Weed profiling fields of herbicide tolerant maize in the Mthatha region, Eastern Cape Province

Abstract

Maize cultivation has been a historical agricultural practise of most indigenous people in southern Africa for centuries. Although the origins of maize are rooted to south-central Mexico in North-America, it has for centuries been an important subsistence and economic crop for southern Africa. In South Africa, maize provides a source of income for thousands of commercial and smallholder farmers countrywide. The continued increased demand for maize production has led to the development of different cultivars to assist farmers to achieve better crop management towards good yields, and subsequent, profitable margins. At the same time, weed management and control, among others, remain a big challenge for maize farmers. Although weeds associated with herbicide tolerant (HT) maize have been studied to a greater degree globally, South Africa still lacks fundamental research to better understand their existence and prevalence in HT maize agroecosystems. This calls for the development of knowledge to determine what weeds interfere with HT maize cropping systems. This will enable farmers to choose best practice and methods for weed control that would minimize or even prevent negative impacts on the environment, while maintaining their fields for sustainable maize production. The study aimed to profile HT maize weed populations in rural farming communities of the selected areas of Oliver Tambo District Municipality, Eastern Cape Province, South Africa. This was done by documenting and comparing their composition (abundance, diversity, and richness) and functional traits between maize fields and their margins. In addition, the different herbicide protocols administered to control weeds in the HT maize fields were also profiled. Field surveys were conducted during the 2017/18, 2018/19 and 2019/20 planting seasons. The HT maize varieties planted were NK603xMON810 (2017/18-season one) and MON89034 (2018/19; 2019/20-seasons two and three, respectively). Weed surveys were conducted at four sites, namely Baziya Makaula, Baziya Mission, Baziya Jojweni and Tsolo. The farming method or practice was the same in all four sites in terms of HT maize varieties used and herbicide protocols followed across the three seasons. The sampling in each field and the field margin (where possible) was carried out in 3x3 m quadrats at intervals of 10 m along a transect line. A total of 88 weed species, representing 31 families, were recorded from 244 plots for both maize fields (MF) and field margins (M) combined. The three families that contributed the most weed species were Asteraceae, Fabaceae and Poaceae. There were three weed species which contributed individual dissimilarity of >5% between maize fields and field margins, and all being dominant in the latter (Cynodon dactylon (L.) Pers, Richardia brasiliensis Gomes and Sonchus nanus O.Hoffm). Weed species trait composition differed significantly (p=0.031) between most sites - and pre-ploughing, post-ploughing and seed sowing surveys (p<0.001) - indicating that local rather than regional factors affect trait composition. However, trait richness for pre-ploughing, post-ploughing and seed sowing surveys did not differ significantly (p>0.05). Weed species richness differed significantly over the three seasons (p<0.05). The findings confirm that maize fields and field margins are characterised by different weed assemblages and diversity. No similarity in species composition was confirmed (p<0.05). Similarity in trait richness was confirmed as it did not differ significantly (p>0.05). Weed species and functional type richness differed significantly with lower richness in post-ploughing than pre-ploughing surveys. Therefore, the herbicide application effectiveness was supported. No clear evidence was gathered to suggest that herbicide application contributes to weed shift, as weed species and functional types did not differ significantly in composition (p>0.05) for pre-ploughing, post-ploughing and seed

sowing surveys. By studying the weed composition in HT maize, the study has successfully provided preliminary evidence of the traits that could be characteristic of HT maize weeds with the ability or predisposition to become herbicide tolerant. The findings suggest that effective weed control using herbicides in HT maize fields is possible, although this still needs further testing across numerous seasons. Furthermore, additional investigation and monitoring of various environmental parameters likely to be responsible for some of the unexplained observations is necessary to fully understand weed mechanisms, both spatially and temporally.