Introduction
Purple witchweed is a hemiparasitic flowering weed belonging to the Orbanchaceae family (Matusova et al. Reference Matusova, Rani, Verstappen, Franssen, Beale and Bouwmeester2005; Mohamed and Musselman Reference Mohamed, Musselman and Labrada2008). It is globally widespread and damaging (Ejeta and Gressel Reference Ejeta and Gressel2007; Oswald Reference Oswald2005; Parker Reference Parker2009), and is particularly prevalent in sub-Saharan Africa (Gethi and Smith Reference Gethi and Smith2004; Mohamed and Musselman Reference Mohamed, Musselman and Labrada2008; Rodenburg et al. Reference Rodenburg, Demont, Zwart and Bastiaans2016). Purple witchweed severely affects production of sorghum [Sorghum bicolor (L.) Moench], corn (Zea mays L.), millets [Eleusine coracana Gaertn. and Cenchrus americanus (L.) Morrone], tef [Eragrostis tef (Zucc.) Trotter], rice (Oryza sativa L.), and even sugarcane (Saccharum officinarum L.) (Addisu and Feleke Reference Addisu and Feleke2021; Atera and Itoh Reference Atera and Itoh2011; Atera et al. Reference Atera, Itoh, Azuma and Ishii2012; Kountche et al. Reference Kountche, Al-Babili and Haussmann2016; Parker Reference Parker2012; Spallek et al. Reference Spallek, Mutuku and Shirasu2013). The weed drastically reduces agricultural productivity for small-scale subsistence farmers in sub-Saharan Africa, including Ethiopia, and is considered the most devastating biological barrier to cereal production (Omanya et al. Reference Omanya, Haussmann, Hess, Reddy, Kayentao, Welz and Geiger2004).
Several studies have highlighted the widespread infestation of purple witchweed in Ethiopia. Gebreslasie et al. (Reference Gebreslasie, Tessema, Hamza and Nigussie2016) reported moderate to severe infestation throughout Tigray. Degebasa et al. (Reference Degebasa, Tessema, Bekeko and Belete2022) found it to be the dominant species in eastern and western Hararghe, Oromia. In Benishangul Gumuz, purple witchweed poses a significant challenge to sorghum production across almost all districts. The impact on sorghum production in Ethiopia is significant and widespread. Yield losses due to purple witchweed infestation range from 65% to 100% (Bayable and Marcantonio Reference Bayable and Di Marcantonio2013; Ejeta et al., Reference Ejeta, Babiker and Butler1993; Haussmann et al. Reference Haussmann, Hess, Welz and Geiger2000; Degebasa et al. Reference Degebasa, Tessema, Bekeko and Belete2022; Tesso et al. Reference Tesso, Gutema, Derassa and Gebisa2007). In Benishangul Gumuz, purple witchweed is the primary factor affecting sorghum production (Mesfin and Girma Reference Mesfin and Girma2022). The detrimental effects extend beyond Ethiopia to other countries in eastern and West Africa (Ejeta and Gressel Reference Ejeta and Gressel2007). These points underscore the necessity of developing effective strategies to manage and control purple witchweed to mitigate its devastating impact on sorghum production in the Benishangul Gumuz region and other affected areas in Ethiopia. The use of resistant crop varieties has been proposed as a practical and cost-effective long-term strategy for managing purple witchweed (Hearne Reference Hearne2009; Mandumbu et al. Reference Mandumbu, Mutengwa, Mabasa and Mwenje2019). Therefore, this study aimed to identify sorghum genotypes that are resistant to purple witchweed.
Materials and Methods
Plant Materials
The sorghum genotypes in this study were selected from landraces collected from farmers’ fields in Ethiopia, specifically in Benishangul Gumuz and some parts of western Oromia regions. The study included 49 genotypes and four released varieties (Table 1). Resistant checks, Berhan and Framida, were obtained from the Melkassa Agricultural Research Center. The use of check varieties allowed for effective assessment of resistance to purple witchweed. Purple witchweed seeds used in this study were collected over 3 yr (2019 to 2021) from heavily infested sorghum fields in various districts of Assosa Zone, including Bambasi, Abramo, and Ura. The seeds were stored in glass jars and kept in the dark at room temperature until needed for the trials.
Sources of 49 sorghum genotypes used in the study.a
a Abbreviations: AmR, Amhara Region/Ethiopia; AsARC, Assosa Agricultural Research Center; BGR, Benishangul Gumuz Region/Ethiopia; ETSL, Ethiopian sorghum landrace; MARC, Melkassa Agricultural Research Center/Ethiopia; OrR, Oromia Region/Ethiopia; S/N, serial number.
For the pot trials, 48 sorghum genotypes were used, including both susceptible and resistant checks. Assosa-1 and Adukara served as the susceptible checks, while Berhan was the resistant check. From these initial tests, 33 genotypes were selected for further evaluation in a specially designed purple witchweed sick plot at the Assosa Agricultural Research Center. This phase included another resistant check, Framida. To validate the pot and sick plot trials, seven sorghum genotypes, including resistant checks Berhan and Framida and promising resistant landraces ETSL102969, ETSL102970, and ETSL102975, alongside susceptible checks Assosa-1 and ETSL102957, were evaluated in farmers’ fields at three locations in Assosa, Benishangul Gumuz Region.
Study Sites, Trial Design, and Procedures
The trials were established at the Assosa Agricultural Research Center located between 10.0432°N and 34.5687°E, 1,553 m asl, in the Assosa Zone of the Benishangul Gumuz Region. The area receives a mean annual rainfall of 1,177 mm and has a mean temperature of 26.8 C.
The pot trials were laid out in a randomized complete block design with two replications in 2020 and 2021 under Lath-house conditions (temperature, 21–28 C; 12-h light/dark photoperiod; watered twice per week). A mix of sand, peat, and compost (1:3:1 volume/volume) filled 96 round plastic pots with a diameter of 27 cm at the top, 22 cm at the bottom, and height of 28.5 cm. Each pot received 4 mg of purple witchweed seeds, which were covered with a thin layer of soil mix (up to 5 cm depth). After a 10-d preconditioning period for the purple witchweed seeds, six sorghum seeds of each genotype were sown and later thinned to three plants per pot. The pots were not fertilized to enhance purple witchweed emergence.
In 2022, 33 sorghum genotypes were evaluated in purple witchweed sick plots at the center. The site was plowed twice with a tractor, and furrows spaced 70 cm apart were prepared with a furrow maker. The trial was laid out in a randomized complete block design with two replications. Furrows within each plot (2 m × 1.40 m) were uniformly infested with purple witchweed seeds collected during the 2021 cropping season. These seeds were covered with a thin soil layer and preconditioned for 10 d. Sorghum genotypes were then sown in the furrows at a rate of 10 kg ha−1. Aside from purple witchweed, other weeds were hand-weeded as observed, and recommended fertilization such as 100 kg ha−1 NPS (19 kg nitrogen, 38 kg P2O5, and 7 kg sulfur) at sowing and 50 kg ha−1 urea after thinning was followed.
For validation, a trial was designed in a randomized complete block design with three replications, using farmers’ fields as replications. Seven selected sorghum genotype were tested using plot sizes of 4.20 m × 4.05 m for each genotype. Weeds, except purple witchweed, were manually removed, and recommended fertilization was followed.
Data Collection and Analysis
Data were collected on both sorghum and purple witchweed parameters. For sorghum, the data included days to 50% anthesis, days to maturity, plant height, number of leaves, biomass, and dry matter (grams per pot). For purple witchweed, recorded data included emerged purple witchweed height and count at weekly intervals from the 7th to 12th week after crop emergence (WACE). Additionally, we measured purple witchweed biomass and dry matter.
To determine the maximum aboveground purple witchweed, we followed the methods suggested by Rodenburg et al. (2006). The area under purple witchweed number progress curve (ASNPC) was calculated as suggested by (Haussmann et al., Reference Haussmann, Hess, Reddy, Welz and Geiger2012) as follows:
$$ASNPC = \mathop \sum \nolimits_{i = 0}^{n - 1} ({{{Y_i} + {Y_{\left( {i + 1} \right)}}} \over 2})\left( {{t_{\left( {i + 1} \right)}} - {t_i}} \right)$$
where n is the number of purple witchweed recording dates, Y i is the purple witchweed count at the i th assessment date, and t i is the number of days after sowing at the i th assessment date.
ANOVA was carried out using the lmer() package in R software (R Core Team 2023), where sorghum genotypes were the fixed effect, while years, replication, and errors were random effects. Residual analysis was performed using the shapiro.test package to ensure the normal distribution of residuals. The randomized complete block design model used was as follows:
$${Y_{ij}} = \mu + {\alpha _i} + \;{\beta _j} + {\varepsilon _{ij}}$$
where, Y ij is the observed value for the experimental unit in the j th replication (r) assigned to the i th genotype, j = 1, 2… r and i = 1, 2,.., µ is the overall mean, α is the effect due to the i th treatment, β is the effect due to the j th block, and ε ij is the error term, where error terms are independent observations from an approximately normal distribution with mean equal to zero and constant variance σ 2 ε .
An independent sample t-test assessed the significant differences in the performance of various sorghum genotypes against purple witchweed. Treatment means were separated using Tukey’s HSD procedure at a 5% probability level. Additionally, sorghum genotypes were hierarchically clustered based on the number of purple witchweed plants that emerged per sorghum plant using the Euclidean distance matrix. Ward’s linkage method was utilized with MINITAB software version 14.
Results and Discussion
Response of Sorghum Landraces to Purple Witchweed
Pot Trial
The mean number of emerged purple witchweed plants per pot across all sorghum genotypes was 12.13, and the mean number of purple witchweed plant per sorghum plant was 4.04 (Table 2). This indicates there was an adequate level of infestation to determine the resistance of sorghum genotypes to purple witchweed.
Response of sorghum landraces to artificially infested purple witchweed grown in pots at Assosa, Benishangul Gumuz, Ethiopia.a,b,c
a Abbreviations: CV, coefficient of variation; MSD, minimum significant difference; WACE, weeks after crop emergence.
b Means within the same column followed by the same letter are not statistically different according to Tukey’s MSD (α = 0.05).
c Purple witchweed plants were counted at 12 wk after crop emergence. Experiments were conducted in 2020–2021.
The pot experiment results showed a highly significant difference (P < 0.0001) in the response of sorghum landraces to purple witchweed infestation (Table 2). The average number of emerged purple witchweed plants per pot ranged from 0.25 for the Berhan variety (resistant check) to 29.75 for the ETSL102973 landrace. Similarly, the number of purple witchweed plants per sorghum plant ranged from 0.08 for Berhan to 9.92 for ETSL102973. These findings suggest that more purple witchweed plants emerged from the pots that contained ETSL102973 sorghum seeds compared with pots that contained Berhan seeds, indicating its susceptibility to purple witchweed (Table 2 and Figure 1). On the other hand, fewer purple witchweed plants emerged in pots with sorghum landraces ETSL102969 and ETSL102970 compared to other landraces (Table 2; Figure 2).
Top five sorghum genotypes with lowest emerged purple witchweed count per sorghum plant from 2020 to 2022 at Assosa, Benishangul Gumuz, Ethiopia. Error bars indicate standard error of uncertainty in the average number of purple witchweed plants per sorghum plant. Tukey’s minimum significant difference = 1.02.
Reaction of sorghum genotypes to area under purple witchweed number progress curve (ASNPC) at Assosa, Benishangul Gumuz, Ethiopia
Purple Witchweed Sick-Plot Trial
The purple witchweed sick-plot experiment revealed significant variation (P < 0.05) among sorghum landraces in the mean number of emerged purple witchweed plants per plot at 12 WACE. Purple witchweed emergence per plot varied depending on the sorghum genotype that was also planted, ranging from 3.0 for Berhan and ETSL102969 to 148.5 for ETSL102944. Similarly, the number of emerged purple witchweed plants per sorghum plant ranged from 0.22 for ETSL102966 to 5.48 for ETSL102954 (Table 3; Figure 1). The results indicated that purple witchweed emergence (3.0 per plot) was lowest in pots with the resistant check Berhan and sorghum landrace ETSL102969. Purple witchweed emergence was also lower in pots with sorghum landraces ETSL102970, ETSL102975, ETSL19001, and ETSL100053than in the sick-plot trial with the resistant check variety Framida. These results are consistent with the findings from previous pot trials (Tables 2 and 3).
Response of sorghum landraces to purple witchweed in an artificially infested sick plot in 2022 at Assosa, Benishangul Gumuz, Ethiopia.a,b
a Abbreviations: CV, coefficient of variation; MSD, minimum significant difference.
b Means within the same column followed by the same letter are not statistically different according to Tukey’s MSD (α = 0.05).
The study showed that the resistant sorghum genotypesmature early, with maturity periods of 125 d for Berhan, 142 d for ETSL102970, and 144 d for ETSL102969 (Table 3). This finding aligns with reports by Ayana et al. (Reference Ayana, Bantte and Tadesse2019), that early maturing sorghum genotypes show resistance to purple witchweed. Franke et al. (Reference Franke, Ellis-Jones, Tarawali, Schulz, Hussaini, Kureh, White, Chikoye, Douthwaite, Oyewole and Olanrewaju2006) also found that earlier maturing sorghum genotypes responded positively to purple witchweed stress. Additionally, the purple witchweed-resistant sorghum genotypes in this study ranged in height from 102 cm for ETSL102969 to 140 cm for Berhan. They also had fewer leaves per plant, ranging from four for ETSL102969 to six for Berhan (Table 3).
Figure 2 illustrates those genotypes ETSL102970, Berhan, and ETSL102969 had lower ASNPC values than other genotypes, indicating slower or less severe emergence of purple witchweed in these resistant sorghum genotypes. The cluster analysis grouped these genotypes into one group (Figure 3). Conversely, genotypes ETSL102957 and ETSL102944 exhibited higher ASNPC values, suggesting a higher incidence and more rapid emergence of purple witchweed around these susceptible genotypes. The resistant checks Berhan and Framida also demonstrated low ASNPC values, confirming their resistance to purple witchweed infestation. These findings further support the potential resistance of sorghum genotypes Berhan, ETSL102969, and ETSL102970 to purple witchweed infestation in Assosa, Benishangul Gumuz Region (Figures 1 and 3).
Cluster analysis showing the relationship among sorghum genotypes for their resistance to purple witchweed at Assosa, Ethiopia.
Validation Trials in Purple Witchweed Hot-Spot Farmer’s Fields
As illustrated in Table 4, the validation trial confirmed that the fewest purple witchweed plants per plot emerged among the resistant check Berhan and the sorghum landrace ETSL102969. Additionally, fewer purple witchweed plants grew among the sorghum landrace ETSL102970 compared to the resistant check Framida. The number of purple witchweed plants per sorghum plant was also low among Berhan, ETSL102969, and ETSL102970 plants. Conversely, the largest number of purple witchweed plants grew among the susceptible checks ETSL102957 and Assosa-1. These findings indicate that Berhan, ETSL102969, and ETSL102970 exhibit strong resistance against purple witchweed infestation. Overall, this validation trial confirmed that ETSL102969 and ETSL102970 sorghum landraces have comparable or superior resistance to purple witchweed compared with resistant checks Berhan and Framida.
Validation of purple witchweed-resistant sorghum landraces in farmers’ fields in 2023 at Assosa, Benishangul Gumuz, Ethiopia.a,b
a Abbreviations: CV, coefficient of variation; MSD, minimum significant difference.
b Means within the same column followed by the same letter are not statistically different according to Tukey’s MSD (α = 0.05).
Furthermore, the promising sorghum landraces ETSL102969 and ETSL102970 demonstrated higher yields than the resistant checks in the validation trial. The larger seed size and white seed color of sorghum landrace ETSL102969 are particularly desirable traits among local farmers (Alemu et al. Reference Alemu, Begna and Bekele2024; Legesse et al. Reference Legesse, Demelash and Seyoum2019). These traits are beneficial for breeding programs because they can be combined with the purple witchweed resistance trait to develop sorghum varieties with both resistance and preferred seed characteristics. Incorporating these traits into breeding programs increases the likelihood of obtaining F-generations with both white color and large seed size, along with resistance to purple witchweed. This approach will not only benefit farmers but also improve productivity and enhance market value for sorghum in the region.
Practical Implication
This research reveals that Berhan, ETSL102969, and ETSL102970 showed strong resistance against purple witchweed. Additionally, the white seed color and large seed size of ETSL102969; a trait highly preferred by farmers, make this genotype more suitable for the breeding program aimed at developing purple witchweed-resistant and locally preferred sorghum varieties to improve yields and food security in regions plagued with purple witchweed. Genotypes such as Berhan, ETSL102969, and ETSL102970, when incorporated into breeding programs, could lead to more robust and resistant sorghum crops, thereby benefiting farmers in purple witchweed-affected regions.
Acknowledgments
We thank staff members of the Assosa Agricultural Research Center for providing a vehicle for collecting purple witchweed seeds, and for helping with field work.
Funding
Funding for this research was provided by the Ethiopian Institute of Agricultural Research.
Competing Interests
The authors declare they have no conflicts of interest.
Open access
