Agronomical and phytochemical investigation of Hyssopus officinalis

Hyssop (Hyssopus officinalis L.) obtained from various commercial sources was grown for three years (1990-1992) in Finland. Yield characteristics, flower colour, volatile oil content/composition and its antimicrobial quality were studied. For comparison, Scottish-grown hyssop was included in oil and quality determinations. The description of the flower colour given by the seed firms was not a reliable indicator of the true colour in many cases. Oil yield was satisfactory and oil composition was rather uniform. Only one different chemotype was identified, this was derived from a Romanian seed source. There was considerable variation in herb yield between plants from different sources. The total fresh herb yield was 0.5-3.2 kg/m : , the dry leaf yield was 67-326 g/m 2 . Seed germination was satisfactory (76-99%), offering opportunities for seed production of varieties with different characteristics.


Introduction
Due to increasing interest in growing and using herbs in the northern parts of Europe, several re- search projects have been carried out in both Finland and Scotland during the last few years (Galambosi et al. 1991, Hay et al. 1988,  Svoboda et al. 1990).Hyssop was one of the 40 herb species which were studied in a five year research project at Puumala, Southern Finland.This species proved to be both cold and frost tolerant with good dry matter yield and volatile oil content (Galambosi et al. 1989).The flowering tops and leaves ofhyssop are used as flavours in the food and drink industry and in various cosmetic products (Genders 1980).It is also a traditional medicinal plant (Bonar 1985, Fleischer and Fleischer  1988), an excellent plant for attracting bees (Hooper 1984) and an attractive garden ornamen- tal (Sanecki 1985).Several types, differing in flower colour, flowering time and leaf shape are available commercially; alba (white flowers), grandiflora (large flowers), rosea (rose flowers) and rubia (red flowers) (Simon et al. 1984).The seed samples are often mixed and it is quite difficult to obtain uniform plant populations for specific re- quirements, such as decorative flower production, honey bee forage production, high volatile oil yield and uniform quantitative oil composition.
The objective of this study was to test in Finland 13 different seed samples of various geographical origins for the variability of colour, growth, fresh and dry matter yield and seed production.For com- parison, hyssop grown in Scotland was included in oil yield and quality determinations.In addition, the antimicrobial activity of the oil was tested against a group of 25 bacterial species.

Growth conditions
The plants were grown at South Savo Research Station, Mikkeli, Finland (grid reference 61°44 N, 27°1 8 E) during 1990-1992 and in the herb garden of the Scottish Agricultural College (55°28 N, 4°3 3 W) during 1990-1992.The meteorological data for Mikkeli are presented in Figures 1 and 2. The origin of seed samples and the colour of the plants grown in Mikkeli are given in Table I.Hyssop seeds Nos.14-21 were obtained from Poyntzfield Nursery, Black Isle, Scotland, and were of French origin.The seeds were sown in pots (5x5 cm diameter) filled with fine peat on 26 April 1990.The pots were kept in a plastic greenhouse and the seedlings transplanted to the field on 8 June.One year old plants of the varieties Nos. 2, 3 and 4 were transplanted into experimental plots from Puumala on 4 June 1990.The density of planting was four plants per m .

Cultivation and fertilization
The soil in Finland was a stony till, pH 6.2.The experimental plots were fertilized before planting (N 35, P 120 and K7O kg/ha) with further N (15 kg/ha) three weeks after planting.The same basic fertilizer mixture was applied at the beginning of the second year.The plants were irrigated twice during the first and once during the second growing season.
In Scotland, the experimental plots were located on a sandy loam soil of pH 5.8.No artificial fertilizers were applied.Farmyard manure was added each autumn.Nos.19-21 were grown in a polytunnel throughout the whole season.

Harvesting and drying
In Finland, plants were harvested each summer in August, during the full flowering period.From each of the varieties 10plants were cut and the following characteristics were determined: colour of flowers, plant height, fresh and dry weight, and leaf:stem ratio.In Scotland, individual, well-established 3 year old plants were collected randomly during the full flowering period.Fresh samples were dried at 35°C and the stems were separated from the leaves through a 3 mm diameter screen.Five plants from each variety were grown for seeds.The seeds were harvested at the end of the vegeta- tion period (25 September 1990, 10 September 1991, 24 September 1992).The hand cut plants were dried at room temperature (18-22°C) and the seeds were crushed either by hand, or by using an experimental harvester (Hege 125 C, Germany).The germination tests were carried out each year 3 months after harvesting, using the top paper method in 9 mm Petri dishes, at 20-23°C day and 17-19°C night temperature, with 4 x 50 seeds per variety.

Distillation of oil
Dried leaves and flowering tops were steam dis- tilled for 2 h using British Pharmacopoeia distilla- tion apparatus (BSI 1985).The quantity of oil ob- tained was measured and the oil was then transferred to glass vials with Teflon-lined caps and stored in a refrigerator at -2 to 6°C until analysed by GC.

GC analysis ofoil
GC was carried out using a United Technologies Packard 439 GC connected to a Hewlett Packard Integrator 3390A.The following operating condi- tions were used: Carbowax 20M column, 25m x 0.32 mm; carrier gas N 2; injection temperature 250°C; flame ionisation detector temperature 250°C; oven temperature initially 50°C, rising to 200°C at 5°C/min; sample size 0.2 1; splitter 1:100.
Standard oil components for comparison were ob- tained from Roth (Karlsruhe, Germany).

Antibacterial properties of volatile oils
For the determination of antibacterial properties of the volatile oil from hyssop, wells were punched in pre-seeded Isosensitest agar plates and to each well was added 15 ml volatile oil (Deans and Ritchie  1987).This was allowed to diffuse into the agar prior to incubation at 25°C for 48 h, after which zones of growth inhibition were measured with vernier calipers.Three wells per plate were made and two replicate plates tested per organism.

Results and discussion
Variation in colour of hyssop flowers Of the 13 seed samples grown in Mikkeli, two had no colour indication.No. 3 proved to be a red and No. 7 a blue coloured hyssop (Table 1).
Six samples had the colour as advertised on the commercial packing or as the original mother plant.The colour of the flowers of five samples was dif- ferent from the advertised description: Nos 1, 10, 11, 12 and 13.Since all seeds were from commer- cial sources, the quality control clearly needs to be improved.

Growth characteristics
The meteorological conditions during the experimental years did not differ significantly from the long term average (Figs.1-2).Frost damage was observed only in acquistion No. 7, which resulted in lower plant weight during the second growing sea- son (Table 2).This indicates the frost tolerance of hyssop (Galambosi et al. 1989).The lower pre- cipitation during September was advantageous for seed ripening.
The differences in the plant height and weight during the consecutive seasons were clear; the av- erage height of the one year old plants was 48 cm, of two year old plants 58-59 cm and of three year old plants 67 cm (Table 2).The smallest variety was No. 3 (Romania) reaching 42 and 47 cm dur- ing the first and second year respectively.This variety had a typically compact habit.The tallest varieties were consistently Nos. 4,6and 7, reaching 62-75 cm of height.However, these plants were loose in habit and the heavy rains often caused lodging.
Table 3.Total fresh weight yield and dry leaf yield of hyssop varieties at different ages (Mikkeli, 1990-1991)

Herb yield
The results of the plant weights presented in Table 2 show significant variability between the sources.The average weight ofplants grown from seeds was 313 g and 386 g during the first and second year, respectively.The transplanted one year old plants were significantly heavier during the second and third season, reaching 407 g and 586 g of fresh weight per plant, respectively.The total fresh weight yield varied between 0.5 and 3.2 kg/m depending on plant age (Table 3).The lowest yield was obtained from the variety No.
3, producing 0.5 (first year), 1.1 (second year) and 1.3 (third year) kg/nr fresh weight.The highest yields were derived from No. 6, producing 1.9,2.7 and 3.2 kg/m 2 of fresh weight throughout the three seasons.
The leaf dry weight (Table 3) varied from 67 to 367 g/m 2 .Consistent with plant vigour and devel- opment, the lowest yield was obtained from the variety No. 3 and the highest yield from the variety No. 6, but the differences were less pronounced than those of fresh yield.The dry matter content varied from 22-30%.The leaf: stem ratio was about 1:1, the first year plants having less, the second and third year old plants having more stems (Table 4).The marketable leaf and flower dry weight yield calculated from the dry matter content and leaf:stem ratio was 10-14% of the total harvested fresh yield.The lowest dry weight yield (10%) was measured in tall, loose, pink and white varieties, due to their high stem and twigs contents.  Mikkeli 1990-1991).

Seed quality
The germination tests showed that the hyssop vari- eties were consistently of good seed quality.The Table 5. Seed germination and thousand seed weight (TSW) of hyssop varieties (Mikkeli, 1991).average germination capacity of seeds was 87% (Tables 5 and 6).The lowest results were achieved by white hyssop (76%), being a late flowering type.
The seed production could be easily mechanized.There were 10-15% differences in the germination of the hand crushed and mechanically crushed seed.The differences could be eliminated by optimization of the harvest times.The average thousand seed weight of hyssop varieties was 1.0 g, ranging between 0.75 g and 1.1 g (Table 5).The pink hys- sop (No. 2) had the lowest seed weight.
Quantitative variations in the volatile oil yield Volatile oil contents of leaves and flowering tops varied for 0.4 to 1.4% (Table 7).Blue varieties grown in Finland had a range of oil yield: 0.7-1.08%.The blue coloured hyssop No. 7 had the highest oil content (1.36%).Mixed coloured vari- eties had a range of 0.94-1.2%,white 0.6-0.7%,pink and red 0.6-0.8%.There were no differences between one and two year old plants.
Oil contents of Scottish-grown blue coloured plants showed values of 0.4-1.4%.There was no difference between plants grown in the open field or in the polytunnel.No clear pattern was observed in the oil content fluctuations, but this high variabil- ity within the individual plants suggests the possibility of improving the oil yield (0.3-1.6%) (JOULAIN  and Ragault 1976, Hilal et al. 1978, Mechraz  et al. 1989).The leaf oil content from samples grown both in Finland and in Scotland was above average.Khodzimatoc and Ramazanova (1975) report unusually high oil contents: about 1.6% from red, 2.2% from blue and 3.7% from white varieties.
In our experimental plants, these levels of oil content were not realized, neither were there any significant differences in oil yield between various types of colour, although the relatively low oil contents of the white and pink varieties have to be noted.
Maximum oil contents were found during the full flowering period, with stems containing a negligible amount ofoil (Kapelev 1986,Timchuk et al.  1986).The results emphasize the importance of selection and of fertilizer use for improved types in aromatic plant species (Svoboda et al. 1990).Hyssop can be an appropriate crop for the northern areas, with very good quality and quantity of the final product.

Quantitative variations in the volatile oil
The volatile oil composition showed similar results for both Finnish and Scottish material, with one exception; red hyssop originating from Romania (Table 7).This oil had significantly higher amounts of germacrene-D (22-23%) and pinocarvone (26- 28%) compared with the other samples.Iso-pinocamphone, pinocamphone, (3-pinene, pinocarvone, germacrene D and 1,8-cineol were the main com- ponents of the oil and accounted for 75-85% of total 1.5 0.7 19c (S)  1.00 0.9 52.8 11.9 6.1 1.7 0.7 20c (S)  0.50 oil.Two major components, pinocamphone and iso-pinocamphone constituted about 50% of total oil and their representative proportions varied in a manner which could not be explained in the exist- ing experiments.Variations in plants derived from different geographical origins emphasize the importance of further selection studies.
Our results are in agreement with earlier reports (Lawrence 1980, Steinmetz et al. 1980, Lawrence, 1984, Timchuk et al. 1986, Galambosi et   al. 1989, Mechraz et al. 1989, Schulz and Ståhl  1991).Only one reference (Khodzhimatov and  Ramazanova 1975) mentioned different hyssop chemotypes, with high amounts of 1,8-cineol, li-nalool, a-terpineol, a-terpinyl acetate and bomyl acetate (identified by GC analysis).It is possible that hyssop collected in the Taskent area showed intraspecific chemical differences.Pinocamphone, iso-pinocamphone, camphor and thujone are res- ponsible for the toxicity of the oil (Steinmetz et al.  1980).The experiments were conducted with rats and both the oil and individual components caused nerve and muscle damage, resulting in epilepsy.The above mentioned monoterpenes are very volatile and their chemical structure may change under different environmental conditions and through postharvest handling.These facts could partially explain the variability of individual oil components A Acinelobacler calcoacelica; B Aeromonas hydrophila; C Alcaligenesfaecalis; D Bacillus subtilis; E Beneckea nalriegens; F Brevibacterium linens; G Brocolhrix thermosphacla; H Citrobacter freundii; I Clostridium sporogenes; J Enterobacter aerogenes; K Enterococcus faecalis; L Erwinia carotovora: M Escherichia coli; N Flavobaclerium suaveolens; O Klebsiella pneumoniae; P Lactobacillus plantarum; Q Leuconostoc cremoris; R Micrococcus luteus;S Moraxella sp.; T Proteus vulgaris; U Pseudomonas aeruginosa; V Salmonellapullorum; W Serralia marcescens; X Staphylococcus aureus; Y Yersinia enlero- colitica.
described by different authors.Incorrect identifica- tion of compounds, using GC methods and standard comparison, is also possible.

Antibacterial characteristics of volatile oils
There was a varying response to the volatile oil in terms of antibacterial properties (Table 8).There were clearly some antibacterial constituent(s) present since a number of bacteria showed noticeable growth inhibition, including Acinetobacler cal- coacetica (spoilage organism), Aeromonas hydrophila (an environmental organism found in water courses which can also be a pathogen of fish), Breviobacterium linens (spoilage organism found in soft cheese), Brocothrix thermosphacla (spoilage organism found in pork sausage), Klebsiella pneumoniae (human pathogen) and Serratia marcescens (secondary opportunist pathogen).
There is no obvious explanation as to why hyssop plants of different geographical origin gave strong inhibition against some bacteria but not others, and equally, why certain bacteria were only susceptible to a number of hyssop volatile oils: the chemical analysis of the oils did not reveal wide variation in the components present.

Table 1 .
Identity and origin of varieties/populations of hyssop grown in Mikkeli. .

Table 4 .
Dry matter content and leaf/stem ratio in hyssop varieties at different ages

Table 7 .
Main components of hyssop oil derived from whole plants grown in Finland (F) and Scotland (S).
= one year old plants; b = two year old plants; c = three year old plants.# = trace amounts. a

Table 8 ,
Antibacterial properties of hyssop volatile oil (inhibition zone diameter in mm; diameter of well, 4 mm, included).