EFFECTS OF DIFFERENT SUBSTRATES ON BASIL SEEDLINGS QUALITY ( Ocimum basilicum L . )

In this study five different substrates were used, such as: compost; mixture of compost, Lumbrikus H and garden soil; mixture of compost and Lumbrikus H; mixture of compost, Lumbrikus H and peat Galicina and Seedling Klassman substrate. Basil seedling was produced in containers according to "speeding" system. The studies have shown that the best quality of basil seedling of variaties Genovese and Lattuga is achieved when the mixture of substrates Compost, Lumbrikus H and Galicina peat are applied in the volume proportion of 50% : 30% : 20%.

application in the vegetable and fruit growing production a long time ago (M a r k o v i ć et al., 1992; H a n i ć , 2000).In our country medicinal, aromatic and seasoning herbs seedlings are still produced in a classical way -the so-called bare root (plants with unprotected root system) in cold and warm seedbeds.Garden soil is most frequently used as a substrate, and in this kind of production plants suffer from stress during transplantation, and as a result take longer time for rooting and continuation of growth and development.
Contemporary production of medicinal species of herbs in the world is conducted in different containers according to the system of protected root system, with the use of different growing substrates (M c G i n n i s et al., 2004; A t i y e h et al., 2000; F e r n a n d e s et al, 2004; T e s i et al., 1995).The selection of suitable substrate is highly significant for the production of seedlings (B u r e s et al., 1993).
The purpose of this study is to find the most favourable substrate for container production of basil.
The experiment was conducted in the green-house of the Faculty of Agriculture of Belgrade during 2005.The basil seedling was produced in containers according to "Speeding" system in three repetitions.Containers made of styrofoam, 53 cm x 31 cm x 5,5 cm in size, with 40 cells each were used.The volume of one cell is 76 cm 3 .On the surface of 1 m 2 there are 148 cells.
Sowing basil seeds in containers, previously filled with disinfected, substrates was conducted on March 26, 2005 with two seeds in each cell.After cropping one plant was left in each cell.During the experiment normal measures of seedling care were applied: watering, shading and ventilation.
Seedling production lasted 56 days when the plants were in the phase of 3-4 pairs of leaves.
Prior to the analysis (measuring), the plants underwent the procedure of "tempering".By random sampling method, 32 plants were selected from each variant of substrates.In analyzed plants height (cm), fresh and dry root (g) and root volume (cm 3 ) were measured by the method of submerging root into a measuring cup with water.
The analysis of experimental results was conducted by means of descriptive and analytical statistics with statistical package STATISTICA.
Of the central tendency indices arithmetical mean ( X ) and mediane (M e ) were calculated.Data variability was quantified through interval of variation (I v ), standard deviation (S), standard error ( ) x S and coefficient of variation (C v ).
With the aim of rendering objective conclusions on the influence of applied substrates on the quality of basil seedlings, homogenity of variances analyzed treatments was tested with Levene's test as well as the congruence of data distribution with the normal distribution model with Kolmogorov-Smirnov's, Lilliefors' and Shapiro-Wilks' test.
Testing the differences between the treatments was conducted by analysis of variance (ANOVA), lsd-test, t-test, Kruskal-Wallis' and Mann-Whitney's test.
Since the quality of the substrates was observed through the effects on six characteristics, discrimination effect was used to determine synthetical rank, expressed by I-distance.

Results and Discusion
Based on the presented results through the main statistical indices in Tab. 1, the lowest average values for all analyzed parameters of the quality of basil seedling, variety Genovese, were obtained on Compost (substrate A).The greatest effect was shown in average values of the tested parameters with the use of Compost + Lumbrikus H + Galisina peat (substrate D).
Similar results were obtained by T e s i et al. (1995) by growing basil on the substrate of Cultural F of dark peat, with the use of different treatments, densities and doses of fertilizers.
The obtained average values for analyzed parameters: dry plant mass and dry root mass were in accordance with the studies of M c G i n n i s et al. ( 2004) who experimented with different doses of fertilizers and the share of vermincompost in substrates.
Coefficients of variation (C v ) for all substrates and all analyzed indices have values less than 30%, which means that empirical series contain homogenous data.
T a b.Distribution of extreme values coefficients of variation for characteristics and varieties is shown in tables 3 and 4.
In the variety Lattuga, the highest average values (tab.2) were obtained with the mixture: Compost + Lumbrikus H + Galicina peat (substrate D), while the lowest values were obtained on Compost (substrate A).
In comparison to the results of research of T e s i et al. 1995.somewhat higher average values were obtained for the tested parameters.
Also, in the variety Genovese, coefficients of variation (C v ) for all substrates and all analyzed indices have values less than 30%, which means that the empirical series contain homogenous data.The distribution of extreme values of coefficients of variation for characteristics and varieties is shown in tables 3 and 4.
Objective conclusions on the effect of substrates on the quality of basil seedlings can be drawn based on the results of testing hypotheses on the equality of average values of characteristics, through which their quality is shown.The established hypotheses can be tested with parametrical and non-parametrical tests.The assumption for the application of the statistical method of analysis of variance (ANOVA) that each subset must be normally distributed was tested with Kolmogorov-Smirnov's, Lilliefors' and Shapiro-Wilks' test.The test results show that this prerequisite was not fulfilled in all cases.
Another assumption that the samples derive from the main sets with equal variances was tested with Levene's test.
The obtained results (tab. 3 and 4) show that the variances are nonhomogenous for all tested characteristics, except for the plant height of the basil of variety Lattuga and dry root mass in the plants of variety Genovese.Therefore, the assumptions of parametrical methods of testing were disrupted.
Bearing this in mind, testing the differences of treatments was conducted with non-parametrical test, but also with parametrical tests.
The results of ANOVA and the adequate non-parametrical test, Kruskal-Wallis' test, are identical and show that the average values of all tested characteristics of the seedling quality depend highly significantly on the applied substrate.However, in comparison with the results of individual tests (lsd, t and Mann-Whitney) the effects of some substrates on individual characteristics of quality do not differ at all, or differ significantly, but not highly significantly.
Therefore, as shown in table 3, with the use of substrate D and E plants of variety Lattuga are obtained, which do not differ significantly in their height.
Based on the results of lsd-test, we can conclude that with the use of substrate A, the same average mass of fresh root is obtained as with the use of substrate B or C, as well as that the substrates B and C have the same effect on the dry root mass of the variety Lattuga.
The relations of effects of substrates on other tested characteristics are shown in tables 3 and 4.

Root volume Min
C-E no significant p>0.05 significant 0.01<p<0.05 The order of the substrates based on average values is not identical for all analyzed quality characteristics, so total discrimination effect was used to determine synthetical rank in the work, expressed with I-distance.The rank determined, based on I-distance, is the same for both varieties.
The best quality of seedlings was achieved with the use of substrate D, followed by E, then C, then B and the lowest quality of seedling was achieved with the use of substrate A.

C o n c l u s i o n
In the production of medicinal, aromatical and seasoning herbs seedlings, the existing technologies are developed and new technologies are discovered.
With the use of natural peat, compost and other materials improved by adding different materials of organic and mineral origin in the form of available nutrients and growth biostimulators.R e z i m e U radu je ispitivano pet različitih supstrata kao što su; kompost; mešavina komposta, Lumbrikusa H i baštenske zemlje; mešavina komposta i Lumbrikusa H; mešavina komposta, Lumbrikusa H i treseta Galicine i Seedling Klassman supstrat.

T a b. 3 .
-Results of analytical statistics for the variety

Slavica Jelačić 1 ,
11. P u t i e v s k y , E., D u d a i , N., L e w i n s o h n , E., R a v i d , U. (2001): Cultivation and production of new species in the mediterranean, World Conference on Medicinal and Aromatic Plant, Abstract-Map Hungary, 57.Budapest 12. T e s i , R., G h i s e l l i , L., T a l l a r i c o (1995): Ricerche sulla coltivazione del basilico in contenitore Colture Protete, N. 12, p.61-66.Received September 5, 2005 Accepted November 15, 2005 UTICAJ RAZLIČITIH SUPSTRATA NA KVALITET RASADA BOSILJKA (Ocimum basilicum L.) Nada Lakić 1 , D. Beatović 1 i Ana Vujošević 1
1. -Results of descriptive statistics for variety Genovese