Determining plant spatial arrangement, plant population, and mineral nutrient requirements of Neoratanenia brachypus (Harms) CA in two contrasting agroecologies: Towards N. brachypus domestication

Abstract

Neorautanenia brachypus (Harms) CA Sm. is a wild, tuberous, perennial herb that many smallholder farmers have adopted as a drought survival feed for livestock in the south-eastern lowveld of Zimbabwe. To enable the commercial production of N. brachypus, there is a need for its domestication. Therefore, the objective of this study was to determine the plant spatial arrangement, plant population, and mineral nutrient requirements of N. brachypus to facilitate its domestication. A systematic review which assessed the effect of N. P and K fertilisers on the growth and yield of root and tuber crops (RTCs) showed variable responses ranging from no responses, positive and negative responses depending on soil fertility, climatic conditions, plant spacing and population density, species of RTC and other factors. There is no or very little research that has been done on plant spatial arrangement, population, and nutrient requirements of N. brachypus to enable agronomic recommendation domains to be produced towards its domestication, motivating this study. The first experiment was organised as a 3 × 2 factorial in a randomised complete block design (RCBD) with inter-row spacing (0.45, 0.60, and 0.75 m) and in-row plant spacing (0.30, and 0.45 m) as factors, replicated three times, at a high rainfall and low fertility arenosol sandy soil (Zvimba site) and a low rainfall high fertility eutric vertisol soil (Chikombedzi site) over two growing seasons (2020/21 and 2021/22). The second experiment employed a nutrient omission design, to evaluate the effects of 145 kg ha-1 N, 128 kg ha-1 P2O5 and 121 kg ha-1 K2O straight fertilisers and their combinations (N&P, N&K, P&K and N, P&K) on the growth and yield of N. brachypus. The third experiment, a 4 × 4 factorial in an RCBD, tested the effect of application rate of NPK (7% N, 14% P2O5, 7% K2O) viz., 0 kg ha-1, 75 kg ha-1, 150 kg ha-1, 300 kg ha-1 and four ammonium nitrate (AN, 34.5% N) fertiliser application rates viz., 0 kg ha-1, 75 kg ha-1, 150 kg ha-1, 300 kg ha-1). The results showed that, at the sandy infertile, Zvimba site, there were significant interactions (P < 0.05) between inter-row and in-row spacing on tuber length, diameter, and yield, which showed that the effect of in-row-spacing on tuber length, diameter and yield depended on inter-row-spacing. Tuber length, diameter, and yield did not significantly differ with an increase in inter-row spacing at 0.45 m in-row spacing. In contrast, tuber length, diameter, and yield were highest at the intermediate inter-row spacing (0.60 m) and were significantly lower at the narrow (0.45 m) and wide (0.75 m) inter-row spacing at 0.30 m in-row spacing. The highest tuber length, diameter and yield were obtained at 0.60 m inter-row x 0.30 m in-row with a plant population density of 55 555 plants ha-1 which is recommended, from the results of this study, for the arenosol sandy soil Zvimba site. At the more fertile deep black clay eutric vertisol soil vii

Chikombedzi site, averaged across in-row spacing, tuber length (121.17 mm), diameter (24.27 mm) and yield (2150 kg ha-1) was significantly higher (P < 0.01) in the narrow inter-row (0.45 m) spacing than the intermediate inter-row (0.6 m) spacing and the wide (0.75 m) inter-row spacing. Therefore, the narrow inter-row (0.45 m) spacing is recommended for the heavy soil site of Chikombedzi, in combination with an in-row spacing of between 0.3 m and 0.45 m depending on farmer circumstances with respect to the availability of seed, use of machinery and other factors. The results of the nutrient omission trial showed that the application of 145 kg ha-1 N significantly reduced tuber yield (P < 0.05) by 63% and 62.5% at the Zvimba site and at the Chikombedzi site, respectively, compared to the control without fertiliser. Application of 128 kg ha-1 P2O5 produced a similar yield to the control at the Zvimba site and increased yield by 34% compared to the control treatment at the Chikombedzi site. There was no significant difference in tuber yield between the control and the 121 kg ha-1 K2O treatment at the Zvimba and Chikombedzi sites. Combinations of N&P, N&K, P&K, and N, P&K straight fertilisers did not significantly increase tuber yield when compared to the control at the Zvimba and Chikombedzi sites. There was a significant interaction (P < 0.05) between compound 7%N, 14%P2O5, 7% K2O fertilizer and top-dressing ammonium nitrate (AN) (34.5% N) fertiliser application rates at the arenosol soil Zvimba site and the eutric vertisol Chikombedzi site on N. brachypus tuber yield. The interactions showed that the achievement of maximum tuber yield depended on the specific combination of rate of application of basal NPK fertiliser plus top dressing AN fertiliser at each site. At the sandy arenosol Zvimba site, the highest tuber yield was attained when 0 kg ha-1 compound NPK × 75 kg ha -1 AN was applied and when 150 kg ha -1compound NPK fertiliser plus 150 kg ha -1 AN fertiliser was applied. At the deep black clay eutric vertisol Chikombedzi site, the highest tuber yield was attained when 150 kg ha -1 compound NPK fertiliser plus 150 kg ha -1 AN fertiliser was applied. However, the interactions did not show a clear trend of how N. brachypus yield was responding to combinations of basal NPK application plus AN top-dressing fertiliser application with most of the treatment combinations, except the ones that attained the highest yield, producing tuber length, diameter and yield similar to the no fertiliser control. Therefore, more research is needed to understand the fertiliser requirements of N. brachypus, which like other wild RTCs, showed that it might require little or no fertiliser