Seed Morphometry and Germination of Four Edible Species of Passiflora spp. Conserved in a Gene Bank

Abstract

The Passifloraceae family is one of the most representative in tropical America, with food, pharmaceutical, and ornamental importance. This study evaluated seed morphometry and germination of eight accessions of four Passiflora edible species, P. edulis; P. ligularis; P. quadrangularis; and P. tripartita var. mollissima, by studying accessions conserved several years in the gene bank (−15 °C) and recently collected accessions. Four experimental phases were carried out as follows: (1) morphometric characterization of seeds with qualitative and quantitative variables; (2) evaluation of germination under two thermal regimes (20 °C/30 °C and 25 °C); (3) application of six pre-germination treatments to overcome dormancy; and (4) tetrazolium tests. In phase 1, P. quadrangularis stood out for its unique morphological characteristics according to multivariate analysis. In phase 2, the alternating thermal regime (20 °C/30 °C) promoted the highest germination. In phase 3, the germination response was specific to each species: mechanical scarification in P. edulis (85.7%), KNO₃ (0.5%) in P. ligularis (35.7%), control in P. quadrangularis (71.1%), and gibberellic acid (GA₃ 400 ppm) in P. tripartita (71.4%). The tetrazolium phase 4 identified the viability status of the seeds. It is concluded that the differences in morphometry and germination reflect the intrinsic characteristics of each species, highlighting the importance of specific protocols for their germination. This study provides tools to optimize the conservation and regeneration of Passiflora spp. germplasm under ex situ conditions, as a genetic base to be utilized in the future.

1. Introduction

The Passifloraceae family is one of the most representative in tropical America, with 520 species recorded in the region; Ecuador and Peru share about 80 species [1,2,3,4]. Four species of passion fruits present the greatest economic importance for their ornamental, pharmaceutical, and food properties: P. edulis Sims (maracuyá), P. ligularis Juss. (granadilla), P. tripartita var. mollissima Juss (Poir) (taxo), and P. quadrangularis L. (badea) [2,3]. Passiflora are propagated sexually (seeds) and asexually (cuttings); however, the sexual method is more widely used, due to its simplicity and low cost [5]. Official statistics for Ecuador report production data for Passiflora edulis and P. ligularis, and not for P. tripartita var. mollissima and P. quadrangularis [6]. In this sense, P. edulis production averages 53,643 t in 7681 ha, with an average of 7.03 t/ha (years 2019–2024); export values reached USD 32 million in 2024 [7]. On the other hand, P. ligularis is commercially cultivated in smaller areas, reaching in 2025 values of USD 1.8/kg in local markets and up to USD 12.7/kg in European markets. In contrast, P. tripartita var. mollissima and P. quadrangularis still remain restricted in Ecuador to small areas (<1 ha) with no official records in recent agricultural surveys [6]. At the international level, Brazil led global Passiflora edulis production in 2020 with 690,364 t [8]; statistics for the other Passiflora are not available. As a consequence, the lack of important commercial use for some of these species, in addition to its adaptation to specific ecological niches [4], puts these species at risk of loss in in situ conditions, which is why their conservation in gene banks is essential [9].

Gene banks conserve plant genetic resources for food and agriculture through collection and conservation, among other activities [10]. One of the methods of conservation is the storage of orthodox seeds under subzero temperatures, which requires monitoring the germination of the conserved samples [11]. Seed germination is defined as the morphological transformation accompanied by physiological processes that result in the elongation and expansion of the embryo, leading to a healthy seedling [12]. The evaluation of germination percentages provides a starting point for further viability studies [13]. During germination, important environmental factors such as light and temperature must be taken into account [14,15]. In the case of Passiflora spp. there is great variability in germination, due to the genetic background of accessions or to the conditions in which the tests are carried out [16].

Passiflora have an exogenous dormancy, i.e., a combination of mechanical dormancy due to the hardness of the testa, as well as chemical dormancy due to germination inhibitors present in the seed coat or the mucilage [15,17]. Therefore, it is common in some plant species to apply pre-germinative treatments to break seed dormancy, such as the use of gibberellic acid (GA₃), in varying concentrations among others [18,19].

Germination and pre-germination treatments for the four Passiflora species have been previously conducted with freshly harvested seeds for P. edulis: [20,21,22,23]; for P. quadrangularis: [17,24,25,26]; for P. mollisima: [27,28]; and for P. ligularis: [29]; or similar germination studies involving several Passiflora species such as P. quadrangularis and P. mollissima in [24] or P. ligularis and P edulis in [30,31], among others. However, to our knowledge, there is no research that compares newly collected seed materials of Passiflora with those preserved in gene banks for several years.

This study involves four edible species of Passiflora spp., collected in Ecuador and conserved in the gene bank of the National Institute of Agricultural Research (INIAP), aiming to describe the seed morphology of the species; to understand the differences in germination of newly collected seed materials with those preserved in the gene bank for several years; and to establish a germination protocol for the four species’ previous regeneration of the materials.

2. Materials and Methods

2.1. Study Area and Samples

This study was carried out in the seed laboratory of the National Department of Plant Genetic Resources (DENAREF) at the National Institute of Agricultural Research (INIAP), which is located at 3050 m a.s.l. in the province of Pichincha, Ecuador.

Seeds of four Andean species of passion flowers were used with two accessions per each species: P. edulis (INIAP-ECU-9296 and INIAP-ECU-28735), collected in the provinces of Imbabura and Pichincha (Sierra region); P. ligularis (INIAP-ECU-18039 and INIAP-ECU-28732), from Carchi and Pichincha (Sierra region); P. quadrangularis (INIAP-ECU-9293 and INIAP-ECU-28734), from the provinces of Guayas and Manabí (Coastal region); and P. tripartita var. mollissima (INIAP-ECU-11894 and INIAP-ECU-28731), from Chimborazo and Pichincha (Sierra). They were stored in the gene bank at −15 °C in plastic-coated aluminum bags.

The accessions P. edulis (INIAP-ECU-9296), P. ligularis (INIAP-ECU-18039), P. qua-drangularis (INIAP-ECU-9293), and P. tripartita var. mollissima (INIAP-ECU-11894) have been conserved in the gene bank for 27, 16, 27, and 30 years, respectively. The accessions P. edulis (INIAP-ECU-28735), P. ligularis (INIAP-ECU-28732), P. quadrangularis (INIAP-ECU-28734), and P. tripartita var. mollissima (INIAP-ECU-28731) are more recently collected materials that have been conserved in the gene bank for four years under the same conditions. Twenty seeds of each accession were randomly selected for the morphometric characterization and for the germination tests and seven seeds were used for each observation for each treatment to evaluate the influence of temperature and pre-germination stimulants.

Due to the reduced availability of Passiflora seed samples conserved at INIAP’s gene bank, a modified germination protocol with less seeds was used, as described by Monteros, Tacán [32]. Germination with small samples conserved at the gene bank (usually wild species) are conducted with samples of 50 seeds or less [33]. However, accessions with low numbers of seeds must be regenerated [11].

2.2. Morphometry

For each accession, 20 seeds were randomly selected and evaluated. Quantitative variables, length (mm), width (mm), and thickness (mm), were measured with a digital caliper Mitutoyo 500-196-20B (PRC); the weight (g) was measured with an OHAUS EXPLORER analytical balance. In addition, the qualitative variables such as shape, margin, apex, base, and ornamentation were assessed based on the taxonomic keys of Mezzonato, Mendonça [19], and Pérez, Tillett [34]. These qualitative variables plus the quantitative variable seed area (mm²) were recorded with the captured seed’s images in a Zeiss stereo microscope (Stemi 508 model) with Zen LT 3.0 software.

2.3. Seed Disinfection

The protocol used was developed by Morillo [35]. In short, the seeds were washed with running water and liquid soap, three times for 5 min; the last rinse was performed with distilled water. Subsequently, inside the laminar flow chamber, each sample was left with 1% povidone-iodine for 60 min and rinsed with sterilized distilled water. Then, it was disinfected with 70% alcohol for 1 min and rinsed with sterilized distilled water. Finally, the seeds were immersed in a diluted solution of 1% sodium hypochlorite plus Tween 20 ^® for 10 min and a final rinse of 3 to 5 min. This disinfection process was carried out for each accession sample before subjecting the seeds to germination tests with temperature and stimulant treatments.

2.4. Seed Germination Under Two Temperature Conditions

Germination tests were carried out in glass tubes of 18 mm diameter and 150 mm high, with 1% SIGMA agar to evaluate the effect of two programmed temperatures in Memmert Shell lab germinators. The first temperature was 30 °C/20 °C (with a 12 h light/12 h dark photoperiod), similar to that used by Angelini, Clemente [36]; Gutiérrez, Miranda [15]; and Miranda, Fischer [5]. The second temperature was a constant 25 °C (constant throughout 24 h), similar to that used by Delanoy, Van Damme [27]. For both temperatures, the relative humidity was maintained at 60%, and the light intensity at 58.824 µmol m⁻² s⁻¹.

Seed germination was evaluated every 15 days as suggested by Rao, Hanson [37]. Only seeds with a radicle length of 5 mm were considered germinated, as proposed by ISTA [38]; the germination percentage (%) was determined by using the formula proposed by ISTA [38] (Equation (1)).

$$Germination percentage={\displaystyle \frac{Nº of seeds germinated}{Nº of seeds set to germinate}}\times 100$$

(1)

2.5. Seed Germination with Pre-Germination Treatments

After identifying the temperature that had the best effect on germination, the next phase was to evaluate six pre-germination treatments: (1) Control; (2) Gibberellic acid (GA₃) at 200 ppm, similar to that proposed by Angelini, Clemente [36], Cárdenas Hernández [18], and Cárdenas, Carranza [30]; (3) GA₃ at 400 ppm similar to Cárdenas Hernández [18], Cárdenas, Carranza [30], and Carranza, Castellanos [17]; (4) Potassium nitrate (KNO₃) at 0.2%; (5) KNO₃ at 0.5% [7]; and (6) mechanical scarification with a cut in the lateral part of the seeds [15,17,24,30,36]. For the GA₃ and KNO₃ treatments, the seeds were soaked for 48 h in the stimulant solutions. Germination of the samples was evaluated every 15 days, as suggested by Roa Nieto [39]. The formula proposed by ISTA [38] was used to evaluate the germination percentage (%) (Equation (1)).

2.6. Viability Test with Tetrazolium

To complement the results obtained from the germination tests, a seed viability test was carried out using a 10% diluted solution of 2,3,5-Trifenyltetrazolium Sigma-Aldrich (St. Louis, MO, USA). For this, five seeds were used for each accession, and the samples were placed in a universal UN Memmert oven at 40 °C for 1 h and 30 min, similar to what was performed by Peralta [40].

2.7. Statistical Analysis

The results obtained in the morphological characterization were analyzed by means of descriptive statistics and multivariate analysis with Principal Component Analysis (PCA) and Cluster analysis, based on five quantitative variables by using Ward’s method and Gower’s distance in the statistical program Infostat Inc. (Vernon Hills, IL, USA) [41].

For the germination tests applying different temperatures and stimulants, four observations were defined per treatment, each consisting of seven seeds of the eight accessions corresponding to the four species of Passiflora. A completely randomized trial (CRD) was applied. The data obtained was analyzed with Microsoft Excel ^®.

The Scheffé multiple comparison test at the 5% significance level allowed the evaluation of differences among study groups in Section 2.4 and Section 2.5 [42].

3. Results

3.1. Seed Morphometry

Appendix A.1 includes the summary of statistics for the different quantitative variables including measures of central tendency and coefficient of variance CV (%) for each accession. The two accessions of P. quadrangularis presented higher values in all quantitative variables similar to the one reported by Pérez, Tillett [34]; the only exception was seed thickness, where P. tripartita (INIAP-ECU-28731) presented a higher value. The accession with the lowest values was P. tripartita var. mollissima (INIAP-ECU-11894).

Appendix A.2 shows that there is a high variation in the qualitative variables: Apex, Margin, Base, and Ornamentation for all species. The Base variable presented five different categories with respect to the eight accessions evaluated. The variable Shape shares similarities among three of the four species (P. ligularis, P. quadrangularis, and P. tripartita). The variation that exists within each species is due to the genetic diversity present, since all the accessions were collected in different localities. The morphological traits that showed the greatest variation in this study were seed width, ranging from 2.95 mm (P. tripartita, INIAP-ECU-11894) to 7.9 mm (P. quadrangularis, INIAP-ECU-28734), and seed area, with values from 11.18 mm² (P. tripartita, INIAP-ECU-11894) to 59.68 mm² (P. quadrangularis, INIAP-ECU-28734, and INIAP-ECU-9293). In addition, seed shape varied between obovate and elliptical, and ornamentation exhibited different patterns for each species.

Principal Component Analysis (PCA) showed that the quantitative variables, weight, length, width, and area, and the qualitative variable apex, explained the first principal component; while thickness, margin, and ornamentation explained the second, jointly accounting for 98.6% of the morphological variability. The relationships among variables revealed significant differences between P. quadrangularis and the other three species (P. edulis, P. ligularis, and P. tripartita var. mollissima), with thickness and seed apex being independent from the other variables, suggesting a distinct genetic basis. In the PCA plot, the accessions of P. edulis, P. ligularis, and P. tripartita var. mollissima clustered in the left quadrant, whereas P. quadrangularis was separated on the right, with accession INIAP-ECU-28731 standing out for its greater thickness. (Figure 1).

The cluster analysis with quantitative data resulted in a dendrogram where the four Passiflora spp. formed four groups (Figure 2). Group 1 includes INIAP-ECU-28732 and INIAP-ECU-18039 (P. ligularis) and INIAP-ECU-28735 (P. edulis), which are similar in thickness, length, width, area, and weight. Group 2 comprises INIAP-ECU-11894 (P. tripartita) and INIAP-ECU-9296 (P. edulis), showing similarity in thickness, length, and weight but differing in width and area. Group 3 consists solely of INIAP-ECU-28731 (P. tripartita), with thickness being the most prominent variable, while qualitative traits differ in the character base. Finally, Group 4 includes the two P. quadrangularis accessions (INIAP-ECU-9293 and INIAP-ECU-28734), differing mainly in width and area (Appendix A.1).

According to

Table 1. Gower distance of the eight accessions of Passiflora spp. represented in Figure 2, corresponding to the cluster analysis.

	INIAP-ECU-28735	INIAP-ECU-9296	INIAP-ECU-18039	INIAP-ECU-28732	INIAP-ECU-9293	INIAP-ECU-28734	INIAP-ECU-11894	INIAP-ECU-28731
INIAP-ECU-28735	0.00
INIAP-ECU-9296	0.33	0.00
INIAP-ECU-18039	0.26	0.40	0.00
INIAP-ECU-28732	0.26	0.39	0.30	0.00
INIAP-ECU-9293	0.71	0.75	0.72	0.67	0.00
INIAP-ECU-28734	0.87	0.90	0.87	0.84	0.50	0.00
INIAP-ECU-11894	0.48	0.41	0.46	0.53	0.85	0.99	0.00
INIAP-ECU-28731	0.49	0.48	0.53	0.45	0.73	0.79	0.63	0.00

, the accessions INIAP-ECU-11894 and INIAP-ECU-28734 present the greatest distance with a Gower value of 0.99, representing the highest dissimilarity among the eight accessions. These correspond to P. tripartita var. mollissima and P. quadrangularis, respectively, with the divergence mainly explained by polarized data in their variables. In particular, the quantitative variables width, area, and weight showed the greatest differences, since seeds of INIAP-ECU-11894 (P. tripartita) are the smallest, while those of INIAP-ECU-28734 (P. quadrangularis) are the largest (Appendix A.1). Conversely, the smallest distance (0.26) was recorded between INIAP-ECU-18039 and INIAP-ECU-28735, due to several similarities in quantitative traits such as length, thickness, and weight.

In

Table 2. Correlation values and corresponding significance (p values) for seed traits measured from 8 accessions of 4 Passiflora species.

Correlation
Control Variables		Weight	Thickness	Length	Broad	Area
Weight	Correlation	1.000
	Significance (unilateral)
Thickness	Correlation	-	1.000
	Significance (unilateral)
Length	Correlation	0.867	-	1.000
	Significance (unilateral)	0.003 *
Broad	Correlation	0.959	-	0.899	1.000
	Significance (unilateral)	0.000 *		0.001 *
Area	Correlation	0.961	-	0.914	0.996	1.000
	Significance (unilateral)	0.000 *		0.001 *	0.000 *	-

* p values under 0.05%.

, it can be observed that correlations were found between the variables: weight, length, width, and area, which achieved a high degree of significance (p < 0.001). A high correlation is observed between the variables of width and area of the seeds (r = 0.996), correlations are also observed between area and weight (r = 0.961), width, and weight (r = 0.959), but not between thickness and the other variables analyzed, probably because it is a variable that measures the thickness of the testa, which is not related to the other variables that involve the entire seeds.

3.2. Seed Germination Under Two Temperature Conditions

Figure 3 shows that temperature one [T1: 20 °C/30 °C (12 h/12 h)] obtained better response in contrast to the second temperature [T2: 25 °C (24 h)] for all species. In the case of P. edulis, P. quadrangularis, and P. tripartita var. mollissima species, it was observed that the effect of temperature 1 (T1) was effective for the samples that had been conserved for the shortest time (4 years). However, for the species P. ligularis the effect of T1 was positive for the sample conserved for 16 years and not for the sample conserved for 4 years.

It was observed that 60 days after sowing with temperature 1, P. edulis (INIAP-ECU-28735) (Figure 3a) had the best results of the study with 71.44% germination. In second place came P. quadrangularis (INIAP-ECU-28734) (Figure 3c) with 36.00% and in third place came P. tripartita (INIAP-ECU-28731) (Figure 3d); all of these accessions were conserved for four years. On the other hand, P. ligularis (INIAP-ECU-18039) (Figure 3b) was the only species in which the oldest sample (16 years) obtained a higher germination than the most recent (INIAP-ECU-18039), which did not germinate. The only species that obtained a response to the temperature 2 treatment (T2) was P. quadrangularis (INIAP-ECU-28734) with 11.00% of germinated seeds at day 60 (Figure 3c), although germination was higher with T1.

The Scheffé multiple comparison test at 5% for the treatments that had a result on the effect of temperature shows statistical differences for both 45 and 60 days. For the germination percentage variables (%), four levels of significance were identified. Range (a) corresponds to the first temperature with P. edulis (INIAP-ECU-28735). The second range (b) is shared by P. quadrangularis (INIAP-ECU-28736) and P. tripartita var. mollissima (INIAP-ECU-28731). In the third range (c), there is a combination between temperatures with respect to the mean values: P. ligularis (INIAP-ECU-18039) at the first temperature and P. quadrangularis (INIAP-ECU-28734) at the second temperature. This proved to be the only treatment that responded to the constant temperature factor, that is, only P. quadrangularis (INIAP-ECU-28734) in temperature treatment 2 obtained a response in germination capacity (Appendix A.3).

3.3. Seed Germination with Pre-Germination Treatments

This trial, which evaluated six pre-germination treatments, determined that for P. edulis (Figure 4a), the accession INIAP-ECU-28735 (four years of conservation) had the highest response for all the proposed treatments. The mechanical scarification obtained 86% germination at day 60 after sowing, followed by the treatment of GA₃ with 200 ppm concentration which reached 62% at 60 days. On the other hand, the accession P. edulis INIAP-ECU-9296 (27 years of conservation) obtained germination responses in only two treatments (Figure 4a), being the best mechanical scarification, reaching a germination percentage of 62% at 60 days after sowing, and the GA₃ treatment with 200 ppm achieved 14%. This accession had not germinated with any of the two temperature treatments; however, it did germinate when pre-germination treatments were applied.

Granadilla (P. ligularis) with its two accessions, INIAP-ECU-18039 (four years of con-servation) and INIAP-ECU-28732 (16 years), showed different germination responses to all their treatments (Figure 4b), showing differences between accessions of the same species. Previously, the most recent seed sample (four years old) did not show germination, but the oldest accession (16 years old) did germinate when two temperatures were evaluated (Figure 3). In this experiment, the oldest accession already showed germination in response to the application of potassium nitrate and the newest one to the application of GA₃; however, the germination percentages are low: INIAP-ECU-18039 (36%) germination with 0.2% KNO₃; 33% with the scarification treatment; and 29% with the 0.5% KNO₃ treatment. In accession INIAP-ECU-28732, germination occurred only in three treatments: GA₃ 200 ppm (29%), GA₃ 400 ppm (29%), and mechanical scarification (14.3%).

Badea, P. quadrangularis, had the lowest response to pre-germination treatments of all species. Figure 4c shows that the accession INIAP-ECU-9293 (17 years of conservation) obtained 0% germination for all treatments, while for INIAP-ECU-28734 (four years), 71% germination was observed in the control sample and 14% germination for the 0.5% KNO₃ treatment. These two treatments were the only ones to present germination percentages different from 0%. Therefore, the INIAP-ECU-9293 accession, according to these results, would not be viable after 27 years of conservation.

In the taxon (P. tripartita var. mollissima), the oldest accession (30 years old), INIAP-ECU-11894, responded to the mechanical scarification treatment with 21% germination (Figure 4d), but not for the other treatments in which germination was null. On the contrary, in the accession of the same species INIAP-ECU-28731 (four years old), there was a response to all treatments. The highest germination percentage belonged to the GA₃ 400 ppm treatment with 71%, followed by the control treatment with 43%; on the other hand, the other treatments showed germination values ranging from 14% to 28%.

The Scheffé test at 5% for the treatments showed results, where statistical differences were observed for the two evaluation dates in the three variables considered in the study, yielding three significance ranges (Appendix A.4).

The highest significance range presented the best results within the study. The highest belonged to the mechanical scarification treatment for P. edulis (INIAP-ECU-28735), which achieved a germination percentage of 85.71% at 60 days after sowing. The first range (a) ranged from germination percentage values at 60 days after sowing, with 38.10% for the 0.2% KNO₃ treatment for P. ligularis (INIAP-ECU-28732), to 85.71% for the same accession with the mechanical scarification treatment. The second significant range (b) ranged from values >0% to <38.10% for the 0.5% KNO₃ treatment in P. edulis (INIAP-ECU-28735). The third range (c) referred to all treatments that showed no response, 0% germination.

3.4. Tetrazolium Test

The results of the tetrazolium tests showed that there is a difference between the accessions with the longest conservation time and those with the shortest time. The intensity of the tetrazolium staining results can be observed in Appendix A.5. The oldest accession, INIAP-ECU-9296 (P. edulis), presents a low degree of staining, in contrast to the more recent accession, INIAP-ECU-28735, which presents a higher intensity of staining. For the two accessions of P. ligularis (INIAP-ECU-18039 and INIAP-ECU-28732), there is similar staining in terms of coverage and intensity (moderate) on the cotyledons and embryonic axes. There is a high contrast between the two accessions of P. quadrangularis (INIAP-ECU-9293, older), reflecting a low level of staining on the seeds, contrary to the accession INIAP-ECU-28734 (more recent) which presents a high intensity of staining of embryonic axes. Finally, the accessions of P. tripartita var. mollissima have a medium or diverse staining intensity for the most recent accession INIAP-ECU-28371. On the contrary, the accession INIAP-ECU-11894 (older) presented low or no staining.

The tetrazolium staining results correlate with the germination percentages established and summarized in Figure 3 and Figure 4; that is, seeds that did not germinate have little or no staining while accessions that germinated stained better with tetrazolium.

4. Discussion

4.1. Seed Morphometry

Morphological characteristics indicate that there are several similarities between the seeds of P. edulis, P. ligularis, and P. tripartita var. mollissima, as described by Pérez, Tillett [34]. Likewise, principal component analysis (Figure 1) evidenced these similarities between P. edulis (INIAP-ECU-9296 and INIAP-ECU-28735), P. ligularis (INIAP-ECU-18039 and INIAP-ECU-28732), and P. tripartita (INIAP-ECU-11894) and INIAP-ECU-28731). However, some morphological variability was identified among accessions of the same species, which is attributed to the genetic diversity of the samples, since each accession comes from a different geographical origin. Martín and de Pascual [43], and Mezzonato and Mendonça [19] point out that the morphological proximity between individuals can be interpreted as taxonomic similarity between individuals. Only in the case of P. quadrangulis (INIAP-ECU-9293 e INIAP-ECU-28734) are the differences more marked, in terms of area and ornamentation variables. On the other hand, the present study also revealed morphological differences among the four species analyzed, with ornamentation being the most distinctive trait. This finding is consistent with previous studies that have reported seed coat surface ornamentation as a useful taxonomic identifier within the genus Passiflora. de Jesus Silva and Souza [44] evaluated the following species from the active Passiflora gene bank at EMBRAPA: P. coccinea Aubl., P. edulis Sims, P. gibertii N.E.Br., P. maliformis L., P. morifolia Mast., P. setacea DC., P. suberosa L., and P. tenuifila Killip. The study demonstrated that seed coat ornamentation can serve as a reliable taxonomic identifier within the genus Passiflora [45].

Meneses [46] alluded to the need to quantify the differences between individuals by means of statistical distances; the smaller the distance between individuals, the more similar they are considered. In the cluster analysis (Figure 2), the Gower distance values of the four species studied demonstrate the most dissimilarity between P. tripartita (INIAP-ECU-11894) and P. quadrangularis (INIAP-ECU-28734) (0.99 distance), and the shortest distance between P. ligularis accessions (INIAP-ECU-18039) and P. edulis (INIAP-ECU-28735), with a value of 0.26 attributed to similarities such as Direction and Shape. Multivariate associations have been described for species such as Passiflora cincinnata by Assis, Sousa [47]. In summary, there are closer morphological relationships between the species P. ligularis, P. tripartita var. mollissima, and P. edulis in contrast to P. quadrangularis, which is grouped alone. These variables could be determining factors in water uptake and germination capacity across different plant species [48], as analyzed by Castillo and Melgarejo [45], who examined seed length, shape, and surface pores in Passiflora seeds from three different sources. The correlations between morphometric traits and germination must be further explored.

4.2. Seed Germination Under Two Temperatures

Authors such as Angelini and Clemente [36] agree that the alternating 20/30 °C regime simulates natural environmental conditions, facilitating the metabolic activation of the embryo. Pereira and Dias [49] emphasize the crucial role of temperature in enzyme activation during germination, while the humidity of the medium, high in the agar (90% humidity according to Gutiérrez [50]), was also a determining factor. Overall, the results confirm that the alternating regime is more effective than the constant temperature of 25 °C, especially in recent accessions such as P. edulis (INIAP-ECU-28735), which achieved the highest germination percentage. Germination percentages obtained under alternating temperatures (20/30 °C) coincided with Gutiérrez and Miranda [15], who reported 94% germination in P. edulis (this experiment reported 86%) and 30% in P. ligularis (this experiment reported 21% with no scarification and 36% with KNO₃ 0.2%).

In P. quadrangularis, previous studies such as those by Carranza and Castellanos [17] and Marostega, and Araujo [25] documented low germination, 54% and 50%, respectively, for this species, which coincides with the low germination obtained in the present study: accession INIAP-ECU-28734, conserved for four years, showed 36% germination with alternating temperature (20/30 °C) and 11% with the constant temperature of 25 °C. Carranza and Castellanos [17] described the poor germination of this species in commercial nurseries.

Storage conditions also play a key role in viability; Ayala and Tatis [51] and Agüero and Pereyra [52] point out that the time of conservation and the origin of the material have a significant influence on seed viability. These results suggest that, in addition to the dormancy of the species, storage time significantly affects the viability of seeds, limiting their germination capacity even under favorable thermal conditions, e.g., accession INIAP-ECU-9293, with 27 years of conservation, germination did not show in either of the two temperature treatments evaluated, even though viability was still observed by using the tetrazolium test. Martín and Sánchez [53] and Gutiérrez and Magni [54] mention the storage of seeds at low temperature and humidity in gene banks as a fundamental role, in order to maintain viability closer to the original for a longer period of time.

4.3. Seed Germination with Pre-Germinative Treatments

Pre-germinative treatments showed specific responses by species and accession. For P. edulis, mechanical scarification was the most effective treatment, reaching 86% germination in INIAP-ECU-28735, in agreement with Gutiérrez and Miranda [15], who obtained 94% germination with a testa puncture technique using a pin. However, the values obtained were lower than those reported by Cárdenas and Carranza [30], who reached up to 89% in conditions of greenhouse facilities (oscillating temperature between 18 and 34 °C; and relative humidity between 61 and 99%), suggesting that factors such as substrate or environmental conditions could also influence the germination, such as the presence of mycorrhizae, as suggested by Ramírez Gil and Agudelo [55]. Another factor to consider is the genetic variation in Passiflora species which can influence germination capacity and viability [45,56].

In P. ligularis, although Cárdenas and Carranza [30] reported percentages higher than 80% using doses of 0.1% and 0.2% KNO₃ and GA₃ 200 ppm in winter conditions (18 °C/34 °C; RH% 61% to 99%), the results were lower in this study. This could be attributed to the hardness and lignification of the testa, which maintains dormancy and hinders water absorption, as explained by Gutiérrez [50]. The Badea, P. quadrangularis, showed low general response to pre-germinative treatments, coinciding with Carranza and Castellanos [17]. Pre-germinative treatments with GA₃ at 400 ppm, KNO₃ at 0.4%, and the method described by Roa Nieto [39] were used as control treatments. However, as in the aforementioned study, low germination rates were observed due to the physical dormancy present in the seeds. In particular, the INIAP-ECU-28734 accession only germinated in the control treatment, and to a lesser extent when 0.5% KNO₃ was applied. Additionally, the observed differences between this study and previous experiments may be partially explained by the random selection of seeds and the relatively small sample size, as highlighted by Ribeiro-Oliveira and Ranal [57] who demonstrate that small sample sizes can significantly influence inferences about germination processes.

For P. tripartita, the results are comparable to those of Delanoy and Van Damme [27], who obtained 27% germination by mechanical scarification, while in this study, 21.4% was obtained in the accession INIAP-ECU-11894.

The establishment of a germination protocol for the four Passiflora species will enable the potential use of the evaluated accessions in hybridization programs, thereby contributing to the development of elite materials, particularly promising for passion fruit (P. edulis) and sweet granadilla (P. ligularis) [58]. Consequently, this approach may lead to the generation of new varieties with improved agronomic value, supported by phenotypic and genotypic characterization of chromosomal regions associated with genes conferring resistance to biotic and abiotic stresses (genotyping), as well as with medicinal properties [59].

It is important to note that the accessions that have been conserved for less time (four years) also have not achieved high germination percentages, which indicates dormancy in Passiflora spp. This type of dormancy can be physical, i.e., the testa does not allow the passage of water to the embryo [13] or it can also be caused by the effect of residual mucilage in the seeds. Chen and Zhang [60] mention that mucilage can make it difficult for the embryo to carry out gas exchange to initiate the germination process. It is also important to consider that the degree of maturity at the time of collection can affect the germination performance (Ortega [61]).

Finally, the tetrazolium test did not allow differentiating between dormant and non-dormant seeds, but it allowed to identify that the samples with a longer period of conservation present lower viability than the most recent ones, as it had already been pointed out by Gutiérrez and Miranda [15] and Vega and Campos [62]. INIAP-ECU-9293 did not germinate with our germination treatments; however, it presented a stained embryo, and consequently, it is viable. In this case, embryo rescue could be used, as suggested by Pramanik, Sahoo [63], as an alternative to multiply the accession. It is important to emphasize that, in order to obtain reliable results in this test, it is essential that the tetrazolio penetrates adequately into the viable tissues, allowing the reduction to deep red formazan [64].

5. Conclusions

The results show morphological and germination variability among Passiflora species, which demonstrates the need to study specific protocols for each of them. Alternating heat treatment (20/30 °C) was the most effective, as well as pre-germinative treatments, for all species. Differentiated protocols can be useful for monitoring seeds conserved in a gene bank.

Older accessions of Passiflora spp. conserved in the INIAP gene bank (17 to 30 years) have decreased germination; most likely, they have suffered deterioration of viability due to the cumulative effect of physiological aging or adverse factors not established during storage. However, since the gene banks manage several species, it is difficult to determine and use the most appropriate protocols for germination for either cultivated or wild species. Then, the tetrazolium test is an important complementary method to verify the viability of seeds. In any case, the oldest Passiflora accessions in this study need to be regenerated as a priority to avoid their loss.

As a general rule, it is necessary to carry out periodic germination and viability tests in gene banks so that accessions with germination percentages below the international standard of 85% can be properly regenerated and ready for future use.