N in single bands 15 cm from the plants produced high late season yield of tomatoes and combined placements in wider bands or four bands produced high fruit yield and leaf concentrations of N, P, K, and Mn of fresh market tomatoes (Wyatt et al., 2001). Parish et al. (1997), found inconsistent response of several vegetable crops to fertilizer placement. They found differences in plant size and vigor in spring turnips with broadcast treatments appearing superior but yield at first harvest and total yield was lower, neither cabbage nor turnip responded to placement treatments. Thus, most fertilizer placement studies have been conducted using vegetable crops with economic production above ground.

Although several fertilizer studies have been conducted with sweetpotatoes, very few have been directed towards fertilizer placement effects on yield and /or nutrient uptake. Earlier work suggests that storage root yield was improved with half the fertilizer applied prior to planting, either directly under the ridge or between 10 and 15 cm from the plants on a level with slightly below the depth at which the plants were set; the remainder being applied two to three weeks after planting (Steinbauer and Kushman, 1971). Tsuno and Fujise (1968) found that storage root yields were increased when K was banded 30 or 40 cm deep into the soil layer. The objective of this experiment was to evaluate the impact of fertilizer placement treatments on vine and storage root yield, root dry matter and elemental leaf and root concentration of sweetpotato.

‘Whatley/Loretan’ sweetpotato slips were grown on a Norfolk sandy loam soil (fine, silicious, thermic typic Paleudult with a pH of 6.2, organic matter <1%, cation exchange capacity of of 4.94 mm L/kg and total N 0.08%, Mortley and Ntibashirwa, 2012) at the George Washington Carver Agricultural Experiment Station in Tuskegee, AL located at latitude 30° 27’ N and 85° 42’ W with an elevation of 470.34 m above sea level. Temperature and rainfall during the growing season averaged 32°C and 158 mm, respectively. A randomized   complete   block   design,  with  10  treatments  and  4 replications in 3-row plots measuring 1.2 m × 6.1 m (of which the center row was used for data collection) was used. Soil analyses results showed a pH level of 6.2, P and K levels medium and Ca, high. The fertilizer sources were ammonium nitrate (34% N), triple superphosphate (34% P) and muriate of potash (60% K) which were applied at rates 135, 35 and 133 kg ha^-1 respectively. Treatments were a combination of banded (Ba), broadcast (Br), combined broadcast/banded (Br/Ba) two days before planting, combination of banding with one or two side-dressings (Ba/Sdr₁, Ba/Sdr₂) at, wide bands (WBa) and an untreated check. Banded fertilizers were placed 10 cm on each side of the plants except WBa which were placed 15 cm on either side or 10 cm deep.

All broadcast fertilizer were evenly applied over each plot area and incorporated manually. All banded treatments except those in wide bands (WBa) were applied in double bands 10 cm on each side of the plants and 10 cm deep. Plants were side dressed similar to Ba being applied as close to the plants as possible (10 cm) 40 days after planting and a second side dressing was applied at 56 days after planting. The plants were subjected to standard cultural practices and moisture was supplied through drip irrigation.

Sweetpotato slips were planted using 0.30 m in-row spacing. At nine weeks after planting, leaf samples consisting of the fifth fully expanded leaf from the end of the vine of five plants were collected from each plot. The leaves were cleaned by dipping in tap water, followed by three successive deionized water rinses. Leaf samples were dried at 65°C for 48 h and ground to pass a 1 mm screen. At harvest, storage root samples (25 g) were collected from four randomly selected US #1 storage roots from each plot and dried at 65°C for 48 h and ground to pass through 1 mm screen. All samples were analyzed for N, P, K, Ca, Mg, Mn, Cu, and B by the Plant Testing Laboratory at the Auburn University, Alabama (located 32° 35' 36.0852' N and 85° 29' 42.5868‘ W) Storage root yield and foliage weight data were obtained from roots and foliage harvested 101 days after transplanting. Storage roots were graded according to the guidelines of the National Sweepotato Collaborators Group (NSCG, 2009) into marketable (US#1, canners and jumbo) and non-marketable (culls and cracks), and root and foliage yields determined. All fields, leaf and storage root analyses data were subjected to analysis of variance and to Waller-Duncan’s Bayesian K-ratio t test at K=100 (SAS, 2009).

Growth responses

Sweetpotato vine yield increased under all placement treatments except with 1/3Br compared to the unfertilized control (Table 1).  Differences in vine yield among comparable Ba and Br treatments (2-5) were not evident. However, plants receiving 1/3 of the fertilizer Ba combined with 1/3 Br (treatment 7) produced significantly higher vine yield than when a similar rate was Ba with one side-dressing, or 1/3 the full rate Ba with 2/3 applied as two side-dressings (treatments 8 and 9). Thus, vine yields produce by all Ba placement treatments at lower quantities were as great as or greater than that of the full rate Br (treatment 6).

Total storage root yield increased substantially in response to placement treatments except with the 2/3 or full rate Br (treatments 5 and 6), when compared with that of the untreated check. Storage root yield tended to be greater with 2/3 Ba (treatment 4) when compared with either a similar or  full rate Br. Most placement treatments increased marketable yield markedly over that of the untreated check (Table 1). Marketable yield was greater when with 2/3 of NPK Ba compared with either the 2/3 Br or full rate Br; the highest marketable root yield was obtained when Ba was combined with two side-dressings. There was no difference in yield of non-marketable root yield among treatments. However, it should be noted that non-marketable yield made up only 5 and 3.6% of total yield among plants receiving treatments 4 and 9, respectively, compared to 12% for treatment 6. The number of storage roots produced per plant was not significantly lower in treatments involving higher (2/3) rates Br, Br/Ba, or the full rate Br.

Yield of US#1s increased significantly when 2/3 NPK was banded compared to either 2/3 Br, 1/3Ba/Br, or the full rate Br (Table 2). It is worthy of note also that the 2/3 Ba treatment produced similar yields of US#1s as that of 1/3 Ba with 2/3 Sd, or 2/3 Ba in wider bands. All fertilizer placement treatments produced significantly greater yields of US#1 storage roots than the untreated check. The percent of US#1s was significantly higher among plants compared to either the full or 2/3 the quantity Br. Although placement treatments significantly affected the yield of US#1s, they produced very little effect on yield of canners and jumbos.

The results show that differences in vine yield were not evident among the comparable broadcast or banded NPK treatments. However, these results do suggest that a combined Br/Ba NPK fertilizer at two-thirds the recommended rate will enhance sweetpotato vine yield compared to Ba with either one or two side dressings. Total yield of US # 1 storage roots increased markedly by banding 2/3 the recommended rate at planting versus either the same rate or full rate Br. This suggests that economically optimum production of US # 1 storage roots can be obtained if the fertilizer is banded rather than when broadcasted at planting. In some countries and even in certain ethnic communities in  the  United  States, sweetpotato is grown for its leaves to be used as a green vegetable like spinach because of its nutrient content or storage roots are used for ethanol production. The optimum fertilizer placement to meet most of the above sweetpotato production characteristics and uses appears to be the 2/3 banded rate. Although there were relatively few clear differences among comparable Br and Ba or combined treatments in uptake of nutrients, the data suggest that the tissue concentrations were enough for yield as no deficiency symptoms were evident.

The authors have not declared any conflict of interests.

The authors appreciate the contribution of the Washington Carver Agricultural Experiment Station, Tuskegee University. This research was supported by funds from USDA/NIFA Evans-Allen Grant No. ALX-SP-1.