Plant, Cell and Environment (1980) 3, 261-269.

Characterization and comparisons of five Na-fixing

Azolla-Anabaena associations, I. Optimization of growth
conditions for biomass increase and N content
in a controlled environment*
GERALD A. PETERS, ROBERT E. TOIA, JR, WILLIAM R. EVANS, DEBORAH K. CRIST, BERGER C.
MAYNE & REBECCA E. POOLE Charles F. Kettering Research Laboratory, 150 East South College Street,
Yellow Springs, Ohio 45387, U.S.A.

Received 28 January 1980; accepted for publication 20 February 1980

Abstract. Biomass increase, C and N content, C2H2 m~^ s~', the Azolla species all doubled their biomass in
reduction, percentage dry weight and chlorophyll a/b 2 days or less and contained 5-6% N on a dry weight
ratios were determined for clones of Azolla caroliniana basis.
Willd., A.filiculoides Lam., A. mexicana Presl., and A.
pinnata R.Br. as a function of nutrient solution, pH,
temperature, photoperiod, and light intensity in con- Introduction
trolled environment studies. These studies were sup- Azolla, a genus of floating aquatic ferns with six extant
plemented by a glasshouse study. Under a 16 h, 26°C species, is distributed throughout the world's tropical
day at a light intensity of 200 /zmol m"^ s~' and an 8 h, and temperate fresh water ecosystems (Svenson, 1944;
19°C dark period, there was no significant difference in Moore, 1969). All species normally contain a N2-fixing
the growth rates of the individual species on the five cyanobacterium of the Nostoc-Anabaena group as a
nutrient solutions employed. Growth was comparable symbiont. Both partners are photosynthetically com-
from pH 5 to pH 8, but decreased at pH 9. Using the petent (Peters, 1975; Ray et al., 1979), and the cyano-
same photoperiod and light intensity but constant phyte can provide the associations with their total N
growth temperatures of 15-40°C, at 5°C intervals, the requirement (Peters & Mayne, 1974b).
individual species exhibited maximum growth, nitro- Azolla offers a significant potential as a N source in
genase (N2ase) activity and N content at either 25° or rice production. This is documented by their tradi-
30°C. There was no difference in the temperature tional use for this purpose in the Far East (Moore,
optima at pH 6 and pH 8. The tolerance of the indivi- 1969; Dao & Tran, 1979; Liu, 1979) as well as by
dual species to elevated temperature was indicated to current field studies in the U.S. (Talley, Talley & Rains,
be A. mexicana> A. pinnata> A. caroliniana> A. fili- 1977; Rains & Talley, 1979; Talley & Rains, 1980), in
culoides. At the optimum temperature, growth rates India (Singh, 1977; 1979) and at the International Rice
increased with increasing photoperiod at both pH 6 Research Institute (IRRI) in the Philippines
and pH 8 but N^ase activity was usually highest at a (Watanabe et al., 1977). Significantly, since one or
16 h light period. At photon flux densities of 100, 200, more species of Azolla occur naturally in almost every
400 and 600 /imol m " ' s"', during a 16 h light period geographical area where lowland rice is grown, tradi-
and optimum growth temperature of the individual tional practices and most field studies have been res-
species, N2ase activity was saturated at less than 200 tricted to the endemic species. Except for the recent
//mol and ggrowth at 400 /imoll ^ ~'' No
N field studies with two of the four New World species in
/
California (Talley et al., 1977; Rains & Talley, 1979;
interacting effects of light and pH were noted for any
Talley & Rains, 1980) the use and evaluation of the
species, nor were light intensities up to 1700 /xmol m"^
potential of Azolla as an alternative N source for rice
s~' detrimental to the growth rate or N content of any
has been based almost entirely on A.pinnata R.Br., one
species in a 5 week glasshouse study with a natural 14.5
of the two Old World species.
h light period and a constant temperature of 27.5°C.
Using the optimum growth temperature, a 16 h light In order to evaluate the potential of Azolla, it is
period, and a photon flux density of at least 400 l essential to have background information on the en-
vironmental tolerances of individual species, or
* Contribution No. 703. strains, and eventually the response of physiological
Correspondence: Dr G. A. Peters, C. F. Kettering Research processes, including N2 fixation, photosynthesis, and
laboratory, 150 E. South College Street, Yellow Springs, Ohio
45387, U.S.A.
respiration, in the various species, to environmental
conditions. These approaches should provide guide-
0140-7791/80/0800-0261$02.00
©1980 Blackwell Scientific Publications lines for the selection of Azolla populations for field

261
262 G. A. PETERS et al.

testing in a given geographical area, as well as charac- desired temperature around the flasks. Light was pro-
teristics which might eventually be combined through vided by cool white fluorescent lamps of normal, high
breeding programmes. Towards this end, we have (HO), and very high (SHO) output supplemented with
selected individual populations of four Azolla species incandescent bulbs. Light intensities were controlled
for detailed studies, acknowledging that differences in by number, as well as type, of fluorescent and incan-
the growth tolerances of widely separated populations descent lights used. Intensities were measured as pho-
of an individual species may equal or exceed differ- tosynthetically active radiation (PAR) using a Lamda
ences between species. The results reported here define LI-190S quantum sensor.
conditions of nutrient composition, pH, temperature,
light intensity and photoperiod yielding maximal bio-
mass increase in five Azolla populations. Fresh weights and doubling times
Prior to initial weight measurements the plant material
was gently blotted several times between pieces of
Materials and methods absorbent paper to remove adhering liquid. For final
weight determinations the plant material was collected
Organisms on Whatman No. 1 filter paper in a Buchner funnel and
The identification of the Azolla species is somewhat then blotted as above. The growth rate was usually
tenuous and their taxonomy is subject to revision. expressed as the reciprocal of the doubling times, i.e.,
Voucher specimens of the species employed in this days"'.
study have been deposited at the New York Botanical
Gardens and at the University of Michigan Her-
barium. Living specimens have likewise been provided Dry weights and earbon and nitrogen analysis
to the New York Botanical Gardens and to Dr J. W. Dry weights were determined after drying the plant
Hall, University of Minnesota. The A. caroliniana material to constant weight at 60°C for 24 h. Dried
Willd. is an isolate which was tentatively identified in plant material was pulverized prior to carbon and
1973 (Peters & Mayne, 1974a). Drs D.W. Rains & S.N. nitrogen analysis and these values are expressed as a
Talley at the University of California, Davis kindly percentage of the dry weight. These analyses were
provided cultures of the other Azolla. The A. filieu- determined with a Perkin-Elmer Model 240 Elemental
loides Lam. is from a population on Kauai in the Analyzer equipped with a Control Equipment Corpo-
Hawaiian Islands. The A. pinnata R.Br. is the Malay- ration MC-341-HA microinjector using approxi-
sian population being maintained at IRRI, the Philip- mately 2.5 mg samples of the dried material weighed
pines. The A. mexicana Presl. (SB) is from the Sutter on a Cahn Model G electrobalance. Dinitrodurene
Basin of California. The A. mexieana Presl. (Gl) is was used as a standard.
from one of the populations collected by D.W. Rains
in Guyana. This isolate appeared to be morphologi-
cally distinct from the other New World species and C2H2 reduetion assays
had been designated as a possible strain of A. micro- Azolla plants were incubated for 30-45 min under an
phylla Kaulfaus. However, during the preparation of atmosphere of 10% C2H2:0.03% CO2 in argon in sealed
this manuscript Dr J.W. Hall informed us that studies vials using the media, pH, temperature and light inten-
of sporulating material indicated the isolate from sity employed for their growth. Ethylene was measured
Guyana was not A. mierophylla Kaulfaus but rather as described previously (Peters & Mayne, 1974b),
another population o^ A. mexicana Presl. using a Gow-Mac 750 gas chromatograph.

Growth conditions Speetrophotometrie determinations

All plants were freed of epiphytic photosynthetic con- Chlorophyll was determined according to Wintermans
taminants as described previously (Peters & Mayne, & DeMots (1965). Protein was determined by the
1974a), and while not axenic, were handled with sterile method of Lowry et al. (1951) after treating the
technique. Plants were grown on sterile media in samples as described previously (Ray et al., 1978).
gauze-cotton capped Fernbach or Erlenmeyer flasks. Phosphate was determined according to the procedure
Filtered, humidified air was introduced into the of Fiske & Subbarow (1925). Iron was determined
flasks to maintain the pCOj- When necessary, media colorimetrically using bathophenanthroline (Diehl et
volumes were maintained by the addition of sterile al, 1965) after digestion of the plant material with
H2O. H2SO4 and HNO3.
Nutrient solutions and control of pH are described
in Results. Media and buffers were sterilized separa-
tely. Diurnal temperatures of 25± TC during a 16 h Results
day and 19± 1°C during an 8 h night were maintained
in a large walk-in growth room. Constant growth tem- Nutrient solutions
peratures were maintained by circulating water at the Doubling times of the Azolla species were determined
 OPTIMIZATION OF GROWTH CONDITIONS FOR AZOLLA SPECIES 263

Table 1. Composition of five N-free nutrient solutions used to Mayeux & Evans, 1966), Although there were no pro-
grow Azolla nounced differences in the growth rates of the indivi-
dual species on the five nutrient solutions (data not
Nutrient solution* shown) each species exhibited slightly better growth on
B C D either the 40% Hoagland's solution or the IRRI
medium than in any of the other solutions, A 7-10 day
Imol dm ^ Stock
K2SO4 0,5 1,0
growth interval allowed the use of a readily manage-
CaCl2-2H2O 1,0 2.0 4.0 1.0 1,0 able inoculum and circumvented diminished growth
MgSO4-7H2O 0.40 0.8 1.6 1.6 0.8 associated with: (1) overcrowding in culture flasks and
0.6 (2) a decrease in pH of the nutrient solutions at longer
o.O4 growth intervals.
NaCl 0.25 0.5 1.0 — 0.5
KH2PO4 0.18 0.35 0,7 — 0.35 Growth rates for the Azolla species were subse-
K2HOP4-3H2O 0,03 0,05 0,1 — 0.05 quently determined again using these two media types.
KCl 1,0 2.0 4.0 — In general, the doubling times for the individual species
Sequestrenet 0.2 0.4 0.8 — 0.4 were either the same or those on the IRRI medium
Micronutrient stockj 2,0 2.0 2,0 2,0 2,0 were slightly superior (Fig, 1). Further, the N content
ppm of Macronutrients of the individual species was slightly higher on the
K 48 96 192 39 96 IRRI medium as were rates of N2fixation(C2H2 reduc-
Ca 40 80 160 40 40 tion) determined after 7 days of growth. Therefore, the
Mg 9.5 19 38 40 19
Na 5.8 11.5 23 14 12 N-free IRRI medium was employed in all subsequent
P 6.3 12.5 25 19 12.5 studies.
Fe 0.8 1,5 3 2 1.5
Cl 115 230 460 71 89
S 13 26 52 67 58 Effect ofpH
Initial studies with various concentrations of several
* A, B and C: 20°,,, 40% (Peters & Mayne, 1974a; Ray et al., buffers in IRRI medium during growth intervals of
1978) and 80"„ Hoagland's solution, respectively; D, IRRI
medium (Watanabe et al., 1977) based on an N-free modifica-
6-13 days indicated that a 10 mmol dm ~ ^finalconcen-
tion of a solution employed for rice culture (Yoshida, Forno tration of the buffers employed was sufficient to main-
& Cock, 1971); E, composite medium tain the desired pH, either pH 5 or pH 6, during a 7-day
t Stock solution is 2g of Sequestrene 330 Fe Iron Chelate growth period. Two buffers were then selected for use
containing 10% Fe in 52 cm^ of" water at each pH and compared with unbuffered controls to
X See text
minimize the possibility of the buffer itself, as opposed
to pH, affecting Azolla growth. Growth rates of the
after 7, 14 and 21 days of growth on the five N-free individual species on one of the two buffers used at
nutrient solutions whose compositions are given in each pH are compared to the unbuffered control in
Table 1. Fig. 2. Buffers used in addition to those shown in Fig, 2
Micronutrients were held constant in thefivemedia were: pH 5, citric acid; pH 6, Bis-Tris, 2,2-Bis-(hydrox-
types, using the A-5 micronutrients of Allen (1968) in
which Co(NO3)2 was replaced with C0CI2 (Johnson, 0-6r

0 5- T

0-4
o 0 4
a)
1
^• 0 5
J 03
c

•§ 0 2

? 0-2I-
X3
3
O

01-
Control succ MES MOPS TR(C CHES
pH6 pH7 pH8 pH9
pH5
Figure 2, Comparison of growth rates ± SD of the Azolla species -)n
A.car. Afil. A.mex. (SB) A.mex. (GI) A. pin. IRRI medium buffered at pH 5, 6, 7, 8, or 9. Values represent
duplicate determinations in each of two experiments. Finai buffer
Figure 1, A comparison of growth rates (expressed as 1/doubling concentration was 10 mmol dm~^. Other growth conditions as for
time in days) ± SD of the Azolla species on two media types. Growth Fig, 1. Buffers: SUCC, Succinic Acid; MES, 2(N-Morpholino)eth-
conditions: pH, 5-6; temperature, 26±1719±1 C; photoperiod, ane Sulphonic Acid; MOPS, Morpholino propane Sulphonic Acid;
16/8 h; light intensity, 200 /imol m"^ s~'; aeration, humidified, TRIC, N-tris(hydroxymethyl)methyl glyeine; CHES, 2(N-cyclohex-
filtered air; duration, 7 days. Duplicate samples, 2 experiments. 40% ylamino)ethane Sulphonic Acid. A. caroliniana, •; A. filieuloides, m;
Hoagland's solution, open bars; IRR! medium, hatched bars. A. mexieana (SB), A; A. mexieana (Gl), •; A. pinna ta, • .
264 G. A. PETERS et al

ymethyl)-2,2',2"-nitriloethanol; pH 7, PIPES, Pipera- obtained using a constant temperature. Temperature

zine-N-N'-bis (2-ethane-Sulphonic Acid); pH 8, tolerances of individual species may be slightly differ-
Bicine, N,N-bis(2-hydroxymethyl)glycine; and pH 9, ent if a diurnal temperature cycle is employed.
AMP, 2-Amino-2-methyl-l-propanol. Of the buffers As seen in Fig. 3 there is a good correlation between
employed, only Bis-Tris was clearly inhibitory to growth rates, C2H2 reduction and percentage N con-
Azolla growth. All species exhibited growth compar- tent from 15°C up to the optimum growth temperature
able to that obtained with unbuffered medium (i.e., for the individual species. The somewhat lessened cor-
initial pH 5.6 and final pH >4.5) from pH 5 to 8, but a relation above the optimum temperature, with C2H2
decrease occurred at pH 9 (Fig. 2). reduction declining more abruptly than the doubling
Except for the two buffers used at pH 9, where the time or percent N content, is a result of the experimen-
pH drifted to as low as 8.3 at the end of the 7-day tal design. Doubling times and percentage N content
period, all buffers maintained the pH within 0.2 pH are based on an integrated value over the 7-day period
units of the initial value. Supplementary experiments whereas C2H2 reduction assays reflect the activity pre-
have indicated that the decreased growth at pH 9 is sent during a single time point on the seventh day. In a
probably not attributable to either of the buffers them- related study the various species were maintained at
selves, nor to a reduced phosphate concentration in the 40°C but doubling times and C2H2 reduction rates
medium, but rather may be due to a decreased avail- were determined after 1, 2, 3, 4 and 7 days. Except for
ability of iron. An analysis of A. filiculoides after A.filiculoides, in which there was an 80% decrease in
growth at pH 6 and pH 9 for 7 days at high light C2H2 reduction activity during the first 24 h, all species
intensity (600 /zmol m"^ s"') revealed that the Fe per showed maximum C2H2 reduction rates after 24 h and
mg dry weight was approximately three times greater a decline at each subsequent interval. No species exhi-
at pH 6 than at pH 9. bited more than 10% of the maximal rate on the fourth
day, and none exhibited any activity on the seventh.
Although none of the species were capable of sustained
Temperature growth at 40° the tolerance to 40°C was A. mexicana
Figure 3 a-c shows the effect of temperatures from (GI)>^. mexicana {S%)>A. pinnata> A. caro-
15X to 40°C at pH 6 (MES) and at pH 8 (Tricine) on liniana>A. filiculoides. This is generally consistent
the growth rate, relative C2H2 reduction (the results are with the data shown in Fig. 3. Growth and N content
entirely comparable when expressed on the basis of of the two A. mexicana varieties and A. pinnata are
fresh weight, chlorophyll, or protein), and the percent greatest at 30°C, decreasing moderately at 35 C, while
N for the Azolla species studied. These results were in A. caroliniana they are greatest at 25''-30°C, drop-

A. caroliniana A. filiculoides A. mexicana [SB) A. mexicana {G\) A. pinnata

15 25 35 15 25 35 15 25 35 15 25 35 15 25 35
Temperature (°C)
Figure 3. (a) Growth rates, (b) C2H2 reduction on basis of fresh weight, and (c) percentage N as a function of temperature
(constant) at pH 6 (closed symbols) and 8 (open symbols) ± SD for the Azolla species studied. Other growth conditions as for Fig
1. Total replicates: 4 at 15", 20°, 35° and 40"C; 8 at 25° and 30°C.
 OPTIMIZATION OF GROWTH CONDITIONS FOR AZOLLA SPECIES 265

ping appreciably at 35°C, and for A, filieuloides they (a)

are greatest at 25°C. The percentage dry weight and the
C/N ratio exhibited a temperature dependence in all 0 6
species, generally being appreciably higher at the tem-
perature extremes than at the optimum. For example,
in A. mexieana grown at pH 6, the percentage dry
weight at 15°, 30° and 40X was 13.7 ±1.68, 5.83 ±0.38
and 11.21 ± 1.62, respectively, while the C/N ratio at 3 0-4
these three temperatures was 12.65 ±0.49, 6.80 ±0.29,
E
and 16.30 ±0,88, respectively. The effect of tempera-
ture on the percentage dry weight was less pronounced c
in some of the other species. The chlorophyll a/b ratio JQ
was relatively constant from 15° to 30°C in all the 1,0-2
species. In A. filieuloides grown at pH 6, for example,
the chlorophyll a/b ratio at 15°, 20°, 25° and 30°C was
4.25 ± 0,30, 4,34 ± 0,11,4,24 ± 0,36 and 4,11 ± 0.28, re-
spectively. Since the Azolla plant contains both chloro-
phyll a and b., while the endophyte contains only
chlorophyll a (Peters & Mayne, 1974a), the consistency 8
of this ratio suggests that the composition of the sym-
biotic association remains constant in this temperature
range. The value of this ratio declined in some species - 6
at 35°C and in all species at 40°C, It has not been
0)
determined whether this reflects a decrease in the
amount of the endophyte or a preferential degradation X
of the Azolla plant's chlorophyll a at the higher growth U
CM

temperatures. There was essentially no difference in the "o

results obtained at pH 6 and pH 8 in any of the species E
at the temperatures used. A. earoliniana, and to a lesser
extent the A. mexieana varieties, exhibited antho- 0
cyanin formation at 15°C, At 40°C A. filieuloides was II lumination { h )
bleached. A, earoliniana a bright red and A. mexieana
Figure 4. (a) Comparison of growth rates ± SD of the Azolla species
(SB) and A. mexieana (Gl) were brownish-green and as a function of three photoperiods with plants grown on IRRI
reddish-brown, respectively. In this study the A, pin- medium buffered at pH 6 (closed symbols) and pH 8 (open symbols).
nata remained green at all temperatures as did the Temperature: 25°C for A. earoliniana, ?ind filieuloides; 30^C for A.
other species at 20°-35 C, mexieana, and pinnata. Other growth conditions as for Fig. 1. Total
replicates: 6. [Star (<^) indicates the average doubhng time for all
species with natural illumination and photoperiod in glasshouse
study.] (b) C2H2 reduction ± SD on the basis of fresh weight as a
Photoperiod function of three photoperiods for the Azolla species grown at pH 6
The growth rate of all species increased with increasing (closed symbols) and 8 (open symbols). Symbols as for Fig. 2, (For
statistical analysis of these data see Appendix 1).
photoperiod (Fig. 4a), With a 12 h light period all
species doubled their biomass every 2,4-3,6 days; with
constant illumination, every 1.6-2.0 days. While the Light intensity andpH
rate of C2H2 reduction was usually somewhat higher at
the 16 h light period (Fig, 4b), the dry weight as a Growth rates of the Azolla species were compared at
percentage of the fresh weight, the percentage N con- pH 5 (succinate), pH 6 (MES) and pH 9 (CHES) at 100
tent and C/N ratio were comparable in the individual and 600 /imol m"-^ s"'. This corresponds to approxi-
species at the 16 h and 24 h light period. mately 8000 and 48,000 lux, respectively, based on a
conversion factor of 12.5 ^mol m"^ s"' per klux
(McCree, 1972). There was no indication of an interac-
Light intensity tion of light and pH for any of the species (Fig, 6).
While the growth rate of all species increased with light Increasing the light intensity from 100 to 600
intensity up to 400 //mol m"^ s"' (Fig, 5a), C2H2 m"^ s"' resulted in approximately the same increase in
reduction activity was saturated at 200 ^umol m"^ s"' growth rate at all three pH values in all the species and
(Fig. 5b). Previous studies have shown that in A. earo- did not alleviate the diminished growth observed pre-
liniana, C2H2 reduction is saturated at lower light in- viously at pH 9 and 200 /imol m-^ s"' (Fig. 2). At 600
tensities than CO2 fixation (Peters, 1976). The percent- ;umol m-^ s"' the doubling times for all species were
age dry weight, chlorophyll a/b ratios, percentage N, between 1.6 and 2.0 days at both pH 5 and pH 6. C2H2
and C/N ratios for the individual species at pH 6 and pH reduction activity showed a trend comparable to that
8 showed little variation as a function of light intensity. for growth in all the species, since, as shown in Fig. 5 a
266 G. A. PETERS et al.

0-6
\

0-5
I
o
T3 0-4
I
g> 0 - 3 \

.a
3
O
0-2

01

pH5 pH6 pH9

Figure 6. Comparison of growth rates ± SD for the Azolla species
grown on IRRI medium at pH 5, 6 and 9 as a function of high (600
^mol m""-^ s~'. closed symbols) and low (100 /zmol m"-^ s~' open
symbols) light intensity. Other growth conditions, etc. as for Fig. 4.
Symbols as for Fig. 2.

above (star. Fig. 4a), the average doubling times for all
100 190 420 600 species over the entire study fell between those
Light intensity (ymol rrf^s" obtained with the 12:12 and the 16:8 photoperiods.
Figure 5. Comparison of growth rates + SD (a) and C2H2 reduction Table 2 shows the percentage dry matter, chlorophyll
activities on a fresh weight basis ± SD (b) as a function of light a/b ratios, percentage N on a dry weight basis and the
intensity at pH 6 (closed symbols) and pH 8 (open symbols) for the C/N ratio for the individual species at the end of the
Azolla species using a 16 h light period. Other growth conditions as fifth week.
for Fig. 4. Symbols as for Fig. 2.

& b, 100 //mol m"2 s"' is not saturating for either Optimized growth eonditions for laboratory culture
growth or C2H2 reduction. Employing the optimum constant temperature for
growth, i.e., 25''C for A.filiculoides and A. caroliniana
Glasshouse study and 30 C for the others (see Fig. 3a), all species
doubled their biomass in 2 days or less and contained
The Azolla species were grown on IRRI medium at pH 5-6% N when grown on IRRI medium with a light
6 and a constant temperature of 27.5°C in a glass- period of at least 16 h and a light intensity of at least
house. The natural photoperiod during the 5 week 400 /imol m"2 s"' (Table 3). The doubling times under
study from May 11 to June 15 averaged 14.5 h of light. these conditions are comparable on medium buffered
Light intensities varied from less than 100 to about at pH 5, 6, 7 and 8. Table 3 also summarizes the values
1700 jumol m~^ s"' between sunrise and sunset and as a obtained for percentage dry matter, C/N ratio and
function of cloud cover. Doubling times, determined at chlorophyll a/b ratio during these two final studies in
weekly intervals, were quite constant for all species, which the Azolla species were grown under optimal
falling between 2.5 and 3.0 days (Table 2). As shown conditions.

Table 2. Doubling times ± SD, percentage dry matter, percentage N,

C/N ratio and Chl a/b ratio and for the Azolla species during a
glasshouse study. The values for the doubling times reflect five deter-
minations in duplicate at weekly intervals. The other values are based
on individual samples at the end of the fifth week

Doubling Times
Species (Days) % Dry Matter %N C/N Chl a/b
A. ear. 2.75 + 0.14 4.77 4.85 8.61 4.26
A.fil. 2.65 + 0.11 5.16 5.97 7.06 4.07
A. mex. (SB) 2.64 + 0.13 5.22 5.85 7.35 4.52
A. mex.(Gl) 2.54 + 0.10 5.34 5.53 7.60 3.91
A. pin. 2.61 ±0.13 5.92 5.34 7.46 3.98
 OPTIMIZATION OF GROWTH CONDITIONS EOR AZOLLA SPECIES 267

Table 3. Doubling times, percentage dry matter, percentage N, C/N ratio and Chl
a/b ± SD in Azolla species grown under 'optimal laboratory conditions' (see Text).
Each value is based on four samples

Doubling time 0
Species (Days) (, Dry Matter %N C/N Chl a/b
A. car. 1.95 + 0.10 5 .58 + 0.30 4.79 + 0.20 8.66 + 0.30 4.35±0. 17
A. fil. 1.83+0.10 5 .26 + 0.20 6.24 + 0.42 6.73 + 0.36 4.05 + 0.08
A. mex. (SB) 1.92 + 0.11 5 .24 + 0.19 6.05 ±0.30 6.99±0.19 3.96 + 0. 18
A, mex. (GI) 1.87±0.10 5 .37 + 0.55 5.16 + 0.41 8.20 + 0.62 4.12 + 0.49
A. pin. 1.78 ±0.08 5.05 + 0.37 4.90 + 0.86 8.12 + 0.74 3.77 + 0. 17

Discussion iron deficiency. Interestingly, infieldstudies in Cahfor-

nia, where the paddy water was pH 8.0-8.5, supple-
To our knowledge this is the first study to characterize mental iron was required to support Azolla growth
and directly compare growth responses as a function of (Talley ^ra/., 1977).
controlled environmental variables in clones of four of Of the environmental variables studied, only con-
the six known Azolla species. As indicated previously, stant growth temperature was clearly species specific.
caution is required in the identification of the New Since different populations of individual species and
World Azolla species. In the absence of sporulating different photoperiods, light intensities, etc. have been
material, identification is extremely tenuous since mor- employed by others, caution is required in comparing
phological aspects of an individual species can vary these results with those obtained by others. Eor exam-
considerably (c.f Pieterse, DeLange & Van Vliet, ple, Watanabe et al. (1977) used a diurnal temperature
1977). cycle along with a 12 h photoperiod and found that the
A number of N-free nutrient solutions having a optimal growth of A. pinnata occurred at a 32°/24°C
relatively wide variation in ionic strength have been day-night period and that growth dropped to half the
used for Azolla growth (Becking, 1972; Olsen, 1972; maximal rate at 35°/27°C. Our A, pinnata clone exhi-
Ashton, 1974; Peters & Mayne, 1974a; Brotonegoro bited maximal growth rates at 30°C and only dropped
6 Abdulkadir, 1976; Watanabe et al., 1977; Ray et al., 25-30% at 35°C. Hoist & Yopp (1979) showed that
1978; Hoist & Yopp, 1979; Talley & Rains, 1980; also under a 14 h photoperiod and a light intensity of 13
see Moore, 1969). While either 'too high or too low' a klux, the growth rate of A, mexieana at 21°C was
concentration of macronutrients reportedly produced nearly twice that at 30°C with almost no growth at
'marked effects' on the growth of A. filiculoides (Ash- 35°C and death at 40°C. This is contrary to the results
ton & Walmsley, 1976) and full strength Knop's solu- presented here which indicate the optimal growth tem-
tion killed A. pinnata after several weeks (Moore, perature for A. mexicana is 30"C. Moreover, Talley et
1969), no pronounced differences were found in the al. (1977) have reported that in field studies A. mexi-
doubling times of the five Azolla species with the five cana withstands higher temperatures than A. filicu-
nutrient solutions employed in this study. The reasons loides. While Ashton & Walmsley (1976) stated that
for selecting IRRI medium for use in studies on the South African populations of A. filiculoides grew from
effect of selected environmental variables on growth 5°C to 45°C with a maximal growth at 27.5°C in
and nitrogen fixation were presented in the results. controlled environmental studies, Talley & Rains
A number of investigators have reported that Azolla (1980) found growth and N content of a California
species are capable of growth from pH 4 to pH 8 or 10 population of A, filieuloides to be greatest under a
with the best growth occurring between pH 5 and pH 8 25715°C to 30720°C diurnal temperature cycle with a
(Nickell, 1958; Olsen, 1972; Ashton & Walmsley, 1976; 12 h light period at 500 jumol m~^ s"'.
Watanabe et al,, 1977; Hoist & Yopp, 1979). Hoist & The present study shows that under the conditions
Yopp (1979) used Tris, MES or citrate, in addition to employed, growth rates of each species continue to
phosphate as buffers and found that ^. mexicana grew increase with increasing light periods. Moreover, the
well from pH 4.2 to 8.0, with a small drop at pH 8.8. In data suggest that total N increases with increasing light
the other studies the pH was either adjusted daily or no periods only by virtue of total biomass increase.
details were provided on how the pH was maintained. Diminished growth rates and/or N2ase activity of
Diminished growth has been shown to occur above pH individual species at high hght intensities have been
7 in ^. caroliniana (Olsen, 1972) and A. pinnata reported (Ashton & Walmsley, 1976; Talley et al,,
(Watanabe et al., 1977), but this is largely overcome by 1977) and high light intensities have been suggested to
increased levels of chelated iron. The present study be the factor most responsible for senescence (Hoist &
demonstrates that each of the Azolla species exhibit Yopp, 1979). Moreover, Ashton & Walmsley (1976)
growth comparable to that of an unbuffered control reported interacting effects of light intensity and pH of
(initial pH 5-6; final pH 4-5) when buffered at pH 5,6, the medium on the growth of A. filiculoides. At low
7 or 8 but that growth is diminished at pH 9; on the light (15,000 lux) growth was greatest at pH 5, at high
basis of a preliminary study the latter is attributed to light (60,000 lux) it was greatest at pH 9-10 and at
268 G. A. PETERS et al.

intermediate light (30,000 lux) growth was comparable Ashton, P.J. & Walmsley, R.D. (1976) The aquatic fern Azolla and
at pH 5 and pH 9-10. None of these effects were its Anahaena symbiont. Endeavour, 35, 39-43.
observed here. In all species growth was saturated at Becking, J.H. (1972) Symbiosen: stickstoff-bindung. Fortschritte der
400 /zmol m " ' s"' and N2ase activity was saturated Botanik, 34, 459-^67.
Becking, J.H. (1979) Environmental requirements of Azolla for use
between 100 and 200 /zmol m"^ s"'. In the supplemen- in tropical rice production. In International Rice Research Insti-
tal glasshouse study, with light intensities up to 1700 tute, Nitrogen and Rice. pp. 345-373. IRRI, Los Banos, Laguna,
pxno\ m"-^ s~' on clear days, the grow1;h rates (Table 2) Philippines.
and N content of the individual species were quite Brotonegoro, S. & Abdulkadir, S. (1976) Growth and nitrogen-fix-
ing activity oi Azolla pinnata. Annales Bogorienses, 6, 69-11.
comparable to those obtained under optimal growth Dao, T.T. & Tran, Q.T. (1979) Use of Azolla in rice production in
conditions in the laboratory. Vietnam. In International Rice Research Institute, Nitrogen and
High plant density (Becking, 1979; Hoist & Yopp, Rice, pp. 395-^05. IRRI, Los Banos, Laguna, Philippines.
1979) and a decrease in the pH of growth solutions can Diehl, H., Smith, G.F., McBride, L. & Cryberg R. (1965) The Iron
diminish growth rates. In order to obtain maximal Reagents: Bathophenanthroline, Bathophenanthroline disulfonic
acid, 2,4,6-Tripyridyl-S-triazine, Phenyl-2-pyridyl Ketoxime, pp.
growth rates plant crowding must be precluded and the 1-64. The G. Frederick Smith Chemical Company, Columbus,
pH controlled. This study has shown that the Azolla Ohio.
species studied can all double their biomass in 2 days or Fiske, C.H. & Subbarow, Y. (1925) The colorimetric determination
less while maintaining an N content of 5-6%. A com- of phosphorus. Journal of Biological Chemistry, 66, 375-400.
Hoist, R.W. & Yopp, J.H. (1979) Studies of the Azolla-Anabaena
pilation of previously reported and/or estimated symbiosis using Azolla mexicana, I. Growth in nature and labora-
growth rates and N contents of laboratory and field- tory. American Fern Journal, 69, 17 25.
grown A. caroliniana, A.filiculoides, A. mexicana and Johnson, G.V., Mayeux, P.A. & Evans, H.J. (1966) A cobalt require-
A. pinnata (Becking, 1979), as well as additional ment for symbiotic growth of Azollafiliculoidesin the absence of
reports for A. mexicana (Hoist & Yopp, 1979) and A. combined nitrogen. Plant Physiology, 41, 852-855.
Liu, C. (1979) Use of Azolla in rice production in China. In Interna-
filiculoides (Talley & Rains, 1980), indicate that the tional Rice Research Institute, Nitrogen and Rice, pp. 375-394.
values reported here equal or significantly exceed IRRI, Los Banos, Laguna, Philippines.
them. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. (1951)
Protein measurement with the Folin phenol reagent. Journal of
Biological Chemistry, 193, 265-275.
Acknowledgments McCree, K. J. (1972) Test of current definitions of photosynthetically
active radiation against leaf photosynthesis data. Agricultural
This research was supported by NSF/ASRA Grant Meteorology, 10, 443-453.
PFR 77-27269. The authors also wish to thank S. Moore, A.W. (1969) Azolla: Biology and agronomic significance.
Botanical Review, 35, 17-34.
Dunbar and M. Zurmehly for preparation of figures Nickell, L.G. (1958) Physiological studies with Azolla under aseptic
and photographs, and D. Patten for manuscript prepa- conditions. I. Isolation and preliminary growth studies. American
ration. Fern Journal, 48, 103-108.
Olsen, C. (1972) On biological nitrogen fixation in nature, particu-
larly in blue-green algae. Comptes Rendus des Travaux du Lahora-
Appendix 1. Statistical analysis of photoperiod data toire Carlsberg, 37, 269-283.
Peters, G.A. (1975) The Azolla-Anabaena azollae relationship. III.
The differences in growth of the individual species as a function of
Studies on metabolic capabilities and further characterization of
photoperiod were statistically significant at the 95% confidence level
the symbiont. Archives of Microbiology, 103, 113-122.
except for the difference between the 12 h and 16 h photoperiod for
Peters, G.A. (1976) Studies on the Azolla-Anabaena azollae sym-
A. caroliniana at pH 6 and 8 and the two /I. mexicana strains at pH 8.
biosis. In Proceedings of the 1st International Symposium on
There was no significant difference at this confidence level in the
Nitrogen Fixation, Vol. 2 (eds W.E. Newton & C.J. Nyman), pp.
growth of any species at the two pH values within a given photo-
592-610. Washington State University Press, Pullman, Washing-
period. Considering all the species collectively their growth at pH 6
ton.
and 8 at each photoperiod was statistically significant from the other
photoperiods but there was no significant difference in the growth at Peters, G.A. & Mayne, B.C. (1974a) The Azolla, Anabaena azollae
the two pH values within a given photoperiod. There was no statisti- relationship. I. Initial characterization of the association. Plant
cally significant difference in the rate of C2H2 reduction for any Physiology, 53, Sn~Sl9.
species at pH 6 versus pH 8 at any photoperiod. The rate of C2H2 Peters, G.A. & Mayne, B.C. (1974b) The Azolla, Anabaena azollae
reduction at 12 h and 24 h was statistically significant from that at 16 relationship. II. Localization of nitrogenase activity as assayed by
h at both pH 6 and 8 when all species were considered. Statistically acetylene reduction. Plant Physiology, 53, 820-824.
significant differences within a species were limited to 12 h versus 16 Pieterse, A.H., Lange, L. De& Van Vliet, J.P. (1977) A comparative
h at pH 8 for A.filiculoides, 16 h versus 24 h at pH 6 for A. mexicana study of Azolla in the Netherlands. Acta Botanica Neerlandica, 26,
(SB) and 16 h versus 24 h at pH 6 and 16 h versus 24 h as well as 12 h 433-449.
versus 24 h at pH 8 for A. pinnata. Rains, D.W. & Talley, S.N. (1979) Use of Azolla in rice production
in North America. In International Rice Research Institute,
Nitrogen and Rice, pp. 419-431. IRRI, Los Banos, Laguna, Philip-
pines.
Ray, T.B., Peters, G.A., Toia, R.E., Jr & Mayne, B.C. (1978) The
References Azolla-Anabaena relationship. VII. Distribution of ammonia-
Allen, M.M. (1968) Simple conditions for growth of unicellular assimilating enzymes, protein, and chlorophyll between host and
blue-green algae on plates. Journal of Phycology, 4, 1-3. symbiont. Plant Physiology, 62, 463^67.
Ashton, P.J. (1974) The effect of some environmental factors on the Ray, T.B., Mayne, B.C., Toia, R.E., Jr & Peters, G.A. (1979) The
growth of AzollafiliculoidesLam. In The Orange River, Progress Azolla-Anabaena relationship. VIII. Photosynthetic characteriza-
Report (ed. E.M.V. Zinderen Bakker, Sr), pp. 123-138. Institute tion of the association and individual partners. Plant Physiology,
for Environmental Sciences, University of the Orange Free State, 64,791-795.
Bloemfontein, South Africa. Singh, P.K. (1977) Multiplication and utilization of fern Azolla
 OPTIMIZATION OE GROWTH CONDITIONS EOR AZOLLA SPECIES 269

containing nitrogen-fixing algal symbiont as green manure in rice Watanabe, I., Espinas, C.R., Berja, N.S. & Alimagno, B.V. (1977)
cultivation. // Riso, 26, 125-137. Utilization of the Azolla-Anabaena complex as a nitrogen fertilizer
Singh, P.K. (1979) Use of Azolla in rice production in India. In for rice. International Rice Research Institute Research Paper
International Rice Research Institute, Nitrogen and Rice, pp. Series, 11, 1-15.
407-^18. IRRI, Los Bafios, Laguna, Philippines. Wintermans, J.F.G.M. & DeMots, A. (1965) Spectrophotometric
Svenson, H.K. (1944) The New World species of Azolla. Ameriean characteristics of chlorophylls a and b and their pheophytins in
Fern Journal, 34, 69-84. ethanol. Biochimica Biophysica Acta, 109, 448-453.
Talley, S.N., Talley, B.J. & Rains, D.W. (1977) Nitrogen fixation by Yoshida, S., Forno, D.A. & Cock, J.H. (1971) Laboratory Manual
Azolla in rice fields. In Genetic Engineering for Nitrogen Fixation for Physiological Studies of Rice, p. 61. International Rice
(ed. A. Hollaender), pp. 259-281. Plenum Press, New York. Research Institute, Los Bafios, Laguna, Philippines.
Talley, S.N. & Rains, D.W. (1980) Azolla filiculoides Lam. as a
fallow-season green manure for rice in temperate climate.
Agronomy Journat, 72, 11-18.