Intercropping Withcassava Proceedings of an International Workshop Held Attrivandrum, India, 27 Nov-I Dec1978 Ectors: Ecward Webec Barry \Estel Anc Ariyn Campdeh

Home , Cassava

Intercropping WithCassava Proceedings of an international workshop held atTrivandrum, India, 27 Nov-i Dec1978 Ectors: Ecward Webec Barry \esteL anc ariyn CampDeH

ARCHIV 38421

IDRC-142e The International Development Research Centre is a public corporation created by the Parliament of Canada in 1970 to support research designed to adapt science and technology to the needs of developing countries. The Centre's activity is concentrated in five sectors: agriculture, food and nutrition sciences; health sciences; information sciences; social sciences; and communications. IDRC is financed solely by the Government of Canada; its policies, however, are set by an international Board of Governors. The Centre's headquarters are in Ottawa, Canada. Regional offices are located in Africa, Asia, Latin America, and the Middle East.

Weber, E. Nestel, B. Campbell, M. IDRC, Ottawa, CA Central Tuber Crops Research Institute, Trivandrum, IN IDRC-142e

Intercropping with cassava : proceedings of an international workshop held at Trivandrum, India, 27 Nov-i Dec 1978. Ottawa, Ont. IDRC, 1979. 144 p.: ill.

/IDRC publication!, !conference paper!, !cassava!, /intercropping!, !crop diversification!, !agricultural research!, /research centres!, !Latin America/, !Asia!, !Africa/ - !agricultural management!, !plant protection!, /soil fertility!, /economic aspects!, !CIAT!, !IRRI/, !agricultural statistics!, bibliography.

UDC: 631.584:635.23 ISBN: 0-88936-231-9

Microfiche edition available IDRC- 142e

Intercropping with Cassava

Proceedings of an international workshop held at Trivandrum, India, 27 Nov-i Dec 1978

Editors: Edward Weber,' Barry Nestel,2 and Marilyn Campbell3

Cosponsored by the Central Tuber Crops Research Institute (Indian Council for Agricultural Research) and the International Development Research Centre

1Senior Program Officer, Agriculture, Food and Nutrition Sciences Division. IDRC. 2Consultant to the AFNS Division of IDRC. 3Communications Division, IDRC. Contents

Foreword Edward Weber and Barry Nestel 5 Multiple cropping cassava and field beans: status of present work at the International Centre of Tropical Agriculture (CIAT) M. Thung and J.H. Cock 7 Intercropping with cassava in Central America Raüf A. Moreno and Robert D. Hart 17 Cassava intercropping in Brazil Marcio Carvaiho Marques Porto, Pedro Alves de Almeida, Pedro Luiz Pires de Mattos, and Raymundo Fonséca Souza 25 Intercropping systems with cassava in Kerala State, India C.R. Mohan Kumar and N. Hrishi 31 Cassava intercropping patterns and management practices in Indonesia Suryatna Effendi 35 Cassava intercropping patterns and management practices at Tamil Nadu Agricultural University, Coimbatore, India C.R. Muthukrishnan and S. Thamburaj 37 Cassava intercropping patterns and practices in Malaysia W.Y. Chew 43 Intercropping with cassava in Africa W. N. 0. Ezeilo 49 Cassava and cassava-based intercrop systems in Thailand Sophon Sinthuprama 57 Cassava intercropping research: agroclimatic and biological interactions H.G. Zandstra 67 Soil fertility implications of cropping patterns and practices for cassava Jerry L. McIntosh and Suryatna Effendi 77 Agroeconomic considerations in cassava intercropping research J.C. Flinn 87 Agronomic implications of cassavalegume intercropping systems Dietrich E. Leihner 103 Crop protection implications of cassava intercropping RaülA.Moreno 113 Discussion summary 129 Bibliography 133 Participants 141

2 Multiple Cropping Cassava and Field Beans: Status of Present Work at the International Centre of Tropical Agriculture (CIAT) M. Thung and J. H. Cock International Centre of Tropical Agriculture, Cali, Colombia

The cassava program at the International Centrethe cassava crop, in such a manner that it did not of Tropical Agriculture (CIAT) has basically been compete for light but utilized the light not intercept- developing monoculture systems for cassava pro- ed by the cassava. duction. However, an agroeconomic survey in Col- Not only has multiple cropping been suggested as ombia showed that about 40% of cassava is grown a means for maximizing capture of solar energy, but in mixed culture (Diaz and Andersen 1977). The also as a probable control method for diseases and authors furthermore estimated that in Latin Americapests. This pest control could be of tremendous im- the figure of 40% is probably fairly representative. portance for the small farmer with limited resources Okigbo (1976) estimated that in Africa about halfwho cannot use purchased inputs to control disease the cassava is grown in mixed cropping systems. In and pests. Hence we have started to investigate the general, the mixed cropping systems are used by theadvantages of mixed cropping in this context. smaller farmers from the lower income groups. In Whilst planning the trials to investigate multiple the last 2 years, the CIAT Cassava Program has cropping systems we became aware that experimen- started serious investigations on cassava intercrop- tal design was often necessarily extremely complex. ping systems. Many of the trials done previously on multiple crop- The cassava crop is among the most efficient cal- ping systems have used a number of different treat- orie producers (250 kilocalories ha-1 day1) of thements and cropping systems; however, the results major crops (Coursey and Haynes 1970). The pro- have been difficult to interpret because of the enor- tein content of cassava is, however, very low and mous number of interactions. In our trials we have hence in CIAT we decided to concentrate on cassava tried to study the effects of varying only one param- grain legume mixes, so as to give a combination that eter and have observed this factor in detail. The au- might more nearly satisfy dietary requirements. thors feel that this approach may not necessarily lead The emphasis of the research has been on cassavato immediate optimum solutions to multiple crop- and Phaseolus vulgaris as CIAT has expertise on ping systems but that it will eventually give the in- both these crops. We have not tried to develop mul- sight necessary to design successful multiple crop- tiple cropping systems as such, but rather have at- ping systems. tempted to define what parameters are important in determining the yield of the two crops when grown together and later use this knowledge to design more efficient systems of production. Experimental Site The cassava crop is generally slow to cover the ground, with complete cover occurring some 3 All experiments were carried out at CIAT head- months after planting. Changing the spatial arrange- quarters, which is located 3° north of the equator, ment from rectangular to square can further delayat 1050 m above mean sea level, on a heavy fertile the time to full ground cover without apparently re- alluvial soil. Mean annual temperature is 24 °C with ducing cassava yield (Cock et al. 1978). The cassava little seasonal change; the rainfall is bimodal with crop is, however, highly sensitive to early compe- two wet and two dry seasons with a total of approx- tition (Doll and Piedrahita 1974). The bean crop imately 1100 mm per year. Beans were irrigated as develops very rapidly and often completes its growth necessary during the dry periods. The characteristics cyclein 90daysorless. This ledus tothehypothesis of the different cassava clones and grain legumes are that it should be possible to grow a bean crop under shown in Table 1.

7 Table 1. Characteristics of cassava and dry bean varieties used in the experiments.

Cassava: Manihot esculenta (Crantz) M Mex 11 A germ-plasm collection variety; has a medium early growth vigour, single erect stemmed and late branched; a good and stable yield performance. M Mex 59 A germ-plasm collection variety; has a moderate early growth vigour but early branched type; heavy foliage stand 3 months after germination; also a good and stable root producer. M Col 113 A germ-plasm collection variety; has a poor early growth vigour, early branched type; heavy foliage stand 3 months after germination.

Bean: Phaseolus vulgaris L. P 302 A black bush bean with a growth habit II, indeterminate erect type; growth cycle 80 days at CIAT conditions; a good yielder. P566 A bush black bean with growth habit II, indeterminate erect type, growth cycle less than 80 days at CIAT conditions. P498 A black bean with growth habit IV, indeterminate prostrate type, growth cycle more than 90 days at CIAT conditions.

Planting Date 100

When the relative date of planting of beans and 80 cassava (P 302 and M Mex 11 respectively) was al- tered, the relative yield of the two crops changed markedly (Fig. I). The yield of beans was greatest 60 when the beans were planted before the cassava and lowest when planted after. The yields of cassavaa 40 were completely in the opposite direction. The maximum reduction of cassava yield was 25% when the 20 beans were planted 2-4 weeks before the cassava. Planting cassava and beans at the same time gave the highest Land Equivalent Ratio (LER) (1.7). The LER was corrected for total time of land use, which -6 -4 -3 -2 -1 0 0 2 3 4 was greater when beans were planted before the cas- Planting date of casnava sava (Fig. 2). The yield of fresh cassava was 35 Intervala (In weaka) of bean planting date In relatIon to planting date of cenanve t/ha and of beans (14% moisture) 2.9 t/ha whenFig. 1. Yield of cassava (root dry matter) and beans (14% planted at the same time; these high yields suggest moisture content) at different relative planting dates. great potential for cassavabean mixed cropping systems. Soybeans (var. ICA-Tunia) were planted 1 month Plant Population before, at the same time, and 1 month after cassava. The medium vigour late branching M Mex 11 and To study different plant populations using a tra- the vigorous early branching M Mex 59 cassava ditional randomized block design it was necessary clones were used. When soybeans were planted 1 either to use a restricted number of plant populations month after cassava, yield was essentially zero andor a very large experiment so as to eliminate border when planted 1 month before the cassava, soybean effects. To avoid these problems the cassava plant yields were the same in mono or mixed culture. population was varied by systematically changing When planted at the same time as the cassava, yieldsthe distance between plants in the row. Hence each of soybeans were slightly reduced (Fig. 3). Yield ofpopulation is bordered on one side by a slightly cassava roots was very variable due to a high percent lower plant population and on the other side by a of root rot caused by very heavy rain just before har- slightly higher plant population. vest. The total harvested biomass of cassava was, In a monoculture system, the yield of the two bean however, very greatly reduced by planting soybeans varieties P 302 and P 498 were not significantly dif- before cassava. It appears once again that the opti-ferent and were not insignificantly increased by in- mum planting date is both crops at the same time creasing the plant population up to 40 plants/m2. The (Fig. 4). yield of P 498 in association was not affected by

8 LER

2.0

1.8

1.6

1.4

I .2

6 4 3 2 1 0 1 2 3 4 6

Weeks before Weeks after

Bean planting date dative to cassava Fig. 2. Land equivalent ratio (LER) of cassavahean associations at different relative planting dates (cassava, M Mex II;bean, cv P302).

60 I!liik4" 25

One month 0 month before the cassava after the cassana Soybean planting internal Fig. 3. Dry soybean yield at maturity when soybean was planted before, at the same time, or after cassava (monoculture yields, 3.4 tI/ia). changes in plant population, but P 302 showed a sava density increased (Fig. 6). The maximum yield tendency for yield to increase with increasing plantof cassava in monoculture was 39.4 metric tonnes population. The bean yields in association with M (t) per hectare with M Mex 11, and 26.0 t per hectare Col 113 were always higher than with M Mex 11 with M Col 113. The cassava yield of M Mex 11 (Fig. 5). M Col 113 is less vigorous than M Mex 11 increased as cassava population increased either in in the early stages. Bean yields decreased as the cas- monoculture (Fig. 7) or in association (Fig. 8),

9 M Meo 11 II Moo 59

100

40 'S Cassava total biomso Cassava total bio 55

0-0 Soybean Soybean 20

0 One month Planting date One month One month Planting date One month before cassava of cassava after cassava before cassava of cassava after cassava Planting date of soybean Fig. 4. Relative yield of cassava and soybean as affected by different planting date of soybean.

P 302 Pusblo 280 280

Monocul tus 260 260 Monocul tees

240 240

180 180 ILSD 0.05 auociotion joi*ociaflon

160 Auociot.d with MCo1 113 Auociotsd 140 withMColIl3 140 LSD 0.05 Bon dinilty Astociatsd with M Msx 11 120 120 Asaociotsd with Msx 11

. C.V. 11% C.V. 11% 100 I 100 10 20 30 40 10 20 30 40

Oson population p1/rn2 Fig. 5. Dry bean yield as affected by plant density of beans, alone and in association with cassava. whereas the yield of M Col 113 tended to declineimproves yield (CIAT 78). Intercropping M Col 113 when the plant population increased. The yields ofwith the more vigorous P 498 did not give the same M Col 113 when intercropped with P 302 were result, probably because the competition of the P sometimes higher than in monoculture of M Col 113 498 with the cassava reduced LAI below the opti- itself. This apparent anomaly is probably due to a mum and hence reduced yield. Cassava yields were reduction of early growth, which reduces leaf area not affected by bean population. index (LAI) to the optimum level and consequently The cassava yields in association with P 498 were

10 70 -

50 - P 498

I 5I5 6.2 8.0 11 14.0 18.5

M Col 113 70 -

50 - P 302

30-

55 6?9 9'.5 6.2 8.0 11 14.0 18.5 Cassava densities x 1000 plants/ha Fig. 6. Dry bean yield as affected by cassava densities. less than those with P 302; however, P 498 gave was no correlation between final cassava yield and higher yields in association than P 302. Hence it may bean yield. This suggests that cassava types can be be difficult to select bean varieties that yield well in selected that cause minimum reduction in bean yield association but do not greatly depress the cassavabut also yield well. Cock et al. (1978) have shown yield. The highest LERs were in fact obtained by the that late branching is a desirable character for high combination of P 302 and M Mex 11 (Fig. 9). It yields in monoculture; late branching varieties pre- should be noted that the growth cycle of P 302 is 80 sumably will give less shade to the grain legume in days and P 498 is more than 90 days. These LERsthe early stages. were obtained at normal monoculture densities for When 20 cassava clones were planted with a 125- cassava (10 000 plants/ha) and beans (250 000 day soybean (var. ICA-Tunia) there was a negative plants/ha). correlation between cassava root yield and soybean yield (Fig. 11). This suggests that with a short season (less than 90 days), grain legume, cassava types Genotypic Variation can be selected that are both high yielding and also allow enough light to pass to enable the grain legume Twenty different cassava clones were planted 2 to yield well; however, in the case of longer season weeks before the black bean P 566. The bean yield grain legumes this cannot be achieved as the low was negatively correlated with the top growth of light interception of cassava for a period of 4 months cassava 3 months after planting (Fig. 10) but there is detrimental to its final yield.

11 t/ha 40 Cassava monocul ture

M Mex 11

N Col 113 30

I I 5.5 6.9 9.5 6.2 8.0 11.0 14.0 18.5 Cassava densities x. 1000 plants/ ha Fig. 7. Cassava yield as affected by plant densities.

Pest Control The populations of lace bug and whiteflies were less in the intercropping system than in the cassava The monoculture system has been criticized be- monoculture with or without plant protection. Horn- cause the genetic and stand uniformity results in worm populations were lower in the intercropping continuous pest susceptibility (Pimentel 1961; system than in the cassava monoculture without in- Southwood and Way 1970; Nickel 1973). Multiple secticide (Table 2). cropping systems on the other hand are praised be- Similar results were obtained in the beans: Dia- cause the diversity of vegetation within the crop area brótica and Ceratoma and thrips populations were can be used to give integrated pest management lower in the intercropping system compared to the (DeLoach 1970; Dempster and Coaker 1974; Altieri monoculture system and the lowest insect popula- et al. 1978). tion was found in the intercropping system when in- To study the plant protection potential of the insecticides were applied (Table 3). tercropping system, a field experiment was set up The yield of cassava was reduced little by mixed and the insect population was measured in a cassava cropping with beans in the no insecticide treatment bean intercropping system both with and without and the bean yield was essentially the same in the insect control systems. mixed and monocropping system (Fig. 12). Hence,

12 ft 5.. M Mex 11

4.3 .; 100 P 302 0U C

ft 498 80

0 4.3

.ft >' 120 43 M Col 113 4.3 ft P 302 ft . 100 ft ft ft 80

P 498 60

69' 9'.5 6.2 8.0 11 14.0 18.5 Cassave densities x 1000 plants/ m2 Fig. 8. Cassava yield when zntercropped with P 302 and P 498 as affected by different cassava densities.

LER 1.4

1.3

1.2 M Mex 11/P302 o M Mex 11/P498 1.1 M Col 113/P302 x M Col 113/P498

0.1 ss 62 69 00 to\110 ,.io 180

0.9

Cassava densities 0.8 x 1000 plants/ha Fig. 9. Land equivalent ratio qf different combinations as compared to best inonocu/ture yield of cassava and bean.

13 %

120 r - 0.495* -

L 2 * U

o I S C I

2 100 I- S S S S I S S S o I S I S -D I 1i 901- S

C 0

0 250 300 350 400 450 500

Average stem and leaf weight of 20 cassava varieties at 3 months (g DM/plant) Fig. 10. Bean yields relative to monoculture as affected by different cassava genotypes.

S 80

A -0.347 U 0 S C V : 49.4 - (0.4)X 60 0

a) 50 S C (0 a) S .0 S a), 0 S S 40 S

(0 S S a) 30 S S S S S S S 20 S

5 15 25 35 45 55 Cassava fresh root yield (t/ha) Fig. ii. Soybean yields relative to monoculture as affected by different cassava genotypes.

14 Table 2. Insect population in cassava monoculture vs. intercropping system cassava-bean, with and without insecticide treatment.

% in Cassava monoculturea Cassava intercmpped5 population reduction by With Without With Without different Insect insecticide insecticide insecticidec insecticide system Lace bug, 3 leaves/plant 4.5 5.6 2.7 3.8 32.1 White fly 3 leaves/plant 7.2 6.5 3.3 4.5 30.0 Shootfly/plot 2.1 2.0 1.5 2.2 Hornworms/plot 0.5 0.8 0.5 0.6 30.0 Homworm eggs/ parasitized plot 50.7 52.4 51.2 45.5 acassava variety M Mex 11. 5Bean variety P566 with M Mex 11. cEndosulfan

Cassava 24 Bean HIGH INPUT

LOW INPUT

18 2600

16 2200

14 1800

12 1400

10 1000 IMono. IAss. Mono. Ass. SYSTEM Fig. 12. The effect of different input on yield of cossava-bean in association.

15 Table 3.Insect population in bean monoculture vs. intercropping system cassava-bean, with and without insecticide treatment.

% in Cassava monocuttur&' Cassava intercroppedb population reduction With Without With Without by different Insect insecticidec insecticide insecticide insecticide systems Leaf hoppers (Empoasca kraemeris) /20m2 89 229 80 216 6 Chrysomelides/20 m2 3 6 3 4 34 Trips/inch2 2 14 6 12 14 Scaphitophius/m2 25 35 20 36 aCassava; M Mex II. bP566xMMex 11. CEndosulfan in the case of low inputs, the LER was very closeplanting appears preferable. It appears possible to to 2 (1.95) whereas in the insecticide-treated plotsselect for cassava varieties that are high yielding and it was 1.89. This data suggest that under minimumalso combine well with short growth cycle legumes; input technology, multiple cropping will show the however, if the legume growth cycle is greater than greatest advantage. 90 days this may be difficult. Both bean and cassava plant populations could be varied considerably with little effect on the LER, and normal monoculture Conclusions densities gave good mixed cropping yields.

Cassava can successfully be grown with short The relative advantage of mixed cropping with growth cycle grain legumes. High yields of both beans and cassava was greatest at low input levels crops can be obtained and LERs of the order of 1.5 of insecticides, suggesting that mixed cropping sys- are quite feasible. Planting date is a very important tems may be most suitable for small-scale producers determining factor in the system, and simultaneous with limited resources to purchase inputs.