Sorghum (Sorghum bicolour (L.) Moench) is an important staple crop in Sub-Saharan Africa (SSA) that can meet the increasing demand of food  . Although, sorghum consumption is high in most SSA countries, the grain yield at the farm level is low due to the effect of biotic and abiotic stresses   .
The obligate root-parasitic flowering plant Striga hermonthica affects the lives of over 100 million people and infests about 40% of arable land in the savanna region  . Striga causes 75% of its damage before it emerges above the ground making its control more difficult  . Mechanical and chemical control options are less effective because they affect Striga after it has already attached and damaged the host  . Different control measures such as hand weeding, crop rotation trap crop, catch crops, intercropping, fertilizers and herbicides have also been suggested but with limited success. The many herbicides that have been tried have not been effective, and are costly and in most cases may not be available to resource-poor farmers in SSA.
In Eritrea, Striga hermonthica affects the majority of farmers especially in the western part of the country, where continuous mono-cropping is practiced  . A report by the African Agricultural Technology Foundation  indicated that 30,000 to 90,000 tonnes of grain sorghum is lost annually due to Striga in Eritrea. Annual yield losses due to Striga in neighbouring countries, for example, Sudan, Ethiopia, Kenya and Uganda are estimated at 1,060,000, 500,000, 50,000 and 40,000 tonnes respectively  . To minimize such yield losses, there is a need to devise control measures against the parasite.
In the past, crop improvement efforts have concentrated on host plant resistance as means of breeding against Striga. The use of resistant variety is considered to be more efficient and practical option for controlling Striga infestation. However, conventional breeding against the parasite has been slow and arduous  . A combination of host plant resistance mechanisms with molecular marker assisted selection (MAS) application will most likely yield promising results as shown in previous experiments  .
Several mechanisms of resistance to Striga in sorghum have been reported that pro- bably operate singly or in various combinations   . Using in-vitro laboratory tech- niques, four specific mechanisms of resistance to Striga which included low production of germination stimulant, low production of the haustoria initiation factor, hypersensitive response, and incompatible response were reported in cultivated sorghums and some wild accessions   .
Low germination stimulant variants of sorghum produce insufficient amounts of the exudates required for germination of conditioned Striga seed. Reduction in amounts of germination stimulants produced by host plants provides the means to reduce numbers of seeds germinating  . Low or no stimulant production by cereal roots has been reported to be a mechanism of host plant resistance to S. hermonthica infections   . Sorghum variants that produce low levels of the germination stimulants have been found to be resistant to Striga in field tests  . Highly susceptible sorghum variants appeared to be high producers of the germination stimulants  . This study tested the germination stimulant production reaction of landraces from Eritrea and that of commercial cultivars and identified genotypes with low levels that may be described as having resistance to Striga.
2. Materials and Methods
2.1. Plant Materials
Seeds of Striga hermonthica were obtained from Kenya Agricultural and Livestock Research Organization (KALRO) sub-station Kibos. They were collected in 2011 from sorghum growing fields at Kibos (00˚04'S, 34˚48'E, 1214 m altitude) using standard protocols  . At the time of use the Striga seeds were 4 years of age. Sorghum land- races were sourced from National Agricultural Research Institute (NARI) of Eritrea which was collected from sorghum growing zones of Gashbarka, Anseba, Southern zo- ne and Northern red sea regions of the country  . Elite backcross lines, improved va- rieties and commercial checks were included in the experiment as indicated in Table 1.
2.2. Striga Seed Conditioning
Striga hermonthica seeds, to respond for a germination stimulant, have to be conditioned by exposing them to favorable moisture and temperature for two weeks  . To condition Striga seeds, they were initially surface disinfected for 5 minutes in a mix of 1% sodium hypochlorite containing 0.02% (v/v) Tween 20  . Floating seeds and debris were discarded. The remaining seeds were rinsed using sterile distilled water and
Table 1. Summary of sorghum germplasm used in the study.
NARI = National agricultural research institute, ICRISAT = International Crops Research Institute for the Semi- Arid Tropics.
later air dried under laminar flow hood. Moistened double layer of 90 mm diameter Whatman no.1 filter papers were placed in a 90 mm sterile petridish. The air dried Striga seeds were sprinkled on the glass- fiber discs (Whatman GF/C) so that each disc had 20 - 30 Stiga seeds and then incubated at 30°C for 14 days   .
2.3. Experiment Setup
The experiment was conducted in laboratory and screen house at BecA-ILRI Hub, Nairobi, Kenya. Each sorghum accession was planted in a screen house in a 10 cm diameter pot containing sand that was sterilized in a preheated oven at 85˚C for 30 minutes. Each pot carried 8 - 10 plants which allowed harvesting at least 1 gram of root. Planting was done at the same date where Striga seeds were placed in an incubator for conditioning to synchronize for maximum stimulant production which occurs during the early stage of root development  . The seedlings were grown for two weeks. The two weeks old sorghum seedlings were then gently removed from the pot and the roots washed.
For testing germination of Striga seeds, the washed roots were cut in to small pieces of about 0.5 cm and 1gram was weighed. Four radial rows of fiber-glass-discs cont- aining conditioned Striga seeds were arranged around 1.5 cm diameter aluminum foil ring centered on double layer of Whatman no.1 filter paper moistened with 3 ml of double distilled water in a 90 mm petridish  . Then 1 gram of the cut root pieces was placed in the aluminum foil ring and 3 ml of double distilled water added to defuse root exudates across the filter paper as described by   . GR24 and double distilled water were used as positive and negative controls, respectively. The Petri dish was then sealed using parafilm then wrapped with aluminum foil and placed at 30˚C for 48 hours in an incubator for Striga germination  . GR24 is a synthetic germination stimulant which is available commercially, is a chemical analog of strigolactones. The stock was prepared as 100mg of GR24 in 10ml of acetone and then diluted with sterile distilled water, a 1 litter stock solution (100 mg・L−1) was made and used at a final concentration of 0.01 mg・L−1.
2.4. Data Recording and Analysis
Following after 48 hours of receiving the Striga germination stimuli, Striga germination count was done under dissecting microscope by counting the number of Striga seeds in each fiber glass discs that had germinated as described by  . A seed was considered as germinated if the radicle was seen protruded through the seed coat.
Percentage germination of Striga hermonthica seeds were calculated for each treatment. Analysis of variance (ANOVA) was carried out using Genstat®15th Edition
(http://www.vsni.co.uk). Treatment means were separated using the least significance difference test at 5% level. Statistical analysis for percent Striga germination data was performed after logarithmic transformations using the formula (log (X + 1), where X is the original individual observation)  . Correlations between percent Striga germination and distance from the source of Striga germination stimulant were also performed.
3. Results and Discussion
All sorghum accessions used in this study germinated well in the pots. This enabled the harvesting of at least 1gram of root from each accession which was required as source of Striga germination stimulant in the study. As defined by Ramaiah et al. 1990,  the term stimulant, refers to that component of the sorghum root exudates that germinate the strain of Striga hermonthica. Analysis of variance for Striga germination revealed that highly significant differences (P < 0.001) were observed among the sorg- hum accessions tested for their ability to cause Striga germination with a range of 11.8 to 40.6% (Table 2). Striga seeds germinated at different levels along the radial position in the petri dish in all sorghum variants, indicating the presence of different levels of germination stimulants. This is in agreement with the work of Karaya et al. 2012, who studied the variability of Striga germination stimulant levels in maize  .
Accession EG1168 stimulated the highest germination of Striga seeds (40.3% ± 4.9) compared to the rest of accessions. On the contrary accession, EG830 induced the lowest level of Striga germination (11.85% ± 2.4). Such low Striga germination percent may indicate a potential for resistance to Striga. No Striga germination was observed in the negative control (double distilled water) while the positive control GR24 exhibited 43.73%, which was not significantly different from germination observed with sorghum accession EG1168 (40.6% ± 4.9). However, all the rest of the sorghum accessions indu- ced significantly lower Striga germination compared to the GR24. Similar results were reported by  .
The top 10 genotypes induced less than 18% Striga germination (Figure 1), while the commercial checks, IS9830, SRN39 and Framida caused 22.46, 22.67 and 23.27% ger- mination, respectively. No significant differences were observed among these commercial checks. However, Striga germination in at least one of the tested landraces, namely accession EG830 had significantly lower (Prob ≤ 0.05) germination than that of the commercial varieties. The five sorghum accessions with the lowest Striga germination were EG830, EG1076, EG473, EG 1261 and EG546 which caused Striga germination percentages of 11.85, 13.05, 14.68, 15.32 and 15.74, respectively. Even though these five accessions did not show total immunity against Striga seed germination, as there is no reported complete resistance to Striga so far in sorghum  , the expression of low percentage level of stimulant production was an indication of their high level of resistance to Striga. Low Striga germination suggests low germination stimulant production. Low level of germination stimulant produced by host plant may result in reduced number of germinated Striga seeds. However, low germination could also be due to some germination-inhibitory compounds produced by the sorghum accessions that may interfere with the germination response sequence of conditioned Striga seeds as reported by  .
The level of Striga germination and the distances from which stimulants where released is shown in Figure 2. Germination percent was high near to the source of stimu-
lant, which suggests that the higher the concentration of the stimulant, the higher the Striga germination percent. As the distance from the source of Striga stimulant increased, the germination percent was significantly reduced to below 15%. In this study,
Table 2. Levels of Striga germination percent exhibited by the sorghum accessions tested.
*** = highly significant (P < 0.001), L.S.D = least significant difference, CV = coefficient of variation, AN = Anseba, GB = Gash Barka, NRS = Northern red sea, S = South.
Figure 1. Percent Striga seed germination category of sorghum accessions and their control.
Figure 2. Correlation between Striga percent seed germination and the distance (mm) from the source of Striga germination stimulant.
the highest germination was recorded on discs which were nearer to the source of stimulant compared to those farther off. Highly significant (P < 0.001) and positive correlation coefficients were observed between Striga germination and the distances from the source of the stimulant. An indication that the closer the Striga seeds to the source of stimulant the higher the amount of seeds stimulated to germinate and vice versa. This result corroborates previous work on variation in Striga germination stimulants production in maize  . Similarly, reports by  indicated that germination stimulant produced by the host plant is mainly exuded in a distance close to radius from the root apex. Support for this spatial relationship between host roots and Striga seed germination as a function of the distance from the host root to where germination stimulant is active to elicit germination was documented  .
The regression equation y = −1.4576x + 42.67 in Figure 2 implies that for every unit increase of distance from the stimulant, the germination percent of the Striga seed is expected to decrease by about 1.4576 percent. The negative slope of the fitted line in Figure 2 also suggests that decrease in Striga germination percent were associated with increased distance from the source of Striga germination stimulant. The high coeffi- cient of determination (R2 = 0.998) indicates the variation in germination percentage was almost all explained by the variation in the distance of concentration of Striga germination stimulants.
In sorghum, four compounds of root exudates which include sorgoleone, sorgola- ctone, strigol and a water-soluble compound with a quantitative biosynthetic pathway are reported as germination stimulants  . Therefore, it is possible that these stimulants were also produced by the accessions used here.
Low Striga germination levels observed in some of the accessions tested in this study may be due to low production of germination stimulant, which is one of the best known mechanisms of resistance in Striga  . This low germination stimulant production is of special interest in breeding for resistance to Striga in sorghum. Low induction of seed germination has been successfully used in sorghum breeding for resistance to Striga hermonthica  . Ejeta and coworkers selected sorghum lines with reduced induction of germination in their breeding programs  . A wide range of sorghum of low stimulant lines has shown resistance in the field which indicates the usefulness of low stimulant form of resistance  . Identification of genotypes with low germination stimulant from the current study will play a crucial role in the improvement of sorghum cultivars for Striga resistance. Since the identified accessions are landraces which are adapted to the local environmental conditions of the country, they can be included directly in the sorghum breeding program for Striga resistance.
The accessions with low Striga germination stimulant producers identified in this study, namely EG830, EG1076, EG473, EG 1261 and EG546 caused lower germination percent of Striga compared with the commercial controls. These accessions may be useful potential sources of resistance to Striga as such or in a backcross breeding pro- gram. It would be interesting to confirm whether the mechanical type of Striga resistance that has been mapped using Quantitative Trait Loci (QTL) and reported elsewhere  would be found in genotypes with low stimulants production. In order to consolidate this resistance, these accessions of low stimulant production could be crossed with the already identified backcrosses with intro-gressed Striga resistance QTL from a previous study  . Such resistance to Striga in sorghum, resulting from a combination of two mechanisms, would be more durable and stable across ecological zones than one based on single gene resistance sources.
This project was supported by the BecA-ILRI Hub through the Africa Biosciences Cha- llenge Fund (ABCF) program. The ABCF Program is funded by the Australian Department for Foreign Affairs and Trade (DFAT) through the BecA-CSIRO partnership; the Syngenta Foundation for Sustainable Agriculture (SFSA); the Bill & Melinda Gates Foundation (BMGF); the UK Department for International Development (DFID) and; the Swedish International Development Cooperation Agency (Sida).
 Mohamed, A., Ali, R., Elhassan, O., Suliman, E., Mugoya, C., Masiga, C.W., Elhusien, A. and Hash, C.T. (2014) First Products of DNA Marker-Assisted Selection in Sorghum Released for Cultivation by Farmers in Sub-Saharan Africa. Journal of Plant Science and Molecular Breeding, 3, 2050-2389.
 Mohamed, A.H., Gamar, Y.A., Elhassan, O.M., Elgada, M.H. and Assar, A.H.A. (2011) Performance of Selected Sorghum Genotypes under Rain-Fed Areas of Sudan. Life Science International Journal, 5, 2119-2127.
 Mohamed, A.H. and Gamar, Y.A. (2011) Climate Change: Selection of Sorghum Genotype with Wide Adaptation, AG-17, for Rain Fed Areas of Sudan. International Journal of Agricultural Research, 1, 143-155.
 Tadesse, Y., Tesfamichael, A., Kiambi, D., Folkertsma, R., Hash, C.T., Ngugi, K., Mutitu, E., Negusse, A., Mussie, W., Mugoya, C., Masiga, C.W. and Santie de Villiers (2015) Marker-Assisted Introgression Improves Striga resistance in an Eritrean Farmer-Preferred Sorghum Variety. Field Crops Research, 173, 22-29.
 Ejeta, G., Butler, L.G., Hess, D.E. and Vogler, R.K. (1991) Genetic and Breeding Strategies for Striga Resistance in Sorghum. In: Ransom, et al., Eds., Proceedings of 5th International Symposium on Parasitic Weeds, Nairobi, 539-544.
 Haussmann, B.I.G., Hess, D.E., Omanya, G.O., Folkertsma, R.T., Reddy, B.V.S., Kayentao, M. and Geiger, H.H. (2004) Genomic Regions Influencing Resistance to the Parasitic Weed Striga hermonthica in Two Recombinant Inbred Populations of Sorghum. Theoretical and Applied Genetics, 109, 1005-1019.
 Karaya, H., Njoroge, K., Mugo, S., Ariga, E.S., Kanampiu, F. and Nderitu, J.H. (2012) Determination of Levels of Striga Germination Stimulants for Maize Gene Bank Accessions and Elite Inbred Lines. International Journal of Plant Production, 6, 209-224.
 Heller, R. and Wegmann, K. (2000) Mechanisms of Resistance to Striga hermonthica (Del.) Benth in Sorghum bicolor (L.) Moench. In: Haussmann, D.E.H.B.I.G., Koyama, M.L., Grivet, L., Rattunde, H.F.W. and Geiger, H.H., Eds., Breeding for Striga Resistance in Cereals, Margraf-verlag, Weikersheim, 19-28.
 Ayongwa, G.C., Stomph, T.J., Emechebe, A.M. and Kuyper, T.W. (2006) Root Nitrogen Concentration of Sorghum above 2% Produces Least Striga hermonthica Stimulation. Annals of Applied Biology, 149, 255-262.
 Berner, D.K., Winslow, M.D., Awad A.E., Cardwell, K.F., Mohan Raj, D.R. and Kim S.K. (1997) Striga Research Methods—A Manual. International Institute of Tropical Agriculture PMB 5320, Ibadan, Nigeria.
 Tesfamichael, A., Githiri, S.M., Kasili R.W., Skilton, R.A., Solomon, M. and Nyende, A.B. (2014) Genetic Diversity Analysis of Eritrean Sorghum (Sorghum bicolor (L.) Moench) Germplasm Using SSR Markers. Molecular Plant Breeding, 5, 1-12.
 Berner, D., Carsky, R., Dashiell, K., Kling, J. and Manyong, V. (1996) A Land Management Based Approach to Integrated Striga hermonthica Control in Sub-Saharan Africa. Outlook on Agriculture, 25, 157-164.
 Rodenburg, J., Bastiaans, L., Weltzien, E. and Hess, D.E. (2005) How Can Selection for Striga Resistance and Tolerance Based on Field Observations Be Improved? Field Crops Research, 93, 34-50.
 Hess, D.E., Ejeta, G. and Butler, L.G. (1991) Research into Germination of Striga Seeds by Sorghum Root Exudates. Proceedings of the 5th International Symposium of Parasitic Weeds, Nairobi, 24-30 June 1991, 217-222.
 Vogler, R.K., Ejeta, G. and Butler, L.G. (1996) Inheritance of Low Production of Striga Germination Stimulant in Sorghum. Crop Science, 36, 1185-1191.
 Haussmann, B.I.G., Hess, D.E., Reddy, B.V.S., Welz, H.G. and Geiger, H.H. (2000) Analysis of Resistance to Striga hermonthica in Diallel Crosses of Sorghum. Euphytica, 116, 33-40.
 Mohamed, A., Rich, P.J., Housley, T.L. and Ejeta, G. (2001) In Vitro Techniques for Studying Mechanisms of Striga Resistance in Sorghum. Proceedings of the 7th International Parasitic Weed Symposium, Nantes, 3-8 June 2001, 96-100.