Our field trial was conducted during the 2024 cropping season, on an experimental farm in Liboudi, Mbankomo sub-division, Centre region, Cameroon. The site is located at latitude 3˚49'49.54"N and longitude 11˚27'17.79"E and 705 m above sea level. The area is characterized as a humid rainforest zone and the soil is clay loam. The total annual rainfall for 2024 was 1611.8 mm while the total rainfall during the period of experimentation (April to December 2024) was 1630.6 mm.

Poultry manure was obtained from Henri and Freres Poultry farm in Yaoundé, Cameroon. The poultry manure was composted for 6 weeks before applications ( Table 1 ). NPK fertilizer (20:10:10) was obtained from the fertilizer unit of the Ministry of Agriculture and Rural Development, Yaoundé, Cameroon. NPK fertilizers, contain nitrogen (N), phosphorus (P), and potassium (K). They play a crucial role in plant nutrition and growth. Nitrogen (N) is essential for leaf and shoot growth, promoting vigorous plant development and contributing to chlorophyll production. Phosphorus (P) is crucial for root development, flowering, and fruit formation, supporting the plant’s ability to uptake nutrients and thrive. Potassium (K) enhances plant resistance to stress, improves water regulation, and contributes to strong stems, increasing overall plant vigor. Optimizing NPK applications involves understanding the specific needs of different crops and considering factors like soil conditions, growth stages, and application methods for maximizing yield and ensuring sustainable practices [20] .

Soil samples were collected with soil auger at a depth of 0 to 20 cm from different locations of the site and bulked into a composite sample. The composite soil sample was air dried, passed through 2 mm sieve, and then analyzed for its physicochemical properties before planting and after harvest. Particle size distribution was performed to determine the soil textural class using the modified hydrometer methods of [21] [22] ( Table 2 ).

Chemical analysis of the soil was according to methods of [23] for soil pH and Macro-Kjeldahl digestion method of [24] for soil nitrogen. Available soil P was determined by Bray II method [25] , soil organic carbon was determined by the method of [26] . Determination of exchangeable cations (Ca²⁺, Mg²⁺, K⁺, Na⁺) and calculation of cation exchange capacities (CECs) was carried out according to the methods of [27] [28] ( Table 3 ).

Note: Means of the same main effect within a row followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

Note: KCl = Potassium chloride; N = Nitrogen; C = Carbon; P = Phosphorus; Ca = Calcium; Mg = Magnesium; K = Potassium; Na = Sodium. Means of the same main effect within a row followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

The planting material consisted of white and red cultivars of X. sagittifolium minituber seeds of mean weight 40 g produced from acclimatized vitroplants under the shed in the plant physiology Laboratory of the Higher Teacher Training College (HTTC), University of Yaoundé I, Yaoundé, Cameroon according to the protocol of [29] [30] ( Figure 1 ).

The experiment was a 4 × 2 factorial arrangement in a randomized complete block design and replicated three times. The site was ploughed, ridged and marked

out into two main blocks, one for each cultivar. Each of these main blocks was further subdivided into 3 sub blocks, which represent the three replicates. Each sub block was divided into five experimental plots, thus a total of fifteen plots were used for each cultivar. Each gross plot measured 4 m × 3 m (12 m²) with a net plot of 2 m × 2 m. The treatments comprised three rates each of application of Poultry Manure (0, 25, and 30 t∙ha⁻¹) and NPK fertilizer (0, 125, and 150 kg∙ha⁻¹). A total of fifteen treatment combinations and three replications were used. The Poultry manure was incorporated into the soils on the experimental plots in a single application based on the treatment combinations, at 2 weeks before planting to ease decomposition, while the NPK fertilizer was applied to the cocoyam stands according to treatment allocation at 4 weeks after planting (WAP) using the ring placement method. Each minituber was planted per hole at a depth of 15 cm and at a spacing of 0.5 m × 1.0 m resulting to twenty-five plants per plot and a total of 375 plants per cultivar. All plots were kept weed free by manual weeding after every 2 weeks [7] .

Five cocoyam plants were randomly selected from each of the net plots, tagged and then used for the determination of average plant height (cm), average number of leaves and average leaf area at 1, 2, 3, 4, 5, 6, 7, and 8 months after planting (MAP).

Leaf Area of Cocoyam = 0.917 (LW)

where L and W are length and width of the cocoyam leaf.

Yield parameters like number of tubers per plant, length (cm) and girth (cm) of tubers, and tuber weight (g) were assessed in 5 plants per treatment after 9 months at physiological maturity as follows:

The following procedure was used to determine Fe, Zn, Cu, Mn, Ca, Mg, K, Na and P contents.

P was extracted as above that is digested in concentrated HNO₃ and analyzed using Murphy Riley reagent also known as the phosphomolybdate blue method. It involved the formation of a blue-colored complex between phosphorus, ammonium molybdate, antimony potassium tartrate, and ascorbic acid in an acidic environment. The intensity of the blue color was directly proportional to the phosphorus concentration and measured using a spectrophotometer. P was read colorimetrically at 860nm using a UV-VIS spectrophotometer (PerkinElmer).

Proteins were extracted according to the modified method of [15] . 1 g of leaves were ground in chilled mortar (4˚C) with 5 ml of Tris-Maleate buffer (10 mM pH 7.2). The crude homogenate was centrifuged for 25 min at 10,000 g using “Beckmann-Coulter microfuge 20 R centrifuge”. The supernatant was removed, stored at −20˚C for future use as crude extract for protein and enzyme assays. Proteins were quantified according to method described by [32] as follows:

The optical density (OD) of protein was read at 595 nm using a spectrophotometer (Shimadzu UV-1605 UV-visible) against a blank containing distilled water. For each extract the experiment was carried out in triplets and the protein content was expressed in mg-equivalent of BSA per Fresh Weight [32] . This in reference to the extrapolation of the standard curve obtained under the same conditions as the samples while using BSA (Bovine Serum Albumine) at 0.1 mg/ml.

Determination of Crude Protein: The powdered tuber samples were analyzed for crude protein content according to the Kjeldahl’s method described in the Association of Official Analytical Chemists [33] .

Protein Digestion: Five grams of the sample was weighed in an ash less filter paper and put into 250 ml digestion flask. Then, 3 g of a catalytic mixture tablet (75 g of CuSO₄ and 0.7 g of K₂SO₄) and 15 mL of 98% H₂SO₄ were added into a digestion flask. The whole mixture was subjected to heating in a digestion chamber until transparent residue (clear light green) content was obtained, it was allowed to cool. After cooling, the digest was transferred into a 100 mL volumetric flask and made up to the mark (100 mL) with distilled water and then distilled using distillation apparatus.

Protein Distillation: Before use, the distillation apparatus was steamed for 15 min. After which, 100 ml conical flask containing 20 ml of 40% boric acid and 2 or 3 drops of Tashiro’s indicator was placed under the distillation apparatus with its outlet tubes inserted into the conical flask. The digest was washed down with distilled water followed by addition of 4 drops of phenolphthalein and 20 mL of 40% (w/v) NaOH solution. The distillation was continued until about 25 mL of distillate was trapped into the boric acid plus indicator solution changed from red to light grey, showing that all the ammonia liberated had been trapped. That means the digest in the condenser was steamed through until enough ammonia gas captured by the boric acid.

Titration: The solution in the receiving flask was titrated with 0.1 mM HCl to a brown color. After titration, the % of nitrogen was calculated as follows:

$\text{Nitrogen}\left(\%\right)=\frac{\left(\text{Vs}-\text{VB}\right)\times \text{mMHCl}\times 0.014008}{\text{Wt}\text{. of sample}}\times 100$

where Vs = Volume (mL) of HCl required to titrate sample; VB = Volume (mL) of acid required to titrate the blank; mMacid = Molarity of acid; and Wt = Weight of sample (g). Then, percentage of crude protein in the sample was calculated from the % nitrogen as:

Crude protein (%) = %N × F

where, F (conversion factor) is equivalent to 6.25 [33] . A blank was run through along with the sample and triplicate analysis was conducted for samples.

All results obtained were descriptively analysed. The results were presented as graphs, histograms or tables using the Excel Microsoft software 2013. Results were analyzed statistically by comparison of means with one-way ANOVA using Turkeys test and determination of least significant difference (LSD) at P < 0.05 with the Software Package SPSS 17.0 in windows. Correlations between different parameters were analysed using sigmaplot version 11 while principal component analysis (PCA) was done by the XLSTAT 2007 Ink version 10.0 software.

All growth parameters (plant height, number of leaves per plant and leaf area) analysed were generally more significant in the white cultivar cocoyam (white cv) plants than in the red cultivar cocoyam (red cv) plants for all treatments ( Table 4 & Table 5 ). After six months of growth there was a significant decrease in all

Note: map = Months after planting; PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); NS = Not significant; * = Significant at 5% level of probability. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

Note: map = Months after planting; PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); NS = Not significant; * = Significant at 5% level of probability. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

growth parameters for all treatments in both white and red cv X. sagittifolium plants. The poultry manure treatments (30 t∙ha⁻¹) showed the most significant increase in all growth parameters analysed from two months after planting to six months after planting among all treatments; meanwhile, within the same growth period, the control treatments (0 t of Poultry manure ha⁻¹ and 0 kg of NPK ha⁻¹) recorded the least increase in average plant height and leaf area for both white and red cultivars of the cocoyam plants ( Table 4 & Table 5 ).

After a period of 9 months, yield parameters were assessed. The four yield parameters (tuber number per plant, tuber weight, tuber length and tuber girth) analysed after harvest were generally greater in white cv cocoyam plants than in red cv cocoyam plants for all treatments. Poultry manure treatments (30 t∙ha⁻¹) showed the most significant yield parameters, while the control treatments (0 t of Poultry manure∙ha⁻¹ and 0 kg of NPK.ha⁻¹) had the least yield parameters for both cultivars ( Table 6 & Table 7 ).

Note: PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); * = Significant at 5% level of probability using Turkeys test. Means of the same main effect within a row followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

Note: PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); * = Significant at 5% level of probability using Turkeys test. Means of the same main effect within a row followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

At 6 months of growth, leaves of poultry manure treated plants (30 t∙ha⁻¹) expressed the most significant protein content (5.74 ± 0.38 mg eq BSA∙g⁻¹ FW) for white cv cocoyam plants followed by the same poultry manure treatments (30 t∙ha⁻¹) for red cv (4.91 ± 0.52mg eq BSA∙g⁻¹ FW). The different treatments of red cv X. sagittifolium plants did not show any significant difference in protein content after 2 and 4 months of planting ( Table 8 ).

Note: map = Months after planting; PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); NS = Not significant; * = Significant at 5% level of probability. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

The crude protein content values in the harvested red tubers were significantly higher than those of the white tubers for all the treatments evaluated. In the white cv tubers, crude protein content values were more significant in NPK (150 kg∙ha⁻¹) treatments at 8.94% than in the control (0 kg∙ha⁻¹) at 6.79%. Red tuber crude protein content value was very significant in poultry manure (30 t∙ha⁻¹) treatments with 18.14% ( Table 9 ).

Note:PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); * = Significant at 5% level of probability. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

Mineral nutrient content assay in harvested tubers, showed that 8 out of the 9 minerals assayed were more significant in red tubers than the white tubers of X. sagittifolium. K was more significant in white cv X. sagittifolium tubers than in the red cv X. sagittifolium tubers. Poultry manure treatments of 30 t/ha (PM2) recorded the most significant Fe, Mg, Ca and Na contents for both cultivars. There was no significant difference in the content of the trace elements Zn and Cu, for all treatments in both cultivars of X. sagittifolium ( Table 10 & Table 11 ).

Note: PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); NS = Not significant; * = Significant at 5% level of probability using Turkeys test. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

Note: PM1 = Poultry manure (25 t∙ha⁻¹); PM2 = Poultry manure (30 t∙ha⁻¹); NPK1 = NPK fertilizer (125 kg∙ha⁻¹); NPK2 = NPK fertilizer (150 kg∙ha⁻¹); NS = Not significant; * = Significant at 5% level of probability using Turkeys test. Means of the same main effect within a column followed by the same letter are not significantly different at 5% of probability level = NS, and different letters are significantly different at 5% of probability level.

At harvest, the correlation circle in the white cv X. sagittifolium plants indicate an explanation for principal component 1 axis (F1) of 85.56% while principal component axis 2 (F2) explained 8.25% of variance. The combined effect of both principal component axes 1 and 2 explained 93.80% of the total variance. In red cv X. sagittifolium plants, principal component 1 axis (F1) recorded 79.02% while principal component 2 axis (F2) had 13.70% of variance, with a combined total effect for both axes of 92.72% ( Figure 2 ).

In both white and red cv X. sagittifolium plants, positive correlations exist between parameters of growth, yield and soil exchangeable bases as indicated in factorial diagrams of principal component analysis (PCA). High positive correlations between sodium and magnesium exchangeable bases with average leaf number were observed while exchangeable potassium bases were highly correlated with yield parameters in both cultivars. In white cv plants, PM1 (25 t∙ha⁻¹) and NPK2 (150 kg∙ha⁻¹) samples were closely associated to each other while in red cv plants, control and NPK1 (125 kg∙ha⁻¹) samples showed a close association ( Figure 3 ).