Weed ora in arable elds of eastern Ethiopia with emphasis on the occurrence of Parthenium hysterophorus

T TAMADO & P MILBERG*

Department of Plant Science, Alemaya University, PO Box 138, Dire Dawa, Ethiopia and Department of Ecology and Crop Production Science, Swedish University of Agricultural Sciences, Box 7043, SE-750 07 Uppsala, and *Department of Biology-IFM, Linkoping University, SE-581 83 Linko ping, Sweden Received 25 January 2000 Revised version accepted 23 May 2000

Summary
An exploratory weed survey was carried out on 240 crop elds in eastern Ethiopia to assess the inuence of some environmental and crop management factors on weed species composition and distribution, and to investigate the association of the recently introduced Parthenium hysterophorus L. with other components of the weed ora. A total of 102 weed taxa belonging to 36 plant families were recorded. Asteraceae, Poaceae and Fabaceae were the most abundant families based on the number of species recorded. Digitaria abyssinica (Hochst. Ex A. Rich) Stapf was the most frequent species (63%) and Parthenium hysterophorus the second most frequent (54%). The latter species was ranked as the most important weed by 90% of the farmers in the lowlands while 86% of the farmers in the highlands ranked the former species as the worst weed. According to a partial canonical correspondence analysis (pCCA), altitude, rainfall, month of planting, number of weedings and soil type were the major environmental/crop management factors inuencing the species distribution in the study area. The rst pCCA axis clearly structured highland and lowland weed species while the second axis distinguished those species that grow on sandy soils. Parthenium hysterophorus has, in a very short time period, emerged as one of the most troublesome weed species in eastern Ethiopia.
Keywords: Ethiopia, Parthenium hysterophorus, pCCA, weed communities.

Introduction
Weeds are the most underestimated pest in tropical agriculture, but they have inuenced human activities more than other crop pests. Although crop yield losses from weeds vary from crop to crop and from region to region, because of various biotic and abiotic factors, it has been estimated that weeds cause a yield loss of about 10% in the less developed countries and 25% in the least developed countries (Akobundu, 1987). There is no detailed information available about the amount of crop yield losses due to weeds in Ethiopia. Furthermore, the relative
Correspondence: T Tamado, Department of Plant Science, Alemaya University, PO Box 138, Dire Dawa, Ethiopia.

Blackwell Science Ltd Weed Research 2000 40, 507521

507

508

T Tamado & P Milberg

importance of common weed species in the major crops and cropping systems is not well documented (Stroud & Parker, 1989), especially in eastern Ethiopia. The presence of each weed population in an arable eld is the result of ecological reactions to previous management practices, soil characteristics of the site and the regional climate (e.g. Froud-Williams et al., 1983; Dale et al., 1992; Salonen, 1993; Andersson & Milberg, 1998). Weed populations also reect the eects of local weather conditions on recruitment, survival and competitive ability (Milberg et al., 2000). Cultural, chemical and biological weed control activities can also exert a strong selective inuence on the weed populations (Chancellor, 1985). Thus, knowledge of the weed community structure is an important component of weed management, and is essential in setting priorities for both weed management and research. Several exotic weed species have recently been introduced to Ethiopia. Among these, Parthenium hysterophorus L. is now recognized as a serious problem of crop and rangelands, particularly in eastern Ethiopia (Seifu W/Kidan, 1990; Frew et al., 1996; ICRA, 1997). Common names of this species are parthenium weed (Australia), bitter weed, carrot weed, broom-bush and congress weed (India), false ragweed and ragweed parthenium (USA), whitetop, escobar amarga and feverfew (Caribbean) (Navie et al., 1996). Parthenium hysterophorus probably originated around the Gulf of Mexico or in central South America and has spread to central and South Africa, southern USA, India, the south China region, Australia and the South Pacic (Haseler, 1976). Apart from typical weedy characteristics, such as competitiveness and the ability to develop and reproduce in a wide range of conditions (Chippendale & Panetta, 1994), many parts of P. hysterophorus plants are reported to have an allelopathic eect on other plants (Dayama, 1986; Swaminathan et al., 1990; Adkins & Sowerby, 1996). Furthermore, P. hysterophorus is toxic to animals (More et al., 1982; Tudor et al., 1982) and causes severe human health problems (Khosla & Sobti, 1979; Chippendale & Panetta, 1994). This weed is believed to have been introduced to eastern Ethiopia with army vehicles from Somalia during the EthiopianSomalian war of 197677. It is locally known as `Klidnole' (living alone) or `Faramsis' (sign your land) possibly indicating its allelopathic potential and strong competitiveness. In view of its recent introduction, however, the ecological distribution of P. hysterophorus in crop elds of eastern Ethiopia is not well known. The objectives of this study were: (i) to determine the prevalence and agro-ecological distribution of weed species in major crops; (ii) to assess the inuence of some environmental factors and crop management practices on weed species composition and distribution; and (iii) to investigate the association of P. hysterophorus with other components of the weed ora.

Materials and methods

Study area

Eastern Ethiopia covers an area of about 266 195 km2 and makes up 21% of the total area of the nation (CSA, 1996). Most of the region is arid or semi-arid, characterized by low and erratic rainfall, and not suitable for crop production. Rainfed crop production is mainly concentrated at altitudes greater than 1500 m. The major crops in eastern Ethiopia are sorghum [Sorghum bicolor (L.) Moench], maize (Zea mays L.) and common bean (Phaseolus vulgaris L.), accounting for about 44%, 33% and 3%, respectively, of the estimated total cultivated area of 340 000 ha

Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

509

(CSA, 1996) and are often grown as intercrops. Groundnut (Arachis hypogaea L.) is also an important cash crop in semi-arid parts of eastern Ethiopia.
Weed survey

A survey of the weed ora in major crops in eastern Ethiopia was carried out during the main rainy season from June to September 1998. Two hundred and forty crop elds ranging in altitude from 1140 m to 2660 m with an annual rainfall of 400 to over 1000 mm were surveyed. The survey areas based on administrative zones are shown in Fig. 1. The elds surveyed represented the major agro-ecological zones and were selected as follows. First, a stratied sampling method based on altitude was chosen: low altitude (<1600 m), medium altitude (16001900 m), high altitude (>1900 m), a classication commonly used in relation to sorghum research in Ethiopia. Second, representative districts within each altitude zone were selected, then representative peasant associations within each district, villages within association, and farmers within villages. Third, elds were selected regardless of size, and on the grounds of accessibility (adjacent to road) and whether it carried the required crop or crop combination. In each eld, four to six 4-m2 quadrats were used following an inverted W pattern (Thomas, 1985). A preliminary species vs. area analysis suggested this sample size to be appropriate to detect almost all of the weed species in the small elds (average farm size is 0.65 ha; CSA, 1996). Percentage cover (the ground area covered by the vertical projection of above-ground plant parts) was estimated by eye for each of the weed species in each quadrat. Identication in the eld was based on weed identication guides (Terry & Michieka, 1987; Stroud & Parker, 1989). Weed species that were dicult to identify in the eld were tagged, pressed and identied in the herbarium of the Alemaya University. Nomenclature follows Flora of Ethiopia and Eritrea, volumes 2, 3, 6 and 7 (Hedberg & Edwards, 1989, 1995; Edwards et al., 1995, 1997).

Fig. 1 Areas surveyed in eastern Ethiopia showing the administrative zones.

Blackwell Science Ltd Weed Research 2000 40, 507521

510

T Tamado & P Milberg

Data on major environmental factors and crop management practices believed to inuence the weed ora in general and P. hysterophorus in particular in each eld were collected by observation (soil type, topography, type of crop), measurement (altitude), interviewing farmers (number of ploughings before planting, month of planting, fertilizer use, number of weeding before the survey date) or from secondary sources (administrative zone, rainfall). Environmental and crop management variables of nominal type (administrative zone, soil type, topography, crop type, fertilizer use) were converted into binary dummy variables that take the value 1 if the eld belongs to the category or 0 if it does not. Altitude, number of ploughings, month of planting, number of weedings and rainfall were quantitative variables and hence measured on an interval scale. Survey date was used as a covariable in the data analysis. A summary of the description of the environmental and crop management variables used in this study is given in Table 1.
Interviews with farmers

Farmers were asked 11 questions on weeds and weed control in general and on P. hysterophorus in particular. Because answers tended to be relatively consistent within each village, only a subset (n 148) of the farmers whose elds had been surveyed were interviewed. The answers reported here stem from the following ve questions: `List and rank the major weeds you encounter in your eld', `Are you aware of P. hysterophorus?', `How and when do you think P. hysterophorus was introduced to this area?', `In which place is P. hysterophorus most problematic: cultivated land, grazing area, forest, or other (specify)?' and `List benecial uses of P. hysterophorus (if any) in order of importance'.
Data analyses

The data on weed species were summarized using: (1) frequency the number of elds in which a species occurred expressed as percentage of the total number of elds surveyed; and (2) eld uniformity the number of sampling quadrats in which a species occurred in a eld expressed as percentage of the total number of samples. An initial detrended correspondence analysis (DCA) was performed on weed species cover (%) data to determine the length of the rst DCA axis. Its length (4.5 standard deviations) suggested strong unimodal species responses along the environmental gradients (Ter Braak & Prentice, 1988), and we therefore decided to use partial canonical correspondence analysis (pCCA), considering dates of the survey as a covariable. A pCCA using percentage cover of P. hysterophorus as the only explanatory variable was also performed to ordinate the distribution of other weed species along the P. hysterophorus cover gradient. Cover values of P. hysterophorus were square-root transformed to dampen the inuence of outliers. The CANOCO 4.0 computer software was used for data analyses (Ter Braak & Smilauer, 1998) and default options were used, except that we downweighted rare species to reduce their unduly large inuence in both analyses.

Results and discussion

Weed communities

A total of 102 weed taxa belonging to 36 plant families were recorded (Appendix 1). Seedlings and small plants of Rhynchosia, Orobanche and Kohautia could not be identied to the species
Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

511

Table 1 Summary of environmental and crop management variables used in partial canonical correspondence analysis (pCCA)

Variable (scale) Administrative zones (nominal) West Hararge East Hararge Dire Dawa Jijiga Altitude (m above sea level) Soil texture (nominal) Clay Sandy Intermediate Month of planting (interval) March (1) April (2) May (3) June (4) July (5) Topography (nominal) Plain Sloping Fertilizer use (nominal) Yes No Number of ploughings (interval) 1 2 3 Number of weedings (interval) 0 1 2 3 4 Type of crop (nominal) Sorghum/groundnut intercrop Sole sorghum Maize/common bean intercrop Sorghum/maize intercrop Sole maize Sorghum/common bean intercrop Maize/groundnut intercrop Sole groundnut Annual rainfall (interval) Date of survey (used as a covariable)

Code

No. of elds or range

W Hararge E Hararge Dire Dawa Jijiga ALTITUDE Clay Sand Intermediate soil MONTH OF PLANTING

57 116 30 37 11402660 76 73 91 15 19 61 48 81 31 164 76 47 193 13 66 155 19 04 2 94 121 21 2 18 90 18 38 48 8 4 16 17

Plain Slope Fertilizer NO. PLOUGHINGS

NO. WEEDINGS

Sorghum/groundnut Sorghum Maize/bean Sorghum/maize Maize Sorghum/bean Maize/groundnut Groundnut RAINFALL

level. The most important families, based on the number of taxa, were Asteraceae (16), Poaceae (12) and Fabaceae (10). These three families were also the most important in small-scale farming in highland Peru, central Mexico and northern Zambia (Afors, 1994; Becker et al., 1998; Vibrans, 1998). These families are very species rich so it is not surprising that they contain many weeds. Although annuals were more common (73 taxa) than perennials (27 taxa), four of the ve most frequent species were perennials (Appendix 1).
Blackwell Science Ltd Weed Research 2000 40, 507521

512

T Tamado & P Milberg

Pulschen (1990) recorded 444 species, belonging to 60 plant families, from central Ethiopia. This large number of species came from 12 dierent annual crops assessed, as compared with four in this study. Moreover, the central Ethiopian highlands receive more rainfall and have better potential for crop production than eastern Ethiopia. About half of the species recorded during the survey are non-native, and it is noteworthy that two of the weeds ranked by farmers as most troublesome (Table 2) were quite recently introduced (P. hysterophorus and Argemone ochroleuca Sweet). Furthermore, Xanthium spinosum L. and Chenopodium opulifolium Schr. were also reported by the farmers to be recent introductions to their areas (unpubl. data from interviews with 148 farmers). Hence, there is reason to be aware of the danger of introduction of new species. One such species, not recorded in the surveyed elds but noted at low altitude during this survey, is Verbesina encelioides (Cav.) A. Gray (Asteraceae). The weed ora in crop elds of eastern Ethiopia was dominated by rather few species, which is a common phenomenon in intensied farming systems (Phillips, 1992; Salonen, 1993; Weber et al., 1995). Only 11 species occurred with a frequency 30%. Digitaria abyssinica (Hochst. Ex A. Rich) Stapf, P. hysterophorus, Cyperus rotundus L., Cynodon dactylon (L.) Pers. and Commelina benghalensis L. had the highest frequencies and eld uniformities. The ranking of weed species based on frequency and eld uniformity was slightly dierent (Appendix 1). The species that deviated most in ranking, and had the largest uniformity/frequency ratio, was Galinsoga parviora Cav. Hence, this species was much more uniformly distributed in elds than other species. Weed control measures in maize and sorghum in eastern Ethiopia are usually hand-hoeing or inter-row oxen cultivation. As there is no use of herbicides and only limited use of inorganic fertilizers, agrochemicals exert only a marginal selection pressure on the weed ora. In southern Spain, Pujadas Salva & Hernandex Bermejo (1988) found a very diverse ora that varied considerably depending on the level of various agricultural inputs. For instance, in non-irrigated

Table 2 Percentage of farmers, living at dierent altitudes in east Ethiopia, that ranked a weed species as rst, second and third most important in their elds. The 2nd and 3rd columns do not add up to 100% because some farmers listed only one weed species
Altitude (m above sea level) <1600 (n = 30) First Cyperus spp. Cynodon dactylon Xanthium abyssinicum Parthenium hysterophorus Digitaria abyssinica Bidens pilosa Galinsoga parviora Amaranthus graecizans Snowdenia polystachya Tagetus minuta Striga spp. Argemone ochroleuca Oxalis latifolia Acanthospermum hispidium Xanthium spinosum 3 90 7 Second 33 3 10 Third 7 17 17 16001900 (n = 67) First 4 28 37 24 3 3 Second 10 3 45 27 1 3 7 Third 3 3 6 6 3 1 First 4 86 2 2 6 >1900 (n = 51) Second Third 37 12 6 6 2 12 6 2 6 6 2 4

Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

513

low input elds of annual crops they recorded 334 species, but found only 79 species in irrigated high input elds. When asked to list important weed species in their area, farmers in our survey mentioned only 15 species in total. Ninety per cent of the interviewed farmers in the lowlands ranked P. hysterophorus as the most important weed while 86% of the farmers in the highlands ranked Digitaria abyssinica highest (Table 2). Cyperus spp. were mentioned frequently by farmers at all altitudes. The latter is a weed taxon of almost world-wide distribution (Haiger et al., 1982) while Digitaria abyssinica is geographically restricted to eastern parts of the African continent and Sri Lanka (Hedberg & Edwards, 1995). Would it have the potential to become a serious weed elsewhere if introduced to new areas? Though some weeds, such as Striga hermonthica (Del.) Benth., Striga asiatica (L.) Kuntze, Argemone ochroleuca and Xanthium spinosum, had low frequency (Appendix 1), they were considered to be problematic weeds by farmers (Table 2). Thus, high frequency does not indicate the economic or sociological importance of a weed species, as some weeds have other uses, such as feed for livestock, which can be especially important in the lowlands. Three parasitic weeds (Striga hermonthica, Striga asiatica and Orobanche spp.) were recorded in the survey. The two Striga species were the most problematic in some parts of Gelemsso in west Hararge and in most parts of Fedis in east Hararge. Farmers in Gelemsso where Striga spp. occurred were forced to abandon growing sorghum and maize due to the plant parasite, while farmers in Fedis used intercropping sorghum and/or maize with groundnut to reduce the infestation.
Weed community analysis

The pCCA ordinated the weed data in relation to the environmental variables. The rst canonical axis, as well as all canonical axes, were highly signicant (P < 0.01 in both Monte Carlo Permutation tests; 199 permutations under the reduced model), indicating a strong relationship between the species and the environmental variables considered. The variation in weed species composition was related to 23 environmental and crop management variables and the species and environmental correlation in the rst and second axis was 0.92 and 0.85 respectively. The rst two ordination axes explained 49.9% of the total variance in the weighted averages for percentage cover of the species with respect to the environmental variables (Table 3). Although all the taxa were included in the analyses, only the most abundant (i.e. total cover) and important species are illustrated in Fig. 2 (n 40) and Fig. 3 (n 49). The rst ordination axis (k1 0.497; Fig. 2) extracted by the pCCA was most closely associated with altitude as indicated by the high interset correlation of 0.866 (Table 3). Thus, the rst axis clearly structured highland and lowland weed species. Galinsoga parviora, Bidens pachyloma (Oliv. & Hiern.) Cuf., Medicago polymorpha L. and Erucastrum arabicum Fisch & Mey were typical highland species, while P. hysterophorus, Boerhavia erecta L., Portulaca oleraceae L., Xanthium abyssinicum Wallroth, Tribulus terrestris L. and Euphorbia heterophylla L. were typical lowland species. However, altitude represents a complex gradient and inuences other environmental and crop management variables. As a result, high rainfall, sloping topography and early planting had strong and positive association with altitude (Fig. 2). The second ordination axis (k2 0.326; Fig. 2) most clearly distinguished species of sandy soil as indicated by the high interset correlation of 0.634 (Table 3). Acanthospermum hispidium DC.,
Blackwell Science Ltd Weed Research 2000 40, 507521

514

T Tamado & P Milberg

Variable Administrative zones West Hararge East Hararge Dire Dawa Jijiga Altitude Soil texture Sandy Intermediate Clay Month of planting Topography Plain Slope Fertilizer use Number of ploughings Number of weedings Type of crop Sorghum/groundnut Sorghum Maize/bean Sorghum/maize Maize Sorghum/bean Maize/groundnut Groundnut Annual rainfall

Axis 1

Axis 2

0.340 0.142 )0.365 )0.152 0.866 )0.353 0.243 0.080 )0.542 )0.368 0.368 0.326 0.178 0.494 )0.149 )0.316 0.094 0.270 0.205 0.083 )0.046 )0.127 0.589

)0.031 0.492 )0.472 )0.163 0.070 0.634 )0.283 )0.302 )0.057 )0.001 0.001 )0.105 )0.072 )0.082 0.249 )0.347 )0.007 )0.052 0.088 )0.018 )0.058 0.440 0.086

Table 3 Inter-set correlations of environmental and crop management variables with the rst two ordination axes from a pCCA of weed survey data from eastern Ethiopia. See Fig. 2

Ocimum basilicum L., Cleome monophylla L. and Oxygonium sinuatum (Meissn.) Dammer were the weed species mainly associated with sandy soils, where groundnut crop and sorghum/ groundnut intercrop are most common (Fig. 2b and Table 3). Fields on clay and intermediate soil types were relatively similar in species composition. Typical weeds for lowlands, where sorghum crops prevailed, were P. hysterophorus, Euphorbia heterophylla, Launea cornuta (Oliv. & Hiern) C. Jerey and Euphorbia hirta L. This supports the notion that soil type structures regional weed oras (e.g. Dale et al., 1992; Hallgren et al., 1999). Several of the most frequent species, such as Digitaria abyssinica, Tagetus minuta L., Commelina benghalensis and Datura stramonium L. were located near the origin of the ordination diagram (Fig. 2) indicating their ability to thrive under very diverse agro-ecosystems. This is also an indication that they may become serious weeds in other situations into which they could potentially be introduced to. Rare species, such as Schkuhria pinnata (Lam.) Thell., Kohautia spp., Barleria eranthemoides R. Brown and Antirrhinum orontium L. (not shown in Fig. 2), were on the edge of the ordination diagram, which could be due either to their preference for relatively extreme environmental conditions, or to a few recordings that by chance occurred at sites with extreme conditions. Therefore, rare species were down-weighted in the analyses and they are not shown in the ordination gures. Among the environmental variables considered, number of ploughings before planting and maize/bean intercrop had a short arrow and a centroid near the origin, respectively (Fig. 2), as
Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

515

Fig. 2 Ordination diagram of species (a) and environmental variables (b) of a pCCA of the weed cover data in 240 crop elds in eastern Ethiopia (keys to the environmental variables and species names is given in Table 1 and Appendix 1 respectively). The variables with interval scale are represented by arrows (upper case text) and the nominal variables by centroids (lower case text). Only the most abundant species are indicated.

Blackwell Science Ltd Weed Research 2000 40, 507521

516

T Tamado & P Milberg

Fig. 3 A pCCA ordination diagram of the weed species along the density gradient of Parthenium hysterophorus (key to the species names is given in Appendix 1). Rare species, with an weighted abundance <10, are not indicated.

well as low interset correlations (Table 3). Consequently, they were less important in explaining the variation in weed species.
Parthenium hysterophorus

Parthenium hysterophorus was the second most abundant species on crop land (Appendix 1), despite being rare at high altitudes, and was perceived as one of the most troublesome weeds (Table 2) within only 20 plant generations from its introduction. From the interviews conducted during the study, farmers in east Hararge, Jijiga and Dire Dawa generally believed that P. hysterophorus was introduced into their area during the EthiopianSomalian war of 197677 by army vehicles, while farmers in west Hararge had dierent views on its introduction such as with grain, livestock and road construction (Table 4, unpubl. data), indicating its likely secondary dispersal from east Hararge. Of the 148 farmers interviewed, 93% in the lowlands, 82% in the mid-altitudes and 35% in high altitudes were aware of the problem of P. hysterophorus in crop lands and/or grazing areas. Of the farmers who were aware of the species, 93% in the lowland and 67% at mid-altitudes considered
Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

517

Table 4 Farmers' views on Parthenium hysterophorus in eastern Ethiopia

Altitude (m.a.s.l.) <1600 Awareness of the weed status of Parthenium (%) Yes No Means of introduction to their area (%) With grain With army vehicles With livestock With road construction Parthenium is an important weed on (%) Grazing land only Both grazing and crop lands Roadsides only (n = 30) 93 7 (n = 28) 54 46 (n = 28) 7 93 16001900 (n = 67) 82 18 (n = 55) 15 29 18 38 (n = 55) 33 67 >1900 (n = 51) 35 65 (n = 18) 44 12 44 (n = 18) 11 6 83

P. hysterophorus to be the most important weed both in crop lands and grazing areas. In the highlands almost all farmers considered it only as a roadside weed. These gures show that this weed is currently more important at low and mid-altitudes than at high altitudes. However, as reported by Haseler (1976) the initial occurrence of P. hysterophorus in a new area usually occurs along roadsides and it is from this foothold that it spreads extensively into agricultural land. We therefore believe that it has the potential to become a serious weed on arable land even in the highlands. No benecial aspects of this weed were reported by the interviewed farmers. The pCCA with density of P. hysterophorus as the sole environmental variable showed that Corchorus trilocularis L., Blepharis ciliaris L., Striga hermonthica, Euphorbia heterophylla, Euphorbia hirta, Boerhavia erecta and Portulaca oleraceae were positively associated with a high density of P. hysterophorus (Fig. 3). In contrast, Galinsoga parviora, Medicago polymorpha, Cyperus esculentus L., Amaranthus hybridus L., Polygonum nepalensis Spreng., Sonchus asper (L.) Hill and Rumex bequaertii De Wild were typical of elds lacking P. hysterophorus. The most frequent species such as D. abyssinica, C. rotundus, C. dactylon and T. minuta were located near the origin of the pCCA score, indicating a potential inertness to competition by P. hysterophorus.

Conclusions
This survey has ranked the most abundant and troublesome weed species in small-scale farms in eastern Ethiopia information that is vital for setting research and development priorities concerning weed management. Furthermore, it has documented the current distribution of the recently introduced P. hysterophorus, a species that has spread rapidly and now aects the livelihood of numerous small-scale farmers.

Acknowledgements
We thank Torsten Andersson for discussions at the planning stage and for critically reading the manuscript and Sigurd Hakansson for generously sharing his weed experiences. We are also most grateful to Abdulrazak Abdulahi for assisting in the eld and in plant identication. Financial
Blackwell Science Ltd Weed Research 2000 40, 507521

518

T Tamado & P Milberg

support was provided by the Swedish Agency for Research Co-operation with Developing Countries (SAREC).

References
ADKINS SW & SOWERBY MS (1996) Allelopathic potential of the weed, Parthenium hysterophorus L., in Australia. Plant Protection Quarterly 11, 2023. AFORS M (1994) Weeds and weed management in small-scale cropping systems in northern Zambia. Crop Production Science 21, Department of Crop Production Science, Swedish University of Agricultural Sciences, Uppsala, Sweden. AKOBUNDU IO (1987) Weed Science in the Tropics. Principles and Practices. Wiley, Chicester, UK. ANDERSSON TN & MILBERG P (1998) Weed ora and the relative importance of site, crop, crop rotation, and nitrogen. Weed Science 46, 3038. BECKER B, TERRONES F & HORCHLER P (1998) Weed communities in Andean cropping systems of northern Peru. Angewandte Botanik 72, 113130. CHANCELLOR RJ (1985) Changes in the weed ora of an arable eld cultivated for 20 years. Journal of Applied Ecology 22, 491501. CHIPPENDALE JF & PANETTA FD (1994) The cost of P. hysterophorus in the Queensland cattle industry. Plant Protection Quarterly 9, 7376. CSA (1996) Distribution of Total Number of Households, Total Crop Land Area and Crop Land Area per Household by Size of Holdings. Agricultural Sample Survey, Central Statistical Authority, Addis Ababa, Ethiopia. DALE MRT, THOMAS AG & JOHN EA (1992) Environmental factors including management practices as correlates of weed community composition in spring seeded crops. Canadian Journal of Botany 70, 19311939. DAYAMA OP (1986) Allelopathic potential of Parthenium hysterophorus L. on the growth, nodulation, and nitrogen content of Leucaena leucocephala. Leucaena Research Reports 7, 3637. EDWARDS S, MESFIN TADESSE & HEDBERG I (1995) Flora of Ethiopia and Eritrea, Part 2 (Canellaceae to Euphorbiaceae), Vol. 2. The National Herbarium (Addis Ababa, Ethiopia) and Department of Systematic Botany (Uppsala, Sweden). EDWARDS S, SEBSEBE DEMISSEW & HEDBERG I (1997) Flora of Ethiopia and Eritrea (Hydrocharitaceae to Arecaceae), Vol. 6. The National Herbarium (Addis Ababa, Ethiopia) and Department of Systematic Botany (Uppsala, Sweden). FREW M, SOLOMON K & MASHILLA D (1996) Prevalence and distribution of Parthenium hysterophorus L. in eastern Ethiopia. In: Proceedings First Annual Conference of the Ethiopian Weed Science Society, November 1993, Addis Ababa, Ethiopia, 2425. FROUD-WILLIAMS RJ, DRENNAN DHS & CHANCELLOR RJ (1983) Inuence of cultivation regime on weed oras of arable cropping systems. Journal of Applied Ecology 20, 187197. HAFLIGER E, KUHN U, HAMET-AHTI L, COOK CDK, FADEN R & SPETA F (1982) Monocot Weeds 3. Documenta, CIBA-GEIGY, Basel, Switzerland. HALLGREN E, PALMER MW & MILBERG P (1999) Data diving with cross-validation: and investigation of broad-scale gradients in Swedish weed communities. Journal of Ecology 87, 10371051. HASELER WH (1976) Parthenium hysterophorus L. in Australia. Pest Articles & News Summaries (PANS) 22, 515517. HEDBERG I & EDWARDS S (1989) Flora of Ethiopia (Pittosporaceae to Araliaceae), Vol. 3. The National Herbarium (Addis Ababa, Ethiopia) and Department of Systematic Botany (Uppsala, Sweden). HEDBERG I & EDWARDS S (1995) Flora of Ethiopia and Eritrea 7 (Poaceae), Vol. 7. The National Herbarium (Addis Ababa, Ethiopia) and Department of Systematic Botany (Uppsala, Sweden). ICRA (1997) Strengthening Farmer Participatory Research and Development in Jijiga Zone: The Case of Moisture Stress Reduction in the Plains and Soil Fertility Management in the Hills. International Center for Development-oriented Research in Agriculture, Wageningen, The Netherlands.

Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

519

KHOSLA SN & SOBTI SN (1979) Parthenium a national health hazard, its control and utility A review. Pesticides 13, 121127. MILBERG P, HALLGREN E & PALMER MW (2000) Interannual variation in weed biomass on arable land in Sweden. Weed Research 40, 311321. MORE PR, VADLAMUDI VP & QURESHI MI (1982) Note on the toxicity of Parthenium hysterophorus in livestock. Indian Journal of Animal Science 52, 456457. NAVIE SC, MCFADYEN RE, PANETTA FD & ADKINS SW (1996) The biology of Australian weeds 27. Parthenium hysterophorus L. Plant Protection Quarterly 11, 7688. PHILLIPS MC (1992) A survey of the arable weeds of Botswana. Tropical Pest Management 38, 1321. PUJADAS SALVA A & HERNANDEZ BERMEJO JE (1988) Floristic composition and agricultural importance of weeds in southern Spain. Weed Research 28, 175180. PULSCHEN L (1990) Composition, synecology and sociological structure of the agrestal ora in Shewa Province, Ethiopia. Angewandte Botanik 64, 445456. SALONEN J (1993) Weed infestation and factors aecting weed incidence in spring cereals in Finland a multivariate approach. Agricultural Science in Finland 2, 525536. SEIFU W/KIDAN (1990) Parthenium hysterophorus L. A Recently Introduced Noxious Weed to Ethiopia. A Preliminary Reconnaissance Survey Report on Eastern Ethiopia. East Hararge Ministry of Agriculture, Ethiopia. SWAMINATHAN C, RAI RS & SMESH KK (1990) Allelopathic eects of Parthenium hysterophorus L. on germination and growth of a few multi-purpose trees and arable crops. International Tree Crops Journal 6, 143150. STROUD A & PARKER C (1989) A Weed Identication Guide for Ethiopia. Food and Agriculture Organization of the United Nations, Rome, Italy. TER BRAAK CJF & PRENTICE IC (1988) A theory of gradient analysis. Advances in Ecological Research 18, 271317. TER BRAAK CJF & SMILAUER P (1998) CANOCO for Windows: Software for Canonical Community Ordination. Microcomputer power, Ithaca, NY, USA. THOMAS AG (1985) Weed survey system used in Saskatchewan for cereal and oil seed crops. Weed Science 33, 3443. TERRY PJ & MICHIEKA RW (1987) Common Weeds in East Africa. Food and Agriculture Organization of the United Nations, Rome, Italy. TUDOR GD, FORD AL, ARMSTRONG TR & BROMAGE EK (1982) Taints in meat from sheep grazing Parthenium hysterophorus. Australian Journal of Experimental Agriculture and Animal Husbandry 22, 4346. VIBRANS H (1998) Native maize eld weed communities in south-central Mexico. Weed Research 38, 153166. WEBER G, ELEMO K & LAGOKE STO (1995) Weed communities in intensied cereal-based cropping systems of the northern Guinea savanna. Weed Research 35, 167178.

Blackwell Science Ltd Weed Research 2000 40, 507521

520

T Tamado & P Milberg

Appendix 1 Weed species of crop elds in eastern Ethiopia, their family, frequency, eld uniformity and life form
Species Digitaria abyssinica (Hochst. Ex A. Rich) Stapf Parthenium hysterophorus L. Cyperus rotundus L. Cynodon dactylon (L.) Pers. Commelina benghalensis L. Tagetus minuta L. Leucas martinicesis (Jacq.) Ait.f. Amaranthus graecizans L. Heliotropium zeylanicum (Burm.) Lam. Bidens pilosa L. Launaea cornuta (Oliv. & Hiern) C. Jeffrey Xanthium abyssinicum Wallroth Oxygonum sinuatum (Meissn.) Dammer Galinsoga parviora Cav. Guzotia scabra (Vis.) Chiov. Datura stramonium L. Commelina latifolia A. Rich Solanum nigrum L. Corchorus trilocularis L. Acanthospermum hispidium DC. Sida acuta Burm. F. Euphorbia hirta L. Erucastrum arabicum Fisch & Mey Cleome monophylla L. Euphorbia heterophylla L. Medicago polymorpha L. Snowdenia polystachya (Fresen.) Pilg. Portulaca oleracea L. Chenopodium opulifolium Schr. Plantago lanceolata L. Trichodesma zeylanicum (L.) R. Br. Hibiscus trionum L. Cyperus esculentus L. Anthriscus sylvestris (L.) Hoffm. Anagallis arvensis L. Tribulus terrestris L. Oxalis latifolia H. B. K. Sonchus asper (L.) Hill Argemone ochroleuca Sweet Acalypha crenata A. Rich. Setaria verticillata (L.) Beauv. Crotalaria incana L. Rumex bequaertii De Wild. Conyza bonariensis (L.) Cronq. Boerhavia erecta L. Blepharis ciliaris L. Senna occidentalis (L.) Link Ocimum basilicum L. Solanum dubium Fresen. Gynandropsis gynandra (L.) Briq. Xanthium spinosum L. Eragrostis cilianensis (All.) Vign. Ex Janchen Polygonum nepalensis Spreng. Galium hamatum L. Justica schimperi (Hochst.) Dandy Lysimachia ruhmeriana Vatke Alternanthera pungens Kunth Convolvulus arvensis L. Code Digi aby Part hys Cype rot Cyno dac Comm ben Tage min Leuc mar Amar gra Heli zey Bide pil Laun cor Xant aby Oxyg sin Gali par Guzo sca Datu str Comm lat Sola nig Corc tri Acan his Sida acu Euph hir Eruc ara Cleo mon Euph het Medi pol Snow pol Port ole Chen opu Plan lan Tric zey Hibi tri Cype esc Anth sil Trib ter Oxal lat Sonc asp Arge och Family Poaceae Asteraceae Cyperaceae Poaceae Commelinaceae Asteraceae Lamiaceae Amaranthaceae Boraginaceae Asteraceae Asteraceae Asteraceae Polygonaceae Asteraceae Asteraceae Solanaceae Commelinaceae Solanaceae Tiliaceae Asteraceae Malvaceae Euphorbiaceae Brassicaceae Capparaceae Euphorbiaceae Fabaceae Poaceae Portulacaceae Chenopodiaceae Plantaginaceae Boraginaceae Malvaceae Cyperaceae Apiaceae Primulaceae Zygophyllaceae Oxalidaceae Asteraceae Papaveraceae Euphorbiaceae Poaceae Fabaceae Polygonaceae Asteraceae Nyctaginaceae Acanthaceae Fabaceae Lamiaceae Solanaceae Capparaceae Asteraceae Poaceae Polygonaceae Rubiaceae Acanthaceae Primulaceae Amaranthaceae Convolvulaceae Frequency (%) 63 54 45 43 43 40 39 35 31 30 30 28 28 28 27 27 23 22 22 21 21 20 19 18 18 17 16 14 12 12 11 11 10 10 10 9 9 8 8 8 8 7 7 7 7 7 6 6 6 6 5 5 5 5 5 5 5 5 Uniformity (%) 39 37 33 25 28 23 21 18 15 16 14 14 17 21 16 11 16 9 8 13 9 10 10 9 8 9 7 8 5 5 6 5 7 6 4 4 5 3 3 3 3 3 2 3 4 2 2 4 2 2 2 2 3 2 2 2 2 2 Life form P grass A herb P sedge P grass P herb A herb A herb A herb A herb A herb P herb A herb A herb A herb A herb A herb P herb A herb A herb A herb P herb A herb A herb A herb A herb A herb A grass A herb A herb B herb A herb A herb P sedge A herb A herb A herb P herb A herb A herb A herb A grass A herb P herb A herb P herb A herb A herb A herb P shrub A herb A herb A grass A herb A herb P shrub A herb P herb P herb

Crot inc Rume beq Cony bon Boer ere Blep cil Ocim bas Sola dub

Poly nep

Alte pun Conv arv

Blackwell Science Ltd Weed Research 2000 40, 507521

Weeds in eastern Ethiopia

521

Appendix 1 (Continued)
Species Solanum incanum L. Alchemilla abyssinica Fresen. Avena fatua L. Glycine wightii (Wight & Arn.) Verdc. Bidens pachyloma (Oliv. & Hiern) Cuf. Alysicarpus quartinianus A. Rich Amaranthus hybridus L. Orobanche spp. Flaveria trinervia (Spreng.) C. Mohr Mollugo nudicaulis Lam. Zinnia peruviana L. Striga asiatica (L.) Kuntze Oplismenus compositus (L.) P. Beauv. Verbena ofcinalis L. Dolichos formosus A. Rich. Ipomoea eriocarpa R. Br. Striga hermonthica (Del.) Benth. Antirrhinum orontium L. Corrigiola littoralis L. Asystasia schimperi T. Anderss. Kohautia spp. Physalis philadelphica Lam. Droguetia iners (Forssk.) Schweinf. Senecio abyssinicus Oliv. & Hiern Rhynchosia spp. Cassia mimosoides L. Indigofera schimperi Jaub. & Spach. Polygonum barbatum L. Polygonum aviculare L. Scleranthus annuus L. Lamium amplexicaule L. Barleria eranthemoides R. Brown Ruellia patula Jaquin Cucumis cifolius A. Rich. Schkuhria pinnata (Lam.) Thell. Fimbristylis hispidula (VA herbl.) Kunth Tragus racemosus (L.) All. Urochloa panicoides P. Beauv. Sporobolus ioclados (Trin.) Nees Digitaria ternata (A. Rich) Stapf Crotalaria laburnifolia L. Eleusine indica (L.) Gaertn. Euphorbia schimperiana Scheele Equisetum arvense L. Code Family Solanaceae Rosaceae Poaceae Fabaceae Asteraceae Fabaceae Amaranthaceae Orobanchaceae Asteraceae Aizoaceae Asteraceae Scrophulariaceae Poaceae Verbenaceae Fabaceae Convolvulaceae Scrophulariaceae Scrophulariaceae Caryophyllaceae Acanthaceae Rubiaceae Solanaceae Urticaceae Asteraceae Fabaceae Fabaceae Fabaceae Polygonaceae Polygonaceae Caryophyllaceae Lamiaceae Acanthaceae Acanthaceae Cucurbitaceae Asteraceae Cyperaceae Poaceae Poaceae Poaceae Poaceae Fabaceae Poaceae Euphorbiaceae Equisetaceae Frequency (%) 4 3 3 3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Uniformity (%) 1 1 1 1 1 1 2 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 0.1 0.1 0.1 0.2 0.2 0.2 0.1 0.1 0.2 0.2 0.3 1 1 0.1 0.1 0.1 0.2 0.1 0.2 0.3 1 0.3 Life form P shrub A herb A grass P herb A herb A herb A herb PAR herb A herb A herb A herb PAR herb P grass P herb P herb A herb PAR herb A herb A herb A herb A herb A herb P herb A herb P herb A herb P shrub A herb A herb A herb A herb P shrub A herb P herb A herb A sedge A grass A grass P grass A grass A herb A grass A herb P fern

Bide pac Amar hyb

Stri asi

Stri her

P, perennial; A, annual; B, biennial, PAR, parasitic.

Blackwell Science Ltd Weed Research 2000 40, 507521