Echinochloa colona (junglerice)

E. colona is a cosmopolitan weed common in crops (mainly rice, maize and vegetables), gardens, roadsides, disturbed sites, waste areas and pastures. It also grows along waterways, on the margins of lakes and ponds, in swamps and wetlands, and in other damp habitats. It has the potential to invade natural areas and completely outcompete native vegetation. In Australia, the USA, South and Central America, it is ranked among the top environmental weeds (USDA-NRCS, 2014). In Australia, this species has invaded wetter habitats, including endangered swamp tea tree (Melaleuca tamariscina subsp. irbyana) thickets (Queensland Department of Primary Industries and Fisheries, 2011). 

The generic name is derived from the Greek echinos, hedgehog, and chloa, grass, in reference to the spikelets in many species that are covered with hard bristles. The specific epithet comes from the Latin colonus, farmer; it is commonly misspelled colonum. Michael (1981; 2009) noted that the correct spelling of the specific epithet is colona not colonum because the name was derived from the non-classical Latin adjective colonus-a-um. The adjective colona having the feminine ending '-a' must be accepted when combined with Echinochloa.

 

Annual, with fibrous, rather shallow roots. Culms stout, usually reddish-purple, erect, ascending or decumbent, often branching from the base, often rooting at the lower nodes, 20-60 cm tall, sometimes nodes conspicuously swollen and usually geniculate, compressed, lower internodes often exposed. Sheath 3-7 cm long, compressed, keeled, glabrous, ligule absent; leaf blades light green, sometimes with transverse purple bands, flat, glabrous, elongate, 4-10 cm long, 3-8 mm wide, margins occasionally scabrous, apex acute. Panicle erect or nodding, green or purple-tinged, 5-15 cm long. Racemes numerous, 2-4 cm long, spreading, ascending, sometimes branched, the lower ones up to 1 cm apart, the upper ones crowded.

Spikelets green tinged with purple, crowded, arranged in ca 4 rows, about 3 mm long, rarely with a short point up to 1 mm long. First glume, 1.2-1.5 mm long, 3-nerved, nearly half as long as the spikelet; second glume, 2.5-3 mm long, 7-nerved; the first lemma is similar to the second glume, first palea ovate, ca 2 mm long, glabrous; second lemma, broadly ovate, ca 2 mm long, glossy. Caryopsis whitish, broadly ovate, 1.7- 2 mm long, flat on one side, convex on the other (Wagner et al., 1999).

E. colona is smaller, branches more at the base and has a more spreading or open type of growth than E. crus-galli (Williams, 1956a).

Seedlings have rolled leaves with pointed tips. The blades and sheaths are usually, but not always, green. There are no auricles or ligules and stems are circular in cross-section. The lowermost leaf sheath has a few hairs but most other leaf sheaths are smooth. The usually flaccid leaf blade has faint striations, a white midrib and smooth margins, at least in the upper part. Young plants have erect leaves thickened at the base and culms are sometimes flat and spreading (Zimdahl et al., 1989).

The absence of a ligule, the purplish-tinged leaves and the neatly 4-rowed racemes are characteristic of E. colona.

E. colona is considered native to tropical and subtropical Asia, but its origin still remains uncertain. It is now widespread throughout the tropical and subtropical regions of the world. In the warm regions of Asia, Africa and Australia, it is widely distributed and occurs most commonly at low altitudes but it does extend to 2000 m above sea level (Holm et al., 1977; Lazarides, 1980). It is predominant on damp, fertile, heavy-textured soils in areas which are seasonally rather than permanently flooded (Lazarides, 1980). Tadulingam and Venkatanarayana (1985) reported that E. colona invariably occurs in rich soils. In the USA, it grows from Virginia to Missouri, south to Florida and Texas and in south-eastern California (Hitchcock, 1950).

In Kenya, E. colona is often the dominant weed of rice. It occurs in cotton and other crops, especially under irrigated conditions (Michieka, 1991).

It prefers sunny or lightly-shaded places, moist or soggy but not long-inundated soils; when the soil desiccates it persists in moister places. It is often abundant in Indonesia (Soerjani et al., 1987).

In Malaysia, it is common in dryland field crops, wetland rice, plantation crops and vegetables; in open damp sandy or clay soils. It is among the three most serious weeds in jute, groundnuts, rape and vegetables in many countries (Barnes and Chan, 1990).

It is an annual grass in upland and paddy fields of the southernmost parts of Japan (Morita, 1997).

E. colona ranked among the 10 most important weed species in 251 fields in the dry and intermediate rice-growing zones of Sri Lanka in 1994-95 (van Mele et al., 1997).

In western Polynesia, E. colona occurs occasionally to commonly in waste places and in crops such as banana, cassava and vegetables, particularly in wet areas (Whistler, 1983).

It is one of the five most important weeds in the eastern plains of Colombia (Bastidas Lopez, 1996).

Determining the date of introduction of E. colona is very difficult mainly because its native distribution range remains unclear and it has been widely cultivated as fodder and forage crop in many tropical and subtropical regions of the world. In the West Indies, E. colona was first reported in 1814 in Cuba. By the 1880s, this species is listed as a common weed in St Croix Island, St Thomas, Puerto Rico, Jamaica, and Cuba (US National Herbarium). 

E. colona propagates primarily by seed. However, the lower creeping part of the plant has the capacity to regenerate and multiply through cut portions (Arora et al., 1976). When nodes come in contact with the soil, rooting occurs and new shoots develop. These, when separated from the mother plant, can give rise to independent plants (Sen, 1981). Chun and Moody (1986) reported that the lower nodes in a flowering stalk were always able to produce adventitious roots. The ability to produce adventitious roots was greater in a younger stalk than in an older stalk.

E. colona seeds are disseminated by wind (Sen, 1981), irrigation water (Kaul, 1986) and animals (Majumder, 1962). Man, in the process of cutting and hauling grass for animal feed, may disperse the seeds over some distance (Sen, 1981). In Costa Rica, E. colona, together with other weeds, such as Oryza rufipogon, O. latifolia, Cyperus iria and several broadleaved weeds, are carefully controlled in rice fields, but reproduce heavily in irrigation canals, shedding their seeds in irrigation water and, thus, re-infesting commercial fields (Rojas and Aguero, 1996).

Many weeds not native to an area are present as impurities in seed. It is suspected that the spread of E. colona in many rice-growing areas of Malaysia is due to contaminated rice seed (Baki, 1981). Others reporting dissemination of E. colona as a contaminant of crop seed include Subba Rao and Prasad (1972), Disthaporn et al. (1998) and Rao and Moody (1990).    

E. colona is a cosmopolitan weed that usually grows in cultivated areas, waste grounds, ditches and fields. Holm et al. (1977) reported that E. colona is associated with 35 crops in more than 60 countries but it has never been reported in temperate cereals, fruits or vegetables. It is one of the most troublesome tropical annual weeds and forms an important association with rice as well as with other crops in many countries. According to Holm et al. (1977), it is the second most important weed of rice.

Waterhouse (1993) listed E. colona as a major weed in beans, cassava, cempedak (Artocarpus polyphema), cocoa, coconut, coffee, groundnut, maize, rice, sapodilla (Achras sapota), sorghum, soyabeans, sugarcane, tobacco and tomato.

E. colona is the most commonly reported weed of rice in South and South-East Asia (Moody, 1988). It is the most important weed in upland rice in Latin America (Gonzalez et al., 1983). In West Africa, it grows in a wide range of soil moisture conditions from swampy or hydromorphic soils to dry land (Akobundu and Agyakwa, 1987). In India and Pakistan, E. crus-galli and E. colona are the most important annual weeds in summer (kharif) crops (Shad and Siddiqui, 1996). It is one of the most commonly reported weeds of rice in Nepal (Moody, 1986) and one of the four most commonly reported weeds of grain legumes in the Philippines (Moody, 1984).

As well as being an agricultural weed, it grows along waterways, on the margins of lakes and ponds, in swamps and wetlands, and in other damp habitats in Australia (Queensland Department of Primary Industries and Fisheries, 2011).

There are numerous records in the literature which associate pests with alternative weed hosts.

E. colona has been reported as a host plant of a series of viruses, among them Maize sterile stunt virus [Barley yellow striate mosaic virus] (Greber, 1984) and Maize yellow stripe virus (Mahmoud et al., 1996). In Aragua state, Venezuela, Nava et al. (1996) found Potato virus X (PVX) and a geminivirus, which was detected by nucleic acid hybridization, in E. colona.

Root bugs, Blissus spp., mainly attack plants from the Poaceae and species of Blissus were associated with the weeds E. colona and Digitaria sanguinalis (Leon Tamayo 1997). Vasquez and Sanchez (1991) investigated the life cycle of B. leucopterus under laboratory and greenhouse conditions and found that E. colona was the food plant most favoured by the lygaeid.

In Cuba, Gutiérrez et al. (1991) described aspects of the biology and population dynamics of the pentatomid rice pest, Oebalus insularis, which uses E. colona as a host plant. Madhusudhan and Gopalan (1989) studied the biology of the rice pest, Stenchaetothrips biformis, on different ages of cv. TN 1 seedlings and E. colona; the life cycle was completed on both hosts. In India, the biology of the coreid rice pest Leptocorisa oratorius was studied on E. colona in a greenhouse by Dhuri and Mazagaonkar (1986) and on rice by Shah (1989). E. crus-galli and E. colona served as important alternative food plants. The presence of these weeds in rice fields during fallow periods allowed the pest to survive.

Shrivastava and Mathur (1986) studied the host specificity and biology of Nephotettix virescens and N. nigropictus on 14 host plants. Both species completed their life cycles on E. colona. Intensive cultivation of semi-dwarf rice cultivars has made the overlapping of two crops of rice a common practice. Such circumstances might cause greater survival and propagation of tungro virus in these rice cultivars, their stubble and in wild rices. Some weed hosts such as E. colona might also act as link hosts between two crops, especially in places where a single crop is grown per year. Rice is a preferred host for N. virescens over weeds, whereas N. nigropictus prefers weeds to rice. Thus in areas where cropping systems overlap, N. virescens might play a greater role in tungro perpetuation. In single cropping areas, N. nigropictus might transmit tungro virus among weeds or wild rices in the off-season (Anjaneyulu et al., 1981).

Israel and Prakasa Rao (1968) reported that weeds such as E. colona, E. crus-galli, Paspalum distichum, Brachiaria ramosa and Cynodon dactylon growing in and around rice fields serve as foci of infestation for insect pests such as gall midge, rice hispa, rice leaf miners, plant hoppers and rice bugs. The first showers at the start of the monsoon season bring about the growth of weeds, which serve as suitable hosts for these insect pests which later migrate into rice seedling nurseries or the standing crop at vulnerable stages. Destruction of the weeds will help to minimize the population of these insects.

Sugimoto and Nugaliyadde (1996) reported that nymphal development of Leptocorisa oratorius takes place in paddy fields before rice flowering, when the fields were infested with adults whose ovaries had already matured by feeding on panicles of weeds (E. crus-galli, E. colona and Digitaria ciliaris) or early-flowering rice plants. Therefore, weeds should be controlled to eliminate panicle-bearing weeds in and around paddy fields before the onset of rice flowering to prevent advanced ovarian maturation in the adult pest population, which migrate to paddy fields to oviposit.

In contrast, Litsinger et al. (1987) noted that weeding during the first month after upland rice establishment forced insect pests such as armyworms, which prefer grassy weeds, on to the rice. Soil insects will move to the rice crop and clean culture also may force termites to attack rice. Pantua and Litsinger (1988) reported that yellow stem borer moths spent more time in grassy areas than in the rice field during the day.

Balal et al. (1988) noted that weed-free rice fields usually have more stemborer-infected rice plants than weedy fields since the grassy weeds attract stemborers. Ostrinia nubilalis attacked E. crus-galli and E. colona.

Yik and Birchfield (1977) determined the life cycle of the nematode Meloidogyne graminicola from infective second-stage larvae to mature egg-laying females as 16 days on E. colona under laboratory and greenhouse conditions at 26°C. Large galls were formed on E. colona seedlings 4 days after being sown in infested soil. Soomro (1989) reported that significantly more nematodes (M. graminicola) invaded the roots of E. colona than of rice, yet the roots of the weed were larger than uninfested rice roots, thus proving to be more tolerant and competent compared to rice. Due to invasion by the nematode, the root growth of both hosts was reduced, the morphology of the roots altered and the laterals grew in clusters.

Treca (1985) noted that the seeds of weeds such as E. colona and E. stagnina attracted birds to rice crops. Control of weeds such as E. crus-galli and E. colona indirectly controlled birds and the spread of grain-sucking insects such as Leptocorisa sp., Pachybrachius pallicornis and Nezara viridula (Anon., 1981).

E. colona has been reported to be an alternate host of a number of diseases, insect pests and nematodes.

Allelopathy

There is some evidence for allelopathic effects of leachates or extracts of E. colona on rice, Vigna mungo (Porwal and Mundra, 1992) and sorghum (Dharmaraj et al., 1988). In the latter study, root leachates from E. colona caused less reduction in germination than several other weed species, 44 and 52% germination for shoot and root leachates, respectively. These differences appeared to be correlated with levels of phenolics. Almodóvar-Vega et al. (1988a, b) reported that E. colona caused adverse allelopathic effects on pumpkin and decreased plant growth of aubergine seedlings through decreased stem length and dry weight.

Conversely, there is evidence for allelopathic effects of some crops on E. colona. Germination, root length and fresh weight were reduced by acidic and neutral fractions of root exudates of sunflower (Park et al., 1992). Population and biomass were reduced 40 and 30%, respectively, by rice residues, and linseed residues reduced germination (Khan and Vaishya, 1992). Germination and seedling growth were inhibited by stem, leaf and root extracts of sorghum (Kim et al., 1993). Incorporation of plant stubble from seven pigeon pea cultivars reduced weed densities in the subsequent tomato and pepper crops grown at Juana Diaz, Puerto Rico, but not at Lajas. Greatest reduction in density was from 219 to 78 weeds per 0.5 m² with Line 84. The weed species suppressed by pigeon pea in the tomato crop included E. colona. Semidey and Bosques-Vega (1999) concluded that weed suppression was due to allelopathy.

Aqueous extracts from Lantara camara twigs inhibited, to various degrees, the germination and seedling growth of rice-associated weeds, including E. colona (Bansal, 1998). However, extracts from Blumea lacera (Oudhia et al., 1998) and Datura stramonium (Oudhia and Tripathi, 1998) had no detrimental effects on E. colona.

E. colona has been reported as a host plant of a series of viruses, among them Maize sterile stunt virus [Barley yellow striate mosaic virus] (Greber, 1984).

Yik and Birchfield (1977) determined that the life cycle of the nematode Meloidogyne graminicola from infective second-stage larvae to mature egg-laying females, was 16 days on E. colona under laboratory conditions at 26°C. Large galls were formed on E. colona seedlings 4 days after sowing in infested soil.

Vasquez and Sanchez (1991) investigated the life cycle of Blissus leucopterus under laboratory and greenhouse conditions in Colombia and found that E. colona was the food plant most favoured by the lygaeid.

In Cuba, Gutierrez et al. (1991) described aspects of the biology and population dynamics of the pentatomid rice pest Oebalus insularis, which uses E. colona as a host plant. Madhusudhan and Gopalan (1989) studied the biology of the rice pest Stenchaetothrips biformis on different ages of variety TN1 seedlings and the weed E. colona; the life cycle was completed on both hosts. The biology of the coreid rice pest Leptocorisa oratorius on E. colona in a greenhouse in India was studied by Dhuri and Mazagaonkar (1986) and by Shah (1989) on rice in India. Studies confirmed that E. crus-galli and E. colona served as important alternative food plants. The presence of these weeds in rice fields during fallow periods allowed the pest to survive.

Shrivastava and Mathur (1986) studied the host specificity and biology of Nephotettix virescens and N. nigropictus on 14 host plants. Both species completed their life cycle on E. colona.

E. colona causes substantial yield reductions because of its severe infestations, rapid growth and great competitive ability. Competition between E. colona and crops has been studied by a number of authors and they have demonstrated that E. colona is a strong competitor for nutrients and water. E. colona has been listed as a common weed in rice, maize (Kapoor and Ramakrishnan, 1975), cotton (Panwar and Malik, 1991), mung-bean (Kumar and Kairon, 1990), bush-bean (Phaseolus vulgaris), pole-bean (Premalal et al., 1998), taro (Colocasia esculenta var. esculenta; Lugo et al., 1998), and castor-bean (Munoz et al., 1988) among others.

Fischer et al. (1997) reported that rice cultivars differed in their competitiveness against E. colona. Average yield losses ranged from 27 to 62% under saturating E. colona infestations of up to 5.9 t DM/ha. Leaf area index, tiller number, and canopy light interception recorded in competition, and not much before 40 days after emergence, correlated positively with rice competitiveness. Ni et al. (2000) found that biomass at tillering was the best predictor of rice competitiveness against weeds.

In dry seeded rice, weed seeds and crop seeds germinate at about the same time and the weeds, being more vigorous, almost smother the crop. Of all the weeds emerging during the upland period, E. colona is the most dominant with weed emergence continuing for 30-35 days after sowing. Mukhopadhyay et al. (1972) reported that grasses made up 85-89% of the total weed population and 90-96% of the total dry matter in unweeded dry seeded rice. E. colona, the predominant grass species, reduced yields by 74-98%.

Suriapermana (1977) reported that season-long competition of E. colona with transplanted rice cv. IR34 caused 43% yield reduction compared with 31% loss with competition from Monochoria vaginalis and 55% loss in the unweeded check plot where all the test weed species and the natural weed population competed simultaneously against rice.

In 1994 and 1995, Roldan (1995) conducted an experiment to determine the effect of different populations of E. colona - the commonest grass weed - on the yield of dry seeded rice cv. PSBR C16. Yield from the hand-weeded control (weeded at 20, 33 and 45 days after crop emergence) was 3.3 t/ha in 1994 and 2.3 t/ha in 1995. In 1994, yield losses ranged from 15% when there was season-long competition from five E. colona plants/m² to 36% with competition from 40 E. colona plants/m². Losses in 1995 were 3 and 30%, respectively, for the same densities. The natural population of E. colona (50 plants/m² in 1994, 101 plants/m² in 1995) caused 30% yield loss in 1994 and 48% yield loss in 1995 compared to 49 and 83%, respectively, for the unweeded control. In Bulacan, Philippines, Mercado and Talatala (1977) reported that a natural population (280 plants/m²) of E. colona reduced dry seeded rice yield by 76%.

In pot experiments, seed cotton yield based on the average of weed densities (5, 10 and 20 plants/pot) was reduced 74% by E. colona. The weeds, which reduced significantly the shoot nitrogen and potassium contents of cotton, and accumulated twice as much of these nutrients had no marked effect on phosphorus. Cotton leaf area index was reduced by 75%. Intra- and inter-specific interferences decreased the weed-specific detrimental effect on cotton. Comparable seed yield losses in cotton were obtained with five E. colona per pot. Cotton produced the maximum number of bolls and weights of stalks and bolls when kept free from weeds for approximately 5 weeks after sowing (Guantes and Mercado, 1975).

Bush bean (Phaseolus vulgaris cv. Top Crop) yield was reduced 48, 37 and 53% by root, shoot and full competition of E. colona, respectively. Full competition reduced the biomass of E. colona by over 50%. Pole bean (cv. Kentucky Wonder Green) yield was not affected by weed competition (Premalal et al., 1998).

E. colona grows along waterways, on the margins of lakes and ponds, in swamps and wetlands, and in other damp habitats. It is regarded as an environmental weed in parts of Queensland, the Northern Territory and Victoria (Queensland Department of Primary Industries and Fisheries, 2011).

In south-eastern Queensland E. colona is ranked among the top 200 environmental weeds and has invaded wetter habitats, including the critically endangered swamp tea tree (Melaleuca irbyana, formerly Melaleuca tamariscina subsp. irbyana) thickets. It is also a common weed of arid wetland areas in central Australia (i.e. in the Northern Territory and South Australia).

E. colona produces a highly palatable fodder that is relished by animals and is considered to be one of the finest fodder grasses (Cope, 1982). It can be cut three to four times during the wet season when flower heads are in full bloom. The common practice in many countries in South-East Asia is to remove whole plants out of rice fields, where it is regarded as a weed. The collected material is then fed to ruminants. Farmers in Myanmar consider E. colona to be the best grass for feeding to milking animals and will go long distances to obtain it (McKerral, 1923).

Sen and Mabey (1966) found that the quality of E. colona herbage is high. The crude protein content decreased from 13.8% at 4 weeks to 10.3% at 10 weeks. The crude fibre content was low, 22.6 and 25.6% at 4 and 12 weeks, respectively. The contents of phosphorus and calcium were adequate for cattle requirements.

During famines, humans eat the seeds (Rhind, 1945; Lazarides, 1980). Some forms of this variable plant are cultivated in tropical Africa and Asia for the grain, which is used for food (Wagner et al., 1999). Sawa (E. colona) is grown in India, Nepal and Sikkim. Primitive races of sawa grown in India differ from wild E. colona only in the tardy disarticulation of their spikelets at maturity (de Wet, 1992).

In Java, young shoots of barnyard millets are eaten as a vegetable. Also, the seed of E. colona is used as a feed for caged birds (Partohardjono and Jansen, 1996). In Rajasthan, India, the seeds are boiled in water and used as a substitute for rice. The seeds are also ground into flour, sometimes being mixed with maize or blackgram, and made into bread (Gammie, 1902) or porridge (Abdelmuti, 1991).

Chemical Control

According to Lorenzi (1986) the following herbicides and combinations give good control of E. colona under different crop situations: acetochlor, asulam, asulam + diuron, atrazine + metolachlor, atrazine + simazine, bromacil, butachlor, cyanazine + MSMA, dalapon, diuron, diuron + MSMA, diuron + paraquat, 2,4-D + glyphosate, EPTC, haloxyfop-methyl, clethodim, imazaquin, metribuzin, oryzalin, pendimethalin, sethoxydim and vernolate. Ampong-Nyarko and De Datta (1991) summarizing herbicide use in rice, confirmed E. colona susceptibility to butachlor, butralin, cinmethylin, chlomethoxyfen, fenoxaprop, glyphosate, molinate, oxadiazon, oxyfluorfen, paraquat, pendimethalin and thiobencarb but indicated no control with bentazone, 2,4-D and dimethametryn.

Other herbicides reported to control E. colona in rice are cyhalofop butyl, fenoxaprop, fluazifop, oxadiazon (Galinato et al., 1999), acifluorfen, propanil, thiobencarb + propanil (Estorninos and Moody, 1988), anilofos + 2,4-D (Avudaithai and Veerabadran, 2000), bispyribac sodium (Chin, 1999; Palis et al., 1999a), clefoxydim (Schöfl et al., 1999), fentrazamide + propanil (Fursch, 1999), flufenacet (Palis et al., 1999b), oxadiargyl (Dario and Gallo, 1999), pretilachlor (Angiras and Rana, 1998) and quinclorac (Lopez Martinez et al., 1999).

Fischer et al. (1993b), after economic analysis of treatments, showed that under heavy weed infestations, in a system where rice cannot be flooded early and continuously, three postemergence herbicide applications (propanil + butachlor + pendimethalin applied at 9 days after emergence, thiobencarb + propanil + picloram +2,4-D at 18 days after emergence, and fenoxaprop-ethyl + bentazone at 44 days after emergence) followed by spot applications of paraquat were justified.

Mukhopadhyay et al. (1972) applied the following treatments in rice cv. IR8, sown as an upland crop: (a) hand-weeded three times, (b) propanil applied, (c) as (b) + one hand weeding, (d) nitrofen at a lower rate, (e) nitrofen at a higher rate and (f) as (d) + one hand weeding. Grasses, especially E. colona, which comprised 80-89% of weed populations were effectively controlled by herbicide treatments, especially by (d), (c) and (f).

In Raipur, India, Kolhe and Tripathi (1998) found that preemergence applications of anilofos and thiobencarb each supplemented by one hand weeding at 35 days after sowing were comparable to hand weeding twice in reducing weed dry matter from 1241-1317 kg/ha in the unweeded control; to 124-202 and 193-241 kg/ha, respectively, and enhancing rice yields from 9-11 q/ha in the unweeded control, to 32-38 q/ha and 33-37 q/ha, respectively. Preemergence application of anilofos or thiobencarb followed by postemergence application of cyhalofop-butyl was more effective than 2,4-D.

Saini and Angiras (1998) conducted a field experiment during the kharif seasons of 1994 and 1995 in Palampur, India, in maize containing a weed flora dominated by E. colona, E. crus-galli, Cyperus iria, C. esculentus, Commelina benghalensis and Ageratum conyzoides. Grain yield in the unweeded control was 67% lower than that of the most effective treatment of atrazine. In Turkey, Uremis et al. (2000) reported that the control range of metolachlor + atrazine was broader than that of pendimethalin for weed control in maize.

Bohorquez and Salazar (1975) reported that in the coastal areas of Ecuador, weed competition accounts for losses of 20-40% of cotton yields. Excellent control of the major weeds (E. colona, Leptochloa filiformis [L. mucronata] and Portulaca oleracea) was obtained with preemergence applications of fluometuron + prometryn, followed by slashing 55 or 65 days after sowing. Yields were superior to those from mechanically weeded controls. Herbicide costs were offset by the increased yields. Haloxyfop-methyl effectively controlled E. colona in cotton (Panwar, 1991). Herbicide activity was improved by the addition of surfactant. Best yields were obtained with haloxyfop-methyl + pendimethalin with or without one hoeing 45 days after sowing.

Punia and Malik (1999) reported that a mixture of trifluralin and pendimethalin plus one hoeing 30 days after sowing was more effective in controlling E. colona in groundnut than herbicide alone. Other herbicides reported to give good weed control in groundnut are oxyfluorfen, fluchloralin (Patel et al., 1997b), haloxyfop-R-methyl ester and fluazifop-butyl (Lueang-a-papong and Singhara Na Ayudhaya, 1999).

Herbicides suggested for control of E. colona in sugarcane include diuron, dalapon, trifluralin, paraquat, metribuzin, ametryn + metribuzin and diuron + metribuzin (Diaz et al., 2000). Fuentes et al. (2000) reported that amicarbazone controlled weeds, including E. colona, slightly better than diuron when applied preemergence and gave equivalent control when applied postemergence.

In India, Balyan and Malik (1998) achieved 80% control of E. colona in soyabeans with a pre-plant application of trifluralin. In the kharif season of 1996, the most effective level of control (88.9%) of a broad-spectrum of weeds, including E. colona, was achieved following a tank mix application of fenoxaprop-P + lactofen (Kolhe et al., 1998). In the rainy seasons of 1994 and 1995 in Haryana, India, fluazifop provided an average of 95% E. colona control. Chlorimuron failed to provide any control of E. colona. Chlorimuron antagonized E. colona control by fluazifop when the two herbicides were applied in a tank mix combination (Balyan and Pahwa, 1998).

Noldin et al. (1998) conducted field studies in Texas, USA, from 1992 to 1994 to evaluate herbicides for red rice (Oryza sativa) and Echinochloa spp. (primarily E. crus-galli and E. colona) control in soyabeans. Early-season Echinochloa spp. control with trifluralin, pendimethalin and pendimethalin + imazaquin applied preplant incorporated; metolachlor, dimethenamid and metolachlor + imazaquin applied preplant incorporated or preemergence; imazapic + imazaquin, and imazapic + imazethapyr applied preemergence; and sethoxydim and quizalofop-P applied postemergence was 90 to 100% in at least 2 of 3 years. However, Echinochloa spp. control decreased for all treatments later in the season. Pendimethalin applied preplant or in mixture with imazaquin injured soyabeans by 14 to 34% in 2 years. Trifluralin preplant incorporated, dimethenamid preplant incorporated or pre-emergence, imazaquin preplant incorporated, metolachlor + imazaquin preplant incorporated or preemergence, and imazapic + imazethapyr injured soyabeans 12 to 41% in at least 1 of 3 years.

Sequential application of a reduced rate of soil-applied trifluralin with postemergence fluazifop or a reduced rate of soil-applied trifluralin or pendimethalin followed by hand hoeing at 35 days after sowing provided better control of a broad spectrum of weeds in soyabeans than a single application of a postemergence herbicide applied at reduced or recommended rates (Chhokar and Balyan, 1999).

Quizalofop-ethyl was most effective in controlling grass weeds, including E. colona, in soyabeans (Singh and Tomar, 2000) but oxyflurofen caused phytotoxicity to germinating soyabean plants resulting in a significant reduction in grain yield (Singh et al., 1996).

In Ethiopia, Zewdie and Tanner (1999) found that pendimethalin + bromoxynil + MCPA was the most effective treatment for controlling both grass (Sorghum arundinaceum and E. colona) and broadleaf weeds in wheat and resulted in a significant increase in grain yield from 0.83 t/ha in the unweeded control to 2.97 t/ha.

In Kharagpur, India, during the winter seasons of 1993-95, Choubey et al. (1998) found that there was a significant increase in wheat yields with an increase in the frequency of irrigation from 0.6 to 0.8 IW/CPE. Hand weeding and preemergence application of pendimethalin were superior to the mechanical weeding and unweeded treatments with respect to crop yield. The predominant weeds included Enydra fluctuans, Physalis minima, Digitaria sanguinalis and E. colona. With increase in irrigation levels from 0.6 to 1.0 IW/CPE ratio, there was significant increase in the population and dry matter of weeds. Hand weeding and pendimethalin were equally effective in minimizing the population and dry matter production of weeds.

In jute, where the dominant weeds were E. colona, Paspalum distichum, Eclipta prostrata, Ageratum conyzoides, Commelina benghalensis, Sida cordifolia, Cyperus iria and Cyperus rotundus, application of fluchloralin 3 days before sowing + one hand weeding at 35 days after sowing produced lowest weed dry matter, highest mean fibre yield of 2.64 t/ha and highest net return of the herbicide treatments. While the hand weeding treatment produced the lowest weed dry matter and the highest fibre yield (2.87 t), the net return was markedly lower than most of the herbicide treatments (Mishra and Bhol, 1996).

Lugo et al. (1998) reported that oxyfluorfen gave 87-96% control of weeds, including E. colona, in tannia (Xanthosoma sp.); ametryn gave 83-89% control. Yields were higher when oxyfluorfen was used.

Oxyfluorfen and sethoxydim gave 83 and 97% control of grasses, including E. colona, in cilantro (Coriandrum sativum), while oxyfluorfen gave 82% control of grasses in spiny coriander (Eryngium foetidum). However, yields of both crops were highest in the hand-weeded plots (Lugo and Santiago, 1996).

For clusterbean (Cyamopsis tetragonoloba), yields obtained with fluchloralin were similar to those for the weed-free check and superior to pendimethalin (Bhadoria et al., 1996).

Pendimethalin was effective in controlling weeds, including E. colona, in onion (Allium cepa) (Verma and Singh, 1997).

Fluchloralin + one hand weeding effectively controlled weeds in blackgram and significantly increased number of pods/plant, seeds/pod, grain yield and economic returns (Yadav et al., 1997).

Efficacy of the nonselective herbicide, glyphosate, in controlling E. colona was greatest with well-watered (100% of field capacity) plants that were placed under cool (20/25°C) and humid (92% RH) conditions. Efficacy was least when applied to plants under severe water stress (29% of field capacity) that were placed under hot (35/30°C) and less humid (65% RH) conditions. The level of irradiance did not alter efficacy (Tanpipat et al., 1997).

Herbicide Resistance

Biotypes of E. colona vary in their susceptibility to herbicides. Variation in the susceptibility of E. colona biotypes to butachlor and thiobencarb has been observed in the Philippines (International Rice Research Institute, 1986). By repeated use of an herbicide, a resistant population may develop.

Repeated use of propanil for weed control in rice has led to the evolution of propanil-resistant E. colona biotypes. Valverde (1996) reported that propanil resistance occurred in all Central American countries and in Colombia, the most resistant populations requiring over 50 times the normal rate for control. E. colona resistance to propanil has also been reported from Venezuela (Ortiz et al., 2000) while there are indications of resistance in Senegal (Haefele et al., 2000). Resistance has been attributed to a difference in herbicide metabolism between resistant and susceptible biotypes related to aryl acylamidase levels and enzymatic activity. Resistant biotypes have elevated levels of aryl acylamidase, which enable them to metabolize propanil to 3,4-dichloroaniline at a much higher rate than susceptible biotypes. Leah et al. (1994) observed that E. colona resistance to propanil, expressed as aryl acylamidase activity, increased as growth stage advanced.

Fischer et al. (1993a) found that the propanil-resistant biotypes of E. colona had greater leaf area and dry matter accumulation and were taller than a susceptible biotype but Garro et al. (1992) reported that there were no significant differences in plant height, leaf number, inflorescence number and tiller number per plant between tolerant and susceptible plants.

As a result of resistance to propanil, late postemergence applications of fenoxaprop, an acetyl CoA carboxylase inhibitor, have been used to control E. colona. However, in Costa Rica populations of E. colona resistant to fenoxaprop have developed after three seasons (Riches et al., 1996). Hoagland et al. (2004) discuss the propanil resistance associated with elevated levels of aryl-acylamidase, and the potential of compounds which act as inhibitors of aryl-acylamidase to be applied as synergists with propanil to increase herbicidal activity.

In Trinidad and Tobago, Bridgemohan and Bridgemohan (2014) report the presence of an E. colona biotype resistant to herbicides, but not necessarily to fenoxaprop-P. In Iran, Elahifard et al. (2013) report the development of biotypes resistant to ametryn and metribuzin.

Several methods have been developed to control herbicide-resistant E. colona populations. Key components of resistance management are the use of mixtures and sequences of herbicides with different modes of action and contrasting chemistry (Jutsum and Graham, 1995). Riches et al. (1997) reported that propanil performance was improved by the addition of low rates of pendimethalin. Field studies during 1994-96 in dry-seeded upland rice in southern Costa Rica compared propanil + fenoxaprop-P-ethyl with pendimethalin. Pendimethalin provided good control and resulted in reduced weed control costs.

Caseley et al. (1996) reported that either aryl acylamidase or cytochrome P450 inhibitors could be used to synergize propanil phytotoxicity against resistant biotypes of E. colona but mixtures with both types of synergist were most potent. Mixtures of propanil with carbamate (for example, carbaryl) or organophosphate insecticides reduce selectivity in rice by inhibiting aryl acylamidase activity. These insecticides synergized propanil activity against resistant biotypes of E. colona. Selectivity in rice and synergy against resistant biotypes have been achieved by mixing propanil with the herbicide piperophos.

Glasshouse studies with pot-grown plants indicated that propanil- and fenoxaprop-P-resistant biotypes could be controlled selectively in rice with quinclorac, pendimethalin, thiobencarb, bispyribac-sodium, pyriminobac-methyl and pyribenzoxim. Clomazone also provided excellent control of E. colona, but it caused transient damage to rice and its vapour affected some non-target plants. Cyhalofop-butyl was selective in rice and effectively controlled propanil- but not fenoxaprop-P-resistant E. colona (Caseley et al., 1997).

Widderick et al. (2013) report that in 2007, the world’s first population of glyphosate-resistant E. colona was confirmed in Australia. Since then, over 70 populations have been confirmed as resistant in the subtropical region. Sequential treatments with different herbicides can help with control. Widderick et al. say that sequential application of glyphosate followed by paraquat provided 96-100% control across all experiments, irrespective of the growth stage, and the addition of metolachlor and metolachlor+atrazine to glyphosate or paraquat significantly reduced subsequent emergence. Herbicide treatments have been identified that provide excellent control of small E. colona plants, irrespective of their glyphosate resistance status. It is recommended that these tactics of knockdown herbicides, sequential applications and pre-emergence herbicides should be incorporated into an integrated weed management strategy in order to greatly improve E. colona control, reduce seed production by the sprayed survivors and to minimize the risk of the further development of glyphosate resistance.

Thornby et al. (2013) discuss ways of delaying evolution of glyphosate resistance in E. colona in cotton farming systems. The most robust approach to delaying or preventing resistance is using non-glyphosate knockdowns and strategic tillage to control glyphosate survivors. Werth et al. (2012) say that presence of a glyphosate resistant grass has the potential to increase weed control costs by Aus$40 to Aus$90/ha/yr as well as increasing fallow costs by up to Aus$70/ha/yr. in Australian cotton. The authors discuss weed management challenges for the Australian cotton industry and explore management options for glyphosate tolerant and resistant weeds. Widderick et al. (2012) examine non-chemical options for managing the seed bank of glyphosate-resistant weeds including E. colona, and report on the impact of different tillage and stubble treatments on emergence and persistence

Rapid and accurate detection of herbicide-resistant weeds is key to resistance management decision making. Kim et al. (2000) developed rapid detection methods for discriminating between resistant and susceptible biotypes of E. colona to propanil and fenoxaprop at various growth stages from seed to flowering.

Cultural Control

E. colona is easy to uproot and destroy in the younger stages. It can be controlled by cultivation and early, deep flooding (Williams, 1956b).

Ghobrial (1981) found that grain yields of rice increased as the number of hand weedings increased. The best yield (4.1 t/ha) was obtained with three to four weedings at 15, 30, 45 and 60 days after rice emergence. Oxadiazon gave good residual weed control of the dominant weed species, including E. colona, for 6 to 7 weeks. The combination with continuous flooding from 6 weeks after rice emergence gave excellent weed control during the entire season and yields comparable to or even better than frequently-weeded rice.

Ghorai and Bera (1998) found that weed biomass in pointed gourd (Trichosanthes dioica) was reduced by 17-69% by mulching and this maintained a low weed status for 8-9 months. Mulch at 16 t/ha increased crop yields by 163-323%, compared to the no-mulch treatment and also increased individual gourd weight significantly. The major weed species observed included Cyperus rotundus, E. colona, Digitaria spp., Cynodon dactylon, Ageratum spp., Chenopodiumalbum and Melilotus alba.

Biological Control

Control of E. colona by inoculation with its pathogens was investigated in the greenhouse in Japan by Tsukamoto et al. (1997). Exserohilum monoceras [Setosphaeria monoceras] exhibited high herbicidal activity against E. colona. In the Philippines, E. monoceras killed seedlings of E. colona without affecting rice (Zhang et al., 1996). When adequate dew was provided, 100% mortality occurred over the broad dew-period temperature range of 20-30°C. The minimum dew period to achieve 100% mortality was 16 hours (Zhang and Watson, 1997a). An inoculum dose of 5.0 x 107 conidia/m² was required to obtain 100% mortality of E. colona seedlings, the 1.5-leaf stage being the most susceptible. Increasing inoculum density increased weed control efficacy on younger or older Echinochloa seedlings (Zhang and Watson, 1997b). Eusebio and Watson (2000) found that mixtures of fungal pathogens were compatible, virulent and synergistic often producing superior control than when they were applied alone. Tosiah et al. (2010) report severe damage of E. colona by applications of Exserohilum monoceras, with maize the only crop tested becoming infected.

Mycoflora reported on E. colona, and potential for biocontrol, are discussed by Karuppiah and Seetharaman (2006) and Karuppiah et al. (2006). Chauhan et al. (2010) discuss the potential of post-dispersal seed predation for reducing weed populations, and suggest that seed predation of species including Digitaria ciliaris and E. colona could contribute to ecologically-based weed management in rice.