Path analysis reveals simultaneous increase in vitamin C and soluble solids in guava progeniesAbstract
A simultaneous increase of the antioxidant vitamin C (VITC) and in the soluble solids/acidity ratio (TSS/ACI) may enhance guava quality and stimulate consumption. The present study aimed to estimate phenotypic correlations and perform path analysis using a causal diagram for VITC and TSS/ACI (dependent variables) and 11 other independent variables in progenies from the crosses [‘Pedro Sato’ × ‘Purple Guava’], [‘Pedro Sato’ × ‘Paluma’], and [‘Paluma’ × ‘Purple Guava’], to guide selection and breeding of new guava cultivars. A total of 295 progenies from the three crosses were evaluated over four production cycles for 13 physicochemical traits. Significant positive correlations between VITC and TSS/ACI were found, ranging from 0.140 to 0.227. Significant positive phenotypic correlations (‘r’) were estimated for VITC × TSS/ACI and significant negative correlations among VITC and TSS/ACI with total fruit yield and mean fruit weight. The direct effects of VITC to TSS/ACI and TSS/ACI to VITC were consistently positive and significant across all analyses, while the ‘r’ were not consistent, highlighting the limitations of the phenotypic correlations in explaining cause-and-effect relationships. The direct effect estimates indicate that a simultaneous increase in VITC and TSS/ACI is possible, though it may have a negative impact in total yield and mean fruit weight. Indirect selection for increased VITC can be achieved through selection for high TSS, or vice versa.
Keywords:
phenotypic correlation; Proc Calis; Psidium guajava
Resumo
Um aumento simultâneo no teor de vitamina C (VITC) — um importante antioxidante — e na relação sólidos solúveis/acidez (TSS/ACI) — que determina a doçura — pode favorecer o consumo de goiaba. Este estudo teve como objetivo estimar correlações fenotípicas e realizar análise de trilha utilizando um diagrama causal para VITC e TSS/ACI (variáveis dependentes) e outras 11 variáveis independentes em progênies provenientes dos cruzamentos ‘Pedro Sato’ × ‘Goiaba Roxa’, ‘Pedro Sato’ × ‘Paluma’ e ‘Paluma’ × ‘Goiaba Roxa’, com o objetivo de orientar a seleção e o melhoramento de novas cultivares de goiabeira. Foram avaliadas 295 progênies provenientes dos três cruzamentos ao longo de quatro ciclos de produção para 13 características físico-químicas. Correlações positivas significativas (‘r’) entre VITC e TSS/ACI foram observadas, variando de 0,140 a 0,227. Foram estimadas correlações fenotípicas positivas significativas para VITC × TSS/ACI e correlações negativas significativas entre VITC e TSS/ACI com a produção total de frutos e o peso médio dos frutos. Os efeitos diretos de VITC para TSS/ACI e TSS/ACI para VITC foram consistentemente positivos e significativos em todas as análises, enquanto as ‘r’ não foram consistentes, destacando as limitações da correlação na explicação das relações de causa e efeito. As estimativas dos efeitos diretos indicam que um aumento simultâneo em VITC e TSS/ACI é possível, embora possa resultar em redução da produção total e do peso médio dos frutos. A seleção indireta para aumento de VITC pode ser realizada por meio da seleção para altos teores de sólidos solúveis, ou vice-versa.
Palavras-chave:
correlação fenotípica; Proc Calis; Psidium guajava
1. Introduction
Guava (Psidium guajava L.) is a tropical fruit crop of great economic and nutritional importance, widely cultivated in Brazil and in several tropical and subtropical regions worldwide. Bioactive compounds from species of the genus Psidium have shown significant pharmacological properties, becoming promising alternatives for the development of therapeutic strategies for numerous pathologies (Rocha et al., 2025). Within the Myrtaceae family, which comprises nearly 2,800 species, guava stands out for its broad adaptability to edaphoclimatic conditions, hardiness, and relative tolerance to pests and diseases (Colombo and Cavichioli, 2019). Guava fruits exhibit high variability in shape, color, and texture depending on the cultivar, and are notable for their high nutritional and functional value, containing vitamins, minerals, and antioxidant compounds such as carotenoids, polyphenols, and flavonoids (Upadhyay et al., 2019).
Vitamin C (ascorbic acid) is particularly noteworthy among guava compounds, with concentrations may range from 200 to 400 mg per 100 g of pulp, depending on the genotype, maturation stage, and cultivation conditions. The ascorbic acid content in guava can be up to six times higher than that of orange (Youssef and Ibrahim, 2016). Humans do not synthesize vitamin C, and the recommended daily intake is 95 and 110 mg day−1 for women and men, respectively (Nielsen et al., 2021).
Total soluble solids (TSS), which mainly represent sugars in fruits, and titratable acidity (TA), which reflects organic acid content, are important indicators of guava quality and flavor (Lima et al., 2002). TSS can serve as an indicator of fruit sweetness, making them more appealing to consumers. The TSS-to-TA ratio is a key attribute for fresh fruit consumption, though it is less relevant for industrial processing (Du Preez and Welgemoed, 1990). Lima et al. (2002) reported TSS values ranging from 9.2 to 10.9 °Brix and TA from 0.40 to 1.04%, resulting in TSS/TA ratios around 10 and 25, depending on genotype and growing conditions.
Correlation analysis is widely used to investigate associations among breeding traits, allowing inferences about their potential relationships. However, this method does not distinguish between direct and indirect effects, which may compromise selection accuracy (Cruz et al., 2014). To overcome this limitation, path analysis, proposed by Wright (1921), has been successfully applied to various crop species. This statistical technique decomposes correlation coefficients into direct and indirect effects, allowing a more precise understanding of causal relationships among traits (Cruz et al., 2014).
In guava, Kherwar and Usha (2016) demonstrated that correlations between bioactive compounds and antioxidant activity do not always reflect true causal relationships, highlighting the need for path analysis to identify actual effects. Their results showed that compounds such as pectin, lycopene, and anthocyanins exert direct and positive effects on antioxidant activity, whereas ascorbic acid, although positively correlated, exhibited a negative direct effect.
Studies aiming to simultaneously increase vitamin C content and the soluble solids/titratable acidity ratio within a single plant are very limited in fruit crops and nonexistent in guava. Moreover, correlation and path analyses, are two statistical tools that can facilitate indirect selection — that is, selecting for a trait that is easier and less costly to measure to improve another of interest (Cruz et al., 2014).
In this context, the objective of the present study was to estimate phenotypic correlations among 13 physicochemical variables in progenies from three crosses and to decompose these correlations into direct effects within a causal diagram. Vitamin C and the TSS-to-TA ratio were considered dependent (endogenous) variables, while the remaining traits were treated as independent (exogenous) variables. The ultimate goal is to guide the selection and development of new guava cultivars.
2. Materials and Methods
2.1. Plant material and physicochemical analyses
The experiment was conducted at the Bebedouro Experimental Station of Embrapa Semiárido, located in Petrolina (latitude 09º09’00”S, longitude 40º22’00”W, altitude 360m), Pernambuco, Brazil. The soil at the Bebedouro site was classified as a Dystrophic Red-Yellow Argisol.
Three guava populations were evaluated and were obtained through manual crosses among commercial cultivars: ‘Pedro Sato’ × ‘Purple Guava’ (PSR) (n = 110), ‘Pedro Sato’ × ‘Paluma’ (PSP) (n = 95), and ‘Paluma’ × ‘Purple Guava’ (PR) (n = 90). The hybridizations were confirmed using phenotypic markers to ensure the authenticity of the crosses. The segregating plants were grafted (under field conditions) onto the rootstock ‘BRS Guaraçá’, which is resistant to root-knot nematodes (Santos et al., 2022). Parental plants were also planted and employed as controls. The planting spacing was 4.0 m × 2.5 m, with drip irrigation and management practices consistent with commercial cultivation in the region, including pruning (for canopy formation) and fruit harvesting. Evaluations were carried out, in the same plants, over four consecutive production cycles, from 2023 to 2025. Fruits were harvested according to the maturation stage of each plant, with five to eight ripe fruits collected per plant. After harvest, fruits were placed in labeled paper bags and transported to the laboratory for physical and chemical analyses.
Morphoagronomic and physicochemical fruit traits were measured: (1) yield per plant (YPP) = estimated by multiplying the total number of fruits by the mean fruit weight (MFW); (2) MFW = average of five to eight fruits per plant, measured using a semi-analytical balance; (3) number of fruits per plant (NFP); (4) fruit length (FL) and (5) fruit diameter (FD) = determined using a digital caliper in a sample of up to eight fruits; (6) firmness = assessed with a penetrometer by taking two readings on opposite sides of the fruit (kilogram-force, kgf); (7) seed weight (PSEED) = obtained after washing, drying, and weighing the seeds on an analytical balance; and (8) pulp and peel weight (C+P) = estimated as the difference between the mean fruit weight and seed weight. Chemical analyses included: (9) total soluble solids (TSS) = determined with a handheld refractometer; (10) titratable acidity (ACI) = determined by titration with 0.1 N NaOH (Correa et al., 2011); (11) TSS/ACI ratio = calculated by dividing total soluble solids by titratable acidity; (12) vitamin C (VITC) = determined using the Tillmans method (Correa et al., 2011); (13) β-carotene (BETA) and (14) lycopene (LYCO) = estimated spectrophotometrically using specific equations for each pigment (Nagata and Yamashita, 1992); and (15) anthocyanin (ANTO) = determined spectrophotometrically at 535 nm after acid-alcohol extraction (Tan et al., 2022; Ribeiro et al., 2026).
2.2. Statistical analyses
All 13 variables were treated as independent and paired in all possible combinations to estimate simple (phenotypic) correlations within the progenies of the three crosses (PR, PSR, and PSP) and in a combined analysis of all progenies. Correlation significance was tested at the 5% probability level using the ‘t-test’, and all analyses were performed with the PROC CORR procedure in SAS University, version 9.4 (SAS Institute, 2014).
The path analysis diagram considered vitamin C (VITC) and the soluble solids/titratable acidity ratio (TSS/ACI) as dependent (endogenous) variables, while the remaining traits were treated as independent (exogenous) variables. Path analyses were performed separately for each cross as well as for the combined dataset. The correlation matrix obtained from SAS PROC CORR was used for path analyses rather than the “maximum likelihood” matrix estimated by SAS PROC CALIS. The variables total fruit number, seed weight, total soluble solids, and titratable acidity were excluded from the path analysis due to multicollinearity, as they contributed to the estimation of other variables, such as mean fruit weight.
Estimates of direct effects and model tests (“modeling info”) were obtained using PROC CALIS, version 9.4 (SAS Institute, 2014), while the coefficients of determination (R2) were calculated using the software Genes (Cruz, 1998). Student’s t-tests for direct effects were performed in PROC CALIS, version 9.4 (SAS Institute, 2014), since software Genes does not provide this estimate. Indirect effects on the dependent variables were not emphasized, as the main interest in path analysis lies in direct effects.
3. Results
Sample sizes for estimating correlation coefficients ranged from 215 (anthocyanin) to 295 (TSS/ACI) in the PR population, from 316 (anthocyanin) to 440 (TSS/ACI) in the PSR population, from 312 (lycopene) to 320 (TSS/ACI) in the PSP population, and from 531 (anthocyanin) to 1055 (acidity) in the combined analysis of the three populations. Phenotypic correlation estimates were obtained using the classical Pearson’s correlation formula, which showed only slight deviations from those estimated by the maximum likelihood (ML) method (data not shown) in PROC CALIS (SAS). Although ML estimation is suitable for handling missing data, it reduces sample size. Even so, the sample sizes in the present study exceed those recommended by Gnambs (2023) that suggested at least 40 individuals for unbiased estimation of the Pearson’s correlation coefficient.
3.1. Correlation analyses in three segregating populations of guava
In the ‘Paluma’ × ‘Purple Guava’ (PR) population, significant positive correlations were observed between vitamin C × total soluble solids (0.285) and vitamin C × TSS/ACI (0.223). Significant negative correlations were found for vitamin C × total yield (−0.234), vitamin C × fruit diameter (−0.285), vitamin C × β-carotene (−0.234), and vitamin C × lycopene (−0.155), whereas correlations with the remaining traits were not significant (Table 1, above the main diagonal). In the ‘Pedro Sato’ × ‘Purple Guava’ (PSR) population, significant positive correlations were observed for vitamin C × total soluble solids (0.297) and vitamin C × TSS/ACI (0.255). Significant negative correlations occurred for vitamin C × total yield (−0.151), vitamin C × mean fruit weight (−0.164), vitamin C × fruit diameter (−0.171), and vitamin C × peel + pulp (−0.139). Correlations with other traits were not significant (Table 1, below the main diagonal).
Estimates of phenotypic correlation coefficients by the PROC CORR in SAS among 13 traits# of F2 progenies from the crosses ‘Paluma’ × ‘Purple Guava’ (PR) (above the main diagonal) and ‘Pedro Sato’ × ‘Purple Guava’ (PSR) (below the main diagonal).
In the ‘Pedro Sato’ × ‘Paluma’ (PSP) population, significant positive correlations were detected only for vitamin C × total soluble solids (0.149) and vitamin C × TSS/ACI (0.140). Significant negative correlations were found for vitamin C × mean fruit weight (−0.114), vitamin C × fruit diameter (−0.154), vitamin C × β-carotene (−0.143), and vitamin C × lycopene (−0.210), with nonsignificant correlations for the other traits (Table 2, above the main diagonal). In the combined analysis of the three populations (PR, PSR, and PSP), significant positive correlations were obtained for vitamin C × total soluble solids (0.261) and vitamin C × TSS/ACI (0.227). Significant negative correlations were observed for vitamin C × total yield (−0.109), vitamin C × mean fruit weight (−0.135), vitamin C × fruit diameter (−0.170), vitamin C × peel + pulp (−0.111), vitamin C × lycopene (−0.108), and vitamin C × β-carotene (−0.108), while the remaining correlations were not significant (Table 2, below the main diagonal).
Estimates of phenotypic correlation coefficients by the PROC CORR in SAS among 13 traits# of F2 progenies from the crosses ‘Pedro Sato’ × ‘Paluma’ (PSP) (above the main diagonal) and all progenies from the crosses ‘Paluma’ × ‘Purple Guava’, ‘Pedro Sato’ × ‘Purple Guava’, and ‘Pedro Sato’ × ‘Purple Guava’ (below the main diagonal).
In the PR population, significant positive correlations for TSS/ACI were observed with total soluble solids (0.437) and vitamin C (0.223). Significant negative correlations were obtained with total yield (−0.245), acidity (−0.788), and lycopene (−0.198), with no significant correlations for the remaining traits (Table 1, above the main diagonal). In the PSR population, significant positive correlations were detected between TSS/ACI × total soluble solids (0.554) and TSS/ACI × vitamin C (0.225), and significant negative correlations for TSS/ACI × total yield (−0.174) and TSS/ACI × acidity (−0.500), with nonsignificant correlations for the remaining traits (Table 1, below the main diagonal).
In the PSP population, significant positive correlations were observed for TSS/ACI × total yield (0.120), TSS/ACI × total soluble solids (0.318), and TSS/ACI × vitamin C (0.140). A significant negative correlation was found only for TSS/ACI × acidity (−0.752), while other correlations were not significant (Table 2, above the main diagonal). In the combined analysis of the three populations, significant positive correlations were obtained for TSS/ACI × total soluble solids (0.461) and TSS/ACI × vitamin C (0.227), and a significant negative correlation for TSS/ACI × acidity (−0.558), with nonsignificant correlations for the remaining variables (Table 2, below the main diagonal).
3.2. Path analyses in three segregating guava populations
Estimates of direct effect coefficients of independent variables on vitamin C and the TSS/ACI ratio were largely consistent between PROC CALIS and Genes, indicating that both employ solutions derived from the matrix equation X′Xβ = X′Y.
In the PR population, significant positive direct effects on vitamin C were observed only for the TSS/ACI ratio (0.160), while significant negative effects were found for total fruit yield (−0.185), fruit diameter (−0.548), and β-carotene (−0.195) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 55 degrees of freedom and a determination coefficient of 0.22. In the PSR population, significant positive direct effects on vitamin C were observed for fruit length (0.237), TSS/ACI ratio (0.195), and peel + pulp (0.771), whereas significant negative effects were found for mean fruit weight (−1.00) and β-carotene (−0.122) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 78 degrees of freedom and a determination coefficient of 0.15.
Causal diagram showing estimates of direct effects for 11 independent/exogenous variables on the dependent/endogenous variable vitamin C (VITC) and soluble solids to acidity ratio (TSS_ACI) in progenies from guava crosses ‘Paluma’ × ‘Purple Guava’ (PR), ‘Pedro Sato’ × ‘Purple Guava’ (PSR), ‘Pedro Sato’ × ‘Paluma’ (PSP), and all progenies (PR, PSR, and PSP). **, *, ns = significant at 1%, 5%, and non-significant by the t-test (n–2 degrees of freedom).
In the PSP population, significant positive direct effects on vitamin C were detected only for the TSS/ACI ratio (0.121), and significant negative effects for lycopene (−0.138) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 45 degrees of freedom and a determination coefficient of 0.10. In the combined analysis of the three populations, significant positive direct effects on vitamin C were observed for the TSS/ACI ratio (0.217) and peel + pulp (0.493), while significant negative effects occurred for mean fruit weight (−0.562), fruit diameter (−0.137), and β-carotene (−0.112) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 78 degrees of freedom and a determination coefficient of 0.15.
For the PR population, significant positive direct effects on the TSS/ACI ratio were observed only for vitamin C (0.167), while significant negative effects were detected for total fruit yield (−0.227), mean fruit weight (−1.517), peel + pulp (−1.485), and lycopene (−0.197) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 55 degrees of freedom and a determination coefficient of 0.17. In the PSR population, significant positive direct effects on the TSS/ACI ratio were observed only for vitamin C (0.205), and significant negative effects only for total fruit yield (−0.163) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 78 degrees of freedom and a determination coefficient of 0.10.
In the PSP population, significant positive direct effects on the TSS/ACI ratio were observed for vitamin C (0.124) and peel + pulp (0.712), while a significant negative effect was found for mean fruit weight (−0.571) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 45 degrees of freedom and a determination coefficient of 0.10. In the combined analysis of the three populations, significant positive direct effects on the TSS/ACI ratio were observed for vitamin C (0.224) and β-carotene (0.100) (Figure 1). The “modeling info” chi-square test was significant (p < 0.001), with 78 degrees of freedom and a determination coefficient of 0.10.
In the PR population, progenies with simultaneous increases in vitamin C and TSS were identified: a) cycle 3, progenies PR102 (290.18; 11.6) and PR41 (312.50; 11.8); b) cycle 4, progenies PR102 (462.96; 11.9), PR17 (462.96; 12.5) and PR46 (416.67; 13.2); c) cycle 5, progenies PR22 (279.72; 10.7), PR79 (262.24; 11.1) and PR94 (262.24; 10.5). In the PSR population, progenies with simultaneous increases in vitamin C and TSS were identified: a) cycle 3, progeny PSR78 (290.18; 11.3); b) cycle 4, progenies PSR10 (416.67; 12.8), PSR118 (416.67; 14.8) and PSR57 (462.96; 13.6); c) cycle 5, progenies PSR11 (262.24; 10.7) and PSR123 (384.62; 13.5); d) cycle 6, progenies PSR118 (450.00; 13.1), PSR20 (390.63; 11.3), PSR43 (300.00; 12.1) and PSR98 (359.38; 11.8).
In the PSP population, progenies with simultaneous increases in vitamin C and TSS were identified: a) cycle 3, progenies PSP62 (267.86; 11,4) and PSP74 (312.50; 11.4), b) cycle 4, progeny PSP05 (393.52; 13.50), c) cycle 5, progenies PSP108 (297.20; 10.4), PSP11 (411.76; 10.1), PSP111 (297.20; 9.2), PSP75 (297.20; 10.7) and PSP91 (297.20; 11.1), d) cycle 6, progenies PSP110 (325.00; 11.8), PSP118 (265.63; 11.2) and PSP34 (325.00; 11.5).
4. Discussion
Guava is an important source of vitamin C for human nutrition. This nutrient is required daily by the human body. Therefore, increase vitamin C content is a major breeding objective in guava breeding programs. On the other hand, the soluble solids/acidity ratio (TSS/Acidity) indicates fruit sweetness, a key sensorial attribute for fresh consumption. Studies aiming to simultaneously increase both traits through genetic selection remain limited in major fruit crops, particularly in guava. The present study is therefore pioneering in applying both correlation and path analyses for vitamin C and TSS in guava. It is important to highlight that these statistical tools can facilitate indirect selection, enabling breeders to improve a complex or costly trait by carrying out selection for another trait that is easier and less expensive to measure.
In the progenies of the three guava crosses, the soluble solids/acidity ratio (TSS/Acidity) showed significant positive phenotypic correlations with soluble solids and vitamin C, ranging from 0.140 to 0.460, and significant negative correlations with total fruit yield (YPP), mean fruit weight (MFW), and fruit diameter (FD), ranging from –0.114 to –0.285. These results indicate that higher TSS/Acidity ratios are associated with increased soluble solids and vitamin C contents, but with reduced yield and acidity. Lakade et al. (2011) reported positive correlations between TSS/Acidity and total yield (0.611) and MFW (0.695), corroborating only the positive association between vitamin C and TSS observed here.
Lakade et al. (2011), studying 10 guava genotypes in a single cycle, reported significant positive phenotypic correlations between ascorbic acid (vitamin C) and soluble solids (0.905), vitamin C and total fruit yield (0.844), and vitamin C and mean fruit weight (0.650), corroborating only the vitamin C × TSS correlation of the present study. Meena et al. (2020) also found phenotypic correlations between vitamin C × TSS (0.838) and vitamin C × MFW (0.836), partially supporting these results. Conversely, Kherwar and Usha (2016) observed a significant negative correlation between vitamin C × TSS/Acidity (–0.870), which contrasts with the present findings as well as those of Lakade et al. (2011) and Meena et al. (2020). Altogether, these results may indicate the presence of population-specific phenotypic associations among these traits in guava.
All path diagrams were significant according to the chi-square test (p < 0.001), although the coefficients of determination (R2) were low, ranging from 10% to 22% for vitamin C and TSS/Acidity. This indicates that, despite statistical significance, the causal models explained only a modest portion of the variance in the dependent variables. In contrast, Kherwar and Usha (2016) reported a higher R2 (98%, based on residual effects for bioactive compounds in guava.
In all path analyses for vitamin C, the significant positive direct effects consistent were estimated to TSS/Acidity, whereas non-consistent were observed all other traits, except firmness and anthocyanin. For the TSS/Acidity ratio diagram, significant positive direct effects were consistent for vitamin C, whereas non-consistent were observed all other traits, except firmness and anthocyanin. As discussed in the seminal work of Dewey and Lu (1959), discrepancies between correlation coefficients and direct effects are expected, since path analysis decomposes the correlation into direct and indirect effects, minimizing confounding influences among traits.
Kherwar and Usha (2016) reported a strong direct effect of the TSS/Acidity ratio on bioactive compounds, including vitamin C, corroborating the estimates from this study. Similarly, Mishra and Nandi (2018) found high positive direct effects of soluble solids and strong negative direct effects of fruit length, diameter, total branch number per plant, mean fruit weight, and yield per plant on vitamin C content in tomato, supporting our findings that production-related traits negatively influence vitamin C levels.
The direct effect estimates in this study indicate that a simultaneous increase in soluble solids and vitamin C, as well as in the TSS/Acidity ratio, is achievable; however, this may lead to reductions in total yield and mean fruit weight. Alternatively, indirect selection for high soluble solids content—a trait that can be easily and rapidly measured using a refractometer—may represent an efficient strategy to enhance vitamin C concentration, which is more laborious and costly to quantify.
5. Conclusions
Significant positive phenotypic correlations were estimated for VITC × TSS/ACI and significant negative correlations among VITC and TSS/ACI with total fruit yield and mean fruit weight.
The direct effects of VITC to TSS/ACI and TSS/ACI to VITC were consistently positive and significant across all analyses, while the correlations were not consistently, highlighting the limitations of the phenotypic correlations in explaining cause-and-effect relationships.
The direct effect estimates indicate that a simultaneous increase in VITC and TSS/ACI is possible, though it may have a negative impact in total yield and mean fruit weight. Indirect selection for increased VITC can be achieved through selection for high TSS, or vice versa.
Acknowledgements
This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil – Finance Code 001. C.A.F.S. has a researcher productivity CNPq grant.
Data Availability Statement
The entire dataset supporting the results of this study was published in the article itself.
References
-
COLOMBO, R.H. and CAVICHIOLI, F.A., 2019. Custo de produção na cultura da goiaba. Revista Interface Tecnológica, vol. 16, no. 1, pp. 470-479. https://doi.org/10.31510/infa.v16i1.587
» https://doi.org/10.31510/infa.v16i1.587 -
CORRÊA, L.C., SANTOS, C.A.F., VIANELLO, F. and LIMA, G.P.P., 2011. Antioxidant content in guava (Psidium guajava) and araçá (Psidium spp.) germplasm from different Brazilian regions. Plant Genetic Resources, vol. 9, no. 3, pp. 384-391. https://doi.org/10.1017/S1479262111000025
» https://doi.org/10.1017/S1479262111000025 -
CRUZ, C.D., 1998. Programa GENES: aplicativo computacional em estatística aplicada à genética. Genetics and Molecular Biology, vol. 21, no. 1, pp. 135-138. https://doi.org/10.1590/S1415-47571998000100022
» https://doi.org/10.1590/S1415-47571998000100022 - CRUZ, C.D., CARNEIRO, P.C.S. and REGAZZI, A.J., 2014. Modelos biométricos aplicados ao melhoramento genético 3. ed. Viçosa: Editora UFV.
-
DEWEY, D.R. and LU, K.H., 1959. A correlation and path-coefficient analysis of components of crested wheatgrass seed production. Agronomy Journal, vol. 51, no. 9, pp. 515-518. https://doi.org/10.2134/agronj1959.00021962005100090002x
» https://doi.org/10.2134/agronj1959.00021962005100090002x -
DU PREEZ, R.J. and WELGEMOED, C.P., 1990. Variability in fruit characteristics of five guava selections. Acta Horticulturae, vol. 275, no. 275, pp. 351-360. https://doi.org/10.17660/ActaHortic.1990.275.42
» https://doi.org/10.17660/ActaHortic.1990.275.42 -
GNAMBS, T., 2023. A brief note on the standard error of the pearson correlation. Collabra. Psychology, vol. 9, no. 1, pp. 87615. https://doi.org/10.1525/collabra.87615
» https://doi.org/10.1525/collabra.87615 -
KHERWAR, D. and USHA, K., 2016. Genetic variations, character association and path analysis studies in guava (Psidium guajava L.) for bioactive and antioxidant attributes. Indian Journal of Plant Physiology / Official Publication of the Indian Society for Plant Physiology, vol. 21, no. 3, pp. 355-361. https://doi.org/10.1007/s40502-016-0230-7
» https://doi.org/10.1007/s40502-016-0230-7 - LAKADE, S.K., TAMBE, T.B., DHOMANE, P.A. and GHARGE, V.R., 2011. Correlation studies in guava. Asian Journal of Horticulture, vol. 6, pp. 159-161.
-
LIMA, M.A.C.D., ASSIS, J.S.D. and GONZAGA-NETO, L., 2002. Caracterização dos frutos de Goiaba e seleção de cultivares na Região do Submédio São Francisco. Revista Brasileira de Fruticultura, vol. 24, no. 1, pp. 273-276. https://doi.org/10.1590/S0100-29452002000100061
» https://doi.org/10.1590/S0100-29452002000100061 - MEENA, L.K., BHATNAGAR, P., SINGH, J., CHOPRA, R. and SOLANKI, P., 2020. Correlation amongst yield and quality attributes of guava fruit in response to foliar feeding of zinc and iron in Vertisols of Jhalawar district. The Pharma Innovation Journal, vol. 9, no. 6, pp. 349-351.
- MISHRA, A. and NANDI, A., 2018. Correlation and path coefficient analysis for quality traits in tomato (Solanum lycopersicon L.). Journal of Pharmacognosy and Phytochemistry, vol. 7, no. 1, pp. 1733-1738.
-
NAGATA, M. and YAMASHITA, I., 1992. Simple method for simultaneous determination of chlorophyll and carotenoids in tomato fruit. Nippon Shokuhin Kogyo Gakkaishi, vol. 39, no. 10, pp. 925-928. https://doi.org/10.3136/nskkk1962.39.925
» https://doi.org/10.3136/nskkk1962.39.925 -
NIELSEN, C.W., RUSTAD, T. and HOLDT, S.L., 2021. Vitamin C from seaweed: A review assessing seaweed as contributor to daily intake. Foods, vol. 10, no. 1, pp. 198. https://doi.org/10.3390/foods10010198 PMid:33478115.
» https://doi.org/10.3390/foods10010198 -
ROCHA, T.S., MARQUES, E.J., NASCIMENTO, C.M., SILVA, M.C., BONESS, H.V.M. and VALE, V.L.C., 2025. Bioactive potential and chemical profile of extracts of Psidium oligospermum Mart. ex DC.(Myrtaceae). Brazilian Journal of Biology, vol. 85, pp. e290748. https://doi.org/10.1590/1519-6984.290748 PMid:40105537.
» https://doi.org/10.1590/1519-6984.290748 -
RIBEIRO, T.S., FREITAS, S.T., COSTA, A.C.S.S., BOITEUX, L.S. and SANTOS, C.A.F., 2026. Genetic parameters and identification of transgressive segregation for lycopene and anthocyanin contents in fruits from guava (Psidium guajava) crosses. Journal of Crop Improvement, vol. 40, no. 2, pp. 193-204. https://doi.org/10.1080/15427528.2026.2639084
» https://doi.org/10.1080/15427528.2026.2639084 -
SANTOS, C.A.F., COSTA, S.R., BOITEUX, L.S., GRATTAPAGLIA, D. and SILVA-JUNIOR, O.B., 2022. Genetic associations with resistance to Meloidogyne enterolobii in guava (Psidium sp.) using cross-genera SNPs and comparative genomics to Eucalyptus highlight evolutionary conservation across the Myrtaceae. PLoS One, vol. 17, no. 11, pp. e0273959. https://doi.org/10.1371/journal.pone.0273959 PMid:36322533.
» https://doi.org/10.1371/journal.pone.0273959 - SAS INSTITUTE, 2014. SAS® Studio, Version 9.4, SAS OnDemand for Academics Cary: SAS Institute Inc.
-
TAN, J., HAN, Y., HAN, B., QI, X., CAI, X., GE, S. and XUE, H., 2022. Extraction and purification of anthocyanins: A review. Journal of Agriculture and Food Research, vol. 8, pp. 100306. https://doi.org/10.1016/j.jafr.2022.100306
» https://doi.org/10.1016/j.jafr.2022.100306 -
UPADHYAY, R., DASS, J.F.P., CHAUHAN, A.K., YADAV, P., SINGH, M. and SINGH, R.B., 2019. Guava enriched functional foods: therapeutic potentials and technological challenges. In: R.B. SINGH, R.R. WATSON and T. TAKAHASHI, eds. The role of functional food security in global health London: Academic Press, pp. 365-378. https://doi.org/10.1016/B978-0-12-813148-0.00021-9
» https://doi.org/10.1016/B978-0-12-813148-0.00021-9 - WRIGHT, S., 1921. Correlation and causation. Journal of Agricultural Research, vol. 20, pp. 557-585.
-
YOUSSEF, M. and IBRAHIM, R., 2016. Molecular markers associated with high vitamin-C content in guava. Journal of Agricultural Chemistry and Biotechnology, vol. 7, no. 3, pp. 49-55. https://doi.org/10.21608/jacb.2016.40744
» https://doi.org/10.21608/jacb.2016.40744
Edited by
-
Editor:
Takako Matsumura Tundisi
