O. barthii is a plant of wetlands and a natural weed of lowland rice. It has not so far spread significantly outside its native distribution of western, eastern and southern Africa, but has the potential to do so. It is of concern in Mali...
O. barthii is a plant of wetlands and a natural weed of lowland rice. It has not so far spread significantly outside its native distribution of western, eastern and southern Africa, but has the potential to do so. It is of concern in Mali, where it threatens to spread locally into new rice-growing areas, and is also of concern to the USA, where it could add to the serious existing problems from hybrid ‘red rice’.
The name Oryza barthii was for some time wrongly used for the perennial O. longistaminata, but Clayton (1968) clarified the nomenclature so that it now applies strictly to the annual wild rice of Africa, which has also been known as O. breviligulata. Clayton (1968) also proposed that O. stapfii should be merged into this species. Bardenas and Chang (1966) studied the morphology of a range of samples of O. barthii and ‘O. stapfii’; the latter had apparently been distinguished on the basis of its shorter spikelets (8-9 mm, compared with 9-11 mm in O. barthii), but in all other characteristics O. stapfii could not be separated and they concluded that they should be treated as synonyms. O. barthii does in any case show considerable genetic diversity (e.g. Gezahegn et al., 2010).
The relationships between O. barthii, O. longistaminata and the cultivated species O. glaberrima have been much discussed. Nayar (1968; 2012) has suggested that O. sativa may have been introduced to West Africa much earlier than is usually assumed, and that O. barthii could derive from hybridization between O. sativa and O. glaberrima. However, the predominant assumption is that O. barthii derived from the perennial O. longistaminata, and O. glaberrima in turn from O. barthii (e.g. Khush, 1997; Li ZhiMing et al., 2011). All these species have the same chromosome number and AA genome.
Occasionally referred to as a perennial, but O. barthii generally behaves as an annual aquatic grass growing in tufts. Culms 60-120 cm tall (rarely more), 3-8-noded, rather weak, erect or geniculately ascending, producing aerial roots from the lower nodes, terete, spongy, striate, smooth, glabrous. Leaf-sheaths scarious, striate, somewhat tight when young, later loose and usually wrinkled, smooth, glabrous, produced into short auricles at the mouth. Ligule 3-6 (rarely -9) mm long, thinly membraneous, truncate, or rounded. Leaf-laminae 15-45 by 0.4-1.3 cm, linear to very narrowly elliptic (always broadest around the middle), apex acute, intense green, flaccid, glabrous, smooth on the lower, asperulous on the upper surface; midrib not very distinct, pale. Panicle 20-35 x 3-7.5 cm, obdeltoid, rather dense, many-flowered, erect or more rarely somewhat nodding; rhachis stout, angular, sulcate, glabrous, smooth; branches usually erect or obliquely ascending, often appressed to the rhachis, angular, scaberulous. Pedicels 1-6 mm long, stout, striate, scaberulous or smooth, glabrous. Glumes reduced to a tiny 2-lobed rim, remaining at the pedicel. Spikelets 8-10.5 mm long (excluding the awn), up to 3.4 mm wide, deciduous, obliquely inserted on the pedicel, oblong to oblong-semi-elliptic in lateral view, pale green to straw coloured. Sterile lemmas 2.5-4.5 mm long, equal in length or nearly so, similar in shape, lanceolate or narrowly triangular in lateral view, acute, chartaceous-coriaceous, smooth or dorsally asperulous. Fertile lemma slightly shorter than the spikelet, cymbiform, coriaceous, with 2 longitudinal lateral grooves; flanks stiffly hispid to glabrous, inconspicuously tessellate; keels rounded, usually stiffly ciliate mainly towards the apex; apical callus usually dark purplish; awn (6.5)8-16(-19) cm long, stiff, salmon-pink to purplish when fresh, terete or obtusely angular, densely covered with short forwardly directed bristles. Palea about as long as the lemma but much narrower, similar in texture and indument, with the apex drawn out in a short blunt purplish point.
O. barthii is an African species, not reliably known to occur elsewhere. There are some records from Japan (GBIF, 2013), but Chu (1970) refered to ‘samples of O. breviligulata maintained in the National Institute of Genetics’ in Shizuoka Province and it seems likely that these samples were in cultivation only. Within Africa it is especially widespread in West Africa, but is also recorded in most countries of eastern and southern Africa.
The distribution in this summary table is based on all the information available. When several references are cited, they may give conflicting information on the status. Further details may be available for individual references in the Distribution Table Details section which can be selected by going to Generate Report.
The main risk of long-distance introduction comes almost exclusively from contaminated rice crop seed, but this will depend on the volumes of rice exported out of African countries. At present this is negligible, and is likely to remain so for the next decade, according to projections by Wailes and Chavez (2012), with the exception of Egypt, where the weed so far is not recorded. Hence the risk of introduction can be said to be low.
O. barthii grows in stagnant water, slowly flowing water or pools, shallow ponds, seasonally flooded land, rice fields and similar habitats, often forming pure dominant stands but usually scattered with other aquatic grasses, up to 1500 m altitude. It is also found in mopane or savanna woodland, savanna or fadama (Flora Zambesiaca, 2013; Knowledgebank, 2013; PROTA4U, 2013).
Chromosome number 2n = 24 (Lu BaoRong et al., 1997), sharing this and the AA genome with O. sativa, O. glaberrima and O. longistaminata. O. barthii hybridises readily with O. glaberrima (Lu et al., 2003) but not so readily with O. sativa (Aladejana et al., 2008).
O. barthii is an annual species reproducing by seed, which are shed as they mature. It is predominantly inbreeding, with an outcrossing rate of 5–20%. It is a short-day plant (PROTA4U, 2013).
Crossing with O. sativa is often unsuccessful (Chu et al., 1966), so there is not the problem of hybridization with cultivated rice to produce weedy progeny which shed their seeds early, as happens between O. sativa and O. rufipogon. Furthermore, crosses with O. glaberrima tend to show genetic weakness (Chu and Oka, 1972).
Seeds may show dormancy up to 5 months after shedding. In a laboratory experiment, seeds germinated best from a depth of 2.5 cm but two weeks after sowing no differences were observed between seeds in the 0 to 7.5 cm horizon (WARDA, 1979). More than 70% of O. barthii seeds emerged from 15 cm depth (Rijn and Verhagen, 1980). It is not clear whether these results were obtained in moist or in saturated soils. The optimum temperature for germination is 32°C but germination can occur down to 20°C. Simulated moisture stress and hydrogen ion activity adversely affected germination, indicating high sensitivity to moisture stress and solute concentrations (Obadoni et al., 2004).
Physiology and Phenology
Kiran et al. (2013) compared a range of rice species for their photosynthetic capacity and other characteristics. O. barthii was inferior to O. sativa in most respects, especially in carotenoid content.
Seeds of O. barthii were still viable after 12 months burial in a paddy soil (DFID, 2002).
Hamaoka et al. (2103) studied the genetic variations in dry matter production, nitrogen uptake and nitrogen use efficiency in the AA genome of Oryza species grown under different nitrogen conditions and found O. barthii to be less dependent on nitrogen than O. sativa..
O. barthii is clearly dependent on wet conditions for normal development and maturation. It prefers clay or black cotton soils.
Natural dispersal occurs via water flow, especially via irrigation systems. Tuor et al. (2001) reported irrigation water and the soil seed bank as being major sources of infestation, but not nursery beds and seed supplies, which are of only minor importance.
Vector Transmission (Biotic)
Local movement could occur via livestock but there are no documented instances.
Accidental introduction may occur locally by movement in water, livestock or vehicles. Longer-distance movement could occur in contaminated crop seed, but in the absence of a significant export market out of virtually all infested countries, this risk is minimal.
Intentional introduction may occur for research purposes but is unlikely to lead to accidental infestations.
O. barthii was identified as a major weed problem in rice in Ghana, Cote d’Ivoire, Mali and Sierra Leone (DFID, 2002). Other sources note a similar status in Nigeria, Niger and Chad (Gaouna et al., 2011). Although less often cited as a problem in eastern Africa, Johnson et al. (2000) noted that following control of O. longistaminata by puddling and transplanting, ‘the problem of O. barthii remains.’
Dense infestations can cause massive losses of cultivated rice yield. For example, Tuor et al. (2001) reported losses in rice of more than 2 t/ha, or almost 40% caused by O. barthii infestations at densities of 70 tillers/m2 in Mali. Also in Mali, weeded plots of rice yielded 46-98% more paddy than unweeded plots infested with O. barthii (Vallee, 1980). In Senegal, Davies (1984) reported up to 97% yield loss due to O. barthii.
As well as acting as a competitive weed, O. barthii it may cause indirect damage as a reservoir for important rice diseases and pests, such as the African rice gall midge and rice yellow mottle virus (DFID, 2012).
O. barthii can be an important component of natural pastures, as reported by Kulich et al. (1985) from Zambia, and as such has a positive economic impact.
O. barthii has potential as a source of useful genes for transfer to cultivated O. sativa or O. glaberrima. To this end, it was included in genetic studies by Aliyu et al. (2013). Specifically, it has potential as a source of resistance to blast (Pyricularia grisea) (Eizenga et al., 2004), sheath blight (Rhizocyonia solani) (Prasad and Eizenga, 2008), bacterial leaf blight (Xanthomonas oryzae) (Kaushal and Ravi, 1998), the nematode Heterodera sacchari/schactii (Reversat and Destombes, 1998), stem borer attack (AfricaRice, 2013) and rice tungro virus (Jain et al., 1989).
Harlan et al. (1989) recorded that O. barthii is used as famine food in some parts of Africa.
O. barthii provides a food source for waterbuck in Benin (Kassa et al., 2008) and for Kafue lechwe antelope in Zambia (Rees, 1978).
Among Oryza species occurring in Africa, O. brachyantha is distinguished from O. barthii by its narrower spikelets, which are less than 2 mm wide, wheras O. punctata is distinguished by its shorter spikelets, which are less than 7 mm long. Weedy ‘red rice’ types originating from crosses between O. sativa and O. rufipogon are distinguished by their longer ligules, which are at least 10 and up to 17 mm long. The closest relative of O. barthii is the cultivated O. glaberrima, which is distinguished by its persistent, smooth (not hispid), normally awnless spikelets.
Due to the variable regulations around (de)registration of pesticides, your national list of registered pesticides or relevant authority should be consulted to determine which products are legally allowed for use in your country when considering chemical control. Pesticides should always be used in a lawful manner, consistent with the product's label.
Cultural Control and Sanitary Measures
Recommendations for cultural control of wild rices in the Senegal River basin include: a) ploughing and irrigation to induce germination and then cultivation to kill the seedlings. Irrigation should be in the hot season at the end of February or beginning of March. Repeating the operations 2-3 times destroys most seeds in the top 10 cm of soil; b) growing an early variety of rice, such as Sintiane Diofior, which is sown during April/May and can be harvested in mid-October, before the flowering or ripening of wild rices. Two harvests of variety Sintiane Diofior, followed by irrigation, are recommended before another variety is sown; c) growing a longer-season variety such as Paugern, which is sown in May and can be harvested at the end of October. The early rapid growth of Paugern smothers weeds and gives results similar to those obtained with Sintiane Diofior except that larger areas can be treated. As with Sintiane Diofior, two harvests are recommended, followed by irrigation (Couey, 1966).
Other suggestions by Wirjahardja et al. (1983) include: a) direct sowing of pre-germinated seed; b) puddling soil before broadcasting crop seed; c) row-seeding, so that weedy rice between rows can be recognised and removed; d) growing cultivars with purple foliage, so the green weedy seedlings can be recognised; e) short-season varieties (as described above); f) deep ploughing; and g) crop rotation.
In Mali, yields were increased by 250% in 1975 and by 9-135% in 1976 over untreated controls when land was ploughed after flooding or under drained conditions, or before further cultivations. When these preparations were followed by various weed control treatments the yield was increased by 200% or more (Valee, 1980).
There are no herbicides that provide clear selectivity against O. barthii in rice, though Deuse et al. (1978) obtained some selectivity with oxadiazon at 0.6 and 0.75 kg/ha, and Wirdjahardja et al (1983) refered to some selectivity with molinate. In a recent review, Rodenburg and Johnson (2013) confirmed that O. barthii is not controlled by propanil, butachlor, oxadiazon, thiobencarb or pendimethalin, but molinate, quinclorac, and pretilachlor may be partially selective.
Pre-irrigation and delayed sowing allows for effective control by glyphosate.
Simulated moisture stress and hydrogen ion activity have been shown to adversely affect germination, indicating high sensitivity to moisture stress and solute concentrations. Treatments that would increase the soil pH could therefore limit the spread of O. barthii (Obadoni et al., 2004).
Parker and Dean (1976) achieved selectivity with several herbicides in combination with seed ‘safeners’ such as naphthalic anhydride, but this technique was never used widely in the field. The more recent equivalent is the use of imidazolinone-resistant rice varieties such as Clearfield, though it has been pointed out (e.g. DFID, 2002; IRRI, 2013) that any out-crossing from the crop could lead to the build-up of equally herbicide-resistant weed. Fortunately this may be less of a problem with O. barthii, as it apparently crosses less readily with O. sativa than the forms of wild rice derived from O. rufipogon; however, because crossing of O. barthii with O. glaberrima occurs more readily, it is possible that the risk with the new NERICA (‘New Rice for Africa’) rice varieties, which have been derived by crossing O. sativa and O. glaberrima, could be greater, but no evidence for this has been seen so far. If this approach is used, IRRI (2013) suggested the following precautions to be followed:
- Do not grow herbicide-resistant rice in consecutive years in the same field.
- Rotate herbicides and use of herbicide mixtures to ensure complete control of weeds.
- Do not allow escaped weeds to go to seed by pulling out the escaped weeds.
- Use integrated weed management programs that include effective cultural practices (land preparation, tillage practices, water management, etc.) and rotation of crop establishment methods.
Johnson DE, Riches CR, Kayake J, Tuor F, 2000. Wild rice in sub-Saharan Africa: its incidence and scope for improved management. In: Proceedings of Global Workshop on Red Rice Control, Varadero, Cuba. 30 August - 3 September 1999 [ed. by Proceedings of Global Workshop Red Rice Control, on \Varadero, \Cuba]. Rome, Italy: Food and Agriculture Organisation (FAO) of the United Nations, 87-93
Kaushal P, Ravi P, 1996. Crossability of wild species of Oryza with O. sativa cvs PR 106 and Pusa Basmati 1 for transfer of bacterial leaf blight resistance through interspecific hybridization. Journal of Agricultural Science, 130(4):423-430
Khush GS, 1997. Origin, dispersal, cultivation and variation of rice. Plant Molecular Biology, 35:25-34
Johnson DE, Riches CR, Kayake J, Tuor F, 2000. Wild rice in sub-Saharan Africa: its incidence and scope for improved management. [Proceedings of Global Workshop on Red Rice Control, Varadero, Cuba. 30 August - 3 September 1999], Rome, Italy: Food and Agriculture Organisation (FAO) of the United Nations. 87-93.