29.

2/2004

December 2004

International Rice Research Notes

International Rice Research Institute IRRI home page: http://www.irri.org Riceweb: http://www.riceweb.org Riceworld: http://www.riceworld.org IRRI Library: http://ricelib.irri.cgiar.org IRRN: http://www.irri.org/irrn.htm

Copyright International Rice Research Institute 2004

NOTICE: This journal is copyrighted in the name of the International Rice Research Institute (IRRI). IRRI has the exclusive right to reproduce or authorize reproduction of the copyrighted work, to prepare derivative works based upon the copyrighted work, and to distribute copies of the copyrighted work to the public for sale or other transfer of ownership. You may copy, duplicate, or otherwise reproduce any of the articles or portions of the articles in the copyrighted work; but you must acknowledge the journal and its copyright owner as the source of the copyrighted work. You may not modify, translate, or use the data herein contained to prepare a derivative work without the prior written consent of the copyright owner.

Contents
IRRN BEST ARTICLE AWARDS
6 Wedding rice not thrown but sown

Cover photos: A. Javellana and Dr. Kam

Plant breeding

10 Molecular strategies to use nuclear male sterility in

plant hybrid breeding Xinqi Li , Longping Yuan, Jinhua Xiao, and S. McCouch

14 Development of TGMS lines and two-line rice

hybrids through a shuttle breeding program between IRRI and China Tongmin Mou, Chunhai Li, Junying Xu, S.S. Virmani, and D.L. Sanchez genes to nontarget insects K. Bashir, T. Husnain, and S. Riazuddin

12 HPLC profiles of polyamines in the leaves and

anthers of a new thermosensitive genic male sterile rice line D. Vijayalakshmi and U. Bangarusamy

15 Response of transgenic rice expressing two Bt

Genetic resources

17 Giria high-yielding, slender-grain, sheath blight

and bacterial blightresistant variety for shallowwater conditions in West Bengal S.K. Sinha, S.N. Sen, and A. Biswas

22 Two outstanding rice varieties developed through

selection from naturally occurring genetic variation in Sri Lanka D. Sumith de Z. Abeysiriwardena

18 Varietal turnover and seed exchange: implica-

tions for conservation of rice genetic diversity on-farm A. Sirabanchongkran, N. Yimyam, W. Boonma, K. Rerkasem, K. Coffey, M. Pinedo-Vasquez, and C. Padoch

24 Dorfak: an aromatic, high-yielding, short-duration

variety with good cooking quality for the irrigated lowlands of Iran M. Nahvi, M. Allahgholipour, A. Jauhar Ali, M.S. Mohammed Salehi, H. Rahim Saroush, H. Dorosti, A. Erfani, F. Padasht, and F. Alinia

20 Invasion of weedy rice in rice fields in Thailand:

problems and management C. Maneechote, S. Jamjod, and B. Rerkasem

December 2004

Pest science & management

26 Long-term effects of fertilizers and herbicides

on a nematode population under a rice-wheatcowpea system A.P. Singh and Nand Ram

40 Neural network approximation of sampling yieldeffort curves of rice invertebrates Wenjun Zhang, Yanhong Qi, and A.T. Barrion

28 A method for transforming rice stinkbug counts

in rice Mu Mu Thein, V.S. Singh, S. Chander, and N. Kalra

43 Between-habitat movement of rice arthropods and

its ecological role Wenjun Zhang and Yanhong Qi

30 Weed control in direct, dry-seeded rice in

India: comparison of seedbed preparation and use of pendimethalin S.K. Sharma, D.K. Pandey, K.S. Ganagwar, and O.K. Tomar

45 A generator of a web map of biological interactions

for rice invertebrates W.J. Zhang, H.X. Wu, and Y.H. Qi

32 A Web-based software for randomization tests of

cluster analysis of invertebrate biodiversity in a rice ecosystem Wenjun Zhang, Runjie Zhang, and Dexiang Gu

47 Pathogenic variability of rice blast isolates in Nepal

B. Chaudhary, S.M. Shrestha, and R.C. Sharma

34 Detection of Xanthomonas oryzae pv. oryzae by

NCM-ELISA in naturally infected rice plants Y. Suryadi and T.S. Kadir

49 Brown planthopper injury vs mechanical injury

to the rice plant E. Rubia-Sanchez, T. Watanabe, and Y. Suzuki

36 Detrimental effects of niclosamide 250EC at

preseeding in direct-seeded rice culture R.C. Joshi, M.S. Desamito, A.R. Martin, L.S. Sebastian, and J.B. Coupland

51 Ecology of jungle rice (Echinochloa colonum), a

weed of the rice agroecosystem: a case study in Bilaspur (Chhattishgarh) V. Dubey

38 Predatory behavior of mirid bug, Cyrtorhinus

lividipennis, on rice plants with different nitrogen regimes Z.X. Lu, X.P. Yu, and C. Hu

Soil, nutrient, & water management

56 Impact of integrated plant nutrient management on upland rainfed rice cultivation H. K. Senapati, S. S. Sahoo, B. Jena, and D. Sahoo

60 Effects of long-term application of manure and

fertilizer on the upland rainfed rice-cropping system H.K. Senapati and P.C. Senapati

57 Impact of organic manure and inorganic

phosphatic fertilizer on yield and nutrient uptake in a rice-rice cropping system R. K. Kaleeswari and S. Subramanian

61 Effect of silicon sources and fertility levels on

transplanted rice P.C. Sudhakar, J.P. Singh, and Kalyan Singh

IRRN 29.2

63 Fertilizer optimization in rice by eliminating

hidden nutrient deficiencies M.R. Latha and V. Murugappan

67 Nitrogen fertilizer increases protein and reduces

breakage of rice cultivar Chainat 1 M. Leesawatwong, S. Jamjod, B. Rerkasem, J. Kuo, and B. Dell

65 Crop yield, disease incidence, and insect pest

attack in relation to N dynamics in rice R.S. Rekhi, R. Singh, R.K. Goel, and J. Singh

69 Participatory irrigation management for efficient

water use and enhanced rice productivity in Tamil Nadu, India B. Chandrasekaran, A. Nandhagopal, K. Palanisami, S.D. Sirakumar, V. Balasubramanian, and R. Rajendran

Crop management & physiology

71 Effect of low light on yield and physiological

attributes of rice M. Lakshmi Praba, M. Vanangamudi, and V. Thandapani

78 A simple screening technique for salinity tolerance

in rice: germination rate under stress D. Sumith de Z. Abeysiriwardena

73 Nursery technology for early production of robust

rice seedlings to transplant under integrated crop management R. Rajendran ,V. Balasubramanian, V. Ravi, K. Valliappan, T. Jayaraj, and S. Ramanathan

80 Osmohardening: a new technique for rice seed

invigoration S.M.A. Basra, M. Farooq, K. Hafeez, and N. Ahmad

75 Influence of high- and low-temperature

treatments on seed germination and seedling vigor of coarse and fine rice M. Farooq, S.M.A. Basra, K. Hafeez, and E.A. Warriach

Socioeconomics

82 Status of rice seed production in Sri Lanka:

some bottlenecks M. Wijeratna and W.N. De Silva

85 Farmers participatory diagnosis of the flood-

prone deepwater rice-cropping system in eastern India U.P. Singh women in storing grains P. Sumathi and M.N. Budhar

83 Farmers participatory evaluation of salinetolerant rice varieties L. Ponnusamy and N. Thenmathi

86 Postharvest technology of rice: role of farm

84 Deepwater agroecosystem characterization

and system diagnosis of farmers problems in Bihar, India U.P. Singh

89 Extension service and rice yield gap in Nigeria,

1980-2002 O.I. Oladele

December 2004

Agricultural engineeering

91 Subsurface drainage to increase rice productivity

in a saline environment S.K. Gupta, T.V. Satyanarayana, M.V. Manjunatha, and V.B. Kuligod

94 NOTES FROM THE FIELD

98 INSTRUCTIONS TO CONTRIBUTORS

Editorial Board Jagdish K. Ladha, Editor-in-Chief Gary Jahn (pest science and management) Zhikang Li (plant breeding; molecular and cell biology, China) Swapan Datta (plant breeding; molecular and cell biology, Los Baos) Sushil Pandey (socioeconomics; agricultural engineering) Renee Lafitte (crop management and physiology) Abdelbagi Ismail (soil, nutrient, and water management; environment) Edwin Javier (genetic resources)

Production Team Tess Rola, managing editor Editorial Bill Hardy Graphic design Emmanuel Panisales Diane Martinez Editorial assistance Diane Martinez

IRRN 29.2

IRRN BEST ARTICLE AWARDS

Wedding rice not thrown but sown

(Left to right) Anothai Sirabanchongkran, Mou Tongmin, G. Nageswara Rao, Abdul Rashid, Ashutosh Pathak, and Ramani Kumar Sarkar

arrying well is what some ethnic Karen farmers in the northern Thai village of Tee Cha do in an unexpected way. According to an awardwinning scientific paper, families with marriage ties outside of the village grow rice crops that are better because they are more genetically diverse. This finding points toward the importance of kinship as a pathway for seed exchange, wrote the authors of Varietal turnover and seed exchange: implications for conservation of rice genetic diversity on-farm.Not only are marriage relations important routes of exchange, but many norms of exchange are deeply embedded in cultural practice. For example, Tee Cha villagers cite a rule that requires children to maintain a specific family variety after their parents death. These findings are important because conservation initiatives that attempt to increase varietal exchange without under-standing local practices might actually undermine the local systems of exchange that are crucial to the maintenance of diversity. The paper won the IRRN Best Article Award in the Genetic Resources category, one of seven awards with which International Rice Research Notes, now in its 29th year of publication by the International Rice Research Institute (IRRI), is marking International Year of Rice 2004. Its first author, Anothai Sirabanchongkran, a researcher in the Faculty of Agriculture at Chiang Mai University in Thailand, will receive US$500, as will the first authors of the other winning papers. Papers that have not yet been published in IRRN will appear in this issue (see box for previously published papers).
December 2004

Many on-farm conservation projects focus on encouraging continued planting of local varieties, and much effort is placed on stopping farmers from abandoning these varieties, stated the Thai and U.S. collaborators, who were funded by the Collaborative Crop Research Program of the McKnight Foundation. But it was found that the rate of abandonment or variety turnover was high, even when the number of local varieties grown in the village remained stable. One of the projects central findings is that in areas outside Thailands main riceproduction regionstotaling some one-fifth of the countrys rice landrice genetic diversity and the success of rice farming are closely linked, said Anothai. Genetic variation in the local rice germplasm is not static, but is continuously being renewed and

enhanced by farmers seed management, with seed exchanges among farmers within and between villages identified as one key process. This has led us to investigate more closely the dynamics of varietal turnover, seed exchange networks, and the relevant social processes. Shuttle breeding Winning the award in the Molecular and Cell Biology category was Development of TGMS lines and two-line rice hybrids through a shuttle breeding program between IRRI and China. First author Mou Tongmin, professor of plant breeding at the Plant Science and Technology College of Huazhong Agricultural University in Wuhan, has been researching genetics and hybrid rice breeding for 17 years, including a stint in 2002 as a research fellow at IRRI under tropical hybrid pioneer Sant Virmani.

IRRN Best Article Award winners circle

Genetic Resources: Varietal turnover and seed exchange: implications for conservation of rice genetic diversity on-farm. A. Sirabanchongkran, N. Yimyam, W. Boonma, and K. Rerkasem, Faculty of Agriculture, Chiang Mai University, Thailand; K. Coffey and M. Pinedo-Vasquez, Department of Ecology, Evolution and Environmental Biology, Columbia University, USA; and C. Padoch, Institute of Economic Botany, New York Botanical Garden, USA (this issue p 18-20) Molecular and Cell Biology: Development of TGMS lines and two-line rice hybrids through a shuttle breeding program between IRRI and China. Tongmin Mou, Chunhai Li, and Junying Xu, National Key Laboratory for Crop Genetic Improvement, Huazhong Agricultural University, China; S.S. Virmani and D.L. Sanchez, IRRI (this issue p 14-15) Socioeconomics: Advance estimation of rice production in India from weather indices. G. Nageswara Rao, Y.S. Ramakrishna, A.V.R. Kesava Rao, and G.G.S.N. Rao, Central Research Institute for Dryland Agriculture, Hyderabad, India (Dec 2003) Soil, Nutrient, and Water Management: Boron deficiency in calcareous soils reduces paddy yield and impairs grain quality.
IRRN 29.2

A. Rashid and M. Yasin, Land Resources Research Program, National Agricultural Research Center (NARC); M. Ashraf, Crop Sciences Division, Pakistan Agricultural Research Council; and R.A. Mann, Rice Research Program, NARC, Islamabad, Pakistan (June 2004) Pest Science and Management: Pseudomonas strain GRP3 induces systemic resistance against sheath blight in rice caused by Rhizoctonia solani. A. Pathak, A. Sharma, and B.N. Johri, Department of Microbiology, College of Basic Sciences and Humanities; and A.K. Sharma, Department of Agro-forestry, College of Agricultural Sciences, G.B. Pant University of Agriculture and Technology, Pantnagar, India (June 2004) Crop Management and Physiology: Chlorophyll fluorescence parameters as indicators of submergence tolerance in rice. R.K. Sarkar, D. Panda, D.N. Rao and S.G. Sharma, Central Rice Research Institute, Cuttack, India (June 2004) Plant Breeding: Santosha high-yielding variety for rainfed lowland developed through participatory breeding for Bihar, India. R. Thakur, N.K. Singh, S.B. Mishra, A.K. Singh and K.K. Singh, Rajendra Agricultural University, Samastipur; and R.K. Singh, IRRI-India Office, New Delhi, India (Dec 2003) 7

His research team developed the first indica thermosensitive genic male sterile line and the first japonica photoperiod-sensitive genic male sterile line. The resulting japonica two-line hybrid rice was introduced in the Chinese province of Hubei in 1990 and now covers a planted area of 33,000 ha. In 2002-03, the resulting indica two-line hybrid was certified for use in Hubei, where it now covers 10,000 ha. Indian meteorologist G. Nageswara Rao was the first author of the Socioeconomics category winner, Advance estimation of rice production in India from weather indices. Dr. Rao, who currently works under the United Nations Development Programme as an assistant professor and head of the Meteorology Department at Arba Minch University in Ethiopia, collaborated on the paper with former colleagues at the Central Research Institute for Dryland Agriculture in Hyderabad, India. The paper presents a model that estimates, using monthly rainfall data and the southern oscillation index (indicating El NioLa Nia fluctuations), Indias total annual rice production 6-8 months before the harvest. These advance estimates are very useful for the national government and planners as they make policy decisions about food security and the export and import of rice, explained Dr. Rao. The author of several publications on various aspects of meteorology, Dr. Rao received the Young Scientist Award from the Indian governments Department of Science and Technology for his work on long-range forecasting of monsoon rainfall and water resources in the Godavari Basin. Another widely published scientist will receive the $500 award for the paper Boron deficiency in calcareous soils reduces paddy yield and impairs grain quality, which won in the Soil, Nutrient, and Water Management category. First author Abdul Rashids 200 publications include the only soil science book produced in Pakistan. Internationally recognized for his expertise in micronutrient nutrition of crops, the chief soil scientist at Pakistans National Agricultural Research Center is on the editorial board of the European Journal of Agronomy and has won several coveted awards and honors.
8

Multidisciplinary research As rice soils in Pakistan are calcareous [high in calcium carbonate, or lime] and low in organic matter, we suspected that boron deficiency might be a cause of low rice yields, Dr. Rashid explained. Our research established that rice grown in the low-boron, calcareous soils of the major rice-growing areas of Pakistan not only suffers severe yield losses, but its grain quality is also impaired. We found that applying boron in such situations improves yield, cooking quality, and farmers incomesubstantially. This significant accomplishment was made possible by a multidisciplinary research team including soil fertility specialists, a plant physiologist, and an agronomist. The award for Pest Science and Management went to the paper Pseudomonas strain GRP3 induces systemic resistance against sheath blight in rice caused by Rhizoctonia solani. First author Ashutosh Pathak is a microbiologist now working for a water purification firm in the Indian state of Uttaranchal while serving as a faculty member and adviser in food technology at Allahabad University. This study started in 2000, when Prof. Johri had gone through all the details of sheath blight occurrence and losses due to this disease, recalled Dr. Pathak, referring to Bhavdish N. Johri, the coauthor under whom he received his doctorate from G.B. Pant University of Agriculture and Technology in Pantnagar, Uttaranchal. He assigned to me this work in integrated pest management. Dr. Pathak and his colleagues at G.B. Pant screened several bacterial cultures and found that the Pseudomonas fluorescence strain GRP3 exhibited both biocontrol and plant growthpromoting rhizobacteria properties. We carried this study out on susceptible and moderately resistant varieties, he explained. Susceptible varieties shifted to the moderately resistant or resistant group of rice cultivars, and moderately resistant varieties became resistant. The prize in the Crop Management and Physiology category went to a paper by a group funded by the Indian Council of Agricultural Research at the Central Rice Research Institute in Cuttack, Orissa.
December 2004

The group has been working on developing a suitable technique to screen rice varieties exhibiting submergence tolerance, explained Ramani Kumar Sarkar, senior scientist and first author of Chlorophyll fluorescence parameters as indicators of submergence tolerance in rice. The traditional technique of submerging rice plants and then looking for survivors has helped in identifying submergence tolerance, but it results in the loss of valuable materials during screening. So we tried a nondestructive technique based

on chlorophyll fluorescence parameters to differentiate between tolerant and susceptible genotypes. Finally, the winner in the Plant Breeding category was Santosha high-yielding variety for rainfed lowland developed through participatory breeding for Bihar, India. The first author of the paper was R. Thakur, who collaborated with four of his colleagues at Rajendra Agricultural University, in the Indian state of Bihar, and with R.K. Singh, the recently retired IRRI liaison scientist for India.
(Source: Rice Today,October-December 2004)

Rice is

Life
9

IRRN 29.2

Plant breeding

Molecular strategies to use nuclear male sterility in plant hybrid breeding

Xinqi Li and Longping Yuan, China National Hybrid Rice Research and Development Center, Changsha 410125, China; Jinhua Xiao and S. McCouch, Department of Plant Breeding, Cornell University, Ithaca, NY 14853, USA

The exploitation of plant heterosis is an effective approach to increase food production. Heterotic hybrid varieties in major crops such as rice, wheat, and cotton can show more than a 20% yield advantage over the best conventional ones under the same cultivation conditions. Difficulties in breeding elite male sterile lines and producing commercial hybrid seed hamper the development of hybrid crops. Nuclear male sterility in plants includes both spontaneous and engineered sterility. Spontaneous nuclear male sterile mutation is a common phenomenon in nature. The frequency of such mutations is high; it is easily induced from ionizing radiation and chemical mutagents. The most spontaneous nuclear male sterility is usually controlled by a pair of recessive genes (msms/MFMF msMF; msms/ MFms 1 msms:1 MFms). The major advantages of using such a muta-tion are (1) the single ms gene that expresses complete male sterility under all rice genetic backgrounds can be easily transferred into any varieties with desirable characteristics, (2) almost all varieties can restore nuclear male sterile lines, (3) no negative effects caused by male sterile cytoplasm and no unitary cytoplasm problems are
10

observed, and (4) the male sterility is stable in the environment, unlike photoperiodsensitive genic male sterile (PGMS) lines or thermosensitive genic male sterile (TGMS) lines. Thus, spontaneous genic male sterility can be the best genetic tool for hybrid seed production. The fact that nuclear male sterility in many crops does not permit effective production of population with 100% male sterile plants seriously limits its use in hybrid seed production. Cigan and Albertsen (1998) suggested the use of an inducible promoter to multiply ms lines: take the native promoter off the male fertility gene and replace it with a promoter that functions only in response to a chemical (such as antibiotic tetracycline) and transfer it into the male sterile line. The transgenic line should become fertile when the inducing chemical is present. For hybrid seed production, the chemical will not be applied. Three new strategies were suggested here to multiply male sterile lines by means of cloning and transformation of recessive male fertility genes of spontaneous genic male sterility or through artificial male fertility gene transformation.

Use of the MF gene through plastid transformation Plastid transformation offers a means of directing foreign DNA to a specific region of the genome, allowing incorporation of a specific region of the plasmid DNA only. In addition, the number of copies of the transgene is greatly amplified as a consequence of the multiple copies of plastome within the cell and a very high level of transgene expression can be achieved. The transgene can be contained within the plant by the lack of release through pollen, since, in most higher plants, plastids are transmitted maternally. When plastid transformation technology becomes a practical approach for gene transfer, it is hoped that the MF gene can be transferred into the plastid genome of male sterile plants and the expression of male sterility will be suppressed. The transgene should be male fertile then. Since the transgene will have the same nucleus as the male sterile plants, they can be used as the male parent to reproduce the nuclear male sterile lines when crossed with the male sterile plants. Thus, male sterile plants are reproduced (Fig. 1). As the transgene is used only to reproduce male sterile
December 2004

lines, the hybrid seeds in rice production will not belong to transgenic plants. Therefore, if this system works, it will be an ideal way to explore crop heterosis. Use of the MF gene combined with transiently active promo-ters Transiently active promoters are substantially active only during a well-defined phase of plant growth. The transiently active promoters being used in this project will be a promoter active in late embryogenesis (Fig. 2). We want to combine a lethal gene such as RNase or the ribosomal inhibitor protein gene to this promoter, then splice it with the male fertility restoration gene, and transfer them into male sterile plants. The expression of the lethal gene spliced with the restoration gene can be limited to the complete maturity stage. So, the transgene should be fertile at the heading stage and should be killed by the lethal gene at maturity. Any seeds with the lethal gene will be killed after maturation because of the transiently active promoters functioning (i.e., seeds with the male fertility gene will be killed). The other seeds without the male fertility gene will survive and have the same genetic background as the female parent (i.e., male sterile plants). Immature seeds with the lethal gene can survive before maturity. We can harvest the immature seeds as the male parent. The yield for the multiplication of male sterile lines is not very important. For example, the seed production field ratio for the multiplicaIRRN 29.2

ms plants msms msms MF in plastic ms line multiplication

ms line

msms

Hybrid seed production Hybrid seeds

Conventional lines MFMF ms: male sterile gene MF: male fertile gene

Fig. 1. Use of the MF gene through plastid transformation.

ms plants

Transgenic maintainers

Mature seeds die, immature seeds survive as maintainers

msms msms
P + lethal gene + MF

P + lethal gene + MF

msms

msms ms line

Fig. 2. Use of transiently active promoters (P: transiently active promoter active only in late embryogenesis, ms: male sterile gene, MF: male fertile gene).

ms plants Spray inducer

ms ms X X Y Y ms line

ms ms A---MF---A A---MF---A Y Y X X
No inducer

ms ms

X X Y Y

A---MF---A A---MF---A

Male fertile line

Fig. 3. Use of the recombinase gene and its repressor gene (A: lox; X: recombinase gene; Y: Xs repressor; ms: male sterile gene; MF: male fertile gene).

tion of the ms line to hybrid seed production to hybrid rice production is around 1:100:10,000. Use of the MF gene through the gene switch system The Cre/lox system can snip out a stretch of DNA in the

plant genome. The recombinase gene can be controlled by a repressor gene whose protein prevents the recombinase gene from being switched on. And, a stimulant, such as the antibiotic tetracycline, can block the repressor protein. This is the so11

called gene switch system (Fig. 3). In this study, the male fertility gene will be spliced into a DNA sequence to each end that can be recognized by Cre recombinase. This stretch of DNA and the gene switch system will be transferred into the male sterile (ms) lines. Thus, we can obtain a transgenic plant with a homozygous male fertile gene and male sterility gene and

the gene switch system. The transge-nic line should be fertile and should reproduce seeds like a conventional variety. When it is treated with tetracyline, the repressor protein will be blocked. Then, the recombinase gene can work and snip out the male fertility gene. The fertile transgenic line becomes male sterile and can be used as the female parent for

hybrid seed production. If this strategy is combined with the transiently active promoter system, the transgenic spontaneous nuclear male sterility line might be more useful and practical in hybrid seed production. Reference
Cigan AM, Albertsen MC. 1998. Reversible nuclear genetic system for male sterility in transgenic plants. USA Patent No. 6,399,856.

HPLC profiles of polyamines in the leaves and anthers of a new thermosensitive genic male sterile rice line
D. Vijayalakshmi, Plant Physiology Division, Sugarcane Breeding Institute, Coimbatore; and U. Bangarusamy, Department of Crop Physiology, Tamil Nadu Agricultural University, Coimbatore 641003, Tamil Nadu, India E-mail: viji tnau2003@yahoo.co.in

Hybrid rice is successfully cultivated in India and, in recent years, the country has released commercial rice hybrids having at least 1 t ha 1 higher yield than inbred checks. The cytoplasmic male sterility (CMS) system, also known as the three-line system, is presently the most widely used system for producing F 1 rice hybrids. Although effective, the system is cumbersome because CMS lines require specific maintainer and restorer lines and possess wild abortive (WA) cytoplasm that is vulnerable to biological stresses. On the other hand, the thermosensitive genic male sterility (TGMS) system does not require restorer lines, and the seeds of TGMS lines are multiplied by selfing when exposed to the right temperature at the critical growth stage (in the tropics, 515 d
12

after panicle initiation) (Maruyama et al 1990). But plants are constantly challenged by unfavorable fluctuations in their environment and the accumulation of lowmolecular-weight metabolites such as polyamines is a universal response to cope with various forms of environmental stresses such as extreme temperature (Vernon et al 1993). It is important to understand the physiological basis for fertility transformation in TGMS lines. An attempt has been made to study the role of polyamines in fertility transformation of the most stable TGMS line, TS29. The TS29 line, along with a known thermo-insensitive variety (CO 46), formed the basic material for the experiment. Plants were grown in plastic pots in a glasshouse. These were transferred to a growth chamber when they

reached stage III of panicle development (when panicles were 1 mm long and covered with hairs). The plants were kept inside the phytotron up to stage VII (filling phase of pollen, floret and panicles reach full length, and color turns green). The biochemical analyses were done immediately after stage VII. The temperature treatments, given in the phytotron to imitate the diurnal variation of the summer season to induce male sterility, are specified below. Sowing date was adjusted, such that TS29 and CO 46 reached the panicle initiation stage during the 1st week of December and January. The mean maximum and minimum temperatures for December and January were 28.6 and 19.9 C, respectively. The polyamines in the leaf and young panicles of fertile and sterile TS29 line at stage VII
December 2004

Time (h) 08001000 10001200 12001600 16000800

Temperature (C) 34 35 36 35

Relative Light humidity intensity (%) (M m2 s1) 80 70 60 60 600 800 800 600

HPLC profile of polyamine contents (mol g 1 ) in leaves and young panicles of TS29 and CO 46 at stage VII of panicle development. Treatment Leaves of fertile plant Leaves of sterile plant Leaves of CO 46 plant Panicles of fertile plant Panicles of sterile plant Panicles of CO 46 plant
a

Spermidine 2.13 nda 3.55 0.05 nd 0.68

Putre- Cadavarine scine nd 3.84 nd nd 0.03 nd 0.39 0.33 0.58 nd nd 0.04

of panicle development were analyzed by the benzoylation method as reported by Serrano et al (1995). The retention time for the standards putrescine, spermidine, and cadavarine was 2.18, 3.46, and 5.56 min, respectively, and the respective area covered was 9,345%, 11,104%, and 14,734%. Distinct differences were observed in polyamine contents of TS29 under sterile and fertile conditions (see table). The fertile plant leaves showed higher amounts of spermidine (2.13 mol g 1 ), followed by cadavarine (0.39 mol g 1 ). The polyamine putrescine was not detected. In sterile plant leaves, putrescine (3.84 mol g 1 ) was found to be the major polyamine, followed by cadavarine (0.33 mol g1). Spermidine was not

nd = not detected.

g1) in the leaves and panicles, respectively. It appeared that the presence of spermidine favored fertile pollens, while putrescine favored sterile pollens. Vernon et al (1993) indicated that plants tended to produce low-molecularweight metabolites in response to environmental stresses. Feng and Cao (1993) also observed increased putrescine content under sterility and increased spermidine content under fertility conditions in rice variety C407S. References
Feng JY, Cao DM. 1993. Changes in fertility and polyamines in young panicles of photosensitive nuclear sterile paddy rice. J. Nanjing Agric. Univ. 16(2):107-110. Maruyama K, Araki H, Amao E. 1990. Enhancement of outcrossing habits of rice plant of mutation breeding. Gamma Field Symp. 29:11-25. Serrano M, Martinez MC, Riqueime F, Romojaro F. 1995. Endogenous levels of polyamines and ABA in pepper fruits during growth and ripening. Physiol. Plant. 95:73-76. Vernon DM, Tarcyzynksi MC, Jensen RG, Bchnert HJ. 1993. Polyamines in plants. Plant J. 4:199-205.

detected. The amounts of polyamines present in the panicles were less than those in the leaves. In the fertile panicle, only spermidine (0.05 mol g 1 ) was detected, whereas, in the sterile panicle, putrescine (0.03 mol g 1) was detected. The leaves and panicles of CO 46 showed a similar trend as seen in the TS29 fertile plant, with slightly higher contents of spermidine (3.55 mol g 1 and 0.68 mol g 1) and cadaverine (0.58 mol g 1 and 0.04 mol

IRRN 29.2

13

Development of TGMS lines and two-line rice hybrids through a shuttle breeding program between IRRI and China
Tongmin Mou, Chunhai Li, and Junying Xu, The National Key Lab for Crop Genetic Improvement, Huazhong Agricultural University (HAU), Wuhan 430070, China; and S.S. Virmani and D.L. Sanchez, IRRI E-mail: tongmin@public.wh.hb.cn

Based on the discovery of photoperiod-sensitive genic male sterile (PGMS) rice mutant Nongken 58S (Shi 1981), rice scientists in China began work on the two-line system of hybrid rice breeding in the early 1980s. Subsequently, some thermosensitive genic male sterile (TGMS) rice mutants, such as Annong S-1 and 5460S, were developed in 1988 through screening of breeding materials and radiation, respectively (Lu et al 1998). The two-line system was established in China by using PGMS and TGMS lines. This method was considered more efficient than the threeline system in exploiting heterosis to increase yield. More than 10 environment-sensitive genic male sterile lines have been used to breed more than 32 commercial two-line hybrids in China. The area planted to two-line hybrids, which gave 5% higher yield than three-line hybrids, was 2 million ha in 2002. In the tropics, TGMS lines may be more useful than PGMS lines because of the significant temperature variation among seasons. A shuttle breeding program between China and IRRI was conducted in the past 5 years to develop and evaluate TGMS lines and two-line rice hybrids. Several useful TGMS lines have so far been developed.
14

Three TGMS lines TM104S and TM105S from HAU and IR73827-23S from IRRIwere evaluated in the IRRI phytotron. IR73827-23S was also evaluated under natural conditions at Wuhan, China. The results (Table 1) showed that IR73827-23S was completely sterile at 24.9 o C (mean) and partially sterile at 23.9 o C or less. This implies that the critical sterility point (CSP) of IR73827-23S was between 24 and 25 oC in the phytotron for 12 d. TM104S was almost sterile at 24.9 o C and partially sterile at 23.9 oC. Its CSP may be at around 25 o C (mean). TM105S was completely sterile under 23.9 o C and 24.9 oC and partially fertile at 22.9 oC or less. The CSP of TM105S may be between 23 and 24 oC (mean) in the phytotron. The pollen fertilities of IR73827-23S and TM105S at 21.9 oC were lower than those at 22.9 oC because their tolerance for lower temperature was less physiologically. The results of field evaluations at Wuhan showed that IR7382723S was completely sterile when the daily mean temperature was higher than 23.5 o C, almost sterile from 22.0 to 23.5 oC for shorter than 3 d in the sensitive stage, and partially fertile at temperatures lower than 23.5 oC for longer than 4 d (see figure). This means that the CSP under

natural conditions was 23.5 oC in 3 d. The stage sensitive to temperature ranged from 11 to 5 d before heading. IR73827-23S was crosstested by more than 50 conventional inbred lines or restorers of three-line hybrid rice. The combinations were evaluated in observed yield trial nurseries (2-m 2 area per plot) in Hainan and Wuhan in 2003. Several heterotic combinations were identified (Table 2). The yields of three combinations reached 11.67 12.26 t ha 1 , surpassing the three-line control by 10.8 16.4% in Hainan. Yields of six combinations were 7.509.33 t ha 1, 8.535.1% over that of the three-line control in Wuhan, Hubei. The yields in Hainan were higher than those in Wuhan because of the higher fertilization level in Hainan.
Table 1. Fertility expression of three TGMS lines in the phytotron for 12 d, IRRI, 2002. Time 0500-0800 0800-1100 1100-1500 1500-1900 1900-0100 0100-0500 Mean IR73827-23S TM104S TM105S 21 23 25 23 21 19 21.9 Temperature (oC) 22 23 24 25 26 27 24 25 22 23 20 21 22.9 23.9 Pollen fertility (%) 25.2 57.7 25.7 81.1 66.0 14.2 24.4 48.8 0.0 24 26 28 26 24 22 24.9 0.0 4. 6 0.0

December 2004

Pollen sterility (%) 100 90 80 70 60 50 40

Daily mean temperature (oC) 37.0 33.5 30.0 26.5 23.0 19.5
Pollen sterility Temperature

16.0 12.5 9.0 5.5

Sept 10

Date of flowering The fertility expression of IR73827-23S in a field at Wuhan, China, in 2003.

June 20

June 30

June 31

Table 2. Yield trials of some two-line hybrids, China, 2003. Location Hainan Combination IR73827-23S/Luhui 6 IR73827-23S/TMQ30349 IR73827-23S/R615 Shanyou 63 (check) IR73827-23S/Mianhiu501 IR73827-23S/Minghui63 IR73827-23S/TMQ30349 IR73827-23S/R524 IR73827-23S/TMQ30506 IR73827-23S/TMQ30525 II-You 725 (check) Growth duration (d)a 134 137 135 141 133 130 130 130 132 128 129 Yield (t ha1) 11.80 12.26 11.67 10.53 8.23 7.50 8.33 8.80 9.33 7.73 6.91 Increase over check (%) 12.0 16.4 10.8 19.0 8.5 20.6 27.4 35.1 11.9

References

Wuhan

Shi MS. 1981. Preliminary report of breeding and utilization of late japonica natural double purpose line. J. Hubei Agric. Sci. 7:1-3. Lu XG, Virmani SS, Yang RC. 1998. Advances in two-line hybrid rice breeding. In: Virmani SS, Siddiq EA, Muralidharan K, editors. Advances in hybrid rice technology. Proceedings of the Third International Symposium on Hybrid Rice, 14-16 Nov 1996, Hyderabad, India. Los Baos (Philippines): International Rice Research Institute. p 89-98.

Days from seeding to maturity.

Response of transgenic rice expressing two Bt genes to nontarget insects

K. Bashir, T. Husnain, and S. Riazuddin, National Centre of Excellence in Molecular Biology (CEMB), 87-W, Canal Bank Road, University of the Punjab, Lahore, Pakistan

Different transgenic lines of indica basmati rice (Basmati 370) expressing cry1Ac, cry2A, or cry1Ac + cry2A genes were selected on the basis of molecular analysis and insect bioassays. During the first year (2001), 396 plants of five
IRRN 29.2

different lines, along with the untransformed control, were sown following a randomized complete block design at CEMB, while a total of 3,000 plants of four transgenic lines expressing two Bt genes were randomized following a

split-plot design at CEMB and at Rice Research Institute, Kala Shah Kaku, Lahore, during the second year (2002). These lines showed a high level of resistance to yellow stem borer and rice leaffolder throughout the entire growth
15

Sept 20

May 11

May 16

May 21

May 31

July 22

July 10

Aug 11

Aug 21

Aug 31

May 1

May 6

Aug 1

2.0

16

December 2004

season (Bashir et al, unpubl. data). The specificity of cry genes for target insects is a great advantage in planning for field testing of transgenic lines containing such toxins. It is interesting to check the response of insects belonging to insect orders other than Lepidoptera and Diptera to transgenic lines expressing Bt genes at a high dose. A common method to check the response of transgenic lines to such insects is to compare the damage caused by these insects in transgenic lines and a control (Fitt et al 1994, Sims 1995, Orr and Landis 1998, Bashir et al 2004). In our studies, 15 plants per plot per line were selected to study the extent of damage caused by nine different insects of three orders (Hemiptera, Homoptera, and Orthoptera) belonging to five different families. These insect species were first identified in the experimental field and the damage they caused was characterized. The number of leaves damaged by any of these

insects per plant was counted and expressed as a percentage of total leaves. Data were processed by analysis of variance, followed by Duncans multiple range test. No statistical difference was observed between the damage in the transgenic lines and that in the control in any case. During the first year, the average damage in the transgenic lines was 7.84% as compared with 7.40% in the untransformed control. The following year, damage in transgenic lines was 14.4%, while that of the untransformed control was 14.0% (see figure). These results showed that transgenic lines expressing one and two Bt genes have no adverse effects on the presence of nontarget insects belonging to the insect orders Hemiptera, Homoptera, and Orthoptera. References
Bashir K, Husnain T, Fatima T, Latif Z, Mehdi SA, Riazzudin S. 2004. Field evaluation and risk assessment of transgenic indica basmati rice. Mol. Breed. 13:301-312.

Damage (%) 16 14 12 10 8 6 4 2 0 2001 Year Damage caused by nontarget insects in transgenic lines and an untransformed control (means statistically similar according to ANOVA, followed by Duncans multiple range test). 2002 Transgenic Control

Fitt GP, Mares CL, Llewellyn DJ. 1994. Field evaluation and potential ecological impact of transgenic cottons (Gossypium hirsutum) in Australia. Biocont. Sci. Technol. 4:535-548. Orr DB, Landis DL. 1997. Oviposition of European corn borer (Lepidoptera: Pyralidae) and impact of natural enemy populations in transgenic versus isogenic corn. J. Econ. Entomol. 90(4):905-909. Sims SR. 1995. Bacillus thuringiensis var. kurstaki CryIA(c) protein expressed in transgenic cotton: effects on beneficial and other non-target insects. Southwestern Entomol. 20(4):493-500.

Genetic resources

Giria high-yielding, slender-grain, sheath blight-and bacterial blight-resistant variety for shallow-water conditions in West Bengal
S.K. Sinha, S.N. Sen, and A. Biswas, Rice Research Station, Chinsurah, Hooghly, India

Table 2. Grain and straw yields (t ha1) of Giri and check varieties in front-line demonstration trials conducted in West Bengal. Season/ location 1999 kharif Midnapore 2000 kharif Dinajpur Hooglhy Birbhum Nadia 2001 kharif Malda Murshidabad Nadia Burdwan Jalpaiguri
a

Mean grain yield (t ha1) Giri Checka

% increase/ decrease over checks 38 8 3 14 13 12 4 10 99

Mean straw yield (t ha1) Giri Checka

% increase/ decrease over checks 38 15 15 14 4 20 97

Table 1. Characteristics of Giri. Plant height (cm) Flowering duration (d) Plant type Number of tillers per plant Panicle type Panicle exsertion Awning Apiculus color 1,000-grain weight (g) Kernel length (mm) Kernel breadth (mm) L/B Kernel appearance Milling (%) Head rice (%) 130 115120 Erect, semitall, nonlodging 911 Intermediate Complete Absent Nonpigmented 21 6.21 1.86 3.34 White 63.6 59.2

3.50 3.90 1.81 4.65 1.41 3.23 4.45 3.38 5.53 4.23

2.53 3.61 1.87 4.09

b

6.11 4.79 2.01 1.92 6.58 6.51 5.46 5.19

4.43 4.18 2.01

b

2.86 3.98 3.51 5.05 2.14

5.78 6.28 4.56 2.62

Checks were Mugai and Swarna in 1999 and 2000, respectively. In 2001, checks were BR11 at Nadia, a local variety at Jalpaiguri, and Swarna at other sites. b Damaged by floods.

IRRN 29.2

17

Rice cultivation in lowlands is complex because of the presence of several biotic and abiotic stress factors. Submergence tolerance and incidence of different pests and diseases, including constraints to the adoption of control measures in waterlogged fields, are special problems for this ecosystem. In addition, superior grain quality, especially long slender grains, is lacking, except for variety Sashi, which was recently released from this research station by Sinha et al (2001). A hybridization program began in 1984-85 using IR36, a variety with multiple resistance developed from the International Rice Research Institute, and Bhasamanik, a local landrace with wide adaptability and good grain quality grown under lowland conditions. Giri, developed from this cross, was nominated in

1994 for the All India testing under shallow-water conditions (Table 1). Based on overall performance, the variety (as IET14400) was recommended in 1998 for shallowwater conditions and was released in 2002 by the State Variety Release Committee for West Bengal after extensive testing through front-line demonstration trials in farmers fields (Table 2). The results showed its superiority over check entries, including Swarna, which is very popular in eastern Indian states, particularly West Bengal. Swarnas popularity is on the decline because of its susceptibility to sheath blight and other diseases. Out of 1,261 entries tested, the Directorate

of Rice Research has identified Giri as resistant to sheath blight and bacterial leaf blight. Moreover, it can tolerate submergence. Its long slender grain makes its cultivation remunerative to farmers. Its straw yield was also higher than that of check varieties, bringing about additional income from increased total biological yield. The variety has been released for lowland cultivation in West Bengal. Reference
Sinha SK, Mandal BK, Chatterjee SD. 2001. Sashi (CN842-15-5, IET14105), a high-yielding, long-slender grain variety for shallow water conditions in West Bengal, India. Int. Rice Res. Notes 26(2):33-34.

Varietal turnover and seed exchange: implications for conservation of rice genetic diversity on-farm
A. Sirabanchongkran, N. Yimyam, W. Boonma, K. Rerkasem, Faculty of Agriculture, Chiang Mai University, Chiang Mai 50200, Thailand; K. Coffey and M. Pinedo-Vasquez, Department of Ecology, Evolution, and Environmental Biology, Columbia University, New York; and C. Padoch, Institute of Economic Botany, The New York Botanical Garden, Bronx, New York Email: yim1975@yahoo.com

The loss of rice genetic diversity has been recognized as an issue of global concern. The Conference on Biological Diversity specifically mentions on-farm conservation of this diversity as an important strategy to ensure continued availability of these resources for both local communities and breeding programs (UN 1992). Northern Thailand lies in the heart of the primary center of diversity of rice (Chang 1976). Farmers in northern Thailand have long made use of a great diversity of rice varieties in their upland and paddy farms. Over the past few decades, many agricultural systems in this region have changed as farmers gained access to improved varieties and chemical fertilizers as well as to new markets for cash crops. The adoption and use of new technologies has varied widely among communities and farmers and the landscape of northern Thailand now features an array of rice production systems. There is also great variation in the importance of local varieties and genetically diverse seed lots (Brush 1995). In this study, we test our hypothesis that seed turnover is an important component of
1

farming systems rich in biodiversity. We suggest that this finding conflicts with conservation initiatives that focus on preventing changes in varietal preference. We also explore the relationship between marriage patterns and networks of seed exchange. We discuss potential conflicts between common interventions (such as seed fairs and community seed banks) and local exchange systems that are embedded in cultural practices. This paper focuses on data from Tee Cha, one of the three villages from which data have been collected. It is a Pwo Karen village established more than 200 y ago in the Salween watershed near the border between Thailand and Myanmar. The village comprises 39 households with a total population of 158. The primary agricultural system in the village is upland shifting cultivation (with a 6 7-y fallow cycle) of rice and other swidden crops for subsistence. More than 20 traditional rice varieties are grown each year in the village, with individual households growing 3 to 5 varieties, depending on household requirements and biophysical limitations.

Many on-farm conservation projects focus on encouraging continued planting of local varieties and much effort is placed on stopping farmers from abandoning these varieties. But it was found that the rate of abandonment or variety turnover was high, even when the number of local varieties grown in the village remained stable. In Tee Cha, abandonment and acquisition of varieties appear to be the norm (Table 1). When considering all of the varieties, an average of 22% 1 of farmer households that grew a variety in one year failed to plant that same variety the next year. However, the number of varieties grown in the village remained stable throughout the 3-y period. Results were similar for the other two study villages. This example displays the fluidity of germplasm use in villages that retain high levels of agricultural diversity. It also highlights the importance of understanding the dynamics of this movement if conservation initiatives are to be effective. To understand the pathways of germplasm movement, we investigated the connection between village exogamy (an exogamous

This percentage was calculated by comparing each variety a farmer grew in year one of the study with each variety a farmer grew in year two and then repeating the calculation for the interval between years two and three. Twenty-two percent is the average for all farmers in the study for intervals one/two and two/three.

18

December 2004

Table 1. Varietal turnover in Tee Cha, 2001-03. Variety namea Farmers growing the variety (no.) 15 7 14 8 2 1 13 6 2 3 1 3 1 1 2 1 6 23 3 1 Growers as percent of villageb (N = 36) 42 19 39 22 6 3 36 17 6 8 3 8 3 3 6 3 17 64 8 3 Acquiredc Abandonedd

Table 2. Village exogamy and the use of rare varieties, 2001-03. Marriage type Households (no.) 26 10 36 Households growing rare varietiesc (no.) 12 3 15

Bue Bang Bue Chu Bue Gua Bue Goal Bue Kee Bue Khu Koo Bue Mue Ta Bong Bue Paw Low Bue Pho Lae Bue Tho Lae Bue Houy Singh Bue Ka Sed Bue Sor May Pa Ai Bang Pa Ai Chai Pa Ai Chair Pa Ai Goal Pa Ai Khupe Pa Ai Ki Kai Pa Ai Na

5 4 4 3 0 0 2 3 1 0 1 0 1 1 1 0 2 6 1 0

4 2 3 3 0 0 0 0 0 1 0 0 0 0 1 0 3 3 0 0

Exogamousa Endogamousb Total

a Exogamous marriages contain at least one individual from outside of the village. bEndogamous marriages are between two individuals from inside the village. c For this research, a rare variety is defined as a variety that is grown only by one or two households in the village.

a In the Karen language, Bue designates nonglutinous varieties and Pa Ai designates glutinous varieties. bTotal number of farmers who grew the specified variety at least once during the 3-y study. cNumber of farmers who started to grow the variety after not growing it for more than 1 y during the 3-y study. dNumber of farmers who grew the variety and then stopped for at least 1 y during the 3-y study.

household consists of a married couple with one or both individuals immigrating from outside of the village) and the use of common rice varieties. We hypothesized that farmers with strong kinship connections outside of the village would be more likely to grow less common varieties. The results show a significant difference in the number of rare varieties (defined here as a variety that is grown only by one or two households in the village) grown by endogamous and exogamous households (Table 2). This finding points toward the importance of kinship as a pathway for seed exchange. It also suggests that ties outside of the village might be valuable sources of local varieties.
IRRN 29.2

The discovery that rice variety turnover is a prominent feature of seed systems in villages that maintain high varietal diversity is consistent with recent studies of other crops such as maize (Louette 1998). Analysis of data from the other two villages in our study also supports this conclusion. Results like these have motivated contemporary conservation initiatives to consider populations of varieties as metapopulations of fields interconnected by varietal exchange. This conclusion enables workers to look beyond static presence/ absence tallies and focus on the networks of exchange that make varieties available to farmers. A common reaction to this inference is to attempt

to increase the exchange of varieties by hosting seed fairs or establishing communitylevel seed banks. As the exogamy example above illustrates, intricate local systems of exchange are already in place. Not only are marriage relations important routes of exchange, but many norms of exchange are deeply embedded in cultural practices. For example, Tee Cha villagers cite a rule that requires children to maintain a specific family variety after their parents death. These findings are important because conservation initiatives that attempt to increase varietal exchange without understanding local practices might actually undermine the local systems of exchange that are crucial to the maintenance of diversity. Even in a case where local varieties match the preferences of local farmers, it should not be assumed that increased access to local varieties invariably results in increased diversity. In-depth studies of villages targeted for conservation are required to ensure that introduced methods of seed exchange do not threaten local systems.
19

References

Brush SB. 1995. In-situ conservation of landraces in centers of crop diversity. Crop Sci. 35:346-354. Chang TT. 1976. The origin, evolution, cultivation, dissemination, and diversification of Asian and African rice. Euphytica 25:425-441. Louette D. 1998. Traditional management of seed and genetic

diversity. In: Brush SB, editor. Genes in the field: on-farm conservation of crop diversity. Boca Raton, Fl (USA): Lewis Publishers. UN (United Nations). 1992. Convention on Biological Diversity. Rio de Janeiro (Brazil): United Nations Conference on the Environment and Development.

Acknowledgments

The authors thank the Collaborative Crop Research Program of the McKnight Foundation and the farmers of Tee Cha, Mae Rid, and Mae Moot for their support.

Invasion of weedy rice in rice fields in Thailand: problems and management

C. Maneechote, Plant Protection Research and Development Office, Department of Agriculture, Chatuchak, Bangkok 10900, Thailand; S. Jamjod and B. Rerkasem, Agronomy Department, Faculty of Agriculture, Chiang Mai University, Chiang Mai 50200, Thailand E-mail: mchanya@asiaaccess.net.th

Wild rice (Oryza rufipogon Griff.), a relative of cultivated rice (Oryza sativa L.), is common in natural habitats throughout Thailand (Chitrakorn 1985). Characteristics of wild rice include seed shattering, black seed coat, awns on seeds, red pericarp, and sensitivity to photoperiod (Oka 1988). The oldest farmers say that wild rice has always been a common feature of rice fields in Thailand. They and younger farmers as well recall that, even 20 y ago, it was found mainly in natural habitats on the edges of farm ditches, deep depressions, and swampy places, but rarely in rice fields. Weedy rice with characteristics of O. rufipogon has been recognized as a weed in rice fields, but the level of infestation had been quite low, until recently. In 2001, weedy rice was observed in two areas of the Central Plain of Thailand. The first area was in Kanchana20

buri, northwest of Bangkok, where high-yielding varieties (HYVs) are wet-seeded in a double-rice cropping system. The weedy rice flowered approximately 2 wk earlier and its panicles were 2030 cm taller than the rice crop. In Prachinburi and Nakhon Nayok, northeast of Bangkok, weedy rice was found in deepwater rice (25 m water depth), where seeds of mostly traditional deepwater varieties are dry-sown to germinate after the initial rains (Fig. 1). A crop-cut survey found grain yield decreasing linearly with increases in percent infestation (Fig. 2). In addition to yield loss, weedy rice also lowered the price because of the admixture of grains with red pericarp. A monitoring of more than 100 fields in two villages of Kanchanaburi Province from 2001 to 2003 revealed an increasingly more serious and expanding weedy rice problem. From

just three infested fields in May 2001, more than 90% of the fields were affected by October 2003. Infestation ranged from 10% to 90%. To date, weedy rice is also found in other provinces of Central ThailandPathum Thani, Suphan Buri, Nakorn Pathom, and Saraburi. Farmers classify weedy rice into different groups according to appearance: Khao Hang (rice with awn), Khao Deed (jumping rice), and Khao Daeng (red rice) (see table). The occurrence of weedy rice in cultivated rice is very difficult to control with herbicides because of the genetic similarity between the two types. Hand removal of weedy rice seedlings in the field at the early stages is also impractical because they are indistinguishable from cultivated rice at this time. Where weedy rice panicles emerge 1 or 2 wk earlier and are taller than cultivated rice, a popular control method among farmDecember 2004

Some characteristics of three types of weedy rice. Characteristic Khao Hang Khao Deed Khao Daeng

% seed 50 100 0 shattering Plant Taller than As tall as Taller than/ height rice crop rice crop as tall as rice crop Pericarp White White Red and red and red Seed Black, Yellow Brown coat yellow, brown Awn Yes Yes No on seeds and no

Fig. 1. Weedy rice in direct-seeded rice cultivation in Kanchanaburi (a) and Nakhon Nayok (b), Thailand. Note: panicles of cultivated rice were covered with the taller weedy rice and could not be seen here.
Rice yield (t ha-1) 8

20

40 Infestation (%)

60

80

100

Fig. 2. Decline in rice yield with increasing weedy rice infestation (percentage of panicles with at least one of the following characteristics: taller than the rice crop, panicles with almost 100% shattering, spikelets with awn). Each symbol represents a 1 X 1-m2 sample in a farmers field (of HYV, mainly Supanburi 1 or Chainat 1) in Samsib Harb and Kok Trabong villages, Kanchanaburi Province, Thailand.

ers in the study villages is topping off panicles with sickles. This, however, seems to have driven the weedy rice to mimic cultivated rice more closely in terms of heading time and plant height. The best way to prevent weedy rice from becoming invasive is to make sure that weedy rice seed is not introduced into rice fields. Clean seed must be used. However, farmers in the study area, as in other parts of Thailand, rarely ever sow certified seed. Farmers generally keep some of last years crop as seed or they may get it from other farmers apparently clean crop. We have found up to 4,000 weedy rice seeds in 1 kg of an apparently clean seed crop (Boonchoui, unpubl. data). It has been estimated that only two red rice seeds kg1, planted in a red ricefree rice field, could produce 100 kg of red rice ha1 within three seasons (Noldin 2000). Sowing seed contaminated by weedy rice is likely the primary cause of invasiveness of rice fields.

IRRN 29.2

21

To control weedy rice that has become established in rice fields, different methods were tested: 1. Preplanting herbicide application to kill weedy rice and other weeds that emerge before sowing certified clean rice seed. Removal of panicles prior to seed set was also practiced. 2. One fallow with flooding to eliminate the seed bank. Clean seeds and hand pulling were practiced the following season. 3. Transplanting rice with clean seeds to facilitate removal by hand. Preplanting herbicide was applied during land preparation.

4. Spray-topping with herbicides to reduce seed bank accumulation but achieve some yield. The control of weedy rice in rice fields should benefit from understanding the conditions that have led to this sudden invasiveness. References
Chitrakon S. 1995. Characterization, evaluation and utilization of wild rice germplasm in Thailand. PhD thesis, Hokkaido University, Japan. Noldin JA. 2000. Red rice status and management in the Americas. In: Baki BB, Chin DV, Mortimer M, editors. Proceedings of wild and weedy rice in rice ecosystems in Asia: a review. Los Baos (Philippines): International Rice Research Institute. p 21-24.

Oka HI. 1988. Origin of cultivated rice. Tokyo (Japan): Scientific Societies Press/Elsevier. 254 p.

Acknowledgments

This work is supported by the Thailand Research Fund and the McKnight Foundation. The authors are grateful to Mr. Sophon Piyasirinond of Aventis Crop Science (Thailand), who drew our attention to herbicide-resistant Ya Khao Nok, which turned out to be weedy rice; Mr. Samruey Sametup, Mrs. Malai Ruengrueng, Mr. Prasert Jaimwong, Mr Somjit Raksai, Mr. Prayad, Mrs. Bang-on Kuklor, and the other farmers in Kanchanaburi for their cooperation and advice.

Two outstanding rice varieties developed through selection from naturally occurring genetic variation in Sri Lanka
D. Sumith de Z. Abeysiriwardena, Rice Research and Development Institute, Batalagoda, Ibbagamuwa, Sri Lanka

Crop improvement is based on creating genetic variability and using it through selection. Natural outcrossing and mutation are the main forces that create natural genetic variability in rice. As rice is a self-pollinated crop, incidences of natural outcrossing are rare and mutation is also one in a million. Therefore, the occurrence of natural genetic variability is rare and, as a result, use of such genetic variability has not been given much emphasis in rice. However, finding rare genetic combinations is possible in populations where genetic variability has
22

occurred naturally. Two rice varietiesLanka Samurdhi and Bg 250having such rare genetic combinations have been developed through selection from naturally occurring genetic variation in Sri Lanka. Thus, the importance of naturally occurring genetic variability cannot be ignored and should be given due consideration in rice improvement. Lanka Samurdhi is a semidwarf variety with Basmati-type grain quality traits and high yield potential. Sri Lankan plant breeders failed to combine the superior

grain quality and high yield potential with the semidwarf character through artificial hybridization and mutation breeding. The potential of natural outcrossing in getting the desired genetic combination was thus harnessed. Natural outcrossing was successful and, finally, Lanka Samurdhi, which bred true to type, was identified. The grain yield of Lanka Samurdhi was comparable with that of the nonaromatic standard checks and was much higher than that of Basmati 370, the Basmati-type check (Table 1). Lanka
December 2004

Samurdhi was comparable with Basmati 370 in terms of grain quality but superior to it in other agronomic characteristics (Table 2). It has strong aroma even under conditions of high temperature and high humidity. This variety has officially been released for cultivation in Sri Lanka and there is an increasing demand for it among commercial farmers. Plant breeders have likewise been successful in developing an ultra-short-duration variety that matures in around 80 d, particularly during the dry season, with a yield potential of more than 6 t ha 1, through selection from naturally occurring genetic variation in farmers fields. One of the major problems in rainfed rice cultivation in most of Sri Lanka is water shortage during maturity of even the presently grown shortest maturity group (3 mo). Thus, most of the rainfed areas and some of the irrigated areas go under fallow during the dry season. An ultra-short-duration variety that can escape from drought can overcome this problem. This miracle variety, which has been named Bg 250, is a nonlodging, short plant type with high-yielding ability (Table 1), acceptable grain quality, and resistance to leaf blast, bacterial leaf blight, thrips, and brown planthopper (Table 2). It is also responsive to a high fertilization rate. Data from adaptability trials have shown that this variety is suited to drought- and floodprone areas. In the former, it can escape from drought if planted early; in the latter, it
IRRN 29.2

Table 1. Average grain yields (t ha 1 ) of Lanka Samurdhi, Bg 250, and check varieties in researchers and farmers fields during the wet and dry seasons. Variety typea Location/season Basmati-type, superior grain quality Lanka Samurdhi Researchers fields Wet season Dry season Farmers fields Wet season Dry season 3.63 a 4.40 b 4.88 a 6.27 a Nonaromatic checks 3.88 a 5.95 a 4.50 a 6.39 a Basmati checks 0.60 b 1.91 c 2.49 b b Ultra-short-duration Bg 250 Bg 750 (check) 2.86 b 2.49 b 3.28 b

4.32 a 3.83 a 4.72 a

a Within a variety type X location X season combination, means followed by the same letter are not significantly different at the 5% probability level. b = data not available.

Table 2. Agronomic and grain quality characteristics of Lanka Samurdhi, Bg 250, and their respective check varieties, Basmati 370 and Bg 750. Variety Characteristic Basmati-type, superior grain quality Lanka Samurdhi General Plant type Semidwarf Lodging Resistant Tillering High Maturity duration (d) 118123 Response to N fertilizer High Pericarp color White Yield potential High Reaction to pests and diseasesa Rice gall midge Sa Brown planthopper MR Thrips MS Leaf blast MR Bacterial leaf blight MS Grain quality Brown rice (%) 81.6 Total milled rice (%) 74.0 Head grain (%) 64.4 Grain length (mm) 7.2 Grain shape Long slender Chalkiness Low Translucency High Gelatinization temperature (C) Low Amylose content (%) 17.6 Grain elongation ratio 1.7 Degree of aroma Strong Basmati 370 Ultra-short-duration Bg 250 Bg 750

Tall Susceptible Low 102108 Low White Low S S S S S 76.8 73.0 25.9 7.8 Long slender High Low Intermediate 21.0 2.0 Strong

Semidwarf Resistant High 7485 High White High MR/MS R/MR MR R/MR MR 78.2 74.4 65.9 b Intermediate bold Low Intermediate Intermediate High

Tall Susceptible Low 7986 Low Red Low S MR/MS MS MS MR 78.3 74.7 64.5 Intermediate bold Low Intermediate Intermediate High

a R = resistant, S = susceptible, MR = moderately resistant, MS = moderately susceptible. b = data not available.

can escape from the flood if planted after the flood has receded, yielding well before the subsequent dry period.

This variety is expected to be officially released for cultivation in Sri Lanka before the end of 2004.
23

Dorfak: an aromatic, high-yielding, short-duration variety with good cooking quality for the irrigated lowlands of Iran
M. Nahvi, M. Allahgholipour, A. Jauhar Ali, M.S. Mohammed Salehi, H. Rahim Saroush, H. Dorosti, A. Erfani, F. Padasht, and F. Alinia, Rice Research Institute of Iran, P.O. Box 1658, Rasht, Iran

Rice is a diet staple, next to wheat, for Iranians, especially in the two northern provinces of Guilan and Mazandaran. In 2003, rice was cultivated on 615,000 ha. Total production was 3.3 million t and productivity averaged 5 t ha 1 . Rice production and productivity have decreased over the past 2 y primarily because of the wide-scale adoption of improved high-yielding varieties. Before 2001, the area grown to traditional landraces (comprising mostly premiumquality Sadri rice) remained stable because farmers refused to adopt improved varieties on account of their poor cooking quality. Iranian consumers gave high preference to good cooking quality. Traditional landraces such as Hashemi, Musa Tarom, Binam, and Salari, which possess strong aroma and intermediate amylose and gelatinization temperature, and which become soft and flaky upon cooking, were cultivated on more than 60% of the total rice area. However, these tall landraces have low yields (3.0 t ha 1 ) and are susceptible to blast and lodging. Therefore, one of the major breeding objectives in the last two decades has been to combine the premium-quality features of these landraces with the high yield, dwarfness, and disease and insect pest resistance of improved varieties.
24

Dorfak (line no. 424) is a single-cross derivative between Sepidrood, a highyielding improved variety (male parent), and Salari, a traditional landrace with premium cooking quality but poor yield (female parent). The cross was made in 1988 and the segregating population was handled using the pedigree breeding approach. Single-plant selections in the early generations were based on early, synchronous flowering habit, increased number of tillers, high yield potential with reduced plant height, and resistance to blast disease. However, selections based on grain quality began in the F 5 generation (1994). In 1995, of the different pedigree families selected, line 424 was identified as high-yielding, semidwarf, resistant to blast, and having good cooking quality. In preliminary and advanced yield trials, it was compared with improved varieties such as Khazar and Sepidrood. Multienvironmental trials were conducted at three sites in 1996-98 and the pooled analysis of these trials averaged over sites and years revealed highly significant yield differences among entries. Line 424 performed consistently, giving a yield advantage of 17.4% over improved varieties (range of 5.56.3 t ha 1 ) (Table 1). It had a 34.5% yield

advantage over traditional check variety Binam at three on-farm trial sites (data not shown). Named Dorfak, line 424 was released for cultivation in Guilan Province in 2002. Dorfak is a semidwarf (plant height, 110 cm), aromatic, short-duration variety (120 d) with 17 productive tillers, 115 grains per panicle, and good grain and cooking quality features (Table 2). Rice kernel characteristics clearly showed that Dorfak is a transgressive segregant, especially for kernel length and lengthwidth ratio. However, for all other grain quality traits, Dorfak was much closer to the traditional landrace parent Salari. Dorfak has become popular with farmers and consumers alike.

Table 1. Yield performance of Dorfak in adaptive research trials conducted at three different sites from 1996 to 1998. Site Av yield (t ha1) (over 3 y) Dorfak 6.1 5.9 5.8 5.9 Khazara 5.1 4.7 4.9 4.9 % increase over check 16.4 20.3 15.5 17.4

Rasht Fouman Astaneh Mean

a

Improved check variety.

December 2004

Table 2. Agronomic and grain quality characteristics of Dorfak in comparison with its parents. Characteristica Agronomic traits Plant height (cm) Productive tillers plant1 (no.) Maturity (d after sowing) Grains panicle1 (no.) Grain quality traits Total milled rice recovery (%) Milled head rice recovery (%) Broken rice (%) Shape Aroma (present/absent) Grain length (mm) Kernel length (mm) Kernel width (mm) Kernel length/width Cooked kernel length (mm) Cooked kernel width (mm) Cooked kernel length/width Elongation ratio (lengthwise) Amylose (%) Gel consistency (gel length in mm) Gelatinization temperature (alkali spreading value score) Dorfak 110 17 120 115 68.4 53.9 14.5 Long slender Present 11 7.9 1.9 4.16 12.1 2.9 4.17 1.53 23.3 58.2 3.0 Salarib 141.9 12.8 119 97.2 70.5 53.5 17.0 Long slender Present 9.8 7.4 1.9 3.89 12.3 2.4 5.10 1.66 21.7 60.0 3.3 Sepidroodc 89.3 16.5 124 91.6 69.5 40.4 29.1 Long slender Absent 9.8 7.3 1.9 3.84 11.1 2.7 4.11 1.52 27.0 27.0 7.0

a All traits measured according to the IRRI Standard evaluation system for rice (1996). bSalaritraditional landrace check. cSepidroodimproved check variety.

IRRN 29.2

25

Pest science & management

Long-term effects of fertilizers and herbicides on a nematode population under a rice-wheat-cowpea system
A.P. Singh and Nand Ram, Department of Soil Science, G.B. Pant University of Agriculture and Technology, Pantnagar 263145, India E-mail: nandram@fastmail.fm

Nematodes are ubiquitous. They respond to soil inputs and management practices and can be used as a measure of the biological response of soils to these manipulations (Bohlen and Edwards 1994). Fertilizers and herbicides are regularly being used to enhance crop production. But information on the behavior of nematodes under the continuous use of these agricultural inputs in a fixed crop sequence over the years is meager. This study was carried out during the 30th year of the long-term fertilizer experiment initiated in 1971 with the rice-wheat-cowpea system at Pantnagar (29 N, 79 3 E) on an Aquic Hap1udoll in northern India. The climate is humid subtropical (Nand Ram 1995). Details of the treatments are shown in Table 1. Fertilizers, farmyard manure (FYM), and herbicides were applied in this permanent field experiment that used a randomized block design. These inputs were regularly applied only to rice and wheat; cowpea has been grown as a fodder crop without any agricultural input since the beginning of the experiment. For rice and wheat, annual NPK rates at optimal

dose were 240, 52, and 70 kg ha l applied as urea, single superphosphate, and muriate of potash, respectively. In the 100% NPK + FYM treatment, FYM was incorporated to the soil (15 t ha l) once a year before wheat sowing. In the 100% NPK + Zn treatment, Zn was used as a foliar spray in both rice and wheat until 1992. But, in 1993, soil application at 50 kg ZnSO 4 ha l was followed. Treatments with 100% N, 100% NP, and 100% NPK + hand weeding (HW) were Zn-free up to 1992, after which Zn was applied in the soil at 50 kg ZnSO 4 ha l to correct a

severe Zn disorder in rice. In all treatments with Zn, application at 50 kg ZnSO4 hal was done when the soil became Zn-deficient. Butachlor and pendimethalin were used as herbicides to control weeds in rice and wheat, respectively. But, in the control and in the 100% NPK + HW treatments, weeds of both crops were removed by hand. Since the inception of this long-term experiment, no studies on nematodes have been conducted. For the first time, in June 2000, composite surface soil samples consist-

Table 1. Details of agricultural input treatments for rice and wheat. Treatment Control 100% Na 100 NPa 100% NPK 100% NPK + FYM 100% NPK + Zna 100% NPK + HWa Crops Rice Wheat Rice Wheat Rice Wheat Rice Wheat Rice Wheat Rice Wheat Rice Wheat Fertilizer rates (kg ha1) N 0 0 120 120 120 120 120 120 120 120 120 120 120 120 P 0 0 0 0 26 26 26 26 26 26 26 26 26 26 K 0 0 0 0 0 0 33 37 33 37 33 37 33 37 FYM (t ha1) 0 0 0 0 0 0 0 0 0 15 0 0 0 0 Weed control treatment Hand weeding Hand weeding Butachlor Pendimethalin Butachlor Pendimethalin Butachlor Pendimethalin Butachlor Pendimethalin Butachlor Pendimethalin Hand weeding Hand weeding

a Zinc was applied into the soil in 1993 and 1997 (soil was Zn-deficient). In the 100% NPK + Zn treatment, Zn was used as a foliar spray in both crops until 1992.

26

December 2004

ing of plant roots were collected at 015-cm depth of the rhizosphere to investigate the effect of the continuous use of fertilizers and herbicides on the nematode population. Nematodes, irrespective of kind, were isolated by vigorously stirring 250 cm 3 of soil with 4 L of distilled water, following the nematode counting procedure of Dropkin (1980). Soil samples were analyzed for pH, electrical conductivity (EC), and organic carbon. The nematode population under the different treatments ranged from 349 to 843 per 250-cm 3 of soil (Table 2). The highest number of nematodes was observed in the control, followed by the 100% NPK + HW treatments, where hand weeding was used and no herbicide was applied for about 3 decades. The use of herbicides in rice and wheat had a significant detrimental effect on the nematode population. This may be attributed to better weed management achieved by the use of herbicides as either weeds may
Table 2. Effect of different treatments on the nematode population. Treatment Nematode population per 250 cm3 of soil 843 626 613 606 349 592 773 7.5 Decrease in population (%) 25.7 27.7 28.1 58.6 29.8 8.3

Control 100% N 100 NP 100% NPK 100% NPK + FYM 100% NPK + Zn 100% NPK + HW CD (5 %)

serve as hosts for different nematodes (Kavanagh 1974) or herbicides may be toxic to the nematodes (Weischer and Muller 1985). Fertilizer application also suppressed the nematode population. As compared with the control, fertilizer inputs reduced the nematode population by 8.358.6% under different treatments. The addition of fertilizers either increased nematode mortality in the soil or enhanced root stock resistance to nematode infection (Melakeberhan et al 1997). Moreover, the combined use of optimal NPK fertilizers and FYM (100% NPK + FYM) proved harmful to the nematode population. The number of nematodes declined by 257 per 250 cm 3 of soil under integrated nutrient management (100% NPK + FYM) from 606 per 250 cm 3 under optimal NPK fertilizer alone (100% NPK). Gaur and Meher (1994) also reported good control of nematodes by FYM application. In addition, cowpea was planted as a fodder crop, relying on the residual fertility brought about by the different treatments. Therefore, the better growth of cowpea under 100% NPK + FYM (Nand Ram 1995) might have also contributed to a nematode population reduction as cowpea can effectively control several species of Meloidogyne. In the experimental field, soil pH, EC, and organic C ranged from 7.63 to 7.83, and 0.310.39 dS m l , 0.49-1.56%c1.56%, respectively. Soil pH (r = 0.069) and EC (r =

0.600) had a negative and nonsignificant relationship with the nematode population. Dadhwal et al (1987) reported the same findings. Soil organic C showed a negative and significant relationship (r = 0.883**) with the number of nematodes. References
Bohlen PI, Edwards CA. 1994. The response of nematode trophic groups to organic and inorganic nutrient inputs in agroecosystems: defining soil quality for a sustainable environment. Soil Sci. Soc. Am. Spl. Publ. 35. p 235-244. Dadhwal KS, Venkataraman S, Sengupta TK, Laskar S. 1987. Nematode population in acid soil. J. Indian Soc. Soil Sci. 35:141-142. Dropkin VH. 1980. Introduction to plant nematology. New York: John Wiley and Sons. p 58-59 Gaur HS, Meher HC. 1994. Integrated control of plant parasitic nematodes in tomato nursery bed and field application of a nematicide, fertilizer and organic manure. Afro-Asia J. Nematol. 4:203-206. Kavanagh T. 1974. The influence of herbicides on plant disease. II. Vegetables, root crops and potatoes. Sci. Proc. Royal Dublin Soc. B. 3:251-265. Melakeberhan H, Bird GW, Gore R. 1997. Impact of plant nutrition of Pratylenchus penetrans infection of Prunus avium rootstocks. J. Nematol. 29:381-388. Nand Ram. 1995. Long-term effect of fertilizers on crop production and soil properties in a Mollisol. Tech. Research Bull. 124. Experimental Station, G.B. Pant University of Agriculture and Technology, Pantnagar, Uttar Pradesh, India. Weischer B, Muller J. 1985. Side effects of plant protection products on nematodes and their antagonists. Berichte Uber Landwirtschaft. Sonderheft. 198:159-176.

IRRN 29.2

27

A method for transforming rice stinkbug counts in rice

Mu Mu Thein and V.S. Singh, Division of Entomology; and S. Chander and N. Kalra, Unit of Simulation and Informatics, Indian Agricultural Research Institute (IARI), New Delhi 110012, India E-mail: s_chander2000@yahoo.com

Most statistical procedures presuppose a normal distribution, with variance independent of mean insect population. However, in field counts of a majority of insect species, variance varies as a power function of the population (Taylor 1961). Such data have to be transformed before analysis to eliminate the dependence of the variance on the mean population. This study aimed to assess the effectiveness of aggregation parameter b-based transformation in breaking the dependence of population variance on the mean density of rice stinkbug Leptocorisa varicornis F. and to compare its performance with routinely used logarithmic transformations. The present investigation was carried out using rice variety Pusa-834 during the rainy seasons of 2001 and 2002 at IARI. The experiment involved two sets (Set 1 and Set 2) in the first year and one set (Set 3) in the second year, each set containing 40 2 m 1.5-m quadrats. One-monthold seedlings were transplanted in well-puddled fields on 18 Jul 2001 and 20 Jul 2002 at 15 X 20-cm spacing. NPK was applied at 120-60-40 kg ha 1 and the crop was irrigated as required. No pesticide was applied. The populations of nymphs and adults of L. varicornis were recorded on
28

10 randomly selected hills in every quadrat at 5-d intervals after flowering until the pests disappeared. The population counts so obtained were multiplied by 3 to derive a population estimate for a 1 m 1-m area as each quadrat contained about 30 hills (Ruesink and Kogan 1982). The mean population (M) and variance (S 2) were determined for combined nymphal and adult populations observed at each observation. Taylors power law (Taylor 1961) was fitted to mean population and variance data of Set 1 and sampling parameter a and aggregation parameter b were determined. The original pest counts (X) of each of the 40 quadrats for every observation were transformed by log (X + 1) and

X 1b/2 transformations. The value of b obtained from the data of Set 1 was used for transforming the original pest counts of sets 2 and 3. The mean and variance were worked out for transformed counts with each method of transformation. The appropriateness of these transformations was examined by computing simple correlation coefficient r between the variance and mean of original counts as well as transformed counts. The chi-square test (Barletts test) for testing the homogeneity of sampling variances in each set was also carried out (Gomez and Gomez 1984). The values of a and b were found to be 1.1429 (log a = 0.058) and 1.5048, respectively, in Set 1 (see figure). A

Log variance (S2) 1.3 1.0 0.7 0.4 0.1 0.2 -0.5 -0.5 Set 1 Log S2 = 1.5048 log M + 0.058 R = 0.9038

-0.3

-0.1

0.1

0.3

0.5

0.7

Log mean population (M)

Fig. 1. Determination of Taylors power law parameters for rice stinkbug Leptocorisa varicornis.

December 2004

Table 1. Population mean and variance of original and transformed counts of Leptocorisa varicornis infesting Pusa 834, 2001 rainy season. a Set 1 Sample Original count (X) Log (X + 1)transformed count M 0.218 0.316 0.411 0.484 0.271 0.379 0.126 S2 0.054 0.061 0.068 0.103 0.059 0.071 0.029 X1b/2transformed count (b = 1.5048) M 0.572 0.593 0.977 1.070 0.843 0.875 0.389 S2 0.341 0.608 0.270 0.305 0.294 0.345 0.261 M 0.97 1.5 1.95 2.77 1.55 1.05 0.32 Original count (X) Set 2 Log (X + 1)transformed count S2 1.056 1.538 2.726 7.513 1.741 1.279 0.277 2 = 109.2 (P<.01, 6 df) r = 0.911 (P<0.01, 5 df) M 0.236 0.341 0.393 0.470 0.335 0.248 0.095 S2 0.053 0.054 0.074 0.098 0.070 0.056 0.022 2 = 81.7 (P<0.01, 6 df) r = 0.954 (P<0.01, 5 df) X1b/2transformed count (b = 1.5048) M 0.626 0.872 0.926 1.041 0.798 0.655 0.257 S2 0.326 0.276 0.316 0.324 0.360 0.336 0.239 2 = 2.18 (P>.05, 6 df) r = 0.455 (P>0.05, 5 df)

M 1 2 3 4 5 6 7 Chisquare Correlation 0.9 1.4 2.0 3.0 1.2 1.8 0.45 2 = 119.8 (P<0.01, 6 df) r = 0.883 (P<0.01, 5df)

S2 1.015 1.579 2.692 9.974 1.533 1.856 0.408

2 = 15.13 (P<0.05, 6 df) r = 0.941 (P<0.01, 5 df)

2 = 7.21 (P>0.05, 6 df) r = 0.254 (P>0.05, 5 df)

M = mean population, S 2 = variance, df = degrees of freedom.

Table 2. Population mean and variance of original and transformed counts of Leptocorisa varicornis infesting Pusa 834, 2002 rainy season.a Set 3 Sample M 1 2 3 4 5 6 7 Chisquare Correlation 1.07 1.55 2.02 2.97 1.22 1.75 0.52 Original count (X) S2 1.097 1.843 3.256 9.563 2.692 3.167 1.589 2 = 68.98 (P<0.01, 6 df) r = 0.881 (P<0.01, 5 df) Log (X + 1)-transformed count M 0.261 0.345 0.399 0.482 0.261 0.350 0.108 S2 0.051 0.057 0.079 0.104 0.069 0.083 0.049 2 = 13.14 (P<0.05, 6 df) r = 0.834 (P<0.05, 5 df) X1 b/2 transformed count b = 1.5048 M 0.703 0.876 0.931 1.054 0.668 0.813 0.267 S2 0.314 0.281 0.323 0.337 0.358 0.381 0.261 2 = 2.002 (P>0.05, 6df) r = 0.478 (P>0.05, 5 df)

M = mean population, S 2 = variance, df = degrees of freedom.

value of b greater than 1.0 revealed an aggregated pest distribution on the crop. The mean and variance of original bug counts as well as log (X + 1)-transformed counts in the three sets were highly correlated (Tables 1 and 2). The
IRRN 29.2

logarithmic transformation thus proved ineffective in reducing the dependence of variance on stinkbug mean population. However, the aggregation parameter b-based transformation eliminated the dependence of variance on

mean population as correlation coefficient values between the mean and variance of b-transformed counts in three sets were nonsignificant (r = 0.253, 0.455, and 0.478, respectively). The chi-square values for the homogeneity test of sample variances of original counts in three sets (119.8, 109.2, and 67.9, respectively) were significant, thus indicating that sample variances were not homogeneous (Tables 1 and 2). Likewise, chi-square values for the homogeneity test of sample variances of log (X + 1)-transformed counts in three sets (15.13, 81.75, and 13.14, respectively) were also significant, reflecting nonhomogeneity of the variances. However, chi-square values for homogeneity of sample variances of X 1 b/2 -transformed counts in three sets (7.21, 2.18, and 2.0, respectively) were nonsignificant,
29

showing homogeneity of the sample variances. The aggregation parameter-based transformation of the rice bug field counts thus proved effective in normalizing the sample variances. Often, logarithmic transformation is used for normalizing population variance without caring for its validity, which may produce spurious results. However,

the use of aggregation parameters for the transformation of insect counts will be useful in overcoming this lacuna. References
Gomez KA, Gomez AA. 1984. Statistical procedures for agricultural research. New York (USA): John Wiley & Sons. 680 p.

Ruesink WG, Kogan M. 1982. The quantitative basis of pest management: sampling and measuring. In: Metcalf RL, Luckman WH, editors. Introduction to insect pest management. New York (USA): John Wiley & Sons. p 315-352. Taylor LR. 1961. Aggregation, variance, and mean. Nature 189:732-735.

Weed control in direct, dry-seeded rice in India: comparison of seedbed preparation and use of pendimethalin
S.K. Sharma, D.K. Pandey, K.S. Ganagwar, and O.K. Tomar, Project Directorate for Cropping System Research, Modipuram, Meerut 250110 (Uttar Pradesh), India E-mail: pdcsr@vsnl.com

Under direct seeding in drybed conditions, weeds are a major constraint to rice productivity because rice and weeds germinate almost simultaneously. Grassy weeds are also more difficult to remove by hand because their morphology is similar to that of rice. Weeds have reportedly reduced yield of dryseeded rice by 1560% (Moody 1982). Direct-seeded rice covers 26% and 28% of the total rice area in South Asia and India, respectively (Pandey and Velasco 1999). Thus, appropriate weed control measures will be of immense importance to the farming community. This study aimed to evaluate the effect of weed management practices on the productivity and profitability of aerobic rice, particularly in places where labor is scanty and costly.
30

A field experiment was conducted during the 2001-02 kharif at the project site in Uttar Pradesh on sandy loam soil (Typic Ustochrept). Treatments were laid out in a splitplot design replicated thrice. Variety Pant Dhan 12 was sown at 60 kg seed ha1 under aerobic conditions. Two seedbed practices (stale seedbed and traditional seedbed) were used as the main plot. To prepare the stale seedbed after the rabi crop harvest in April, the field was irrigated once. Two cross harrowings + one cultivator + one planking were then performed. The field was left for germination of weed seeds for about 15 d, after which the field was finally prepared by one shallow plowing with the cultivator, followed by planking. In the traditional method, at sowing time, two harrowings

+ two cultivator + one planking were done to prepare a well-pulverized seedbed. Five weed control methods (hand weeding twice, herbicide + one hand weeding, criss-cross seeding + one hand weeding, criss-cross seeding + one hand weeding + herbicide, and unweeded check) were assigned to subplots. A uniform dose of fertilizer was applied basally 37 kg N, 26.4 kg P, 49.8 kg K, and 4.6 kg Zn ha 1 . Nitrogen was applied in two 53-kg doses at 25 d after sowing (DAS) and at 5560 DAS. The herbicide (pendimethalin 35% EC at 4 L ha 1 ) was sprayed 12 d after seeding. Hand weeding was done at 25 DAS and at 5055 DAS. Weed samples were collected from two places at random in a 1-m2 area and weed dry weight (oven-dry basis) was assessed.
December 2004

Table 1. Effects of weed control on grain and biomass yield, net returns, weed number, and weed dry weight in direct-seeded rice under dry conditions. Grain yield (t ha1) 2001 M1 = stale seedbed M2 = traditional seedbed CD (P = 0.05) T1 = hand weeding twice T2 = pendimethalin + CCSa + one hand weeding T3 = CCS + one hand weeding T4 = pendimethalin + CCS + one hand weeding T5 = unweeded check CD (P = 0.05)
a

Treatment

Biomass yield (t ha1) 2002 2.84 2.5 ns 3.88 4.88 0.00 4.58 0.00 0.58 2001 11.90 10.90 ns 12.00 12.60 11.50 13.30 7.70 2.85 2002 6.59 5.83 ns 9.25 11.16 0.00 10.63 0.00 1.44

Net returns (US$ ha1) 2001 224 163 277 337 172 360 179 2002 23 13 135 280 342 245 295

Weeds (no. m2)

Weed dry weight at flowering (g m2) 2001 209.30 256.77 ns 53.71 11.76 285.23 8.94 805.57 59.68 2002 292.20 387.33 ns 150.84 43.44 495.84 69.67 916.67 175.72

2001 125.47 155.47 22.25 31.50 1.50 172.35 1.50 495.35 21.30

2002 171.93 193.73 ns 54.17 14.67 335.67 32.17 477.50 59.68

4.31 3.84 0.26 4.96 5.32 3.84 5.52 0.73 0.10

CCS = criss-cross sowing, ns = nonsignificant.

The most dominant weeds found in the experimental field were Cyperus rotundus, Echinochloa colonum, Echinochloa crus-galli, Cyperus iria, Eclipta alba, Setaria glauca, Scirpus spp., Monochoria spp., and Celosia argentea. Among the weed species, Celosia argentea and Cyperus rotundus were not satisfactorily controlled by pendimethalin. Regardless of land preparation method, the application of pendimethalin + one hand weeding and criss-cross sowing + one hand weeding + pendimethalin resulted in significantly fewer weeds and less weed biomass at flowering compared with the unweeded check and the treatment with one hand weeding + criss-cross sowing. In the first year, stale seedbed preparation gave significantly higher grain yield than traditional seedbed preparation. Among weed control treatments, pendimethalin + one hand weeding + criss-cross sowing and pendimethalin + one hand
IRRN 29.2

weeding gave similar yields but these were significantly superior to two hand weedings. Biomass yield followed the same trend. Net returns (US$ ha1) were higher in the stale seedbed. Pendimethalin + one hand weeding and pendimethalin + criss-cross seeding + one hand weeding gave similar net returns, but these were significantly higher than two hand weedings. Similar results were reported by Behera and Jena (1998).
Behera AK, Jena SN. 1998. Weed control in direct seeded, rainfed upland rice (Oryza sativa). Indian J. Agron. 43(2):284-290. Moody K. 1982. Weed control in dry seeded rice. In: Proceedings of the Workshop on Cropping Systems Research in Asia. Manila (Philippines): International Rice Research Institute. Pandey S, Velasco L. 1999. Economics of direct seeding in Asia: patterns of adoption and research priorities. Int. Rice Res. Notes 24(2):6-11.

Table 2. Cost of operation and total cost (US$ ha1). Treatment Cost of operation (US$ ha1) 57 68 71 55 Total cost (US$ ha1)a 337 343 366 349

Stale seedbed Traditional seedbed Hand weeding twice Pendimethalin + one hand weeding
a

Covers land preparation ($40 ha 1 ), sowing ($12.67 ha 1 ), insecticide ($22.22 ha 1 ), fertilizer ($77.77 ha1 ), irrigation ($83.33 ha 1 ), harvesting + threshing + cleaning ($53.33 ha1), land revenue ($5.55 ha1), weed control (hand weeding twice at $71 ha1 and herbicide + hand weeding once at $55 ha1).

References

31

A Web-based software for randomization tests of cluster analysis of invertebrate biodiversity in a rice ecosystem
Wenjun Zhang, Runjie Zhang, and Dexiang Gu, Research Institute of Entomology, Zhongshan University, Guangzhou 510275, Peoples Republic of China E-mail: zhangwenjun@scientist.com

As an important analytical tool, cluster analysis is widely used in ecological research e.g., community classification, biological evolution analyses, and biogeographic comparisons (Krebs 1989). Dozens of algorithms for cluster analysis have been developed for common use or special purposes (Zhang and Fang 1982). Most of them, however, do not use appropriate statistical tests in the computation procedures. Thus, we are not able to evaluate statistically the confidence of the classifications in the cluster analysis. Classical statistics can be used to answer the above question when statistical assumptions on data have been met. For example, are the individuals randomly sampled from the population of interest? Do clustered individuals come from different populations or groups that share equal population standard deviations or means? Do the values coincide with a normal distribution or with other known distributions (Manly 1997)? Unfortunately, these assumptions are usually not met in most ecological studies. The randomization method, always featuring fewer statistical restrictions, provides an effective method for testing the statistical significance of ecological data and has been used in ecologi32

cal studies (Manly 1997, Zhang and Schoenly 2001). In this paper, the algorithm of cluster analysis with a randomization test procedure was developed and implemented as the Web-based software StatTestCluster. It contains four procedures: weighting and standardization of the data, calculation of distance measures, hierarchical clustering, and a randomization statistical test. Assume that the data (abundance, occurrence, etc.) for the ith sample and jth taxon are a ij , i = 1,2,...,n; j = 1,2,...,m. Weights can be set for each taxon to denote its relative importance (e.g., in integrated pest management, the more economically important the pest is, the larger weight should be set for it). The weight of the ith taxon is w i , and it should meet w i >0, i=1m wi =1. Thus, the input data will be a ij =w i a ij ,i=1,2,..,n; j=1,2,..., m. The input data can be standardized by mapping them to the range [1,+1] or [0,1] (Zhang and Fang 1982). Data can only be standardized if a quantitative distance measure is chosen. Fourteen distance (or similarity) measures, including measures for continuous values, Euclidean distance, Manhattan distance, Chebyshev distance, correlation coefficient, and angular

cosine; measures for multiple discrete values, linkage coefficient, colinkage coefficients 13; and measures for Boolean values, point correlation coefficient, quadratic correlation coefficient, angular cosines 12, and Jacarrd coefficient, are available for selection by users (Zhang and Fang 1982, Qi and Zhang 2003). At first, each sample makes a class. Five algorithms define between-class dissimilarities: nearest neighbor, farthest neighbor, class average, centroid clustering, and sum of square deviations (Zhang and Fang 1982, Krebs 1989). Between-class distance for these algorithms is respectively defined as the shortest distance between samples, the longest distance between samples, the square root of the average squared distance between two samples in two classes, the distance between weighted centers (each center is the mean of samples in the same class), and the sum of squared deviations. For this last distance measure, assume that the sum of squared deviations for class P and Q is p and q, respectively. Combine P and Q into a new class R, and, if the sum of squared deviations for class R is r, then the distance between P and Q is r-p-q. Select classes P and Q for which this distance is smallest and combine them into a
December 2004

new class, R. For each cycle of clustering, the number of classes will decrease until all samples are combined into the same class. The randomization test procedure is described by the following steps: assume that a class, C, is divided into two classes, A and B, which contain n A and n B samples of class C, respectively. The n A+n B samples of class C randomly reallocate into two classes with nA and n B samples. Calculate the expected distance between the two pseudoclasses and compare whether it is greater than the distance between the classes A and B. Repeat the simulation many times, calculate the number of times the expected distance is greater than the distance between classes A and B, and take the percentage as the p value. The p value is used to make a statistical test. The threshold p value for the test can be defined as 0.1, 0.05, etc. If the calculated p value is less than the p threshold, then class C can be divided into classes A and B with statistical significance; if the calculated p value is greater than the p threshold, then class C cannot be reliably divided into classes A and B. The software StatTest Cluster provides this statistical test of classifications so that users can decide whether the current classification should be trusted or not. The software was programmed in Java Applet and HTML to enable interactive operations on the Internet, to allow easy access to users worldwide, and for use in teaching programs and online research. Before using the software, users must have
IRRN 29.2

the Web browser Java enabled. Algorithm details can be found by clicking on Hint on the Applet window. A window will pop up to show algorithm details, data file format, and references. Users must enter the number of samples, the number of taxa (indices), the number of randomizations (e.g., 100 or 1,000), the threshold of statistical significance level (i.e., 0.1 means a 90% degree of confidence), and specify the cluster data file by clicking on Open Clustering File. A dialog box will pop up waiting for a data file to be chosen. Finally, the software can be run and the results window and cluster tree window will appear after the computation is finished. The cluster data file is a plain text file with an extension .txt. The current classification in the output file can be considered statistically significant if P the threshold significance level. By clicking on Save, users can save cluster results as a plain text file. The cluster tree can be rotated by clicking on Rotate on the cluster tree window. If an error occurs (e.g., the user forgot to enter parameter values), a warning window will appear. In this study, 10 samples of rice invertebrates were collected from Guangdong rice fields and these samples were classified using StatTest Cluster (Euclidean distance, class average, nontaxa weighting, standard deviation standardization, 100 randomizations, a threshold significance level of 0.1). Results

are indicated as follows. The cluster tree (omitted) can be used, but we should focus on the significance of the statistical test on each classification in the following:

In this cluster analysis, as indicated by the P values (labeled by the symbol *), the following classifications are considered to be statistically significant (1 2 4 5 6 7 9 10) (1 2 4 5 6 7) + (9 10), (2 4 7) (2 4) + (7), (1 6) (1) + (6), (2 4) (2) + (4), (9 10) (9) + (10). In classical cluster analysis, P values were not given so we could not determine whether the classification was reliable or not. References
Krebs CJ. 1989. Ecological methodology. New York: HarperCollins Publishers. Manly BFJ. 1997. Randomization, bootstrap and Monte-Carlo methods in biology. 2nd ed. London (UK): Chapman Hall.

33

Qi YH, Zhang WJ. 2003. CorreDetector: a network sharing software for correlation analysis of information and data. J. Chin. Soc. Sci. Tech. Inf. 22:266-268. Zhang YQ, Fang KT. 1982. Introduction to multivariate statistics. Beijing: Science Press.

Zhang WJ, Schoenly KG. 2001. A randomization test and software to compare ecological communities. Int. Rice Res. Notes 26(2):48-49.

Acknowledgment
This project was supported by the National Natural Science Foundation of China, through 30170184; the Foundation of Scientific Research for Personnel from Abroad, through the Ministry of Education of China 2000.

Detection of Xanthomonas oryzae pv. oryzae by NCM-ELISA in naturally infected rice plants
Y. Suryadi, Department of Protein Engineering and Immunology, Indonesian Center for Agricultural Biotechnology and Genetic Resources (ICABGR), Jl. Tentara Pelajar 3A, Bogor 16111, Indonesia; and T.S. Kadir, Indonesian Research Institute for Rice (IRIR), Jl. Raya 9, Sukamandi Subang 41256, West Java, Indonesia E-mail: yshid@yahoo.co.uk

Bacterial blight (BB) caused by Xanthomonas oryzae pv. oryzae (Xoo) is an economically important bacterial disease (Mew 1987). Effective control techniques need to be developed to improve crop production in the country. The spread of BB occurs through plant debris, weeds, water, and seeds. Hence, early detection and identification of the pathogen are crucial in an integrated disease management program. This can be done through seed health testing and plant quarantine. Conventional techniques are currently being used in disease diagnoses, particularly in developing countries. These techniques are still applicable, though they are less effective and are time-consuming. Serological and molecular approaches have shown effectiveness because they are accurate and fast. However, adoption of the latter techniques very much de34

pends on the availability of equipment and affordability of chemicals. Serological assays such as the enzyme-linked immunosorbent assay (ELISA) and its variant need to be developed and made readily applicable for field users. Previous trials indicate that ELISA was effective in detecting plant pathogens such as Ralstonia solanacearum (Smith et al 1995). This study aims to develop a polyclonal antibody (pAb) against Xoo for practical use in BB detection. Naturally infected rice plants were collected and isolated from rice fields in West Java and Central Java during the 2003 planting season. A laboratory experiment was conducted at ICABGR-Bogor and IRIR-Sukamandi. Each sample was stored in a freezer until required. Infected leaves were grown on Wakimoto agar medium for 3 d, adjusted to a concentration

of 109 cfu ml1. Bacterial whole cells of Xoo isolates, fixed with 2% glutaraldehyde, formalin, or untreated antigen, were used to immunize a white New Zealand rabbit. Two intramuscular injections were made on the rabbit with cells of Xoo in a 5-mL saline solution mixed with either the same volume of complete or incomplete Freunds adjuvant. Crude antiserum was pooled and partially purified using ammonium sulfate precipitation following serial dialysis processes using phosphate buffer (pH 7.2). pAb was measured by a spectrophotometer using an absorbance value of 280 and 260, assuming that optical density (OD)280/260 = 1.4 is equivalent to protein of 1 mg mL 1. Collected samples of Xoo were further tested by nitrocellulose membrane (NCM)ELISA following the protocol of Fuentes (1993). One hundred L of whole cells diluted
December 2004

in 0.05 M sodium carbonate buffer (pH 9.6) was directly dropped onto nylon membrane (Hybond N). The membrane was dipped in TBS blocking solution (Tris buffer sulfate, 2% skim milk, and 2% TritonX) following three washings in TBS for 15 min. The membrane was coated in goat antirabbit conjugated with alkaline phosphatase diluted in TBS buffer for 1 h, then it was washed again with TBS for 15 min. After addition of NBT/BCIP substrate, the membrane was incubated for 30 min. The reaction of the color substrate on spotted samples was observed. Successful detection based on serological techniques depends on the specificity and sensitivity of the antibody. pAb Xoo derived from isolate race 4 from Sukamandi (pAb 4) has been evaluated and it revealed a positive reaction against all isolates tested. Two other pAbs (pAb 3, pAb 8) showed weak reactions to Xoo samples (Table 1). The titer of pAb-Xoo produced in this study was

Table 1. Effectivenessa of pAbs against Xoo samples. Xoo sample Polyclonal antibody (dilution 1:2,000) pAb 4 ++ ++ ++ pAb 3 ++ + + pAb 8 + + +

Pure culture Plant sap (leaf extract) Plant sap (seed extract) Control (water)
a

++ = strong reaction, + = weak reaction, = no reaction.

checked by plate ELISA. A high titer of antibody was obtained (as much as 2,048), with protein purity (OD 280/260) ranging from 1.15 to 1.32. The pAb was able to detect Xoo antigen from the crude extract and pure cultures of samples fixed with glutaraldehyde, formalin, or untreated cells. No other cross reaction was observed with other different plant pathogenic bacteria such as Ralstonia solanacearum, Pseudomonas syringae pv. glycinea, or Xanthomonas campestris pv. glycinea. The minimum detectable concentration of Xoo antigen was approximately 10 4 cells mL 1 . The samples containing plant sap and pure cultures of Xoo extracted from diseased plants infected by BB from various locations in East Java showed a positive reaction to pAb-Xoo. This study indicated that pAbs developed in ICABGRBogor, which had been previously used in the detection of rice plant samples infected by Xoo using an indirect ELISA test, gave similar results. Table 2 shows the results of this study for the detection of Xoo by NCM-ELISA in naturally infected rice plants from various hosts and locations. An application of the newly developed pAbs from local Xoo isolates could serve as material stock for the routine detection of Xoo in various parts of the rice plant using serological NCM-ELISA. To further evaluate these pAbs, a similar screening of antibodies to provide cheap and effective pAbs in the region is necessary.

Table 2. Detection of Xoo from naturally infected rice plants by NCM ELISA using pAb 4. Isolate samples Xoo1 Xoo2 Xoo3 Xoo4 Xoo5 Xoo6 Xoo7 Xoo8 Xoo9 Xoo10 Xoo11 Xoo12 Xoo13 Xoo14 Xoo15 Xoo16 Xoo17 Xoo18 Xoo19 Xoo20 Host cultivars Ciherang Cimelati Sintanur Way Apo Buru IR64 Ciherang Ciherang Widas BP 140 Way Apo Buru Hybrid rice IR42 IR64 Memberamo Local variety Local sticky rice IR64 Ciherang Fatmawati Sadang Source (locality)a Ciranjang Ciranjang Muara, Bogor Ciranjang Ciranjang Ciranjang Sukamandi Sukamandi Sukamandi Sukamandi Sukamandi Pagaden Pagaden Pagaden Pagaden Cisalak Cisalak Batang Batang Batang Reaction to pAbb ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++ ++

a All in West Java, except Batang, which is in Central Java. b++ = strong.

Using NCM-ELISA, the pAb developed from Xoo was shown to be effective in detecting BB samples. It could be used for further detection of rice Xoo isolates from a wide locality. References
Fuentes S. 1993. Detection of sweet potato viruses using NCM-ELISA techniques. Lima (Peru): International Potato Center. Mew TW. 1987. Current status and future prospects of research on bacterial blight of rice. Annu. Rev. Phytopathol. 25:359-382. Smith R, Jones AP, Elphinstone JG, Forde SMD. 1995. Production of antibodies against Pseudomonas solanacearum, the causative agent of bacterial wilt. Food Agric. Immunol. 7:67-69.

IRRN 29.2

35

Detrimental effects of niclosamide 250EC at preseeding in direct-seeded rice culture

R.C. Joshi, M.S. Desamito, A.R. Martin, and L.S. Sebastian, Department of Agriculture-Philippine Rice Research Institute (DA-PhilRice), Maligaya, Muoz Science City, Nueva Ecija 3119, Philippines; and J.B. Coupland, Science Research Foundation, University Gate East, Park Row, Bristol, BS 5UB, United Kingdom E-mail: rcjoshi@philrice.gov.ph; joshiraviph@yahoo.com

During the field evaluation of the potential low-dose metaldehyde formulation against golden apple snail (GAS), Pomacea canaliculata (Lamarck), we consistently observed that only microplots treated with niclosamide 250EC had adverse effects on the growth and establishment of direct-seeded rice (DSR). This prompted us to verify field observations under screenhouse tests at the PhilRice-Central Experiment Station in Nueva Ecija, Philippines. Commercial synthetic molluscicides such as niclosamide 250EC and metaldehyde are used among lowland irrigated transplanted rice farmers in Asia. However, niclosamide 250EC is preferred over metaldehyde formulations because of its quick-kill action on GAS. Unfortunately, niclosamide 250EC is lethal to nontarget beneficial water-borne organisms such as frogs, fish, etc. Hence, in the future, it is important for rice farmers to practice ecologically sustainable GAS management options that are safe to the rice crop, their health, and their environment. Healthy seeds of rice variety IR64 were selected using the flotation method. The seeds were soaked in dis36

tilled water for 24 h and then incubated for 12 d in petri dishes lined with Whatman filter paper no. 42. The trays were filled with 1-in-thick soil. Before sowing of seeds, the trays were saturated with water; 25 sprouted seeds were sown per tray (34.3 26.7 11.4 cm), in five rows, each row having five seeds. Each treatment was replicated four times, with a total of 100 seeds. Two GAS of 15 1 mm were introduced per tray after sowing of seeds. All treatments were made as sprays prior to seeding. The following were the treatments: niclosamide 250EC at 6 mL, 12 mL, and 15 mL in 15 L of water. Two kinds of control were maintained that were sprayed with water. The positive control had no niclosamide 250EC treatment but two GAS. The negative control had neither niclosamide 250EC nor GAS. Water from the trays was drained after 24 h. All the experimental conditions mirrored those in DSR farmers fields. After 10 d, data on the number of rice seeds that emerged above the soil, rice seedling shoot and root lengths, phytotoxicity, retarded/abnormal growth of rice seedlings, and snail mortality were recorded.

The most visible serious effects of niclosamide 250EC on rice seedlings were on seedling height (see table and Figs. 24). The seedlings had low and uneven emergence and were stunted when treated with niclosamide 250EC. Belowground effects on rice seedlings were a marked reduction in root growth and development at all concentrations of niclosamide 250EC tested as preseeding treatment (Figs. 1 and 4). In the negative control, rice seedling height and root length were the highest (see table and Fig. 1). No phytotoxicity symptoms were observed in the field or screenhouse tests. Our screenhouse results confirmed the field observations. This has important implications for the establishment of
Effect of niclosamide 250EC on the shoot growth of DSR seedlings, PhilRice-CES, 2004. Treatment 6 mL niclosamide in 15 L water 12 mL niclosamide in 15 L water 15 mL niclosamide in 15 L water Control (+) with GAS and no niclosamide Control () without GAS and no niclosamide
a

Rice seedling height (cm)a (mean S.D.) 5.1 2.1 4.5 2.0 4.3 2.5 10.1 3.5 12.5 3.0

Average of 100 rice seedlings in each treatment.

December 2004

Root length (cm) 14 12 10 8 6 4 2 Control (+) with GAS 6 ml niclosamide in 15 L water 12 ml niclosamide in 15 L water 12 ml niclosamide in 15 L water Control (+) without GAS 0

Fig. 3. Control rice seedlings (without GAS).

Fig. 1. Effect of niclosamide 250EC on root length of DSR, PhilRice-CES, 2004.

Fig. 2. Effects of 15 mL of niclosamide in 15 L of water on germinating rice seedlings in direct-seeded rice.

Fig. 4. Rice shoot and root lengths (cm) in control (untreated) and lengths in treatment with niclosamide 250EC.

a healthy DSR crop. Once farmers establish their crops as DSR and they use niclosamide 250EC as a preseeding treatment, they will experience very poor and

uneven crop establishment. This condition would also expose seedlings to GAS damage for a much longer time. It is therefore important to share this new finding with

DSR farmers in farmer field schools since DSR is increasing in popularity in many Asian countries and in countries where GAS has also become invasive.

IRRN 29.2

37

Predatory behavior of mirid bug, Cyrtorhinus lividipennis, on rice plants with different nitrogen regimes
Z.X. Lu and X.P. Yu, Zhejiang Academy of Agricultural Sciences, Hangzhou 310021, China; and C. Hu, Institute of Applied Entomology, Zhejiang University, Hangzhou 310029, China

The mirid bug, Cyrtorhinus lividipennis, is a plant-feeding and predatory insect, preferring plant- and leafhopper eggs and young nymphs (Shepard et al 1987). It is now one of the important predators of brown planthopper (BPH) Nilaparvata lugens and whitebacked planthopper (WBPH) Sogatella furcifera (Heong et al 1991). Its populations were highly correlated with planthoppers maintaining a high density in the field even in the absence of BPH, but Laba and Heong (1996) found no preference for BPH and WBPH eggs. Tritropic interactions among plants, herbivores, and natural enemies can be affected profoundly by host-plant quality, architecture, and distribution (Boethel and Eikenbary 1986). C. lividipennis population density increased significantly in the field with higher N fertilization through improved nutrient conditions in rice plants (Meerzainudeen and Kareem 1999). Through a series of laboratory experiments, we tried to quantify the influence of N on the predatory behavior of C. lividipennis in rice. Rice plants (IR64) under four N regimes200, 100, 50, and 0 kg ha 1 (labeled 200N, 100N, 50N, and 0N, respectively)were studied. Ammonium nitrate was applied at different rice growth
38

stages. An electronic chlorophyll meter (SPAD 502, Minolta Camera Co., Osaka, Japan) was used to assess tissue N. By pooling the data from all plant growth stages, a relationship between leaf N content (N) and leaf SPAD readings was found to be N% = 0.1151 SPAD 1.2772 (F = 162, P< 0.001). This linear model was used to predict the N content of plants for all the experiments. The insects employed in this experiment were successively mass-cultured on host plants with low (0N) and high (200N) N regimes and were labeled 0NGi and 200NGi for BPH and 0NFi for and 200NFi for mirid bug, respectively (i is the number of successive generations on rice plants with the same N regime). Gravid BPH females were confined in a mylar cage on 45-d-old rice plants (main stem and one primary tiller per pot) with four N regimes for oviposition for 24 h. Four potted rice plants with four N regimes (200N, 100N, 50N, and 0N) were arranged randomly in a plastic tray covered by a cylindrical plastic cage (height, 80 cm) with fine nylon mesh on top and one 3cm-long window on the side wall for insect infestation. Seventy-two hours after infestation by five newly molted mirid bug females confined in a plastic tray, the healthy and

attacked BPH eggs were counted by dissecting the leaf sheaths and midrib tissues under a binocular microscope. To determine the predatory preference for BPH eggs on the different parts of the rice plant, we attached the sachets on the leaf sheath and leaf blade at 5 cm above the soil and at 3 cm from the collar for oviposition by gravid BPH females. The predatory capacity of mirid bug on BPH eggs was qualified by detecting the number on BPH eggs consumed by one newly emerged female adult in 24 h under different egg densities on rice plants with low and high N regimes. Cyrtorhinus lividipennis can complete its life cycle on the rice plant (Yu et al 1996). However, its predatory performance was positively affected by the N content of the host plant. Rice plants with higher N content were apparently softer, with more succulent tissues than in plants with lower N content. Female adults preferred to attack the BPH eggs in plants with higher N content than in plants with low N in both the greenhouse and darkroom (P<0.0001) (Fig. 1). They also preferred feeding on BPH eggs in the midrib of the rice leaf blade (P<0.0001). Very few eggs laid in the leaf sheath were attacked on both the 0N and 200N rice plants. FurtherDecember 2004

No. of eggs attacked 15

R2 = 0.7111 P<0.0001

15 A 10
R2 = 0.3261 P<0.0036

10

0 1.0 15

1.5
R2 = 0.4734 P<0.0002

2.5

3.0 C

3.5

0 1.0 15

1.5
R2 = 0.6365 P<0.0001

2.5

3.0 D

3.5

10

10

0 0.8

1.6

2.4

3.2

4.0

0 0.8

1.6

2.4

3.2

4.0

Nitrogen content in rice plants (%)

Fig. 1. Number of BPH eggs attacked by C. lividipennis populations on rice plants with different N contents. A: 200NF2 population in the greenhouse. C: 200NF2 population in the darkroom. B: 0NF2 population in the greenhouse. D: 0NF2 population in the darkroom.

more, the number of BPH eggs attacked in the leaf blade was markedly higher on the 0N rice plants than on the 200N rice plants (P = 0.0245). However, no significant differences in predatory preference were found between 0N and 200N rice plants (P = 0.0686), while a markedly higher predatory rate in total eggs, more than twice, was recorded on the 0N plants than on the 200N plants (P<0.0001). The functional responses of C. lividipennis populations reared with BPH eggs on rice plants with low and high N fertilizers were typically Hollings Type II (Fig. 2). The bugs preying on BPH eggs on 0N rice plants had significantly higher attack rates than those on 200N rice plants.
IRRN 29.2

Meanwhile, no observable differences in handling time were found between bug populations. These suggest that the high N fertilizer decreases the predatory capacity of C. lividipennis on BPH eggs. IR64 had changes in morphological and histological structures, physiological properties, and chemical components, such as a softer and thicker sheath, more succulent plants, higher N content in the sap, and bigger canopy size, leading to modified volatiles and other chemical and visual cues after application of N fertilizer. Such modifications resulted in the enhanced attractiveness of the rice plants to mirid bugs. Similar results obtained from the greenhouse and the dark-

room proved that increased predatory preference for BPH eggs by mirid bugs was attributed mainly to the many cues derived from rice plants with higher N content. A possible explanation for the great number of eggs attacked by mirid bugs on high-N rice plants may be that Cyrtorhinus, being plant feeders and hopper predators, seek host plants with higher nutrient contents. Compared with the leaf sheath of low-N plants, the tissues of the midrib in leaf blades are softer and more suitable for feeding by mirid bugs, so hopper eggs on the midrib in leaf blades are attacked easily. This is why significantly more BPH eggs on midribs of leaf blades were consumed. Unfortunately,
39

No. of eggs attacked 40 30 20 10 0

100

200

300

400

100

200

300

400

Density of BPH eggs

Fig. 2. Functional responses of female C. lividipennis adult populations to BPH eggs on rice plants with different N regimes. A: 0NF2 C. lividipennis on 0N rice plants. B: 200NF2 C. lividipennis on 200N rice plants.

this characteristic does not make the mirid bug an excellent biological agent because it contributes to lower predatory effectiveness on target eggs in the field and in nochoice experiments with a high N regime. The bigger size and more sources of cues could mislead the orienting, foraging, accessing, and locating of mirid bugs in nontargets or it could extend the range to find targets. Based on this study, the predatory capacity of mirid bugs can be enhanced on host plants treated with high-N fertilizers.

References

Boethel DJ, Eikenbary RD. 1986. Interactions of plant resistance and parasitoids and predators of insects. Chichester (UK): Ellis Horwood Limited. 224 p. Heong KL, Aquino GB, Barrion AT. 1991. Arthropod community structures of rice ecosystems in the Philippines. Bull. Entomol. Res. 81:407-416. Laba IW, Heong KL. 1996. Predation of Cyrtorhinus lividipennis on eggs of planthoppers in rice. Indonesian J. Crop Sci. 11(2):40-50. Meerzainudeen M, Kareem AA. 1999. Effect of biofertilizers on BPH (Nilaparvata lugens) and its biocontrol agents in rice ecosystem. In: Vistas of rice research. Tamil Nadu Agricultural University, India. p 469-474.

Shepard MB, Barrion AT, Litsinger JA. 1987. Friends of the rice farmer: helpful insects, spiders, and pathogens. Manila (Philippines): International Rice Research Institute. Yu XP, Heong KL, Hu C. 1996. Effect of various nonrice hosts on the growth, reproduction, and predation of mirid bug, Cyrtorhinus lividipennis Reuter. In: Heong et al, editors. Rice IPM Conferenceintegrating science and people in rice pest management. Manila (Philippines): International Rice Research Institute. p 56-63.

Neural network approximation of sampling yield-effort curves of rice invertebrates

Wenjun Zhang and Yanhong Qi, Research Institute of Entomology, Zhongshan University, Guangzhou 510275, Peoples Republic of China; and A.T. Barrion, formerly of the Entomology and Plant Pathology Division, IRRI E-mail: zhangwenjun@scientist.com

Biodiversity studies in ecology and agriculture often begin with the analysis of sampling curves. To measure the completeness of sampling, a sampling yield-effort curve
40

can be drawn that plots the number of taxa sampled against the sample size (Colwell and Coddington 1994, Zhang and Schoenly 1999a,b). This curve is a step

function with a slope that should decrease as sample size increases and as fewer taxa remain to be sampled. Many models or methods were developed to fit these
December 2004

functions. Most of these models, however, yielded fixed errors. Neural network methods, always with any desired accuracy, were widely used to fit functions in engineering and related research (Bian and Zhang 2000, Hagan et al 1996, Zhang and Qi 2002). Therefore, we expect a better goodness of fit for sampling yieldeffort curves with neural networks. The back propagation (BP) neural network and the radial basis function (RBF) neural network algorithms are introduced and tested in this study to provide an effective tool to fit the sampling yield-effort curves and to document the sampling information, based on data sets of invertebrates sampled in tropical irrigated rice fields. Some conclusions on rice invertebrates are obtained from the fitted functions of neural networks. Four sampling data sets of rice invertebrates were sampled on the IRRI farm at different dates, which were represented as Mar18, Apr15, Sep17, and Oct08, each with 60 suction samples. The taxabased data sets were lumped into families using biodiversity software LUMP (Schoenly and Zhang 1999) and bootstrapped with bootstrap methods (Colwell and Coddington 1994, Zhang and Schoenly 1999a,b). The BP neural network is a multilayer architecture (Bian and Zhang 2000, Hagan et al 1996). For the two-layer BP network, as used in this report, the transfer function of the neuron in the hidden layer is a sigmoid functionuj = 1/ (1 + exp (Bx)), j = 1,2,,N
IRRN 29.2

and the transfer function of the neuron in the output layer is a linear function. The Levenberg-Marquardt rule was used to train the twolayer BP network. It was developed and trained to fit function and make pattern recognition and extrapolation. Two learning procedures are included in BP network training. The first one is the positive propagation process in which the input signal is transferred layer by layer and practical outputs of every neuron are computed. The second is the back propagation in which the errors between practical and expected outputs are progressively computed layer by layer, and weights are adjusted according to the errors. The RBF network is made up of two layers, a hidden layer of radial basis neurons and an output layer of linear neurons (Bian and Zhang 2000, Hagan et al 1996). The transfer functions of neurons in the hidden layer are radial basis functions, typically Gaussian kernel functions: u j = exp((x c j ) 2 /(2 j 2 )), j = 1,2,,N, where u j is the output of the jth hidden neuron, x is the input, cj is the centralized value of the Gaussian function, j is the standardized constant, and N is the number of neurons in the hidden layer. The weights and biases of each neuron in the hidden layer define the position and width of a radial basis function. The linear output neuron forms a weighted sum of these radial basis functionse.g., linear output y is the linear combination of outputs of hidden neurons: y = Nj

= 1

w j u j . With the correct weight and bias values for each layer and enough hidden neurons, a radial basis network can fit any function with any desired accuracy. There are two stages in the learning procedure of the RBF network. The first step is determining the centralized values c j and standardized constants j according to inputs, and the second step is solving the weights w j of the output layer based on the least square sum principle. The BP and RBF networks are trained with bootstrapped taxon data of four sampling sets. Results revealed that both networks fit the taxa richness-sample size curves well (Fig. 1). The BP architecture with five neurons is enough to fit these functions, but a longer time was required to reach the desired accuracy. An asymptote was ideally obtained by extrapolation of the trained BP network. RBF networks were quickly trained with the desired accuracy, but more neurons (60) were required, although the number of neurons was automatically produced by the RBF network. The required extrapolation of taxa richness-sample size function cannot be obtained from the trained RBF network. Similar results were produced for family richness-sample size relationships. The total number of rice invertebrate taxa for the four sampling sets Zmar18, Zapr15, Zoct08, and Zsep17, extrapolated from the asymptote of the trained BP network function, was 140, 149, 147, and 144, respectively, while the observed taxa rich41

42

December 2004

ness was 126, 141, 140, and 131, when sample size was 60. The total number of invertebrate families was 69, 73, 80, and 77, extrapolated from the asymptote of the BP network function, as compared with the observed family richness of 66, 71, 75, and 75 in a sample size of 60. For comparison, we tested the neural network algorithms with a mathematical function, the Arrhenius b model, N = aS , where N is the number of taxa when sample size is S. The results showed that, with the least square sum error and mean square error, the BP network and RBF network fit the curves much better than did the mathematical function (see table). Similar results were obtained in the analysis of family richness. The BP network and RBF network can fit a nonlinear function with any desired accuracy. In addition to the interpolation, the BP network can extrapolate the function and thus the asymptote of the yield-effort curve can be drawn. It will take longer for the BP network to train the network and the results will always be less stable than those of the RBF network. The RBF network will require more neurons to fit the functions and, generally, it cannot be used to extrapolate the function. Compared with mathematical functions, both networks are much better in fitting the sampling curves and documenting sampling information. The total numbers of rice invertebrate taxa for the four sampling sets, extrapolated from the trained BP network, were between

Comparison of goodness of fit to taxa richness-sample size curves derived using various methods. BP extrapolated taxa Mar18 Apr15 Oct08 Sep17
a

References

BP fitted MSEa 0.0220 0.3270 0.0090 0.0110

RBF fitted SSEb 0 0 0 0

Arrhenius fitted MSEa 5.2722 25.5010 17.4749 15.4121

140 149 147 144

MSE = mean square error. bSSE = square sum error.

140 and 149. Numbers of invertebrate families extrapolated from the BP network are between 69 and 80. An appropriate number of neurons should be specified in training the BP network. Excessive neurons will make the BP network overlearned; thus, the resultant extrapolation will not be reliable. However, function approximation with the desired accuracy will not be achieved when fewer neurons are used in the BP network. Unlike most other models with the property of self-learning, which allows the network to find any intrinsic mechanism (which is invisibly stored in the trained network) uncovered in sampling data, no assumptions are needed to fit sampling yield-effort curves using the BP network and RBF network.

Bian ZQ, Zhang XG. 2000. Pattern recognition. 2nd ed. Beijing: Tsinghua University Press. Colwell RK, Coddington JA. 1994. Estimating terrestrial biodiversity through extrapolation. Phil. Trans. Roy. Soc. London B 345:101-108. Hagan MT, Demuth HB, Beale MH. 1996. Neural network design. PWS Publishing Company. Zhang WJ, Schoenly KG. 1999a. IRRI Biodiversity Software Series. II. COLLECT1 and COLLECT2: programs for calculating statistics of collectors curves. IRRI Technical Bulletin No. 2. Manila (Philippines): International Rice Research Institute. 15 p. Zhang WJ, Schoenly KG. 1999b. IRRI Biodiversity Software Series. IV. EXTSPP1 and EXTSPP2: programs for comparing and performancetesting eight extrapolation-based estimators of total taxonomic richness. IRRI Technical Bulletin No. 4. Manila (Philippines): International Rice Research Institute. 13 p. Zhang WJ, Qi YH. 2002. Functional link artificial neural network and agribiodiversity analysis. Biodivers. Sci. 10(3):345-350.

Acknowledgments
This project was supported by the National Natural Science Foundation of China, through 30170184, and the Asian Development Bank, through RETA 5711. We thank the following people who participated in the rice invertebrate investigation, identification, and data sheet processing: I. Domingo, V. Magalit, L. Datoon, A. Rivera, E. Hernandez, E. Revilla, E. Rico, R. Abuyo, J. Reyes, B. Aquino, and A. Pulpulaan.

Between-habitat movement of rice arthropods and its ecological role

Wenjun Zhang and Yanhong Qi, Research Institute of Entomology and School of Life Sciences, Zhongshan University, Guangzhou 510275, Peoples Republic of China E-mail: zhangwenjun@scientist.com

Natural enemies of rice insect pests harbored in heterogeneous habitats around rice fields have been identified as the natural agents for sustainable rice pest management for a long time (Way and Heong 1994). However, the movement of rice arthropods must be investigated to analyze the sensitivity of the natural enemies in tracking insect pests. In this study, the movement of rice arthropods along a rice transect is detected by using an algorithm for detecting boundaries of ecological transects in order to find and confirm mechanisms for natural and sustainable control of rice insect pests by their natural enemies. A transect along the eastwest direction and traversing 6 ha of rice fields and uncultivated habitats on the IRRI upland farm was established (Fig. 1). Transect sites were spaced at approximately 5-m intervals and were sampled for arthropods and vegetation every 2 wk in the growing season, using a suction sampler with a 0.16-m 2 enclosure and pin-frame sampler, respectively. Edges in the landscape were distinguishable as roads (o), borders between rice and natural vegetation (rw), rice bunds (b), and rice-rice bund-rice (rbr), etc. Laboratory sorting treated different life stages (immatures, adults) separately. Data from the
IRRN 29.2

records were stored as a siteby-taxa matrix. Nine transects were obtained for arthropod categories and sampling dates (predaceous, parasitoid, and herbivorous arthropods, each with three transects at three sampling dates). An improved algorithm (to be published) was used to detect boundaries of transects in this study. More boundaries could be detected by using the improved algorithm compared with a previous one (Cornelius and Reynolds 1991, Zhang and Schoenly 1999). We set the parameter values in the algorithm as follows: distance measure = Euclidean distance, minimum window width=2, step length of window width=2, randomizations=5,000, and confidence level=99%. Pooled results are analyzed here. The results showed that the boundaries, the direction of movement, and the speed of movement of rice arthropods can be sensitively detected. Herbivorous arthropods were more sensitive to the visible boundaries such as roads, rice bunds, etc., compared with parasitoids and predaceous arthropods (Figs. 1 and 2). It seems that the herbivorous arthropods, with a speed of 512 m d 1, move by 200 m in the 36th day after transplanting. However, the boundaries of parasitoids

were even found across the entire transect on the 8th day after transplanting, and the boundaries of predaceous arthropods were almost dispersed across the entire transect on the 22nd day after transplanting (Fig. 2). We may find from the dynamic distributions of boundaries that, in response to the movement of herbivorous arthropods, parasitoids and predaceous arthropods moved in the opposite direction along the transect (Fig. 2). The spatial distribution of herbivorous arthropods is largely dependent on the distribution of visible boundaries in the transect compared with parasitoids and predaceous arthropods. Herbivorous arthropods move more slowly than did parasitoids and predaceous arthropods across different habitats. Moreover, parasitoids move faster than predaceous arthropods as a whole. Parasitoids and predaceous arthropods can move in the opposite direction against the movement of herbivorous arthropods along the transect. It can be concluded that, in heterogeneous habitats, the parasitoids and predaceous arthropods may quickly detect and track the herbivorous arthropods and could significantly control the population of herbivorous arthropods. It will be effective to maintain
43

Fig. 1. A rice transect with 51 sites at 5-m intervals. Edges are distinguishable as roads (o), borders between rice and natural vegetation (r-w), rice bunds (b), and rice-rice bund-rice (rbr), etc.

Fig. 2. Distributions of boundaries for arthropod transects with different trophic levels and growing periods. Minimum window width = 2, step length of window width = 2, randomizations = 5,000, confidence level = 99%. Pooled results for Euclidean distance measure are listed.

44

December 2004

diverse habitats around the rice field to provide harbors for various natural enemies. References
Cornelius JM, Reynolds JF. 1991. On determining the statistical significance of discontinuities within ordered ecological data. Ecology 72:2057-2070. Way MJ, Heong KL. 1994. The role of biodiversity in the dynamics and management of insect pests of tropical irrigated rice: a review. Bull.

Entomol. Res. 84:567-587. Zhang WJ, Schoenly KG. 1999. IRRI Biodiversity Software Series III. BOUNDARY: a program for detecting boundaries in ecological landscapes. IRRI Technical Bulletin No. 3. Manila (Philippines): International Rice Research Institute. 17 p.

dation of China, through 30170184. We are grateful to the following people who participated in the rice arthropod investigation, identification, and data sheet processingA.T. Barrion, I. Domingo, V. Magalit, L. Datoon, A. Rivera, E. Hernandez, E. Revilla, E. Rico, R. Abuyo, J. Reyes, B. Aquino, and A. Pulpulaan.

Acknowledgments
This project was supported by the Asian Development Bank, through RETA 5711, and the National Natural Science Foun

A generator of a web map of biological interactions for rice invertebrates

W.J. Zhang, H.X. Wu, and Y.H. Qi, Research Institute of Entomology, Zhongshan University, Guangzhou 510275, Peoples Republic of China E-mail: zhangwenjun@scientist.com

In a rice habitat, dozens, hundreds, or even thousands of invertebrate taxa interact with each other. A complex web of biological interactions is thus generated. If we assume that 60 taxa are found in a rice habitat and they interact with each other, a total of 1,770 pathways will be generated on a web map. It is obviously impractical to draw this map by hand. However, we have not found a fast tool to generate the web map until now. For this reason, we developed a network software to automatically generate a web map of biological interactions (direct or indirect) for rice invertebrates. A web map can be drawn by a computer with investigated data or pathway information included in an html file. Fourteen distance measures can be chosen for generating pairwise taxa with significant parametric or nonparametric correlations on the basis of randomization tests
IRRN 29.2

(Manly 1997, Qi and Zhang 2003, Zhang and Schoenly 2001) or classic statistical tests (Zhang and Fang 1982). Two choices are provided to generate web pathways: (1) the first is to compute significant correlations between taxa and (2) the second is to include the pathway information in the html file. If users choose to compile the pathway information in the html file, for instance, the following sentence should be included inside the tags <applet></applet>, which are a complete string of a comma-delimited list of all pairwise taxa or habitat-taxa: <param name=lines value=Rice Field-Coleoptera/150,Rice FieldDiptera/150,Rice FieldOdonata/150,Rice FieldHemiptera/150,Rice FieldHymenoptera/150,Rice Field-Araneae/150,Rice Field-Ephemeroptera/ 150,Rice Field-Lepidoptera/

150,Rice Field-Cyproida/ 150,Rice Field-Cyclopoida/ 150,Rice Field-Arthropleona/ 150,Rice Field-Orthoptera/ 150,Rice Field-Acari/150,Rice Field-Mesogastropoda/ 150,Rice Field-Symphypleona/150,RiceField-Thysanoptera/150,Rice Field-Blattodea/150,ColeopteraCyclopoida/400,DipteraHemiptera/400, DipteraCyclopoida/400, HemipteraSymphypleona/400,Hym e n o p t e r a - A c a r i / 400,Arthropleona-Acari/ 400"> In this web information, Rice Field is the habitat name and should be linked to every taxon. The pathway lengths are 150 and 400 for different pathways. A delimited text file should be prepared to be used by the algorithm. In this data file, the first row will be sample ID numbers and the first column will be taxon names. The values are the in45

vestigated individual numbers. Fourteen distance (or similarity) measures, including measures for continuous valuesEuclidean distance, Manhattan distance, Chebyshev distance, correlation coefficient, and angular cosine; measures for multiple discrete values, linkage coefficient, co-linkage coefficients 13; and measures for Boolean values, point correlation coefficient, quadratic correlation coefficient, angular cosines 12, and Jacarrd coefficientshall be available to users (Zhang and Fang 1982, Qi and Zhang 2003). The pairwise taxa with significant correlations will be calculated by the algorithm (Qi and Zhang 2003), and used to construct web information. The Pearson correlation and partial correlation will be tested with the classic t test and the other measures will be tested with randomization tests.

The software was programmed in Java Applet and HTML to enable interactive operations on the Internet and to allow easy access to users worldwide. It can be used in teaching programs and online research. Before using the software, users must have the Web browser Java enabled. Algorithm details can be found by clicking on Hint on the Applet window. A window will appear to show algorithm details, data file format, and references. If users have chosen to produce web information by computation, they must choose the distance measures, enter the number of taxa and the number of samples, and, if a distance measure other than the Pearson correlation or partial correlation is chosen, users must also enter the number of randomizations (e.g., 100 or 1,000) following the distance measure selected. The re-

quired statistical significance level (i.e., 0.01, which means a 99% degree of confidence) and habitat name should also be entered. Specify the data file by clicking on Open Data File. A dialog box will appear, waiting for a data file to be chosen. Finally, the software can be run and the results window and web map window will be shown after the computation is finished. If an error occurs (e.g., the user forgot to enter parameter values), a warning window will appear. All these operational procedures are easy to follow (Fig. 1). In the web map window, users can drag the taxon box with the mouse to find all links to this taxon. As an example, we use a rice invertebrate data file with 75 families and 60 samples. The automatically generated web map is shown in Figure 2.

Fig. 1. Window for input parameters.

46

December 2004

Fig. 2. Web map window.

References

Manly BFJ. 1997. Randomization, bootstrap and Monte Carlo methods in biology. 2nd ed. London (UK): Chapman & Hall. Qi YH, Zhang WJ. 2003. CorreDetector: a network sharing software for correlation analysis of information and data. J. China Soc. Sci. Tech. Info. 22(Suppl.):266-268.

Zhang YQ, Fang KT. 1982. Introduction to multivariate statistics. Beijing: Science Press. Zhang WJ, Schoenly KG. 2001. A randomization test and software to compare ecological communities. Int. Rice Res. Notes 26(2):48-49.

Acknowledgment
This project was supported by the National Natural Science Foundation of China, through 30170184, and the Foundation of Scientific Research for Personnel from Abroad, through the Ministry of Education of China 2000.

Pathogenic variability of rice blast isolates in Nepal

B. Chaudhary, National Rice Research Program, Hardinath, Nepal; and S.M. Shrestha and R.C. Sharma, Institute of Agriculture and Animal Science, Rampur, Nepal

Blast, caused by Pyricularia grisea Sacc., is a major disease of rice (Oryza sativa L.) in Nepal. It causes seedling loss (Chaudhary and Sah 1998) and adversely affects plant growth and yield (Chaudhary 1999).
IRRN 29.2

The fungus is highly variable, making disease management complicated. This study was undertaken to understand the variability of blast isolates in order to deploy resistance genes properly.

Diseased leaves were collected on diverse genotypes from different locations. Twenty-five monoconidial isolates were obtained and maintained in agar slants during 2000 following the proce47

dures of Bonman et al (1986). Isolates were multiplied on prune agar plates and inoculated at 10 5 mL 1 on 35-leaf stage seedlings of 10 differential varieties in a greenhouse in 2001 (Table 1). The 10 seedlings of each differential, replicated thrice, were raised in aluminum trays, filled with farmyard manure, and fertilized with 150-22-0 kg NPK ha 1 . After inoculation, the tray was kept in a cage covered with water-soaked jute bags at 25 C for 72 h. High relative humidity was maintained. Disease was scored on the 7th day of inoculation using a 05 scale and isolates were grouped into different

pathotypes. The frequency of virulent phenotypes (FVP) was estimated using FVP = total number of compatible reactions/number of isolates used for inoculation. Relative virulence frequency (RVF), the sum of the compatibility frequencies of all isolates from one cultivar on each of the 10 differentials, was also calculated. The 25 isolates resulted in 15 pathotypes (Table 1). The most virulent isolates from Lumle (1,3001,500 m asl)N22-1L, N22-2L, N22-4L, KJ-1L and KJ-3Linduced compatible reactions on 9 out of 10 differentials. Isolates from Khumaltar (1,360

m) were more virulent than those from terai and inner terai. A majority of the isolates represented the same pathotype within the same geographical and host origin, suggesting location- and genotype-specific virulence. The FVP ranged from 0.0 to 1.0, showing Laxmi to be incompatible with all isolates (Table 2). FVP was 0.2 on the pyramids of Pi-1 + Pi-2, Pi-1 + Pi-2 + Pi-3, and Pi-1 + Pi-2 + Pi4, indicating that Pi-1 + Pi-2 was as effective as the threegene pyramids. The other pyramids, Pi-2 + Pi-3 and Pi-1 + Pi-4, were compatible with 32% and 52% of the isolates, respectively. The unusually

Table 1. Reactions of Pyricularia grisea on rice differential genotypes evaluated in the greenhouse at Rampur, Nepal, 2001. Isolate 1 K59-1L N22-1L N22-2L N22-3L N22-4L KP1-L KP-2L KJ-1L KJ-3L NR56-11H CI-1R Tsu-2L K59-5L KP-3L NR256-2R Tc-4Kh Tc-7Kh Tc-10Kh Tc-5Kh K59-4L SM-1R Ma-1R Jh-2R Kan-1R Dul-1L
a

Test host genotypes 2 R S S MR S R MR S S R R R R R R R R R R R R R R R R 3 MR S S S S S S S S MR R S S S R S R R R S R R R R R 4 MR S S S S MR S S S R R MR R R R R R R R S R R R R R 5 MR S S MR S MR MR S S R R MR R R R R R R R MR R R R R R 6 R S S MR S MR MR S S R R MR R R R R R R R MR R R R R R 7 S S S S S S S S S S S S S S S S S S S S S S S S S 8 R S S MR S S S S S R R S R S R S S S S S R R R R MR 9 S S S S S S S S S S S S S S S S S S S S S S S S S 10 R R R R R R R R R R R R R R R R R R R R R R R R R

Host

Pathotype

Sc S S S S S Patong S Patong S S R R MR MR S Patong R S R S S S R R R R S

K59 N22 N22 N22 N22 Kinandang Kinandang Kalo Jangali Kalo Jangali NR1556 CI5309 Tsuyuake K59 Kinandang NR10293 Taichung 176 Taichung 176 Taichung 176 Taichung 176 K59 Sona Masuli Masuli Jhingling78 Kato51 Dular

I II II III II IV V II II VI VII VIII IX X VII X XI XII XII XIII VII VII VII VII XV

L = Lumle (Lumle Agricultural Research Station, 1,300-1,500 m); R = Rampur (Agricultural Research Station, Rampur, 228 m); Kh = Khumaltar (Disciplinary Divisions, Kathmandu, 1,360 m); H = Hardinath (National Rice Research Program, Hardinath, 93 m). b1 = C105TTP-4L23 (Pi-4 + Pi-?); 2 = BL121 (Pi-1 + Pi-2); 3 = BL142 (Pi-1 + Pi-4); 4 = BL23-181-2 (Pi-2 + Pi-3); 5 = BL123-56-2 (Pi-1 + Pi-2 + Pi-3); 6 = A57-115-8 (Pi-1 + Pi-2 + Pi-4); 7 = CO 39 (Pi-a?); 8 = Lijiangxintuanheigu; 9 = Masuli (Mayang Ebos 80*2/Taichung 65); 10 = Laxmi (IR833-6-1//IR1561-149-1/IR1737). cLesion type measured on a 05 scale: 02 = resistant (R), 3 = moderately resistant (MR), and 4-5 = susceptible (S).

48

December 2004

Table 2. Frequency of virulent phenotypes in 25 Pyricularia grisea isolates collected in 2000 at Lumle, Rampur, Khumaltar, and Hardinath, Nepal, and tested in 2001. Genotype C101TTP-4L23 BL121 BL142 BL23-181-2 BL123-56-2 A57-115-8 CO 39 Lijiangxintuanheigu Masuli Laxmi
a

Relative virulence frequency 10

Known resistance genes Pi-4 + Pi-1 + Pi-1 + Pi-2 + Pi-1 + Pi-1 + Pi-a?a Pi-? Pi-2 Pi-4 Pi-3 Pi-2 + Pi-3 Pi-2 + Pi-4

Frequency of virulent phenotypes 0.60 0.20 0.52 0.32 0.20 0.20 1.00 0.56 1.00 0.00
8

Unknown Unknown

Unknown.

lower FVP on Lijiangxintuanheigu (LTH) could mean the presence of avirulence genes for LTH in the Nepalese pathogen population. The RVF of isolates from Kalo Jangali (Oryza rufipogon) and N22 was markedly higher than that of the other isolates (see figure). Isolates recovered from Kinandang Patong, K59 (Pi-t), Taichung 176, Tsuyuake (Pi-K m), and Dular (Pi-K) had an intermediate RVF. The isolates from the rest of the genotypes had a low RVF. The results indicate that the Nepalese blast pathogen populations varied in virulence. The compatibility of all

isolates with Masuli and CO 39 indicates that these genotypes, either in mixture or alone, could be used for spreader rows to build up inoculum in the blast nursery. The most virulent isolates from O. rufipogon and N22 could be useful for screening rice germplasm in the greenhouse. The incompatibility of the isolates with Laxmi suggests that this variety could be a potential donor parent in breeding for blast resistance. References
Bonman JM, Vergel de Dios TI, Khin MM. 1986. Physiologic specialization of Pyricularia oryzae in the Philippines. Plant Dis. 70:767-769.

KJ

KP

Tc Du Cl Host of origin

SM

Jh

Relative virulence frequency of Pyricularia grisea isolates recovered from 14 rice genotypes (KJ = Kalo Jangali, N22, KP = Kinandang Patong, K59, Tc = Taichung 176, Tsu = Tsuyuake, Du = Dular, NR56 = NR1556, CI = CI5309, NR256 = NR10293, SM = Sona Masuli, Ma = Masuli, Jh = Jhingling78, Ka = Kanto51) during 2000 and tested during 2001.

Chaudhary B. 1999. Effect of blast disease on rice yield. Nepal Agric. Res. J. 3:813. Chaudhary B, Sah DN. 1998. Efficacy of Beam 75 WP in controlling leaf blast disease at the seedling stage of rice. Nepal Agric. Res. J. 2:42-47.

Brown planthopper injury vs mechanical injury to the rice plant

E. Rubia-Sanchez, T. Watanabe, and Y. Suzuki, National Agricultural Research Center, 3-1-1 Kannondai, Tsukuba 3058666, Japan

Insect pests cause injury to the rice plant in various ways. The brown planthopper sucks assimilates from the phloem (Sogawa 1982) and reduces plant height and organ dry weights (Rubia-Sanchez et al
IRRN 29.2

2003). Mechanical injury is the removal of leaves or tillers using scissors. The manner by which injury is inflicted may vary, but the response of the rice plant to injury may be similar. Hence, it is important

to compare the effects of BPH injury with that of mechanical injury to better understand plant response to BPH feeding. We chose two cultivars, Nipponbare, which belongs to japonica rice, and Taichung
49

Native I (TN1), which belongs to indica rice, to determine if there is a difference between the two cultivars in plant response to injury. Twenty-six-day-old Nipponbare and TN1 were subjected to the following treatments in the glasshouse (25/20 C day/night T, 14L:10D photoperiod, and 70% RH): T1control (healthy intact plant), T2cutting one primary tiller, T3cutting the two lowest leaves of the main tiller, T4cutting all the leaves of the main tiller, T5 cutting two upper leaves of the main tiller, T6exposing the plant to 16 fourth-instar BPH for 10 d, and T7cutting the main tiller. Except for T1, all treatments were applied 10 d after the introduction of BPH (T6). The experiment was arranged in a randomized complete block design with four replications. Two weeks after treatment, plant height, number of tillers, leaf area, and plant dry weights were recorded. Data were analyzed using ANOVA and the control means were compared with the other treatment means using Dunnetts test.

Unlike with removal of some leaves, BPH-infested plants did not recover from BPH injury 2 wk after treatment (Table 1). BPH feeding decreased plant height, leaf area, and leaf, stem, and root dry weights of the two cultivars. Similarly, the removal of the main tiller also decreased plant growth of the two cultivars (Table 1). There seems to be a similarity in plant response to injury between BPH-infested plants and plants without the main tiller, even though the method by which plants were injured differed. Removal of the main tiller took out a huge amount of assimilates at one time, whereas, in the case of BPH injury, the assimilates were slowly taken out of the plant. For BPH-plant interaction, production of new tillers was not an immediate response by rice plants to BPH injury (Table 2). For stem borer-plant interaction, plants compensate for stem borer injury by producing new tillers (Rubia-Sanchez et al 1996). Two weeks after treatment, IR64 had a 23% increase in the number of tillers in re-

sponse to stem borer injury, whereas TN1, another indica cultivar, had an 18% decrease in the number of tillers in response to BPH injury. Although Rubia-Sanchez et al (1999) showed that healthy tillers could translocate assimilates to BPH-injured tillers, the present study suggests that the amount of assimilates may not be sufficient to produce more tillers 2 wk after BPH injury. Failure to produce new tillers in response to BPH injury may be due to the effect of BPH feeding on root growth. BPH injury decreased root dry weight (Table 1). The roots transport nutrients and assimilates to form new tillers. Regardless of the treatment applied, BPH injury or mechanical injury, Nipponbare was more susceptible to injury than TN1 (Tables 1 and 2). For example, BPH reduced plant height and leaf area of Nipponbare by 28% and 63%, respectively, whereas those of TN1 decreased by 6% and 37%, respectively. Removal of one primary tiller reduced leaf area of Nipponbare by 46%

Table 1. Effects of various treatments, including BPH injury, on plant height, leaf area, and leaf, stem, and root dry weights.a Treatment Healthy Removal of one primary tiller Removal of two lowest leaves of the main tiller Removal of all leaves of the main tiller Removal of two upper leaves of the main tiller BPH feeding Removal of the main tiller F ratio Prob>F
a

Plant height (cm) Nipponbare 61.9 60.6 59.6 53.0* 48.3* 44.4* 37.4* 20.63 <0.0001 TN1 55.0 54.2 54.1 44.1* 48.6* 51.8 43.4* 11.23 <0.0001

Leaf area (cm2) Nipponbare 162.7 87.1* 109.8 116.8 117.2 59.9* 55.9* 7.51 0.0002 TN1 141.1 130.7 139.4 115.5 160.9 89.5* 97.6* 9.41 <0.0001

Leaves (mg) Nipponbare 61.2 33.8* 39.8* 29.2* 43.8* 24.8* 18.2* 12.99 <0.0001 TN1 53.0 49.2 50.8 40.2* 56.8 33.2* 32.2* 11.82 <0.0001

Stems (mg) Nipponbare 75.0 47.2* 60.2 46.2* 62.8 33.8* 23.8* 12.12 <0.0001 TN1 68.0 54.0 70.0 52.8 71.8 46.0* 33.8* 8.15 0.0001

Roots (mg) Nipponbare 58.0 37.2 38.2 39.0 60.8 21.8* 23.2* 6.87 0.0004 TN1 99.7 93.7 77.3 60.3 59.0 36.3* 48.3* 3.77 0.0105

In a column, means followed by an asterisk are significantly different from those of the healthy control.

50

December 2004

Table 2. Effects of treatments, including BPH injury, on the number of tillers 2 wk after treatment.a Treatment Healthy Removal of one primary tiller Removal of two lowest leaves of the main tiller Removal of all leaves of the main tiller Removal of two upper leaves of the main tiller BPH feeding Removal of the main tiller Stem borer feedingb F ratio Prob>F
a b

BPH injury Nipponbare 3.75 2.25* 3.00 3.00 3.25 2.25* 3.00 4.8 0.0031 TN1 3.25 3.00 3.75 3.75 4.00 2.75* 3.00 3.8 0.01

Stem borer injuryb IR64 3.30 0 3.20 3.00 2.30 4.30

borer-injured plants. Lastly, cultivar Nipponbare was more susceptible than TN1 to any type of injury. References
Rubia-Sanchez E, Heong KL, Zalucki M, Gonzales B, Norton GA. 1996. Mechanisms of compensation of rice plants to yellow stem borer Scirpophaga incertulas (Walker) injury. Crop Prot. 15:335-340. Rubia-Sanchez E, Suzuki Y, Arimura K, Miyamoto K, Matsumura M, Watanabe T. 2003. Comparing Nilaparvata lugens (Stl) and Sogatella furcifera (Horvath) (Homoptera: Delphacidae) feeding effects on rice plant growth processes at the vegetative stage. Crop Prot. 22:967-974. Rubia-Sanchez E, Suzuki Y, Miyamoto K, Watanabe T. 1999. The potential for compensation of the effects of the brown planthopper Nilaparvata lugens Stl (Homoptera: Delphacidae) feeding on rice. Crop Prot. 18:39-45. Sogawa K. 1982. The rice brown planthopper: feeding physiology and host plant interactions. Annu. Rev. Entomol. 27:49-73.

In a column, means followed by an asterisk are significantly different from those of the healthy control. Data from Rubia et al (1996).

and TN1 by 7%; plant dry weights of Nipponbare and TN1decreased by 39% and 10%, respectively. TN1 grew faster and had a greater photosynthetic rate and N content than Nipponbare (RubiaSanchez et al, unpubl.), which probably makes TN1 less susceptible to injury. Interestingly, root dry weights of both cultivars injured by BPH decreased by about 60%. This shows the important effect of BPH feeding on root growth. Our earlier study showed that BPH also decreased the root N content of many cultivars

(Rubia-Sanchez et al 2003). BPH feeds near the base of the rice plant so that its feeding may interfere with the transport of nutrients from the roots to the shoots and vice versa. To conclude, except for main tiller removal, the effects of BPH injury on plant growth of Nipponbare and TN1 2 wk after treatment were more significant than with leaf removal or primary tiller removal. BPH-injured plants did not immediately recover from injury by the formation of new tillers, unlike stem

Acknowledgment
This work is financially supported by the Japan Science and Technology Agency.

IRRN 29.2

51

Ecology of jungle rice (Echinochloa colonum), a weed of the rice agroecosystem: a case study in Bilaspur (Chhattishgarh)
V. Dubey, Department of Botany, CMD College, Bilaspur, Chhattishgarh, India

Chhattisgarh in India is a ricegrowing state, with about 80% of its population living in villages that directly or indirectly depend on rice cultivation. The types of rice culture used are direct seeding and transplanting (dry seeded rainfed or wet seeded lowland and/or upland). Bilaspur in Chhattishgarh has a very high temperature, thus assuming an ecological importance with respect to vegetation of cultivated crops, especially rice. There is a great need for research that will prevent losses caused by weeds and will enable management of weeds in direct seeded rice using the most economical and feasible method. These objectives can be achieved through a better understanding of the ecology of different weeds infesting direct-seeded rice. The present study aimed to study some ecological aspects of jungle rice (Echinochloa colonum Link), the most common and obnoxious weed of direct-seeded rice. The investigation included a general survey of the weed flora, phytosociology and phenology of the weeds, and autecological observation of E. colonum. Knowing the peculiarities in the life history of the plant and its epharmony with its environment may provide a tool to alter that epharmony to mans advantage in terms of
52

better rice yield. The study observed the morphology of the seed, seed germination, and growth performance. Phytosociology is the study of detailed structure and description of different plant species or a set of individual plant species grouping together and having mutual relationships among themselves and to their environment. It is through a series of complicated interactions and reactions that the phytosociological interrelationship of any plant community is established. Various climatic, edaphic, and biotic factors combine to determine the floristic composition of a particular plant community. In case of cultivated rice, the natural floristic composition is very much disturbed when we continually manipulate vegetation by cultural means such as tillage or hoeing or by use of herbicides or fire. Weeds appear to be those plants that occupy the earliest stages of secondary succession. Quantitative phytosociological characters such as percentage frequency, density, and abundance were studied following the methodology described by Ashby (1948) and Misra (1968). The plant species were identified preserved in herbarium sheets. Seasonal ecological phenomena are described by superimposing the phenological calendar over the civic calen-

dar. Observations of the different phases of a life cycle such as seed germination, vegetative growth, flowering, fruiting, and seed maturation in their temporal occurrence are used to prepare a phenological calendar. Such phonological observations recorded in the field can provide good information about the ecological behavior of an organism. The phenology of weeds was studied through several visits to the rice fields at regular intervals during the growth period and observations were recorded in the form of a phenogram. (Fig. 1). Jungle rice (E. colonum ), popularly known as sawan, is a troublesome weed. It was the dominant weed species, showing monthly variation in quantitative phytosociological characteristics. E.colonum exhibited 68% and 86% frequency in September and October, which was maximum among all weeds. Its density increases as rice matures, thus making competition severe. E. colonum was found to be the most common and uniformly distributed throughout the rice fields in all study sites having direct-seeded rice culture. The increasing peaks of frequency, density, and abundance revealed that September and October are critical periods for competition with rice. Not only was there more severe competition; quantity
December 2004

Table 1. Percentage frequency (F), density (D), and abundance (A) of weed species in rice fields.a July D 12.0 1.4 1.03 0.50 0.48 0.12 0.04 0.24 0.5 0.40 0.2 0.6 0.28 0.22 0.52 0.14 0.4 0.22 0.36 0.04 2.2 1.63 1.0 2.54 10.72 1.36 1.06 2.8 0.70 0.06 0.54 0.84 0.54 0.12 1.00 0.20 0.24 0.32 0.42 0.44 0.1 0.4 1.72 0.06 0.06 0.28 0.32 0.26 0.4 0.12 0.4 0.26 0.32 0.06 0.04 8 14 12 12 6 6 16 10 10 4 0.08 0.12 0.16 0.16 0.06 6.06 0.08 0.1 0.12 0.04 1.00 1.03 1.33 1.33 1.00 1.00 0.50 1.0 1.20 1.00 100 3.4 2.52 1.55 2.91 1.50 4.72 4.66 2.70 1.50 3.33 2.00 2.0 2.00 2.62 2.75 1.0 3.33 4.52 1.00 1.00 2.0 2.28 1.85 2.85 1.2 2.85 2.16 1.77 1.00 1.00 10. 72 48 60 24 26 10 4 12 24 18 12 50 36 1.86 18 1.55 14 1.57 22 2.36 8 1.16 26 1.53 40 6.35 10 22 4 12. 0 2.91 2.20 2.08 1.84 1.20 1.0 2.0 2.08 2.22 1.66 1.20 0.82 100 52 68 14 20 22 4 24 30 16 22 36 20 26 22 24 8 12 6 10 26 6 8 22 12 12 10 14 6 22 16 8 10 12 6 12.5 1.44 1.72 0.16 0.38 0.32 0.04 0.38 0.42 0.30 0.50 0.58 0.26 0.38 0.42 0.5 0.14 0.24 0.06 0.1 1.30 0.08 0.12 0.36 0.2 0.12 0.14 0.20 0.08 0.32 0.32 0.12 0.12 0.24 0.12 12. 5 2.76 2.52 1.14 1.9 1.45 1.00 1.58 1.40 1.87 2.27 1.61 1.30 1.46 1.90 2.08 1.16 2.00 1.0 1.0 5.00 1.33 1.50 1.63 1.66 1.0 1.4 1.42 1.33 1.45 2.00 1.5 1.2 2.00 2.00 100 56 86 22 12 24 10 32 42 46 36 42 28 12 18 24 26 22 32 8 20 14 16 20 13.0 1.44 1.84 0.6 0.14 0.3 0.14 1.16 0.12 0.78 0.4 0.58 0.34 0.2 0.2 0.62 0.4 0.36 2.96 0.14 0.24 0.24 0.16 0.32 13. 0 2.57 2.13 2.72 1.16 1.24 1.40 3.62 1.0 1.69 1.11 1.38 1.21 1.66 1.11 1.40 1.53 1.63 2.00 1.75 1.20 1.71 1.00 1.60 100 44 64 12 28 16 16 16 26 8 14 12 8 28 20 26 4 22 22 16 18 14 12 18 6 6 8 6 8 14 A %F D A %F D A %F D A %F D 13.0 1.74 1.30 0.26 0.62 0.30 0.4 0.36 0.6 0.1 0.2 0.24 0.1 1.00 0.2 0.5 0.04 0.22 0.22 1.78 0.32 0.18 0.24 0.3 0.06 0.08 0.88 0.1 0.14 0.42 August September October November A 13.0 3.95 2.03 2.16 2.1 1.87 2.5 2.25 2.30 1.25 1.42 2.00 1.25 3.57 2.5 1.92 1.0 1.57 1.57 2.51 1.77 1.28 2.0 1.66 1.0 1.33 1.00 1.66 1.75 3.0

IRRN 29.2

Name

%F

Oryza sativa Paspalum scrobiculatum Echinochloa colonum Cyanotis axillaris Crotalaria albida Cieome cheledonii Cassia tora Cyperus rotundus Cyperus iria Cyperus flavidus Cyperus haspens Fimbristylis ferruginea Aeschynomene aspara Kyllinga triceps Scirpus gressus Eriocolon quinquangulare Euphorbia hirta Eclypta alba Aneilema nudiflorum Ischemum rugosum Sida acuta Sida veronaecifolia Sida rhombifolia Abutilon indicum Aegeratum conizoides Striga euphrasiodes Phyllanthus madraspatensis Alternanthera sessilis Echinochloa crus-galli Commelina benghalensis Aerua lanata Centella asiatica Euphobia microphylla Amaranthus spinosus Achyranthus aspara Corchorus aestuans Partheniun hysterophorus Xanthium strumarium Indigoferra linifolia Coldenia procumbens Desmodium triflorum

100 40 42 18 24 4 22 18 20 8 30 10 12 16 16 16 10 12 38 6 6 14 14 14 14 10 14 12 18 4 4

53

Croton bonplendianum Ammania baceifera Oxalis corniculata Heliotropium supinum Rungia spp. Chrozophora indica Spheranthus indicus Celosia argentea Pistia strateiodes Cynodon dactylon Ludwigia parviflora Marsilea quadrifolia Oryza sativa (red rice) Polygonum glabrum Dactyloctenium aegypticum Echinochloa glabrescens Cyperus difformis

Table 1 continued...

Results are expressed as av of samples from 10 localities.

and quality of grain was likewise reduced. E. colonum germinates and emerges in July with rice after the rains. Flowering starts in the first week of September and the seed matures from September to October, simultaneously with rice. This is the reason jungle rice seeds are often harvested with the crop and the weed inoculum is retained for the next crop. Autecological investigations reveal that jungle rice is an emergent aquatic annual grass, appearing as a selfsown plant in rice field with water. Its very small size and rotundate-elliptic shape help in seed dispersal. This explains its widespread distribution throughout the rice fields. These seeds get easily mixed with the crop seeds, thus retaining the weed inoculum. Seed production per plant was very high, varying greatly with different environmental conditions (Table 2). A large seed output and reproductive capacity indicate its success in the ecosystem. Longevity and the prolonged viability suggest that, once an area of land is infested by it, the potential exists for continued reoccupation over time. Freshly collected seeds exhibit a period of innate dormancy, which seems to be adjusted to the environmental conditions to which the seeds are exposed. Potassium nitrate treatment proved successful in breaking dormancy. The plant could not tolerate waterlogged conditions. Hence, jungle rice is characterized by high plasticity in its vegetative growth, rapid early growth phases,
December 2004

54

36 4 2 20 16 8 12 10 8

0.74 0.04 0.2 0.20 0.40 0.24 0.28 0.24 0.16

2.05 2.00 1.00 1.00 2.5 3.0 2.33 2.00 2.00

10 1.20 8 2.00 16 4 12

0.3 0.04 0.16

36 0.16 10 24 6 12 10 8 24 12 18 1.87 24 14 28 1.00 32 1.33 34 0.12

0.78 0.6 0.4 0.12 0.20 0.24 0.20 0.24 0.24 0.44 0.60 0.42 0.60 0.80 1.12

6.5 6.00 1.66 2.00 1.66 2.40 2.50 2.24 2.00 2.44 2.50 3.00 2.14 2.50 3.29

10 12 10 14 2 18 6 18 14 22 32

0.6 0.24 0.16 0.48 0.02 0.22 0.20 0.70 0.40 0.40 1.00

6.00 4.00 1.60 3.42 1.00 2.00 3.33 3.88 2.85 1.8 3.12

6 32 4 2 32 22 6 12 6 6 64

0.08 0.62 0.12 0.04 0.32 0.32 0.20 0.12 0.16 0.06 0.64

1.33 1.93 1.00 2.00 2.00 1.45 4.00 1.88 2.66 2.00 3.55

and a prolonged vegetative phase before and during reproduction. This allocates a large proportion of resources to seed production. Thus E. colonum seems fit in the category of a competitive ruderal, according to Grimes concept of plant strategies and resource allocation. Some of the traits that correlate E. colonum with r selection are Variable and unpredictable climate High seed output (8,00043,000 seeds plant 1) Density-dependent mortality Mortality at the early stage (otherwise it sur vives) Population size variable with respect to time and not in equi

Table 2. Vegetative and reproductive characteristics of Echinochloa colonum.a Maturity (d) 90-105
a

Height (cm) 4770

Seeds plant1 (no.) 9,00042,327

Seed weight (mg) 2.262.30

Reproductive capacity 10551489

Results expressed as av of five populations from 10 localities.

librium; can recolonize each year Interspecific and in traspecific competi tion Short life span (usu ally <1 yr) Selection favors rapid development, early re production, small body size, and single reproductive period in a life span Results in overall pro ductivity Thus E.colonum is a competitive ruderal showing optimum resource capture and

seed production. This enables it to survive in the agroecosystem of Bilaspur. References

Ashby E.1948. Statistical ecology II. A reassessment. Bot. Rev.14:220-234. Grime JP. 1979. Plant strategies and vegetation process. New York: Wiley. Misra R. 1968. Ecology workbook. New Delhi: Oxford & I. B. H. Publishing Co. United States Department of Agriculture (USDA). 1980. Rice. Agricultural statistics. Washington D.C.: U.S. Government Printing Office. p 19-28.

ERRATA
A, Swarup and N.P.S. Yaduranshi. Response of rice and wheat to organic and inorganic fertilizers and soil amendment under sodic water-irrigated conditions. 29(1) June 2004, p 49-51. p. 49, first sentence, last paragraph: 5-d-old seedlings should be 35-d-old seedlings. p. 50, Table 3, some entries are incorrect. The table should look like this. Corrected values are in boldface.
Table 3. Physicochemical properties of surface soil (015) after 8 y of experimentation, Karnal, India, 1994-2001. Treatment Rice T1 (control) T2 (N120) T3 (N120 P26) T4 (N120 P26 K42) T4 (N120 P26 K42 Zn4.5) T6 (T4 + FYM 10 t ha1) T7 (T4 +gypsum 5 t ha1) T8 (T4 FYM 10 t ha1 + gypsum 5 t ha1) T9 (T4 + pressmud 10 t ha1) T10 (T4 + pressmud 10 t ha1 + gypsum 5 t ha1) CD (P=0.05) Wheat Control N120 N120 P26 N120 P26 K42 N120 P26 K42 N120 P26 K42 N120 P26 K42 N120 P26 K42 N120 P26 K42 N120 P26 K42 8.50 8.53 8.48 8.52 8.50 8.40 8.20 8.30 8.30 8.30 0.09 0.26 0.27 0.27 0.26 0.26 0.41 0.37 0.40 0.41 0.40 0.05 90 146 146 146 144 158 148 159 155 157 8 11.0 8.5 18.5 22.2 20.5 22.0 18.9 23.3 23.8 23.2 3.1 220 250 240 282 279 296 291 297 295 292 22 0.64 0.65 0.65 0.66 1.12 1.03 0.69 1.11 0.96 0.95 0.14 18.7 13.5 13.5 14.2 14.1 12.1 11.7 10.3 11.5 10.1 0.8 pH Organic C (%) Available nutrients (kg ha1) N P K DTPA extractable Zn (mg kg-1) SAR

IRRN 29.2

55

Soil, nutrient, & water management

Impact of integrated plant nutrient management on upland rainfed rice cultivation

H. K. Senapati, S. S. Sahoo, B. Jena, and D. Sahoo, Department of Soil Science and Agricultural Chemistry, Orissa University of Agriculture and Technology, Bhubaneswar 751003, Orissa, India

The poor tribal farmers of Kandhamal District, Orissa, India, practice the age-old method of agriculture. Their fertilizer consumption is only 4 kg ha 1. They cultivate local low-yielding rice varieties under a monocropping system in the upland, hill slopes where nutrient leaching is a major problem. To remove sole dependence on chemical fertilizers, integrated plant nutrient management (IPNM) approach is introduced in the area. Experiments were conducted during 2000-01 kharif on seven selected farmers fields (each farmer treated as a replicate). Intercropped dry-seeded rice variety JHU-11-26 and black gram PU-30 were sown in each field on the same date. Chemical fertilizers, farmyard manure (FYM), and biofertilizer combinations were applied as different treatments using randomized block design. Azotobacter (10 kg ha1) was applied with FYM before sowing. The recommended doses of P and K (each at 30 kg ha 1 ) were applied to all treatments. The soils were sandy loam to silty loam, moderate to slightly acidic (ph 5.86.4), and classed under red soil (Alfisol). Organic carbon con-

tent varied from 0.25 to 0.79%; with 163178 kg available N, 8.516.5 kg P, and 105240 kg K ha1. The results revealed that the treatments had almost similar effects in all farmers fields. Incorporation of intercropped black gram with application of Azotobacter and 5 t FYM ha 1 significantly increased grain yield, producing the highest mean yield of 1.67 t ha 1 with 40 kg supplemental N (Table 1). With the IPNM schedule, the increase over control ranged from 74.6 to 117.0%, more than that achieved with 100% recommended dose of fertilizer.

Black gram incorporation alone without chemical N also increased yield over the control by 23.3%, due to N 2 fixation by the legume and incorporation as manure. Values of agronomic efficiency, physiological efficiency, and apparent recovery were highest with the IPNM schedule (Table 1). The highest benefitcost ratio of 0.346 with 1.346 return per rupee invested were observed with the N40 + black gram + Azotobacter treatment. Treatment with N 40 + black gram + Azotobacter + FYM resulted in highest N uptake during tillering and panicle initiation and by the

Table 1. Effects of IPNM on yield, agronomic efficiency, physiological efficiency, apparent N recovery, and economic yield of upland rice. Treatment Mean grain yield (t ha1) 0.77 1.30 0.95 1.34 1.45 1.67 0.17 Increase over control (%) 69.0 23.4 74.0 88.3 117.0 Agronomic Physiological Apparent BenefitReturn efficiency efficiency N cost ratio per rupee (kg kg1 N) (kg kg1 N recovery invested removed) (%) 8.8 18.0 14.3 17.0 22.5 46.9 40.0 43.5 44.7 45.9 18.8 45.0 32.7 38.0 49.0 0.156 0.182 0.052 0.250 0.346 0.177 0.843 1.182 0.948 1.250 1.346 1.177

Control (N0) 100% RDFa N60) N0 + black gram N40 + black gram N40 + black gram + Azotobacter N40 + black gram + Azotobacter + FYM CD (0.05)
a

RDF = recommended dose of fertilizer.

56

December 2004

grain and straw (Table 2), which directly influenced grain yield. The highest available N content was also observed in this treatment, indicating an improvement in the fertility status of the soil. Intercropping of rice with black gram, green manuring along with Azotobacter, and addition of FYM are the most suitable IPNM approaches for rainfed upland rice in the tribal zones.

Table 2. Effects of IPNM on N content and uptake in rice and N status of soils after harvest. Treatment N in plant (%) Tillering 1.43 1.72 1.56 1.70 1.72 1.76 0.06 Panicle initiation 1.40 1.64 1.53 1.66 1.68 1.72 0.05 N uptake (kg ha1) Grain 9.2 16.4 11.8 16.8 18.4 21.0 2.4 Straw 3.2 7.3 5.1 8.7 9.2 11.0 1.8 Total 12.4 23.7 16.9 25.5 27.6 32.0 3.5 Available N in soil (kg ha1) 165 172 172 175 183 199 16

Control (N0) 100% RDF (N60) N0 + black gram N40 + black gram N40 + black gram + Azotobacter N40 + black gram + Azotobacter + FYM CD (0.05)

Impact of organic manure and inorganic phosphatic fertilizer on yield and nutrient uptake in a rice-rice cropping system
R.K. Kaleeswari and S. Subramanian, Department of Soil Science and Agricultural Chemistry, Agricultural College and Research Institute (ACRI), Tamil Nadu Agricultural University, Madurai, India

The complementary use of organic manures and chemical fertilizers augments the efficiency of both substances to sustain a higher level of rice productivity. An increase in grain and straw yields has been reported with increasing levels of Gliricidia leaves because of the overall improvement in soil properties such as water and nutrient retention. Field experiments were conducted at the ACRI research farm during the 2000-01 wet seasons (Aug-Dec) to study the influence of organic manures and inorganic phosphatic fertilizers on the ricerice cropping system. The experimental soil had the following characteristics: sand 66.14%, silt 8.23%, clay 21.92%, pH 7.3, electrical conductivity (EC) 0.24 dS m 1,
IRRN 29.2

organic carbon 4.8 g kg 1 , available N 146 kg ha1, available P 9.50 kg ha1, and available K 250 kg ha 1 . Twentyfour treatment combinations used organic sources: M1 = no organics, M 2 = farmyard manure (FYM), M3 = poultry manure (PM), and M4 = green leaf manure Gliricidia sepium (GLM) at 12.5 t ha1, and inorganic P sources: P1 = single superphosphate (SSP), P2 = Udaipur rock phosphate (URP) at levels L1-0, L2-30, and L3-26.4 kg P 2O5 ha 1. The experiment was set out in a factorial randomized block design replicated thrice (net plot size, 5 4 m). The nutrient contents of organic manures used in this experiment are GLM: 2.90% N, 0.50% P, 2.80% K; FYM: 0.54% N, 0.28% P, 0.57% K; and poultry ma-

nure (PM): 3.03% N, 1.04% P, 1.16% K. All plots also received 100 kg N ha 1 (urea) and 80 kg K ha 1 (muriate of potash)a basal dose of 50% N and a full dose of P and K were applied at transplanting, with the remaining 50% N applied in two equal splits at 30-d intervals. The P fertilizer doses were applied as per treatment requirements. To study the residual effect, the original layout was maintained without any disturbance throughout the study period. Each plot was divided into two portions: in one, fertilizers and manures were added as per treatment schedule (applied plots); the other portion was unfertilized (residual plots). In the diacid digest of plant samples, N was determined by the micro57

Kjeldahl method, and, in the triacid digest of plant samples, P and K were determined by the vanadomolybdate yellow color method and flame photometric method. Incorporation of organic manure significantly increased rice yield over plots with no manure. Maximum grain yield (5.06 t ha 1 ) was obtained by M 4, followed by M2 (4.54 t ha1) and M3 (4.33 t ha 1 ). Minimum grain yield (3.68 t ha 1 ) was obtained in no-manure plots (M 1). Incorporation of Gliricidia leaves at 5 t ha 1 significantly increased grain yield of rice by 8.68% over green manure application (Turkhede et al 1996). Irrespective of source, application of P fertilizers gave significantly higher grain yield than the control (N- and Kapplied plots). Grain yield was increased by P application and other studies reported higher values with SSP than with rock phosphate (Rani et al 1994). The interaction effect of manure with P sources on grain yield was found to be significant. The maximum grain yield was registered by GLM + SSP (5.32 t ha 1 ), closely followed by FYM + SSP (Table 1). The uptake of N, P, and K by the rice grain through GLM was maximum, followed by FYM. During the 2000 wet season, the increase in N, P, and K uptake over the no-manure plot was 49.0%, 28.1%, and 47.1%, respectively. In the residual plots, the corresponding increases in nutrient uptake were 63.7%, 46.6%, and 43.2% during the 2001 wet season. The result58

Table 1. Effect of organic manure and inorganic P levels on grain yield (t ha1) of rice. Application ratea Manure N 0 0 0 67.5 67.5 67.5 387.5 387.5 387.5 362.5 362.5 362.5 0 0 0 67.5 67.5 67.5 387.5 387.5 387.5 362.5 362.5 362.5 P 0 0 0 35 35 35 130 130 130 62.5 62.5 62.5 0 0 0 35 35 35 130 130 130 62.5 62.5 62.5 K 0 0 0 71.25 71.25 71.25 145 145 145 350 350 350 0 0 0 71.25 71.25 71.25 145 145 145 350 350 350 P (as SSP) Grain yield 2000 WSb 2.64 3.79 4.61 3.82 5.05 5.57 3.76 4.64 5.38 4.38 5.66 5.92 2.36 3.34 4.07 3.68 4.23 4.87 3.41 4.04 4.73 4.07 4.98 5.33 2001 WS (Applied) 3.25 4.47 5.40 4.12 5.20 5.80 4.05 4.96 5.50 4.73 5.81 6.04 3.00 4.12 4.90 3.97 4.89 5.25 3.73 4.61 5.09 4.45 5.23 5.51 2001 WS (Residual) 2.55 3.47 4.20 3.22 4.20 4.93 3.15 4.06 4.64 3.75 4.82 5.18 2.47 3.31 4.01 3.07 4.13 4.35 2.87 3.68 4.26 3.57 4.33 4.61

M1 M1 M1 M2 M2 M2 M3 M3 M3 M4 M4 M4 M1 M1 M1 M2 M2 M2 M3 M3 M3 M4 M4 M4

0 30 60 0 30 60 0 30 60 0 30 60 0 30 60 0 30 60 0 30 60 0 30 60

Treatment means M1 M2 M3 M4 3.68 4.54 4.33 5.06 4.04 4.90 5.51 3.79 4.15 4.75 0.11 0.14 0.10 4.21 4.87 4.66 5.30 3.17 5.11 5.69 2.99 4.74 5.19 0.07 0.08 0.06 3.34 3.98 3.78 4.38 4.14 4.74 3.86 4.31 0.02 0.03 0.02

P as SSP 100:0:503.72 100:30:50 100:60:50 P as URP 100:0:503.38 100:30:50 100:60:50 Interaction Manure P sources Manure P levels P sources P levels
a

M1 = 0 t ha1; M2 = 12.5 t FYM ha1; M3 = 12.5 t PM ha1; M4 = 12.5 t GLM ha1. bWS = wet season.

ing increase in nutrient uptake by the grain with the addition of GLM could be attributed to the release of soil nutrients due to intermediate acids produced during decomposition, in addition to the chelating effect of complex intermediate organic molecules

(Trivedi et al 1995). Among the organic manures studied, the contribution toward yield and nutrient uptake of rice grain was in the order GLM>FYM>PM. The higher CaCO 3 content of PM (10%) might have neutralized the organic acids produced durDecember 2004

Table 2. Effect of organic manure and inorganic P levels on nutrient uptake (kg ha1) by rice grain. N uptake Application ratea Manure Inorganic 2000 WS 50 50 50 50 50 50 50 50 50 50 50 50 27.3 44.1 54.4 36.4 62.2 71.8 36.2 60.6 72.8 42.0 72.5 78.1 2001 2001 WS WS (applied) (residual) 30.9 47.2 57.9 40.5 64.9 76.9 41.5 65.5 75.2 46.6 76.2 81.2 N uptake 2000 WS M1 M1 M1 M2 M2 M2 M3 M3 M3 M4 M4 M4 0 0 0 67.5 67.5 67.5 387.5 387.5 387.5 362.5 362.5 362.5 0 0 0 35 35 35 130 130 130 62.5 62.5 62.5 0 0 0 71.25 71.25 71.25 145 145 145 350 350 350 100 100 100 100 100 100 100 100 100 100 100 100 0 30 60 0 30 60 0 30 60 0 30 60 50 50 50 50 50 50 50 50 50 50 50 50 22.2 32.9 36.8 35.3 49.4 59.2 33.6 49.2 60.5 38.7 59.0 67.3 2001 2001 WS WS (applied) (residual) 28.6 42.5 50.6 38.7 58.7 66.3 37.4 57.5 66.7 43.1 64.4 71.9 22.3 31.8 39.1 29.2 46.6 53.6 27.8 44.3 53.9 33.9 51.9 58.2 2000 WS 5.3 10.9 13.0 8.6 12.9 16.0 7.7 11.0 14.4 11.1 15.5 17.9 22.9 32.5 39.9 29.6 50.4 63.5 31.0 51.9 61.6 36.0 61.3 67.2 2000 WS 7.0 12.9 18.6 9.8 14.4 21.4 8.7 13.4 16.2 12.6 12.9 23.7 P uptake 2001 2001 WS WS (applied) (residual) 8.9 15.0 20.7 11.9 18.3 24.6 11.0 15.9 19.4 15.5 22.8 26.6 P uptake 2001 2001 2000 WS WS WS (applied) (residual) 7.4 13.2 17.3 10.8 17.2 19.7 9.6 14.4 17.5 13.5 18.4 21.2 5.7 9.8 13.5 7.4 13.6 14.9 6.7 11.9 14.3 10.3 13.9 16.2 4.2 6.9 12.7 8.2 10.2 12.9 7.9 9.5 12.1 9.8 12.8 15.3 6.3 11.1 15.1 8.5 13.8 19.7 7.8 12.3 15.6 11.1 17.3 21.6 2000 WS 6.7 9.5 16.9 9.2 13.3 16.5 8.7 11.9 14.6 10.8 15.9 18.9 K uptake 2001 2001 WS WS (applied) (residual) 8.1 17.7 19.4 11.2 15.4 18.6 11.0 14.3 16.7 13.3 17.7 21.3 K uptake 2001 2001 WS WS (applied) (residual) 6.0 9.8 16.5 10.6 13.3 15.1 9.3 12.2 14.7 14.3 15.0 17.2 4.5 7.2 12.8 7.4 10.9 12.2 6.7 9.4 11.9 8.9 11.8 13.6 5.7 8.3 14.4 8.0 11.4 13.8 7.8 10.7 13.4 9.9 13.9 17.4

M1 M1 M1 M2 M2 M2 M3 M3 M3 M4 M4 M4

0 0 0 67.5 67.5 67.5 387.5 387.5 387.5 362.5 362.5 362.5

0 0 0 35 35 35 130 130 130 62.5 62.5 62.5

0 0 0 71.25 71.25 71.25 145 145 145 350 350 350

100 100 100 100 100 100 100 100 100 100 100 100

0 30 60 0 30 60 0 30 60 0 30 60

Treatment means M1 M2 M3 M4 100:0:50 100:30:50 100:60:50 100:0:50 100:30:50 100:60:50 36.3 52.4 52.2 59.6 35.5 59.9 69.3 32.5 47.6 55.9 NS 1.97 1.39 42.9 57.7 57.3 63.9 39.9 63.5 72.8 36.9 55.8 63.9 NS NS NS 31.4 45.5 45.1 51.4 29.9 49.0 58.1 28.3 29.1 34.1 NS 4.41 3.12 11.3 13.8 11.9 15.6 9.5 13.4 19.9 8.1 12.6 15.3 NS NS 0.97 13.7 17.1 14.6 19.7 11.8 17.9 22.8 10.3 15.8 18.9 1.07 NS 0.83 10.3 12.9 11.4 15.1 8.4 13.6 17.9 7.5 12.3 14.7 0.96 NS 0.93 9.4 11.7 10.8 13.9 8.7 12.6 16.7 7.5 9.9 13.3 NS 1.02 0.72 12.9 14.0 13.0 16.5 10.9 16.3 19.0 10.0 12.6 15.9 NS 1.60 NS 8.8 10.6 9.9 12.6 7.9 11.1 14.8 6.9 9.8 12.6 NS 1.27 NS

P as SSP

P as URP

Interaction Manure P sources Manure P levels P sources P levels

a

M1 = 0 t ha1, M2 = 12.5 t FYM ha1, M3 = 12.5 t PM ha1, M4 = 12.5 t GLM ha1. bWS = wet season.

IRRN 29.2

59

ing decomposition, thereby limiting the availability of acids for P dissolution (Mahimairaja et al 1995). P dissolution from inorganic P sources can be improved by incorporation of GLM. Hence, for the wetland ecosystem, the combined application of GLM at 12.5 t ha 1 and inorganic P fertilizers can be recommended.

References

Mahimairaja S, Bolan NS, Hedley MJ. 1995. Dissolution of PR during the composting of poultry manure: an incubation experiment. Fert. Res. 40:93-104. Rani P, Duraisamy P, Ramasamy S. 1994. Relative effectiveness of rock phosphate vs superphosphate on rice yield and quality. Madras Agric. J. 1(6):337-338. Trivedi BS, Bhatt PM, Patel JM, Gami RC. 1995. Increasing efficiency of fertilizer P through addition of

organic amendments in groundnut. J. Indian Soc. Soil Sci. 43(4):627-629. Turkhede AB, Chaudhari BT, Chore CN, Tayde RD, Bobade PN. 1996. Effect of green manuring and fertilizer levels on yield of lowland paddy. J. Soils Crops 6(2):181-189.

Effects of long-term application of manure and fertilizer on the upland rainfed rice-cropping system
H.K. Senapati and P.C. Senapati, All India Coordinated Research Project (AICRP) for Dryland Agriculture, Orissa University of Agriculture and Technology (OUAT), Phulbani 762001, Orissa, India

The cultivation of upland rainfed rice is risky and expensive. Tribal farmers in Orissa plant it on hilltops, hillocks, and hill slopes. Nutrient-use efficiency is very low and moisture stress further reduces yield. Farmyard manure (FYM) and green leaf manure (GLM) are used to increase yield and maintain soil fertility. We conducted an experiment to select an integrated nutrient management practice suitable for the rainfed ricecropping system. Organic and inorganic sources of nutrients were used. The trials were conducted during kharif from 1994 to 2001 in a permanent plot at the AICRP site. There were eight experiments nine treatments replicated thrice (27 plots, 10 8 m2 each). Short-duration (85 d) rice variety Zhu 11-26 was grown
60

continuously in kharif and a local variety of horsegram was grown in rabi. The experimental design was a randomized block design. Soil at the experimental site is an Alfisol with sandy loam texture, pH 5.2, 0.32% organic carbon, and 165 kg available N, 4.4 kg P, and 91.3 kg K ha1. Rice was grown under rainfed conditions with combinations of inorganic and organic mixtures of chemical fertilizers, FYM, GLM (0.45% N), and Cassia siamea (0.36% N). (See table for specific combinations in each treatment.) The results revealed that all treatments were superior to the control. The highest mean grain yield of 2.01 t ha 1 was obtained with FYM supplying 30 kg N ha 1 and half the recommended dose (20 kg ha 1 ) each of P and K (T6). Yield was

on a par with the 100% recommended fertilizer dose (T 2) and 50% N as FYM + 50% as chemical fertilizer along with full doses of P and K (T9). The yields obtained with GLM (T4 and Cassia T 5) were similar to that of T2. The highest increase over the control was observed with FYM (T 6) (151%), followed by T 2 (150%) and T 9 (149%). The highest agronomic efficiency (40.5 kg kg1 N) and water-use efficiency (1.42 kg ha 1 mm) were found with the FYM treatment (T 6). With the continuous addition of FYM and GLM, organic carbon content as well as organic matter of the soil increased, thereby influencing nutrient availability and improving soil fertility.

December 2004

Long-term effects of manurial practices on rice yield, agronomic efficiency, water-use efficiency, and fertility status of the soil. Treatmenta Mean rice yield (t ha1) 0.81 2.00 1.45 1.36 1.44 2.01 1.67 Increase over control (%) 150 81 70 80 151 108 Agronomic Water-use Organic efficiency efficiency (%) (kg kg1 N) (kg ha1 mm) 20.0 21.6 18.6 21.3 40.5 14.5 0.55 1.08 0.75 0.86 0.87 1.42 0.89 0.31 0.34 0.35 0.47 0.46 0.51 0.48 Available nutrients (kg ha1) N 175.8 246.0 234.4 222.6 218.7 281.2 212.6 P 4.8 12.8 11.4 9.2 8.9 10.6 8.6 K 58.1 130.7 118.3 108.0 107.1 119.1 101.6

T1 T2 T3 T4 T5 T6 T7

T8

T9

Control 100% RDF (60-40-40 kg NPK ha1) 50% RDF (30-20-20 kg NPK ha1) Gliricidia (30-20-20 kg NPK ha1) Cassia (30 N) + 20-20 kg PK ha1 FYM (30 N) + 20-20 kg PK ha1 T3 + T4 30 N (CF) + 30 N (G) + 40-40 kg PK ha1 T3 + T4 30 N (CF) + 30 N (C) + 40-40 kg PK ha1 T3 + T4 30 N (CF) + 30 N (F) + 40-40 kg PK ha1

1.74

117

15.6

0.83

0.46

205.7

8.4

104.6

1.99

149

19.9

1.16

0.51

289.0

10.5

128.6

a RDF = recommended dose of fertilizer, FYM = farmyard manure, CF = chemical fertilizer, G = Gliricidia, C = Cassia.

Effect of silicon sources and fertility levels on transplanted rice

P.C. Sudhakar, J.P. Singh, and Kalyan Singh, Department of Agronomy, Institute of Agricultural Sciences, Banaras Hindu University (BHU), Varanasi 221005, India

Even with the balanced use of NPK in long-term studies, high yield levels in rice could not be maintained because of other macro- and micronutrient deficiencies and deterioration in the soil physical ecosystem. Rice is a known silicon (Si) accumulator and it benefits from silicon nutrition
IRRN 29.2

(Takahashi 1995). There is an acute disposal problem with respect to industrial wastes such as basic slag (16% Si) and fly ash (30% Si). In addition, there are also problems with agricultural wastes (e.g., rice straw, 4% Si) because of the reduction in animal power on the farm. Our study aimed to

evaluate different industrial and farm wastes as a source of Si for sustained rice yields. The field experiment was conducted at the BHU research farm in Uttar Pradesh. The soil at the experimental field was Gangetic alluvial (Ustochrept) with pH 7.3. It was moderately fertile, being
61

low in organic carbon and available N and medium in available P and K. The experiment was laid out in a splitplot design with four fertility levels (F 1: 80-40-40-16-0.25 kg N-P-K-S-Zn EDTA; F 2: 120-6060-24-0.50; F 3: 160-80-80-320.75; and F 4 : 200-100-100-401.00) in main plots and three Si sources (S1: basic slag, S2: fly ash, and S3: rice straw compost applied at 120 kg Si ha1), including a control (S 0) in subplots replicated thrice. The nutrients N, P, K, S, and Zn were applied in the forms of urea, diammonium phosphate, muriate of potash, elemental sulfur, and Zn EDTA, respectively. Coming from different sources, Si was applied as per treatment requirements. Rice variety BPT5204 (Samba Mansuri) was used in the experiment. To calculate dry matter production, four randomly selected hills from suitable rows were cut at the collar region. Shoots were oven-dried at 60 C for 48 h and weighed. The weight thus obtained was recorded as dry weight hill 1 (g) after dividing total dry weight by the total number of hills (4) in the sample. Grain yield was recorded (kg plot 1) after threshing, winnowing, cleaning, and drying. It was then computed to kg ha1. The difference between bundle weight and grain yield gave straw yield (kg plot 1), which was computed to kg ha 1. The fertility levels significantly affected dry matter production hill1 up to F3 in all growth stages, except at tillering during both years (Table 1). Grain and straw yields increased with an in62

Table 1. Effect of fertility levels and Si sources on dry matter production (g hill1) at different growth stages of rice.a Treatment Fertility level F1 F2 F3 F4 SEM LSD (P = 0.05) Silicon source S0 S1 S2 S3 SEM LSD (P = 0.05)
a

Tillering 2001 3.8 5.0 5.1 5.4 0.9 ns 4.2 5.3 4.9 4.8 0.4 ns 2002 3.7 4.3 4.7 4.9 0.60 ns 4.1 4.6 4.4 4.4 0.3 ns

Late jointing 2001 6.3 7.3 8.1 8.5 0.4 1.3 6.4 8.1 7.9 7.6 0.4 1.1 2002 6.9 7.2 7.5 7.7 0.2 0.6 6.4 7.7 7.6 7.6 0.1 0.4

Flowering 2001 16.5 18.9 20.0 21.8 1.0 3.5 17.8 20.2 19.9 19.3 0.5 1.5 2002 16.4 18.8 20.1 21.7 0.7 2.5 17.5 19.7 19.7 19.4 0.5 1.4

At harvest 2001 55.9 59.5 62.3 62.7 0.6 2.2 57.7 61.23 60.9 60.7 0.4 1.2 2002 52.5 56.7 59.3 60.9 0.5 1.8 55.0 58.5 58.4 57.5 0.2 0.6

See text for treatment details. ns = not significant.

Table 2. Effects of fertility levels and Si sources on grain and straw yields (kg ha1) of rice. Treatment Fertility level F1 F2 F3 F4 SEM LSD(P = 0.05) Silicon source S0 S1 S2 S3 SEM LSD(P = 0.05)
a

Grain yield 2001 4,393 4,905 5,026 5,092 76 263 4,626 4,986 4,860 4,943 85 249 2002 4,203 4,625 5,005 5,066 103 358 4,116 5,006 4,817 4,960 89 259 Pooled 4,298 4,765 5,015 5,079 64 198 4,371 4,996 4,839 4,951 62 175 2001 5,630 7,353 7,741 7,902 131 454 6,817 7,414 7,207 7,188 120 350

Straw yield 2002 5,601 6,791 7,136 7,217 83 241 6,391 6,847 6,799 6,708 63 217 Pooled 5,615 7,072 7,439 7,559 73 224 6,604 7,130 7,003 6,948 73 207

See text for treatment details.

crease in fertility levels up to F 4, but the differences were significant only up to F 3 (Table 2). The positive effect on growth attributes can be explained by the fact that nutrients, mainly N, P, and S, are needed to produce leaves and tillers, which together contribute to dry matter production (Ishizuka 1971). Nitrogen, P, and K, along with S and Zn,

have a marked effect on increasing rice yields by promoting growth and better movement of photosynthates from the leaf to the grain (Dong et al 1981). The application of Si through any of the sources significantly increased dry matter and grain and straw yields relative to the control (no Si). Although Si sources were statistically on a par among
December 2004

themselves, basic slag tended to be better than rice straw compost, followed by fly ash. The maintenance of photosynthetic activity and the more efficient use of light, coupled with translocation of assimilated products to the sink (Rani and Narayanan 1994) because of Si fertilization, could be the reasons for the increased dry matter production. The yield increase through Si application is largely attributed to the advantage gained in grain filling and grain weight because of better translocation of photosynthates (Rani et al 1997). Among Si sources, basic slag performed relatively better

than did fly ash and rice straw compost. It has higher Si availability than these two sources (Ma and Takahashi 1991). Silicon application increases both grain and straw yields of rice under both low and high fertility levels. Basic slag, fly ash, and rice straw compost as sources of Si have shown positive effects on rice grain and straw yield, with basic slag giving a better performance. Therefore, these Si sources, preferably basic slag, can well be used to increase and sustain rice yields while solving disposal problems.

References

Dong CT, Liu ZG, Zou BJ, Zhu C, Zhang CL, Liang W. 1981. Study on the nutrition of rice. 2. Effect of zinc and silicon on increasing rice yield. Liaoning Agric. Sci. Liaoning Nongye Kexue 4:13-18. Ishizuka Y. 1971. Physiology of the rice plant. Adv. Agron. 23:241-315. Ma JF, Takahashi E. 1991. Availability of rice straw silicon to rice plants. Soil Sci. Plant Nutr. 37(1):111-116. Rani AY, Narayanan A. 1994. Role of silicon in plant growth. Agro Annu. Rev. Plant Physiol. (B & A) 1:243262. Rani YA, Narayanan A, Devi VS, Subbaramamma P. 1997. The effect of silicon application on growth and yield of rice plants. Ann. Plant Physiol. 11(2):125-128. Takahashi E. 1995. Uptake mode and physiological function of silica. Sci. Rice Plant 2:58-71.

Fertilizer optimization in rice by eliminating hidden nutrient deficiencies

M.R. Latha and V. Murugappan, Department of Soil Science and Agricultural Chemistry, Center for Soil and Crop Management Studies, Tamil Nadu Agricultural University, Coimbatore 641003, India

Intensive rice cultivation with modern high-yielding varieties using high-analysis fertilizers has induced secondary and micronutrient deficiencies in soils. Many times, these hidden nutrient deficiencies are not identified and corrected before field experiments are conducted to define fertilizer recommendations. These hidden nutrient deficiencies, if not identified and removed beforehand, will limit crop yield and the optima thus defined using the experimental data will be inappropriate. This is one of the reasons for the recently observed plateau in the average
IRRN 29.2

yield growth of rice. This paper describes a systematic approach of experimentation, which enables identification and removal of such hidden nutrient deficiencies, if any, at experimental sites used for evaluating fertilizer optima for rice. The study was conducted in a farmers field in Ramapuram village in Thanjavur District (black calcareous soil belonging to the Kalathur series), representing the major rice-growing tract of Tamil Nadu, in the Cauvery Delta Zone. The study was conducted in four phasesincubation study, greenhouse study,

field experiment, and on-farm evaluation. The incubation experiment was carried out to learn whether any of the applied plant nutrients, P, K, S, Zn, Cu, Mn, and B, react abnormally with the experimental soil (Hunter 1980). This was accomplished by adding a series of concentrations of all these nutrients to a sample of soil, incubating them for a specific time, and subsequently extracting them. If, in any case, the amount extracted at zero level of addition exceeds three times the critical level, that particular nutrient was considered sufficient for rice growth. In other cases, the
63

optimal amount of a nutrient was derived using the nutrient sorption curves obtained by plotting the amount of nutrient extracted (y axis) against the amount added (x axis) and extrapolating to the point corresponding to three times the critical level on the y axis. The results of incubation studies indicated that, among the nutrients, P, K, and Zn were found to limit yield in the experimental soil as the level of each of them extracted at zero level of incubation was less than three times that of the critical level: 10.0 mg Olsen P kg 1 (Anonymous 1980), 0.18 meq NH 4OAc-K 100 g 1 (Anonymous 1980), and 1.20 mg DTPA-Zn kg 1 (Savithri 1978). Since N was generally found to be limiting in all soils, it was also included in the optimum nutrient treatment (ONT). It consisted of those nutrients that were deficient in the soil as indicated by the incubation study: N = 50 mg kg1, P = 12 mg kg1, K = 0.045 meq 100 g1, and Z = 3.2 mg kg1. In the second phase, a greenhouse study was conducted using the missing element design (Hunter 1980). There were 14 treatments in this experiment replicated four times in a completely randomized design. Treatment 1 was the ONT. Treatments 213 received the same amount of each nutrient as the ONT, except for the nutrient under study. Treatment 14 was the check. Sorghum (CO 27) was sown as the test crop. The seedlings were maintained in plastic tumblers fed with the respective nutrient solution as per treatment re64

quirements and harvested after 3 wk. The relative dry matter yield (relative to the yield in the ONT) in each treatment was calculated. The exclusion of N, P, K, and Zn was found to significantly lower relative dry matter yield compared with that of the ONT (Table 1). But the other nutrients, such as Ca, Mg, S, Cu, Fe, Mn, Mo, and B, did not show a significant influence on dry matter yield of sorghum. This greenhouse study confirmed that N, P, K, and Zn were deficient in the experimental soil. In the third phase of the study, in the field, 12 treatments were replicated four times. The treatments were 0, 140, 170 (N 2), and 200 kg N ha1; 0, 12, 24 (P2), and 36 kg P ha1; and 0, 20, 35 (K2), and 50 kg K ha1. N2P2K2 + Zn formed the ONT. All the treatments received the optimum amount of 6.5 kg Zn ha 1 , except the ONT-Zn treatment. The blanket recommendation was included for comparison. Rice (ADT36) was grown as the test crop during the kharif season (June-September 1999). From the plot of grain yield data on a quadratic polynomial surface, the optimum amounts of fertilizers that ensure balanced fertilization were evolved (Table 2). In addition, 30 kg ZnSO4 ha 1 was found to be the optimum amount of Zn for rice in this soil. The blanket fertilizer is suboptimal with respect to all the nutrients in the experimental soil. The economic optimum rates of nutrients developed by the above systematic procedure were verified in on-

farm trials at two locations Ramapuram and Manankorai villages in Thanjavur. The soils of these locations belonged to the same soil series, the Kalathur. The economics of balanced fertilization through a systematic approach was compared with blanket fertilizer recommendations (Table 3). Rice yields as well as profits were higher with balanced fertilization achieved by using the newly evolved
Table 1. Relative dry matter yield of sorghum CO 27 for different treatments in the greenhouse study. Treatment Dry matter yield (g 5 plants1 ) 2.57 2.50 2.34 1.57 1.71 1.93 2.37 2.44 2.83 2.48 2.66 2.58 1.92 1.27 2.22 0.29 0.60 Relative yield (%) 100 97 91 61 66 75 92 94 110 96 103 101 75 49

Optimum nutrient treatment (ONT ) ONT + Ca ONT + Mg ONT N ONT P ONT K ONT + B ONT + Cu ONT + Fe ONT + Mn ONT + Mo ONT + S ONT Zn Control Mean SE Critical difference (P = 0.05)

Table 2. Optimum rates of nutrients (kg ha1) in comparison with blanket fertilizer recommendations for rice in an experimental soil in the Cauvery Delta Zone, Tamil Nadu, India, 1999. Experiment N Kharif rice (ADT36) 175
a

Economic optimuma P2O5 54 K2O 50 N 120

Blanketb P2O5 K2O 38 38

With 30 kg ZnSO4 ha1.bWith 25 kg ZnSO4 ha1.

December 2004

fertilizer optima as compared with those attained with blanket fertilizer recommendations (Table 3). There was an average yield increase of 23%
Table 3. Economics of balanced fertilization vs current blanket recommendations for rice in the Cauvery Delta Zone, Tamil Nadu, India, 1999. Ramapuram Treatment Manankorai

Yield Profit Yield Profit (kg ha1) in Rs (kg ha1) in Rs (US$) (US$)a

Balanced fertilization by systematic approach 7,054 25,056 6,831 24,164 (557) (537) Blanket fertilizer 5,827 21,043 5,444 19,511 (468) (434) Difference because of balanced fertilization 1,227 4,013 1,387 4,653 (89) (103) Critical difference 127 116 (P = 0.05)
a

due to the adoption of newly evolved optima, which, in monetary terms, was Rs 4,333/(US$98) ha1 more than that of the blanket recommendations. Thus, the fertilizer optima developed through the systematic approach, which envisages identifying and accounting for any hidden nutrient deficiencies before experimentation to define fertilizer optima, are realistic and are the key to sustainable agriculture. References
Anonymous. 1980. Proceedings of the TNAU and TNDA seminar on responses of crops to application of P and K and soil fertility evaluation, Tamil Nadu Agricultural University, Coimbatore. Hunter AV. 1980. Laboratory and greenhouse techniques for nutrient survey studies to determine the soil

amendments required for optimum plant growth. Orange City, Florida (USA): Agro Services International, Inc. Savithri P. 1978. Studies on micronutrient fertilization on the availability of nutrients in the soil and their uptake in a cropping system. PhD thesis. Tamil Nadu Agricultural University, Coimbatore.

Acknowledgments
The authors sincerely acknowledge the financial assistance provided by the Potash and Phosphate Institute of Canada under its India Programme for this investigation.

US$1 = Rs 45.

Crop yield, disease incidence, and insect pest attack in relation to N dynamics in rice
R.S. Rekhi, and R. Singh, Department of Soils; and R.K. Goel and J. Singh, Department of Plant Breeding, Genetics, and Biotechnology, Punjab Agricultural University (PAU), Ludhiana 141004, India E-mail:rsrekhi04@yahoo.com

The decline in rice yield on the Indo-Gangetic plains is a matter of great concern (Abrol et al 2000). We evaluated available soil N, plant N content and uptake during rice growth, and incidence of major diseases and insect pest attacks in an ongoing longterm experiment (since 1982) at the PAU farm (30 562 N and 75 272 E). The experimental site is characterized as semiarid subtropical, with soil pH of 8.1, electrical
IRRN 29.2

conductivity of 0.18 dS m 1 , 0.21% organic carbon, 3 kg NaHCO 3-extractable P, and 93 kg NH 4Oac-extractable K ha1. The fertilizer treatments were a control; application of recommended optimum (100%) and supra-optimal (150%) amounts of N, NP, NPK, NPKZn; half N through Sesbania aculeata green manure (GM) and the rest through urea; and one-third N through farmyard manure (FYM) and the rest through

urea plus recommended amounts of PKZn. The recommended amounts of N, P, K, and Zn were 120, 30, 30, and 3 kg ha1, respectively. Nitrogen was applied in three equal amounts at transplanting and at 3 and 6 wk after transplanting, whereas all of P and K were drilled at puddling. The experiment was laid out in a randomized block design with six replications. The available N in soil estimated by alkaline KMnO 4
65

oxidation (Subbiah and Asija 1956) increased at 20 d after transplanting (DAT) and declined thereafter (Fig. 1a). Treatment effects revealed a higher amount of available N at 180 kg N ha 1 , followed closely by the treatments of integrated use of FYM and GM. The higher soil N was reflected in the higher N content and uptake by crop plants in these treatments (Fig. 1b, c). Further, plant N content at 40, 60, and 80 DAT showed a significant correlation (r) of 0.95**, 0.92**, and 0.95** with available N at 20, 40, and 60 DAT, respectively. Though N content and uptake were higher at 180 kg N ha 1, this did not give higher grain yield. Grain yield was highest under the integrated use of FYM (see table). The incidence of diseases and pests, evaluated according to the Standard evaluation system for rice (IRRI 1996), showed that the application of 180 kg N ha 1 resulted in a higher incidence of sheath blight (ShB) caused by Rhizoctonia solani and sheath rot (ShR) mainly caused by Sarocladium oryzae and Fusarium moniliforme. There was a significant correlation between ShB/ShR incidence and plant N content at 40 DAT (r = 0.87**/0.86**) and 60 DAT (r = 0.82**/0.89**). The application of 180 kg N ha1 also resulted in a higher incidence of whitebacked planthoppers and leaffolders and was significantly correlated (r = 0.88** and 0.85**, respectively) with plant N content at 40 DAT. We conclude that supraoptimal fertilization leads to a higher incidence of diseases
66

Available N (kg ha-1) 120 a

100

80

60 N content (%) 2.5 b

Control N120 N180 GM + N60 FYM + N60

2.0

1.5

1.0 N uptake (kg ha-1) 100 c 80 60 40 20 0

20

40

60 Days after transplanting

80

100

120

Fig. 1. Effects of fertilizer and manure treatments on (a) available N in the soil, (b) plant N content, and (c) N uptake by rice.

December 2004

Effects of different fertilizer and manure treatments on the reaction of rice cv PR114 to some major diseases and insects, and rice yield. Reactionb to diseases Treatmenta Sheath blight (incidence) 3.7 4.8 5.2 5.5 5.8 5.9 5.9 6.2 6.5 6.7 6.7 (9.0) Sheath rot (severity) 20.2 (26.6) 45.7 (42.4) 44.3 (41.6) 37.5 (37.0) 45.8 (42.3) 59.9 (51.0) 45.2 (42.2) 48.8 (44.3) 62.9 (52.6) 58.3 (49.6) 50.4 (45.2) (10.3) Insect pests Hopperburn area (%) 0.0 (0.0) 8.3 (14.9) 11.7 (0.9) 10.3 (0.5) 6.2 (11.5) 25.8 (0.6) 17.8 (22.6) 6.7 (0.5) 23.3 (0.2) 25.0 (0.3) 23.7 (28.7) (9.0) Leaffolderdamaged leaves (%) 0.0 (0.0) 42.5 (40.3) 50.8 (45.4) 43.3 (40.8) 28.3 (31.8) 58.3 (49.8) 60.8 (51.1) 38.3 (37.4) 50.8 (45.6) 61.7 (52.3) 64.2 (53.3) (12.1) Mean grain yield, (t ha1)c

and insect pests. A balanced nutrient application at the optimum level is important to attain high rice yield. References
Abrol IP, Bronson KF, Duxburry JM, Gupta RK, editors. 2000. Long-term soil fertility experiments in ricewheat cropping systems. RiceWheat Consortium Paper Series 6. New Delhi (India): Rice-Wheat Consortium for the Indo-Gangetic Plains. IRRI (International Rice Research Institute). 1996. Standard evaluation system for rice. 4th ed. Los Baos (Philippines): IRRI. Subbiah BV, Asija GL. 1956. A rapid procedure for estimation of available nitrogen in soils. Curr. Sci. 25:259-260.

Control N120P30K30Zn3.0 N120P30K30 N120P30 FYM + N80P30K30 N180P45 GM + N60P30K30 N120 Nl180 N180P45K45 N180P45K45Zn4.5 LSD (P = 0.05)
a

3.19 5.75 5.83 5.45 6.07 5.49 5.27 5.19 5.27 5.58 5.52 0.39

Subscripts are nutrient doses in kg ha 1 . bAccording to the Standard evaluation system for rice (IRRI 1996). Numbers in parentheses are angular transformed values. GM = green manure (Sesbania aculeata), FYM = farmyard manure. cMean of 3 y (2001-03) WS.

Nitrogen fertilizer increases protein and reduces breakage of rice cultivar Chainat 1
M. Leesawatwong, S. Jamjod, and B. Rerkasem, Agronomy Department, Faculty of Agriculture, Chiang Mai University, Chiang Mai 50200, Thailand; J. Kuo, Centre for Microscopy and Microanalysis, University of Western Australia, Crawley 6009, Australia; B. Dell, School of Biological Sciences and Biotechnology, Murdoch University, Perth 6150, Australia E-mail: g4368102@cm.edu

Rice breakage during milling is a serious problem in many parts of the world as it reduces farmers income (Leesawatwong et al 2003). Breakage usually occurs in the middle region of the endosperm. Breakage during milling may be alleviated in some cultivars if the density of storage protein could be increased in the endosperm in the region of breakage, as cultivars with higher protein content are less vulnerable to breakage (Nangju and De Datta 1970). Nitrogen fertilizer has the potential to increase protein content in
IRRN 29.2

some cultivars (Perez et al 1996). The objective of this study, therefore, was to determine whether increasing the accumulation of head rice protein, through the use of N fertilizer, would reduce breakage during milling of Chainat 1 (CN1; a Thai high-yielding cultivar). Rough rice was obtained in the 2001 wet season from an N fertilizer experiment with nil and 120 kg N ha1 at flowering, a factorial in a randomized complete block design with three replications, at Chiang Mai, Thailand. Treatments producing head rice N

concentrations, by the Kjeldahl method, of 1.3% (nil fertilizer) and 2.0% N (120 kg N ha1) were chosen for investigation. There was no difference in rough rice yield between the two treatments (data not shown). The rice was dried at room temperature to 14% wet-basis moisture content, dehulled (sheller series P-1, Ngek Seng Huat Ltd., Thailand), and the resulting brown rice polished for 30 s (miller series K-1, Ngek Seng Huat Ltd., Thailand). The percentage of breakage (percentage of broken rice as total milled rice
67

Effect of N treatment on head rice N concentration (%) and percentage of breakage of CN1.a N treatment 0 kg N ha1 120 kg N ha1 at flowering
a

Head rice N concentration (%) 1.3 b 2.0 a

Breakage (%) 20.9 b 10.8 a

Different letters indicate significant difference by LSD (P = 0.05).

Storage protein (arrows) accumulation (positive staining with 1% amido black 10B) in the lateral portions of (a) low- and (b) high-N brown rice grains of CN1. A: aleurone layer, S: starch granule.

68

December 2004

weight) was determined. Three brown rice grains were fixed and embedded in glycol methacrylate, a watersoluble resin (OBrien and McCully 1981). To investigate differences in storage protein accumulation between the two treatments, sections (2.5 m thick) were cut in the region of endosperm breakage, stained for storage protein by 1% amido black 10B, a protein-binding dye, and examined with bright-field light microscopy. The use of N fertilizer at flowering increased head rice N concentration from 1.3% to 2.0% N and reduced grain breakage by 50% (see table). Storage protein in endosperm sections, taken in the breakage region, stained black with amido black 10B and was mostly present as protein bodies between the starch granules. Visually, storage protein was much more abundant in this region in rice from the 120 kg N ha 1 treatment than that in rice grown without N fertilizer (see figure). Similar results on protein accumulation were observed in head and broken rice (data

not presented). Storage protein did not accumulate uniformly in all regions of the endosperm. Furthermore, rice that was more vulnerable to breakage had fewer cell layers containing protein in the region of breakage. The results suggest that both the total amount and distribution of protein within the endosperm can affect breakage. Although the mechanism for reducing breakage is unknown, studies have shown that protein increases hardness and thus could make the rice more resistant to breakage during milling. Not all rice cultivars are anticipated to respond in the manner shown for CN1 as there is some evidence that some rice cultivars can preferentially accumulate protein in the region of breakage at low fertilizer N (unpubl. data). Further work is required to identify the range of cultivars for which agronomic treatments such as N fertilizer could enhance head rice yield, thereby increasing income to farmers. Research is progressing toward the use of N fertil-

izer, not only for promoting crop yield but also for improving rice quality. References
Leesawatwong M, Jamjod S, Rerkasem B, Pinjai S. 2003. Determinants of a premium priced, special quality rice. Int. Rice Res. Notes 28(1):34. Nangju D, De Datta SK. 1970. Effect of time of harvest and nitrogen level on yield and grain breakage in transplanted rice. Agron. J. 62:468474. OBrien TP, McCully ME. 1981. The study of plant structure. Australia: Termarcarphi Pty. Ltd. Perez M, Juliano BO, Liboon SP, Alcantara JM, Cassman KG. 1996. Effects of late nitrogen fertilizer application on head rice yield, protein content, and grain quality of rice. Cereal Chem. 73:556-560.

Acknowledgments
KiThe authors wish to acknowledge the Thailand Research Fund and the McKnight Foundation for financial support. The first author is a recipient of a Royal Golden Jubilee PhD scholarship. The Centre for Microscopy and Microanalysis generously provided facilities for histology and microscopy studies.

Participatory irrigation management for efficient water use and enhanced rice productivity in Tamil Nadu, India
B. Chandrasekaran and A. Nandhagopal, Tamil Nadu Rice Research Institute, Aduthurai, India; K. Palanisami and S.D. Sivakumar, Tamil Nadu Agricultural University, Coimbatore, India; and V. Balasubramanian and R. Rajendran, IRRI

Rivers, lakes, and wells are the main sources of irrigation water in India. The country has exploited its rivers as efficiently as possible during the last five decades. The number of dams constructed has increased from less than 300 in the 1960s to 4,300 today. The number of major and medium irrigation projects increased considerably during the 1960-95 planning periods. In spite of the impressive buildup of water resources, current per capita availability of water in India is only 2,200 m 3, compared with 65,000 m 3 in Japan and 62,000 m3 in the U.S. With increasing population, per capita availability of water in India may even be halved by 2025. In addition, poor transmission of water (50% wastage) is responsible for the suboptimal operation of many irrigation projects, resulting in low efficiency (<35%) of a few major projects. The tank-irrigated area of 3.56 million ha in 1975 has now been reduced by 22% because of improper maintenance and siltation of tanks. Overexploitation of groundwater is also posing serious water-use problems for agriculture and other sectors. Data from the United Nations Development Programme (UNDP) show that rice output per liter of water was 0.50.7 g at research sta-

tions but only about 0.2 g in farmers fields, suggesting a potential to develop and promote improved technologies for efficient water management in irrigated agriculture. Adaptive research trials (ART) conducted in 1999-2001 in Tamil Nadu, India, on improved water management indicated a water savings of 11% in rice, 15% in maize, 14% in tomato, and 13% in onion. The community bore-well system tested in different cropping systems recorded more water savings and economic returns in the rice-soybeansesame system than in the other cropping systems (Chandrasekaran et al 2001). Informal or formal farmer organizations are critical for the efficient management of water. One approach being promoted in India is the organization of farmers into a water users association. This paper discusses the results of community-based, on-farm irrigation management trials conducted in Tamil Nadu, India. To determine the requirements and benefits of improved irrigation management, an experiment under the Operational Research Project was conducted at two different sites of the Cauvery Delta Zone (CDZ), Rajendram and Vaidyanathanpettai, in 1992-97. Two treatments control (continuous flooding)

and improved irrigation management (intermittent)were studied as a nonreplicated trial in these villages. The practices followed under improved irrigation management were maintaining a thin film of water from 3 to 7 d after transplanting (DAT) and then irrigating to 5-cm water depth 1 d after the disappearance of ponded water from previous irrigation. Seepage was reduced by forming bunds around the fields. The mean data over a 5-y period on the amount of water applied and the corresponding rice grain yield at the head, middle, and tail end of the irrigation command area are presented in the figure. Water use in improved water management (intermittent irrigation) vs farmers irrigation (continuous flooding) was 1,0101,040 vs 1,2001,260 mm ha 1 at the head, 902955 vs 1,0501,110 at the middle, and 900940 vs 850875 at the tail end of the irrigation system (SWMRI 1998). The improved irrigation management reduced water use at the head and middle reaches and spared more water for farmers at the tail end compared with the farmers system. Under the farmers practice of continuous flooding, an inadequate water supply at the tail end area led to low rice yields, whereas the improved

IRRN 29.2

69

Irrigation water (mm) 1200 1000 800 600 400 200 0 1200 1000 800 600 400 200 0 Head Middle Sluice reach Tall
IWM- improved water management FP- farmers practice

Grain yield (t ha-1) 8 a 6 4 2 0 c 8 6 4 2 0 Head Middle Sluice reach Tall d b

The research data revealed that all farmers in an irrigation system could increase rice productivity through the judicious management of irrigation water and equal sharing of water from the head to the tail end area. Efficient soil and crop need-based irrigation technologies have been developed for rice and rice-based cropping systems. Widespread dissemination of these technologies will require full farmer participation and policy support. References
Chandrasekaran B, Ravisankar N, Karunanithi S, Bhuvaneswari J, Palanisami K. 2001. Role of community bore wells in Cauvery New Delta areas for conjunctive use. Paper presented at a training program, CET, WTC, Tamil Nadu Agricultural University, Coimbatore, India. SWMRI (Soil and Water Management Research Institute). 1998. Annual report. Coimbatore (India): SWMRI, Tamil Nadu Agricultural University.

Effect of participatory irrigation management on water use and rice productivity in CDZ, Tamil Nadu, India, 1992-97 (mean of 5 y). Rajendram (a,b); Vaidyanathanpettai (c,d).

irrigation management permitted judicious use of water by all farmers and resulted in increased rice yields by 9% at the head, 11% at the middle, and 40% at the tail end

(see figure). Thus, the amount of water saved at the head and middle reaches was economically shared by the tail end users, which ultimately increased rice productivity.

70

December 2004

Crop management & physiology

Effect of low light on yield and physiological attributes of rice

M. Lakshmi Praba, M. Vanangamudi, and V. Thandapani, Department of Crop Physiology, Tamil Nadu Agricultural University (TNAU), Coimbatore 641003, India. (Present address: Crop, Soil, and Water Sciences Division, IRRI, DAPO Box 7777, Metro Manila, Philippines.)

The rice crop in India occupies about 40 million ha of land, mostly raised during the wet season (July-October). However, yield in this season is very low, averaging around 1.2 t ha 1. The major cause of impaired yield is reduced light intensity due to overcast skies prevailing during this season (Murty et al 1975). Since sunlight is critical in photosynthesis, it has been well documented that, under low sunlight, dry matter accumulation is impaired. Information on key physiological parameters is therefore necessary to understand the nature of yield manipulation in the wet season. This study aimed to determine the influence of induced low light at different growth stages of the rice crop and the consequent changes in its physiological and yield components. Rice variety CO 43 was grown in pots during the 2002 wet season in a factorial randomized block design at the glasshouse of the Department of Crop Physiology, TNAU. Thirty-day-old seedlings were planted in 30-cm-diameter 26-cm-high mud pots. Normal sunlight was estimated to be around 60 klux. Shade treatment was given by
IRRN 29.2

covering the plants with white gada cloth (around 29 klux), giving around 50% of normal light. Three treatments were imposedT 1 = shading from transplanting to panicle initiation (PI), T 2 = shading from PI to flowering, and T 3 = shading from flowering to harvest with five replications. Five days after anthesis, the top three leaves were excised from the treated and the control plants. Total chlorophyll a and b fractions were determined as per the method of Yoshida et al (1972). Soluble protein content was determined using Folin-Phenol reagent (Lowry et al 1951), and enzyme nitrate reductase (NR) activity was measured as per the method of Nicholas et al (1976). Yield and yield compo-

nents were determined following standard procedures. Table 1 shows the effect of low light on various physiological attributes. With reduction in light intensity, chlorophyll content showed a progressive increase in all treatments. The absolute quantities were highest in treatment 2, which significantly differed from the control. Similar results of increased chlorophyll under shade from 10% to 43% were reported by Singh et al (1988). Liu et al (1984) suggested that the increase in chlorophyll is due to the tendency of the rice crop to enrich the assimilatory system to produce more photosynthates and recommended that high total chlorophyll and low a-b ratio

Table 1. Effect of low light on contents of chlorophyll (mg g1 FW), soluble protein (mg g 1 FW), and enzyme nitrate reductase (NR) activity ( mol NO2 g1 FW h1) in rice variety CO 43 at flowering stage. Treatment Control T1a T2 T3 Means of treatments LSD (5% level) CV (%) Chlorophyll a 1.96 2.00 2.8 2.05 2.28 0.352 11.6 Chlorophyll b 0.69 0.73 1.10 0.86 0.89 0.334 28.7 Total chlorophyll 2.65 2.73 3.91 2.91 3.19 0.448 34.2 a/b 3.12 2.82 2.72 2.56 2.70 ns 11.6 Soluble protein 25.00 18.25 16.04 11.75 15.34 4.098 16.7 NR activity 650 436 325 429 397 64.49 10.2

a T 1 = shade from transplanting to panicle initiation, T 2 = shade from panicle initiation to flowering, T3 = shade from flowering to harvest. ns = nonsignificant.

71

could be used as selection parameters to screen varieties for efficient photosynthesis at low light. A significant reduction in soluble protein content (36% and 43% in T2 and T3, respectively) was also observed. Protein synthesis in leaves is dependent initially on the grain reserves and, later, on carbon fixation by the leaf itself. A similar reduction in soluble protein by 27.6% was reported in shaded barley by Blenkinsop and Dale (1974), who reported that, when photosynthesis is prevented by shade, the amount of protein attained is low and overall turnover is biased in favor of breakdown so that the protein level decreases. Singh et al (1988) also reported a high correlation between yield reduction and ribulose biphosphate synthetase activity in rice. The results of NR activity show that it decreased markedly in T2. A reduction in NR activity at the reproductive stage was reported in shade-treated wheat plants by Prakash et al (1982). He suggested that, under shade, inhibition of NR was associated with an increase in peroxidase activity. Biochemical investigations strongly showed that the stages from PI to flowering and from flowering to harvest were more highly vulnerable to reduced light intensity than the early stage (transplanting to PI). Similar results were also reported in rice by Voleti et al (1991). There was a distinct difference between control and shaded plants in terms of yield (Table 2). Grain yield per hill was highest in the control (3.26 g) and lowest in T 3
72

(1.1 g), which is 33% of control yield. This is due to a lower reduction in spikelet fertility (31%) than that of the control (70.2%). The effects of low light on spikelet fertility and the consequent grain yield reduction were also observed in T 1 and T 2. The reduction was less in T 1. Yield reduction caused by low spikelet fertility and low grain number under low light was also reported by Patro and Sahu (1986), who also reported that the adverse effect of low light was less with plants exposed to shade during the vegetative stage than with those exposed during the reproductive stage. Low light also significantly reduced total dry matter per plant (TDMP) (Table 2), which could explain the lower harvest index in treatment means compared with the control. The reduction in TDMP was contributed by the reduced nitrogen metabolism as evident from the decreased NR activity and soluble protein content. Similar reports came

from Nayak and Murty (1980). There was no significant reduction in 1,000-grain weight. We conclude that low light has a detrimental effect on the physiology and yield of the rice plant. The adverse effect is more predominant at the reproductive stage than at the vegetative stage. Also, physiological parameters such as chlorophyll and soluble protein can be taken as selection indices for evaluating varieties grown under low light. We suggest that the rice crop should be exposed at least for 40 sunny days, particularly at heading, to obtain high fertility and optimum harvest index. References
Blenkinsop PG, Dale JE. 1974. The effect of shade treatments and light intensity on ribulose-1-5diphosphate carboxylase activity and fraction I protein level in the first leaf of barley. J. Exp. Bot. 25(88):899-912. Liu ZQ, Liu ZY, Ma DP, Zeng SF. 1984. A study of the relationship between chlorophyll content and

Table 2. Variability in dry matter and yield and yield-attributing characters of rice variety CO 43 under different light conditions. Treatment TDMP (g hill1) Grain yield per hill (g) 3.26 2.51 (77) 1.63 (50) 1.72 (33) 1.72 (53) 0.394 12.5 1,000 grain weight (g) 22.36 20.96 19.70 19.40 20.02 ns 15.7 Harvest index Panicles (no.) Spikelets Filled panicle1 spikelets (no.) panicle1 (no.) 50 50 48 52 50 7.01 11.8 35 30 21 16 22 4.16 10.9 Spikelet fertility (%) 70.2 60.2 43.4 31.0 45.0 7.67 35.8

Control T1a T2 T3 Treatment means LSD (5% level CV(%)

9.16 7.05 (77)b 4.67 (51) 3.84 (42) 5.18 (53) 0.697 8.02

0.36 0.35 0.34 0.28 0.32 0.070 13.4

3.6 3.4 3.6 3.0 3.3 1.67 10.2

a T 1 = shade from transplanting to panicle initiation, T 2 = shade from panicle initiation to flowering, T3 = shade from flowering to harvest. ns = nonsignificant. bNumbers in parentheses indicate percentage of control.

December 2004

photosynthetic rate in rice. Acta Agron. Sin. 10:57-62. Lowry OH, Rosenberg NJ, Fan AL, Randall RJ. 1951. Protein measurement with Folin reagent. J. Biol. Chem. 193:265-271. Murty KS, Nayak SK, Sahu G. 1975. Effect of low light stress on rice crop. In: Proceedings of the symposium on crop plant responses to environmental stresses. Almora: VPKAS. p 74-84. Nayak SK, Murty KS. 1980. Effect of varying light intensities on growth parameters in rice. Indian J. Plant Physiol. 23:309-316.

Nicholas JC, Harper JE, Hageman RH. 1976. Nitrate reductase activity in soybeans. I. Effects of light and temperature. Plant Physiol. 58:731735. Patro B, Sahu G. 1986. Effect of low light at different stages of the crop on sink size. Oryza 23:123-125. Prakash V. 1982. Effect of genotypes, moisture stress and shading on the activities of nitrate reductase and peroxidase in wheat leaves. Plant Physiol. Biochem. 9(1):41-47. Singh VP, Dey SK, Murty KS. 1988. Effect of low light stress on growth

and yield of rice. Indian J. Plant Physiol. 31(1):84-91. Voleti SR, Singh VP, Nayak SK. 1991. Effect of low light on physiological characters and consequent seed germination of rice cultivars. Plant Physiol. Biochem. 18(2):65-67. Yoshida S, Forno DA, Cock JH, Gomez KA. 1972. Laboratory manual for physiological studies of rice. Los Baos (Philippines): International Rice Research Institute. 70 p.

Nursery technology for early production of robust rice seedlings to transplant under integrated crop management
R. Rajendran and V. Balasubramanian, IRRI; V. Ravi, K. Valliappan, and T. Jayaraj, Soil and Water Management Research Institute (SWMRI), Thanjavur, Tamil Nadu, India; and S. Ramanathan, Tamil Nadu Agricultural University, Coimbatore, Tamil Nadu, India

The recently introduced integrated crop management (ICM) for rice has modified certain crop management practices in the system of rice intensification (SRI) developed in Madagascar to enhance rice productivity and increase farmers profits. It involves transplanting of young (14- to 15-d-old) seedlings at wider spacing, mechanical weeding, intermittent irrigation, etc. Among these, the use of young seedlings is the single most important practice contributing to high yields. Planting of 15-dold seedlings at one seedling per hill produced higher rice yield than planting 21-d-old seedlings in North Sumatra, Indonesia (Makarim et al 2002). In the conventional wetbed nursery, seedlings do not attain the expected size (15- to 20-cm height) even at 2530 d after seeding (DAS)
IRRN 29.2

because of invariably high seeding rates and poor nursery management. In Tamil Nadu, India, the seed rate used for the wetbed nursery is as high as 80100 kg per 800-m2 nursery area to plant 1 ha, and the resultant seedlings are naturally thin and delicate. Therefore, we decided to develop a modified rice mat nursery (MRMN) for producing robust, healthy rice seedlings in 15 d by evaluating different soil + manure + rice husk mixes as a medium for MRMN. A few of the nursery management details such as thickness of the medium, lining of the seedbed, seed rate, and nutrient needs were standardized by nonreplicated preliminary trials. Three field trials on MRMN were conducted at SWMRI, Thanjavur, and at the Tamil Nadu Rice Research Institute during Feb-Mar 2003

and May 2003. The experimental design used was a randomized block with seven treatments and three replications for the first trial (at both locations) and four treatments and five replications for the second trial (SWMRI). The plot (nursery bed) size was 1.5 m 2. In the first trial, seven treatments involving various seedbed media were tested (Table 1). The prepared nursery media were blended with well-powdered diammonium phosphate (DAP) at 0.5 g kg 1 of medium. White polyethylene sheets were laid over a 1.2-m-wide raised bed (5 cm high) of the required length. The enriched medium was poured inside a prefabricated wooden frame placed over the polyethylene layer. Pregerminated seeds of rice variety ADT43 were sown on 26 Feb 2003 at the rate of 5 kg
73

Table 1. Growth characteristics of 15-d-old rice seedlings in the modified rice mat nursery, SWMRI farm, Thanjavur, and TRRI farm, Aduthurai, Tamil Nadu, India, Feb-Mar 2003. Treatment SWMRI, Thanjavur Root length (cm) 5.7 9.7 4.7 6.6 5.3 9.9 5.5 1.8 Seedling height (cm) 20.5 18.3 14.3 17.6 14.5 20.8 16.7 3.0 Leaves seedling1 (no.) 4.5 5.0 3.0 5.2 3.4 4.2 3.8 1.3 Leaf length (cm) 13.7 11.0 8.7 12.2 9.2 12.9 9.2 1.7 Seedling vigor index 5.2 5.4 4.0 5.2 4.9 6.1 4.3 naa Root length (cm) 6.7 7.1 5.1 8.0 5.9 10.4 6.4 0.6 TRRI, Aduthurai Seedling height (cm) 21.0 19.0 12.9 18.2 15.8 21.6 17.0 0.6 Leaves seedling1 (no.) 4.1 4.6 3.4 4.3 3.9 4.5 4.1 0.1 Leaf length (cm) 12.4 11.5 9.1 12.0 12.7 13.1 9.4 0.5 Seedling vigor index 5.8 6.0 5.6 6.1 5.9 6.4 5.6 0.5

Soil alone Press mud alone Farmyard manure (FYM) alone Soil + press mud (2:1 w/w) Soil + FYM (2:1 w/w) Soil + press mud (1:1 w/w) Soil + FYM (1:1w/) LSD (5%)
a

na = not statistically analyzed.

Table 2. Growth characteristics of 15d-old rice seedlings grown in the modified rice mat nursery, TNRRI farm, Aduthurai, Tamil Nadu, India, May 2003. Treatment Root Seedling Leaves Seedling length height seedling1 vigor (cm) (cm) (no.) index 5.7 3.8 15.7 11.5 3.6 3.3 5.6 4.0

Soil alone Soil 70% + farmyard manure 20% + rice husk 10% Soil 50% + press mud 50% Soil 90% + rice husk 10% LSD (5%)
a

7.5 6.0 0.8

17.0 14.1 1.1

3.9 3.8 0.5

8.2 6.1 naa

na = not statistically analyzed.

dry seed per 100 m 2 (equal to 50 g dry seed or 100 g wet seed per m 2). The seeds were properly covered with the medium and water was sprinkled uniformly to saturate the medium. The seed frame was carefully removed and the sowing process was continued. The mat nursery was watered 23 times a day up to 5 d after sowing. After
74

the 5th day, water was let in between the mat nursery beds to keep the beds soaked. On the 9th day, 0.5% urea solution was sprinkled on the nursery beds. Water was drained 2 d before lifting and transportation of seedling mats to the main field for transplanting. Ten randomly selected seedlings were carefully lifted on the 14th day for seedling growth observations. Seedling vigor index was calculated using seedling dry weight and % germination of seed (Abdul-Baki and Anderson 1973). In the second trial, the treatments were further modified based on observations during the first trial. Another experiment was conducted in May 2003 at the TNRRI farm, this time, with only four treatments (Table 2). ADT43 was sown on 2 May 2003. The same nursery management practices and recording of observations were done. To assess seedling characteristics, seedlings were pulled out on 16 May 2003.

Rice seedlings grown with soil + press mud (1:1 w/ w) exhibited the best seedling characteristics (Table 1), with root length of 9.9 cm and 10.4 cm and seedling height of 20.8 cm and 21.6 cm, respectively, at Thanjavur and Aduthurai. Seedling vigor index was 6.1 at Thanjavur and 6.4 at Aduthurai for the soil + press mud (1:1 w/w) mix. This treatment was found suitable for producing goodquality seedlings in MRMN. In the TNRRI revised trial, seedlings grown in soil + press mud (1:1 w/w) medium produced the longest roots (7.5 cm), with a seedling height of 17.0 cm and a seedling vigor index of 8.2. Root length was 6.0 cm and seedling vigor index was 6.1 in the soil + rice husk medium, followed by soil-alone treatment (Table 2). Thus, either soil + press mud or soil + rice husk were suitable for producing robust and healthy seedlings in 15 d. Overall, MRMN reduces the cost of producing seedDecember 2004

lings that may be planted in 1 ha of main field by about US$35, representing a 50% savings compared with a conventional wetbed nursery (data not shown). Soil alone or the soil + rice husk mixture (9:1 w/w) or soil + press mud mixture (1:1 w/w) could serve as a good medium for the mat nursery. The seedlings reach a height of 1820 cm with four leaves in 15 d. The seedling mat can be easily transported

to the main field for transplanting. The MRMN reduces the land requirement (by 88%) and water use (by 55%), and saves on seed cost (8590%), fertilizer (90%), and labor (34%). It is therefore highly economical and very affordable to farmers. References
Abdul-Baki AA, Anderson JD. 1973. Vigor determination in soybean seed by multiple criteria. Crop Sci. 13:630-633.

Makarim AK, Balasubramanian V, Zaini Z, Syamsiah I, Diratmadja H, Arafah IGPA, Wardana IP, Gani A. 2002. System of rice intensification (SRI): evaluation of seedling age and selected components in Indonesia. In: Bouman B, Hengsdijk H, Hardy B, Bindraban PS, Tuong TP, Ladha JK, editors. Water-wise rice production. Manila (Philippines): International Rice Research Institute and The Netherlands: Plant Research International. p 119-127.

Influence of high- and low-temperature treatments on seed germination and seedling vigor of coarse and fine rice
M. Farooq, S.M.A. Basra, K. Hafeez, and E.A. Warriach, Department of Crop Physiology, University of Agriculture, Faisalabad 38040, Pakistan E-mail: tulip_farooq@yahoo.co.in

Rapid and uniform crop stand is a prerequisite for betterquality produce. If seeds germinate erratically over a long time, seedling growth will not be uniform and plants will mature over a wider period. Seed invigoration treatments are therefore developed to improve seed performance during germination and emergence. Dry-heat treatment of seeds is used to control external and internal seedborne pathogens such as fungi, bacteria, viruses, and nematodes (Nakagawa and Yamaguchi 1989), and to break seed dormancy (Zhang 1990). In general, the high temperature in this treatment reduces seed viability and seedling vigor, but the optimum temperature for breaking dormancy promotes seed germination and seedling emergence in cereal crops (Lee et al 2002) and cotton (Basra et al 2003). The deIRRN 29.2

gree of promotion of seed germination by dry-heat treatment showed wide intraspecific variation (Herranz et al 1998). Nonlethal heat shock treatment before radicle emergence effectively promoted the germination of barley, wheat, and cotton seeds (Dell Aquilla and Di Turi 1996, Dell Aquilla et al 1998, Basra et al 2003). Presowing chilling treatments are being used effectively alone or with other invigoration techniques to shorten the period between planting and emergence and to protect the seeds from abiotic and biotic stresses during the critical phase of seedling establishment (Basra et al 2002). Our study aimed to study the influence of highand low-temperature treatments on the germination and seedling vigor of both coarse and fine rice. Seeds of coarse rice cultivar KS-282 (initial

seed moisture 7.89%) and fine rice Super Basmati (8.06% seed moisture) were used. For the dry-heat treatment, 250 g of seed of each cultivar were incubated at 40 C for 72 h or 60 C for 24 h in an oven (EYLA forced-air oven, WFO600 ND, Rikakikai Co. Ltd., Tokyo, Japan). Seeds were incubated in tightly covered glass jars to avoid evaporation losses. In the chilling treatment, a weighed quantity (250 g) of both coarse and fine rice was sealed in polythene bags and placed in a refrigerator at 19 C for 72 h. Seeds were sown in petri dishes between layers of moist filter paper at 27 C in the incubator. Germination was observed daily following the AOSA method (AOSA 1990). Germination index (GI) was calculated as described in AOSA (1983):
75

GI =

No. of germinated seeds Days of first count

No. of germinated seeds Days of final count

The energy of germination (GE) was recorded on the 4th day after planting. It is the percentage of germinating seeds 4 d after planting relative to the total number of seeds tested (Ruan et al 2002). After washing in distilled water, 5 g of seeds were soaked in 50 mL of distilled water at 25 C. Electrical conductivity of steep water was measured 0.5, 1.0, 1.5, 2.0, 6.0, 12.0, and 24.0 h after soaking using a conductivity meter and expressed as S cm 1 (Ashraf et al 1999). The results showed that presowing temperature treatments significantly (P < 0.05) affected the germination vigor of both coarse and fine rice seeds (see table). Significantly higher values of time to start germination, mean germination time, and time to get 50% germination in fine rice were observed in seeds subjected to dry-heat treatment at 60 C for 24 h compared with other treatments, including the control (see table). However, all treatments resulted in lower final germination percentage compared with that of the control. Statistically maximum GI and GE were noted in fine rice seeds subjected to dry-heat treatment at 40 C for 72 h. For coarse rice, a similar trend was noted the highest final germination percentage (FGP) was noted in seeds subjected to chilling (statistically similar to that of the control); however, dryheat treatment resulted in FGP lower than that of the control.
76

The dry-heat treatments resulted in lower electrical conductivity of seed leachates compared with that of the control in fine rice seeds. The chilling treatments resulted in a higher rate of seed leachates (see figure, a). In coarse rice, presowing dry-heat treatments resulted in a higher rate of seed leachates compared with the control. The highest rate of seed leachates was noted in seeds treated with dry-heat at 60 C for 24 h. However, chilling resulted in decreased electrical conductivity of seed leachates compared with that of the control (see figure, b).
EC of seed leachates 600 A
Control Chilling

The effect of dry-heat treatment on seed germination depended on dry-heat intensity and duration of exposure. Both coarse and fine rice had the same germination vigor. Dry heat at 60 C for 24 h resulted in reduced vigor compared with that of the control (see table). This may be due to membrane deterioration as indicated by the higher EC of seed leachates (see figure, b). Similar results were reported in rice (Lee et al 2002) and legumes (Sacheti and Sacheti 1996). However, both high- and low-temperature treatments could not invigorate rice seeds as de-

450

Heat treatment at 40oC for 72 h Heat treatment at 40oC for 24 h

300

150

0 500 B 375 250 125 0

0.5

1.5

12

24

Soaking period (h)

Effect of presowing high- and low-temperature treatments on the EC of seed leachates in (a) fine and (b) coarse rice seeds.

December 2004

Effect of high- and low-temperature treatments on the germination vigor of coarse and fine rice. Treatment Time to start germination (d) 4b 4b 4b 6a 1.033 4b 4b 4b 6a 1.033 T50 (d) MGT (d) Final Germination Energy germination index of percentage germination (%) 92.50 a 76.92 c 84.00 b 69.23 d 5.461 92.30 a 94.00 a 84.61 b 38.46 c 6.112 11.38 a 9.460 b 11.23 a 5.830 c 1.031 13.76 a 13.00 a 11.00 b 3.830 c 1.003 38 b 30 c 53 a 0d 4.273 84 a 76 b 46 c 00 d 5.213

Control Chilling Heat treatment at 40 C for 72 h Heat treatment at 60 C for 24 h LSD at 0.05 Coarse Control rice Chilling Heat treatment at 40 C for 72 h Heat treatment at 60 C for 24 h LSD at 0.05

Fine rice

5.43 b 5.35 b 3.85 c 8.13 a 1.001 3.59 b 3.65 b 4.00 b 7.30 a 0.997

5.16 b 5.20 b 4.80 b 7.4 a 0.471 4.20 c 4.50 c 4.90 b 7.0 a 0.334

Numbers not followed by the same letter differ significantly at P = 0.05.

IRRN 29.2

77

picted by the GI and FGP values in both types of seed and by the GE in coarse rice. This trend was confirmed by the EC of seed leachates, which indicated that, because of better membrane repair, a lower rate of seed leachates resulted in chilled coarse rice and dry heat-treated seeds at 40 C for 72 h in fine rice seed (see figure). This indicates a genetic difference between the two rice types, confirming earlier observations that rice cultivars respond differently to dry-heat treatments (Lee et al 2002). In the case of the EC of seed leachates, chilling resulted in a higher rate in fine rice, whereas a lower rate was seen in coarse rice (see figure), probably the result of better membrane repair in coarse rice and membrane deterioration in fine rice, as reported earlier (Dadlani and Seshu 1990). Contrasting results were observed in seeds subjected to dry-heat treatment at 40 C for 72 h compared with

chilled onesdry-heat treatment at 40 C for 72 h resulted in a higher EC of seed leachates in coarse rice because of better membrane repair and a lower rate of seed leachates in fine rice. This again indicates genetic differences between the two rice types. Presowing temperature treatments have a significant effect on the germination and seedling vigor of both coarse and fine rice. In our study, both types of rice responded differently to temperature treatments. Chilling treatment in coarse rice and dry-heat treatment in fine rice have potential for vigor enhancement; further research is needed to explore the exact protocols. References
Ashraf M, Akhtar N, Tahira F, Nasim F. 1999. Effect of NaCl pretreatment for improving seed quality of cereals. Seed Sci. Technol. 20(3):435-440. AOSA (Association of Official Seed Analysts). 1983. Seed vigor testing

handbook. Contribution No. 32. AOSA. AOSA (Association of Official Seed Analysis). 1990. Rules for testing seeds. J. Seed Technol. 12(3):1-112. Basra SMA, Zia MN, Mehmood T, Afzal I, Khaliq A. 2002. Comparison of different invigoration techniques in wheat (Triticum aestivum L.) seeds. Pak. J. Arid Agric. 5(2):11-16. Basra SMA, Ashraf M, Iqbal N, Khaliq A, Ahmad R. 2003. Physiological and biochemical aspects of presowing heat stress on cotton seed. Seed Sci. Technol. 33(3). (in press) Dadlani M, Seshu DV. 1990. Effect of wet and dry heat treatment on rice seed germination and seedling vigor. Int. Rice Res. Newsl. 15:21-22. Dell Aquilla A, Corona MG, Di Turi M. 1998. Heat-shock proteins in monitoring ageing and heatinduced tolerance in germinating wheat and barley embryos. Seed Sci. Res. 8:91-98. Dell Aquilla A, Di Turi M. 1996. The germination response to heat and salt stress in evaluating vigor loss in aged wheat seeds. Seed Sci. Technol. 23:551-561. Herranz JM, Ferrands P, Martinez SJJ. 1998. Influence of heat on seed germination of seven Mediterranean leguminosae species. Plant Ecol. 135:95-103. Lee SY, Lee JH, Kwon TO. 2002. Varietal differences in seed germination and seedling vigor of Korean rice varieties following dry heat treatments. Seed Sci. Technol. 30:311-321. Nakagawa A, Yamaguchi T. 1989. Seed treatments for control of seed-borne Fusarium roseum on wheat. JARQ 23:94-99. Ruan S, Xue Q, Tylkowska K. 2002. The influence of priming on germination of rice (Oryza sativa L.) seeds and seedling emergence and performance in flooded soils. Seed Sci. Technol. 30:61-67. Sacheti U, Sacheti U. 1996. Effect of extremely high temperature on the germination of seeds of some leguminous species and semi-arid areas of Oman. Indian J. Ecol. 23:29-33. Zhang XG. 1990. Physiochemical treatments to break dormancy in rice. Int. Rice Res. Newsl. 15:22.

A simple screening technique for salinity tolerance in rice: germination rate under stress
D. Sumith de Z. Abeysiriwardena, Rice Research and Development Institute, Batalagoda, Ibbagamuwa, Sri Lanka

Currently available screening procedures for tolerance for salinity in rice have their own limitations (Ponnamperuma 1977, Aslam et al 1993). Hence, an investigation was carried out to develop a lowcost, simple, reliable, and efficient laboratory procedure to screen rice cultivars for this trait. The hypothesis tested was whether cultivar differences in ability to sustain seed viability and show normal germination when presoaked in saline solutions with high concentrations for several days under laboratory condiSeed germination (%) (transformed data) 100

tions are associated with level of salinity tolerance in the field. A two-factor (cultivar and time of soaking in a saline solution) factorial experiment was laid out in a randomized complete block design with three replications under laboratory conditions. Ten cultivars with a known reaction to salinity consisting of four tolerant (Pokkali, Nona bokra, At354, and At 401) and six susceptible (Bg450, Bg94-1, Bg350, Bg403, Bg304, and Bg300) varieties were used in the experiment.

80

60

Pokkali Nona bokra At 354

40

At 401 Bg 450 Bg 403

20

Bg 304 Bg 94-1 Bg 300 Bg 350

Seed soaking time in saline solution (d)

Fig. 1. Seed germination rates of 10 rice cultivars with increasing days of seed soaking in saline solution with EC 45 dS m1 in the first experiment. Bars represent standard error.

Seeds of cultivars with initial germination within the 96100% range were soaked in a common salt solution with a high concentration specific electrical conductivity (EC) was as high as 45 dS m1. Samples of 50 seeds were removed from the salt solution at 0 (control), 3, 6, and 9 d after soaking, thoroughly washed with tap water, and allowed to germinate on wet blotting paper in petri dishes. After 5 d, the percentage of normal seed germination in each petri dish was recorded. The experiment was repeated once again without the 3-d soaking-time treatment. Percentage data were transformed using arcsine transformation before analysis. In the analyses of variance of both experiments, the two-way interaction effect of cultivar time of soaking was found to be significant. The interaction effect was thus further studied using response curves plotted with transformed data (Figs. 1 and 2). For all cultivars, seed germination decreased as days of presoaking in saline solution increased. However, the rate of decrease in germination varied among cultivars. Interestingly, germination rates of tolerant cultivars (Pokkali, At354, At401, and Nona bokra) were found to be much higher than those of susceptible cultivars in both experiments. Pokkali had the
December 2004

78

IRRN 29.2

79

highest germination rate within the tolerant group. Thus, Pokkali could be ranked highly tolerant and Nona bokra, At401, and At354 could be ranked tolerant. In the susceptible group, based on the germination rates, Bg450 appeared to be moderately susceptible; Bg94-1, Bg300, Bg403, and Bg350 appeared to be susceptible; and Bg304 appeared to be highly susceptible to salinity. These results are in perfect agreement with the known reactions of these cultivars to salinity in the field. In addition, mean comparisons among cultivars at 6 and 9 d of soaking showed that the 9-d soaking treatment would be the best in clearly identifying cultivars highly tolerant, tolerant, and susceptible to salinity (Figs. 1 and 2). Thus, the proposed laboratory screening procedure appeared simple, effective, reliable, and comparatively less expensive and less cumbersome.

Seed germination (%) (transformed data) 100

80

60
Pokkali Nona bokra

40

At 354 At 401 Bg 450 Bg 403

20

Bg 304 Bg 94-1 Bg 300 Bg 350

0 0

3 6 Seed soaking time in saline solution (d)

Fig. 2. Seed germination rates of 10 rice cultivars with increasing days of seed soaking in saline solution with EC 45 dS m 1 in the second experiment. Bars represent standard error.

References

Aslam M, Qureshi RH, Ahmed N. 1993. A rapid screening technique for salt tolerance in rice (Oryza sativa L.). Plant Soil 150:99-107.

Ponnamperuma FN. 1977. Screening rice for tolerance for mineral stresses. IRRI Res. Pap. Ser. 6. 21 p.

Osmohardening: a new technique for rice seed invigoration

S.M.A. Basra, M. Farooq, K. Hafeez, and N. Ahmad, Department of Crop Physiology, University of Agriculture, Faisalabad 38040, Pakistan E-mail: tulip_farooq@hotmail.com

Direct seeding could be an attractive alternative to transplanting of rice (Balasubramanian and Hill 2002), but poor germination, uneven crop stand, and high weed infestation are among the main constraints to its adoption (Du and Tuong 2002). Seed priming is an effective technique for rapid and uniform seed germination of several cereal crops (Basra et al 2004, Farooq et al 2004). Seed hardening, also called wetting and drying or hydration/dehydration, is done by repeated soaking and drying of seeds in water (Pen Aloza and Eira 1993). Osmoconditioning is a special type of seed invigoration that has been used for slow hydration of seeds in aerated, low-water-potential solution (Bradford 1986). In earlier studies, hardening (Basra et al 2003) and osmopriming (Lee and Kim 1999) were found to be effective invigoration tools in rice. Seeds are hardened in tap or distilled water (Lee et al 1998); furthermore, seed priming is performed in a single cycle of wetting and drying (Lee and Kim 1999). These studies aimed to explore the possibility of rice seed priming by two cycles of wetting and drying (osmohardening) like seed hardening in low-water-potential solutions.
80

Seeds of coarse rice cultivar KS282 and fine rice cultivar Super Basmati were used. A weighed quantity of seeds (250 g) was soaked in solutions having 22.2 g CaCl 2 L1, 16.4 g NaCl L1, 30 g KNO3 L1, and 20.74 g KCl L1 or tap water at 27 C for 24 h, followed by redrying to initial moisture under shade with forced air. The osmotic potential of all these solutions was 1.25 MPa (except tap water for hardening). The cycle was repeated twice (Basra et al 2003). The ratio of seed weight to solution volume was 1:5 (g mL1) (Ruan et al 2002). Control and treated seeds were sown in plastic trays (25 in each) having moist sand, replicated four times, and placed in a growth chamber (Windon, England). Day and night lengths were kept at 15 and 9 h, with 30 and 24 C, respectively. Relative humidity was maintained at 70%. Emergence was recorded daily according to the seedling evaluation described in AOSA (1990). Maximum emergence percentage was seen in CaCl 2osmohardened seeds. This was statistically similar to that of hardened and NaClosmohardened seeds. Maximum root-shoot ratio was recorded in the control, which was similar to that of NaCl-

and KCl- osmohardened seeds. The significantly highest seedling fresh weight was observed in CaCl 2-osmohardened seeds, followed by NaCl-osmohardened seeds (see table). Minimum seedling fresh weight was noted in the control; all treatments resulted in higher seedling fresh weight. Maximum seedling dry weight was obtained in seeds osmohardened with KNO 3, followed by CaCl 2osmohardened seeds. The maximum emergence percentage was noted in the control, which was identical to that of CaCl2- and KCl-osmohardened seeds, whereas the minimum emergence percentage was observed in seeds osmohardened with KNO 3 (see table). The highest root-shoot ratio was noted in the control, which was similar to that of CaCl 2-, KCl-, and NaClosmohardened seeds. The lowest value of this ratio was seen in hardened seeds, which was identical to that seen in seeds osmohardened with KNO3 (see table). The significantly highest seedling fresh weight was noted in the control; all the seed treatments resulted in lower fresh weight, but minimum fresh weight was recorded in seeds osmohardened with KNO 3. Maximum seedling dry
December 2004

weight was obtained in hardened seeds, followed by untreated seeds. However, minimum dry weight was recorded in seeds osmohardened with KNO 3. An improvement in seedling vigor was observed following osmohardening treatments (see table). This is independent of the nature of the salt use in both coarse and fine rice as earlier reported by Ruan et al (2002) (in the case of coarse rice). More vigorous seedlings were observed in and CaCl 2-osmohardened hardened seeds than in all other treatments as indicated by the higher values of emergence percentage and seedling fresh weight in both coarse and fine rice (see table 1). Ruan et al (2002) found that CaCl 2- osmoprimed seeds resulted in increased vigor when treated rice seeds were direct-seeded under flooded conditions. KCl-osmohar-dened and untreated seeds in both types of rice germinated on the same day (see table). In coarse rice, the performance of KClosmohar-dened seeds did not support the earlier findings of Du and Tuong (2002), who reported that KCl was the best of all the salts when used on direct-seeded rice. Osmopriming and hardening can thus successfully be integrated to enhance the vigor of coarse and fine rice. Osmohardening with CaCl 2 and simple hardening performed better than all other treatments, including the control, in both coarse and fine rice.

Effects of seed treatment on seedling vigor of coarse and fine rice.a Treatment FEP (%)b Root/shoot Seedling Seedling fresh weight dry weight (g) (g) 1.23 e 2.1 d 3.31 a 3.15 b 2.86 c 2.91c 0.112 3.99 a 3.00 c 3.15 b 2.97 c 1.98 e 2.83 d 0.1212 0.84 d 1.97 c 2.15 b 1.94 c 2.44 a 2.091b 0.095 2.28 b 2.33 a 2.07 c 2.06 c 1.59 d 1.99 c 0.0986

Fine rice

Coarse rice

Control Hardening CaCl2 NaCl KNO3 KCl LSD at 0.05 Control Hardening CaCl2 NaCl KNO3 KCl LSD at 0.05

41.415 e 80.23 a 83.19 a 76.97 ab 66.75 d 72.19 c 4.342 92.47 a 84.00 b 90.40 a 79.10 bc 58.75 d 86.25 ab 6.234

1.66 a 0.97c 0.90 c 1.31 a 0.98 c 1.26 ab 0.213 1.45 a 1.05 c 1.37 a 1.27 ab 1.07 c 1.27 ab 0.1782

Numbers not followed by the same letter differ significantly at P = 0.05. bFEP = final emergence percentage.

References

AOSA (Association of Official Seed Analysts). 1990. Rules for testing seeds. J. Seed Technol. 12:1-112. Balasubramanian V, Hill JE. 2002. Direct seeding of rice in Asia: emerging issues and strategic research needs for the 21st century. In:. Pandey S, Mortimer M, Wade L, Tuong TP, Lopez K, Hardy B, editors. Direct seeding: research strategies and opportunities. Manila (Philippines): International Rice Research Institute. p 241-256. Basra SMA, Farooq M, Khaliq A. 2003. Comparative study of pre-sowing seed enhancement treatments in fine rice (Oryza sativa L.). Pak. J. Life Soc. Sci. 1:5-9. Basra SMA, Farooq M, Tabassum R, Ahmed N. 2004. -amylase activity, sugar content, germination and seedling vigor of primed coarse rice seed (Oryza sativa L.). Seed Sci. Technol. (in press) Bradford KJ. 1986. Manipulation of seed water relations via osmotic priming to improve germination under stress conditions. Hortic. Sci. 21:1105-1112. Du LV, Tuong TP. 2002. Enhancing the performance of dry-seeded rice: effects of seed priming, seedling rate, and time of seeding. In: Pandey S, Mortimer M, Wade L, Tuong TP, Lopez K, Hardy B, editors. Direct seeding: research strategies and opportunities. Manila (Philippines): International Rice Research Institute. p 241-256.

Farooq M, Basra SMA, Tabassum R, Ahmed N. 2004. Physiological and biochemical aspects of seed vigor enhancement treatments in fine rice (Oryza sativa L.) cv. Basmati 385. Seed Sci. Technol. (in press) Lee SS, Kim JH. 1999. Morphological change, sugar content, and amylase activity of rice seeds under various priming conditions. Kor. J. Crop. Sci. 44:138-142. Lee SS, Kim JH, Hong SB, Yun SH. 1998. Effect of humidification and hardening treatment on seed germination of rice. Kor. J. Crop. Sci. 43:157-160. Pen Aloza APS, Eira MTS. 1993. Hydration-dehydration treatments on tomato seeds (Lycopersicon esculentum Mill). Seed Sci. Technol. 21:309-316. Ruan S, Xue Q, Tylkowska K. 2002. The influence of priming on germination of rice (Oryza sativa L.) seeds and seedling emergence and performance in flooded soils. Seed Sci Technol. 30:61-67.

IRRN 29.2

81

Socioeconomics

Status of rice seed production in Sri Lanka: some bottlenecks

M. Wijeratna and W.N. De Silva, Department of Agricultural Economics, Faculty of Agriculture, University of Ruhuna, Mapalana, Sri Lanka

82

December 2004

In many rice-growing seasons, the government of Sri Lanka has been encountering the problem of inadequate rice seed supply. To solve this problem, the government welcomed private seed producers and encouraged farmers to produce their own seed. The Department of Agriculture has set quality standards for seed producers to follow. A seed certification center was established at Bata-ata, in the southern district of Hambantota, to test the rice seed samples. A series of case studies was conducted in Matara and Hambantota districts to identify the bottlenecks in seed production and evaluate the seed certification procedure. The Bata-ata seed certification center was chosen for the study. The investigations were done during the major rice cultivation season (maha) in 2000-01. The results of the study show that a significant percentage of the samples tested did not conform to accepted

seed quality standards. In fact, the rate of rejection was very high among private growers32,712 kg of seed out of 42,021 kg were rejected. The accepted percentage was only 22.2 (9,309 kg). Even on government farms, only 58% (20,008 kg) of the seed was accepted. These findings reveal that seed producers have not maintained the required quality standards for rice seed. This has affected the quality of seed produced and this neglect on the part of rice farmers depressed its market price. The main limitations have to do with low laboratory standards pertaining to percentage of moisture, mixing of rice types, high impurity content, immature grains, damaged grains, discolored grains, and milling recovery. The seed producers were not prepared to do proper seed production management in the field to ensure good seed that will pass quality standards. To remedy this situation, it is essential that an intensified ag-

Status of seed certification involving different institutes in Sri Lanka. Institute Total Quanno. of tity samples tested (kg) 17 52 39 13 11 34,747 146,616 111,549 19,516 27,593 QuanPertity centage accepted accepted (kg) 20,008 55,132 30,846 4,305 6,109 57.6 37.6 27.7 22.1 22.1

Government farms Hungama MPCSa Tissa MPCSa Sarvodaya Ruhunu farmers associationTissa DoAb Hambantota Private growers
a

20 20

54,366 42,021

10,619 9,309

19.5 22.2

Multipurpose Cooperative Society. bDepartment of Agriculture.

ricultural extension program for seed producers be developed, focusing on appropriate cultural practices to produce high-quality rice seed. Acknowledgment The authors acknowledge the National Research Council for its assistance under Research Grant No. 99-38.

Farmers participatory evaluation of saline-tolerant rice varieties

L. Ponnusamy and N. Thenmathi, Central Institute of Brackishwater Aquaculture, Indian Council of Agricultural Research, 75 Santhome High Road, Chennai 600028, Tamil Nadu, India

Promising salt-tolerant lines should undergo on-farm evaluation to validate recommendations for their wider adoption (Jauhar Ali et al 1999). Low rice yield caused by soil salinity and water stress was assessed by farmers through a participatory rural appraisal (PRA) under a World Bank-funded National Technology Project (NATP) called Institution Village Linkage Program for Technology Assessment and Refinement in the Coastal AgroEcosystem of Tiruvallur District of Tamil Nadu. We conducted the present investigation with farmers to assess the performance of rice varieties (Trichy 1 and Trichy 2) grown by farmers and to evaluate their salinity tolerance. PRA tools were used for problem identification, field trials, monitoring and evaluation, and sending farmers feedback to the research and extension system. The field study was carried out in Kattur, Tiruvallur, Tamil Nadu, during samba (Aug 2001 to Jan 2002). The farmers grow White Ponni, ADT43, and BPT5204 and are dependent on rainfed tank irrigation. These varieties are not salttolerant. Weather is moderately warm, with rainy periods between October and December. Soil of the village is moderately drained clay loam with a pH of 7.4. The soil has low available N, meIRRN 29.2

dium P, and high K content. A matrix ranking was used to compare the three rice varieties. Ranking was carried out by a group of farmers and triangulated (cross-checked) with other group extension functionaries. Farmers preferred White Ponni because of its stable yield, profitability, taste, and slender grain (Table 1). Though BPT5204 gave better yield and profit, it was more susceptible to pests and diseases. However, these varieties did not tolerate saline conditions. At the same time, whenever there is not enough rainfall during the northeast monsoon, water stress reported during the end of crop growth results in low rice yield. Near the sea coast, where soil salinity is high, the yield of White Ponni was low. Farmers thus felt the need for saline-tolerant, shortduration rice varieties to mitigate the problems of water stress and salinity. Seeds of saline-tolerant varieties Trichy 1 and Trichy 2 were obtained from the Agricultural College and Research Insititute, Tamil Nadu Agricultural University. Onfarm trials were conducted in 10 farmers fields (area was 0.4 ha each). White Ponni was taken as the control. The grain yield of Trichy 1 increased by 8% and that of Trichy 2 by 4.8% compared with that of the control.

Table 1. Matrix ranking of existing rice varieties by farmers in the study village.a Characteristic White Ponni BPT5204 ADT43

Yield 5 3 2 Yield stability 5 3 2 Risk in cultivation 1 5 3 Profitability 5 5 3 Seed availability 5 5 5 Susceptibility 1 5 4 to pests and diseases Tolerance for 5 2 4 weed infestation Tolerance for 4 1 3 water stress Labor use 5 2 2 Compatibility 4 1 2 with weather Taste 5 3 4 Digestibility 4 5 4 Grain size Slender Short, bold Bold Straw yield 5 3 4 Parched rice5 3 4 making quality Duration (d) 155160 140 115120 Lodging/ Lodging Nonlodging Nonlodging nonlodging type
a On a scale of 15, where 1 = least/minimum and 5 = maximum/higher/more/best. Yield scores are based on visual assessment by farmers.

Table 2. Evaluation of saline-tolerant varieties. Item Farmers Trichy 1 Trichy 2 variety (White Ponni) 268 98.2 15.2 3.75 5.17 23.23 228.33 578.31 349.98 2.53 255 89.7 24.3 4.05 5.54 26.13 224.58 524.48 299.89 2.33 250 95.4 22.4 3.93 5.23 32.75 200.77 507.75 306.98 2.52

Panicles m2 (no.) Filled grains panicle1 (no.) 1,000-grain weight (g) Grain yield (t ha1) Straw yield (t ha1) Per day productivity (kg) Returns (US$ ha1)a Cost of cultivation Gross returns Net returns Benefit-cost ratio
a

US$1 = 48 rupees.

83

Similarly, the respective straw yields (7.15% and 1.06%) and productivity (12.48% and 40.9%) of Trichy 1 and Trichy 2 were significantly higher than those of the control (White Ponni). But the net returns from growing White Ponni were higher because of its fine grain, whereas the Trichy varieties fetched a low market price because of their coarse grain. Further, one-way analysis of variance was carried out with these three rice varieties to study the differences in mean yield performance. The

results showed that yields of these three varieties differed significantly from each other at the 2% level of significance. Duncans multiple range test was conducted to achieve multiple comparisons. The analysis revealed that the varieties are different. Trichy 1 (4.5 0.05) had better yield than Trichy 2 (3.39 vs 0.98) and White Ponni (3.75 vs 0.44). The farmers were also convinced of the satisfactory performance of Trichy 1 and Trichy 2. We need welldesigned on-farm trials to evaluate the performance of

Trichy 1 and Trichy 2 under saline and water-deficit conditions. Reference

Jauhar Ali A, Somasundaram E, Nanamohammed SE, Rajagopalan R, Rajukkannu K, Subramanian M. 1999. Rice in coastal saline soils of Tamil Nadu. In: Vistas of rice research. Aduthurai, Tamil Nadu: Tamil Nadu Rice Research Institute.

Acknowledgment

This project is supported by the Indian Council of Agricultural Research under the National Agricultural Technology Project Fund.

Deepwater agroecosystem characterization and system diagnosis of farmers problems in Bihar, India
U.P. Singh, Department of Agronomy, Institute of Agricultural Sciences, Banaras Hindu University, Varanasi 221005, Uttar Pradesh, India E-mail: u_psingh@sify.com/upratap@banaras.ernet.in

The deepwater rice (DWR) environment is so variable and complex with regard to hydrology that it cannot be duplicated at the research center, where technology is generated. Consequently, the technology does not perform well in farmers fields. In certain areas, on-station technology may be adopted by farmers, but it fails completely in other areas. There is therefore a need to classify the different agroecosystems for on-station and on-farm research. The DWR area in Bihar constitutes about 1.1 million ha, with productivity averaging 1.0 t ha1. This area falls under one of the most complex, diverse, fragile, and risk-prone environments.
84

An agroecosystem analysis of flood-prone DWR environments of north Bihar showed that rice is the only crop that thrives well in deepwater conditions. Rice is followed by late wheat or boro rice. Mixed cropping is practiced in limited locations in some areas. The usual practice in deepwater chaurs is to broadcast rice with mungbean, sesame, sorghum, maize, or Capsularis jute. However, in the low-rainfall season, the cropping pattern is DWR, followed by winterseason cereals, pulses, and oilseeds, depending on available soil moisture status. Normally, in very deepwater areas, water remains yearround and transplanting is

practiced. In deepwater areas, cropping patterns vary according to rainfall or flood patterns. Rice-fish farming is also practiced in deepwater areas. After the harvesting of rice (Dec-Jan), local fish such as Mangur, Singhi, Bhakur, Rohu, and Katla are also harvested. No supplemental food was given to the fish. A system diagnosis of farmers problems was done through participatory rural appraisal and problem ranking/prioritization. The exercise showed lower and unstable production potential of the DWR ecosystem. The key problem pointed out by the farmers was poor DWR yield. The biophysical causes identified were irregular hydrolDecember 2004

ogy, lack of agro-technology, lack of suitable varieties, and weed problems. Among these primary causes, irregular hydrology was most critical as this often resulted in waterlogging or flash floods. The socioeconomic constraints include the low economic status of farmers, their lack of knowledge, and their small landholdings. Economic status was directly influenced by cash on hand, credit facilities, and technical know-how. Appropriate crops/varieties, better drainage facili

ties, and improved agro-technology were the solutions recommended to increase the productivity of deepwater areas. Rajshree, Vaidehi, and Barh Avrodhi for medium deep, waterlogged areas (up to 0.5 m water depth); Sudha, Sabita, and Jalpriya for deepwater (0.51.0 m); and Varidhi, Jalmagna, and Jalnidhi for very deepwater (>1.0 m) were identified as promising varieties for the DWR ecologies. They have drought and submergence tolerance, and they can grow

long. Suitable varieties for preflood crop mixtures and postflood environments should additionally be identified or introduced to further increase productivity. The performance of drainage facilities can be improved by diverting excess water to canals that would then be used for irrigation purposes or for operating fishponds. Appropriate agro-technology such as rotational rice-fish farming should be promoted.

Farmers participatory diagnosis of the flood-prone deepwater rice-cropping system in eastern India
U.P. Singh, Department of Agronomy, Institute of Agricultural Sciences, Banaras Hindu University, Varanasi 221005, Uttar Pradesh, India E-mail: u_psingh@sify.com/upratap@banaras.ernet.in

The flood-prone deepwater rice (DWR) areas of eastern India are spread over 3.9 million ha. Here, rice is grown as rainfed dryland or under shallow flooding for the first 13 mo and then flooded to depths exceeding 50 cm for a month or longer. About 12% of the rice-growing area of eastern India is under the DWR ecosystem. The main DWR areas are located in eastern Uttar Pradesh, Bihar, West Bengal, Assam, and Orissa, where productivity is very poor (1 t ha 1 ). Deepwater areas, called chaurs or tals, are situated in remote areas and their environment is very complex and unpredictable, often times subjected to stresses of water stagnation. A
IRRN 29.2

farmer participatory evaluation was done at two sites Biraul chaurs (north Bihar) and Suraha tal (eastern Uttar Pradesh)to identify and develop ecologically sustainable and economically viable technologies that can improve the production potential of DWR farming systems. Agroecosystem analysis and system diagnosis of farmers problems were done in 11 villages with 48 farmers. Problems and issues related to DWR ecosystems were prioritized. On the basis of problem ranking and participatory rural appraisal, the common problems encountered in DWR growing environments were waterlogging, flash/intermittent floods, lack of suit-

able varieties/cropping patterns, lack of improved management practices, and socioeconomic/bureaucratic difficulties. Identification of appropriate crop/DWR varieties, sustainable and improved agricultural technologies, suitable cropping/farming systems, and better drainage facilities were the solutions mentioned. Based on the solutions identified, farmers participatory on-farm and onstation activities were designed and evaluated. A large number of strains/cultivars/varieties were tested in different DWR growing conditions0.5 m, 0.51.0 m, and > 1.0 m floodwater depth. Nine varieties Rajshree, Vaidehi, and Barh
85

Avrodhi for medium deep, waterlogged (up to 0.5 m); Sudha, Sabita, and Jalpriya for deepwater (0.51.0 m); and Varidhi, Jalmagna, and Jalnidhi for very deepwater (>1.0 m water depth)were identified as promising. These varieties gave superior performance in various onfarm trials, with 1525% higher yield over local varieties (Table 1). Rajshree, Sabita, and Barh Avrodhi had better performance under delayed planting in medium deepwater conditions (Table 2). Sabita and Barh Avrodhi were promising in a double-transplanting situation, whereas Sudha and Sabita fared well under random/haphazard planting. Pusa Baisakhi and NP 28 were found to be suitable mungbean varieties for the preflood crop mixture as a DWR intercrop. Application of 20 kg N ha 1 at the onset of the first monsoon and line sowing increased yield by 13%. Improved varieties with improved management practices and need-based plant protection measures gave 30% more yield than the farmers management practices. Preemergence application of butachlor and 2,4-D at 1.5 and 0.5 kg ha 1 , respectively, was effective in reducing weeds in preflood environments. Vaidehi and Barh Avrodhi proved to be better weed competitors. In preflood, mixed cropping, on-farm evaluation, average productivity of DWR + Capsularis jute crop mixture was 3.19 t ha 1 (rice yield equivalent). This was fol86

lowed by DWR + mungbean (2.44 t ha 1 ). The DWR + Capsularis jute combination gave as high as 4.37 t ha 1 yield equivalent, a substantial gain in this environment. During the postflood environment, paira cropping of Bakala (Vicia faba) registered the highest yield equivalent (3.47 t ha 1 ), followed by lentil, linseed, and mustard. Among the cropping systems, the highest mean yield equiva-

lent (3.54 t ha1) was given by DWR-lentil, followed by DWR-wheat and DWRlinseed (Table 3). After DWR harvest, the possibility of cultivating boro (dry-season) rice was assured both in farmers fields and at research stations, especially in waterstagnant areas. Gautam, Rasi, Pusa 2-21, IR64, Sarjoo 52, Pant 12, and Prabhat were evaluated as promising cultivars for the boro season.

Table 1. Varietal performance in farmers participatory on-farm evaluation under deepwater rice ecosystems (mean of 3 y). DWR Farmers subparticipating ecosystem (no.) Medium deepwater Deepwater Very deepwater 10 Mean peak water depth (cm) 34.552.0 Barh Avrodhi Madhukar 2.52 2.30 Jalpriya 1.75 Jalmagna 1.49 Sudha 1.83 Jalnidhi 1.59 Grain yield (t ha1) Vaidehi 2.59 Chakiya 59 1.60 Varidhi 1.69 Rajshree 2.47 Sabita 1.90 Kariywa local 2.30 Kariywa local 1.66 Jaisurya local 1.58

13 11

74.3103.3 122.7145.5

Table 2. Effect of planting dates and varieties on grain yield (t ha1) under the flood-prone, medium deepwater rice ecosystem (on-station pooled data of 3 y). Planting date Varieties Rajshree Sabita 3.03 1.78 1.22 Barh Avrodhi 2.34 1.36 1.76 Madhukar 2.15 1.30 0.75 Vaidehi 2.93 2.10 0.85 BKP 246 3.10 1.43 0.90

30 Jul (30-d-old seedlings) 2.40 30 Aug (60-d-old seedlings) 1.75 30 Aug (30-d-old seedlings) 1.25 Difference between Date mean at the same level of variety Variety mean at the same level of planting date

LSD (P = 0.05) 0.38 0.31

December 2004

Table 3. Farmers participatory on-farm evaluation under different flood-prone DWR-based cropping systems (mean of 3 y). Cropping system Farmers participating (no.) 5 5 5 6 DWR + sesame a b c 1.23 0.15 2.14 DWR-bakala 1.54 0.55 DWR-wheat 3.47 Yield (t ha1)a

Preflood mixed cropping with DWR Postflood paira cropping with DWR DWR cropping system DWRboro rice

DWR + mungbean a b c 0.98 0.36 2.44 DWR-lentil 1.56 0.25 2.56 DWR-lentil

DWR + jute fiber a b c 0.96 1.12 3.19 DWR-linseed 1.53 0.18 2.43 DWR-linseed 1.5 0.34 3.20 DWR-boro (IR64) 1.12 7.15 10.0

DWR only 1.41 DWR-mustard a b c 1.52 0.09 1.00

1.51 1.21 3.39 DWR-boro (Sarjoo 52) 1.20 6.80 9.70

1.45 0.52 3.54 DWR-boro (Gautam) 1.38 6.85 9.94

DWR-boro (Pant 12) 1.50 6.50 9.60

a = DWR, b = mixed crop/second crop, c = yield equivalent in terms of DWR at prevailing market prices.

Postharvest technology of rice: role of farm women in storing grains

P. Sumathi and M.N. Budhar, Regional Research Station, Tamil Nadu Agricultural University, Paiyur 635112, Krishnagiri District, Tamil Nadu, India

The importance of grain storage in rice production is often ignored. Inadequate storage facilities and improper storage methods can cause considerable losses to rice farmers. Total postharvest losses in food grain account for 9.3% of total production, 6.5% of which are losses incurred during storage alone (Gandhi 1983). Minimizing these losses can increase food grain supply, thereby making headway in feeding millions of hungry people. Farmers must therefore learn how to store rice properly, especially during the transitory period, to protect the grains from the effects of weather or from insects and pests. Womens role in this respect is yet to be recognized, in spite of their significant inIRRN 29.2

volvement in food processing and food storage. Women farmers play distinct and well-accepted roles in all the activities of rice cultivation, and 93% of the farm women are themselves actively involved in storing rice grains (Sumathi and Budhar 2003). They receive less than 5% of extension services worldwide, their priorities are rarely reflected in agricultural research or national policies, and their role as agricultural producers is still largely unrecognized and has not been addressed (LEISA 2002). We conducted a study among women farmers to find out the role of farm women in grain storage using different storage practices. We also aimed to identify the needs of farm women with

respect to effective storage of grains. Data were collected at Tirupattur taluk of Vellore District in Tamil Nadu, India. Agriculture is the main source of livelihood of the people in this area and rice is grown on about 72,000 ha in three distinct seasons. A sample of 100 women farmers was selected randomly from five villages (20 from each village). Most of them cultivate small farms (0.11 ha) and some who have more land (>1 ha) live in clusters. Through a well-structured interview schedule, data on the adoption of different storage methods and feedback regarding problems in adoption were collected. Among 12 storage practices identified, four (see Table 1)drying grains before storage, stacking bags
87

horizontally, keeping bags on planks and leaving passages, and using metal traps for rodent controlwere found to have been adopted by a majority of the farm women. The possible reason for the better adoption is that all these practices are cheaper and require fewer skills. Half of the respondents used polythenelined bags and rat poison (zinc phosphide). The low adoption might be due to the practice being risky, dangerous to domestic animals, expensive, and requiring technical guidance. The rest of the practices were found to have very low levels of adoption because they are capital-intensive and necessitate considerable skill in practicing them. A higher percentage of nonadoption was found in other storage practices for various reasons: high cost, complexity, availability of local substitutes, lack of knowledge on storage structures, and inadequate technical guidance. Women with big farms had a higher percentage of adopting storage practices such as the use of impregnated bags, pretreatment of storage structures, tin-cone plates for storage structures, and anticoagulants for rat control. These women farmers can afford to practice these storage methods. However, the percentage of usage of metal storage bins was high among women with small farms since these structures were provided at subsidized rates. A majority of the small farm women (76%) wanted financial assistance in the form
88

Table 1. Distribution of female respondents, by storage method adopted. Recommended practice Dry grains before storage Use impregnated bags Use polythene-lined bags Keep bags on planks and leave passages Pretreatment of storage structures Stack bags horizontally Use metal storage bins Use tin-cone plates for storage structures Fumigate with ethylene dibromide ampules Use metal traps for rodent control Use anticoagulants for rat control Use prebaiting and poison baiting for rats with zinc phosphide Small farmers No. 50 26 36 5 50 15 9 1 43 7 21 % 100 52 72 10 100 30 18 2 86 14 42 Big farmers No. 50 6 25 40 15 50 3 21 4 45 18 34 % 100 12 50 80 30 100 6 42 8 90 36 68 Total No. 100 6 51 76 20 100 18 30 5 88 25 55 % 100 6 51 76 20 100 18 30 5 88 25 55

Table 2. Distribution of female respondents, by suggestion to overcome adoption problems. Suggestiona Financial assistance by providing credit Providing subsidy Ensuring timely supply of materials Ensuring timely technical guidance Arranging special campaigns on storage structures Ensuring adequate supply of materials Developing low-cost storage structures
a

Small farmers No. 38 36 20 18 17 12 10 % 76 72 40 36 34 24 20

Big farmers No. 16 12 32 26 8 35 13 % 32 24 64 52 16 70 26 No. 54 48 52 44 25 47 23

Total % 54 48 52 44 25 47 23

Multiple responses.

of credit to enable them to buy better storage structures such as metal bins, whereas 72% asked for a subsidy for the same purpose. Some respondents (34%) wanted special campaigns on storage structures (Table 2). On the other hand, women with big farms recommended an adequate (70%) and timely supply of materials (64%). Half of these farm women needed timely technical guidance on postharvest technology of paddy (52%). The results show an increased involvement of farm women in agriculture and other allied sectors. These

women farmers have more responsibilities, especially in the Indian agricultural system. Hence, for a balanced and sustainable growth of rural India, womens role in the developmental process should be recognized and adequately supported. References
Gandhi NK. 1983. Stepping up rural warehousing. Kurukshetra 31(22):18-20. LEISA. 2002. Women in agriculture. Low External Input Sustainable Agric. 4(4):4-5. Sumathi P, Budhar MN. 2003. Extent of womens involvement in Indian rice cultivation. Int. Rice Res. Notes 28(1):74-76.

December 2004

Extension service and rice yield gap in Nigeria, 1980-2002

O.I. Oladele, Japan International Research Center for Agricultural Sciences, Development Research Division, Tsukuba, Ibaraki 305-8686, Japan E-mail: oladele@jircas.affrc.go.jp

Rice is cultivated in virtually all agroecological zones in Nigeria. In 2000, 6.4% of the 25 million ha of land cultivated to various food crops was grown to rice, with average national yield of 1.47 t ha 1 . Though rice contributes a significant proportion of the food requirements of the population, production capacity is far below national requirements. To meet the increasing demand, milled rice was imported to bridge the gap between domestic demand and supply. Several approaches have been adopted to develop agriculture in Nigeria with a focus on rice and other staple food crops. Extension delivery had gone through several typologies, among which are the farm settlement scheme, national accelerated food production program, Operation Feed the Nation, River Basin Development Authority, green revolution, Pilot Agricultural Development Project (PADP), and, later, the multistate ADP. A major feature of the Nigerian agricultural extension service in the recent past is the involvement of nongovernment organizations in extension delivery in Nigeria (Oladele et al 2004). The trend of rice research in West Africa (including Nigeria) shows that rice growing started in about 1850, with expansion occurring from 1950 to date, which has been the exclusive preIRRN 29.2

serve of national programs, research institutions from developed countries (CIRAD and ORSTOM), and international agricultural research centers (WARDA, IITA, and IRRI) (Virmani et al 1978). Gomez et al (1979) identified two types of yield gapsthat between experiment stations and farmers fields and that between potential and actual farmer yields that constitute the true research-extension gap due to a combination of biological, technical, and socioeconomic constraints. Evenson (1997) reported four yield levels actual, best-practice, researchpotential, and sciencepotential yield. Associated with these yields are extension, research, and science gaps. While yield potential has been increasing because of the release of new varieties and, lately, exploitation of biotechnology, yield in the field is yet to catch up with expected yield. Despite the voluminous research on soil, agronomic conditions, biotechnology, varieties, and stress tolerance, a gap still exists between potential and actual yield, notwithstanding the complementary relationship between the physical sciences and socioeconomic research. This paper therefore examines how extension service influenced the rice yield gap in Nigeria from 1980 to 2002.

Data for the study (actual yield, potential yield, and extension activities in Nigeria from 1980 to 2002) were collected from secondary sources. The potential yield for the irrigated ecosystem is 57 t ha 1 ; rainfed lowland, 2.55.0 t ha1; and upland, 1.5 4.5 t ha 1 (WARDA 2002, Dalton and Guei 2003). The average of the range of potential yields was used in subsequent analysis. The other variables of the study were the number of operating agricultural development projects (OADPS), extension funding (EFUND), number of extension agents (NEA), number of farm visits by extension agents (FVEA), number of farm trials and demonstrations (NFTD), number of farmers trained (NFT), number of small-plot adoption techniques established (SPAT), number of women in agricultural groups formed (WIAG), number of extension zones (ZONES), number of extension blocks (BLOCKS), number of extension cells/circles (CELL/CIRCLE), number of subject matter specialists (SMS), number of block extension supervisors (BES), extension agent to farmer ratio (EARATIO), research intensity (RI), extension intensity (EI), research funding (RFUND), number of researchers (NRSCH), and yield gap (YGAP). To explain the yield gap using four functional
89

Multiple regression analysis explaining the yield gap. Functional form Constant (regression coefficient) t value OADPS t value NFTt value BLOCKS t value EARATIO t value SE R R2 F Linear 2.0919.51** 1.236E-021.93 1.324E-044.38** 1.741E-041.38 0.3504.37** 0.04 0.91 0.82 21.36 Double log 1.199.02** 8.68E-021.81 4.73E-040.044 2.86E-021.95 9.50E-024.44** 0.06 0.92 0.83 23.12 Semi log 3.239.72** 0.221.79 5.29E-030.19 7.28E-021.97 0.244.50** 0.15 0.91 0.83 22.63 Exponential 0.7413.65 4.48E-031.39 4.22E-052.77* 4.45E-050.70 9.01E-022.23* 0.08 0.81 0.66 8.71

This then implies a need for concerted efforts to achieve proper monitoring and implementation of extension services, effective contact with farmers, and less bureaucracy. Only then can food security be ensured. References
Dalton TJ, Guei RG. 2003. Productivity gains from rice genetic enhancements in West Africa: countries and ecologies. World Dev. 31(2):359-374. Evenson RE. 1997. The economic contributions of agricultural extension to agricultural and rural development. In: Swanson BE, Bentz RP, Sofranko AJ, editors. Improving agricultural extension. Rome: Food and Agriculture Organization. Gomez KA, Herdt RW, Barker R, De Datta SK. 1979. A methodology for identifying constraints to high rice yields in farmers fields. In: Farm-level constraints to high rice yields in Asia. Manila (Philippines): International Rice Research Institute. p 27-47. Oladele OI, Koyama O, Sakagami JI. 2004. Africa in search of extension system: experience from Nigeria. J. Food Agric. Environ. 2(1):276-280. Virmani SS, Olufonile JO, Abifarin AO. 1978. Rice improvement in tropical Anglophone Africa. In: Rice in Africa. Proceedings of a conference held at IITA, Ibadan, Nigeria, 7-11 March 1977. New York: Academic Press. WARDA (West Africa Rice Development Association). 2002. Annual report. Bouak, Cte dIvoire. 84 p.

forms, from the list of 19 variables, principal component analysis (PCA) was used to determine which variables can be fitted into the regression equation. The extraction method of PCA identified four variables to be fitted into the regression equation: OADPS (0.957), NFT (0.485), BLOCKS (0.509), and EARATIO (0.914). The lead equation chosen for the four functional forms fitted to the data was the linear function as it has the least standard error of the parameter estimate and high coefficients of determination (R and R 2). Two of the variables were significant at 5%; NFT and EARATIO had more impact on the yield gap than OADPS and BLOCKS. While the first

two variables represent actual contact (training and dissemination of research results to farmers), the remaining two emphasize the operational setup of the extension agency. While there was a positive relationship between yield gap and NFT, there was an inverse relationship between yield gap and EARATIO. There is a need to increase the effective extension contact between extension agents and farmers in Nigeria. This will help increase the overall output and productivity of rice. The findings, particularly the extraction of the four extension variables as causal factors for the yield gap, suggest that transitory growth in rice production is a potential problem in Nigeria.

90

December 2004

NOTES FROM THE FIELD

Vanishing populations of Oryza minuta Presl. in Pangil, Laguna, Philippines

M.C.N. Banaticla and M.S.R. Almazan, IRRI

Oryza minuta Presl., a wild rice species known to be resistant to bacterial blight and blast and green and brown planthopper, is endemic to the Malesian floristic region, specifically in Thailand, the Philippines, Indonesia, and Papua New Guinea. While it has been widely used in breeding and other experimental studies, its ecology and population dynamics remain poorly understood. In 1980, Vaughan documented six populations of O. minuta located along a stream in Barrio Balian, Pangil, Laguna. Seed samples from these populations were collected and later conserved at IRRIs International Rice Genebank. On 18 Feb 2004, we surveyed the same site and found that all the populations reported by Vaughan (1980) no longer existed. The small irrigation stream where the three populations once stood has been converted into a cemented canal (Fig. 1). During the visit, a road was being constructed near the former location of the other three populations. We encountered five clumps of O. minuta in different locations along the main stream, but unlike the vastly spreading populations described by Vaughan (1980),

they appear in small clusters of 23 plants (Fig. 2). Two clusters were partly strangled by Mikania cordata, a weedy herbaceous vine. The exclusive occurrence of O. minuta along the stream indicates its dependence on water flow for seed dispersal. It is possible that a mother plant or population located farther upstream serves as the seed source. It would be interesting to track this pre-

sumed mother population and learn more about its biology. In a period of 24 years, the populations of O. minuta in Barrio Balian, Pangil, Laguna, have greatly dwindled, leaving only traces of individual plants. Livestock raising, farming, and road construction destroy the natural vegetation within the vicinity. Such human activities probably played a major role in the

Fig. 1. Cemented canal where O. minuta populations once thrived.

Fig. 2. A cluster of O. minuta found in the study area.

94

December 2004

alarming disappearance of O. minuta populations in the area. Further studies on the conservation and population dynamics of O. minuta in Pangil, Laguna, are needed to prevent the local extinction of this species.

Reference

Vaughan DA. 1980. Oryza minuta observations in its natural habitat. IRRI trip report, Philippines.

Current status of rice pests and their management in Assam, Indiaa discussion with extension agents
Z. Islam, K.L. Heong, and M. Bell, IRRI; and L.K. Hazarika, D.J. Rajkhowa, S. Ali, B.C. Dutta, and M. Bhuyan, Assam Agricultural University, Jorhat 785013, India

2 h discussion sessions on each of the major groups of pests (insects, weeds, diseases, and rodents) during an integrated pest management training course implemented at Assam Agricultural University (AAU) in Jorhat on 5-15 Jul 2004. The discussion sessions were led by AAU experts and 19 trainees involved in rice pest management education, research, and extension participated. The gists of these sessions are presented. Insect pests Although many insects feed on rice plants, only a few occasionally reach pest status in Assam. The potential insect pests include rice hispa [Dicladispa armigera (Olivier)], yellow stem borer [Scirpophaga incertulas (Walker)], rice bug (Leptocorisa spp.), brown planthopper [Nilaparvata lugens (Stl)], leaffolders [Cnaphalocrosis medinalis (Guene) and Marasmia spp.], caseworm [Nymphula depunctalis (Guene)], and thrips [Stenchaetothrips biformis (Bagnall)]. So far, reliable yield loss estimates are not available and their exact pest status remains uncertain. The important potential pest is rice hispa, which is endemic in Sibsagar, Lakhimpur, Nalbari, Borpeta, Cachar, and Karimganj districts. It is more abundant during sali and ahu seasons than in boro. In outbreak situations, high yield losses may occur in specific fields. Several species of stem borers are present and yellow stem borer is considered predominant; however, conflicting opinions arose with re-

Rice is the most important crop in Assam, India, grown on about 70% of the total cultivated land (3.64 million ha) in the state (MA 2003). In fact, about 70% of the rice area of the seven northeast Indian states is under Assam. Rice is mostly grown in the low-lying deltas of the Brahmaputra and Barak rivers; some rice is also cultivated in upland situations in the northern hill region. The sali (monsoon crop: transplanted lowland rainfed rice and deepwater rice) rice is dominant, occupying 68.8% of the total rice area of 2.54 million ha, followed by ahu (summer rice) (18.3%), and boro (dry-season/winter) rice (12.9%). To understand Assams current pest problems in rice, and their management, we arranged four
IRRN 29.2

spect to the relative importance of the other species. The incidence of brown planthopper, thrips, leaffolders, and caseworm seems to be increasing, while that of earcutting caterpillars [Mythimna separata (Walker)] and swarming caterpillars [Spodoptera mauritia (Walker)] has decreased significantly compared with their incidence in the pre-green revolution period. The incidence of brown planthopper is relatively greater in the boro, particularly in the Barak River Valley. The rice bug is a major problem in the ahu season, particularly in the early ahu crop. Caseworm is a localized pest and occurs more in the sali season. Root and panicle aphids, white grubs, and rice bugs are major insect pests of upland rice, including the jhum (slash-and-burn) system. So far, insecticide use is minimum and there is some use of botanicals (extracts of plants). However, the extent of insecticide and botanical use and whether farmers are deriving any benefit from their use are yet to be determined. The AAU identified an effective muscardine fungus against rice hispa, whose effect is comparable with that of insecticide. It affects eggs, larvae, and adults but is more efficient in killing eggs. A mass culture method has been developed, but production and marketing have yet to be taken up by either the private or public sector. Diseases Assam is very rich in rice genetic diversity. Many traditional cultivars possess genes
95

of resistance to most rice diseases. Many diseases infest rice plants in Assam, but the most important ones are sheath blight, bacterial blight, and blast in the sali; blast and sheath rot in the ahu; sheath rot and sheath blight in the boro; and ufra and root-knot in deepwater rice. The extent of average yield losses attributed to diseases is not known. However, it is assumed that 1520% yield losses may occur in some infested fields, although average losses would be much less. Farmers seem to be not very much aware about diseases, unlike what they know of other pests. Although several cultural and chemical options are available for the management of each of the major diseases, it is not known what percentage of the farmers use these technologies, if at all. Some fungicides are available at the grassroots level, but their use is very low. However, most of the popular modern rice varieties possess resistance to or tolerance for blast, and some degree of tolerance for sheath blight and bacterial blight (Table 1). Weeds The agroclimatic conditions of Assam favor the rapid growth of weeds. The type of rice culture, season, soil, and cropping pattern influence the weed flora in rice fields. The most important weed flora in the different rice cultures are listed in Table 2. Among rice cultures, the weed problem is most severe in dry-seeded rice (upland rice), followed by wet-seeded (sprouted seeds) and transplanted rice. Trans96

Table 1. Resistance status of 12 most popular rice varieties of Assam, India. Variety according to rank Season Blast Sali Sali Sali Ahu Ahu Boro Sali Sali Sali Sali Boro Boro T S MR MR T MR T T MR MR MR MR Sheath blight T S T MS MS T T T T T T T Reaction to major diseases Sheath rot MR T MR Bacterial blight MS T T T T T T T T T Stem rot MS T MS Remark Neck blast (R)

Ranjit Mahsuri Bahadur Luit Disang Joymati Aghoni Keteki Manoharsali Andrewsali Joytiprasad Bishnuprasad

Neck blast (T) MS MS

MR MR MS

R = resistant, MR = moderately resistant, MS = moderately susceptible, T = tolerant.

planted rice has fewer weed problems because puddling, followed by transplanting of rice seedlings, gives rice plants a head start and enables farmers to use water for weed suppression. The weed flora in rice are dominated by grasses, followed by broadleaves and sedges (Barua and Gogoi 2002). However, alternate wetting and drying conditions that prevail in any culture may result in severe weed problems. The diverse weed flora in deepwater rice emerge in three flushes before, during, and after the floods. Weeds may reduce rice yields by 1540%, if not controlled at all. In general, farmers are aware of harmful effects of weeds and they make serious efforts to control weeds using direct and indirect methods. The direct methods practiced by Assamese farmers include manual weeding by using simple tools, the use of simple rotary weeders, and herbicides. So far, manual weeding is the most common, whereas the use of herbicides is very

limited. Indirect methods involve tillage, the use of water, transplanting of rice seedlings, etc. Rodents To date, 34 species of rats and mice under 12 genera were recorded in northeast India, including Assam (Singh et al 1994). The dominant species are Bandicota bengalensis in all states of northeast India; Rattus rattus khyensis in Mizoram; R. bowersi in Meghalaya and Miaoram; Rattus nitidus nitidus in Arunachal Pradesh, Meghalaya, and Sikkim; R. rattus tistae in Arunachal Pradesh, Manipur, Meghalaya, Mizoram, Sikkim, and Tripura; and Niviventer niviventer in Meghalaya. It is believed that the outbreak of Bambusa tulda and Dendrocalamus longispathus is mainly associated with the flowering of the major bamboo variety Thingtam, while that of Melocanna bambusoides is associated with the flowering of variety Mautam (Kumar and Pathak 2002). Three rat outbreaks coincided with
December 2004

Table 2. Major weed flora in different rice cultures in Assam, India. Weed species Directseeded Grasses Cynadon dactylon Digitaria setigera Echinochloa crus-galli E. stagnina Eleusine indica Eragrostis unioloides Hackelochloa granularis Hymanachne acutigluma Isachne himalatica Leersia hexandra Paspalum conjugatum Paspalum scrobiculatum Sacciolepis interrupta Sedges Cyperus iria C. pilosus C. rotundus Eleocharis acutangula E. dulcis Fimbristylis littoralis Fimbristylis spp. Scirpus juncoides S. maritimus Other weeds Ageratum houstonianum Cuphea balsamona Eichhornia crassipes Fissendocarpa linifolia Ludwigia adscendens Melochia corchorifolia Mimosa pudica Monochoria vaginalis Sagittaria guayanensis Salvinia natanes Sphenoclea zeylanica Spilanthes paniculata v v v v v v v v v v v v v v v v v v v v v v v Abundance as major weed in different rice cultures TP Ahu TP Sali Wetseeded TP Boro DWR

v v

v v v

v v

v v v

v v v

(relative abundance of 59%, 7%, 19.1%, 14.5%, and 5.8%, respectively). Rice tiller cut by rats was estimated at 5.7% in sali, 3.3% in ahu, 5.0% in boro, and 10.0% in deepwater. Tiller damage at the vegetative phase and at panicle initiation, dough, and ripening stages in the sali season was estimated at 2.4%, 3.5%, 7.4%, and 5.7%, respectively. Farmers use local traps, rodenticides, and cats and dogs for rodent control mainly in the homestead. But they rarely make an effort to control rats in rice fields. References
Barua IC, Gogoi AK. 2002. Diversity of weed flora in upland direct seeded rice ecosystem in Assam. In: Proceedings of the UGC-sponsored State-level Seminar on Biodiversity of Assam and Its Conservation. Karimgaj, Assam (India): Karimganj College. p 108-116. Kumar D, Pathak KA. 2002. Bamboo flowering and rodent outbreak in the northeast hill region of India. In: Glimpses of rodent research in India. AICRP on Rodent Control. Jodhpur (India): All India Coordinated Research Project. p 40-43. MA (Ministry of Agriculture). 2003. Basic statistics of north east region, 200203. Directorate of Economics and Statistics, MA, Government of India. Singh YP, Kumar D, Gangwar SK. 1994. New records of some rodent species from the states of North Eastern hill region. Rodent Newsl. 18(5):12-13.

v v v v v

v v v v v v v

v v

v v v

DWR = deepwater rice, TP = transplanted.

bamboo flowering in 1911, 1929, and 1956, leading to famine in hilly regions of Assam and other northeast Indian states and forcing migration from the hilly regions to the valleys. The flowering of Thingtam bamboo also occurred in 1977, but extensive crop damage was not observed, which was attributed to control measures set by the government. The next flowering of Mautam bamboo is expected between 2004 and
IRRN 29.2

2007. The Indian Council of Agricultural Research started the All India Coordinated Research Project on Rodent Control in 1977 and gradually established 10 centers, two of which are located in Shellong (Meghalaya) and Jorhat (Assam). Recent studies in the river valleys in Assam revealed that B. bengalensis is the dominant species in rice fields, followed by B. indica, M. booduga, and R. sikkimensis

97

Agricultural engineering

Subsurface drainage to increase rice productivity in a saline environment

S.K. Gupta, Central Soil Salinity Research Institute, Karnal, Haryana; T.V. Satyanarayana, Acharya NG Ranga Agricultural University, IDNP, Bapatala, Andhra Pradesh; M.V. Manjunatha, University of Agricultural Sciences, Agricultural Research Station, Gangavathi, Karnataka; and V.B. Kuligod, University of Agricultural Sciences, Agricultural Research Station, Bheemarayanagudy, Karnataka, India E-mail: skgupta@cssri.ren.nic.in

The average productivity of rice in Indiaat less than 2 t ha1is, by all standards, low (MIB 2003). Irrigated rice yields in the range of 2.54.0 t ha 1 in many places are far from their anticipated productivity of 6.07.5 t ha 1 . Such low yields, in many cases, can be due to salt-related soil degradation. As such, a rising water table, appearance of salt efflorescence at the soil surface, and declining rice yields cause farmers, researchers, and policymakers to worry because those problems can jeopardize the future of irri-

gated agriculture and bring about downtrends in the economy. Based on the outcome of baseline and soil surveys, it has been hypothesized that large areas going out of cultivation and low yields can be attributed to the high water table, accompanied by salt-related land degradation. Studies began under an Indo-Dutch Network Project at 10 locations in four irrigation commands covering two states of India to assess the yield-salinity relationships for rice and to test improved drainage techniques

to reverse or mitigate salt-related problems. Based on the reconnaissance, soil, and baseline surveys, appropriate subsurface drainage systems were designed at each of the 10 study sites to control the water table and facilitate leaching (see table). The drain spacing varied from as low as 27 m to 150 m, whereas the depth of the drains was relatively shallow (<1.5 m) in all cases. A total of 193 ha were treated with drainage (see table). Since the area covered under these studies has been under

Increase in rice yield and cropping intensity as a result of salinity control through subsurface drainage. State Location Area treated (ha) Drain spacing and depth in parentheses (m) Soil salinity, ECe (dS m1) and rice yield (t ha1) Before drainage ECe Andhra Pradesh Karnataka Konanki Uppugunduru Islampur (PP) Islampur Phase II Vaddarathi Gundur Siddapur Gangavathi Sindhanur Gorebal 13.0 12.0 26.0 14.4 4.9 11.5 6.9 2.9 62.0 40.0 30, 60 (0.91.1) 30, 45, 60 (1.21.35) 20, 30, 30, 50, 60 (1.01.2) 30, 50 (1.11.2) 27 (0.9) 27 (0.9) 27 (0.9) 27 (0.9) 150 (0.75) (1.7)e 5.7 4.8 12.0 8.4 6.5 Yield 3.7 4.3 1.4 (k) 1.4 (r) 1.9 (k) 2.3 (r) 3.5c 2.8c 2.4c 4.0c 2.2 2.3 ECe 2.8 3.4, 2.9, 2.3a 6.0 0.6 2.7 1.6 0.6 2.6 0.9 After drainage Yield 5.6 5.5, 5.6, 6.5a 5.6 (k) 6.0 (r) 3.0 (k) 4.0 (r) 8.4 8.1 7.3 7.9 3.7, 6.77.5b 7.2 Cropping intensity (%) Before drainage 70 90 77.0 58.2 0d 0d 0d 0d 143 After drainage 130 165 50.1 59.4 200d 200d 200d 200d 191

a First, second, and third year. bFirst year and range during subsequent years. k = kharif (Aug-Dec), r = rabi (Jan-May), = average salinity was quite high due to large barren patches, with salinity ranging from 20 dS m1 (Gundur) to 57 dS m1 (Siddapur). cYield of rice in a few plots where salinity was relatively less. dOnly for barren patches constituting about 90% of the land, = not calculated/not reported/not available, PP = pilot project. eInterceptor pipe drain.

IRRN 29.2

91

the control of the farmers, they actively participated during all phases of the project (Hanumantaiah et al 2003). Farmers were allowed to follow normal cropping practices. Besides the regular monitoring of the subsurface drainage system, crop performance was studied by conducting crop-cutting tests at fixed grid points where the soil samples were also drawn. Yield increments due to treatment effects were calculated by averaging seasonal yield over various fields at all the sites. The predrainage soil salinity and crop yield data at these points were used as baseline data to assess the rice yield increase due to treatment effects. Yield was also related to changes in soil salinity over the years after the implementation of subsurface drainage. Cropping intensity was determined by collecting land-use data. In a few cases, crop cutting and soil surveys were extended to areas where a drainage system was not implemented.
Crop yield (t ha-1) 8 7 6 5 4 3 2 1 0 0 2

Thus, crop performance evaluation was spread over an area of 350 ha, including the treated area. Results revealed that rice yield in farmers fields at all the study sites was low with uncultivated barren patches where salt efflorescence was clearly visible. The rice yields in all cases were negatively correlated with soil salinity (EC e). The farmers perception that declining yields are due to soil salinity was shown to be valid. Predrainage surveys revealed that, despite continuous flooding in lowland rice fields, the salts were not leached because of the high water table, resulting in inadequate drainage conditions. Compared with rice yield under normal land in the region, there appeared to be considerable scope for yield improvement in degraded lands. The monitoring and evaluation of the subsurface drainage systems revealed that the water table remained deeper; EC of the soil decreased and crop yield and

4 6 ECe (dS m-1) at harvest

10

Rice yield as affected by soil salinity.

cropping intensity increased in the drained area (see table and figure). The increase in rice yield as a result of declining EC at two sites depicted in the figure attests to the drainage-induced land reclamation. It was observed that high yields exceeding 6 t ha1 can be obtained only if soil salinity at harvest is below 3.0 dS m 1 . In a related study at Gohana in Haryana, it has been observed that an optimum Basmati crop can be obtained only if soil salinity is 2.5 dS m 1 or less. Since Basmati is a relatively saltsensitive genotype, the results obtained in these studies seem reasonable. It could then be concluded that the drainage intervention effectively reclaimed the land, even though the drain spacing at a few sites was unusually wide (see table). From the drainage point of view, the most critical situation is that of the Sindhanur drainage project where drain spacing was 150 m and the average drain discharge was quite low. Even then, the land could be reclaimed for rice cultivation within 1 year. It seems that the cultivation of submerged lowland rice aided in the leaching of salts within a period of 1 y. Theoretically, taking a saturation water content of 0.5 m3 m3 of soil and an observed average drain discharge of 0.3 mm d1, a downward flow velocity of 0.6 mm d 1 could leach the first 20 cm of the soil in about 1 y. For a shallowrooted field, that is an acceptable depth. Other crops would require a longer reclamation period.
December 2004

92

Rice production increased because of an increase in both productivity and cropping intensity (see table). The increase in cropping intensity is attributed to the cropping of barren uncultivated patches as a result of land reclamation and to the second crop of rice, which could be grown as a result of improved physicochemical conditions of the soil and general improvement in water management. In areas where rice is not grown as a second crop, other crops could be cultivated, thus rais

ing the overall productivity of the area. This study proved that salt-related soil degradation due to irrigation is a major threat to the sustainability of rice cropping under semiarid conditions, especially at locations where groundwater is at shallow depth and possibilities to leach accumulated salts are limited. The existing saline soils in irrigation commands of India can be reclaimed through irrigated rice cropping, provided some form of drainage is put up to

maintain the salt balance. Such an approach would raise rice productivity in irrigation commands. References
Hanumantaiah CV, Satyanarayana TV, Gupta SK. 2003. Participatory management in lift irrigation: a success story from Andhra Pradesh. Water Energy Int. 60: 32-37. MIB (Ministry of Information and Broadcasting). 2003. India 2003: a reference manual. Publication Division. MIB, Government of India, New Delhi. 938 p.

IRRN 29.2

93

NOTES FROM THE FIELD

Vanishing populations of Oryza minuta Presl. in Pangil, Laguna, Philippines

M.C.N. Banaticla and M.S.R. Almazan, IRRI

Oryza minuta Presl., a wild rice species known to be resistant to bacterial blight and blast and green and brown planthopper, is endemic to the Malesian floristic region, specifically in Thailand, the Philippines, Indonesia, and Papua New Guinea. While it has been widely used in breeding and other experimental studies, its ecology and population dynamics remain poorly understood. In 1980, Vaughan documented six populations of O. minuta located along a stream in Barrio Balian, Pangil, Laguna. Seed samples from these populations were collected and later conserved at IRRIs International Rice Genebank. On 18 Feb 2004, we surveyed the same site and found that all the populations reported by Vaughan (1980) no longer existed. The small irrigation stream where the three populations once stood has been converted into a cemented canal (Fig. 1). During the visit, a road was being constructed near the former location of the other three populations. We encountered five clumps of O. minuta in different locations along the main stream, but unlike the vastly spreading populations described by Vaughan (1980),

they appear in small clusters of 23 plants (Fig. 2). Two clusters were partly strangled by Mikania cordata, a weedy herbaceous vine. The exclusive occurrence of O. minuta along the stream indicates its dependence on water flow for seed dispersal. It is possible that a mother plant or population located farther upstream serves as the seed source. It would be interesting to track this pre-

sumed mother population and learn more about its biology. In a period of 24 years, the populations of O. minuta in Barrio Balian, Pangil, Laguna, have greatly dwindled, leaving only traces of individual plants. Livestock raising, farming, and road construction destroy the natural vegetation within the vicinity. Such human activities probably played a major role in the

Fig. 1. Cemented canal where O. minuta populations once thrived.

Fig. 2. A cluster of O. minuta found in the study area.

94

December 2004

alarming disappearance of O. minuta populations in the area. Further studies on the conservation and population dynamics of O. minuta in Pangil, Laguna, are needed to prevent the local extinction of this species.

Reference

Vaughan DA. 1980. Oryza minuta observations in its natural habitat. IRRI trip report, Philippines.

Current status of rice pests and their management in Assam, Indiaa discussion with extension agents
Z. Islam, K.L. Heong, and M. Bell, IRRI; and L.K. Hazarika, D.J. Rajkhowa, S. Ali, B.C. Dutta, and M. Bhuyan, Assam Agricultural University, Jorhat 785013, India

2 h discussion sessions on each of the major groups of pests (insects, weeds, diseases, and rodents) during an integrated pest management training course implemented at Assam Agricultural University (AAU) in Jorhat on 5-15 Jul 2004. The discussion sessions were led by AAU experts and 19 trainees involved in rice pest management education, research, and extension participated. The gists of these sessions are presented. Insect pests Although many insects feed on rice plants, only a few occasionally reach pest status in Assam. The potential insect pests include rice hispa [Dicladispa armigera (Olivier)], yellow stem borer [Scirpophaga incertulas (Walker)], rice bug (Leptocorisa spp.), brown planthopper [Nilaparvata lugens (Stl)], leaffolders [Cnaphalocrosis medinalis (Guene) and Marasmia spp.], caseworm [Nymphula depunctalis (Guene)], and thrips [Stenchaetothrips biformis (Bagnall)]. So far, reliable yield loss estimates are not available and their exact pest status remains uncertain. The important potential pest is rice hispa, which is endemic in Sibsagar, Lakhimpur, Nalbari, Borpeta, Cachar, and Karimganj districts. It is more abundant during sali and ahu seasons than in boro. In outbreak situations, high yield losses may occur in specific fields. Several species of stem borers are present and yellow stem borer is considered predominant; however, conflicting opinions arose with re-

Rice is the most important crop in Assam, India, grown on about 70% of the total cultivated land (3.64 million ha) in the state (MA 2003). In fact, about 70% of the rice area of the seven northeast Indian states is under Assam. Rice is mostly grown in the low-lying deltas of the Brahmaputra and Barak rivers; some rice is also cultivated in upland situations in the northern hill region. The sali (monsoon crop: transplanted lowland rainfed rice and deepwater rice) rice is dominant, occupying 68.8% of the total rice area of 2.54 million ha, followed by ahu (summer rice) (18.3%), and boro (dry-season/winter) rice (12.9%). To understand Assams current pest problems in rice, and their management, we arranged four
IRRN 29.2

spect to the relative importance of the other species. The incidence of brown planthopper, thrips, leaffolders, and caseworm seems to be increasing, while that of earcutting caterpillars [Mythimna separata (Walker)] and swarming caterpillars [Spodoptera mauritia (Walker)] has decreased significantly compared with their incidence in the pre-green revolution period. The incidence of brown planthopper is relatively greater in the boro, particularly in the Barak River Valley. The rice bug is a major problem in the ahu season, particularly in the early ahu crop. Caseworm is a localized pest and occurs more in the sali season. Root and panicle aphids, white grubs, and rice bugs are major insect pests of upland rice, including the jhum (slash-and-burn) system. So far, insecticide use is minimum and there is some use of botanicals (extracts of plants). However, the extent of insecticide and botanical use and whether farmers are deriving any benefit from their use are yet to be determined. The AAU identified an effective muscardine fungus against rice hispa, whose effect is comparable with that of insecticide. It affects eggs, larvae, and adults but is more efficient in killing eggs. A mass culture method has been developed, but production and marketing have yet to be taken up by either the private or public sector. Diseases Assam is very rich in rice genetic diversity. Many traditional cultivars possess genes
95

of resistance to most rice diseases. Many diseases infest rice plants in Assam, but the most important ones are sheath blight, bacterial blight, and blast in the sali; blast and sheath rot in the ahu; sheath rot and sheath blight in the boro; and ufra and root-knot in deepwater rice. The extent of average yield losses attributed to diseases is not known. However, it is assumed that 1520% yield losses may occur in some infested fields, although average losses would be much less. Farmers seem to be not very much aware about diseases, unlike what they know of other pests. Although several cultural and chemical options are available for the management of each of the major diseases, it is not known what percentage of the farmers use these technologies, if at all. Some fungicides are available at the grassroots level, but their use is very low. However, most of the popular modern rice varieties possess resistance to or tolerance for blast, and some degree of tolerance for sheath blight and bacterial blight (Table 1). Weeds The agroclimatic conditions of Assam favor the rapid growth of weeds. The type of rice culture, season, soil, and cropping pattern influence the weed flora in rice fields. The most important weed flora in the different rice cultures are listed in Table 2. Among rice cultures, the weed problem is most severe in dry-seeded rice (upland rice), followed by wet-seeded (sprouted seeds) and transplanted rice. Trans96

Table 1. Resistance status of 12 most popular rice varieties of Assam, India. Variety according to rank Season Blast Sali Sali Sali Ahu Ahu Boro Sali Sali Sali Sali Boro Boro T S MR MR T MR T T MR MR MR MR Sheath blight T S T MS MS T T T T T T T Reaction to major diseases Sheath rot MR T MR Bacterial blight MS T T T T T T T T T Stem rot MS T MS Remark Neck blast (R)

Ranjit Mahsuri Bahadur Luit Disang Joymati Aghoni Keteki Manoharsali Andrewsali Joytiprasad Bishnuprasad

Neck blast (T) MS MS

MR MR MS

R = resistant, MR = moderately resistant, MS = moderately susceptible, T = tolerant.

planted rice has fewer weed problems because puddling, followed by transplanting of rice seedlings, gives rice plants a head start and enables farmers to use water for weed suppression. The weed flora in rice are dominated by grasses, followed by broadleaves and sedges (Barua and Gogoi 2002). However, alternate wetting and drying conditions that prevail in any culture may result in severe weed problems. The diverse weed flora in deepwater rice emerge in three flushes before, during, and after the floods. Weeds may reduce rice yields by 1540%, if not controlled at all. In general, farmers are aware of harmful effects of weeds and they make serious efforts to control weeds using direct and indirect methods. The direct methods practiced by Assamese farmers include manual weeding by using simple tools, the use of simple rotary weeders, and herbicides. So far, manual weeding is the most common, whereas the use of herbicides is very

limited. Indirect methods involve tillage, the use of water, transplanting of rice seedlings, etc. Rodents To date, 34 species of rats and mice under 12 genera were recorded in northeast India, including Assam (Singh et al 1994). The dominant species are Bandicota bengalensis in all states of northeast India; Rattus rattus khyensis in Mizoram; R. bowersi in Meghalaya and Miaoram; Rattus nitidus nitidus in Arunachal Pradesh, Meghalaya, and Sikkim; R. rattus tistae in Arunachal Pradesh, Manipur, Meghalaya, Mizoram, Sikkim, and Tripura; and Niviventer niviventer in Meghalaya. It is believed that the outbreak of Bambusa tulda and Dendrocalamus longispathus is mainly associated with the flowering of the major bamboo variety Thingtam, while that of Melocanna bambusoides is associated with the flowering of variety Mautam (Kumar and Pathak 2002). Three rat outbreaks coincided with
December 2004

Table 2. Major weed flora in different rice cultures in Assam, India. Weed species Directseeded Grasses Cynadon dactylon Digitaria setigera Echinochloa crus-galli E. stagnina Eleusine indica Eragrostis unioloides Hackelochloa granularis Hymanachne acutigluma Isachne himalatica Leersia hexandra Paspalum conjugatum Paspalum scrobiculatum Sacciolepis interrupta Sedges Cyperus iria C. pilosus C. rotundus Eleocharis acutangula E. dulcis Fimbristylis littoralis Fimbristylis spp. Scirpus juncoides S. maritimus Other weeds Ageratum houstonianum Cuphea balsamona Eichhornia crassipes Fissendocarpa linifolia Ludwigia adscendens Melochia corchorifolia Mimosa pudica Monochoria vaginalis Sagittaria guayanensis Salvinia natanes Sphenoclea zeylanica Spilanthes paniculata v v v v v v v v v v v v v v v v v v v v v v v Abundance as major weed in different rice cultures TP Ahu TP Sali Wetseeded TP Boro DWR

v v

v v v

v v

v v v

v v v

(relative abundance of 59%, 7%, 19.1%, 14.5%, and 5.8%, respectively). Rice tiller cut by rats was estimated at 5.7% in sali, 3.3% in ahu, 5.0% in boro, and 10.0% in deepwater. Tiller damage at the vegetative phase and at panicle initiation, dough, and ripening stages in the sali season was estimated at 2.4%, 3.5%, 7.4%, and 5.7%, respectively. Farmers use local traps, rodenticides, and cats and dogs for rodent control mainly in the homestead. But they rarely make an effort to control rats in rice fields. References
Barua IC, Gogoi AK. 2002. Diversity of weed flora in upland direct seeded rice ecosystem in Assam. In: Proceedings of the UGC-sponsored State-level Seminar on Biodiversity of Assam and Its Conservation. Karimgaj, Assam (India): Karimganj College. p 108-116. Kumar D, Pathak KA. 2002. Bamboo flowering and rodent outbreak in the northeast hill region of India. In: Glimpses of rodent research in India. AICRP on Rodent Control. Jodhpur (India): All India Coordinated Research Project. p 40-43. MA (Ministry of Agriculture). 2003. Basic statistics of north east region, 200203. Directorate of Economics and Statistics, MA, Government of India. Singh YP, Kumar D, Gangwar SK. 1994. New records of some rodent species from the states of North Eastern hill region. Rodent Newsl. 18(5):12-13.

v v v v v

v v v v v v v

v v

v v v

DWR = deepwater rice, TP = transplanted.

bamboo flowering in 1911, 1929, and 1956, leading to famine in hilly regions of Assam and other northeast Indian states and forcing migration from the hilly regions to the valleys. The flowering of Thingtam bamboo also occurred in 1977, but extensive crop damage was not observed, which was attributed to control measures set by the government. The next flowering of Mautam bamboo is expected between 2004 and
IRRN 29.2

2007. The Indian Council of Agricultural Research started the All India Coordinated Research Project on Rodent Control in 1977 and gradually established 10 centers, two of which are located in Shellong (Meghalaya) and Jorhat (Assam). Recent studies in the river valleys in Assam revealed that B. bengalensis is the dominant species in rice fields, followed by B. indica, M. booduga, and R. sikkimensis

97

INSTRUCTIONS TO CONTRIBUTORS IRRN welcomes three types of submitted manuscripts:research notes, mini reviews, and notes from the field. All manuscripts must have international or pan-national relevance to rice science or production, be written in English, and be an original work of the author(s), and must not have been previously published elsewhere. By submitting the manuscript, the author automatically assigns the copyright of the article to IRRI. Research notes Research notes submitted to IRRN should report on work conducted during the immediate past 3 yr or work in progress advance rice knowledge use appropriate research design and data collection methodology report pertinent, adequate data apply appropriate statistical analysis, and reach supportable conclusions. Routine research. Reports of screening trials of varieties, fertilizer, cropping methods, and other routine observations using standard methodologies to establish local recommendations are not ordinarily accepted. Preliminary research findings.To reach well-supported conclusions, field trials should be repeated across more than one season, in multiple seasons, or in more than one location as appropriate. Preliminary research findings from a single season or location may be accepted for publication in IRRN if the findings are of exceptional interest. Preliminary data published in IRRN may later be published as part of a more extensive study in another peer-reviewed publication, if the original IRRN article is cited. However, a note submitted to IRRN should not consist solely of data that have been extracted from a larger publication that has already been or will soon be published elsewhere. Multiple submissions. Normally, only one report for a single experiment will be accepted.Two or more items about the same work submitted at the same time will be returned for merging. Submitting at different times multiple notes from the same experiment is highly inappropriate. Detection will result in the rejection of all submissions on that research. Manuscript preparation. Arrange the note as a brief statement of research objectives, a short description of project design, and a succinct discussion of results. Relate results to the objectives. Do not include abstracts. Up to five references may be cited. Restrain acknowledgments.Limit each note to no more than two pages of double-spaced typewritten text (approximately 500 words). 98 Each note may include up to two tables and/or figures (graphs, illustrations, or photos). Refer to all tables and figures in the text. Group tables and figures at the end of the note, each on a separate page. Tables and figures must have clear titles that adequately explain contents. Apply these rules, as appropriate, to all research notes: Methodology Include an internationally known check or control treatment in all experiments. Report grain yield at 14% moisture content. Quantify survey data, such as infection percentage,degree of severity, and sampling base. When evaluating susceptibility, resistance, and tolerance, report the actual quantification of damage due to stress, which was used to assess level or incidence. Specify the measurements used. Provide the genetic background for new varieties or breeding lines. Specify the rice production systems as irrigated, rainfed lowland, upland, and flood-prone (deepwater and tidal wetlands). Indicate the type of rice culture (transplanted, wet seeded, dry seeded). Terminology If local terms for seasons are used, define them by characteristic weather (dry season,wet season,monsoon) and by months. Use standard, internationally recognized terms to describe rice plant parts, growth stages, and management practices. Do not use local names. Provide scientific names for diseases, insects, weeds, and crop plants. Do not use local names alone. Do not use local monetary units. Express all economic data in terms of the US$, and include the exchange rate used. Use generic names, not trade names, for all chemicals. Use the International System of Units for all measurements. For example,express yield data in metric tons per hectare (t ha-1) for field studies. Do not use local units of measure. When using acronyms or abbreviations, write the name in full on first mention, followed by the acronym or abbreviation in parentheses. Use the abbreviation thereafter. Define any nonstandard abbreviation or symbol used in tables or figures in a footnote, caption, or legend. Mini reviews Mini reviews should address topics of current interest to a broad selection of rice researchers, and highlight new developments that are shaping current work in the field. Authors should contact the appropriate editorial board member before submitting a mini review to verify that the subject is appropriate and that no similar reviews are already in preparation. (A list of the editors and their areas of responsibility appears on the inside front cover of each IRRN issue.) Because only 1-2 mini reviews can be published per issue, IRRN will require high quality standards for manuscripts accepted for publication.The reviews should be 2000-3000 words long, including references.Refer to the guidelines for research notes for other aspects of writing and content. Notes from the field Notes from the field should address important new observations or trends in rice-growing areas, such as pest outbreaks or new pest introductions, or the adoption or spread of new crop management practices.These observations,while not the result of experiments, must be carefully described and documented. Notes should be approximately 250 words in length. Refer to the guidelines for research notes for other aspects of writing and content. Review of manuscripts The IRRN managing editor will send an acknowledgment card or an email message when a note is received.An IRRI scientist, selected by the editorial board,reviews each note. Depending on the reviewers report, a note will be accepted for publication, rejected, or returned to the author(s) for revision. Submission of manuscripts Submit the original manuscript and a duplicate, each with a clear copy of all tables and figures, to IRRN. Retain a copy of the note and of all tables and figures. Send manuscripts, correspondence, and comments or suggestions about IRRN by mail or email to: The IRRN Managing Editor IRRI, DAPO Box 7777 Metro Manila Philippines Fax: +63 (2) 580-5699; 891-1174 E-mail: t.rola@cgiar.org

December 2004

DAPO Box 7777, Metro Manila, Philippines

Printed matter