Harvesting of Scenedesmus obliquus in

Wastewaters: Auto- or Bioflocculation?

Alain Lavoie and J. de la h u e

Dkpartement de Biolqie et Centre de Recherche en Nutrition, Universitk
Laval, Qkbec, Qukbec, Canada GtK 7P4
Accepted for publication November 26, 1986

Autoflocculation and bioflocculation are considered to be Moellmer6 and more recently by Sukenik and Shelef7 who
the most promising means for the economical harvesting postulated the following mechanism: increasing pH in algal
of microalgae. We have therefore studied these phenom- cultures, either by COPconsumption by algal photosynthesis
ena with cultures of Scenedesmus obliquus produced
during biological tertiary wastewater treatment. The or by direct addition of alkali, leads the culture medium to
quantity of extracellular polymers produced during age- a supersaturation state with respect to calcium and phos-
ing of the cultures proved insufficient to initiate bio- phate ions. Such a supersaturation causes an initial nuclea-
flocculation while the concentration of Ca2+and POa3- of tion of calcium phosphate precipitation which is promoted
the treated effluent were t o o low t o induce auto- by the algal cells serving as a solid surface. In the presence
flocculation. It has been shown, however, that the algae
sediment more readily upon ageing, possibly as a result of excess calcium ions, the calcium phosphate precipitate is
of increased cell density. The use of density gradients positively charged and therefore adsorbed on the negatively
made with Percoll (a colloidal solution of silica particles) charged algal cells agglomerating them and promoting algal
allowed measurement of the true cell density and flocculation.
showed that this increases when cultures enter the de- The second phenomenon, based on algal physiology, is
clining growth phase. The quality of the biomass thus
harvested is, however, considerably impaired, protein called bioflocculation. Two mechanisms were proposed to
content decreasing from 62.7% (dry wt) during the ex- explain it. Schuessler, in 1967,* suggested that biopolymers
ponential growth phase (day 5) to 14% at the end of cul- (principally polysaccharides) produced by the algae could
tures (day 21). be acting as a flocculant. He noted a higher production and
excretion of these products at the end of the exponential
INTRODUCTION growth phase, where maximal algal flocculability is ob-
served. Pavoni and co-workers' also showed the ability of
Although chemical flocculation leads to a good har- extracellular polymers from algal cultures to flocculate inor-
vesting of Scenedesmus obliquus, it would represent an ganic colloid suspensions. The second mechanism proposed
expensive step in the biological tertiary treatment of waste- suggested that bacteria use algal extracellular products for
waters. Moreover, the physiological state of microalgae growth and induce algal flocculation as they do in the acti-
plays an important part in the flocculation process. The vated sludges of secondary wastewater treatment. 'c-'~
maximal algal recovery by chemical flocculation is obtained Since natural flocculation can be considered as the most
at about the end of the exponential growth phase and de- promising means of economically harvesting microalgae,
creases rapidly as soon as the cultures enter the declining we wanted to determine whether this phenomenon occurred
growth phase. The accumulation of extracellular polymers in Scenedesmus obliquus cultures and could be used to
at this growth stage could be acting as a protective colloid harvest the algal biomasses produced during biological ter-
masking the algal surface charges and leading therefore to a tiary wastewater treatment and intended for animal feeding.
reduction of flocculant efficiency. Moreover, a better knowledge of the physiological reaction
There are only few data published on the natural floccu- of Scenedesmus obliquus is needed to enable maximal floc-
lation of microalgae. Bogan and c o - ~ o r k e r s working
,~ on culation efficiency by natural or chemical processes as well.
the removal of phosphorus by the filamenteous green alga
Stigeoclonium stagnatik, noted floc formation when grow-
MATERiALS AND METHODS
ing this species with air bubbling. A similar phenomenon
was observed by Golueke and Oswald,s for microalgal cul-
Algal Cultures
tures photosynthetically active in a shallow pond during
warm and sunny days. The strain of Scenedesmus obliquus used was isolated
These flocculation phenomena are quite different. The from the secondary effluent of the Valcartier wastewater
former, called autoflocculation, was first studied by treatment plant (Quebec, Canada). This freshwater Chloro-

Biotechnology and Bioengineering, Vol. 30, Pp. 852-859 (1987)

0 1987 John Wiley & Sons, Inc. CCC 0006-3592/87/070852-08$04.00
phycea, commonly found in wastewaters, is well adapted to agitated, three in the darkness and three under illumination
this environment and has been proven to be a suitable or- (210 pE/m2/s). After I , 2, and 4 h, the pH and the fraction
ganism to perform a tertiary biological treatment of waste- of algae still in suspension were measured in the upper-
waters. l 3 most 250 mL supernatant of two samples (1 dark and
Algae were grown in a lOOO-L, stainless-steel tank con- 1 light) in the same way as for flocculation tests. All these
taining 400 L secondary effluent from a wastewater treat- tests were done in the middle part of the light period of the
ment plant located at Valcartier near Quebec. This type of day-night cycle.
medium has the following general characteristics: N-HN,' ,
9.5 mg/L; N-N02- + N-N03-, 2.5 mg/L; P-PO.,-,
Cell Density
2.2 mg/L; total suspended solids, 10 mg dry wt/L; chemi-
cal oxygen demand, 30 mg 02/L; pH, 7.0. The 21-day For the study of the influence of microalgal cell density
batch cultures were carried out at room temperature (20°C) on sedimentation, cell density was measured with stepwise
under 350 pE/m2/s illumination with a light: dark cycle of density gradients of Percoll. I6 Mixing Percoll with the cul-
14: 10. Air-bubbling (4 h-I) provided sufficient CO, for mi- ture medium leads to low osmolality ( 5 2 0 mOS/kg H20)
croalgal photosynthesis. Intermittent mixing (30 min every solutions. This prevents algal cell plasmolysis and allows
2 h) with a propeller (?4hp motor) prevented any sedimen- the measurement of true cell density. The gradient prepara-
tation of algal cells. Evaporative losses were replenished tion techniques have been described elsewhere.l6
daily with distilled water. Ammonium (N-NH4+) and ortho-
phosphate (P-PO:-) ions were measured daily using a
Extracellular Products
Technicon Autoanalyzer (Industrial Methods 98-79 w/a
and 94/70 w/b, respectively) according to the APHA.14 The quantity of extracellular products in algal cultures
Algal growth was monitored by transmittance at 678 nmI5 was measured according to Lewin. l7 Following two succes-
and dry weightI3 measurements. sive filtrations of a 100-mL culture sample first over a
Five different stocks of secondary effluent were used and Whatman 934 A-H then through a Millipore 0.45-pm filter,
were identified by their sampling time: 84-06-21,84-07-19, the filtrate was concentrated to 10 mL by boiling. After
84-10-12, 84-12-01, and 85-01-08. cooling, 30 mL of ethanol (95%) were added to the sample
which was then stored at 4°C overnight. The white preci-
Flocculation and Sedimentation pitate formed was centrifuged, washed with ethanol,
vacuum-dried at room temperature and weighed. Results are
Effect of pH expressed in milligrams of extracellular polymers per liter
of culture.
The effect of pH on algal flocculation was studied in a
standard jar test apparatus. The pH of six 1-L suspension
Extracellular and Capsular Carbohydrates
samples was adjusted with 1.ON NaOH or H2SO4 to values
ranging from 6.0 to 12.0; samples were then stirred at A 15-mL culture sample was centrifuged for 15 min at
100 rpm for 1 min, at 50 rpm for 14 min, and left still for 2000g (IEC Clinical Centrifuge). The supernatant was im-
15 min. Samples were immediately withdrawn at the 0.75-L mediately used for extracellular carbohydrate measurement
level for the determination of the remaining algae after floc- by the anthrone method according to Jermyn.I8 The same
culation and sedimentation. Flocculation efficiency was es- technique was used for another culture sample previously
tablished as: heated at 100°C for 7 min in order to measure total carbo-
hydrate production.
T - To
E = x 100 The capsular fraction was the difference between extra-
100 - To cellular and total carbohydrates. Values are expressed in mg
where E is the percentage of algae removed from culture; T of glucose/L.
is the algal Concentration transmittance at 678 nm after the
flocculation test; and To is the algal concentration after the Biomass Composition
flocculation test in the control jar.
Nitrogen, carbon and hydrogen contents of algae were
Effect of Calcium and Phosphate measured by combustion of lyophilised algal samples (0.2
to 1.0 mg dry wt) in a CHN analyzer (Hewlett-Packard).
The effect of calcium and phosphate on flocculation was
Algal protein content was calculated by multiplying the
studied by adding KH2P04or CaC1: to obtain final concen-
total nitrogen values by 5.6, a factor obtained from amino
trations of 0 . 2 d and 2.5mh4, respectively, in the samples,
acid analysis (unpublished results).
prior to the flocculation tests.

Effect of Light and Algal Cell Density Bacterial Counts

In order to determine the effect of light on the sedimen- The changes in bacterial counts were followed by the
tation of algae, six 1-L suspension samples were left un- plate count method. Starting with a dilution series of algal

LAVOIE AND DE LA NOUE: HARVESTING OF SCENEDESMUS OBLIQUUS 853

cultures (lo-'- 0.1-mL samples were spread on nutri-
ent agar in triplicate. The agar plates were incubated for
48 h at 20°C19and colony forming units were counted.

RESULTS

Algal Culture Growth

Figure 1 shows the typical growth pattern of Scene- pH 10

desmus obliquus batch cultures. Although ammonium was

exhausted by day 6, phosphate (P-P04-3) uptake continued
up to day 9. Phosphate exhaustion coincided with the de-
clining growth phase and the stationary phase was attained
one or two days later. It can be seen, however, that the algal
biomass concentration increased continuously up to day 22.
o]
.-.
j
-.-.-.-.-.-.-._.-..o-.-.
* . , , , ,pH 7.5-8.7
Influence of pH and Culture Age on Microalgae
Flocculation 50

~-._~,._._i_._.C.-C.~.-.-.-.-.-.
The influence of the age of the culture on the floccu- loo/
0 , ,

lability of microalgae for pH values ranging from 7 to 12 is

shown in Figure 2. Without any control, the pH of the 50
cultures reached stable values comprised between 7.5 and
8.7. Effective flocculation was observed at pH 11 and 12 0
4 8 12 16 20
during the early days of culture. However, after day 6, there
Time (days)
was no flocculation at pH values between 7 and 11 and
efficiency was below 50% at a pH value of 12. Figure 2. Influence of pH and culture age on Scenedesmus obliquus
natural flocculation: (-----) onset of stationary phase (water stock:
84-06-21).

1000 1 N H ~ Phosphate ion concentration appears to be an important

0 PO, factor in the flocculation process. Figure 3 shows that
harvesting efficiency for pH value of 12 decreased as micro-
algae exhausted PO:-. Moreover, a better flocculation effi-
ciency was observed at the stationary phase for cultures
showing higher residual ortho-phosphate concentrations
[Fig. 3(A)].

1
120

80 7

40
5
%
0 , .
2
-120 3
0
Plll

-80
d
3

ln
300 - 40
ul 200
m
g 100 -0
0 4 8 12 16 20
0 2 4 6 8 10 12 14 16 18 2 0 22
Time (days)
Time (days)
Figure 3. Importance of P 0 4 3 - residual concentrations on auto-
Figure 1. Growth of Scenedesmus obliquus on wastewater (water stock: flocculation of Scenedesmus obliquus cultures at a pH value of 12, for
84-07- 19). water stocks (A) 84-06-21 and (B) 84-07-19.

854 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 30, NOVEMBER 1987

 A B
Influence of Calcium and Orthophosphate Transmittance (96)
Concentration on Scenedesmus obliquus
Autoflocculation
Figures 4(A) and 4(B) show how 2.5mM of Ca2+ or 0: , - , , , . , , , ~
J, , . , , , , , ,
0.2mM of PO:- influenced the autoflocculation efficiency Capsutar sugars (mg glucose/L)

for pH values up to 11. One can see that the addition

of calcium or phosphate significantly raised the auto-
flocculation efficiency for pH values of 10 and 11;the simul-
taneous addition of these two products allowed a good Extracellular sugar (mg glucose/L)

autoflocculation (80%)at pH values of 8 and higher. More-

over, the age of culture markedly influenced the auto-
flocculation phenomenon since the efficiency quickly
Extracellular proteins (mg/L)
decreased as soon as the culture entered the declining
growth phase (day 9), despite the addition of calcium and
phosphate.
Extracel lular polymers ( mg /L )
100-

Bioflocculation
In order to determine whether bioflocculation enables the 0; ( , , , , . , , , J, , , , , , , , , , ,
0 4 8 12 16 20 0 4 8 12- 16 20
efficient harvesting of Scenedesmus obliquus, we monitored Time (days)
the appearance of products capable of inducing bio-
Figure 5. Appearance times of products responsible for bioflocculation
flocculation during algal culture. Figure 5 shows that the induction in Scenedesrnus obliquus cultures: ( .1 ) onset of stationary phase
quantities of extracellular polymers varied during the course for water stocks (A) 84-10-12 and (B) 84-12-01.
of incubation and did not appear to be correlated with the
age of the cultures. The polymer concentrations measured in
different cultures never exceeded 120 mg/L, even if the carbohydrate and nonprotein material. The amount of ex-
values obtained include considerable quantities of non- tracellular proteins, variable for different cultures, was
comprised between 0 and 7 mg/L in all the experiments.
Capsular and extracellular sugars appeared at the onset of
A B
the declining growth phase.
While the amount of extracellular sugar showed few varia-
tions during the cultures, that of capsular material increased
0 4 , . . . , 4 , . , , , , ,~
continually until the end of culture. The maximal concen-
100 trations obtained during this study were 5.2 mg/L for extra-
cellular and 25.4 mg/L for capsular sugars.

Microalgal Sedimentation
The efficiency of sedimentation of Scenedesmus obliquus
was closely related to the physiological state of the algae as
shown in Figure 6. Although maximal efficiency varied be-
tween cultures, observed tendencies were consistent, sedi-
mentation increasing with declining growth phase. One can
also see that light influenced this phenomenon. When cul-
tures were kept under darkness, more algae sedimented and
more rapidly. Although illumination caused the pH to rise to
100
10, no floc formation was observed under these conditions.
?=-\ h
pM PO:

’J ‘0, 1
.r-.-.-...--.-.+, .‘ I-.-.-.-.‘
I Cell Density
Higher microalgal sedimentation with culture ageing is
4 8 12 16 20 4 6 8 10 12
not related to flocculation phenomena (no floc formation);
Time (days)
we therefore followed the progression of cellular density
Figure 4. Importance of calcium and phosphate for autoflocculation of during two batch cultures in order to find out whether this
Scenedesrnus obliquus and the influence of culture age on this phenome-
non: ( 1) onset of stationary phase for water stocks (A) 84-10-12 and (B)
parameter could explain microalgal behavior. Figure 7
84-12-01; (0)control; addition of (0)Ca2+; (M) PO.,-,(0) Caz+ and shows the distribution of algae in different density classes
Po:-. and their relative abundance in each of these classes for the

LAVOIE AND DE LA NOUE: HARVESTING OF SCENEDESMUS OBLIQUUS 855

 --
30 30 25 25 5 5 5 5 5 5
60 60 60 60 15 15 10 15 10 10 15
10 10 15 1 5 80 4 0 25 25 25 25 20 25 5 5 5 5 5
4 0 25 15 20 25 35 35 30 3 0 3 0 30 2 0 20
35 15 15 20 20 30 30 3 0 35 35 40 40
25 25 15 5 5 15 15 10 10 15 10
5 112
0
5 5 5 5
15 15
5 5
15 15
5 1 0
15 20

T
t
15 10
17 7 0 60 10 20 20
66 15 30 34 35 40 25 20 15 5 5 5 5 5
17 66 20 10 20 40 40 30 30 25 25 25 20 20 10

0 I Y 35 20 10 2 0 25 30
10 20 10 S 15
5 5 5 10
3 0 3 0 30
15 20 15
10 10 10
30 40 50 4 0
15 10 5
10 10 5
10
10
8t 1'2 1'6 20 4 8t 1'2 1'6 20 5 10 10 10 10 15 15 20 2 0 30
. . .
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
Time (days)
Time (days)
Figore 6. Effect of culture ageing on algal sedimentation in (A) darkness
and (B) under illumination. Sedimentation duration was (0)1 h, (0) 2 h, Figure 7. Relationship between the density of Scenedesmus obliquus
and (m) 4 h. Water stocks were 84-12-01 (bottom) and 85-01-08 (top). The cells (their distribution (%) in different density classes) and sedimentation
arrow ( T ) denotes the onset of declining growth phase. efficiency in darkness (expressed as algal recovery (%)). Sedimentation
duration: 4 hours. Water stocks were (A) 84-12-01 and (B) 85-01-08.The
arrow ( $ ) denotes the end of the exponential growth phase.

entire duration of 21-day cultures. One can see that Biomass Composition
cell densities showed lowest values when algae were experi-
encing exponential growth phase and that the algal popu- Table I1 summarizes the variations in chemical com-
lation was distributed into two or three density classes position of two experimental populations of Scenedesmus
obliquus. One can see that the protein content of microalgae
(1.090 g/mL or less). As the cultures aged, classes of higher
density appeared and algal populations became more and dropped quickly at the end of exponential growth phase. The
carbon content of biomass decreased lightly in the early
more diversified (from four to eight density classes).
stationary phase, but gradually rose afterwards. It can also
Sedimentation under illumination being always less effi-
be noted that the hydrogen content increased during station-
cient than under dark conditions, despite identical initial cell
density in both cases, density measurements were made with
ary phase.
the cells remaining in suspension under both conditions.
Bacterial Counts
Table I shows that decantation in the dark allowed a marked
sedimentation of algae with density higher 1.090 g/mL The typical changes in bacterial counts during S . obliquus
while illumination prevented such a selective sedimentation. batch culture on secondary effluent are illustrated in

Table I. Influence of light on the sedimentation of Scenedesmus obliquus: distribution of algae (%)
in different density classes.

Density (g/mL)

~~

Culture A 17.5% 35% 17.5% 10% 10% 10%

Supernatant L" 12.5% 37.5% 37.5% 12.5% - -
Supernatant Db 100% - - - - -
Culture B 25% 33% 18% 8% 8% 8%
Supernatant L 11% 33% 33% - 11% 11%
Supematant D 100% - - - - -
Culture C 25% 33% 25% 8.5% - 8.5%
Supernatant L 11% 33% 33% - 11% 11%
Supernatant D 50% 50% - - - -

"This is the algal distribution in the supernatant after a 4-h sedimentation under illumination.
'This is the algal distribution in the supernatant after a 4-h sedimentation in darkness.

856 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 30, NOVEMBER 1987

 Table 11. Influence of culture age on chemical composition of Scenedesmus obliquus

Age Percent C Percent H Percent N Percent protein

(days) @') (%I (%)

7 53.9 9.7 9.2 51.2

11 46.6 5.8 4.2 23.6
15 46.7 7.4 3.4 19.0
19 49.5 22.5 3.0 16.8
End of exponential growth D-7; beginning of stdonary phase D-10.
5 52.5 7.5 11.2 62.7
8 51.7 8.8 9.9 55.4
14 46.3 14.8 4.3 24.1
22 53.0 24.0 2.5 14.0
End of exponential growth D-8; beginning of stationary phase D-l 1.

Figure 8. The number of bacteria clearly decreased as the efficiency obtained at this pH was below 45%. The low
algal population was growing exponentially and reached lo3 calcium content of our culture medium (6 mg/L) explains
colony forming units/mL, a low value subsequently main- such a result. In fact, a minimal calcium concentration of
tained. 40 mg/L is reported as necessary for maximal auto-
flocculation.' This implies that calcium addition would be
DISCUSSION required to induce autoflocculationin most of the secondary
effluents as the calcium content of freshwaters is
Autoflocculation ca. 25 mg/L.21 In warm climates, evaporation can be high
enough to increase the calcium concentration, hence the
The autoflocculation phenomenon has been usually asso- induction of autoflocculation.5-7
ciated with intense photosynthetic activity. The resulting The simultaneous addition of 2 . 5 d calcium and 0.2mM
alkaline conditions induce a coprecipitation of magne- orthosphosphate allowed autoflocculation induction from
sium, calcium, phosphate, and carbonate with algal cells." pH 8.0 with an improved and constant efficiency for pH 9,
Sukenik and Shelef7 identified Ca2+and PO-: as the main 10, and 11 [Figs. 4(A) and 4(B)]. These results agree with
causes of this phenomenon, the respective concentration of those of Sukenik and Shelef; this was expected since we
these ions determining the pH value required to induce auto- used the ion concentrations proposed by these authors to
flocculation. allow maximal autoflocculation at pH values higher than
The poor recovery of algae obtained during our experi- 8.5. The separate addition of CaZ+or POP- allows identi-
ments with pH values lower than 12 (Fig. 2) suggests that fication of the limiting ion. An example of calcium limi-
one of these ions, if not both, was limiting. Moreover, it can tation is shown in Figure 4(A). One can see that the addition
be seen from Figure 3 that flocculation at pH 12 was closely of Ca" alone, while culture phosphate concentration was
related to the residual phosphate concentration in the culture still considerable (D-5), caused a significant improvement
medium. However, the PO:- concentrations measured on of flocculation efficiency at a pH value of 9. Nevertheless,
days 5 and 6 could allow autoflocculation induction at pH from day 8, while orthophosphate concentration was re-
values higher than 9.3 provided that sufficient Ca2+concen- duced, the simple addition of Ca2+was insufficientto induce
trations were present.' Nevertheless, no flocculation was algal flocculation under the same pH condition. An inverse
observed for pH values lower than 11 while the maximal relationship can be seen for a culture that presented low
phosphate concentrations. It can be seen in Figure 4(B) that
the simultaneous addition of calcium and phosphate was
160 rewired in order to obtain a significant flocculation at pH 9,
10, both. ions being limiting. Moreover, adding Ca2+ alone
was insufficient, even at pH 10, while adding orthophos-
: I \
.- phate was sufficient to maximize flocculation under this pH
condition.
Even in the presence of Ca2+and PO:- concentrations
capable of maximal autoflocculation induction (pH > 8.0),
the percentage of algae recovered by sedimentationdropped
abruptly as soon as the cultures entered the declining growth
5 10 15 20 phase. This behavior is similar to that observed during stud-
Time (days)
ies on chemical flocculation of microalgae'~~ and shows the
Figure 8. Changes in bacterial population during Scenedesmus ob[iquus importance of algal cell physiology for the flocculation
culture. Water stock was 84-10-12. process.

LAVOIE AND DE LA NOUE: HARVESTING OF SCENEDESMUS OBLIQUUS 857

Bioflocculation Sedimentation and Cell Density
The results obtained during the study of the influence of Although the results obtained point out that Scenedesmus
pH on microalgal flocculation (Fig. 2) as well as the con- obliquus did not experience the required conditions for bio-
centrations of extracellular sugars and polymers found in the flocculation under our culture conditions, the sedimentation
cultures (Fig. 5 ) show that Scenedesmus obliquus har- of algal cells appeared to be closely related to their physio-
vesting could not be effected by bioflocculation, under our logical state (Fig. 6). In fact, the percentage of algal re-
culture conditions (batch cultures using secondary effluent covery was lowest during the exponential growth phase,
with excess C02provided through air bubbling). Since the increased rapidly during the declining growth phase and
pH of our cultures did not exceed values of 7.5-8.7, we can remained about constant at the stationary phase.
infer from the work of de la Noue et a1.22that the availability The variation of sedimentation in the dark (duration of
of CO, was sufficient under the present conditions. In fact, 4 h) followed a similar pattern to that of cell density class
according to Pavoni and co-workers,' bioflocculation occurs modifications (Fig. 7). The increase in recovery efficiency
when pH is higher than 10 and its efficiency improves with was coincident with the appearance of algal cells with a
culture ageing when the lysis of old cells contributes by density higher than 1.090 g/mL, when they entered the
increasing the amount of extracellular polymers responsible declining growth phase. The highest recovery efficiency
for floc formation. attained was observed when there was no longer any density
Our results show that no significant algal flocculation class lower than 1.090 g/mL. This factor, however, should
could be induced after the eleventh day of culture whatever not play a major role in harvesting systems for larger scale
the medium pH (Fig. 2). Moreover, the higher extracellular operation due to the importance of other factors and to
polymer concentration measured after 2 1 days of culture differences in culture modes.
never exceeded 120 mg/L for all our experiments, a value The nonselective sedimentation observed under illu-
markedly inferior to that of 500 mg/L reported by Pavoni mination (Table I) was responsible for the lower algal recov-
and co-workers.' ery (Fig. 6 ) and can be explained by two factors. The first
Although massive production of extracellular poly- one, entirely physical, is the induction of thermal con-
saccharides by any microalga is possible, blue-green and vection currents. The second one, purely physiological, is
filamenteous species are the most suitable organisms.23 oxygen production via algal photosynthesis. Oxygen micro-
Comparing our results with those of Pavoni and co-workers' bubbles appears on the algal cell surfaces and makes the
is not an easy task, since they did not identify or describe the algae float.
algae they used. These authors concluded that extracellular
polymers are responsible for bioflocculation by comparing
Algal Chemical Composition
their results to those they obtained previously for bacterial
population^,^^ without considering algal autoflocculationor The rapid decrease of algal protein content when cultures
sedimentation. entered the declining growth phase is common for Chloro-
Capsular and extracellular sugars appeared at the onset of phy~eae.,~ The consumption of intracellular stocks of nitro-
the declining growth phase for Scenedesmus obliquus. gen and protein excretion are two mechanisms that can
These results agree with those obtained by Samson25who explain this phenomenon.
studied the physiology of Oocystis sp., another fresh- The hydrogen content increase that accompanied the car-
water Chlorophycea. This author reported that Oocystis bon content restoration is indicative of the accumulation of
grown on synthetic medium excretes proteins only a few photosynthetic products ,26 mainly saccharides, that could
days after the beginning of the stationary phase. The pres- have been responsible for the increase in cell density.''
ence of extracellular proteins in our cultures from the very Work is under progress to clarify this point.
beginning, is explained by the use of secondary effluent as
culture medium.
CONCLUSIONS
The appearance of capsular sugars coincided with the loss
of flocculation efficiency reported in Figures 4(A) and 4(B) The amounts of extracellular polymers produced during
as previously reported during chemical flocculation stud- the ageing of cultures as well as the low Ca2+ and PO:-
ies.' The protection of algal surface charges by bio- concentrations in the secondary effluents used appear to
polymers, as proposed by Tenney et a1.,3 could therefore be insufficient to allow bio- and autoflocculation respec-
have been the principal cause for the slight flocculability of tively of Scenedesmus obliquus under our laboratory con-
Scenedesmus obliquus on attaining the stationary phase. ditions. The exhaustion of phosphate is in any event one
The low bacterial population observed in Scenedesmus of the main purposes of biological tertiary treatment, and
obliquus batch cultures, grown on secondary effluent in the use of calcium as a flocculant gives harvested biomass
the laboratory, eliminated the possibility of alga-bacteria of insufficient quality for feeding purposes. Consequently,
co-flocculation as reported by Eisenberg et al." with autoflocculation does not appear a suitable process for har-
Ankistrodesmus sp. grown on raw sewage. These authors, vesting Scenedesmus obliquus if biomasses are intended for
studying bioflocculation of microalgae in oxidation ponds, feeding purposes.
noted that flocs were composed of algal cells incorporated Although the algal sedimentation without flocculation im-
into the matrices of filamentous bacteria. proves with the ageing of cultures, this harvesting process

858 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 30, NOVEMBER 1987

remains unstable and very slow. Moreover, the protein con- 7. A. Sukenik and G. Shelef, Biotechnol. Bioeng., 26, 142 (1984).
tent of algae decreases quickly as soon as the cultures enter 8. R. G. Schuessler, “An investigation of the chemical flocculation and
autoflocculation of algae,” M.S. thesis, University of Notre Dame,
the declining growth phase. If the harvested biomass must South Bend, IN, 1967.
be of a fair quality for animal feeding purposes, chemical 9. J. L. Pavoni, S. W. Keiber, and G. T. Boblitt, “The harvesting of algae
flocculation, especially with chitosan as reported else- as a food source from wastewater using natural and induced floccu-
where,2 presently appears to be the best way to harvest lation techniques,” Conference on the use of wastewater in the produc-
Scenedesmus obliquus cultures at least until new polymeric tion of food and fiber, Oklahoma City, OK, March 6-8, 1974.
10. J. Bauld and T. D. Brock, J . Phycol., 10, 101 (1974).
flocculants are tested and proved to be safe for such a pur- 11. D. M. Eisenberg, B. Koopman, J. R. Benemann, and W. J. Oswald,
pose. However, the results obtained indicate the possibility Biotechnol. Bioeng. Symp., 11, 429 (1981).
of a significant reduction of the flocculating agent concen- 12. K. Wolter, Mar. Ecol. Prog. Ser., 7 , 287 (1982).
tration by acting before the microalgae enter the declining 13. J. de le Noiie, G. A. Picard, J. C. Piette, and C. Kirouac, Water Res.,
growth phase, i.e., before the appearance of capsular sugars 14, 1125 (1980).
14. APHA, AWWA, and WPCF, Standards methods for examination of
and at a time when the protein content of the biomass is water and wastewater 15th ed. (WPCF, Washington, DC, 1980).
maximal. 15. C. Sorokin, Handbook of phycological methods, J. R. Stein, Ed.
(Cambridge University Press, Cambridge, 1973), pp. 321-343.
This work was supported by grants from the Natural Sciences and
16. A. Lavoie, J.-L. Mouget, and J. de la Noiie, J . Microbiol. Methods,
Engineering Research Council of Canada (NSERC) and the Fonds
4, 251 (1986).
F.C.A.R. from the Ministry of Education of QuCbec (team grant).
17. R. A. Lewin, Can. J. Microbiol., 2, 665 (1956).
We wish to thank C. Dallaire for the drawings, D. Ni Eidhin for
18. M. A . Jermyn, Anal. Biochem., 68, 332 (1975).
help in the writing of the final version, and G. Gagnon for typing
19. M. J. Pelczar and E. C. S . Chan, Laboratory exercises in micro-
the manuscript.
biology, 4th ed. (McGraw-Hill, New York, 1977).
20. J. Benemann, B. Koopman, J. Wiseman, D. Eseinberg, and R. Goe-
bel, Algae Biomass, G. Shelef and C. J. Soeder, Eds. (Elsevier, Am-
References sterdam, 1980), pp. 457-495.
21. G. K. Reid and R. D. Wood, Ecology of inland waters and estuaries.
1. A. Lavoie, J. de la Noiie, and J. B. SCrodes, Can. J. Civ. Eng., 11, 2nd ed., (Van Nostrand, New York, 1976).
266 (1984). 22. J. de la Noiie, L. Cloutier-Mantha, P. Walsh, and G. Picard, Biomass,
2. A. Lavoie and J. de la Noiie, J. WorldMaricul. SOC., 14,685 (1983). 4, 43 (1984).
3. M. W. Tenney, W. F. Echelberger, R. G. Schuessler, and J. L. Pavoni, 23. B. G. Moore and R. G. Tisher, Science, 145, 586 (1964).
Appl. Microbiol., 18, 965 (1969). 24. J. L. Pavoni, M. W. Tenney, and W. F. Echelberger, J . Water Pollut.
4. R. N. Bogan, 0.E. Albertson, and J. C. Pluntze, J . Sanit. Eng. Div., Control Fed., 44,414 (1972).
1, 20 (1960). 25. R. Samson, “CharactCristiques physiologiques de la croissance et da la
5 . C. G. Golueke and W. J. Oswald, J. Water Pollut. Control Fed., 37, production de polysacchaides chez neuf esgces d’algues endogknes
471 (1965). isolCes des effluents du traitement secondaire de l’usine d’tpuration de
6. W. 0. Moellmer, “Factors controlling autoflocculation for algae and Valcartier,” thtse de M.S. UniversitC Laval, QuCbec, Canada, 1980.
algal nutrient removal,” Ph.D. dissertation, University of California, 26. F. P. Healy, Crit. Rev. Microbiol., 3, 69 (1973).
Berkeley, CA, 1970. 27. R. L. Oliver and A. E. Walsby, Limnol. Oceanogr.. 29, 879 (1984).

LAVOIE AND DE LA NOUE: HARVESTING OF SCENEDESMUS OBLIQUUS 859