2.1. Plant Biometrical and Growth Indices, Seed Yield, and Germinability
As regards the biometrical parameters (
Table 1), both bulb planting time and density showed a significant effect on the number of stalks per plant, as well as their height and diameter, inflorescence (umbel) diameter, and leaf area index (LAI).
The flower stalk number per plant decreased with the planting time delay, from 4 units under the 20 December planting to 2.7 with the 21 February planting. This trend is consistent with the reports of research carried out by El-Helaly and Karam [
8] in Egypt.
Moreover, the number of flower stalks decreased from 3.6 with two bulbs per m
2 to three under the five bulbs per m
2 density. The stalk height also showed a decreasing trend with the planting delay, with 13.6% reduction from the earliest to the latest planting time. Moreover, the bulb density increase from two to five per m
2 caused the stalk height reduction from 112.7 to 105.9 cm. The flower stalk diameter decreased from 41.9 mm with the first planting to 35.5 mm under the last one. The increase from the lowest to the highest bulb density resulted in 10.4% reduction of the stalk diameter. Our research findings are consistent with those recorded in previous research [
8,
9,
20].
The umbel diameter was also adversely related with the planting time, as the highest value was recorded under the earliest planting (81.2 mm) and the lowest was detected with the latest (63.8 mm). The highest umbel diameter was recorded with mid-November planting and the lowest in mid-January, as the earliest planting enhanced flower stalks growth under favorable climate conditions and provided the umbels with a higher nutrient supply, leading to flowering anticipation [
20].
Mollah et al. [
21] reported that, out of five planting dates ranging from 1 October to 30 November at 15-day intervals, the 15 November planting time led to the tallest plants and the highest number of stalks (4.65) and umbels per plant, as well as the highest umbel diameter, though it did not statistically differ from 30 October and 30 November plantings.
In the present research, the decreasing trend of flower stalk diameter with the bulb density raise (from 42.2 to 37.8 mm) confirmed previous findings [
7].
The earliest planted crops showed a higher leaf apparatus and dry weight accumulation compared to that attained by the later crops, as can be observed in
Table 2. This result is due to the longer growth time span of the crops planted on 20 December, which also accumulated a higher biomass amount. Similar results were obtained in previous research carried out on storage onion cultivar Ramata di Montoro for bulb production in southern Italy [
4].
Leaf area and dry matter per plant decreased with the bulb density rise, from 935.3 cm
2 per plant and 73.8 g, respectively, under two bulbs per m
2 to 838.8 cm
2 per plant and 59.6 g, respectively with five bulbs per m
2. Conversely, leaf area and dry weight per m
2 were enhanced by the bulb density increase, from 0.28 m
2·m
−2 and 147.6 g, respectively, under the lowest density to 0.63 m
2·m
−2 and 298 g, respectively, with the highest one. Contrastingly, Caruso et al. [
4] did not record the significant effect of plant density on leaf area in bulb-oriented onions.
The two experimental factors, bulb ‘planting time’ and ‘density’, showed significant effects on yield components, ratio between seeds and umbel weight, and seed germinability (
Table 3).
Both seed yield per plant and per hectare were significantly affected by the interaction between bulb planting time and density (
Figure 1 and
Figure 2). Indeed, in both cases the two earliest planting times (20 December and 10 January) showed the best results at any bulb density, whereas the latest planting (21 February) resulted in the lowest yield per plant. Similarly, the two lowest bulb densities (2.0 and 2.5 bulbs per m
2) led to the highest yield per plant at any planting time, but they were not significantly different from 3.3 plants per m
2 density in correspondence with the two latest plantings. Otherwise, yield per hectare showed the highest values, with 3.3 bulbs per m
2 under the earliest planting time, and even with 5.0 bulbs per m
2 upon the second planting, whereas it increased from 2.0 to 5.0 bulbs per m
2 in the two latest crop cycles.
The interaction between bulb planting time and density was also significant on the seed number per plant and per square meter (
Figure 3 and
Figure 4). The former variable showed the same trends as described for yield per plant, whereas the trends of seed number per square meter only differed from the previous yield indicator because of the increase from 2.0 up to 5.0 bulbs per m
2 under the two earliest plantings and up to 3.3 bulbs per m
2 upon the two latest plantings.
The mean seed weight was significantly affected by the interaction between planting time and plant density (
Figure 5). The two earliest planting times resulted in the highest values with 2.0 to 3.3 bulbs per m
2, whereas they gave only better results than the latest planting correspondently to 5.0 bulbs per m
2. Moreover, upon 20 December planting, the lowest bulb density led to higher mean seed weight than the two highest densities; under 10 January planting, only the 5.5 bulbs per m
2 treatment was less effective than the other densities, and no significant difference between the treatments were recorded in correspondence with the two latest planting times.
Several correlations between the yield, growth, and biometrical variables examined were significant (
Table 4). In particular, the seed yield per plant was positively correlated with both the seed number (r = 0.99 at
p ≤ 0.01) and mean weight (r = 0.89 at
p ≤ 0.01), whereas the seed yield per ha was not significantly correlated with the seed mean weight. Moreover, seed germinability showed significant positive correlations with all the mentioned parameters, except for LAI, which suggests the enhancement of this seed feature with the improvement of plant growth and productive conditions.
In previous research [
22], the 10 cm spacing along the row was found more suitable both for obtaining higher onion seed yield and lower intensity of crop management labor in Sudanese environmental conditions, compared to 2.5 and 5.0 cm.
Mollah et al. [
21] reported that, out of five planting dates ranging from 1 October to 30 November at 15-day intervals, the 15 November planting time resulted in the highest values of leaf area index (3.97) and seed yield (1.13 t·ha
−1), both due to seed number per umbel (300) and mean seed weight (4.2 mg), as well as seed germination rate (84%); the lowest values were obtained with the two plantings practiced in the first half of October.
Khodadadi [
23] in Iran found that 6 November led to taller plants and higher seed yield of onions compared to 5 March. Aminpour and Mortazavibak [
24] reported that Texas Early Grano 502 cultivar in Isfahan showed the highest number of umbels and seed yield (1.4 t ha
−1) when planted on 22 September. The latter date was found as the most appropriate for onion vegetative growth under the Karaj temperate climate, whereas the 21 November planting best fitted the Bangladesh tropical conditions [
25].
El-Helaly and Karam [
8] observed on cultivar Giza 20 grown in Egypt that the highest number of seed stalks, umbels, and seed yield per plant, as well as total seed yield, were associated to 15 November planting and the lowest to 15 January.
In other research [
26], plant height, as well as the number of flowering stalks and of umbels per plant, decreased from 22 September to 5 November; however, seed yield was best affected by the planting dates ranging from 22 September to 6 October or from 21 October to 5 November, depending on the cultivar, and showed the same trend as the number of capsules per umbel and of seeds per capsule. Notably, yield increase resulted from the reduction of flowering stalk number and concurrent rise of umbel diameter, whereas the decrease recorded on 5 November was caused by adverse high temperature during July–August which were unfavorable for pollination, seed setting, and development [
26]. In a different area [
27], flower pollination decreases due to spring high temperatures and insecticide use which negatively affect seed setting and production. Khokhar [
6] reported that the time from planting to inflorescence emergence and to floret opening increased linearly from the earliest planting date of 5 December (148 and 187 days, respectively) to the latest of 5 March (167 and 206 days, respectively); the highest seed yield was obtained under the earliest planting (5 December), when the conditions were more favorable for inflorescence initiation, with flowering occurring in early June and seed ripeness in mid-August. Mohamedali and Nouri [
28] recorded the optimal planting time between mid-October to mid-November for onion seed yield, and the decrease of seed production upon later sowing was mainly caused by flower abortion increase.
The most effective temperature range for promoting the inflorescence primordia initiation is 9 °C to 13 °C [
29]; however, cultivars differ in their optimal temperature requirements depending on the location to which they are adapted. Indeed, varieties from the northern or southern Russia have an optimal temperature range of 3 °C to 4 °C or of 9 °C to 10 °C, respectively. Though some authors [
30] reported no significant effect of photoperiod on inflorescence initiation, in other research the time to this phase occurrence under inductive temperatures (9–13 °C) was reduced by increasing the photoperiod from 8 to 20 h per day in several cultivars [
31,
32,
33]. In this respect, Brewster [
32] reported the need of 86-day and 38-day exposure under 8 and 20 h photoperiod, respectively, in cultivar Rijnsburger; however, other varieties showed the requirement of over 14–16 h and, in cultivar Senshyu, the low sensitivity to photoperiod. In other investigations [
6], the varieties showed different behavior, being inducted to inflorescence initiation under a short photoperiod (8 h d
−1) at 11 °C to 13 °C, or under short to intermediate daylength (11 h d
−1 to 14 h d
−1) at 11 to 19 °C.
Following primordia initiation, a wider temperature range of 10 °C to 20 °C favors the emergence of inflorescence [
34], which is fastest at 20 °C [
35]. Over 20 °C, flower stalks fail to emerge [
30], even upon the planting of sets with well-developed inflorescence initials [
36]. Indeed, temperature exceeding this threshold halts inflorescence emergence directly when applied under short day conditions [
30], and indirectly under long days as it promotes bulbing [
37].
Once an inflorescence has been initiated, the rate of stalk elongation increases with rising temperature from 10 °C to 30 °C and a lengthening photoperiod to 14–16 h per day [
33,
35,
38,
39], both in over-wintered and spring-sown onion cultivars [
40]. Contrastingly to each other, in a research study, Bertaud [
35] found that the inflorescence emergence rate was faster under a 14 h photoperiod at 20 °C, whereas in other investigations 15 h at 10–16 °C was the most effective treatment [
33].
Brewster [
38] also reported that from 20 °C to 30 °C, both the growth of emerged flower stalks and the seed ripening show a 4–5 week anticipation compared with plants kept at 15 °C to 20 °C; however, under 16 h day length in plants with developed inflorescence initials, the inflorescence number is reduced, thus resulting in a larger proportion of shoots forming bulbs rather than inflorescences. Brewster [
34] recorded the opening of florets 11 days earlier at 22 °C as compared to 16 °C, and Van Kampen [
41] found that the temperature of 30 °C during the early stalk development can result in flower abortion because of assimilate competition with the bulbing process, though this is combined with the appropriate cultivar day length requirement.
The percentage of flower stalks with well-developed inflorescence initials and the floret number per umbel showed decreasing and increasing trends, respectively, upon increasing mean temperature from 14 °C to 23 °C and day length from 11 to 17 h per day [
40,
42], with the latter temperature and photoperiod trends, the time from planting to inflorescence initiation, spathe opening, flowering, and seed ripening decreases being a linear function of both climate parameters. Contrastingly, short days (10 h) can even halve the number of emerged inflorescences compared to plants grown under long days (14 to 18 h), also reducing flower stalk growth and flower number per umbel [
29]; however, in some cultivars the number of plants with emerging inflorescences are increased by a 10 h photoperiod [
36].
The ratio between seeds and inflorescence weight showed a decreasing trend with the planting time delay, with the highest value of 20% under the first planting being not significantly different from the second one, and the lowest being of 12.5% with the last crop cycle (
Table 3). Moreover, this variable value decreased when the bulb density raised from two to five bulbs per m
2 (from 17.5 to 16.8%, respectively).
As far as seed germinability is concerned (
Table 3), the planting time delay caused significant decreases, as the highest value was recorded under the 20 December planting (98%) and the lowest with the latest planting performed on 21 February (74%).
Bulb density affected the seed germinability, which showed the highest value of 89% under the two bulbs per m2 treatment and the lowest of 82% with five bulbs per m2.
In previous investigations [
43], the germination rate ranged between 80 and 90%. According to Maciel and coworkers [
44], the genotype plays an important role on seed germinability in interaction with temperature, which is deemed crucial even by other authors [
45]. Mollah et al. [
21] reported that the 15 November planting time resulted in the highest seed germination rate (84%), whereas 1 and 16 October had the lowest.