The field experiment was conducted during three consecutive growing seasons (2022–2024) at the experimental station of Karshi State University, located in the Karshi district of the Kashkadarya region, Uzbekistan (approximate coordinates: 38°52′N, 65°48′E). The region is characterized by an arid continental climate with rapid temperature accumulation in early spring.

The soil of the experimental site is classified as irrigated meadow-alluvial (local classification), characterized by a groundwater table depth of 2.0–3.0 m. The soil texture is medium loam with low salinity. Pre-plant soil analysis (0–30 cm depth) indicated a low humus content of 0.9–1.1%, total nitrogen of 0.08–0.10%, available phosphorus of 18–22 mg kg^-1, and exchangeable potassium of 200–215 mg kg^-1 (Figure 1).

The study evaluated five ultra-early watermelon (Citrullus lanatus) hybrids: Montana F1, Talisman F1, Dolby F1, Hollar-Crimson Sweet, and Krimstar F1. These hybrids were selected for their potential adaptability to early-season stressors and rapid maturation rates.

A randomized complete block design (RCBD) with three replications was employed. The total area of each experimental plot was 73.5 m². Seeds were sown during the first decade of March (March 1–10) using the “thermos” method. This technique involves a double-layer polyethylene film cover to optimize soil temperature and moisture retention, a practice widely cited for enhancing seedling emergence in early-season cucurbits (Díaz-Pérez, 2022; Zhang et al., 2024).

The experiment followed a factorial arrangement with two factors: watermelon hybrid and fertilization strategy.

Five ultra-early hybrids were evaluated under identical microclimatic conditions, while fertilization method constituted the main management factor.

In both fertilization treatments all plots were managed under the same temporary double-layer film cover; thus, the film acted as a common background technology, while the only experimental factor was the fertilization strategy (broadcast versus localized per nest).

The temporary film cover was applied uniformly to all plots and was therefore not treated as an experimental factor.

The planting scheme followed a double-row system: (280 + 70)/2 × 70 cm, resulting in a plant density of approximately 8,160 plants ha^-1.

To assess nutrient use efficiency and yield response, two distinct fertilization strategies were compared:

The ranges reported for per-nest NPK rates reflect minor adjustments based on hybrid-specific plant density and nest spacing, ensuring comparable nutrient supply per unit area across treatments.

Based on the established plant density, localized per-nest fertilization resulted in a 30–40% lower total mineral nitrogen input per hectare compared with the conventional broadcast fertilization practice.

Phenological observations were recorded according to standard agronomic protocols. Germination, flowering (male and female), and fruit setting dates were documented when 10% (onset) and 75% (full stage) of the plant population reached the respective stage.

Biometric measurements were conducted at the flowering and fruit ripening stages on 10 representative plants per plot. Parameters included main vine length (cm), number of leaves per plant, and leaf area (m²). Leaf area was estimated using linear dimensions according to established methodological guidelines for cucurbits.

Total marketable yield was harvested in June. Fruit quality was assessed by measuring the Total Soluble Solids (TSS) content using a hand-held refractometer (expressed in Brix %), which is a key indicator of watermelon internal quality and consumer acceptance.

Data from the 2022–2024 growing seasons were subjected to analysis of variance (ANOVA) in a randomized complete block design to test the main effects of hybrid, fertilization method and year, as well as their interactions. A combined ANOVA across years was performed, treating year as a random factor. Because the interaction between year and treatment was not significant for the studied traits, the results are presented as three-year means. Statistical processing followed the methodology of Dospekhov (1985), and treatment means were separated using Fisher’s Least Significant Difference (LSD) test at p < 0.05.

Hybrid and fertilization strategy were treated as fixed effects, while year and block were considered random factors.

Mean comparisons were performed using Fisher’s LSD test at P ≤ 0.05, which is appropriate for factorial field experiments with a limited number of treatments.

In this study, nutrient-use efficiency was assessed indirectly through the ratio between marketable yield and total applied mineral nitrogen per unit area, allowing comparison of fertilization strategies under identical growing conditions.

The vegetative development of ultra-early watermelon hybrids was significantly influenced by both the genotype and the fertilization method. The biometric measurements taken during the fruit ripening stage revealed that the localized (per nest) application of organo-mineral fertilizers promoted more vigorous growth compared to the conventional broadcast method.

As shown in Table 1, the main vine length across all hybrids ranged from 218 cm to 292 cm under the optimized fertilization regime, whereas under the control conditions, it ranged from 195 cm to 268 cm.

Differences among fertilization strategies were statistically significant (p < 0.05), while variation among hybrids reflected genotype-specific growth potential.

The hybrid Krimstar F1 exhibited the longest main vine (292 cm) under localized fertilization, which differed significantly from the broadcast treatment within the same hybrid (p < 0.05).

Similarly, the assimilation surface area (leaf area) — a critical parameter for photosynthetic efficiency — was positively affected by the localized nutrient application. Leaf area per plant ranged from 0.74 to 0.82 m² under localized fertilization, compared with lower values under broadcast fertilization, corresponding to an increase of approximately 8–12%.

Montana F1 and Talisman F1 showed the highest leaf area values, indicating stronger vegetative development under identical early-season growing conditions.

The analysis of yield components indicated that the “thermos” method combined with nest-fertilization significantly accelerated crop maturation and increased marketable yield. Because the film cover was applied uniformly across all treatments, differences in crop performance are attributed to fertilization strategy rather than to thermal protection effects.

The appearance of the first marketable fruits in the localized fertilization treatments occurred 3–5 days earlier than in the broadcast fertilization plots, which is agronomically relevant for ultra-early production systems.

Total marketable yield varied significantly among the studied hybrids (Figure 2). The optimized fertilization strategy resulted in yields ranging from 22.7 to 26.1 t ha^-1, whereas the conventional method produced between 19.8 and 23.4 t ha^-1.

The highest productivity was recorded for Krimstar F1 (26.1 t ha⁻¹) and Montana F1 (25.4 t ha⁻¹) under localized fertilization. LSD analysis confirmed that yield differences between fertilization strategies were statistically significant at p < 0.05.

These results indicate that placing nutrients directly in the root zone enhances crop productivity under ultra-early cultivation conditions, consistent with improved fertilizer efficiency.

Fruit quality, particularly sugar content expressed as total soluble solids (TSS), is a key determinant of consumer acceptance. The results indicate that increased yields under localized fertilization did not lead to a reduction in fruit quality.

A common concern in intensive vegetable production systems is the potential “dilution effect,” whereby increased yields may be accompanied by a reduction in fruit sugar concentration. In the present study, however, higher yields obtained under localized fertilization were not associated with a decline in Total Soluble Solids (TSS). Across the evaluated hybrids, TSS values remained within a relatively narrow range, indicating that fruit sweetness was maintained despite yield enhancement.

The graphical results (Figure 3) provide the primary quantitative evidence for TSS stability, while the photographic assessment of fruit internal quality (Figure 4) offers a representative visual illustration of flesh color and refractometer-based measurements at harvest maturity. These images should be interpreted as supportive, qualitative documentation rather than standalone evidence of treatment effects.

The maintenance of TSS under localized fertilization may be related to a more balanced nutrient supply in the root zone, particularly adequate potassium availability during the fruit-filling stage. Nevertheless, since the present study did not directly quantify carbohydrate partitioning or enzymatic activity related to sugar metabolism, this interpretation should be considered indicative rather than mechanistic.

The primary objective of ultra-early watermelon cultivation is to overcome the thermal limitations of early spring. In this study, earlier harvest (3–5 days) was observed under localized fertilization while all treatments were grown under identical temporary film-cover conditions.

Therefore, differences in earliness cannot be attributed solely to the film cover itself, which served as a background technology rather than an experimental factor.

Díaz-Pérez (2022) reported that plastic mulches and low tunnels effectively increase soil temperature and accumulate growing degree days (GDD), which are critical for cucurbit establishment in cool seasons. Our findings are consistent with this general understanding of plasticulture systems, although the present experiment did not include a no-cover control to directly quantify the independent effect of film cover on phenology.

Zhang et al. (2024) emphasized that optimizing root-zone temperature through film covering enhances root metabolic activity and nutrient uptake. In the present study, enhanced vegetative growth observed under localized fertilization likely reflects improved nutrient availability under favorable early-season thermal conditions, rather than a distinct film-cover effect.

A key finding of this study is that localized (per nest) fertilization maintained, and in some cases exceeded, the yield obtained from conventional broadcast application, despite a significant reduction in total mineral nitrogen input. Conventional broadcasting often leads to low Nitrogen Use Efficiency (NUE) due to leaching and volatilization, a problem highlighted by Wang et al. (2022) in arid zone agriculture. Similar conclusions regarding the limited yield response to increased nitrogen inputs have been reported by Giraldo-Sanclemente et al. (2025).

In our experiment, the optimized localized treatment reduced the mineral nitrogen dose by approximately 30–40% compared with the conventional broadcast recommendation, while maintaining or even increasing total marketable yield. This outcome indicates a higher agronomic efficiency of applied nitrogen, as reflected by greater yield per unit of mineral N input.

Although nitrate leaching and soil nitrogen dynamics were not directly measured in this study, a reduction in applied mineral nitrogen is generally associated with a lower potential risk of nitrogen losses in irrigated arid environments.

Therefore, the observed yield response suggests that localized fertilization may contribute to more sustainable nutrient management under ultra-early watermelon production systems.

Our data suggest that placing nutrients directly in the “nest” (rhizosphere zone) ensures a higher availability of nutrients during early root development. This observation is consistent with Hou et al. (2024), who reported that reduced chemical fertilizer combined with organic amendments can improve watermelon yield.

However, potential effects on soil microbial diversity or nutrient cycling were not assessed in the present study and should be considered as prospective research directions rather than confirmed mechanisms.

The differential performance of hybrids highlights the importance of genotype selection for specific environmental conditions. Krimstar F1 and Montana F1 consistently exhibited greater vine length and higher yield under localized fertilization.

These differences reflect genotype-specific responses to nutrient management rather than direct evidence of physiological mechanisms such as heterosis or hydraulic conductance, which were not measured in this study.

Nevertheless, the observed stability of these hybrids across seasons suggests that they possess favorable traits for ultra-early production under arid continental conditions.

Similar varietal differences have been reported by Luna et al. (2024). While such studies indicate that genetic traits may influence water and nutrient relations, the present results should be interpreted primarily at the agronomic performance level.

The ability of Krimstar F1 to achieve high yields (26.1 t ha⁻¹) under reduced mineral nitrogen input indicates its suitability for resource-efficient production systems, rather than a direct quantification of genetic NUE.

A common trade-off in agronomy is the “dilution effect,” where higher yields are associated with reduced sugar or nutrient concentrations in fruits. In the present study, no statistically significant decline in Total Soluble Solids (TSS) was observed between fertilization strategies, despite higher yields under localized fertilization.

TSS values remained within a narrow range (7.4–7.5%) for hybrids such as Montana F1, indicating that fruit sweetness was maintained.

These results suggest that yield improvement achieved through localized fertilization did not adversely affect internal fruit quality.

Potassium is known to play a role in sugar translocation (Song et al., 2022). Although potassium dynamics were not directly quantified, balanced NPK application in the root zone may have supported assimilate transport during fruit filling, contributing to the maintenance of TSS levels.