Cotton crops versus weeds : when is the competition period?

Cotton crops

versus

weeds:

when is the competition period?

In A frica, cotton crops

are often first weeded when

weed competition is already

high. Is there a crucial period

during which weed infestation

is particularly harmful?

Three years of research

have provided an answer

to this question.

I

n agriculture, it is essential to determine when crops w ill be the most and least har­ med by weeds. DECOIN (1992) noted that some weed infestations can have a less

P. Y. DOUTI

A nié-M ono Research station BP 1, Anié, Togo

This study w as carried out in collaboration with K. DJAGNI (CIRAD-IRCT, Togo) and E. JALLAS (CIRAD-CA, France).

serious im pact on crops than th e ir visual appearance would indicate. For NIETO, quo­ ted by GURNAH (1974), cropfields have to be kept clear of weeds during "critical per­ iods" of the crop growth cycle in order to a c h ie v e m a x im u m y ie ld s . CAUSSANEL (1989) d efines this as the p e rio d d u rin g w h ic h the presence o f weeds can lead to measurable yield losses. This is the prime period for conducting weeding operations, w ith the f o llo w in g factors also h a v in g a considerable influence: the environment, the extent of weed infestation in the cropfield, the weed flora composition, crop density, etc. For exa m p le , a c c o rd in g to KASASIAN & SEEYAVE (1969), in Jamaica, Trinidad and Tobago, the critical period for weed competi­ tio n is 2-4 weeks after sow ing for French beans (Phaseolus vulgaris) and 3 weeks after sowing for sweet potatoes (Ipomoea batatas). According to the same authors and under the

Brachiaria lata.

harmful w eeds

Experiments with cotton crops

This study was carried out from 1988 to 1990 at

t h e A n i é - M o n o re s e a rc h s ta tio n in T o g o , on vertisol, w ith 1 100 to 1 200 m m annual rainfall (F igu re 1). The tr ia ls w e re c o n d u c te d in p lo ts c r o p p e d w i t h r a i n f e d c o t t o n ( G o s s y p i u m

h irs u tu m ), in v o lv i n g v a r ie tie s w i t h a 1 6 0 -d a y

cycle (Stam F, 1988; Stam 45-E, 1989; Stam F-59, 1990). The cotton plants w e re sown on ridges at 0 .8 0 m X 0 .3 0 m s p a c in g (th e o re tic a l p la n tin g rate: 41 6 6 6 plants/ha). The c u ltiv a t io n p e rio d was fro m July to D ecem ber. Rainfall d u rin g this period was 556 m m (1988), 771 m m (1989) and 4 7 5 m m (19 90 ). T he p lo ts w e re f e r t iliz e d and treated w ith pesticides.

Methods

The m ethod for de term in ing critica l periods in the c ro p g ro w th c y c le , fro m the studies o f NIE TO , PETERS, W EARER & D A W S O N , as q u o te d by CAUSSANEL (1989), was based on the results of an exp erim e nt in v o lv in g tw o series o f trials aimed at analysing y ie ld in terms o f c o m p e titio n tim e. In the first series, weeds w e re a llo w e d to g ro w f r e e l y f o r p r o g r e s s iv e l y lo n g e r p e r io d s a fte r s o w in g before the y w e re c o n tro lle d , in ord e r to determ ine early com p etitio n (Ce). In the second series, the cropped plo t was kept weed-free from s ow ing until progressively later dates, in order to assess late c o m p e titio n (Cl).

Experimental design

A randomised Fisher blo c k design w ith 8 repeats was used for this study. There w e re fo u r ridges, 10 m lo n g X 0 .8 0 m apart, on each basic p lo t.

Early competition treatm ents:

The plots were kept weed-free from the indicated first weeding date.

W1 W 2 W 3 W 4 W 5 W 6 Plot code W 7 W 8 W 9 i r Sowing date 14 I I I I I I I I 21 28 35 42 4 9 56 63 70 Date of the first weeding (number of days after sowing)

Late competition treatm ents:

The plots were kept weed-free from sowing until the indicated last weeding date. Plot code W 1 0 W11 W 1 2 W 1 3 W 1 4 W 1 5 W 1 6 W 1 7 W 1 8

r

i

I

I

1

1

I

I

I

I

Sowing date 14 21 28 35 42 4 9 56 63 70 Date of the last weeding (number of days after sowing)

The plots were weeded every 14 days after the first weeding or before the last Plot W 1 9 was weeded throughout the crop cycle

Plot W 2 0 was never weeded

Plot W21 was weeded 20 days and 40 days after sowing Plot W 2 2 was treated with a pre-emergence herbicide then weeded

Figure 2. Description of the trial plots.

O n l y t h e t w o c e n t r a l r id g e s o n e a c h p l o t w e re m o n ito re d . A ll tests ov e r the 3-year study p e r i o d w e r e c o n d u c t e d at t h e s a m e s ite . H o w e v e r , a n e w r a n d o m d i s t r i b u t io n o f fie ld treatments was do ne each year.

Treatments

The different treatments were as follows (Figure 2): - plots assigned to the early c o m p e titio n groups w e r e w e e d e d f r o m th e d a te in d ic a t e d in the p r o to c o l. Subsequent w e e d in g w as d o n e every 14 days until c rop d e v e lo p m e n t was considered sufficient;

- for the late c o m p e titio n group, the plots were kept weed-free from sow ing until the date o f the last w e ed ing established in the pro to col. The first w e e d in g began 14 days a fte r s o w in g and was repeated every 2 weeks thereafter;

- t h e w e e d - f r e e c o n t r o l p l o t W 1 9 ( r e g u la r m aintenance) was w e e d e d 7 days after sowing, then w eekly. W e e d in g was stopped w hen the cot­ ton plants had gro w n sufficiently;

- p lo t W 2 0 (no m a in te n a n c e ) was not w eed ed d u r i n g t h e c r o p c y c l e , w h e r e a s p l o t W 2 1 (recom m ended m aintenance) was weeded for the f i r s t t i m e o n d a y 2 0 , a n d t h e n w e e d e d and earthed-up on day 40;

- p lo t W 2 2 was treate d w ith a p re -em erg en ce h e rb ic id e c o m b in in g d ip r o p e tr y n (24 0 g/l) and m e to la c h lo r (160 g/l), at a dose o f 1 600 g/ha of total a c tiv e in gred ie nt. A fte r the h e rb ic id e too k e f f e c t , t h e p l o t w a s m a n u a lly w e e d e d w h e n necessary.

A t each weeding, flora inventories are conducted, w h ile noting weed abundance. Final cottonseed yields w ere also determ in ed.

harmful w eeds

same conditions, weed infestations can be v e ry h a r m fu l to to m a to ( L y c o p e r s ic u m

esculentum) crops w ith in the first 30 days

fo llo w in g transplanting. Weed competition can also have a serious impact on other crops over extended periods, e.g. yam (Dioscorea spp.) throughout its grow ing cycle (KASA- SIAN & SEEYAVE, q u o te d by G U R N A H , 1974). In Sudan, the critical period for most crops is 4-6 weeks after sow ing (CROW - THER, quoted by H A M D O U M & TIG ANI, 1977). NIETO et al. (1968) demonstrated that w e e d c o v e r th re a te n e d m a iz e cro p s in Mexico w ith in the first 30 days post-emer­ gence. In Togo, this period occurs 10-24 days after sowing (SCHMID et a!., 1983). In rain- fed rice cropfields in southern Togo, weed competition is most serious 21-30 days after sowing (BOYODA, 1991).

In addition, on the basis of agronomic and economic analyses conducted in five villages in Togo from 1983 to 1988 (COUSINIE et al., 1989), herbicide experiments were carried out in smallholders' cotton cropfields from 1987 to 1989 (FAURE et al., 1988; TONATO

et al., 1989 & 1990). The results revealed that

farmers generally weeded their fields once or twice (extension services often recommend tw o weedings), but these operations were often too late. Since the technical guidelines established for cotton cropping were insuffi­ cient, a 3-year study was undertaken (1988-1990) to identify the weed flora and the criti­ cal period for weed competition with respect to cotton.

Weed flora

identification

Over the study period, 37 weed species were identified in 1988, 42 in 1989 and 46 in 1990 (Table 1).

There were always substantially higher num­ bers of broad-leaved species as compared to Cyperaceae and Poaceae species. However, the latter group was generally dom inant in terms of overall plant numbers (quantitative dominance).

Brachiaria lata (Schumacher) C.E. Hubbard,

the most common weed in 1988 and 1989, was dominated by Echinochloa colona (L.) Link in 1990. In 1988, B. lata was so dom i­ nant that the weed p opulation was consi­ dered monospecific in this trial.

Table 1. N u m b e r o f species inventoried.

W eed group 1988 1989 1990

Poaceae & Cyperaceae 9 12 13

Broad-leaved species 28 30 33

Total 37 42 46

Phyllanthus amarus. Photo CIRAD-AMATROP

Apart from changes in weed flora concerning

B. lata and E. colona, Cynodon dactylon (L.)

Pers., w hich was relatively scarce in 1988, flourished in the following years. Some culti­ vation practices prom ote the propagation of this stolon- and rhizom e-bearing plant: ploughing, cross-harrowing (disking), etc. Infestations of Cyperus esculentus L., Cyperus

ro tu n d u s (L.) and D ig it a r ia h o r iz o n t a lis

W i l l d e n o v also in c re a s e d . W e e d flo r a patterns therefore changed over time, notably an in cre a se in C. d a c t y lo n , w h ic h was generally responsible for the intra-block hete­ rogeneity noted in 1990.

The main weeds encountered yearly over the 3-year study, in order of importance, were: - 1 9 88: B. lata and P h y lla n th u s amarus Schumacher & Thonning;

- 1989: B. lata, C. esculentus, C. dactylon and Phyllanthus amarus;

- 1 9 9 0 : E. c o lo n a , B. lata, C. d a c ty lo n ,

harmful w e ed s

A marked influence of

competition periods

Gross cottonseed yields (observed and rela­ tive) obtained in the 1988 and 1989 trials are given in Table 2. They were calculated in relation to a weed-free control (W19). The 1990 results were not taken into account in the analysis of means, because of the high heterogeneity induced by weed flora changes (dominance of C. dactylon in patches) and by the loss of plants during their growth cycle. The mean cottonseed yields per plot for the 1988-1989 period indicated a production potential of around 2 000 kg/ha of cotton­ seed, for the varieties studied, in the weed- free control (W19), whereas the unweeded control (W20) yielded only one tenth of this (200 kg/ha).

In the early competition group, mean cotton­ seed yields were equal to that of the weed- free c o n tro l w h e n the firs t w e e d in g was performed 14, 21 or 28 days after sowing (plots W1-W3). There were significant yield differences relative to the weed-free control when this first w eeding was done 35 days after sowing or later.

For the late competition group treatments, the results did not differ significantly from the weed-free control when weeding was stop­ ped 4 2 -7 0 days after sow ing (plots W 1 4 - W18). Under our experimental conditions, only plots in which weeding was stopped 14, 21, 28 or 35 days after sowing (plots W 10- W 1 3) gave significantly lower yields than the weed-free control.

Cyperus esculentus. For the treatment with two weedings (E21 ), at

Photo H. M e rlier 20 and 40 days after sowing, and the treat­ ment with a herbicide treatment and weeding (W22), the cottonseed yields did not differ significantly from that obtained on the weed- free control plot.

Competition thresholds

The early weed competition threshold is de­ fined as the date of the first w eeding after which a significant difference in crop yields, relative to a permanently weed-free control, is recorded. The late weed competition thres­ hold is the date of the last weeding beyond which a significant difference in crop yields, as compared to the weed-free control, is not observed.

T a b le 2. C o tto n s e e d y ie ld s o b ta in e d fo r th e d iffe re n t tre a tm e n ts (m eans o f 1 9 8 8 a n d 1 9 8 9 results, fo r w h i c h y ie ld s w e r e v e ry s im ila r, in kg /h a a n d p e rc e n ta g e re la tiv e to th e w e e d -fre e c o n t r o l W 1 9).

T reatm ent M e a n y ie ld Result Significance 1 9 8 8-1 9 8 9 (kg/ha) % o f W 1 9 Early competition W1 1 788 97 Results W 2 1 731 94 n o t s ig n ific a n tly W 3 1 644 89 d iffe re n t fr o m c o n tro l Late competition W 1 4 1 728 94 W 1 5 1 639 89 W 1 6 1 757 95 W 1 7 1 780 96 W 1 8 1 800 98

W eed -fre e c o n tro l

W 1 9 1 845 100 S m a llh o ld er p lo t

W21 1 645 89

Plot w ith h e rbicide

W 2 2 1 705 92 Early competition W 4 1 375 75 Results W 5 1 411 76 s ig n ific a n tly W 6 1 051 57 d iffe re n t from W 7 969 53 c o n tro l W 8 806 44 W 9 676 37 Late competition W 1 0 579 31 W11 953 52 W 1 2 1 239 67 W 1 3 1 488 81 U n w e e d ed c o n tro l W 2 0 233 13 M ean 1 356 C o e ffic ie n t o f variation 11.5

D u n n ett's Test, 5% sig n ifica n ce threshold, relative to c o n tro l W 1 9.

harmful w e ed s

Under our experimental conditions, the early weed com petition threshold (Te) occurred 28-35 days after sowing and the late weed com petition threshold (Tl) was 35-42 days after sowing. The c ritic a l period for weed competition (Cw) was estimated as the time in te rv a l b e tw e e n these tw o th re s h o ld s , i.e. 28-42 days after sowing (Figure 3). NIETO

e t al. (1 9 6 8 ) stated th a t w e e d s s h o u ld

theoretically be destroyed during this period.

Te: early competition threshold; Tl: late competition threshold; Ce: early competition; Cl: late competition; W p : permanent weeding; Pc: critical period for weed competition

J

2 000-1 'ro 1 5 0 0 - >-1 000-Wp

Conclusion

Now that this critical period has been establi­ shed, it would be of interest to clarify certain aspects of weed c o m p e titio n w ith cotton crops.

Several maintenance conditions should be defined and compared:

- how should cotton crops be kept weed-free during the critical period?

- is it important to weed before and after this period?

- are three successive weedings efficient: a first weeding w ithin 28 days after sowing, a second during the critical period, and a third after 42 days?

- is the technique involving treatment with a p re -e m e rg e n c e h e r b ic id e f o l l o w e d (if necessary) by manual weeding sufficiently effective?

Com m ellna benghalensis. Photo H M erlier

Critical period for weed

competition in cotton

When are the critical periods for

cropfields

weed competition in other countries?

500---1

---- 1

----1

---- 1

----1

----1

---- 1

----1

----1

---- 1

7 14 21 28 35 42 49 56 63 70 Days after sowing

Figure 3. Cottonseed yields in plots subjected to early competition (Ce) and those subjected to late competition (Cl). Determination of critical period (Pc) for weed competition in cotton cropfields.

In the present study, the critical period for weed competition was approximated for cot­ ton crops in Togo (28-42 days after sowing), and the detrim ental effects of early weed competition was demonstrated. The graphs in Figure 3 highlight a serious cottonseed yield loss due to early weed c o m p e titio n and a much smaller loss due to late competition. The differences recorded in 1990 indicate that these conclusions are not definitive, as the e n v iro n m e n t and spatial d is trib u tio n of weed species are im portant factors that require further consideration.

The c r itic a l p e rio d varies in d iffe re n t co u n trie s . T he results o f a study o f SCHWERZEL & T H O M A S (1971) in Z im b a b w e (1967-1969) showed that the critical period for weed com p etitio n in out-of-season irrigated cotton fields was 2-4 weeks post-emergence, w h ile in rainfed cotton cropfields this critical period was as late as 6-8 weeks post-emergence. In C o lum b ia, according to P E RD O N O quoted by SCHWERZEL & T H O M A S (1971 ), this period extends fro m th e 2 0 ^ to th e 4 5 ^ da y p o s t-e m e rg e n c e fo r c o tto n . These results are c lo s e to th o s e o b ta in e d in T o g o . P E R D O N O c o n f ir m e d th a t w e e d c o m p e t it io n is no lo n g e r d e t r im e n ta l to c o tto n c ro p s a fter 45 days. In T a n z a n ia , th e c r it ic a l p e rio d fo r c o tto n ranges fro m 30 to 75 days after s o w in g (SAKIRA, q u o te d by G U R N A H , 1 9 74 ), w h erea s in M ah a ra s h tra State in India, SHELKE & BHOSLE (1990) found that this period ranged from 20 to 60 days after sowing.

harmful w eeds

Bibliography

BO Y O D A T.B.K., 1991. Contribution à l'étude de la nuisibilité des adventices en riziculture pluviale au Togo. Direction de la recherche agronomique de Lomé et u ni­ versité du Bénin. Lomé, Togo, 103 p.

CAUSSANEL J.-P., 1989. N u is ib ilité et seuils de n u is ib ilité des mauvaises herbes dans un système de culture annuelle : situation de concurrence bispécifique. Agronomie 9 :219-240.

C O U S IN I E P., D J A G N I K., FAURE G ., 1 9 8 9 . L'agriculture togolaise en zone cotonnière : de l'analyse a u x r e c o m m a n d a t io n s . R a p p o r t de s y n th è s e des campagnes agricoles 1985 à 1989. CIRAD-IRCT, Lomé, Togo, 25 p.

D E C O IN M ., 1 9 9 2 . ITCF et le d é s h e rb a g e des céréales : l'im p a s s e ne passera pas. P h yto m a 4 4 3 : 31-32.

FAURE L., JALLAS E., D O U T I P.Y., FAURE G., D J A G N I K., 1 9 8 8 . R a p p o rt a n n u e l 1 9 8 7 . Exp é rim e n ta tio n h e rb ic id e en m ilie u réel. Projet de r e c h e r c h e - d é v e lo p p e m e n t. C IR A D -IR C T , s e c tio n s a g ro n o m ie et a g ro -é c o n o m ie , station d 'A n ié -M o n o , Togo, 20 p.

G U R N A H A .M ., 1974. C ritical weed c o m p e titio n perio ds in annuel crops. Proceedings o f the 5th East African Weed control conference, 1974, p. 89-98.

H A M D O U M A . M . , T IG A N I K .B ., 1 9 7 7 . W e e d control problems in the Sudan. Pans 23 (2) : 190-194.

KASASIAN L., SEEYAVE J., 1969. Critical periods for weed competition. Pans 15 (2) : 208-212.

N IETO J.H., B R O N D O M .A ., G O N Z A L E Z J.T., 1 9 6 8 . C r itic a l p e rio d s o f th e c ro p g ro w th c y c le fo r competition from weeds. Pans 14 (2) : 159-166.

SCHMID W., DOSSEKOU M „ KOCH W., WALTER H „ 1983. Aspects de la m odification de la flore adventice dans le systèm e de p r o d u c t io n a g r ic o le du T o g o .

In Comptes rendus de la deuxièm e conférence bisan­

nuelle de la Société ouest-africaine de m alherbologie (SOAM), 17-22 october 1983, A b id ja n , Côte d 'iv o ire . SOAM, IDESSA, Bouaké, Côte d'ivoire, p. 19-33.

SCHWERZEL P. J., T H O M A S P.E.L., 1971. W eed competition in cotton. Pans 17 (1) : 30-34.

SHELKE J.H., BHOSLE R.H., 1990. Determination of c r itic a l p eriods o f c ro p w e e d c o m p e titio n in rainfed cotton. Journal o f Maharashtra Agricultural Universities

15 (2) : 257-258.

T O N A T O S.H., JALLAS E., DOUTI P.Y., FAURE G „ DJAGNI K., TREKOU K., 1989. Rapport annuel 1988. E xp é rim e n ta tio n h e rb ic id e en m ilie u réel. Projet de r e c h e r c h e - d é v e lo p p e m e n t. C IR A D -IR C T , s e c tio n s a g ro n o m ie et a g ro -é c o n o m ie , station d 'A n ié -M o n o , Togo, 21 p.

TONATO S. H., DOUTI P.Y., COUSINIE P., DJAGNI K., 1990. Rapport annuel 1989. Expérimentation herbicide en m ilie u réel. Projet de rech e rc h e-d é v e lo pp e m en t. CIRAD-IRCT, sections a g ro n o m ie et a g ro -é con o m ie , station d'Anié-M ono, Togo, 21 p.

Abstract... Résumé...

Resumen

P. Y. DOUTI — Cotton crops versus weeds: when is the competition period?

In Togo, weed control operations are often late in cotton fields. However, previous studies have shown tha t fields must be well maintained in order to obtain good yields. An experiment carried out fro m 1 9 8 8 to 1 9 9 0 r e v e a le d the c ritical p e rio d w h e n w eed competition is the most serious for cotton crops, i.e. 2 8 -4 2 days after so w in g. These results w ill he lp in d e t e r m in in g w e e d control alternatives to precisely target field control operations.

Keywords: cotton, weed, competition, weeding, Togo.

P. Y. DOUTI — Cotonnier contre mauvaises herbes : quelle est la période de concurrence ?

Au Togo, les premières interventions de sarclage en culture cotonnière paysanne sont souvent tardives. Or, des t ra v a u x an térieurs ont montré que l'entretien est le premier facteur limitant pour l'obtention de bons rendements. Une expérimentation conduite de 1 9 8 8 à 1990 a permis de déterm iner la période critique au cours de laquelle la

Echinochloa colona concurrence des mauvaises herbes est la plus défavorable. Cette pério­ de se situe entre 28 et 4 2 ¡ours après le semis. La détermination de la période critique permet d'identifier des alternatives d'entretien pour un meilleur ciblage des interventions au champ.

Mots-clés : cotonnier, mauvaise herbe, concurrence, désherbage, Togo.

Photo T. Le Bourgeois

P.- Y. DOUTI — Algodón contra malezas: ¿Cuál es el periodo de competencia?

En Togo, las p rim e ra s intervenciones de escarda en el cultivo algodonero campesino suelen ser tardías. Ahora bien, unos trabajos anteriores demostraron que el mantenim iento es el primer factor limitativo para la obtención de buenos rendimientos. Una experiencia llevada a cabo de 198 8 a 1 9 9 0 permitió determinar el periodo crítico d u r a n t e el cual la c o m p e te n c ia de las m a le z a s es la más desfavorable. Dicho periodo se sitúa entre 2 8 y 4 2 días después de la s ie m b ra . Esta d e te rm in a c ió n del pe riod o crítico p e rm ite pues e n c o n t ra r a lt e r n a t iv a s de m a n t e n i m ie n t o p a ra d e f in ir más precisamente las intervenciones en el campo.

Palabras clave: algodón, maleza, competencia, escardadura, Togo.