Survey of Cotton Weeds in Middle Awash of Ethiopia

BACKGROUND

Cotton (Gossypium hirsutum L.) is a very important and widely cultivated fiber crop in the world, which is produced for its lint which is the primary commercial product that generates income for cotton producers and fabric for the textile industries (Basra 2002 and MOI, 2015). In Ethiopia, cotton is produced in small and large-scale commercial farms. Suitable arable land for cotton production in the country was estimated at about 3,000,810 hectares (MOA, 2010). Total annual national production was estimated at about 148,649MT raw cotton (EIAR, 2016).

The major cotton growing areas of Ethiopia include the Awash River basin, Northern and Southern Omo, West Gambella, North Bale and Northwest of Metema and Humera. Also the largest potential areas for cotton production were identified in the Western and Southern parts of the country (MOA, 2010). The major pest problem of cotton production in Ethiopia are lack of effective insect, disease and weed pest management practices(WARC 2000). According to Benedict and Altman (2001) infestation level of any specific pests largely varied from year to year and place to place.

Yield of cotton was greatly reduced due to the natural occurring mixed weed population, which the seed cotton yield loss 62.43 - 96.21% was recorded; when weeding was completely denied throughout the crop growing season (Esayas et al., 2013). Losses caused by weeds in Arkansas, USA were estimated 34 million dollars annually (Smith, 2000). Survey conducted in Middle Awash during 2000 cropping season indicated that the infestation level was very high for most of the weed species, thus broad leaf, grass and sedges. Higher weed densities were recorded at flowering and near harvesting growth stage of cotton resulting in reduction of yield and harvest efficiency (WARC 2000). Weed species such as Xanthium strumarium that was not economical weeds are critical weeds in cotton fields (Esayas and Abraham, 2000). Similarly a seed cotton yield loss of 35.03-88.13% and 56.45-94.44% occurred when weeding was delayed for 60 and 75 days, respectively. So it could be shown that the major yield loss occurred up to 75 days during the cotton growth period (Esayas et al., 2013). Seed cotton yield was increased with the advancement of weed-free period (Esayas et al., 2013). On the other hand, the longer weeds were allowed to grow and compete with the crop, the higher the seed cotton yield reduction (Esayas et al., 2013). Bishnoi et al., (1993) reported that weed free field from 20 days after sowing produced highest seed cotton yields (2798 kg ha^-1) compared to unweeded control (1614 kg ha^-1). Khan and Khan (2003) reported that grassy weeds cause 15 – 40% and broad leaf weeds 15 – 30% yield losses in cotton crop. Keeley and Thullen (1991) reported that 16 and 26% of yield losses occurred when Bermuda grass was permitted to compete with cotton for 12 and 20 weeks, respectively. Also yellow nut sedge (Cyperus esculentus L.) infestation in cotton field was reduced up to 34% seed cotton yield (Mofett and Mcclosky, 1998).Tillage systems and fertilization determines the composition and abundance of weed species in crop fields and can be helpful in understanding how particular weed species increase or decrease, in terms of numbers and diversity, and how crop management can contribute to the suppression of weeds (Travlos et al., 2018)

Weed flora in different locations differ widely in their diversity depending upon environmental and soil conditions and hence the appropriate identification of the weed flora species in the cotton field is essential for the use of effective weed management practices. Therefore, survey of weed flora species composition, distribution and frequency in irrigated Middle Awash potential cotton production areas is essential for a comprehensive understanding of the weed problem that poses negative impacts on cotton production in the study area.

METHODOLOGY

The study was done during 2016 cropping season, between June to September, to determine weed densities, distribution and intensity in irrigated cotton growing areas of middle Awash. The study was carried out in two districts of Middle Awash Rift valley of Ethiopia Amibara (Werer, Badhamo and Melkesedi, Ambash and Billen) and Gewane (Amasabure, Deble, Eglay, Galeallo and Galeallo bora) known by their high potential of irrigated cotton farming. From each irrigated cotton growing districts, 5kebeles and from each kebele, 10-20 farms were randomly selected at every 1-5 Km intervals. The survey was done at seedling and flowering cotton growth stage. The study focuses on small scale farms, private sector and Commercial farms of cotton. During weed sampling, small and large scale areas were considered.

The weed sampling was randomly selected within trisected diagonal in the form of “X” pattern thereby systematically walked in each sample field and the data were counted within 1m²quadrant and a total of 5 samples were taken per hectare.The number of samples per hectare was determined by species in the areas and condition (Pohlan, 1984).Specimens of all weed species encountered in the sampling areas was collected, tagged and pressed in the field using a newspaper and herbarium presser. Field notes were documented by the colour of the flower, fruit, fragrance or any special features of the plants collected. Plant specimens in the field were identified using the available plant identification guides. Identification in the field was based on weed identification guides (Stroud A and Parker C. 1989). Those species that were difficult to identify in the field was tagged, pressed, and sent to the National herbarium of Addis Ababa University and identified.

The data was analyzed by using descriptive statistics to determine weed species composition of abundance (A), dominance (D), frequency (F), and similarity index (SI). The determinations were illustrated by the use of the following formulae (1-4) that was described by Taye and Yohannes (1998).

Frequency: F = X/N x 100…….. (1)

Where, F = frequency, X = number of occurrences of a weed species, N = sample number.

Abundance: A= W/N……… (2)

Where, A = abundance, W = number of individuals of a weed species, N = sample number

Dominance: D= A/ A*100 …… (3)

Where, D = dominance, A = total abundance of all species.

Similarity Index (SI): (Epg)/ (Epg+Epa+Epb)x100 … (4)

Where, SI= similarity index; Epg = number of weed species found in all locations; Epa = number of species only in location a; Epb = number of species only in location b.

RESULTS AND DISCUSSION

Weed species composition

A total of 27 weed species within 17 families was recorded in the cotton-growing areas of Middle Awash (Amibara and Gewane) at seedling and flowering stages of cotton growing periods (Tables 1 and 2). From these weed species, 21 were recorded in both locations and weed species only found in Amibara and Gewane were one and six respectively. The prevalence of weed species at cotton seedling stage was higher than that of flowering stage in both locations. In this study the ten most dominant families of the highest diversity according to the represented weed species were Poaceae, Amaranthaceae, Euphorbiaceous, Malvacea, Convolvulaceae, Asteraceae, Aizoaceae, Cyperaceae, Solanaceae and Chenopodiaceae. Most of the weed species (74 %) were erect annual broad leave herbs, (18.5) grasses, (3.7) Sedges and the rest were annual or perennial climbers or perennial shrubs.

Table 1: Number and proportion of weed species within the seventeen diverse families
No	Family	Life Form	Number of Species	Percent Flora
1	Poaceae	Grass	5	17.85
2	Amaranthaceae	Herb	3	10.74
3	Euphorbiaceae	Herb	3	10.74
4	Malvacea	Herb	2	7.14
5	Convolvulaceae	Herb	2	7.14
6	Asteraceae	Herb	2	7.14
7	Fabaceae	Shrub	1	3.57
8	Cyperaceae	Sedge	1	3.57
9	Cucurbitaceae	Shrub	1	3.57
10	Tiliaceae	Herb	1	3.57
11	Commeliaceae	Herb	1	3.57
12	Solanaceae	Herb	1	3.57
13	Capparaceae	Herb	1	3.57
14	Zygophyllaceae	Herb	1	3.57
15	Chenopodiaceae	Herb	1	3.57
16	Papaveraceae	Herb	1	3.57
17	Aizoaceae	Herb	1	3.57

Frequency and Abundance

The result of weed survey revealed that the dominance level of individual weed species varied across locations and crop growth stages. Asteraceae family has a higher dominance percentage at flowering and lower dominance percentage at seedling in Amibara district and no dominance at Gewane district, while Tiliaceae family has higher dominance percentage at flowering and lower dominance percentage at seedling in Gewane district and lower dominance percentage at Amibara district in both stages

The frequency occurrence of individual weed species ranged from 4 to 80 and 12 to 62 at seedling and flowering stages, while the infestation level based on weed dominance at seedling and flowering cotton stages ranged between 1 to 15, 1 to 37 respectively occurred at high weed density(Tables 1 and 2). This result is supported by Taye and Yohannes (1998), weed species having frequency and dominancy levels below 5.0% and 0.05%, respectively, weeds occurred rarely and at low density. In both Amibara and Gewane districts, higher densities of weed species were recorded during seedling stage. The range of weed species per sample was higher in both locations Amibara and Gewane during cotton seedling stage. In previous studies on different crops such as field pea, faba bean, barley, wheat and teff, weeds had a positive and significant relationship among the weed species abundance, dominance and frequency (Taye and Yohannes, 1998 and Kedir et al., 1999 a b).

The weed flora in cotton fields of Amibara and Gewane were infested by a number of weed species, these weed species infestation of cotton fields can be attributed due to the use of Awash River as irrigation which brings weed seeds and fertile soils from the highland parts of Ethiopia which are required by the crops that create conducive environment for germination and growth of weeds. The average of weed species frequency value over locations, were ranged from 2-80% in both growth stage and districts. The highest weed frequency was recorded by Digitaria abyssinica (78%) followed by Corchorus trilocularis (74%) and Eriocloa fatmensis (72%) at seedling stage in Amibara districts. The lowest frequency value was recorded by Polypogonmon speliensis L. (8%) followed by launaea cornuta, and Acalypha crenata (11 and 12%) respectively. The abundance value of the weed species varied from 1 to 10 plants m-2. The highest abundance value (10 plants m-2) was recorded by Echinocloa colane followed by Digitaria abyssinica (9 plantsm-2), Corchorus trilocularis and Eriocloa fatmensis (7 plants m-2). Whereas, the least abundance value (1 plants m-2) was recorded by Commelina benghalensis and Datura stramonium followed by Abutilo hirtum(Lam.), Ipomea ariocarpa and Cucumis dipsaceus (2 plants m-2) (Table 1). The highest weed frequency value at cotton flowering stage was recorded by (62%) Digitaria abyssinica and followed by (60%) Zaleya pentandra(L.) and (50%) Eriocloa fatmensis. Whereas the least weed frequency was recorded by (10%) launaea cornuta and Cucumis dipsaceus and followed by (12%) Amaranthus spinosus and Sorghum arundianaceum. The abundance value ranges from 2 to 38 plants m-2. The highest weed abundance value were recorded by 38 plants m-2 Xanthium strumarium and followed by 22 plants m-2 Cyperus esculentus and the least abundance value was(1 plant m-2)Datura stramonium and (2 plants m-2) Abutilo hirtum(Lam.), Commelina benghalensis and Digera muricata.

The highest frequency value of Gewane location at seedling stage is recorded by Zaleyapentandra (L.) (80%) followed by Echinocloa colane and Eriocloa fatmensis (78%), whereas the least frequency value recorded by Abutilo hirtum (Lam.) (4%) and followed by Argemone mexican (7%) and Gynandropsis gynandra (10%). The weed abundance at seedling stage was varied from 1 to 36 plants m-2. The highest weed abundance value (36 plants m-2) was recorded by Cyperus esculentus and followed by 17 plants m-2 Zaleya pentandra (L.). The least abundance value was recorded by Commelina benghalensis and Datura stramonium (1 plant m-2) and followed by Amaranthus spinosus, Abutilo hirtum(Lam.), Cucumis dipsaceus and Gynandropsis gynandra (2 plant m-2).

The highest frequency value at flowering stage was (74%) recorded by Digitaria abyssinica followed by Echinocloa colane (62%) and Zaleya pentandra (L.) (60%), whereas, the least frequency level is recorded by Prosopis juliflora (6 %). Weed abundance at flowering stage varied from 2 to 41 plants m-2. The highest weed abundance value was recorded by (41%) Digitaria abyssinica and followed by Corchorus trilocularis (38%) and Echinocloa colane (28%), whereas the least weed abundance was recorded by Ipomea ariocarp (2%) followed by Amaranthus spinosus, Convovulus arvensis and Sorghum arundianaceum (4%).

Therefore, this study showed that Echinocloa colane, Eriocloa fatmensis and Digitaria abyssinica weeds has higher frequency and dominance both at seedling and flowering cotton growing stage at Amibara district, whereas Digitaria abyssinica and Zaleya pentandra (L.) were the highest dominant weed at seedling and flowering cotton growing stage at Gewane district and these weeds are the major environmental and economic problems in the cotton field in the study area. Similar findings were done by (Roger N. et al 2015 and Gidesa A. et al 2016); they reported that if the specific plant species had higher frequency and dominance value, it indicates the economic importance of it.

	Table 2. Amibara Weed Frequency, Abundance and Dominance at Seedling and flowering cotton growing stage in 2016
			Frequency		Abundance		Dominance
No	Weed species	Family	At Seedling	At Flowering	At Seedling	At Flowering	At Seedling	At Flowering
1	Echinocloa colana	Poaceae	70	40	10	7	11	7
2	Digitaria abyssinica	Poaceae	78	62	9	16	9	15
3	Zaleya pentandra(l.)	Aizoaceae	56	60	4	11	5	10
4	Eriocloa fatmensis	Poaceae	72	50	7	8	9	7
5	Amaranthus spinosus	Amaranthaceae	62	12	5	3	5	1
6	Abutilo hirtum(Lam.)	Malvacea	24	0	2	0	2	0
7	Polypogon monspeliensis(L.)	Poaceae	8	0	2	0	1	0
8	Xanthium strumarium	Asteraceae	48	34	5	38	5	34
9	Convolvulus arvensis	Convolvulaceae	50	20	3	0	3	0
10	Prosopis juliflora	Fabaceae	30	0	2	0	3	0
11	Ipomea ariocarpa	Convolvulaceae	44	16	2	2	2	1
12	Cyperus esculentus	Cyperaceae	60	30	9	22	11	20
13	Launaea cornuta	Asteraceae	11	10	9	2	3	2
14	Cucumis dipsaceus	Cucurbitaceae	34	10	2	2	4	1
15	Sorghum arundianaceum	Poaceae	30	12	2	2	2	2
16	Corchorus trilocularis	Tiliaceae	74	34	7	4	7	4
17	Hibiscus trionum	Malvacea	40	15	4	2	9	1
18	Phyllanthus rotundifolius	Euphorbiaceae	56	0	4	0	4	0
19	Commelina benghalensis	Commeliaceae	26	0	1	0	2	0
20	Digera muricata	Amaranthaceae	38	0	3	0	2	0
21	Acalypha crenata	Euphorbiaceae	12	0	3	0	2	0
22	Datura stramonium	Solanaceae	24	26	1	1	1	1

	Table 3. Gewane Weed Frequency, Abundance and Dominance at Seedling and flowering cotton growing stage in 2016
			Frequency		Abundance		Dominance
No	Weed species	Family	At Seedling	At Flowering	At Seedling	At Flowering	At Seedling	At Flowering
1	Echinocloa colana	Poaceae	78	62	12	28	9	13
2	Digitaria abyssinica	Poaceae	74	74	10	41	7	24
3	Zaleya pentandra(l.)	Aizoaceae	80	60	17	24	13	9
4	Eriocloa fatmensis	Poaceae	78	32	13	7	10	5
5	Amaranthus spinosus	Amaranthaceae	22	26	1	4	2	2
6	Abutilo hirtum(Lam.)	Malvacea	4	0	1	0	1	0
7	Polypogon monspeliensis(L.)	Poaceae	48	0	3	0	2	0
8	Xanthium strumarium	Asteraceae	70	0	6	0	3	0
9	Convolvulus arvensis	Convolvulaceae	36	48	3	4	1	1
10	Prosopis juliflora(Sw.)	Fabaceae	0	6	0	2	0	1
11	Ipomea ariocarpa	Convolvulaceae	58	16	3	2	2	1
12	Cyperus esculentus	Cyperaceae	76	42	36	11	15	4
13	Cucumis dipsaceus	Cucurbitaceae	20	0	1	0	1	0
14	Sorghum arundianaceum	Poaceae	30	24	2	4	1	1
15	Corchorus trilocularis	Tiliaceae	30	56	2	38	2	37
16	Hibiscus trionum	Malvacea	66	0	4	0	1	0
17	Phyllanthus rotundifolius	Euphorbiaceae	68	0	7	0	5	0
18	Commelina benghalensis	Commeliaceae	38	0	2	0	1	0
19	Digera muricata	Amaranthaceae	52	0	5	0	3	0
20	Acalypha crenata	Euphorbiaceae	10	0	1	0	1	0
21	Datura stramonium	Solanaceae	40	58	2	11	2	9
22	Gynandropsis gynandra	Capparaceae	10	0	1	0	1	0
23	Tribulus terrestris	Zygophyllaceae	34	0	7	0	4	0
24	Amaranthus viridisHook.F	Amaranthaceae	42	0	6	0	4	0
25	Chenopodium ambrosiodes	Chenopodiaceae	48	0	3	0	2	0
26	Euphorbia microphylla H.	Euphorbiaceae	10	0	3	0	2	0
27	Argemone mexican	Papaveraceae	7	0	3	0	2	0

Table 4. Similarity Index of Weed species composition of cotton at Seedling stage at Amibara in 2016
	Badhamo	Wadullela	Ambash	Melke-Sedi			Billen
Badhamo	100	81.81	81.81	78.26			75
Wadullela		100	90	85.7			81.81
Ambash			100	85.7			85.7
Melke-sedi				100			78.26
Billen							100

Table 5. Similarity Index of Weeds of cotton at flowering stage at Amibara in 2016
	Badhamo	Wadullela	Ambash	Melke-Sedi			Billen
Badhamo	100	75	78.26	81.81			85.7
Wadullela		100	72	81.81			78.26
Ambash			100	78.26			81.81
Melke-sedi				100			85.7
Billen							100

Table 6. Similarity Index of Weeds of cotton at flowering stage at Gewena in 2016
	Amasabure	Deble	Eglay	Galeallo	Galeallo bora
Amasabure	100	86.67	81.25	86.67	76.47
Deble		100	72.22	76.47	68.42
Eglay			100	72.22	65
Galeallo				100	68.42
Galeallo bora					100

Table 7. Similarity Index of Weeds of cotton at flowering stage at Gewena in 2016
	Amasabure	Deble	Eglay	Galeallo		Galeallo bora
Amasabure	100	73.33	64.7	73.33		64.7
Deble		100	68.75	73.33		64.7
Eglay			100	68.75		61.11
Galeallo				100		74.7
Galeallo bora						100

Table 8. Similarity Index of Weeds of cotton at Amibara and Gewena in 2016
	Seedling Stage
Survey Sites	Amibara	Gewena
Amibara	100	71.42
Gewena	71.42	100
	Flowering Stage
Amibara	100	68.75
Gewena	68.75	100

CONCLUSION

Based on the weed survey, a total of twenty seven different weed species from seventeen families were identified. The importance of each weed species was determined by calculating the frequency, abundance and dominance values. This experiment identified a large and diversified weed species that were found in the study area. The most dominant weed families according to the frequency and number of weed species were Poaceae, Amaranthaceae, Euphorbiaceous, Malvacea, Convolvulaceae and Asteraceae. Weed species composition were similar between districts and across districts. The study shows that the two districts have similar weed species composition and hence, it is possible to apply the same weed management practices for the study areas.

REFERENCE

Basra, A. S. (2002). Cotton Fibers Developmental Biology Quality Improvement, and Textile Processing (ed.). CBS. New Delhi. 387 p.

Benedict, J.H. and D.W. Altman (2001). Commercialization of transgenic cotton expressing insecticidal crystal protein. p. 137-201. In J.J. Jenkins and S. Saha (ed.) Genetic improvement of cotton: Emerging technologies. Science Publ., Enfield, NH.

Bishnoi, L.K., Panwar, R. S Malik, R.K. and Rathi, S.S. (1993). Effect of varieties and weed free maintenance period on weed competition in cotton. In: Integrated weed management for sustainable agriculture. Proc. of the Indian Society of Weed Science, International symposium, held on the 18-20 November. Hisar, Haryana, India; 3:182-183.

EIAR (Ethiopian Institute of Agricultural Research) (2016). National Cotton Research Program Strategy for Fifteen Years (2016-2030). Developing national cotton commodity long term (15 years) research strategy consultative workshop by cotton research core team members. October 24-29, 2016. Addis Ababa, Ethiopia. Pp. 26-33.

Esayas T. and Abraham G.H. (2000). Quantitative and qualitative survey of weeds growing in association with cotton in the Middle and the Lower Awash. Progress report for the period of 2000/2001: Werer Agricultural Research Center.

Esayas T., Abraham G.H. and Mashila D. (2013). Ethiopian Journal Applied Science Technology 4: 41-49.

Esayas T., Abraham G.H. and Mashila D. (2012).Ethiopian Journal of Applied Science Technology 3 (1): 57- 69.

Gidesa A, Tadesse T, Dabi A (2016). Quantitative Determination of Weed Occurrence on Upland Rice of Bambasi, Ethiopia. Ecology and Evolutionary Biology 1(3): 53-56.

Kedir N., Feyissa T. and Tilahun G. (1999a). Results of weed survey in the major barley- and wheat growing areas of the Bale highlands. Arem 5: 85-95.

Kedir N., Feyissa T. and Tilahun G.(1999b). Results of weed survey in the major field pea and faba bean growing-areas of the Bale highlands. Arem. 5: 109-121.

Keeley, P.E. and Thullen, R.J. (1991). Growth and interaction of bermudagrass (Cynodondactylon) with cotton (Gossypium hirsutum L.). Weed Sci., 39(4): 570-574.

Khan, N.U. and Khan, S. U. (2003) Integrated weed management in upland cotton. Pak. J. Weed Sci. Res., 9(3 and 4):185-192.

MOA (Ministry of Agriculture) (2010). Plant Health Regulatory Directorate List of Registered Pesticides. April, Addis Ababa, Ethiopia.

Mofett, J.E. and Mcclosky, W.B. (1998). Effects of soil moisture and yellow nut sedge (Cyperus esculentus) density on cotton (Gossypium hirsutum L.). Weed Sci., 46(2): 231-237.

MOI (Ministry of Industry). 2015. National Cotton Development strategy for Ethiopia (unpublished).Addis Ababa, Ethiopia.

Pohlan, J. (1984). Arable farming and weed control. Institute of Tropical Agriculture, Plant Production Section, German Democratic Republic.141 pp.

Roger N, Micheal DK Owen, Swanton CJ (2015). Weed Abundance, Distribution, Diversity, and Community Analyses. Weed Science 63: 64-90

Smith, K.L. (2000) Controlling weeds in cotton. In: Proceedings of the 2000 Cotton Research Meeting. AAES: Special report, 33-37.

Stroud A and Parker C. (1989). A Weed identification guide for Ethiopia.FAO, Rome, Italy.

Takim FO and Amodu AA (2013). Quantitative Estimate of Weeds of Sugarcane (Saccharum Officinarum L.) Crop In Ilorin, Southern Guinea Savanna of Nigeria. Ethiopian Journal of Environmental Studies and Management 6 (6): 611-619.

Taye T. and Yohannes L. (1998). Qualitative and quantitative determination of weeds in teff in west Shewa Zone, Arem, 4: 46-60.

Travlos IS, Cheimona N, Roussis I and Bilalis DJ (2018). Weed-Species Abundance and Diversity Indices in Relation to Tillage Systems and Fertilization. Front. Environ. Sci. 6:11. doi: 10.3389/fenvs.2018.00011

Werer Agricultural Research Center (WARC), (2000). Strategies and priorities for cotton research.