RESEARCH ON ALGAE, BLUE-GREEN ALGAE, AND PHOTOTROPHIC NITROGEN FIXATION AT THE INTERNATIONAL RICE RESEARCH INSTITUTE (1963-81), SUMMARIZATION, PROBLEMS , AND PROSPECTS1 ABSTRACT This paper provides general information on one of the major research areas of IRRI'S Soil Microbiology Department. Con- clusions made after an extensive survey of the literature on the role of blue-green algae in rice cultivation are summa- rized. Microbiological research at IRRI on algae, blue-green algae, and phototrophic nitrogen fixation is reviewed. The paper also provides information on problems encountered and prospects for future research. Date 4 J A M 1983 I i . 'By P. A. Roger, visiting scientist (ORSTOM, 24, Rue Bayard-75008, Paris, France), and I. Watanabe, soil microbiologist, IRRI, Los Baños, Laguna, Philippines. Submitted to the IRRI Research Paper Series Committee February 1982. RESEARCH ON ALGAE, BLUE-GREEN ALGAE, AND PHOTOTROPHIC NITROGEN FIXATION , AT THE INTERNATIONAL RICE RESEARCH INSTITUTE (1963-81), SUMMARIZATION, PROBLEMS, AND 'PROSPECTS I It has long been recognized that flooded rice soils are .still not well known. No systematic maintain a degve of nitrogen fertility. Long-term analysis has correlated 'their presence or fertility trials have indicated that there is a absence with environmental factors. major input of nitrogen (35-50 kg N/crop) into e ' flooded rice soils. Much of this input is attrib- e Ecological studies of BGA in submerged soils uted to nitrogen fixation by: are limited by problems in methodology, pri- marily in estimating algal biomasses quan- e heterotrophic microorganisms associated titatively. Fragmentary quantitative mea- with rice roots, surements indicate that N2-fixing BGA popu- lation densities vary from a few to 10719 e azolla, dry soil; biomasses vary from a few kilo- I I grams to 24 t (fresh weight)/ha. In a BGA, and favorable environment a NZ-fixing algal bloom may contribute 30-60 kg N/ha to the photosynthetic bacteria. ecosystkm. Nitrogen fixation has been a major research area Among the physical factors affecting the since the establishment of soil microbiology re- seasonal fluctuations of the phytoplankton, search at IRRI in 1963. Equal importance has been light is the most important. It affects given to the study of the different groups of ni- the algal biomass qualitatively and quan- trogen fixers. This paper summarizes research con- titatively. BGA may be regarded as a low- ducted on the photosynthetic free-living nitrogen light species; in areas of high incident fixers. light intensities, BGA develop only when protected by a sufficiently dense rice Early IRRI studies of biological nitrogen-fixing canopy. In areas of moderate incident activity (NFA) used 15N isotope. In 1968 the in- light intensities, during rainy or cloudy expensive and sensitive acetylene reduction assay weather, light deficiency under a dense (ARA) method to detect nitrogen fixation was in- rice canopy may limit BGA growth. To a troduced, stimulating research activities bn NFA. lesser extent temperature and water regime As a result of discussions during the 1978 Nitro- may also influence BGA growth in the wet- gen and Rice Symposium at IRRI, research priority. land field. was given to NF'A by azolla, BGA, and heterotiophic microorganisms in the rice root zone. IRRI was Among the biotic factors capable of lim- asked to compile all relevant information on the iting BGA growth in rice fields, only graz- BGA that relates to rice. ing by invertebrates has been documented. Evidence exists that pathogenicity and an- The information was compiled in 1979 and Blue- tagonisms may affect BGA in the field but ' green algae and rice was published in 1980. those have not yet been demonstrated. Since 1979, the relative importance given to IRRI e Among soil properties, pH is the most im- research on BGA and phototrophic NFA has in- portant factor determining the atgal flora creased. Trials for evaluating phototrophic NFA composition. BGA prefer environments that have continued and research on relations between are neutral to alkaline. A positive cor- BGA and the rice plant and on factors limiting BGA relation between pH and occurrence of has started. BGA is common. Next to pH, available phos- phorus content of the soil is the most de- cisive factor favoring BGA growth. Little BLUE-GREEN ALGAE AND RICE I 1 information is available on the effect of other soil properties on BGA. Based on the literature review presented by Blue- green algae and rice (Roger and Kulasooriya 1980) e Agronomic practices for growing rice influ- the folLowing general, statements can be made. ence BGA growth. I l - Land preparation and management have only e The wetland rice environment is favorable iqcidental effects . for the growth of blue-green algae (BGA). The relative occurrence of BGA, however, - Pesticides -- depending on their nature, ' varies greatly and they are not always pre- their concentration, and the algal -- I sent in rice soils. Reasons €or their thete- strains can have inhibitory, selec- rogenous and sometimes limited distribution tive, or stimulatory effects. BGA are 4 IRPS No. 78, June 1,982 generally more resistant to pesticides field varies within large limits and seems than other algae and tolerate pesticide to be more affected by nitrogen fertilizers levels recommended for field application. than heterotrophic nitrogen fixation. BGA Insecticides are generally less toxic to epiphytism makes a limited contribution to BGA than other pesticides and have the the nitrogen input in shallow-water rice, secondary beneficial ef feet of suppres- but this contribution has agronomic signif- sing the population of algae grazers. icance in deepwater rice. - Among chemical fertilization practices, e BGA have benefited rice plants by the pro- phosphorus application and liming of duction of growth-promoting substances. The acidic soils have a beneficial effect on additive effect Ôf algalization in the pre- BGA growth. The effect of nitrogenous sence of a high level of nitrogen fertili- fertilizers is not well known. BGA growth zer was interpreted as an index of this G was inhibited by mineral nitrogen in flask growth-promoting effect, but such an intser- cultures but that may not occur to the pretation has not been demonstrated in the same extent in the field. The effect of field and should be treated with caution. nitrogen fertilizers on BGA in the field Beneficial effects,of BGA on rice, such as has received little attention. This is increasing phosphorus availability, de- surprising in view of the observation that ~ creasing sulfide injury, and preventing the BGA inoculation produced an increase in .. growthpf weeds, have also been reported. grain yield even at high levels of fer- '- tilizer N. From experiments conducted e Because BGA are recognized as one of the without algal inoculation, a depressive most important nitrogen-fixing agents in effect of nitrogen fertilizer on algal NFA flooded rice soils, many trials have stud- has been established. Other nutrients (Mo, ied the increase in rice yield by algal in- Fe, Mg, K, etc.) are required for optimal oculation (algalization). growth of BGA, but their ecological implica- tions as liming factors, or as factors affect- - Most of the experiments were on a black ing the composition of the algal community in -box basis and examined only the grain wetland rice fields, have not been documented. yield effect of an agronomic practice - (algalization); the intermediate effects Organic manure, depending on the nature were not studied. Little information is and mode of application, may favor or de- available on the qualitative and quanti- press BGA growth. Plant residue incorpo- tative evolution of the nitrogen-fixing ration, which produces by-products of an- algal flora, the evolution of the photo- - aerobic decomposition toxic to algae, synthetic NFA, and the nitrogen balance seems less beneficial to BGA than surface in inoculated paddy soils. Pot experi- application. ments may be suitable for qualitative studies, but they overestimate the ef- e In physiological studies on BGA in wetland fects of algal inoculation. On the other rice fields, much emphasis has been placed hand, mast of the field experiments have on the NFA, whereas the study of in situ been conducted over one growing season productivity and photosynthetic activity only and may have underestimated the ef- has been neglected. A conceivable role of fects of algalization. The advantages of CO2 depletion as a limiting factor for a a slow nitrogen release might not be ap- ARA when large biomasses occur, and the in- parent in the first crop after algal ino- fluence of the concurrent pH increase on cula tion. nitrogen losses by volatilization, need to be documented. - e Algal NFA has most frequently been studied Algalization has been reported to have a by ARA measurement. .This method is liable beneficial effect on grain yield in seve- to give misinterpretation of quantLtative ral countries. There are also reports, results, but it is convenient and reliable however, that indicate failure of algali- for qualitative studies when the measure- zation in widely different agroclimates. ments are brief, the problems 'of gas dif- Little is known about the limiting fac- fusion and greenhouse effects are minimize tors for algalization. Among the soil minimized, and statistically valid sa- properties, a low pH and low available pling methods are adopted. phosphorus content are the only ones well documented. Knowledge of the relation be- Estimates of the amounts of nitrogen fixed tween soil properties and the establish- by BGA vary from a few to several kilograms ment of the algal inoculum is certainly a per crop. The average value of the reported major gap. Among detrimental biotic fac- estimates (30 kgicrop) seems to constitute tors, only grazing by zooplankton has a satisfactory reference value when envi- been studied. Low temperature, heavy ronmental factors favor BGA growth. The re: rains, and cloudy weather have also been lative contribution of BGA as a percentage reported to limit the establishment of of the total nitrogen fixed in the rice the' algal inoculum. IRPS No. 78, June 1982 5 I . - Algalization, when effective, has been As a general conclusion, a beneficial role .of BGA reported to increase the size of the rice in wetland rice fields appears positive. The abun- plant, its nitrogen content, and the num- dance of a sometimes repetitive literature on this ber of tillers, panicles, spikelets, and subject clearly indicates that researchers have filled spikelets per panicle. The better felt the importance and the potentialities of BGA grain yield has' been used to assess the in ric,e cultivation. Unfortunately the ecology of effect of -algal inoculation. From the re- BGA in rice fields ,and their modes ,of action on ports on field experiments, mainly in In- the plant are skill poorly understood. This limits dia, it appears that algal inoculation, the use of BGA as a biofertilizer. where effective, causes about 14% rela- tive increase in yield, corresponding to about 450 kg gtainlha per crop. PHOTOTROPHIC NITROGEN FIXATION , RESEARCH AT IRRI (1963-81) - A higher increase in grain yield was ob- served when algalization was done in com- IRRI work with BGA since 1964 has ranged from eco- bination with lime and phosphorus, and logical studies to field tests of cultural practi- sometimes with molybdenum application. It ces. A summary of papers on algae, blue-green al- appears, however, that there is no sig- gae, and photosynthetic. nitrogen fixation pub-' nificant difference in yield increase, lished by IRRI authors is presented in Appendix 1. , I strictly due to algalization in the pre- / sence or absence of non-nitrogen fertili- zers and that the increase in yield h e Ecological and. physiological studies on algae to non-nitrogen fertilizers is generally higher than that due to algalization. Studies on the,changes in soil algae and BGA after flooding of the soil were done in 1963 and 1964. - Results concerning the effects of algali- Populations were assessed by recording growth af- zation in the presence of nitrogen ferti- ter incubation in a liquid medium and applying the lizers are controversial. Several reports, most probable number estimation. Studies with a , indicate failure of algalization in the soil with pH 6.6 (Table 1) indicated that Chloro- presence of nitrogen fertilizer. A large- phyceae, mainly unicellular , were dominant. Among . scale experiment in India, however, indi- the BGA, Qscillatoriaceae were generally emore- cates a beneficial effect of algalization abundant than Nostocaceae. The population of ' even at high levels of nitrogen. tocaceae reached its peak 48 days after flooding but was only of the order of 1,7OO/g of soil. Sim- - There is evidence that algalization pro- ilar results were obtained with an acidic soil (pH duces both a cumulative and residual ef- 4.7,). From these data it was concluded that a ma- fect attributed to a buildup of, the soil jor role of BGA in nitrogen fixation in the soils nitrogen, organic matter, and the algal studied was doubtful. However, because no measure- flora. However, little is known about the ment of the NFA was done and the methodology used effects of algalization on soil ,proper- for enumeration was demonstrated to underestimate ties and soil microflora. algal populations.(Reynaud and Roger 1977), this conclusion may be erroneous, particularly in the case of the soil with a pH of 6.6. Another knowledge gap concerns the compa- rison between algalization and management practices that enhance the growth and ac- Table 1. Algal populations developing in flooded tivity of indigenous natural populations clay sgilipH 6:6; OM 2.0%; total N 0.14%). (From of nitrogen-fixing BGA. In some, cases, IRRI L1965-1. management can make algalization unneces- sary but it is necessary where efficient No. of days Chloro- Oscillate- ] Nos toca- strains are absent in khe soil. The submerged ghvceae riaceae cese search for highly efficient strains is O 500 200 200 still at a theoretigal level and the re- canrnended inoculum remains a soil-based 13 1900 1900 200 mixture of strains. 48 1700 1700 1700 - $9 1300 O O In India algalization in rice fields has I proceeded a little beyond the 'stage of go,, 1200 200 1O 0 fundamental research. A method of produ- cing algal inoculum, easily adoptable by farmers, has been developed and recommen- dations for field inoculation have been Occurrence and, abundance of total algae and BGA fn I developed; Inoculum production and alga- flooded Maahas clay at different stages of rice lization technology indicated a cost- growth were studied in the greenhouse in J968 benefit ratio of 1 to 10 and an addition- (Fig. 1). High numbers of algae and BGA in the 2- al income of about 300 Indian rypeeslha to 7-cm soil layer indicated a high concentration per crop in 1979. To our knowledge, such of spores in the soil. Algal population was gener- trials are still confined to India. ally highest at mximum tillering. Application of 6 IRPS No. 78, June 1982 Log (na)/g soil I I Algae Blue-green algae No plant 0-2cm 'I 2 3 4 I 2 3 4 I 2 3 4 I 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 Growth stage of the rice plant Fig. 1. Algal population during the rice-growing period (L = light, D = dark; depth of soil 0-2 cm and 2-7 cm; stages of rice: 1 = transplanting, 2 =maximum tillering, 3 = panicle initiation, 4 = maturity (IRRI 1968). ammonium sulfate had a depressive effect on nitro- chemicals that are less phytotoxic to rice plant gen-f ixing BGA. Of 308 nitrogen-fixing isolates, was undertaken in 1969. Of the commercially avail- 267 belonged to the Nostoc species, 38 to Anabae- able algicides and herbicides, monoquat and di- -na, 1 to Stigonema, 1 to Tolypothrix, and 1 to chlone, tested at the rate of 0.5 and 2.5 kg Scytonema. Gloeotrichia and Aphanothece, which are a.i./ha, were effective at both rates (IRRI 1970). cmmon at present on IRRI farm, were not recorded. Dichlone 50 WP was found efficient (IRRI 1971) and much less harmful to rice plant than copper sul- Measurements of specific ARA among 12 strains gave fate. values ranging from 0.5 to 5.4 nmol C2Hq/min per mg protein with an average value around 2 Further experiments (IRRI 1972) indicated that (IRRI [1968]). These values are in agreement with fungicides like benomyl and maneb can also effec- those reported in the literature (see Reynaud and tively control algae. The best control resulted Roger 1979). from Dichlone 50, WP at 6 kglha. A laboratory study of the effects of light inten- Evaluation of phototrophic NFA in paddy fields sity on BGA showed that regardless of the age of the culture, the maximum NFA was obtained at 10 The first IRRI trial to estimate phototrophic NFA klux. At 5 klux NFA was about 80% of the maximum. was in 1968. Soil from pot experiments was incu- Activity of a field-grown Gloeotrichia was satura- bated in test tubes under 'an atmosphere enriched ted at about 10 klux. At 5 klux, 87% of the maxi- with I5N. The data (Table 2) show that photo- mum NFA was obtained, suggesting that NFA under 5 trophic NFA was dominant in this soil and that the klux is similar to that under light-saturated con- addition of nitrogen fertilizer remarkably de- ditions (IRRI 1976). pressed the amount of nitrogen fixed. Soils with- out nitrogen fertilizer had an estimated NFA cor- 'The possibility of a detrimental effect of algae responding to 30 kg N/ha per month when exposed to on direct-seeded rice was reported in 1964 (IRRI light (IRRI [1968]). It has been reported, how- [196'5]). Algae may form a soil surface mat, which ever, that small-scale experiments favor the germinating rice seedlings have difficulty pene- growth of BGA and may largely overestimate photo- trating. Seedlings that do penetrate the mat are trophic NFA (Roger and Kulasooriya 1980). often uprooted when the mat breaks free from the soil and floats to the surface of the water. The acetylene reduction assay. (ARA) was used to estimate NFA in the floodwater during the 1970 wet Among the algae that are detrimental to rice, BGA season. ARA was generally higher in nonplanted can be considered incidental. The most harmful ge- fields than in planted fields (Fig. 2). This was nera are the filamentous or reticulated colonial probably due to the shading effect of rice plant types, most frequently Chlorophyceae (Roger and when incident light intensities were low because Kulasooriya 1980). Copper sulfate is the most com- of the cloudy weather of the wet season. This hy- monly used algicide. A search for more effective pothesis was supported by measurement of ARA of IRPS NO.7 8, June 1982 7 the floodwater during the 1971 dry season which &ne formed (nmol/ml per hour) showed a higher activity than during the wet sea- son (Fig. 3) and no significant difference between the planted and nonplanted fields (IRRI 1972). Diurnal variations of phototrophic NFA were stud- ied in 1973. A bell-shaped curve was observed in a soil inoculated with Aulosira fertilissima 3 weeks before measurement (Fig. 4a). Measurement of the & activity of laboratory-grown fertilissima and Anabaena dpiroides placed in the field showed a dissimetrical curve with a maximal activity in the 2 morning and a negligible activity in the afternoon (Fig. 4b). Such curves are characteristic of algal cultures not adapted to high light intensities and placed under direct sunlight (Roger and Kulasoo- riya 1980). I Ethylene formed (nmol/ml per hour) 2.0 I I C 7 Weeks after transplanting Fig. 3. Nitrogen-fixing activity, measured by C2H2-C2H4 assay, in floodwater of planted IR20 and non- planted fields, 1971 dry-season (IRRI 1972). Aulosira hrhhwm Weeks after transplanting 40 Fig. 2. Nitrogen-fixing activity, measured by C2H2-C2H4 assay, in floodwater of planted IR20 and non- planted fields, 1970 wet season (IRRI 1971). ARA assays were made in 1974 in rice fields at two sites in Albay province, Philippines-.- The sites Time represented 2 distinct soil types Puro clay Fig. 4. Diurnal changes of acetylene reduction in loam (PH:. 5.2, O.M: 3.08%, total N: 0.19%, avail- a field; a, acetylene reduction in an Aulosira able P: 11.5 ppm, no nitrogen fertilizer applied) fertilissima-treated plot planted to IR20 (activi- and Santo Domingo loamy sand (pH: 6.7, 0.M: 2.6%, ty detelfmined 3 weeks after inoculation); b, a,cety- total N: 0.14%, available P: 11.7 ppm, nitrogen lene reduction by laboratory-grown N2-fixing algae fertilizer applied). A peak of nitrogenase activi- . placed in the field (IRRI 1974) ties equivalent to 48 g N2/ha per hour for the Puro soils and 6.4 g N2/ha per hour for the San- to Domingo soils were obtained at 44 and 41 days. after transplanting, respectively. Thereafter the In situ ARA measurement in long-term fertility activity decreased until harvest. The absence of trials of IRRI (IRRI 1976) showed that photosyn- NFA in samples covered with black cloth suggested thetic NFA was important in nonfertilized plots, that BGA were the principal agent of nitrogen fix- particularly during the later stages of rice ation in these soils. The lower rate of NFA in the growth where values corresponding to O. 1-0.2 .kg N less acidic soil (Santo Domingo) was probably due fixed/ha per day were recorded. Continuation of ' to an inhibitory effect of nitrogen fertilizers. the measurements during the dry season (IRRI 1977) The estimated amount of nitrogen fixation by BGA indicated that phototrophic NFA was higher in the ranged from 18.5. to 33.3 kg N/ha per crop season dry season than in the wet season (Fig. 5). At the for Puro soils and from 2.3 to 5.7 kg N/ha per end of the dry season, when algal biomass was high, cropping season for Santo Domingo soils (IRRI total in situ NFA was 5,700 nmol C2Nq/m2 per day. 1975). This activity decreased to 660 nmol C2Hq/m2 per 8 IRPS No. 78, June 1982 day when floodwater and BGA were removed and re- In the field, frames (1 m x 1 m) were installed and placed by deionized water. This indicated that pho- 3 treatments -- without black cloth cover and insec- todependent NFA in floodwater was the major compo- ticide, without black cloth but with insecticide, and nent: In both seasons, the removal of algae greatly with insecticide and black cloth -- were compared. reduced NFA (Fig. 6). By extrapolating measurements, Weeding was manual with the least disturbance of the daily algal ARA was estimated to be 200 mmol C2H4/ surface soil. The plots that received insecticide m2 in the wet season (163 days) and 300 mmol C2H4/ and were exposed to light had a statistically signif- m2 during the dry season (168 days). Daily ARA icant increase of total nitrogen content of the sur- associated with the rice plant was 90 mmol C2H4/m2 face.3 mm of soil, amounting to 7 kg N/ha. This ac- in the wet season (IR26) and 50 mmol C2H4/m2 in the cumulation represents the residue of newly fixed dry season (IR36) (IRRI 1978). The relative con- nitrogen, which had not undergone downward movement, tribution of BGA, photosynthetic bacteria, and he- plant uptake, or loss (IRRI 1981). terotrophic bacteria was assessed using propanil, a herbicide that acts as a potent inhibitor of BGA. Unfortunately, propanil very significantly enhanced the population of N2-fixing purple sulfur bacteria. c Both photosynthetic bacteria and BGA had a substan- tially higher activity than rhizosphere microflora. 100 The results also suggested that the potential con- - tribution of photosynthetic bacteria may even ap- 50 proach that of active BGA (IRRI 1979). O A nitrogen balance technique was also used to eval- B75 wet seam TR 1976 dry HseDo son W uate phototrophic nitrogen fixation. Total nitro- gen content of the soil was measuied after two Algol wt (gkhamberl crops in a pot experiment (IRRI 119681). A larger amount of nitrogen was measured in the soils kept in the light than in those kept in the dark. E$ M fertilizer P$$ Nitrogen balance studies conducted in 1977 suggest- 1 NW NW NW (( ed an input of 37 to 50 kg N/ha per crop into flood- ed paddies. From greenhouse experiments it was Cp, (pmol/chamberdailyl I- I suggested that most of the net nitrogen fixation was due to the phototrophic microorganisms and that the rice plant stimulated NFA (IRRI 1978). An attempt was made to ascertain if there is an ac- cumulation of nitrogen in the surface (0-3 mm) layer 10 I I I I I I I I I I of flooded soil during a rice crop. Any accumula- J J A S O N D J F M A M tion detected would presumably represent nitrogen Fig. 5. Nitrogen-fixing (NF) activity and derived from phototrophic nitrogen fixation at soil algal mass in long-term fertilizer experi- surface. Greenhouse experiments proved the feasi- ments at IRRI, 1975-76. NW = no water, ND = bility of.this principle. Covering pots with black no determination, TR = transplanting, HD = cloth eliminated the surface accumulation of nitrogen. heading, HV = harvest (IRRI 1977). Table 2. Nitrogen fixation in various soil treatments (IRRI Lï96gÏ)). Incubated in light Incubated in dark a/ Pretreatment- Aton % excess Estimated amount of Atom % excess Estimated amount of b/- fixed nitrogen 15&l fixed nitrogen (kg/ha per month) (kg/ha per month) CI P a , no plant-;, .395 32. O .O38 3.2 PKD, no p- lant-(-I I .493 32.8 .o20 1.2 PKL IR8 .459 32.6 .O26 1.9 PKD IR8 .514 36.0 .O06 0.5 PKL Peta .517 38.1 .O47 3.4 PKD Peta .410 27.2 .O27 1.8 NPKL No plant .O43 5.7 ..O 48 4.6 NPKD, No plant .o20 2.0 O00 0.0 NPKZ, IR8 .O75 6.4 .O17 1.5 , NPKD IR8 .O70 5.1 .O24 1.8 NPKL Peta .151 13.3 ..O 31 2.5 NPKD Peta .O89 5.6 O00 0.0 alTreatment of 'Maahas clay soil in greenhouse pots from which soil samples were obtained 2 months after transplanting IR8 and Peta. b/Determined after incubati of soil in glass tubes for 1 month in the green- house. CIL = light (good, growth of indigenous algae). "7D = dark (no algal growth observed),. IRPS No. 78, June 1982 9 Epiphytic BGA and epiphytic phototrophic NFA on wetland rice, aquatic weeds, and deepwater rice. Observations in 1976 on deepwater rice roots indi- cated a significant NFA CARÄ) on aquatic roots. Specific ARA ranged from 37 to 260 nmol C2H+/h per g fresh weight and increased with the d&ameter (age) of the roots. A preferential colonizztion around the axils of the emerging lateral root by 6- +ALG4€ Nostoc sp. was observed (IRRI 1977). 5- Detailed studies on epiphytism were initiated in 4- 1979. Visual and microscopic observations, algal and bacterial counts, and ARA assays were made on 3- shallow-water rice, submerged weeds, and deepwater rice to study epiphytic growth of BGA. Nature of algal epiphytism. Comparing the three different hosts. it was found that eD_iDh.Yt-iSm and the associated ARA on wetland rice at seedling, tillering, and heading stages, and on the sub- Jun Jul Aug Sep oct ' MV merged weed Chara was predominantly due to Gloeo- trichia colonies that were visible to the naked TR HD M RT PL eye. On the other hand, the epiphytic algae on 1976 wet season wetland rice at maturity, on the submerged weed Najas, and on deepwater rice could be observed only under the microscope and the dominant algae 3 were Nostoc, Calothrix, and Anabaena. On deepwater -ALGAE rice plants, Anabaena and Calothrix were found on 2- the aquatic roots and leaf $heaths; Nostoc colo- nies were frequently present at the points of la- teral branches of roots. A unique finding in these studies was that BGA al- so exist inside the cavities of senescent leaf sheaths. Algal growth inside the cavities of dead or senescent leaf sheaths was also found in a Thailand deepwater rice area, not only on deepwa- ter rice but also on submerged grasses. This "en- dophytism", however, in addition to being not con- fined to rice, was not present in living, healthy tiss ues . A frequent observation was that older parts of' the =- hosts supported more numerous epiphytic BGA. Older parts appear to have more rough surface. Also -ra supported more epiphytic algae than Najas, pro- I Dec I' Jan ', Feb I' Mar I' Apr 'I May I bably because,of the rough corticated surface of TR IR36 Hq HV pr TL Chara leaves. Even old, rough nylon strings, but TR IR26 HD HV not new, smooth ones, supported epiphytic algal PL growth. 1977 dry season Fig. 6. Acetylene reduction rate in rice plots From these facts, it was concluded that both epi- at IRRI, with and without algae. TR = trans- phytism and "endophytism" are possibly related to planting, HD = heading, HV = harvesting, RT = abiotic effects, of which a mechanical effect in ratoon, PL = plowing (IRRI 1978). relation to the roughness of the host surface ap- pears to be of major importance. . . Epiphytic nitrogen-fixing activity. Rates of light ARA on wetland rice gradually diminish from seed- Blue-green algae and the rice plant ling to maturity, mainly due to the concomitant decrease of Gloeotrichia epiphytism and the reduc- Compared with the bacteria-plant association, the tion of available light (Table 3). In deepwater BGA-plant association has been less studied. Ob- rice, also, there was a decrease in specific ARA ' servations o? algal colonies on submerged portions (nmol/g fresh weight of host) in the light from of the rice shoot led us to study the possible as- heading to maturity, but this was compensated by sociation between BGA and rice. A second topic has an increase in the host biomass, so,t hat a con- been the availability to the rice plant of nitro- stant ,activity per plant was observed at both gen fixed by BGA. ,:. these stages. 10 IRPS No. 78, June 1982 Table 3. Epiphytic ARA on rice and weeds (Kulasooriya et al 1980). Dominant nmol/g fresh weight per hour pmol/mZ per hour Sample nitrogen-fixing Light Dark (total) b/ Rice at seedling stage 614.0 n.d.- 5i.n- Gloeo trichia Rice at tillering stage 37.5 10.1 15.0 Nostoc Rice at heading stage 1.9 O. 67 1.2 Calo thrix Rice at matuFity 1.1 1.1 2.5 Deepwater rice at he,ading Nostoc 15.9 4.9 125 Cala thri x Deepwater rice at maturity 9.7 1. ,o 125 Chara Gloeotrichia 36.6 3.3 8.g’ I Nostoc Calothrix Najas Nostoc 26.7 1.8 Mo no Cho r ia Calot hrix 1.8 1.3 Cyperus 4.4 2.5 J S’For a weed biomass of 2 t/ha. b/n.d. = no data. Table 4. The I5N - I4N ratio in deepwater and shallow-water rice (IRRL 1981). Total N 1” 15N/14N ratio Site Water Light (Bx depth exposure (mg/po t> (mglpot) x 100) A B Bangkhen Shallow + 718 111 15 u, + Huntra Semideep 979 86 9 + Bangkok Deep 952 97 10 + IRRI Shallow 502 194 39 Shalloy - 5 04 230 46 + Deep 737 20 1 27.2 Deep - 5 30 1.14 21.5 The results indicated that the contribution by epi- the soil surface. In,deep water, the submerged por- phytic N2-fixing microorganisms on wetland rice tion of some plants was covered with black cloth to fields is low but the epiphytic BGA play an impor- prevent epiph tic algal growth on the rice plant. tant role in inoculum conservation, from which the Total N and uptake to the aboveground portions regeneration of periodic algal blooms can take are shown in Table 4. The 15N/I4N ratio was always place. Epiphytic nitrogen fixation on deepwater lower in shallow water, indicating that in deep rice, on. the other hand, makes a substantial con- water more nitrogen from a source other than soil tribution to this ecosystem (10-20 kg N/ha), main- was absorbed by the plant than in shallow water. ly due to the greater biomass available for colo- If the nitrogen in the plant comes partly from biol- nization by epiphytic algae. , ogical nitrogen fixation by epiphytic EGA, the deep- water rice, of which the submerged portion was cov- Role of epiphytic BGA in the nitrogen nutrition of ered with black cloth, must have a lower l5N/I4N deepwater rice. To evaluate the importance of epi- ratio. The ratio of measurements on the whole phytic nitrogen fixation in deepwater rice, and the plants was not in agreement with that prediction. availability of fixed nitrogen for the plant, I5N No significant difference was observed, perhaps be- experiments were conducted at IRRI and in Thailand. cause of an abnormal growth of the deepwater rice plant when covered with black cloth. But the compa- In one trial (IRRI 1981) I5N dilution technique was rison of 15N/I4N ratio of various parts of deep- used in the field. Pots containing soil with 15N water rice was indicative. The least enrichment of ammonium salt were placed in shallow water (5 cm) I5N in the deepwater rice which had submerged por- and deep water. Some pots in shallow water were tions exposed’to light was found in the aquatic covered with black cloth to prevent algal growth on roots and submerged leaf sheaths. These portions