Professional Documents
Culture Documents
DOI 10.1007/s10529-005-1779-9
Key words: mycelial cultivation, optimization, Pleurotus ostreatus, response surface analysis, whey
permeate
Abstract
A novel approach to utilizing whey permeate, the cultivation of mycelia of the edible mushroom Pleurotus
ostreatus, is introduced. Response surface analysis (RSA) was successfully applied to determine the com-
bination of substrate concentration, temperature and pH that would result in a maximal mycelial extension
rate under solid state cultivation. The conditions to maximize the mycelial extension rate were predicted to
be 44 g lactose l)1, pH 6.0 and 24.2 C. Subsequent verification of these levels agreed with model
predictions.
Whey permeate and mycelial strain For this experiment, ranges of temperature and
pH for growth of P. ostreatus were based on liter-
Dried whey permeate powder was dissolved in ature values (Hatvani 2001). Because whey per-
distilled water to obtain differing concentrations meate has not previously been used for the
of lactose (Table 1). Since the purpose of the re- mycelial cultivation of P. ostreatus, we performed
search was to provide information about the preliminary experiments to estimate the maxi-
treatment of raw cheese whey permeate with the mum growth rate of P. ostreatus in various con-
mushroom mycelia, no additional nutrients were centrations of whey permeate using a modified
added. Commercial agar (Becton Dickinson and bisectional method (Chapra & Canale 1998). The
Co, Sparks, Md., USA) was added to the differ- radial extension rates of P. ostreatus at different
ent concentrations in the ratio of 1.5% (w/v) and substrate concentrations were then fitted to an
the solutions were then mixed and autoclaved at equation suggested by Shi et al. (1999) (Figure 1).
120 C for 20 min. Before pouring the solution The r-squared value for this modified equation
into petri dishes, the pH was adjusted by addi- was 0.92 and the standard error of estimate was
tion of 0.5 M HCl or 0.5 M NaOH as needed to 0.09. Analysis of model parameters was signifi-
meet the experimental parameters in Table 1. cant for all terms at 5% a level. Thus, it was con-
The solution was then poured into petri dishes cluded that the model adequately described the
and allowed to solidify for solid state cultivation effect of whey permeate concentration on mycelial
of the mycelia. growth of P. ostreatus. The substrate concentra-
Pleurotus ostreatus (KCTC 6735, Korean Col- tion that maximized the mycelial growth rate was
lection for Type Cultures) was transferred to pet- assumed to be 50 g lactose l)1, which was used as
ri dishes containing PDA media and incubated at a center point for RSA model building.
25 C for 4 days. Mycelial agar discs (5 mm) A total of 19 trials, including a center point,
were used as inocula for subsequent experiments. were run to approximate the response surface for
the mycelial production of P. ostreatus. To find a
maximum response (i.e. highest radial extension
Analysis of mycelial growth using response surface
rate) in the RSA, increasingly complex equations
methodology
from linear to partial cubic were sequentially tes-
ted to model the data obtained from the trials in
Growth of mycelia was on PDA medium in petri
Table 1. When the data were analyzed using the
dishes. Growth was measured as the diameter of
various models, the p-value of regression was sig-
the colony in four different places along lines
nificant at the 5% a level while lack of fit was
crossing at right angles and the average value
not significant at the same confidence level only
was used as the size of the colony.
for the partial cubic model (Equation 1).
The radial extension rate estimated was as-
sumed to be the mycelial growth rate under the g ¼ 104 þ 1:5x1 þ 27:9x2 þ 2:5x3
solid state culture at given conditions (Sang et al.
4:6 101 x1 x2 1:2 102 x1 x3
2001). These rates were then used to perform
response surface analysis (RSA) along with their 3:1 101 x2 x3 8:5 104 x21
ð1Þ
corresponding environmental conditions and 1:9x22 2:8 102 x23 þ 2:9
obtain the optimal conditions that would maxi-
mize the mycelial growth rate. The experiment 104 x22 x23 þ 2:8 103 x1 x2 x3
(Table 1) was based on a 32 orthogonal design þ 3:4 102 x1 x22
(pH, temperature and substrate concentration
at two levels each) with a center point being where
replicated 5 times as previously described g: experimental value of the mycelial growth
(Adinarayana et al. 2003, Lee et al. 2003). This rate (mm d)1)
type of design was used to minimize the number of xk: independent variable k (k = lactose
trials needed to obtain statistically relevant results. concentration, pH and temperature in order)
1539
Independent variables
Linear 1 40 5 20 0.7
design 2 60 5 20 0.5
3 40 7 20 0.7
4 60 7 20 0.4
5 40 5 30 0.8
6 60 5 30 0.9
Fig. 1. Observed and predicted mycelial growth rate of Pleu- 7 40 7 30 0.2
rotus ostreatus at different lactose concentrations: (d) ob- 8 60 7 30 1.0
served mycelial growth rate; (–) model predictions. An 9a 50 6 25 1.3±0.2
equation suggested by Shi et al. (1999) was used to fit the Quadratic 10 40 6 25 1.3
radial extension rates of P. ostreatus with different substrate
design 11 60 6 25 0.8
concentrations.
12 50 5 25 1.1
13 50 7 25 1.1
The regression coefficient and residual standard 14 50 6 20 0.8
deviation of the partial cubic model were 0.92 15 50 6 30 0.6
and 0.17, respectively. This indicated that the Validation 16 44.1 6 24.2 1.7±0.1
partial cubic model was an accurate representa-
tion of the data and that curvature existed within
the defined response surface. Therefore, this show a definite peak and constant contour lines,
equation was used to analyze the response sur- suggesting that the control variables involved are
face and determine a set of conditions that not interdependent. The third interaction (i.e.,
would maximize mycelial growth rate by setting lactose concentration temperature), however,
the partial derivatives of the equation to 0 with was significant on the mycelial growth at the 5%
respect to the independent variables. The condi- a level. This indicates an interaction where the
tions determined as providing maximal mycelial values obtained for lactose concentration and
growth rate were 44.1 g lactose l)1 concentration, temperature will vary based on the values of one
pH 6.0 and 24.2 C. Maximum growth rate un- another. This relationship is shown in Figure 3
der these conditions was estimated to be and the interaction between the variables mani-
1.4±0.5 mm d)1. fests itself in the form of an elongated ellipse.
Statistical analysis of the variables involved in The rounded ridge, inside the design boundary,
the model show that incubation temperature was runs diagonally on the plot from lower right to
significant for mycelial growth rate at 1% a level. the upper left.
The three-way interaction of the variables was In order to verify the accuracy of the model
not significant at the 5% a level, suggesting that predictions, an additional five trials of the experi-
all three variables were not simultaneously inter- ment were run under the optimal growth condi-
dependent. Of the three possible two-way inter- tions predicted by the model (24.2 C, pH 6.0
actions among the variables, two of the and 44.1 g lactose l)1); afterwards, residual plots
interactions (lactose concentration pH and pH for all experimental values were examined for
temperature) were not significant at the 5% a any weakness in using the model (Draper & John
level. Figure 2, for example, shows the relation- 1988). These trials resulted in an average growth
ship of the non-significant interactions between rate of 1.7 mm d l)1 with a standard deviation of
the pH and temperature. These response surfaces 0.1, which was similar to the model output. The
1540
residual plots for the model and the experimental conditions for P. ostreatus mycelia when grown
data set showed no patterns or trends (Figure 4). using whey permeate as a substrate.
A check of the constant variance assumption
(Draper & John 1988) could also be addressed
because a random plot of residuals meant homo- Conclusions
geneous error variances across the observed val-
ues. Therefore, it can be concluded that the The results of this experiment suggest that whey
equation was able to accurately model the permeate could provide a viable growth substrate
response surface of and predict optimal growth for the cultivation of P. ostreatus mycelia. The
References