C E Y L O N C O C O N U T Q U A R T E R L Y Volume XV January-June, 1964 Nos. 1/2 CONTENTS PAGE 1. ANNUAL REPORT OF THE COCONUT RESEARCH INSTITUTE FOR 1963 . . 1 2. REPORT OF THE DIRECTOR 4 3. REPORT OF THE CHEMIST 11 4. REPORT OF THE BOTANIST 23 5. REPORT OF THE SOIL CHEMIST 26 6. REPORT OF THE AGROSTOLOGIST 41 7. REPORT OF THE CROP PROTECTION OFFICER 50 8. REPORT OF THE BIOMETRICIAN 58 9. REPORT OF THE SOIL SURVEY UNIT 64 10. REPORT OF THE CHIEF ADVISORY OFFICER 75 11. REPORT OF THE PLANTING OFFICER 81 12. REPORT OF THE WELFARE OFFICER 84 13. REPORT OF THE PUBLICATIONS OFFICER 86 14. REPORT ON THE ESTATES I. BANDIRIPPUWA ESTATE . . . . . . . . 88 n . RATMALAGARA ESTATE . . . . . . . . . . . . . . 91 Price: Rs. 2-50 Annual Subscription: Rs. 4-50 Date of Publications—3rd March, 1965. PRINTED BY THE COLOMBO APOTHECARIES' CO. , LTD. , COLOMBO ANNUAL REPORT OF THE COCONUT RESEARCH INSTITUTE FOR 1963 The present report is the 35th Annual Report of the Coconut Research Institute, which was established by Ordinance No. 29 of 1928 dated December 1928. t REPORT OF THE CHAIRMAN On January 1963, the Coconut Research Board consisted of the following:— Ex-Officio Members Director of Agriculture, Mr. A.V. Richards, M.Sc. (Calif.), B.Sc. (Lond.), Dip.Agric. (Cantab.) A.I.C.T.L. (Trinidad). Treasury Representative: Mr. G.M. Sparkes. Coconut Rehabilitation Commissioner: Mr. B. Mahadeva. Chairman, Low Country Products' Association: Mr. Francis Amarasuriya. Director, Coconut Research Institute: Dr. M.L.M. Salgado, Ph.D. (Cantab.), B.Sc. (Lond.), Dip.Agric. (Cantab.). Nominated Members Nominated by the Honourable Minister from Senators and Members of Parliament. Mr. Wijebahu Wijesinghe, M.P. Mr. W.I. Hugh Fernando, M.P. Nominated by the Planters' Association of Ceylon. Mr. C.T. Van Geyzel, J.P. Mr. B. Warusavitana. Nominated by the Low Country Products' Association. Senator Thomas Amarasuriya, O.B.E. (Chairman, Coconut Research Board). Mr. C.A.M. de Silva. r Nominated by the Honourable Minister to represent the Small-holders. Mr. A.F. Samarasinghe. Mr. L.W.A. Fernando. During the period 1963 Mr. W.I. Hugh Fernando, M.P. was elected Speaker of the House of Repre­ sentatives. The place of Mr. Hugh Fernando on the Coconut Research Board was taken by Mr. S.D.R. Jayaratna, M.P., Senator Thomas Amarasuriya the Chairman, Coconut Research Board resigned from 1 the Board when he was elected President of the Senate and his place was taken by Mr. C. Selwyn Samarawcera who was nominated by the Low Country Products' Association. With the resignation of Mr. G.M. Sparkes, from the public service, the post of the Treasury Representative on the Board remained vacant. Mr. C.A.M. de Silva was unanimously elected Chairman of the Coconut Research Board. Meetings.*—Four meetings of the Coconut Research Board were held during the year on 16th March, 15th June, 14th September and 19th December. Three Special Meetings of the Coconut Research Board were held on September 26th, November 22nd and December 4th. Committees Administrative Committee (Personnel at 1st January 1963): (1) Chairman of the Board. (2) Mr. G.M. Sparkes. (3) Mr. B. Warusavitana. (4) Mr. C.A.M. de Silva. (5) Mr. B. Mahadeva. (6) Mr. A.V. Richards. (7) Mr. W.I. Hugh Fernando. (8) Director, Coconut Research Institute. The 42nd, 43rd, 44th and 45th meetings of the Administrative Committee were held on 25th January, 26th April, 26th July and 25th October 1963, respectively. Estate and Experimental Committee (Personnel at 1st January 1963): (1) Mr. Francis Amarasuriya. (2) Mr. A.V. Richards. (3) Mr. B. Warusavitana. (4) Mr. C.T. Van Geyzel, J.P. (5) Chairman of the Board. (6) Mr. L.W.A. Fernando. (7) Mr. Wijebahu Wijesinghe, M.P. (8) Director, Coconut Research Institute. Three Co-opted Members: (1) Mr. X. Jobin. (2) Mr. M.M. Kumarakulasingham. (3) Mr. R.H. de Mel. The 43rd, 44th, 45th and 46th meetings of the Estate and Experimental Committee were held on 16th February, 15th May, 17th August and 16th November, 1963 respectively. Extension Committee (Personnel at 1st January 1963): (1) Mr. C.T. Van Geyzel, J.P. (2) Mr. Francis Amarasuriya. (3) Mr. C.A.M. de Silva. (4) Mr. B. Mahadeva. (5) Mr. A.F. Samarasinghe. (6) Mr. L.W.A. Fernando. (7) Director, Coconut Research Institute. The 35th, 36th and 37th meetings of the Extension Committee were held on 9th February, 10th August and 8th November, 1963 respectively. Editorial Committee (Personnel at 1st January 1963): (1) Mr. Francis Amarasuriya. (2) Director, Coconut Research Institute. (3) Dr. D.V. Liyanage. (4) Mr. Wijebahu Wijesinghe, M.P. (5) Mr. A.F. Samarasinghe later Mr. L.W.A. Fernando. C.A.M. D E SILVA, Chairman, Coconut Research Board. REPORT OF THE DIRECTOR 1. STAFF Staff of the Coconut Research Institute at the end of 1963 was as follows:— Administrative Division Director—Dr. M.L.M. Salgado, Ph.D. (Cantab.), B.Sc. (Lond.), Dip.Agric. (Cantab.). Chief Administrative Officer and Secretary to the Board—Mr. S.C. Kahawita; B.Com. (Lond.). Assistant Secretary—Mr. T.T.A.J.C. Samarasinghe, LL.B. (Ceylon): Soil Chemistry Division Soil Chemist—Dr. D.A. Nethsinghe, D.Phil. (Oxon.), B.Sc. (Ceylon), A.R.I.C. Research Assistant—Mr. T.S. Balakrishnamurthi, B.Sc. (Ceylon). (On overseas Study leave). Chemistry Division Chemist—Dr. W.R.N. Nathanael, Ph.D. (Lond.), M.Sc. (Lond.), F.R.I.C. Research Assistant—Vacant. Botany Division Botanist—Dr. D.V. Liyanage, Ph.D. (Manch.), B.Sc. (Lond.). Research Assistant—Mr. M.A.P. Manthriratne, B.Sc. (Lond.). Agrostology Division Agrostologist—Dr. K. Santhirasegaram, Ph.D. (Adelaide), B.Sc. (Ceylon). Research Assistant—Mrs. N. Rajaratnam, B.Sc. (Ceylon). Planting Division Planting Officer—Mr. P.D.L. Fernando. Assistant Planting Officer—Mr. C.W.S. de Silva. Advisory Division Chief Advisory Officer—Mr. CA. Wickremasuriya, B.Sc. (Ceylon). Crop Protection Division Crop Protection Officer—Mr. U.B.M. Ekanayake, B.Sc. (Ceylon). Biometry Division Biometrician—Mr. V. Abeywardena. Research Assistant—(Statistics)—Mr. J.K.T. Fernando, B.Sc. (Ceylon). 4 Dr. K. Santhirasegaram who was away in Australia on post-graduate study returned to Ceylon, having obtained the Ph.D. degree of the University of Adelaide. Dr. W.R.N. Nathanael, Chemist was elected a Fellow of the Royal Institute of Chemistry. Mr. T. Balakrishnamurthi proceeded on post-graduate study leave to the University of Aberdeen, Scotland. 2. MEETINGS The Director attended the following meetings besides those of the Coconut Research Board and its Committees:— (i) Annual General Meeting of the Planters' Association of Ceylon. c (ii) Meetings of the District Planters' Associations of Chilaw/Negombo and Kurunegala Planters' Associations. (iii) Central Board of Agriculture. (iv) Animal Husbandry Sub-Committee of Central Board of Agriculture. (v) Committee appointed by Hon. Minister of Agriculture on Food and Milk Production on Coconut Estates. (vi) Seminar on Coir at Small Industries Corporation, Velona, Katubedde, Moratuwa. 3. RESEARCH CONFERENCES The following subjects were discussed at Research Conferences held during the year:— (i) "Problems of Extension Work on Coconuts" led by Mr. CA. Wickramasuriya, Chief Advisory Officer. (ii) "Problems of Replanting of Senile Coconut Palms" led by Dr. D.V. Liyanage, Mr. R.H. de Mel, Co-opted Member of the Estates and Experimental Committee was present and partici­ pated in the Discussions. (iii) "Some Recent Soil and Land Use Surveys of Coconut Development Schemes in Ceylon" led by Mr. K.S.O. Perera, O.I.C., Soil Survey Unit. 4. VISITORS The Visitors during the year included the following:— Dr. L. Horton, Assistant Director, Tropical Products Institute, London. Mr. J. Legrund, Commercial Attache, French Embassy. Mr. D. Rhind, Adviser in Agricultural Research—Department of Technical Co-operation, London. Mr. W.V.D. Peiris, Coconut Improvement Officer, F.A.O. Dr. W.D. Raymond, Tropical Products Institute, London. Mr. A.H. Green, Plantations Group, Unilever Ltd.—spent 3 days at the Institute. Dr. Karl Maramorosch, Boyce Thompson Institute, New York. Members of the Animal Husbandry Sub-Committcc of the Central Board of Agriculture. The Chinese Agricultural Delegation. Miss Lyndon Moore, Agricultural Economics Research Institute, University of Oxford. Dr. H.S. Hopf, Scientific Secretary of the Tropical Pesticide Research Committee. Mr. H.A. Hassan, Director, Botanical Gardens, Assovan, Egypt. Mr. Kuo Ying, Counsellor of the Embassy of the People's Republic of China in Ceylon. 5 5. TRAINEES Mr. D.E. Nwbsisi, Manager, Eastern Nigeria Development Corporation Coconut Estate. Messrs. Hassan Bin Abdulia and Othman Bin Mohamed from the Department of Agriculture, Malaya. Mr. Frederic Lewis, Nominee of the Government of Cuba. Mr. Atiqur Rahaman, Manager, Industrial Estates, East Pakistan Small Industries Corporation, studied copra processing and coconut cultivation work. 6. FIELD DAYS Two Field Days were held at Ratmalagara Estate and the Isolated Seed Garden on 7th January 1963 for the members of the Low Country Products' Association and on 18th July 1963 for the members of the Kurunegala Planters' Association. 7. PUBLICATIONS The Director contributed articles on "Coconuts in 1962" and "Activities of the Coconut Research Institute in 1962" to the Annual Report of the Planters' Association of Ceylon. An article by the Director entitled "Diversification of the Coconut Economy" was contributed to the "Times of Ceylon" Agricultural and Engineering Supplement. 8. NOTES ON REPORTS OF DIVISIONS The following notes draw attention to points of interest relating to the detailed reports of the Technical and Extension work of the Institute:— Chemistry (i) Desiccated Coconut:—Consequent on certain complaints received by the Ceylon Coconut Board from overseas buyers, regarding the problem of oil exudation from certain shipments of Ceylon Desiccated Coconut packed in paper bags, some ad hoc experiments have been conducted to establish the factors responsible for this phenomenon. On the basis of chemical, physical and microscopical examination of parallel samples of Ceylon and Philippine Desiccated Coconut, the tentative conclusion has been that oil exudation results from some sort of cell-Wall degradation or breakdown. It is probable that this is associated with certain critical temperature combinations and/or limits for sterilization and drying. The problem continues to be investigated. In preference to the hydraulic press method, a more satisfactory technique using thick quality / (No. 3) filter paper sandwiched between metal plates has been evolved for routine testing of oil exudation. This method is now being used regularly in the quality testing laboratory of the Ceylon Coconut Board. (ii) Grated Coconut Meat:—The grated meat from the fresh coconut kernel is used in the Ceylon household for culinary purposes. Further, it has been estimated that the annual per capita consumption of coconuts is 140. This would mean that the annual domestic consumption (for cooking and other uses in the household) would be in the region of 1,400 million nuts. 6 For cooking purposes, it is customary to press out manually the "milk" from the grated meat using a certain amount of hot water. With a view to obtaining an estimate of the efficiency of this extrac­ tion—(particularly as regards oil and protein), a simple experiment simulating the household method was carried out. The samples of grated meat obtained during three successive manual pressings were then analysed chemically. The results have shown that there is a colossal waste of valuable oil and protein in the "pol-kudu", that is discarded as waste. In terms of weights—on the basis that 1,200 nuts give 560 lbs. (a candy) of copra, (6% moisture) with an oil content of 68.0% (dry basis) it has been computed that the loss of oil represents 21.3% of the oil per nut. The corresponding total weight loss on 1,400 million nuts con­ sumed locally would thus amount to 39,000 tons per annum. In other words, the economic loss to the industry (without taking into account the protein) would be in the region of 43.5 millions. If the Central Bank figure of 990 million nuts of annual domestic consumption is adopted, then the corresponding economic loss would be 30.8 million. (Calculated on the basis that the value of a ton of oil is Rs. 1,100/-). (iii) Vinegar:—(a) The Vinegar Generator installed at the new factory Nainamadama which was started off in June 1962, continued to operate very satisfactorily during the year .The factory was inspected at regular intervals and the acid strength of the finished vinegar was found to be consistently over 5.5%. The maize cobs originally used in the packing chamber were found to last the guaranteed period of one year without any form of degradation or breakdown. (b) A satisfactory method was evolved for the preparation of caramel (of high tinctorial efficiency), for colouring vinegar. A bottle of this caramel has been found adequate to colour 20 gallons of vinegar to a dark wine-tint. This method has been demonstrated to some vinegar-makers and it is now being applied at the Nainamadama factory for colouring their "Generator" Brand Vinegar. (c) An experiment was successfully carried out to test the efficacy of common salt in retarding the growth of mycodermal film in bottled vinegar. It was found that a minimum concentration of 2.0% salt dissolved in the vinegar would completely arrest the formation of any mycodermal suspensions. As an effective substitute for sterilization, common salt at a concentration of 2% is now being used at the Nainamadama Factory. Botany Work on hybridisation and selection, progeny trials and planting techniques were continued during the year. The three progeny trials at Marandawila, Walpita and Bandirippuwa were maintained during the year. Nuts of crosses done during the previous year, viz. dialled, testing male transmitters, study of segre­ gation of certain characters, etc., were harvested and some of them planted in the nursery. Altogether 375 seedlings were planted at Pottukulama Sub-Station in November 1963. A programme of work was undertaken to study whether a uniformly high-yielding king-coconut palm without periodicity in bearing, and possibly early bearing, could be developed by crossing selected palms within and between forms. Crossing of typica X typica and typica X punrilla where the male parent is a prepotent was continued at six stations during the year. A technique to increase the quantity of pollen collected from an inflorescence was devised, based on a method developed by Jamaican workers. The pollen bank was well patronised by 8 private estates and 3,856 samples of pollen issued to them. Preliminary studies on culturing embryos were undertaken. 1,880,644 selected seed nuts were supplied to the Planting Division during the year. Soil Chemistry Seven long-term field experiments on manurial problems were continued in 1963—in addition to the observation trials on magnesium deficiency and leaf scorch problems. The results of field experiments on bearing palms at Bandirippuwa and Ratmalagara give support to our current recommendation on NPK manuring (in 1960 we recommended an increase in the level of NPK manuring to about double the rates previously recommended). The experiment comparing organic with inorganic fertilizers on a light sandy loam at Marandawila Estate, Bingiriya has shown, that organic fertilizers offer no advantage over the cheaper inorganic fertilizers. In the experiment on methods of fertilizer placement at Walahapitiya Estate, Nattandiya, there was little difference between the surface application of fertilizers in a circular strip 3 feet wide 3 feet away from the palm, and the considerably more expensive traditional circular trenches. Broadcast applications proved to be less effective. A second field trial on adult palms using radioactive phosphorus confirmed that placement in the entire area round the palm up to a distance of about 6 feet from the bole was more efficient than placement in a 3 feet wide circular strip 3 feet away from the palm, or placement in the centre of the squares. Chemical analysis of leaflet samples from palms affected by "Leaf Scorch" in the Southern Province has shown that these palms are deficient in Boron and Magnesium. The possible association of the problem with nutrient deficiencies cannot yet be overruled, and further trials on this aspect are in progress. Agrostology Soil Fertility:—All coconut soils so far studied showed acute deficiencies of N.P.K. except those from Bandirippuwa where phosphate gave positive response only, at the early stages. However, soils sampled from the newly acquired sections are showing phosphate response as to be expected from the mineral composition of the soils. In experiments carried out to study minor nutrient status of the Gonapinuwela soils where coconuts are showing peculiar symptoms, boron applied as sodium borate deca hydrate at 3 lbs./acre gave signifi­ cantly high yields of Medicago sativa. However this deficiency of B has also been noted with other coconut soils (cf. cinnamon sand at Horrekelle and latcritic gravel at Mattegoda) where the coconuts are not affected as in the South. The M. sativa plants growing in the Gonapinuwela soils developed symptoms which could not be explained on the basis of nutritional deficiencies; and examination of the root system showed decay of the tips and lesion scattered over the entire system. The presence of a parasitic nematode Hemicycliophora longicaudata in the roots has been reported by the Crop Protection Division. 8 Pasture under Coconuts:—Obeservations on the effect of Brachiaria brizantha and B. milliformis on coconuts and ancillary experiments on the performance and suitability of these species have shown that: (a) The pasture species would compete with coconuts for soil moisture and nutrients. It is suspected that competition for soil moisture would be more severe than for nutrients particularly under low rainfall conditions. (b) Both species yield to the same extent under similar conditions. (c) B. milliformis responds more to nitrogen manuring than B. brizantha. (d) B. milliformis is dominant over B. brizantha in the light microclimate under coconuts and the level of nitrogen manuring in these experiments. (e) The herbage yield of B. milliformis in particular is less during the N/E monsoon compared to the S/W monsoon. This appears to be associated with early maturity during the former season due to shorter day length at that time. Crop Protection The Parasite Breeding Station at Mylambavely, Batticaloa was opened in April 1963 and Mr. Edwin Dharmaraju, the Colombo Plan Entomologist, left in May on the completion of his assignment in Ceylon. Over five million- parasites of the Coconut Caterpillar were produced and over two million were released to fifty estates. Life cycle studies were carried out on Microbracon brevicornis and Persierola nephantides and field studies on the pest-parasite complex carried out. Until now no field recoveries of Microbracon brevicornis have been made and the factors responsible for this are under study. The Microbial insecticide Baithane L—69, based on Bacillus thuringiensis was found to be effective against the larvae of Nephantis serinopa. A number of insecticides have been tested in the Laboratory against the Red Weevil, Rynchophorus ferruginens and telodrin was found to be the most effective against the larvae and pupae of this pest. During 1963, much emphasis was laid on the study of the "Leaf Scorch" problem in the Gonapinu- wela area. Studies on its mode of spread in the field and the symptomotology are in progress. Two weak parasitic fungi, Pestalotiopsis palmarum (Cke) Petrak and Sydow., have been isolated from the leaves of affected palms; but are not likely to be the primary causes of the condition; Histopathological studies have revealed that there is a blackening of the cortical cells and a blockage of the xylem vessels of the roots. A nematode, Hemicycliophora longicaudata, Loos(Criconematidae)* has been found to be associated with roots of affected palms. H. longicaudata, Loos is found to occur in large numbers in the soil of the Gonapinuwela area. It has not been possible to ascertain, as yet, whether this is the primary cause of Leaf Scorch, and investigations on this aspect are in progress, in addition to studies on the micro-organisms associated with the roots and soil and the environmental conditions necessary for the spread of "leaf- scorch". Mycological studies on Bud-rot and Nut-fall of coconut were conducted. * It has been renamed Caloosia Longicaudata, Loos. Biometry 10 The Biometrician besides assisting the Research Divisions in their statistical work as well as the designing of experiments, carried out the following items of research:— „ (a) A study was made of the patterns of the within year variations of coconuts crops vis-a-vis the patterns of variations of the crop components. (b) Nut water nutrients data were re-examined in relation to applied fertiliser dosages as well as yields and some new trends were observed. Soil Survey The Soil Survey of Puttalam North area, mainly comprising the Vanatavillu Middle Class Scheme which was commenced in 1962 was continued in 1963. Soils have been classified into three series, consisting of red, yellow, and bleached yellow series, the latter lying in the Valley of the Villus. A Soil Survey of lands along the Padawiya-Pulmoddai Road was carried out. Three soil types were observed—gravelly soils on the ridges, a deep alluvium and a clay hydromorphic association. Only the second type was recommended as being suitable for coconut cultivation. Advisory and Extension Work Besides the normal advisory work and inspections under the Fertilizer Subsidy Scheme, 354,469 seedlings covering an acreage of 4,764 acres, and 118 tons of fertilizer and aldrex were issued in connection with the Citronella Subsidy Scheme. Nurseries:—2,096,096 seednuts were planted in the 14 Planting Division Nurseries and 454,610 1 . seednuts for the Citronella Subsidy Scheme in the 2 Nurseries at Medamulana and Mapalana. M.L.M. SALGADO, Director, Coconut Research Institute. REPORT OF THE CHEMIST 1. RECOVERY OF COCONUT OIL DURING EXPELLER PRESSING Monthly records kept at an oil mill in the Kurunegala District have shown that on an average a candy (560 lbs.) of copra (on expeller pressing) may be expected to yield 336 lbs. of oil (60.0%) and 191 lbs. of poonac (34.1 %). In other words, the gross operational losses on every candy would amount to 33 lbs. or 5.9 per cent of the weight of copra taken. In milling practice, it is generally estimated that 6.5 candies of copra would give a ton of oil and 0.5 ton of poonac, which are not widely different from the out-turns that have been recorded at the above mill. Analyses of samples of comminuted copra (before and after drying) as prepared for expeller pressing showed that at least 2 per cent of the losses are due to dryage during the preparation of desi-copra. Besides "copra dust", of course a certain amount of sediment poonac and oil are also recovered from the filter presses, which are normally reported as losses. Records are being kept at this mill in order to obtain quantitative estimates of these recoveries. Expeller Efficiency Samples of poonac were drawn from the three expellers (all of Indian manufacture), installed at this mill, and these were analysed with the following results:— Sample % Moisture %OIL Sample % Moisture Wet basis Dry basis Expeller No. 1 5.02 15.34 16.15 Expeller No. 2 5.06 17.82 18.76 Expeller No. 3 6.15 15.88 16.92 Mean 5.41 16.35 17.28 The residual oil contents in all three samples should be reckoned high for good expeller poonac. Regarding expeller efficiency—on the basis that a candy of copra [6.0% moisture and 68.3% oil (dry basis)] contains 359.52 lbs. of oil, the actual oil recovered per candy (viz. 336 lbs.) would represent 93.5 % of the potential oil contained in the copra. 2. DESICCATED COCONUT (a) Consequent on certain complaints received by the Ceylon Coconut Board from overseas buyers, regarding the problem of oil exudation from certain shipments of Ceylon Desiccated Coconut packed in paper bags, some ad hoc experiments have been conducted to establish the factor or factors responsible for this phenomenon. 11 CHEMICAL EXAMINATION OF CEYLON AND PHILIPPINE DESICCATED COCONUT (Samples supplied by the Ceylon Coconut Board on 27.4.63) TABLE I (a) 1 2 3 4 5 6 7 8 Sample % Passing through B.S. 10 mesh test sieve % Moisture %OIL Free Fatty acidity on Pressed oil (% Laurie) % Total Sugars (as Sucrose) % Protein % Total Mineral Matter Sample % Passing through B.S. 10 mesh test sieve % Moisture Wet basis Dry basis Free Fatty acidity on Pressed oil (% Laurie) % Total Sugars (as Sucrose) % Protein % Total Mineral Matter Sample % Passing through B.S. 10 mesh test sieve % Moisture Wet basis Dry basis Free Fatty acidity on Pressed oil (% Laurie) Wet basis Dry basis Wet basis Dry basis Wet basis Dry basis C E Y L O N (CO P H I L I P P I N E (PO 6 9 . 2 8 9 . 5 3 . 6 4 . 6 6 7 . 7 6 6 . 7 7 0 . 3 7 0 . 0 0 . 1 4 0 . 2 1 5 . 8 4 . 9 6 . 0 5 . 2 6 . 0 6 . 2 6 . 7 6 . 7 1 . 4 1 . 8 1 . 6 1 . 9 TABLE I (b) 1 | 2 Sample MINERAL ELEMENTS Sample On Original DCN(mg.per \00g.) O/i Oil-free Meal (Per cent) Sample % Mois­ ture K P Ca Mg N Mois­ ture % K P Ca Mg N Sample % Mois­ ture Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Mois­ ture % Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry C E Y L O N ( C , ) P H I L I P P I N E (PO 3 . 6 4 . 6 ____ 3 8 9 531 4 3 7 5 6 5 1 5 2 1 6 4 1 7 0 1 7 4 2 7 19 3 0 2 0 121 8 9 1 3 6 9 5 9 5 3 1 0 0 0 1 0 7 0 1 0 7 0 1 0 . 9 6 . 1 1 . 3 1 1 . 7 7 1 . 4 7 1 . 8 8 0 . 5 1 0 . 5 4 0 . 5 7 0 . 5 8 0 . 0 9 0 . 0 6 0 . 1 0 0 . 0 7 0 . 4 1 0 . 3 0 0 . 4 6 0 . 3 2 3 . 2 1 3 . 3 4 3 . 6 1 3;. 5 5 On the basis of chemical, physical and microscopical examination of parallel samples of Ceylon and Philippine Desiccated Coconut, the tentative conclusion has been that oil exudation results from some sort of cell-wall degradation or breakdown. It is probable that this is associated with certain critical temperature combinations and/or limits for sterilization and drying. In Ceylon, the pared kernel is sterilized for 1J minutes at a temperature of 210°F, whereas in the Philippines the corresponding sterili­ zation temporature is only 150°F for an immersion time of 10 minutes. The results of chemical examination of two samples of desiccated coconut used in this study are charted in Tables I (a) and I (b). Using a hydraulic press some preliminary oil exudation tests were also carried out on the two samples, and the results are summarised in Table II. TABLE II Sample Oil exuded (as % of weight taken) 400 Ibs./sq. inch for 5 mins. Oil exuded (as % of weight taken) 800 Ibs./sq. inch for 5 mins. TOTAL Ceylon ordinary (fine) 30.18 (%F.f.a. = 0 . 4 2 ) 6.47 (%F.f.a. = 1.03) 36.65 Philippine ordinary (fine) 27.15 (%F.f.a. = 0.21) 4.68 (%F.f.a. = 0.84) 31.83 The figures though not conclusive would appear to be in conformity with the theory that the higher exudation of oil from Ceylon D.C., is to be linked with some sort of cell-wall breakdown or weakening during sterilization and or drying. The experiments are being continued. (b) As the hydraulic press method was found unsuitable for testing over low pressure ranges, this method was subsequently abandoned. Instead, a more sensitive and satisfactory plate-method has been evolved for testing oil exudation from desiccated coconut. The method consists in spreading out evenly an accurately weighed quantity (5 grms.) of desiccated coconut between two Whatman No. 3, thick, absorbent quality, 15 cm. filter papers with cellophane backing. In order to ensure uniformity in particle size, only that fraction of the sample that passes through a B.S. (10 mesh) sieve is weighed out initially for the test. The filter papers (with the sample) arc then sandwiched in the horizontal position between two heavy metal plates (approximately 7" square) and subjected to incremental pressures by loading the top plate with standard weights. The weight of the top plate is of course taken into account in computing the applied pressure. After subjecting the sample to a specific pressure for exactly one hour, the desiccated coconut is carefully scraped off and the filter papers weighed. The increase in weight expressed as a percentage of the dry weight of desiccated coconut taken represents the oil exudation at each specific pressure. The observation has been made that the moisture content of the samples is one of the factors that could vitiate the results. Apart from dehydrating all samples for test to bone-dryness in a vacuum desiccator, it has also been found that incremental pressures for the lower range 10-5-40 constitute a more sensitive index for the comparison of sample differences than the higher range 20-10-60 which was used earlier in these tests. The results of 5 bone-dry samples tested in this manner over the lower pressure range are charted in Table III (a). 808O—? 13 TABLE III (a) Oil Exudation from Desiccated Coconut (Pressed between IS cm. filter papers) 1 2 3 SAMPLE WEIGHT PRESSURE (oil exuded as % of bone-dry weight) WEIGHT (lbs./27.4 sq") (ibs.isq tor 1 hour) / II III IV V Mean C.V. (CU) (CW) (CP) (CD (Ci) (%) Plate only = 10 (P) 0.36 13.86 14.00 14.65 16.55 17.37 15.29 9.26 P + 5 = 15 0.55 11.68 10.27 8.35 11.07 7.68 9.81 15.77 P + 10 = 20 0.73 8.27 7.49 6.00 6.68 4.70 6.63 18.54 P + 15 = 25 0.91 5.69 5.77 4.67 4.48 3.00 4.72 21.32 P + 20 = 30 1.09 4.81 4.06 3.80 3.84 2.13 3.73 23.53 P + 25 = 35 1.28 3.09 2.82 2.44 3.08 1.74 2.63 19.23 P + 30 = 40 1.46 2.55 2.88 2.40 2.44 1.25 2.30 24.06 Total Oil Exuded (%) 49.95 47.29 42.31 48.14 37.87 45.11 9.79 The figures in column 3 show that the coefficient of variation between the five samples (when the load is 10 lbs. for 1 hour) is 9.3%. For the same five bone-dry samples the following figures have been recorded for an initial load of 20 lbs., when applied for an hour. The data indicate the relative sensitivities of the samples to the lower and higher pressures rsepectively. Samples Oil Exuded (subjected to a load of 20 lbs.) (as % of dry wt.) CU CW CP CT 16.73 16.31 16.06 16.69 20.26 Coefficient of Variation (%) 8.98 Coefficient of Variation for CU, CW, CP and CT = 1.51%, whereas for the same four samples at the lower pressure it works out to 7.27%. It will be seen that the inclusion of sample (C) blurs the variation factor, as it is exceptionally high from the point of view of oil exudation. As those experiments are being continued the figures in Table III (b) have been computed to show the variations over the different pressure ranges for comparison with future samples. 14 This plate-method of testing oil exudation is being recommended for use in the routine examination of desiccated coconut. It is being currently employed in the D.C., quality testing laboratory of the Ceylon Coconut Board. TABLE m (6) Oil Exudation from Desiccated Coconut—(Pressed between IS cm. filter papers) 1 2 WEIGHT (lbs.121 A sq") SAMPLE (oil exuded as % of bone-dry weight) WEIGHT (lbs.121 A sq") I (CU) II (CW) III (CP) IV (CT) V (Cx) Mean C.V. (7.) (A) Plate only (P) = 10 13.86 14.00 14.65 16.55 17.37 15.29 9.26 (B) „ ^ , 1 1 hour each P + 5 J 25.54 24.27 23.00 27.62 25.05 25.10 6.08 (Q p P 1 1 0 } l h o u r e a c h 33.81 31.76 29.00 34.30 29.75 31.72 6.66 (D) C — plus (P + 15) — 1 hour 39.50 37.53 33.67 38.78 32.75 36.45 7.49 (E) D — plus (P + 20) — 1 hour 44.31 41.59 37.47 42.62 34.88 40.17 8.65 (F) E — plus (P + 25) — 1 hour 47.40 44.41 39.91 45.70 36.62 42.81 9.27 (G) F —plus (P + 30) — 1 hour 49.95 47.29 42.31 | 48.14 37.87 45.11 9.79 15 3. GRATED COCONUT MEAT Composition of D.C. Sample (ICP/H) % Moisture — 3.89 % Oil (wet basis) — 67.24 % Oil (dry basis) — 69.96 % Dry oil-free meal — 30.04 16 The grated meat from the fresh coconut kernel is used in the Ceylon household for culinary purposes. Further, it has been estimated (Central Bank of Ceylon) that the nut equivalent of the annual domestic consumption of coconuts is in the region of 990 millions. For cooking purposes, it is customary to press out manually the "milk" from the grated meat using a certain amount of hot water. With a view to obtaining an estimate of the efficiency of this extraction —(particularly as regards oil and protein), a simple experiment simulating the household method was carried out. The samples of grated meat obtained during three successive manual pressings were then analysed chemically. The complete results obtained in the study arc charted in Tables IV (a) to IV (c). The results clearly show that there is a colossal waste of valuable oil and protein in the "pol-kudu" that is discarded as waste. In terms of weights—on the basis that 1,200 nuts give 560 lbs. (a candy) of copra (6% moisture) with an oil content of 68.0% (dry basis), it has been computed that the loss of oil represents 23.1 % of the oil per nut. The corresponding total weight loss on 990 million nuts consumed locally would thus amount to approximately 30,407. tons per annum. The economic loss to the industry (without taking into account the protein) would be in the region of Rs. 33.5 millions. [Calculated on the basis that the value of a ton of coconut oil is Rs. 1,100]. The problem of averting these losses would hinge on the possible preparation of some form of emulsified coconut oil as a suitable substitute for coconut milk. Such a product would of course have to be sufficiently cheap for regular household use. 4. ESTIMATION OF CRUDE FIBRE A.G. Norman (J. Agric. Sci., 1935, 25, pp. 529-40) has shown that crude fibre is a variable mixture of cellulose and lignin, neither of which is a constant fraction of the quantity present in the original sample. During the acid digestion—starch is hydrolysed, but cellulose and lignin are very slightly changed. During the digestion with alkali, cellulose is slightly changed, lignin is attacked to a considerable but variable extent, and the proteins are almost completely removed. Norman thinks that any empirical method should include the whole of the lignin, and has suggested as a possible alternative, acid hydrolysis with a correction for the proteins. As it is proposed to carry out a series of crude-fibre estimations on desiccated coconut and other products of the coconut kernel, a few preliminary experiments were carried out to test the reproducibility of results obtained with the above empirical method. Norman's suggestion about acid hydrolysis with a correction for proteins was tried out, but was found to give inconsistent values. A sample of desiccated coconut with the analytical composition given below, was subjected to different treatments and then examined for crude fibre with the results charted in Table V. CHEMICAL EXAMINATION OF GRATED COCONUT MEAT AFTER SUCCESSIVE MANUAL PRESSINGS USING HOT WATER 1 . 2 3 4 5 6 7 Sample % Moisture % Oil % Total Sugars (as Sucrose) % Protein Total Mineral Matter % Dry Weight Inter-rela­ tionships (Succes­ sive Pressings) Sample % Moisture Wet basis Dry basis As% of tot. in fresh meat Wet basis Dry basis As% of tot. in fresh meat Wet basis Dry basis As% of tot. in fresh meat Wet basis Dry basis As% of tot. in fresh meat Dry Weight Inter-rela­ tionships (Succes­ sive Pressings) GK (fresh) 44.7 36.1 65.4 100.0 3.9 7.1 100.0 4.4 8.0 100.0 1.9 3.4 100.0 100.0 GK (calculated on air- dry basis) 5.7 61.7 65.4 100.0 6.7 7.1 100.0 7.5 8.0 100.0 3.2 3.4 100.0 100.0 GKI 5.7 56.2 59.6 46.3 6.3 6.7 47.9 7.4 7.9 50.0 4.5 4.8 71.8 50.8 GKII 5.7 50.4 53.4 30.1 5.4 5.7 29.6 6.5 6.9 31.9 3.5 3.7 40.3 36.9 GK1II 5.7 46.7 49.5 23.9 3.8 4.0 17.7 5.6 5.9 23.2 1.3 1.4 13.0 31.6 GK — Original freshly grated meat. 1 GKI — Grated meat after FIRST manual expression with hot water. GK II — Grated meat after SECOND manual expression with hot water. GK HI — Grated meat after THIRD manual expression with hot water. Note:—Samples GK I, GK II and GK III were air-dried before analyses. (Original grated sample prepared from five dead-ripe FALLEN NUTS) TABLE IV (a) (Moisture, Oil, Sugars, Protein and Total Mineral Matter) CHEMICAL EXAMINATION OF GRATED COCONUT MEAT AFTER SUCCESSIVE MANUAL PRESSINGS USING HOT WATER (Original grated sample prepared from five dead-ripe FALLEN NUTS) TABLE IV (b) MINERAL ELEMENTS—On oil-free meal (Per cent) 1 2 3 4 5 6 7 8 Sample % Moisture K P Ca N Dry oil- free meal Inter-rela­ tionships (Per cent) Sample % Moisture Wet Dry As% of tot. in Fresh Nut Wet i Dry As% of tot. in Fresh Nut Wet Dry As% of tot. in Fresh Nut Wet Dry As% of tot. in Fresh Nut Wet Dry As% of tot. in Fresh Nut Dry oil- free meal Inter-rela­ tionships (Per cent) IGK 4.8 1.S6 1.64 100.0 0.59 0.62 100.0 0.07 0.07 100.0 0.41 0.43 100.0 3.53 3.71 100.0 100.0 IGKI 4.8 1.15 1.21 43.7 0.50 0.52 50.0 0.07 0.07 58.6 0.37 0.39 53.7 2.96 3.11 49.6 59.2 IGKO 4.8 1.00 1.05 31.8 0.35 0.37 29.0 0.06 0.06 42.9 0.34 0.36 41.6 2.27 2.38 31.8 49.7 IGKm 4.8 0.57 0.60 16.9 0.24 0.25 19.4 0.05 0.05 32.9 0.27 0.28 30.0 1.78 1.87 23.3 46.2 IGK — Original freshly grated meat. I GK I — Grated meat after FIRST manual expression with hot water. I GK II — Grated meat after SECOND manual expression with hot water. IGKTH — Grated meat after THIRD manual expression with hot water. 4r CHEMICAL EXAMINATION OF GRATED COCONUT MEAT AFTER SUCCESSIVE MANUAL PRESSINGS USING HOT WATER TABLE IV (c) MINERAL ELEMENTS—Calculated on respective samples of Grated Meat (Mg. per 100 g.) 1 2 3 4 5 6 7 Sample % Moisture K P Ca Mg N Sample % Moisture Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry I GK (fresh) 44.7 314 568 118 213 13 24 82 149 710 1284 I GK (calculated on air-dry basis) 5.7 536 568 201 213 23 24 141 149 1211 1284 I G K I 5.7 461 489 198 210 26 28 149 158 1184 1256 I GK n 5.7 461 489 162 172 26 28 158 j 168 1046 1109 IGKIII 5.7 287 304 117 124 24 25 133 j 141 890 944 I GK — Original freshly grated meat. I GK I — Grated meat after FIRST manual expression with hot water. I GK II — Grated meat after SECOND manual expression with hot water. I GK III — Grated meat after THIRD manual expression with hot water. (Original grated sample prepared from five dead-ripe FALLEN NUTS) CHEMICAL EXAMINATION OF GRATED COCONUT MEAT AFTER SUCCESSIVE MANUAL PRESSINGS USING HOT WATER (Original grated sample prepared from five dead-ripe FALLEN NUTS) TABLE IV (d) (Weight Characteristics of Components of De-husked Nuts used for Experiment) 1 2 3 4 5 Afut rVn Weight of Nut Weight of Grated Weight of Nut- Weight of Shell (by llUt J1V, (grms.) Meat (grms.) Water (grms.) difference) (grms.) 1 1065 457 288 320 2 872 439 168 265 3 1036 417 308 311 4 712 390 114 208 5 734 321 185 228 Mean 884 405 213 266 TABLE IV (e) (Moisture and Dry Matter in Grated Meat during Successive Pressings) 1 2 3 4 5 6 Sample % Moisture in the samples as prepared for the Experiment Wet and Dry Weights of the portion used for determination of the losses in Dry Matter during successive Pressings Wet Weights of Portion recalcu­ lated on 44 .7% moisture basis Wet and Dry Weights obtain­ able from 1000grms. of fresh meat (44.7 %) (moisture) Dry Weights as"/0of original from 1000grms. i Wet Weight Dry Weight Wet Weight Dry Weight GK (fresh) GK I GKII . . GKIII . . 44.7 57.9 66.1 66.9 832 555 502 ' 442 460 234 170 146 832 423 307 264 1000 508 369 317 553 281 204 175 100.0 50.8 36.9 31.6 TABLE V Crude Fibre Estimations on Desiccated Coconut (Sample ICP/H) 1 2 3 4 5 6 No. No. of Extractions No. of Disintegra­ tions Acid Alkali % FIBRE (Original D.C.) (Pet. Ether) No. of Disintegra­ tions Digestion Digestion Wet basis Dry basis 1 2 3 4 5 6 one one two two three three nil nil one one two two yes yes yes yes yes yes yes yes yes yes yes yes 12.53 12.61 3.76 3.98 3.61 3.27 13.04 13.12 3.91 4.14 3.76 3.40 7 On oil-free meal obtained from routine oil determination 3.75 3.90 The figures clearly show that the particle size of the material is a critical factor in the values obtained for crude fibre. With good duplicate agreement the unground material obviously gives spuriously high figures. Between the samples ground once and twice, the values are not so widely divergent, though definitely lower with increased diminution of particle size. In the proposed series of fibre estimations the procedure to be followed would be to use the dry oil-free material after passing through a pre-determined standard mesh test sieve. 5. VINEGAR (a) The Vinegar Generator installed at the new factory at Nainamadama which was started off in June 1962 continued to operate very satisfactorily during the year. The factory was inspected at regular intervals and the acid strength of the finished vinegar was found to be consistently over 5.5%. The maize cobs originally used in the packing chamber were found to last the guaranteed period of one year without any form of degradation or breakdown. (b) A satisfactory method was evolved for the preparation of caramel (of high tinctorial efficiency), for colouring vinegar. A bottle of this caramel has been found adequate to colour 20 gallons of vinegar to a dark wine-tint. This method has been demonstrated to some vinegar-makers and it is now being applied at the new Nainamadama factory for colouring their "Generator Brand" Vinegar. (c) An experiment was successfully carried out to test the efficacy of common salt in retarding the growth of mycodermal film in bottled vinegar. It was found that a minimum concentration of 2.0% salt dissolved in the vinegar would completely arrest the formation of any mycodermal suspensions. 21 As an effective substitute for sterilization, common salt at a concentration of 2% is now being used at the Nainamadama factory. 6. POT CULTURE EXPERIMENT (a) A fifth Sand Pot Culture Experiment on 288 seedlings was laid down during the year. It is proposed to carry out preliminary tests on the relative merits and demerits of using leaf-punch samples (side by side with the entire laminae) for diagnostic analyses. It is proposed to restrict leaf punch analyses in this experiment to K and N only. (6) Work on this experiment was focussed principally on the analyses of plant samples from previous Pot Culture Experiments. 7. MISCELLANEOUS WORK (a) 32 samples of copra received from the Botanist were analysed for moisture and oil contents during the year. (b) Analyses and reports were made during the year on various samples of copra, desiccated coconut, poonac, sediment poonac and acetifying toddy. (c) A sample of farmyard manure received from Marawila was analysed and reported' on for its manurial value in respect of nitrogen, phosphorus, potassium, calcium and magnesium. W.R.N. NATHANAEL, Chemist, Coconut Research Institute. 22 REPORT OF THE BOTANIST 1. PROGENY TRIALS The three progeny trials at Marandawila, Walpita and Bandirippuwa were maintained during the year. The Walpita trial takes the form of two Cubic Lattices (A and B), and unselected, open- pollinated progenies of 125 unselected seed parents are tested in each Lattice. The analysis of variance of yield of nuts and weight of husked-nuts for two years of Lattice A (age of palms—12 and 13 years) showed that the differences between treatments in yield of nuts were not significant, but the differences in weight of husked-nuts were sigiuficant (P<0.05). The adjusted mean weight of husked-nuts per palm per year varied between 102.2 lb. and 41.8 lb. with a population mean of 71.9 lb., and that indicates the relative variation in breeding values between the seed parents. From this analysis it was possible to classify the 125 seed parents according to their breeding values, and use some of them for further breeding work. The offspring of the two best seed-parents 55 and 237 have given a mean yield of copra more than 40 per cent of the population mean, and further the coefficients of variation within progenies have been relatively low. These two palms could be classified as prepotent provisionally—provisionally as the progenies are only 13 years old and have not reached the optimum bearing age yet. Pollen from palms 55 and 237 were collected regularly for controlled pollination work and 25 of the best palms of the Marandawila Trial, classified according to their breeding values were selected for pollination work. 2. HYBRIDISATION BETWEEN VARIETIES AND FORMS The programme of work on breeding and selection of coconut palms was continued during the year. Nuts of crosses done during the previous year, viz. diallel, testing male transmitters, study of segregation of certain characters, etc., were harvested and some of them have been already planted in the nursery. 453 nuts have been harvested from one diallel cross involving 8 typica palms: two of the palms have given relatively few nuts when used as the female parent, but the same palms when used a$ the pollen parent have given a sufficiently large number of nuts. 536 nuts were collected from the crosses to test male transmitters: the pollen parents being 6 selected palms from the Isolated Seed Garden. Two types of cross typica X typica and pumila X typica were done. Altogether 375 seedlings were planted at Pothukulama in November, 1963. The king-coconut palms (var. aurantiaca, form thembili) grown in Ceylon show a marked periodicity in bearing, which character is presumably associated with the predominantly inbreeding system of the palm, similar to the dwarf (nana) variety. A programme of work was undertaken to study whether a uniformly high-yielding king-coconut palm without periodicity in bearing, and if possible early-bearing, could be developed by crossing selected palms within and between forms. A number of progenies of various crosses planted since 1950 are under observation and some of their yield data for 1963 are given below. The 22 progenies of typica X pumila have given a mean yield 23 of 166 nuts and a husked-nut weight of 297.9 lb. per palm, and 7 progenies of pumila X typica have given 162 nuts and a husked-nut weight of 312.4 lb. per palm. 10 progenies of typica X typica have given 182 nuts and a husked-nut weight of 322.0 Ib. per palm. Although these F, palms are only 13 years old, their yields are unusually high. Since the number of progenies of each cross is relatively small conclusive results cannot be drawn from them. Plantings done subsequently, particularly the experi­ mental block planted at Pothukulama in 1962 should give a better picture of the relative differences in yield between some of the above types of palms. Crossing of typica X typica and typica X pumila where the male parent is a prepotent was continued at six stations during the year. 25,171 nuts were harvested from crosses done in 1962. 16,028 seedlings of the above crosses were issued to the industry. Pollen Bank:—A technique to increase the quantity of pollen collected from an inflorescence was devised, based on a modification of a method developed by Jamaican workers. Anthers of coconut flowers were separated from the perianth lobes and Were oven dried at 40°C for 4 to 5 hours. Thereafter pollen was separated from the dehisced anthers by sieving. Pollen grains showed a germination of about 60 per cent, agamst 80 per cent for the control i.e., pollen collected from spikelets in the laboratory under atmospheric conditions. The drop in germination is compensated by the large increase in pollen collected. It was possible to collect 9.0 gm. of pollen from a prepotent palm using this new technique, whereas previously the average collection from the same palm was 1.0 gm. of pollen. The pollen bank was well patronised by 8 private estates and 3.856 samples of pollen have been issued to them, each sample being sufficient to pollinate three inflorescences once. The influence of pollen parent on endosperm weight:—An experiment was carried out from 1961 to 1963 to study the effect, if any, of the pollen parent on some nut characters like weight and volume of husked-nut and particularly thickness and weight of endosperm. The data of this experiment will be published in due course. 3. EMBRYO CULTURE Preliminary studies on cuituring coconut embiyos were undertaken and when a successful technique has been developed, it is proposed to study whether the embryos of certain genetical combinations leading to heterosis develop faster than those not exhibiting such a phenomenon and related problems. Coconut embryos were dissected from the endosperm, surface sterilized with ethyl alchol and grown on White's nutrient medium, supplemented with coconut water. A few embryos were successfully estab­ lished in culture, but their growth retarded after 4 weeks. Absence of a culture room affected the progress of these studies. 4. CYTOLOGY The somatic chromosome number of 14 pairs of twins of coconut seedlings were determined largely to find out whether haploidy is associated with twins. All the plants examined had the usual chromosome number 2n = 32, and haploids were not detected. Repeated attempts to induce polyploidy in coconuts either by colchicine or alternate heat and cold treatments have not been successful. A pasture grass grown under coconuts and commonly referred to as Brachiaria brizantha has a chromosome complement of 2n = 36, whereas the published figure for the species is 2n = 54. 24 5. PLANTING TECHNIQUES Hedge planting:—The observation trials of a new system of close planting for selective thinning are being continued and it is yet too premature to analyse the yield data as the palms are relatively young. About S acres at the Seed Garden were planted in May 1963 with a planting distance of 32 feet between rows and 12 feet within the row. Depth of planting:—The field trial where four depths of planting a coconut seedling—6, 12, 18 and 24 inches—are compared has completed 7 years. The design is randomised block lay-out with 9 palms per plot and 5 replications. The seedlings were manured regularly from the date of transplantation. At the end of the year, palms in flower for each treatment were as follows: 6 inches depth—82.2 per cent, 12 inches depth—64.4 per cent, 18 inches depth—82.2 per cent and 24 inches depth—SO.0 percent. Th? differences between treatments were not significant. Thus, planting seedlings at the four depths indicated above has not made any difference with respect to their flowering-age, although it is commonly believed that deep planting delays flowering. Size of seed hole:—-This experiment planted in November, 1955, has shown that the size of seed-hole cut to transplant a coconut seedling has no effect on the initial flowering of the palms (vide. Annual Report for 1961). The yield data collected in 1962 and 1963 showed that the difference between the four treatments with respect to yield of nuts and husked-nut weights were not significant. The relative yield data are indicated below yield; per palm includes non-bearing palms as well. The soil in the block is mostly light gravel with a sandy loam in a smaller area. It is apparent that under the conditions of the experiment, the size of seed hole prepared to transplant a coconut seedling is immaterial, provided the plants are manured regularly and the land is cultivated. Treatment 1962 Yield per palm No. of nuts Wt. of husked nuts (lb.) 1963 Yield per palm No. of nuts Wt. of husked nuts (Ib.) 1 X 1 x 1ft. seed hole 17 31.8 23 39.0 3 X 3 X 3ft. seed hole 23 36.4 28 41.4 Cruciform type 23 41.2 28 45.4 Post-hole borer type 17 31.3 24 41.3 6. SEED SUPPLY SCHEME 1,880,604 selected seed-nuts from selected palms were supplied to the Planting Division during the year. Six new estates were examined and four of them were selected as suitable for seed collection. 49,897 palms have been selected for seed collection by the end of the year. 7. MISCELLANEOUS Field Day:—Field days were held at the Isolated Seed Garden and at Rathmalagara Estate on two separate days for members of the L.C.P.A. and Kurunegala P.A. Publications:—The following paper was submitted for publication in the Ceylon Coconut Quarterly. Methods of under planting in senile coconut plantations by D.V. Liyanage. Staff:—Mr. M.A.P. Manthriratne, Snr. Technical Assistant was promoted Research Assistant from 16th March, 1963. 25 D.V. LIYANAGE, Botanist, Coconut Research Institute. REPORT OF THE SOIL CHEMIST SUMMARY Seven long-term field experiments on manurial problems were continued in 1963—in addition to the observation trials on magnesium deficiency and leaf scorch problems. Pre-treatment records were maintained during the year for two new experiments—one on nitrogen quality, and the other on the problem of round-the-year immature nut fall. The results of field experiments on bearing palms at Bandirippuwa and Ratmalagara give further support to our current recommendations on N.P.K. manuring (in 1960 we recommended an increase in the levels of N.P.K. manuring to about double the rates previously recommended). The experiment comparing organic with inorganic fertilizers on a light sandy loam at Marandawila Estate, Bingiriya has shown that organic fertilizers offer no advantage over the cheaper inorganic ferti­ lizers. We still receive occasional queries as to whether it would not be more advantageous to use organic sources of nitrogen and phosphorus in our fertilizer mixtures. This experiment confirms our view that it would be economically disadvantages to do so. In the experiment on methods of fertilizer placement at Walahapitiya Estate, Nattandiya, there was little difference between the surface application of fertilizers in a circular strip 3 feet wide and 3 feet away from the palm and the considerably more expensive traditional circular trenches. Broadcast applications proved to be less effective. A second field trial on adult palms using radioactive phosphorus confirmed that placement in the entire area round the palm up to a distance of about 6 feet from the bole was more efficient than placement in a 3 feet wide circular s'rip 3 feet away from the palm, or placement in the centres of squares. Chemical analysis leaflet samples from palms affected by "leaf scorch" in the southern province has shown that these palms are deficient in boron and magnesium. The possible association of the problem of leaf scorch with nutrient deficiencies cannot still be over-ruled, and further trials on this aspect are in progress. Laboratory studies have indicated that except on sandy soils, dolomite can be applied to palms at the same time as the usual N.P.K. fertilizer mixtures without loss of nitrogen provided that the fertilizers are well incorporated into the soil. A. FIELD EXPERIMENTS 1. 3 x 3 X 3 N.P.K. Factorial Experiment (Bandirippuwa Estate) (See C.R.I. Annual Report 1949 and 1951 for details of treatment and design). The 28th year of this experiment was completed in 1963. The yields for 1963 are the highest recorded since 1955. The main effects of N, P and K for 1963 are given in Table I. Potash continues to show a highly significant effect, while phosphate has given a significant response for the second year in succession. 26 Potash at 2.25 lbs. KgO/palm biennially (4$ lbs. muriate of potash) gave an increase of 102 lbs. copra per acre over the lower level of 1.5 lbs. K 2 0 . The response to phosphate at 1 lb. PaCWpalm biennially was 181 lbs. copra/acre over the control. None of the interactions showed significance. It is interesting to note that although these palms are now well over 60 years old, they still continue to give high yields- over 4,400 nuts/acre when fertilizers are applied. TABLE I Yield data (main effects) for 1963—33 N.P.K. Experiment, Bandirippuwa (66 palms/acre) Treatment Lbs. Copra per acre Nutsl acre y Copra Nuts Difference per acre Copra Nuts lbs. Copra out-turn nutsl candy Diffe­ rence in nutsl candy No (0.0 lbs. N) 2,033 4,087 100 100 _ 1,126 Ni(0.51bs.N) 2,071 4,277 102 105 + 38 + 190 1,156 + 30 N 2 (1.01bs.N) 1,988 4,205 • 98 103 - 45 + 118 1,185 + 59 Po(0.01bs.Pj0 5) 1,909 3,885 100 100 1,140 — Pi(1.01bs.P 2O 5) 2,090 , 4,305 109 111 + 181* + 420** 1,154 + 14 P 2 (2.01bs.P 2 O 6 ) 2,093 4,380 110 113 + 184* + 495** 1,172 + 32 Ki (0.75 lbs. KaO) 1,770 3,839 100 100 1,215 — K 2 ( l . 51bs.K 2 0) 2,110 4,295 119 112 + 340***+456*** 1,140 -75 K 3 (2 .25 lbs. K 2 0 ) 2,212 4,435 125 116 + 442*** +596**° 1,123 -92 Significant difference at P. 05 = 154 lbs. copra/acre and 278 nuts/acre. * — Significant at P. 05. **— Significant at P. 01. *** — Significant at P. 001. 2. Manurial X Cultivation Experiment (Ratmalagara Estate) The 20th year of this experiment was concluded in June 1963. Here too the yields for the year 1962-63 are the highest recorded since 1955. Both phosphate and potash showed significant effects (Table II)—phosphate at 1.0 lbs. PtOs per palm biennially giving nearly 800 lbs. copra per acre more than the control plots. With phosphate applications a yield of over 4,200 nuts per acre was obtained. None of the interactions showed statistical significance. 27 TABLE n Yield data for 1962-1963, Manurial X Cultivation Experiment, Ratmalagara (60 palms/acre) Difference Copra Diffe- _ ~ Nut si % per acre out-turn rence Treatment Copra ' .. , _ r .. 4 t . . t . r acre Copra Nuts Copra Nuts Nuts/ innutsj per acre r , , r lbs. candy candy P 0(0.01bs.PiOi) 1,404 2,634 — — 1,051 — P (1.01bs.P £O 5) 2,198 4,244 + 794***+ 610*** 1,107 + 56 C 0 (no ploughing) 1,749 3,391 — — 1,085 — C (ploughing) 1,853 3,590 + 104 +199 1,085 0 C (ploughing) Significant difference P. 05 = 142 lbs. copra/acre and 269 nuts/acre. K 0 (0.0 lbs. K 2 0 ) 1,673 3,339 — — 1,117 — K , (1.0 lbs. K 2 0 ) 1,843 3,565 + 170 +226 1,083 -34 K 2 (2.0 lbs. K 2 0 ) 1,886 3,555 + 213* +216 1,058 -59 Significant difference .at P . 05 = 174 lbs. copra/acre and 330 nuts/acre. * — Significant at P . 05. *** — Significant at P . 001. 3. 3 X 3 X 3 N .P .K . Experiment on Young Palms (Ratmalagara Estate) (See C.R.I. Annual Report 1959 for details of treatment and design). This experiment was planted in 1948. During the early period (up to the flowering stage) only nitrogen and phosphorus gave significant responses. The effect of potash reached significance at bearing stage. The yield data for the main effects of N , P and K given in Table III (a) once again show that all three major nutrients have produced highly significant increases in crops. The analysis of variance has also indicated a highly significant N P interaction ( P . 001), while the N K and P K interactions were also found to be significant at the 5% probability level. The N P , N K and P K interactions are seen in the two-way tables —Table III (b). The response to potash which is linear increases with increasing levels of nitrogen—the combination giving highest yields being K 2 N 2 (3 lbs. each of muriate of potash and ammonium sulphate per palm annually). At lower levels of potash, the higher nitrogen dosage ( N 2 ) depresses yields. This confirms the earlier results obtained in the 3 3 N . P . K . experiment at Bandirippuwa. There is also a positive N P interaction—the combination N 2 P 2 (3 lbs. ammonium sulphate, 3 lbs. Saphos phosphate) giving the highest yields. The combinations N 2 P 2 K 2 (equivalent to 9 lbs. C.R.I, mixture "A") has given 4,778 nuts per acre, whereas the control plots have given only 2,474 nuts per acre, and the combination N J P J K J (4i lbs. C.R.I, mixture "A") gave 3,768 nuts per acre. 28 TABLE m (a) Yield data for 1963—3s N.P.K. Experiment on Young Palms, Ratmalagara Estate (55 palms/acre) Treatment Lbs. Copra per acre % Diffe­ rence i r , . ,, Nuts in lbs. Copra] acre % Diffe- rence in nutsl acre Copra out-turn Nutsl candy Diffe- rence in nutsl candy N 0 1,523 100 — 2,768 100 _ 1,018 N, 1,860 122 + 337*"* 3,358 121 + 590 1,011 - 7 N 2 1,827 120 + 304*** 3,456 125 + 688 1,059 + 41 Po 1,188 100 — 2,147 100 1,012 P, 2,022 170 + 834*** 3,666 171 + 1,519 1,015 + 3 P 2 2,000 168 + 812*** 3,769 176 + 1,622 1,055 + 43 K 0 1,582 100 — 3,042 100 1,077 K x 1,695 107 + 113 3,099 102 + 57 1,024 - 53 K 2 1,932 122 + 350***3,442 113 + 343 997 - 80 Significant difference at P. 05 = 132.6 lbs. copra/acre. Significant at P. 001. TABLE HI (b) lbs. Copra per acre * 0 tf2 Po 1,246 1,705 1,617 Pi 1,231 2,220 2,127 p 2 1,085 2,140 2,256 N0 JV2 K 0 1,493 1,692 1,561 Kx 1,423 1,874 1,787 K 2 1,650 2,013 2,132 ^ 0 K 0 1,105 1,773 1,868 K, 1,072 1,975 2,038 K 2 1,385 2,317 2,094 29 TABLE m (c) lbs. Copra per acre " N0K0 • N0Kt • - P 0 1,355 1,152 1,230 Pj 1,366 1,655 2,093 P 2 1,759 1,464 1,627 N,K2 Po 952 1,163 1,580 Fi 2,103 2,172 2,387 P 2 2,022 • 2,287 2,072 N2K0 N2K, N2K2 P 0 1,008 902 1,345 Pj 1,852 2,098 . • 2 » 4 . 7 . 1 — P 2 1,823 2,362 2,582 4. Manurial Experiment on Organics vs. Inorganics and frequency of manuring (Co-operative Experiment at Marandawila Group, Bingiriya). (See C.R.I. Annual-Report 1959 for details of treatment and design). The 5th year of this experiment was concluded in June 1963. Table IV gives the yield data (adjusted by co-variance analysis) for the different treatments during the year 1962-1963. This year too all treat­ ments were found to be significantly better than the control (no manure). But none of the other treatment differences showed statistical significance. For the 5th year in succession it has been shown that organic manures offer no advantage over the cheaper inorganic fertilizer even on the sandy loam soil. Biennial application of fertilizers in full circles round the palm has been found to be as effective as annual application in half circles. It is possible that any- advantages of annual application was not realised as only half the root system was utilised for the uptake of fertilizers applied each year. The experimental treatments will therefore be altered to full circle application of fertilizers both annually and biennially. 30 TABLE IV Yield data for 1963, Experiment on Organics vs. Inorganics, Marandawila Estate (Copra yield adjusted by co-variance analysis) (60 palms/acre) Treatment lbs. Copralacre 7. Difference in lbs. Copralacre Out-turn nuts/candy Control . . 2,317 100 — 1,103 Inorganics annually 2,758 119 441 1,057 Inorganics in 2 years 2,752 119 435 992 Organics annually 2,758 119 441 1,003 Organics in 2 years 2,760 119 443 1,037 Cattle manure with supplements 2,850 123 533 1,012 5. 4 X 4 X 4 N.P.K. Experiment on adult palms (Bandirippuwa Estate) (See C.R.I. Annual Report 1960 for details of treatment and design). This experiment has now completed three years. The fourth annual manuring was done in November 1963. The copra yields for the 2nd year (1962) showed a significant positive response to nitrogen. In the third year significant positive responses were obtained for both nitrogen and potash. The yield data for 1963 (adjusted by co-variance analysis) is given in Table V. The phosphate response showed no significance, nor did any of the interactions. Both nitrogen and potash have given the highest yields at the higher levels of application (7£ lbs. sulphate of ammonia and 3 lbs. muriate of potash per palm respectively). TABLE V Yield data for -1963—43 N.P.K. Experiment on adult palms, Bandirippuwa (Copra yield adjusted by co-variance analysis) (66 palmsjacre) Treatment Ibs.Copral acre % Difference in lbs. Copra/acre Nuts/acre Copra out-turn in nuts/candy N 0 (0.0 lbs. N) 1,824 100 — 3,873 1,176 ^ ( 0 . 5 lbs. N) 2,013 110 + 189** 4,116 1,188 N 2 (1.0ibs'.N) 2,089 115 + 265** 4,451 1,192 N 3 (1 .51bs.N) 2,111 116 + 287** 4,679 1,214 P 0 (0.0 lbs. P A ) 1,986 100 _— 4,213 1,191 M 0 . 5 1 b s . P A ) 1,939 98 - 47 4,399 1,172 P 2 (1.0 lbs. P , 0 6 ) 1,961 99 - 25 1,467 1,198 P 3 (1.5 lbs. P A ) 2,149 108 + 163 4,342 1,212 K 0 (0.0 lbs. K 2 0 ) 1,829 100 — 4,085 1,210 K x (0.5 lbs. K 2 0) 2,015 110 +186** 4,224 1,190 K 2 (1.01bs.K 2 O) 2,070 113 + 241** 4,398 1,217 Kj (1.5 lbs. K 2 0 ) 2,122 116 + 293** 4,413 1,158 ••SignificantatP.Ol. Significant difference P. 05 — 138 lbs. Copra/acre. It is interesting to note that in the old 3 B N.P.K. trial at Bandirippuwa the potash response showed significance from the third year onwards, while the effect of nitrogen was statistically significant only in the second year. 31 http://M0.51bs.PA 6. Methods of placement and liming experiment on adult palms (Co-operative experiment at Walahapitiya Estate, Nattandiya). (See C.R.I. Annual Report 1961 for details of treatment and design). The third annual manuring was done in October 1963. During the course of the year the plots receiving lime were given a further applications of lime at the rate of 15 cwts. per acre, so that these have now been treated with a total of 45 cwts. lime per acre. The lime has reduced the acidity of the top soil (0-9") in these plots from pH4.2 to pH 6.8. But the sub-soil (9-18") still remains quite acid at pH 4.3. Liming produced no significant effects on yields at the end of the second year of this experiment. Further applications of lime will be done to reduce the acidity of the sub-soil as well. Broadcast application of fertilizers was less efficient than placement round the palm. Surface appli­ cation of fertilisers in a circular strip 3 feet wide and 3 feet away from the palm showed little difference to the considerably more expensive traditional circular trench system of placement. The yield data (adjusted by co-variance analysis) for 1963 is shown in Table VI. It is still premature to draw any definite conclusions. TABLE VI Yield data for 1963—Experiment on fertilizer placement and liming, Walahapitiya Estate (Yields adjusted by co-variance analysis) (50 palms/acre) Treatment lbs. Copra/acre V / o Difference lbs. Copra/acre O (no fertilisers) B (broadcast) C (Circular trenches) S (Spread on surface in circular strip) 1,406 1,428 1,570 1,528 100 101 111 109 22 164*** 118*** *** Significant at P. 001. Significant difference P. 05 = 52 lbs. Copra/acre. L„ (no lime) (lime) O (no fertilizers) F x (3£ lbs. N.P.K. mixture) F 2 (7 lbs. N.P.K. mixture) 1,474 1,490 1,406 1,498 1,518 100 101 100 106 108 16 92 112 7. 4 X 4 X 4 N. P. K. Experiment on young palms, Potliukulama Estate (See C.R.I. Annual Report 1961 for details of treatment and design). This experiment which was planted in December 1960 completed its second manurial year in 1963. As from 1963, the annual fertilizer dosage was applied in two split doses in June and December. At the end of 1963, four palms had flowered. To facilitate weed control, particularly illuk, a mixed legumi­ nous cover crop consisting of pueraria, centrosema and calapogonium was sown in 1962. By the end of 1963 the cover crop was well established. With the quick and healthy growth of the palnis, red weevil has become a problem. Four palms were lost in 1963 through red weevil attack. 32 A number of palms were attacked by the fungus disease, Helminthosporium. There was no asso­ ciation of the disease with any particular manurial treatment. Measurements of heights and girths of palms, and counts of total leaf number and new leaf production were kept at six-monthly intervals. Statistical analysis of these records showed that both nitrogen and potash had a positive significant effect in leaf production. Nitrogen had a significant effect in increasing the height and girth of palms. But, unlike in the experiment on young palms at Ratmalagara, phosphate has not had a response yet, although available phosphate in the soil is low. Whether this is due to poor availability of saphos phosphate under these soil conditions, or whether the chemical method of assessing available soil phosphate did not give a correct picture of the actual situation, has to be investigated. 8. Observation trials on yellowing palnis, Walgama Estate, Rukmale and Mattegoda Estate, Polgasowita. (See also C.R.I. Annual Reports for 1960, 1961 and 1962). These trials which were commenced in 1957 completed their sixth year in 1963. All yellow pain's which had been subject to applications of magnesium sulphate or dolomite have now been restored 10 & healthy green colour. As in the previous year, magnesium caused a marked increase in nut yields. At Walgama Estate affected palms receiving N.P.K. fertilizer only gave 20 nuts/palm in 1963, whereas those treated with magnesium as well gave 46 nuts/palm. Green palms which showed no visual symptoms of magnesium deficiency gave 46 nuts/palm when treated with N.P.K. fertilizers only. The addition of magnesium increased these yields to 62 nuts/palm. Magnesium deficiency is likely to limit nut production even before any visual symptoms of deficiency occur. We therefore recommend the application of dolomitic limestone as a long-term preventive measure against magnesium deficiency in certain areas even though no visual symptoms have been observed. 9. Leaf Scorch problem, Gonapinuwela area (See also C.R.I. Annual Reports for 1961 and 1962). During 1963, the studies on nutritional aspects of the problem of "Leaf Scorch" was confined to soil applications of boron since only the group of palms treated with borax showed no signs of further deterioration. Moreover, leaf analysis had indicated that palms in this area arc deficient in boron. But no conclusions can be drawn yet as to whether the problem is of nutritional origin. Another interesting observation was that the group of palms which had been previously subject to spraying with copper, iron and manganese sulphates showed temporary improvement. Whether this Was due to copper acting as a fungicide is a matter for further investigation, particularly since the Commonwealth Mycological Institute has reported that visual symptoms observed in connection with this problem are probably caused by weakly parasitic and saprophytic fungi attacking the weakened leaves. In 1963, about 250 diseased palms and 250 healthy palms were selected at Kirimetiya Estate for treatment with various combinations of major and minor elements, and also lime. The objects of this experiment arc to determine (a) whether by the application of any particular nutrient combination diseased palms can be restored to normal health, and (£>) whether healthy palms can be made less suscepti­ ble to disease by the application of any particular nutrient treatment. 33 B. LABORATORY INVESTIGATIONS 1. Radio-Isotopes (P 3 2 ) project—Fertilizer placement trials (See also C.RJ. Annual Reports for 1960, 1961 and 1962). A second field trial on placement using radioactive phosphorus was carried out on adult bearing palms (over 40 years old) grown on a sandy loam soil at Marandawila Estate, Bingiriya. The palms had been subject to regular broadcast application of fertilizers for the past eight years. The following methods of placement were compared:—(a) Basin round palms extending to a distance of 5 feet 8 inches from the bole of the palm, (b) traditional 3 feet wide circular trench, 3 feet away from the base of the palm, and (c) circles of 5 feet 6 inches radius in squares between palms (distance between palms was about 26 feet X 28 feet). The area in which radioactive phosphorus was applied (96 square feet) were similar for all three methods of placement. This ensured the elimination of any complications which may result due to isotopic exchange. Depth of application was 4 inches, or less if there was any danger of root damage when removing the surface soil. The experimental units were single palms, and each placement was replicated five times. To minimise errors due to individual palm variability and soil heterogeneity, the following precau­ tions were taken. Each replicate of the three different treatments were grouped together so as to form a block—palms receiving the different treatments being as close together as possible within each block. The experimental palms were separated from each other by guard palms. Palms of similar bearing capacity and growth were selected for treatment comparisons within each block. Each experimental palm was treated with a solution of 66 grammes dibasic ammonium phosphate labelled with 5 millicuries P 3 2 in 10 litres water. This solution was uniformly sprinkled in the entire area of placement. In the case of placement (c) at the centres of squares, the full phosphate dosage was given to each of the four circles in the squares surrounding the experimental palm since the fertilizer applied to each square would be equally accessible to four palms. Leaflets were sampled from the sixth frond of each experimental palm at intervals of one week, four weeks and thirteen weeks after the P 3 2 application. The leaflet samples were dry ashed and the phosphate precipitated and redissolved to eliminate interfering sources of radioactivity. P 3 2 was deter­ mined by the liquid counting technique, and total phosphate by the colorimetric vanado-molybdate method. The experimental results are summarised in Table I (6). The efficiency of the different methods of placement is judged by the amount of radioactivity taken up by the palms (measured in terms of specific activities). In the first week, there was no noticeable, difference between the placement. After four weeks, there is a clear trend in favour of placement in the entire area round the palm up to a distance of 5 feet 8 inches from the bole (placement A), while after thirteen weeks the superiority of placement (A) has increased to about 100% over placements (B) and (C). The variation between palms do not permit any distinction to be made between placements (B) and (C). These results confirm those obtained in the first experiment. (C.R.I. ANNUAL REPORT 1962). 34 TABLE I B Radio Isotopes Project—(P 3 2) Marandawila Estate, Placement trial on adult coconut palms (Soil type Loamy sand, Fertilizers broadcast annually for last 8 years). Specific activities (counts per minute/mg'. P) in 6th frond leaflet REPLICATE NOS. 1 2 3 4 5 Mean Time after P 3 2 Application (Weeks) Method of Placement A. Basin (5' 8* radius from bole outwards) B. Traditional Circular Trench (3' wide 3' away from bole) C. Centres of Squares 1 4 13 20.5 156 415 7.2 51.2 121 25.6 134 317 1 4 13 5.2 34.8 127 3.3 23.6 103 3.70 44.8 120 1 4 13 15.8 111 376 16.1 87.7 216 5.8 42.6 199 1 4 13 16.3 84.9 198 30.0 79.9 150 1.2 11.2 29.7 1 4 13 6.0 60.1 173 1.1 13.8 54.6 3.1 22.8 60.2 1 . 4 13 12.7 89.4 258 11.5 51.2 129 7.9 51.1 145 This study indicates that contrary to popular belief the density of absorbing roots is much higher in the area immediately surrounding the palm up to a distance of about 6 feet from the bole even under conditions which would be considered ideal for the development of a wide spread root system (light, well drained sandy loam soil and regular broadcast application of fertilizers). The extent to which more efficient placement would help to reduce costs of production is a matter for further investigation. 2. Chemical analysis of leaflets from "Leaf Scorch" problem palms Table II B (a) gives the means and ranges of chemical analysis of leaflets sampled from the first and sixth fully opened fronds of ten healthy and ten diseased palms at Kirimetiya Estate, Elpitiya. All samples were taken from healthy green fronds. We are grateful to the Institute de Researches pour les Huiles et Oleagineux, Paris, for this chemical analysis. Examination of the individual palm analysis using the rank-sum test of Wilcoxon revealed that the leaflets sampled from frond 1 of diseased palms had a lower content of magnesium and boron, and a higher content of nitrogen and phosphorus. In frond 6, the only difference observed was a lower content of manganese and magnesium in samples from diseased palms. The differences observed are however not large. Table II B (b) gives the mean chemical analysis of leaflets sampled from the 1st and 6th fronds of ten individual palms at Bandirippuwa Estate. It is seen that these have a generally higher content of magnesium and boron than the leaflets of both healthy and diseased palms at Kirimetiya Estate. Magnesium deficiency is known to occur widely in coconut palms grown on the lateritic soils of the heavy rainfall areas of the Southern and Western province, and under acute conditions is associated with an intense yellowing of leaves. In the Western province there are coconut estates which have been suffering from acute magnesium deficiency for over ten years, but the palms did not die, or exhibit any of the other symptoms of "Leaf Scorch". Although leaf analysis has shown diseased palms at Kiri­ metiya Estate to be more deficient in magnesium than the healthy areas, the palms exhibited no visual symptoms characteristic of magnesium deficiency (in other coconut lands in the Gonapinuwala area, both the usual symptoms of both magnesium deficiency and "Leaf Scorch" has been observed to occur simultaneously). Leaf analysis has shown that palms in the Western province (Mattegoda and Walgama Estates, where magnesium deficiency occurred) have a low boron content (less than 10 ppm. B)—but they showed no symptoms of leaf scorch. It appears therefore that although leaf analysis has indicated diseased palms to be deficient in magnesium and boron, the problem of leaf scorch cannot be the direct result of any nutrient deficiency-. The possibility that nutrient deficiency has weakened the palms and made them susceptible to attack by some parasitic organisms (e.g. nematode, fungus) cannot be over-ruled. Field, investigations on these lines are in progress. 36 TABLE n B (a) Chemical analysis of leaflets sampled from palms affected by "leaf scorch" and healthy palms at Kirimetiya Estate Means of samples taken from ten individual palms % ppm P K Ca Mg S Na B Mn Al Fe C M Zn Mo Frond 1 Healthy Mean Range 1.40 1.29- 1.50 0.136 0.124- 0.153 2.155 1.532- 2.357 0.231 0.184- 0.291 0.178 0.142- 0.213 0.135 0 .126- 0.150 0.226 0 .173- 0.281 12 8 .38- 15.47 87 55- 115 5.8 1.0- 14.0 56 5 1 - 61 10 5 .42- 11.97 42 27- 85 0.21 0 .18 - 0.30 Diseased Mean Range 1.61T 1.31- 2.12 0.152t 0.130- 0.167 2.235 1.633- 2.769 0.226 0.170- 0.297 0.151* 0.117- 0.190 0.146 0 .129- 0.183 0.217 0 .131- 0.277 11* 9 . lO- l l . 6 3 85 59- 126 14 1.5- 52.0 56 45- 70 9 7.OS- 10.76 51 30- 100 0.60 0 . 2 0 - 2.00 Frond 6 Healthy Mean Range 2.13 1.92- 2.24 0.151 1.34- 1.73 1.907 1.011- 2.34 0.273 0 .201- 0.449 0.205 0.155- 0.280 0.183 0.157- 0.213 0.344 0.234- 0.451 9 7 .92 - 10.93 118 69- 180 24 5 . 0 - 65.0 67 62- 75 8 7 .05- 9.93 29 2 3 . 0 - 40.0 0.22 0 . 2 0 - 0.30 Diseased Mean Range 1.95 1.87- 2.28 0.152 1.33- 1.69 2.069 1.458- 2.574 0.247 0.174- 0.350 0.174* 0.127- 0.254 0.173 0 .144- 0.206 0.292 0.137- 0.427 10 8 .57- 10.69 85* 52 - 139 9 1 .2- 24.0 62 47- 71 7 5 .69- 9.06 36 12 .0- 70.0 0.45 0 .20 - 1.90 fSignificantly higher than samples from healthy palms. •Significantly lower than samples from healthy palms. TABLE II B (6) Mean of samples taken from ten individual palms at Bandirippuwa Estate % ppm N : P K Ca Mg S Na B Mn Al Fe Cu Zn Mo' Frond 1 Mean Range 1.71 1.41- 1.91 0.155 0.127- 0.192 2.164 1.34- 2.80 0.165 0 .131- 0.229 0.218 0 .176- 0.288 — 16 14.2- 19.9 28 15.44 21 7.31 85 54-' 139 — — — Frond 6 Mean Range 2.35 2 .21 - 2.64 0.166 0.149- 0.192 1.22 0 .58- 2.10 0.230 0 .144- 0.374 0.277 0.166- 0.411 — 13.5 1 1 . 1 - 17.6 33 25- 57 — — — — 3. A study on the effect of frond age on leaf composition Leaflets were sampled from the 1st, 6th and 14th fully opened fronds of ten healthy adult palms randomly selected from a block at Bandirippuwa Estate. The chemical analysis of these samples for the five major elements.—N, P, K, Ca and Mg are given in Table III B. Frond 1 had a consistently lower N content, while K decreased with increasing age of frond. Ca and Mg increased with frond age, but P showed no significant variation. TABLE in B Chemical composition of leaflets in fronds of different age (% mean on samples from ten palms) N P K Ca Mg Frond No. 1 Mean . . . . 1.715 0.155 2.164 0.165 0.218 Std.Dev. . . 0.148 0.019 0.512 0.034 0.036 Frond No. 6 Mean . . . . 2.353 0.166 1.216 0.230 0.277 Std.Dev. . . 0.133 0.013 0.502 0.082 0.069 Frond No. 14 Mean . . 2.303 0.154 0.868 0.325 0.282 Std.Dev. . . 0.190 0.0176 0.460 0.052 0.075 4. Studies on the effects of crushed dolomite on fertilizer and soil nitrogen As a safety measure against possible nitrogen losses, the application of dolomite to coconut palms is done separately, at a later date after the usual N.P.K. fertilizer mixtures containing sulphate of ammonia has been applied. This practice involves extra expenditure. A simple laboratory experiment was therefore carried out to obtain some quantitative estimate of nitrogen losses when dolomite is mixed with fertilizers both in the presence and absence of soil. On mixing crushed dolomite with CRI general coconut fertilizer mixture (containing equal propor­ tions of ammonium sulphate, saphos phosphate and muriate of potash), the loss of nitrogen was negligible even after seven days provided that the mixture was kept dry. When exposed to the atmosphere under laboratory conditions, the mixture had absorbed about 6% moisture and lost 22% nitrogen after 7 days. The loss of nitrogen after one day was 8%. Under conditions of storage where the atmosphere was moisture saturated, the loss of nitrogen after one week was 60%, with a moisture absorption of 12%. The addition of dolomite to a damp mixture of a clay soil and N.P.K. fertiliser in proportions similar to field applications did not cause any noticeable loss of nitrogen even after 6 weeks. Under similar conditions, about 10% N was lost from a sandy soil, and 6% N lost from a sandy loam. Dolomite raised the soil pH from about 4.2 to 7 in the loam and clay soils, and from 5.5 to 7.4 in the sand. 39 In the clay soii, dolomite application had little effect on nitrification, but in the sand and sandy loam, nitrate N increased from 10 to 90 ppm. and 2 to 200 ppm. respectively. The results show that except on sandy soils, dolomite can be applied at the same time as the usual N.P.K. mixture provided that the fertilizers arc well incorporated into the soil. The effects of enhanced nitrification (resulting from lime application to acid soils) on the nitrogen response in coconut yields is a matter which merits further investigation. C. MISCELLANEOUS 1. Premanurial records on yields were maintained on individual adult palms on a block of about 30 acres at Mawatte estate, Dankotuwa, with a view to laying down a new field experiment to test the relative efficiency of different forms of nitrogenous fertilisers (urea, ammonium sulphate, nitrochalk). 2. Individual palm records on mature nuts, female flowers and immature fallen nuts were maintained on a block of about 1,000 palms on an estate at Bingiriya where round the year immature nutfall had been reported. We propose to carry out a simple field trial on this block to see whether the immature fall of nuts is associated with any nutrient deficiency. 3. (i) The following papers were presented at the Annual Sessions (1963) of the Ceylon Association for the Advancement of Science by members of the division:— (a) "Estimation of calcium and magnesium in coconut leaflets by the EDTA method" —by M.A.T. de Silva. (b) "Maintaining fertility on coconut land"—by D.A. Nethsinghe. (ii) Dr. D.A. Nethsinghe presented a paper on "Conserving Fertility on Coconut Lands" at a Symposium on the "Management of Resources for Conservation'and Production" held by the Soil Conservation Society of Ceylon. (iii) An article on "Manuring of Young Palms" by D.A. Nethsinghe was published in the Ceylon Coconut Planters' Review, Vol. Ill, No. 2. D. PERSONNEL Mr. T.S. Balakrishnamurthi, Research Assistant, left for U.K. on overseas study leave under a Colombo Plan scholarship in September 1963. He is reading for a post-graduate degree in Soil Science at the University of Aberdeen. Mr. Ivan Soysa, Field Assistant, resigned from his post in October 1963. Mr. Stanley Fernando, Laboratory and Field Attendant was promoted to the post of Field Assistant in December 1963. D.A. NETHSINGHE, Soil Chemist, Coconut Research Institute. 40 REPORT OF THE AGROSTOLOGIST 1. INTRODUCTION The Agrostologist returned from overseas study leave and resumed duties in July. During his stay in Australia, the Australian Government, the Waite Agricultural Research Institute, the CSIRO Division of Tropical Pastures, Queensland, and various Government Departments, Associations and Clubs extended to him very warm welcome and friendship which made his stay pleasant, profitable and success­ ful. To all of them he expresses his most sincere thanks. The old phytosolarium is still out of commission and as a result work on soil fertility has been greatly reduced. It is with great reluctance we record here the fact that manuscripts submitted to the Director as far back as 1958 are yet to be published. It is also felt necessary to point out that as the section on pasture trials would reveal, much information is being lost due to lack of facilities to carry out chemical analysis. The cadre on 31.12.63 was as follows:— Agrostologist:— K. Santhirasegaram, B.Sc. (Gey.), Ph.D. (Adl.) Research Assistant:— D.T. Rajaratnam, (Mrs.) B.Sc. (Cey.) Technical Assistants:— G.C.M. Goonesekera, Dip. Agric. (Pera.) D.E.F. Ferdinandez. Vacant D.C. Ellewela K.C. Muttuchchamy Y.G. Fernando, (Miss) W.S.C. Perera W.H.J. Fernando A.G.K. Silva S.F.A.:— L.F. Assistants:— L.F. Attendants:— Tractor Driver:— K.A.D.W. Jinadasa K.P.C. Fernando R.M. Dayaratne A.A. Fernando A. Dassanayake J.M.J. Jayamanne W.P.T. Perera S.A. Ratnayakc P.J.E. Fernando D. Amarasinghc M.I. Marikkar 4 | 2. SOIL FERTILITY STUDIES (a) Follow up studies at Horakelly and Ratmalagara These experiments reported in detail in the 1962 Annual Report have been completed and will be published. (6) Keenekelle Soils Preliminary studies in these have been completed. The soil is deficient in N, P and K. The results are being written up for publication. (c) Pothukulama Soils Studies on these too have been completed and the results are being written up for publication. (d) Gonapinuwela Soils Detail studies on this have been commenced. The first two experiments have been on micro-nutrients to furnish information to the Soil Chemist for his detailed field studies of the cause of the symptoms exhibited by coconuts in this area. Experiment I This was a 2 5 factorial experiment of Fe, Cu, Zn, Mn, B and Mg with two replicates and planted to Medicago sativa. At the early stages the plants responded to applications of Cu and B; but at the later stages these responses were less marked and led to significant depression in yield. At this stage however the plants showed symptoms which could not be explained in the light of past experience. In view of the fact that the coconuts were also showing unusual symptoms the possibility of the test plants also being affected by organisms were considered and the shoots and roots examined. The root system was found to decay at the tips and black lesions were found along the system. They were further examined by the Crop Protection Officer who reported the presence of a parasitic nematode Hemicycliophora longicaudata (Loos) in the roots. Experiment II This was a 4 X 2 factorial of 4 levels of B (0, 3, 6 and 12 lbs. Sodium borate/acre) and Ca as CaC0 3 10 cwts./acre with 4 replicates of all treatments and planted to M. Sativa. At all harvests Ca increased yields significant at the 0.1 % level, and growth from the second harvest onwards in absence of Ca was almost nil; and for the purpose of evaluating response to B, only the CaC0 3 treatments were considered. The yield/plant (gms. dry wt.) at the various stages of growth (for the 4 levels of B) are presented in Table I. 42 TABLE I Age in days *o *s * 6 Bn 39 0.07 0.22 0.20 0.12 48 0.13 0.36 0.50 0.34 38 0.57 1.07 1.03 0.84 90 0.41 1.69 1.47 1.31 112 0.16 0.63 0.60 0.52 • Total 1.34 3.97 3.80 3.13 Mean 0.26 0.79 0.76 0.62 B increased yields significantly (0.1.%) at the first level of application (3 lbs./acre). Thereafter there had been no significant responses, though a consistently increasing depression with increase in the rate of application was obvious. There was no serious effect due to the nematode in this experiment though the soils for both experi­ ments were drawn from the same sample. While the first .experiment was commenced soon after the soils" were sampled this experiment was planted six months later during which time the soil remained air dry. This drying may have killed the nematodes. H. longicaudata is a noncyst forming type (Ekanayake). • . It fs pertinent at this point to discuss the results of these experiments in relation to the leaf scorch symptoms in the South. This soil is deficient in boron, but boron deficiency had been detected in other soils (cinnamon sand at Horakelly and lateritic gravel at Mattegoda) employing similar techniques and same-test plant; but the coconuts in those areas do not show symptoms similar to that, exhibited at Gonapinuwela. At Horakelly incipient deficiency of Cu and marked response to added S have been recorded. It therefore appears highly unlikely that scorching of the coconut leaves at Gonapinuwela is caused by a deficiency of Cu, B and or S. It may be that due to some other primary cause, these or some other nutrient (s) is made unavailable either by non absorption or subsequent precipitation within the plant. As for the primary cause the root systems both with the coconuts and lucerne suggest the nematode to be a causative organism. The recent report of Rhind would well find support in these obser­ vations and merit some detailed studies. That secretions of an organism could upset nutrient availability one may site the inbalance of K and Mg caused by Fusarium infection of the bole of oil palms-in-.West Africa; (Hasselo personal communication). The prevalence of these symptoms in and around husk heaps could well mean that the soil conditions beneath the heaps favour the existence of effective population of the causative organisms. ( e ) Bandirippuwa Soils., . . . . . . The soils of the main estate were studied earlier by Paltridge and Santhirasegaram where response to P was rather unusual and Salgado had explained this as due to heavy application of bone meal at an.earlier..date. It.was therefore decided to study the soils from the newly acquired areas with a view tp assess the phosphate.status. 43 3. PASTURE STUDIES (i) Comparison of 3 grasses (P{) B/E: This has now been concluded in its original form. Each plot is now divided into 6 sub-plots of 6 palms each. Two sub-plots (random) in each plot receive nitrogen as follows:— (i) ^ original level of N application (£ cwt. (NH4)2S04/acre/Season). (ii) Original level of N application (1 cwt. (NH4)2S04/acre/Season). (iii) Double original level of N application (2 cwts. (NH4)2S04/acre/Season). Palm manuring is being carried out annually in accordance with the estate practice. This is to study the competition for N between coconuts and the various grasses and also the yield of the grasses at the three levels of N application. (ii) Cultivation X type of manuring (P2) B\E This experiment is now concluded. A detailed account of the results obtained have been reported in 1962. The plots are however used for the layout of experiments to study the growth and yield of B. miliiformis under various frequencies of N application (P 1 5 ) and different frequencies and height of defoliation (P 1 6 ) . (iii) Cultivation experiment (/*4) R/E The 1963 data further supported the findings reported in 1962. The experiment is being continued. (iv) Intensity of grazing trial (P 5) R/E The data on nut yield further supported the findings presented in 1962. There was still no definite trend due to the level of grazing. Due to M. miliiformis replacing B. brizantha in one of the plots and the large difference between replicates in the yield of herbage it was decided that no further information could be derived from the experiment in its original size. It was therefore decided to reduce it while maintaining the essentiality as follows:— (i) Weeds only—£ acre. (ii) B. brizantha ungrazed—\ acre. (iii) B. brizantha grazed 2 beasts/acre—\ acre. (iv) B. brizantha grazed 16 beasts/acre—\ acre. (v) Same as (iii), total fertilizer doubled. (vi) Same as (iv), total fertilizer doubled. The rest of the area is split into smaller plots to study the effect of doubling the level of manuring of N, P and K as a factorial (P 2 1). (v) Type of grazing Experiment (P6) R/E This experiment is being continued and the 1963 data support the findings reported in 1962. (vi) Pasture X Manurial experiment (P 7) B\E The system of grazing in this experiment has been changed from the October/November season when all pasture plots are grazed simultaneously with varying number and size of animals, depending 44 on the amount of herbage available so that grazing could be completed at the same time in all plots. The earlier system of grazing where animals were moved from replicate to replicate was followed with a view to obtain some information of animal performance. Since no worthwhile information was obtained over the three years, the new system was adopted. All plots were sampled for herbage yields prior to commencement of grazing. Coconut yields The yield of nuts and copra/acre for 1963 were adjusted with the 1959 pre-experimental data and analysed. Significant effects were recorded for nitrogen and potash while there was some advantage due to phosphates. Pasture species also had significant effects. There was no interaction among nutrients and between nutrients and pastures. Nitrogen Nitrogen increased number of nuts/acre particularly between N 2 and N 4 ( P > 0 . 5 ) while yield of copra/acre was not significant to the same extent. (Table II). TABLE O Number of nuts and copra (lb.)/acre and copra (oz.)/nut at the 3 levels of N N of nuts Copra (lbs.) Copra (oz.)fnut 2827.52 - 1457.24 8.40 2958.78 1547.58 8.35 N 4 3312.64 1680.56 ' 7.00 L.S.D.(P>0.5)* 309.76 179.46 •Highest value for any comparison. Similar findings have been reported by Eden etal (1963) for nitrogen. In the experiment reported by Eden etal (Loc. cit) the levels of N applied were 0, 0.5 and 1.0 lb. per palm every two years, while the levels of N applied in this experiment are 0.72, 1.44 and 2.88 lbs. over the same period. While Eden etal too have had response to N occasionally particularly at the early years of their experiment and that the data of this experiment now considered is only for 1963, the levels of application of N are however such that these results throw some doubt on their conclusions that benefit from nitrogen manuring was small or non existent. It is of particular interest to point out that while fertilizer application Was biennial in the experiment quoted by Eden etal, in the present experiment nitrogen is applied broadcast biannually. It may well be that this nutrient would have greater effects on yield with frequent applications. Potash Marked response was obtained from both number of nuts and Copre/acre. (Table in) . 3080—4 45 TABLE TH Number of nuts and copra (lbs.) per acre for the two levels of K No. of nuts Copra (lbs.) 2870.40 1466.46 3195.52 1657.13 L.S.D. (P>0 .5 ) 261.08 143.03 Pastures The estate control (weeds) plots yielded highest number of nuts and copra per acre. Between the two pastures, B. brizantha had a more severe effect than B. miliiformis particularly on the yield of copra. (Table IV). TABLE IV Number of nuts and copra (lbs.) per acre for the 3 types of vegetation No. of nuts Copra (lbs.) Estate control 3368.42 1747.87 B. brizantha 2628.48 1322.68 B. miliiformis 3101.44 1614.84 L.S.D.(P >0 .05 )* . . 350.87 114.91 •Highest value for any comparison. In another experiment (Px) reported in 1962 it was shown that in the same area these two pastures had a very slight beneficial effect on number of nuts per acre compared to the weed control. It must be pointed out that in that experiment (P^ the palms received manure around the base in addition to the broadcast fertilization and that the amount of weed growth was very much more compared to this experiment. The systematic delay in grazing of the replicates which would have had deleterious effects on soil moisture may be a further contributary factor in the present experiment. The new grazing system may eliminate" this and a more realistic picture would emerge in time. Herbage Yields The sampling made prior to grazing in November were analysed. The yield of herbage was signifi­ cantly low in the weed controls compared to the pastures, while there was no appreciable difference between the two pasture grasses. Only nitrogen had any significant effect on the yield of herbage which increased linearly with increase in the amount of N, particularly in the pasture plots. There was no obvious trend in the control plots (Table V) and there was no significant interaction either. 46 TA