The results from detailed scanning electron microscopy images processed using FijiJ software revealed that most of the African marigold pollen grains belonged to the oblate spheroidal group, followed by the suboblate and rarely oblate groups, according to the classification of [23] (Fig. 1). The pollen grains of the marigold plants were isopolar and radially symmetrical with tricolpi apertures. The spines on the walls of the pollen grains were pyramidshaped and often conical, with a wide base at the bottom and a blunt to pointed apical tip.
Descriptive statistics for pollen morphometrics were calculated to determine the general characteristic features of the marigold pollen grains (Table 2). The results revealed lower variance and range for all the traits under study, except for the perimeter of the pollen grains. The lengths of the polar and equatorial axes ranged from 22.64 to 39.57 µm and 25.31 to 49.74 µm, respectively, while the perimeter of the pollen grains ranged from 82.33 to 133.14 µm. Furthermore, the results revealed that the data pertaining to the polar axis followed a normal distribution, as the values obtained for skewness were near zero. In contrast, all the other traits studied had skewed distributions.
Table 2
General descriptive statistics of African marigold pollen grains
| Perimeter (µm) | P: Polar axis (µm) | E: Equatorial axis (µm) | Length of ectoaperture (µm) | P/E ratio |
Mean | 101.42 | 29.39 | 32.86 | 4.51 | 0.90 |
Range | 82.33- 133.14 | 22.64–39.57 | 25.37–49.74 | 1.86–10.61 | 0.64–1.05 |
Sample Variance | 93.39 | 9.36 | 11.74 | 2.79 | 0.01 |
Skewness | 0.62 | 0.22 | 0.59 | 0.95 | -0.91 |
Analysis of variance was used to assess the statistical significance of differences in pollen morphological traits among various flower forms of marigold. The results revealed significant differences among the different flower forms for the traits studied (Table 3). Petaloid male sterile flowers exhibited significantly greater values for almost all the traits than did the other flower forms. The semidouble fertile form displayed the lowest values for perimeter and both polar and equatorial axis lengths. Petaloid male sterile flowers were found to be on par with the double-fertile form for the equatorial axis and the apetaloid double-fertile and double-fertile form for the polar axis. Furthermore, Duncan’s multiple range test was used to analyze the relationships and differences between the genotypes. Significant variations were observed for all the traits among the genotypes (Table 4). The genotype IIHRM 2–3 SF was significantly different for most of the traits under study except for the equatorial axis. The genotype IIHRMO 2330 was significantly different in terms of the trait perimeter of the pollen grains. Across the genotypes, the trait P/E ratio did not vary significantly. The pollen grains, classified according to Edrtman’s classification, revealed three shapes: oblate spheroidal, suboblate and rarely oblate. Moreover, across the genotypes, oblate spheroidal-shaped pollen grains were dominant, followed by suboblate pollen grains.
Table 3
Pollen morphometric features of different flower forms in African marigolds
Flower forms | Perimeter (µm) | P: Polar axis (µm) | E: Equatorial axis (µm) | Length of ectoaperture (µm) | P/E ratio |
Single fertile | 101.26b | 28.26c | 32.70b | 4.86b | 0.87c |
Semidouble fertile | 92.04d | 27.35c | 30.21d | 5.14b | 0.91b |
Double fertile | 103.62b | 30.18b | 34.09a | 3.88c | 0.89bc |
Apetaloid sterile with single fertile | 96.54c | 27.54c | 30.64cd | 3.72c | 0.90bc |
Apetaloid sterile with double fertile | 96.25c | 30.39b | 31.92bc | 3.75c | 0.95a |
Petaloid sterile with single fertile | 114.55a | 32.16a | 35.26a | 6.36a | 0.91b |
F value | 33.58* | 17.27* | 14.96 * | 14.89 * | 4.88 * |
SE (m) ± | 1.39 | 0.49 | 0.56 | 0.28 | 0.01 |
SE (d) | 1.97 | 0.69 | 0.79 | 0.40 | 0.02 |
CD (1%) | 5.11 | 1.78 | 2.03 | 0.78 | 0.05 |
(* p ≤ 0.01) |
Table 4
Morphometric analysis of pollen grains from African marigold genotypes
Genotypes | Perimeter (µm) | P: Polar axis (µm) | E: Equatorial axis (µm) | Length of ectoaperture (µm) | P/E ratio | Shape of pollen |
IIHRMO12 | 98.76 de | 26.70e | 31.85cd | 6.36a | 0.84e | Oblate spheroidal, suboblate and oblate |
IIHRMO53 | 103.77c | 29.81cd | 33.54b | 5.15b | 0.89bcd | Oblate spheroidal and suboblate |
IIHRMO 2330 | 92.04f | 27.35e | 30.21e | 4.91bc | 0.91bcd | Suboblate and oblate |
IIHRMY 1–4 | 100.59cd | 29.02d | 33.37bc | 3.66d | 0.88cde | Oblate spheroidal and suboblate |
IIHRMO 2335 | 101.96cd | 30.10bcd | 32.58bc | 4.81bc | 0.92ab | Oblate spheroidal |
IIHRMY 2 − 1 | 108.32b | 31.40ab | 36.33a | 3.62d | 0.87de | Oblate spheroidal, suboblate and oblate |
IIHRMO 2356 SF | 96.25e | 30.39bcd | 31.92cd | 3.81d | 0.95a | Oblate spheroidal |
IIHRMO 2337 SF | 96.54e | 27.54e | 30.63de | 3.84d | 0.90bcd | Oblate spheroidal and suboblate |
IIHRM 2–3 SF | 114.55a | 32.15a | 35.26a | 4.18cd | 0.91abc | Oblate spheroidal, suboblate and oblate |
SE (d) ± | 1.89 | 0.65 | 0.75 | 0.41 | 0.02 | |
CD 5% | 3.72 | 1.28 | 1.47 | 0.80 | 0.04 | |
*superscript letters denote grouping through Duncan Multiple Range test |
Tocher’s method of D2 statistics was used to estimate the genetic distance between the marigold genotypes belonging to the same flower form as well as other forms. For the investigation, whether the genotypes belonging to the same flower form were grouped together, the results indicated that the genotypes associated with all the traits studied were clustered into five distinct clusters (Table 5). Five genotypes that belonged to the double-fertile and semidouble fertile groups were grouped under the first cluster, whereas the remaining four genotypes were individually separated into four different clusters without exhibiting a definite trend.
Table 5
Mahalanobis distance between African marigold genotypes based on their pollen morphometric traits
| Cluster I | Cluster II | Cluster III | Cluster IV | Cluster V |
| IIHRMO 2330 | IIHRMO 2335 | IIHRMY 1–4 | IIHRMO 2356 SF | IIHRMY 2 − 1 | IIHRMO 53 | IIHRMO 2337 SF | IIHRMO 12 | IIHRM 2–3 SF |
IIHRMO 2335 | 9.16 | | | | | | | | |
IIHRMY 1–4 | 2.44 | 7.24 | | | | | | | |
IIHRMO 2356 SF | 10.79 | 3.65 | 9.29 | | | | | | |
IIHRMY 2 − 1 | 8.21 | 8.56 | 4.45 | 9.67 | | | | | |
IIHRMO 53 | 6.30 | 11.53 | 8.98 | 8.36 | 14.71 | | | | |
IIHRMO 2337 SF | 7.67 | 5.43 | 5.44 | 12.03 | 15.34 | 13.52 | | | |
IIHRMO 12 | 13.44 | 8.99 | 10.42 | 15.16 | 11.07 | 10.63 | 12.02 | | |
IIHRM 2–3 SF | 7.40 | 7.64 | 12.61 | 13.71 | 12.78 | 13.60 | 13.93 | 13.81 | - |
The variability observed for pollen morphological traits across the genotypes is represented in a box plot for each trait (Fig. 2). A symmetric distribution was recorded only for the polar axis, while an asymmetric distribution was recorded for all the other traits for all the genotypes. The IIHRM 2–3 SF genotype exhibited the highest variability for all the traits under study, followed by the IIHRMO 2330 variety, and the lowest variability was observed for the IIHRMO 2356 SF genotype. Furthermore, for the trait polar and equatorial axis genotype IIHRMO 2330, outliers were recorded, although the range of variability was narrower.
The observed frequency distribution for pollen traits was found to deviate from the expected normal distribution, confirming the results revealed by the box plot. Instead of a single population normally distributed above the mean, two groups were observed within the pollen population of a single genotype distributed over two different mean values for the polar and equatorial axes, suggesting the possibility of pollen selection. Two groups of IIHRMY1-4 and IIHRMO 2337 SF were observed for the equatorial axis, while two groups of IIHRMO 2330 and IIHRM 2–3 SF were observed for the polar axis, revealing variability within the genotypes (Fig. 3).