Effects of three antibiotics (cefotaxime, rifampicin and gentamicin) were tested on in vitro shoots cultures of tansy (Tunacetum vulgare L.). These antibiotics were selected because endophytic bacteria isolated from the in vitro shoot cultures of tansy showed that all bacteria were Gram-negative and their growth was reduced by these three antibiotics. Five isolates were Enterobacteriaceae, three were fluorescent Pseudomonas, and two were aerobic bacteria. Increased concentrations of antibiotics caused usually linear or quadratic changes on the initiation of shoot growth, shoot number, growth rate, and shoot height. These changes and changes in pH of the culture media were tansy genotype-dependent following treatments with gentamicin. Also, the treatment with rifampicin or cefotaxime showed a genotype-dependent effect, because they resulted in significantly higher percentage of rooted plants in one of the three tansy genotypes tested. The growth rate and length of shoots were reduced in the media containing both gentamicin and rifampicin, but less so than in media containing both gentamicin and cefotaxime.

Fungi and Gram-negative bacteria were isolated from inside the roots of fieldgrown goat's rue (Galega orientalis). Fungi were isolated from three plants out of a total of 45 tested. Two multinuclear Rhizoctonia solani isolates were identified to the anastomosis group 5 ( R . solani AG-5-Gal) using pairings with known AG test cultures. One fungal isolate was identified to Phoma chrysanthernicola. Gram-negative bacteria were isolated from three plants out of 25 tested. They were identified using classical methods, the BIOLOG identification system based on the utilisation of 95 different carbon sources, and the MIDI system for the analysis of whole cell fatty acids. The two latter systems were computer-associated and utilised an extensive reference library of isolates. One bacterial isolate was identified as Enterobacter agglomerans and two isolates as Pseudomonas marginalis.R. solani AG-5-Gal reduced the emergence of Lupinus luteus, L . polyphyllus and french bean (Phaseolus vulgaris) and the growth of broad bean (Viciafaba), L. luteus and french bean, but did not cause obvious damage in goat's rue and pea (Pisum satiuum). However, R. solani AG-5-Gal was re-isolated from the roots of all the test plant species following inoculation. P. chrysanthemicola reduced the emergence of L. polyphyllus and the growth of goat's rue, french bean and broad bean, and it was re-isolated from all of the test plant species (except for french bean) following inoculation.All the bacteria reduced the emergence of french bean, but not that of goat's rue and pea, when applied to the soil. When the roots were dipped into bacterial suspension, all the bacteria damaged french bean and L . polyphyllus. Additionally, P. marginalis JV3 damaged goat's rue and red clover. The pathogenicity of the fungi and bacteria were not changed when they were doubleinoculated in pairs, except for R. solani AG-5-Gal and P. marginalis JV2 which reduced the emergence of goat's rue when inoculated together but not when inoculated separately.

Ten bacteria were isolated from the in vitro shoot cultures of different tansy (Tanacetum vulgare) genotypes. All isolates were Gram-negative. Five isolates belonged to Enterobacteriaceae, three isolates were strains of fluorescent Pseudomonas, and two isolates were strains of other aerobic bacteria. The combined treatment with gentamicin (50 mg·L–1) and rifampicin (25 mg·L–1) prevented the growth of all bacteria, whereas none of the antibiotics (ampicillin, cefotaxime, rifampicin, gentamicin, or streptomycin) controlled all bacteria when used alone. The antibiotics ranged based on their increasing adverse effects on shoot cultures as follows: rifampicin, cefotaxime, and gentamicin. Increased antibiotic concentration reduced the initiation and growth rates of shoots, roots and calli with some exceptions. Low concentrations of gentamicin and cefotaxime both increased the growth rate and shoot number per plant. Rifampicin stimulated the root growth of some of the biotypes tested. The growth rate of calli increased significantly in the presence of both gentamicin (25 mg·L–1) and rifampicin (25 mg·L–1) in the growth media, whereas the use of either antibiotic alone reduced the calli growth relative to the control. Viability of tansy protoplasts was enhanced by rifampicin and to a lesser extend by gentamicin and cefotaxime. The growth of bacteria isolated from tansy tissue culture can be prevented with a combined gentamicin and rifampicin treatment. The growth rate of plants decreased slightly when grown in the media supplemented with the same antibiotic combination. However, the growth retardation was not permanent, and the plants recovered and grew vigorously when transferred to antibiotic free medium.

M . -L . S A V E L A A N D M . U O S U K A I N E N . 1994.Meristem-tip cultures of apple rootstock 'YP' were started at different times of the year over a period of 2 years and the contamination of the cultures was monitored during five subcultures. Bacterial contaminants were isolated to pure cultures, identified by the API test system and appropriate additional tests, and the sensitivity of the most common isolates to different antibiotics was determined. Of the 216 strains isolated from the initiation cultures, 78?40 were pseudomonads, coryneforms or enterobacteria. Only three bacterial contaminants were found at the multiplication stage. A greater part of the contaminants were likely to originate from the stock plant. Rifampicin (at 50-200 m g 1-I) and cefotaxime (at 25C1500 m g 1-I) were found to be bactericidal against many isolates, b u t differences between species and strains were found.