LEAST— COST RATIGNS FOR BROILERS - A LINEAR PROGRAMMING APFROACH BY MOJISOLA ABOSEDE ADESIDA B.Sc. (Agric. Econs) University of Ibadan 3 * A thesis in the Department of Agricultural Economics Submitted to the Faculty of Agriculture and Forest Resources Management in partial fulfilment of the requirements for the degree of DOCTOR CF PHILOSOPHY of the UNIVERSITY CF IBADAN September 1979 111 - Abstract It has been established that Nigeria has a food problem especially where protcin intake is concemed. The poultry industry has b'odh identified = s the quiekest means af expanding protein supply and lowering its cost f , within the short run (.10-12 weeks for broilersj. However, feeds account for S5-75 percent of the total costs of production. Moreover, the nume- rous problrans faeing the feed industry coupled with the poor quality of feeds produced haue greatly limited the profitability and rapid expansion of the industry. The linear programming (l_.P.) tool was utilized to formulate least-cost diets which made use of locally available ingredients. The scarcity and rising costs of the grains (maize and guinoa-corn) which provide over GO percent by weight of broiler feeds prompted the use of cassava flour as an cnergy providing Substitute. Feeding trials were carried out to test the efficiency cf the least-cost diets. The objectiv/es of the study are (1 ) To use L.PNtool to formulate different least-cost rations which meet specific nutritional specifications for broilers, Ing readily available feed ingredients. Cassava and soya- n are being tested as energy and protein providing substitutes respectively. (2) To compare the least-cost formulated diets with the diets used by some commercial farms. (3] To find the optimum killing age/weight. - iv - (4 ) To find the rate of Substitution of cassava flour for maizc and guinea-corn in the ration for broilers. (5 ) To determine the economics of using different levels of A cassava flour in the rations for broilers. Experimental results shovi/ed that Starter diets with 24 percent Protein and 5 percent fibre level were better than those with 26 percent protein and 3 percent fib re levels. The computerised Starter and finisher diets tested were chcopqr and were found to perform better than the commercial diets. For the cassava based diets, analysis of the experimental results showed significant (P <^0.01, differences in Feed Conversion Efficiency (F.C.E.) in both Starter and finisher diets in which guinea- corn and maize were replaced. For weight gain, significant (P 0.01) differences were found only in Starter and finisher diets in which cassava replaced maize. For feed intake, significant differences (P ^ O.Ol) occurred only in Starter diets in which cassava replaced maize. The diets that caused significant differences were those in which the cassava contents were very high (25-40 percent) and they performed paorest. Even though growth is suppressed due to reduced feed intake caused by the powdery nature of the feeds, it is pertinent to note that diets with 40 percent cassava are still highly tolerable to the birds. Analysis of the weight response as cassava level increases showed that the decrease in weight gain was more rapid when cassava was being substituted for by maize rether than by guinea-corn. This could be V - attributed to the availability of nutrients or the amino-acid balance crf the guinea-corn based diets. Carcass qualities of the birds viere not taken into consideration because they are not highly rated in this socie.t.y.......................................... A The diets were further investigsted to see hov; the nutrients contents and energy-based ingredients influenced performanca, using the multiple linear regression model. The square root and quadratin functions were fitted but the quadratic forms gave the lead equations using the laid down criteria. Feed, protein, energy and the amino-acids intakes proved to be significant explanatory variables for the live- weight gain in the birds. Marginal Analysis was performed on scme selected functions. The elasticit^of produntion for energy and protein showed increasing retums to scale in the Starter and finisher diets st tho m:tt value of inputs. As higher levels of inputs are used, dimi- nishing retums is likely to set in. The elasticity of Substitution exhibits a unitary one also at the mean value of inputs. A percentage increase in the energy content of the feed results in an equal percentage decrease in the protein level of the diet. The extent of Substitution is limited by the requirement of the birds. Optimum quantities of the energy-based ingredients to produce the Optimum broiler weight gain were determined. Production surfaces, isoquants and isoclines were produced for selected functions of the energy-based ingredients. The rate of Substitution between guinea-com/cassava and maize/cassava were found to be declining with increasing level of output as more of cassava and less - vi - rf maize or guinea-corn are used. Estimates üf revenue over feed costs for the various diets were conputed. It was discovered that non-significant differences between riet without cassava was not synonymous with equal revanuc yielding diets. I- aeneral, the camputerised diets without cassava gave hiaher revenue . _ . < 2 - 3 rran the commercial diets. For the diets in which cassava replaced ths rrains, the revenue accruing to the farmer decreased as the perccntage cassava content increased. The revenue from guinea-corn diets were -owever higher than in the maize diets. Diets with 10 percent cassava -ad higher or equal revenue with the commercial diets. Diets with higher cassava levels were costlier because cassava is costlier than the grains. It is however envisaged that priccs of cassava may fall in the near future cecause of increases in production. Revenue from the diets was thereforo octained using var ,ing costs of diets as cassava price varies. When cassava was made to assume the same price with guinca-corn, all the cmoutarised diets except that with 30 percent cassava levcl had higher revenues than the commercial diets. The revenue increased as the cassava crires were reduced but the diets with 30 percent cassava gave the lowest revenue all the time. Optimum killing age determined suggested that rrrilers be sold at eleven weeks for most of the diets except those in "ich five and 10 percent cassava replaced guinea-corn. The implications of this study are that efforts to improve returns — pojltry farmers must be focussed on the cost and quality of feeds. V l l Partic.ular attention must be paid to cheap sources of protein, carbo- -ydrate and oils. There is a very high potential for the use of cassav/a if its adoption becomes a reality in the future. Further investigations are necessary in testing the least-cost tiets with the existing various breeds of broilers. Comparison can also be made of the use of soyabean and groundnut cake as a protein - viii - Apknowledgements In preparing the materials for this thesis, I received help from -any people in tho Department of Agricultural Economics and Teaching and Research Ferm, both of the University of Ibadan and Pfizer Livestock Feeds. Although it is not possible to includc-. the name of overybady involved, yet I am particularly grateful tc the followingt (i) Dr. 0. ügunfowora, Reader in the Department of Agricultural Economics, University of Ibadan. He supervised my work, made helpful suggestions and painstakingly road through and corrected the drofts of the thesis. He elso provided the means of funeling the resc« S T (ü) Professor D. L. Fetuga, Department of Animal Science, University of Ibadan, who readily offered guidance and useful suggestions, carefully read through and corrected the drafts of 'che thesis. ‘.iii) Professor 3. U. ülayide, Deputy Vice-Chancellor, University of Ibadan his interost in the study and for proviriing tty^.iatn3 of funding the resesreh. iv) Pj'vi'"" sor G. M. Babatunde, Head of tho Department of Animal f i'neu, University of Ibadan for his expert advice and for oporjcg his experimental pens for use in this study. Dr. J. 3. Flinn, farmer Farming Systems expert at I.I.T.A. Ibadan for his useful suggestions. Ho also offerred assistanca to makr. the Linear Programming Model work. - ix - (vi) Dr. J. K. Olayemi, Reader in the Deptartment of Agricultural Economics, University of Ibadan for his guidance at crur.ial stages of the study and for provirling the means of funding the research. (vii) Dr. J. A. Akinwumi, lecturer in the department of Agricultural Economics for reading thraugh and correcting the drafts of the thesis. (v/iii) The staff of Teaching and Research Farm, University of Ibadan especially Mr. F. A. Roberts, the Farm Director, M. T. Oke, and Mr. J. A. Bolarinwa who Drovided the exoerimental materials (ix) The staff of Pfizer, Mr. H. A. Aderibigbe, Mr. Ladokun and Mrs. Akintola who mede available some scarco feed ingredients and useful Information. (x) The Office uf the Department of Agricultural Economics von Aya Akilepa and Mr. Qpeogun for typiinn;~ tne thesis sc ‘.veil. (xi) Mrs. Wiütkes Osofisan and Mrs. Oni both of the Bumputing Oentre University of Ibadan for writing up and ruoning the programmes used in the study. - X - (xii) Doctors Adene and Kasali önd Mr. Njoku of the Faculty of Vetarinary Medicine for providing veterinary Services throughout and whenever needed. (xiii) Relations who gave me all the moral Support needed to see me through. My deepest appreciation goes to my husband 'Biyi, and our daughter ’Tutu who patiently lived through the demanding period of this study and gave me all the encouragement to sustain me. - xi - CERTIFICATION BY SUPERVISORS We certify that this work was narried out by Mrs. M. A. Adesida in the Department of Olabisi Ogunfowora, B.Sc. (Lond.), M.Sc. (Reading), Ph.D. (Iowa) CONTENTS Ffetge Title ..................... i Dedication ................. ii Abstracts.... ............. iii Acknowledgement ............ vi'ii Certification by Supervisor xi Table of Contents ......... xii List of Tables ............ xix List of Figures ........... xxiv CHARTER ONE INTRCDUCTION .................... 1 1.1 The Problem ............... 1 1.1.1 Deficiency in protein intake 1 1.1.2 Expansion of protein sirpply 3 1.1.3 Feed costs and guality .... 4 1.2 The need for this study .... 9 1.3 Objectives of thn study .... 10 1.4 Plan of Thesis ............ 10 I -A^TcR LINEAR PROGRAMMING AND FEED Cü.iPOUNDING - A THEORETICAL REVIEW .................................................. f 2.1 Previous Works in Computer-formulated poultry industry 12 2.2 The model for lecst-cost ration formulation ........ 14 - xiii - 2.2.1 The model ......................... H 2.2.2 Problems of Diet Formulation ..... 17 (a) Tho feed compounder'5 problem 17 (b) The livestack farmer's problem ........A 19 >APTER THREE :H LEAST-COST RATION COMPOSITIONS EMFLOYED IN THE STUDY 20 3.1 The basic matrix .......... i ..... . 20 3.2 Restrictions in the model .. ............... 20 3.2.1 The weight constraint ..... 22 .... 3.2.2 The ingredient canstraints . 22 £ ........ (a) Energy based ingredients a . . ................... ... 22 (b) Protein based ingredients ........................... 23 (c) Mineral and vitamin based ingredients ............. 23 3.2.3 Nutritional restrictions ................................. 24 (a) Energy . 24 Cb) Protein ..... (c) Minerals ..... Cd) Amino-acids .. (*) Fibre ....... (f) F a t ......... 31 5.2.4 Concluding remarks................. ..................... 32 Solutions and discussion of the model .................... 32 5.3.1 Stability of the mix of ingredients to changes in their prices in the Starter ratinn ................. ......... 36 - xiv - Page CHAPTER FGÜR IV FEEDING TRIALS 4.1 INTRODUCTION .......................... 39 4.2 Experimental diets ............... . 39 4.2.1 Starter diets for Experiments I and II : : : : : : 39 4.2.2 Finisher diets ........................ ...... x ..... 42 4.3 The madsl ............................. 42 4.4 Experimental design and setting V 44 4.5 Broiler birds and their management ..:.:. : 45 4.6 Limitations cf tha experimentr 46.... (a) Fecd ingredicnts 46 (b) Housing ... 46 (c) Method of measuxrKoment 47 4.7 Experiments and performanee comparisons .................. . 47 4.7.1 Experiment I: Comparison of eight computerised Starter diets of \yarying preteio fibre and cassava contents 47 (a) Breed and stonking r-,t e ............................ 4B Cb) Mortality ................................... ....... 4B (c) Performance comparisons ............................ 49 4.7.2 Experiment II: Selen;ing the best pair of computerised starter and finisher diets ............................ 49 (a) Breed and stocking r a t e ............................ 51 (b) Mortality .......................................... 51 (c) Performance comparisons of two selected Starter diets 52 (d) Conclusions on starter diets ....................... 55 - XV - 'age 4.7.3 Finisher diets ............................ ............... 55 Performance comparisons............... .................. 56 4.7.4 Performance comparison of the computerised Starter and finisher diets v/ith two commercial diet3.............. 4.0 Empirical estimation of parameters affecting broiler diets 60 (1) Functional forms employed ........... ............... 62 (2) Regression results .... ............... ............... 65 (a) Criteria for selecting the ’lead' equation ............ 63 (b) Effect of fsed intake an liveweight gain ............. 64 (i) Methodolagy ............ 65 (iij Estimating procedures............................... 05 (iii) Empirical results ..................................... 66 4.3.4 Comparison of marginal tph:ysical product with feed conv/orsion efficiency...... ........................... ........... 70 4.9 Summary and conclusions.................................. 72 CHARTER FI VE V CASSAVA AS A Sl E FOR MAIZE AND GUINEA-CORN IN POULTRY (BROILER) 73 5.1 Intrüduj ction ........................ 73 5.2 Choraicteristics ..................... 74 5.3 Factors affecting cassava production . 74 5.4 Price comparisons................................. ....... 5.5 Previous studies .......................................... 5.5.1 The review.... ........................................... 77 5.5.2 The offect of Processing an the nutrient composition of cassava ................................................... 83 - xvi - Page 5.5.3 Summary of works on cassava...... .................... 8 ‘ 5.6 L.P. Solutions for the inclusion of cassava in Starter diots ..................... ............. ............. 8! 5.7 The composition and costs of Starter diets with varying prices of cassava ....................... ........ 87 5.7.1 Substitution between maize and cassava .... 89 5.7.2 Substitution betweer nuiriea-com and cassava 80 5.8 Feeding trials ............... 89 5.8.1 Experimental diets ............ .... ^ ....... 90 ...... (a) Cassava replacing guinea-corn (Starters) 90 (b) Cassava replacing maize (Starters^.............. 92 (c) Cassava replacing guinea--corn (finisher) ....... 95 (d) Cassava replacing maize (finishers) ...... ...... 95 5.9 Experimental .......................................... 98 5.9.1 Statistical analyses nf data ......................... 100 (a) Experiment III ................................. 100 (b) Experiment I V ................................. 100 (c) Experiment V ................................. 105 v C H > Experiment VI ............ .................... 108 5.9.2 Goncldsions .......................................... 1 nn 5.10 Weight response as cassava level increased ........... 112 CHARTER SIX VI AN EVALUATION DF THE TECHNICAL AND ECONOMIC PERFORMANGES CF VARIOUS NUTRIENTS AND INGREDIENTS IN BROILER DIETS 115 6.1 Single nutrient effects 116 - xvii - Rage 6.1.1 Effect of protein intake on liveweight g a i n ............ 116 6.1.2 Effect of energy intake on liveweight gain ............. 120 5.1.3 Effect of Amino-acids on liveweight gain ............... 121 6.1.4 Marginal analysis .......... .... . 428 6.2 Nutrient combination effects 1 34 5.2.1 Effect of energy and protein Dn üveweight 1 3 ; 5.2.2 Effect of lysine, methionine and cystine on liveweight gain 137 S ' 5.2.3 Marginal analysis ....................................... 138 6.3 Substitution between guinea-corn, maize and cassava ...... 143 6.3.1 Maize and cassava ....................................... 143 a X „ 5.3.2 Guinea-corn and cassava .. ................................ 156 5.4 Economic analysis of the diets^........................... 165 6.4.1 The optirnum marketing age/weight......................... 157 S. 4.2 Composition of net revenue from diets.................... 170 (i) Compi .'terised diets with varying fibre and protein 170 lcvels............................................. 170 (ü) Diets in which cassava replaced guinea-corn ......... 172 (iii) Diets in which cassava replaced maize ............... 176 7HAPTER SEVEN T U SUMMARV AND CONCLUSIONS ..................................... 181 7.1 Summary of the study ................ 181 7.1.1 Regression analysis..................... 1®4 7.1.2 Economic analysis of the diets............... 185 - xviii - Page 7.2 Conclusions ............................................... . 1 '; 7.3 Recommendations............. 188 7.4 Suggestions for furthcr resaarch ........................ 189 REFERENCES.......................................... .. 190 APPENDIX ........ .............................. ...... , 200 ......................... ............ / 1 - xix - LIST CF TABLEB ' -3LES Page 1.1 Available average calories and pretein supply per capita in the 12 States of Nigeria 1960/59 ..... 2 1.2 Price ('./Ton) of livestock fseds in Ibadan, Nigeria, 1972-1973 ...................... 6 3.1 The basix matrix ........................... 21 3.2 Performance of broilsr rations containing higher levels of energy ranging from 1,760 kcals/kg - 3,159 kcals/kg. 25 3.3 Mineral requirement of chicks ...... ̂ ..................... 27 3.4 Amino-acid requirement of chick^fc»......................... 29 3.5 Costs and compositions of the Starter and finisher .rations excluding cassava ... 3? Effect of changing costs of ingrodients on the percentage composition and costs of Starter ration, average producer costs, 1976 prices versus U.I. costs, 1977 prices .................. 35 Stability cf the mix of ingredients to changes in their prices in Starter rations....................................... .. 37 4.1 Nutriant composition and cost of eight Starter diets ........ ■\0 4.2 Composition and cost of the tv;o finisher diets without cassava 4r3 4.3 Performance comparison of ton Starter diets ............ . 50 4.4 Performanca comparison of tvvo Starter diets .................. 53 4.4.1 ANOVA table for weight gains ................................. 54 4.4.2 ANOVA table for feed intake .................................. 54 4.4.3 ANOVA table for feed conversion efficiency ................... XX TABLE Rege 4.5 Performance comparison of two finisher diets with two commercial diets ......................................... 56 4.5.1 ANOVA table for weight gains .... .................... 57 4.5.2 ANOVA table for feed lntake ....... ........... ............ 57 4.5.3 ANOVA table for feed conversion efficiency .... ̂ ..... 5i7 ........ 4.6 Performance comparison of the conputerised startcr aind finisher diets with two commercial diets — 59 4.7 Empirical results of weight response to foed intake in Starter diets . ..............^ .................... 67 4.0 Empirical results of weight response to feed intake in finisher diets . ..... ....... 69 5.1 The effect cf including cassava on the cost and composition of starter ration - U.I. prices .......................... 86 5.2 The composition and costs of Starter diets with varying prices of cassava ................................................ 00 5.3 Composition of starter diets with cassava substituting for guinea-com ............ ................................. 91 5.4 Composition and costs of starter diets with cassava rcplacing maize in the diets ............................ .......... 53 5.5 Composition and costs of finisher diets with cassava replacing guinea-corn ................................... 96 5. 3 Composition and costs of finisher diets • with cassava rcplacing maize 97 TABLE Paqe • • - j P r j 5.7 Performance comparisons of Starter diets in which guinaa-corn is replaced by OrQQ percent of cassava.......... ........ 101 5.7.1 ANOV/A tablo for weight gains........... .................... 102 5.7.2 ANOV/A table for feod intake ......................... 102 5.7.3 ANOV/A table for feed canversion efficiency ...... ........... 1^2 5.8 Fterfcrmance comparisons of Starter diets in which cnssava (0-30) percent) replaced maize in the dietjjp .............. 1®3 5.8.1 ANCN/A table for vyeight gains................................. 1®^ 5.8.2 ANCVA table for feed intake...... .£v....................... 10-r 5.8.3 ANOV/A table for feed conversion efficiency ...... ........... 101 5.9 Performance comparisons of finisher diets in which 0-40 percent cassava replaced guinea-com in the diets ............... 106 5.9. 1 AN0VA table for weight gains............................... . 107 5.9.2 ANOV/A table for fead intake ................................. 107 5.9.3 ANOV/A table for feed cenversion eff iciency.................. 107 5.10 Performance comparisons of finisher diets i.n which 0-40 percent cassaua replaced maize in the diets ...................... 1®9 5. 10.1 ANOV/A table for weight gains................................ H g 5.10.2 ANOV/A table for feed intake....... ................... ...... 11 o 5.10.3 ANOV/A table for feed conversion efficiency ................. 110 5.1 Effects of protein intake on liveweight gain ............... 119 5.2 Effects of energy intake on liveweight gain ............... 122 5.3 Effects of lysine intake on liveweight gain. ............. . 12' 5.4 Effect of methionine plus cystins intake on liveweight gain 127 - xxii - TABLE Page G.5 Combined effect of protein and energy on liveweight gain .... 136 6 .6 Combined effect of lysine and methioninc plus cystine on liveweight gain ........................................... 139 6.7 Combined effects of maize and cassava ....................... 1 *5 6.8 Maize intake, cassava intake and average weekly liveweight gain coordinate points on the production surface........ 1'r8 6.9 Derivation of points of the isoclines on the production surface 131 6.10 Combined effects of guinea-corn and cassava on liveweight gain 15® 6 . 1 1 Guinea-com intake, cassava intake and average weekly liveweight gain coordinate points on the production surface ........ 6.12 Derivation of points of tho isoclines on the production surfac 5.13 Revenue over feed cnsts at G, 8, 10 1 1 & 1 2 weeks of age for diets in which cassava replaced guinea-corn (f/bird) ..... f68 5.14 Revenue over feed costs at 6 , 8, 10, 11 &12 weeks of age per diets in which cassava replaced maize (0/bird) .......... 169 5.15 Summary of net revenue over feed costs per bird for diets in \ y experiment II for the whole of the reering period ....... 171 5.16 Summary of revenue over feed costs per bird for seven diets in which cassava replaced guinea-corn in the diets and two commercial diets . ................ 173 5.17 Summary of net revenue over feed costs per bird at varying costs of feed (varying costs of cassava) when cassava is replaced by guinea-com in the diets...... .......... 17 5 - xxiii - TABLE Page S.18 Summary of net revenue ovcr feed costs per bird from seven diets in which cassava replaced maize in the diets and ' ■ two commercial diots..................................... . 177 . ? < P S. 19 Surmary of net revenue over feed cests per bird at varying costs of feed (varying costs of cassava) ivhen cassava is replaced by maize in the diets.......... ................. . 17 9 LIST CF FIGIJRES Figures ?*32. 1.1 Economic input/output relationships 8 5.1 Weight response as r.assava level increases in the Starter diets ................................. 447 6.1 Isoquants and isoclines fcr maize and cassava 149 6.2 Isoquants and isoclines for guinea-corn and ~ .................................................................................................. ” e r CHARTER ONE INTRODUCTION The Nigerian govemment, in ordor to reduce the importation of poultry Products and to increase the amount of dietary protein available, initiated a rapid agricultural developmont Programme in the 19G2--60 plan. Ssteps taken included the production of breeding flocks, ths establishme.n.t. of hatcheries and feed depots in major towns as Service centres in tbhe. couLntry. Greater efforts were made in subsequent National Plans and a large number of entrepre­ neurs ventured into poultry production. IJnfortunately, most producers wont out of business due to low margin cf profit and, in most cases, losses sustainoc This was due, among other factors, to t;hi3 fact that feed costs were very high and inspite of the high costs, the feed were of poor quality. ̂ ̂ The Problem 1.1.1 Deficiency in Protein Intake It is an established fact that the protein intake of the average Nigerian either from plant or animal origin is inadequate. For instance, the FAO s u r v o y ä1'9/s h n w e d that about 90 percent of the protein in the diet of the average Nigerian is derived from vegefcables, pulses and nuts, thus leaving only 10 percent to be supplied by animals. In fact, the supply of vegstable protein is adequate only in the Northern part of the country and the supply of animal protein is alsc higher there than in the Southern parts. Table 1.1 shows the available average calories and protein supply per capita in the former 12 States of Nigeria for 1968/69. TABlE 1.1: Ava lieble average- calcries and proteir. .oueply per capita in- the 12 States of Nigeria 1960/'59 * Source: Hanoi} «d fron food balanca sheet tables (9-22) jf ül:-yidek 3. 0., et al. 59/. - 3 - These figures do not measure up to the FAO redömmendations of 2,500 kilocalories of energy and 65 ;grams of protein per day except for Mid-West, Benue-Plateau and East-Central States which have 77 groraa . 82 gms. and 67 grems of protein per day and 2,B81, 3,961, 3,091 kilocalories of energy per day respectively. Two other Southern States, Western and South-Eastem, have energy supplies greater than the recDmmended figure but their protein intakes are below Standard. The protein/calorie percentage figures are higher in the Northern States than in the Southern States.^^^ The FAO estimated per caput consumption of all types of meat (excluding offals) to be 9.2 kg. per year, 7.1 kg. per year and 5.2 kg. per year in the North, West and East respectively. ✓"fliis is very small compared with the FAO recommendations of 65 grams of protein per day out of which 35 grms. should be of animal origin. The FAO 1969-~ report also show protein shortage in the diet of the average Nigerian. It observed that the low average intake of animal products could be due to either their non-availability or the fact that they are too costly for certain income groups. The workings of economic forces show that it is due to both. If demand exceeds supply, denoting unavailabili price rises so high that some income groups cannot afford it. 1.1.2; Expansion of Protein Supply The high price of animal products prevents the low income eamers from using the sources of animal protein which otherwise should have been made available cheaply to them to improve their nutrition. The quiekest means of increasing animal protein supply within the short run is through rapid expansion of the poultry industry. The poultry industry is characterised by a short economic cycle (10-12 weeks for broilers) and should be relied upon to provide a significant Proportion of the animal protein needs of the rapidly growing population. The industry requires a package of cheap and best quality feeds, best breed of birds with good inten­ sive management practices in a commerclnl set up. Unfortunately, poultry production in Nigeria is characterised by high costs, small profit margins and the need for adequate finsncing.6™0'/ Cattle, sheep and goats are important in the Northern parts of the country where extensive naturQl grassland areas are available. In the South, extensive arable pasture no longer exists because of high population pressure on land and the dense Vegetation. ™ a « * u t h would therefore have to rely mostly on poultry and piggery industries for most of their protein requirementE since pigs and poultry convert grain seeds and waste products unfit for human consumption into meat.^^ Within poultry production however, retums to capital come quicker from broiler production than egg production because broiler production is characte­ rised by lower capital turnover than in egg production, V®®* ^1* and in 13-12 weeks thousands of broilers can be produced. However, since feed accounts *or up to 65 percent of the costs of production, the cost and quality of feeds ~ave greatly limited the profitability of and thus the rapid expansion of the broiler Enterprise. '.1.3; Feed Costs and Quality The most important factor determining profit levels in intensive livestock - 5 - enterprises is the cost af feeds. Ikpi —32-/ gave the percentage contributian af feed costs in brailer proriuction as 64.37. Heady —27*■/ and Blagburn i—n/- gave it as lying between 65-75 percent. Feed prices keep rising as the demand far feed increases. Oyenuga 62/ attributed the rise in the prices of feed to steep rise in the prices of the basic feed ingredients such as maize, groundnut cake and fish meal which rose by 114, 100 and 355 percent respectively between 1969 and 1974. Table 1.2 below shows the price trend of livestock feeds in Ibadan (Oyo State) from 1972-78. From this teble, prices of the various kinds of livestack feeds increased in the ränge 11 pereent to 164 percent within the seven year period. Fatteners and weaners mash increase respectively by about 111 and 126 percent whilst pig breeders mash increased by 119 percent within the seven year period. Growers mash increased by 130 percent . Layers mash by about 144 percent, broiler finisher by about 134 percent and broiler Starter by about 155 percent. The highest increase was obtained in the price of chicks mash which rose by about 164 percent. It would appear therefore that the most reasonable cause crf action is a combination of measures designed to reduce the cost of feeds and improve the quality of diets such that a high level of animal performsnce and feed effi­ ciency could be attained. The main economic problem is coneerned with the relation between total feed input and meat output. It is characteristic of broilers like all other - 6 - TA3LE 1 2* Pricc ('i/fan') of ;livcstock feeds in Ibadan, üyo State, Nigeria, 1972-1976 ‘ 1 Source; PfizG* ProcLnts Limited, n.o. - not ^vailabli'. - 7 - growing stock that their efficiency ss feed convorters declines with age. It is most profitable to seil off the birds or kill them at the point where the margin over total feed costs per bird is at its maximum. This is the point where the value of extra meat produced with corrtinuous feeding is equal to the extra cost of feed. Figure 1.1 below illustrates the point further. The figure shows the value of extra meat production rssulting from every additional gjcanr- of feed, assuming the prices of poultry meat and feed to be constant. The MVP curve shows the diminishing marginal retum relationship -betweenfeed input and meat output and the feed cost line is horizontal showing a constant price. where MC = MR and here the net retums to feed are maximised and can be converted to an Optimum killing age or weight (CN). If, for example, the price of feed increases, the optimum kiliing age will be shortened to the point where the dotted line intersects the MVP curve (ON). Similarly if feed is more efficiently converted into meat, or if the price of poultry meat rises, the MVP moves up to the dotted curve and this will increase the optimum killing age to the new point of intersection between the two curves (OM-,), Efficiency of feed conversion obviously depends upon the quality of feed, poultry management, and genetic ability of the birds to convert feed into meat. Sincc feed is the major item of cost, it is svident that an improved ratio (kg. of feed required to produce 1 kg. of poultry meat) will result in a significant reduction in the costs of production. An increase in feed efficiency obviously has greater impact in reducing the cost per kilogr; - e - liveweight whon the price of feed is high. 1.2: The Need for This Study The need for this study arose, therefore, from the desire to expand poultry production to improve the nutritional status of the average Nigerian by making poultry production possible at low costs. Some of the possible Solutions to the expansii ie poultry industry lie in reducing the unit cost of production through efficient ration formula- tion. But since feed costs represant a significant proportion of the total cost of production, rnuch attention producing good quality feeds at the lowest prices possible Reduction in feed costs could be accomplished by making use of locally availcble ingredients, substituting some ingredients for others in response to changing price structure, achieving management and technical efficiency in workable procurement and distri- bution arrangements for ingredients and feeds, respectively. In particular, the scarcity and the rising costs of maize which provides over 60 percent by weight of broiler feeds calls for a search for other energy providing Substitutes. Accordingly, the exploration of the possibility of substituting cassava for maize as the major sources of energy in broiler ration could be regarded as a worthwhile exercise. Moreso when it is realised that the production of guinea-com is not yet well organised and the competition * between human beings, livestock, textile and starch industries for maize makes ..ü - ths locally produccd maize so soaroo and costly. Casnave is baing ccnsidered because a lot of work has been dons by research institutior.s both in its production and uss in livestock faeds. The economic aspect has howover, been neglected. - \ 1.3: Dbjoctivos of the Study . The ob.iectives of this study ero~: O v l) To uso Linear Programming (l_P) toolvbo ̂ ormuLofce different leert cost rations which meet specific ̂ JteHfci'itionEl seocifications fer broilers, using roadily available feed ingredicnts in Niger: n. Soyabsan and cassava are^bs^g testsd os protoin and cnergy Substitutes, respectivoly. 2) To comparo the ieast-cosc formulated/car.;:utcrisod dints vv:.th tho diets uss^^Neome commercial Farns by -feeding tho different diets to hj^^hers. 3) To finfl%he Optimum r* — • •. 4) Tu^fVid thc rate of Substitution of cassava flcur fev r.aize aridguii.oa-com in the ratio. .5 fer broilers. o dstermine the economics of using different lovels of rassava flour in the rations of broilers. 1.4: Plan jof Thesis The socond chapter traces tho historicval dsveiaf ..ent cf e.-rpiri'Y for- lulated diets for livestock. The ralationship bsb.v’cr tham and the präsent study arc highlighted. Tho tool or technique used is th - ~hly e>'~ "inod. 11 The model for least-cost ration fonnulation is stated with the Problems inv/clved. The third chapter discusses the least-cost ration campositians employed in the study. It shows also hov/ the nutrient requirements are miet to ensure a balanced diet. Chapter four discusses cassava as a component of animal feed, focussing in particular on the advantages and disadvantages of'ats use. The results of the Substitution of cessava for maize and guinea-com in the L.P. model are also presented here. In chapter five, the model used e feeding trials is discussed including the experimental design and setting. The management, diseases and general condition of the experimental broilers are discussed thus revealing the limitotions of the study. Performance romparisons of all the experimental diets arc made. Chapter six presents the major results of the study. Empirical estima- tion of p3rameter& affecting broiler diets dominated the chapter. Nutrient effects either singly or in combination with eaoh other and the extent of depcndence an amino-acids for growth are examined. An economic analysis is performsd for the diets and Linear Programming is compared with marginal analysis as Optimum decisian criteria. Chapter sev/en gives the summary of major finriings, general conclusion and the policy recommendations. CHAPTER TWO LINEAR FROGRAMMING AND FEFD C t t W liNDING - A THEGRETIGAL REVIEW 2.1: Previous V/prks in Computer-formblated A lat of work has been done in the field of I .ng formula- tion of poultry feeds. Anwar —2/ used L.P. tool to compute least-cost chick Starter ratian according to the ratio between protein quality value and energy content. He incorporated Gross Protein Value Units and the romaining five percenic was contributed by the different feed additives. He supplemented the proteil rieh feedstuffs with lysine and methionine which are the most limiting amino-acids. The protein-rich feedstuffs which were supplemented were ponoidered as different ingredients having higher prices, since the GFVs increased more than for the unsupplemented ones. With soya- bean as the protein source, he found lysine to be ineffective whilst supple- mentation with methionine was economical, both in the reduction of the cost of the ration and the incroase in weight gains. Anwar also emphasized the measure of the protein quality of the feed. 25/ Gifcson, et al.— devoted their study to a detailed examination of the methods and procedure involvec! in the use of L.P. and an electronic- Computer to detarmine several types of leest-cost poultry rations inc.luding chick starte turkey Starter, layer and broiler rations. They alsc examined the effects of changing feed ingredient prices on the least-cost rations. They suggested that the results of an L.P, analysis are only as reliable as the input data which are used. This study did not only confirm the usefulness of ir^^xTi sciplinary cooperation in agriculture but the results alsc haue important implications for both poultry nutrition research and their applications to commercial feedmill operations. however, the. only ghortcoming in the study jls the failuro to reeegnise . that least-cost rations are not necessarCiill)vy the most economic in terms of profitability. This could haue been ascertained through feeding trials designed to test the birds' performance on and the lcvels Qf profit from the various least-cost rations. Heady, Balloin and McAlexander --1 estimated and fitted broiler pro­ duction functions and specified least-cost rations over two weight range3. Growth isoquonts indicating the possible combinations of two major feed ingredients (corn and soyabean) which result in a fixed gain lev/el were pre- dicted. They also determined the Optimum level of feeding as well as the Optimum market weight. Flinn •Q— Jl SeRstiirmated broiler production functions as well as response surfaces, using diets of varying protein levels. He performed carcass quality tests and determined various economic indices such as optimal slaughter weight, and least-cost feed rations. Least-cost feed rations were determined by flnding least-cost input combinations that produce the optimal slaughter - 14 - wsight. Ho found that least-cost inputs howevcr, depend solely on the ratir cf input prices and the biological characteristics of the response ?urface. This study uses predetermined least-cost computerized rations in the feeding trials for the purposes of estimating such economic indices. < 2 ~ 2.2: The Model for Least-Cost ration Formulation Due to recent advances made in poultry nutrition research, it is becominc increasingly difficult to find the ration which is least-cost and at the same time, meets the recommended nutritional requirsments. A well balanced poultry ration may inv/olve 20 to 30 nutrient require- ments and many fe3d ingredients whose prices are constantly changing. Througn lang experience, it may be possible to determine the low-cost ration by trial and error methods, but this is costly and time-consuming and there is no guarantee that the selected ration is the least-cost one. (2, 8, 13, 15, 22, It has however, been demonstrated in various studies 23, 25, 30, 31, 30, 57 , 76) t..h at. the .L .P_. t. ool. i. s a very powerf_u l. , rapi. d. and efficient technique as far as formulating least-cost rations is concerned. 2 .2 .1 :The Supposo there are n ingredients X , X-, ..... Xn such as maize, cassava, ' *-> » guinca-corn, blood meal, fish meal, groundnut cek-3, etc. available for for­ mulating the ration, and their prices are C^, (J0 , ..... Cn, then the problem reduces to minimising an objective function: - 15 - Z =» 7 .C.X. .................... (eq. 2.1) where Z is the cost of the formulated ration 0^ is the net price per unit of activity (there are n different activities) X, is the level at which each activity is to be ' S ? Usually, the rations have spscified nutrient requirements or restrictions to meet the needs of the birds. If the percentage of nutrients such as meta- bolizable energy, protein, amino-acids, vitamins, fibre, calcium and ( phosphate is less or greater than the specified levels, it might reduce the quality of the ration. These restrictions constitute the constraints in L.P. model and they take the form of linear inequalities which means that the total requirements for any nutrient must be equal to or less than the total amount of that nutrient available in all the ingredients included in the ration. The statemi that the protein level in the ration must be less than or equal to the percentage protein in all the ingredients can be exprossed as followsv .nJ- • Protein a . .x . (eq. 2.2) J-1 J J Similarly for energy, tha relationship can be expressed as follows. n , Energy, ^ ^Vb. . x . (eq. 2.3) 3 5 i<1 J - 16 - For many nutrients that ara involvod, the problem can be stated as n Minimize Z C .X . (eq. 2.1) j=--1 W J Subject to di ,a. . x . (eq. 2.4)Ci 1J J (eq. 2.5) x j > Where, ith < > > ' d^ is the level of ths nutrient a.. or b. , is per unit content of the i ' ' nutrient in the j . ingredient •**vJ J 4*h X . is the level at which the j . ingredient cnmss into the Programme. X . D specifies^hat there must ba nc negative activity, i.e., there mtuusst bte na negative amount of ingredient. Ssveral modificatians and/or refinernents of the basic model could be cons- trunted to make the Solutions morc nutritionally acceptable. For example, models could bn structured to permit a speoified leval of an ingredient / nutrient, or a percentage ränge could be specifiod for certain nutrients. In symbolic terms, a model, which specifies oxactly 0.01 porcent of a given nutrient/tonne of ration can be oxpressed as follows: n .. 0.01 = fs a. ,x. <*v~ i j J (eq. 2.6) J=1 ün the other hand, a model which specifies that a nutrient ingredient should be included in a given ration at the ränge of 2 - 5 percent per tonne could be represented as follows: 0.02 ^ a , ,X. (eq. 2.7) - jT, ij J n .. 0 . 0 S > (eq. 2.8) Equation (2.7) ensures that the ingredient/nutrient is included at a level not lower than 2 percent while equation (2.8) ensures that the nutrient/ ingredient is not included in the « p a t a level higher than 5 percent All the Symbols are as explained in the basic models. 2.2.2: Problems qf_ Dlet_Fqrrw|latiqn Numerous problems face the farmer when formulating lcast-cost diets for livestock from time to time. There aro two categories of farmers in livestock production namely (i) the feed compounder and (ii) the farmer who raises the livestock b^'feeding the computerised diets for rneat, eggs or milk production. Problem^ of diet formulation can therefore be groupsd into those from these two categories of farmers: (a) The feed compounders problems (i) The most important problem of the feed compounder is that which concern the nutrient composition of feedstuffs. There is no doubt that there is considerable variability in the quality of feedstuffs from one batch to the - 13 - ather. This makes it compulsory that each batch of feedstuff be analysed. The available storage facilitiss even influence the nutrient availability □f ingredients. (ii) A second Problem facing the compounder in setting restrictions for a the nutrients supplied to them. For instanco, the amino-acid centent of a feed is an important guide tO' quality but relatively little is known about the availability of individual amino acids in feed ingredients to the animals. Chemical tests that should cients are very expensive and time consuming. (iii) Thirdly, random variations in the cost of ingredients depending on tho magnitudo of change, will affect the least-cost mix. The compounder therefore has to obtain a new least-cost composition for every change in cost outside the minimum and maximum ränge allowed by the Linear Program. (iv) Fourthly, the compounder must decide on what his Optimum goal should be. He can ehoose either cost minimisation or Profit' maximisation depending on his assets and facilities. For instance, cost minimisation is the aim of a farmer who has a small to medium size mill and depends vsry largely on the ability to handle small quantities of several diets. He must also have constant access to L.P. to Scan the various recent available feed ingredients and their e xisting prices to meet each particular order. Whereas, a compounder selling feedstuffs tc farmers and has amplB storage, milling and mixing capacity 19 - to provide for the demands of all customers with respect to quantity and type of diet, is concemed with getting the maximum return an capital investment. Lastly, the compounder must maintain the business by satisfying the demands of farmers. His gfeatest problem in this respect is that farmers generally do not buy in a perfectly competitive way. Also, for long term planning as to what his mill output should be, he depends very largely on the volume of sales which may not Show tha correct picture. t v V (b) The livestock Farmers. FVcblems The livestock farmers,prnblem begins from the fact that he buys a cheaper mix which is supposed to meet statod nutrient Standards but he does not know its Interpretation in terms of their performa nces with the animals. His ma.jor piroblem is to maximise retums over food costs, especially in intensive live­ stock enterprises where food costs represent a large part of total costs. Increasing or docreosing the nutrient contents of a diet does not mean that proportional changos will take place in the intake of the animal or that the increase or decrease in cost of the diet may be more than offset by the increase»o? decrease in pcrformanoe. Tho fermer is thus mainly con- cerned about the influsnco of nutrient intake on animal performance. However, animal performance depends on a numbo.r of nutritional, genetic and environ­ mental factors. CHARTER THREE LEAST-COST RATION COMPOSITIONS EMPLOYED IN THE STUDY 3.1 The Basic Matrix Tabls 3.1 shows the various feed ingredients that are possible companents in a typical broiler starter/finisher ration, thcir prices per ton and their < 5 - T nutrient composition. Niheteen available ingredients were considered in this study. They include maize, guinea-com, and cassava which form the bulk of poultry rctions and •supply a high percentage of energy and reasonable amount of protein and amino-cvcids. They are readily availuble locally, but their supply is seasonal. The qualities and therefore their nutrient compositions also vary depending on the variety and efficiency of storage. Groundnut cake, blood meal, meat and bone scrap, fish meal and soyabean meal, supply the ma.jor part of protein and oil in the diet. Most of these have high peroentages of essential amino acids - lysine and methionine. The minerals are tc be supplied by Ad Vit. Most crf the Nigerian feedstuffs have besn evaluated for their nutrient campo- sitions. (20,62,63, 64). Qifferences in the estinates are uttributable to varietal differences, local Processing technioues and storage practices. How- ever, the coefficients used in this study were obtaineri from series of analysis carried out by Pfizer Livestock Feed and the Faculties of Agriculture in sorm; Nigerian Universitios. 3.2: He_s_tyictip_ns. in the, Model* There are tnree sets of restrictions: These are (a) the quantity of -*The basic matrix is used as refcrence point in this discussion. - 3-1 - TAflLE 3, H TVC LINEAR ffimRA>.MT>C frlSIC MATR1& | Gui­ 0. Slood Bkreors­ Bans ■Ohyaaltls U Folaolh CnUlic-- G Vo aBto ­ Mm Drloü Wheat Ad SUyn. nea nut Salt ns *Uooranl#> Ysaat Offslr Vt asnt Soys üea Ihnls ■»*" flo auerva corn Cokfe Waal Gir Um] nsa­ R ipl Scrsp — C, -220 -110 -70 -45 -60 -130 -600 —360 -224 - 56 - 73 -140 - 60 -ISO* W 1-1100 -265 -200 -320 PTyzar Priesa =2 -200 -120 -60 -40 -60 -600 -1300 -220 - 52 - 70 -1S3 - 44 -isc: -4130 -2SJ ---- i h ■ • i LB£PR Dl- LmEcRi- f STTEARP- SFIH E»fRl. E Quality ffslQht 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 %’00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Min. Flbr* 3.00 3.00 Flbrs 9.00 5.00 2.90 3.00 0.72 •6.0 1 t. 0.36 ,.S5 V 1.84 11.64 1.29 3.21 1.42 |>:.*TvrrvT pe— îcnge Prctsln 22 18 11.20 49.9C 73.50 28.50 64.83 31.46 10.70 44.80 16.28 60.00 60.00 9.52 31.20 Protein 26 22 11.20 49.90 75.50 28.50 64.00 äi.«e 10.70 44. 83 16.20 60.00 60.00 9.60 51.20 2.21 Energy (IC) koslu/kg. 2930 2000 3400 2833 2050 1646 2900 1914 1986 1850 2133 3360 2728 3624 Fat 4.9 2.5 3.4C 16,40 0.97 7.20 7.30 10.42 6.74 1.23 4.00 3,8.1 7.94 0.92 Fat 7.5 4.3 3.4(i 16.40 0.97 7.20 7.50 10.42 6.74 1.23 4.00 3.83 7.94 0.92 Cslciu* 3.6 0.8 0.07 0.47 0.63 0.34 32.60 36.24 S.Gfl 22.20 13.26 0.28 0.21 0.12 3. Cd 0.23 C.6S Calei ua 1.5 1.2 0.07 0.47 0.63 0.34 32.66 36.24 »■“ 13.26 0.28 0.21 0.12 C.Cö 0.23 0.60 fVn'>phc7rua • 0.6 0.4 0.36 0.67 0.22 0.56 17.42 0.08 3.31 17.90 7.17 0.74 1.34 1.82 0.30 0.78 0.13 ^ o s o h o r u a O.B 0.0 0.36 0.67 0.22 0.50 17.42 C.00 3.31 17.90 7.17 0.74 1.54 1.02 0.30 0.78 0.13 Lysins 1.20 0.9 0.36 2.03 7.22 1.47 S.ß5 2.33 0.77 3.36 0.67 99.00 0.43 2.74 0.jS7 Lyi l n a 1.30 1.30 0.36 2.03 7.22 1.47 5.65 2.33 0.77 3.56 0.67 99.00 C.43 2.74 O.Off Vatnlonina o.co 0.35 0.14 0.47 0.32 0.53 2.35 0.79 0.35 1.32 0.28 99.00 0.12 0.66 o.txa Vsthlonlna 0.60 0.5C 0.14 0.47 0.32 2.35 0.79 0.35 1.52 C.28 99.00 0.12 0.66 0.C43 Cystins 0.25 0.20 0.13 0.56 0.92 Ü.J2 1.20 0.29 0.33 0.51 0.13 0.69 - C/Stlna 0.45 0.35 0.15 0.5R 0.92 □ .ja 1.20 0.29 0.53 0.51 3.13 0.69 " tf In. Tryptophan 0.25 0.18 0.12 0.62 O.Sj Q,3B 0.65 0.21 0,78 3.07 0.67 0.022 Usm . Tryptophan 0.33 C.25 0.12 0.6* D.SJ 0.36 o.as 0.21 0.78 0.07 0.67 0.022 riFTDlETT fT-.Tr.T'̂TPfl Guinea Cnrn 0.65 0.73 OruunPnut CaKa 0.30 0.30 '•caphata 3.CJ3 Q.C3 1.00 |i uallty Salt 0.003 0.003 1.00 * Wml G O'n« Scrap Q.076 0.075 1.00 Ui«. Palm Kernel Uoal 0.15 0.15 1.00 Orlcd Ytait 0.25 0.04 1.00 «►aat offal4jR 0.05 0.07 ImMllty *4 Vit ̂ 1.00O.OCo 0.005 1.00 ’W a '.'.ilzo 0.73 0.30 * 1.00 Scvn̂|n 0.23 - 1.00 • 0.30 0.40 1.00 f l f f f f f f f-f f u s. - 22 - the total rnix of ingredients which is the weight constraint; (b) the ingredient level specifications and (c) the nutrient requirement specifica- tion. These constraints aro discussed fully below. 3.2.*: The weight Constraint The weight constraint fixes the quantity of the mnix within which the nutrient requirements and the ingredient specifications must be met. The model was constrained to produce sxactly one tonne of the mix. 3.2.2: The Ingredient Constraints Cortain restrictions were placed on the levels of inclusion of availabl ingredients to conform with proved nutritional requirements and to tako account of their availability and costs. (a) Energy Oasod Ingredients: Guinea-corn and/or maize for instanco should not oxceed 71 pcrcent of the mix because the birds should get limited energy supply from carbohydrata. Besides, therc. is high demand for these ingredients for human consumption and due to weather uncertainties, yields fluctuate and occasional shortages occur in supply. These render the grains very costly components of livestock feeds. Cassava is thus being included in the mix up to the 30 pereant level in the Starter and 40 percent level in the finishcr rations. This is not to say that cassava is cheaper nor is it more readily availablc but it is being tosted as a close substitute so that at least, produaers have other choices whc-'i ‘:1t is absolutely impossible to use the grair.s. Bcsidos. th -.ountry is trying hard to proserve its foreign exchange and very soar. may pT"-e r ’-trictions on the importation of maize. - 23- - Alna, the fact that maize is being imported at s chsaper pries now is what makes casssva sasm costly. (b) Protein Gesjjd Ŝ>3redicpts: The mnriel was structured to ensure a balance bstwnon prutein supply from both plant and animal sources since this maintains a balance in the proportion vegetable and animal protein. This was acccmplishcd in the modal by setting the maximurr levcl of groundnut cake at 30 percent vfhilst that nf fish meal was at six parcent. Blood meal was limited to IG percent becauza of its low digestibility and the unbalanced nature af its supply of amiidj -acids. Palm and its rsported non-palatability for chickens at high levels. Meat and bone meal was limited to the level of 0.75 percent because of its very unbale.nced calcium-phosphorus ratio which could upset mineral balance if ailowed to enter the mix at high levels. (c) Minsrel and Vitamin _Dased Ingredients; Allowance was made for the non--availab: phusphorus from grains and vegetoble sourcos by insuring that tha rs ntain reascnable parcentages of oystor Shell or mcat and bone scrap or borie meal to provide rsadily available calcium and phosphorus. Their maximum levels respectively are 3, 7.5 and 3.5 percent. To improvz the palatabilitv of the dist and provide sufficient minerals and Vitamins, the modcl spscified that salt arid Ad Vitamins should be 0.3 and 0.3 percent respectively for the Starters. 0.5 porcent level of Ad Vitamin was specifisd for the finishers. - 24 - Columns EJ. and B0 of Table 3.1 Show the restrictions specified in the I c . basic matrix for the inclusions of various ingredients in thn Starters and finishers respoctively. 3.2.3: Nutritional Restrictions The nutrient content and intake of the rations .3 the performance of intensive livestock entarprises.—5/ (a) Energy: The anergy lovel of tho foed influances the rate of live- vveight gains —33/ . The largest single dietary need of an animal is the energy source for physiological processes such as movement, rnspiration, absorption of nutrionts, circulation, exeretion, nervous System, temperature regulation and reproduction. The feed intal44 12,519 23.2 2.97 2,528 1.52 K . : P . 12,432 21.1 3.23 2,211 5.10 11,286 10.1 3.49 1,396 5.46 10,344 15. 1 3.74 j Source: Tcble was obridcod fror, "Poultry Production by Card and Naishoim . Ficivrcs arc csrnvortad tc metric. The above results in Table 3.2 confirci the conclusiona that feeds with nigher enargy content have better fer-o corrworsion efficicnc'*. Thercfore. for the starrer ration, a minimutr. onargv lavel of 3,000 kcals. per kg. was specifiod. - 2 6 - For the finisher ration, minirnum levels of 2,90C and 2,000 kcals./kg. were specified. Low encrgy leuels havs been specified benause of existing high environmental temperatures. The mear. tempcroture being about 08 F. It also helps to reduce the cost of thc ration since the encrgy-providing ingrediEnts are the most costly. • / £ * (b) F’roteins: The protem nceds of birds should be de?ermined in terms of supplying adequate quantitative and proper balance y r t h e essential amino- acids from the main protein constituents - fish meal, groundnut cakc, soya- bean meal, blood mcol, etc. Typical broiler rations were seid to contnin 22-24 percent protein in the U.K. (Card and fJeishlm) Rao fed 18, 22 and 26 percent protein rations on pure bred broilor chicks from 0-10 wceks and finished them with 19 and 22 percent protein rations. He found body weights to increase with increase in protein lovel but ths incroases were not significantly different. He thorefore recommcnded 22-24 percent protein in the Starter ration and 19 per- cent protein » M T i‘rinisher ration. Flinn .so fed rations with different protein levels on broiler birds an-; found no significant differences in thc weight gains using 20 and 22 perceant protein levels in the finisher rations. For the purposes of thia study, two levels of protein, 24 and 26 per­ cent were specified for the Starter rations whilst 20 and 22 percent were chosen for the finisher rations in accordance with the works done on protein in Nigeria. £ f & - 2 6 - For the finisher ration, minirnun levels of 2,90C and 2,000 kcals./kg. viere spccified. Law cncrgy leucls havs been specified because of existing high environmental temperaturcs. The mear. temporature being about 00°F. It also helps to reduce the cost of thc ration since the enorgy-providing ingredients are the most costly. (b) Proteins: The protein ncods of birds should be dcteDrrmm:ined in terms of supplying adequate quantitative and proper balance of the essential omino- acids from the mein protein constituents - fish mecl, groundnut cake, soya- bean meal, blood meal, etc. Typinal broiler rations were seid to contnin 22-24 percent protein in the U.K. (Card and Neishim) Rio iFed 18, 22 and 26 porcont protein rations on pure brod broiler chicks from 0-1C weeks and finished them with 19 and 22 percont protein rations5,. Hs f ouritd body waights to increase with increaso in protein lovel but the increases wers not significantly different. Hc thorefore recomrr.onded 22-24 percent protein in the Starter ration and 19 per­ cent protein in^ha finisher ration. Füinn also fed retions with different protein levels on broiler birds and found no significant differences in thc weight gains using 20 and 22 percent protein levels in the finisher rations. For the purposes of this study, two levels of protein, 24 and 26 per­ cent were specified for the starter rations whilst 20 and 22 percent were choscn for the finisher rations in accordance with the works done on protein in Nigeria. % & & & (c) Minerals: Calcium and phosphorus ov- thc most important minerals in chicks nutrition. Net only nust birds receivo ndequate amounts, but they must consume them in suitable proportions. The calcium phosphorus retio shnuld hc v/ithin the ränge cf 1.5:1. For both thc Starters and finishers, a minimum levnl of 1.5 and 1 were specified respectively. For phosphorus, the ranga was 0.4 - ü.8 percent. Reauiremsnts are givor. in saf. ty nargins duc bo losses during Proces­ sing and storngs which rcsult in variations inStha nutrient contonts of feed. Belcw is a tablc showing racommondsd levcl :hcsc minerals: XA§t§. Ü l?. : Minnrnl- requireinent,, Minerals ; Gterting Chicks j Growing Chicks sne! Broilers • and Broilers ^ ----------------------1 --------------------- Calcium % «\ ^ ; i.q ! o.a < v I : Phosphorus -/-> 0.5 0.5 (Auctilable ) Source: Tablc was compiled from "Paulcry Production” by Card and Neishoini and "Puultry focdi.og and Management in the Tropics" by M. L. icntt '■°e Other minerals such as sodium, petassium, magnesium, chlarine, iodine, Iron, manganese, copper, molybder.ium ’inc. solenium and cobalt are required only as constituents of vitetmiri Birds eannot synthosisc vitamin B,p using an inorganic source of cobalt. Calcium, phosphorus, sodium, rhlorine and potassium are the major elements required in large qunntities. Calcium is abuut nne pereant of the diet and tho rost are needod in traoe3. Lack of calcium, phosphorus anc' magnesium rcsults in poor mincrali-- sation of the bone. The murked deformity nf the Skeleton may occur to ciivc rickets. Phosphorus is essential in anargy mctabclism, it is a canstituent of nucleic acids and activity of scveral enzyne systams. Calcium is important in blood clotting and musclc contraction. Deficiency in any of the minerals leads to poor growth and body dehydra o (d) Amino—acids: The intake of nutrients such as the essential amino- acids - lysine, cystine, methionijno, tryptophaie, affnet the growth rate of animals. Protein synthesis hie birds requircs that all amino-acids noeded t O to make up tho protein lyc pljesent in tho body at riaarly the same time. When an essential amine-aeid is ebsar.t, nc protein is synthesized at all. Carcasses of animals fod roiticr*ö doficiant in amino-acids usunlly conto in more fat. Amir acids 0£ficicncy-^ilsr rcsults in poor Fsathering. The so called essential amino-acidssare those whioh cannot bo synthesizod in the body and therefore havo to. ,bbo prouioed in tho diet. Lysine, cystine, methienine and tryptophane are the criticsl amino-acid.- oecause they are difficilt to supply in proper amounts fron feed preteins. Cystine can hewsver, bo synthesized from methio- nine. Methionine supplemcntction of broiler diets at varying protein lcvels was found to improve the efficiaicy of feed conuersion at all ages up to ton wecks but less markodly with advancing age. & However, with female chicks, growth was stimulated to a limited axtont only aff four w:d of diet •: cf_ lner. s Diet, A r Lysine 5 1.13 4 1.02 Mcthic rüne 2 3.47 2 0.41 Cystine 1.5 0.35 1.6 0.31 Tryptophan i • -i 0.23 « j* J.21 Applicable Protein 23.3 20.5 lcvel Source; Table was abridged from "Poultry .Productic,-,'* hy E. Card and Maiden C. Noisheim (11J 30 - The 'ablas whict follow indicate aimilar ret -nmendations as abovc. AtiinsrflßM jm1.5 21.0 19.5 19.0 . lao-acids as Perc ine of the Diets Lysine 5 1.125 1.075 1.05 1.025 0.975 0.95 Mothion'ine 2 0.42 0.41 0.39 0.38 Oystine 1.6 0.3)5 ' D.335 ’J. 320 0.31 C. 3.15 Tryptophan 1.0 Ü.22E 0.21 0.205 0.195 0.19 Source;: Table was nbridg^d prom "Nutriant rnguiremmt of chickens around ' the World" by M. L. Scott («1 TABLE 3.4.3: $ S p I L E 0 3 Amino-Acid e •.v.-cks c.-O wea!c3 4 1 - Lysine QS) ^ 1.2:3 1.10 Methionine ('/aj 0.06 .'.75 Or Methicnlne (‘̂)+ 0.46 G. 40 Oystine (p) 0.40 0.35 Tryptophan 0.23 0.2G Source: Table was abridged from Table 1 of pp. 203 in "Poultry Feeding and Management in the ' rnpies" b> f.l.L. Scott in "Animal Production in the Tropics"’ c.dited by LogsJjL, Oyenuga and Babatui ide» - 31 - From the abcve recommondations, a rangs of 1.25 ~ 1.85 percent was stated for lysine, C.5 - 1.0 pereant for methionine, D.3 - 0.4 percent for tryptophan for the startt rationa. For the finisbers, a ränge of 0.10 - 9.25 percent was specified For tryptophanQp.20-9.45 percent for cystine, ü.5 - 0.8 percent for methionin, 0.9 - 1.3 percent for lysine. The model also made allowance for the Provision of lysino and methionine from synthetic sourecs lf necessary. (e) Fihrej_ Birds cannet tolorate high lcvol of fibro in their diets. High fibro content rcsults in lower nutrient intake and digestibility of the proteins. This reduecs the extent of utilization of the associoted amino-acids. The ränge specified for both Starters and finishers was between 3 and 5 percent. (f) Fat: Fats ere potent sourecs of onergy. They previde 2-3 times metabolizable enorgy a/^rains. Rations high Oy- . cakv up and rio not flow readily. The usual practical limi ß~ß p reeni' of the diet. With special technologies pel- leted diets may heve ?-ö perceni jf added fat .11/ Partinularly with the inclusion of oassava up to th 4J oercant levci, the. ränge specified was betwoon SK 5 *• 5.Ü percent for both the etortor and the finishcr rations. This reduces dryness of the feori. Fats not only serve as energy sources, but they also supnly tho essen­ tial fatty acids particularly linoleic e.cid which is very important in chick growth. The fat content of tho birds diets offect their growth rate particulai the sourcos nf the fat which determine its availability of use to the birds. - 32 - Fats with unsaturated fatty-acids may not ba stähle, which leads to oxidative rancidity and thus destruction nf vitamins A. D and E. It is however very important that chicks get considarable nart of their energy sapply from fat and not mainly from curbohydratc. 3.2.4: Concluding Remarks Nutrient doficiencies cari rasult ir. high, mortality rate, reduced livestock pcrformance and consequently loy.wsi laovdd%l, of profit. It is therefore very important to provide all nutritional factors when formulating diets. The most important classes of nutrients which are crucial to chicks are the protein and essential amino-acids, enorgy lsvel, fibre, mineral, fat ond .vitamin conto: The exact quantitative lcvels of eoch of those nutrients which will give moximum growth rate ar; no+- knnwn but roasonable reliable quantitative dato havo been nbtainod from exijorimmts. This is the reason why the model speci- fied o rc.nge of valu^s, 'minimen, and maximum levcls for the nutrients in Tablc 3.1. h3: Soluti.no anr iL3cu.3Si.Tn of the Model Usiiag the nutritional .uiu Lngrodient constraints specified in Table 3.1, Table 3.5 shows the cost and cn Position of the starter and finisher rations excluding cassava fror» the ing: -dient base. ’.Vifhin the framework of tho nutritional requiraments, the lutri. nt composition of tha ingredients and the cost per unit of oach ingrediGnts spocificd, the rations have been compounded at the least costs of 'lö'l.S? and i'175.29 for ths starter and finisher rations respuctivcly. - .33 - TABLE 3.5: Costs and cr--~ vdLtloris of thö Starter rnd finishör rations xcludinq uasaava - 34 - The finisher ration as expectodf is rh^aper than the Starter ration, by (v16.38 per ton. The protain lavols for the Starter and finisher diets are 24 and 21 percent respectively. Calcium phosphorus ratin was 1.2:1 sind 1:1 in the Starter and finisher respectively. Lysi.nc leval was 1.25 for the startai and 1.3Q for the finisher Methionine-cystine nnmbination was 0.78 for the Starter and 0.79 for the finisher. In maeting the netritional specifications at least-cost, the Starter ration comos up with a higher onergy ration of 2900 metabolizable cnergy (l«l,E.) kcals./kg. than the finisher which has only 2300 M.E. kcals./kg. The diffcrences in composition lin in the fact that the Starter ration includes groundnut cako and ment Ctny^urie scrap which \re excluded in the mix for the finisher ration. On the other hand, the finisher includes bone rneal, oyster Shell and dried yoast whilst the ntartor ration excludes them. They both had methionine supplementatior from rhe synthetic source. These lsvels of nutrients and ingredisnts r.icot tne specified Issels of requirements for chicks. (See Tablq 3.1). In Tniaiiik.3.£ below, the. cffect cf different pries ievels (C-, and C2 ) on the composition an: cc c of starrer ration, is examined. The first set of costs G,, represents t. >vorogc feed prrJucers costs of ingredients obtained fram Pfizer ar.d Ulnoq.in nonmercial sd mixars in September 1976. The seconc. sot of costs. C,-., is that of the University of Ibadcn Teaching and Research Farm os of Juno 1977. Table 3.S shows that the cost of the feed increased by 17.4 pcrcont duc to the increase in prices of ingredients from set to set C2 * The commercial farm’s prices are lower than U.I. prices - 35 TABLE 3 . 6 t Effrct of cbanglng costs i. tcrcnirnti or thc pcrcentagc ccr.ipooitinn and cnats of stt~rt«r ration, average produccr roste, 1976 priccs versus U.I. costa, 1977 priccs Ingredionts an: Nutrients Av. Producer U.I. Costs Percsntrge Oosts. r Change- BrewerTs gre ins MaiZo Rice Pran Blood Mli.1 Sayabe t Syi -.th.'tfid Ölethionine Moat and 3onc Groundnut Cake Salt V.'heat Offals Ad Vit. Oried Yeast Gone Woal Oyster Shell Guinea norn - 50.34 Gest per ton of mix 1 iü'. 331 s‘167. 309 17.4 Protolr-’ a - 26 Ftit 5.ß 5.0 - Fibre & i s. it 4. o 9.5 Energy •0.54 2854.70 48.6 niwthir^iine C.5' 0.50 - Lysine a nti • i-».- 1.25 ~ Tryptophane • • r ~i- * 0.31 24 Phos;pphh arus ■J t **• V . C*.-nj - Calci;uirurin . 5 1.00 100.0 U/st:iinne 0.31 0.38 { 12.4 for cortain fecd inrrredients because thcy obtoin them directly from suppliers «hereas U.I. obtair.s its supply frem contractors. In addition, thare has been an increase in the general prine levol within the period af September 197S and June 1977. The tryptnphan, calcium and cystine contunts incroased with 1977 prices by 24,130 and 12.4 pcccent respoctivoly. The conposition cof the ration under the two price structures showo markod Variation. Guinea-corn replncod mair.e v complctely in ration whilst rice.bran and blaud meal were oxeluded. In ration C-, also, dried yoast, bone meal and q^ster Shell werc complctely oxeluded whilst soyabcan, synthetic me Ine mejat and bone meal and ground- nut cake were changed by 174, 28.7, 6dv Gr>d 03.3 percent respectively. The fibre and energy levcls also changobl by 9.6 and 48.5 pcrccnt respectively. 3.3.1 3ja_b_ility of the Mix c" Inqrodients to changes iji Their Prices in the Startür Ration Thero is a ränge of prices mithin which each of the ingredients in the mix renoin in thie solution. Outside this ränge, that is, if the price of the ingredient below the minimum prine er rises abeve the maximum price stated, then it might. bo 1 otr.lly substituted for or its quantity decreased or increased thus changirrj the Optimum mix. Table 3.7 bol ~w shev/s the lower and upper price ränge at which each of the ingrodients ramain in the solutirn. The opportunity costs of each of the cxcluded ingredients rro shown also. The opportunity cost is the penalty for including a unit of tho oxeluded ingredient in the solutien. Ingredients with equality constraints have no price effect on the stability of the mix. IVheat - 37 - TABLE 3.7: otabildfcv _gf_tfac ^ i * .gf .Jjraggctfettts tt&shangcs- i n^hgir;_priccs in.startcrJratdgns - — — Gange Over Which Price per Solutions Remein StableIncluded Ingredients Amount ton % . A1 *4 Lower V* Dried Yeast 1.03 150 1B6 Soyabean 30.00 150 - Groundnut Cake 3.06 200 / 102 225 Meat and Bone Scrap 1.32 250 s V'249 287 Bone Meal 0.70 60 . S / 12 33 Brewer's Grains 5.00 4 Q ^, ' 194 - Oyster Shell 2.49 13S S 96 229 Salt 0.30 - - WhcatOffals 5.00 7L - Ad Vit 0.60 | 1355 — - Guinea-corn 50.34 > V A 220 207 257 Syn. Methionine G. 15 \ 1980 — 2933 Excluded Ingredienvtts Opportunity Costs (i.) Fish Meal 442.89 Ptaize & 16. 02 Syn. Lysine 2323.22 Dried Milk Powde»' 104.70 Blood Meal 32.19 Di-Calcium Phosphate 331.59 - 38 - offels, brcwar's grains and soyabear cnulrf takc up einy price abovc 70, 194 and 55 naira respectiv/ely. Also tho cpportunity cost of increasing the maizs contcnt of thn mix for example, by one oei-cent is (.'15.02. This is to say that the cust of one tunne of feed will increase fron ('187.31 to >7202.33. The compoeition and costs of the experimental diets aro discussed in CHARTER FOIJfi FEEDING TRIALS 4.1 Intr.iy iutdün It is one thing for the c. Tnnuterised diets to be least-cost, but it is another thing for them to sntisfy thr orofit maximisation objective of the fanner. For instance, the diets may bs ton bulky(fibrous] and undigestibls for the birds 30 that very little of the nutrients is absorbed. Also, the diets mny be impalatable such that feed intnke is considerably reduced. Ferd3ng trials were therefore carrieb out 4n Order tc compare the birds1 performr' -3 on the forroulated .diets with the existing commercial diets. For the purpose of this study, Spinal perfornanoe refers to the feed- grnwth relotionship. Cercas3 chfcracSeristics of broilers ars not highly rated in this society and ns such, they havs not been oonsidered in the study. 4.2 Experimental Diets 4.2.1: Starter _Diets_ for Experiments I and II Eight Starter diets were formulated using varying protein (24 and 2f percent), fibre (3 and 5 percentJ, and cassava (C and 5 percent) specifi- cations. The nutrient composition and costs of the different Starter diets are shov/n in Table 4.1. Protain and fibre lavels are known to he major factors affecting the performance of birds. High fibrous diets uro too bulky and intolerable for birds. Besides, high fibre content. renders tha feed undigestable and hinders the absorptiun and utilisatiun cf tht. important nutrients. The levels of TABLE 4.1: PErTENTAfE COMPOSITCMH CF CDMPJTERI5ED OTARTEn OIET Ingredlents and Diet 1 Ölet 2 Ölet 3 Oiet 4 Diet 5 Diet 6 Diet 7 Oiet 6 Nutrients 1. Wheat offals 3.22 4.44 3.00 5.00 .uu 5.00 5.00 2. Soyabean 30.00 30.00 30.00 30.00 30.00 30.00 30.00 3. Malze 56.56 50.94 53.00 45.90 20.64 23.60 4. Bane Me&l 0.91 0.64 0.9B 1.10 0.60 1.20 0.70 5. Syn. Methionine 0.16 0.12 1.17 0.12 0V. ) 0.15 0.19 0.15 i 6. Brewer*s grein 2.47 4.56 2.00 ß.QÜ 6.00 6.00 6.00 7. Meat & Bone Screp 1.75 1.90 2.50 3.20 2.20 2.76 1.00 1.30 I 0. Salt 0.30 *0.30' 0.30 0.30 Q. 30 0.30 0.30 0.30 9. Oyster Shell 1.20 1.43 0.60 1.20 1.16 1.30 1.30 1.50 10. Ad. Vit. 0.60 0.60 0.60 0.60 0.60 0.60 0.60 11. Dried Yeast 2.83 5.00 1.70 3.00 0.70 1.03 12. Groundnut Cake 3.44 < ; > 2.10 2.90 3.10 13. Cassava 5.00 5.00 5.00 5.00 14. Blood Meal 0.30 ö 15. Syn. Lyslne 0.07 0.00 16. Guinea carn 20.20 45.00 30.00 50.30 Cost per.tan *213.36 *2 CS. 09 *266.60 *262.67 *253.60 *241.60 *199.34 *187.01 1. Protein 26 24 26 24 26 24 26 2. Fibre C g ? 3.16 3.16 3.16 4.26 4.66 4.96 4.76 3. Fat 5 J 5 5 5.0 5.0 5.0 5.0 5.0 9 r 4. Calcium .20 1.20 1.20 1.20 1.2C 1.20 1.20 1.20 5. Lysine 1.25 1.36 1.25 1.42 1.25 1.25 1.25 1.25 6. Methionine 0.50 0.5 0.5 0.5 0.5 0.5 0.5 0.50 7. Phospherus 0.6 0.0 0 .6 0.0 0 0 0.0 0.6 0.00 6. Cystins 0.32 0.34 0.3 0.33 0.32 0.34 0.33 0.35 9. Tryptophan 0.27 0.30 0.27 0.29 0.27 0.30 0 .20 0.31 10. Energy 3202.0 3300.7 3274.0 3285.7 3139.1 3007.5 3120 3002 - 41 - protein and fibre chosen are such that they brocket the adequate ronges which may influence the performance of the birds on the various diets. Cassava flour is being tested as a substitute for the groina. The effects of the varying protein, fibre and cassava levels are discussed below. Diets with higher protein levels are less costly than diets with the lower protein levels. This is due to the fact that changes in the Optimum mix of ingredients occura only among the energy providing ingredients which are the costly components. Also, in all cases, there was greater need to use synthetic methioniite (very costly component) in the mix to attain the specified level at lower protein levels than at higher levels. Groundnut cake is included in the mix only with feeds containing higher protein levels but in diets with lower protesin levels, only soyabean is included and because it is cheaper than groundnut c.ake, it is included at the maximum level specified before groundnut cake comes into the mix. Diets including cassava in the mix as substitute for part of the groins are mora costly than their counterparts without cassava. For instance, diet 3 costs f1266.68 per tonne whereas diet 1 costs only {1213.36 per tonne. Diet 3 Substitutes 5 percent cassava for only 3.56 percent maize in diet 1 and the cost difference is *153.32 per tonne. The difference between the cost of maize and cassava is {.70 per tonne. The fibre content specified affects substitutior in the ingredients which havu high. : ibre contents. 3uch affected ingredients are wheat offals and dried brewers groins. These ingredients came in at their maximum levels in the diets with higher fibre content. Other mineral based and vitamin supplying ingredients come into the mix at the various levels so - 42 - as ta make up far the deficiencies of the fibrous, protein and carbohydrate ingredients included in the Optimum mix. 4.2.2: Finisher Diets Two finisher diets were formulated for the second expariment. Their costs and campositions are shown in Table 4.2. Two protein levels, 20 and 22 percent were specified. It is pertinent to note that these formulations were clone mainly to test the two protein lev/els in the finisher diets. Maize was however totally excluded from the Optimum mix due to the fact that it is costlier than guinea-com and it is not superior to guinea-com in terms of nutrient composition. The major difference in the Optimum mix is that diet 10 which has the lower protein level includes blood meel wnich is excluded in diet 9. Also, diet 9 includes synthetic lysine whereas diet iSf excludes it. Diet 10 is however cheaper than diet 9. This oonfarmu ’tc expectations since the diet with a lower protein level should bs cheaper'. 4.3 The Model The net revenue to the poultiy farmer depends mainly an the rate at which tha feed is converted infcc .Hveweight gains, the quantity of feed required per unit liveweight guin (feed conversion efficiency), and the Optimum market weight consistent with the Optimum profit objective. Growth rate as mentioned before is a function of various factors such as: i - 43 TABLE 4.2: Ccirxia^hLan und. 'nnst„Qf. iihe_tva.:£inl5hgr .djfctajrttthout JiQScava Ingredients & Nutrients Diet 9 Diet 10 Guinea-com 63.36 Groundnut cake 15.70 - 1 . 3 9 Brewer's grein 7.52 8.00 Whcatoffals 5.42 ^ ) 7.00 Blood meal 5.64 Dried Yeast 4.00 Oyster Shell y i . 4 ö 1.56 Bone meal ^ Q j r 1.4 4 1.45 Synthetic Methionine 0.03 0.44 Ad. Vit. 0.50 0.50 Salt 0.30 0.30 Synthetic Lysine 0.26 Cost per tonne of mix «203.23 «187.15 Proteirv^x Fib^EL 4.50 4.60 Lysint 1.2£ 1.10 Fat 3.65 3.90 Calcium 0.80 0.80 Cystine 0.30 0.26 Methionine 0.50 0.50 Tryptophan 0.23 0.23 Phosphorus 0.40 0.40 Energy 2882.38 2906.52 - 44 - i ) the quality of the diet, ü ) the genetic characteristics of the experimental birds, iii) the environment which involves experimental conditions such as temperature, space (stocking density) and Ventilation, iv ) management, which involves the feeding watering, sanitation practices and state of health of the birds. A In the feeding trials, all the above factors were kept con stant except the quality of the diets. In symbolic tenns, the relationship between weight gain, feed intake aed other factors affecting growth can be expressed as follows: X , X , ...X - f ' .̂Xn- XrM-1* T J f f V V l ' . p’ p+1* q where, G = growth rate 1 n nutritional factors which determine the quality of the diet n+1 ....Xm genetic characteristJfes “ V , ....XP ....xq Management. V f Through experimental design, X^ ^ .X could be assuned constant, and the weight gain-^. s. e. d. intake model reduceä to G = f(X^, ..... , Xn). Feeding s were carried out to establish in quantitative terms the true V relationshitp of the above model for the various comcuterizäd and canmercial diets I and II. 4.4 Experimental Design and Setting The design of axperiment used is a randomised block layout. One of the two most important conditions in a randomised design is that there should - 45 - be two or more treatment levels which may differ either qualitatively or quantitatively. The second condition is that the experimental subjects should be randamly assigned to the treatment levels. Thia particular design has a very useful advantage in that it is not campulsory far the number crf birds to be equel under each di6t. This applies to caaea where occasional deaths occur or in case of disease outbreaks. However, the same nunber of birds were randomly assigned to each diet initially. Each replicate had the same floor space (1 .8 sq ft. per bird) and all birds received the same medicatian throughout the duretian of the . Birds were fed and watered ad libit _ , s. At the end of each week, group measurements of the feed incake and weight gains were taken. Group measurements had to be taken because wing tagg were not available to differentiate one bird from the other. Also, because the deep litter System of management was used, it was not possible to obtain individual Birds and fcheir Management Prior to the \.rrival of chicks, c^re vas taken tn ensure that all equipment were clean and in good operating condition. Heating was done at least 24 hours before chicks arrived especially when a new consignment arrived during the wet season. Adequate feeders and waterers wtre provideri for the birds. Waterers were filled n few hours ahead of arrival so the water was at room temperature when the chicks arrived. Antibiotics and Vitamins were administered in the drinking water for the first one week to give tht. chicks a good Start in life. Overhead lights were provided for the first f9w days when chicks arrived in Order to help the chicks find feed and water. Thereafter, only the attrec- tion lights under the brocdar boxes were prcvided. This vas to prevent crowding and straying. Feed and water were provided at all times. The birds were vaccinated as advised by the veterinary doctor. At day old-intre-occular vaccine against new castle- asft'<± at 4 wwskc muscölpr.vodpine -anjiin^ife^/IJoulfpox-.wc:TT'p-!jivpn to the b ird s. A fte r Clearing o ff Gcch b$tch of bj-rHs, the-î p1 *so wae^t^orp^ighly cleened and disinfectod. 4.6 Limitation uf the Experiments Three important things oroved to be limiting in the oonduct of the experiments:- 4.6.1: Feed Ingredients -eed ingredients were not stored under optimum conditions so that sametimes they have beer, vnrustöd by wcovils. This affects the nutriant contents of the ingredients. Feed samplus wäre however taken for analysis and were found ta contaii i the atatad proteirt, fibre and fat contents for each of the diet3c A 4.6.2: Houaing Ideal 1'',m «ftnr-wi-'h bntr.h nf birds, ̂ the .hpusc should be thoroughly washed and disinfected all over. The house should then be left for at least four weeks before ■earing another batch in the same house. Unfortunately, this was not po3Sible on the Teaching and Research Form in view of the numerous research pro iects beina carried out and the pressure for research time and - 4 7 - space by every user. The resting period in these axperiments was two weeks after thorough cleaning and disinfection. Despite these thorough cleaning the continuous cropping in the houses resulted in e build up of diseases in the houses and these were passed on to each new flock reared in the houses. It was therefore very difficult to prevent incidence of diseases in the experimental birds. However, the effects were reduced to a minimum by administering drugs es preventive and curative measures. Post-mortem examinations were done as soon as deatths occurrecd to deter- mine the cause of the deaths. 4.6.3: Methods of Measurements The problem here as mentioned earlier is that birds could not be weighed individually because there were no wing tags to identify them. Group weighing has been practised in variotis studies by Cent,-1—55/ Heady 27/ . \ / and Dillon — • and the»/ have shown that the method does not adversely affect the results, but rather helps in the regression analysis by eliminating certain Statistical Problems like autocorrelation even though it results in the loss of some degrees of freedom. 4.7: Experiments and Performance Comparisons 4.7.1: Experiment 1: Comparlson of Eight Conrputerised Diets of Varying Protein, Fibre and Cassave Contents The compositions of eight Starter diets tested were earlier presented as Table 3.0. The composition of the two cormercial diets tested as control are not availeble because the feed compounders neuer want to reveal their formulae. The labels found on tho v/arious types of commercial feed only Show the ingredients '.ised and the proximate analysis of the percentage levels of protein, fibre and fat in the feed. The ab^ective of this experiment was to compare the performance of the birds fed each of the Starter diets with varying levels of protein,fibre anc - m - cassava. Two protein levels - {24 and 25 percent), two fibre levels - (3 and 5 percent) and cassava levels - (0 and 5 percent) were the different combinations of the sight diets. These diets hed the following levels of the different components: Diets Protein Fibre (1)’ 24 3 (2) 24 -i 0 ) 2 4 5 (4) 24 5 (5) 26 3 (6) 26 3 (7) 26 (8) 26 J (a) Breed and Stocking ! 'fj * The pcrticu^r breud - Ln thv- >.p:;*'i:r7'nt is the Cobb broiler. The Cobb is a dominantly weit. „ f Sered L:lr .• K t ras b-en bred specially for broiler cuplitivs l't »fT^oted for i: - :.if- grovt': rate, early featheqring and gooc: feed conversicriNraie. However- ‘.v umbs very easily to disease outbreaks. Six hundred u’ay old broiler cbicku wera randoraly distributed to ten pens such that nen nontcinsd 6C birds. Each pen was r .ndomly allocated to each of the cen axperinenial distr (eight oomputerised and two cammercial diets >» ' (b) Mortality A recorcl of daily mortality was kept for each of the pens throughout the experimental psriod. Mortality rote «s nine percent in the whole flock even though the highest numoer of deaths occ.urred in only two of the pens. - 49 (c) Performance Comparison of Eight Starter and Twg Conriercial Pieta« The expe­ rimental * results are presented below: Records of feed intakt and liveweight gain are shown in Appendix A, Figures in Table 4.3 below are mean values for the various treatments. The t-statistic is used to test for significant differences among the means. Table 4.3 summarises the performance comparisons of the birds an the Starter diets based on efficiency yardsticks of average weight gain, average feed intake for the period of six weeks and the efficiency of feed conversion. Column two of the table Shaws that there were significant [PjC 0.01) differences between the average total gained by the birds an the various diets used. Fron column three an - ̂ \ cer be deduced that commercial II starter required the least ~uar^&y of fcod to prt on one kilogrom liveweight and the t-test shows that there •vom sinni^ioanf (P-4! O.Ol) differences in the feed intake and feed convars'on efficiency. Diets 3, 4, 5 and 6 contsined five pereant leve.l of oassevn and they hav'e beer» consumed in the same manner as the other diets^ Diet 5 perfornoo best nmung che eight computerised diets in terms of fê ti efficiency (3. '.‘j •nri this w-s -ignificantly (P ,£ O.Ol) different from the value? r“rordert for the other diets. 4.7.2: Experiment H : Selecting the best erir of Computerised Starter and Finisher Diets The mein objective of this experiw-nt was to test starter and finisher diets concurrently so that it would be possible to select the best pair in terms of least-cost and best response. «r 50 - TABLE 4.3: Pcrfomancc companincn of tcn Starter nliots Diets erc as listod in Tabl.a ä. 1 abovs. '"^Significant nt P 0.01 - 51 - Duc to lack of space anr! funds, only two Starter diets and two finisher diets warn tested with two conmercial diets as control. Two Computer!seri stnrter diets 2 and 7 of Tr.ble R.3 vvere selected frnm the first Experiment because they had lower ccjsvs Irtan the re st ancJ performed best among the diets without cassaya. The compcp jSlticns of the two camputerised finisher diets are shewn in Tablr: d. 2. fal Breed and Stockincj Rate — - ™ — -— — \ y The samc breed cf broilers 'was used as in axperiment 1. Three hundred day-old broiler chicks werc randomly riistributed into 1? pens such that each pen had 25 birds. The two chor.cn camputerised Starter diets v/ere allocated to eight of the pens such that ' ■ four replicates. T‘u: two oommercial diets werc allocated to the ranajmni. feur pens such t!-)at each had two replicates. This is also <£■ ls*cl< crf = pr.ee arid funds. After six weeks the birds were transferrari to the- ccrnpji-' >r* ?. : and commerciol finisher diets and roared to the age of 12 weeks. There werc Four replio.’tes of th~ Starter and finisher rctions. The .•.hart • ;iw sh. Tables 4.4 and 4.5 Show the performanca camperison of tho Stertor and finishrar diets, respectively. Analysis of variance test~ werra performed on columns 2-4 of Tablo 4.4 (see ANOVA tables 4.4.1-4.4.3) which show that there were 3ignificant (P ^-0.05) differmees bctweeri the avergga weight gains, average fcerl intake and feed convorsion officiency for thn airds or ;nr;h diGt. To detect the treatment causing thoso differencos C’itorion uf lsrf was used. Using the criterion of weight gains5,, signil ; oa.f u if fu rra n c G E (P-^»'l.ül) wer*/ shown between dicts 2 and 7t IX a n d ' ü’4nd . II ; -ir !;o deduced therefore that diet 7 was the best out of the sw.rir <>'r\ - in terms of weight gains. With feed intake, howevor, signifikant (P . Ü1) difforences were prominent in diets 2 and 7, 2 and H , I and II and I and ?. Diets 2 and T and 7 and I were not consumed significantly morr thon .-,rch other. Thra most acceptable +lsd means least significont difference. It is derived stati3tically using the formula iy J 2 | , t^^ (n-l)(k-l) d.f. where r = number of observations per moan n = number of repl’icatcs, k <= number of treatments S, 2 = error variance K = level of significanc.o Differences between two peirs of means which are greater than d are significont. TABLE 4.4: Prjrfomcncn Conparison of thc ntart.cr — — TABLE 4.4.1; jtubli r er weicht gains # ^SigniFicant at P ̂ 0.01 - 55 - starter diet to the birris was therofore conrnercial II. In terms of feed conversion efficiency, significant (P n.Ql) difforances occurred betwesn diets 2 and H , I and II and 7 and II. The best dict was Starter diet 7, followed by diet 2 whinh v/arc not significantly (P-^D.Ül) different frofn sctr.h other. Üanmercial II diet perrormed poorest. fdj Conclusions ori Star'car Diets ---------- “ -------- --- ---- The starter diets vvhich ineludcd five pereant cassava performed as well as those withuut nassnva. Diets wit. 3$ pcrcent protein and 5 percent fibre lev/els wert hei ter than thos.. percent protein and 3 percent fibre levels. Tf>c- two oomputerisnd .^^rter diets in the senond experiment wern the cheapest and they wäre fuNftr! to perform bc-.ctar than the commercial diets. There is no doubt tliet piffurencGS oocur in the performanco of the diets and this could t-c ̂ arived *.n economic terms. In achieving this, the costs of the v/erious f eetis and the voljp of weight gained are considered. This aspect is rieai lch^.'ith in a later chnpter. 4.7.3: •F •i *nui%*s*hv-\*- r' ‘D* i::'•t* :• The rasults for fchj firisher diets are summarised in Table 4.5. The figures in the table are moan values for the varinas trsatments. Analysis of vnriance tests v/ore performad on cnlumns 2-4 of Table 4.5 (See Anova Tables -4.5.1 — 4.5.3) which show thet there were no significant (P^0.05) differences between the a\r rt'.gc weight gains, average feed intake and feed conversion afficiency for the birds on ench diet. - 5ß - TABLE 4.5; P- jrf.ggnannp cqrmrlqqn ,,af b>'o f j J « P . JfiqflfclCg- ^1-1 TABLE 4.5.1: fci&iß-för. k.vlghLni»ir^. Source Sum of Degrae of Mean Variation Squares Freedom Squar** P F0.ü5(3,8) F0.01(3.8) Treatment 0.071 3 0.024 1.14^Q -• 4.07 Error 0.157 8 -. 021 7.59 Total Q.23C 11 _____________ r_|— - --- L ^ j TABLE 4.5 .2: ‘VCA./'A i\.ll ?*A « C » .ata. iium ̂* Source Sum of Dogree of X ) F0.05(3,8) Variation Squares Freedcv’ /" \ & (ar*'j r f o .o i (3,o ) Treatment 0.301 Ü.1C 1.449 4.07 Error •0.551 0.069 7.5G Total 0.852 s_11_ TABLE 4.5.33: £NOt/.r» .r fJEO convr.rii -s• .i?ff.-dr.,ncy — f t ' Source of ^ Sum of Oegreea of Mann ^0.05(3,1 Variation Squaru Frcodum Sauere F F0.0l(3,( Treatment 0.0723 3 0.241 1.346 4.07 Error 1.429 b 0.179 7.59 Total 2.152 11 - — ....»-«,__________ ________________ - SG - Althoucrh thc differences between the performanca of these diets are not significant (P <1 D. 05), it coulci bc observee thafc slight variations still occurrcd. Columns three and fojr of Toble 4.‘5 Show that the commercial diets needed more fced to put on on? kilogram uf weight. Thi.s inniplied that the computerised diets used lass feod and since they happeniaa!d \tno be choaper than the commercial diets, irhey .-/ould aiv/o moro '"etum^^o^the farmer. The rcsults confirmed that .'he diets /Rr? nt rritionall nced. Howevar , it io cf intorest fco d^termine the best combination of Starter and finisher diets. The anelysis for this is presented in Table 4.6. Comparison is mainly on the besi s of feed conversion officiency. Column four shows that the best pair of diets is using starte*’ 7 with finisher 9. Also, all the compu­ terised Starter and finisher combinstions proved better in F.fJ.E. than the fcwo commercial diets. is pertinent to note that nonn of the computerised ....... ........ ... . ■ \ĵ k --- ---- Prices of compLöyri^'J diets wer- increased by 22.48 percent to make up for thc? overheau uVr y.-ts which the amn.ercial diets included. The figure was recammenpmi\jy OguiiliTvortt et. al. 'v5ß;. The cost* derived are thus: : 250/ton C230/ton Commarcial T <200 Commsmial IT JQ1 - 59 - TABLE 4.6: PsTformence comparisön of the ccepurnrissd startar and finishEr diets with two commercial diets ■r *3 au Tc ülss 4.1 and 4.2. Starter and finisner diets in experiment II enntained cassavn flour. 4.8: Enpirical Estimatic: 1 of F~tû eters Affegting Broiler Oiets Regression analysis hcis beer, performerf in order to investigate the functional relationship betv/ecn the feed intake lovels of the various Starter and Finisher diets and livewoigbt jains. Experiments 1 fcpauits shown in Appendix ( b) v/ere usnd für the rsgression analysi s- > Solutions to econometric Problems and thg choice of funcfctonal forms are discussed below. Violation of the assumption of non^^toregression in tho linear regres­ sion analysis poses a problem. Her utoeccorrelation occurs because each weekly weight of the birds is depe’ndsent an the weight of the birds at the end erf the previour v/eek. A]f^eon, since the birds wero fed in the same manner over the experimental peafeJ^ ii i /• r,.ost likely that a group will consis- tently be above or belang at/Rragc-, resultiny in succ.essive observational error which is pc^^^JSiy norrGlatBd ” "• ■ . The con snouences of such auto— correlated are givon py fV:.rtsoyianris J-/ as follows: (a) Alt houyh the cstimabed Parameters are statistically unbiased, their value in any single sarnole is not corrsct. (b) The varionre1 of the ranci.-- er* or disturbancs term may be seriously undurestimated if the disturbance terms are autocorrtj- lcted. In particplar, the undersstimation of the variance of the error term will be more serious in the caso of positive auto- correlation. Henco the varianep of the error term will be seriously underestimatod, and constjquently, thu variarices of the Gstimatcci Parameters will bo underostimated, especially if the method nf ordinary leest—souaras is applied, c. The variances of thE estinnfod parameters aro underestimated whon ordinary least squarcs muthod nf o^timntion is gsed. Therofore, with a false smallor variance, the reliability of tus is naage- rated. An estimato nun th~roforo be reger-'c reliable when However, Hcady and DilIon— - suggestod on U ̂of removing autocorrelatii by reducing the analysis tn a static (timeless) ono by regressing total woight gain for the birds ;3n total feucl intnke. This woulri only lead to a reduction in the degreo" of freedom, but th.: remaining degrees of freedor., are still enough to onablt the estimation of reliable substitutjnn rates and coefficiont This is iXeotly thu -omo thing as regressing the observations an weekly livewe ight ja ins on wmrkl/ fead intake vnluos. for the purposo of this study, average values of the replicates were used insteed Qf the individual obsorvations because of the nature of the axperiments. Firstlv, the birds are fed ad libitum, and it becomes necessary that sither the time intervals of taking readings are fixed or the quantity of feed consumed is predetermined. It is however, not possible to control both at the same time. . Usually, obasrvations on feed consump- tion and weight gains are takon at fixud intervnls of time such as a weck er month. Weekly observations an rinde in this study. Secondly, the birds are seif fed with full time- arcess to ths feed and so the amount of feed consumed is determinod by the par» . ular uird thus making it « random variable which is endogenous. It is therefcire measured with error äö4 5, e estimated production coefficients will be biased. Thirril/, individual gain in weights of birds could not be measured *•;.:> e there was no mau\/ of identäfying each bird and the amount of feed consurrnd by eaef ifcd r* the weight gnineri by each bird. 4.3.1: Functional forma employed — ■ jka» • • m t a > m ̂ ^ ..‘t Two type., of alnobrnic equetio5hß0 were fitted to the experimental data namely the Querir -ic. and Square rn*at functions. Quariratio and Square root funr.tions be.vo iaorlines tnpi converge tu n pnint, 0.1i'»ving r.pecification cf ans ration c a c t e n t with maximurn meat prc'uption. Also, the ouadratic function allowr, dî iS,n*jwiing and negative marginal products, and also defincs 2 maximurr. A li&rt > M ,v> of the quadrutio form is that it imposss linear ly leert -sing i products which may be a poor apprnximation to tho true oiol-jgj of th.v Produktion process. Heady and Dillon ?sJfound the quadratic function to be acceptcble. The isociuants of both the quadratic and squara root functions are not -o"vnntnti.c to the input exos but rather they intersect the iriput axes so trat certain output levels can be attained using either of tho inputs only. The only differencs in the ohape of isoclines of the quadratic and square root functions is that the isoclines of the square root functian pass through - 63 - the origin so that it compromises between the Cobb-Oouglas and Quadretic functions. The isoclines of the squaro root do not specify a fixed ndx of resources for attaining different output levels as do the isoclines of the Cabb-Douglas function, and it does not impose linear isoclinoo os the Quadretic function. 4.6.2: Regression Results (a) Criteria for Selectlng the Lead Equation To deterndne whether the regression results & good or bed, the fol- lowing factors haue to be considered: (i) R. 2 - This reveals the goodness of fit. It shows the contribution of the regressors to the explanation of variability in the dependent variable. 9imply, it gives the percentage of the dependent variable explained by the independent variables. (ii) Dubin-Watson statistic (DW) - This test helps to reveal whether the assumption of non-autoregression is violated or not. The value obtained in the regression results should be between the upper and lower values obtained fram the table. DW ^ dL Autocorrelation occurs. DW dll No autocorrelation dL < DW dU Indeterministic where dL > Lower table value of DW statistic dU « Upper table value of DW statistic. - 64 - (iii) Signs - The estimated parameters must bear signs which corres­ pond to the type of correlation between the dependent and explanatory variables from apriori knowledge. Explanatory variables are either positively or negatively related. (iv) F test - The F value in the regression result is compared to the table F value and if signifioant, means that the explanatory variables make signifioant contribution to the dependent variable. (v) Standard error or t-test — These two serve the same purpose of deter- rrdning the significance of individual explanatory variables. The mein difference is ,iust that the Standard error gives a rough estimate. The Standard error value is signifioant so long as it is about half the regression coefficient. The t-statistic which makes use of the Standard error provides the test-statistic which shows whether a particular regression coefficient is signi- ficantly different from Zero or not. If F is signifioant but the t-tests are not signifioant, then multi- collinearity poses a big problem. In a single linear regression, the F and t tests coincide and they serve the same purpose. The significance of the tests reveals that the individual explanatory variables are making significent contributions to the dependent variable. (b) Effect of Feed lnteke on liveweight gain This section enables proper camparison of the computerised and commer- cial diets. The relationship that exists between feed intake and liveweight gain are establisbed in functinnal Morris by regressing average weekly feed intake an liveweight gain. As established in sectior d.Z of this chapter, if genetic and environmental Factors are well controliod, variations in weight gain3 are dua largsly tn the level af feed intnku. rne^iowever, is a function cif its nu tri ent nontents. Here, tne level af Feed intake is the subject of discussion. The nutrient cantonts are in a later section. The relationship bctv/oen feed intake and ̂ v^//eight gnin thus established vn.ll enhance th... pwssibility of forecaefeing vvhat quantity of feed produces an output of broiier rrnat as well as comparing the marginal analysis with foed convorslon officienc nical analysis). (i) Ntethodolpgy Weekly rocords of feec! mii4 * $ and liveiVv. ight gains vier , fcakcn as rav/ data. Average weekly f'qur^s per birr warf; derived. However, to salve th; Problem of autooorrrlata k i.tiiah occurs in a tirnes seriös cxperiment such as described earlinr i i. h Jntroduction of this section, the method of first difference v/as applied to the oumjlative volues of these average v/eekly figures of feed intake and liveweight gain. This in effcct is the same as using the average v/eekly figures. (ii) Estimating Procedures The modol to be testet war a feod response mrdel whosi- mathemetical representntion ca.n be written es Fellows: W - f(x, V) v/here W = Average v/eekly liveweight gain, - 6 6 - X = Average weekly fned intake V = Error tarn. In estimntinn the Parameters of the foeri response modal, two functional forms were tricd namaly:- Quadrafcic: W - a -5- bX *:• cX y (eq. 4.1) i Square root: W a bX + cX' + y (aq. 4.2) From the time-seriss data r,ollnotcd ouring the Experimentr the parameters of the feed response model wen «stifte *.ad for^tfrs commiputerised and commercial Starter and finisher diets. The empiriofflVitesults are presanted below: (iil) Empirical Resulti' For all the different diets, the quadratic < urm gava the "lead" equation in terms of the critsria listed in section 4.8.2q obove. The functions for the Starter diets aj^^rfcso/ited in Tobls 4.7 . In these diets, feed intake was found to be an important explanntory variable as measurec! by the F-tests (significant at P ^-.ül) and the percentagns of i:he variabilities in liveweijht gair. expleined f R^) 2 . The R values rangod from 72-69 percont. This high explan?tory power for feed intakas in these diets msans thot the nutrient balanoe arid availability Situation were such that intake was not affected and mast of the feed consumed was used directly for growth purp-ses. The O.W. test statistic shows no cutocorrolation in r.ll tho diets v/hich sugganta thot the oquations con be used for inferential purposes. F'i'or.i npriori cnowledge, X t (feed intake) - 67 - Jtible 4/7: Empirir.nl resul’to of weight response to fceri intako in Starter diets denotes significance at P <, 0.05 All othar variables arn as previously defined. - 6 0 - conforms with the expectations of positive correlotion with the regressors and (livev/eight gain). One oxpects that livewaight geiin should increase as fest! intake increases. The functions for the finisher dists aro presented in Table 4.0. In thfcse finisher diets, feeri intake was also found to be an imoortant explanntory variable as rneasured by the F•-tests ( sigriifipa'g t ^ t P G.01-Ü.05) and 02 values. rC vnlues ranged from 83-92 percent g pt for commercial I diet which had 54 peroent. In this cranmernial I ’iiet, whero R was as low as 54 percent, it would apper.r h->f grnwttwsSp was influenced by same other faetnrs other t h m feecl int ~kr. per These other factors could include reduced digestibilit• i td hmc|K"Si\erad ovailabiiity of nutrients or even an unbalanced pattem _-f r̂rj.ny eicxHn. This Situation is not noticeable in ths computerisec! rlif?ts (nn^ fei.ul Ti) where restrictions in the model used in formulating therr hs^-^^oen speoified such that at least a bclanced pattern of amino-ecids isN^xhtained. The regressnr ! the correct sigp except in computerised finisher in eguötions 4.9 and 4.10. diet 10. The Durt^g^itson t••o s t statistic is u.-.deterministic ^„ There is doubt therefr.cS hü to whethor serial correlation oen.irs among the residuals. The estimSjr^^guations ore hawever, still useful for predictive purposes. The commercial startsr and finisher diets were sirnilar in the way they influ­ enced liveweight gains in terms of accounting for vrriabilities in liveweight gain and their marginal physical productivitiBS (ivlPP] are disoussed belaw. - e9 Table _4. S: Emjjirical Restilts nf Weigl it Response to_ F_eeri_ InJ^ke. J p rÄnÄsjTG.T. Piets. 4.8.4: CcjTgjjEyison af ‘toiiiinnl Riysin»! JYod dct_ with Fe3c^f^anv^3ion_ pFFijjienc The marginal physi.-.al produrtivities cf eech of the diets in obtained by differentiating the osLifik.ting quations of euch of the die^c 4.3 - 4.10 with respect tc feeri tninkss. TrF.C.tZ. oalues are thosc ^^w£ved in Tables 5.2 - 5.3. The marginal phyiiral »roductivity is nomparcr/T u-i Food Convar- sion Efficienr Ln the sen « tnat 1 oth tv •poasurl.nrj Jbh*1 additional weight gain if on? .mrr d.logran ui focd is c :r, ‘med. MPP is ü^faineci fr ĵy for eonii of t+r^S^fation W = a + bX + CK2. d, Starter 2 •— « 0.75 - 22.K g ................... 4.3.1 Starter 7 - O.d^-NJSst .................. 4.4.2 Finisher 3 ^ => ß - 3.02BX...... 4.5.3 Finisher 10 -jV ^ -0.35 + 113.S0X................. 4.8.4 Comm I Starter = °*7S “ 111-26X................. 4.7.5 Cjmm II Stey*fĉ r * *ĵ = 0.33 - 04.1CX .......... ...... 4. C.6 Corsn TI v V dW - 0.54 - 0.06X ................. 4.9.7 CIA Comm II Finisher rJW * 0.51 -Ü.10X 4.10.0 The MPPs and the corresponcJing FCEs for the various diets are presented below: 1 Diets MPP F.C.E. F.C.E. Computerised Diets Starter diet 2 0.7493 2.73 0.37 Starter diet 7 0.4840 2.42 0.41 Finisher diet 9 0.3651 3.01 0.33 Finisher diet 10 0.3524 3.15 < * * Averages 0.4877 2.83 0.35 v ^ Commercial Diets Commercial I Starter 0.7210 2.©5 0.35 Commercial II Finisher 0.5408 0.28 Commercial I Finisher 0.3342 3.20 0.31 > Commercial II Finisher 0 . 5 1 4 ^ 3.63 0.28 Averages 0.5276 3.32 0. 30 Marginal Physical Rroductivity (MPP) ia a measure of the increase in liveweight gain when an additional unit of feed is consumed. Qh the other hand, Feed Conversion Efficiency (FCE) is a technical measure of the quantity of feed required to produce one kilogram of liveweight. Taking the inverse of FCE, makes it comparable outright with MPP. On the average, an intake of one kilogram of both the commercial and computerised diets produces approximately 0.5 kilogram of liveweight gain. Comparison of the MPP with FCE values for these diets Show some variations. The MPP for the commercial diets is higher than that for the computerised diets but the FCE for the computerised diets is much better - 7 2 - than that for the commercial diets. For the computerised diets, FCE Shows that ono kilogrem of feed produces 0.35 kilogrem of liveweight gain wherees for the commercial diots it produces only 0.30 kilogram of liveweight gain. Summary and Conclusions It has bcen established that feed intake is an important explanatory variable in livewoight gains as shown by the values (54~92 percent) and F-test (statistically significant P 0.01) of the estimated equations, The computerised diets compare well with existing commercial diets, as shown by FCE and MPP values (differences in FCE were found not to be statistically significant). Tho computerised diets however have an edge over the existing commercial diets beceuse they have better FCE and they are cheaper and would therefore increase the net revenue accruing to the farmer. (Prices in page 58). CHARTE-* FIVE CASSAVA AS A S*.*3S rXTIJTE f-OR MAIZE AND GUINEA-CORN IM PUULTRY ' t’ßnÖILEny FEED " c.1 Incrpduction Up tili now, grains uv* haen tf mein energy saurces tn poultry foeds. The demond for grein? for i ruman consumption is so high t io supply in the ccuntry cannot meet it. This has lod to high impbrtation of corn in vsry large quantities to meet the rinmand for both human and animal consump- tion. In addition, there is also the need for inrdiuusstrial manufacture of dextrin for the production of glucose ns v/ell as starch for the textile industry. This high demand for maize haä sS l PeLd to very stcep rises in the price of maize in the last five yn Search for alternative sourcss of energy in compound feeds havc sbowr, that cassava and sweet potato have consider'bin potential. This study tested cassava as a substitute for the grains. Cassava, despite its low cultivaticn has been chosen as a substitute because of its other good characteristics which are disr.ussed below. Emphasis 'w ûld bo placed mainly on finding out the rate of Substitution betv/een the ^^nd cassava 03 well as the economics of the use of cassava. The prevailing price of cassava is higher than that of the grains and this condition may not be so in future if production increases and use is made of the uxistieg improved technology. This has prompted the adoption of parametric programming to tsst the effsct of varying cassava prices on the total cost of a given weight of broiler ration. The response of the birds with increasing levels of cassava in the diets is also of utmost interest. The returns nccruing to the farmar witri the use of casseva bcisud diets will be conpared to ths grr.in basrad diets. The Optimum aambination of maize/ cassava and guin^i~corii/cassavn will also be determined. 5.2 Characteristics — ..... — Cassava, aparr from being an .11 saason crop, h^SlVoAyv/ery high photo- synthetic potential thus makine it fchs highe st enargy yielriing per hectars crop. ^ Thn high yield/hectarc makes of the cheap sources of carbohydrate. It is uasily propagated ffeoi^cuttinns and very resistant to pests, weeds and diseases, It has the ability to tolerate drought and poar soils and poses fewer storag- problcms than the greins. It is thcrefore intnresting to exploro the possibilities of using cassava es a close Substi­ tute for the mojor sourcss of cnergy in livestock fccds which are as of naj largsly me.de up of maize and guinea-oom. The prob1 nm of toxicity is being ovsrcome mainly via plant breeding techniques which sqlsct for low hydracyanic ocid yield and through efficient Processing tochrilques whic.h climinatn the toxic constituents. Processing mey howevc to the c.ost of the product. jsava has many alternative uses, most important of which is its use as staple foorj in almest all parts of Nigeria. It is . also being used for industrial purpnses, for examplc. in textil:’ industries and fcr making starch. 5 . Pantors^/yfj^-c^tir^ Qsssave Production Despite eli the qualities of cassava, it is only recently tlvit morc. - 75 - systematic arid sophicticated mnthods of cassava culticotion have been smployed.—9/' w3*9 •/ The :nain Factor retardinn its extensive use as livestock fsed is the prn.scnco of cyanognnatLc glucosidas which mnke it toxic to 40/ animals.*-- Fortunately, hov/cvar, is the fact thot it is water solublo and breaks down under high tonporatures. Elimint tion of the tsVanogenatic glucosidas is the nim of treditional mnthods of proccssi^^ihich entsils eithnr soaking for many days befors sundrying or roasting or grating and fermenting for three days before ransting. Cassava hss been discredited as human or anima1 foad becauso of its low protein content. It is essen- tially an enurgy food and indicatians nov; aro t hat the futurc food shortage in developing countries may give cossav«, ynev; importance so tliat farmers may shift to proclucirig it on a larg^^ale. Cassava is highly soil-depleting and farmers reeort to its production only when soils can no longer give good harvest for other croos. Also, it givss a lower incomn A to the pmducers but it componsates by g4 LH^/^Lghor yields per unit of hectaro. fJchsr factors affecting the cwlra^ntion of cassava include low levcl of knowledge of r'gronomical practioes and cassava potentiale coupled with unorganised market­ ing syst^K^^ the cassava roots. : i. d Price c amparisrns As of now, cassava fiour seems to be more costly thar, the grains. A faw factors whief could be shooting jp the costs of cassava fiour are diffi cultics- in handling, starege, Processing and marketing of the root crop. - 76 - Processing has the greatest relevance to the high pries of cassava The peels of the root contain a phosphorylase liberation :.if the enzyme lynamarase. The enzyme allows free Hydrogen cyanide (HGN), the toxic substance in cassava to bo liberated from the cyanogenetic glucosides present in the tubers. The rind of the tubers therefore must not be present in the root maal during Processing. This Processing stage may therefore nacessitate more labour. Most of the Processing to date is by fnanual traditional methocls. The develop­ ment of machinery for the bulk Processing of cassava may reduce the cost and increase the volume of flour available for livestock feed production. Sinne labour is. c very costly factor of production iri our economy, it is highly centributory to the high cost of cassava flour. All that is neederJ then is mors efforts on research as regards the development of high yielding varieties coupled with better agronomical practices and Processing techniques. Maw that mest imported products are being bann d or restricted to consorve our foreign exchange earnings, maize may coma to bs affected. As of now, maize is largely imported at a much lower cost compsred to the cost of cassava flour and non-imported maize. Also, seasonal production is reflected in the cost of cassava because cassava p: \ces vary from one season to another. üntil oxesss cassava left over after human consumption can be processed and stored For livestock use, seasonal price variations will continus. With great opportun!ties opened to in research, it is hoped that UNIVERSITY OF IBADAN LIBRARY - 77 - largor quantitins woulc? be proriuced and marie a.vailable at lower costs. This is possib? 9 if research findtings aro rully adup'ced and incorporated into peasant farminr; and large acalr farming Systems. Thus its inclusion in the ’.diet will ort infl.ita t!ir. cost of the feed unroalistically. 5.5 fYevir î Stujies 5.5.1: The TT?yiej>: A lot of work has been done on tho fBeding of' bvc to both pigs and poultry. Tttire aro variouc form? of dry commerci&l feoriing producta from the cossava plant. They are in Chips, pellet^ roctangular bars, broken roots, cubes and cassava rueal which is in f'tfh pnwder form. The refuso er .vaste s y is another product and the leaf n/jo L which i?s the driod aerial part (ar only leavesj of the cassava plant. This study fccuses ' i i ecrn the cossava meal. The nutritive content of cassavt NOeries ording to variety, ngc ant.t the Processing tecnnology. ̂ ••■'•^^■»»/aniöes arc the mnst sirablc elements of the plant. The '/ariea between 0.01 to C.24 per-:.ent in fresh tubers with the bitter« varix Lir,r- r.ontaining 0.G2 — 0.03 osreant and the sweet onss having ^ss; thari 0. )1 percent. Freo hydrngen cynnide is liberated from the cyanogenetic glucosides by tho action of the enzync linomarase which is naturally present in tho plant. Glucosides and linamarase come into contact only wher- i:hi plant tissue is damaged. 7/ «—7-0x/ - 70 - The fesding of oaasava to 001 iltry dr.bv-; bcck to 1935 >;̂ n Tabaynyorv7-?^/ / incorporated o product drrri'id fron» thr cxLracxi: r, of easstivo starch at the Ck 30 nnd 6C pereant ü -vals i.ito chickun diets. Ha founr! t'mt fssd consunption and 12-week body vveight decretised c.s thr level of the cassavc. byiprnduct in the diet increased. The next plane of work war; in 1941 by Mcf.'illan and Oudley.-Ahoy fed cnicken diets cn> »tairing 2W and 40 perj^fe^ssava root Flour and did not notice deletirious* cffects on I ho w m of thr hirds. Howuver, thoy concl'«.!:"■ that t!ie higher 1 ovel of gub Statut! un oroduced e reriuction in wsight rjein. Klein and Barlowen (1054) aFfirmorKa^their own study that cassava Flour contained a Factor thc.t ditTiinishqjTdesn^1 oonnumption. They roconmendeti thnt cnssava Flour shoulri be used a ev/b•l1gs not higher than 10 percont oer.ausc higher lovols wer'- reported to degroaso weight gaiM and feeri pfficisncy ml The works of Randon, et al;* ranlYvY. -J the finriings in 1959. These wäre also \ ~>/il the vimvs uf '/nt and Perffpar in 105 nnd in addition they expressed their firidinns r.hat nfet/erse pffonts of high levels of cassavr occur mainly in the I N w v/eeks of life. Vajt 1966) therefore affirmod that bi’oilcr^y^^'innr." !_• cassava at levels higher than 1h percent only after the ItoEütrh ur-’k. So ^9c4, Vcr-v rin;' Stute 7-5/ nbserved thut weight gains wert higher with cassava ne': lots t r i wit'i thr. meal, and that it is the excessiv/ely fine nature cf c fl.:» t t - tifl*«nnrecl thr iTtnke negabively. ChOu and Müller (19?2) ■.onfir .•'■ l thet cessnve. pel Irts coult! bi? used up to to 50 perccnt 1',/ci without any ndvarso effnet providud that the diets were duly - 7 9 - balanccd with rogord tt ■ ther nucri- . dthar works which proved that the powdsry nvture of cas:*vvv. ’Luut dccreases weight gains are tf\/ fcbose of Montilla, et al. . Tiic first ono was in 1969, Thcyincor­ porated , 15, e; ui X pr*ro*- . .-»ssavo root flour (sunnri^d fnr about 36 hours) into rhi-’-t mtiuna. Uv the sixth week, thay found decreases in both weight qain: aoc! fi'-d efficiency as tt ie eassc-. vn level incraases in the rtiono. In 197.. / they ••■'dvaric.rd in their seccndNetudy by U3ing the sarne levela nf Substitution, and nddinq tc cj i fne diets, five pereant > > > animal fat and fi,.= peroent sogar cane r.pflcteks with the via« to oli— minating the poivdery nsturs of th- Ab the eighth weak, they foun thnt nc significant diPfarfinco • ln nnnr.umption and woigl t innreese, or feed effioienny wäre detec!^1 fitetween the treatments. Also, Feed nosts wert reduced by 7.6 jA ' £ parccnt for the chicks which receiverJ ratxons contrining 16 an^V»-. parv.ent oussave. In 1975. fchoy oarried aut a third stuef/ v/hic.!* iirfers f. .he second on3y in the vnrifity li?" naaaS« Jsed. Ratn-sr than using the 1 sweet1 cessave, thcy tjand tho '' sA^r.as: ave root flour to roplace the c o m in tha diet parii’..! ly' b S “ nnreent anu tatally by 37 percc-nt of the ration. In both cesej.t ■-... ja,;-' mollcocs and animal fat wäre used to renlace 11 perr.mi: •' t!v c c t to eliminatn tha powdery natcrc. At the fourth weck only th, .'7 pereant cossava dict was significuntly pnorsr. The ßuthors • . ributed this t- ‘‘V 'C. • >. • .bent of thr- casseva variety used. - a a ­ lt is pertinent to nots thr.t tKo ca sra.wc fl nur i:sad Ln th-os« Experi­ mente was merely eundripd. Thp i. t .c 'ü x u of riochiariir.c as a moderater oF toxic effects of cuss* ■ pr« moi m poultry hos becn acknowleriged by many authrrs. 1/ . Adscibola rovaal«. i tb^t adnn:; notniouinc rnay ber reguirrd to imprevp the quality and utilix.itri .Mi of J:V ntary protoin anci •'.von in a properly bßlancnd diot, it nr / sewr: For the doto.'^esfeion of pruasic arid v/) lieh is rejnosnd in the hydrolysia yf lin^raarv; and - V loteustralin. Hs dre/J attention to the neec! tc^ Tato responses to r.iddecl methionine in rntions tu the levuls of pf^strin in the diat: an • BU " to the mturs PalntablUty V 4* . fort. He rtn»3Kl th»t methionine shares its role with othclE^ SViTMLi1fur-doncrr. such as cystinr , thiosulfate and elcmental sulf^^^yMMethii rvine is prüflrred because it is an essential amino—nr.id nfcic! wnen rnctaboliseri, it yields evstine and cysteino. The present stmiy acknov.Tcrincs thest Facts and pruvides For then hy n. • Ifyj.nr- levcl- cf inclusion of thcsc amino-acids in the riirts. hur 1.0U'■^^»av4, Ross —1 J?/ feu jp tr 50 percent oasoava to poultry observ^^n";: i an c but niontt Elv/aya significant irprovsment in bady waight .# . *i tr. c v n vmoo supplnmsnted with •?. 15 - "i. 20 percent methio- nine. 7 .• c . viludcrj chat yn the rntion was well balanced /.Ith protein and mst‘ iunine, up to .* n>.rennt cnssavs rnot meeil sarisfectorily replaced corn in the diet. ••Vifhv.jf nnthianine supplemantation thoy Found deterio- ration in weight gain at threo v/eaks of ege and signifiannt differences - 81 - in fccd efficiency whcn cassaea lcv 1s p-iofc-vdcd 20 pnrcent. Thn arldition of nolasses and scybeon c&J had no btnsficial effect, which provei: occording to the auth ea. v, of thr additinn of animal fat and soybean flour. "J. tghf gvi " ‘?,3**nd the smoun«:. of oassays flour increasss although the nif/o-runc: . ignifi? -uf; rj.y at aossevti levelc frcr» 37.3 - 45.r'. Thcy poncluded i:l- if the fas:’ balariced far energy and protein, cassave f’J.iir -dü1;J b? incorpnrvntad into chicks diet up to the .TO percent lev .3. withoLrc aPfocting^ .vnight increase. Phueh and Hutagsluny —65 / tw^tad rations vVritr. 19,22 and 25 percent levej.s of protein end 8, 20 i-j 48 percent lav/als of cassava flour an brailoro from 3-6 weekr of ago. After the nixth werk, the protein levels of the diets v/ero chenged to 17, 26 end 23 percent. V/ith 20 percent ca$sava, they foun-.l that the peroentay oßvaass yiold (dry bR3is) and carcass yisld protein, were significantly higher and fat. Production was lower than that v/ith cassava levels ab ovo 20 pßi*cent. l?/hcn the cassava ipcreased ab07.3 the 20 percent lovel, riiyetibility of the protein was reduced and that of fat was increaseri. Armas an1} Chicen (1975) ~ raplaced c o m in broilar diets with cassave flour at thu 18, wnri fv1 percent levels. The diets wäre made to be isocslurir and isopj- /teinaef ous. They fourc' nn aignificant diffe- rences in weight grins er-1 fi-jsd afficinncy although the dieh v/ith ;34 percent cassava tu? i c lnv> ./night gain. The fnet that the diets contained ß or 16 percent sniual protein, or were suppüemented with 0.3 percent methionine and U. 3 peo’ctnt ly sine did not effect the results. differcnces in their chnminai oonpositior. r.oused oy ago, time of hervest and methnds of ,3roce3sing. 5.5.3: Sümmary of Works pr ssay« The first picoos of work vverr. en faedinig - 1 1.. 1f uec so sa ac jvv*- ir ve« fi usc uo;r moal to chioks and they worc fcr. ir t<: infiuencicfr’ f30'I consupption nncjn- tiv/ely. N»unrJn;. c<.« ce wero excluded from the 24 percent protein diets. In tha 24 pereant protein diets, the innlusion of cassav/c causeri 7.-1, 16.6, 54.9, 14.3, 53.5, 5.5, 54.0 percentago changes in oystnr sh oll: , muat agd bonc,lysine, guinaa- corn.methionine and bone-meal rospvptively. Ohanges in guinea-corn and maize. the energy-based innredionts are duu to their S u b s t i t u t i o n for the amount of cassavn iricludod. Gheî s: ,c .t in the synthotic amino-acids ars dus to the low content of these affiincr-aciris present in cassavs. Substitution in the mineral-bssuOrt v£ngru.:isncs is to msorn that the specified levrls arc met inS^e onrrsct Proportion. For the above roasans'for nhsnrnr. in compositions, substitutions in the 26 percent protein diets occurred in oyster Shell, rneat and bonc, nuineo com, bono-fneal, driec! yeast and gruundnut cake. The peroentage uhenyes warn , 61.6, 11.1, 21.9. 61.6 anc! 7.6 respectively. Bqfeta, jpations previded the essential nutrients et thu lavels speoiflBd for thr broiler Starter rotions. The lc/el of innlusion of synthecic methionine jwnges botween 7.15 - 0 .19 and this is consistent with the lav/ols epaed :?ied in the studius roviewed abovn. B6 - TABLE 5.1: affecfc* af iraclurjinn cass v ■ nn th' •vjrt i3 enrapositior nf* otartur rciticn - 0.1. prires Percsntagr ror.,pr»~i tinn Incjredients and ftjH -»im- im Protein 24' Minimum Protein 2.6/3 Nutrients „'. .'xtriout Gessavo > „/ ' Change V/ith a OC. 33» l\jCi "ua '3 Chan^c f - ...... ..........i ...........> 1 r ............ .................... Brewer's grsins I - 5 .0 0 j # 1 . Soyabean 3 0 .0 0 “ I - ; ^ j Oyster Shell 7 .4 2 .4 9 7 .4 S / 8- * | Malze f IG. 5 ^ ^ Meat and Bone 5 4 .9 2 . *•-' 1 .3 2 5 1 . a Syn. Lysins i Guinea--corn 4 5 -1 1 5 0 .0 4 I n . i Syn. Methionine J * 0 .5 n . 15 0 .1 5 Dons mea? 5 4 .0 0 .2 2 0 .7 0 2 1 .9 Salt - 0 .3 0 0 .3G IVheat Bran - ■ s .n n s .n o - Ad Vit - 0 . 5 0 1. - Casscva 5 . r i : - - Dried Ysast i. 60 1 .0 3 5 1 .6 G. nut oake 2 . .36 3 .0 5 7 .0 0 r . : r . i t . ........................ Cost of 1 < W mix 2253. v.fS . 1 9 9 .3 4 2 2 .0 0 S'5 4 1 . S S »‘ '1 0 7 .0 1 2 2 .5 • -» ^ «fr >. * • t -• . Protein 2 4 24 2 5 | 2 5 - Fibre 4 .1 4 .0 1. /4 4 .5 4 . 6 2 .2 Fat< N > 5 .0 5.0 - F r) Cr*i . *X ■ - Calcium 1 .5 - i 1 .3 1 .5 - Lysine 1 .2 5 1 .2 5 i . » 1 .2 5 - Methionine Ü.5C l 0 .5 0 \ 0 .5 0 - Phosphorus 0 .8 0 o.sn - mix withiii the rr.nor; of 0-30 percent. The different prices o\c3ssave> isod are 125, .'150, '175, ’20n , .225 and '250 per tonne. These prices are lower than the existing prioc of cassava usad in the basic solution which .is J2Ci/tonne. Tn the Solutions, it is expecteo thet csssave should substituce for guinea- corn mai;.' in vurying proportions. Tnis is beoausc tho priccs of mai^o and guinc -coiti differ; so erc thoir nutrient compcsitions. Maxxe costs per tonne whemr.s thr. pries or guinua-com is . '220 per trwrv?. T;.L*lc 5,2 .-»hcK*> tu. costs and nompusitions of Starter diets wiien varying price tf c"-:;saua a m osed in the progremming exercisns. 88 TA0LE 5.2: Thr. composifclon cid nosfc s nf ntr.rtcr dirts v/i'th vsrying prices of rasr.vw Cassava replacing i;. Car j assnva rcplarjing Maize ......— ............. . .. I ■ ...... ......... ~.... Ingredionts P n I 0 E 13 • U F G A S S A Cassava Guinea-com Malze Brewer*s greins Synthetir: Methior.ir Blood moal 3yn. Lysins Palm kerncl meal Salt G. nut coke Ad Vit l ieot & 3one screp vVheat Off als ".54 •;DST rF FEEO/TQfJNT • - ■ — — *-«i ......... / ... • - * • • H. k:i7.v. an :i0u.T4 ;:iS‘i.6S 1T8C.36 ."194. 94 i 199.93 I .‘107.96 >•‘203.20 v.206.52 . . . . . - ^ ...... -5- The mix of ingrediantn r: :r' -*c the snmc for prices nf cassov/a at 0125 and ' 15j per tonne but .!••• cgsGs of tho fand-' v/.tr/ >•••.•-̂-> passava price was fixed at i: 125/tonne fch& ft:'.; cosfcs ..'130.36 and with :rcv’ price fixed at .‘'150/ tonne th* feer* cosfcs 137.86. :-r Optimum solution retivains the 3smr; for a jangt. nf pric.es of cassava frorti ■'200-.''251. However, ths >.osts of the read arc .'199.93, 203.23, 520C.S2 frrr nGssaua prices nf ' 20'.?, 3225 and C25G per tonne respectivsly. - 89 5.7.1: Substitution b^twoen> Mcdza ; J Cassava At cass.--.ve pri-.as uf ’2nn, r _ , - > J '.'250 per tonne, the solution remai is hh • samn bi.it with varying .. .«.sta cf 199.93, ‘.'293.2t1, .206.62 psr tenne rospectivnl/. Cessave m incluOoc' up to 13 percent levol only. Vvhrjn the priao or ccssavr. was love-red to ."175 per tonno, the mix includod cassava ut n higher level of PP perccntAi the cost of t'.' feod reduced to 194.94. per tonne. At na eci, prires of 125 and 0150 per tonne, ths Optimum mix remains ti.iQ same with the post of fencl falling to t"1?9.36 and .' 127'.36 per tonne. Cassava however, comos into the solution at the mexirr. ir. Tovel of 30 pereant. ...................... < § r 5.7.2: Substitution ^ i f ^ n and Cassava The Solutions with roiosat to the jor.tpusicians and costs of feed diff r hnwevor. H V ^ ition is betwarn cassava and guinsa-corn rather thon mr.ine. ij^^sava does not .tarne irtn trr eptinum mix if its cost is higher thon .'150 per ton ip. At the orice uf 150 per tonre, cassave cyifirs intn the soDutim at the 27 percent levol and et .' 125 per tan.r , it r. orte 3 inte the ".b< at the maximum lovcl of 30 perce >is ootvte rf i:.-”- fc ad euer jene as ihr price of cassava innreases. ’. it r-h'sdir,r "Arielc As stresasd earlin- in cheptcr fe^r, ths loast-rost diet is not necessaril/ t‘ ... lnasu-time rtr the mast öfficient riiot. Also it - 9 0 - may nnt be Lho dict that givcs t'n: mrxiir.nm net pmfit. In ths prvvious sections, ccssav has beer. made to rsplncc rned.cc and qi dnna-rnrn in the diats. It is thereParo vrr .- importavit in rcoasure the responses of birds to such cr-csav/u ivjsod diots ns done in bhc oarlier studies reviowcd. Fonciing trials w.cre onrrierl oat also to clarif^ ̂ ie nofcion that caasava■beisei diota usually cepross .vowth and Pant it does, whother it is nor. jnomic&l to usc it. 0n<:. ochT oblpptive is to conparc the conoutorised riicts with two cnr.Ynonly osod conrnercisl diets. 5.3.1: Experimental diets In fnrnulating tho expriment t? s, the existing price of cassevr. of • 32'". per tonn- wos t «sechs A total of 14 Starter on^*i ^5 finisher diets warn *ompeteriscrf vor th'j cxperiments. It v/os th. obiective in the formtil .fcian of tho / > storter diets to hnv/a ' 5, 1 15, El, 25 and 30 pereant oassava replace gainea-co» and meizo r- spnctiv/ely. These lov'clö of c.nssava wäre jiven eguelity resfcraints sc as to ensore that aacsava is includerj at the exact l^k/als in thn v/orions diets. This is bacavse the cnst of cassavaNC3 so high that it dnes not uomo in fr5sl\ into the mix if given c minimurn or maxinun constraic* . and jlP it comas in at all, it does not nnter at the nxerr integer Itvc.-is specified. (a) Csssaya roplacin.- guLien corn (Startersj: The solotion with regard to cassava replnrincr g' ine» -am is pr ■> ■••d.xi J..-. Tsble 5.3. The mix changsd as the cassava lev :1 in-rovisad i the diets. Because of the - 91 - Cnroposltion and costs cf startur diat? wi th cessava subntituting for TADLE 5.3: guinca-ddrrl. Ingrcdients and Nutrients GCC 1 GCC 2 GCC 3 GCC 4 nco c DCC 7 Cassnyn 5.30 10.00 15.00 Guinea-com 49. 15 42.09 36.03 Brewer’s grain.s 5.X 5.3 1 5.0o 3yn. methionine 0.23 •».23 0.23 Blocd Meal 5. 3"> 4.93 >. 5 Ffelm kernol mn/r.l 15. o:: 15.00 15. 1> Sr.lt 0.33 J.30 Gro.ind nut cekr 14.05 17.65 SpSs Ad Vit 0.C0 0.60 10.03 V/haat Bran Ceat Fr Bon;.* scr~p 3.57 3.44 4.55 Synthetic Lysine 0.20 0.25 0.20 J.26 1.25 5.00 4. 36 4.50 4.50 4.50 >.d1 .02 0.82 0.03 U.04 1.03 1.25 1.25 1.25 1.25 1.25 1.25 0.50 ■ .53 n-. • • < 0.59 0.50 G.50 0.75 0.74 '. ;e 0.73 n.cö 0.77 0.27 0.27 n, "■ > 0.27 0,26 0.25 24. «X 24.00 24.00 24.0.1 24. OG 24.00 ZSO. . 29* ü, . . 29.Ä ... 2909 . 2900* • • . . 4. * ., • ...290C Co.st of ane 199. n 205.57 214.04 221.51 228.90 237.1ü cori nf feed 19: .SO 291.5? 206.54 211.51 215.48 222.10 with these 194.1 12C .5/ 199.04 231.51 203.98 20-7.1C pri.se s of 191.S.; •m.5? 191.54 191.51 191.48 192.10 cassoya 109. i •i 3(7.5? 104.34 181.51 178.93 177.10 -i-073 1-7 indicate ntartsr rfiets in which csssava replacnd guinoa corn as the ma.̂ ar cvi’rqy aoumn. - 92 - low protein content nf cassava, tho cjroundnut cako lovel incroasad with increascs in tho cassava level. 'Vhen cnssava innroased from 0 io 5 pereant, groundnut cako incroosesd by 1.73 percent. For svery other 5 percent cnssava lncrep.sn up to 25 purccnt level, there was a 3.6 percent incrouso in tbv groundnvt oaks level. The Inst: 5 percont increase i • cassava up tn the 30 pereant level produced only an increase of 1.C pereant in tho groundnut ceke lovol. Ihn rpverse was tha Situation with Blood meal. Decreasos in tha bluod meql level were the same for the first and last addition of 5 pereant cassava into tho diet. Increasing cassava from 0 tc 5 percent end 25 to 3J pOrcent each resulted in e decreaso of 0.82 and 0.S3 percent respcctlvciy. However, the addition of 5 percent cassava up to tha 25 percervc levei resulted in a constant increase of 1.37 pereant of bloori meal. Such Vife tho pattem of chenge in the levels of \ meat and bone scrap. Tu fcjalancc tha omino-acid oontants of the diets, synthetic lysint J ivals increasod in the diets at an average of 0.05 purccnt for cvarf}-?' nercent incroiese in cassava. For' che first and last 5 pereant irtorenci in cassava, gu±nea~corn feil by 5.2 percent and successdvä 5 pcrceni* increase up tn 25 percent, there was o decrcase of 7,05 percent in tha guinea-com level. The rate of Substitution is 0.74. TIic cocts of the diutä .increr sod as the cassava level increascd. (b) Cassava replaoingraizt (startors): Tha solution with regerd to cassava replncing maize in the diet is presuntsd in Table 5.4. The - 93 - TAÜL5 5.4: Oemposildnn and costs nf 'stortdr dicts with csssava replacing mraizn in the dicts *-• » • Ingredients and Nutrients r.n mc s MC 3 ’C4 MC5 MC£ MC? - » • * • Gassava 0 G.on 10.00 .30.00 r.bizo 55.77 49. 11 41.06 14.55 Drewer*s greine 3.53 4.37 4.75 4.27 3yn. Methioninn 0.25 0.25 0.25 0.22 Blood Deal 10.51 3.07 7.12 3.32 0.30 Palm kerncl Mcel 15.00 15.00 15. JO 15.9 0 ^ ^ .0 0 15.00 3alt 0.30 r» op 0.39 .511^ 0.30 0.30 Grcundnut cake 11.97 13.07 1'\7? ^*£<01 25.99 30.0G Ad Vit 0.60 ). .T 0.50 Ü' ß*fiC Viheat Dran 1.30 ' !crt & Bone scrap 2.19 2.63 2.03 1.30 0.23 Synthetic. Lysine 1.12 0.23 B•o *n •e ’ '' .’ .■a*a••l•* ** * * * **■* • -* •* * 90.20 ’ - * -* *-* * •* r~* • " * • *• •* * *•* Tryptophan I- J « 0C.C9.. ‘ T0.2v2} 0.23 1.24 25 0.25 0.25 Fibre 3.45 \ 3.53 3. C0 3.97 4.11 s.24 4.11 Fat /* • »c-:>*. 4.50 4.5-i 4.51 4.50 Calcium A ■ '.60 . • oO 3.50 U.G2 0.25 Lyoine / \ !Qp 1.30 1.X 1.25 1.25 1.25 1.25 fviethionino 0.50 0. 50 ■\5C 0.51 0.50 0.5Ü 0.50 Riosphorus { 15 perccnt level of cassava at 0.03 perccnt L ..'•ot.siffij .23 .X'-r-i-O. I r the perccnt cassava diet, to rnake up for nw lysine content v cassava. Substitution in the ?nnrgy bt.cn ingrndient resulted in a Subs­ titution rate of ''.73. The first 5 ptcv.ont cassava increase resulted in a deemase of pereant in the naiy.e centerit. Subsequent increases in cassava up tu tha 25 percent level resulted in an average of 7.3 perccnt denreasr, TI*.* last addition vT cassava from 25 to 30 pcrcent resulted in a 5.52 perccnt decrease in malze level. The costs of the diots als increasc-’1 as the cassava level increased in the diets. ~ 95 - ( c) Oissuva Replaninp^ Ojinca-ncrn (Finishers): Tnble 5.5 gives the «-umposition of the finisher diets in which cassava replaccd guiriea- curn at varying levels fron Q to 40 percent. In the Finisher diets, the cassava levels were increased to 35 and 40 percent levels. Equo- lity restraint was also used so thnt the stated levels wcrc includsd in the diets. These a m showri in Tablcs 5.5 and 5.5. Changes in the protein components of the fand were very i.iinor. Groundnut cake was included in the rntion as from 20 percent cassava levcl. Increases v/ero 0.99 percent with sach 5 pereant inercase in cassava level. Increase from 05 percent to 40 percent cassava results in a 3.1 per­ eant inercase in groundnut cake- m^baximam levnl of groundnut cake in the diet being 7.71 p e r o e n t ^ ^ i s low level of groundnut cake was hav/ever compensated for in dthe diets which t>.X)k up B'icod mesl at the maximum level pcrm.lt tso In the energy bospd inirradlents, the rat® >' Substitution of cassava for guiaan-̂ Q'sr. v-.c 0.01.With ?r.ch 5 p e e m t iniirsnae in cassava level, therc- was e. :k -.raas if ö-a-jt •* percent in rho lsvel of guigea^cam in the eiet . lig ; /••■ riw :ions necurred in the mineral cor:pon&g%6 such us b-:-nc m»-#;'' > oc' . der Shell. Lysins was complßtely excludec1 and other ingretixe••jm-.ined constant. The costs of the fand increased as the cassava ievr.~ increased. (d) Cassavo Rlejalacinĝ j/oi^e (.Fini sh vrs): The composition and cost oF diets in which Cassava partiell/ ‘-eplnoed malze appear in Table 5.5. Slight changes occurred in the pr' tcin oomponents of the feed. For TABLE 5.5: Composition and costs of finishm tiieia w'th cassave replacing guinca-corn & Ingredients Nutrients GCC Q GCC .9 L30C 10 GCC 11 GCC 12 GCC 13 GCC 14 GCC 15 GCCPrices Oassava 0.00 5.00 10.00 15. Pf 20..' 25..V 30,5$ 35. og 40. 0 Guinea-corn 56.21 53.51 44.86 39.2* « '•' / ?ß. 33 14.05 6.5' Palm kernel manl 15.00 15.00 15.00 15.' 'i i~j. y 1$. 15.00 15.0: Oystor Shell 0.51 LJ. 44. 0.4C ü.3e o.r-j p 0.44 0.54 0.41 Dried Yeat 4.00 4.00 4.00 4. tJLi 4. Üi A « Ui •> 1 4. OG 4.00 4.01 3one Msal 1.35 1. 42 1.76 1.7t Brewer,s gruins 5.12 5.60 7.00 7.0C V.'heat Bre.n ?.'X' 7. PO 7.00 7.0C Salt 0.30 0.35 0.30 0.31. Ad Vit 0.5T . 5P 0.50 0.5C giood Maal ö .7l: 9.92 10.00 10.00 9.47 t3. nut oakc *• 4.61 7.71 Synthetic Meth. 0.23 0.24 ri p/i 0.24 . 0.24 0.2a 0.24 0.25 0.24 Calci' Ji.n 0.81 0.34 1 j , UC i U.92 0.93 1.02 1.11 0.80 1.20 Tryptcphan 0.20 Vüfl9 0.20 0.20 0.20 0.20 0.20 0.20 Fibre 5.00 5.00 5.00 4.90 4.91 4.84 4.77 4.74 Protein 21.02 V 20.45 20.10 20.33 20.24 20.16 20.10 20.5U r^hosphorus 0,o£ir 0.80 0.80 0.60 0.00 0.80 0.80 0.80 Fat y M . 3.50 3.45 3.34 3.23 3.14 4.04 3.02 Methionine 50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 Cystinu 0.23 '1.27 ' i. 2 2.27 1.27 0.26 0.26 Lysine 1.30 !.’• 0 1.7 1.29 1 30 1. 5C 1.30 1.30 1.30 ____ .. ?8p0 2300 _ □ _ . 23Ü.J _ _ 2000 20;>J _... 2300 _ 2800 2300 Cnst one ‘•.320 175.29 i79-16 133. ■JJ 187.Gl 191.33 196 32 230.69 205.35 210.2C ton of fecc .vith cassavat: :;27Q 176.66 V .09 T ’9.51 131.30 10-3. 52 135.69 187.85 190.2: prices . 220 17*.10 173. itj 172. *>i 171.35 •|71 03 170.69 170.35 170.2C stated. •:-i70 171.66 16 .OS 164,51 161 .36 158. 52 155,69 152.05 150.2C •::i20 169.i: 1 “1.•9 167.01 151.36 145. 02 140.69 135.35 13Q.2C GC00--15 denotes the finisher diets in whicn nassavu partieO ly r^pla-ed guinea-corn in the diet. - 97 - TAOLE 5.6: Cbmposition and c.osts of firiishnr ’dir.t. .vith caasava replacing malze - Tngredients Nuirisnts Pricts MC 3 MG 9 Müll! Mt n MC 12 i-i:i3 MC15 iV C 1 C Oassnva - 5.00 io.no 15. an ?' U 00 00 35.00 40.00 i.taize 53.31 30.02 41.84 33.4! 27.17 14. 6-4 0.60 2.52 P-kernsl menl 15.00 15.00 15.00 15.CG 15.00 3 15.00 15.00 15.00 Oyster shcl!9 0.06 .. - -- V Drisd Yeafet - - . < 1.6.; 1.40 2.22 3.04 Bone Meel 1.7S 1.70 1.10 1.53 1. :■ •• 1.77 1.75 1.72 Brewer*s grains 5.1t' O. 15 7.00 7, ^ ?. jr 7.00 7.30 7.00 Wheat Bran Salt 0.30 0.3.' . o. i 0.30 0.30 0.30 ‘Ad Vit 0.50 0.50 o.?j s y . O.50 .50 0.50 0.50 Blcori deal 9.69 9.07 fror 9.a.'; t... t-..- .1 3. X 8.52 G. 25 G. nut coiie £.41 5.14 j,0\cio irj,r}4 1 1 .&- 12.52 13,15 13, SO 14.43 Smy .»n. mmc *thioninr U.23 0.24) 0.22 >:.27_ 0.24 • • . • « » » - * * * • « t » » • •• -• . . . .'V.26 £L25 0.24 Calcium Methinn. i.Pf) 0.30 G.oO ll.LVJ f _ » f.* 1 1.00 1.01 1.04 Tryptophan 0.1G 0.20 : i.20 0.21 9.22 0.23 ribre 4,40 4.54 4.66 4.7?. 4.72 4.7fr 4.73 4.01 t Protein 2-1.20 21.22 21.96 22.00 22.00 22. X 22.00 22.X Phosphorus 2.»: ).77 ?5 8.30 O.CjÜ i 0.00 0.30 4.12 ‘ 4 . .V' 4. no 3.87 3. r 3 .5 t- 3.44 3.30 at Methionins ■ . ~ 0.5 5.5 *n. 1 >r.. rt c 0.5 0.5 Cystine^ 0.70 * , ,X1 0.25 0.29 o.r* 0.29 0.29 .28 0.28 Lysins 1-3*. 1.30 1..,ü 1.30 1.31 1.70 1.37 1.30 1.30 ^.ergy........ 2600 2G0CI ^l>LI j 2 X 0 2 X 0 2fu. •_ 2300 2300 Cost üf 1 ' 193. 76 194.99 19.3.64 190.95 2'3 .4.7 204.' ■■ 5*i ■ 209.15 211.70 tan of_ C’270 192.49 191.34 191.45 191.49 191.7.4 191.30 191.65 191.70 êeri with "22 j 1C9.99 173.54 133.35 161.43 173. 175. X 174.15 171.70 187.40 ':hnsQ pricetP 101.67 175.43 171.40 165.34 151.50 153.55 151.'70 of cassavaOl?’' 184.95 175.64 136.95 161.43 154.34 146. SG 139.15 121.70 - 90 - nach 5 pereant incrcase in the oassava tn/cl, blm:! mcal dccreased ijy 0.10 pcrcent. Front 10 tu 15 pereant cassava lcvel, the dexrcesr is C.75 pcrcent in blond meal. Front 3 percent up t' 40 perernt cassava lavol. decreascs ere ebout 0.25 pereant. Für grounclnut enka, there was no dis-- ccrnibla pattem of inernaso 33 the cassava luvrls increcsed. Howevor, front ü-4ü percent cassava lovcls, thr: yrounrdinut cak̂kü» lIesvccllss of inelusion variod front 6.41 tn 14.43 porcent. Slight changes opeurred in the mineralvtonpunents such as bone meal and driscl ynast. Dried yeost whs cXimineted from the cliets with ' —20 percanc cassava. dyster Shell 'was also exeluded front tha mix for all cassava basec diefcs. Th- rctii:so«£f^ Bsiu.b!sbtsititutinn of cassava for mnize was 0.7G with riecresses of aba^>^|:- pereant in thr rtaize component for every increass in the aassayf loyal. Slight nhanrias ocourrsd in the methionino levcls whereas synthatic Iv-sinr; was :xciuded complctnly. The m s t s of thr riiets inci-'sssed ns tho cessr-vn level irtproased. 5.5 hxperitr... . ifal Fo^r^pnsec.ttive sxpr-rrts wäre ’a.rr! ad out using broiler chickens of tha opbb strair. Thr rirs!- t*n crooriments (experiments ITland TSj) involvac! trials with start.’r c-inks (->*3 weaks), v/hils the last two experimonts (V and VT) inv-jlved <'x ishers (iS--‘l2 wnrksj. In the first 1:wu . xperimer,1 s, birds w°ra randomly distributed in the pens such that oac-n pur. '•nii birds. Euch di et heb fnur replicate groeps of 3” bir is roaking n total of 120 birds pnr diät. ~ 99 ■■ In the experimant ■jXIIdicts GCC1 - GCC? w e m Computer formulatad such that cassr.vn Flour p rtially replacod guinea-com at varying levels of 0, 5, 10., 15, 2ü, 25 and 30 pereant respentivsly. Twa commercial diets derioted commercial I and III wäre? also used for comparison. The composifcinn nf the linear programmed diets used in this trial appear in Tabls 5.3. In experiment IV. the set up was exaccly ehe 2,nc es in experimo:nt III except that the diets were progranmad inr ^such a! way that cassavr. ■ flour replnocd malze rather than guinea-*qnrn in the trst diets. The corflpositioi iS of tho diets used in this ^rperiment ar<3 giv/nr in Table 5.4. Also n different commoroiol eilet vias used instead of commercial II in this experimont and is acoordingly denr.ted an commercial III. Experiment V and VI compared the responsiveness of birds to x/ario 's finishor dict in which cassp.v/* rcplaned guinea~corn at levels 0, 5,1.'., 1 0 , 2Hf/9gJV30f b5 und parcent (Experiment V; or meize at the srmc levels (experiment VT}. In nach of these exper’iments two nom.:inrc.iâ firUwnar diets 1 nnf III were used es Standards against which iUäV^neer Programm:, . Ir-^3 were compnred. ihu birds used were also vobb broilers which ■.»*•*• r> rnr.uomisad into pens such that thero were 20 birds per prn. Thero -cre fuur replioe.tes pur diet giving a total of SC birds per diet. In all of the above Experiments, weekly records of chang1" in body weight and fee,: consi nptien were kept. Records of daily mortrlity were also kept. ~ 100 - 5.9.1: Statistical anc.ly.si3 qf Östa Rcsults (a) fxpnrirn.cn t III The rr5..!lts Fon this arc prnsonted in Tnblr 5.'7. Tho tcchnique of onolysis of ' verianoo .wns usod. on .oolnmns 3--C of Tnhlc 5.7 (Sfee An ovo tablos 5.7.1 - 5.7.3) which show that thnra wäre no signifinant (P ].n5) differentes bohwnm üv avesrage v/cight gains an 1 n\xxr,rjre fa'öd intako, but thero warn fnr food canvarsion sffir.iency far the b irr's on eaoh diot. To detar.t tho treatmsnts onusing thesc significant (p ".' 5) dif-- ferencos in only the feed convr.rsion officicncy, the criterion of least significont differonco was usqri. It was discovored hfvjt 'c rv.f d , GGC 5 and GDC 7 warn enusirjg tho differnncos. These cüfta for-rNir .••cd prurcst in tsrms of FGE. Howover, pniring the di etc / a\rrSJ sanrpnring them showed tl>at signiFinant (P 11. CI) diffRrcnc^^^pJrrred betwoen diots 3CC 1 and UGC4, GCC1 and GGn5, GCC1 and-^!*. tho. GCÜ2 and GCC5, GT.C2 and 3037, GGC3 and GCC4, GCC3 and GQT.5, f ' jrirl GGC7. Other signifinant differences occurred beiiw^/ diets 3f d, GTXJ5 and GCC7 and conrnprcisl diots I and I T L ^ s S (b) Experiment I'- The ra» prr si'mmarisor! i» Tabla 5.3. - 101 - TA3LE 5.7: P'rfcm^ncu: cnmprri3ons crfr Stertor dicts iri which guinect-rcom is r-placc-d by. ” - 3,-i pure ent: of cassava l'/eight onusr- on Efficiency • •pi l rUrt Cornmercial Onmmerr.ial III *3 \ Mnan Standard Error ^Compositiun of diuts ore i T ü b ls 5 .3 . 102 - Table 5.7.1: AMCA/A Tabl. . f - r «fcigM. :*uns * * *— •*•*** ̂ * * 4 1 * * ...... . .... — ' ' Source of Sum of Degress of Mcan F, . 05 (3,24) Variation Squares Freedom Square» F 'n.ui (0, 24 1. « k A. 4 • • #•#«. ...... . ». . ....* * • Treatment 0.2D o 3... 25 A a s 2.27 Blacks o.on 3 0 .27 ^ ^ i . 3 £ Error 9.2G r>/y ö. i '11 Total 1.25 )5 o n ., , p_# . , , . ^ . ... ... , . ...... J Table 5.7.2: ANOVA Tnble vor FcndJ tInteako t v y Sourc.e of Sum of pejrecp. dt i,1 iean Fü.05fy, 24) Variation Squeres • Fredo ■ ■3 nur y.(;: F0. : !l(ö, 24) ..................... ( r i Treatment 0 . » J B 2.35 i Dlacks 1. 3 '•* • 3. 35 3 Error 24 ..-J19 T rtal ab t _ . . i . Vfa* b• lü• • .5. W..7.4. .. ai'JQ\/A T? oic f r Food Gonvorsion Efficiency .......7 ? »• * I > . « • Source er Sum of üoetroes r_»f ■ oar .3.25(3 ,24) Variation Squares Freedom •.v-.jarcs ' F7.01(3, 24} Treatment G. 594 . rfi»ö r n OOr» Blacks n oq *KJQ 3.159" Ti, 36 Error 5.2711 24 . Zna * Total 15.225 i • *i -• Significant at P ^ . i'.\ - 103 - Parformgncn mmparisana of st'.xt';r diats in v.'!ii.:h -visssva TA0LE 5.3: (0 - 30 perernt) mpl. ;d : 'ti.e dir.te - 104 - Table 5. 3.1s ANIMS Tnb\ . F/r >v ifjht jgöins - 105 - Analysis of variance technique was usec! an colunns 3-5 of Table 5.8 (See ANO/A Tables 5.8.1 - 5.8.3) which showed that there were signi- ficant (P ^ 0.05) differences between the average weight gains, and feed cenversian efficiency for the birds on each diet. The criterion of lsd was used to detect the treatments causing the differences. In the weight gained by birds, significant (P 0.01) differences occurred mainly with diets 6 and MC 7 and commercial H I when compared with the other diets. Diets NC 6 and MC 7 performed poorest whilst commercial H I diet was best. Compering the FCE of the biLrds foir each diet, significant (P 41 0.01) differences were caused by diets MC 5, MC 6 and MC 7 which were the poorest and then commercial I H diet which was the best. The results of the experiment are summarised in Table 5.9. Analysis of variance technique used on columns 3-5 of Table 5.9 are summarised in ANCN/A Tables 5.9.1 - 5.9.3. The tests showed that signi** ficant (P 0.05) differences occurred only in the feed conversion - 106 - * TAQLE 5 9* Pt;rfarncrv:r cls iparisuns of Finisher dioto in which 0-40 pnreent -.osacivn rcplcccd guir .--i ;orn in the diets 1 3 i <2 ‘‘Diets Cassava Average Avereyö Feed/lVeight Levc'ln Weipht gain Fand Intake Fectl Conversion (kg.) (kg. ) Efficiency ■ A SCE ........... .................. GCC a 1.04 4.513 4.363 GCC 9 5 0.005 4.555 5.650 GCC 10 10 0.9 G5! N.493 4.551 GCC 11 15 0.557 1.237 4.94-4 3 ? 9 GCC 12 20 0.879 4.600 5.233 GCC 13 25 O.G^n 4.4,17 5.246 GCC 14 30 A r\«w‘ • . c.. 4.780 G. 536 §> GCC 15 = 1,305 4. \?1 3.327 i GCC 15 Ä s 7*30 4.634 3.33J Commcrcial : .927 V . l 7 4.495 Ocmmerciel III 1.1*5: 4.235 3.651 Mean .. :05 -.40/ 5.131 Standard Error I * . '} '■ 0.20' n.aei; Compositions of dieis dn Tabla 5.5. 1Q'7 - Tabin 5.9.1? ANCV"'. Tpble: f: r '.'cight ö.iinc - 10B - Efficiency of tho birds on each diab. To detcrrnine tho diets r.ausing the significant (P l) diffsrences in FdC of fch* birds nn thase ciiets, tho Isc' statistic tost was porfermüd, binnifi: d: difforences occurred neinly with diets >'•*’ 1?, GCC 10 GO'J 1C- and commercial III whan cornparod wich the üther diets. Diets GT/J 10 ylÄ Q X 15 pcrformed poorest whilst diets GG’Z 1? and commercial III wäre best. (d) Exporii.'ion t VI ........... The results are summarised in Table 5.1*!. Analysis of variancs tochnique usndj m columns 3-5 le 5.10 are summarised in ANOVA Tobles 5.15.1 • 5.10.3. The tust3 shnwed significant (P ^ C.01] differences in the average^ vvr̂ Lght gains and F c E of thr birds on each diot. The Isb shatisbic tsat shuwed that the diets ecu sing the signi­ ficant (P -C. J.• j1 } clif-K-.r'iricos ir tho everage woight g iins and feed corvarsiori efftffeQcy of th.3 birds are MC IC, MC 15 and fÄO 16 which perfomed pqbfesti.. 3o aluo dir! comrnörciol III diet which porformed best :\n ̂ Änjns of woight gains c.nr fand r.cnversinn "ficiency. 5.9 . ̂ Coj-.olii sioi iS The rnsults of the expr-rimonts indicated significant (P ü.ül], [P 0.05) difForyncciS iri the Food Corwcrsion Efficiency of the birds. * - 109 - TAE3-E 5 1 i* Pcrfarmnnc conparisons af ■finisher riiets in which 0-40- Piment cnssnva rcplacnd nvaizo in thc clicts / Table 5.10.1: AflCVA T^blc* for Weigi-.fc Ge-ins 'SirniTicant at P *£_ . 1, - 111 - As regerds wuiyht gains, signifir.ant (P ^ U.Cl) differuncns warn fo'.ind cnly in Starter and finisber dint3 in which cassova rcplaced maizn. In eil the dicts (stercar and finisher) differenr.e^cotkd bc observec! in tha avaregc. wripht gcios, feoä intnkc anri rjonversxon of the bird= althnngh tha aroes whero the.se diffe: ps hevr beep significont werc hxgnligntnd above. The dists Gausing signifionnt > liff erenrcs wäre tha di ata thot pcrformed pcorgSo which ivere thoae in v;hich thr cassave. ctntnnt.s werrj vcry high («X - 40 perccnt). In formulating these dints. the f-at anttfeapt \.us konstant and there was no cd di ti cun of supplBrasrvtary fat^ or oil to rerJoce the pov/ictory nature cansed by hign cae ä-ava ce.1; iton/t [V^uq ht and Stutf"*S-/{ 1954), Che u and Müller1*2 /■ % / v (19'?2), Montillr- et ai - J. 7his pnwdcry nature of the diets reduced ferd i Lake $p^:hö birds. T!si« in turn ratluoRC* thn nutrient intoi'S and e onsogde,ntrv lad to reducot! yrowth rare. In the uxots whern theiru r;ere no significan i. differences in weight gaips’1 (Start or und finisher diets in whirh ocssava replaced yuinea- j there i3 an indication thafc nutrients wäre equally avnilablu to the b.irris in adeguetr- and olmosi the snroe qunntities bsceuse the dints wäre campoundcd to be outrition&lly balanced. The main differonoes would then br in thr: retums over fseri costs for eanh of the diots. This is estimatu.:; in the uoxt ohcipter. Although, grovvth is sjppresac.d • -er. e-‘ding tn the nation of various - 112 - authors, ib is pertinent to nots tnat even the diet with 4ü percent levcl of r.QSsau'a is still highly tolerable to tho birds. 5.10 Wcight iksponso as Cassavc Invel increased The notian thet hns been bald to this time ir> Fnct that higher lovels of c&ssava in ehe foed imp.Lrs greu wtb rate. The samc has boen nbsnrved in ths scriea of ujxxpocrir.itmtd performnei in this study as shown in Figura 5.2. Th-' dists wäre br.Riocac! nutritionally with cunino ecicis as srggrsted by Ohou and tfjllar 12/' as well os with the other nutrients. The diots also oamo; out tn ba isoproteinnceous and isnoaloric ns suggnstoJ by v'ariu-.fc» Authors (Armas and Chicco,-• Olsen, ut al. 57/ . Th•uit_«oe Vthors concluded thrt up to 50 percent •.nsisava could be used für chisks without anv deleterious effects. Üthor fectors whien co'Hd by responsib]c for decreasing woight gains 1 / aro thercrfore, the S-EfJ contents af tho ca3say..t used, (Adegboln - , Enrique? and ̂ s s , Klein and Barlowen , Montilla, et al “• 1975) and the pbufdfeTy characteristia naturc of cassavn, (Chou and Müller — and Stute Randen c\ al % f.tontilla, et al — ^ (1970). All the authors with thr notic i thrt it is the HCN and powdery nature of nassava tnat affects grov/th sorgest the use cf cassavn at levcls higher than 15 percent only aftur the Pcmrth week. Since other authors (Adcgbola, Gedeih.". ei' al.) have proveu riethionine as a moderat L, i <. t.r . . : (eq. 6.2) where Protein>ifffek»:. onci thr. other terra s aro as defined corliex Two functionai equ:• nrf :r5 nsuiraoted arc srated in the iraplicit form as Quadrat:iicc & '• ^ br,P + \l (eq. 6.3) ' t„ Square Root W + 4. y (eq. 5.4) From the data obtaineri fron the experiments described in chapter five (Sec Appendix C) the recression pararneters vvere estimated for the startcra and finishers. The raethod of first differenre was applied to oumulotive woekly figur’es in nrder to elininate autoregressive distur- bances. The figures used therefore revert to the weekly values. The protein enoff J.cient is oxpected to bo positive since intake of more - 110 - pfotein is cxpoctcri to result in incroased woight gains. The empirical results follow. (b) Empirical Results The quariratic functions havc been sslected as tho Jo.Ja equations for the f jllowinp r aasen s:*- (j.) Tho aötimatiTp connhlun does noi seriously contradir.t theorrti.nal and a priori exprrtations as to the signs af the regressicn cnefr'i; ia.nts. (^i) Mary of tho rogrossion qnerfitöpts are signifikant. (iii) Ins poßfricient er multijriL.’ ^t?trrnir;ation (n‘~ j is such that the für oticr. previrte» a ganr1 fit to Lho data as measured by F-tost and, (iv) The rosidi t1 a^c not serial.'- anrrolated as tosted by tho Durbin-Wntson tcct streistins. Tho rcsults ara prescntch in Toble 5.1. In all instdtaccs, the protein ooufficients hn.d the sxpectod positive signs and werc significent ai the one percent levsl of probability axcept in the Starters and finishers whcre guinoe-com was replacod by casssve (experir.ients III and V). This irnplies that protein intakc is a significant explanatory variable as far as liveweight gain in birds is concerned. Protein in the Starters explained 54 percent and in the finishers 30 percent of the total variations in liveweight gain. These low values of r) could bc d"o to the fast that protein is not the only - 119 - TA3LE S. 1: Effects of protoin intake Qn liveweight gain > * • ■ Equat- Experi­ Dependcnt Constant Independent C*. j* 0,vD5 ian No. ment Variable Term Variables R dL d(J ---- - . f. * ki • S.5 III W S2u. 5 ; U.! Ou44 0.39« 1.03 1.33 1.40 (n .00099} 6.5 IV W -102£.5 12.G2 -0.02 0.53 1^.55' 1.10 II 11 (d. ?3(i'.r-PS)" v IV 3<"-f>3. 0 ?.!jo -0.0C317 0.< V 1.03 rei> • ow-O, « II I l VI '!■! i -51i3ö. 14.29 ~ü. 0005 ■ S r '■.90 1.52 1.41 1.52 (7.01J- (n. c5ir~r j........ i.......... » * * Experiment III • Starter dicts in v/hir;/ eva roplened nuini — corr. " IV " ” " M Oaizi ’* V Finisher " guinea-corf- VI " ” 9 rmir-.A O Figuros in parsnthoscs a m standarr rrrnrs. dL - Lower table valne nf Dv-rjsfl n~%t3on ciU •• Upper tabl . valun of Durhin-wetson + Dsnotos signifiernre at P ..01 ++ Denntes signifiuance at P *»■ o 15 is the cosfficient of multiple dstermination D.W. is the Durbin Uatsnn cdäculated vai' fco F or telsrance lsve|. being insufficient for further computation. - 120 - source of liveweight gain and in cddition, it nould bc that not all the protoin taken was avsilsble for tho birds* mstabalism. The Durbin- Watson tests Show abSence of autocorrelation except in equation 6.6. 6.1.2; Effect of Enpr^jy intakg _ orr^ivcoweight^ ̂ ain Energy in the diots ie suppiied also by the different ingrodicnts v/hir.h make up the fsed. Energy velues used wäre therefore obtained by using the Proportion of onergy per kilv.gran of feed and the total feed consumed by the birds. Average -.v ;ek‘J. fig. tres werc regrsssed on average weekly liveweight gains. (Se^ Appendix j c T (a) Estimating Prooeduros It has beon steted earlier in sention 5.1.1 (a) that weight gain is a functinn of feed intsk„e 0 < c W f(X / ..... V] Di. L /\ - f(p, e , (eq. 6.9 whare E -- Energy X •= Fand infc-ake A tlD Ar.ids Other vt es ere as previously defined. " W ^ f(E,/ V) (eq. S.10) Two functional forms of equation 6.10 are eetimated. These are quadra- tic and square root and thoir inplicit forms are expressed as follows:- Quadratic W « b O b„1 E b^dt " + vv (eq. S. 11)4 Square rojc V « bQ b^E > b2Ee + v (eq. 6.12 - 121 - The regrassiori Parameters are also estimated for the Starters and finishers separately. The energy cnefficient is expected to bs positive as livsweight gain should increase with increasing energy intake. The nrnpirical results are as shovvn below. (b) Ejnpirical Resuljbs The quadrati»- f-tstiooo hnve been seloctesd for the reasons steted earlier in seati.v. 6. 1.1 (aj. Thn results are presanted in Table 6.2 It is only in the ase ef Starter diets in whinhN^fcsava repleced guinea-com that the regrnssisn porainotar is not statistically signifi- ennt. Thn c.oefficisnts boar th~. expert® positiv'!’, sinns. These Show that energy intake contribi tgs ainr^^^^ntly is tht v/ariatians in livswsight gain of the birds. Thi1 s' MN s> "n ifii-med tna F—tests which are signifikant st the ono pdrcqnt 1 .-vel excnpl for Starter and finisher dints in which cesaavo rrplaaei1 crvinea-com (experiments I I I and v ). However, energy cxplainscf •'►i piTcrnt in rhe Starters whereas it explained <14 pereant ln ths finishers of the total varistions in livowaight gains. These vsl'es higher than these for protein .cre low. This coole! be beot»ae o n y is not tho only Sr rno oP livsweight gain or probat!/ that samt: of 'the onergy intcJ • « • ( »- « 4 Experiment IIT ~ St.artar diucs in '/'hil f eassava replam ? r-jinsa-corn IV - Starter '.'iets in wbi'jd-, ce :3avc. ropiarer- :naizc V - Finisher diets in whict ■ '.asae.va roplaaed guinua-corn VI - Finisher Uiots Oivr.pava t- pleced malze Figures in parenthescs arv Standard error.-; dL - Lover table v/."luc of Dui -bin- ̂ fcson dU — Upper table v/aiue of Qui'bin—w'atäon <■ Dcnotes significannn ut P JL. H- Denotos significonr.o et P ri.~>5 R is tho coefficient of multiple detarmin::tinn D.W. is the Ourbin-'.Vatson calculated vclue F is tho F-test statistic^ (a) Ommission of rogressien term is due to F or loleran- :e levsl being insufficiont for further cumputation. UJ -rl - 123 - amino-acids mnst li!;oly to be limiting in pnultry feedstuffs (Fetuga, 21 / , et a*lr\1975j. It is the reason why it had boen necessary to Supple­ ment thom with synthetic sourccs in the focri. (i) Ly sine Lysine intake values wäre calculated from the average foed intake valuos (See Appendix n). (a) Estimc ting Prccer irc3 The ralatiortship botween livnweight gair: %nd amino-acids intake had b m n establis/.iin the last section. / « ... V) X .= v(p, 2, A, V) vvhere P -- f(L, MC, V). ■ ^ A is r.ampnsed mairily of lysinr- £rö mothionine cystine. = f(L. V: / A (eq. S.17) vvhcro MC = pv.is cystine intaks L = LyfsKz;no* • *>tht?y Symbols are as previously defined. Tvvo Funct'ional FagjEitt.is nF nquotion S. 17 are estime.tod for the starter­ end finisher? ßeparntely. The lysine coefficient is expected to bc positi liveweight nuin shoulc! increase with ineraasing lysine intake. (b) Empirical Results The empirical results ctg presentnd in Table C.3. The quadratic forms are the lec.d aquations. 124 - < ? r TABLE j.3: Effects of lysine intoko on livcweight gain .... Independent Ec;uat- Experi- Dependent Constcnt Variable ion No. ment Variable term D.W. ot c 0.05 L dL dU ö. 13 III w -16.°. 24 553.03 -120.70 0.013 0.17 1.03 1.33 1.40 ('420.17) (435.73) C.1S IV * r 103,'.91 -457’ 0 14.5 1.21 ( PfJ 1.9 ) ‘* (190.7)*’ 5.20 -■ ■•294 . 3 . 39.06 1.5'' t • | 1 The olasticity r w r;l3 in.orrasing retiims tn scnls as far as prntein is concemec' and much "*«r so thct it is in the starte period. The fact that it is greater thnn one r.lsc indinotes that the intakn of one unit - 130 - of protein rosults in a morn then proportinnate inoresss in the weicht gain of the bird. This ocnurs at the mcan value of input*' hüwevnr. - As morc inputs are eued, denressing rstuma could set in. (b) Protein in Finisher Dißts For the brnilor finisher, the production functi s expressed in equation 6.Ö vvhich is stutod *is W - -5150.7? 14.29P - 0.006J (ref. eq. 5.3) MP =• dP r: 14.23 - 0.G17P (eq. 0.31) At- --t-h-e- -m-a--x-i-m-u-m- -v-v-a-i-g-h-t- -g-a-i-n-,- -M--P > “ * Iirf ÄdP. ü ' aI * '* / . r>;Then P « V-*-T"\'. > •- üfeOl.nSfüE- grrrm» - 0.C41 kqs. (ii) Maximum 3utjj.it: "Th- output contributod by protein is obtainod b> sr’ 3tit.itinc; tiio jaiuu of P inte; equation 5.3 stated above. V7 4^g/ö0.77 1F 11.St - 6005. 30 336.23 gramj » 0.335 legs. Tnir is thr maximum contribuhion of protei.- to the total weight gain. (iii) The elasticity qf Pro,ijc'~i ~i Ep As defined for the Starters it ran be expressed es - 131 - F 14.29 P - 0 , 0 1 7 _________p P “ -5169.77 + 14.29? ̂ 0.0005h - 3.59 > - 1 The elastidity of production also shows increaslng retums to scals in the use of protein during the finishing period. It is pertinent to note that this elasticity is derived at the mean value of input. There could be a point when decreasing retums will set in. The intake of a percentage increase in protein results in a more than proportionate increase in the weight gain of the bird. The proportionate increase in the weight gain of the bird is however greater in the Starter than in the finisher. At greater values of protein intake, say 1600 mgs, decreasing retums will set in. This is shown as follov. E 11432 - 10000P “ - 5169.77 + 11432-5440 = 8§£. 23 - 0.67 < 1 (c) Energy in Broiler The productioi as expressed in equation 6.14 which is stated as W - - 1027.; (ref. eq. 6.14) (i) Marginal Physical Product (MP); Taking the first differential of equation 6.14 gives (eq. 6.32) At the maximutn weight gain MP = 0 Then, E tnjBös; - 4333-33 K«ns/kg. -132 - (ii) Maximum Weijjht Gain cantributed by ener-gy is obteineri by substi- tuting tho value of E obtained abovs into equation G,14. V/ = 1027.20 + 4576.67 - 2253.33 = 42C. 14 granjs - 0.426 kgs. o ier, it can 1327.2 + 1.04E - Q.00012E whare E is thc mean value of enurgy. input E 2166.67 k-.als/kg. Ep - 8.21 > 1 . This olasticity nf production revoals increasing roturns to scale for energy intake tot) during the starting period. Sinne the elasticity is greater than one, the intaku of one unit of Energy rcsults i.i a more than proportionate inorcaso Ln tho weight gain af the bird. This also is thc Situation at thc mean .r-1 -lt- cf the .input. Lnr production functicn is as exprossed in equation S.1S whin'- is iter.q... as W = -4340.97 ^ .90t - Q. J0GÜ4E2 (ref. eq. 6.16) The first differential of equation 6.16 giv/es - 134 - 6.2 Nutriant Combinotion Effects 6.2.1: Effect of Enorgv and Protein an Liveweinnt Gain In the previnus scotion it was established that enorgy and pirotein individuolly, cxplain only a sinall part of the variations in liveweight gain. It has thersfore becamc necessary to Highlight the effecte of the two graups of nutrients an livsweight gain. Most of the ingredients making up the livostock Pcr.d havo .r.lso Leen gröupnc! into thesc major nutrients sources. iz) Estimatinn Procedures Livowoight guin .an t*«refard b” said ln dopend on thesc two ma ior groups or nutrients. This can be expressed as follou^: 7 » f(E, P,/.(.... V) (sq. 6.34) where ‘7 - Liveweigbt gain E =— uE.»n vergy intake Bf otein intnko Error tem. The two functional equatinns estimatod aro stated in the implinit form as Ouadratis W - b -l- b„E b^P b^EP -V brE2 + b„oP2+ V (cq. 6.35) Square root '7 -- b^E b^P + b^E2P^ + b^E -i- bj-P' + ^ 6‘^6) Using the calc iated data in Appendix G, energy and protein intake yalues worc rogresscJ an liveweigbt gain The* enorgy and protein coefficients ore expneted to bc positive sincet coro cf their intakes should result in - 135 - incroased weight gains. The empirical results follow. (b) Empirical Results The guadrakic functions haue been selected es the laad equations for ressans stated earlier. The results aro presented in Table 6.5. metabolism. Anothcr factor may be due to tho fr.ct that intakc levels were used rather than the percentage levels in the diets which have been by Flinn, et al.-^ in _ _ Jrom a nutritional view pr.int, the protein level of ths diet is a potentially mare accurate causal variable then protein intake. It was not possible tc use protein level ns an explanatcry variable in this annlysis ber-ause the studies 2 wäre conductcd with isoproteinaceous or isonitrrgeneous diets. The R vrlues for t*.eacconibined effects of pr stein and cnergy rJid not differ from the velves far their individual effects. This may be because protein - 136 - TABLE 5.5 : COMBINED EFFECT CF PROTEIN ANÜ ENERGY CJN LIVEWEIGHT GAIIM • ft— • • •>«. • ft— • • • •- • »- • • ftA«. » 4.1 ( Experi Dependent Conotent Independent _Variables- I F D.W. — 0.05 Number ment Variable term P dL dÜ „ . — . . „j 5.3/’ III VV 607.7 fa) -0.012 0.00009 0.033 0.42 1.81 1.26 1.5o (0.015) (0.0001) S.3C IV W -1836.5 12.67 . -0 . 12 H ^ 14. &' 1. io r. 11 (4.73)*' (0. 3.39 V w -6606.6 9.1 . 0.35 s -D.00047 2.3 2.33 tt II(3.96)’" (0.27) (0.0002C)+ 6.40 VI w -4456.8U 12.45 M - 0.00000 0.31 | 4.3o+ 1.59 1.35 1.59 (e.23) (C.OO'XJj 1 ..... . Experiment III - Starter diets in whioh cassevn replaced guinea-com " IV - Starter diots in whinh cassava replaced maize " V - Finisher diets in which cassav/a replaced guinea-corn " VI - Finisher diets in which cassave. replaced maize Figures in parentheses are Standard errors dL - Lower table v/alue of Durbin-Watson dü - Upper tnblc value of Durbin-'.Vatson + Denotes signiflc.ance at P -H- Denotes significance at P <, 0.05 2 R is tho coefficient of multiple doteiTnination D.W. io tho Durbin -Watson cnlculated valuo F is the F-tnot Statistin (n) Ümmiosion of regrosslon term io due to F or telorance lovnl being insufficient for further computation. - 137 - contributas to the snergy* The. Durbin-Watson statistic tests ehow QbsencG of autocarrolation except in oquation 6.38. 6.2.2: Effect of Lysine. Methionins and Cystine on LivcweighTtE bGa in The 'eint effect of the nost important arnino-ncids is cOnsiderod in this section. This enables the dctormination of the ratcs of substi- totion as well as tho elesticity of Substitution. (a) Estin*tinc; Procedures L : . civeight gsin oan be said to be dopendernt also on the amino-acids intakes o f tho birds. This is cxpres 5SOd HS follows: IV = f(L, r!Cj wheru .... L - Lysinr. intake MC Methionine and Cystin intake Cthor terms are as previously dofined. The twc functionn.l equations estimatec! are st^fc^vLn the implicit form as Quadratic bfj b^L + b^iC + bgLMC b^L* bgMC^+V (eq. 6.41) 1 ,t ^ Ä- , Square root W - + b^L + b^MC b^L*MCa + b^L^ + b^C'^ V(eq. 6.42) Using the calculatcd data in Appendix C, lysino and methianine plus cystine intake values vvcre regresseri on liveweight gain. The lysine and methionine plus cystine coeffiuier.ts arc expected to be positive sincc more of their intakes should resilt in increased weight gains. The cmpirical results are as statod below: - 138 - (b) Empirlcal Results The quadratic functions have been sclected as the lead oquations. The results were as follows: (Table ö.6) In all instar ices except in oxporiment V, the ly sine and methionine plus cystinc terms which nre includod in the cquecions shovv the expected oxcept in equation 6.44 positive signs. The Durbin—VV&tson teste show non-autoregressionj^. Rcsponsiblc for the -Immission of sonie terms and low Fl values Dssibly the fact that iritr.ko levols were used r.rther than the percentagu lovels in the di-ts. As discussed in Flinn ct al *Y-*J /' , Lysine levol may be a more , ............. . . . v T acoi rare csa*a.usnl variable than lysine It is not possible tc rogress lysine levcl on liveweight gain buc.ause it v/as constant through- cut the diets. If these constant lrvols are rogressed on liveweight gsin, multicollinearity would be introduced into the model. Thero was little or no Variation in the a^thionino and cystine levels in the diets too. The regression Parameters are significa.ot except for Experiment III. In experiment I]II, only 3 pcrcent of the variabilities in liveweight in gain is explain-.dr Thoy explained 55. 15, 3'^experiments IV, V and VI respectivoly. This implies that thu smino-acirls contribute significantly to liveweight in the birds but that thay are not the sole sources of liveweight gain. Also, it is not all the intakes that arc available for the birds1 metabolism. 5.2.3: Marginal Analysis The marginal analysis concept has been discussed earlier in section 6.1.4 obovo. 139 - TABLE 6. S: COMBINED EFFECT CF LYSINE AND f.CTHIONINE PLCJ6 CYSTINE ON LIVEWEIGHT GAIN ßruatinn Expari— Dopendent Cnnstant Independent Variables Mumbar rnant Variable terxn . 2 R2 F D.W. 0.05 MC dL dU 6.43 III W -735.6 1193.1 (a) -203.1 -102.2 0.03 0.24 1.81 1.26 1.56 (2614.7) (521.4) (166.4 £.44 TV w -17D5.5 (a) 3777.p (a) 224.3 0.56 10.2+ 1.20 H If (1373) ■' (165.2) 5.45 u V» -4)147.96 -273.70 4311.19 («) 0.15 1.46 2.27 »I II (313.5) (2373.69) 6.46 VI w -326''.'* 1511.73 (Q) -429.79 . 400.11 74.40 0.30 3.37 1.7S 1.35 1.51 (1 '93.73) (172.86)' {264.03) (4:48.84) Experiment III - Starter diets in which cassavo " IV - Starter diets in which cassava V - Finisher diets in whinh cossava replaceri guinsa-oorn VI - Finisher diets in which cassava replaced maize Figures in parentheses are Standard errors dL - Lower table value of Durbin-Watson dü - Upper table value of Durbxn-Watson Denotcs signifi.-.ance at P < 0.01 +•?* Denotes significancs at P ^.05^, 2 H is the coeffioient of multiple determination D.W. is tho Durbin-Watson calculat* d value F is tho F-test statistic (n) Qnmission nf regrussigfi türm is duc tn F or tolerance level bcing insuffioient for fort her cumputetion. - 140 - In this seotion, tho conc.nt is rnads usc of in dotermining tho marginal rctns uf tjbstitüuxon bntweon thess various nutrients and the elastioity of Substitution nf one nutricofc für another. (a) Ensrriy and Protein (i) Mni^inal Ratos of Substitution (i'.IRß): ’rhe • n a r n t nAr l of Substi­ tut- ion is defir.rd by th.c docr- *se ii fhc use of ono ni^Mcnt brought about by a unit. incroe.sn in ths uss nf tho other. This concent nssumas ha fact that moximum livoweight gairi can ba obtained by tho birbs with ynridus cnmbinatims of protoin BPW enorgy or tho amino- aoids. Thn intyrost höre 1'v>f. in finding tho Proportion of onc nutriant that will replabc one unit of another nutrient. To obtain tho MP.S for protoin, partial derivatives nf tho pruduction function of\^-5.39 «ro rieterminod. VV =. -Qöfjß.e 9. l i , l 3.36E •• Q.i»Nj7PE (rof. * ß. 39) The partial derivative VLth rejspue-o.t tc prutuin gives ä , 9. U- - fJ.hüiK7E (eg. C.4?) The psrti Lai de rivative- with rcspart tn onergy Ls gl van ns follows: dE n .3 6 - Ü .00G 47P (cg. ö .4 ß ) Tc obtain tho quantities nf onnryy and profcein that will give meximum liveweight gain squations 5.47 and 5.43 a m cquated to zero. Solving yields E « 19361.7f2 kcals/kg. P = 765.95 graiiiS. - 141 - Marginal rate of Substitution is obbeinRc! st the mean valuc of thfesef- inputs whicb are E - 3c"GO. 051 kcals/kg, and P = 302.98 groms. '•JUS is obtair.od from tho ratio of the partial derivates h rospect to protein and energy and is giver as follows: _ e dE _dW d\V --£3.15 - 3.003GC47E (sq. 5.40) oP dP dE n.3c • H.30043P - -25.23. This means that a ur.it inercasn in ensrgy results in a noro then Propor­ tion-?, to dorrcase in protoin. Tho implication is that a high onergy diet will result in a low protoin diet. 3inco protein supplying ingredients are choaper thatj tho energy supplying onos, it will pay a fermer bettor to uso low ennrgy diets and high protoin öiets. There is howsvar a limit^J^^icw low the energy lo»/cl anc! how high the protein lovel could be. s is bocaooj - diet wich too high calories will reduce feed in1»4= given up for a quantity □f onergy then the Former can mak: usc cr thia nrivantage up to the point nt weich further substitutions becomr detrimental to the birds. The elnsticity of Substitution af anorgy For protnirwLs oxprossed as dE P 'Ep dP • | (ec;. 6.41) P ond E are the mean ve.lues of inputs -jn d protein obtained in sertien (iJe nbovo. P v 332 . 93 *~s “ '25‘2J *963J . 35 ZsK?' The olusticity of Substitution ofx cnergy for protein is unitary since it is exaetly equal to onr. o pementagc increase in the cnergy level results in an equal pemöntage clocreasr in the protein level within certnin limits. In making usr of the adventages of substituting energy for protein it must be borne in minri thr.t they meist be substituted in equal proportic.^. However, the uxtent or Substitution is limited by the oirds* rfcquirements. ( b ) Amir lO-Aqids For tho amino-acids fhq above concepts do not apply because they are essential amino-acida nd aru renuired in definite proportions in tho birds’ metabolian. fho '.lucstion of one substituting for another therefore doos not arise. - V.3 - 6.3 Substitution^ pctwasn Guirsaj-Gorn,^ i.Vaize and Gassava In this study, emphasis has bccn placed cn fho oxtent of substi­ tution of cassava for eithrr mal or guinea-oorn. This scction is to prcdint gain or growtli isuquants indioating thc possibl-, combinations of thc ingrcdicnts which will rnsvlt in n fixod gain 1 uthcr ob.ioctivss inc^'idc prodicting their marginal rates of Substitution in prod"cing a perticular Icvol of gain and prodicting isoclincs indica- ting thc ingrcdinnt ccnbinations Por particular gain lovcls which havn th"3 samc rote of Substitution. Mouel Tra wcijht gair>3 obtsained in thc birds Ls assi.imed to bo dopendent on thc energy prov/iding ingyedients guinea-corn and malze which cassava is also substituting for in ths diats. / A 6.3.1: Geize and Gassave Tal Estimotincr ProKedurcs The relafi.'.riship batwenr thu ^ngredionts end weight gain is exprass>eed inp wie Lmplicit furm a* W - f(K, Ga, f....... , V) (oq. 6.4?) Whcre ' i - i_i jcwaight gain Mz - Maize intako Ga = Gaisgva intako V Errpr ‘.crm. - 144 - Two functicnol cquations, tho quadratic and square root ore astimated. These? are exprossed as Kollows: Quadratic W - h , -v L y y + b X a b ^ t -5- byla" bgivfie 4 \j {oq.6.<0B] Square m n t W ■= b^J’ bgCn b3Mz^ b^Ca“ bgrnr/Ca' +\j (ap.6.A9) Malze and cassava infcakcs wore calnulated fron their amounts ln fced consumod (Src Appendix D'}. These values wore rogressed on livewoight gsin and tha psramotsrs cstirnated using thu nei+iod of first difference o n tho cunulativc values. (b) Empirica] Rosults The quadratic funotions ha solncted ns the load cquations since thev satisfied the. rri.i^<£* ;tntr.r parlier. Thaao are with respect to ex.Äwtsd signs of the incfÖp: idsnl variables, thn magnitude of R^, si gnificancc of regressive pan ek: ers and Durbin-Watson tost- statistics. The square root functions ha.! low.’-.r f. veiuos and signs which worc nontrary to expeccations. Mure corras wäre oxcxudod also. The rnsults were as follnws: (Table C. >}. - 1 4 5 - TABLE 6.7: GOMBINED EFFEKTS CF MAIZE AND CA3SAVA Independent Variables ex. - 0.05 ! iiuftiün Experi- Dcpendent lonstant tnrm Mz Ca 2 D.W. dL dU iMi 'mtnr ment Variabio Ci. L>" Starter3 \,V • 2062.74 * ■;. H7 0.07 •0- J023 -0JJ071 0,60 9.4 1.98 1.25 1.56 (1.01)" (3.17) ( ■ .-'.1094 (i.i.or3i)'f . i.ÜI Finishers :V 1^52.37 1 .CU e 1 .4 -0.30052 0.6’. 3.29’ 2.29 1.31 1.5- (o.eoy: (0.9'. l) (0.9*103^) ;'t Starters and !,7 345.31 0.34O.ÜK Ü. 10 -0.00006 J. 70 22.3“ 1.76 1.51 1.6rFinisi i ?rs (C. H 0 + (o.ia) (r;. J'104 (C.:T'009) n Coefficient of multiple determinatl-un F - o'_e ti.3i.ic - Caleulated Curbir-,vat9on value ru - Lower abla valuc cf Ourbin-Wrcaon statistic dU -- Upper tat1 r> value of Jurbin- -Vatson statistic + - Significant at P V. 0. ’l - Significant at P-C. ^.05 Figurus in perontheses are Standard errors. Tha rogrossion coafficiants arn stetistically significant exccpt in th* casc whcra the startr.rs and finishers are cnrnbinnd. In all cascs, the mai.~- and cassnvn nncfficionts had the axpacted signs, and thc sigrdfioance of the F-tusts nt onc prrcant lovol show thnt thc energy-providii ig ingrorJior4-. r siggificant variable thc modal. Hov.'svar, in the Starters, th: irvq-'M'i ntn explaincd SC pi^Sönt and in tha Finishers, S1 pereant, '/hon tto whcle raarinn ^^■’.i.-jd is considered, thoy axple.ined ?•» pcrr.cnt. Thorc is evidcnce that thoro are other oxplanotory variables not stntcd in thn modal auch as other nutrient Sapplying ingredinnts. availabä lity of ntiit'itJntn,gonntir. oomposition of bird and physinal conch’ct of thu GXpcrim r. There is absence of auto- corrolation, thorefore the estimat rcliablo. (c) The Rasponsa Snrfaco The productior.i s™urfaqp d aLstima~ted j.y obtaining all thc possiblo combinations of thoj t-.vr Crigrfedicnts whioh will rosiilt in n fixed gsin lavsl. Equ‘tien 6, 52 is roarranged in thc form ax bx : x. to conform to equation 6.54. Equation o. 5? jtfr>qmr.s 2 0.00511 » Q - ( .IC • Fi.■YT1OS M7.)Ca + ('!.3tf.Jz • 1.0C!Cm\1z‘“ -!• 345.31-V/) Thon, e - • . J0.111 b * f •. 1 r.- - G***ms ‘ ) [ 34.v Q , ■ i ß ‘4 3 4 5 * 3 1 - ' / •:. ' ■- i Solving for Ca gives Ge "b i/b'" - 4-ac (aq. 6.54) - 147 •(0.1G—CI.QOOOGMz) J (n. 13-0. ddOOSUz)'' - 4(n.TOji 1 ) (0 .34?.1z~3.a0Q04Mz+34S 1 -W) (cq. S.50J The results fcr vcrying lcvml: nF -lai/ intak " wel1l as livoweight • gain arc prcsen toc* in Tnblu 0. response1 is presonted in Figure 5.1. It pmsarits t!u vari--u9 isoquftn..^ ncl isaciinos for vary- ing gnin lev/els anr1 MRS rospuofcii/t rLy. U ) Isoqjjeurbs: Thn iso-luants/SVjwyi in Figurr. 5.1 arc not asymptotic to the input (ingrediin.) ytes! This iniplies that certain output levels con bt tainari Prf 'Ti one input ul ent- whcn the other is at the zero leval, The isoqyunts src linear anc! inPcr that thorc is no limit to the levol tjf ^pfcut and output whinh is profitable*^. It also indi- cates that of tho two ingrcdionts shoula bn uscd in producing tho o u t p u t F o r the output levcl of 0.D kilogram gri'-, dt»; iso­ quant Show.“ declining marginal rate, of oubstit’rfcion na moro of cassav/a and lass nf malze is includeri in the fand. Mi- et poj.nt A « fl.41 whareas at paint 8 it is 0. £"7, This re seit i . significant bocause looking at tho LF Outputs of Tatl'-? 5.- csssevo and maize Substi­ tute fnr ear.h othsr, although cfissaua is forced in at thosc levels. If Cassova is not forced in, only r.iaizo would haue beon included in the solution. - K 3 - TABLE ßjfl: HAIZE INTAKE, CASSAVA INTAKE AND AVERAGE WEEKLY l iv e Ve ig h t ga'in cootoinate* roiNre' ü'! the‘ m m uTTi'a'i' sijreace - 149 - Fig. 6-1: ISOQUANTS AND ISOCLINES FOR MAIZE AND CASSAVA M AIZE INTAKE IN Kgs- CASSAVA IN TA K E IN Kgs. - 150 - (ii) Isoclines: The positively slopuel linon in Figurc o. 1 arc isoclines which connect points of equal slope u r äqual marginal rotes of Substi­ tution betwecm the two ingreüiunts. Tn* points on the isoclines nre obteinnd by squnting thu ratin r»f ■■K\rr_ : o.l productivities l̂o the. inverse pricc rntiov The isoclines n.r- 1 inci-.r, positively slopod ahd arn not forr.nd through the origin- of thu inpuh pinne. Gijjn^^ha oonstant priqt ratios are greatsr thun ;.: ra, tb: isorlini^ inicrsoct thr- casssva axis. Since thev do ihis uv qi* .otities 'crneter than zürn, they cannot serve is scale lineu. nxprojioi, psths} Urev infor that the Propor­ tion of rcsourocs m:.:- Tu.ngo üs hi TiySNEvvols cf output or gain are po/ utteined if the faetrn er sw. rt. riv r-' xo .-hang Gut the changu isat c consii .t r: : t. ‘-ii Ny** linonr iinu. Tnblc t-.D shows the points of eqi ■ 1 rrarg: subuH tutior or prico ratio. ( d) Analy s ^ ^ ^ N In this SSC^^cNthr.. ooncept c4' ioar final analysis i ■ applied to the ingredients es oppliod tu the notrients. (i) (tar̂ in'il Kysioel Prüduei . für trr.izo end casssvn.arc derived by obtainJprf'tne partial dorivetns if squaticvi •••.• .vith respect to malze and caasave. Thr production fun^tion is cxprv.nsi ;<• es '.7 = 345.3l-rT.34Mi: •!- !.irc.* . . •- . «J. 1 -0. >J0Qö'e?lz fcef. 5.52) Pertie.1 derivative with respor- to moj is giver. as dvV dihz =- •■'.34 - l .‘CJr (eo. £ . 5 5 J Tablo 6.9 D ’G S ’2 [•*2 S H :.i‘L G ’O LJ sovdbns NoiicjncDtfci 3H± i-JO S3JIT3DSI 3Ki dD SjX'lLJd ’dD NÖnVAIU3ci •G‘D 319V1 - 1S8 - Partial derivative with respent fco cassava is rri dW C.10 'J.00022 Ca - 0.00005 M? (uq. 5.5?) dCa (ii) Maximurn jLivav/cight Gei r : ••.'t-'.if.-d t>y first finding hho quanti- kies nf mni*R and ca-'savn f r maxie,un live'.v:-ifjhgnin and ijlJerT substi-- titutinr: hack in the original prudoc.kiin runction. The^fc^quantities arc obtainud by setting cquacions >ncJ i. .52 \p xero and solving simeltancou;: Ly to riiv° rosnav- •- 1.44 Kgs. "iai2* - 2.28 Uga. These value3 arc' tl 3.^ cid tu lad t r' *. :?tion S. 52. 345.31 t C.3A( >"■ . V* ■ & $ . ' [ & v •( .';cnn)(440)(440) -- . r m <) (22r> v - "'.95 kgs. This maximum liv aweighi^SAn ’Si b nnnpared to ihr ogtimuj livrweight nein. (iiij Optima’̂ Livnv/aiyUt gain: The Optimum livsw n.ght gain differs fron the maxirv^' ̂üvi: . . irjht gain becauso it tukes the priues nf tho inputs an*! nutpfc*« irito considv-ratior. in detominirg the state at whioh pro~ dection 3ho .tid terminato. Thet is the reasan fnr aquating the. ratio of rrLinal physied producta to the inverse retio of input/output pricos es stated bolow. Ratio of nquation 5.53 and Ü.62 cre set equa]. to inverse ratio of pricos as fallaws. - 153 - (eq. 6.50) VVhcrc cost Df ma izr 'CSu/tori Eqt-iatinn 6. ü. 3*1 n. iß“ Solving rw; These valuos arc substi teteh >, nquation 5.52 .*nd th. Optimum live- woight gai.i Tbtci.voe’ io y gre-tns. di3ct!33^/'.i 3c*ctian 5.4. Ac a higher broiler prico er 1 nwer ingre-- dient price ehe optimal broiler wo.ight changes. The Optimum liveiveight ge in eit e hicjhav bruiler pries of .-'3003 per ton is obtained as follaws. - 154 - ?517 □ .34 - o.üooqp Mz - p.coqqs > ' p m if.'ifl !* uV.;G' I22' Ca - ö . V ’S.ii-.' Jf* (oq. 6.63) Solving equati m 6. >/i \lds riaizn 3.93.'? kgs. Ca&3ayr. ■- CS. 2 5 6 k n . Substitutino th< vt val ue« int;, - c , i n. per ton and nassave pricc is reducod fromOi32G to 239'• per ton. "* ' " The ratin af MP equated to tho invursc ratio of pricos is aivsn as 20. y : 0.34%>*4.C20r3 Wz - C.JCQCS Ha r . ,.1 Q^ K ^ O. jnn?2 Co' .1.30302 Efc ‘ 300, (ec,. o.S.J 220"' Solvlncf BQuatd.c*7 yiBlcJa Fvteira -= 4.030 tags. Cassava - 2.253 kgs. Substit'itirig Lhasa v/oluo: into equation 6.52 giv/ey 105. .07 incja. - 1.3566 kgs. 155 - This Optimum broilcr ^cdght gaic .nlan -.herjgss but the changa is nogligiblo in this caae. (O.SO grams) fiv/1 Ontiniiffri qu&ntities nf ingrodiants für a givon liveweight nain: Tn uptain tt ‘ Optimum qurntitiub of :l> ig-rxliunts fnr a q^vten liveweight gnin, the proacdvr.. is cs follows: Equation 5.52 is rsdnccd te -;be T orr, : it becomes O.OOQI1C>a2 - fn.nnnr’̂ Mz + .isJCa + ( . V ie \'i 343. .31 - » ) (cq. 6.C2) For ta giwnn 1 i vov/eight gain nF . X11 )(0.34Mz~0.OOCKMMz -fl5ö.fl9 ..........C 1' ’' ̂ ' ....... ................... .... . ........ ‘~ U.00C2C (eq. 6.60) If mi ri ii 2 . 4 0 ’-'.Vatson tusts. ahow nbocnct^of1"̂ - Q^h e regression coefficients aro not staristieelly signifipant. It ds only with either the Starters or the finishere tha.t tho cosff.icient for guinea-com is significant (equation 6.67 and 6.Ü9}, The signs are as r.xpectod except for equation G. £7v/hich is for all the Starters. For the Starters, the square torrns show positive signs and there is a positive interaoti i ^ for quinea-com and cassave. Howove-r, the F-tests are significant at the uno pereant leveG of probability showing a joint contributic :he onergy providing inyrodicits to tho liveweight gains I m kh e The? 'J.fnt. &xc?2alnsd SD aerxzrnt a S ehe i/ariätöili t i a s in livsvvsight gain of the birris in the Starters, whilst Gb percent is exo^ainod in the finishers and only 59 percent when the starter3 and finishers ore cnr.ibined. This ncinfirrns thet thorc arr other cxplanatory variables in the modal. (c) The Response ourfr-ce Tho prori’jc tion Surfaoe was ebtained by sstinating oll tho possibln combinations nf tio tv;c :u igrodionts whinh will re seit in a fixed gain level 158 - TADLE 6.10: C0M8INED EFFECTS OF GUINEA-CORN AND CASSAVA ON L Independent Variables Expcri- Dopendent Constant D.W. Number ment Variable torm GC Ca GC Ca2 GC Ca . - ~ . . . . . ... G. -6? Starters IV 5G3.: 0.57 ••2.30 0.0075 0.0032 0.0033 0.03 21.95' 2.34 1.56 (1.351 (2.12) ( 0.00078 ( 0:^)01 7^H ' (0.0024 ) 0.66 Finishers 1.51 1.47 -0.00025 -O.OÜ028 -0.00052 0.60 10.87 2.26 1.31 1.5? (0.03 )'X' ( 1.02) (d.ÜO£2l7J ( 1 00027) (0.0004) tartors c.0 and l*i«l 361.02 0.32 -CfcOGGOo -0.00002 0.59 19.94 j.VC 1.51 | 1.65 Finishers (o. lü f (0.15 .' JOf-3) ( \ •00009) ' * • • • • .1 ? R - Confficient of multiple dct.Tr'minstien F - F-3tatistic D.W. - Celculated Derbin-Watson value dL - Lower tablc value of Durbin-'Vatson Statistin dü - Upper tablc value üf Ourbin-V/atson statistic + - Significant at P 4 0.01 •K- - Significant Qt P < 0.05 Figuros in parentheses erc Standard errors. 159 - Equation 6.£9 is reeirrangRc* in thr for. 'f fix1- bx c. It becomes -G.00002 Ce" + 0.13 Cö (ü.32 <3C - i .fi;m03 GO2 -!• 361.02 - W) (cq. 6.7 ) Thon, c - -0.00002 b - r.. 13 c -=> (C.32 GC 0.00003 GV? 361.02 - V). Solving for cessava with Rquation ö. 5t, gives 2 . V 2 •01-13). +. . .(Ü.13J - 4/.^.00002}.(.32 GC - pjtf8S?.jC . + .361 ..02 - W) (eq. 6. 2(-0.00:K;2) Tho rcsults fnr varying guinoa--corn A f* as well tis liveweight gain are presentod in Tablß 6.11. Th., response surfscc is prcsenlcd in Figure 6.2. It prosonts tho vajrücrjs isoqunnts and isanlincs fnr verying gain lcivols ranging frum 0.5C kg. to 1.13 kg. and f.BS ranging from 0.625 to 1.10. vTbe isoquants s n w in fiu rc 6.2 are not asymptotic to tb^^iput (ingroriinntj axrj. Thio signifies that certain output loyale i V y attf-ined fror onc input alone when tho other is at rhe zoro \ The isoquants are .non-linear nnd are downward slopi-sg^^y^. j inriicctcs that the rate uf Substitution of cassava for guinc5--ctr"‘: rjsolines for a givon output with nore of cessnvn anc! less of guinee-corn. 27/ For the output levol of 0.90 kg. gnin, MRS at A wf- runs at point B, it is -2.20. CASSAVA INTAKE IN Kgs- - 161 TABLZJi.11: GUINEA-CORM INTAKE, CASSAVA INTAKE AND AVERAGE »VEEKLY LIVEVOGHT GAIN GC30RCINATE' TOlKffS GN THE mCDUCTICN ..... ........... SURFÄGE "...... .... t i.« » «. • Guinea-ccrn Rats of gain (icgs.) per -Vcsk intakc in kc?3. •*'. 50 0.70 • ,9C k.4. • • • * . «- • » d £ _ ^ftssava intakn n • 0 1.35 5.777 ' v 0.337 3.5 -0.10 2. 125 ^.210 7.582 1.0 0.40G 6. o0S 1.5 - Q . 5 ^ ^ 1. oon 5.350 2.0 0.153 3.072 2.5 -'.:.523 1.101 3.0 j.400 3.5 3.10, • 4 ^ " - 162 - Isoclines The pesitively sloped linos in fiaure. 3.2 are tho car^'DQnding isoclins3 for guinea-cnrn and cassava since thsy ronnccc of oqual MRS nr inverse prioe. ratio.sf of thn two inpredients. Tahl■ ■ t.12 shows thr points jf equal pricc ratiu nr marginal rr.tns of subsfrfY ’tion. The isoclinos are linear, and dn nat pass ohrough the origln. This indicstcs that thr; mix of ingredients shrjuld nhange aSHxitput xs axpande er produci. er factor price;. cheng,i butNfic- th-»e rate of change “houia ir constant. The isoclines caonuf snrve as v?calc lines because thrsy intcvsect tho maize cixis at quantij/^u^^i lIdc;u.>e than zero (i.r. at negative values) \' (c) J.’arcjiried̂ Analysis The concept o rrrinni enalyai s is applied also to the oase in whioh cassava substituied far guint-a >jom in !~h.- r.anc >r ingr©dionts es w q3 [ione the nutrients. (i) W 1 Piy/siral Products: Fron; eg1 lötion ö.&■' W r, 361.02 ■: 0.3200 V. 130a - 0.0000300 ‘ 0.00002 0a (raf. oq.06E?) dr dGC - 0 .22 - O.i'OOOc- GCEj (eq. 6,72 tfV (nq. 6.73) dba - 0 .1 3 - a.OOOruj 0a - 163 - TABLE 3.12: DERIVATION CF POINTS CF THE ISOCLINES ON THE PRODUCTION “ ...... ......... SlrfÄö e... - 164 - ■Sottina oquations S.72 nnrl a.73 to zcro, and solving öimultanGOusly yields GC - 5.33 kg. Ga a 3.25 kg. (ii) Mcxi.nun livuweight p '-rin :i = obtainod by substitutir® abovc valucs into equatinn 6. 69 . ? ** Maximum livawcight gain = 361.32 1X6.53 422.5 - 853.23 - 211.2*S ' - 1425.5 grams - 1.43 kg. This tiioximuni livewright grin is higher thnn in tha aase when maize is used. It also sllowa a highur ralt io of cassave to guinna—corn which is 3:5 wherens fer r^ssave to maize it is 1:5. Tl • c higher v/eight gair may net bo ustificc' by thha h:igher oost of cassava whinh is includsd in the diut. (iii) pjatjjnurn auanti tios of passaya and guinea-corn required f or a j^ven liyev/pig_ht galti: First, the Optimum quontiti.es of guinsa-corn and cassava for Optimum livuweighv gein re obtained by oqueting vSbe retio of marginal products obtainud in cquotions £. 74 and 6. 1?5 3 to thn inverse re tio of priccs 0.32 - q. oüooe an 22p n7i7 "•* bVonhocbe.. (oq. 5. 74.)32C Or D.32 ~ j.'jr&e GO 320 (iS. 75) 2200 and, - 155 - 0.13 - 0.00004 C 320d (G. 715) 2200 Sclving cquetions G. 75 snc! Vf! simultanoousl/ yiclds 1.361 Krs cf guinea-com, and 3.633 kgs af passava. For y giv/on livnvi. ight gain of 3.30 kn. thc Optimum qynntities af GC and Cu ar? obtaincd by iubutitutincj k? * '00 mos inta thc. farm Cg ~ -(J.13) + J o n iS? . (v.roXVKo.326C - . .02-300* (eq. 5.97) -O.ÜP004 Wh an guinoa-com is 1.561 Ugs, oassava' ̂ t«S0 kgs. Optimum ouantitias of GC and Ca for 0.0 kg. liv/cwoighjbkgain ere 1. cV 1 kga. and 6.SC kga. resp'. tfvely. The quentities for meximum gain . .re quite different for tho quantitius for the O p t im u m livoweight grin and thess i n turn differ fron thc quantitins for a givon livoweight gain ef 0.30 kg. To obtain the seune^üvev/aighr gain nf 0 3C: kn . iorn of cassava and guinea-corn arc^e^Sured thon fcrr mai"c -n,. cejcv. (iv) C]ptimuir J^Cf^'aight g?ir. To obta l- tr, -Ls '•u Optimum quantitios of c.as$oyn * .a rjulnca- ooir. ob+’ainer ar.lior ^rom oquations 6.75 and 5.76 eite fei 1 bank into eu jatior: .,. Gf. 1661.33, Cc « 3S33 gram W «^611 .*.002 +-}- 0.32GC + 0.1 X a •i.‘Jjü03 CR* - O.G-TT? GC" W = 351.02 + 531.63 ; 472.29 - 32.83 - 253.92 li> - 1018,14 gram? = 1,013 kg. At this optimal hraller weight, tho Lix-da nru eben ; wenks cid and could serve as roastors. The Optimum hroilnr reicht gain is lowe^than thc - 1t6 6• - maximum livoweight gain whereas when cnssava substitutcs for malze, the maximum weight sjain is lower than the’ Optimum wsight gain. The optimal livoweight gain whbn neasava and guinoa-curn are uscd arc lower than thc optimal broiler .,-.-jioht whnn cessava r»nd maize are uscd. 6.0 Economic Analysis af thc Diet̂ ' The, * . „ ~ - to make maximum possiblo profil;. In an oconomy wherr little value is place.1 an carcnss quality of oirvis, ravcnueis dctormined by the total weight attainec' durinc; tnu mariner nrrioti and th. cost and quantity of feeü rcquirc':!. Rovonue Wl1.:lci 1 or’c dopend mainly on tho rate -t which foed gr?ts oonvertcd to liveweiu fc gain5 (growth rate), thc "tity of feen required per unit livt.-weight gain (fcsd conversion officienoy). tho c.ost of the feed, and th" market value of the birds. Therofore. to determine which dir K is best or ;o find the Optimum mar^o'j/g‘age , the crite” i • l uscd '.re (i) Th»^y : prouiding '-ha v . i growch r ;tc, which givos the feer. .••*Mver*icliy. (iv) tho net revorv i .■ over foscl costs. For a meanincif'.il ana"ySis, th. Starter and fl .vj fher diets have been merged so that thr v/hnlo roaring otricri Ls •- '.nsl-'orcd. 6.4.1: The Optimum Marketing jsge/weicht To be able to know wbioh ago is best for the farmer to seil off birds, revenue ovor foed costs hnve bcon obtaineri at different ages and marketcble weights. Fivc stagos at 6, 0, 10, 11 and 12 wecks were conoidored in this unr.ly.sis. Th(3 valuc er the weight ge>in up to a stagc is obtrinod by tho birds '.veight multipliod by the market velue. The quantit/ of Pcec! taksn by the bird up to that is multiplicd by tho unit c.ost. Tho differenec botwoen tho rovenue nnd cost of feed at eaoh stagc is thus obtninod. TabiG>6. 13 shws rovenue over feed costs at eaoh of fivc stages of yrcwS^ For diets in which cassava replacud guinea-oorn. It can be .ibaprvrad from tho C columns that for most of tho diets, the maximum profit margln is obtained st clevcn weeks of ago. It io unly in diats GGHI and 3, GCC2 and 9, GCC3 and 10 that a highur rovenue is obtair.od at tho 12 weeks of agc. It is thoroforc best for tho ff'armer to seil off the birds at 11 wecks of ago. This short •ing perlod a3 er v ahlas tho Fermer to have maro batcher iffVotainks t ? roiso within c year- rh* - i.ocreasns the profi* of J!■th*s« fnrrner P'.r •,?nr. Tabin 6.14 Shows tho rovenue over feedI n#tfX5at th. . f iyvo sitages stated ec-, .uor for ■ U-ts ir which cas- savo rijjlace d maizu. The maximum revem, u n: 7. it coern for the diets is also uotainod al the slovan weeks of * . -i for the cassava based diets, the birds yield the higbsst i «t m A an oorliur age. It pays the farmer to scll off tno birds ob clever »/£-•' vhen thoy are just about 1.0 kg. weight. Thi.s is the- poinf at wuich the valuc of — 168 — TABLE 6.13: Revenue over feed costs eit 6, 8. 10 and 12 wceks cf age for diets in which cas eplace’d guinea-corn ( /bird ) 6 Wecks 8 Wneks V- Wocks 12 Wecks i. • • • • « Diets‘ a1 b1 C 1 Gr- h2 r‘2 Ü * 4 a5 Revenue '5 Vslue’ Val •Je Revenue Valuc Valuc» Value Revenue Valuc Valuc Revenue over ov .ernf af of of feed cf over nf Df over • feed bird: feed feed bird eosts feed fand feedcosts feed feedesst cost ÜDC1G8 .371 1.403 1 . 1 1 2 3.621 2.215 1.594 0.90? 2.102 1.15? 3.406 2.239 1.321 3.771 2.45 l»T.2CS 0.364. 1.236 0.672 . 2.35*' 1.75 0.958 1.995 1.173 3.246 2.073 1.36? 3.607 2.24 rn::r«i;.' 0.4/ iS 1.459 1.054 0.663 2.200 1.54 i .a iB 2.113 1.216 3.305 2.090 1.414 3.584 2. l? «.« KV,. 11 ■'.44 1.291 0.051 0.G91 2.094 1 .< m A f 1/1 1.0SÖ 1.236 3.098 1.352 1.41 / 3.17? 1. V- h c i v m p 1.45 1.225 0.775 3. 729 2.035 1 .& ’? 1.J2.1 1.672 1.241 3.133 1.922 1.529 .3.159 A1 •33 u:r.(/;i3 4.. 128 1 . J49 0.931 2.079 1 .36£j 1.058 1.04 • 1.263 3.006 1.743 1.507 3.197 1. •09 r;i •07/; 14 ).423 1.065 . .342 1.234 1.781 1.C49 2 .97s 1.727 1. 604 2.660 1.OS 0( u.ihl. 1 n.SGO 1.463 0.093 1.402 1.903 1.412 3.495 2.083 1.81? 3.507 1.59 l ,i Mnm. T U l).o93 1.654 0.951 1. ''24 2.255 1.849 4.149 2.300 2 .376 4.205 1.83 Diets OCR 1-7 arc Starter diets in which cassava replaced guinea-corn GCCO—14 are finishcr diets in which cnssava replaced guinoa-corn j..;. a i5t5/ 'Cent of computorisod diets weve increased by 22.48,» • - to makc up 'er overhead chargee 169 - TABLE 6.14: Rcvonuc ovar fced costs at 6, 3, 10 and 12 weeks of nga for diots in which cnsnnvr ucitimni/a fa/bird ) ....... fc • ■ • • • ■ 6 Wcokr B V/oeks 12 Woaks Diets 2 a5 b5 o Value Value Revenue Val uc Value Value Revenue of of nver of of of ovar feed bird fuedcosts fno>.! fced bird Fced costs MC 1 & 3 0.425 1.344 0.919 Q. 733 1.324 3.21P 1.095 1.477 3.302 1.63 MC 2 & 9 0.451 1.470 1. 125 0.751 1.35 2V9 1.930 1.591 3.421 1.63 MC 3 & 10 0.432 1.412 0.93 0.72C 1.137 J. 275 2.133 1.548 3.36? 1.82 MC 4 & 11 0.441 1.439 0.998 C. 751 1.152 3.22 2.J56 1.50-* 3.401 1.5? . MC 5 & 12 0.429 1.253 0.334 0.726 1.152 3.255 2.103 1.526 3. Xi? 1.00 MC 6 & 13 9.464 0.902 0.433 0.753 1.093 3. 0.12 1. :-/J9 1.437 2.997 1.56 MC 7 & 14 0.43? 1.157 0.725 0.7G1 1.155 3.00 1.045 1.304 2.440 1.14 Cnmm. I 0.500 1.463 0. 995 0.944 2 1.412 3.495 2.C33 1.017 3.50? 1.69 Corm. IIx 0.693 1.654 0.961 1.135 1.349 4.149 2.309 7.376 4.206 1.63 -i-, , , , , tv.Diets MO i y 1 arc Starter diets in which ctrssavo repl ;cod maize MC 0 •• i4 are finisher dirts in which cassavo re ilanad mnize Coat of m mputerisnd dints viere increased by 22.40,-. -f ln mako up for ovorhead charges hc oc“ aiercial dicts included. - 170 - wcight added is greatur than crr equal to the vclue of feed taken. The fermer also has ths opportun!ty of having more batchos in a yeor. In thc Nigcrian Situation however. domand is . or heavier/older birds. 6.4.2: Camjporison of Net Revenue frevn diets The next Seriös of nnalyse^-* drei with thn conipnrison3_c^nct revenuo over the nntire rearing poriod f'nr all the experimonts. (i) Ĉ ofnputr.rised̂ di nt3 plth varying fibr • and proteiji levc ls In thn first sec of diets snown in tahle 6. -:5. nix different diets emerge frorn the varioul combine tions of t!u startcr s! er diets of Experiment II which was di so nnst 3v r Ir ».'vipinr four. Thore are four diets in which oanh s fort er i-‘ .. ..„le.i < ach of thc finisher. The romaining two are nommercid diets. T'ible o .15 sunrr.arises the net n-vonue over feed cusen for the four differc. b^pairs of Startern and f i n i a n d thc two commercial ctic ts undorTNnsideratisn, Using thc nriterion, of net revenuo over fu.;j oosts, l? tan be seen fron: the fntla below t̂ iat comöinations of ither Starter 7 vvith finisher 9 or Sc are tr.s Lest, follownd by c.o.*>binatiorn of sturtcr 2 with either finisher 9 or 10. Comnercials T and II fol.lov: third and fourth respectiveXy. In general, thc -omputorised diets gave higher net revenue over feed costs. It is p e r t i n e n t t o tote höre t h a t l a c k of S t a t i s t i c a l s i g n i f i - c a n c e in t e r . h n i c n l o f f i c i o : iuy p a r a m e t o r r . cf two diets does n o t r.eces- s a r i l y m e a n that both diets arc aqua 1 l y oconomir.elly goorl (ns reported 171 TABLE 6.15: Summary of Net Revenue over feed costs per bird for diets in cxperiment II for th tho. ranring-period — ........ *r - — -- ! 6-1 D I E T S Feed cost: Feed consumed j Value of IVeight gained Value of Net Rever.ae for 12 weeks Ifend eonsumed J diets Weight gained over feed costs N/kg. kg. i^for 12 weeks kg. 2 Starter 2 6 1 .6Ü 1.29 '.60 9 Finisher . < n 3.39 2.64 2 Starter . 2C\ 1.63 1.29 2. SO 10 Finisher 1.23 3.69 2.57 7 Startoi ■i. 2C 1.36 1.69 1.Si 9 Fin-i sher 0.25 3.69 0.92 2.64 7 Starter 1 . 66 0.45 0.77 1.69 2.96 w Finisher ’J. 23- 3.69 0.35 1.17 2.57 CUMMERCIAL I Starter 0.33 T.dl o.so 0.64 1.11 2.27 Finisher 0.30 A 3.93 1 .1 0 1.20 2.64 Starter 0.34 2 .2 0 0.76 n.65 1.43 C0MMERCIAL III . 2.01F.i. ni. sh,er D.3Ö 3.30 1 .0 1 1.37 2. 'S • - • t. • 4 • #.l. • % • • •i « . .••••• • • * • A. 4. • -:-Prices of compjtcrised diets worc increased by 22. 4CT/J to make up for the overhead chargss which the commernial dints i icluded. The figure was the average of overhend costs of four modcl manufactviring plant? given in Ogunfownr et ol. 56/ •H-Prien -if 2 .20/k A u Lghl vas usod. It is the rries used an the Univorsity Teaching ar.d Research I nrni. 1 7 2 - in feeding Bxperimunts in chaptcr V), nrithi.r does it hold that a diet statistically signii'icantly Superior to enother diet in an Efficiency yardstick is necessarily bottcr economically. For cxample, diets 7 and 9 had better fl .d Efficiency indices than diets 2 and 10 yet, their combinations ore th;: •»••v.o in tcros of ntl rovenue ovor fned costs. Tbat is, cambining c’.ict 7 with icts and 1ü gdvc thr same profit as when diet 2 is combincd with diets 9 and 10. However',,.4iets 9 and 10 yijld the same revenue sven though their pretein cont&iqts vary. Diet 9 had 22 pereant protein wherens diet 10 hnri 2!. prrcont protein whilst their fibre levels were the same. Diet 7 performed better than diet 2 oven though its protein contnni of 24 peroent is lower than thet of diet 2 which had 25 persent. It :an b» inferred then that a 24 percent protein Starter diet is better than ^ 2 t percent protein diet. (ii) Diets in which cassava replacod guinea-corn The summary of net revenues over fend costs ’ for Experiments of Starter and fini3her diets in which cassava replaced guinea-corr is presentod in Table 6. 16. öoth startor ar.d finisher diets in which guinca-corn replaced cassava the various levels were corisidcred along with two r.ommercial diets. Diets iXC ') - GOC 1? had cassava I ca/c I s of 0, 5, 10, 15, 20, 26, 30, 35 and 40 percent rcspoctively. Finisher dints GCC 15 and GCC 1G ars sxcludod from the anelysis si ~c thoro aro ru startrr diets to match. It was observed in chapher five that signifiennt differences occurrer! in the startor diots ond the finisher diets. .. . - 173 - TA8LE 6.1G: Summary of Net Rovenuo over fced costs per bird far seven diets in which ci guinee-corn in the diets and two r.ammnroial dints 3V 6 D I E T 3 Fecd Valuc wsight Volue of evenue Fced Cnssava ÄT:'- costs ;onsumed of fced gained on diets weight foed cost; level nonsumod gained."/kg. kg. kg. G0C 1 0 . Starter 0.235 1.50 0.3' 1 0.674 1.403 1 . 1 1 2 0 .2 1 4.540 G. 954 2 .«aBw 1.334 2.45 GCC Ü Finisher 1.04 GCC 2 Starter 0.244 1.49 0.364 • i. 552 1.236 3.372 o 2.24 GCC Finisher 3.22 4.555 ! . 0"? 0.305 1.771 2.769 GGC 3 13,' Starter 0.253 1.60 0.4O5 1.455 1.0545GC 13 Finisher 9.224 4.493 1.005 ^^ 0V E 6^ . 125 2.17 1.119 GGC 4 15,, Starter 0.25? ]. 50 53? 0.351occ 11 Finisher 93 <".237 857 1. S 0.912 1.76 O X 5 20"'. Starter 1.225 0.775CS" 1? F inishor 1.934 0.056 1.63 Ü X 5 Starter n.2B 0.G13 1.349 0.901 GGC 13 25:'. Finisher 0.24 5.040 1.848 n . 790 1.59 GCD 7 ■' -f ' Stertor 0.29 ♦ J# ‘ 0.434 C.642Ren 14 Finisher 3.24e> 0.727 1.599 0.421 Ca.amcrr.ial 1 Starter 0.55F. C.5G5 1.463 0.095 Finishor 1.254 0. 929 2.044 (J. 790 1.69 Commcv'.ial III Starter 3.393 0.593 Ü.752 1 .664 G.9G1Finishor Ü.39C 1.635 1. 1G 2.552 0.356 1.03 -i- Prices of romputerised diets .vere increascd by 22.4£f,. to make up for the nverhead charges which thr eommaptiai diets included. Tho figure". wes the average of averhead costs of four rnodnl rnnnufucturing plants givrn in Ügunfowora, et al.,56/ Price of kg. livaw ight was ur>ed. It is the pricc usod on the University Teaching and arm luadnn - 174 - Tn this economic anciysis, it ran bn secn that tho net profit varies as tho c&ssava cont ■'•' V ' fche <1 i.r I veries. Thero is evidence thni revenue ctecrqoscs : aurr.entag3 cassava content of tho dict increases. Diots CCC 2 and 9 (S*/- caasnva) yi^ld tho high»st revsnuo of *'2.24 per birri whon comparing dr* is with cassava. The revenuo decreased to {-1.06 per bird in diets GCC 7 and 14 which •* in 30 percent cassava. However, diets ^ontnining up to 1n | 5 "cassava had higher V or c ;ual rovonue with the confnojv-tinl diets. 1 . ib reason is that these commercial diefr arc s ict ..ci U?r thn.-y4jlw computorisad dicts and * their higher veioht gains ; "v.,i.iH t nffseth the costs. Diots with higher cassava levels are cosLiier bcoaese cassava is costlier than the grains. Pf-f it iev/els rrr thox'efprc obtaine:! using different noots of the diets as c&ssr.va prdcewriqs. Computations .ire shown in Tablc 6 7 (e) Revenue with yaryinn cassava pr\rco When cassava wn3 made to assvme the sann' prico with guinea—corn ( '220 por ton), all tho canputarised iliers hoc. higher not prafits than the commercial dict3 exeopt for diet-.-: OCR ° and 9 and GCC 7 and 14. Diets QCC2 & 9 iavo oqual net profit.) wit • .-aramcrf.ial I. At this cassava prico and lower prices of 17Ü and "17. pr- tan, diets GCC 7 & 14 gave tho lowest rovenur,. How-vor, revenuo 'btoinod v;hi-n lower cassava prices aro usocl aro alv/ays higher than Por higher prices of cassava. Por instanec, diets GCO 3 and 1 > gave tho highnst revenuo of 72.25 per bird - 175 - TA8LE 5.17: Summarv af Net Revenue over feed costs per bird at varying costs of feed (Varying costs üf cassava ) wheri cassava is replaced by guinea--com in the diets "3̂ ' . . . “ - •- “ “ ... 1* PC ost or »f" ♦f en re' idM ei «t« ♦* 4 Value... ...o...f feed cons...u.med ValuG cf Not Revenue aver Varying priccs of nsssava Feed at varying cassava priccs feed costs at varying D I E T S • Cufis'jmed Woight ^ " T gained cassava priccs (©--4 ) kns. Cassava .-*220 .'170 i2r. :22'.i •220 ;:i7 n ri2P levols GCG 1 Starter 0.235 0.235 LJ..- —D ro 2. 423 2 .4 2 3 2 .4 2 3G00 Finisher 0.215 0.215 7.215 4.543 GCC 2 Starter U« C.ÖO 0.235 oon r; 1.632 1.70 1.710Finisher 0.213 0 .2 10 0 .2 tV 4 .555 GCC LI Starter ü. 241 ,1.235 0 .2 2 9 1 .6 0ic.r 2.246 2.202 1 .3 1 9‘.3TJC 10 Fin:i. : r 0 .2 12 0.206 0.2DC 4.4 GCG Starte: 0.244 P. 235 C.72C gcg 11 1G,’ Finis1 !■>• 0 .2 1 1 .201 1 .0 7 2 1.930 1.904 G. 192 GGG S Starter 0.247 J.23520.' O.PTJ 0.190 1 .7 3 3 1.3E0 1.939GCC 12 Finisher GCC <3 Starter 0.250 0.235 occ 25,' 1. R7C 1.955 2 .0 5 013 Finisher 0.2U9 0. 194 GCG 7 Starter 0.253 0.235 0 .2 1 7 1 .4 5 1,294 1.40-5 GCC 14 3Gf;. Finisher ■ ‘J.2D9 0. 191 IQ . 172 4 .7 0 0 1.524 1 .7 2 Corrmiorcial StarterFinisher 4 . 179 1.69 l mmcrcial III StarterFinisher 1.03 Prlci * of computariead dl re inercased by 22.40/j tu mnke up far th E overhead charges which the commercial diets innluded. TI in fiijum wns Uin avoru orhund nsts af four model manufacturing plants given in Ogunfownra, et al..®®' H Prlco uf n/k«. 1 t wna usud. It. is tho pricu used nn tha üni^orsity Teaching and Research Farm, Ibadan - 176 - at csssava price of '220 par ton. F v diets GCC 7 and 14 which gave the laast net rovohue, it gavc 1.62 per birc! when casscva costs >.‘120 per ton and >'1.29 per bird when cessavr enats "22rJ per ton. (iii) Diets in which cassevr rrplaced - vir.. This onalysis is with diets in which cassave s;.. atacl Tor mui "" rether than guinea-corn. Tlu dicts wcr>' those in ./hieb nassave replacoci maize in the Starters and finishers describef oarlicr in chapter five. The summary of net revenue ovor foed cos*.? — for expnriments four and six is presented in Tablc 6.1 Diets NC Q -• N7J 15 find cassavs lcvels of 0j 5, 10, 15, 20, 25, ÖJ, 35 and 40 pL-rocnt respectivcly. Finisher diets PTC 15 and ,'C 16 are exclurJed^fro” the nnalysis bocause there arc no Starter diets of equal caasay^i nentent to ivrnir.h. The results höre are similar to those obtained for gui oa-corn in Tablc 6.1£ except for the fact thnt tho net nevenue in tnese diets nre lower than those in which cassevc rcp?.aced guir» a r.-npn rsther than mcize. This could be due to th<_ faot' Wirt guinea--' orc basec! diets arc choaper than the maize basod diSet>s beceunc g.in i. is chrapcr than maize. It could also be the fact ui'Ot per-fonnanoe is butter with guinea-corn based diets than the maize o- : diets. For tho maize basod diets, diets f.C 5 and 12 gavr* tt • highe.*;4, revenun of '1.84 per bird whilst diets MC 7 and 14 gav the Inst revenue of 1.14 per bird. Here, diets -,vith cassava lev/els of vp to SC: pereant had highor or equel revenue with the commorcial diets. - 177 - It can be sa8n also tbat it is the higher nost cif the commercial diets relative tc the rumputcrised dicts that reduced tho revenue accroing to the farmer from these o.jmmercial diets. This Shows that an attempt at a critical assessmmt of feed pruduction costs may reduce feed prices and thus increasc the profit manjins nf poultry producers, and encooragc the Expansion of the industry. 0f purtic-ular importanco ’ s the indi- cation that cassava can be osed up to a fairly high 1 without ■ adversely affnrting the perf or.n- .nc. - cf the birds. (e.) Rj^enue with varying cassav. oriccs Since diets canteinlnc hinher ca levels are costlior hecausr p cassava is costlior than ttu grains, it bsscro.rormes necessary to view whet hoppens to net rhvanuc if cassava pri£ces ore lowored. Profit levels were thereforc obtained using different costs of the; diets as cassava pricc varies. Compotations are shewn in Table 6.19. When cassava assumes lower prices, the costs of the read arc cenoiderably reduced thus raising tho net rovonuo as showr. in Table 6.19. For diets MC 2 and 9 for instance, revenue rises from 01.73 per bxrrl when cassava costs 032Q/ton (Table 6.13) to 01-81 per bird .vhen cassava costs !‘120 per ton (Table 6.19). Howevcr, ̂ fie rovenuc when mal e is u3nd to replace casseva are much lower than when guinea-corr, is .. ;od (compare Tables 6.13 and 6.19). The revonuo for the Computer.1 -cd r’ie'jy lower prices of cassava are higher than the revenues givon by the c'manercial diets axcept for diets 2 & 9 and MC 7 and 14. MG 2 anv! 9 ntwever givß a higher net revonue tbgn 178 - TABLE 6.18_: Summciry of Net Revenue over feed costs per bird fron seven diets in which cussava rcplaced maise in the. diets and two commercial diets 5 D I E T S (7-5) Caslava Valuc of Feecf" Feod foed ’.Veight Valuc of Net Revenue Levels costs consumcd consumed gnined weight over fbGd costson diets gained :'/kg. kg. MT 1 0> Starter I. 25-: 1 .66 0. 425 9.511 LI. 919 MC g Finisher 0.23? 4 . 33 1 , nc~

Xv** 0.275 1.56 V. 574 1.263 0 .834 MC 12 i-inishür J. 247 4.535 C. 974 2 .143 1.01 1.04 MC 5 25, S ’:artor 3.233 0 .410 0 .902 0 .430MC 13 Finisher 0.253 0 .953 2.097 1. 123 1.56 VC 7 Starter 0.292 0 .526 1.157 9 .725 MC 14 c ü; Finisher 0.253 0 .583 1.233 0.411 1.14 Commercial Starter 0.33 1.558 0 .6 5 5 1.463 0 .3 9 5Finisher 0.30 1.254 C.929 2 .0 4 4 0 .7 9 1.59 Commercial III Starter 0.390 1.74 0.693 0 .752 1.S54 0.951Finisher 0,398 4.235 1.585 1 .15 2 .552 0 .0 5 6 1.83 ' Pricos of c. sed diets wore increased by 22.4f».'a to pake up for the overhead chargns which the c ’ diets included. The figurc was thd.avcroge of overhead costs of four model rnanufn plants given in Ogunfowora, et al. P@f -J- Pric.e of livoweight weu used. It is the pricc used Dn the University Teaching and Rose. Ibadan TABLE 6.19: Sunmary of Net Revenue over fesd costs per bird at varying casts of feed (varying costs of cas_ cassava is rcplaced by maizo in the diets 3 i s D I E T S Dosts of feed et FeedConsumed Value of feed consumed Value of Net Revenue over feed costs varying pricos of cassava 0 at varying cassava prices Weight at varying cassava prices (ß-4 ) kgs. *— *-*■ sained Cassavn levels ..'220 i'170 2120 0220 C12C i 220 0170 ri20 Cf' KC 1 Starter 0.256 0.25S 1.477 3.302 1.025 1.025 1.825 MC 0 Finisher 0.2.37 J.237 ' - .CT* MC 2 Sfaarti ;r C.±33 n.esr J.2' ’ 3.317 1 .766 1.785 1.805 MC 9 Finishrr •..230 ü.rr:7 1 rr/ 3.295 1. > 24 1.8Gd 1.902 iS7- 3.401 1 .939 1.99G 2.055 2ÜT, 3.406 1.995 2.070 2.145 2SF/o 2.999 1.732 1.017 1.900 30/ 3.945 2 .220 1. '096 1.187 1.275 ciimmer- CIAL I 3.507 1.69 OOMvUlJ.- C I Al III 4.205 1.83 Pr lau i uf c/npiilcriouc \ r nercus’.J by 22.‘■.‘ff,* to mcke np for tho overhnd Chargen Prien of 22 "'0/k~ li 1 hi willc11 t:l • i i : pi > ability of substituting cassava flour f'or thn tr ■’.ditian^i irriis used in broj.lor fcods, and the nr.onumics j" the resulrlnq diots. Chapter I oefinrx? tnt probla,i aroa as rogards the deficienc-y of protein i; hb- ava/Spe Hignrian diot and the arens to bc examined concoming liveScdc!? feeds and quality. The broiler industry was soloctt.ad bgjj^so of its st-hort economic, cyclo whinh cillowcd sevoral runs to be oarriet out to tost various food mixes. The ob.jactives nf tho study ao .vsIj ch the plcn of tho W.csis wer'.' statod also. Chnpcni II gc vo a reviev: nf previous works in computerisod poultry diuts and ra briof theorv of tue Linear Programming Technique. The model für least-coöt ration foruulcitian was oxpressed highlighting tt oroblemo of diet farmulacinn. - 102 Chapter H I discussnri tha basic metrix in tho linear programming model, thc restr "vions in thc model and the Solutions obtained by changinn the restrictions in the matrix as desired. In chapter IV, the experimental diets were discussed as well as the feeding trials. Certain Statistical Problems associated with thc estimation of prnduction functions from nutrition experiments were dealt with. This reengnizns thc prosance of corrclated errors arising from repeated obser nfinns on the same experimental unit and the random nature of feed intake resulting from ad libitum feeding. The data werr therefore transformed appropfrLately fo correct for such Problems. In chapter V, the compnrisan between cascava and thc grains was made with respceict fco tiieir uharactoriststiics and prir.es. Previous works on thc use Df nassiv/a bgslici diets for poultry were also revieweci. Considcring tha Foc t that lenst-cost diets arc not nc?cessarily the most officiowft terms of animal performance, foeding trials were corrled out with the aim Df comparing their effects with those of the commercial n^ets an animal porformanco. Six feeding trials were carriod out to camparu the six different categories of camputerised diets with some of the existing commGir.iel diets. Th^ first experiment tested eight Starter diets with various combinations of two protein levels, two fibre levels and the inclusion of cassava at the five porcent level. Comparison of these diets was based only on feed conversion efficiency and the diets with lower (2^ ) protein but high (S;4 ) fibre levels provod better. - 1G3 - Tho second experiment was bascd on the first ons in that two of the startcr diuts were scloctcd anc' fed to tho hirds beforc tssting two other computeriscd finishcr diots nn thsm. Therc was no significant differences (P ^ ü.Q5) betwesn tha finishnr diets and thair combination with diet 7 (24'̂ protein and 5, fibrn) Starter proved bottcr than when combined ».ith diet 2 startor (26>; prrteir, snd 'S'j fibre). experiments three to six, fnur groups of diets were formuiated. The first group had startsr diots in which cassava ievels of G, 15, 20, 25, and 30 percent were forced to rsplace guinaa-corn (GCOI-GGC?). The socond group also startar diots in which the samo lovols of cassava wäre forced tu rcplace rreize (f.r.1-MC7). The third and fourth groups had finisher diets in which 0, 5, 10, 15, 20, 25, 30, 35 and 40 percent leveis of cassav/a were farcod ^Vreplace guinea-com (GG08-GGC15)and Malze (MC3-MG16) rnspecti^elV*,^ Conparisons of these diets were based only on yrowth perfortnaince i nrdices such as average total weight gain (ATG), avr.’age total fend intakc (ATFJ, and feed conversion efficiency (FGE). No '.nclŝt of carcaas quality was rietermined bccauso it was not considored an impertant factor iri our cconomy. Tn rimonts three and four in which cassava replaccd guinca- corn and maiz. .> respectively in tho diots et six different levels, analysis of vrrianoe tosLs showed that there wore significant differences (P 'C. 0.05) in the diets. The least significant differenco (lsd) Statistical tert prnvrv; nnmmornial diet III to be the best wherc-as diet K ’C 7 wes ocor- t Ponv porformance of diots GOC 6 G MGS and GCG7 G !,o7 'vos nttributüd i:o tl high content of cassava in the diets which gave thcm Q powdery nsturc and inhibitod thcir censumption, Experimente five and six tested finisher diets in which cassava replaced guinsa- corn and malze in the diots respectively. Analysis of variance tests showed that thore wert? signifikant (P 0.05) differenecs in the diets. 7.1.1: Regression Analysis In -haptnr six of the study, the diets werc fi'urther mvestignted to reveal hnw the: prinr.ipal nutrient companento SAenced animal performance. Regression analysis was thoref?oürtu> cesarried out on the experimental rasul te nbtained from the fundier trinls far both Starter and finisher computerised diets as well as thr sommernial diets. Analyses wore on feed intake. single? nutrient and nutrient combihation effects on liveweight gain as well -SyiF'E effects of the cnergy-bosod ingredients on liveweight gain. Liveawfetiight gair was regressed on protein intake, onergy intake, lysine intake and methionine plus cystinc intake as single nutrinnts: then on protein and energy intakt. lysine and methio- nine plus cystino, intake as combined nutrients. For the energy-based ingrndients liveweight gain was regressed on maize* intake and cassave intake, tfr n guinea-corn intake and cassava intake. Both the quadrotic and square roct equations were Fitted but the quadrßtic ^orms which p.rformod botter werc selected as the lead equations in that thay 3atisfied the laicl down oriteria. Examination of the quarirctic functiona.1 equations showed that feed intake is an important cxplnnatory variable as far as liveweight gain is concemed and it exploir d murt than half of the total variations in - 165 - thc performenr.e nf thc birdo. For the single nutriont cffects, protein, Energy and the amino- acids proved to ba impurtant sxplanatory variables as far as liveweight gain is cnncemed. In tho finishars, protcin and nnergy explained lass than half of thc. total Variations whereas in the Starters, they explained morn thnn half nf thc total variations in the birds1 porformance. For thc nutriont combineition affacts, the functional equations showed thefc protein and energy and the amino-acide arc significant explanntcry variables in liveweight gairis. Protein and energy combined explained pernent in thn Starters and 31 pcrcent in the finishars whoreas the -amino-neide altog-.b .er explained 56 percent in the Starters and 33 pereant in tho finishars. vovt perccntages explained by thc nutrients was attributed to the fast that not all the nutrients were mada available for the birds,- ’ motabolism. uterginal analysie was aarriad ent on somo seleoted functions. Tho elnsticity of productigh Tot- enenjy and protein showod increasing rrtur.ns to scalo in thc sknrtsr and fininher r’i- fs. Tho Optimum broiler woight . s < gain was detprminad as wnl_ thc quantities of malze, guinea-corn and oassava^^^ütain il*. Rx’uduction surfaces, isaquants and isoclines were calci; fsted for sulectcd functions. 7.1.2: Economic snaxy.sis qf„tne_jliets Using the results of the feeding trials, ostimates of net revenue aver feed cast for the different diots wert computcrf. It was discovered that non-signifiennt differencos between dists was not synenymous with - 186 - cqual rcvenue yielding diets. Tor instanco, diet 7 had bettor feed officirancy than diot 2 and di i ; 9 wag bettor than diet 10 yet, combining Starter diot 7 with finishor diets 9 and 10 gavc the same rcvenue as combining Starter diet 2. v/ith finishcr diets 9 and 10. In general, the cnmputerised diets gave higher revenuc the.n the cammercial diets. For the diets in which cassava replaced quinoa-com and maize, tho rcvenue nccruing to the farmer dccreesed ns tho percentage cassava content of the diet incroased. The rcvenue nncruing fram the diets in which cas­ sava replaced guinaa -cern are however higher than the revenuos from diets in which cassava replcced malzn. Diets containirig up to 10 pcrccnt cassava had higher or squal reven :h tho commcrcial diets. Diets with higher cassava levels aro costlier bccause cassava is costlier than the grains. Profit levals were therevfere obtained using dif- feront costs of diets pries varios. V.'hen cassava was made to assume the samc pries with inoa-com and malze, all the computerisnd diets except. diets GCH 2 ii. S and CCC 7 and 14. MC. 7 !ahd 14 had higher net Profit with commercial I diet. Net profits increased as cassava prices warn redunod but diets GCG 7 arid 11 and MD 7 & 14 gave the lowest net Profit trv tim- . Optini ..x r* /ting aj. ietermined suggestea that broilors bc? sold at olevut. waer-s wnrr therc was maximum revenuo over feed costs for diets containing cassa*:. except diets GCCl and 8, GCC2 and 9 and. GCC 3 and 10. - 187 - 7.2 Conclusion s The importanec cf the rolo played by the animal scientist and the econonist is well illustrated by the biological and estimational prrOlems involved in this study. The results of thi? study rnveel that any efforts to improve the retums to poultry, fnrmors must be fncussed on the cost and quality ef foeds. Particular attention must be paid to the ingredients included in the feed. Chcap sources of protein and oils such as soyabean and palm komel meal have beer» usod at high levels in this study to reduce the cost of feods. There wcra diets in which groundnut cake was at very low levels (GCC 12~1C) or cvcn^Äfipletcly eliminated (GCC 8-11) and these dicts did not perform poorer than those containing high levels of groundnut cake. Fish meal was not included in any of the computerised diets. Fish meal, apart from buing an imported feed item, has experi- enced stesp pricc inerßeses and at certnin poriods had been unavailable. The study mphoais on thn importance of bolancing the diets nutritionallLyv ubslinqg the technique of Linear Programming and taking into considerasttiion the rnatrictions c.|- the nutrients for each dass of feed. Cassava flour has beert fciUf.d tr: partly Substitute for the grains without advorsely offentiny ti pcrformancs of the birds. The high cost of cassavo flour however reduces the net revenue to the farmer when it is used at high levels of 20 percent and above. If the cost of cassava flour can be roduccd the-n thn food costs ivould bo roducod and the returns of the farmer would increase. Cassava grows well and uses marginal soils. - 188 - Efforts arc being made to raiso thc yield further. Nov. that it has bocn provcd to be usable up to 20 pcrcent as energy source fesd mixcrs in Nigeria can be said to haue- a waiting alternative in case thc cheapor grein becomcs loss availabls. c Rcsults of tho rogrussion analyois showed that protein, unergy and somc essential atr.ino-acids such as lysine, methionino and cystine arc important m tho growth porformance of broilors but they explained ortly a little Proportion which raisud a questiun on their availability f• r the birds' motabolism. Marginal analysJ^shovvod the maximum contri- betion of nach nutrient bot the r.xtnnt.^S^'M ch ane cauld Substitute for tho other depends on tho rnsts tff/tho sources of these nutrients arid tho neods of tho birds. & ?. "! Roep. »nondati on s From the rosults1 uf uxpr.rii.r s-ito test ing varying pre i. in and fibre levcls, prar, tioal fcroilor produdars woulci bc advised to start birds with dicta\dl^high protein od fibre lcvels and finisn the oirds on rJiots with lewer protein percontagr anc! high fibre content. If tho grains aru to be 1 oplticed by cassove, tho levcl of inclusion of cassav/o shnulr not nxeced 2Q pornLiit. This is tho point at which re'/snoo, . r.: -r tc tho farmer would not bo adversely affected. Encouragement sh.jult ' given tu fsed production in pellet form if cassava is to bo i.ssd at higher lcvols. - 189 - It would be better if tho uss of guinna--corn can bc strcsscd murc. in poi itry diets than baforci. The «experimental results provee! that guinca-r.orn performs better than mnize when nombinod v/ith cassaua in poultry riiets. Government must now intensify resi.arch and production octivities in rospoct of both cassavn and guxnea-corn. Ine resutts suggested also that broiler producers shauld sc.ll off birds at eleven weeks nf agc if cassavn Substitutes Por the yroliinns. This i:s tho point at which maximum revenu- is obtained. Tho extuVnt. to which tho weight and the quality of tho poultrv meac satisl^^fchc requiromonts of consumors was huwever not taken intn uccount. Government should fund rcscarch into the dotaileri Chemical anclysis tif local feodstuff and tho nutrient availability of the birds’ matRbtilisrii, . V, A Suggjc s^jjns for further rosoarch Düring the r'<&!%■■ of this study somr. diots wore formulated in which soyabeari flin43" was introduocii ac rhu pi'otein source. The nxperi-- ment was howeyer not dooigned to moasura tho affocts of using oither soyaboan ms«L urily 03• groi '• nlnut cakc as pretoin sources in the feed. It would bc. nf int rust to know which of the two ingredients is better or if perhaps thu. euvlu be perfect Substitutes. Iri vic*w of the uaidous broiler breeds existing in the market now, it would bc worthwhili» to find aut the effects of the various computerised feods an the performance of each broiler breed inpluding our local stock. - 190 F I E F E R E N C E S 1. Adcgbola, A. A., Mcthioninc as an additive to cassava based diets. Proceedings of a workshop hcld ut the Univnrsity uf Guelph lOth - 20th April 197?. International Developm snt R osoarnh Centre. Cassava as animal fand ndited by Nestel B and • «.• • /• > .» — »- • *.« • k • • « . i■ • Graham, M. 2. Anwor, U. R. Linear Pro^iinminci Method for computinn least-cost ohick startur rotin; i aucor-d.ing t^ jfri ratio botweon gross protein quality valuc units (.GFWs ) and energy content. University of Alberta, 137G. World 's Poultry Science Jourpal, Vol. 2 3. Armas, A. B. and Chicno. C. {r. 1373. Evaluacion da la harina de yuca (Manihot oöakmtei) an racioncs para pollos de ungorde. A^ü^n. Trop. (Maraoa/J, 23(g ), 593-599. 4. Oabatunde, G. M.,Ör^ugaf B. L. and Qyonuge, V. A. (1971) The effects of s/avinij tho dictury caloric profcein ratio on the per- snc.u characteristics and cercass quality of growing pigs the- tropics. Animal Production 13: 695-702. 5. Babatdnde, G. M. nnd Fetuga, G. ... (1976). Determination of the minimum crude protein roquirements cf broiler Starters and finishers in the cropics. Nie/. J. Anim. Production 3: 126- - 191 Baker, F. Nasr, H., Morrir.e, F. and Bruce, J. (l95ü) Bacterial breokdown îf structurai siarchcs and starch producta in tho digestive tract af ruminant and non-ruminant mammals. J. Pothv>Brct. 62, pp. 517-63!8. Barrious, F. A. and Bressani, D. (l957) Composioion quimica de la raizy du la hojas de algunns variedades de yuca Manihot. Turrialba 17(3), J1A-320. B. Box. G. E. P. (195t). The exploratiun and exploitotion of response surfaces: ocinu gc-noral cons_i_ddefrcaitfiH tons. Biomctrics 10: 15-öG 9. Benny, J. M. (19(39) Mechanist of Tapioca Mnlaysian Crop Diversificetion 0'^^^ Rpn;i it Wo. 13. 10. Blagburni,, C. H. Farm rji urning am* l\tenapement London, Longmans Green & C^^19^5. \ N 11. Card, Leslie E., Aleisheim Maiden C. Poultry Production 3rd oditip'n^ i-.ic.Graw Hill Book Coy, Inc. 1971 12. Chou.h K. § , 5nd Müller, Z. (1972) Completc Substitution of maize >; tapioca in broiler rations. In Australian Poultr^ Science^ Convention^ _ Auckland, _1972: proceedings, New Zealand World Poultry Science Association, 149-1S0. 13. Church, D. C. , Brown, W. G., and Ralston, A. f. (1963) Evaluation of cattl;-. fnttening rations f ormulated with linear Programming tecHniqucs. J. Animal Scipneo 22i 090-903. - 192 - 14. Clayton, E. S. (1D6'7) Thu econand cs of the_ pajltry industry. Longmano, Green and Co. Ltd. Ist Edition. 15. üont, J. 0. (1972) Proptical fead fonnulation by linear Rroqrarnmina. Praceudinns, international summer school on C om p u ters and rosuarch in animal nutrition and veterinary medicine 151-163. 16. Dnut J. B. and ..nsey, H. (1967) Linear Prag^T^hing and animal njutrition. Crosby-Lockwocd, LondoNi/ 17. Enriquez, F. Q. anri Ross, E. (1967) The valuc of cassava root msal for chicks. Poult. Sei. 46, 622-626. 1fl. Food and Agriculture Organisation of the United Nations (ISS1)) Indicativc World Plan for Agric. Dovpi. Rome Italy. 19. F A 0, Agric. Devpl in Nigeria 1965-Qü (Rome 1956) / A . 20. Fetuga, 0. L., [19711 An assessment qf the protein quality of somq Niger i-a fqqds and foeriincj^stuffs^ _in_thc nutrition rr/rat- vi-j'-. Unpublishod Ph.D. Thesis, U.I. 21. Fetug^.'ö. L. , rV:-tundü, 9. r... a: .! Oyenuga, V. A. (1975) Protei lävf-lr Li ciots foi* r.cropean pigs in the tropics. I. 7 . -f-;- r; c 0f •.u thteninti supplemuntatlon on the protein r: -c-, ivor., v bs .J-• .>igu. Anim. Prod. 20, 133-146 ci3:iur . 0. an» '7 . . m . W. (l‘J53) Linear Programming •ppj-ioc to i eed nixing under different price conaitions j. Farm Er.ons. MoX. 35 Nu. 4 - 193 - 23. Flinn, J. C., Qroartwitch, fl. A, Young, L. G., and Ashton, G. C. (lb72) Exploration of 1! «ast-cost diets for grovving hogs. University of Guelph workinc pnpc:; Mn AE/72/l5/l972 24. Gadclha, J. A., Campos, .J. and Moyrose, V. (19£9) Farelo de raspe do mandioca Experi.cutipp, 9(4), 111-132 25. Gilson, J. C., Yeh, M. H. and Hodgson, G. G. (1953) The use of Linear Programming t - ueterminc lcast-cost poi Itry rations Fac. of .Afjric. and Home Econs. The University of Manitoba Winnipeg. Technical SullotiiV'iMÜH 7. 2S. Guttridgc, D. G. A. an:! I.iorcjan, J. T. (1959) Mothionine supplc- nusntation of broiler uiLfWltASs £at var/inn ]ovel3. World1 s ------- Poultry S e i J. Goc-uoc. Vol. 15 Na. 4 . 27. Hcady, E. ü., Gellcun and l.'cAlcxander. Production functions for least-cost broiler retions. 5th edition. functions 1961 Iowa State University Press, Ames, Iowa. 23. Heady/E. Oillon, J. L. (l9ol) .^ricul_tural Production unetions I.v-.va utntr‘ University Press, Amos, Iowa, 29. Hu bi ng, 3, T. , Pnuah, 3. H. and Hcw, V. F. (1973J The utilizatiun of cassave- (tepieoa)_ in livestcck fcpdincj Papers presontod at tfic Third International -Symp. on Tropical Root Crops, Nigeria II7A, Ibadan 2-9 Dg c . 1973. 33. Hutten, R. F. and Allison, J. Fl. (1957) A linear pngramming model for development of feod fomulas under well-operating conditions J. Farm Ecens. dol. 39: 94—111. - 194 - 31. Hutton, R. F., and McAlaxandor, R. H. (1957) A 3implificd fecd mix modcl. J. Farm tr.ons. Vol. 39 No 3 F&rt II pp. 714-730 32. Ikpi, A.I.E. (1972) Ineorno Oppurtunitios in Broiler Proftuction: An experimental appreach. Unpublished M.Sc. Thesis. Ooptirtment of Agricultural Economics, 33. Jaap f'i. George (1973) Growth rote of broiler chickens. World*s Poultry Sei. Jourru-il. Ort.-Dec. 1970 Vol. 26 No. 4 pp 804 3* Johnson, R. M. anr! Raymond, V/. D. ( The Chemical composition of co.tc tropica 1 focri plants. IV Manioc Trap Sei. 7(3) pp 109-115 35. Kaizman, I. (1965) Sulving fecd Problems through linear program­ ming. J. Ferra Esons. 38(2): 420-429 30. Klein, IV. and Barlowan, von G. (}954) Tapioca Mehl in aufzuchfuttor Arch. Gofiuqchlkd 10, 415-523 37. Krxi* :;oyiano*B, A. (1973). Thcory of Ecunometrics The McMillan Ltd. London. 3G. Kupglu/i, 0. 0. (19/6) The ec':.lor.ücs of a private poultry industry Misccllanenua study. Dopt, of Agric. Economics, U.I. 39. Mconünzic, H. C. and Gcduoil, 1. E. 11956) Linuer Programming and the eoat of pig i h.tc'iiric* retinns. J.Fann Econs. Vol. 10:4 40. Mahcnrirsrrthan, T. (1971) Potential of tapioca (Manihot Utilissima puhl o:.. a livestock Feed - A review. Mal. Agric. J. . 4T.(l) G 426 pp. 77-39. 195 - 41. Makinde, M. A. (19?4) The structuro and economics of commercial ecjg production -_A ca sc study^ of Olaggun enterprises Mi sc.study Dopt, o'f Agric. Economics, U.I. 42. Matanmi, I. 0. (1978): A rejyiuw of work done an putrient requircment of poultry in Nigeria. World Poultry Science Association ( 1978) J,Nigerian Branch” pp 16-20 43. Mba, A. U., A. 0. Bollo, V. A. Oyonuga and F/J lS Olubajo (1974) Sr The effects of varying the levals of proteins and red nalm oil on the utilisation of poultry feeds by broiler chicks. Niß.. AgrtLc. J. 11: 1 44. McCiuskey, W. H. (19G2) Poultry ßroj^css in Nigeria Oregon State University. Corvallio, Oregon, U.S.A. 45. Maust, L. E. , Scott, M/’̂ y a n d Pond, W. G. (1972b) The metabolisable enorgy of iri.ee bran, cassava flour and bluckycye cowpeas for grotoing chickcns. Poult. S_ci. 51 pp 1397-1401. 46. McMillan, A, fl. and Dudloy, F. J. (’941) Pctato meal, tepioca ji^-ü arid town .VwStr in chickcn retiono. Harper Adams * -IJtilitx jfoult.̂ J. 26(9), 191-104. 47. J., Maner, J. H. Op a:: >. H., Pnrtcla, R. and Calls, J.T. (1970). Nutritive valuc nf different tropical sourccs of onergy. J. Anim. Sei. 31 (•;) pp. 200, abs. 206 43. Mentills, J. J. , Mur.doz, 0. fl. and Wiedenhcrfer, H. (i960) Utilizacion ■ le la harina dr tuborcuto du yuca (Manihot oxulanta), en rarionos inioiaderas pore pcllns dc> ungorde. Arph. Latinnom. Nutr. 19(4), 331-303 - 196 - 49. (1970) qp. eit. In comun. cient, XIV congr. Mundial Avicult, Madrid, Espana, 985 p. 50. ---- , Castillo, P. P., and ’.Viedenhofer, (1975) Efect: de la incorporacion de harina de yuce amarga an raciones pare pol los de angorde. Agrar.. Trqp. 25, 259-266. 51. fAjllor, Z., Chou, K. C. and Nah, K. C. (1974) Cassava as a total Substitute for ccrcals in livestock and poultry rotions. Proceodings of the 1974 tropical products institute Conference 1-5 April 85-95. 52. Nestel, B. (1973) Current utilisatior, and futurc potential for caesava in chronic cassavn toxicity: Proc. of an interdisci- plinary workshop, 29-X Jan. pp. 11.26. 53. Obiorha, 0. (19^5J The ruplaccment valuo of processed cas~'. in r finisher rations. Nig.^ J. Anim. Prori. Vnl. 2 PP* 54. Dlomu,i, J. Mw " i fei :v 1 in and onergy lcvsls for finisbing ’jt __in p tropical anvirunmrant. J. y'.nxra. Prug. 4(in pro3s) CK C l ^ X 0. •. . . .n.' Bin.!, H. R.( 1969a): Amino-acid supplninoi>Lai.ie'' nf n. riujoca meal in chick diets. Poult. Sei 4Ü, 1949 -9**33. 19(3?b. The metabolisable Indent nnd feeding vuiuo of mandioca meal in dict for chicks. Poult. üci. 43, 1445-1452, 1963. Metabolisable energy and feeding value of mandioca meal Poult ..Sei. 47, 1703 197 - 56. Ogunfowcra, et al. (1975) Live stock Fpod study_ rspart. Study toam commissioned by the Federal Livestack Department, Kaduna. 57. Ocunfeworü , C. and Fctuga, 8. L. (1975J Least-cost ratien formulation for pir production. Nig. J. of Anim. Prod. No. 2 Val. 2 Dec. 1975. 58. Olayido, S. 0. and Ogunf oworr-( 0. Optimum- timing in broilcr production Bulletin of Rural Economics and Scciology Vol 5 No. 2 59. Qlayidc, S. 0. ct al. A ̂ quantitative analysis of_ feedI ^quire^ents, supplios and demands in Nigeria, 1958-1905. Federal fviinistry of Agric. G Natural Resources, Lggos, Nig. 1971. 60. ülayi.-'e, S. G. (1972) Comm. broilor growing structurc of costs and returns. mcäl report AER No. 72.1, Department of Agricultural Economics & Extension, U.I. 61. Omolc, T. A. i ehr nutrition of layors. Cassava gs a n i m a l fcec r V c . t workshop helri et the University of Guolph 16-2l April. International Devpt. Research tro. University cf Guelph. 62. Uyafaga, V. /*. ( . ’• j toiguriai _ fguds and feedinrg stufFs. Ibadan s Jnivor&it> .-ress Ibadan, Nigeria. 53. ---------... ... (lr'v; j Ihr -,'impcsitior, and nutritive value of oertöin . <- ding oruffs i" Nigeria. I. Roots tubers and grecn leeyns. European J. exp. jlcjric. 23(9C’), 01-95 54- ------------- s (1351] Nutritive value of cereal and cassava diets for growing and fgttoning pigs in Nigeria. Brit. J. Nutr. 15, 327-330. - 198 - 65. Phuah, C. H. anc! Hutagnlung, R. I. (1974) Effect of levels of dietury protein and cassava on performance and body compc- sition of chickens. Mölay. Acjric^ Research 3,99-106. 66. Panda, B., Reddy. V. R. A review cf work done in India on nutrient requiroments of chickens. Pool Division, Indian Vet Res. Inst. Izatnagar U.P. Indio Pouit. Sei. J. 32, No. 4, NPv. 1976. 67. Rendon, M. Ounit r, H. and Marinz 0. (1969) Utilizacion de la yuca ( tanihot] oxulenta' nngorde de pollos asaderos. Hey. ISA, 4(3). 159-1 0 60. Scotti,. M. L., Poultry foê ,ü management in the tropics. Animal Production Editcd by Loosli, Enbatunde and Fctuga. Ibadan Univcrsity Press. 69. -, i'Jutricnt requiroments of chickens and around the worlrfT 70. Serres, H. and Tillon, J. P. (1973) Oassava for pig foeding. I. Possibilities and use limits. Rev.^E lec. Mod. Vet., Pays. Top. 26(2) pp. 225-223. 71. Squibb, R. L. and M. K. iVylri (1951) Effect cf yuca meal in baby chink rfitions. Turrialba |(s): 193-299 Cited in A.I.D. 1972. 72. Tabfiyoyong, T. T. (1935J The valuo of cassava. rofuse meal in the F r„ • • . ; : ratirma fer growing chicks. Phillip. Agric. 24, 509 - 199 - i. ’ 73. Voght, H. (l9Sß) The use of tapioca meal in poultry rations. Worlds Poultry Sei. J. 22 pp. 113-12b 74. and Pc mar, '.V. (1963) Der Einsatz von Thpioca-und Maniokamahl ir. geflugcl mastfuttcr Arch. Goflugclkd 27, 431. 75. find Stute, K. Prugung von Tapiocapellets im Geflugolmast-Allemifutte: Arch. Geflügelte'. 28, 342 76. V/augh, E. V. [1951} The minimom-co. dairy feed. J. Farm Econs. - 200 - APPJ5HDIX A Table 1: Records o.f average vevkly feed irtake and live- - 201 - APPENDIX A Continued. - 202 - APPEMDIX A Continuod* - 203 - APPENDIX E Energy lwel of fe«d = 2 ,8 6 1.4 KcaWKg, - 204 - APPENDIX £ Contd. Sners>^ levei of feed = 2,Qr?i .9 Kcais/Kg, - 205 - APP^.DJX £ (Coutd.) fcergy lfvcl of feed 2,tb:<.U - 206 - APPENDIX 3 (Contd.) Energy Lev^l o f the fe e d - 2 ,9 0 6 .2 K cals/K g , - 207 - APPENDIX 3 (Contd.) T 1 -Diet Av. Weekly live- Av. Weel-JLy weight gain (Kg) Fced intake (Kg) Comm.I 0.025 0-15 0.01+ 0 , 7 < ? - 0.08 0 . 1 7 & - ? 0.1T 0.17 0 ^ 7 0.20 ’ O.J+Ö 0.03? * P 0 , 7 • 0,00 ! o.it 0 . 2 y \ 0.35 0 .18 0.1+8 C.sO Comm. II 0.027 0.17 I r o.o67 0.21+ o . m C..3 2 0 .13 2 0.1+1+ 0.151» Ü.?0 0.167 0.1+9 0.026 0.17 0.06? 0.2? 0.11 c.1+5 i 0 .1 2 « c.?o 0.11+ 0.5 0 0 .16 j 0.?2 - 200 - APPENDIX B (Ccntd.) Diet Comn. I Av. W&ekly live- Av. Weekly F&ed Finisher weight gain (Kg) Intake (Kg) - 209 - APPJ5HDIX C Starters (3rd Ebcpt.) APPi-JilaX 0. Startors (3rd Ajcpt.) Contd. - 211 API'EI.TjIX C Cotd. Starters (l>th Expt.) - 212 - APPkuLIX C. Geiitd» Starter (l*th fixpt.) Weiĝ .t I'eed Protein Energy Lysine Methionine Gain Intake Intake Intake Intake + Cystine (gms.) (gms.) (gras.) (Cal^ries) (gms.) Intake (gms.) 10. 581.9 1339.9 321.6 3885.7 3 g - 101*5 1*58.1 1293 320.5 371*9.7 1008..-. 11. 61*9.9 4335.ft 320.5 3872.7 ö J.67 141*1.- ’ 511*. 7 11*88.7 367.3 1*317:V 1.86 1.163 12. 551.7 11*36.3 3U*. 8 ^ 6 . 7 1.76 1.120. 1*16.7 1275.8 306.2 3699.3 1.59 1095 13. 333.2 1281 ° ’ 3 V 3797.8 1.61* 1£>21 11*. 35U.6 1111*. 7 3232.6 1.1*0 0-81*7 31*5.9 1216.7 292.0 3528.1- 1.52 fl-901*. Coram.I 587.1 131U.7 339.1 4 561.3 1-377.1* 330.6 C. II 61*8.3 1i*51 • 6 632.2 1518.S — - 2 1 3 - APFEKIUX C. Cor.td. Finishers (Uth Expt.) S - 2 U - APPEi.DlX C. Contd. Finishers (i+th Expt.) Wt. Feed Protein Energy Gain Lysine Cystine + Intake Intake Intake Intake Kethionine (gms.) (gms.) (gms.) (C alories) (gms.) (gms.) 11. 735 1+095 8fci+ 111+66 5-3» - " 3.235 650 31+92.1 06.8 9777.9 3 % 2.759 12. 862.ii 1+181.6 88.;. 3 117C8.5 S - U + 3.303 67U-7 31+15.3 720.6 9 5 6 2 . ^ O . a 2.698 13. 910 3830 8Ji1.1 1+.98 3.026 hSh.2 315'+ 691.7 :0 2.1+89 1U. 931.6 3666.6 ^ 10272.1 2.896 « » • l o S 1+.77 C ovjt,.'. 793.3 3609.9 e i i X 1C331.7 I+.80> > 2.915 1089 3930 C .III 685.3 3786.8 V i 1020 3890 750 W . 3 • 15. 1020 801.3 11216.8 5.21 3.165 886.8C* ^ 3773.8 8,30.2 10566.6 j s S 1+.91 2.981 16. U 5 W C 3776.9 830.3 10575.3 1+.91 2.981+ 512.3 3118.5 666.1 8731.8 i+.05 2.961+ 17- 8i|0 3l4+o 758.2 9632 1+.1+7 2.683 760 3200 701' 8960 1+.16 2.1+96 13. 990 ;-5‘-5 78JU. 3 9962 1+.63 2.781 1610 3230 7*10« 6 901+1. i+.20 2.519 I - 215 - APPENDIX C. Contd. Finishers (3rd Expt). Diets Weight Feed Protein Energy Lysine XeXhicrine Gain Intake Intake Intake Intake + Oysiine (ffas.) (gm£.) (gms.) (Calories) (gms.) In pake 1. 1173.8 51+52.1* 111*6.1 15266.7 7.09 1i3C7 1318.5 5296.8 1113.1; 11+831 6.89 Q * U81+ 2. 787-7 5621*. 2 1169.6 157U7.8 7.31 ^ UU+3 1021 5320.6 1106.7 11+893.2 1*203 w 3. 1151.1 5373.7 1098.9 1501+6.1* 7.rc 1+200. 1130.9 5373-7 10969.3 1 SOl+öSjr* 6.99 M91 b. 1000 51Ö0.U 101+1.3 6.68 3988 700 U531+.3 92j.5 ^^^j61+ 5.93 353 J i 5. 1000 51+36.7 H O S 0 X 152*2.0 7.07 L186 967.2 5289.6 11+810.9 6.86 W 3 6. 91+7.1* 5300.u IO?. .8 11*81*1.1 6.89 1+081.+ 101+1+.8 1059.9 11+653.5 6.80 1+P29- 7. JUl-i.Lt 5517.7 1112.6 15U1+9.6 7.17 1+-21+6 890 4 5516.1 1112.1 1510+5.1 7:17 i+2tf7 8. 979.9 ► 5671;. 3 11t0.5 15885 7.36 1+312. 11Q5.3 5231 105U4 11*61*6.8 6.80 J575 9- 837.U 5301.7 1086.1 II48I4L.8 6.89 8p 29 9U5 5630.7 115^.3 1.5766 7.32 1*079 ' 10. 89U.8 5222.14 1110.0 1a622.7 6.79 ^ 2 5 1171-U 5192.8 1101+.5 1L;>;».<3 6.75 1*402 ■ 11. 91+7.JU 5833.7 i • • - ' • • r.56 j U608 1021.3 5652.8 1102.7 15627 7-35 111+65 - 216 - AP?ü.3)IX C . Gor,«.d . Finishers (3rd. Expt.) - 217 - APPKTODIX D Starters (3rd Üxpt.) Weight Feed. G. G o m Maize Cassava Cassava Gain Intake Level Intake («cs.) (gms.) (gms.) (gms.) % (gms.) 1. 828.8 1837.3 1017.1 0 0 661.5 1903.2 1053.6 0 0 2. 667.7 1700.6 836.0 85.0 671.6 16 6 1.2 8 16.7 -r 83 n - j 0 ? 3- 760.6 1812.9 763.1 10 181.3 813.5 1081*. 2 793.1 10 188. 1* k. 760.6 1878.9 658.2 281.8 756.5 1853.2 65C 15 278.3 5. 1005 2053.2 5 7 ^ C 20 1*10.6 521.3 1903.3 5T3 : 20 380.7 6. 755.7 "879.5 .393.0 25 U69.9 779.5 2000 1*16.2 25 500.0 7. 803 295.4 30 605.1* 710.9 < ^ 7 1 .2 273.9 30 561.1* 6. 6 1 6 . ^ 1871 101*3. 0 61*7^ 172l*.l* 5 0 1. 0 9. 852.5 181*1.5 9 u , . .- 9 2.1 75U.9 2069.6 J0 16.1* 5 10 3 .5 10. 803.5 181*6.6 773.0 10 181*.7 - 218 - APPENDIX D Cpntd. Starters (3rd Expt.) - 219 - APPENDIX D Contd. Finishers Uth Expt. Weight Feed G. C o m Maize Cassava Cassava Gain Intake Level Intake I♦ (gms.) (gms.) (gms.) % (gms.) ■ 1 1. U169-5 23U3.7 ^ 0 ° < ? 325 U. h 1629.3 0 < j r 2. 3916.9 1979.U 196.0 3397.1 1695.8 167.9 > 5 3. 3929.6 1762.8 10 393.0 3283.5 1U73.0 10 328.U !+• 3711.6 15 556.7 3h61.6 13^. 7 V 15 519.2 5. U168.5 »353.5 10 833.7 3506.1 - k 1138.1+ 20 7 0 1.2 / C 6. 10UO.7 25 988.1 395r 827.3 25 785.5 7. i+5 2 8.3 91U.3 30 1 358.5 3587.9 / 2h» h 30 1076.u 8. U370 61 i+.c 35 1529.5 31+87.7 U90.0 35 1220.7 9. U212.Ö 276.8 Uo 1685.1 3390.1 222.7 Uo 1356.0 03 -J O N - 0 - 0 CD V / . __________________________________ - 220 - Ä PPE lijy jX D i'ontd. Finishers l*th Expt. Weight Feed G. C o m Maize Cassava Cassava Gain IntcJce Level Intake Cans.V (gms.) (gms.) (gms.) % (gms.) 10. 850 U195 _ 2257.3 0 y- L» ■ 6Uk 31U3.7 - 1691.6 0 0 ^ 11. 862.U i-l 181.6 - 17U9.6 20i+.8 650.8 3U92.1 - 1676.9 17U.6 12. 862. U U181.6 - 1 7 1 + 9 .^ k 10 Ul 8.2 67U.7 3M5.3 - i u a ^ 10 3U1.5 13. 910 3830 - ^ 0 . 6 15 57U-5 USU.2 3150 1053.U 15 U72.5 1lu 931.6 3668.6 99v. 8 20 733.7 793.8 3689.9 \OC2.6 20 738.0 15. 1020 U006 Ä - 8L1.3 25 • 1001.5 886.6 - 792.5 25 9U3.5 16. h5h.2 ^ p . 9 - 560.0 30 133.1 ^3118.5 - j/j 2.8 30 935.6 17. Qko 3Wo - 298.6 35 120^.0 760 3200 - -77.8 35 1120.0 18. 990 3565 - 89.0 ho 11*26.0 810 3230 - 81. ij ho 1292.0 — - 221 Ai/PiINDIX I) Contd. Starters 1+th übcpt. Weight Feed. • 0. Com Malze Cassava Gain Intake Level Cassava (gms.) (gms.) (gms.) (gms.) % Intake(gms. 1. 557.14 1215.3 672.8 0 0 627.6 1377.6 762.6 - 0 A 0 2. 370 1136.7 558.8 - % ^ 56.8 536.7 1395 685.8 129.6 3. 536.2 1295.7 5145.14 $ 1 0 129.6 5U0 • 11423.7 599.2 - V 10 11*2.1* 1*. 1472.7 1638.h 591.5 - 15 253.3 357.9 350.7 I4S6.7 15 195.6 - - 5. 3148.3 1253.9 p - 20 250.8 351.7 1 Ul 2.2 395.0 4 - 20 282.1* 6. U5o 13U6.7 281.6 - 25 336.7 1+68.1+ 1187.7 21*8.3 - 25 296.9 7. 208.1 11*1.9 - 30 290.7 213.5 592.3 11+5.3 - 30 297.7 8. 585 1533.3 - 855.1 0 0 592.U < T lE50i .1i - 837.3 oc 0 9- 571 .lA> 1563.6 - 767.9 5 78.18 505.2 11481.5 - 727.6 5 7U.10 1010. 501.9 1339.9 - 5 0 0 . 9 10 13H.0 1456.1 1293- - 551.3 10 129.3 11. 6149.9 1335.1* - 1*61.7 15 200.3 5114.7 11*88.7 - 51U.6 15 223.3 12. 551.7 1I436.8 391.2 20 287.1 1*16.7 1275.8 - 3 W . k 20 255.2 13. 333.2 1261 - 257.1 25 320.3 367.8 1309.6 - 262.8 25 327.1+ 1U- 35U.6 111U.7 - 162.2 30 33U.U 3145.9 1216.7 - 1 7 7 . 0 30 365.0 - 222 - APPEiiDIX I) Contd. Finishers (3rd Zocpt.) (1 T------------------ Weight j Feed i G. C o m Malze Cassava Cassava Gain Intake Level Intake (gms.) (gms.) (gms.) (gms.) % (gms.) 1. 1173.8 5652.6 3066.8 — 0 04 1316.5 5296.8 2977.3 - 0 0 2. 787.7 5626.2 2660.6 - 281.2 1021 5320.8 2687.5 - 266.0 3. 1151.1 5385.3 2615-9 - 538.5 1130.9 5373.7 . 2610.6 10 537.6 6. 1000 5180.6 2036.9 / 15 777.1 700 6596.3 180K.6 15 689.1 5. 1000 5636.7 1765.3 $ 20 1087.3 987.2 5289.6 1717.5 IV “ 20 1057.9 6. 967.6 5300.6 1 3 9 5 .0 ^ - 25 1325.1 1066.8 5233.6 1378.V - 25 1308.6 7- 766*6 5517.7 - 30 1655.3 890 5516.1 - 30 1656.8 8. 975.9 5676.3 V - 3053.3 0 0 1105.3 5231 \ - 2816.8 0 0 9. 837.6 - 2565.9 5 265.1 965 5630.7 - 2753.9 5 281.5 10. 896.8 ' '5222.6 - 2185.1 10 522.2 1171.-6 5192.8 - 2172.7 10 519.3 11. 967,6 5833*7 - 1950.8 15 875.1 1021.3 5652.8. - 1890.3 15 8 6 7 .9 12. 1021.3 5626.5 - 1528.2 20 1 1 2 6 .9 863.6 5307.8 - 1662.1 20 1061.6 13. 1191.6 5561.2 1163-7 25 1385.3 * 1010.9 i 6361.5 1 * - • 1293.3 30 1560.6 16. 1006.3 1;1 5359.3 j 795.3 3c 1607.8 1166.7 j 5626.9 j - 836.7 30 1687.5