conditions and growth phase, but are generally of the order of 10−^6 g O 2 (g DW)−^1 s−^1.
Maximum values typically occur early in the exponential growth phase (e.g. Dubuis et
al., 1995). In operational terms, it is the oxygen uptake rate (OUR), the product of qo and
biomass concentration, X, which is most relevant. Maximum OUR typically coincides
with maximum (viable) biomass level, at the onset of the stationary growth phase.
Reported data (e.g. Bond et al., 1988; Ho et al., 1995) typically refer to small scale
systems, with relatively low biomass concentrations (≤12 g DW L−^1 ). However, assuming
an upper qo limit of about 0.6 mmol O 2 (g DW)−^1 h−^1 , an OUR of up to 12 mmol L−^1 h−^1
can be expected at a biomass level of 20 g DW L−^1.
In strongly aerobic cultures, the rate of metabolic heat evolution (Qmet) is directly
proportional to the oxygen consumption rate (Cooney et al., 1969), viz.
(1)where V is the broth volume. Accordingly, lower oxygen consumption rates in plant
systems result in correspondingly lower metabolic heat loads than in microbial systems.
Metabolic heat generation rates of approximately 138 J (g DW)−^1 h−^1 were reported by
Hashimoto and Azechi (1988) in a 20 m^3 N. tabacum chemostat culture (6.34 m^3 wv). At
an average biomass level of 17 g L−^1 , this corresponds to a modest metabolic heat load of
652 W m−^3. No data are available for overall heat transfer coefficients in aerated, agitated
plant cell systems. However, Koloini (1990) reports values of between 100 and 200 W
m−^2 K−^1 for viscous (xanthan) fermentation broths. Assuming similar values for a well-
designed, jacketed plant bioreactor, metabolic heat removal should not constitute an
operational difficulty, even allowing for less intense agitation conditions and the reduced
driving force associated with a typical cultivation temperature of 25°C.
Critical dissolved oxygen (DO) concentrations (ccrit) are generally assumed to be in the
range 15–20% of air saturation. Values of 16% (Tate and Payne, 1991) and 25% (Dubuis
et al., 1995) have been reported for C. roseus and Coffea arabica, respectively. However,
the ccrit value for metabolite synthesis may be substantially higher than that for biomass
production. Recent studies of C. roseus (Schlatmann et al., 1994b, 1995) indicate a
critical DO value of 43% for ajmalicine production, although suspensions cultivated at
high (85%) and low (18%) DO levels exhibited no significant differences in primary
metabolism. Moreover, under oxygen limited conditions, biomass yields (g DW per g C-
source) have been shown to be approximately constant (Tate and Payne, 1991) or to fall
only slightly (Pareilleux and Vinas, 1983).
Expressed in terms of saturation-type (Monod) kinetics, the apparent oxygen
saturation constant, Km,app (Payne et al., 1991) has been reported as 0.009 mmol L−^1 for
C. roseus (Pareilleux and Vinas, 1983) and 0.078 mmol L−^1 for C. arabica (Dubuis et al.,
1995) suspensions with a mean aggregate diameter of 2 mm. Values are line-dependent
and are also affected by system morphology and mass transfer characteristics. An oxygen
maintenance coefficient of 6.26 mmol O 2 (C-mol biomass)−^1 h−^1 was determined for N.
tabacum suspensions (van Gulik et al., 1992). Assuming a molecular weight of
approximately 30.7 (C-mol basis) for batch-cultured N. tabacum cells, this corresponds to
a value of 0.204 mmol O 2 (g DW)−^1 h−^1. However, on the basis of viable biomass alone,
Ho et al. (1995) reported values of between 0.115 and 0.358 mmol O 2 (g DW)−^1 h−^1 for N.
tabacum, accounting for 52% and 87% of respiration, respectively. The higher value was
Bioreactor design for plant cell suspension cultures 427