Environmental Project, 982 – Quantification and Identification of Active Microorganisms in Microbial Plant Protection Products

Quantification and Identification of Active Microorganisms in Microbial Plant Protection Products

4 Discussion

4.1 Quantification of active micro-organisms and comparison to available information on abundance
- 4.1.1 Bacillus thuringiensis
- 4.1.2 Fungi and actinomycetes
4.2 Identification of active micro-organisms
4.3 Contaminating bacteria

4.1 Quantification of active micro-organisms and comparison to available information on abundance

4.1.1 Bacillus thuringiensis

Within an acceptable range, all the MPPP based on Bt contained the number of micro-organisms specified by the producers. However, Løschenkohl et al. (2003) reported both lower (2.6 x 10⁸ CFU ml^-1 for Bactimos) and higher (1.1 x 10¹⁰ CFU ml^-1 for Vectobac 12 and 2.2 x 10¹⁰ ml^-1 for Dipel) number of active micro-organisms. This could indicate variation in the products over time or problems with the technique of quantification.

For the Bt products the toxicity was indicated as International Toxic Unit (ITU), which FAO and WHO recommend. According to FAO (2004), toxicity is determined by bioassays of bacterial preparations of Bt against mosquito larvae using internationally recognized reference powders. Three standard powders have been used for testing B. thuringiensis subsp. israelensis (Bti): IPS78, IPS82, and HD968 (Glare and O'Callaghan 2000). In the bioassay, the Bti products are tested against the Bti reference powder, using early fourth-instar larvae of Aedes aegypti (strain Bora Bora). The toxicity of IPS82 and HD968 has an arbitrarily toxicity of 15,000 ITU mg^-1 powder and 4740 ITU mg^-1 powder, respectively, against this insect strain (FAO 2004, Glare and O'Callaghan 2000). When the active micro-organism is a B. thuringiensis subsp. kurstaki (Btk) it should be tested against the reference strain Btk HD1 (Glare and O'Callaghan 2000), using neonate larvae of the cabbage looper Trichoplusia ni. A significant problem exists with the determination of ITU of a product as the supply of reference powders eventually is exhausted. For example, the Pasteur Institute, Paris, France, made IPS82 available. The Pasteur Institute has, however, scaled down these activities and the reference is no longer available. FAO recommends the tests to be conducted by independent laboratories and the test results as well as the reference powders made available. Due to the present need of references, activities are in progress in the international community of producers and scientists to make new reference strains and powders available (Hansen 2004).

The toxicity test with bioassays using reference powder and living insects has received criticism. Toxicity is influenced by product particle size (Skovmand et al. 1997). Additionally, the larval stage, the food fed to the larvae prior and during the bioassay, temperature, and light regime affect the results of the bioassay (Skovmand et al. 1998).

4.1.2 Fungi and actinomycetes

In seven products out of the nine MPPP based on fungi and actinomycetes that we tested, we found a lower abundance of active micro-organisms than indicated by the producers (Table 3.1). The lower abundance of active micro-organisms detected can be due to methodological problems such as:

incomplete separation of clumping spores prior to plate spreading, which results in more than one spore giving rise to a single CFU
low culturability meaning that some spores do not form a colony at the present incubation conditions.
Alternatively, the lower abundance of micro-organisms can be due to:
lower abundance of spores in the product
low viability meaning that some spores are dead and unable to be revived.

The fungal MPPP were all treated as prescribed by the producers (Table 2.1), but still incomplete separation of spores might have lowered the CFU. The extraction and separation of Bt was optimized, but still clumping was observed (Appendix a). Hence, incomplete separation of the spores is indeed likely. This does not, however, explain the difference between batches, which most likely is not due to methodological problems but reflect real differences in abundance and culturability of the active micro-organisms.

The lower abundance observed might be a problem for the efficacy of the products when used. For routine control of the quantity of active micro-organisms of both bacteria and fungi, separation of spores prior to CFU determination should be optimized by testing combinations of separation treatment. A few were tested on Bt spores (Appendix a), and this gave some indications of the achievements possible. However, it was not tested on MPPP with fungi as the active micro-organism. Direct counting of spores in the microscope circumvents the problem of incomplete spore separation. However, direct counting does not tell us anything about the viability of the spores and hence their potential survival, activity and efficacy upon use.

4.2 Identification of active micro-organisms

All 63 putative Bt isolates were identified as B. cereus group members by PCR-ITS. Upon microscopic examination only one of the isolates did not produce the crystalline inclusions containing d-endotoxin, while the other 62 did and as such were identified as Bt. Probably, the isolate not producing the crystals either did not express the toxin genes or had lost the plasmid carrying the toxin genes. Either way, it has lost its virulence at the present conditions. However, this is not anticipated to cause any relevant change in the efficacy. It does, though, highlight the chance of loss of virulence of the active micro-organisms. As the d-endotoxin production can be a metabolic burden and as the genes are positioned on a plasmid in Bt, and as such has a higher chance of being lost, this loss of virulence should be considered in risk assessment.

The active micro-organism in Mycostop was identified as Streptomyces umbrinus and not S. griseoviridis as stated by the producer. According to DSMZ, possible reasons for this difference in identification can be due to incompleteness of the identification keys. Additionally and most importantly, the identification to S. umbrinus is to a large extent based on DNA sequence homology, which is considered as one of the most important features for identification.

The active micro-organisms in Mycotal and Vertalec were not identified to Verticillium lecanii as expected based on the producer's information, but rather to Lecanicillium muscarium and L. longisporum, respectively. The reason for this is probably a revision in 2001 of the genus. However, it reflects the dynamic nature of microbial taxonomy that is still changing and developing.

The remaining fungal MPPP were identified to the expected species.

The active micro-organisms were identified to species level but not to strain or isolate level. Often the plant protecting activity is strain related and the best control of the MPPP would have included identification to strain level, if possible, and tests of plant protecting efficacy. For the Bt products the toxicity was made likely by detecting the presence of the crystalline inclusions containing d-endotoxin. For the streptomycete and the fungi no such tests of toxicity were performed.

For risk assessment these examples of changes in the species affiliation of active micro-organisms in Mycostop, Mycotal and Vertalec accentuate the necessity to use the exact same strain for experimental evaluation of risks. It also raises questions about the use of the principle of familiarity. If the risk assessment of a micro-organism is based on experience and results of closely related micro-organisms, then a change in identity of the micro-organism or segregation into several species could lead to a need of a re-evaluation of the risk assessment.

4.3 Contaminating bacteria

In the data packages submitted to the Danish EPA the producers have included information on the level of microbial contaminants expected to be present in their products (Table 3.2). In the data package submitted to the Danish EPA on Supresivit it is recommended to include the antibiotic chloramphenicol in the media when enumerating the active micro-organism (Trichoderma harzianum). As chloramphenicol inhibits bacterial growth, this is a clear indication that some bacteria are expected to be present. According to the submitted data a concentration of contaminating bacteria of 50 CFU g^-1 is expected. The number found in the present study was several magnitudes higher (2-13 x 10⁸ CFU g^-1).

From information supplied by the producer of Tri002 and Tri003, it appears that T. harzianum is added to inert montmorillonite clay, which harbors an average of 4.1 x 10⁷ CFU g^-1 bacteria and an average of 4.1 x 10⁷ CFU g^-1 fungi. Hence, it is expected that the end product also contain contaminating bacteria and fungi, which was also reported by the producer. The number of CFU found in this study was actually well below the number reported by the producer. In a study submitted by the producer, no pathogenic bacteria (E. coli, Staphylococcus, H. influenza, S. pneumonia, Vibrio sp., Yersinia sp., Salmonella, Shigella, Bacillus, and Campylobacter) were, however, found in Tri002 and Tri003 (Stasz and Hayes 1997). Batch no. 0031 of Tri003 only harbored 5.24 x 10⁴ CFU g^-1 of the active micro-organism and at the same time a similar number of contaminating bacteria (3.78 x 10⁴ CFU g^-1). This could indicate that something has gone wrong, either in the production of this batch or during storage. In the Tri002 batch no. 0326 and Rotstop the number of contaminating bacteria were just above the detection limit.

The presence of contaminating bacteria can be a disadvantage for the product. The mode of action and efficacy of the MPPP become uncertain when other micro-organisms are present. Large quantities of contaminating organisms will cause difficulties in the risk assessment, as these micro-organisms should be taken into consideration. Especially, when the number of both the active micro-organism and the number of contaminating bacteria fluctuate between batches, as was the case with Tri003, the risks and efficacy of the MPPP is difficult to evaluate.

Obviously, the presence of human, animal and plant pathogenic micro-organisms will be of greatest concern. For several of the products it is reported that no pathogenic bacteria are present. In order to clarify whether the isolated bacteria found in the fungal MPPP are pathogenic, further testing has to be performed, which is beyond the scope of this project.