The study was conducted at the University of Fort Hare Research Farm (32° 46' 282“ S and 26° 50' 067” E, at an altitude of 508 m above sea level). The experimental site receives an average annual rainfall of 535 mm, and has a mean annual temperature of 18.7°C (Nyambo et al., 2020c). The dominant natural vegetation is the Acacia species. Temperature and rainfall recorded during the study period are as given in Figure 1.

A 4 × 3 factorial experiment (12 treatment combinations) was conducted. It was laid out in an RCBD, replicated four times. In each replication, each treatment was represented by 4 pots which had one plant per pot, to give a of total of 12 × 4 = 48 pots per replication, therefore the 48 × 4 replications giving a total of 192 pots or plants. The two factors were (a) nitrogen fertilizer at four levels viz 100, 200, 300 and 400 kg N ha⁻¹ and (b) phosphorus fertilizer at three levels (40, 80, and 120 kg P ha⁻¹). The nitrogen fertilizer levels were selected based on a study by Abdullahi et al. (2013), who recommended using the 200 kg N ha⁻¹, and the phosphorus levels were based on the study by Mendieta-Araica et al. (2013), who recommended 44 kg P ha⁻¹ as the optimum rates for improved biomass yield and nutrient quality Moringa. Half the nitrogen fertilizer was applied at planting, while the other half was applied 4 weeks after transplanting (WAP) in the form of lime ammonium nitrate (LAN, 28), Phosphorus treatments were incorporated into the soil before planting in the form of superphosphate (SSP, 10.5). Potassium fertilizer was applied basal at planting to all pots at a rate of 731 kg ha-1 using potassium chloride (KCl) Mendieta-Araica et al. (2013). Basal fertilizer application was done soon after transplanting.

The experiment was carried out using 10 L plastic pots filled with 8 kg of loam soil. Seedlings were raised onsite from seeds bought from Nala Farm, Johannesburg, South Africa. The seedlings were transplanted 4 weeks after planting (at a plant height of about 30 cm). Data collection was started 4 weeks after transplanting. The basic properties of the soil used in the study are shown as in Table 1.

Non-destructive parameters, such as plant height and stem diameter, were taken at 4, 8, and 12 weeks after transplanting. Plant height was measured using a 1 m ruler, while stem diameter was measured using a Vernier caliper.

Data collected were subjected to statistical analyses using JMP 15.0 statistical software (SAS Institute, Inc., Cary, NC, USA). Means were separated using Turkey's test at p < 0.05 when ANOVA indicated a significant P-value. Significant differences were identified at three probability levels; p < 0.05, 0.01, and 0.001. The study was done over 12 weeks; therefore, time was introduced as an extra factor when analyzing for growth. The statistical model for the parameters measured during the growing period is given in Equation (2);

Where: μ = the general mean

β = the blocking effect

Ai = the effect of nitrogen application rate

Bi = the effect of phosphorus application rate

Ai × Bi = the interaction effect of nitrogen and phosphorus application rate

εijk = the standard error.

The three-way interaction of nitrogen × phosphorus × time was highly significant (p < 0.0001) with respect to plant height, stem diameter and number of leaves (Table 2). Similarly, the two-way interactions of nitrogen × phosphorus; nitrogen × time and phosphorus × time had a significant effect (p < 0.0001) on plant height, stem diameter and the number of leaves. The main factors of nitrogen, phosphorus and time greatly influenced (p < 0.0001) plant height and stem diameter, while the effects of phosphorus and time was significant at p < 0.001 with respect to the number of leaves.

At four WAP, tallest moringa trees were recorded from the 100 kg N ha⁻¹ and 80 kg P ha⁻¹ treatment combination, while the 400 kg N ha⁻¹ and 40 kg P ha⁻¹ of the treatment combination produced the shortest trees. At 12 WAP, moringa trees supplied with 100 kg N ha⁻¹ and 80 kg P ha⁻¹, were 46.8% and 45% taller compared to trees that were supplied with a treatment of 400 kg N ha⁻¹ and 120 kg P ha⁻¹; and 300 kg N ha⁻¹ and 120 kg P ha⁻¹, respectively (Figure 2).

At eight WAP the results indicated that trees that received 100 kg N ha⁻¹ and 80 kg P ha⁻¹ had significantly thicker stems (11.12 mm) compared to trees that received the treatment of 400 kg N ha⁻¹ and 80 kg P ha⁻¹ (7.0 mm). Similarly, trees that received 100 kg N ha⁻¹ and 80 kg P ha⁻¹ and trees treated with 200 kg N ha⁻¹ and 80 kg P ha⁻¹ at twelve WAP had significantly greater stem diameter than trees that received treatment of 400 kg N ha⁻¹ and 40 kg P ha⁻¹ (Figure 3).

The 100 kg N ha⁻¹ and 80 kg P ha⁻¹ treatment combination had 16.9% more leaves than the plants that received 300 kg N ha⁻¹ and 40 kg P ha⁻¹ at 4 WAP. On the other hand, at four WAP, trees that received 400 kg N ha⁻¹ and 120 kg P ha⁻¹ had 40.1% more leaves than trees in the 300 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination (Figure 4).

The interaction between the two factors nitrogen and phosphorus, was highly significant with respect to nutrients (p < 0.0001) and moisture (p < 0.001) (Table 3). The highest moisture content was recorded for the 300 kg N ha⁻¹ and 40 kg P ha⁻¹ treatment combination, while the least was in the treatment combination of 300 kg N ha⁻¹ and 80 kg P ha⁻¹ (Table 3). Ash contents was 21.38% higher in the treatment combination of 200 kg N ha⁻¹ and 80 kg P ha⁻¹, as compared to that of the 300 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination. Crude protein significantly increased with levels of both nitrogen and phosphorus. The highest crude protein was obtained from the 400 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination (38.28%), and the lowest in the 100 kg N ha⁻¹ and 40 kg P ha⁻¹ treatment combination (Table 3).

Interaction of nitrogen and phosphorus was highly significant (P < 0.0001) with respect to base cations and the K: (Ca+Mg) ratio (Table 4). Nitrogen depicted a high level of significance (p < 0.0001) with respect to Ca, Mg, and Na, while K was significant at p < 0.001. Effect of phosphorus was also highly significant (p < 0.0001) on Mg, Na and ratio of K: Ca+Mg, while the content of K was significant at p < 0.001. Only Ca was not significantly affected by P. Highest Ca content was observed in the treatment combination of 100 kg N ha⁻¹ and 80 kg P ha⁻¹, while the lowest was in the treatment combination of 300 kg N ha⁻¹ and 40 kg P ha⁻¹. Treatment combination of 100 kg N ha⁻¹ and 120 kg P ha⁻¹ had 94% Mg content higher than that of the treatment combination of 300 kg N ha⁻¹ and 120 kg P ha⁻¹ (Table 4). The K content was 15% higher in 100 kg N ha⁻¹ and 120 kg P ha⁻¹ compared to that of the 300 kg N ha⁻¹ and 40 kg P ha⁻¹ treatment combination (Table 4).

Highest concentrations of P and Zn were observed in the 400 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination, while the lowest concentrations of P and Zn were from plants treated with 100 kg N ha⁻¹ and 40 kg P ha⁻¹ (Table 5).

Copper was 73.5% higher in treatment combination of 400 kg N ha⁻¹ and 120 kg P ha⁻¹ compared to that of treatment combination of 400 kg N ha⁻¹ and 80 kg P ha⁻¹ (Table 5). The highest level of Fe content was record from 100 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination and the lowest from 200 kg N ha⁻¹ and 120 kg P ha⁻¹ treatment combination (Table 5).

Various studies have attributed the increase in yield and quality of moringa plants to appropriate nutrient supply (Abdelwanise et al., 2015). In the current study, the 100 kg N ha⁻¹ and 80 kg P ha⁻¹ treatment consistently produced significantly taller plants at all three sampling times. Higher doses of both nitrogen and phosphorus fertilizers did not increase the height of the moringa tree. Similarly, stem diameter and number of leaves did not increase with an increase in the dosage of N and P fertilizers. Lack of a consistent growth increase with higher fertilizer doses indicates that the moringa plants did not require higher nitrogen and phosphorus. The lack of response to higher fertilizer doses could be because the experiment was conducted in the hot summer months of December and March 2016 (Figure 1). Nutrient utilization of plants, particularly nitrogen, can potentially be reduced when precipitation is low and temperatures are high (Kering et al., 2011). Similar results were reported by Palada (1996), who showed that higher nitrogen fertilizers rates did not significantly increase the plant height of moringa tree. However, contrasting results were also reported in other studies. In a study conducted in Zimbabwe on sandy soil, Pahla et al. (2014) reported that plant height and shoot and root dry matter contents significantly increased with an increase in fertilizer rates. The same authors reported that the best performance of the moringa trees was observed at a fertilizer combination of 276 kg N ha⁻¹ and 400 kg ha⁻¹ P₂O₅. In a study conducted by Abdullahi et al. (2013) in North central Nigeria, the highest plant height was observed in moringa trees that received 200 kg N ha⁻¹ at 11 WAP on sandy loam soil. This inconsistency of these results could be explained by the differences in the climate, agronomic practices, soil type and environmental factors that significantly influence the nutrient content and growth of the trees.

Moringa is a miracle tree with a great indigenous source of highly digestible proteins, Ca, Fe, and Vitamin C. It contains all the essential nutritional elements essential for livestock and human beings (Gopalakrishnan et al., 2016). The resultant high nutritional contents imply that moringa leaves could be a nutritionally valuable and healthy ingredient for animal and human consumption. These nutrients may not be strictly medicinal but could be valuable in preventing malnutrition.

Crude protein that increased with an increase in fertilizer dose, was within the range of content in sunflower seeds, commonly found in protein concentrates (Mapiye et al., 2010). The direct proportional relationship between nitrogen and phosphorus regarding protein content implies that nitrogen and phosphorus rates increase protein content. These results also agree with Moyo et al. (2011). The results corroborate their capacity of moringa to be used as protein supplements in livestock diets, as reported by other authors (Jayanegara et al., 2019). Similarly, the observed results are in range but slightly lower than the crude protein content reported by Sodamade et al. (2013). In line with these findings, Ogbe and John (2011), reported crude protein contents between 17.01 and 27.51% in moringa. The composition differences can also be due to variations in soil type, climate, agronomic practices, environmental factors and leaf harvesting stage/age (Moyo et al., 2011).

ADF and NDF are important chemical properties that show the fibrosity and energy value of feed. The National Research Council of the National Academies recommends a minimum of 19% ADF and a range of 25–35% NDF for animal feed. According to Janhi et al. (2020), recommended forages should contain ADF and NDF below 45 and 65%, respectively. Regardless of the fertilizer combinations, both NDF and ADF values in the current study were within the recommended range but are slightly higher than the contents reported for other studies. For example, Foidl et al. (2001) reported NDF and ADF contents of 21.9 and 11.4%, respectively. Moyo et al. (2011) also reported much lower content: 11.40% NDF and 8.49% ADF. High temperatures and dry conditions in the summer season of the Eastern Cape Province could explain the higher NDF and ADF. Extremely high temperature and moisture stress reduce forage quality by lowering leaf/stem ratio, intensifying lignification, and increasing ADF and NDF contents (Kering et al., 2011). The differences in NDF and ADF between this study and others are possibly due to differences in agro-climatic conditions, leaf ages, and harvest time.