mination of the conidia. Using a dilute sus-
pension of conidia plated on to a nutritive
agar medium, viable and non-viable spores
can be counted under a microscope follow-
ing incubation for a predetermined time and
temperature (Goettel, 1987; McClatchie et al.,
1994; Hong et al., 1999). Other variations,
such as liquid media (Milner et al., 1997) or
fluorescence staining (Firstencel et al., 1990;
Jimenez and Gillespie, 1990) can also be
used. However, the use of colony-forming
unit (cfu) determination from dilution plat-
ing is not recommended for viability testing
as it is prone to too many inaccuracies
(Jenkins and Grzywacz, 2000).
For microbial inoculants, efficacy is par-
ticularly difficult to measure under standard-
ized conditions, as these products tend to
confer long-term benefits on the crop as a
whole. These effects are not easily measured
over the short term and raise some chal-
lenges to manufacturers of these products.
For fungal pathogens, virulence is gener-
ally verified with a standard bioassay using
the target organism (plant, insect or fungus).
The dose/dosage used should be low to
enable differences in virulence/pathogenic-
ity to be detected, and bioassays should
always be run against a standard batch of
product as the quality of the test organism
can vary over time. For many products, it
may not be necessary to conduct bioassays
on every batch. However, a series of bioas-
says should be carried out initially to estab-
lish that batches do not vary significantly
from each other provided production condi-
tions are standardized. It should then be suf-
ficient to carry out periodic checks on
random batches, provided production condi-
tions do not change. Any changes in the pro-
duction process would necessitate a further
series of assays to identify any resultant
changes in efficacy of the product.
However, given that bioassays are notori-
ously variable and their utility in detecting
small changes in virulence/pathogenicity is
limited (Robertson et al., 1995), the determi-
nation of acceptable limits for batch-to-batch
variation is difficult. Variability in bioassays
occurs as a result of a number of factors,
including the natural variability in the host
organism and hence its susceptibility to the
fungal pathogen, variation in the dose of
the pathogen acquired and the conditions
under which the bioassay is conducted.
Steps can be taken to reduce variability;
these include maintenance of the host
organism (fungus, plant or insect) under
controlled/standardized conditions, health
monitoring of the host population (particu-
larly in respect of plants and insects), the
development of a dosing system that as far
as possible ensures that the dose applied is
completely acquired and standardization of
the bioassay incubation conditions (temper-
ature, humidity, etc.). These measures, in
addition to the use of a standard and a con-
trol against which to compare the test sam-
ple, can reduce the effect of variation in the
host population and bioassay procedure.
Thus, the performance of the test sample
should be considered in relation to that of
the standard – using a ratio, for example –
rather than in absolute terms. All bioassay
data should be archived to build up an
accurate picture of the variation in assays
over time.StorageShelf-life is a critical limiting factor in the
use and acceptance of microbial biocontrol
products. If microbial control is to compete
with chemical-control products, shelf-life
should be as long as possible and, prefer-
ably, should not require refrigeration or spe-
cial treatment (Jones and Burges, 1998). A
number of factors affect the shelf-life of a
fungal product. Nutrition and culture condi-
tions during fermentation have already been
highlighted in earlier sections; in addition,
methods of downstream processing and har-
vesting, product moisture content, formula-
tion and packaging all play an important
role in successful long-term storage (Hong et
al., 2000).
Significant progress has been made over
recent years in the understanding of conidial
longevity (Hong et al., 1997, 1998, 1999;
Burges, 1998b). These studies have shown
that conidia of Metarhizium anisopliaevar.
acridum(= Metarhizium flavoviride) behave in
a similar way to seeds when stored under252 N.E. Jenkins and D. Grzywacz