Introduction

The cape gooseberry (Physalis peruviana L.) is a perennial species that grows wild in the mountainous regions of South America, between 800 and 3 000 m above sea level. Although its center of origin is not fully defined, Peru, Chile, Colombia, and Venezuela are recognized as the most likely areas (Fischer & Melgarejo, 2014). In recent decades, the cape gooseberry has gained economic and commercial importance due to growing international demand associated with its nutraceutical properties and health benefits (Fischer et al., 2014). Its fruits are rich in vitamins A, B, and C, and contain minerals such as calcium, iron, and phosphorus (Álvarez-Herrera et al., 2021).

Currently, Colombia is the world's leading producer and exporter of cape gooseberries, with markets including the Netherlands, the United States, Germany, Brazil, the United Arab Emirates, and Canada (National Association of Foreign Trade [ANALDEX], 2023). However, despite an increase in cultivated area, productivity has decreased by 23.63 %, from 16.1 to 12.3 t∙ha^-1 (Álvarez-Herrera et al., 2021). This phenomenon could be because production is primarily extensive and uses low levels of technification (Rodríguez-Puertas et al., 2022). Because of this situation, several studies have explored strategies to improve yield by increasing planting density and adopting advanced agronomic practices. For example, Quevedo-García et al. (2015) reported yields of 27.7 t∙ha^-1 with a density of 5 000 plants∙ha^-1 in a pigpen-type trellising system, while Peña-Castillo et al. (2024) obtained up to 16.17 t∙ha^-1 with the use of biostimulants in combination with densities of 4 444 and 10 000 plants∙ha^-1.

In addition to optimizing agronomic management practices to increase yields and fruit quality, efforts have been made to introduce the crop to geographically different areas from its origin in order to produce fruit on a commercial scale. However, this has not been fully achieved. In countries such as South Africa, England, New Zealand, and Kenya, the cape gooseberry is cultivated on small plots to supply local markets, with seasonal production, unlike Colombia, which can produce and export throughout the year (Quiroga & Kirschbaum, 2021). These findings underscore the need to evaluate the adaptation and productivity of the crop in new areas using technified production systems. In this context, the objective of this research was to assess the productive potential of five cape gooseberry varieties at two plant densities under greenhouse and hydroponic conditions.

Materials and methods

The study was conducted from March to November 2023 in a greenhouse at the Plant Science Department of the Universidad Autónoma Chapingo (UACh) (19° 29’ 24’’ N and 98° 52’ 27’’ W, at 2 250 m a. s. l.). The minimum and maximum temperatures inside the greenhouse were 14.6 and 35.4 °C, respectively, and an average relative humidity of 65 %.

The cape gooseberry varieties evaluated were: ‘Sacha’, ‘Modificada’ (Ecuador), ‘Chiclayo’ (Peru), ‘CSAEGRO’ (Guerrero, Mexico), and ‘Fitos’ (Venezuela). The seeds used were obtained from previous experimental cycles at the Colegio de Postgraduados (COLPOS), the Colegio Superior Agropecuario del Estado de Guerrero (CSAEGro), and UACh. Four of the varieties produce yellow fruits with diameters of 1.5 to 2.5 cm and 12 to 18 °Brix. The Modificada variety is the only one that produces brown-greenish fruits, 1.0 to 1.8 cm in diameter and 10 to 11 °Brix (Orozco-Balbuena et al., 2021).

The factors studied were population density (with two levels: high [27 639 plants∙ha^-1] and low [13 861 plants∙ha^-1], with planting frames of 0.3 × 1.2 m and 0.6 × 1.2 m, respectively) and variety (‘Sacha’, ‘Modificada’, ‘CSAEGRO’, ‘Fitos’, and ‘Chiclayo’). A full factorial treatment design (2 × 5) with a split-plot arrangement was established, where the large plot corresponded to the density and the small plot to the variety.

Sowing was carried out in March 2023 in 200-cell polystyrene trays. Seedlings were kept in seedbeds for 49 days and then transplanted to the greenhouse using a randomized complete block design with four replicates. The experimental unit (EU) consisted of a row of 25 and 13 plants, depending on the population density (high and low, respectively). The crop was grown under greenhouse and hydroponic conditions, using small-grained red tezontle (≤ 0.5 cm) as the substrate and Steiner's universal nutrient solution (Steiner, 1984) with an electrical conductivity of 2.5 dS∙m^-1.

The variables evaluated were plant height (PH, cm), stem diameter (SD, mm), fruit weight with calyx (FWC, g), fruit weight without calyx (FWWC, g), total yield (TY, g), yield per plant (YPP, g), polar diameter of the fruit (PD, mm), equatorial diameter of the fruit (ED, mm), total soluble solids (TSS, °Brix), total titratable acidity (TTA, %) and vitamin C (VC, mg∙100 g^-1). The variables PH and SD were measured every two weeks from May 21 to August 26, 2023. The variables FWC, FWSC, PD, ED, TSS, TTA, and VC were evaluated every 20 days between August 10 and November 30, 2023 in fruits at maturity stage 5 according to the Instituto Colombiano de Normas Técnicas y Certificación (ICONTEC, 1999). It was verified that all the fruits evaluated did not present any physical or physiological damage.

The PH was measured with a measuring tape (Truper®). The variables SD, PD, and ED were measured with a digital caliper (Daniu®). FWC and FWWC were obtained using a digital scale (SF-400A, Electronic Compact Scale). SST were measured in five fruits per EU using a handheld digital refractometer (ATAGO, PAL-1®). TY (results ratio) was obtained by summing the FWWC from the seven harvests carried out in each EU, and YPP was calculated by dividing TY by the number of plants per EU. TTA and VC were determined according to the AOAC International (2023) methodology. For TTA, 5 g of fruit were blended in 50 mL of distilled water; subsequently, a 10 mL aliquot was taken and titrated with 0.1 N sodium hydroxide using 1 % phenolphthalein as an indicator. For VC, 5 g of fruit were liquefied in 50 mL of oxalic acid, a 5 mL aliquot was taken and titrated with Tillman's solution (DFI-2,6-dichlorophenol-indophenol), the concentration in mg∙100 g^-1 of fruit was calculated from the measured flow rate in the titration.

The data were analyzed using an analysis of variance with a randomized complete block design in split-plot arrangement, using the Statistical Analysis System version 9.0. Subsequently, Tukey's test (α = 0.05) was performed for factors with a significant effect. For variables with significant interactions, comparisons were made between levels of the variety factor within each density level.

Results and discussion

Analysis of variance

Analysis of variance revealed highly significant effects (p ≤ 0.01) of density and variety on SD and PH starting from measurement four (SD4) and one (PH1), respectively. The density × variety interaction was significant (p ≤ 0.05) in PH5, and highly significant in PH1, PH6, and PH7 (Table 1). Quevedo-García et al. (2015) reported that low planting densities favor plant development due to better branch distribution and larger leaf area, while Mora-Aguilar et al. (2006) maintain that differences in SD and PH are due to genetic variability.

Table 1. Mean squares of the analysis of variance of variables evaluated in five varieties of cape gooseberry (Physalis peruviana L.) cultivated at two population densities.

In the FWC and FWWC variables, the density factor showed significant and highly significant effects from the first measurement, as well as highly significant effects for TY and YPP. The variety factor was significant in four of seven harvests for FWC and FWWC, as well as for TY and YPP. The effect of the variety factor could be due to the genetic characteristics of each variety and its differential capacity to produce larger and, consequently, heavier fruits, as well as the number of fruits per plant (Lagos et al., 2007). The effect of the density factor, on the other hand, is related to crop management, since large, branched plants were used, which increases competition for photoassimilates by the fruits. The density × variety interaction was only significant in FWC3 and FWWC3 (Table 2).

Table 2. Mean squares of the analysis of variance of variables evaluated in five varieties of cape gooseberry (Physalis peruviana L.) grown at two population densities.

Álvarez-Herrera et al. (2021) found that cape gooseberry production in greenhouses decreases when plants are subjected to water stress and high temperatures, while Peña et al. (2021) mentioned that yield depends on the genetic characteristics of the cultivar, agroclimatic conditions, and agronomic management. Quevedo-García et al. (2015) indicated that adequate spacing between plants favors crop development because it improves aeration and light capture in the leaves. Regarding the yield per plant (YPP), Criollo et al. (2014) indicated that it depends on the number of fruits produced per plant, which shows bimonthly peaks that gradually decrease until harvests are minimal (Álvarez-Herrera et al., 2021).

For PD and ED, significant and highly significant effects were observed for both factors, except for PD4, PD7, ED1, ED6, and ED7 for the density factor, and for PD2 and ED2 for the variety factor. For TSS, only the variety factor showed highly significant effects (Table 3). The significance in fruit diameter may be due to genetic variability resulting from additive and non-additive effects (Trevisani et al., 2024). Antúnez-Ocampo et al. (2014) indicated that plants with greater vigor tend to produce fruit with a higher concentration of TSS. These authors also maintain that greater nutrient availability is associated with an increase in fruit size. Miranda and Fischer (2021) observed that soluble sugar content increases during fruit ripening.

Table 3. Mean squares of the analysis of variance of variables evaluated in five varieties of cape gooseberry (Physalis peruviana L.) grown at two population densities.

In the TTA, significant effects were only detected in TTA4 for density and in TTA3 for variety. In contrast, the VC showed highly significant effects only for the variety factor starting from VC2 (Table 4). Miranda and Fischer (2021) mentioned that acidity decreases as the degree of maturity increases. Likewise, the VC content is influenced by the fruit's maturity stage (Ávila et al., 2006).

Table 4. Mean squares of the analysis of variance of variables evaluated in five varieties of cape gooseberry (Physalis peruviana L.) grown at two population densities.

Comparison of means for density

Tukey's test showed no significant differences in SD between densities. However, an increase in AP was observed starting at PH3, where the high density surpassed the low density (Figures 1a and 1b). Quevedo-García et al. (2015) reported heights of 139.5 cm with a density of 5 000 plants∙ha^-1 in a traditional trellis system. The PH achieved in the present study (up to 236.31 cm) could be due to competition among plants for light capture at the evaluated densities (13 861 and 27 639 plants∙ha^-1 for low and high density, respectively) and to the simple trellis training system used.

Significant differences favoring low density were detected in FWC, FWWC, TY, and YPP (Figures 1c, 1d, 1e, and 1f). Gastelum-Osorio et al. (2013) reported higher FWC at densities of 8 plants∙m^-1 and higher FWWC at 4 plants∙m^-1. Peña-Castillo et al. (2024) obtained 16.17 t∙ha^-1 with 4 444 plants∙ha^-1 and 750 mL∙200 L^-1 of biostimulant. In the present study, yields were 8.29 and 4.58 t∙ha^-1 for low and high density, respectively.

Regarding fruit diameter, significant differences were only observed in PD2 and ED2 for high density, and in ED3 for low density (Figures 2a and 2b). Quevedo-García et al. (2015) obtained diameters of 2.193 cm with 1,666 plants∙ha^-1, while Peña-Castillo et al. (2024) achieved diameters of 6.2 cm with the application of 750 mL∙200 L^-1 of biostimulant and a planting density of 4 444 plants∙ha^-1. This suggests that lower densities and appropriate management practices can favor the production of larger fruit.

The mean comparison test showed no significant differences between densities in TSS, TTA, and VC (Figure 2c, 2d, and 2e). Kretzschmar et al. (2014) mentioned that high planting densities negatively affect the VC content in cape gooseberry fruits; however, this was not corroborated in the present investigation, as there were no differences between densities.

Comparison of means for variety

The Sacha variety exhibited the largest diameter starting at SD4, with a final diameter of 18.05 mm at SD8, and the greatest height at the end of the evaluations (228.87 cm) (Figures 3a and 3b). This behavior reflects its vigor and adaptation to the agronomic management system used in this study. Mora-Aguilar et al. (2006) reported stem diameters (SD) and stem lengths (SLR) of 72 to 85 cm and 1.30 to 1.33 cm, respectively, at 64 days after transplanting (DAT) in wild cape gooseberry collections, values lower than those observed in the present study. Furthermore, the final plant height of the Sacha, Modificada, and Chiclayo varieties was greater (228.87, 221.27, and 222.63 cm, respectively) than that reported by Orozco-Balbuena et al. (2021) at 52 days after transplanting (94.83, 100.92, and 100.75 cm, respectively). In contrast, the SDs reported for ‘Sacha’ (17-25 mm), ‘Modificada’ (10-18 mm), and ‘Chiclayo’ (17-22 mm) coincide with those reported by these authors. The greater PH could be due to the fact that the varieties were evaluated at higher population densities.

The Modificada variety tended to exhibit significantly lower values for FWC, FWWC, TY, and YPP, while ‘Sacha’ showed the highest YPP (451.82 g) (Figure 3c, 3d, 3e, and 3f). Peña et al. (2021) reported 7.33 t∙ha^-1 with the Celendín ecotype, and Quevedo-García et al. (2015) obtained up to 27.7 t∙ha^-1 with 5 000 plants∙ha^-1 and a pigpen-type trellis system with the Colombia ecotype, and a yield per plant of 574.50 g with 2 500 plants∙ha^-1. In this study, ‘Sacha’ reached an average of 6.26 t∙ha^-1 and 8.29 t∙ha^-1 at low density, comparable to that obtained by Peña et al. (2021).

The Sacha variety also stood out in PD, ED, and TSS (Figures 4a, 4b, and 4c). Peña et al. (2021) reported 19.56 mm ED and 14.97 °Brix in fruits of the Celendín ecotype. Grigolo et al. (2021), for their part, obtained values of 15 °Brix in fresh cape gooseberry fruits. These results are comparable to those obtained in the present study (10 to 16 °Brix).

The Modificada variety showed the lowest values for VC (except in VC1) and TTA3 (Figures 4d and 4e). These results could be due to factors adverse to VC accumulation in the fruit, such as high temperatures inside the greenhouse or irrigation frequency. The VC values (2.49 to 4.74 mg∙100 g^-1 of fresh fruit) are higher than those reported by Bazalar-Pereda et al. (2020) (0.34 to 1.77 mg∙100 g^-1 of fresh fruit), although lower than those obtained by Álvarez-Herrera et al. (2014) (up to 22.5 mg∙100 g^-1 of fresh fruit). This variation may be due to variability between plants (Grigolo et al., 2021) and agronomic management, particularly irrigation frequency, since low frequencies contribute to a higher concentration of VC in the fruits (Álvarez-Herrera et al., 2014).

Interaction analysis

Mean comparisons were performed only for variables where the density × variety interaction was significant (p < 0.05). The analysis focused on evaluating the performance of the varieties within each density level. In PH1, ‘Modificada’ showed the greatest height, with a significant difference (p < 0.05) compared to the other varieties at both densities (Figure 5a). Likewise, ‘Modificada’ was statistically different at the high density in PH6 and PH7 (Figures 5c and 5d), while ‘Fitos’ exhibited different behavior at the low density in PH6 (Figure 5c), and at the low density in PH7, ‘Chiclayo’ had the lowest value, with significant differences (p < 0.05) compared to the other treatments (Figure 5d).

In FWWC3, the varieties did not differ (p > 0.05) at high density, while at low density, ‘Modificada’ and ‘Sacha’ showed different behavior compared to the other variables (p < 0.05), but similar behavior among themselves (p > 0.05) (Figure 6a). For PD1 and PD4, ‘Modificada’ differed significantly from the other varieties at high density (Figures 6b and 6c). In contrast, at low density, ‘CSAEGRO’ and ‘Sacha’ were significantly different from the other variables in PD1 (Figure 6b), while ‘Fitos’ and ‘Modificada’ were statistically inferior at low density in PD4 (Figure 6c). No differences (p > 0.05) were detected between varieties for ED2 at any density (Figure 6d).

According to Orozco-Balbuena et al. (2021), high planting densities increase yield per hectare but reduce fruit quality in terms of weight and size. This could explain the differences between varieties within each density and highlights the genetic capacity of each variety to thrive under different levels of competition.

Conclusions

Low density favored yield, although it reduced plant height. The Sacha variety stood out for its higher yield and fruit quality. Although no significant interaction was observed between variety and density for yield, interactions were detected between plant height and fruit size, demonstrating that the Fitos, Chiclayo, and CSAEGRO varieties respond differently to population density.