The effect of planting date on growth , seed stalk development and yield of sweet fennel

To study the growth and development of sweet fennel (‘ZEFA-Fino’ and ‘ZEFATardo’) seedlings were raised under greenhouse conditions (night temperature 16—18°C) and planted out at 3—4 true leaf stage on May 21st, June 10th, July 7th and July 27th. The effects of covering mulch and short day treatment of seedlings were studied also. First harvests were performed when field measurements showed that the greatestwidth of ‘ZEFA-Fino’ bulbs were on average60 mm. Three harvests were then performed at one week intervals. Plant and bulb weights, bulb greatest width and thickness and, the heights of plant, bulb and stem inside the bulb was measured. The number of bolters were recorded. The bulb width of ‘ZEFA-Fino’ reached 60 mm in 1.5—2 months after planting out depending on planting date. Plant and bulb weights, bulb width and thickness, stem height inside the bulb and the bulb height in the earliest plantings increased as the growth period lengthened. The bulb width of ‘ZEFA-Fino’ was in higher correlation with bulb weight than bulb thickness or height. The ratio between the height of stem inside the bulb and bulb height increased fastest in the lot planted under cover in May and in the lot planted out in June; at the fourth harvest the relationship was 0.8. The largest number of bolters (8 %) was found in the latter lot. The yield estimate for ‘ZEFA-Fino’ at the third and fourth harvest was 170— 360 kg/are depending on planting date and length of the growing season. All lots of ‘ZEFA-Tardo’ except that planted in July bolted quickly after the bulb width had grown to 60—70 mm. The yield quality of ‘ZEFA-Tardo’ was low because of elongated bulbs and great variation in bulb height. Short day treatment (dark period 16h) of the seedlings retarded but did not inhibit bolting of ‘ZEFA-Tardo’. Index words: sweet fennel, planting date, growth, seedstalk development


Introduction
Sweet fennel {Foeniculum vulgare L. var.azoricum), an old vegetable in Mediterranean countries (Mansfeld 1986), is cultivated for late season and winter crop especially in Ita-  ly, but also in other South European and some North African countries (Buishand & Jonge 1984, Vogel 1986).During the last few years an interest in growing sweet fennel has arisen 237 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND in many Middle and North European coun- tries where efforts have been made to find new varieties for outdoor vegetable produc- tion.Several sweet fennel cultivars have been in field tests also in Southern Finland, where the outdoor season lasts from May to Septem- ber/October, the results, however, were not encouraging enough (Suhonen 1970) to be- gin commercial production, but the Swiss cultivar Wadenswil was accepted by home gardeners.
In growing sweet fennel during spring and summer the main problem has been the quick development of seed stalk.As a consequence of this process the "bulb" (the tight bundle of swollen petiole bases) looses its tightness, crispness and commercial value too.Breeding of sweet fennel to diminish the sensitivity to bolting has been done especially in Swit-  zerland, and a year around cultivar ZEFA- Fino was introduced in 1977 (Keller et al.  1977).According to field tests in Switzerland (Keller et al. 1977), in FRG (Andresen &  Frenz 1981), in Denmark (Henriksen 1982), in Norway (Rygg & Roed 1986), in GDR (Vogel 1987) and in Finland (Anon. 1985) 'ZEFA-Fino' yields proper bulbs during spring and summer, but even then there can be many bolters (Vogel 1987) resulting in heavy loss of commercial sales.
To avoid the bolting losses during the cul- tivation of sweet fennel, exact information of seed stalk development would be useful, but the information in literature on this subject is limited.Thus experiments were designed to study the growth pattern, seed stalk development and yield of sweet fennel in early, middle and late season.In early season the effects of a covering mulch and a short day treatment of seedlings were also examined.The experiments were carried out on the test field at our Department in 1987.

Materials and methods
The cultivars in a series of five experiments were ZEFA-Fino and/or ZEFA-Tardo, a late season cultivar (Table 1).Growing recommen- dations given by Keller et al. (1977) and by Buishand and Jonge (1984) were adapted as follows.The young plants were raised in a greenhouse with night temperature between 16-18°C.Seeds were sown in seedbeds on April 10th, May sth, June 6th and June 26th.
The seedlings at cotyledon stage were pricked into peat pots (FP 622) which were filled with fertilized peat.In experiment B two weeks short day treatment (dark period 16 h from 4 p.m. to 8 a.m.) after 6 days from pricking was given.The control plants were grown un- der natural long day.
Sweet fennels were planted out when they had 3-4 true leaves.Before planting (Table 1) plants were hardend outdoors for 3-4 days.
The planting density was 45 cm X 20 cm.The soil of experimental plots was humous coarse silt with pH 6.8.As basic fertilization of 10 kg/100 m 2 a compound fertilizer (N 7 %, P 5 %, K 15 %) was given just before planting.Additional nitrogen fertilization with cal- cium nitrate (2 kg/100 m 2) was performed when the petiole bases began to swell.Dur-  ing dry periods the experimental plots were irrigated.Weeds were controlled mechanically.
During the growth period the plants were healthy and no treatments with pesticides or fungicides were needed.
In experiment A Agryl P-17 sheet was used to study the effects of a covering mulch early in the season.The plants were covered just after planting out and sheet was taken off at the first harvest.The control plants remained un- covered.
First harvest (Table 1) was performed, when the test measurements from the field showed that the greatest width of 'ZEFA-Fino' bulbs were on average 60 mm.The following three harvests were undertaken at one week inter- vals.At harvest the stem of sweet fennel was cut just beneath the bulb.The height and weight of rosettes were measured and the number of leaves were counted.The plants were then trimmed: sideshoots and loose outer leaves were removed, and the petioles of opened leaves were cut 5 cm over the swollen leaf base, only the innermost leaves which were shorter than 7 cm were left uncut.Trimmed bulbs were weighed individually and the greatest bulb width, height and thickness were recorded.The bulb was then split and the height of the stem inside the bulb was measured.If the seed stalk had already grown out of bulb, the plant was classified as a bolter.In estimating salable yield healthy proper bulbs weighing 100-450 g were included.
All experiments were arranged according to split plot design in quadruplicates.The aver- ages presented in Tables 2-6 are means of 24 individual specimens, if no bolters were found.If bolting percentage was 50 % or more, no averages were calculated.

Results
Covering mulch on the vegetation from end of May to beginning of July increased the growth of 'ZEFA-Fino' (Table 2).When the average bulb width in covering treatment was 60 mm, the width of control bulbs was only 49 mm.Also the bulb thickness and height were greater in the covering treatment.The average plant and bulb weights at first and sec- ond harvests in the covering treatment were 53-56 % greater than the weights of control plants, but the relationship between bulb and plant weights were the same.The cover also accelerated the growth of stem inside the bulb, as well as the stem growth in relation to bulb height, however, only 0.5 °7o of plants bolted.
In the control there were no bolters.The cover especially increased the salable yield of first and second harvests.
'ZEFA-Tardo' plants grown under short day regime were less green after treatment and shorter (24 cm) at the end of plant raising period than the control plants (30 cm), but both had 4 opened true leaves on average.At the first harvest the short day treated plants were higher (47 cm) and heavier (104 g), and they had less true leaves (8.1) than the control plants (42 cm, 94 g, 9.9 true leaves).The treat- ment retarded the number of bolters and the growth of stem inside the bulb (Table 3), but it was not effective enough to ensure a proper yield.When the average bulb weight of the short day treated plants was 100 g, there were already 17 % bolters in the stand and a week later the bolting percentage was 70 %.
The bulb width of 'ZEFA-Fino' planted on June 10th and on July 7th reached 60 mm after 1.5 months growth in the field, but the last planting lot needed about 10 days more to reach the same stage (Table 4).The growth of bulb width of 'ZEFA-Tardo' was slightly slower (Table 5).
The bolting percentage of 'ZEFA-Tardo' was very high in lots planted out on June 10th and July 7th (Table 5), and even in the last group there were nearly 50 % bolters at the time of the fourth harvest.Bolters in 'ZEFA-Fino' were found only at the fourth harvest in the lots planted out on June 10th (8 %) and on July 7th (4 %), where the effective tem- Table 2.The effect of covering mulch and harvesting time on the growth of 'ZEFA-Fino' planted May 21st (experiment A).First harvest 44 days (July 34rd) after planting out, subsequent harvests at one week intervals.and significance of differences between means of M and C, Tukey's test.
Table 3.Effect of short day treatment during plant raising and of harvesting time on bolting and on properties of non-bolted bulbs of 'ZEFA-Tardo' planted out May 21st (experiment B), First harvest 44 days (July 3rd) after planting out, subsequent harvests at one week intervals.perature sums were 595°C and 546°C respectively (Figure 1).In the last planting, 'ZEFA-Fino' plants had less true leaves than in the earlier lots, but they were above average in height (Table 4).
The width and the stem height inside the bulb, and also the thickness of 'ZEFA-Fino' bulbs increased with a longer growing period, but the increase in bulb height was significant only for the lot planted out in June(Table 4).The Table 4.The effect of planting date and harvesting time on number of true leaves, plant height and on properties of bulbs of 'ZEFA-Fino'.First harvest 43 days (planted June 10th, experiment Cl), 45 days (planted July 7th, ex- periment C 2) and 56 days (planted July 27th, experiment C 3) after planting out, subsequent harvests at one week intervals.x Values followed by same letter are not significantly different at P = 0.05 by Tukey's test.

Harvest
increases in average plant and bulb weights be- tween harvests were notable especially in the lot planted out on June 10th (Figure 1).The relationship between bulb weight and plant weight varied from 0.47 to 0.62.This ratio was lowest at the time of the first harvests.The highest ratio was obtained for the lot planted on June 10th at the last harvest.The growth of bulbs in the last planting lot was very slow after the first harvest, but there still was sig- nificant increase in bulb weight (73 g) between first and fourth harvest.The increase in ef- fective temperature sum between those harvests was only 46°C.
The salable yield of 'ZEFA-Fino' increased with a lengthening of the growing period, and the greatest salable yields were achieved from stands planted out on June 10th and on July 7th (Figure 1).The yield of 'ZEFA-Tardo' was low because of bolters, and the yield quality was also poor, because of elongated bulbs and great variation in bulb heights especially for the last planting lot (Table 5).
In bulbs of 'ZEFA-Fino' there were significant correlations between bulb weight vs bulb greatest width, greatest thickness and height in all experiments (Table 6).The best corre- lation coefficients were between weight vs width.Using the average values of bulb width and bulb weight presented in Tables 2 and 4 Table 5.The effect of planting date and harvesting time on bolting and on properties of non-bolted bulbs of 'ZEFA-Tardo'.First harvests 43 days (planted June 10th), 45 days (planted July 7th) and 56 days (planted July 27th) after planting out, subsequent harvests at one week intervals.Temp, sum = the sum of effective day degrees over 5°C and in Figure 1, it can be estimated that the bulb size of 100 g was reached when the bulb width was about 64 mm and 200 g when bulb width was about 82 mm.

Discussion
The test cultivars behaved themselves very differently under the prevailing circumstances.'ZEFA-Tardo' proved to be a quick bolter in spring and in summer as was to be expect- ed, but there was also plenty of bolters in au- tumn, when it should have yielded proper bulbs.Our results with 'ZEFA-Tardo' are in agreement with the Danish results (Henriksen 1982), but differ from the Dutch results, which showed no bolters in late season culti- vation (Buishand & Jonge 1984).Short day treatment during seedling phase retarded, but did not inhibit the development of seed stalk of 'ZEFA-Tardo', which agrees with a similar experiment in Denmark (Henriksen 1982).
The non-bolted individuals of 'ZEFA-Tardo' gave small bulbs weighing under 200 g, which had relatively long stem inside the bulb.
In contrast to 'ZEFA-Tardo' bolters were only found with 'ZEFA-Fino' when the har- vest was 2 months after planting out.In sow- ing and planting time experiments carried out in Switzerland, the bolting of 'ZEFA-Fino' af- ter about 2 months growing time under field conditions was under 1 %, except in one lot planted out in beginning of July (Keller et al. 1977).In the Danish experiments there were I-ls % bolters after two months growing time under field conditions, and after 70 days the percentage of bolters was 37 % in a lot planted out at the end of June (Henriksen 1982).In field tests made in GDR the number of bolters in 'ZEFA-Fino' after about two months growing time was usually high, often over 20 %.It was stated that the speed of bolting depends on weather, and in stands planted out during June or during the two first weeks of July a high percentage of bolting can be expected (Vogel 1987).In our tests with 'ZEFA-Fino' the tendency was the same; dur- ing warmest periods of weather, and under cover the growth of stem inside the bulb was fastest, and some individuals bolted.
The field measurements of bulb width of 'ZEFA-Fino' to forecast the average bulb width in yield gave good correlations in ex- periments A and Cl, but not as good in ex- periments C 2 and C 3, where the first harvests were delayed.Obviously there were too long intervals between field measurements.According to our results, the bulb width of 'ZEFA-Fino' is highly correlated to bulb weight, and it is better forecaster for yield than bulb height or thickness.In foretelling the de- velopment of seed stalk the measurement of stem length inside the bulb seems to be a good parameter.Our observations indicate, that the relation 0.7 between stem height inside the bulb and bulb height will predict that in warm weather 'ZEFA-Tardo' is likely to bolt.In 'ZEFA-Fino' the same stem/bulb height re- lation did not give a high number of bolters under the test conditions.A warmer period at this stage could alter the behavior of this cultivar, thus further tests are required.
In our experiments the number of individu- als was insufficient for giving an exact prog- nosis of salable yields.The estimates, howev- er, indicate, that a yield level of 200-300 kg/ are for 'ZEFA-Fino' can be reached in 2-2.5 months growing time under field conditions, if young plants are raised beforehand in a hot house and the weight of 100 g per bulb is ap- proved as the lower limit.The leaf yield will be nearly the same as the bulb yield, or slightly smaller with the swelling of bulbs.
same letter are not significantly different at P = 0.05 by Tukey's test.> Great variation between individuals specially in the last planting lot, limits are given in the table.

Fig
Fig. I. Plant and bulb weights and estimated salable yield of 'ZEFA-Fino' at harvests 1 to 4 in experiment A (planted May 21st) and in experiments Cl, C 2 and C 3 (planted June 10th, July 7th and July 27th).

Table 1 .
Cultivars, sowing and planting dates, length of plant raising period and growing time on field until first

Table 6 .
Correlations between bulb weight and bulb width, thickness and height of 'ZEFA-Fino' in experi- ments A, Cl, C 2 and C3.