The field trial was conducted in Konni region at Wazir, INRAN Striga screening field located around 8 km from Konni and 3 km from the main road (RN1). Konni is located around 417 km from Niamey in the eastern part of Niger with a rainfall distribution of between 500 to 600 mm (INS, 2016). Wazir is an area characterized by an agricultural farming system where farmers grow different cash crops. It is a highly Striga infested area and was used since 1987 for Striga screening purposes by INRAN Niger.

Twenty sorghum genotypes (Table 1) were planted in a randomized complete block design (RCBD) with three replications. Thus, in the twenty genotypes fifteen are documented Striga resistant genotypes and the five remain are Niger local landraces susceptible to Striga but highly preferred by smallholder farmers. In the experiment, the seeds were sow at 3 cm of depth and the row length was 3 m. The rows are separate by 1.80 m, and in each row, 10 hills were planted separated by 0.30 m. Thus, during the planting process 1 g (16000 viable seed) 6 of Striga seed were added in each hill with 5 to 6 sorghum grains. For a better development, after 14 days of growth the extra plants were removed to remain only 3 to 4 plants per hill. After this process, 2 to 3 weeding sessions were done for removing the other weeds except the Striga plants.

The data on sorghum crop components and Striga traits were collected in field. The sorghum data included plant vigor (Vg), 50% days to flowering (50% Flo), number of sorghum plants per hill (NPSorg), number of panicles (NPANI), plant height (HTR) in cm, 1000 grain weight (PoiGR) in g, and grain yield (Kg). The Striga emergence (EMR), the number of Striga plants emerged at 45, 60 and 90 days (SN45, SN60 and SN90) were collected as Striga related traits. The recorded data were transferred and analyzed using SAS (9.4). Thus, a normality test was performed in GENSTAT Edition 14, following by general linear model analysis and principal component analysis using SAS (9.4).

Linear correlation among traits varied from none, significant and highly significant correlations values. Number of Striga plants emerged after planting was negatively and highly correlated with number of Striga plants emerged at 60 days (r = -0.69) and Striga plants emerged at 90 days (r=-0.69). Thus, for the 50% days to flowering (Flo), a high significant correlation was observed with the panicles number, the grain weight, and the grain yield with respectively a correlation value r = 0.91, r = 0.86 and r = 0.84 (Table 2). For the plant height (HTR) a high significant correlation was also observed with the sorghum plant number per hill (NPSorg) with a correlation value r = 0.68. The panicles number (NPANI) had a high correlation (r = 0.802) with the sorghum grain weight (PoiGR). The number of Striga plants emerged at 45 days (SN45) was highly correlated to the number of Striga plants emerged at 60 days (NS60), the number of Striga plants emerged at 90 days (NS90) and also the grains weight (PoiGR) with respective correlation values of r = 0.79, r = 0.81 and r = 0.77. The Striga plant emerged at 60 days was highly correlated at 98% (r = 0.98) to the Striga plants emerged at 90 days and it was also correlated to the sorghum grain yield at 62% (r = 0.62). The number of Striga plants emerged at 60 days (NS60), the number of Striga plants emerged at 90 days (NS90), the sorghum grain weight (PoiGR) and the sorghum plant vigor (Vig) were highly correlated to sorghum grain yield with respectively 68% (r = 0.68), 75% (r = 075), 77% (r = 0.77) and 88% (r = 0.88) of correlation values (Table 2).

From the screening at Konni station, a principal component analysis (PCA) using 11 traits was performed to assess 20 sorghum genotypes from different countries. Among those 11 traits used, only the component 1, the component 2 and the component 3 were identified to have an eigen value higher (>) than 1 under Striga infestation condition at Konni station. Thus, the three components contribute at 42.3% for the component 1, 14.4% for the component 2 and 79.1% for the component 3, in the total phenotypic variation among the 20 genotypes. The selected three Eigen values are explaining around 79.1% of the variation among the collected parameters in the experiment.

Therefore, depending on the correlation with the 11 traits (Sorghum and Striga traits) initially used, several significant correlation matrices were obtained (Table 3). Thus, concerning the sorghum plant vigor (Vg), a negative significant correlations r = 0.61 was found with of among the PC3. For the Striga emergence (EMR) a negative significant correlation of r = 0.36 was obtain in the PC1. In addition, concerning the numbers of Striga plants at 45 days, r = 0.32 of significant correlation value was observed in the PC1. The number of Striga plants at 60, was also significantly correlated with r = 0.41 to the PC1. For the number of emerged Striga plant at 90 days, a significant correlation value of r =0.41was obtain in the PC1. For the number of sorghum plants per hill (NPSorg), 0.36 of significant correlation value was obtain in the PC1. Thus, the flowering days is an important stage in the development and the multiplication process. At this side, PC3 obtain a significant correlation value of 0.71. In addition, from the plant height (HTR), the PC1 observe 0.36 of correlation significance. Thus for the sorghum panicle number (NPANI), 0.41 of significant correlation values was obtained in the PC2. The grain weight (PoiGR) obtained 0.57 of significant correlation value on the PC2. Finally, for the grain yield, 0.55 for the PC2 as significant correlation value was observed.

The component plotting scores gave the results of the first two components, PC1 and PC2 on the genotypes distribution. In addition, each component was distributed into four environments comprising the environment 1 to the environment 4 (Figure 1). The genotypes 04CZ-F5P-52, ICSV1049, MDK, CE-180-33, Framida and Hakori Karoua were positively correlated to the environment 1 of the PC1. Thus, the genotypes El Mota, F2-20 and Eltsedaoua were positively correlated in the environment 2 of the PC1. Concerning the PC2, it is characterized by two environments (Env3 and Env4) distributed negatively and positively along the component. The environment 3 (Env3) was positively correlated to the PC2 and negatively correlated to the PC1 and comprise the genotype P9401, P9403, CE-151-262 and TXN13/BC3F5. The last environment (Env4) contains more genotypes than the others environment (Env1, Env2, Env3) and are negatively correlated to the PC 1 and PC2. The give the distribution of the different genotype contained on the environment 4 (SRN39, Brahan, S35, P9406, P9405, Mota Maradi, and AG8N13/BC3F5).

Concerning the component one and the component three, at this side, the environment 1 of the component 1 contain the genotype SRN39, CE-180-33, El tsedaoua and were positively correlated to the environment one. Thus, in the environment 2 correlated positively there are six genotypes represented by the genotype F2-20, 04-CZ-F5P-52, ICSV-1049, MDK, FRAMIDA, Hakori Karoua. For instance, the environment 3 was positively and negatively correlated respectively to the component 1 and the component 3 with the genotypes P9401, P 9403, Brahan, S35, CE-151-262 and the genotype P9405. In addition, the last environment correlated negatively to the component 1 and the component 3, four genotypes, were found, the SRN39, P9406, Mota Maradi and (AG-8XN13) BC3F5.

For the component two and the component three, the environment 1 was positively correlated to the component 2 and contain the genotype El Mota and El tsedaoua. Concerning the environment 2 of the component 2 was positively correlated with the genotype P9401, P9403, F2-20, CE-151-262, and the genotype (TXN13) BC3F5-41. Thus, the environment 3 is negatively correlated to the component 2 and positively correlated to the component 3 and contain Brahan, S35, 04-CZ-F5P-52, ICSV1049, MDK, P9405, FRAMIDA, Hakori Karoua .

The main objective of the study was to assess 20 sorghum genotypes under Striga infestation at Konni station in Niger for Striga resistance and determine Striga impact on sorghum growth and agronomic parameters. Thus, the normality test performed was significant along the recorded data. In addition, the general linear analysis (GLM) with the different mean squares obtained on the sorghum attributes and the weed parameters, were also significant. Moreover, this obtained result show a large variation among the genotype s and confirm the existence variability between them. The correlation study undertaken sort the interaction between the different variables involved in the experiment to better guide the breeders. For instance, correlation study deeply gave more information on the bidirectional relationship of the characters 10. Thus, it is useful to segregate the genotypes with good yield resistant to Striga and the susceptible genotypes. Therefore, the results will well guide the breeding process to better select the good material for the next hybridization study. So, the correlations found between the Striga emergence date (EMR) and Striga plants numbers at 60 and 90 days (NS60 and NS90), were highly negatively correlated for both of them. The negative correlation is explained by the fact that, when the Striga emergence start early or late it decrease the Striga plant number at 60 and 90 days. Obviously, if the correlations were positive, it should increase the plants numbers at 60 and 90 days. For the 50 % date to flowering, a high positive correlation was observed with the number of panicles (NPANI), the 1000 grain weight (PoiGR) and the yield. At the genotypic level, the result gave more explanation on the relationship between the flowering times and the yield attributes and also confirmed the previous study made by 7. The earlier the flowering time lower will be the yield. So, the positive correlation shows that late flowering is better for obtaining a greater number of panicles, a high grain weigh and high yield contrary to what was found by 3 and 10. This result will orient the breeder to select the best genotypes. In addition, the correlation between the panicle numbers and the grain weight was positively highly significant, meaning, higher panicle numbers result in higher 1000 grain weight. Concerning the number of Striga plants at 45 days (NS45), it was positively highly correlated to the number of Striga at 60 days and 90 days. Thus, when the number of Striga plants at 45 days is high, the number of Striga plants at 60 and 90 days and will be equally high and inversely when the number is less, there will also less number of Striga plants at 60 and 90 days 1. In addition, the 1000 grain weight and the yield were positively correlated which is well known in sorghum crop 12. Therefore, when the grain production is well distributing it will highly positively impact the yield . They also confirmed the high positive correlation between the plant vigor (Vig) and the yield. In addition, the PCA analysis identified three components with an

Eigen values greater than one. Thus, the varieties located in the environment 3 of the component 2 have a higher negative correlation with the component 1. However, in this environment Striga emergence starts earlier compared to in the other s environments 3. That means the genotype P9401, P9403, Brahan, S35, CE-151-262 and P9405 were under high Striga emergence impact with high grain yield. The highest plant height (HTR) were located in the environment 2 of the component 1. In this environment (ENV2), the genotype El Mota, F2-20 and El tsedaoua possess a high height compared to the others genotypes. The component 2 was positively correlated to the number of panicles (NPANI), the 1000 grain weight (PoiGR) and the grain yield in the environment 3 with high Striga emergence impact and concern the genotype P9401, (TXN13) BC3F5-41, CE-151-262 and P9403. In the environment 3 of the component 3 positively correlated to the plant vigor and the 50% date to flowering (Flo), the Striga emergence was high in this area and contains the genotype P9401, P9403, Brahan, S35, CE151-262, 04-CZ-F5P-52, ICSV-1049, MDK, P9405, (TXN13) BC3F5-41, FRAMIDA, and Hakori Karoua. Those varieties are characterized by a better plant vigor and are early flowering genotypes. In addition, the component 3, was also negatively correlated to the number of Striga plants number at 45, 60 and 90 days after the planting date. It concerned SRN39, CE-180-33,

P9406, Mota Maradi and the genotype (AG-8XN13) BC3F5 highly infested by the weed with less grain yield observed. This environment contains a high Striga plants number and confirmed the presence of tolerant genotypes with high grain yield in these area.