Livestock Research for Rural Development 24 (10) 2012 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Determining suitable size of stem cutting for propagating Becium grandflorum

Haftom Gebremedhn and Tesfay Belay

Mekelle Agricultural Research Center, Tigray Agricultural Research Institute,
P.O.Box 492, Mekelle, Tigray, Ethiopia.


Becium grandiflorum is an endemic plant to the highlands of Ethiopia and Eritrea. It is one of the most important honeybee plants in the Tigray region of northern Ethiopia.  Honeybees highly visit the flowers of the plant for pollen and nectar. However, B. grandiflorum is currently threatened due to population pressure and the associated farmland expansion. For rapid propagation of the plant using stem cutting, optimum cutting size was not identified. This study was therefore designed to determine optimum cutting size of stem cuttings of B. grandiflorum.  The study was conducted at Mekelle Agricultural Research Center. Six stem cutting sizes (10cm, 20cm, 30cm, 40cm, 50cm and 60cm) were considered as treatments. They were replicated  three times and design used was randomized complete block design.

The 20cm sized stem cuttings gave maximum flower number while 60 cm long cutting had the least flower number. This indicates that cuttings bigger than 20 cm would result in lower number of flower per plant and thus is not recommended.

Key words: Flower number, honeybees


Favorable and diversified agro-climatic conditions of Ethiopia create environmental conditions conducive for the growth of over 7000 species of flowering plants that supported large number of bee colonies in the nation (Gidey and Mekonen 2010). Tigray region of northern Ethiopia is one of the potential beekeeping areas. In this region apiculture is a good source of income for smallholder farmers. Honey from this region, Tigray white honey, is very popular and has high demand in the local and international markets (UNIDO 2009).  Locally it is the most expensive (Taddele and Nejdan 2008) and often comes from specific districts in the region like Hawzien, Atsby-Wonberta, Adigrat and Woukro (Taddele and Nejdan 2008) (Fig 1). These districts are  the main habitats of Becium grandiflorum (Lam.) (Haftom et al 2011; Alemtsehay 2011). The honey from this bee forage is preferred by consumers because of its attractive color and also pleasant taste. Honey from these districts gets premium price in the local and international markets (Taddele and Nejdan 2008).

B. grandiflorum is one of the regionally important honeybee plants of Tigray region. It is an endemic plant to the highlands of Ethiopia and Eritrea (Bein et al 1996; Fichtl and Admassu 1994; Guinad and Dechassa 2001). It is locally known as Tebeb. It is a shrub or sub shrub. Honeybees frequently visit the flowers of the plant for collecting pollen and nectar (Haftom et al 2012; Fitchel and Admasu 1994). Hence the plant is ranked as the best honey bee forage by beekeepers and other extension workers (Haftom et al, 2011; Alemtsehay 2011).  Despite the fact that Tigrian white honey is high in quality and gets better price, its quantity is limited (Taddele and Nejdan 2008).

The species has also additional value as fuel wood and  human food where the flowers are  plucked and eaten fresh,  broom for cleaning threshing ground, roofing of traditional houses, food flavoring, traditional medicine against malaria, and for soil and water conservation (Guinad and Dechassa 2001; Nurya 2010). This is done by uprooting the plant from the ground. Hence, the density of B. grandiflorum in its natural habitat is declining. Other possible factors for decline in density could be  increased livestock population, natural resource degradation and farmland expansion.

 In addition, natural regeneration of the plant from seed is being restricted by widespread human interference and there are no new B.grandiflorum plantations and the result is there is shortage of bee forage in the region. To minimize the shortage of bee fodder, some beekeepers have started to propagate the plant using stem cutting. Propagation by stem cuttings is the preferred method for propagation of B.grandiflorum because it is the easiest and cost-effective of other options (Verheij 2004). Wider use of the propagation technique was hindered by lack of knowledge on the suitable size of the stem cutting. Besides, no research has been conducted on the technical and economic feasibility of the propagation for large scale purposes.


The major objective of this study was therefore to determine the suitable size of stem cutting of B. grandiflorum for efficient propagation so as to recommend for farmers involved in small and large scale honey bee production.  

Figure 1. Location map of major B.grandiflorum growing areas and white honey producing districts (shaded areas) of Tigray Region of Ethiopia. (Source:

Materials and Methods

Description of study areas

The study was carried-out at Mekelle agricultural research center, Illala site. Illala is located North-east of Mekelle at an elevation of 1970 meters at 250 51’N latitude and 390 61’ longitudes. The predominant soil type in the study area is vertisol. 


Cuttings collection and preparation


 Young and healthy branches were collected from 35 mother plants which had similar performnce and are beleived to come from the same source. Stem cutting of six sizes : 10, 20, 30, 40, 50, and 60 cm were prepared. The cuttings were uniform in diameter and age so as to avoid variability among treatments due to these traits.  


In the nursery 50 plastic sleeves were prepared for planting each cutting size. Plastic sleeves were 15 cms long and 8 cm wide. The sleeves were filled with a soil mixture of silt, sand and manure at the ratio of 20: 2: 5, respectively. Plastic sleeves filled with the soil mixtures were arranged in the open air and were watered with tap water. Cuttings were then planted directly after the preparation of the cuttings.


The cuttings had adequate number of green leafs (5 to 12) to initiate photosynthesis. The green leafs were partially injured /cut/ to stimulate new growth. To avoid the entrance of water during planting and growing period, the cuttings were prepared in such a way that the top ends have a slant surface (angle of 45°) vertically, and contain a minimum of two nodes. Seedlings were transplanted to the experimental plot after 55 days.  




Seedlings were transplanted early in the morning and planted on the same day to minimize moisture stress. All the seedlings from the different cutting size were planted in pits with a diameter and depth of 20 cms. After transplanting the seedlings on the experimental plot, supplementary watering was provided once a week when there was no rain falls from June to September. 


Experimental design


The study had six treatments:  10, 20, 30, 40, 50, and 60 cms long cuttings of the stem. Each cutting size was considered as a treatment. Each treatment was replicated three times and the design was randomized complete block. Plants and rows were spaced 1.5 m while plots and blocks were 2 meters apart. 


Data collection and analysis  


To collect data five plants were taken randomly from each plot.  Data collected included canopy cover, number of flowers and branches/plant, height, and seed yield per plant.


Canopy cover was calculated by using the formula


C.C = (D1 +D2)/2,


where D1 is diameter of the plant towards the larger canopy coverage and D2 is diameter of the plant towards the small canopy coverage and C.C reprsents  canopy cover of the plant in cm. So the canopy cover of the plant will be expressed in terms of the average diameter in cm.


The total number of flowers per plant was calculated by counting the total number of flower heads per plant and number of flowers per flower head. The total number of flowers per plant was then calculated by the formula


 T.F = H.F *N.F.H,


where T.F refers to the total number of flowers per plant, H.F to the number of flower heads per plant and N.F.H is number of flowers per flower head. To know the number of flowers per flower head, a sample of 10 flower heads per plant was taken randomly. 


The height of the plant was measured from ground to the tip of the longest branch with the help of a measuring tape. Actual number of branches and seed per plant were determined by direct counting and recording.


The collected data was statistically analyzed using the one-way ANOVA analysis of variance procedure and least significant difference test was calculated to identify significant difference among the treatments using the statistical programme GENSTAT (GENSTAT©, Version 13). Correlation was done for specific variables using SPSS version 16.

Photo 1. Seedlings and matured plants of B.grandiflorum in  nursery and experimental site (A,  B.grandiflorum seedling in nursery : B, matured B.grandiflorum: and  C,  honeybee visiting B.grandiflorum flowers)

Result and Discussion

The result of variance analysis showed that there was significant difference (P< 0.001) among the treatments.  Mean values for number of branches, canopy cover, and height are also presented in table 1.Cuttings size of 20 cm had the highest number of branches (9.3) and greatest canopy cover (129.6 cm) that was significantly different from all the cutting sizes except 10 cm. Beside plants from cutting size of 60 cm had the smallest branch number (4.1) and canopy cover (103.6 cm). The difference on canopy cover between the treatments might be due to difference in the number of branches. Because canopy cover and number of branches per plant had significant positive correlation (r= +0.41, P<0.001) (Table 2).

Cutting size of 30 cm had the highest plant height (114.7cm) that was significantly different from 10, 20, 40 cm sized cuttings. The correlation between number of branches and plant height was negative and significant (r= -0.27, P<0.01) showeding a weak correlation between the characters. Moreover plant height had weak negative correlation with canopy cover (r = -0.09, P>0.05) (Table 2). Similarly Bal (2005) reported negative but highly significant correlation between plant height and number of flowers per plant on Tartary buckwheat.


Regarding the number of flowers and seeds per plant, highly significant differences (P< 0.001) was found between them. 20 cm cutting size had the highest number of flowers (7859) and seeds (3504) per plant, while 60cm had the smallest number of flowers (5335) and seeds (1682) per plant. This might be due to the difference in the number of branches amongst the treatments. Rajesh (2010) mentioned that the flowering pattern and flowering performance of a plant are considered to be the sum of all the genetic, physiological and morphological traits of a specific variety. John et al (1987) also revealed that a plant with more vegetative growth develops more flowers and seeds. The poor performance of the largest cutting size (60 cm) might be attributed to the woody nature of the cutting that might have converted most of the food materials to lignin which later on resulted to lower rooting and shooting. This result was similar to the study conducted by Kathiravan et al (2009) on Jatropha species.  


The correlation between number of flowers and branches was positive and significant (r= +0.54, P<0.001) (Table 2). Rasoul et al (2012), Rajesh (2010) and Bal (2005) also reported significant positive correlation between number of flowers and branches in Matricaria chamomilla  L., Linum usitatissimum L. and Buckwheat species, respectively. There was no correlation between  plant height and number of flowers per plant (Table 2). Similarly Muhammad et al (2006) reported non-significant but weak negative correlation between plant height and number of flowers in Rosa Species.

Table 1. Mean values for canopy cover, number of branches and flowers, plant height, and seed per plant of B.grandiflorum for different cutting sizes.














114 db

119 cb

112 edc

104 fdc






9.3 a


5.1 cd

4.4 de

4.1 ef















7049 ab

7859 a

6844 bc

5412 e


5335 ef





2870 b

3504 a

2544 bc

2225 cd

2034 de

1682 f




Means in rows followed by the same letter are not significantly different at 0.01 probability level.

 C.C,  canopy cover in cm; N.B, number of branches;, H, height of plants in cm; F.N, total number of flowers per  plant; S.P,  total seed per plant.

Table 2. Correlation matrix between variables: canopy cover, number of branchs, flowers and seeds/plant and plant height.





































**, significant at the P<0.01; ***, significant at p<0.001.

Significantly positive correlation was obtained between number of seeds and the number of branches per plant (r = 0.61, P <0.001) (Table 2) indicating higher association between the two characters. Similarly positive correlation was reported between grain yield and number of branches per plant on soybean (Muhammad et al 2007).

Conclusions and recommendation


We are grateful for the financial and material support provided by the Mekelle Agricultural Research Center of the Tigray Agricultural Research institute. Further we would like to thank the apiculture and sericulture case team members of Mekelle agricultural research center for their technical contribution. We are also very grateful to Kinfe Mezgebe of Mekelle agricultural research center coordinator and Gebre Hadgu from Humera agricultural research center for their very kind support throughout the study.


Alemtsehay T, Yosef M, and Gebremedhin W 2011 Seasonal availability of common bee flora in relation to land use and colony performance in Gergera watershed Atsbi Wembwrta district eastern zone of Tigray, Ethiopia, M.Sc. Thesis. Wondo Genet College of Forestry and Natural Resources, School of Graduate Studies, Hawassa University Wondo Genet, Ethiopia.


Bal K J 2005 Correlation, regression and path coefficient analyses for some yield components in common and Tartary buckwheat in Nepal.


Bein E, Habte H, Jaber A, Birnie A and Tengnas B  1996  Useful trees and shrubs in Eritrea. Technical hand book No 12.


Fichtl R and Admasu A 1994 Honeybee flora of Ethiopia. The national herbarium, Addis  Ababa university, Deutscher Entwicklungsdienst (DED).  Margraf verlag, Germany. 


Gidey Y and Mekonen T 2010 Participatory Technology and Constraints Assessment to Improve the Livelihood of Beekeepers in Tigray Region, northern Ethiopia. CNCS Mekelle University. Volume 2 (1): 76-92.


Guinad Y and Dechassa L 2001 Wild-food Plants in Ethiopia. Reflections on the role of 'wild-foods' and  ‘famine-foods’ at a time of drought. United Nations Development Programme, Addis Abeba, Ethiopia.


Haftom G, Zelealem T,  Girmay M,  and Awet E  2011 Honeybee flora seasonal availability and honeybee seasonal activity and managemental interaction in Debrekidan and Begasheka watershed, Tigray, Ethiopia. Proceedings of the workshop on accomplishments of the operational research project (2008- 2011), April 29-30, 2011, Mekelle, Ethiopia, pp. 72-84


Haftom G N, Zelealem T G and Girmay M W 2012 Honeybee flora and floral calendar of Northen Ethiopia. LAP LAMBERT Academic Publishing GmbH and Co.KG and licensors. ISBN: 978-3-8473-7729-0.


John B A, Gordon R H and Parrish D J 1987 Plant Science. McGraw-Hill publishing Company.126582.


Kathiravan M  Ponnuswamy A S and Vanitha C 2009 Determination of suitable cutting size for vegetative propagation and comparison of propagules to evaluate the seed quality attributes in Jatropha curcas Linn. Natural product Radiance, Vol. 8(2), pp,&162-166.


Muhammad S, Adnan Y, Muhammad A K, Asif A K and Atif R 2006 Short communication correlation studies in Rosa species under faisalabad (pakistan) conditions. Journal of agriculture &social sciences 1813–2235/2006/02–1–58–59.


Muhammad F A M, Muhammad A, Afsari S Q And Abdul G 2007 Assessment of genetic variability, correlation and path analyses for yield and its components in soybean,  Pak. J. Bot., 39(2): 405-413.


Nurya A 2010 Ethnobotanical study of medicinal plants used by local people in Ofla wereda, southern zone of Tigray region, Ethiopia ,Addis Ababa University.


Rajesh K J 2010 Correlation study of flowering performance and pattern with the yield in Linumusitatissimum L. African Journal of Plant Science Vol. 5 (3), pp. 146-151.


Rasoul M, Hamid D and Hossein Z 2012 Interrelationships among flower yield and related characters in chamomile populations (Matricaria chamomilla L.) Journal of Medicinal Plants Research Vol. 6(19), pp. 3549-3554.


Taddele A and Nejdan Y 2008  Beekeeping and honey production in the Tigray region, Ethiopia. Report. African Services Committee Ethiopia. 


UNIDO (The United Nations industrial development organization) 2009 Millennium cities initiative (MCI) investment opportunities for development, Mekelle, Ethiopia. Meklle investment profile.


Verheij Ed 2004 Agrodok 19 Propagating and planting Trees.  ISBN: 90-77073-99-X, NUGI: 835

Received 22 August 2012; Accepted 10 September 2012; Published 1 October 2012

Go to top