This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
INTRODUCTION
Iris dichotoma and I. domestica are herbaceous perennials belonging to the family Iridaceae, with a diploid chromosome number of 2n=2x=32 native to China, and have a long flowering period in summer and high resistance to cold, drought and barren conditions, which make them excellent materials for iris breeding (Zheng et al., 2017). Therefore, I. dichotoma and I. domestica have been attracting the attention of plant breeders (Bi et al., 2012). In nature, I. dichotoma is found with violet, yellow and white flowers, but no red or orange flowers. Its individual flowers stay open for about 4 h in form the late afternoon to the early evening (Luo et al., 2016). To improve its flower colour and flowering time, interspecific crosses were conducted to introgress useful genes from I. domestica to I. dichotoma and this has resulted in some novel varieties (Yang et al., 2013; Lian et al., 2016; Ruan et al., 2017; Xu et al., 2017; Gao et al., 2020). Previous studies concentrated on violet and white-flowered I. dichotoma plants (Chimphamba, 1973; Ruan et al., 2017), whereas I. dichotoma with yellow flowers has not been comprehensively investigated. The F2 generation of I. dichotoma (violet-flowered) and I. domestica shows great variation in flower colours, but not with yellow (Lenz, 1972). Therefore, creating new germplasm with yellow-flowered I. dichotoma is a significant goal of Iris-breeding programs.
Pollen fertility is very important for plant development and reproduction (Consolaro et al., 1996; Conterato et al., 2006). Variations in pollen fertility are usually attributed to abnormal chromosome pairing and abnormal meiotic division (Yabuya, 1991). Abnormal meiosis, such as early chromosome migration, lagging chromosomes, bridge, unequal separation, micronuclei, multiple spindles, nonsynchronous division, unequal separation and so on, was observed in many plants, especially in interspecific hybrids (Wang et al., 2010; Hu et al., 2021). Compared with their parents, interspecific hybrids of Dendranthema had a higher frequency of chromosome bridges and lagging chromosomes (Cui et al., 2007). Approximately 56.5% of the cells of allotriploid lily ‘Cocossa’ demonstrated meiotic abnormalities at telophase II (Zhang et al., 2017). Sometimes, the high frequency of abnormal meiosis was also found in intraspecific hybrids. The hybrids of Hydrangea aspera subsp. aspera Kawakami group × subsp. sargentiana showed a higher frequency of meiotic abnormalities (48.2%) compared with the two parents and resulted in reduced pollen fertility (Crespel and Morel, 2014). Generally, the fertility of interspecific/intraspecific hybrid could be restored by successive backcrosses. So far, there is no information on pollen viability and meiosis of I. dichotoma, I. domestica and their hybrids. Therefore, meiotic behaviour and pollen fertility of yellow-flowered I. dichotoma, I. domestica and their F1, F2 and BC1 hybrid progenies were investigated in this study. Furthermore, the relationships between meiotic abnormalities and crossbreeding were discussed. The knowledge acquired will contribute to the rational use of these genetic resources.
MATERIALS AND METHODS
Plant materials
Yellow-flowered I. dichotoma (Y2) was collected from Pinglu, Shanxi Province, and I. domestica (S3) was obtained from the Shenyang Botanical Garden, Shenyang, Liaoning Province, China, in 2007. Both of them flower from July to late August, and their flowering periods overlap extensively. The populations of F1(Y2 × S3), F2[(Y2 × S3)⊗], BC1-Y[(Y2 × S3) × Y2] and BC1-S[(Y2 × S3) × S3] were established from 2009 to 2014, and 242, 126, 130 and 121 individuals were obtained, respectively. According to their fertility, we selected some of them for this study (Figure 1). Meiosis behaviour and pollen viability were studied in two parents, one F1 hybrid (F1-5), three F2 hybrids (F2-1, F2-2 and F2-3), four BC1-S hybrids (BC1-S-1, BC1-S-2, BC1-S-3 and BC1-S-4) and two BC1-Y hybrids (BC1-Y-1 and BC1-Y-2).
Analysis of male meiosis
For analyses of meiosis, 2 mm, 4 mm, 6 mm, 8 mm and 10 mm length of young flower buds were collected between 8:00 am and 10:00 am in the middle of July 2015 and 2016 (field temperature about 29–31°C), fixed in a mixture of absolute alcohol and glacial acetic acid (3:1) for 24 h and then stored in 70% alcohol at 4°C. The anthers were squashed in 5% carbol fuchsin and examined under a microscope (Motic BA400, China).
Pollen viability and size observation
Flowers were collected at anther dehiscence and the pollens were dispersed in the laboratory for 3 h (relative humidity <30%; temperature 24°C). The pollens were cultured on a solid medium (150 g ⋅ L−1 sucrose + 100 mg ⋅ L−1 H3BO3 + 100 mg ⋅ L−1 CaCl2 + 3 g ⋅ L−1 agar, pH 5.8) at 25°C in an incubator (RLD-350D-4, China) for 2 h. At least 200 pollen grains per individual were observed under a microscope (Motic BA400, China). A pollen grain was considered to have germinated if the length of the pollen tube was equal to or longer than the pollen diameter.
Pollen size and morphological observation
Pollen grains were directly attached to double-sided adhesive tape and coated with gold. For pollen diameter, at least 1,000 pollen grains were measured for each accession. Microscopic observation and image acquisition were conducted using a scanning electron microscope at a magnification of 500× (TM3030, Japan).
RESULTS
Abnormal meiotic chromosome behaviours
During nor mal microsporogenesis, simultaneous meiosis was observed in I. dichotoma, I. domestica and their hybrid progenies: cytokinesis started only after telophase II (after the division of the four nuclei) and resulted in four microspores (tetrad), each containing one nucleus. Figure 2A–2T shows photomicrographs of successive meiotic stages, from interphase I to the pollen stage. One of the striking observations of the interphase I stage in this study was the diffuse appearance of diplotene (Figure 2E, 2F). The diffuse stage in diplotene was observed in I. dichotoma, I. domestica and all of their hybrids. At the beginning of the diplotene, nuclei were filled with extended and fuzzy single chromosome threads, corresponding to the diffuse stage described in many plants, and they stained only faintly or not wholly. Following the diffuse stage, the chromosomes contracted markedly (Figure 2G).
In each stage of meiosis, several frequent meiotic abnormalities were recorded. Across all cells analysed, the abnormal meiosis rates in most hybrid progenies were much higher than that of their parents. Over 50.0% of the cells exhibited abnormalities in F2-1 (57.3%), BC1-S-1 (58.7%), BC1-S-2 (54.5%) and BC1-S-3 (52.7%), in contrast to 26.2% in I. dichotoma, 22.0% in I. domestica, 39.5% in F1-5, 41.4% in F2-2, 34.2% in F2-3, 47.7% in BC1-S-4, 25.8% in BC1-Y-1 and 23.3% in BC1-Y-2 (Table 1).
The meiosis abnormalities rate (%) in I. dichotoma, I. domestica and their hybrid progenies.
Phase abnormalities
Plant materials
Y2
S3
F1-5
F2-1
F2-2
F2-3
BC1-S-1
BC1-S-2
BC1-S-3
BC1-S-4
BC1-Y-1
BC1-Y-2
Metaphase I
288
241
200
130
184
210
162
207
165
80
142
150
Non-congression
114 (39.6%)
94 (39.0%)
83 (41.5%)
57 (43.8%)
80 (43.5%)
85 (40.5%)
98 (60.5%)
126 (60.9%)
78 (42.3%)
30 (37.5%)
40 (28.2%)
27 (18.0%)
Anaphase I
216
170
156
104
155
140
113
120
150
126
115
143
Lagging chromosomes
82 (37.9%)
35 (20.6%)
53 (34.0%)
11 (10.6%)
42 (27.1%)
31 (22.1%)
46 (40.7%)
50 (41.7%)
60 (40.0%)
41 (32.5%)
24 (20.9%)
14 (9.8%)
Bridge
19 (8.7%)
0 (0.0%)
28 (17.9%)
12 (11.5%)
5 (3.2%)
18 (12.9%)
3 (2.7%)
6 (5.0%)
11 (7.3%)
15 (11.9%)
5 (4.3%)
10 (7.0%)
Telophase I
195
152
152
123
121
127
107
102
109
88
120
167
Unequal separation
22 (11.3%)
8 (5.3%)
35 (23.0%)
64 (52.0%)
13 (10.7%)
9 (7.1%)
47 (43.9%)
24 (23.5%)
0 (0.0%)
25 (28.4%)
27 (22.5%)
45 (26.9%)
Micronuclei
7 (3.6%)
0 (0.0%)
3 (2.0%)
0 (0.0)
9 (7.4%)
12 (9.4%)
10 (9.3%)
4 (3.9%)
19 (17.4%)
5 (5.7%)
3 (2.5%)
0 (0.0%)
Metaphase II
160
160
154
230
131
140
101
142
164
142
150
150
Non-congression
1 (0.6%)
14 (8.8%)
24 (15.6%)
67 (29.1%)
28 (21.4%)
19 (13.6%)
26 (25.7%)
24 (16.9%)
30 (18.3%)
22 (15.5%)
10 (6.7%)
5 (3.33%)
Abnormal spindle
3 (1.9)
2 (1.3%)
41 (26.6%)
76 (33.0%)
6 (4.6%)
17 (12.1%)
14 (13.9%)
54 (38.0%)
61 (37.2%)
44 (31.0%)
25 (16.7%)
34 (22.7%)
Anaphase II
220
188
202
221
194
200
214
281
310
235
250
269
Lagging chromosomes
31 (14.1%)
26 (13.8%)
12 (5.9%)
42 (19.0%)
51 (26.3%)
26 (13.0%)
14 (6.5%)
51 (18.1%)
43 (13.9%)
54 (23.0%)
22 (8.8%)
46 (17.1%)
Bridge
6 (2.7%)
0 (0.0%)
14 (6.9%)
8 (3.6%)
11 (5.7%)
10 (5.0%)
0 (0.0%)
0 (0.0%)
17 (5.5%)
6 (2.6%)
0 (0.0%)
3 (1.1%)
Abnormal spindle
5 (2.3%)
6 (3.2%)
17 (8.4%)
46 (20.8%)
27 (13.9%)
11 (5.5%)
64 (29.9%)
30 (10.7%)
71 (22.9%)
16 (6.8%)
6 (2.4%)
2 (0.7%)
Not synchronous division
14 (6.4%)
11 (5.9%)
33 (16.3%)
74 (33.5%)
39 (20.1%)
23 (11.5%)
70 (32.7%)
90 (32.0%)
65 (21.0%)
75 (31.9%)
29 (11.6%)
14 (5.2%)
Telophase II
223
172
200
150
125
183
151
138
152
295
166
121
Unequal separation
43 (19.3%)
42 (24.4%)
76 (38.0)
91 (60.7%)
62 (49.6%)
67 (36.6%)
82 (54.3%)
72 (56.5%)
96 (63.2%)
98 (33.2%)
33 (19.9%)
21 (17.4%)
Micronuclei
8 (3.6%)
0 (0.0%)
1 (4.5%)
1 (0.7%)
4 (3.2%)
12 (6.6%)
24 (15.9%)
6 (4.3%)
2 (1.3%)
30 (10.2%)
19 (11.4%)
12 (9.9%)
Total
1,302
1,083
1,064
958
910
1,000
848
990
1,050
966
943
1,000
Abnormalities rate
26.2%
22.0%
39.5%
57.3%
41.4%
34.2%
58.7%
54.5%
52.7%
47.7%
25.8%
23.3%
During the first meiotic division, the presence of non-congressed chromosomes that did not align on the equatorial plate was observed at metaphase I in all 12 accessions (Figure 3A, 3B). The consequent behaviour of these univalents depended on the genotype and resulted in retarded movement to the poles and the formation of micronuclei. At anaphase I, lagging chromosomes were found in all 12 accessions. In BC1-Y-1 and BC1-Y-2, the univalents joined the large chromosome sets at the poles, although they were retarded in their movement (Figure 3C). In contrast, lagging chromosomes formed new micronuclei in BC1-S-3. In this phase, chromosomal single and double bridges were found (Figure 3D, 3E). In BC1-S-1, chromosome stickiness resulted in thick chromosomal bridges, restricting chromosome separation (Figure 3F). Chromosome bridges were completely absent in I. domestica and appeared only rarely in F2-2 and BC1-Y-1. At telophase I, micronuclei derived from lagging chromosomes and/or acentric fragments were found (Figure 3H). F2-1 and BC1-S-1 showed relatively high numbers (52.0% and 43.9%, respectively) of cells with unequal chromosome segregation (Figure 3G).
During the second meiotic division, abnormalities related to the non-congressing chromosomes were observed in I. dichotoma, BC1-Y-1 and BC1-Y-2 with a low frequency at metaphase II. In other genotypes, this percentage varied between 8.5% and 29.1% (Figure 3I, 3J). Apart from normal spindles that had a linear alignment, abnormal spindles in meiosis II were also found. When abnormal organisation occurred, the most common conformations were a tripolar spindle (Figure 3K), followed by the perpendicular formation of the convergent zone (Figure 3L). At anaphase II, lagging chromosomes (Figure 3M), chromosome bridges (Figure 3N) and tripolar and perpendicular spindles were observed again (Figure 3O–3Q). Another spindle abnormality was the occurrence of multiple spindles in BC1-S-3 (Figure 3R). In F1-5 and F2-3, chromosome separation was asynchronous at anaphase II (Figure 3S, 3T). Chromosome disequilibrium segregation was also found at anaphase II and telophaseII (Figure 3U, 3V). New micronuclei formed at telophase II, which resulted in the formation of polyads (Figure 3Y).
At the tetrad formation, the second meiotic division mostly ran perpendicular to the first, causing the four microspores to be orientated as shown in Figure 4G. Rarely, the resulting four microspores were arranged in a row during this stage. One of the striking observations of the tetrad stage in this study was the sporad configurations in hybrids (Table 2 and Figure 4) including monads, dyads, triads, tetrads and polyads, reflecting low pollen fertility. Monads were mainly observed in F2-1, but not viable (Figure 4A). Dyads and triads were found in most of the individuals at the end of microsporogenesis, indicating the production of unreduced gametes. Triads (composing of one 2n microspore and two n microspores) might result from the formation of disoriented (tripolar) spindles. However, the size of microspores in BC1-S-2 and BC1-S-3 was very variable and several pentads, hexads or abnormal tetrads (containing three large and one small microspores) were observed (Figure 4I–4O).
Distribution (%) of sporad configurations in I. dichotoma, I. domestica and their hybrid progenies.
Materials
No. of cells studied
Monad Rate (%)
Dyad Rate (%)
Triad Rate (%)
Tetrad Rate (%)
Polyad Rate (%)
Y2
245
1.9
4.3
10.6
83.1
0.0
S3
213
0.0
3.8
13.6
78.4
4.2
F1-5
253
0.4
5.1
5.5
50.2
38.7
F2-1
208
10.1
36.5
16.4
35.1
1.9
F2-2
201
6.0
0.0
16.0
72.1
6.0
F2-3
260
0.0
7.7
30.8
57.7
3.9
BC1-S-1
432
2.3
14.6
33.8
45.4
3.9
BC1-S-2
254
1.2
7.1
29.1
51.2
11.4
BC1-S-3
271
5.5
11.1
33.2
46.1
4.1
BC1-S-4
373
0.0
4.6
37.0
53.1
5.4
BC1-Y-1
363
1.7
10.5
34.2
47.9
5.8
BC1-Y-2
319
1.3
15.7
34.5
44.5
4.1
Pollen viability and size variation
The analyses of pollen viability and size in I. dichotoma, I. domestica and their hybrids are shown in Table 3. F2-1 and BC1-S-1 were male-sterile and did not have any pollen. The average pollen fertility of F1-5 (10.1%) was much lower than that of its parents, whereas F2-2 produced few pollen grains, and they rarely germinated. Pollen abortion rates of I. dichotoma and I. domestica (1.9% and 14.5%, respectively) were lower than that of the interspecific hybrids except for BC1-S-4 (13.3%). The mean pollen fertility of BC1-S-2 (1.6%) was the lowest besides F2-1 and BC1-S-1 without pollen. The pollen germination rates of F1-5 (10.1%), F2-2 (2.4%) and F2-3 (21.5%) were lower than that of I. dichotoma (26.1%) and I. domestica (35.1%). In BC1-S-2 (1.6%) and BC1-S-3 (4.8%), almost no pollen grains germinated. BC1-S-4 pollen appeared normal but germinated poorly with only 7.2% of pollen germination. However, BC1-Y-1 and BC1-Y-2 demonstrated higher pollen germination rates than others except for I. domestica (Figure 5).
Pollen germination rates of I. dichotoma (Y2), I. domestica (S3) and their hybrid progenies.
Materials
Pollen in plant
Normal pollen size (μm)
Large pollen (%)
No. of pollens observed
No. of pollens germinated
No. of abort pollens
Abortion rate (%)
Germination rate (%)
Y2
More
65–75
0.5
1,713
447
33
1.9
26.1
S3
More
60–70
0.2
2,706
951
392
14.5
35.1
F1-5
More
60–70
0.8
2,039
205
738
36.2
10.1
F2-1
None
–
–
–
–
–
–
–
F2-2
Little
63–73
0.7
1,354
32
628
46.4
2.4
F2-3
More
60–70
1.1
1,014
218
323
31.9
21.5
BC1-S-1
None
–
–
–
–
–
–
–
BC1-S-2
Little
61–71
1.7
582
9
311
53.4
1.6
BC1-S-3
Little
62–72
1.7
673
32
334
49.6
4.8
BC1-S-4
More
58–68
1.9
1,049
76
139
13.3
7.2
BC1-Y-1
More
65–75
2.0
1,899
526
696
36.7
27.7
BC1-Y-2
More
63–73
2.3
1,916
597
631
32.9
31.2
Large pollen grains: >1.3× diameter of normal-sized pollen.
The normal-sized pollen grain diameters of I. dichotoma, I. domestica, F1-5, F2-2, F2-3, BC1-S-2, BC1-S-3, BC1-S-4, BC1-Y-1 and BC1-Y-2 were 65–75 μm, 60–70 μm, 60–70 μm, 63–73 μm, 60–70 μm, 58–68 μm, 61–71 μm, 62–72 μm, 65–75 mm and 63–73 μm, respectively. In addition to single pollen grains, including regular smaller ellipsoidal and larger spherical pollen grains, a few adherent pollen grains were viable. Furthermore, some large pollen grains (>1.3 times the diameter of normal-sized pollen grain) were observed in I. dichotoma (0.5%), I. domestica (0.2%), F1-5 (0.8%), F2-2 (0.7%), F2-3 (1.1%), BC1-S-2 (1.9%), BC1-S-3 (1.7%), BC1-S-4 (1.7%), BC1-Y-1 (2.0%) and BC1-Y-2 (2.3%). As large grains are an indication of 2n pollen production, 2n gametes might be present in these plants (Figure 6C–6F), which have potential for polyploid breeding, especially in BC1-Y-2, BC1-Y-1, BC1-S-2, BC1-S-3 and BC1-S-4. Although meiotic aberrations in these large pollen producers resulted mainly in triads, monads, dyads and polyads were also observed.
DISCUSSION
The most striking form of the diffuse stage that we observed in these hybrids was the diffuse diplotene stage. To our knowledge, this phenomenon was found in the Iris species for the first time. The same phenomenon has also been reported in Rosa and large genome cereals (Colas et al., 2017) as well as in Larix (Kołowerzo-Lubnau et al., 2015). Especially in interspecific hybrid plants, the diffuse stage has been proposed to be widespread in plant meiosis, affecting restitution from diffuse to fully contracted chromosomes, with chromosomes often entering metaphase before fully contracting (Klášterská and Natarajan, 1974). In this study, it occurred in summer with high temperatures and drought stress. In Larix species, meiosis calmed during the autumn and pollen mother cells remained in the diffuse stage for several months to resist the cold (Ekberg et al., 1968). This possibly reflects that the diffuse stage is influenced by some extraordinary environmental conditions.
The high frequency of meiosis abnormalities was observed in BC1-S-1 (58.7%), BC1-S-2 (54.5%), BC1-S-3 (52.7%), BC1-S-4(47.7%) and F2-1(57.3%), with low pollen-germination rates (0, 1.6%, 4.8%, 7.2% and 0, respectively). Pollen fertility analysis of different cross combinations in this study indicated that meiotic abnormalities, such as non-congression, lagging chromosomes, unequal separation and abnormal spindle fibres orientation, led to low pollen fertility. As previously reported, pollen viability is closely associated with meiosis of pollen mother cells, and meiotic abnormalities affect pollen development (Boff and Chifino-wittmann, 2002; Zha et al., 2007; Farco and Dematteis, 2014; Usandizaga et al., 2020). The relationship between meiotic abnormalities and fertility has been studied in many plant species, including Musa and Brassica (Adeleke et al., 2004; Yang et al., 2020). Meiotic abnormalities are shown to be negatively correlated with pollen fertility (Devi and Borua, 1997).
The pollen fertility of F1-5 (10.1%) was much lower than that of its parents I. dichotoma (26.1%) and I. domestica (35.1%). Backcrossing F1-5 to I. dichotoma resulted in rapid restoration of fertility, and the pollen fertility of BC1-Y-1 and BC1-Y-2 was higher than that of F1-5. However, when backcrossing F1-5 to I. domestica, the BC1 hybrids had a high total meiosis abnormality rate (exceeding 50.0%) and produced no or shrunken pollen with relatively low pollen fertility (0 to 4.8%) compared with F1-5 (10.1%). Low fertility of BCl plants was also reported in Agropyron junceum (Charpentier et al., 1986) although backcrossing can restore the fertility of interspecific hybrids (Yushkina et al., 2009). F1-5 was male fertile and presented only 1–4 unpaired chromosomes. Partial fertility of the interspecific hybrid could be associated with relatively high chromosome association frequency, indicating partial genomic homology between the parental species (Asano, 1984). The abnormal segregation of chromosomes in meiosis I and II can lead to the formation of micronuclei, which either may remain in microspore tetrads or may be eliminated in the form of microcytes for additional cytokinesis. In this case, the formed tetrads will be genetically unbalanced, and the resulting pollen will become dysfunctional or shrunken. Although tetrads were found in F2-1 and BC1-Y-1, they did not grow cell walls and form pollen grains. Therefore, the frequency of 2n pollen was low in I. dichotoma, I. domestica and their hybrids which exhibited a high frequency of dyad and triad. Additionally, microspores in tetrads could originate from an unequal chromosome division, usually leading to sterile pollen (Do Nascimento et al., 2014).
In this study, the lagging chromosomes, unequal separation and nonsynchronous division led to chromosomes slowly moving to cell poles, forming the micronucleus or disappearing in the cytoplasm, and finally led to the production of unbalanced tetrads or polyad and abnormal pollen grains. Similar meiotic aberrations have been reported in Brachiaria (Gallo et al., 2007), Begonia (Dewitte et al., 2010), Hydrangea aspera (Crespel and Morel, 2014) and allotriploid lily ‘Cocossa’ (Zhang et al., 2017). They can be contributed to chromosome pairing problems due to their lack of homology and are the expression of the typical behaviour of univalent or non-paired chromosomes. In some rare cases, unequal and incomplete nuclei separation resulted in monads, such as asymmetric cytokinesis in pollen mother cells in which the entire nucleus was included in one of the cells (Barba-Gonzalez et al., 2005). In this study, the presence of dyads and/or triads at the tetrad stage imply the production of 2n pollen grains. Abnormal spindle fibres orientation is associated with the production of 2n gametes. The formation of disoriented spindle fibres led to a dyad (composed of two 2n microspores) in the case of fused or parallel spindle fibres, and to a triad (composed of one 2n microspore and two n microspores) in the case of tripolar spindle fibres. The production of 2n gametes was observed in interspecific hybrids of potato (Conicella, 1991), Populus (Dong et al., 2015) and strawberry (Luo et al., 2018).
The spontaneous occurrence of unreduced gametes can produce unexpected polyploids such as triploids and tetraploids in both natural and artificial crosses in various plants such as Lilium (Van Tuyl et al., 2005), Tulip (Marasek-Ciolakowska, 2012) and Primula (Kato et al., 2008). When they occur in interspecific hybrids with some extent of viability, 2n gametes can be used to produce sexual progeny through either crossing or selfing (Ramanna and Jacobsen, 2003; Van Tuyl and Lim, 2003). In the case of meiotic polyploidisation, polyploid progenies are raised by crossing with 2n gamete-producing genotypes. The most important advantage of meiotic polyploidisation is that homologous recombination occurs at a high frequency between parental chromosomes during meiosis. Therefore, a large amount of heterozygosity can be transmitted to the polyploid offspring. Because of the function of unreduced gametes, a considerable number of Iris garden varieties with large flowers, improved form and plant habit, and fully fertile were produced through intercrossing by meiotic polyploidisation (Bee et al., 1986). In this study, the selection of individuals producing 2n gametes allowed manipulating the ploidy level of the offspring. Because unreduced pollen is not always viable, as previously demonstrated in Begonia (Dewitte et al., 2010), screening for viable unreduced gametes should be undertaken in the future, as these cells present an interesting tool for polyploidisation and can be used as pollen donors to produce polyploid iris individuals exhibiting larger flowers as well as higher pollen and seed fertility (Barba-Gonzalez et al., 2004; Nikoloudakis et al., 2018).
CONCLUSIONS
In summary, this study analysed the relationship between meiotic abnormalities and pollen fertility in 12 Iris individuals comprising parents and hybrids resulting from their crosses and backcrosses. Meiotic abnormalities negatively affected pollen fertility. The diffuse diplotene stage was observed for the first time in the genus Iris. Poor pollen fertility of BC1-S presented difficulties for further breeding. As all BC1-Y individuals displayed relatively similar flower colour (yellow) and higher pollen fertility than other cross combinations, they are excellent plant material for yellow-flowered iris breeding. In addition, BC1-Y-2, BC1-Y-1, BC1-S-2, BC1-S-3 and BC1-S-4 have potential in polyploid breeding.