Phosphate Rock Fertilization in Tilled and No-Till Low-Input Systems in the Humid Tropics Mwenja P. Gichuru and Pedro A. Sanchez* ABSTRACT Application of phosphate rock is of interest to low-input systems in the humid tropics, because phosphate rock is less expensive than ordinary superphosphate and releases P quickly in acid soils. How- ever, the lack of incorporation in zero-tillage systems may result in low availability of surface-applied P. This study was directed toward determining the effects of these two P sources under no-till and rotovation tillage systems on crop production in a fine-loamy sili- ceous, isohyperthermic Typic Paleudult of Yurimaguas, Peru. A ro- tation of Al-tolerant cultivars of rice (Oryza sativa L.)-rice-cowpea (Vigna unguicukala) was followed for seven consecutive harvests. Grain yields increased with rotovation in the first crop, were not affected by tillage methods during the second and third crops, but decreased with rotovation from the third to the seventh crops. Se- chura phosphate rock at a soil pH of 4.5 was as effective as super- phosphate in supplying available P. A total of 13.9 Mg ha-' of rice and 2.5 Mg ha-' of cowpea grain was produced in seven harvests in newly cleared fields without lime or P application. There were significant responses to P fertilization in one rice crop and in both cowpea crops. On the average, however, rice yield did not respond to P. A single application of 22 kg P ha-' was sufficient to produce 85% of the maximum yield of cowpea for 2 yr. The results with cowpea indicate that broadcast phosphate rock is a good source of P for low-input systems on acid soils where acid-tolerant cultivars are used. HE ULTISOLS of the Amazon basin of Peru are T generally acidic and deficient in P. Although soil fertility problems can be corrected by using liming and fertilization practices similar to those developed for Ultisols of the southeastern United States (Sanchez et al., 1983), socioeconomic constraints often limit the application of fertilizer-based continuous crop pro- duction in areas where farmers still practice shifting cultivation. The low-input approach to food produc- tion is an alternative to shifting cultivation in areas where fallow periods are too short and continuous cul- tivation technology is limited by lack of market in- frastructure. This approach is based on acid tolerant cultivars, minimum tillage, minimum use of pur- chased inputs and maximum nutrient recycling (San- chez and Benites, 1987). Application of phosphate rock is an attractive pos- sibility for low-input systems because it is less expen- sive than superphosphates (Khasawneh and Doll, 1978). The use of acid-tolerant crops may permit more efficient utilization of P from phosphate rock, because the plants will grow under acid conditions that favor dissolution of apatite. In addition to reducing energy costs and promoting soil conservation, surface appli- M.P. Gichuru, Int. Inst. for Tropical Agric., Ibadan, Nigeria; and P.A. Sanchez, Dep. of Soil Sci., North Carolina State Univ., Raleigh, NC 27695-7619. North Carolina Agric. Res. Service Paper J1294. Contribution of the TropSoils Program, in cooperation with the Instituto Nacional de Investigacibn Agraria y Agroindustrial of Peru (INIAA) and supported by the United States Agency for Interna- tional Development. Received 7 Oct. 1987. *Corresponding author. Published in Agron. J. 80:943-947 (1 988). cations of P may also facilitate a gradual transition from shifting to continuous cultivation, while fertil- izer incorporation is difficult in fields that still have fallen logs and tree stumps. Although surface appli- cation of P fertilizers has been found to be comparable to incorporating them into nonacid soils (Singh et al. 1966; Kang and Yunusa, 1977; Hargrove et al., 1982), little information is available on surface application of phosphate rock to annual food crops under acid soil conditions. The objective of this work was to determine the relationships between P sources and tillage methods in a low-input system using acid-tolerant crops. MATERIALS AND METHODS An experiment was conducted on a nearly level field (Y- 208) at the Yurimaguas Experiment Station in the Amazon of Peru. The soil is a fine-loamy siliceous, isohyperthermic Typic Paleudult (Tyler et al., 1978). Initial soil properties are shown in Table 1. The climate is humid tropical with a mean annual temperature of 26"C, mean annual rainfall of 2100 mm with a weak dry season, and an elevation of 182 m. The field was cleared of secondary forest in May 1982, and burned and planted to a uniform crop of cowpea prior to treatment applications. The experimental design was split- split plots in randomized complete blocks with four repli- cations. The main plot treatments were two tillage methods: no-till with surface application of fertilizers, and rotovated with fertilizers broadcast and incorporated to 8 to 10 cm. Rotovation was repeated prior to each subsequent planting. The subplots were P sources: ordinary superphosphate con- taining 90 g kg-' of total P, and finely ground Sechura phos- phate rock containing 130 g kg-' of total P and 23 g kg-' of citrate soluble P (Chien and Hammond, 1978). Its chem- ical composition (130, 328, 3.2, 15.6, 41, and 28 g kg-', P, Ca, Mg, Na, C02, and F, respectively) indicates that Sechura rock is favorable for direct application (Hol€inann, and Breen, 1964; Fassbender, 1967; Chien and Hammond, 1978). Se- chura is one of the most reactive phosphate rocks, and its reserves are calculated to be over 550 million tonnes (Fass- bender, 1967; Leon et al., 1986). Sub-subplot (15-mZ) treat- ments were P rates of 0, 11, 22,44, and 88 kg ha-l applied once, plus an additional treatment of 11 kg ha-l per crop, making a total of 24 treatments. A rotation of upland rice (cv. African0 Desconocid0)-rice- cowpea (cv. Vita 6 or Vita 7)-rice-rice-cowpea was followed. These cultivars are highly tolerant of soil acidity (Nicho- laides and Piha, 1987). Hole spacing was 0.5 by 0.3 m for rice, and 0.5 by 0.1 5 m for cowpea. An average of four seeds per hole was planted using a "tacarpo" (planting stick). Two Table 1. Selected soil properties immediately after slash and burn. Exchangeable Effec- AI soil tive satura- Avail. depth Clay pH AI Ca Mg K CEC tion Pt cm - % - - cmol L-' soil - % mg kg-' 0-15 15 4.5 1.9 1.3 0.4 0.15 3.8 53 14 15-30 19 4.3 3.9 0.5 0.1 0.11 4.6 84 4 30-45 21 4.3 4.1 0.4 0.1 0.09 4.7 88 3 t By the modified Olsen procedure (Hunter, 1974). 943 Published November, 1988 944 AGRONOMY JOURNAL, VOL. 80, NOVEMBER-DECEMBER 1988 weeks after emergence, cowpea was thinned to two plants per hole. In addition, the following fertilizers were uniformly applied to all plots: 50 kg N ha-' as urea, applied to the first rice crop, and 40 kg N ha-' per rice crop thereafter; 100 kg K ha-l as K2S04, per rice crop; 25 kg Mg ha-l as MgSO4.7H2O, per crop; and 1 kg ha-I each of Zn and Cu as sulfates, applied once. No lime was applied, and soil pH remained at about 4.5. Bulk density was determined with undisturbed Uhland cores. Crop residues and weeds were returned to the soil surface. Rice grain and straw yields were obtained from a 6-m2 area devoid of border effects. Cowpea stover yields were estimated from 1-m2 samples just before leaf drop, and cowpea grain yield estimates were made from a 5-m2 area. All grain yields are reported at 14% moisture. Stover samples of whole plants after harvest were dried in a forced draft oven at 70°C. Tissue subsamples were ground and dry-ashed for the de- termination of N, P, K, Ca, and Mg as described by Hunter (1 974). Soil samples were taken from the surface 15 cm in each plot immediately after clearing and burning, and 26, 43, 74, and 91 wk after treatment application. The samples were air-dried and ground to pass a 2-mm sieve. Exchange- able Al, Ca, and Mg were extracted from 5 cm3 of soil with 25 mL of 1 M KC1. Phosphorus and K were extracted by the modified Olsen procedure (Hunter 1974). This extrac- tant measures the reaction products of phosphate rock and does not react with the unaltered rock (Khasawneh and Doll, 1978). Soil pH measurements were made using 1:2.5, soil/ water slurries. All laboratory procedures are those used by the North Carolina State University Tropical Soils Labo- ratory as described by Hunter (1974). The SAS procedures were used as described by the SAS Institute Staff (1 979). Values of LSD are presented only if the overall F-test was significant at P = 0.05. Therefore, the error rate for LSD is experiment-wise (Steel and Torrie, 1980). The best regression lines were fitted using General Linear Models procedure (SAS Institute Staff, 1979). The term "sig- nificant differences" indicates that the means or regression fits being discussed were statistically different at the proba- bility level of 0.05 or less. RESULTS AND DISCUSSION The cowpea uniformity trial produced an average of 2 Mg ha-l of grain. This yield is high and com- parable to cowpea yields obtained with complete lime and fertilization in nearby trials (Nicholaides and Piha, 1987). The high yields are believed to be due to the initial favorable soil chemical properties after slash and burn (Table 1). Source Efects There were no significant differences in grain yield due to P source in any of the seven crops harvested. Rice yields averaged 2.9 Mg ha-' and ranged from 2.1 to 4.1 Mg ha-'; cowpea yields averaged 1.3 Mg ha-' and ranged from 1.1 to 1.4 Mg ha-'. All represent moderate to high crop yields for this environment. The absence of a significant difference between P sources could be due to the effect of the soil acidity in solubilizing the phosphate rock. The two P sources produced similar modified Olsen extractable P values as shown in Table 2. Sechura phosphate rock has been shown ta be highly reactive, but most results are from controlled pot experiments where plants are not grown to maturity (Hoffmann and Breen, 1964; Chien and Hammond, 1978; Leon et al., 1986). These findings indicate the reactivity effectiveness of Sechura under acid field conditions. Since the responses to P appli- cation rates were the same fix the two sources, only yield data from phosphate rock are presented in the rest of the paper. Tillage Eflects Table 3 shows the influence of tillage treatments on grain yields of seven consecutive crops. The first rice crop produced 840 kg ha-' more grain in rotovated than in no-till treatments. This significant increase was probably due to improved soil physical properties and perhaps to a more uniform distribution of the chem- ical inputs, crop residues, and ash fiom burning into the root zone (not measured). The second and third crops, however, showed no significant tillage response. After the third crop, severe yield reduction occurred in the rotovated treatments. 'The fourth, fifth, and sev- enth crops produced significantly more grain in the no-till treatments than in the rotovated ones. Tillage differences were not significant in the sixth crop, but no-till plots tended to yield slightly more than the ro- tovated treatments. The reason for the negative response to rotov •3 S X< S 90 80 70 60 50 o Rice y =86.8 +0.42827 P -0.003167 P2 R2=0.04NS Cowpea y =64.9+1.2326 P-0.009759P2 R2=0.29** 22 44 66 88 PHOSPHORUS RATE ( kg P ha'1) Fig. 3. Effect of applied phosphate rock on relative grain yiield of five rice and two cowpea harvest (** = significant at 0.01 proba- bility level; NS = not significant). Plotted points are means of four replications. (b) Stover o y=0.45+O.OI89P-O.OOOI9P2 R2=0.37* D y =0.69+0.0044 P rz=O.I4NS • y =0.95 + 0.0238P-O.OOOI6P2 R2=0.34* A y =0.60+0,0022 P rz=0.34* ' • y= 0.38+0.00085 P r2=O.I4NS 22 44 66 88 PHOSPHORUS RATE (RgPha"1) Fig. 2. Effect of applied phosphate rock on grain and stover P concentration of five consecutive crop (numbers above regression lines represent crop sequence; *, **, *** = significant at 0.05,0.01, and 0.001 probability levels, respectively; NS = not significant). Plotted points are means of four replications. Numerals 1 to 7 refer to crop sequence. BURESH ET AL.: INHIBITORS FOR UREA-FERTILIZED RICE 947 AJ Menu AJ Tables of Contents (Disc 6) Help Search