Yield and Profitability of Sweet Potato ( Ipomoea batatas (L.) Lam) as a Function of Increasing Levels of Phosphorus and Varieties in Southern Ethiopia

.


Introduction
Sweet potato (Ipomoea batatas Lam.) ranks second following Irish potato in the world root and tuber crop production and the third most important crop after potato and cassava largely grown in East Africa [1]. In Ethiopia, sweet potato ranks frst in total production (37.2%) and second in area coverage (15.21%) next to Irish potato compared to other root and tuber crops cultivated [2].
In Ethiopia, potential yield of sweet potato under research station and on-farm research with improved management practices is ≥50 and 35.0 t ha −1 , respectively [2,3]. However, the average respective yields of the crop at national, regional (SNNP), and Zonal (Wolaita) levels were 42, 25.42, and 37.87 t ha −1 , respectively [4]. Te productivity of sweet potato in Boloso Sore District is 13 t ha −1 which is far below that of research station, national, and regional average yield (BSDAS, unpublished data). Many diverse and complex biotic, abiotic, and human factors contribute to the low productivity of sweet potato in the study area. Some of the production constraints are a lack of good quality planting materials, poor soil management practices, diseases, weeds, insect pests, and limited or no use of chemical fertilizer.
Among the nutrients, P is widely used by plants and is most defcient in tropical soils [5], and most of the Ethiopian soils are low in available P. Hameda et al. [6] reported that diferent P-rates refected a signifcant efect on total and marketable tuber yield, tuber dry matter, average tuber root weight, and average tuber root diameter. Hassan et al. [7] and Dumbuya et al. [8] also found that fertilization of sweet potato plants with P resulted in signifcant increase in total and marketable yield, although they did not quantify the marginal rate of return.
In Ethiopia, some researchers had previously evaluated the responses of sweet potato to diferent application rates of organic (FYM) and inorganic (N and P) fertilizers at different locations [9]. Soil fertility studies conducted over diferent locations for various crops in the country generally have shown good yield response to applied P fertilizers indicating the defciency of P in the soil [5]. To improve the productivity of sweet potato, the best-performing variety and site-specifc P rate for its economic feasibility should be determined. Hence, this study was initiated to evaluate the response and economic proftability of sweet potato varieties to P rates.

Description of Experimental
Site. Field experiment was conducted during the 2016 main cropping season (March 1 to June 30) at Tadissa on a farmer's feld in Boloso Sore District, Wolaita zone, Southern Ethiopia. Geographical coordinate of the site is 7°04ʹ N latitude and 37°41ʹ E longitude with an altitude of 1830 masl. Te soil is sandy loam in texture, moderately acidic, low in available P and total nitrogen, and moderate in cation exchange capacity (Table 1). Te mean annual rainfall is 1520 mm with a bimodal pattern. Te average mean temperature is 20.02°C.

Treatments and Experimental
Design. Treatments consisted of three sweet potato varieties (Awassa-83, Kulfo, and Local) and fve rates of P (0, 10, 20, 30, 40 kg ha −1 ). Te treatments were combined in a 3 × 5 factorial arrangement and laid out in a randomized complete block design with three replications.

Experimental Materials and Procedures.
Before planting, the land was well prepared manually using an oxen plough. Di-ammonium phosphate (DAP) (18% N, 46% P 2 O 5 ) was used as a source of phosphorus that was applied at planting. Similarly, urea (46% N) at a rate of 50 kg ha −1 was used as a source of nitrogen and applied uniformly to all plots at planting by considering the amount of N found in each level of DAP. Vine cuttings having a length of 30 cm were planted at a spacing of 60 (between rows) and 30 cm (between plants) according to the recommendation of Geleta [12]. Te plot size was 2.4 by 2.4 m. A distance of 100 and 150 cm was maintained between plots and blocks, respectively.

Data Collection.
To collect soil data, representative soil samples were taken using an auger from the depth of 0-30 cm at diferent points of the experimental feld before planting to make one composite sample. Te samples were used to determine the following soil parameters: texture, pH, total nitrogen (TN), available phosphorus (AP), and cation exchange capacity (CEC). Texture was analyzed at the Wolaita Sodo soil testing center, southern Ethiopia. Soil pH was determined at the 1 : 2.5 soil to water ratio using a glass electrode attached to a pH digital meter [13]. Total nitrogen and available phosphorus were determined according to the methods of Dewis and Freitas [14] and Olsen and Dean [15], respectively. CEC was determined using Kjeldhal procedure as described by Ranist et al. [16].
Six plants were systematically selected from the middle two rows by considering border efects to collect vine length, vine internode length, and vine girth. Vine length (cm) was measured from the base of the main shoot to the tip. Vine internode length (cm) was measured between the fourth and ffth nodes from the tip. Vine girth (cm) is the mean diameter of the vine measured between the fourth and ffth node from the tip of the main shoot and measured with the aid of calliper.
Root data taken include the number of roots per plant (number of roots produced per plant), storage root length (cm) (vertical length of the root measured from the tip to the scar of separation), storage root diameter (cm) (diameter of the root taken from the middle portion), and the root dry matter weight (obtained from roots cut into smaller pieces and dried in a hot oven at 80°C for 48 hours). Te weight was taken by using a sensitive balance. Marketable yield (t ha −1 ) is the weight of clean, uninfected roots that fall in the size range of 100 to 500 g. Tis was taken by weighing the roots collected from the harvestable plot. Total yield (t ha −1 ) is the sum total of both marketable and unmarketable storage root yields obtained from the harvestable plot. Te collection of data from one year experiment is the limitation of this study.

Economic Proftability.
To consolidate the statistical analysis of the agronomic data, economic analysis was done for each treatment. For economic evaluation, cost, return, and beneft to cost ratios were calculated according to the procedure given by CIMMYT [17]. Total variable cost (TVC) = sum of all variable costs in a given treatment. Storage root yield (SRY) = total yield harvested from one hectare. Adjusted yield (AJY) = SRY × 90%, the average yield adjusted downward by a certain percentage to refect the diference between the experimental yield and the yield of farmers. Total revenue (TR) = AJY × feld price of the storage root; the gross feld beneft for each treatment was calculated by multiplying the feld price by the adjusted yield. Net revenue (NR) = total revenue (TR) − total variable cost (TVC); the fnal line of the partial budget is the net benefts. It was calculated by subtracting the total costs that vary from the gross feld benefts for each treatment. Marginal rate of  [10] return (MRR %) = ΔNR/ΔTVC × 100, which is the marginal net beneft (i.e., the change in net beneft) divided by the marginal cost (i.e., the change in cost).

Statistical Data Analysis.
Data was subjected to analysis of variance (ANOVA) using Statistical Analysis Software (SAS) version 9.2 [18] (SAS, 2009). Signifcantly difering means with respect to the efect of P rates, varieties, and their interaction were separated using the least signifcant difference (LSD) test at 5% level of signifcance as described by Gomez and Gomez [19].

Physicochemical Properties of the Experimental Site Soil.
Te soil analysis results showed that the soil was moderately acidic, sandy loam in texture, low in AP and TN, and moderate in CEC (Table 1). Tis result indicated that the soil requires external nutrient sources such as farmyard manure, inorganic fertilizers, and crop residues for enhancing growth and yield of the crop.

Above Ground Growth of Sweet Potato.
Sweet potato varieties varied signifcantly (P < 0.05) on their vine length, vine internodes length, vine diameter, and vine number, while P had no signifcant efect on these parameters (Tables 2 and 3). Te local variety had the longest vine length (149.94 cm) and vine internode length (4.49 cm) while the shortest length of the indicated parameters was observed from Kulfo (Table 2). Similarly, the highest vine diameter (3.5 cm) and vine number (16.77) were recorded from Awassa-83 and Kulfo, respectively, while the lowest one for the respective parameters (2.61 cm and 9.1) was obtained from Kulfo and local varieties, respectively (Table 2). Te current study results indicated that the local variety had better growth performance than the other varieties. Tis variability might be attributed to their genetic diference on growth traits.
Tis indicates that local varieties can be used as a good vine source especially where production is aimed at producing sweet potato vines. Te vines can also be used as forage for the feeding of ruminants since the vines are rich in their protein and mineral contents were needed in livestock feed [20]. Similar to the current study result, Kathabwalika et al. [21] and Rahman et al. [22] reported that a signifcant diference was observed on vine length among sweet potato varieties that were evaluated. Furthermore, Berhanu and Beniam [23] reported similar results. Te results also indicate that P level had no signifcant efect on the above ground growth parameters. Tis may be due to the low availability of P in the soil and its fxation (Table 1). Rashid and Waithaka [24] reported that phosphorus nutrition did not signifcantly increase vine production of sweet potato. On the other hand, Hassan et al. [7] and El-Sayed et al. [25] stated that application of 20 kg P ha −1 on sweet potato plants resulted in signifcantly superior vine length, number of branches per plant, and leaf area per plant as compared with control treatment. Tis diference may be due to the efective absorption of available phosphorus in the soil.

Number of Roots per Plant.
Number of roots per plant showed signifcant (P < 0.05) variation only with variety diference (Table 3). Te highest (6.49) and lowest (2.39) number of roots per plant was noted from Awassa-83 and Kulfo varieties of sweet potato, respectively (Table 4). Te Awassa-83 variety showed the highest NRRP (6.43) compared to Kulfo and local varieties. Te variation among varieties in the number of roots may be attributed to their genetic diference since variability in genetic makeup in turn expressed on their growth performance [26]. Egbe et al. [27] reported that sweet potato varieties showed signifcant variation in their root number.

Storage Root Length and Diameter.
Te interaction effects of P rate and variety signifcantly (P < 0.05) infuenced root length and diameter (Table 3). Te longest (23.22 cm) and shortest (10.11 cm) storage root lengths were recorded from Awassa-83 variety at P rate of 30 kg ha −1 and Kulfo with 10 kg P ha −1 , respectively ( Figure 1). Similarly, the highest (23.31 cm) and lowest (11.81 cm) root diameters were observed from the local variety at P rate of 10 kg ha −1 and Kulfo at P rate of 40 kg ha −1 , respectively ( Figure 2).
Te variation on storage root growth parameter of sweet potato with their convulsive to P rates might be related to their inherent variability to use nutrients like P. Berhanu and Beniam [23] stated that sweet potato varieties varied on their root growth in response to P fertilizer rates.

Root Dry Weight.
Interaction efect of P rates and varieties was signifcant (P < 0.05) on storage root dry weight (Table 3). Te highest storage root dry weight (18.12 t ha −1 ) was obtained from AW-83 variety at P rate of 20 kg ha −1 followed by the same variety at P rate of 30 kg ha −1 with a mean storage root dry weight of 14.67 t ha −1 (Table 4). Te lowest storage root dry weight (0.64 t ha −1 ) was obtained from a variety of Kulfo at 0 kg P ha −1 (Table 5).
Tis result indicated that varieties of sweet potato showed variation on their storage root dry weight as P rates increase which might be attributed to the fact that crop varieties show variation on their resource utilization which in turn expressed on assimilates allocation and biomass formation [28]. Te variation might be related to the genetic formation of the genotypes which may infuence the strength of their biomass production and the partitioning of assimilate to diferent parts of the plant. Similarly, Belanger [29] reported that variations on tuber growth and biomass partitioning of two potato cultivars were observed as a result of diferent fertilization rates and irrigation treatments.

Marketable and Total Root Yield.
Te highest marketable root yield (30.22 t ha −1 ) was obtained from a variety of AW-83 at P rate of 30 kg ha −1 followed by the same variety at P rate of 40 kg h −1 with a mean marketable yield of 29.13 t     Means followed by the same letters are not signifcantly diferent at the 5% level of signifcance. AW-83 � sweet potato variety named as Awassa-83, LSD � least signifcant diference at 5% probability level, CV � coefcient of variation, and V * P � LSD value for interaction of variety by P rates.
Te results indicate that marketable root yield of the three sweet potato varieties signifcantly increased with increasing P rates up to 30 kg ha −1 and then showed Means with the same letters are not signifcantly diferent at the 5% probability level. CV � coefcient of variation, AW-83 � sweet potato variety named as Awassa-83, LSD � least signifcant diference at 5% level of signifcance, and V * P � LSD value for interaction of variety by P rate.   a declining trend. Tese increments might be related to the important role of phosphorus as an essential component of many organic compounds in the plant such as phosphor-proteins, phospholipids, nucleic acids, and nucleotides that indirectly refected on yield [30]. Te declining of marketable root yield with application of P after 30 kg ha −1 might be attributed to the overdose efect of P that induces defciency of micronutrient [31,32]. Similar results were reported by Hassan et al. [7] who refected that P fertilization of sweet potato plants cause signifcant increases in total and marketable yield. Additionally, [6] indicated that root yield and yield components of sweet potato were increased by increasing rate of P from 6.5 to 20 kg ha −1 .
Te variety AW-83 gave the highest total root yield (33.57 t ha −1 ) at P rate of 30 kg ha −1 while the lowest total root yield (8.76 t ha −1 ) is obtained from the variety of Kulfo at P rate of 0 kg ha −1 (Figure 3). Furthermore, total root yield of the three sweet potato varieties signifcantly increased with increasing P rates up to 30 kg ha −1 and then shows declining trends. Te variability of sweet potato varieties on their root yield in response to P rates might be related to their diferences on root yield attributes and their inherent variation on capturing environmental resources and later on root yields formation. Te reduction of root yield after application of P over the optimum rate might induced defciency of micronutrients especially zinc as a result of high P build-up in the soil [31,32]. Similarly, Hameda et al. [6] report that the increase of P from 0 to 20 kg ha −1 increased root yield, but the increase of P from 20 to 40 kg ha −1 decreased root yield.

Economic Proftability.
Te partial budget analysis showed that the highest net income was found with the dose of 30 kg P ha −1 (34 329 birr) followed by 40 kg P ha −1 (32 311 birr) of the same variety, Awassa-83 (Table 7). Te lowest net income (7,387 birr) occurred when no P was applied on the Kulfo variety.
Te highest marginal rate of return (592.53%) was achieved with the Awassa-83 variety at a P rate of 10 kg ha −1 . Te phosphorus rate of 40 kg/ha was rejected because the treatments with higher cost and lower net beneft than previous successive treatments were considered dominated [17]. Te highest cost-beneft ratio (18.16) is found with the AW-83 variety at a P rate of 10 kg ha −1 and the lowest Kulfo at a P rate of 40 kg ha −1 . Terefore, the combination of 10 kg P ha −1 with the Awassa-83 variety is proftable and is recommended for farmers in the study area and other areas with similar agroecological conditions.

Conclusion
Te combined efect of varietal performance and P rates signifcantly infuence sweet potato growth and yield. Te Awassa-83 variety shows the best performance in most of the parameters measured with respect to the local and Kulfo varieties. Te Asawa-83 variety with P at a rate of 10 kg ha −1 contributes the greatest economic beneft in this study.

Data Availability
Te data used to support the fndings of the study are available from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.