6.40 kg m^− 2. Although yield was not significantly affected by the interaction between concentration and application timing, their combined effects significantly influenced fruit quality attributes. The interaction of 0.4% extract applied at 14DAT + FS+14DBH significantly produced the highest TSS/TA ratio (1.85 vs. 1.26 in control) and total phenolic content (238.9 vs. 173.6 mg 100 g^− 1 FW), while maintaining elevated chlorophyll and carotenoid levels. This treatment also achieved the highest TOPSIS closeness coefficient (Ci = 0.72), identifying it as the optimal application strategy. From a practical perspective, this optimised treatment represents an effective and feasible cultural practice, requiring only three targeted applications at key phenological stages while improving fruit quality and market value. Its adoption may improve fruit nutritional quality and consumer appeal, which could be beneficial for growers. These findings provide a structured basis for decision-making in the use of seaweed-based biostimulants in bell pepper production. Future research should focus on multi-environment validation and economic assessment to support wider commercial adoption.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

D.M: methodology, investigation, formal analysis, writing-original draft, reviewing and editing. T. P. M: Resources, funding acquisition, supervision, writing-reviewing, and editing. T. K. S: Conceptualization, formal analysis, supervision, writing-reviewing, and editing. S. G. N: Writing, reviewing, and editing.

Data availability

All raw data supporting the findings of this study will be made by authors upon request.

Declarations

Ethical statement

The raw data supporting the conclusions of this article will be made available by the authors on request.

Competing interests

The authors declare no competing interests.